Received: from mail-pz0-f61.google.com ([209.85.210.61])
	by chain.digitalkingdom.org with esmtp (Exim 4.72)
	(envelope-from <bpfk-list+bncCMbnveiNHRCZh9DoBBoE-x2PEA@googlegroups.com>)
	id 1PVxHZ-0001mQ-SD; Thu, 23 Dec 2010 18:21:27 -0800
Received: by pzk9 with SMTP id 9sf351248pzk.16
        for <multiple recipients>; Thu, 23 Dec 2010 18:21:19 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=googlegroups.com; s=beta;
        h=domainkey-signature:received:mime-version:x-beenthere:received
         :received:received:received:received-spf:received:received:received
         :date:message-id:to:subject:from:x-original-sender
         :x-original-authentication-results:reply-to:precedence:mailing-list
         :list-id:list-post:list-help:list-archive:sender:list-subscribe
         :list-unsubscribe:content-type:content-transfer-encoding;
        bh=kWDwSNVCw1COPP3RUuLUgHjbCJGy8TXLGEpB3s6Fc9Y=;
        b=ollYkXlPZtLelhfDdNaRb9laqzqWVJ+RfLpijqiOw/ZXNCeN3Kw5sFjRAQYlYB1Yzt
         k/0uNNW7MMYDq7KMfpVP7J9ls8KlQfQbLkv+lEYujFKUkq1ynaAHF0JpjTl6WohLyIey
         k2igdj3gimdRF2sBkByhD/dOTeYnJyWk28yuo=
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=googlegroups.com; s=beta;
        h=mime-version:x-beenthere:received-spf:date:message-id:to:subject
         :from:x-original-sender:x-original-authentication-results:reply-to
         :precedence:mailing-list:list-id:list-post:list-help:list-archive
         :sender:list-subscribe:list-unsubscribe:content-type
         :content-transfer-encoding;
        b=L9yzF4sv3N2MpnJqvNwpMAD+aZ6/e6zxVgTDyXl7UzmonLHAeH+FUotsXabIsj19Ek
         o7io/33qfD3/Xwdf2ZOzTbia6w+0CxElNH1X/G3sbCw+P+sjndLiSQdgk67NkYoUpZYH
         4oFGxplN76dgGYgG9TrzEWcsxum7WUnkNBKhk=
Received: by 10.142.62.31 with SMTP id k31mr463703wfa.9.1293157273812;
        Thu, 23 Dec 2010 18:21:13 -0800 (PST)
MIME-Version: 1.0
X-BeenThere: bpfk-list@googlegroups.com
Received: by 10.142.249.41 with SMTP id w41ls12690672wfh.1.p; Thu, 23 Dec 2010
 18:21:13 -0800 (PST)
Received: by 10.142.177.1 with SMTP id z1mr6654082wfe.2.1293157272692;
        Thu, 23 Dec 2010 18:21:12 -0800 (PST)
Received: by 10.142.177.1 with SMTP id z1mr6654079wfe.2.1293157272120;
        Thu, 23 Dec 2010 18:21:12 -0800 (PST)
Received: from chain.digitalkingdom.org (digitalkingdom.org [173.13.139.234])
        by gmr-mx.google.com with ESMTPS id m3si8774218wfl.5.2010.12.23.18.21.09
        (version=TLSv1/SSLv3 cipher=RC4-MD5);
        Thu, 23 Dec 2010 18:21:12 -0800 (PST)
Received-SPF: pass (google.com: best guess record for domain of nobody@digitalkingdom.org designates 173.13.139.234 as permitted sender) client-ip=173.13.139.234;
Received: from nobody by chain.digitalkingdom.org with local (Exim 4.72)
	(envelope-from <nobody@digitalkingdom.org>)
	id 1PVxHI-0001mC-E2
	for bpfk-list@googlegroups.com; Thu, 23 Dec 2010 18:21:08 -0800
Received: from 128-177-28-49.ip.openhosting.com ([128.177.28.49] helo=oh-www1.lojban.org)
	by chain.digitalkingdom.org with esmtp (Exim 4.72)
	(envelope-from <www-data@oh-www1.lojban.org>)
	id 1PVxHG-0001ls-G4
	for bpfk@lojban.org; Thu, 23 Dec 2010 18:21:08 -0800
Received: from www-data by oh-www1.lojban.org with local (Exim 4.72)
	(envelope-from <www-data@oh-www1.lojban.org>)
	id 1PVxHE-0003pA-Ln
	for bpfk@lojban.org; Thu, 23 Dec 2010 21:21:05 -0500
Date: Thu, 23 Dec 2010 21:21:04 -0500
Message-Id: <E1PVxHE-0003pA-Ln@oh-www1.lojban.org>
To: bpfk@lojban.org
Subject: [bpfk] dag-cll git updates for Thu Dec 23 21:21:04 EST 2010
From: www-data <www-data@oh-www1.lojban.org>
X-Original-Sender: www-data@oh-www1.lojban.org
X-Original-Authentication-Results: gmr-mx.google.com; spf=pass (google.com:
 best guess record for domain of nobody@digitalkingdom.org designates
 173.13.139.234 as permitted sender) smtp.mail=nobody@digitalkingdom.org
Reply-To: bpfk-list@googlegroups.com
Precedence: list
Mailing-list: list bpfk-list@googlegroups.com; contact bpfk-list+owners@googlegroups.com
List-ID: <bpfk-list.googlegroups.com>
List-Post: <http://groups.google.com/group/bpfk-list/post?hl=en_US>, <mailto:bpfk-list@googlegroups.com>
List-Help: <http://groups.google.com/support/?hl=en_US>, <mailto:bpfk-list+help@googlegroups.com>
List-Archive: <http://groups.google.com/group/bpfk-list?hl=en_US>
Sender: bpfk-list@googlegroups.com
List-Subscribe: <http://groups.google.com/group/bpfk-list/subscribe?hl=en_US>, <mailto:bpfk-list+subscribe@googlegroups.com>
List-Unsubscribe: <http://groups.google.com/group/bpfk-list/subscribe?hl=en_US>,
 <mailto:bpfk-list+unsubscribe@googlegroups.com>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable
Content-Length: 2060108

commit 3876745f65dd1f06b1e50a72513c8bcd2a9964a7
Author: Robin Lee Powell <rlpowell@digitalkingdom.org>
Date:   Thu Dec 23 17:53:28 2010 -0800

    Checkin of the old source, just in case.

diff --git a/todocbook/orig/REFG9192009.DOC b/todocbook/orig/REFG9192009.DO=
C
new file mode 100644
index 0000000..078ca89
Binary files /dev/null and b/todocbook/orig/REFG9192009.DOC differ
diff --git a/todocbook/orig/REFG9192009.docx b/todocbook/orig/REFG9192009.d=
ocx
new file mode 100644
index 0000000..7d2bb3b
Binary files /dev/null and b/todocbook/orig/REFG9192009.docx differ
diff --git a/todocbook/orig/catdoc.out b/todocbook/orig/catdoc.out
new file mode 100644
index 0000000..aff2dca
--- /dev/null
+++ b/todocbook/orig/catdoc.out
@@ -0,0 +1,20580 @@
+<h2>Chapter 1<br>
+Lojban As We Mangle It In Lojbanistan: About This Book</h2>
+<h6>$Revision: 4.3 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>What is Lojban?</h3>
+<p>
+<cx "Lojban, history of">  XE "Lojban: history of"   <cx "Logical Language=
 Group, relation to Lojban">  XE "Logical Language Group: relation to Lojba=
n"   Lojban (pronounced =93LOZH-bahn=94) is a constructed language. Previou=
s versions of the language were called =93Loglan=94 by Dr. James Cooke Brow=
n, who founded the Loglan Project and started the development of the langua=
ge in 1955. The goals for the language were first described in the open lit=
erature in the article =93Loglan=94, published in <cite>Scientific American=
</cite>, June, 1960. Made well-known by that article and by occasional refe=
rences in science fiction (most notably in Robert Heinlein's novel <cite>Th=
e Moon Is A Harsh Mistress) </cite>and computer publications, Loglan and Lo=
jban have been built over four decades by dozens of workers and hundreds of=
 supporters, led since 1987 by The Logical Language Group (who are the publ=
ishers of this book).
+<p>
+<cx "Lojban, features of">  XE "Lojban: features of"   There are thousands=
 of artificial languages (of which Esperanto is the best-known), but Loglan=
/Lojban has been engineered to make it unique in several ways. The followin=
g are the main features of Lojban:
+<p>
+<dl compact><dt><dd>Lojban is designed to be used by people in communicati=
on with each other, and possibly in the future with computers.
+<p><dt><dd>Lojban is designed to be neutral between cultures.
+<p><dt><dd>Lojban grammar is based on the principles of predicate logic.
+<p><dt><dd>Lojban has an unambiguous yet flexible grammar.
+<p><dt><dd>Lojban has phonetic spelling, and unambiguously resolves its so=
unds into words.
+<p><dt><dd>Lojban is simple compared to natural languages; it is easy to l=
earn.
+<p><dt><dd>Lojban's 1300 root words can be easily combined to form a vocab=
ulary of millions of words.
+<p><dt><dd>Lojban is regular; the rules of the language are without except=
ions.
+<p><dt><dd>Lojban attempts to remove restrictions on creative and clear th=
ought and communication.
+<p><dt><dd>Lojban has a variety of uses, ranging from the creative to the =
scientific, from the theoretical to the practical.
+<p><dt><dd>Lojban has been demonstrated in translation and in original wor=
ks of prose and poetry.
+</dl><a name=3Ds2><h3>What is this book?</h3>
+<p>
+<cx "reference grammar">  XE "reference grammar"   <cx "this book, goal of=
">  XE "this book: goal of"   <cx "goal of this book">  XE "goal of this bo=
ok"   This book is what is called a =93reference grammar=94. It attempts to=
 expound the whole Lojban language, or at least as much of it as is underst=
ood at present. Lojban is a rich language with many features, and an attemp=
t has been made to discover the functions of those features. The word =93di=
scover=94 is used advisedly; Lojban was not =93invented=94 by any one perso=
n or committee. Often, grammatical features were introduced into the langua=
ge long before their usage was fully understood. Sometimes they were introd=
uced for one reason, only to prove more useful for other reasons not recogn=
ized at the time.
+<p>
+By intention, this book is complete in description but not in explanation.=
 For every rule in the formal Lojban grammar (given in <a href=3Dchap21.htm=
l>Chapter 21</a>), there is a bit of explanation and an example somewhere i=
n the book, and often a great deal more than a bit. In essence, <a href=3Dc=
hap2.html>Chapter 2 </a>gives a brief overview of the language, <a href=3Dc=
hap21.html>Chapter 21 </a>gives the formal structure of the language, and t=
he chapters in between put semantic bones on that formal flesh. I hope that=
 eventually more grammatical material founded on (or even correcting) the e=
xplanations in this book will become available.
+<p>
+<cx "Lojban, stability of">  XE "Lojban: stability of"   <cx "linguistic d=
rift">  XE "linguistic drift"   Nevertheless, the publication of this book =
is, in one sense, the completion of a long period of language evolution. Wi=
th the exception of a possible revision of the language that will not even =
be considered until five years from publication date, and any revisions of =
this book needed to correct outright errors, the language described in this=
 book will not be changing by deliberate act of its creators any more. Inst=
ead, language change will take place in the form of new vocabulary =97 Lojb=
an does not yet have nearly the vocabulary it needs to be a fully usable la=
nguage of the modern world, as <a href=3Dchap12.html>Chapter 12 </a>explain=
s =97 and through the irregular natural processes of drift and (who knows?)=
 native-speaker evolution. (Teach your children Lojban!) You can learn the =
language described here with assurance that (unlike previous versions of Lo=
jban and Loglan, as well as most other artificial languages) it will not be=
 subject to further fiddling by language-meisters.
+<p>
+<cx "structure of this book">  XE "structure of this book"   <cx "this boo=
k, structure of">  XE "this book: structure of"   It is probably worth ment=
ioning that this book was written somewhat piecemeal. Each chapter began li=
fe as an explication of a specific Lojban topic; only later did these begin=
 to clump together into a larger structure of words and ideas. Therefore, t=
here are perhaps not as many cross-references as there should be. However, =
I have attempted to make the index as comprehensive as possible.
+<p>
+<cx "Lojbanistan">  XE "Lojbanistan"   <cx "jokes">  XE "jokes"   <cx "cha=
pter titles, intent of">  XE "chapter titles: intent of"   Each chapter has=
 a descriptive title, often involving some play on words; this is an attemp=
t to make the chapters more memorable. The title of <a href=3Dchap1.html>Ch=
apter 1 </a>(which you are now reading), for example, is an allusion to the=
 book <cite>English As We Speak It In Ireland</cite>, by P. W. Joyce, which=
 is a sort of informal reference grammar of Hiberno-English. =93Lojbanistan=
=94 is both an imaginary country where Lojban is the native language, and a=
 term for the actual community of Lojban-speakers, scattered over the world=
. Why =93mangle=94? As yet, nobody in the real Lojbanistan speaks the langu=
age at all well, by the standards of the imaginary Lojbanistan; that is one=
 of the circumstances this book is meant to help remedy.
+<p>
+What are the typographical conventions of this book?</h3>
+<p><cx "typographical conventions">  XE "typographical conventions"   <a n=
ame=3Ds3><h3>
+<cx "sections of this book">  XE "sections of this book"   <cx "this book,=
 sections of">  XE "this book: sections of"   Each chapter is broken into n=
umbered sections; each section contains a mixture of expository text, numbe=
red examples, and possibly tables.
+<p>
+<cx "examples in this book">  XE "examples in this book"   <cx "this book,=
 examples of">  XE "this book: examples of"   The reader will notice a cert=
ain similarity in the examples used throughout the book. One chapter after =
another rings the changes on the self-same sentences:
+<p>
+<ex "go to the store">  XE "go to the store: example=93  <pre><a name=3De1=
d1>1.1)	mi klama le zarci
+	I go-to that-which-I-describe-as-a store.
+	I go to the store.
+</pre>will become wearisomely familiar before <a href=3Dchap21.html>Chapte=
r 21 </a>is reached. This method is deliberate; I have tried to use simple =
and (eventually) familiar examples wherever possible, to avoid obscuring ne=
w grammatical points with new vocabulary. Of course, this is not the method=
 of a textbook, but this book is not a textbook (although people have learn=
ed Lojban from it and its predecessors). Rather, it is intended both for se=
lf-learning (of course, at present would-be Lojban teachers must be self-le=
arners) and to serve as a reference in the usual sense, for looking up obsc=
ure points about the language.
+<p>
+<cx "example of examples">  XE "example of examples"   <cx "structure of e=
xamples">  XE "structure of examples"   <cx "examples, structure of">  XE "=
examples: structure of"   It is useful to talk further about <a href=3D#e1d=
1>Example 1.1 </a>for what it illustrates about examples in this book. Exam=
ples usually occupy three lines. The first of these is in Lojban, the secon=
d in a word-by-word literal translation of the Lojban into English, and the=
 third in colloquial English. The second and third lines are sometimes call=
ed the =93literal translation=94 and the =93colloquial translation=94 respe=
ctively. Sometimes, when clarity is not sacrificed thereby, one or both are=
 omitted. If there is more than one Lojban sentence, it generally means tha=
t they have the same meaning.
+<p>
+<cx "square brackets, use of in notation">  XE "square brackets: use of in=
 notation"   Words are sometimes surrounded by square brackets. In Lojban t=
exts, these enclose optional grammatical particles that may (in the context=
 of the particular example) be either omitted or included. In literal trans=
lations, they enclose words that are used as conventional translations of s=
pecific Lojban words, but don't have exactly the meanings or uses that the =
English word would suggest. In <a href=3Dchap3.html>Chapter 3</a>, square b=
rackets surround phonetic representations in the International Phonetic Alp=
habet.
+<p>
+<cx "tables, format of">  XE "tables: format of"   <cx "grammatical catego=
ries, use of upper case for">  XE "grammatical categories: use of upper cas=
e for"   Many of the tables, especially those placed at the head of various=
 sections, are in three columns. The first column contains Lojban words dis=
cussed in that section; the second column contains the grammatical category=
 (represented by an UPPER CASE Lojban word) to which the word belongs, and =
the third column contains a brief English gloss, not necessarily or typical=
ly a full explanation. Other tables are explained in context.
+<p>
+<cx "technical terms">  XE "technical terms"   A few Lojban words are used=
 in this book as technical terms. All of these are explained in <a href=3Dc=
hap2.html>Chapter 2</a>, except for a few used only in single chapters, whi=
ch are explained in the introductory sections of those chapters.
+<p>
+<a name=3Ds4><h3>Disclaimers</h3>
+<p>
+<cx "disclaimers">  XE "disclaimers"   It is necessary to add, alas, that =
the examples used in this book do not refer to any existing person, place, =
or institution, and that any such resemblance is entirely coincidental and =
unintentional, and not intended to give offense.
+<p>
+<cx "dictionary, superior authority of">  XE "dictionary: superior authori=
ty of"   When definitions and place structures of gismu, and especially of =
lujvo, are given in this book, they may differ from those given in the Lojb=
an/English dictionary (which, as of this writing, is not yet published). If=
 so, the information given in the dictionary supersedes whatever is given h=
ere.
+<p>
+<a name=3Ds5><h3>Acknowledgements and Credits</h3>
+<p>
+<cx "author of this book">  XE "author of this book"   <cx "this book, aut=
hor of">  XE "this book: author of"   <cx "LLG">  XE "LLG"   Although the b=
ulk of this book was written for the Logical Language Group (LLG) by John C=
owan, who is represented by the occasional authorial =93I=94, certain chapt=
ers were first written by others and then heavily edited by me to fit into =
this book.
+<p>
+<cx "contributors to this book">  XE "contributors to this book"   <cx "th=
is book, contributors to">  XE "this book: contributors to"   In particular=
: <a href=3Dchap2.html>Chapter 2 </a>is a fusion of originally separate doc=
uments, one by Athelstan, and one by Nora Tansky LeChevalier and Bob LeChev=
alier; <a href=3Dchap3.html>Chapters 3 </a>and <a href=3Dchap4.html>4 </a>w=
ere originally written by Bob LeChevalier with contributions by Chuck Barto=
n; <a href=3Dchap12.html>Chapter 12 </a>was originally written (in much lon=
ger form) by Nick Nicholas; the dialogue near the end of <a href=3Dchap13.h=
tml>Chapter 13 </a>was contributed by Nora Tansky LeChevalier; <a href=3Dch=
ap15.html>Chapter 15 </a>and parts of <a href=3Dchap16.html>Chapter 16 </a>=
were originally by Bob LeChevalier; and the YACC grammar in <a href=3Dchap2=
1.html>Chapter 21 </a>is the work of several hands, but is primarily by Bob=
 LeChevalier and Jeff Taylor. The BNF grammar, which is also in <a href=3Dc=
hap21.html>Chapter 21</a>, was originally written by me, then rewritten by =
Clark Nelson, and finally touched up by me again.
+<p>
+<cx "credits for this book">  XE "credits for this book"   <cx "this book,=
 credits for">  XE "this book: credits for"   The research into natural lan=
guages from which parts of <a href=3Dchap5.html>Chapter 5 </a>draw their ma=
terial was performed by Ivan Derzhanski. LLG acknowledges his kind permissi=
on to use the fruits of his research.
+<p>
+<cx "credits for pictures">  XE "credits for pictures"   <cx "pictures, cr=
edits for">  XE "pictures: credits for"   The pictures in this book were dr=
awn by Nora Tansky LeChevalier, except for the picture appearing in <a href=
=3Dchap4.html>Chapter 4</a>, which is by Sylvia Rutiser Rissell.
+The index was made by Nora Tansky LeChevalier.
+<p>
+<cx "reviewers of this book">  XE "reviewers of this book"   <cx "this boo=
k, reviewers of">  XE "this book: reviewers of"   I would like to thank the=
 following people for their detailed reviews, suggestions, comments, and ea=
rly detection of my embarrassing errors in Lojban, logic, English, and cros=
s-references: Nick Nicholas, Mark Shoulson, Veijo Vilva, Colin Fine, And Ro=
sta, Jorge Llambias, Iain Alexander, Paulo S. L. M. Barreto, Robert J. Chas=
sell, Gale Cowan, Karen Stein, Ivan Derzhanski, Jim Carter, Irene Gates, Bo=
b LeChevalier, John Parks-Clifford (also known as =93pc=94), and Nora Tansk=
y LeChevalier.
+<p>
+Nick Nicholas (NSN) would like to thank the following Lojbanists: Mark Sho=
ulson, Veijo Vilva, Colin Fine, And Rosta, and Iain Alexander for their sug=
gestions and comments; John Cowan, for his extensive comments, his exemplar=
y trailblazing of Lojban grammar, and for solving the =93manskapi=94 dilemm=
a for NSN; Jorge Llambias, for his even more extensive comments, and for fo=
rcing NSN to think more than he was inclined to; Bob LeChevalier, for his s=
keptical overview of the issue, his encouragement, and for scouring all Loj=
ban text his computer has been burdened with for lujvo; Nora Tansky LeCheva=
lier, for writing the program converting old rafsi text to new rafsi text, =
and sparing NSN from embarrassing errors; and Jim Carter, for his dogged pe=
rsistence in analyzing lujvo algorithmically, which inspired this research,=
 and for first identifying the three lujvo classes.
+<p>
+<cx "Brown, James Cooke">  XE "Brown: James Cooke"   Of course, the entire=
 Loglan Project owes a considerable debt to James Cooke Brown as the langua=
ge inventor, and also to several earlier contributors to the development of=
 the language. Especially noteworthy are Doug Landauer, Jeff Prothero, Scot=
t Layson, Jeff Taylor, and Bob McIvor. Final responsibility for the remaini=
ng errors and infelicities is solely mine.
+<p>
+<a name=3Ds6><h3>Informal Bibliography</h3>
+<p>
+<cx "bibliography">  XE "bibliography"   <cx "Loglan">  XE "Loglan"   The =
founding document for the Loglan Project, of which this book is one of the =
products, is <cite>Loglan 1: A Logical Language </cite>by James Cooke Brown=
 (4th ed. 1989, The Loglan Institute, Gainesville, Florida, U.S.A.) The lan=
guage described therein is not Lojban, but is very close to it and may be c=
onsidered an ancestral version. It is regrettably necessary to state that n=
othing in this book has been approved by Dr. Brown, and that the very exist=
ence of Lojban is disapproved of by him.
+<p>
+The logic of Lojban, such as it is, owes a good deal to the American philo=
sopher W. v.O. Quine, especially <cite>Word and Object </cite>(1960, M.I.T.=
 Press). Much of Quine's philosophical writings, especially on observation =
sentences, reads like a literal translation from Lojban.
+<p>
+The theory of negation expounded in <a href=3Dchap15.html>Chapter 15 </a>i=
s derived from a reading of Larry Horn's work <cite>The Natural History of =
Negation</cite>.
+<p>
+Of course, neither Brown nor Quine nor Horn is in any way responsible for =
the uses or misuses I have made of their works.
+<p>
+<cx "books about Lojban">  XE "books about Lojban"   Depending on just whe=
n you are reading this book, there may be three other books about Lojban av=
ailable: a textbook, a Lojban/English dictionary, and a book containing gen=
eral information about Lojban. You can probably get these books, if they ha=
ve been published, from the same place where you got this book. In addition=
, other books not yet foreseen may also exist.
+<p>
+Captions to Pictures</h3>
+<p><cx "captions to pictures">  XE "captions to pictures"   <cx "pictures,=
 captions to">  XE "pictures: captions to"   <a name=3Ds7><h3>
+The following examples list the Lojban caption, with a translation, for th=
e picture at the head of each chapter. If a chapter's picture has no captio=
n, =93(none)=94 is specified instead.
+<p>
+<pre><a name=3De7d1>7.1)	<a href=3Dchap1.gif>coi lojban.		coi rodo </a>
+	Greetings, O Lojban!	Greetings, all-of you
+<a name=3De7d2>7.2)	(none)
+<a name=3De7d3>7.3)	<a href=3Dchap3.gif>.i .ai .i .ai .o </a>
+	[untranslatable]
+<a name=3De7d4>7.4)	<a href=3Dchap4.gif>jbobliku </a>
+	Lojbanic-blocks
+<a name=3De7d5>7.5)	(none)
+<a name=3De7d6>7.6)	<a href=3Dchap6.gif>lei re nanmu cu bevri le re nanmu =
</a>
+	The-mass-of two men carry the two men
+	Two men (jointly) carry two men (both of them).
+<a name=3De7d7>7.7)	<a href=3Dchap7.gif>ma drani danfu </a>
+		.i di'e
+		.i di'u
+		.i dei
+		.i ri
+		.i do'i
+	[What sumti] is-the-correct type-of-answer?
+		The-next-sentence.
+		The-previous-sentence.
+		This-sentence.
+		The-previous-sentence.
+		An-unspecified-utterance.
+<a name=3De7d8>7.8)	<a href=3Dchap8.gif>ko viska re prenu poi bruna la san=
tas. </a>
+	[You!] see two persons who-are brothers-of Santa.
+<a name=3De7d9>7.9)	(none)
+<a name=3De7d10>7.10)	<a href=3Dchap10.gif>za'o klama </a>
+	[superfective] come/go
+	Something goes (or comes) for too long.
+<a name=3De7d11>7.11)	<a href=3Dchap11.gif>le si'o kunti </a>
+	The concept-of emptiness
+<a name=3De7d12>7.12)	(none)
+<a name=3De7d13>7.13)	<a href=3Dchap13.gif>.oi ro'i ro'a ro'o </a>
+	[Pain!] [emotional] [social] [physical]
+<a name=3De7d14>7.14)	(none)
+<a name=3De7d15>7.15)	<a href=3Dchap15.gif>mi na'e lumci le karce </a>
+	I other-than wash the car.
+	I didn't wash the car.
+<a name=3De7d16>7.16)	<a href=3Dchap16.gif>drata mupli pe'u .djan. </a>
+	Another example [please] John.
+	Another example, John, please!
+<a name=3De7d17>7.17)	<a href=3Dchap17.gif>zai xanlerfu by. ly. .obu .jy b=
y. .abu ny. </a>
+	[Shift] hand-letters l o j b a n
+	"Lojban" in the U.S. manual alphabet
+<a name=3De7d18>7.18)	<a href=3Dchap18.gif>no	no </a>
+	0	0
+<a name=3De7d19>7.19)	(none)
+<a name=3De7d20>7.20)	(none)
+<a name=3De7d21>7.21)	(none)
+<p></pre><cx "boring legalities">  XE "boring legalities"   <cx "legalitie=
s, boring">  XE "legalities: boring"   <cx "permissions notice">  XE "permi=
ssions notice"   <a name=3Ds8><h3>
+<p>
+</body></html>
+<img src=3Dchap2.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D4=
05>
+
+<h2>Chapter 2
+<br>
+A Quick Tour of Lojban Grammar, With Diagrams</h2>
+<h6>$Revision: 4.1 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>The concept of the bridi</h3>
+<p>
+<cx "bridi, concept of">  XE "bridi: concept of"   This chapter gives diag=
rammed examples of basic Lojban sentence structures. The most general patte=
rn is covered first, followed by successive variations on the basic compone=
nts of the Lojban sentence. There are many more capabilities not covered in=
 this chapter, but covered in detail in later chapters, so this chapter is =
a =93quick tour=94 of the material later covered more slowly throughout the=
 book. It also introduces most of the Lojban words used to discuss Lojban g=
rammar.
+<p>
+<ex "John and Sam">  XE "John and Sam: example=93  <ex "father">  XE "fath=
er: example=93  <ex "hits">  XE "hits: example=93  <ex "taller">  XE "talle=
r: example=93  Let us consider John and Sam and three statements about them=
:
+<p>
+<pre><a name=3De1d1>1.1)	John is the father of Sam.
+<a name=3De1d2>1.2)	John hits Sam.
+<a name=3De1d3>1.3)	John is taller than Sam.
+</pre><cx "relationship, active/static/attributive compared">  XE "relatio=
nship: active/static/attributive compared"   <cx "predication, as a relatio=
nship">  XE "predication: as a relationship"   <cx "bridi, compared with pr=
edication">  XE "bridi: compared with predication"   <cx "predication, comp=
ared with bridi">  XE "predication: compared with bridi"   <cx "brivla, rel=
ation to bridi">  XE "brivla: relation to bridi"   <cx "sumti, relation wit=
h bridi">  XE "sumti: relation with bridi"   These examples all describe re=
lationships between John and Sam. However, in English, we use the noun =93f=
ather=94 to describe a static relationship in <a href=3D#e1d1>Example 1.1</=
a>, the verb =93hits=94 to describe an active relationship in <a href=3D#e1=
d2>Example 1.2</a>, and the adjective =93taller=94 to describe an attributi=
ve relationship in <a href=3D#e1d3>Example 1.3</a>. In Lojban we make no su=
ch grammatical distinctions; these three sentences, when expressed in Lojba=
n, are structurally identical. The same part of speech is used to represent=
 the relationship. In formal logic this whole structure is called a =93pred=
ication=94; in Lojban it is called a =93bridi=94, and the central part of s=
peech is the =93selbri=94. Logicians refer to the things thus related as =
=93arguments=94, while Lojbanists call them =93sumti=94. These Lojban terms=
 will be used for the rest of the book.
+<p>
+bridi (predicate)
+|
+
+
+
+
+John
+is the father of
+Sam
+
+
+
+|
+sumti
+|
+selbri
+|
+sumti (argument)
+</pre><ex "give">  XE "give: example=93  In a relationship, there are a de=
finite number of things being related. In English, for example, =93give=94 =
has three places: the donor, the recipient and the gift. For example:
+<p>
+<pre><a name=3De1d4>1.4)	John gives Sam the book.
+</pre>and
+<p>
+<pre><a name=3De1d5>1.5)	Sam gives John the book.
+</pre>mean two different things because the relative positions of =93John=
=94 and =93Sam=94 have been switched.  <p>
+Further,
+<p>
+<pre><a name=3De1d6>1.6)	The book gives John Sam.
+</pre>seems strange to us merely because the places are being filled by un=
orthodox arguments. The relationship expressed by =93give=94 has not change=
d.
+<p>
+<cx "place structure, definition of">  XE "place structure: definition"   =
In Lojban, each selbri has a specified number and type of arguments, known =
collectively as its =93place structure=94. The simplest kind of selbri cons=
ists of a single root word, called a =93gismu=94, and the definition in a d=
ictionary gives the place structure explicitly. The primary task of constru=
cting a Lojban sentence, after choosing the relationship itself, is decidin=
g what you will use to fill in the sumti places.
+<p>
+This book uses the Lojban terms =93bridi=94, =93sumti=94, and =93selbri=94=
, because it is best to come to understand them independently of the Englis=
h associations of the corresponding words, which are only roughly similar i=
n meaning anyhow.
+<p>
+<ex "underlines">  XE "underlines: example=93  <cx "notation conventions, =
for Quick Tour chapter">  XE "notation conventions: for Quick Tour chapter"=
   <cx "      notation convention for Quick Tour chapter">  XE  selma=92o"u=
nderscore notation for Quick Tour chapter"  cx "=3D notation convention for=
 Quick Tour chapter">  XE  selma=92o"double underscore notation convention =
for Quick Tour chapter"  The Lojban examples in this chapter (but not in th=
e rest of the book) use a single underline (     ) under each sumti, and a =
double underline (     ) under each selbri, to help you to tell them apart.
+<p>
+<a name=3Ds2><h3>Pronunciation</h3>
+<p>
+<cx "pronunciation (quick-tour version)">  XE "pronunciation: quick-tour v=
ersion"   Detailed pronunciation and spelling rules are given in <a href=3D=
chap3.html>Chapter 3</a>, but what follows will keep the reader from going =
too far astray while digesting this chapter.
+<p>
+<cx "vowels, pronunciation of (quick-tour version)">  XE "vowels: pronunci=
ation of, quick-tour version"   Lojban has six recognized vowels: =93a=94, =
=93e=94, =93i=94, =93o=94, =93u=94 and =93y=94. The first five are roughly =
pronounced as =93a=94 as in =93father=94, =93e=94 as in =93let=94, =93i=94 =
as in =93machine=94, =93o=94 as in =93dome=94 and =93u=94 as in =93flute=94=
. =93y=94 is pronounced as the sound called =93schwa=94, that is, as the un=
stressed =93a=94 as in =93about=94 or =93around=94.
+<p>
+<cx "consonants, pronunciation of (quick-tour version)">  XE "consonants: =
pronunciation of, quick-tour version"   Twelve consonants in Lojban are pro=
nounced more or less as their counterparts are in English: =93b=94, =93d=94=
, =93f=94, =93k=94, =93l=94, =93m=94, =93n=94, =93p=94, =93r=94, =93t=94, =
=93v=94 and =93z=94. The letter =93c=94, on the other hand is pronounced as=
 the =93sh=94 in =93hush=94, while =93j=94 is its voiced counterpart, the s=
ound of the =93s=94 in =93pleasure=94. =93g=94 is always pronounced as it i=
s in =93gift=94, never as in =93giant=94. =93s=94 is as in =93sell=94, neve=
r as in =93rose=94. The sound of =93x=94 is not found in English in normal =
words. It is found as =93ch=94 in Scottish =93loch=94, as =93j=94 in Spanis=
h =93junta=94, and as =93ch=94 in German =93Bach=94; it also appears in the=
 English interjection =93yecchh!=94. It gets easier to say as you practice =
it. The letter =93r=94 can be trilled, but doesn't have to be.
+<p>
+<cx "diphthongs, pronunciation of (quick-tour version)">  XE "diphthongs: =
pronunciation of, quick-tour version"   The Lojban diphthongs =93ai=94, =93=
ei=94, =93oi=94, and =93au=94 are pronounced much as in the English words =
=93sigh=94, =93say=94, =93boy=94, and =93how=94. Other Lojban diphthongs be=
gin with an =93i=94 pronounced like English =93y=94 (for example, =93io=94 =
is pronounced =93yo=94) or else with a =93u=94 pronounced like English =93w=
=94 (for example, =93ua=94 is pronounced =93wa=94).
+<p>
+<cx "apostrophe (quick-tour version)">  XE "apostrophe: quick-tour version=
"   <cx "comma (quick-tour version)">  XE "comma: quick-tour version"   <cx=
 "period (quick-tour version)">  XE "period: quick-tour version"   Lojban a=
lso has three =93semi-letters=94: the period, the comma and the apostrophe.=
 The period represents a glottal stop or a pause; it is a required stoppage=
 of the flow of air in the speech stream. The apostrophe sounds just like t=
he English letter =93h=94. Unlike a regular consonant, it is not found at t=
he beginning or end of a word, nor is it found adjacent to a consonant; it =
is only found between two vowels. The comma has no sound associated with it=
, and is used to separate syllables that might ordinarily run together. It =
is not used in this chapter.
+<p>
+<cx "stress (quick-tour version)">  XE "stress: quick-tour version"   Stre=
ss falls on the next to the last syllable of all words, unless that vowel i=
s =93y=94, which is never stressed; in such words the third-to-last syllabl=
e is stressed. If a word only has one syllable, then that syllable is not s=
tressed.
+<p>
+All Lojban words are pronounced as they are spelled: there are no silent l=
etters.
+
+<p>
+<a name=3Ds3><h3>Words that can act as sumti</h3>
+<p>
+<cx "pro-sumti (quick-tour version)">  XE "pro-sumti: quick-tour version" =
  Here is a short table of single words used as sumti. This table provides =
examples only, not the entire set of such words, which may be found in <a h=
ref=3Dchap7.html>Chapter 7 </a>.
+<p>
+<pre>mi	I/me, we/us
+do	you
+ti	this, these
+ta	that, those
+tu	that far away, those far away
+zo'e	unspecified value (used when a sumti is unimportant or obvious)
+</pre><p>Lojban sumti are not specific as to number (singular or plural), =
nor gender (masculine/feminine/neutral). Such distinctions can be optionall=
y added by methods that are beyond the scope of this chapter.
+<p>
+<cx "pointing cmavo (quick-tour version)">  XE "pointing cmavo: quick-tour=
 version"   The cmavo =93ti=94, =93ta=94, and =93tu=94 refer to whatever th=
e speaker is pointing at, and should not be used to refer to things that ca=
nnot in principle be pointed at.
+<p>
+<cx "names (quick-tour version)">  XE "names: quick-tour version"   Names =
may also be used as sumti, provided they are preceded with the word =93la=
=94:
+<p>
+<pre>la meris.	the one/ones named Mary
+la djan.	the one/ones named John
+</pre><p>Other Lojban spelling versions are possible for names from other =
languages, and there are restrictions on which letters may appear in Lojban=
 names: see <a href=3Dchap6.html>Chapter 6 </a>for more information.
+<p>
+<a name=3Ds4><h3>Some words used to indicate selbri relations</h3>
+<p>
+<cx "selbri list for quick tour">  XE "selbri list for quick tour"   Here =
is a short table of some words used as Lojban selbri in this chapter:
+<p>
+<pre>vecnu	x1 (seller) sells x2 (goods) to x3 (buyer) for x4 (price)
+tavla	x1 (talker) talks to x2 (audience) about x3 (topic) in language x4
+sutra	x1 (agent) is fast at doing x2 (action)
+blari'o	x1 (object/light source) is blue-green
+melbi	x1 (object/idea) is beautiful to x2 (observer) by standard x3
+cutci	x1 is a shoe/boot for x2 (foot) made of x3 (material)
+bajra	x1 runs on x2 (surface) using x3 (limbs) in manner x4 (gait)
+klama	x1 goes/comes to x2 (destination) from x3 (origin point) via x4 (rou=
te) using x5 (means of transportation)
+pluka	x1 pleases/is pleasing to x2 (experiencer) under conditions x3
+gerku	x1 is a dog of breed x2
+kurji	x1 takes care of x2
+kanro	x1 is healthy by standard x2
+stali	x1 stays/remains with x2
+zarci	x1 is a market/store/shop selling x2 (products) operated by x3 (stor=
ekeeper)
+</pre><cx "x1, notation convention (quick-tour version)">  XE "x1: notatio=
n convention, quick-tour version"   Each selbri (relation) has a specific r=
ule that defines the role of each sumti in the bridi, based on its position=
. In the table above, that order was expressed by labeling the sumti positi=
ons as x1, x2, x3, x4, and x5.
+<p>
+<cx "words not in the dictionary">  XE "words not in the dictionary"   Lik=
e the table in <a href=3D#s3>Section 3</a>, this table is far from complete=
: in fact, no complete table can exist, because Lojban allows new words to =
be created (in specified ways) whenever a speaker or writer finds the exist=
ing supply of words inadequate. This notion is a basic difference between L=
ojban (and some other languages such as German and Chinese) and English; in=
 English, most people are very leery of using words that =93aren't in the d=
ictionary=94. Lojbanists are encouraged to invent new words; doing so is a =
major way of participating in the development of the language. <a href=3Dch=
ap4.html>Chapter 4 </a>explains how to make new words, and <a href=3Dchap12=
.html>Chapter 12 </a>explains how to give them appropriate meanings.
+<p>
+<a name=3Ds5><h3>Some simple Lojban bridi</h3>
+<p>
+<cx "bridi (quick-tour version)">  XE "bridi: quick-tour version"   Let's =
look at a simple Lojban bridi. The place structure of the gismu =93tavla=94=
 is
+<p>
+<pre><a name=3De5d1>5.1)	x1 talks to x2 about x3 in language x4
+</pre>where the =93x=94es with following numbers represent the various arg=
uments that could be inserted at the given positions in the English sentenc=
e. For example:
+<p>
+<ex "engineering">  XE "engineering: example=93  <pre><a name=3De5d2>5.2)	=
John talks to Sam about engineering in Lojban.
+</pre>has =93John=94 in the x1 place, =93Sam=94 in the x2 place, =93engine=
ering=94 in the x3 place, and =93Lojban=94 in the x4 place, and could be pa=
raphrased:
+<p>
+<pre><a name=3De5d3>5.3)	Talking is going on,
+		with speaker John
+		and listener Sam
+		and subject matter engineering
+		and language Lojban.
+</pre><p>The Lojban bridi corresponding to <a href=3D#e5d1>Example 5.1 </a=
>will have the form
+<p>
+<lx "tavla">  XE "tavla"   <pre><a name=3De5d4>5.4)	x1 [cu] tavla x2 x3 x4
+</pre><cx "cu, omission of (quick-tour version)">  XE "cu: omission of, qu=
ick-tour version"   <cx "cu, use of (quick-tour version)">  XE "cu: use of,=
 quick-tour version"   <lx "cu (quick-tour version)">  XE "cu: quick-tour v=
ersion"   The word =93cu=94 serves as a separator between any preceding sum=
ti and the selbri. It can often be omitted, as in the following examples.
+<p>
+<pre><a name=3De5d5>5.5)	mi tavla do zo'e zo'e
+	I talk to you about something in some language.
+<a name=3De5d6>5.6)	do tavla mi ta zo'e
+	You talk to me about that thing in a language.
+<a name=3De5d7>5.7)	mi tavla zo'e tu ti
+	I talk to someone about that thing yonder in this language.
+</pre>(<a href=3D#e5d7>Example 5.7 </a>is a bit unusual, as there is no ea=
sy way to point to a language; one might point to a copy of this book, and =
hope the meaning gets across!)
+<p>
+<cx "ellipsis (quick-tour version)">  XE "ellipsis: quick-tour version"   =
<lx "zo'e (quick-tour version)">  XE "zo'e: quick-tour version"   When ther=
e are one or more occurrences of the cmavo =93zo'e=94 at the end of a bridi=
, they may be omitted, a process called =93ellipsis=94. <a href=3D#e5d5>Exa=
mple 5.5 </a>and <a href=3D#e5d6>Example 5.6 </a>may be expressed thus:
+<p>
+<pre><a name=3De5d8>5.8)	mi tavla do
+	I talk to you (about something in some language).
+<a name=3De5d9>5.9)	do tavla mi ta
+	You talk to me about that thing (in some language).
+</pre><p>Note that <a href=3D#e5d7>Example 5.7 </a>is not subject to ellip=
sis by this direct method, as the =93zo'e=94 in it is not at the end of the=
 bridi.
+<p>
+<a name=3Ds6><h3>Variant bridi structure</h3>
+<p>
+<cx "sumti placement, variant (quick-tour version)">  XE "sumti placement:=
 variant, quick-tour version"   Consider the sentence
+<p>
+6.1)	mi	[cu]	vecnu	ti	ta	zo=92e
+	seller-x1
+sells	goods-sold-x2	buyer-x3	price-x4
+	I
+sell	this	to that	for some price.
+	I sell this-thing/these-things to that-buyer/those-buyers.
+(The price is obvious or unimportant.)
+</pre><a href=3D#e6d1>Example 6.1 </a>has one sumti (the x1) before the se=
lbri. It is also possible to put more than one sumti before the selbri, wit=
hout changing the order of sumti:
+6.2)	mi	ti	[cu]	vecnu	ta
+	seller-x1	goods-sold-x2
+sells	buyer-x3
+	I	this
+sell	to that.
+	(translates as stilted or poetic English)
+I this thing do sell to that buyer.
+6.3)	mi	ti	ta	[cu]	vecnu
+
+seller-x1	goods-sold-x2	buyer-x3
+sells
+
+I	this	to that
+sell.
+
+(translates as stilted or poetic English)
+I this thing to that buyer do sell.
+<p>
+</pre><a href=3D#e6d{1}>Examples 6.1 </a>through <a href=3D#e6d3>6.3 </a>m=
ean the same thing. Usually, placing more than one sumti before the selbri =
is done for style or for emphasis on the sumti that are out-of-place from t=
heir normal position. (Native speakers of languages other than English may =
prefer such orders.)
+<p>
+<cx "observatives (quick-tour version)">  XE "observatives: quick-tour ver=
sion"   If there are no sumti before the selbri, then it is understood that=
 the x1 sumti value is equivalent to =93zo'e=94; i.e. unimportant or obviou=
s, and therefore not given. Any sumti after the selbri start counting from =
x2.
+6.4)	ta	[cu]	melbi
+	object/idea-x1
+is-beautiful (to someone by some standard)
+	That/Those
+is/are beautiful.
+	That is beautiful.
+
+	Those are beautiful.
+
+<p>
+</pre>when the x1 is omitted, becomes:
+6.5)	 _                _	[cu]	melbi
+	unspecified-x1
+is-beautiful (to someone by some standard)
+	Beautiful!
+
+	It=92s beautiful!
+
+<p>
+</pre><p>Omitting the x1 adds emphasis to the selbri relation, which has b=
ecome first in the sentence. This kind of sentence is termed an observative=
, because it is often used when someone first observes or takes note of the=
 relationship, and wishes to quickly communicate it to someone else. Common=
ly understood English observatives include =93Smoke!=94 upon seeing smoke o=
r smelling the odor, or =93Car!=94 to a person crossing the street who migh=
t be in danger. Any Lojban selbri can be used as an observative if no sumti=
 appear before the selbri.
+<p>
+The word =93cu=94 does not occur in an observative; =93cu=94 is a separato=
r, and there must be a sumti before the selbri that needs to be kept separa=
te for =93cu=94 to be used. With no sumti preceding the selbri, =93cu=94 is=
 not permitted. Short words like =93cu=94 which serve grammatical functions=
 are called =93cmavo=94 in Lojban.
+<p>
+<a name=3Ds7><h3>Varying the order of sumti</h3>
+<p>
+<cx "sumti reordering (quick-tour version)">  XE "sumti reordering: quick-=
tour version"   <lx "se (quick-tour version)">  XE "se: quick-tour version"=
   <lx "SE (quick-tour version)">  XE "SE selma=92o: quick-tour version"   =
For one reason or another you may want to change the order, placing one par=
ticular sumti at the front of the bridi. The cmavo =93se=94, when placed be=
fore the last word of the selbri, will switch the meanings of the first and=
 second sumti places.  So
+<p>
+<pre><a name=3De7d1>7.1)	mi tavla do ti
+	I talk to you about this.
+</pre>has the same meaning as
+<p>
+<pre><a name=3De7d2>7.2)	do se tavla mi ti
+	You are talked to by me about this.
+</pre><p><lx "te (quick-tour version)">  XE "te: quick-tour version"   The=
 cmavo =93te=94, when used in the same location, switches the meanings of t=
he first and the third sumti places.
+<p>
+<pre><a name=3De7d3>7.3)	mi tavla do ti
+	I talk to you about this.
+</pre>has the same meaning as
+<p>
+<pre><a name=3De7d4>7.4)	ti te tavla do mi
+	This is talked about to you by me.
+</pre><p>Note that only the first and third sumti have switched places; th=
e second sumti has remained in the second place.
+<p>
+<lx "ve (quick-tour version)">  XE "ve: quick-tour version"   <lx "xe (qui=
ck-tour version)">  XE "xe: quick-tour version"   The cmavo =93ve=94 and =
=93xe=94 switch the first and fourth sumti places, and the first and fifth =
sumti places, respectively. These changes in the order of places are known =
as =93conversions=94, and the =93se=94, =93te=94, =93ve=94, and =93xe=94 cm=
avo are said to convert the selbri.
+<p>
+More than one of these operators may be used on a given selbri at one time=
, and in such a case they are evaluated from left to right. However, in pra=
ctice they are used one at a time, as there are better tools for complex ma=
nipulation of the sumti places. See <a href=3Dchap5.html>Chapter 5 </a>for =
details.
+<p>
+<cx "passive voice">  XE "passive voice"   The effect is similar to what i=
n English is called the =93passive voice=94. In Lojban, the converted selbr=
i has a new place structure that is renumbered to reflect the place reversa=
l, thus having effects when such a conversion is used in combination with o=
ther constructs such as =93le selbri [ku]=94 (see <a href=3D#s10>Section 10=
</a>).
+<p>
+<a name=3Ds8><h3>The basic structure of longer utterances</h3>
+<p>
+<lx ".i (quick-tour version)">  XE "i: quick-tour version"   <lx "I (quick=
-tour version)">  XE "I: quick-tour version"   <lx "ni'o (quick-tour versio=
n)">  XE "ni'o: quick-tour version"   <lx "NIhO (quick-tour version)">  XE =
"NIhO selma=92o: quick-tour version"   People don't always say just one sen=
tence. Lojban has a specific structure for talk or writing that is longer t=
han one sentence. The entirety of a given speech event or written text is c=
alled an utterance. The sentences (usually, but not always, bridi) in an ut=
terance are separated by the cmavo =93ni'o=94 and =93.i=94. These correspon=
d to a brief pause (or nothing at all) in spoken English, and the various p=
unctuation marks like period, question mark, and exclamation mark in writte=
n English. These separators prevent the sumti at the beginning of the next =
sentence from being mistaken for a trailing sumti of the previous sentence.
+<p>
+The cmavo =93ni'o=94 separates paragraphs (covering different topics of di=
scussion). In a long text or utterance, the topical structure of the text m=
ay be indicated by multiple =93ni'o=94s, with perhaps =93ni'oni'oni'o=94 us=
ed to indicate a chapter, =93ni'oni'o=94 to indicate a section, and a singl=
e =93ni'o=94 to indicate a subtopic corresponding to a single English parag=
raph.
+<p>
+The cmavo =93.i=94 separates sentences. It is sometimes compounded with wo=
rds that modify the exact meaning (the semantics) of the sentence in the co=
ntext of the utterance. (The cmavo =93xu=94, discussed in <a href=3D#s1>Sec=
tion 1</a>.7, is one such word =97 it turns the sentence from a statement t=
o a question about truth.) When more than one person is talking, a new spea=
ker will usually omit the =93.i=94 even though she/he may be continuing on =
the same topic.
+<p>
+It is still O.K. for a new speaker to say the =93.i=94 before continuing; =
indeed, it is encouraged for maximum clarity (since it is possible that the=
 second speaker might merely be adding words onto the end of the first spea=
ker's sentence). A good translation for =93.i=94 is the =93and=94 used in r=
un-on sentences when people are talking informally: =93I did this, and then=
 I did that, and=A0<dots>=85</dots>, and=A0<dots>=85</dots>=94.
+<p>
+<a name=3Ds9><h3>tanru</h3>
+<p>
+<cx "tanru (quick-tour version)">  XE "tanru: quick-tour version"   When t=
wo gismu are adjacent, the first one modifies the second, and the selbri ta=
kes its place structure from the rightmost word. Such combinations of gismu=
 are called =93tanru=94. For example,
+<p>
+<pre><a name=3De9d1>9.1)	sutra tavla
+</pre>has the place structure
+<p>
+<ex "fast talker">  XE "fast talker: example=93  <pre><a name=3De9d2>9.2)	=
x1 is a fast type-of talker to x2 about x3 in language x4
+	x1 talks fast to x2 about x3 in language x4
+</pre><cx "tanru default grouping (quick-tour version)">  XE "tanru defaul=
t grouping: quick-tour version"   When three or more gismu are in a row, th=
e first modifies the second, and that combined meaning modifies the third, =
and that combined meaning modifies the fourth, and so on. For example
+<p>
+<ex "fast-talker shoe">  XE "fast-talker shoe: example=93  <pre><a name=3D=
e9d3>9.3)	sutra tavla cutci
+</pre>has the place structure
+<p>
+<pre><a name=3De9d4>9.4)	s1 is a fast-talker type of shoe worn by s2 of ma=
terial s3
+</pre><p>That is, it is a shoe that is worn by a fast talker rather than a=
 shoe that is fast and is also worn by a talker.
+<p>
+Note especially the use of =93type-of=94 as a mechanism for connecting the=
 English translations of the two or more gismu; this convention helps the l=
earner understand each tanru in its context. Creative interpretations are a=
lso possible, however:
+<p>
+<ex "runner shoe">  XE "runner shoe: example=93  <pre><a name=3De9d5>9.5)	=
bajra cutci
+	runner shoe
+</pre>most probably refers to shoes suitable for runners, but might be int=
erpreted in some imaginative instances as =93shoes that run (by themselves?=
)=94. In general, however, the meaning of a tanru is determined by the lite=
ral meaning of its components, and not by any connotations or figurative me=
anings. Thus
+<p>
+<pre><a name=3De9d6>9.6)	sutra tavla
+	fast-talker
+</pre>would not necessarily imply any trickery or deception, unlike the En=
glish idiom, and a
+<p>
+<ex "social butterfly">  XE "social butterfly: example=93  <ex "butterfly,=
 social">  XE "butterfly: social, example=93  <pre><a name=3De9d7>9.7)	jikc=
a toldi
+	social butterfly
+</pre><ex "Lepidoptera">  XE "Lepidoptera: example=93  must always be an i=
nsect with large brightly-colored wings, of the family <i>Lepidoptera</i>.
+<p>
+<cx "tanru, place structure of (quick-tour version)">  XE "tanru: place st=
ructure of, quick-tour version"   The place structure of a tanru is always =
that of the final component of the tanru. Thus, the following has the place=
 structure of =93klama=94:
+9.8)	mi	[cu]	sutra klama	la meris.
+	I
+quickly-go	to Mary.
+<p>
+</pre><cx "tanru conversion, effect on place structure (quick-tour version=
)">  XE "tanru conversion: effect on place structure, quick-tour version"  =
 With the conversion =93se klama=94 as the final component of the tanru, th=
e place structure of the entire selbri is that of =93se klama=94: the x1 pl=
ace is the destination, and the x2 place is the one who goes:
+9.9)	mi	[cu]	sutra se klama	la meris.
+	I
+quickly am-gone-to	by Mary.
+<p>
+</pre><cx "tanru, and conversion (quick-tour version)">  XE "tanru: and co=
nversion, quick-tour version"   The following example shows that there is m=
ore to conversion than merely switching places, though:
+9.10)	la tam.	[cu]	melbi tavla	la meris.
+	Tom
+beautifully-talks	to Mary.
+	Tom is a beautiful-talker to Mary.
+<p>
+</pre>has the place structure of =93tavla=94, but note the two distinct in=
terpretations.
+<p>
+Now, using conversion, we can modify the place structure order:
+9.11)	la meris.	[cu]	melbi se tavla	la tam.
+	Mary
+is beautifully-talked-to	by Tom.
+	Mary is a beautiful-audience for Tom.
+<p>
+</pre>and we see that the modification has been changed so as to focus on =
Mary's role in the bridi relationship, leading to a different set of possib=
le interpretations.
+<p>
+Note that there is no place structure change if the modifying term is conv=
erted, and so less drastic variation in possible meanings:
+9.12)	la tam.	[cu]	tavla melbi 	la meris.
+	Tom
+is talkerly-beautiful	to Mary.
+<p>
+9.13)	la tam.	[cu]	se tavla melbi 	la meris.
+	Tom
+is audiencely-beautiful	to Mary.
+
+</pre>and we see that the manner in which Tom is seen as beautiful by Mary=
 changes, but Tom is still the one perceived as beautiful, and Mary, the ob=
server of beauty.
+<p>
+<a name=3Ds10><h3>Description sumti</h3>
+<p>
+<cx "descriptions (quick-tour version)">  XE "descriptions: quick-tour ver=
sion"   <ex "talker">  XE "talker: example=93  Often we wish to talk about =
things other than the speaker, the listener and things we can point to. Let=
's say I want to talk about a talker other than =93mi=94. What I want to ta=
lk about would naturally fit into the first place of =93tavla=94. Lojban, i=
t turns out, has an operator that pulls this first place out of a selbri an=
d converts it to a sumti called a =93description sumti=94. The description =
sumti =93le tavla ku=94 means =93the talker=94, and may be used wherever an=
y sumti may be used.
+<p>
+For example,
+<p>
+<pre><a name=3De10d1>10.1)	mi tavla do le tavla ku
+</pre>means the same as
+<p>
+<pre><a name=3De10d2>10.2)	I talk to you about the talker
+</pre>where =93the talker=94 is presumably someone other than me, though n=
ot necessarily.
+<p>
+Similarly =93le sutra tavla ku=94 is =93the fast talker=94, and =93le sutr=
a te tavla ku=94 is =93the fast subject of talk=94 or =93the subject of fas=
t talk=94. Which of these related meanings is understood will depend on the=
 context in which the expression is used. The most plausible interpretation=
 within the context will generally be assumed by a listener to be the inten=
ded one.
+<p>
+In many cases the word =93ku=94 may be omitted. In particular, it is never=
 necessary in a description at the end of a sentence, so:
+10.3)	mi	tavla	do	le tavla
+	I	talk-to	you	about-the talker.
+<p>
+</pre>means exactly the same thing as <a href=3D#e10d1>Example 10.1</a>.=
=20
+<p>
+<cx "cu, need for (quick-tour version)">  XE "cu: need for, quick-tour ver=
sion"   There is a problem when we want to say =93The fast one is talking.=
=94 The =93obvious=94 translation =93le sutra tavla=94 turns out to mean =
=93the fast talker=94, and has no selbri at all. To solve this problem we c=
an use the word =93cu=94, which so far has always been optional, in front o=
f the selbri.
+<p>
+The word =93cu=94 has no meaning, and exists only to mark the beginning of=
 the selbri within the bridi, separating it from a previous sumti. It comes=
 before any other part of the selbri, including other cmavo like =93se=94 o=
r =93te=94. Thus:
+<p>
+<pre><a name=3De10d4>10.4)	le sutra tavla
+	The fast talker
+<a name=3De10d5>10.5)	le sutra cu tavla
+	The fast one is talking.
+<a name=3De10d6>10.6)	le sutra se tavla
+	The fast talked-to one
+<a name=3De10d7>10.7)	le sutra cu se tavla
+	The fast one is talked to.
+</pre><cx "ku (quick-tour version)">  XE "ku: quick-tour version"   <cx "K=
U (quick-tour version)">  XE "KU selma=92o: quick-tour version"   Consider =
the following more complex example, with two description sumti.
+10.8)	mi	[cu]	tavla	le vecnu [ku]	le blari=92o [ku]
+	I
+talk-to	the seller	about the blue-green-thing.
+<p>
+</pre><p>The sumti =93le vecnu=94 contains the selbri =93vecnu=94, which h=
as the =93seller=94 in the x1 place, and uses it in this sentence to descri=
be a particular =93seller=94 that the speaker has in mind (one that he or s=
he probably expects the listener will also know about). Similarly, the spea=
ker has a particular blue-green thing in mind, which is described using =93=
le=94 to mark =93blari'o=94, a selbri whose first sumti is something blue-g=
reen.
+<p>
+It is safe to omit both occurrences of =93ku=94 in <a href=3D#e10d8>Exampl=
e 10.8 </a>, and it is also safe to omit the =93cu=94.
+<p>
+<a name=3Ds11><h3>Examples of brivla</h3>
+<p>
+<cx "brivla, types of (quick-tour version)">  XE "brivla: types of, quick-=
tour version"   The simplest form of selbri is an individual word. A word w=
hich may by itself express a selbri relation is called a =93brivla=94. The =
three types of brivla are gismu (root words), lujvo (compounds), and fu'ivl=
a (borrowings from other languages). All have identical grammatical uses. S=
o far, most of our selbri have been gismu or tanru built from gismu.
+<p>
+<cx "gismu (quick-tour version)">  XE "gismu: quick-tour version"   gismu:
+11.1)	mi	[cu]	klama	ti	zo=92e	zo=92e	ta
+	Go-er
+goes	destination	origin	route	means.
+	I go here (to this) using that means (from somewhere via some route).
+<p>
+</pre><cx "lujvo (quick-tour version)">  XE "lujvo: quick-tour version"   =
lujvo:
+11.2)	ta	[cu]	blari=92o
+	That
+is-blue-green.
+<p>
+</pre><cx "fu'ivla (quick-tour version)">  XE "fu'ivla: quick-tour version=
"   fu'ivla:
+11.3)	ti	[cu]	djarspageti
+	This
+is-spaghetti.
+<p>
+</pre><cx "cmavo as selbri (quick-tour version)">  XE "cmavo as selbri: qu=
ick-tour version"   Some cmavo may also serve as selbri, acting as variable=
s that stand for another selbri. The most commonly used of these is =93go'i=
=94, which represents the main bridi of the previous Lojban sentence, with =
any new sumti or other sentence features being expressed replacing the prev=
iously expressed ones. Thus, in this context:
+11.4)	ta	[cu]	go=92i
+	That
+too/same-as-last selbri.
+	That (is spaghetti), too.
+<p>
+</pre><a name=3Ds12><h3>The sumti =93di'u=94 and =93la'e di'u=94</h3>
+<p>
+<cx "reference (quick-tour version)">  XE "reference: quick-tour version" =
  In English, I might say =93The dog is beautiful=94, and you might reply =
=93This pleases me.=94 How do you know what =93this=94 refers to? Lojban us=
es different expressions to convey the possible meanings of the English:
+<p>
+<ex "beautiful dog">  XE "beautiful dog: example=93  <pre><a name=3De12d1>=
12.1)	le gerku [ku] cu melbi
+	The dog is beautiful.
+</pre><p>The following three sentences all might translate as =93This plea=
ses me.=94
+<p>
+<cx "pleases">  XE "pleases"   <pre><a name=3De12d2>12.2)	ti [cu] pluka mi
+	This (the dog) pleases me.
+</pre><lx "di'u (quick-tour version)">  XE "di'u: quick-tour version"   <p=
re><a name=3De12d3>12.3)	di'u [cu] pluka mi
+	This (the last sentence) pleases me (perhaps because it is grammatical
+or sounds nice).
+</pre><lx "la'edi'u (quick-tour version)">  XE "la'edi'u: quick-tour versi=
on"   <pre><a name=3De12d4>12.4)	la'e di'u [cu] pluka mi
+	This (the meaning of the last sentence; i.e. that the dog is beautiful)=
=20
+pleases me.
+</pre><a href=3D#e12d4>Example 12.4 </a>uses one sumti to point to or refe=
r to another by inference. It is common to write =93la'edi'u=94 as a single=
 word; it is used more often than =93di'u=94 by itself.
+<p>
+<a name=3Ds13><h3>Possession</h3>
+<p>
+<cx "possession (quick-tour version)">  XE "possession: quick-tour version=
"   =93Possession=94 refers to the concept of specifying an object by sayin=
g who it belongs to (or with). A full explanation of Lojban possession is g=
iven in <a href=3Dchap8.html>Chapter 8</a>. A simple means of expressing po=
ssession, however, is to place a sumti representing the possessor of an obj=
ect within the description sumti that refers to the object: specifically, b=
etween the =93le=94 and the selbri of the description:
+13.1)	le mi gerku	cu	sutra
+	The of-me dog
+is fast.
+	My dog is fast.
+
+<p>
+</pre><cx "possession not ownership (quick-tour version)">  XE "possession=
 not ownership: quick-tour version"   In Lojban, possession doesn't necessa=
rily mean ownership: one may =93possess=94 a chair simply by sitting on it,=
 even though it actually belongs to someone else. English uses possession c=
asually in the same way, but also uses it to refer to actual ownership or e=
ven more intimate relationships: =93my arm=94 doesn't mean =93some arm I ow=
n=94 but rather =93the arm that is part of my body=94. Lojban has methods o=
f specifying all these different kinds of possession precisely and easily.
+<p>
+<a name=3Ds14><h3>Vocatives and commands</h3>
+<p>
+<cx "vocatives (quick-tour version)">  XE "vocatives: quick-tour version" =
  <lx "doi (quick-tour version)">  XE "doi: quick-tour version"   <lx "DOI =
(quick-tour version)">  XE "DOI selma=92o: quick-tour version"   You may ca=
ll someone's attention to the fact that you are addressing them by using =
=93doi=94 followed by their name. The sentence
+<p>
+<pre><a name=3De14d1>14.1)	doi djan.
+</pre>means =93Oh, John, I'm talking to you=94. It also has the effect of =
setting the value of =93do=94; =93do=94 now refers to =93John=94 until it i=
s changed in some way in the conversation. Note that <a href=3D#e14d1>Examp=
le 14.1 </a>is not a bridi, but it is a legitimate Lojban sentence neverthe=
less; it is known as a =93vocative phrase=94.
+<p>
+<lx "coi (quick-tour version)">  XE "coi: quick-tour version"   <lx "co'o =
(quick-tour version)">  XE "co'o: quick-tour version"   Other cmavo can be =
used instead of =93doi=94 in a vocative phrase, with a different significan=
ce. For example, the cmavo =93coi=94 means =93hello=94 and =93co'o=94 means=
 =93good-bye=94. Either word may stand alone, they may follow one another, =
or either may be followed by a pause and a name. (Vocative phrases with =93=
doi=94 do not need a pause before the name.)
+<p>
+<pre><a name=3De14d2>14.2)	coi. djan.
+	Hello, John.
+<a name=3De14d3>14.3)	co'o. djan.
+	Good-bye, John.
+</pre><cx "commands (quick-tour version)">  XE "commands: quick-tour versi=
on"   <cx "imperatives (quick-tour version)">  XE "imperatives: quick-tour =
version"   Commands are expressed in Lojban by a simple variation of the ma=
in bridi structure. If you say
+<p>
+<pre><a name=3De14d4>14.4)	do tavla
+	You are-talking.
+</pre>you are simply making a statement of fact. In order to issue a comma=
nd in Lojban, substitute the word =93ko=94 for =93do=94. The bridi
+<p>
+<ex "Talk!">  XE "Talk!: example=93  <lx "ko (quick-tour version)">  XE "k=
o: quick-tour version"   <pre><a name=3De14d5>14.5)	ko tavla
+</pre>instructs the listener to do whatever is necessary to make <a href=
=3D#e14d4>Example 14.4 true; it means =93Talk!=94  Other examples:
+<p>
+<pre><a name=3De14d6>14.6)	ko sutra
+	Be fast!
+</pre><p>The =93ko=94 need not be in the x1 place, but rather can occur an=
ywhere a sumti is allowed, leading to possible Lojban commands that are ver=
y unlike English commands:
+<p>
+<pre><a name=3De14d7>14.7)	mi tavla ko
+	Be talked to by me.
+	Let me talk to you.
+</pre><p>The cmavo =93ko=94 can fill any appropriate sumti place, and can =
be used as often as is appropriate for the selbri:
+<p>
+<pre><a name=3De14d8>14.8)	ko kurji ko
+</pre>and
+<p>
+<ex "Take care!">  XE "Take care!: example=93  <pre><a name=3De14d9>14.9)	=
ko ko kurji
+</pre>both mean =93You take care of you=94 and =93Be taken care of by you=
=94, or to put it colloquially, =93Take care of yourself=94.
+<p>
+<a name=3Ds15><h3>Questions</h3>
+<p>
+<cx "questions (quick-tour version)">  XE "questions: quick-tour version" =
  There are many kinds of questions in Lojban: full explanations appear in =
<a href=3Dchap19.html>Chapter 19 </a>and in various other chapters througho=
ut the book. In this chapter, we will introduce three kinds: sumti question=
s, selbri questions, and yes/no questions.
+<p>
+<cx "questions (quick-tour version)">  XE "questions: quick-tour version" =
  <cx "sumti questions (quick-tour version)">  XE "sumti questions: quick-t=
our version"   <lx "ma (quick-tour version)">  XE "ma: quick-tour version" =
  The cmavo =93ma=94 is used to create a sumti question: it indicates that =
the speaker wishes to know the sumti which should be placed at the location=
 of the =93ma=94 to make the bridi true. It can be translated as =93Who?=94=
 or =93What?=94 in most cases, but also serves for =93When?=94, =93Where?=
=94, and =93Why?=94 when used in sumti places that express time, location, =
or cause. For example:
+15.1)	ma	tavla	do	mi
+	Who?	talks	to-you	about-me.
+	Who is talking to you about me?
+
+<p>
+</pre><p>The listener can reply by simply stating a sumti:
+<p>
+<pre><a name=3De15d2>15.2)	la djan.
+	John (is talking to you about me).
+
+</pre><p>Like =93ko=94, =93ma=94 can occur in any position where a sumti i=
s allowed, not just in the first position:
+15.3)	do	[cu]	tavla	ma
+	You
+talk	to what/whom?
+<p>
+</pre><p>A =93ma=94 can also appear in multiple sumti positions in one sen=
tence, in effect asking several questions at once.
+15.4)	ma	[cu]	tavla	ma
+	What/Who
+talks	to what/whom?
+<p>
+</pre><cx "separate questions (quick-tour version)">  XE "separate questio=
ns: quick-tour version"   The two separate =93ma=94 positions ask two separ=
ate questions, and can therefore be answered with different values in each =
sumti place.
+<p>
+<cx "selbri questions (quick-tour version)">  XE "selbri questions: quick-=
tour version"   <cx "bridi questions (quick-tour version)">  XE "bridi ques=
tions: quick-tour version"   <lx "mo (quick-tour version)">  XE "mo: quick-=
tour version"   The cmavo =93mo=94 is the selbri analogue of =93ma=94. It a=
sks the respondent to provide a selbri that would be a true relation if ins=
erted in place of the =93mo=94:
+15.5)	do	[cu]	mo
+	You
+are-what/do-what?
+<p>
+</pre><p>A =93mo=94 may be used anywhere a brivla or other selbri might. K=
eep this in mind for later examples. Unfortunately, by itself, =93mo=94 is =
a very non-specific question. The response to the question in <a href=3D#e1=
5d5>Example 15.5 </a>could be:
+<p>
+<pre><a name=3De15d6>15.6)	mi [cu] melbi
+	I am beautiful.
+</pre>or:
+<p>
+<pre><a name=3De15d7>15.7)	mi [cu] tavla
+	I talk.
+</pre><cx "speaker-listener cooperation">  XE "speaker-listener cooperatio=
n"   Clearly, =93mo=94 requires some cooperation between the speaker and th=
e respondent to ensure that the right question is being answered. If contex=
t doesn't make the question specific enough, the speaker must ask the quest=
ion more specifically using a more complex construction such as a tanru (se=
e <a href=3D#s9>Section 9</a>).
+<p>
+It is perfectly permissible for the respondent to fill in other unspecifie=
d places in responding to a =93mo=94 question. Thus, the respondent in <a h=
ref=3D#e15d7>Example 15.7 </a>could have also specified an audience, a topi=
c, and/or a language in the response.
+<p>
+<cx "yes/no questions (quick-tour version)">  XE "yes/no questions: quick-=
tour version"   Finally, we must consider questions that can be answered =
=93Yes=94 or =93No=94, such as
+<p>
+<pre><a name=3De15d8>15.8)	Are you talking to me?
+</pre><p>Like all yes-or-no questions in English, <a href=3D#e15d8>Example=
 15.8 </a>may be reformulated as
+<p>
+<pre><a name=3De15d9>15.9)	Is it true that you are talking to me?
+</pre><lx "xu (quick-tour version)">  XE "xu: quick-tour version"   In Loj=
ban we have a word that asks precisely that question in precisely the same =
way. The cmavo =93xu=94, when placed in front of a bridi, asks whether that=
 bridi is true as stated. So
+15.10)	xu	do	tavla	mi
+	Is-it-true-that	you	are-talking	to-me?
+<p>
+</pre>is the Lojban translation of <a href=3D#e15d8>Example 15.8</a>.
+<p>
+<cx "go'i with xu (quick-tour version)">  XE "go'i with xu: quick-tour ver=
sion"   <cx "affirmative answer (quick-tour version)">  XE "affirmative ans=
wer: quick-tour version"   The answer =93Yes=94 may be given by simply rest=
ating the bridi without the =93xu=94 question word. Lojban has a shorthand =
for doing this with the word =93go'i=94, mentioned in <a href=3D#s11>Sectio=
n 11</a>. Instead of a negative answer, the bridi may be restated in such a=
 way as to make it true. If this can be done by substituting sumti, it may =
be done with =93go'i=94 as well. For example:
+<p>
+<ex "healthy">  XE "healthy: example=93  <pre><a name=3De15d11>15.11)	xu d=
o kanro
+	Are you healthy?
+</pre>can be answered with
+<p>
+<pre><a name=3De15d12>15.12)	mi kanro
+	I am healthy.
+</pre>or
+<p>
+<pre><a name=3De15d13>15.13)	go'i
+	I am healthy.
+</pre>(Note that =93do=94 to the questioner is =93mi=94 to the respondent.=
)
+<p>
+or
+<p>
+<pre><a name=3De15d14>15.14)	le tavla cu kanro
+	The talker is healthy.
+</pre>or
+<p>
+<pre><a name=3De15d15>15.15)	le tavla cu go'i
+	The talker is healthy.
+</pre><lx "nago'i (quick-tour version)">  XE "nago'i: quick-tour version" =
  <cx "negative answer (quick-tour version)">  XE "negative answer: quick-t=
our version"   A general negative answer may be given by =93na go'i=94. =93=
na=94 may be placed before any selbri (but after the =93cu=94). It is equiv=
alent to stating =93It is not true that=A0<dots>=85</dots>=94 before the br=
idi. It does not imply that anything else is true or untrue, only that that=
 specific bridi is not true. More details on negative statements are availa=
ble in <a href=3Dchap15.html>Chapter 15</a>.
+<p>
+<a name=3Ds16><h3>Indicators</h3>
+<p>
+<cx "indicators (quick-tour version)">  XE "indicators: quick-tour version=
"   <cx "attitudinal indicators (quick-tour version)">  XE "attitudinal ind=
icators: quick-tour version"   <cx "interjections (quick-tour version)">  X=
E "interjections: quick-tour version"   Different cultures express emotions=
 and attitudes with a variety of intonations and gestures that are not usua=
lly included in written language. Some of these are available in some langu=
ages as interjections (i.e. Aha!, Oh no!, Ouch!, Aahh!, etc.), but they var=
y greatly from culture to culture.
+<p>
+Lojban has a group of cmavo known as =93attitudinal indicators=94 which sp=
ecifically covers this type of commentary on spoken statements. They are bo=
th written and spoken, but require no specific intonation or gestures. Gram=
matically they are very simple: one or more attitudinals at the beginning o=
f a bridi apply to the entire bridi; anywhere else in the bridi they apply =
to the word immediately to the left. For example:
+<p>
+<lx "UI (quick-tour version)">  XE "UI selma=92o: quick-tour version"   <p=
re><a name=3De16d1>16.1)	.ie mi [cu] klama
+	Agreement! I go.
+	Yep! I'll go.
+<a name=3De16d2>16.2)	.ei mi [cu] klama
+	Obligation! I go.
+	I should go.
+
+
+<a name=3De16d3>16.3)	mi [cu] klama le melbi .ui [ku]
+	I go to the beautiful-thing (and I am happy because it is the beautiful
+thing I'm going to).
+</pre><cx "discursives (quick-tour version)">  XE "discursives: quick-tour=
 version"   <cx "metalinguistic words (quick-tour version)">  XE "metalingu=
istic words: quick-tour version"   <cx "but/and equivalence">  XE "but/and =
equivalence"   Not all indicators indicate attitudes. Discursives, another =
group of cmavo with the same grammatical rules as attitudinal indicators, a=
llow free expression of certain kinds of commentary about the main utteranc=
es. Using discursives allows a clear separation of these so-called =93metal=
inguistic=94 features from the underlying statements and logical structure.=
 By comparison, the English words =93but=94 and =93also=94, which discursiv=
ely indicate contrast or an added weight of example, are logically equivale=
nt to =93and=94, which does not have a discursive content. The average Engl=
ish-speaker does not think about, and may not even realize, the paradoxical=
 idea that =93but=94 basically means =93and=94.
+<p>
+<pre><a name=3De16d4>16.4)	mi [cu] klama  .i do [cu] stali
+	I go.  You stay.
+<a name=3De16d5>16.5)	mi [cu] klama  .i ji'a do [cu] stali
+	I go.  In addition, you stay.  (added weight)
+<a name=3De16d6>16.6)	mi [cu] klama  .i ku'i do [cu] stali
+	I go.  However, you stay.  (contrast)
+</pre><cx "evidentials (quick-tour version)">  XE "evidentials: quick-tour=
 version"   Another group of indicators are called =93evidentials=94. Evide=
ntials show the speaker's relationship to the statement, specifically how t=
he speaker came to make the statement. These include =93za'a=94 (I directly=
 observe the relationship), =93pe'i=94 (I believe that the relationship hol=
ds), =93ru'a=94 (I postulate the relationship), and others. Many American I=
ndian languages use this kind of words.
+<p>
+<pre><a name=3De16d7>16.7)	pe'i do [cu] melbi
+	I opine!  You are beautiful.
+<a name=3De16d8>16.8)	za'a do [cu] melbi
+	I directly observe!  You are beautiful.
+</pre><a name=3Ds17><h3>Tenses</h3>
+<p>
+<cx "tenses (quick-tour version)">  XE =93tense: quick-tour version"   <cx=
 "time tenses (quick-tour version)">  XE "time tenses: quick-tour version" =
  In English, every verb is tagged for the grammatical category called tens=
e: past, present, or future. The sentence
+<p>
+<pre><a name=3De17d1>17.1)	John went to the store
+</pre>necessarily happens at some time in the past, whereas
+<p>
+<pre><a name=3De17d2>17.2)	John is going to the store
+</pre>is necessarily happening right now.
+<p>
+<cx "sentences, tenseless (quick-tour version)">  XE "sentences: tenseless=
, quick-tour version"   The Lojban sentence
+<p>
+17.3)	la djan.	[cu]	klama	le zarci
+	John
+goes/went/will-go	to-the store.
+</pre>serves as a translation of either <a href=3D#e17d1>Example 17.1 </a>=
or <a href=3D#e17d2>Example 17.2</a>, and of many other possible English se=
ntences as well. It is not marked for tense, and can refer to an event in t=
he past, the present or the future. This rule does not mean that Lojban has=
 no way of representing the time of an event. A close translation of <a hre=
f=3D#e17d1>Example 17.1 </a>would be:
+17.4)	la djan.	pu klama	le zarci
+	John	[past] goes	to-the store.
+<p>
+</pre>where the tag =93pu=94 forces the sentence to refer to a time in the=
 past. Similarly,
+17.5)	la djan.	ca klama	le zarci
+	John	[present] goes	to-the store.
+<p>
+</pre>necessarily refers to the present, because of the tag =93ca=94. Tags=
 used in this way always appear at the very beginning of the selbri, just a=
fter the =93cu=94, and they may make a =93cu=94 unnecessary, since tags can=
not be absorbed into tanru. Such tags serve as an equivalent to English ten=
ses and adverbs. In Lojban, tense information is completely optional. If un=
specified, the appropriate tense is picked up from context.
+<p>
+<cx "space tenses (quick-tour version)">  XE "space tenses: quick-tour ver=
sion"   Lojban also extends the notion of =93tense=94 to refer not only to =
time but to space. The following example uses the tag =93vu=94 to specify t=
hat the event it describes happens far away from the speaker:
+17.6)	do	vu vecnu	zo=92e
+	You	yonder sell	something-unspecified.
+<p>
+</pre><p>In addition, tense tags (either for time or space) can be prefixe=
d to the selbri of a description, producing a tensed sumti:
+17.7)	le pu bajra [ku]	cu	tavla
+	The earlier/former/past runner
+talked/talks.
+<p>
+</pre>(Since Lojban tense is optional, we don't know when he or she talks.=
)
+<p>
+Tensed sumti with space tags correspond roughly to the English use of =93t=
his=94 or =93that=94 as adjectives, as in the following example, which uses=
 the tag =93vi=94 meaning =93nearby=94:
+17.8)	le vi bajra [ku]	cu	tavla
+	The nearby runner
+talks.
+	This runner talks.
+<p>
+</pre><p>Do not confuse the use of =93vi=94 in <a href=3D#e17d8>Example 17=
.8 </a>with the cmavo =93ti=94, which also means =93this=94, but in the sen=
se of =93this thing=94.
+<p>
+<cx "sumti with tenses (quick-tour version)">  XE "sumti with tenses: quic=
k-tour version"   Furthermore, a tense tag can appear both on the selbri an=
d within a description, as in the following example (where =93ba=94 is the =
tag for future time):
+17.9)	le vi bajra [ku]	cu	ba klama
+	The here runner
+[future] goes.
+	The talker who is here will go.
+	This talker will go.
+<p>
+</pre><a name=3Ds18><h3>Lojban grammatical terms</h3>
+<p>
+<cx "grammatical terms (quick-tour version)">  XE "grammatical terms: quic=
k-tour version"   Here is a review of the Lojban grammatical terms used in =
this chapter, plus some others used throughout this book. Only terms that a=
re themselves Lojban words are included: there are of course many expressio=
ns like =93indicator=94 in <a href=3Dchap16.html>Chapter 16 </a>that are no=
t explained here. See the Index for further help with these.
+<p>
+<cx "bridi, definition (quick-tour version)">  XE "bridi: definition, quic=
k-tour version"   <dl compact><dt>bridi:  <dd>predication; the basic unit o=
f Lojban expression; the main kind of Lojban sentence; a claim that some ob=
jects stand in some relationship, or that some single object has some prope=
rty.
+</dl><cx "sumti, definition (quick-tour version)">  XE "sumti: definition,=
 quick-tour version"   <dl compact><dt>sumti:  <dd>argument; words identify=
ing something which stands in a specified relationship to something else, o=
r which has a specified property. See <a href=3Dchap6.html>Chapter 6</a>.
+</dl><cx "selbri, definition (quick-tour version)">  XE "selbri: definitio=
n, quick-tour version"   <dl compact><dt>selbri:  <dd>logical predicate; th=
e core of a bridi; the word or words specifying the relationship between th=
e objects referred to by the sumti. See <a href=3Dchap5.html>Chapter 5</a>.
+</dl><cx "cmavo, definition (quick-tour version)">  XE "cmavo: definition,=
 quick-tour version"   <dl compact><dt>cmavo:  <dd>one of the Lojban parts =
of speech; a short word; a structural word; a word used for its grammatical=
 function.
+</dl><cx "brivla, definition (quick-tour version)">  XE "brivla: definitio=
n, quick-tour version"   <dl compact><dt>brivla:  <dd>one of the Lojban par=
ts of speech; a content word; a predicate word; can function as a selbri; i=
s a gismu, a lujvo, or a fu'ivla. See <a href=3Dchap4.html>Chapter 4</a>.
+</dl><cx "gismu, definition (quick-tour version)">  XE "gismu: definition,=
 quick-tour version"   <dl compact><dt>gismu:  <dd>a root word; a kind of b=
rivla; has associated rafsi. See <a href=3Dchap4.html>Chapter 4</a>.
+</dl><cx "lujvo, definition (quick-tour version)">  XE "lujvo: definition,=
 quick-tour version"   <dl compact><dt>lujvo:  <dd>a compound word; a kind =
of brivla; may or may not appear in a dictionary; does not have associated =
rafsi. See <a href=3Dchap4.html>Chapter 4 </a>and <a href=3Dchap12.html>Cha=
pter 12</a>.
+</dl><cx "fu'ivla, definition (quick-tour version)">  XE "fu'ivla: definit=
ion, quick-tour version"   <dl compact><dt>fu'ivla:  <dd>a borrowed word; a=
 kind of brivla; may or may not appear in a dictionary; copied in a modifie=
d form from some non-Lojban language; usually refers to some aspect of cult=
ure or the natural world; does not have associated rafsi. See <a href=3Dcha=
p4.html>Chapter 4</a>.
+</dl><cx "rafsi, definition (quick-tour version)">  XE "rafsi: definition,=
 quick-tour version"   <dl compact><dt>rafsi:  <dd>a word fragment; one or =
more is associated with each gismu; can be assembled according to rules in =
order to make lujvo; not a valid word by itself. See <a href=3Dchap4.html>C=
hapter 4</a>.
+</dl><cx "tanru, definition (quick-tour version)">  XE "tanru: definition,=
 quick-tour version"   <dl compact><dt>tanru:  <dd>a group of two or more b=
rivla, possibly with associated cmavo, that form a selbri; always divisible=
 into two parts, with the first part modifying the meaning of the second pa=
rt (which is taken to be basic). See <a href=3Dchap5.html>Chapter 5</a>.
+</dl><cx "selma'o, definition (quick-tour version)">  XE "selma'o: definit=
ion, quick-tour version"   <dl compact><dt>selma'o:  <dd>a group of cmavo t=
hat have the same grammatical use (can appear interchangeably in sentences,=
 as far as the grammar is concerned) but differ in meaning or other usage. =
See <a </a>href=3Dchap20.html>Chapter 20</a>.
+
+</body></html>
+<img src=3Dchap3.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D4=
05>
+
+<h2>Chapter 3
+<br>
+The Hills Are Alive With The Sounds Of Lojban</h2>
+<h6>$Revision: 4.2 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>Orthography</h3>
+<p>
+<cx "isomorphism, audio-visual">  XE "isomorphism: audio-visual"   <cx "au=
dio-visual isomorphism">  XE "audio-visual isomorphism"   <cx "pronunciatio=
n, relation to orthography">  XE "pronunciation: relation to orthography"  =
 <cx "orthography, relation to pronunciation">  XE "orthography: relation t=
o pronunciation"   Lojban is designed so that any properly spoken Lojban ut=
terance can be uniquely transcribed in writing, and any properly written Lo=
jban can be spoken so as to be uniquely reproduced by another person. As a =
consequence, the standard Lojban orthography must assign to each distinct s=
ound, or phoneme, a unique letter or symbol. Each letter or symbol has only=
 one sound or, more accurately, a limited range of sounds that are permitte=
d pronunciations for that phoneme. Some symbols indicate stress (speech emp=
hasis) and pause, which are also essential to Lojban word recognition. In a=
ddition, everything that is represented in other languages by punctuation (=
when written) or by tone of voice (when spoken) is represented in Lojban by=
 words. These two properties together are known technically as =93audio-vis=
ual isomorphism=94.
+<p>
+<cx "Latin alphabet">  XE "Latin alphabet"   <cx "Lojban alphabet">  XE "L=
ojban alphabet"   <cx "alphabet, Lojban">  XE "alphabet: Lojban"   Lojban u=
ses a variant of the Latin (Roman) alphabet, consisting of the following le=
tters and symbols:
+<p>
+<dl compact><dt><dd>	' , . a b c d e f g i j k l m n o p r s t u v x y z
+</dl>omitting the letters =93h=94, =93q=94, and =93w=94.
+<p>
+<cx "alphabetic order">  XE "alphabetic order"   The alphabetic order give=
n above is that of the ASCII coded character set, widely used in computers.=
 By making Lojban alphabetical order the same as ASCII, computerized sortin=
g and searching of Lojban text is facilitated.
+<p>
+<cx "capital letters, use of">  XE "capital letters: use of"   <cx "stress=
, showing non-standard">  XE "stress: showing non-standard"   Capital lette=
rs are used only to represent non-standard stress, which can appear only in=
 the representation of Lojbanized names. Thus the English name =93Josephine=
=94, as normally pronounced, is Lojbanized as =93DJOsefin.=94, pronounced <=
font face=3D"SILDoulosIPA">[=C8dZo sE fin/]</font>. (See <a href=3D#s2>Sect=
ion 2 </a>for an explanation of the symbols within square brackets.) Techni=
cally, it is sufficient to capitalize the vowel letter, in this case =93O=
=94, but it is easier on the reader to capitalize the whole syllable.
+<p>
+Without the capitalization, the ordinary rules of Lojban stress would caus=
e the =93se=94 syllable to be stressed. Lojbanized names are meant to repre=
sent the pronunciation of names from other languages with as little distort=
ion as may be; as such, they are exempt from many of the regular rules of L=
ojban phonology, as will appear in the rest of this chapter.
+<p>
+<a name=3Ds2><h3>Basic phonetics</h3>
+<p>
+<cx "International Phonetic Alphabet (see also IPA)">  XE "International P=
honetic Alphabet (see also IPA)"   <cx "IPA">  XE "IPA"   <cx "phonetic alp=
habet">  XE "phonetic alphabet"   <cx "brackets, use in IPA notation">  XE =
"brackets: use in IPA notation"   Lojban pronunciations are defined using t=
he International Phonetic Alphabet, or IPA, a standard method of transcribi=
ng pronunciations. By convention, IPA transcriptions are always within squa=
re brackets: for example, the word =93cat=94 is pronounced (in General Amer=
ican pronunciation) <font face=3D"SILDoulosIPA">[kQt]</font>. <a href=3D#s1=
0>Section 10 </a>contains a brief explanation of the IPA characters used in=
 this chapter, with their nearest analogues in English, and will be especia=
lly useful to those not familiar with the technical terms used in describin=
g speech sounds.
+<p>
+<cx "pronunciation, standard">  XE "pronunciation: standard"   <cx "standa=
rd pronunciation">  XE "standard pronunciation"   The standard pronunciatio=
ns and permitted variants of the Lojban letters are listed in the table bel=
ow. The descriptions have deliberately been made a bit ambiguous to cover v=
ariations in pronunciation by speakers of different native languages and di=
alects. In all cases except =93r=94 the first IPA symbol shown represents t=
he preferred pronunciation; for =93r=94, all of the variations (and any oth=
er rhotic sound) are equally acceptable.
+<p>
+<cx "pronunciation, IPA for Lojban">  XE "pronunciation: IPA for Lojban"  =
 <cx "Lojban letters, IPA for pronouncing">  XE "Lojban letters: IPA for pr=
onouncing"   <cx "Lojban letters, list with IPA pronunciation">  XE "Lojban=
 letters: list with IPA pronunciation"   <pre>	Letter	IPA	Description
+	'	<font face=3D"SILDoulosIPA">[h]</font>	a unvoiced glottal spirant
+	,	=97	the syllable separator
+	.	<font face=3D"SILDoulosIPA">[/]</font>	a glottal stop or a pause
+	a	<font face=3D"SILDoulosIPA">[a]</font>, <font face=3D"SILDoulosIPA">[A]=
</font>	an open vowel
+	b	<font face=3D"SILDoulosIPA">[b]</font>	a voiced bilabial stop
+	c	<font face=3D"SILDoulosIPA">[S]</font>, <font face=3D"SILDoulosIPA">[=
=A7]</font>	an unvoiced coronal sibilant
+	d	<font face=3D"SILDoulosIPA">[d]</font>	a voiced dental/alveolar stop
+	e	<font face=3D"SILDoulosIPA">[E]</font>, <font face=3D"SILDoulosIPA">[e]=
</font>	a front mid vowel
+	f	<font face=3D"SILDoulosIPA">[f]</font>, <font face=3D"SILDoulosIPA">[=
=B8]</font>	an unvoiced labial fricative
+	g	<font face=3D"SILDoulosIPA">[g]</font>	a voiced velar stop
+	i	<font face=3D"SILDoulosIPA">[i]</font>	a front close vowel
+	j	<font face=3D"SILDoulosIPA">[Z]</font>,<font face=3D"SILDoulosIPA">[=BD=
]</font>	a voiced coronal sibilant
+	k	<font face=3D"SILDoulosIPA">[k]</font>	an unvoiced velar stop
+	l	<font face=3D"SILDoulosIPA">[l]</font>, <font face=3D"SILDoulosIPA">[l`=
]</font>	a voiced lateral approximant
+			(may be syllabic)
+	m	<font face=3D"SILDoulosIPA">[m]</font>, <font face=3D"SILDoulosIPA">[m`=
]</font>	a voiced bilabial nasal
+			(may be syllabic)
+	n	<font face=3D"SILDoulosIPA">[n]</font>, <font face=3D"SILDoulosIPA">[n`=
]</font>,	a voiced dental or velar nasal
+		<font face=3D"SILDoulosIPA">[N]</font>, <font face=3D"SILDoulosIPA">[N`]=
</font>	(may be syllabic)
+]</font>	a back mid vowel
+	p	<font face=3D"SILDoulosIPA">[p]</font>	an unvoiced bilabial stop
+	r	<font face=3D"SILDoulosIPA">[r]</font>, <font face=3D"SILDoulosIPA">[=
=A8]</font>, <font face=3D"SILDoulosIPA">[R]</font>, <font face=3D"SILDoulo=
sIPA">[{]</font>,	a rhotic sound
+		<font face=3D"SILDoulosIPA">[r`]</font>, <font face=3D"SILDoulosIPA">[=
=A8`]</font>, <font face=3D"SILDoulosIPA">[R`]</font>, <font face=3D"SILDou=
losIPA">[{`]</font>
+	s	<font face=3D"SILDoulosIPA">[s]</font>	an unvoiced alveolar sibilant
+	t	<font face=3D"SILDoulosIPA">[t]</font>	an unvoiced dental/alveolar stop
+	u	<font face=3D"SILDoulosIPA">[u]</font>	a back close vowel
+	v	<font face=3D"SILDoulosIPA">[v]</font>, <font face=3D"SILDoulosIPA">[B]=
</font>	a voiced labial fricative
+	x	<font face=3D"SILDoulosIPA">[x]</font>	an unvoiced velar fricative
+	y	<font face=3D"SILDoulosIPA">[=AB]</font>	a central mid vowel
+	z	<font face=3D"SILDoulosIPA">[z]</font>	a voiced alveolar sibilant
+</pre><cx "clarity of sounds">  XE "clarity of sounds"   <cx "sounds, clar=
ity of">  XE "sounds: clarity of"   The Lojban sounds must be clearly prono=
unced so that they are not mistaken for each other. Voicing and placement o=
f the tongue are the key factors in correct pronunciation, but other subtle=
 differences will develop between consonants in a Lojban-speaking community=
. At this point these are the only mandatory rules on the range of sounds.
+<p>
+<cx "rounded/unrounded vowels">  XE "rounded/unrounded vowels"   Note in p=
articular that Lojban vowels can be pronounced with either rounded or unrou=
nded lips; typically =93o=94 and =93u=94 are rounded and the others are not=
, as in English, but this is not a requirement; some people round =93y=94 a=
s well. Lojban consonants can be aspirated or unaspirated. Palatalizing of =
consonants, as found in Russian and other languages, is not generally accep=
table in pronunciation, though a following =93i=94 may cause it.
+<p>
+<cx "sounds, difficult">  XE "sounds: difficult"   <cx "sounds for letters=
, Lojban contrasted with English">  XE "sounds for letters: Lojban contrast=
ed with English"   The sounds represented by the letters =93c=94, =93g=94, =
=93j=94, =93s=94, and =93x=94 require special attention for speakers of Eng=
lish, either because they are ambiguous in the orthography of English (=93c=
=94, =93g=94, =93s=94), or because they are strikingly different in Lojban =
(=93c=94, =93j=94, =93x=94). The English =93c=94 represents three different=
 sounds, <font face=3D"SILDoulosIPA">[k] </font>in =93cat=94 and <font face=
=3D"SILDoulosIPA">[s] </font>in =93cent=94, as well as the <font face=3D"SI=
LDoulosIPA">[S] </font>of =93ocean=94. Similarly, English =93g=94 can repre=
sent <font face=3D"SILDoulosIPA">[g] </font>as in =93go=94, <font face=3D"S=
ILDoulosIPA">[dZ] </font>as in =93gentle=94, and <font face=3D"SILDoulosIPA=
">[Z] </font>as in the second =93g=94 in =93garage=94 (in some pronunciatio=
ns). English =93s=94 can be either <font face=3D"SILDoulosIPA">[s] </font>a=
s in =93cats=94, <font face=3D"SILDoulosIPA">[z] </font>as in =93cards=94, =
<font face=3D"SILDoulosIPA">[S] </font>as in =93tension=94, or <font face=
=3D"SILDoulosIPA">[Z] </font>as in =93measure=94. The sound of Lojban =93x=
=94 doesn't appear in most English dialects at all.
+<p>
+<cx "sounds, complex">  XE "sounds: complex"   <cx "=93ts=94-sound in Russ=
ian, representation in Lojban">  XE "ts-sound in Russian: representation in=
 Lojban"   <cx "=93ch=94-sound in English, representation in Lojban">  XE "=
ch-sound in English: representation in Lojban"   <cx "=93j=94-sound in Engl=
ish, representation in Lojban">  XE "j-sound in English: representation in =
Lojban"   There are two common English sounds that are found in Lojban but =
are not Lojban consonants: the =93ch=94 of =93church=94 and the =93j=94 of =
=93judge=94. In Lojban, these are considered two consonant sounds spoken to=
gether without an intervening vowel sound, and so are represented in Lojban=
 by the two separate consonants: =93tc=94 (IPA <font face=3D"SILDoulosIPA">=
[tS]</font>) and =93dj=94 (IPA <font face=3D"SILDoulosIPA">[dZ]</font>). In=
 general, whether a complex sound is considered one sound or two depends on=
 the language: Russian views =93ts=94 as a single sound, whereas English, F=
rench, and Lojban consider it to be a consonant cluster.
+<p>
+<a name=3Ds3><h3>The special Lojban characters</h3>
+<p>
+<cx "characters, special">  XE "characters: special"   The apostrophe, per=
iod, and comma need special attention. They are all used as indicators of a=
 division between syllables, but each has a different pronunciation, and ea=
ch is used for different reasons:
+<p>
+<cx "apostrophe, definition of">  XE "apostrophe: definition of"   <cx "=
=93 ' =94, definition (see also apostrophe)">  XE " ' symbol: definition (s=
ee also apostrophe)"   <cx "apostrophe, type of letter in word-formation"> =
 XE "apostrophe: type of letter in word-formation"   The apostrophe represe=
nts a phoneme similar to a short, breathy English =93h=94, (IPA <font face=
=3D"SILDoulosIPA">[h]</font>). The letter =93h=94 is not used to represent =
this sound for two reasons: primarily in order to simplify explanations of =
the morphology, but also because the sound is very common, and the apostrop=
he is a visually lightweight representation of it. The apostrophe sound is =
a consonant in nature, but is not treated as either a consonant or a vowel =
for purposes of Lojban morphology (word-formation), which is explained in <=
a href=3Dchap4.html>Chapter 4</a>. In addition, the apostrophe visually par=
allels the comma and the period, which are also used (in different ways) to=
 separate syllables.
+<p>
+<cx "apostrophe, purpose of">  XE "apostrophe: purpose of"   <cx "unvoiced=
 vowel glide, apostrophe as">  XE "unvoiced vowel glide: apostrophe as"   T=
he apostrophe is included in Lojban only to enable a smooth transitionsepar=
ation between vowels, while joining the vowels within a single word. In fac=
t, one way to think of the apostrophe is as representing a unvoiced vowel g=
lide.
+<p>
+<cx "apostrophe, variant of">  XE "apostrophe: variant of"   As a permitte=
d variant, any unvoiced fricative other than those already used in Lojban m=
ay be used to render the apostrophe: IPA <font face=3D"SILDoulosIPA">[T] </=
font>is one possibility. The convenience of the listener should be regarded=
 as paramount in deciding to use a substitute for <font face=3D"SILDoulosIP=
A">[h]</font>.
+<p>
+<cx "period, definition of">  XE "period: definition of"   <cx "glottal st=
op, as pause in Lojban">  XE "glottal stop: as pause in Lojban"   <cx "paus=
e, representation of in Lojban">  XE "pause: representation of in Lojban"  =
 The period represents a mandatory pause, with no specified length; a glott=
al stop (IPA <font face=3D"SILDoulosIPA">[/]</font>) is considered a pause =
of shortest length. A pause (or glottal stop) may appear between any two wo=
rds, and in certain cases =96 explained in detail in <a href=3Dchap4.html>C=
hapter 4 </a>=97 must occur. In particular, a word beginning with a vowel i=
s always preceded by a pause, and a word ending in a consonant is always fo=
llowed by a pause.
+<p>
+<cx "period, optional">  XE "period: optional"   Technically, the period i=
s an optional reminder to the reader of a mandatory pause that is dictated =
by the rules of the language; because these rules are unambiguous, a missin=
g period can be inferred from otherwise correct text. Periods are included =
only as an aid to the reader.
+<p>
+<cx "period, within a word">  XE "period: within a word"   A period also m=
ay be found apparently embedded in a word. When this occurs, such a written=
 string is not one word but two, written together to indicate that the writ=
er intends a unitary meaning for the compound. It is not really necessary t=
o use a space between words if a period appears.
+<p>
+<cx "comma, definition of">  XE "comma: definition of"   <cx "syllable bre=
ak, representation in Lojban">  XE "syllable break: representation in Lojba=
n"   <cx "syllable break, contrasted with pause">  XE "syllable break: cont=
rasted with pause"   <cx "pause, contrasted with syllable break">  XE "paus=
e: contrasted with syllable break"   The comma is used to indicate a syllab=
le break within a word, generally one that is not obvious to the reader. Su=
ch a comma is written to separate syllables, but indicates that there must =
be no pause between them, in contrast to the period. Between two vowels, a =
comma indicates that some type of glide may be necessary to avoid a pause t=
hat would split the two syllables into separate words. It is always legal t=
o use the apostrophe (IPA <font face=3D"SILDoulosIPA">[h]</font>) sound in =
pronouncing a comma. However, a comma cannot be pronounced as a pause or gl=
ottal stop between the two letters separated by the comma, because that pro=
nunciation would split the word into two words.
+<p>
+<cx "comma, main use of">  XE "comma: main use of"   <cx "comma, optional"=
>  XE "comma: optional"   Otherwise, a comma is usually only used to clarif=
y the presence of syllabic =93l=94, =93m=94, =93n=94, or =93r=94 (discussed=
 later). Commas are never required: no two Lojban words differ solely becau=
se of the presence or placement of a comma.
+<p>
+<ex "Old McDonald">  XE "Old McDonald: example=93  Here is a somewhat arti=
ficial example of the difference in pronunciation between periods, commas a=
nd apostrophes. In the English song about Old MacDonald's Farm, the vowel s=
tring written =93ee-i-ee-i-o=94 in English could be Lojbanized with periods=
 as:
+<p>
+<cx "period, example of">  XE "period: example of"   <pre><a name=3De3d1>3=
.1)	.i.ai.i.ai.o
+	<font face=3D"SILDoulosIPA">[/i /aj /i /aj /o]</font>
+	Ee! Eye! Ee! Eye! Oh!
+</pre><p>However, this would sound clipped, staccato, and unmusical compar=
ed to the English. Furthermore, although <a href=3D#e3d1>Example 3.1 </a>is=
 a string of meaningful Lojban words, as a sentence it makes very little se=
nse. (Note the use of periods embedded within the written word.)
+<p>
+If commas were used instead of periods, we could represent the English str=
ing as a Lojbanized name, ending in a consonant:
+<p>
+<cx "comma, example of">  XE "comma: example of"   <pre><a name=3De3d2>3.2=
)	.i,ai,i,ai,on.
+	<font face=3D"SILDoulosIPA">[/i jaj ji jaj jon/]</font>
+</pre><p>The commas represent new syllable breaks, but prohibit the use of=
 pauses or glottal stop. The pronunciation shown is just one possibility, b=
ut closely parallels the intended English pronunciation.
+<p>
+<cx "comma, variant of">  XE "comma: variant of"   However, the use of com=
mas in this way is risky to unambiguous interpretation, since the glides mi=
ght be heard by some listeners as diphthongs, producing something like
+<p>
+<pre><a name=3De3d3>3.3)	.i,iai,ii,iai,ion.
+</pre>which is technically a different Lojban name. Since the intent with =
Lojbanized names is to allow them to be pronounced more like their native c=
ounterparts, the comma is allowed to represent vowel glides or some non-Loj=
banic sound. Such an exception affects only spelling accuracy and the abili=
ty of a reader to replicate the desired pronunciation exactly; it will not =
affect the recognition of word boundaries.
+<p>
+<cx "apostrophe, as preferable over comma in names">  XE "apostrophe: as p=
referable over comma in names"   Still, it is better if Lojbanized names ar=
e always distinct. Therefore, the apostrophe is preferred in regular Lojban=
ized names that are not attempting to simulate a non-Lojban pronunciation p=
erfectly. (Perfection, in any event, is not really achievable, because some=
 sounds simply lack reasonable Lojbanic counterparts.)
+<p>
+If apostrophes were used instead of commas in <a href=3D#e3d2>Example 3.2<=
/a>, it would appear as:
+<p>
+<cx "apostrophe, example of">  XE "apostrophe: example of"   <pre><a name=
=3De3d4>3.4)	.i'ai'i'ai'on.
+	<font face=3D"SILDoulosIPA">[/i hai hi hai hon/]</font>
+</pre>which preserves the rhythm and length, if not the exact sounds, of t=
he original English.
+<p>
+<a name=3Ds4><h3>Diphthongs and syllabic consonants</h3>
+<p>
+<cx "diphthongs, definition of">  XE "diphthongs: definition of"   There e=
xist 16 diphthongs in the Lojban language. A diphthong is a vowel sound tha=
t consists of two elements, a short vowel sound and a glide, either a labia=
l (IPA <font face=3D"SILDoulosIPA">[w]</font>) or palatal (IPA <font face=
=3D"SILDoulosIPA">[j]</font>) glide, that either precedes (an on-glide) or =
follows (an off-glide) the main vowel. Diphthongs always constitute a singl=
e syllable.
+<p>
+<cx "consonants, contrasted with vowels">  XE "consonants: contrasted with=
 vowels"   <cx "vowels, contrasted with consonants">  XE "vowels: contraste=
d with consonants"   For Lojban purposes, a vowel sound is a relatively lon=
g speech-sound that forms the nucleus of a syllable. Consonant sounds are r=
elatively brief and normally require an accompanying vowel sound in order t=
o be audible. Consonants may occur at the beginning or end of a syllable, a=
round the vowel, and there may be several consonants in a cluster in either=
 position. Each separate vowel sound constitutes a distinct syllable; conso=
nant sounds do not affect the determination of syllables.
+<p>
+<cx "vowels, definition of">  XE "vowels: definition of"   The six Lojban =
vowels are =93a=94, =93e=94, =93i=94, =93o=94, =93u=94, and =93y=94. The fi=
rst five vowels appear freely in all kinds of Lojban words. The vowel =93y=
=94 has a limited distribution: it appears only in Lojbanized names, in the=
 Lojban names of the letters of the alphabet, as a glue vowel in compound w=
ords, and standing alone as a space-filler word (like English =93uh=94 or =
=93er=94).
+<p>
+The Lojban diphthongs are shown in the table below. (Variant pronunciation=
s have been omitted, but are much as one would expect based on the variant =
pronunciations of the separate vowel letters: =93ai=94 may be pronounced <f=
ont face=3D"SILDoulosIPA">[Aj]</font>, for example.)
+<p>
+<cx "diphthongs, IPA for">  XE "diphthongs: IPA for"   <cx "diphthongs, li=
st of">  XE "diphthongs: list of"   <pre>Letters		IPA	Description
+ai		<font face=3D"SILDoulosIPA">[aj]	</font> an open vowel with palatal of=
f-glide
+ei		<font face=3D"SILDoulosIPA">[Ej]	</font> a front mid vowel with palata=
l off-glide
+oi		<font face=3D"SILDoulosIPA">[oj]	</font> a back mid vowel with palatal=
 off-glide
+au		<font face=3D"SILDoulosIPA">[aw]	</font> an open vowel with labial off=
-glide
+ia		<font face=3D"SILDoulosIPA">[ja]	</font> an open vowel with palatal on=
-glide
+ie		<font face=3D"SILDoulosIPA">[jE]	</font> a front mid vowel with palata=
l on-glide
+ii		<font face=3D"SILDoulosIPA">[ji]	</font> a front close vowel with pala=
tal on-glide
+io		<font face=3D"SILDoulosIPA">[jo]	</font> a back mid vowel with palatal=
 on-glide
+iu		<font face=3D"SILDoulosIPA">[ju]	</font> a back close vowel with palat=
al on-glide
+
+ua		<font face=3D"SILDoulosIPA">[wa]	</font> an open vowel with labial on-=
glide
+ue		<font face=3D"SILDoulosIPA">[wE]	</font> a front mid vowel with labial=
 on-glide
+ui		<font face=3D"SILDoulosIPA">[wi]	</font> a front close vowel with labi=
al on-glide
+uo		<font face=3D"SILDoulosIPA">[wo]	</font> a back mid vowel with labial =
on-glide
+uu		<font face=3D"SILDoulosIPA">[wu]	</font> a back close vowel with labia=
l on-glide
+iy		<font face=3D"SILDoulosIPA">[j=AB]	</font> a central mid vowel with pa=
latal on-glide
+uy		<font face=3D"SILDoulosIPA">[w=AB]	</font> a central mid vowel with la=
bial on-glide
+</pre>(Approximate English equivalents of most of these diphthongs exist: =
see <a href=3D#s11>Section 11 for examples.)
+<p>
+<cx "diphthongs, classification of">  XE "diphthongs: classification of"  =
 The first four diphthongs above (=93ai=94, =93ei=94, =93oi=94, and =93au=
=94, the ones with off-glides) are freely used in most types of Lojban word=
s; the ten following ones are used only as stand-alone words and in Lojbani=
zed names and borrowings; and the last two (=93iy=94 and =93uy=94) are used=
 only in Lojbanized names.
+<p>
+<cx "consonants, syllabic">  XE "consonants: syllabic"   <cx "syllabic con=
sonants">  XE "syllabic consonants"   The syllabic consonants of Lojban, <f=
ont face=3D"SILDoulosIPA">[l`]</font>, <font face=3D"SILDoulosIPA">[m`]</fo=
nt>, <font face=3D"SILDoulosIPA">[n`]</font>, and <font face=3D"SILDoulosIP=
A">[r`]</font>, are variants of the non-syllabic <font face=3D"SILDoulosIPA=
">[l]</font>, <font face=3D"SILDoulosIPA">[m]</font>, <font face=3D"SILDoul=
osIPA">[n]</font>, and <font face=3D"SILDoulosIPA">[r] </font>respectively.=
 They normally have only a limited distribution, appearing in Lojban names =
and borrowings, although in principle any =93l=94, =93m=94, =93n=94, or =93=
r=94 may be pronounced syllabically. If a syllabic consonant appears next t=
o a =93l=94, =93m=94, =93n=94, or =93r=94 that is not syllabic, it may not =
be clear which is which:
+<p>
+<pre><a name=3De4d1>4.1)	brlgan.
+	<font face=3D"SILDoulosIPA">[br`l gan]</font>
+	or <font face=3D"SILDoulosIPA">[brl` gan]</font>
+</pre>is a hypothetical Lojbanized name with more than one valid pronuncia=
tion; however it is pronounced, it remains the same word.
+<p>
+<ex "Earl">  XE "Earl: example=93  <cx "syllabic consonants, final in word=
">  XE "syllabic consonants: final in word"   Syllabic consonants are treat=
ed as consonants rather than vowels from the standpoint of Lojban morpholog=
y. Thus Lojbanized names, which are generally required to end in a consonan=
t, are allowed to end with a syllabic consonant. An example is =93rl.=94, w=
hich is an approximation of the English name =93Earl=94, and has two syllab=
ic consonants.
+<p>
+<cx "stress, effect of syllabic consonants on">  XE "stress: effect of syl=
labic consonants on"   <cx "syllabic consonants, effect on stress">  XE "sy=
llabic consonants: effect on stress"   Syllables with syllabic consonants a=
nd no vowel are never stressed or counted when determining which syllables =
to stress (see <a href=3D#s9>Section 9</a>).
+<p>
+<a name=3Ds5><h3>Vowel pairs</h3>
+<p>
+<cx "vowel pairs, definition of">  XE "vowel pairs: definition of"   <cx "=
apostrophe, use in vowel pairs">  XE "apostrophe: use in vowel pairs"   <cx=
 "vowel pairs, use of apostrophe in">  XE "vowel pairs: use of apostrophe i=
n"   Lojban vowels also occur in pairs, where each vowel sound is in a sepa=
rate syllable. These two vowel sounds are connected (and separated) by an a=
postrophe. Lojban vowel pairs should be pronounced continuously with the <f=
ont face=3D"SILDoulosIPA">[h] </font>sound between (and not by a glottal st=
op or pause, which would split the two vowels into separate words).
+<p>
+<cx "vowel pairs, contrasted with diphthongs">  XE "vowel pairs: contraste=
d with diphthongs"   <cx "diphthongs, contrasted with vowel pairs">  XE "di=
phthongs: contrasted with vowel pairs"   All vowel combinations are permitt=
ed in two-syllable pairs with the apostrophe separating them; this includes=
 those which constitute diphthongs when the apostrophe is not included.
+
+<p>
+<cx "vowel pairs, list of">  XE "vowel pairs: list of"   The Lojban vowel =
pairs are:
+<p>
+<pre> 	a'a	a'e	a'i	a'o	a'u	a'y
+	e'a	e'e	e'i	e'o	e'u	e'y
+	i'a	i'e	i'i	i'o	i'u	i'y
+	o'a	o'e	o'i	o'o	o'u	o'y
+	u'a	u'e	u'i	u'o	u'u	u'y
+	y'a	y'e	y'i	y'o	y'u	y'y
+</pre><cx "vowel pairs, involving =93y=94">  XE "vowel pairs: involving y"=
   Vowel pairs involving =93y=94 appear only in Lojbanized names. They coul=
d appear in cmavo (structure words), but only =93.y'y.=94 is so used =97 it=
 is the Lojban name of the apostrophe letter (see <a href=3Dchap17.html>Cha=
pter 17</a>).
+<p>
+<cx "vowel pairs, grouping of">  XE "vowel pairs: grouping of"   When more=
 than two vowels occur together in Lojban, the normal pronunciation pairs v=
owels from the left into syllables, as in the Lojbanized name:
+<p>
+<pre><a name=3De5d1>5.1)	meiin.
+	mei,in.
+</pre><a href=3D#e5d1>Example 5.1 contains the diphthong =93ei=94 followed=
 by the vowel =93i=94. In order to indicate a different grouping, the comma=
 must always be used, leading to:
+<p>
+<pre><a name=3De5d2>5.2)	me,iin.
+</pre>which contains the vowel =93e=94 followed by the diphthong =93ii=94.=
 In rough English representation, <a href=3D#e5d1>Example 5.1 </a>is =93May=
 Een=94, whereas <a href=3D#e5d2>Example 5.2 </a>is =93Meh Yeen=94.
+<p>
+<a name=3Ds6><h3>Consonant clusters</h3>
+<p>
+<cx "consonant, definition">  XE "consonant: definition"   <cx "consonant,=
 effect on syllable count">  XE "consonant: effect on syllable count"   A c=
onsonant sound is a relatively brief speech-sound that precedes or follows =
a vowel sound in a syllable; its presence either preceding or following doe=
s not add to the count of syllables, nor is a consonant required in either =
position for any syllable. Lojban has seventeen consonants: for the purpose=
s of this section, the apostrophe is not counted as a consonant.
+<p>
+<cx "consonants, voicing of">  XE "consonants: voicing of"   An important =
distinction dividing Lojban consonants is that of voicing. The following ta=
ble shows the unvoiced consonants and the corresponding voiced ones:
+	  <p>
+<cx "consonants, voiced/unvoiced equivalents">  XE "consonants: voiced/unv=
oiced equivalents"   <pre>UNVOICED		    VOICED
+		p			b
+		t			d
+		k			g
+		f			v
+		c			j
+		s			z
+		x			-
+</pre><p>The consonant =93x=94 has no voiced counterpart in Lojban. The re=
maining consonants, =93l=94, =93m=94, =93n=94, and =93r=94, are typically p=
ronounced with voice, but can be pronounced unvoiced.
+<p>
+<cx "consonant clusters, definition of">  XE "consonant clusters: definiti=
on of"   <cx "single consonants, contrasted with doubled consonants">  XE "=
single consonants: contrasted with doubled consonants"   <cx "single conson=
ants, contrasted with consonant clusters">  XE "single consonants: contrast=
ed with consonant clusters"   <cx "doubled consonants, contrasted with sing=
le consonants">  XE "doubled consonants: contrasted with single consonants"=
   <cx "doubled consonants, contrasted with consonant clusters">  XE "doubl=
ed consonants: contrasted with consonant clusters"   <cx "consonant cluster=
s, contrasted with doubled consonants">  XE "consonant clusters: contrasted=
 with doubled consonants"   <cx "consonant clusters, contrasted with single=
 consonants">  XE "consonant clusters: contrasted with single consonants"  =
 Consonant sounds occur in languages as single consonants, or as doubled, o=
r as clustered combinations. Single consonant sounds are isolated by word b=
oundaries or by intervening vowel sounds from other consonant sounds. Doubl=
ed consonant sounds are either lengthened like <font face=3D"SILDoulosIPA">=
[s] </font>in English =93hiss=94, or repeated like <font face=3D"SILDoulosI=
PA">[k] </font>in English =93backcourt=94. Consonant clusters consist of tw=
o or more single or doubled consonant sounds in a group, each of which is d=
ifferent from its immediate neighbor. In Lojban, doubled consonants are exc=
luded altogether, and clusters are limited to two or three members, except =
in Lojbanized names.
+<p>
+<cx "consonants, position of">  XE "consonants: position of"   Consonants =
can occur in three positions in words: initial (at the beginning), medial (=
in the middle), and final (at the end). In many languages, the sound of a c=
onsonant varies depending upon its position in the word. In Lojban, as much=
 as possible, the sound of a consonant is unrelated to its position. In par=
ticular, the common American English trait of changing a =93t=94 between vo=
wels into a =93d=94 or even a flap (IPA <font face=3D"SILDoulosIPA">[=A8]</=
font>) is unacceptable in Lojban.
+<p>
+<cx "consonants, restrictions on">  XE "consonants: restrictions on"   <cx=
 "consonants, final">  XE "consonants: final"   Lojban imposes no restricti=
ons on the appearance of single consonants in any valid consonant position;=
 however, no consonant (including syllabic consonants) occurs final in a wo=
rd except in Lojbanized names.
+<p>
+<cx "consonant pairs, restrictions on">  XE "consonant pairs: restrictions=
 on"   Pairs of consonants can also appear freely, with the following restr=
ictions:
+<p>
+<dl compact><dt>1)	<dd>It is forbidden for both consonants to be the same,=
 as this would violate the rule against double consonants.
+</dl><cx "voiced/unvoiced consonants, restrictions on">  XE "voiced/unvoic=
ed consonants: restrictions on"   <dl compact><dt>2)	<dd>It is forbidden fo=
r one consonant to be voiced and the other unvoiced. The consonants =93l=94=
, =93m=94, =93n=94, and =93r=94 are exempt from this restriction. As a resu=
lt, =93bf=94 is forbidden, and so is =93sd=94, but both =93fl=94 and =93vl=
=94, and both =93ls=94 and =93lz=94, are permitted.
+<p><dt>3)	<dd>It is forbidden for both consonants to be drawn from the set=
 =93c=94, =93j=94, =93s=94, =93z=94.
+<p><dt>4)	<dd>The specific pairs =93cx=94, =93kx=94, =93xc=94, =93xk=94, a=
nd =93mz=94 are forbidden.
+</dl><ex "James">  XE "James: example=93  <cx "=93y=94, use in avoiding fo=
rbidden consonant pairs">  XE "y: use in avoiding forbidden consonant pairs=
"   These rules apply to all kinds of words, even Lojbanized names. If a na=
me would normally contain a forbidden consonant pair, a =93y=94 can be inse=
rted to break up the pair:
+<p>
+<pre><a name=3De6d1>6.1)	djeimyz.
+	<font face=3D"SILDoulosIPA">[dZEj m=ABz/]</font>
+	James
+</pre><p>The regular English pronunciation of =93James=94, which is <font =
face=3D"SILDoulosIPA">[dZEjmz]</font>, would Lojbanize as =93djeimz.=94, wh=
ich contains a forbidden consonant pair.
+<p>
+<a name=3Ds7><h3>Initial consonant pairs</h3>
+<p>
+<cx "consonant pairs, initial">  XE "consonant pairs: initial"   The set o=
f consonant pairs that may appear at the beginning of a word (excluding Loj=
banized names) is far more restricted than the fairly large group of permis=
sible consonant pairs described in <a href=3D#s6>Section 6</a>. Even so, it=
 is more than English allows, although hopefully not more than English-spea=
kers (and others) can learn to pronounce.
+
+<p>
+<cx "initial consonant pairs, list of">  XE "initial consonant pairs: list=
 of"   There are just 48 such permissible initial consonant pairs, as follo=
ws:
+<p>
+<dl compact><dt>			<dd>bl br
+			cf ck cl cm cn cp cr ct
+			dj dr dz
+			fl fr
+			gl gr
+			jb jd jg jm jv
+			kl kr
+			ml mr
+			pl pr
+			sf sk sl sm sn sp sr st
+			tc tr ts
+			vl vr
+			xl xr
+			zb zd zg zm zv
+</dl><p>Lest this list seem almost random, a pairing of voiced and unvoice=
d equivalent vowels will show significant patterns which may help in learni=
ng:
+<p>
+<pre>			pl pr			fl fr
+			bl br			vl vr
+			cp cf	ct ck cm cn	cl cr
+			jb jv	jd jg jm
+			sp sf	st sk sm sn	sl sr
+			zb zv	zd zg zm
+			tc tr	ts		kl kr
+			dj dr	dz		gl gr
+			ml mr			xl xr
+</pre><cx "voiced consonants, contrasted with unvoiced in allowable conson=
ant pairs">  XE "voiced consonants: contrasted with unvoiced in allowable c=
onsonant pairs"   <cx "unvoiced consonants, contrasted with voiced in allow=
able consonant pairs">  XE "unvoiced consonants: contrasted with voiced in =
allowable consonant pairs"   Note that if both consonants of an initial pai=
r are voiced, the unvoiced equivalent is also permissible, and the voiced p=
air can be pronounced simply by voicing the unvoiced pair. (The converse is=
 not true: =93cn=94 is a permissible initial pair, but =93jn=94 is not.)
+<p>
+<cx "consonant triples">  XE "consonant triples"   Consonant triples can o=
ccur medially in Lojban words. They are subject to the following rules:
+<p>
+<cx "consonant triples, restrictions on">  XE "consonant triples: restrict=
ions on"   <dl compact><dt>1)	<dd>	The first two consonants must constitute=
 a permissible consonant pair;
+<p><dt>2)	<dd>The last two consonants must constitute a permissible initia=
l consonant pair;
+<p><dt>3)	<dd>The triples =93ndj=94, =93ndz=94, =93ntc=94, and =93nts=94 a=
re forbidden.
+</dl><cx "consonant clusters, more than three consonants in">  XE "consona=
nt clusters: more than three consonants in"   Lojbanized names can begin or=
 end with any permissible consonant pair, not just the 48 initial consonant=
 pairs listed above, and can have consonant triples in any location, as lon=
g as the pairs making up those triples are permissible. In addition, names =
can contain consonant clusters with more than three consonants, again requi=
ring that each pair within the cluster is valid.
+<p>
+<a name=3Ds8><h3>Buffering of consonant clusters</h3>
+<p>
+<cx "consonant clusters, buffering of">  XE "consonant clusters: buffering=
 of"   <cx "buffer vowel">  XE "buffer vowel"   <cx "vowel, buffer">  XE "v=
owel: buffer"   Many languages do not have consonant clusters at all, and e=
ven those languages that do have them often allow only a subset of the full=
 Lojban set. As a result, the Lojban design allows the use of a buffer soun=
d between consonant combinations which a speaker finds unpronounceable. Thi=
s sound may be any non-Lojbanic vowel which is clearly separable by the lis=
tener from the Lojban vowels. Some possibilities are IPA <font face=3D"SILD=
oulosIPA">[I]</font>, <font face=3D"SILDoulosIPA">[=F6]</font>, <font face=
=3D"SILDoulosIPA">[U]</font>, or even <font face=3D"SILDoulosIPA">[Y]</font=
>, but there probably is no universally acceptable buffer sound. When using=
 a consonant buffer, the sound should be made as short as possible. Two exa=
mples showing such buffering (we will use <font face=3D"SILDoulosIPA">[I] <=
/font>in this chapter) are:
+<p>
+<pre><a name=3De8d1>8.1)	vrusi
+	<font face=3D"SILDoulosIPA">[=C8vru si]</font>
+	or <font face=3D"SILDoulosIPA">[vI =C8ru si]</font>
+</pre><ex "Amsterdam">  XE "Amsterdam: example=93  <pre><a name=3De8d2>8.2=
)	.AMsterdam.
+	<font face=3D"SILDoulosIPA">[/am ster dam/]</font>
+	or <font face=3D"SILDoulosIPA">[=C8/a mI sI tE rI da mI/]</font>
+</pre><cx "buffer vowel, and stress">  XE "buffer vowel: and stress"   <cx=
 "stress, effect of buffer vowel on">  XE "stress: effect of buffer vowel o=
n"   When a buffer vowel is used, it splits each buffered consonant into it=
s own syllable. However, the buffering syllables are never stressed, and ar=
e not counted in determining stress. They are, in effect, not really syllab=
les to a Lojban listener, and thus their impact is ignored.
+<p>
+Here are more examples of unbuffered and buffered pronunciations:
+<p>
+<pre><a name=3De8d3>8.3)	klama
+	<font face=3D"SILDoulosIPA">[=C8kla ma]</font>
+	<font face=3D"SILDoulosIPA">[kI =C8la ma]</font>
+<a name=3De8d4>8.4)	xapcke
+	<font face=3D"SILDoulosIPA">[=C8xap ckE]</font>
+	<font face=3D"SILDoulosIPA">[=C8xa pI ckE]</font>
+	<font face=3D"SILDoulosIPA">[=C8xa pI cI kE]</font>
+</pre><p>In <a href=3D#e8d4>Example 8.4</a>, we see that buffering vowels =
can be used in just some, rather than all, of the possible places: the seco=
nd pronunciation buffers the =93pc=94 consonant pair but not the =93ck=94. =
The third pronunciation buffers both.
+<p>
+<pre><a name=3De8d5>8.5)	ponyni'u
+	<font face=3D"SILDoulosIPA">[po n=AB =C8ni hu]</font>
+</pre><cx "vowel buffer, contrasted with =93y=94">  XE "vowel buffer: cont=
rasted with y sound"   <cx "=93y=94, contrasted with vowel buffer">  XE "y =
sound: contrasted with vowel buffer"   <ex "bone bread">  XE "bone bread: e=
xample=93  <a href=3D#e8d5>Example 8.5 cannot contain any buffering vowel. =
It is important not to confuse the vowel =93y=94, which is pronounced <font=
 face=3D"SILDoulosIPA">[=AB]</font>, with the buffer, which has a variety o=
f possible pronunciations and is never written. Consider the contrast betwe=
en
+<p>
+<pre><a name=3De8d6>8.6)	bongynanba
+	<font face=3D"SILDoulosIPA">[boN g=AB =C8nan ba]</font>
+</pre>an unlikely Lojban compound word meaning =93bone bread=94 (note the =
use of <font face=3D"SILDoulosIPA">[N] </font>as a representative of =93n=
=94 before =93g=94) and
+
+<p>
+<pre><a name=3De8d7>8.7)	bongnanba
+	<font face=3D"SILDoulosIPA">[boN =C8gnan ba]</font>
+</pre>a possible borrowing from another language (Lojban borrowings can on=
ly take a limited form). If <a href=3D#e8d7>Example 8.7 </a>were pronounced=
 with buffering, as
+<p>
+<pre><a name=3De8d8>8.8)	<font face=3D"SILDoulosIPA">[boN gI =C8nan ba]</f=
ont>
+</pre>it would be very similar to <a href=3D#e8d6>Example 8.6</a>. Only a =
clear distinction between =93y=94 and any buffering vowel would keep the tw=
o words distinct.
+<p>
+<cx "vowels, length of">  XE "vowels: length of"   <cx "buffer vowel, shor=
tening of">  XE "buffer vowel: shortening of"   Since buffering is done for=
 the benefit of the speaker in order to aid pronounceability, there is no g=
uarantee that the listener will not mistake a buffer vowel for one of the s=
ix regular Lojban vowels. The buffer vowel should be as laxly pronounced as=
 possible, as central as possible, and as short as possible. Furthermore, i=
t is worthwhile for speakers who use buffers to pronounce their regular vow=
els a bit longer than usual, to avoid confusion with buffer vowels. The spe=
akers of many languages will have trouble correctly hearing any of the sugg=
ested buffer vowels otherwise. By this guideline, <a href=3D#e8d8>Example 8=
.8 would be pronounced
+<p>
+<pre><a name=3De8d9>8.9)	<font face=3D"SILDoulosIPA">[bo=F9N gI =C8na=F9n =
ba=F9]</font>
+</pre>with lengthened vowels.
+<p>
+<a name=3Ds9><h3>Syllabication and stress</h3>
+<p>
+<cx "syllabication, definition of">  XE "syllabication: definition of"   A=
 Lojban word has one syllable for each of its vowels, diphthongs, and sylla=
bic consonants (referred to simply as =93vowels=94 for the purposes of this=
 section.) Syllabication rules determine which of the consonants separating=
 two vowels belong to the preceding vowel and which to the following vowel.=
 These rules are conventional only; the phonetic facts of the matter about =
how utterances are syllabified in any language are always very complex.
+<p>
+<cx "syllabication, rules for">  XE "syllabication: rules for"   A single =
consonant always belongs to the following vowel. A consonant pair is normal=
ly divided between the two vowels; however, if the pair constitute a valid =
initial consonant pair, they are normally both assigned to the following vo=
wel. A consonant triple is divided between the first and second consonants.=
 Apostrophes and commas, of course, also represent syllable breaks. Syllabi=
c consonants usually appear alone in their syllables.
+<p>
+<cx "syllabication, and names">  XE "syllabication: and names"   It is per=
missible to vary from these rules in Lojbanized names. For example, there a=
re no definitive rules for the syllabication of names with consonant cluste=
rs longer than three consonants. The comma is used to indicate variant syll=
abication or to explicitly mark normal syllabication.
+<p>
+<cx "syllabication, examples of">  XE "syllabication: examples of"   Here =
are some examples of Lojban syllabication:
+<p>
+<pre><a name=3De9d1>9.1)	pujenaicajeba
+	pu,je,nai,ca,je,ba
+</pre><p>This word has no consonant pairs and is therefore syllabified bef=
ore each medial consonant.
+<p>
+<pre><a name=3De9d2>9.2)	ninmu
+	nin,mu
+
+</pre><p>This word is split at a consonant pair.
+<p>
+<pre><a name=3De9d3>9.3)	fitpri
+	fit,pri
+</pre><p>This word is split at a consonant triple, between the first two c=
onsonants of the triple.
+<p>
+<pre><a name=3De9d4>9.4)	sairgoi
+	sair,goi
+	sai,r,goi
+</pre><p>This word contains the consonant pair =93rg=94; the =93r=94 may b=
e pronounced syllabically or not.
+<p>
+<pre><a name=3De9d5>9.5)	klezba
+	klez,ba
+	kle,zba
+</pre><p>This word contains the permissible initial pair =93zb=94, and so =
may be syllabicated either between =93z=94 and =93b=94 or before =93zb=94.
+<p>
+<cx "stress, definition of">  XE "stress: definition of"   <cx "stressed s=
yllable, compared with stressed vowel">  XE "stressed syllable: compared wi=
th stressed vowel"   <cx "stressed vowel, compared with stressed syllable">=
  XE "stressed vowel: compared with stressed syllable"   Stress is a relati=
vely louder pronunciation of one syllable in a word or group of words. Sinc=
e every syllable has a vowel sound (or diphthong or syllabic consonant) as =
its nucleus, and the stress is on the vowel sound itself, the terms =93stre=
ssed syllable=94 and =93stressed vowel=94 are largely interchangeable conce=
pts.
+<p>
+<cx "stress, rules for">  XE "stress: rules for"   Most Lojban words are s=
tressed on the next-to-the-last, or penultimate, syllable. In counting syll=
ables, however, syllables whose vowel is =93y=94 or which contain a syllabi=
c consonant (=93l=94, =93m=94, =93n=94, or =93r=94) are never counted. (The=
 Lojban term for penultimate stress is =93da'amoi terbasna=94.) Similarly, =
syllables created solely by adding a buffer vowel, such as <font face=3D"SI=
LDoulosIPA">[I]</font>, are not counted.
+<p>
+<cx "stress, levels of">  XE "stress: levels of"   There are actually thre=
e levels of stress =97 primary, secondary, and weak. Weak stress is the low=
est level, so it really means no stress at all. Weak stress is required for=
 syllables containing =93y=94, a syllabic consonant, or a buffer vowel.
+<p>
+<cx "stress, primary">  XE "stress: primary"   <cx "cmavo, stress on">  XE=
 "cmavo: stress on"   <cx "brivla, stress on">  XE "brivla: stress on"   <c=
x "names, stress on">  XE "names: stress on"   Primary stress is required o=
n the penultimate syllable of Lojban content words (called =93brivla=94). L=
ojbanized names may be stressed on any syllable, but if a syllable other th=
an the penultimate is stressed, the syllable (or at least its vowel) must b=
e capitalized in writing. Lojban structural words (called =93cmavo=94) may =
be stressed on any syllable or none at all. However, primary stress may not=
 be used in a syllable just preceding a brivla, unless a pause divides them=
; otherwise, the two words may run together.
+<p>
+<cx "stress, secondary">  XE "stress: secondary"   Secondary stress is the=
 optional and non-distinctive emphasis used for other syllables besides tho=
se required to have either weak or primary stress. There are few rules gove=
rning secondary stress, which typically will follow a speaker's native lang=
uage habits or preferences. Secondary stress can be used for contrast, or f=
or emphasis of a point. Secondary stress can be emphasized at any level up =
to primary stress, although the speaker must not allow a false primary stre=
ss in brivla, since errors in word resolution could result.
+<p>
+<ex "Armstrong">  XE "Armstrong: example=93  The following are Lojban word=
s with stress explicitly shown:
+<p>
+<pre><a name=3De9d6>9.6)	dikyjvo
+	DI,ky,jvo
+</pre>(In a fully-buffered dialect, the pronunciation would be: <font face=
=3D"SILDoulosIPA">[=C8di k=AB jI vo]</font>.) Note that the syllable =93ky=
=94 is not counted in determining stress. The vowel =93y=94 is never stress=
ed in a normal Lojban context.
+<p>
+<pre><a name=3De9d7>9.7)	.armstrong.
+	.ARM,strong.
+</pre><p>This is a Lojbanized version of the name =93Armstrong=94. The fin=
al =93g=94 must be explicitly pronounced. With full buffering, the name wou=
ld be pronounced:
+<p>
+<pre><a name=3De9d8>9.8)	<font face=3D"SILDoulosIPA">[=C8/a rI mI sI tI ro=
 nI gI/]</font>
+</pre><p>However, there is no need to insert a buffer in every possible pl=
ace just because it is inserted in one place: partial buffering is also acc=
eptable. In every case, however, the stress remains in the same place: on t=
he first syllable.
+<p>
+<cx "-ng, Lojban contrasted with English">  XE "-ng: Lojban contrasted wit=
h English"   The English pronunciation of =93Armstrong=94, as spelled in En=
glish, is not correct by Lojban standards; the letters =93ng=94 in English =
represent a velar nasal (IPA <font face=3D"SILDoulosIPA">[N]</font>) which =
is a single consonant. In Lojban, =93ng=94 represents two separate consonan=
ts that must both be pronounced; you may not use <font face=3D"SILDoulosIPA=
">[N] </font>to pronounce Lojban =93ng=94, although <font face=3D"SILDoulos=
IPA">[Ng] </font>is acceptable. English speakers are likely to have to pron=
ounce the ending with a buffer, as one of the following:
+<p>
+<pre><a name=3De9d9>9.9)	<font face=3D"SILDoulosIPA">[=C8/arm stron gI/]</=
font>
+	or <font face=3D"SILDoulosIPA">[=C8/arm stroN gI/]</font>
+	or even <font face=3D"SILDoulosIPA">[=C8/arm stro nIg/]</font>
+</pre><p>The normal English pronunciation of the name =93Armstrong=94 coul=
d be Lojbanized as:
+<p>
+<pre><a name=3De9d10>9.10)	.ARMstron.
+</pre>since Lojban =93n=94 is allowed to be pronounced as the velar nasal =
<font face=3D"SILDoulosIPA">[N]</font>.
+<p>
+Here is another example showing the use of =93y=94:
+<p>
+<pre><a name=3De9d11>9.11)	bisydja
+	BI,sy,dja
+	BI,syd,ja
+</pre><p>This word is a compound word, or lujvo, built from the two affixe=
s =93bis=94 and =93dja=94. When they are joined, an impermissible consonant=
 pair results: =93sd=94. In accordance with the algorithm for making lujvo,=
 explained in <a href=3Dchap4.html>Chapter 4</a>, a =93y=94 is inserted to =
separate the impermissible consonant pair; the =93y=94 is not counted as a =
syllable for purposes of stress determination.
+<p>
+<ex "syllabication, variants of">  XE "syllabication: variants of"   <pre>=
<a name=3De9d12>9.12)	da'udja
+	da'UD,ja
+	da'U,dja
+</pre><p>These two syllabications sound the same to a Lojban listener =97 =
the association of unbuffered consonants in syllables is of no import in re=
cognizing the word.
+<p>
+<pre><a name=3De9d13>9.13)	e'u bridi
+	e'u BRI,di
+	E'u BRI,di
+	e'U.BRI,di
+</pre><p>In <a href=3D#e9d13>Example 9.13</a>, =93e'u=94 is a cmavo and =
=93bridi=94 is a brivla. Either of the first two pronunciations is permitte=
d: no primary stress on either syllable of =93e'u=94, or primary stress on =
the first syllable. The third pronunciation, which places primary stress on=
 the second syllable of the cmavo, requires that =97 since the following wo=
rd is a brivla =97 the two words must be separated by a pause. Consider the=
 following two cases:
+<p>
+<pre><a name=3De9d14>9.14)	le re nobli prenu
+	le re NObli PREnu
+<a name=3De9d15>9.15)	le re no bliprenu
+	le re no bliPREnu
+</pre><p>If the cmavo =93no=94 in <a href=3D#e9d15>Example 9.15 </a>were t=
o be stressed, the phrase would sound exactly like the given pronunciation =
of <a href=3D#e9d14>Example 9.14</a>, which is unacceptable in Lojban: a si=
ngle pronunciation cannot represent both.
+<p> <a name=3Ds10><h3>IPA for English-speakers</h3>
+<p> <cx "General American">  XE "General American"   <cx "Received Pronunc=
iation">  XE "Received Pronunciation"   <cx "television">  XE "television" =
  <cx "IPA pronunciation, description">  XE "IPA pronunciation: description=
"   There are many dialects of English, thus making it difficult to define =
the standardized symbols of the IPA in terms useful to every reader. All th=
e symbols used in this chapter are repeated here, in more or less alphabeti=
cal order, with examples drawn from General American. In addition, some att=
ention is given to the Received Pronunciation of (British) English. These t=
wo dialects are referred to as GA and RP respectively. Speakers of other di=
alects should consult a book on phonetics or their local television sets.
+<p> <dl compact><dt><font face=3D"SILDoulosIPA">[=C8]</font>	<dd>An IPA in=
dicator of primary stress; the syllable which follows <font face=3D"SILDoul=
osIPA">[=C8] </font>receives primary stress.
+<p><dt><font face=3D"SILDoulosIPA">[/]</font>	<dd>An allowed variant of Lo=
jban =93.=94. This sound is not usually considered part of English. It is t=
he catch in your throat that sometimes occurs prior to the beginning of a w=
ord (and sometimes a syllable) which starts with a vowel. In some dialects,=
 like Cockney and some kinds of American English, it is used between vowels=
 instead of =93t=94: =93bottle=94 <font face=3D"SILDoulosIPA">[bo/l`]</font=
>. The English interjection =93uh-oh!=94 almost always has it between the s=
yllables.
+<p><dt><font face=3D"SILDoulosIPA">[=F9]</font>	<dd>A symbol indicating th=
at the previous vowel is to be spoken for a longer time than usual. Lojban =
vowels can be pronounced long in order to make a greater contrast with buff=
er vowels.
+<p><dt><font face=3D"SILDoulosIPA">[a]</font>	<dd>The preferred pronunciat=
ion of Lojban =93a=94. This sound doesn't occur in GA, but sounds somewhat =
like the =93ar=94 of =93park=94, as spoken in RP or New England American. I=
t is pronounced further forward in the mouth than <font face=3D"SILDoulosIP=
A">[A]</font>.
+<p><dt><font face=3D"SILDoulosIPA">[A]</font>	<dd>An allowed variant of Lo=
jban =93a=94. The =93a=94 of GA =93father=94. The sound <font face=3D"SILDo=
ulosIPA">[a] </font>is preferred because GA speakers often relax an unstres=
sed <font face=3D"SILDoulosIPA">[A] </font>into a schwa <font face=3D"SILDo=
ulosIPA">[=AB]</font>, as in the usual pronunciations of =93about=94 and =
=93sofa=94. Because schwa is a distinct vowel in Lojban, English speakers m=
ust either learn to avoid this shift or to use <font face=3D"SILDoulosIPA">=
[a] </font>instead: the Lojban word for =93sofa=94 is =93sfofa=94, pronounc=
ed <font face=3D"SILDoulosIPA">[sfofa] </font>or <font face=3D"SILDoulosIPA=
">[sfofA] </font>but never <font face=3D"SILDoulosIPA">[sfof=AB] </font>whi=
ch would be the non-word =93sfofy=94.
+<p><dt><font face=3D"SILDoulosIPA">[Q]</font>	<dd>Not a Lojban sound. The =
=93a=94 of English =93cat=94.
+<p><dt><font face=3D"SILDoulosIPA">[b]</font>	<dd>The preferred pronunciat=
ion of Lojban =93b=94. As in English =93boy=94, =93sober=94, or =93job=94.
+<p><dt><font face=3D"SILDoulosIPA">[B]</font>	<dd>An allowed variant of Lo=
jban =93v=94. Not an English sound; the Spanish =93b=94 or =93v=94 between =
vowels. This sound should not be used for Lojban =93b=94.
+<p><dt><font face=3D"SILDoulosIPA">[d]</font>	<dd>The preferred pronunciat=
ion of Lojban =93d=94. As in English =93dog=94, =93soda=94, or =93mad=94.
+<p><dt><font face=3D"SILDoulosIPA">[E]</font>	<dd>The preferred pronunciat=
ion of Lojban =93e=94. The =93e of English =93met=94.
+<p><dt><font face=3D"SILDoulosIPA">[e]</font>	<dd>An allowed variant of Lo=
jban =93e=94. This sound is not found in English, but is the Spanish =93e=
=94, or the tense =93e=94 of Italian. The vowel of English =93say=94 is sim=
ilar except for the off-glide: you can learn to make this sound by holding =
your tongue steady while saying the first part of the English vowel.
+<p><dt><font face=3D"SILDoulosIPA">[=AB]</font>	<dd>The preferred pronunci=
ation of Lojban =93y=94. As in the =93a=94 of English =93sofa=94 or =93abou=
t=94. Schwa is generally unstressed in Lojban, as it is in English. It is a=
 totally relaxed sound made with the tongue in the middle of the mouth.
+<p><dt><font face=3D"SILDoulosIPA">[f]</font>	<dd>The preferred pronunciat=
ion of Lojban =93f=94. As in =93fee=94, =93loafer=94, or =93chef=94.
+<p><dt><font face=3D"SILDoulosIPA">[=B8]</font>	<dd>An allowed variant of =
Lojban =93f=94. Not an English sound; the Japanese =93f=94 sound.
+<p><dt><font face=3D"SILDoulosIPA">[g]</font>	<dd>The preferred pronunciat=
ion of Lojban =93g=94. As in English =93go=94, =93eagle=94, or =93dog=94.
+<p><dt><font face=3D"SILDoulosIPA">[h]</font>	<dd>The preferred pronunciat=
ion of the Lojban apostrophe sound. As in English =93aha=94 or the second =
=93h=94 in =93oh, hello=94.
+<p><dt><font face=3D"SILDoulosIPA">[i]</font>	<dd>The preferred pronunciat=
ion of Lojban =93i=94. Essentially like the English vowel of =93pizza=94 or=
 =93machine=94, although the English vowel is sometimes pronounced with an =
off-glide, which should not be present in Lojban.
+<p><dt><font face=3D"SILDoulosIPA">[I]</font>	<dd>A possible Lojban buffer=
 vowel. The =93i=94 of English =93bit=94.
+<p><dt><font face=3D"SILDoulosIPA">[=F6]</font>	<dd>A possible Lojban buff=
er vowel. The =93u=94 of =93just=94 in some varieties of GA, those which ma=
ke the word sound more or less like =93jist=94. Also Russian =93y=94 as in =
=93byt'=94 (to be); like a schwa <font face=3D"SILDoulosIPA">[=AB]</font>, =
but higher in the mouth.
+<p><dt><font face=3D"SILDoulosIPA">[j]</font>	<dd>Used in Lojban diphthong=
s beginning or ending with =93i=94. Like the =93y=94 in English =93yard=94 =
or =93say=94.
+<p><dt><font face=3D"SILDoulosIPA">[k]</font>	<dd>The preferred pronunciat=
ion of Lojban =93k=94. As in English =93kill=94, =93token=94, or =93flak=94=
.
+<p><dt><font face=3D"SILDoulosIPA">[l]</font>	<dd>The preferred pronunciat=
ion of Lojban =93l=94. As in English =93low=94, =93nylon=94, or =93excel=94=
.
+<p><dt><font face=3D"SILDoulosIPA">[l`]</font>	<dd>The syllabic version of=
 Lojban =93l=94, as in English =93bottle=94 or =93middle=94.
+<p><dt><font face=3D"SILDoulosIPA">[m]</font>	<dd>The preferred pronunciat=
ion of Lojban =93m=94. As in English =93me=94, =93humor=94, or =93ham=94.
+<p><dt><font face=3D"SILDoulosIPA">[m`]</font>	<dd>The syllabic version of=
 Lojban =93m=94. As in English =93catch 'em=94 or =93bottom=94.
+<p><dt><font face=3D"SILDoulosIPA">[n]</font>	<dd>The preferred pronunciat=
ion of Lojban =93n=94. As in English =93no=94, =93honor=94, or =93son=94.
+<p><dt><font face=3D"SILDoulosIPA">[n`]</font>	<dd>The syllabic version of=
 Lojban =93n=94. As in English =93button=94.
+<p><dt><font face=3D"SILDoulosIPA">[N]</font>	<dd>An allowed variant of Lo=
jban =93n=94, especially in Lojbanized names and before =93g=94 or =93k=94.=
 As in English =93sing=94 or =93singer=94 (but not =93finger=94 or =93dange=
r=94).
+<p><dt><font face=3D"SILDoulosIPA">[N`]</font>	<dd>An allowed variant of L=
ojban syllabic =93n=94, especially in Lojbanized names.
+<p><dt><font face=3D"SILDoulosIPA">[o]</font>	<dd>The preferred pronunciat=
ion of Lojban =93o=94. As in the French =93haute (cuisine)=94 or Spanish =
=93como=94. There is no exact English equivalent of this sound. The nearest=
 GA equivalent is the =93o=94 of =93dough=94 or =93joke=94, but it is essen=
tial that the off-glide (a <font face=3D"SILDoulosIPA">[w]</font>-like soun=
d) at the end of the vowel is not pronounced when speaking Lojban. The RP s=
ound in these words is <font face=3D"SILDoulosIPA">[=ABw] </font>in IPA ter=
ms, and has no <font face=3D"SILDoulosIPA">[o] </font>in it at all; unless =
you can speak with a Scots, Irish, or American accent, you may have trouble=
 with this sound.
+]</font>	<dd>An allowed variant of Lojban =93o=94, especially before =93r=
=94. This sound is a shortened form of the =93aw=94 in GA =93dawn=94 (for t=
hose people who don't pronounce =93dawn=94 and =93Don=94 alike; if you do, =
you may have trouble with this sound). In RP, but not GA, it is the =93o=94=
 of =93hot=94.
+<p><dt><font face=3D"SILDoulosIPA">[p]</font>	<dd>The preferred pronunciat=
ion of Lojban =93p=94. As in English =93pay=94, =93super=94, or =93up=94.
+<p><dt><font face=3D"SILDoulosIPA">[r]</font>	<dd>One version of Lojban =
=93r=94. Not an English sound. The Spanish =93rr=94 and the Scots =93r=94, =
a tongue-tip trill.
+<p><dt><font face=3D"SILDoulosIPA">[=A8]</font>	<dd>One version of Lojban =
=93r=94. As in GA =93right=94, =93baron=94, or =93car=94. Not found in RP.
+<p><dt><font face=3D"SILDoulosIPA">[R]</font>	<dd>One version of Lojban =
=93r=94. In GA, appears as a variant of =93t=94 or =93d=94 in the words =93=
metal=94 and =93medal=94 respectively. A tongue-tip flap.
+<font face=3D"SILDoulosIPA">[{]</font>	One version of Lojban =93r=94. Not =
an English sound. The French or German =93r=94 in =93reine=94 or =93rot=94 =
respectively. A uvular trill.
+<p><dt><dd><font face=3D"SILDoulosIPA">[r`]</font>, <font face=3D"SILDoulo=
sIPA">[=A8`]</font>, <font face=3D"SILDoulosIPA">[R`]</font>, <font face=3D=
"SILDoulosIPA">[{`] </font>are syllabic versions of the above. <font face=
=3D"SILDoulosIPA">[=A8`] </font>appears in the GA (but not RP) pronunciatio=
n of =93bird=94.
+<p><dt><font face=3D"SILDoulosIPA">[s]</font>	<dd>The preferred pronunciat=
ion of Lojban =93s=94. As in English =93so=94, =93basin=94, or =93yes=94.
+<p><dt><font face=3D"SILDoulosIPA">[S]</font>	<dd>The preferred pronunciat=
ion of Lojban =93c=94. The =93sh=94 of English =93ship=94, =93ashen=94, or =
=93dish=94.
+<p><dt><font face=3D"SILDoulosIPA">[=A7]</font>	<dd>An allowed variant of =
Lojban =93s=94. Not an English sound. The Hindi retroflex =93s=94 with dot =
below, or Klingon =93S=94.
+<p><dt><font face=3D"SILDoulosIPA">[t]</font>	<dd>The preferred pronunciat=
ion of Lojban =93t=94. As in English =93tea=94, =93later=94, or =93not=94. =
It is important to avoid the GA habit of pronouncing the =93t=94 between vo=
wels as <font face=3D"SILDoulosIPA">[d] </font>or <font face=3D"SILDoulosIP=
A">[R]</font>.
+<p><dt><font face=3D"SILDoulosIPA">[T]</font>	<dd>Not normally a Lojban so=
und, but a possible variant of Lojban =93 ' =94. The =93th=94 of English =
=93thin=94 (but not =93then=94).
+<p><dt><font face=3D"SILDoulosIPA">[v]</font>	<dd>The preferred pronunciat=
ion of Lojban =93v=94. As in English =93voice=94, =93savor=94, or =93live=
=94.
+<p><dt><font face=3D"SILDoulosIPA">[w]</font>	<dd>Used in Lojban diphthong=
s beginning or ending with =93u=94. Like the =93w=94 in English =93wet=94 <=
font face=3D"SILDoulosIPA">[wEt] </font>or =93cow=94 <font face=3D"SILDoulo=
sIPA">[kAw]</font>.
+<p><dt><font face=3D"SILDoulosIPA">[x]</font>	<dd>The preferred pronunciat=
ion of Lojban =93x=94. Not normally an English sound, but used in some pron=
unciations of =93loch=94 and =93Bach=94; =93gh=94 in Scots =93might=94 and =
=93night=94. The German =93Ach-Laut=94. To pronounce <font face=3D"SILDoulo=
sIPA">[x]</font>, force air through your throat without vibrating your voca=
l chords; there should be lots of scrape.
+<p><dt><font face=3D"SILDoulosIPA">[Y]</font>	<dd>A possible Lojban buffer=
 vowel. Not an English sound: the =93=FC=94 of German =93h=FCbsch=94.
+<p><dt><font face=3D"SILDoulosIPA">[z]</font>	<dd>The preferred pronunciat=
ion of Lojban =93z=94. As in English =93zoo=94, =93hazard=94, or =93fizz=94=
.
+<p><dt><font face=3D"SILDoulosIPA">[Z]</font>	<dd>The preferred pronunciat=
ion of Lojban =93j=94. The =93si=94 of English =93vision=94, or the consona=
nt at the end of GA =93garage=94.
+<p><dt><font face=3D"SILDoulosIPA">[=BD]</font>	<dd>An allowed variant of =
Lojban =93z=94. Not an English sound. The voiced version of <font face=3D"S=
ILDoulosIPA">[=A7]</font>.
+</dl><a name=3Ds11><h3>English analogues for Lojban diphthongs</h3>
+<p>
+<cx "diphthongs, English analogues of">  XE "diphthongs: English analogues=
 of"   Here is a list of English words that contain diphthongs that are sim=
ilar to the Lojban diphthongs. This list does not constitute an official pr=
onunciation guide; it is intended as a help to English-speakers.
+<p>
+<pre>      Lojban	 English
+	ai	 =93pie=94
+	ei	 =93pay=94
+	oi	 =93boy=94
+	au	 =93cow=94
+	ia	 =93yard=94
+	ie	 =93yes=94
+	ii	 =93ye=94
+	io	 =93yodel=94 (in GA only)
+	iu	 =93unicorn=94 or =93few=94
+	ua	 =93suave=94
+	ue	 =93wet=94
+	ui	 =93we=94
+	uo	 =93woe=94 (in GA only)
+	uu	 =93woo=94
+	iy	 =93million=94 (the =93io=94 part, that is)
+	uy	 =93was=94 (when unstressed)
+</pre><a name=3Ds12><h3>Oddball orthographies</h3>
+<p>
+<cx "orthography, non-standard">  XE "orthography: non-standard"   The fol=
lowing notes describe ways in which Lojban has been written or could be wri=
tten that differ from the standard orthography explained in the rest of thi=
s chapter. Nobody needs to read this section except people with an interest=
 in the obscure. Technicalities are used without explanation or further apo=
logy.
+<p>
+There exists an alternative orthography for Lojban, which is designed to b=
e as compatible as possible (but no more so) with the orthography used in p=
re-Lojban versions of Loglan. The consonants undergo no change, except that=
 =93x=94 is replaced by =93h=94. The individual vowels likewise remain unch=
anged. However, the vowel pairs and diphthongs are changed as follows:
+<p>
+<dl compact><dt><dd>=93ai=94, =93ei=94, =93oi=94, =93au=94 become =93ai=94=
, =93ei=94, =93oi=94, =93ao=94.
+<p><dt><dd>=93ia=94 through =93iu=94 and =93ua=94 through =93uu=94 remain =
unchanged.
+<p><dt><dd>=93a'i=94, =93e'i=94, =93o'i=94 and =93a'o=94 become =93a,i=94,=
 =93e,i=94, =93o,i=94 and =93a,o=94.
+<p><dt><dd>=93i'a=94 through =93i'u=94 and =93u'a=94 through =93u'u=94 are=
 changed to =93ia=94 through =93iu=94 and =93ua=94 through =93uu=94 in lujv=
o and cmavo other than attitudinals, but become =93i,a=94 through =93i,u=94=
 and =93u,a=94 through =93u,u=94 in names, fu'ivla, and attitudinal cmavo.
+<p><dt><dd>All other vowel pairs simply drop the apostrophe.
+</dl><p>The result of these rules is to eliminate the apostrophe altogethe=
r, replacing it with comma where necessary, and otherwise with nothing. In =
addition, names and the cmavo =93.i=94 are capitalized, and irregular stres=
s is marked with an apostrophe (now no longer used for a sound) following t=
he stressed syllable.
+<p>
+Three points must be emphasized about this alternative orthography:
+<p>
+<cx "non-standard orthographies, caveat">  XE "non-standard orthographies:=
 caveat"   <dl compact><dt><dd>It is not standard, and has not been used.
+<p><dt><dd>It does not represent any changes to the standard Lojban phonol=
ogy; it is simply a representation of the same phonology using a different =
written form.
+<p><dt><dd>It was designed to aid in a planned rapprochement between the L=
ogical Language Group and The Loglan Institute, a group headed by James Coo=
ke Brown. The rapprochement never took place.
+</dl><cx "non-standard orthographies, Cyrillic">  XE "non-standard orthogr=
aphies: Cyrillic"   There also exists a Cyrillic orthography for Lojban whi=
ch was designed when the introductory Lojban brochure was translated into R=
ussian. It uses the =93=E0=94, =93=E1=94, =93=E2=94, =93=E3=94, =93=E4=94, =
=93=E5=94, =93=E6=94, =93=E7=94, =93=E8=94, =93=EA=94, =93=EB=94, =93=EC=94=
, =93=ED=94, =93=EE=94, =93=EF=94, =93=F0=94, =93=F1=94, =93=F2=94, =93=F3=
=94, =93=F4=94, =93=F5=94, and =93=F8=94 in the obvious ways. The Latin let=
ter =93y=94 is mapped onto the hard sign =93=FA=94, as in Bulgarian. The ap=
ostrophe, comma, and period are unchanged. Diphthongs are written as vowel =
pairs, as in the Roman representation.
+<p>
+<cx "non-standard orthographies, Tengwar">  XE "non-standard orthographies=
: Tengwar"   <cx "Tolkien, and non-standard Lojban orthography">  XE "Tolki=
en: and non-standard Lojban orthography"   Finally, an orthography using th=
e Tengwar of F=E9&eacute;anor, a fictional orthography invented by J. R. R.=
 Tolkien and described in the Appendixes to <cite>The Lord Of The Rings</ci=
te>, has been devised for Lojban. The following mapping, which closely rese=
mbles that used for Westron, will be meaningful only to those who have read=
 those appendixes. In brief, the tincot=E9&eacute;ma and parmat=E9&eacute;m=
a are used in the conventional ways; the calmat=E9&eacute;ma represents pal=
atal consonants, and the quesset=E9&eacute;ma represents velar consonants.
+<p>
+<pre> 	t	tinco		p	parma
+	-	calma		k	quesse
+	d	ando		b	umbar
+	-	anga		g	ungwe
+	-	thule		f	formen
+	c	harma		x	hwesta
+	-	anto		v	ampa
+	j	anca		-	unque
+	n	numen		m	malta
+	-	noldo		-	nwalme
+	r	ore		u	vala
+	i	anna		-	vilya
+</pre><p>The letters =93vala=94 and =93anna=94 are use for =93u=94 and =93=
i=94 only when those letters are used to represent glides. Of the additiona=
l letters, =93r=94, =93l=94, =93s=94, and =93z=94 are written with =93r=F3&=
oacute;men=94, =93lambe=94, =93silme=94, and =93=E1&aacute;re/=93esse=94 re=
spectively; the inverted forms are used as free variants.
+<p>
+<p>
+Lojban, like Quenya, is a vowel-last language, so tehtar are read as follo=
wing the tengwar on which they are placed. The conventional tehtar are used=
 for the five regular vowels, and the dot below for =93y=94. The Lojban apo=
strophe is represented by =93halla=94. There is no equivalent of the Lojban=
 comma or period.
+<p>
+<dl compact><dt><dd>
+
+</body></html>
+<img src=3Dchap4.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D4=
05>
+<h2>Chapter 4
+<br>
+The Shape Of Words To Come: Lojban Morphology</h2>
+<h6>$Revision: 4.1 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>Introductory</h3>
+<p>
+<cx "morphology, simplicity of">  XE "morphology: simplicity of"   <cx "mo=
rphology, definition">  XE "morphology: definition"   <cx "word forms, in L=
ojban (see also morphology)">  XE "word forms: in Lojban (see also morpholo=
gy)"   Morphology is the part of grammar that deals with the form of words.=
 Lojban's morphology is fairly simple compared to that of many languages, b=
ecause Lojban words don't change form depending on how they are used. Engli=
sh has only a small number of such changes compared to languages like Russi=
an, but it does have changes like =93boys=94 as the plural of =93boy=94, or=
 =93walked=94 as the past-tense form of =93walk=94. To make plurals or past=
 tenses in Lojban, you add separate words to the sentence that express the =
number of boys, or the time when the walking was going on.
+<p>
+<cx "morphology, derivational">  XE "morphology: derivational"   <cx "deri=
vational morphology, definition">  XE "derivational morphology: definition"=
   <cx "word forms, as related to grammatical uses">  XE "word forms: as re=
lated to grammatical uses"   However, Lojban does have what is called =93de=
rivational morphology=94: the capability of building new words from old wor=
ds. In addition, the form of words tells us something about their grammatic=
al uses, and sometimes about the means by which they entered the language. =
Lojban has very orderly rules for the formation of words of various types, =
both the words that already exist and new words yet to be created by speake=
rs and writers.
+<p>
+<cx "morphology, conventions for">  XE "morphology: conventions for"   A s=
tream of Lojban sounds can be uniquely broken up into its component words a=
ccording to specific rules. These so-called =93morphology rules=94 are summ=
arized in this chapter. (However, a detailed algorithm for breaking sounds =
into words has not yet been fully debugged, and so is not presented in this=
 book.) First, here are some conventions used to talk about groups of Lojba=
n letters, including vowels and consonants.
+<p>
+<cx "morphology, symbolic conventions for discussing">  XE "morphology: sy=
mbolic conventions for discussing"   <cx "=93V=94, as a symbol for a single=
 vowel">  XE "V: as a symbol for a single vowel"   <cx "=93y=94, as not a v=
owel for morphological discussions">  XE "y: considered not to be a vowel f=
or morphological discussions"   <dl compact><dt>1)	<dd>V represents any sin=
gle Lojban vowel except =93y=94; that is, it represents =93a=94, =93e=94, =
=93i=94, =93o=94, or =93u=94.
+</dl><cx "=93VV=94, as a symbol for a double vowel">  XE "VV string: as a =
symbol for a double vowel"   <dl compact><dt>2)	<dd>VV represents either a =
diphthong, one of the following:
+<p><dt>	<dd>ai ei oi au
+	<p><dt><dd>	or a two-syllable vowel pair with an apostrophe separating th=
e vowels, one of the following:
+<p><dt>	<dd>a'a a'e a'i a'o a'u
+	e'a e'e e'i e'o e'u
+	i'a i'e i'i i'o i'u
+	o'a o'e o'i o'o o'u
+	u'a u'e u'i u'o u'u
+</dl><cx "=93C=94, as a symbol for a single consonant">  XE "C string: as =
a symbol for a single consonant"   <cx "syllabic =93l=94, as a consonant fo=
r morphological discussions">  XE "syllabic l: considered as a consonant fo=
r morphological discussions"   <cx "syllabic =93m=94, as a consonant for mo=
rphological discussions">  XE "syllabic m: as a consonant for morphological=
 discussions"   <cx "syllabic =93n=94, as a consonant for morphological dis=
cussions">  XE "syllabic n: as a consonant for morphological discussions"  =
 <cx "syllabic =93r=94, as a consonant for morphological discussions">  XE =
"syllabic r: as a consonant for morphological discussions"   <cx "apostroph=
e, as not a consonant for morphological discussions">  XE "apostrophe: as n=
ot a consonant for morphological discussions"   <dl compact><dt>3)	<dd>C re=
presents a single Lojban consonant, not including the apostrophe, one of =
=93b=94, =93c=94, =93d=94, =93f=94, =93g=94, =93j=94, =93k=94, =93l=94, =93=
m=94, =93n=94, =93p=94, =93r=94, =93s=94, =93t=94, =93v=94, =93x=94, or =93=
z=94. Syllabic =93l=94, =93m=94, =93n=94, and =93r=94 always count as conso=
nants for the purposes of this chapter.
+
+</dl><cx "=93CC=94, as a symbol for a permissible initial consonant pair">=
  XE "CC string: as a symbol for a permissible initial consonant pair"   <d=
l compact><dt>4)	<dd>CC represents two adjacent consonants of type C which =
constitute one of the 48 permissible initial consonant pairs:
+<p><dt>	<dd>bl br
+	cf ck cl cm cn cp cr ct
+	dj dr dz
+	fl fr
+	gl gr
+	jb jd jg jm jv
+	kl kr
+	ml mr
+	pl pr
+	sf sk sl sm sn sp sr st
+	tc tr ts
+	vl vr
+	xl xr
+	zb zd zg zm zv
+</dl><cx "=93C/C=94, as a symbol for a permissible consonant pair">  XE "C=
/C string: as a symbol for a permissible consonant pair"   <dl compact><dt>=
5)	<dd>C/C represents two adjacent consonants which constitute one of the p=
ermissible consonant pairs (not necessarily a permissible initial consonant=
 pair). The permissible consonant pairs are explained in <a href=3Dchap3.ht=
ml>Chapter 3</a>. In brief, any consonant pair is permissible unless it: co=
ntains two identical letters, contains both a voiced (excluding =93r=94, =
=93l=94, =93m=94, =93n=94) and and an unvoiced consonant, or is one of cert=
ain specified forbidden pairs.
+</dl><cx "=93C/CC=94, as a symbol for a consonant triple">  XE "C/CC strin=
g: as a symbol for a consonant triple"   <dl compact><dt>6)	<dd>C/CC repres=
ents a consonant triple. The first two consonants must constitute a permiss=
ible consonant pair; the last two consonants must constitute a permissible =
initial consonant pair.
+</dl><cx "word classes">  XE "word classes"   <cx "parts of speech">  XE "=
parts of speech"   <cx "cmavo, as one of the 3 basic word classes">  XE "cm=
avo: as one of the 3 basic word classes"   <cx "cmene, as one of the 3 basi=
c word classes">  XE "cmene: as one of the 3 basic word classes"   <cx "bri=
vla, as one of the 3 basic word classes">  XE "brivla: as one of the 3 basi=
c word classes"   Lojban has three basic word classes =97 parts of speech =
=97 in contrast to the eight that are traditional in English. These three c=
lasses are called cmavo, brivla, and cmene. Each of these classes has uniqu=
ely identifying properties =97 an arrangement of letters that allows the wo=
rd to be uniquely and unambiguously recognized as a separate word in a stri=
ng of Lojban, upon either reading or hearing, and as belonging to a specifi=
c word-class.
+<p>
+They are also functionally different: cmavo are the structure words, corre=
sponding to English words like =93and=94, =93if=94, =93the=94 and =93to=94;=
 brivla are the content words, corresponding to English words like =93come=
=94, =93red=94, =93doctor=94, and =93freely=94; cmene are proper names, cor=
responding to English =93James=94, =93Afghanistan=94, and =93Pope John Paul=
 II=94.
+<p>
+<a name=3Ds2><h3>cmavo</h3>
+<p>
+<cx "cmavo, definition">  XE "cmavo: definition"   <cx "selma'o, definitio=
n">  XE "selma'o: definition"   <cx "structure words">  XE "structure words=
"   <cx "articles, cmavo as Lojban equivalents">  XE "articles: cmavo as Lo=
jban equivalents"   <cx "conjunctions, cmavo as Lojban equivalents">  XE "c=
onjunctions: cmavo as Lojban equivalents"   <cx "prepositions, cmavo as Loj=
ban equivalents">  XE "prepositions: cmavo as Lojban equivalents"   <cx "nu=
mbers, cmavo as Lojban equivalents">  XE "numbers: cmavo as Lojban equivale=
nts"   <cx "punctuation marks, cmavo as Lojban equivalents">  XE "punctuati=
on marks: cmavo as Lojban equivalents"   The first group of Lojban words di=
scussed in this chapter are the cmavo. They are the structure words that ho=
ld the Lojban language together. They often have no semantic meaning in the=
mselves, though they may affect the semantics of brivla to which they are a=
ttached. The cmavo include the equivalent of English articles, conjunctions=
, prepositions, numbers, and punctuation marks. There are over a hundred su=
bcategories of cmavo, known as =93selma'o=94, each having a specifically de=
fined grammatical usage. The various selma'o are discussed throughout <a hr=
ef=3Dchap5.html>Chapters 5 </a>to <a href=3Dchap19.html>19 </a>and summariz=
ed in <a href=3Dchap20.html>Chapter 20</a>.
+<p>
+<cx "cmavo, structure of">  XE "cmavo: structure of"   Standard cmavo occu=
r in four forms defined by their word structure. Here are some examples of =
the various forms:
+<p>
+<pre> 	V-form		.a	.e	.i	.o	.u
+	CV-form	ba	ce	di	fo	gu
+	VV-form	.au	.ei	.ia	.o'u	.u'e
+	CVV-form	ki'a	pei	mi'o	coi	cu'u
+</pre><p>In addition, there is the cmavo =93.y.=94 (remember that =93y=94 =
is not a V), which must have pauses before and after it.
+<p>
+<cx "cmavo, simple">  XE "cmavo: simple"   <cx "cmavo, for experimental us=
e">  XE "cmavo: for experimental use"   <cx "experimental cmavo, forms for"=
>  XE "experimental cmavo: forms for"   <cx "cmavo, lack of relation of for=
m to grammatical use">  XE "cmavo: lack of relation of form to grammatical =
use"   A simple cmavo thus has the property of having only one or two vowel=
s, or of having a single consonant followed by one or two vowels. Words con=
sisting of three or more vowels in a row, or a single consonant followed by=
 three or more vowels, are also of cmavo form, but are reserved for experim=
ental use: a few examples are =93ku'a'e=94, =93sau'e=94, and =93bai'ai=94. =
All CVV cmavo beginning with the letter =93x=94 are also reserved for exper=
imental use. In general, though, the form of a cmavo tells you little or no=
thing about its grammatical use.
+<p>
+<cx "cmavo, experimental">  XE "cmavo: experimental"   <cx "experimental c=
mavo, definition">  XE "experimental cmavo: definition"   =93Experimental u=
se=94 means that the language designers will not assign any standard meanin=
g or usage to these words, and words and usages coined by Lojban speakers w=
ill not appear in official dictionaries for the indefinite future. Experime=
ntal-use words provide an escape hatch for adding grammatical mechanisms (a=
s opposed to semantic concepts) the need for which was not foreseen.
+<p>
+<cx "cmavo, diphthongs in">  XE "cmavo: diphthongs in"   The cmavo of VV-f=
orm include not only the diphthongs and vowel pairs listed in <a href=3D#s1=
>Section 1</a>, but also the following ten additional diphthongs:
+<p>
+<pre>		.ia	.ie	.ii	.io	.iu
+		.ua	.ue	.ui	.uo	.uu
+</pre><p>In addition, cmavo can have the form =93Cy=94, a consonant follow=
ed by the letter =93y=94. These cmavo represent letters of the Lojban alpha=
bet, and are discussed in detail in <a href=3Dchap17.html>Chapter 17</a>.
+<p>
+<cx "cmavo, compound">  XE "cmavo: compound"   <cx "compound cmavo, defini=
tion">  XE "compound cmavo: definition"   <cx "compound cmavo, compared wit=
h sequence of simple cmavo">  XE "compound cmavo: compared with sequence of=
 simple cmavo"   Compound cmavo are sequences of cmavo attached together to=
 form a single written word. A compound cmavo is always identical in meanin=
g and in grammatical use to the separated sequence of simple cmavo from whi=
ch it is composed. These words are written in compound form merely to save =
visual space, and to ease the reader's burden in identifying when the compo=
nent cmavo are acting together.
+<p>
+<cx "compound cmavo, recognition of">  XE "compound cmavo: recognition of"=
   Compound cmavo, while not visually short like their components, can be r=
eadily identified by two characteristics:
+<p>
+<dl compact><dt>1)	<dd>They have no consonant pairs or clusters, and
+<p><dt>2)	<dd>They end in a vowel.
+</dl><p>For example:
+<p>
+<pre><a name=3De2d1>2.1)	.iseci'i
+	.i se ci'i
+<a name=3De2d2>2.2)	punaijecanai
+	pu nai je ca nai
+<a name=3De2d3>2.3)	ki'e.u'e
+	ki'e .u'e
+</pre><cx "pauses, before vowels">  XE "pauses: before vowels"   The cmavo=
 =93.u'e=94 begins with a vowel, and like all words beginning with a vowel,=
 requires a pause (represented by =93.=94) before it. This pause cannot be =
omitted simply because the cmavo is incorporated into a compound cmavo. On =
the other hand,
+<p>
+<pre><a name=3De2d4>2.4)	ki'e'u'e
+</pre>is a single cmavo reserved for experimental purposes: it has four vo=
wels.
+<p>
+<pre><a name=3De2d5>2.5)	cy.ibu.abu
+	cy. .ibu .abu
+</pre><p>Again the pauses are required (see <a href=3D#s9>Section 9</a>); =
the pause after =93cy.=94 merges with the pause before =93.ibu=94.
+<p>
+<cx "cmavo, stress on">  XE "cmavo: stress on"   <cx "stress, on cmavo">  =
XE "stress: on cmavo"   There is no particular stress required in cmavo or =
their compounds. Some conventions do exist that are not mandatory. For two-=
syllable cmavo, for example, stress is typically placed on the first vowel;=
 an example is
+<p>
+<pre><a name=3De2d6>2.6)	.e'o ko ko kurji
+	.E'o ko ko KURji
+</pre><p>This convention results in a consistent rhythm to the language, s=
ince brivla are required to have penultimate stress; some find this estheti=
cally pleasing.
+<p>
+<cx "pause, requirement between stressed syllables">  XE "pause: requireme=
nt between stressed syllables"   If the final syllable of one word is stres=
sed, and the first syllable of the next word is stressed, you must insert a=
 pause or glottal stop between the two stressed syllables. Thus
+<p>
+<pre><a name=3De2d7>2.7)	le re nanmu
+</pre>can be optionally pronounced
+<p>
+<pre><a name=3De2d8>2.8)	le RE. NANmu
+</pre>since there are no rules forcing stress on either of the first two w=
ords; the stress on =93re=94, though, demands that a pause separate =93re=
=94 from the following syllable =93nan=94 to ensure that the stress on =93n=
an=94 is properly heard as a stressed syllable. The alternative pronunciati=
on
+<p>
+<pre><a name=3De2d9>2.9)	LE re NANmu
+</pre>is also valid; this would apply secondary stress (used for purposes =
of emphasis, contrast or sentence rhythm) to =93le=94, comparable in rhythm=
ical effect to the English phrase =93THE two men=94. In <a href=3D#e2d8>Exa=
mple 2.8</a>, the secondary stress on =93re=94 would be similar to that in =
the English phrase =93the TWO men=94.
+<p>
+Both cmavo may also be left unstressed, thus:
+<p>
+<pre><a name=3De2d10>2.10)	le re NANmu
+</pre><p>This would probably be the most common usage.
+<p>
+<a name=3Ds3><h3>brivla</h3>
+<p>
+<cx "brivla, definition">  XE "brivla: definition"   <cx "nouns, brivla as=
 Lojban equivalents">  XE "nouns: brivla as Lojban equivalents"   <cx "adje=
ctives, brivla as Lojban equivalents">  XE "adjectives: brivla as Lojban eq=
uivalents"   <cx "verbs, brivla as Lojban equivalents">  XE "verbs: brivla =
as Lojban equivalents"   <cx "adverbs, brivla as Lojban equivalents">  XE "=
adverbs: brivla as Lojban equivalents"   Predicate words, called =93brivla=
=94, are at the core of Lojban. They carry most of the semantic information=
 in the language. They serve as the equivalent of English nouns, verbs, adj=
ectives, and adverbs, all in a single part of speech.
+<p>
+<cx "types and subtypes of words">  XE "types and subtypes of words"   <cx=
 "subtypes of words">  XE "subtypes of words"   Every brivla belongs to one=
 of three major subtypes. These subtypes are defined by the form, or morpho=
logy, of the word =97 all words of a particular structure can be assigned b=
y sight or sound to a particular type (cmavo, brivla, or cmene) and subtype=
. Knowing the type and subtype then gives you, the reader or listener, sign=
ificant clues to the meaning and the origin of the word, even if you have n=
ever heard the word before.
+<p>
+<cx "flexible vocabulary">  XE "flexible vocabulary"   The same principle =
allows you, when speaking or writing, to invent new brivla for new concepts=
 =93on the fly=94; yet it offers people that you are trying to communicate =
with a good chance to figure out your meaning. In this way, Lojban has a fl=
exible vocabulary which can be expanded indefinitely.
+<p>
+<cx "brivla, properties of">  XE "brivla: properties of"   All brivla have=
 the following properties:
+<p>
+<dl compact><dt>1)	<dd>always end in a vowel;
+<p><dt>2)	<dd>always contain a consonant pair in the first five letters, w=
here =93y=94 and apostrophe are not counted as letters for this purpose (Se=
e <a href=3D#s6>Section 6</a>);
+<p><dt>3)	<dd>always are stressed on the next-to-the-last (penultimate) sy=
llable; this implies that they have two or more syllables.
+</dl><cx "brivla, recognition of">  XE "brivla: recognition of"   <cx "bri=
vla form, contrasted with cmavo form">  XE "brivla form: contrasted with cm=
avo form"   <cx "brivla form, contrasted with cmene form">  XE "brivla form=
: contrasted with cmene form"   <cx "cmavo form, contrasted with brivla for=
m">  XE "cmavo form: contrasted with brivla form"   <cx "cmene form, contra=
sted with brivla form">  XE "cmene form: contrasted with brivla form"   The=
 presence of a consonant pair distinguishes brivla from cmavo and their com=
pounds. The final vowel distinguishes brivla from cmene, which always end i=
n a consonant. Thus =93da'amei=94 must be a compound cmavo because it lacks=
 a consonant pair; =93lojban.=94 must be a name because it lacks a final vo=
wel.
+<p>
+<cx "brivla, consonant pairs in">  XE "brivla: consonant pairs in"   <cx "=
consonant pairs, in brivla">  XE "consonant pairs: in brivla"   <cx "=93y=
=94, between letters of consonant pair">  XE "y: letter: between letters of=
 consonant pair"   <cx "consonant pairs, =93y=94 within">  XE "consonant pa=
irs: letter y within"   Thus, =93bisycla=94 has the consonant pair =93sc=94=
 in the first five non-=93y=94 letters even though the =93sc=94 actually ap=
pears in the form of =93syc=94. Similarly, the word =93ro'inre'o=94 contain=
s =93nr=94 in the first five letters because the apostrophes are not counte=
d for this purpose.
+<p>
+<cx "brivla, subtypes of">  XE "brivla: subtypes of"   The three subtypes =
of brivla are:
+<p>
+<cx "primitive roots, gismu as">  XE "primitive roots: gismu as"   <cx "gi=
smu, as a subtype of brivla">  XE "gismu: as a subtype of brivla"   <dl com=
pact><dt>1)	<dd>gismu, the Lojban primitive roots from which all other briv=
la are built;
+</dl><cx "lujvo, as a subtype of brivla">  XE "lujvo: as a subtype of briv=
la"   <cx "compound of gismu, lujvo as">  XE "compound of gismu: lujvo as" =
  <dl compact><dt>2)	<dd>lujvo, the compounds of two or more gismu; and
+</dl><cx "fu'ivla, as a subtype of brivla">  XE "fu'ivla: as a subtype of =
brivla"   <cx "borrowing from other language, fu'ivla as">  XE "borrowing f=
rom other language: fu'ivla as"   <pre>3)	fu'ivla (literally =93copy-word=
=94), the specialized words that are not Lojban primitives or natural compo=
unds, and are therefore borrowed from other languages.
+</pre><a name=3Ds4><h3>gismu</h3>
+<p>
+<cx "gismu, definition">  XE "gismu: definition"   <cx "gismu, rationale f=
or choice of">  XE "gismu: rationale for choice of"   The gismu, or Lojban =
root words, are those brivla representing concepts most basic to the langua=
ge. The gismu were chosen for various reasons: some represent concepts that=
 are very familiar and basic; some represent concepts that are frequently u=
sed in other languages; some were added because they would be helpful in co=
nstructing more complex words; some because they represent fundamental Lojb=
an concepts (like =93cmavo=94 and =93gismu=94 themselves).
+<p>
+<cx "gismu, selection of">  XE "gismu: selection of"   <cx "gismu, as part=
itioning semantic space">  XE "gismu: as partitioning semantic space"   The=
 gismu do not represent any sort of systematic partitioning of semantic spa=
ce. Some gismu may be superfluous, or appear for historical reasons: the gi=
smu list was being collected for almost 35 years and was only weeded out on=
ce. Instead, the intention is that the gismu blanket semantic space: they m=
ake it possible to talk about the entire range of human concerns.
+<p>
+<cx "gismu and cmavo, major">  XE =93gismu: and cmavo, major"   <cx "cmavo=
 and gismu, major">  XE "cmavo and gismu: major"   <cx "learning Lojban, ma=
gnitude of task">  XE "learning Lojban: magnitude of task"   There are abou=
t 1350 gismu. In learning Lojban, you need only to learn most of these gism=
u and their combining forms (known as =93rafsi=94) as well as perhaps 200 m=
ajor cmavo, and you will be able to communicate effectively in the language=
. This may sound like a lot, but it is a small number compared to the vocab=
ulary needed for similar communications in other languages.
+<p>
+<cx "gismu, rules for">  XE "gismu: rules for"   All gismu have very stron=
g form restrictions. Using the conventions defined in <a href=3D#s1>Section=
 1</a>, all gismu are of the forms CVC/CV or CCVCV. They must meet the rule=
s for all brivla given in <a href=3D#s3>Section 3</a>; furthermore, they:
+<p>
+<dl compact><dt>1)	<dd>always have five letters;
+<p><dt>2)	<dd>always start with a consonant and end with a single vowel;
+<p><dt>3)	<dd>always contain exactly one consonant pair, which is a permis=
sible initial pair (CC) if it's at the beginning of the gismu, but otherwis=
e only has to be a permissible pair (C/C);
+<p><dt>4)	<dd>are always stressed on the first syllable (since that is pen=
ultimate).
+</dl><cx "gismu, length of">  XE "gismu: length of"   The five letter leng=
th distinguishes gismu from lujvo and fu'ivla. (It is possible to have fu'i=
vla like =93spa'i=94 that are five letters long, but they must have =93=A0'=
=A0=94; no gismu contains =93=A0'=A0=94.)
+<p>
+<cx "gismu, special">  XE "gismu: special"   <cx "gismu, conflicts between=
">  XE "gismu: conflicts between"   With the exception of five special briv=
la variables, =93broda=94, =93brode=94, =93brodi=94, =93brodo=94, and =93br=
odu=94, no two gismu differ only in the final vowel. Furthermore, the set o=
f gismu was specifically designed to reduce the likelihood that two similar=
 sounding gismu could be confused. For example, because =93gismu=94 is in t=
he set of gismu, =93kismu=94, =93xismu=94, =93gicmu=94, =93gizmu=94, and =
=93gisnu=94 cannot be.
+<p>
+<cx "gismu, source of">  XE "gismu: source of"   Almost all Lojban gismu a=
re constructed from pieces of words drawn from other languages, specificall=
y Chinese, English, Hindi, Spanish, Russian, and Arabic, the six most widel=
y spoken natural languages. For a given concept, words in the six languages=
 that represent that concept were written in Lojban phonetics. Then a gismu=
 was selected to maximize the recognizability of the Lojban word for speake=
rs of the six languages by weighting the inclusion of the sounds drawn from=
 each language by the number of speakers of that language. See <a href=3D#s=
14>Section 14 </a>for a full explanation of the algorithm.
+<p>
+<cx "gismu, examples of">  XE "gismu: examples of"   Here are a few exampl=
es of gismu, with rough English equivalents (not definitions):
+<p>
+<pre><a name=3De4d1>4.1)	creka
+	shirt
+<a name=3De4d2>4.2)	lijda
+	religion
+<a name=3De4d3>4.3)	blanu
+	blue
+<a name=3De4d4>4.4)	mamta
+	mother
+<a name=3De4d5>4.5)	cukta
+	book
+<a name=3De4d6>4.6)	patfu
+	father
+<a name=3De4d7>4.7)	nanmu
+	man
+<a name=3De4d8>4.8)	ninmu
+	woman
+<p>
+</pre><p>A small number of gismu were formed differently; see <a href=3D#s=
15>Section 15 </a>for a list.
+<p>
+<a name=3Ds5><h3>lujvo</h3>
+<p>
+<cx "tanru, explanation of">  XE "tanru: explanation of"   <cx "seltau, co=
mpared with English adjective">  XE "seltau: compared with English adjectiv=
e"   <cx "seltau, compared with English adverb">  XE "seltau: compared with=
 English adverb"   <cx "modifying brivla (see also seltau)">  XE "modifying=
 brivla (see also seltau)"   When specifying a concept that is not found am=
ong the gismu (or, more specifically, when the relevant gismu seems too gen=
eral in meaning), a Lojbanist generally attempts to express the concept as =
a tanru. Lojban tanru are an elaboration of the concept of =93metaphor=94 u=
sed in English. In Lojban, any brivla can be used to modify another brivla.=
 The first of the pair modifies the second. This modification is usually re=
strictive =97 the modifying brivla reduces the broader sense of the modifie=
d brivla to form a more narrow, concrete, or specific concept. Modifying br=
ivla may thus be seen as acting like English adverbs or adjectives. For exa=
mple,
+<p>
+<pre><a name=3De5d1>5.1)	skami pilno
+</pre>is the tanru which expresses the concept of =93computer user=94.
+<p>
+<cx "tanru, combination of">  XE "tanru: combination of"   The simplest Lo=
jban tanru are pairings of two concepts or ideas. Such tanru take two simpl=
er ideas that can be represented by gismu and combine them into a single mo=
re complex idea. Two-part tanru may then be recombined in pairs with other =
tanru, or with individual gismu, to form more complex or more specific idea=
s, and so on.
+<p>
+<cx "tanru, ambiguity of">  XE "tanru: ambiguity of"   The meaning of a ta=
nru is usually at least partly ambiguous: =93skami pilno=94 could refer to =
a computer that is a user, or to a user of computers. There are a variety o=
f ways that the modifier component can be related to the modified component=
. It is also possible to use cmavo within tanru to provide variations (or t=
o prevent ambiguities) of meaning.
+<p>
+<cx "tanru, and creativity">  XE "tanru: and creativity"   Making tanru is=
 essentially a poetic or creative act, not a science. While the syntax expr=
essing the grouping relationships within tanru is unambiguous, tanru are st=
ill semantically ambiguous, since the rules defining the relationships betw=
een the gismu are flexible. The process of devising a new tanru is dealt wi=
th in detail in <a href=3Dchap5.html>Chapter 5</a>.
+<p>
+<cx "tanru, expression of">  XE "tanru: expression of"   <ex "big boat">  =
XE "big boat: example=93  <ex "father mother">  XE "father mother: example=
=93  <ex "mother father">  XE "mother father: example=93  To express a simp=
le tanru, simply say the component gismu together. Thus the binary metaphor=
 =93big boat=94 becomes the tanru
+<p>
+<pre><a name=3De5d2>5.2)	barda bloti
+</pre>representing roughly the same concept as the English word =93ship=94=
.
+<p>
+<ex "paternal grandmother">  XE "paternal grandmother: example=93  The bin=
ary metaphor =93father mother=94 can refer to a paternal grandmother (=93a =
father-ly type of mother=94), while =93mother father=94 can refer to a mate=
rnal grandfather (=93a mother-ly type of father=94). In Lojban, these becom=
e the tanru
+<p>
+<pre><a name=3De5d3>5.3)	patfu mamta
+</pre>and
+<p>
+<pre><a name=3De5d4>5.4)	mamta patfu
+</pre>respectively.
+<p>
+<cx "tanru, ambiguity in">  XE "tanru: ambiguity in"   The possibility of =
semantic ambiguity can easily be seen in the last case. To interpret <a hre=
f=3D#e5d4>Example 5.4</a>, the listener must determine what type of motherl=
iness pertains to the father being referred to. In an appropriate context, =
=93mamta patfu=94 could mean not =93grandfather=94 but simply =93father wit=
h some motherly attributes=94, depending on the culture. If absolute clarit=
y is required, there are ways to expand upon and explain the exact interrel=
ationship between the components; but such detail is usually not needed.
+<p>
+<cx "brivla, from tanru">  XE "brivla: from tanru"   <cx "lujvo, from tanr=
u">  XE "lujvo: from tanru"   <cx "tanru, to lujvo">  XE "tanru: to lujvo" =
  When a concept expressed in a tanru proves useful, or is frequently expre=
ssed, it is desirable to choose one of the possible meanings of the tanru a=
nd assign it to a new brivla. For <a href=3D#e5d1>Example 5.1</a>, we would=
 probably choose =93user of computers=94, and form the new word
+<p>
+<pre><a name=3De5d5>5.5)	sampli
+</pre><p>Such a brivla, built from the rafsi which represent its component=
 words, is called a =93lujvo=94. Another example, corresponding to the tanr=
u of <a href=3D#e5d2>Example 5.2</a>, would be:
+<p>
+<pre><a name=3De5d6>5.6)	bralo'i
+	big-boat
+	ship
+</pre><cx "lujvo, unambiguous decomposition of">  XE "lujvo: unambiguous d=
ecomposition of"   <cx "rafsi, definition">  XE "rafsi: definition"   <cx "=
lujvo, construction of">  XE "lujvo: construction of"   The lujvo represent=
ing a given tanru is built from units representing the component gismu. The=
se units are called =93rafsi=94 in Lojban. Each rafsi represents only one g=
ismu. The rafsi are attached together in the order of the words in the tanr=
u, occasionally inserting so-called =93hyphen=94 letters to ensure that the=
 pieces stick together as a single word and cannot accidentally be broken a=
part into cmavo, gismu, or other word forms. As a result, each lujvo can be=
 readily and accurately recognized, allowing a listener to pick out the wor=
d from a string of spoken Lojban, and if necessary, unambiguously decompose=
 the word to a unique source tanru, thus providing a strong clue to its mea=
ning.
+<p>
+The lujvo that can be built from the tanru =93mamta patfu=94 in <a href=3D=
#e5d4>Example 5.4 </a>is
+<p>
+<pre><a name=3De5d7>5.7)	mampa'u
+</pre>which refers specifically to the concept =93maternal grandfather=94.=
 The two gismu that constitute the tanru are represented in =93mampa'u=94 b=
y the rafsi =93mam-=94 and =93-pa'u=94, respectively; these two rafsi are t=
hen concatenated together to form =93mampa'u=94.
+<p>
+<cx "lujvo, meaning of">  XE "lujvo: meaning of"   <cx "lujvo, multiple fo=
rms of">  XE "lujvo: multiple forms of"   <cx "rafsi form, effect of choice=
 on meaning of lujvo">  XE "rafsi form: effect of choice on meaning of lujv=
o"   <cx "short rafsi form, compared with long form in effect on lujvo mean=
ing">  XE "short rafsi form: compared with long form in effect on lujvo mea=
ning"   <cx "long rafsi form, compared with short form in effect on lujvo m=
eaning">  XE "long rafsi form: compared with short form in effect on lujvo =
meaning"   Like gismu, lujvo have only one meaning. When a lujvo is formall=
y entered into a dictionary of the language, a specific definition will be =
assigned based on one particular interrelationship between the terms. (See =
<a href=3Dchap12.html>Chapter 12 </a>for how this has been done.) Unlike gi=
smu, lujvo may have more than one form. This is because there is no differe=
nce in meaning between the various rafsi for a gismu when they are used to =
build a lujvo. A long rafsi may be used, especially in noisy environments, =
in place of a short rafsi; the result is considered the same lujvo, even th=
ough the word is spelled and pronounced differently. Thus the word =93brivl=
a=94, built from the tanru =93bridi valsi=94, is the same lujvo as =93briva=
lsi=94, =93bridyvla=94, and =93bridyvalsi=94, each of which uses a differen=
t combination of rafsi.
+<p>
+<cx "lujvo, rules for formation of">  XE "lujvo: rules for formation of"  =
 <cx "rafsi, rules for combining to form lujvo">  XE "rafsi: rules for comb=
ining to form lujvo"   <cx "=93y=94-hyphen, and consonant cluster determina=
tion">  XE "y-hyphen: and consonant cluster determination"   <cx "=93 ' =94=
, and consonant cluster determination in lujvo">  XE " ' symbol: and conson=
ant cluster determination in lujvo"   <cx "apostrophe, and consonant cluste=
r determination in lujvo">  XE "apostrophe: and consonant cluster determina=
tion in lujvo"   When assembling rafsi together into lujvo, the rules for v=
alid brivla must be followed: a consonant cluster must occur in the first f=
ive letters (excluding =93y=94 and =93=A0'=A0=94), and the lujvo must end i=
n a vowel.
+<p>
+<cx "=93y=94-hyphen, use of">  XE "y-hyphen: use of"   <cx "=93y=94-hyphen=
, and stress determination">  XE "y-hyphen: and stress determination"   <cx=
 "glue in lujvo, =93y=94-hyphen as">  XE "glue in lujvo: y-hyphen as"   A =
=93y=94 (which is ignored in determining stress or consonant clusters) is i=
nserted in the middle of the consonant cluster to glue the word together wh=
en the resulting cluster is either not permissible or the word is likely to=
 break up. There are specific rules describing these conditions, detailed i=
n <a href=3D#s6>Section 6</a>.
+<p>
+<cx "=93r=94-hyphen, use of">  XE "r-hyphen: use of"   <cx "glue in lujvo,=
 =93r=94-hyphen as">  XE "glue in lujvo: r-hyphen as"   <cx "=93n=94-hyphen=
, use of">  XE "n-hyphen: use of"   <cx "glue in lujvo, =93n=94-hyphen as">=
  XE "glue in lujvo: n-hyphen as"   An =93r=94 (in some cases, an =93n=94) =
is inserted when a CVV-form rafsi attaches to the beginning of a lujvo in s=
uch a way that there is no consonant cluster. For example, in the lujvo
+<p>
+<ex "field rations">  XE "field rations: example=93  <pre><a name=3De5d8>5=
.8)	soirsai
+	sonci sanmi
+	soldier meal
+	field rations
+</pre><cx "rafsi, contrasted with same-form cmavo in meaning">  XE "rafsi:=
 contrasted with same-form cmavo in meaning"   <cx "cmavo, contrasted with =
same-form rafsi in meaning">  XE "cmavo: contrasted with same-form rafsi in=
 meaning"   the rafsi =93soi-=94 and =93-sai=94 are joined, with the additi=
onal =93r=94 making up the =93rs=94 consonant pair needed to make the word =
a brivla. Without the =93r=94, the word would break up into =93soi sai=94, =
two cmavo. The pair of cmavo have no relation to their rafsi lookalikes; th=
ey will either be ungrammatical (as in this case), or will express a differ=
ent meaning from what was intended.
+<p>
+Learning rafsi and the rules for assembling them into lujvo is clearly see=
n to be necessary for fully using the potential Lojban vocabulary.
+<p>
+<cx "lujvo, invention of">  XE "lujvo: invention of"   Most important, it =
is possible to invent new lujvo while you speak or write in order to repres=
ent a new or unfamiliar concept, one for which you do not know any existing=
 Lojban word. As long as you follow the rules for building these compounds,=
 there is a good chance that you will be understood without explanation.
+<p>
+<a name=3Ds6><h3>rafsi</h3>
+<p>
+<cx "rafsi, use of">  XE "rafsi: use of"   <cx "rafsi, uniqueness in gismu=
 referent of">  XE "rafsi: uniqueness in gismu referent of"   <cx "rafsi, m=
ultiplicity of for single gismu">  XE "rafsi: multiplicity of for single gi=
smu"   <cx "rafsi, level of uniqueness of relation to gismu">  XE "rafsi: l=
evel of uniqueness of relation to gismu"   <cx "gismu, level of uniqueness =
of rafsi relating to">  XE "gismu: level of uniqueness of rafsi relating to=
"   <cx "rafsi, selection considerations in making lujvo">  XE "rafsi: sele=
ction considerations in making lujvo"   Every gismu has from two to five ra=
fsi, each of a different form, but each such rafsi represents only one gism=
u. It is valid to use any of the rafsi forms in building lujvo =97 whicheve=
r the reader or listener will most easily understand, or whichever is most =
pleasing =97 subject to the rules of lujvo making. There is a scoring algor=
ithm which is intended to determine which of the possible and legal lujvo f=
orms will be the standard dictionary form (see <a href=3D#s12>Section 12</a=
>).
+<p>
+<cx "rafsi, forms of">  XE "rafsi: forms of"   <cx "rafsi, long">  XE "raf=
si: long"   <cx "lujvo, unreduced">  XE "lujvo: unreduced"   <cx "gismu, ba=
sic rafsi for">  XE "gismu: basic rafsi for"   <cx "5-letter rafsi, definit=
ion">  XE "5-letter rafsi: definition"   <cx "4-letter rafsi, definition"> =
 XE "4-letter rafsi: definition"   <cx "long rafsi, definition">  XE "long =
rafsi: definition"   <cx "unreduced lujvo, definition">  XE "unreduced lujv=
o: definition"   Each gismu always has at least two rafsi forms; one is the=
 gismu itself (used only at the end of a lujvo), and one is the gismu witho=
ut its final vowel (used only at the beginning or middle of a lujvo). These=
 forms are represented as CVC/CV or CCVCV (called =93the 5-letter rafsi=94)=
, and CVC/C or CCVC (called =93the 4-letter rafsi=94) respectively. The das=
hes in these rafsi form representations show where other rafsi may be attac=
hed to form a valid lujvo. When lujvo are formed only from 4-letter and 5-l=
etter rafsi, known collectively as =93long rafsi=94, they are called =93unr=
educed lujvo=94.
+<p>
+Some examples of unreduced lujvo forms are:
+<p>
+<pre><a name=3De6d1>6.1)	mamtypatfu
+	from =93mamta patfu=94
+	=93mother father=94 or =93maternal grandfather=94
+<a name=3De6d2>6.2)	lerfyliste
+	from =93lerfu liste=94
+	=93letter list=94 or a =93list of letters=94
+	(letters of the alphabet)
+<a name=3De6d3>6.3)	nancyprali
+	from =93nanca prali=94
+	=93year profit=94 or =93annual profit=94
+<a name=3De6d4>6.4)	prunyplipe
+	from =93pruni plipe=94
+	=93elastic (springy) leap=94 or =93spring=94 (the verb)
+</pre><ex "supper">  XE "supper: example=93  <pre><a name=3De6d5>6.5)	vanc=
ysanmi
+	from =93vanci sanmi=94
+	=93evening meal=94 or =93supper=94
+</pre><cx "rafsi, short">  XE "rafsi: short"   <cx "short rafsi">  XE "sho=
rt rafsi"   In addition to these two forms, each gismu may have up to three=
 additional short rafsi, three letters long. All short rafsi have one of th=
e forms CVC, CCV, or CVV. The total number of rafsi forms that are assigned=
 to a gismu depends on how useful the gismu is, or is presumed to be, in ma=
king lujvo, when compared to other gismu that could be assigned the rafsi.
+<p>
+<cx "comparatives, use of zmadu in forming">  XE "comparatives: use of zma=
du in forming"   <cx "-er, use of zmadu in forming">  XE "-er: use of zmadu=
 in forming"   For example, =93zmadu=94 (=93more than=94) has the two short=
 rafsi =93zma=94 and =93mau=94 (in addition to its unreduced rafsi =93zmad=
=94 and =93zmadu=94), because a vast number of lujvo have been created base=
d on =93zmadu=94, corresponding in general to English comparative adjective=
s ending in =93-er=94 such as =93whiter=94 (Lojban =93labmau=94). On the ot=
her hand, =93bakri=94 (=93chalk=94) has no short rafsi and few lujvo.
+<p>
+There are at most one CVC-form, one CCV-form, and one CVV-form rafsi per g=
ismu. In fact, only a tiny handful of gismu have both a CCV-form and a CVV-=
form rafsi assigned, and still fewer have all three forms of short rafsi. H=
owever, gismu with both a CVC-form and another short rafsi are fairly commo=
n, partly because more possible CVC-form rafsi exist. Yet CVC-form rafsi, e=
ven though they are fairly easy to remember, cannot be used at the end of a=
 lujvo (because lujvo must end in vowels), so justifying the assignment of =
an additional short rafsi to many gismu.
+<p>
+<cx "rafsi space">  XE "rafsi space"   <cx "rafsi, rationale for assignmen=
ts of">  XE "rafsi: rationale for assignments of"   The intention was to us=
e the available =93rafsi space=94 =97 the set of all possible short rafsi f=
orms =97 in the most efficient way possible; the goal is to make the most-u=
sed lujvo as short as possible (thus maximizing the use of short rafsi), wh=
ile keeping the rafsi very recognizable to anyone who knows the source gism=
u. For this reason, the letters in a rafsi have always been chosen from amo=
ng the five letters of the corresponding gismu. As a result, there are a li=
mited set of short rafsi available for assignment to each gismu. At most se=
ven possible short rafsi are available for consideration (of which at most =
three can be used, as explained above).
+<p>
+<cx "rafsi, possible forms for construction of">  XE "rafsi: possible form=
s for construction of"   Here are the only short rafsi forms that can possi=
bly exist for gismu of the form CVC/CV, like =93sakli=94. The digits in the=
 second column represent the gismu letters used to form the rafsi.
+<p>
+<pre> 	CVC	123	-sak-
+	CVC	124	-sal-
+	CVV	12'5	-sa'i-
+	CVV	125	-sai-
+	CCV	345	-kli-
+	CCV	132	-ska-
+</pre>(The only actual short rafsi for =93sakli=94 is =93-sal-=94.)
+<p>
+For gismu of the form CCVCV, like =93blaci=94, the only short rafsi forms =
that can exist are:
+<p>
+<pre> 	CVC	134	-bac-
+	CVC	234	-lac
+	CVV	13'5	-ba'i-
+	CVV	135	-bai-
+	CVV	23'5	-la'i-
+	CVV	235	-lai-
+	CCV	123	-bla-
+</pre><cx "rafsi assignments, non-reassignability of">  XE "rafsi assignme=
nts: non-reassignability of"   (In fact, =93blaci=94 has none of these shor=
t rafsi; they are all assigned to other gismu. Lojban speakers are not free=
 to reassign any of the rafsi; the tables shown here are to help understand=
 how the rafsi were chosen in the first place.)
+<p>
+<cx "rafsi, considerations restricting construction of">  XE "rafsi: consi=
derations restricting construction of"   There are a few restrictions: a CV=
V-form rafsi without an apostrophe cannot exist unless the vowels make up o=
ne of the four diphthongs =93ai=94, =93ei=94, =93oi=94, or =93au=94; and a =
CCV-form rafsi is possible only if the two consonants form a permissible in=
itial consonant pair (see <a href=3D#s1>Section 1</a>). Thus =93mamta=94, w=
hich has the same form as =93salci=94, can only have =93mam=94, =93mat=94, =
and =93ma'a=94 as possible rafsi: in fact, only =93mam=94 is assigned to it=
.
+<p>
+<cx "rafsi for numbers">  XE "rafsi for numbers"   <cx "numbers, rafsi for=
">  XE "numbers: rafsi for"   Some cmavo also have associated rafsi, usuall=
y CVC-form. For example, the ten common numerical digits, which are all CV =
form cmavo, each have a CVC-form rafsi formed by adding a consonant to the =
cmavo. Most cmavo that have rafsi are ones used in composing tanru (for a c=
omplete list, see <a href=3Dchap12.html>Chapter 12</a>).
+<p>
+<cx "lujvo, fully reduced">  XE "lujvo: fully reduced"   <cx "fully reduce=
d lujvo, definition">  XE "fully reduced lujvo: definition"   The term for =
a lujvo made up solely of short rafsi is =93fully reduced lujvo=94. Here ar=
e some examples of fully reduced lujvo:
+<p>
+<pre><a name=3De6d6>6.6)	cumfri
+	from =93cumki lifri=94
+	=93possible experience=94
+<a name=3De6d7>6.7)	klezba
+	from =93klesi zbasu=94
+	=93category make=94
+<a name=3De6d8>6.8)	kixta'a
+	from =93krixa tavla=94
+	=93cry-out talk=94
+<a name=3De6d9>6.9)	sniju'o
+	from =93sinxa djuno=94
+	=93sign know=94
+</pre><p>In addition, some of the unreduced forms in the previous example =
may be fully reduced to:
+<p>
+<pre><a name=3De6d10>6.10)	mampa'u
+	from =93mamta patfu=94
+	=93mother father=94 or =93maternal grandfather=94
+<a name=3De6d11>6.11)	lerste
+	from =93lerfu liste=94
+	=93letter list=94 or a =93list of letters=94
+</pre><cx "lujvo, summary of form characteristics">  XE "lujvo: summary of=
 form characteristics"   <cx "lujvo form, final letter of">  XE "lujvo form=
: final letter of"   <cx "lujvo form, consonant cluster requirement in">  X=
E "lujvo form: consonant cluster requirement in"   <cx "lujvo form, number =
of letters in">  XE "lujvo form: number of letters in"   <cx "lujvo, recogn=
izing">  XE "lujvo: recognizing"   As noted above, CVC-form rafsi cannot ap=
pear as the final rafsi in a lujvo, because all lujvo must end with one or =
two vowels. As a brivla, a lujvo must also contain a consonant cluster with=
in the first five letters =97 this ensures that they cannot be mistaken for=
 compound cmavo. Of course, all lujvo have at least six letters since they =
have two or more rafsi, each at least three letters long; hence they cannot=
 be confused with gismu.
+<p>
+<cx "hyphens, use of">  XE "hyphens: use of"   <cx "hyphen letter, definit=
ion">  XE "hyphen letter: definition"   <cx "lujvo form, requirements for h=
yphen insertion in">  XE "lujvo form: requirements for hyphen insertion in"=
   When attaching two rafsi together, it may be necessary to insert a hyphe=
n letter. In Lojban, the term =93hyphen=94 always refers to a letter, eithe=
r the vowel =93y=94 or one of the consonants =93r=94 and =93n=94. (The lett=
er =93l=94 can also be a hyphen, but is not used as one in lujvo.)
+<p>
+<cx "lujvo form, requirements for =93y=94-hyphen insertion in">  XE "lujvo=
 form: requirements for y-hyphen insertion in"   The =93y=94-hyphen is used=
 after a CVC-form rafsi when joining it with the following rafsi could resu=
lt in an impermissible consonant pair, or when the resulting lujvo could fa=
ll apart into two or more words (either cmavo or gismu).
+<p>
+<cx "lujvo, and consonant pairs">  XE "lujvo: and consonant pairs"   Thus,=
 the tanru =93pante tavla=94 (=93protest talk=94) cannot produce the lujvo =
=93patta'a=94, because =93tt=94 is not a permissible consonant pair; the lu=
jvo must be =93patyta'a=94. Similarly, the tanru =93mudri siclu=94 (=93wood=
en whistle=94) cannot form the lujvo =93mudsiclu=94; instead, =93mudysiclu=
=94 must be used. (Remember that =93y=94 is not counted in determining whet=
her the first five letters of a brivla contain a consonant cluster: this is=
 why.)
+<p>
+<cx "rafsi, four-letter, requirement for =93y=94-hyphen">  XE "rafsi: four=
-letter, requirement for y-hyphen"   The =93y=94-hyphen is also used to att=
ach a 4-letter rafsi, formed by dropping the final vowel of a gismu, to the=
 following rafsi. (This procedure was shown, but not explained, in <a href=
=3D#e6d{1}>Examples 6.1 </a>to <a href=3D#e6d5>6.5</a>.)
+The lujvo forms =93zunlyjamfu=94, =93zunlyjma=94, =93zuljamfu=94, and =93z=
uljma=94 are all legitimate and equivalent forms made from the tanru =93zun=
le jamfu=94 (=93left foot=94). Of these, =93zuljma=94 is the preferred one =
since it is the shortest; it thus is likely to be the form listed in a Lojb=
an dictionary.
+<p>
+<cx "=93r=94-hyphen, use of">  XE "r-hyphen: use of"   <cx "lujvo form, re=
quirements for =93r=94-hyphen insertion in">  XE "lujvo form: requirements =
for r-hyphen insertion in"   <cx "lujvo form, requirements for =93n=94-hyph=
en insertion in">  XE "lujvo form: requirements for n-hyphen insertion in" =
  The =93r=94-hyphen and its close relative, the =93n=94-hyphen, are used i=
n lujvo only after CVV-form rafsi. A hyphen is always required in a two-par=
t lujvo of the form CVV-CVV, since otherwise there would be no consonant cl=
uster.
+<p>
+An =93r-=94hyphen or =93n=94-hyphen is also required after the CVV-form ra=
fsi of any lujvo of the form CVV-CVC/CV or CVV-CCVCV since it would otherwi=
se fall apart into a CVV-form cmavo and a gismu. In any lujvo with more tha=
n two parts, a CVV-form rafsi in the initial position must always be follow=
ed by a hyphen. If the hyphen were to be omitted, the supposed lujvo could =
be broken into smaller words without the hyphen: because the CVV-form rafsi=
 would be interpreted as a cmavo, and the remainder of the word as a valid =
lujvo that is one rafsi shorter.
+<p>
+<cx "=93n=94-hyphen, use of">  XE "n-hyphen: use of"   <cx "=93n=94-hyphen=
, contrasted with =93r=94-hyphen in requirements for use">  XE "n-hyphen: c=
ontrasted with r-hyphen in requirements for use"   <cx "=93r=94-hyphen, con=
trasted with =93n=94-hyphen in requirements for use">  XE "r-hyphen: contra=
sted with n-hyphen in requirements for use"   An =93n=94-hyphen is only use=
d in place of an =93r=94-hyphen when the following rafsi begins with =93r=
=94. For example, the tanru =93rokci renro=94 (=93rock throw=94) cannot be =
expressed as =93ro'ire'o=94 (which breaks up into two cmavo), nor can it be=
 =93ro'irre'o=94 (which has an impermissible double consonant); the =93n=94=
-hyphen is required, and the correct form of the hyphenated lujvo is =93ro'=
inre'o=94. The same lujvo could also be expressed without hyphenation as =
=93rokre'o=94.
+<p>
+<lx "zei">  XE "zei"   <lx "ZEI">  XE "ZEI selma=92o"   <cx "lujvo, with z=
ei">  XE "lujvo: with zei"   There is also a different way of building lujv=
o, or rather phrases which are grammatically and semantically equivalent to=
 lujvo. You can make a phrase containing any desired words, joining each pa=
ir of them with the special cmavo =93zei=94. Thus,
+<p>
+<pre><a name=3De6d12>6.12)	bridi zei valsi
+</pre><cx "lujvo, from cmavo with no rafsi">  XE "lujvo: from cmavo with n=
o rafsi"   <cx "rafsi, lack of, effect on forming lujvo">  XE "rafsi: lack =
of, effect on forming lujvo"   <cx "cmene, method of including in lujvo">  =
XE "cmene: method of including in lujvo"   <cx "fu'ivla, method of includin=
g in lujvo">  XE "fu'ivla: method of including in lujvo"   <cx "cmavo witho=
ut rafsi, method of including in lujvo">  XE "cmavo without rafsi: method o=
f including in lujvo"   is the exact equivalent of =93brivla=94 (but not ne=
cessarily the same as the underlying tanru =93bridi valsi=94, which could h=
ave other meanings.) Using =93zei=94 is the only way to get a cmavo lacking=
 a rafsi, a cmene, or a fu'ivla into a lujvo:
+<p>
+<ex "X-ray">  XE "X-ray: example=93  <pre><a name=3De6d13>6.13)	xy. zei ka=
ntu
+	X ray
+</pre><ex "Persian rug">  XE "Persian rug: example=93  <ex "rug, Persian">=
  XE "rug: Persian, example=93  <pre><a name=3De6d14>6.14)	kulnr,farsi zei =
lolgai
+	Farsi floor-cover
+	Persian rug
+</pre><ex "hepatitis">  XE "hepatitis: example=93  <pre><a name=3De6d15>6.=
15)	na'e zei .a zei na'e zei by. livgyterbilma
+	non-A, non-B liver-disease
+	non-A, non-B hepatitis
+</pre><ex "Sherman tank">  XE "Sherman tank: example=93  <ex "tank, Sherma=
n">  XE "tank: Sherman, example=93  <pre><a name=3De6d16>6.16)	.cerman. zei=
 xarnykarce
+	Sherman war-car
+	Sherman tank
+</pre><a href=3D#e6d15>Example 6.15  is particularly noteworthy because th=
e phrase that would be produced by removing the =93zei=94s from it doesn't =
end with a brivla, and in fact is not even grammatical. As written, the exa=
mple is a tanru with two components, but by adding a =93zei=94 between =93b=
y.=94 and =93livgyterbilma=94 to produce
+<p>
+<pre><a name=3De6d17>6.17)	na'e zei .a zei na'e zei by. zei livgyterbilma
+	non-A-non-B-hepatitis
+</pre>the whole phrase would become a single lujvo. The longer lujvo of <a=
 href=3D#e6d17>Example 6.17 </a>may be preferable, because its place struct=
ure can be built from that of =93bilma=94, whereas the place structure of a=
 lujvo without a brivla must be constructed ad hoc.
+<p>
+<cx "rafsi, contrasted with words">  XE "rafsi: contrasted with words"   <=
cx "rafsi, contrasted with cmavo in usage">  XE "rafsi: contrasted with cma=
vo in usage"   <cx "cmavo, contrasted with rafsi in usage">  XE "cmavo: con=
trasted with rafsi in usage"   Note that rafsi may not be used in =93zei=94=
 phrases, because they are not words. CVV rafsi look like words (specifical=
ly cmavo) but there can be no confusion between the two uses of the same le=
tters, because cmavo appear only as separate words or in compound cmavo (wh=
ich are really just a notation for writing separate but closely related wor=
ds as if they were one); rafsi appear only as parts of lujvo.
+<p>
+<a name=3Ds7><h3>fu'ivla</h3>
+<p>
+<cx "fu'ivla, use of">  XE "fu'ivla: use of"   <cx "concrete terms, use of=
 fu'ivla for">  XE "concrete terms: use of fu'ivla for"   <cx "specific ter=
ms, use of fu'ivla for">  XE "specific terms: use of fu'ivla for"   <cx "ja=
rgon, use of fu'ivla for">  XE "jargon: use of fu'ivla for"   <cx "plants, =
use of fu'ivla for specific">  XE "plants: use of fu'ivla for specific"   <=
cx "animals, use of fu'ivla for specific">  XE "animals: use of fu'ivla for=
 specific"   <cx "food, use of fu'ivla for specific">  XE "food: use of fu'=
ivla for specific"   <cx "lujvo, unsuitability of for concrete/specific ter=
ms and jargon">  XE "lujvo: unsuitability of for concrete/specific terms an=
d jargon"   The use of tanru or lujvo is not always appropriate for very co=
ncrete or specific terms (e.g. =93brie=94 or =93cobra=94), or for jargon wo=
rds specialized to a narrow field (e.g. =93quark=94, =93integral=94, or =93=
iambic pentameter=94). These words are in effect names for concepts, and th=
e names were invented by speakers of another language. The vast majority of=
 words referring to plants, animals, foods, and scientific terminology cann=
ot be easily expressed as tanru. They thus must be borrowed (actually =93co=
pied=94) into Lojban from the original language.
+<p>
+<cx "borrowing, four stages of">  XE "borrowing: four stages of"   <lx "la=
'o">  XE "la'o"   <cx "borrowings, using foreign-language name">  XE "borro=
wings: using foreign-language name"   <cx "borrowings, Stage 1">  XE "borro=
wings: Stage 1"   There are four stages of borrowing in Lojban, as words be=
come more and more modified (but shorter and easier to use). Stage 1 is the=
 use of a foreign name quoted with the cmavo =93la'o=94 (explained in full =
in <a href=3Dchap19.html>Chapter 19</a>):
+<p>
+<cx "spaghetti">  XE "spaghetti"   <pre><a name=3De7d1>7.1)	me la'o ly. sp=
aghetti .ly.
+</pre>is a predicate with the place structure =93x1 is a quantity of spagh=
etti=94.
+<p>
+<cx "borrowings, using lojbanized name">  XE "borrowings: using lojbanized=
 name"   <cx "borrowings, Stage 2">  XE "borrowings: Stage 2"   Stage 2 inv=
olves changing the foreign name to a Lojbanized name, as explained in <a hr=
ef=3D#s8>Section 8</a>:
+<p>
+<pre><a name=3De7d2>7.2)	me la spagetis.
+</pre><p>One of these expedients is often quite sufficient when you need a=
 word quickly in conversation. (This can make it easier to get by when you =
do not yet have full command of the Lojban vocabulary, provided you are tal=
king to someone who will recognize the borrowing.)
+<p>
+<cx "borrowings, fu'ivla form with categorizing rafsi">  XE "borrowings: f=
u'ivla form with categorizing rafsi"   <cx "borrowings, Stage 3">  XE "borr=
owings: Stage 3"   <cx "fu'ivla, as Stage 3 borrowings">  XE "fu'ivla: as S=
tage 3 borrowings"   Where a little more universality is desired, the word =
to be borrowed must be Lojbanized into one of several permitted forms. A ra=
fsi is then usually attached to the beginning of the Lojbanized form, using=
 a hyphen to ensure that the resulting word doesn't fall apart.
+<p>
+<cx "fu'ivla categorizer">  XE "fu'ivla categorizer"   <cx "rafsi, as fu'i=
vla categorizer">  XE "rafsi: as fu'ivla categorizer"   <cx "fu'ivla, uniqu=
eness of meaning in">  XE "fu'ivla: uniqueness of meaning in"   <cx "borrow=
ings, most common form for">  XE "borrowings: most common form for"   The r=
afsi categorizes or limits the meaning of the fu'ivla; otherwise a word hav=
ing several different jargon meanings in other languages would require the =
word-inventor to choose which meaning should be assigned to the fu'ivla, si=
nce fu'ivla (like other brivla) are not permitted to have more than one def=
inition. Such a Stage 3 borrowing is the most common kind of fu'ivla.
+<p>
+<cx "borrowings, fu'ivla form without categorizing rafsi">  XE "borrowings=
: fu'ivla form without categorizing rafsi"   <cx "borrowings, Stage 4">  XE=
 "borrowings: Stage 4"   <cx "fu'ivla, as Stage 4 borrowings">  XE "fu'ivla=
: as Stage 4 borrowings"   Finally, Stage 4 fu'ivla do not have any rafsi c=
lassifier, and are used where a fu'ivla has become so common or so importan=
t that it must be made as short as possible. (See <a href=3D#s16>Section 16=
 </a>for a proposal concerning Stage 4 fu'ivla.)
+<p>
+<cx "fu'ivla, construction of">  XE "fu'ivla: construction of"   <cx "fu'i=
vla, form of">  XE "fu'ivla: form of"   The form of a fu'ivla reliably dist=
inguishes it from both the gismu and the cmavo. Like cultural gismu, fu'ivl=
a are generally based on a word from a single non-Lojban language. The word=
 is =93borrowed=94 (actually =93copied=94, hence the Lojban tanru =93fukpi =
valsi=94) from the other language and Lojbanized =97 the phonemes are conve=
rted to their closest Lojban equivalent and modifications are made as neces=
sary to make the word a legitimate Lojban fu'ivla-form word. All fu'ivla:
+<p>
+<cx "fu'ivla, rules for formation of">  XE "fu'ivla: rules for formation o=
f"   <cx "fu'ivla, initial consonant cluster in">  XE "fu'ivla: initial con=
sonant cluster in"   <dl compact><dt>1)	<dd>must contain a consonant cluste=
r in the first five letters of the word; if this consonant cluster is at th=
e beginning, it must either be a permissible initial consonant pair, or a l=
onger cluster such that each pair of adjacent consonants in the cluster is =
a permissible initial consonant pair: =93spraile=94 is acceptable, but not =
=93ktraile=94 or =93trkaile=94;
+<p><dt>2)	<dd>must end in one or more vowels;
+</dl><cx "slinku'i test, definition">  XE "slinku'i test: definition"   <d=
l compact><dt>3)	<dd>must not be gismu or lujvo, or any combination of cmav=
o, gismu, and lujvo; furthermore, a fu'ivla with a CV cmavo joined to the f=
ront of it must not have the form of a lujvo (the so-called =93slinku'i tes=
t=94, not discussed further in this book);
+</dl><ex "syllabic pronunciations of consonants, in fu'ivla">  XE "syllabi=
c pronunciations of consonants: in fu'ivla"   <cx "=93y=94, prohibition fro=
m fu'ivla">  XE "y: letter, prohibition from fu'ivla"   <dl compact><dt>4)	=
<dd>cannot contain =93y=94, although they may contain syllabic pronunciatio=
ns of Lojban consonants;
+</dl><cx "fu'ivla, stress in">  XE "fu'ivla: stress in"   <dl compact><dt>=
5)	<dd>like other brivla, are stressed on the penultimate syllable.
+</dl><cx "fu'ivla, consonant clusters in">  XE "fu'ivla: consonant cluster=
s in"   Note that consonant triples or larger clusters that are not at the =
beginning of a fu'ivla can be quite flexible, as long as all consonant pair=
s are permissible. There is no need to restrict fu'ivla clusters to permiss=
ible initial pairs except at the beginning.
+<p>
+<cx "fu'ivla, categorized contrasted with uncategorized in ease of constru=
ction">  XE "fu'ivla: categorized contrasted with uncategorized in ease of =
construction"   <cx "borrowings, Stage 3 contrasted with Stage 4 in ease of=
 construction">  XE "borrowings: Stage 3 contrasted with Stage 4 in ease of=
 construction"   This is a fairly liberal definition and allows quite a lot=
 of possibilities within =93fu'ivla space=94. Stage 3 fu'ivla can be made e=
asily on the fly, as lujvo can, because the procedure for forming them alwa=
ys guarantees a word that cannot violate any of the rules. Stage 4 fu'ivla =
require running tests that are not simple to characterize or perform, and s=
hould be made only after deliberation and by someone knowledgeable about al=
l the considerations that apply.
+<p>
+<cx "fu'ivla, algorithm for constructing">  XE "fu'ivla: algorithm for con=
structing"   Here is a simple and reliable procedure for making a non-Lojba=
n word into a valid Stage 3 fu'ivla:
+<p>
+<dl compact><dt>1)	<dd>Eliminate all double consonants and silent letters.
+<p><dt>2)	<dd>Convert all sounds to their closest Lojban equivalents. Lojb=
an =93y=94, however, may not be used in any fu'ivla.
+<p><dt>3)	<dd>If the last letter is not a vowel, modify the ending so that=
 the word ends in a vowel, either by removing a final consonant or by addin=
g a suggestively chosen final vowel.
+<p><dt>4)	<dd>If the first letter is not a consonant, modify the beginning=
 so that the word begins with a consonant, either by removing an initial vo=
wel or adding a suggestively chosen initial consonant.
+</dl><cx "=93l=94-hyphen, use of">  XE "l-hyphen: use of"   <cx "fu'ivla c=
ategorizer, selection consideration for">  XE "fu'ivla categorizer: selecti=
on consideration for"   <dl compact><dt>5)	<dd>Prefix the result of steps 1=
-5 with a 4-letter rafsi that categorizes the fu'ivla into a =93topic area=
=94. It is only safe to use a 4-letter rafsi; short rafsi sometimes produce=
 invalid fu'ivla. Hyphenate the rafsi to the rest of the fu'ivla with an =
=93r=94-hyphen; if that would produce a double =93r=94, use an =93n=94-hyph=
en instead; if the rafsi ends in =93r=94 and the rest of the fu'ivla begins=
 with =93n=94 (or vice versa) use an =93l=94-hyphen. (This is the only use =
of =93l=94-hyphen in Lojban.)
+		<p><dt><dd>Alternatively, if a CVC-form short rafsi is available it can =
be used instead of the long rafsi.=20
+<p><dt>6)	<dd>Remember that the stress necessarily appears on the penultim=
ate (next-to-the-last) syllable.
+</dl><ex "syllabic pronunciations of consonants, in fu'ivla category attac=
hment">  XE "syllabic pronunciations of consonants: in fu'ivla category att=
achment"   In this section, the hyphen is set off with commas in the exampl=
es, but these commas are not required in writing, and the hyphen need not b=
e pronounced as a separate syllable.
+<p>
+Here are a few examples:
+<p>
+<ex "spaghetti">  XE "spaghetti: example=93   <pre><a name=3De7d3>7.3)	spa=
ghetti (from English or Italian)
+	spageti (Lojbanize)
+	cidj,r,spageti (prefix long rafsi)
+	dja,r,spageti (prefix short rafsi)
+</pre>where =93cidj-=94 is the 4-letter rafsi for =93cidja=94, the Lojban =
gismu for =93food=94, thus categorizing =93cidjrspageti=94 as a kind of foo=
d. The form with the short rafsi happens to work, but such good fortune can=
not be relied on: in any event, it means the same thing.
+<p>
+<ex "maple trees">  XE "maple trees: example=93  <ex "Acer">  XE "Acer: ex=
ample=93  <pre><a name=3De7d4>7.4)	Acer (the scientific name of maple trees=
)
+	acer (Lojbanize)
+	xaceru (add initial consonant and final vowel)
+	tric,r,xaceru (prefix rafsi)
+	ric,r,xaceru (prefix short rafsi)
+</pre><ex "maple sugar">  XE "maple sugar: example=93  where =93tric-=94 a=
nd =93ric-=94 are rafsi for =93tricu=94, the gismu for =93tree=94. Note tha=
t by the same principles, =93maple sugar=94 could get the fu'ivla =93saktrx=
aceru=94, or could be represented by the tanru =93tricrxaceru sakta=94. Tec=
hnically, =93ricrxaceru=94 and =93tricrxaceru=94 are distinct fu'ivla, but =
they would surely be given the same meanings if both happened to be in use.
+<p>
+<ex "brie">  XE "brie: example=93  <pre><a name=3De7d5>7.5)	brie (from Fre=
nch)
+	bri (Lojbanize)
+	cirl,r,bri (prefix rafsi)
+</pre>where =93cirl-=94 represents =93cirla=94 (=93cheese=94).
+<p>
+<ex "cobra">  XE "cobra: example=93  <pre><a name=3De7d6>7.6)	cobra
+	kobra (Lojbanize)
+	sinc,r,kobra (prefix rafsi)
+</pre>where =93sinc-=94 represents =93since=94 (=93snake=94).
+<p>
+<ex "quark">  XE "quark: example=93  <pre><a name=3De7d7>7.7)	quark
+	kuark (Lojbanize)
+	kuarka (add final vowel)
+	sask,r,kuarka (prefix rafsi)
+</pre><cx "fu'ivla, diphthongs in">  XE "fu'ivla: diphthongs in"   <cx "di=
phthongs, in fu'ivla">  XE "diphthongs: in fu'ivla"   <cx "allowable diphth=
ongs, in fu'ivla contrasted with in gismu/lujvo">  XE "allowable diphthongs=
: in fu'ivla contrasted with in gismu/lujvo"   <cx "allowable diphthongs, i=
n gismu/lujvo contrasted with in fu'ivla">where =93sask-=94 represents =93s=
aske=94 (=93science=94). Note the extra vowel =93a=94 added to the end of t=
he word, and the diphthong =93ua=94, which never appears in gismu or lujvo,=
 but may appear in fu'ivla.
+<p>
+<cx "fu'ivla, disambiguation of">  XE "fu'ivla: disambiguation of"   <cx "=
fu'ivla categorizer, for distinguishing fu'ivla form">  XE "fu'ivla categor=
izer: for distinguishing fu'ivla form"   The use of the prefix helps distin=
guish among the many possible meanings of the borrowed word, depending on t=
he field. As it happens, =93spageti=94 and =93kuarka=94 are valid Stage 4 f=
u'ivla, but =93xaceru=94 looks like a compound cmavo, and =93kobra=94 like =
a gismu.
+<p>
+<cx "fu'ivla categorizer, for distinguishing specialized meanings">  XE "f=
u'ivla categorizer: for distinguishing specialized meanings"   For another =
example, =93integral=94 has a specific meaning to a mathematician. But the =
Lojban fu'ivla =93integrale=94, which is a valid Stage 4 fu'ivla, does not =
convey that mathematical sense to a non-mathematical listener, even one wit=
h an English-speaking background; its source =97 the English word =93integr=
al=94 =97 has various other specialized meanings in other fields.
+<p>
+Left uncontrolled, =93integrale=94 almost certainly would eventually come =
to mean the same collection of loosely related concepts that English associ=
ates with =93integral=94, with only the context to indicate (possibly) that=
 the mathematical term is meant.
+<p>
+<ex "integral, mathematical concept">  XE "integral: mathematical concept,=
 example=93  <ex "integral, architectural concept">  XE "integral: architec=
tural concept, example=93  The prefix method would render the mathematical =
concept as =93cmacrntegrale=94, if the =93i=94 of =93integrale=94 is remove=
d, or something like =93cmacrnintegrale=94, if a new consonant is added to =
the beginning; =93cmac-=94 is the rafsi for =93cmaci=94 (=93mathematics=94)=
. The architectural sense of =93integral=94 might be conveyed with =93djinr=
nintegrale=94 or =93tarmrnintegrale=94, where =93dinju=94 and =93tarmi=94 m=
ean =93building=94 and =93form=94 respectively.
+<p>
+Here are some fu'ivla representing cultures and related things, shown with=
 more than one rafsi prefix:
+<p>
+<ex "Bulgarian">  XE "Bulgarian: example=93  <pre><a name=3De7d8>7.8)	bang=
,r,blgaria
+	Bulgarian (in language)
+<a name=3De7d9>7.9)	kuln,r,blgaria
+	Bulgarian (in culture)
+<a name=3De7d10>7.10)	gugd,r,blgaria
+	Bulgaria (the country)
+</pre><ex "Korean">  XE "Korean: example=93  <pre><a name=3De7d11>7.11)	ba=
ng,r,kore,a
+	Korean (the language)
+<a name=3De7d12>7.12)	kuln,r,kore,a
+	Korean (the culture)
+</pre><cx "invalid diphthongs, in fu'ivla">  XE "invalid diphthongs: in fu=
'ivla"   <cx "fu'ivla, with invalid diphthongs">  XE "fu'ivla: with invalid=
 diphthongs"   <cx "fu'ivla, considerations for choosing basis word">  XE "=
fu'ivla: considerations for choosing basis word"   <ex "Navajo">  XE "Navaj=
o: example=93  Note the commas in <a href=3D#e7d{11}>Examples 7.11 </a>and =
<a href=3D#e7d12>7.12</a>, used because =93ea=94 is not a valid diphthong i=
n Lojban. Arguably, some form of the native name =93Chosen=94 should have b=
een used instead of the internationally known =93Korea=94; this is a recurr=
ing problem in all borrowings. In general, it is better to use the native n=
ame unless using it will severely impede understanding: =93Navajo=94 is far=
 more widely known than =93Dine'e=94.
+<p>
+<a name=3Ds8><h3>cmene</h3>
+<p>
+<cx "cmene, purpose of">  XE "cmene: purpose of"   <cx "names, purpose of"=
>  XE "names: purpose of"   <cx "cmene, definition">  XE "cmene: definition=
"   <cx "names in Lojban (see also =93cmene=94)">  XE "names in Lojban (see=
 also cmene)"   Lojbanized names, called =93cmene=94, are very much like th=
eir counterparts in other languages. They are labels applied to things (or =
people) to stand for them in descriptions or in direct address. They may co=
nvey meaning in themselves, but do not necessarily do so.
+<p>
+<cx "cmene, examples of">  XE "cmene: examples of"   <cx "names, examples =
of">  XE "names: examples of"   <cx "cmene, and analyzability of speech str=
eam">  XE "cmene: and analyzability of speech stream"   <cx "cmene, rationa=
le for lojbanizing">  XE "cmene: rationale for lojbanizing"   <cx "names, r=
ationale for lojbanizing">  XE "names: rationale for lojbanizing"   Because=
 names are often highly personal and individual, Lojban attempts to allow n=
ative language names to be used with a minimum of modification. The require=
ment that the Lojban speech stream be unambiguously analyzable, however, me=
ans that most names must be modified somewhat when they are Lojbanized. Her=
e are a few examples of English names and possible Lojban equivalents:
+<p>
+<ex "Jim">  XE "Jim: example=93  <pre><a name=3De8d1>8.1)	djim.
+	Jim
+</pre><ex "Jane">  XE "Jane: example=93  <pre><a name=3De8d2>8.2)	djein.
+	Jane
+</pre><ex "Arnold">  XE "Arnold: example=93  <pre><a name=3De8d3>8.3)	.arn=
old.
+	Arnold
+</pre><ex "Pete">  XE "Pete: example=93  <pre><a name=3De8d4>8.4)	pit.
+	Pete
+</pre><ex "Katrina">  XE "Katrina: example=93  <pre><a name=3De8d5>8.5)	ka=
trinas.
+	Katrina
+</pre><ex "Catherine">  XE "Catherine: example=93  <pre><a name=3De8d6>8.6=
)	kat,r,in.
+	Catherine
+</pre><cx "syllabic consonant, effect on stress determination">  XE "sylla=
bic consonant: effect on stress"   (Note that syllabic =93r=94 is skipped i=
n determining the stressed syllable, so <a href=3D#e8d6>Example 8.6 </a>is =
stressed on the =93ka=94.)
+<p>
+<ex "Cathy">  XE "Cathy: example=93  <pre><a name=3De8d7>8.7)	katis.
+	Cathy
+</pre>
+<ex "Kate">  XE "Kate: example=93  <pre><a name=3De8d8>8.8)	keit.
+	Kate
+</pre><cx "cmene, rules for formation">  XE "cmene: rules for formation"  =
 <cx "names, rules for formation">  XE "names: rules for formation"   <cx "=
cmene, stress in">  XE "cmene: stress in"   <cx "names, stress in">  XE "na=
mes: stress in"   <cx "cmene, unusual stress in">  XE "cmene: unusual stres=
s in"   <cx "names, unusual stress in">  XE "names: unusual stress in"   Na=
mes may have almost any form, but always end in a consonant, and are follow=
ed by a pause. They are penultimately stressed, unless unusual stress is ma=
rked with capitalization. A name may have multiple parts, each ending with =
a consonant and pause, or the parts may be combined into a single word with=
 no pause. For example,
+<p>
+<ex "John Jones">  XE "John Jones: example=93  <ex "Jones, John">  XE "Jon=
es: John, example=93  <pre><a name=3De8d9>8.9)	djan. djonz.
+</pre>and
+<p>
+<pre><a name=3De8d10>8.10)	djandjonz.
+</pre>are both valid Lojbanizations of =93John Jones=94.
+<p>
+<cx "cmene, authority for">  XE "cmene: authority for"   <cx "names, autho=
rity for">  XE "names: authority for"   The final arbiter of the correct fo=
rm of a name is the person doing the naming, although most cultures grant p=
eople the right to determine how they want their own name to be spelled and=
 pronounced. The English name =93Mary=94 can thus be Lojbanized as =93meris=
.=94, =93maris.=94, =93meiris.=94, =93merix.=94, or even =93marys.=94. The =
last alternative is not pronounced much like its English equivalent, but ma=
y be desirable to someone who values spelling over pronunciation. The final=
 consonant need not be an =93s=94; there must, however, be some Lojban cons=
onant at the end.
+<p>
+<cx "cmene, restrictions on form of">  XE "cmene: restrictions on form of"=
   <cx "names, restrictions on form of">  XE "names: restrictions on form o=
f"   Names are not permitted to have the sequences =93la=94, =93lai=94, or =
=93doi=94 embedded in them, unless the sequence is immediately preceded by =
a consonant. These minor restrictions are due to the fact that all Lojban c=
mene embedded in a speech stream will be preceded by one of these words or =
by a pause. With one of these words embedded, the cmene might break up into=
 valid Lojban words followed by a shorter cmene. However, break-up cannot h=
appen after a consonant, because that would imply that the word before the =
=93la=94, or whatever, ended in a consonant without pause, which is impossi=
ble.
+<p>
+<ex "Laplace">  XE "Laplace: example=93  <ex "Nederlands">  XE "Nederlands=
: example=93  For example, the invalid name =93laplas.=94 would look like t=
he Lojban words =93la plas.=94, and =93ilanas.=94 would be misunderstood as=
 =93.i la nas.=94. However, =93NEderlants.=94 cannot be misheard as =93NEde=
r lants.=94, because =93NEder=94 with no following pause is not a possible =
Lojban word.
+<p>
+<cx "cmene, alternatives for restricted sequences in">  XE "cmene: alterna=
tives for restricted sequences in"   <cx "names, alternatives for restricte=
d sequences in">  XE "names: alternatives for restricted sequences in"   Th=
ere are close alternatives to these forbidden sequences that can be used in=
 Lojbanizing names, such as =93ly=94, =93lei=94, and =93dai=94 or =93do'i=
=94, that do not cause these problems.
+<p>
+<cx "names, rules for">  XE "names: rules for"   <cx "cmene, rules for">  =
XE "cmene: rules for"   Lojban cmene are identifiable as word forms by the =
following characteristics:
+<p>
+<cx "cmene, final letter in">  XE "cmene: final letter in"   <cx "cmene, c=
onsonant clusters permitted in">  XE "cmene: consonant clusters permitted i=
n"   <dl compact><dt>1)	<dd>They must end in one or more consonants. There =
are no rules about how many consonants may appear in a cluster in cmene, pr=
ovided that each consonant pair (whether standing by itself, or as part of =
a larger cluster) is a permissible pair.
+</dl><cx "=93iy=94, in cmene">  XE "iy diphthong: in cmene"   <cx "=93uy=
=94, in cmene">  XE "uy diphthong: in cmene"   <cx "diphthongs, specific to=
 names">  XE "diphthongs: specific to names"   <cx "diphthongs, specific to=
 cmene">  XE "diphthongs: specific to cmene"   <dl compact><dt>2)	<dd>They =
may contain the letter y as a normal, non-hyphenating vowel. They are the o=
nly kind of Lojban word that may contain the two diphthongs =93iy=94 and =
=93uy=94.
+</dl><cx "cmene, requirement for pause after">  XE "cmene: requirement for=
 pause after"   <cx "names, requirement for pause after">  XE "names: requi=
rement for pause after"   <dl compact><dt>3)	<dd>They are always followed i=
n speech by a pause after the final consonant, written as =93.=94.
+</dl><cx "cmene, stress in">  XE "cmene: stress in"   <cx "names, stress i=
n">  XE "names: stress in"   <cx "capitalization, use in names">  XE "capit=
alization: use in names"   <cx "capitalization, for unusual stress in names=
">  XE "capitalization: for unusual stress in names"   <cx "capitalization,=
 use of">  XE "capitalization: use of"   <dl compact><dt>4)	<dd>They may be=
 stressed on any syllable; if this syllable is not the penultimate one, it =
must be capitalized when writing. Neither names nor words that begin senten=
ces are capitalized in Lojban, so this is the only use of capital letters.
+</dl><cx "cmene, from Lojban words">  XE "cmene: from Lojban words"   <cx =
"names, from Lojban words">  XE "names: from Lojban words"   Names meeting =
these criteria may be invented, Lojbanized from names in other languages, o=
r formed by appending a consonant onto a cmavo, a gismu, a fu'ivla or a luj=
vo. Some cmene built from Lojban words are:
+<p>
+<ex "One, the">  XE "One: the, example=93  <pre><a name=3De8d11>8.11)	pav.
+	the One
+	from the cmavo =93pa=94, with rafsi =93pav=94, meaning =93one=94
+</pre><ex "Sun, the">  XE "Sun: the, example=93  <pre><a name=3De8d12>8.12=
)	sol.
+	the Sun
+	from the gismu =93solri=94, meaning =93solar=94, or actually =93pertainin=
g to the Sun=94
+</pre><ex "Chief">  XE "Chief: example=93  <pre><a name=3De8d13>8.13)	ralj=
.
+	Chief (as a title)
+	from the gismu =93ralju=94, meaning =93principal=94.
+</pre><ex "Lord">  XE "Lord: example=93  <ex "Lady">  XE "Lady: example=93=
  <pre><a name=3De8d14>8.14)	nol.
+	Lord/Lady
+	from the gismu =93nobli=94, with rafsi =93nol=94, meaning =93noble=94.
+</pre><cx "names, algorithm for">  XE "names: algorithm for"   <cx "cmene,=
 algorithm for">  XE "cmene: algorithm for"   To Lojbanize a name from the =
various natural languages, apply the following rules:
+<p>
+<dl compact><dt>1)	<dd>Eliminate double consonants and silent letters.
+<p><dt>2)	<dd>Add a final =93s=94 or =93n=94 (or some other consonant that=
 sounds good) if the name ends in a vowel.
+<p><dt>3)	<dd>Convert all sounds to their closest Lojban equivalents.
+<p><dt>4)	<dd>If possible and acceptable, shift the stress to the penultim=
ate (next-to-the-last) syllable. Use commas and capitalization in written L=
ojban when it is necessary to preserve non-standard syllabication or stress=
. Do not capitalize names otherwise.
+</dl><cx "cmene, avoiding impermissible consonant clusters in">  XE "cmene=
: avoiding impermissible consonant clusters in"   <dl compact><dt>5)	<dd>If=
 the name contains an impermissible consonant pair, insert a vowel between =
the consonants: =93y=94 is recommended.
+</dl><cx "cmene, proscribed syllables in">  XE "cmene: proscribed syllable=
s in"   <dl compact><dt>6)	<dd>No cmene may have the syllables =93la=94, =
=93lai=94, or =93doi=94 in them, unless immediately preceded by a consonant=
. If these combinations are present, they must be converted to something el=
se. Possible substitutions include =93ly=94, =93ly'i=94, and =93dai=94 or =
=93do'i=94, respectively.
+</dl><cx "Linnaean names, rules for">  XE "Linnaean names: rules for"   <c=
x "scientific names, rules for">  XE "scientific names: rules for"   There =
are some additional rules for Lojbanizing the scientific names (technically=
 known as =93Linnaean binomials=94 after their inventor) which are internat=
ionally applied to each species of animal or plant. Where precision is esse=
ntial, these names need not be Lojbanized, but can be directly inserted int=
o Lojban text using the cmavo =93la'o=94, explained in <a href=3Dchap19.htm=
l>Chapter 19</a>. Using this cmavo makes the already lengthy Latinized name=
s at least four syllables longer, however, and leaves the pronunciation in =
doubt. The following suggestions, though incomplete, will assist in convert=
ing Linnaean binomals to valid Lojban names. They can also help to create f=
u'ivla based on Linnaean binomials or other words of the international scie=
ntific vocabulary. The term =93back vowel=94 in the following list refers t=
o any of the letters =93a=94, =93o=94, or =93u=94; the term =93front vowel=
=94 correspondingly refers to any of the letters =93e=94, =93i=94, or =93y=
=94.
+<p>
+<dl compact><dt>1)	<dd>Change double consonants other than =93cc=94 to sin=
gle consonants.
+<p><dt>2)	<dd>Change =93cc=94 before a front vowel to =93kc=94, but otherw=
ise to =93k=94.
+<p><dt>3)	<dd>Change =93c=94 before a back vowel and final =93c=94 to =93k=
=94.
+<p><dt>4)	<dd>Change =93ng=94 before a consonant (other than =93h=94) and =
final =93ng=94 to =93n=94.
+<p><dt>5)	<dd>Change =93x=94 to =93z=94 initially, but otherwise to =93ks=
=94.
+<p><dt>6)	<dd>Change =93pn=94 to =93n=94 initially.
+<p><dt>7)	<dd>Change final =93ie=94 and =93ii=94 to =93i=94.
+<p><dt>8)	<dd>Make the following idiosyncratic substitutions:
+<p><dt><dd>		aa	a
+		ae	e
+		ch	k
+		ee	i
+		eigh	ei
+		ew	u
+		igh	ai
+		oo	u
+		ou	u
+		ow	au
+		ph	f
+		q	k
+		sc	sk
+		w	u
+		y	i
+		<p><dt><dd>However, the diphthong substitutions should not be done if th=
e two vowels are in two different syllables.
+<p><dt>9)	<dd>Change =93h=94 between two vowels to =93 ' =94, but otherwis=
e remove it completely. If preservation of the =93h=94 seems essential, cha=
nge it to =93x=94 instead.
+<p><dt>10)	<dd>Place =93 ' =94 between any remaining vowel pairs that do n=
ot form Lojban diphthongs.
+<p><dt><dd>Some further examples of Lojbanized names are:
+</dl><ex "William">  XE "William: example=93  <ex "Svetlana">  XE "Svetlan=
a: example=93  <ex "Mao Zedong">  XE "Mao Zedong: example=93  <ex "Lech Wal=
esa">  XE "Lech Walesa: example=93  <ex "Krishna">  XE "Krishna: example=93=
  <ex "Khrushchev">  XE "Khrushchev: example=93  <ex "Johnson">  XE "Johnso=
n: example=93  <ex "Fujiko">  XE "Fujiko: example=93  <ex "Don Quixote">  X=
E "Don Quixote: example=93  <ex "De Gaulle">  XE "De Gaulle: example=93  <p=
re> 	English =93Mary=94		meris.
+			or	meiris.
+	English =93Smith=94		smit.
+	English =93Jones=94		djonz.
+	English =93John=94		djan. or jan. (American)
+			or	djon. or jon. (British)
+	English =93Alice=94		.alis.
+	English =93Elise=94		.eLIS.
+	English =93Johnson=94	djansn.
+	English =93William=94	.uiliam.
+			or	.uil,iam.
+	English =93Brown=94	braun.
+	English =93Charles=94	tcarlz.
+	French =93Charles=94	carl.
+	French =93De Gaulle=94	dyGOL.
+	German =93Heinrich=94	xainrix.
+	Spanish =93Joaquin=94	xuaKIN.
+	Russian =93Svetlana=94	sfietlanys.
+	Russian =93Khrushchev=94	xrucTCOF.
+	Hindi =93Krishna=94		kricnas.
+	Polish =93Lech Walesa=94	lex. va,uensas.
+	Spanish =93Don Quixote=94	don. kicotes.
+	     or modern Spanish:	don. kixotes.
+	     or Mexican dialect:	don. ki'otes.
+	Chinese =93Mao Zedong=94	maudzydyn.
+	Japanese =93Fujiko=94	fudjikos.
+			or	fujikos.
+</pre><a name=3Ds9><h3>Rules for inserting pauses</h3>
+<p>
+<cx "pauses, rules for">  XE "pauses: rules for"   Summarized in one place=
, here are the rules for inserting pauses between Lojban words:
+<p>
+<cx "pause, proscribed within words">  XE "pause: proscribed within words"=
   <cx "pause, between words">  XE "pause: between words"   <dl compact><dt=
>1)	<dd>Any two words may have a pause between them; it is always illegal t=
o pause in the middle of a word, because that breaks up the word into two w=
ords.
+</dl><cx "consonant-final words, necessity for pause after">  XE "consonan=
t-final words: necessity for pause after"   <cx "pause, and consonant-final=
 words">  XE "pause: and consonant-final words"   <dl compact><dt>2)	<dd>Ev=
ery word ending in a consonant must be followed by a pause. Necessarily, al=
l such words are cmene.
+</dl><cx "vowel-initial words, necessity for pause before">  XE "vowel-ini=
tial words: necessity for pause before"   <cx "pause, and vowel-initial wor=
ds">  XE "pause: and vowel-initial words"   <dl compact><dt>3)	<dd>Every wo=
rd beginning with a vowel must be preceded by a pause. Such words are eithe=
r cmavo, fu'ivla, or cmene; all gismu and lujvo begin with consonants.
+</dl><cx "cmene, rules for pause before">  XE "cmene: rules for pause befo=
re"   <cx "pause, and cmene">  XE "pause: and cmene"   <dl compact><dt>4)	<=
dd>Every cmene must be preceded by a pause, unless the immediately precedin=
g word is one of the cmavo =93la=94, =93lai=94, =93la'i=94, or =93doi=94 (w=
hich is why those strings are forbidden in cmene). However, the situation t=
riggering this rule rarely occurs.
+</dl><cx "stress, final syllable, rules for pause after">  XE "stress: fin=
al syllable, rules for pause after"   <cx "final syllable stress, rules for=
 pause after">  XE "final syllable stress: rules for pause after"   <cx "pa=
use, and final-syllable stress">  XE "pause: and final-syllable stress"   <=
dl compact><dt>5)	<dd>If the last syllable of a word bears the stress, and =
a brivla follows, the two must be separated by a pause, to prevent confusio=
n with the primary stress of the brivla. In this case, the first word must =
be either a cmavo or a cmene with unusual stress (which already ends with a=
 pause, of course).
+</dl><cx "=93Cy=94-form cmavo, rules for pause after">  XE "Cy-form cmavo:=
 rules for pause after"   <cx "cmavo, rules for pause after =93Cy=94-form">=
  XE "cmavo: rules for pause after Cy-form"   <cx "pause, and =93Cy=94-form=
 cmavo">  XE "pause: and Cy-form cmavo"   <dl compact><dt>6)	<dd>A cmavo of=
 the form =93Cy=94 must be followed by a pause unless another =93Cy=94-form=
 cmavo follows.
+</dl><cx "pause, and non-Lojban text">  XE "pause: and non-Lojban text"   =
<cx "non-Lojban text, rules for pause with">  XE "non-Lojban text: rules fo=
r pause with"   <dl compact><dt>7)	<dd>When non-Lojban text is embedded in =
Lojban, it must be preceded and followed by pauses. (How to embed non-Lojba=
n text is explained in <a href=3Dchap19.html>Chapter 19</a>.)
+</dl><a name=3Ds10><h3>Considerations for making lujvo</h3>
+<p>
+Given a tanru which expresses an idea to be used frequently, it can be tur=
ned into a lujvo by following the lujvo-making algorithm which is given in =
<a href=3D#s11>Section 11</a>.
+<p>
+In building a lujvo, the first step is to replace each gismu with a rafsi =
that uniquely represents that gismu. These rafsi are then attached together=
 by fixed rules that allow the resulting compound to be recognized as a sin=
gle word and to be analyzed in only one way.
+<p>
+There are three other complications; only one is serious.
+<p>
+<cx "rafsi, multiple for each gismu">  XE "rafsi: multiple for each gismu"=
   The first is that there is usually more than one rafsi that can be used =
for each gismu. The one to be used is simply whichever one sounds or looks =
best to the speaker or writer. There are usually many valid combinations of=
 possible rafsi. They all are equally valid, and all of them mean exactly t=
he same thing. (The scoring algorithm given in <a href=3D#s12>Section 12 </=
a>is used to choose the standard form of the lujvo =97 the version which wo=
uld be entered into a dictionary.)
+<p>
+<cx "lujvo, unambiguity of">  XE "lujvo: unambiguity of"   <cx "lujvo, con=
sideration in choosing meaning for">  XE "lujvo: consideration in choosing =
meaning for"   <cx "linguistic drift in Lojban, possible source of">  XE "l=
inguistic drift in Lojban: possible source of"   The second complication is=
 the serious one. Remember that a tanru is ambiguous =97 it has several pos=
sible meanings. A lujvo, or at least one that would be put into the diction=
ary, has just a single meaning. Like a gismu, a lujvo is a predicate which =
encompasses one area of the semantic universe, with one set of places. Hope=
fully the meaning chosen is the most useful of the possible semantic spaces=
. A possible source of linguistic drift in Lojban is that as Lojbanic socie=
ty evolves, the concept that seems the most useful one may change.
+<p>
+<cx "lujvo, meaning drift of">  XE "lujvo: meaning drift of"   <lx "za'e">=
  XE "za'e"   <cx "za'e, use to avoid lujvo misunderstandings">  XE "za'e: =
use to avoid lujvo misunderstandings"   You must also be aware of the possi=
bility of some prior meaning of a new lujvo, especially if you are writing =
for posterity. If a lujvo is invented which involves the same tanru as one =
that is in the dictionary, and is assigned a different meaning (or even jus=
t a different place structure), linguistic drift results. This isn't necess=
arily bad. Every natural language does it. But in communication, when you u=
se a meaning different from the dictionary definition, someone else may use=
 the dictionary and therefore misunderstand you. You can use the cmavo =93z=
a'e=94 (explained in <a href=3Dchap19.html>Chapter 19</a>) before a newly c=
oined lujvo to indicate that it may have a non-dictionary meaning.
+<p>
+<cx "lujvo, ultimate guideline for choice of meaning/place-structure">  XE=
 "lujvo: ultimate guideline for choice of meaning/place-structure"   The es=
sential nature of human communication is that if the listener understands, =
then all is well. Let this be the ultimate guideline for choosing meanings =
and place structures for invented lujvo.
+<p>
+<cx "Zipf's Law">  XE "Zipf's Law"   <cx "lujvo, dropping elements of">  X=
E "lujvo: dropping elements of"   The third complication is also simple, bu=
t tends to scare new Lojbanists with its implications. It is based on Zipf'=
s Law, which says that the length of words is inversely proportional to the=
ir usage. The shortest words are those which are used more; the longest one=
s are used less. Conversely, commonly used concepts will be tend to be abbr=
eviated. In English, we have abbreviations and acronyms and jargon, all of =
which represent complex ideas that are used often by small groups of people=
, so they shortened them to convey more information more rapidly.
+<p>
+Therefore, given a complicated tanru with grouping markers, abstraction ma=
rkers, and other cmavo in it to make it syntactically unambiguous, the psyc=
hological basis of Zipf's Law may compel the lujvo-maker to drop some of th=
e cmavo to make a shorter (technically incorrect) tanru, and then use that =
tanru to make the lujvo.
+<p>
+<cx "lujvo, based on multiple tanru">  XE "lujvo: based on multiple tanru"=
   This doesn't lead to ambiguity, as it might seem to. A given lujvo still=
 has exactly one meaning and place structure. It is just that more than one=
 tanru is competing for the same lujvo. But more than one meaning for the t=
anru was already competing for the =93right=94 to define the meaning of the=
 lujvo. Someone has to use judgment in deciding which one meaning is to be =
chosen over the others.
+<p>
+<cx "lujvo, considerations for retaining elements of">  XE "lujvo: conside=
rations for retaining elements of"   <cx "lujvo, shorter for more general c=
oncepts">  XE "lujvo: shorter for more general concepts"   If the lujvo mad=
e by a shorter form of tanru is in use, or is likely to be useful for anoth=
er meaning, the decider then retains one or more of the cmavo, preferably o=
nes that set this meaning apart from the shorter form meaning that is used =
or anticipated. As a rule, therefore, the shorter lujvo will be used for a =
more general concept, possibly even instead of a more frequent word. If bot=
h words are needed, the simpler one should be shorter. It is easier to add =
a cmavo to clarify the meaning of the more complex term than it is to find =
a good alternate tanru for the simpler term.
+<p>
+<cx "lujvo, and the listener">  XE "lujvo: and the listener"   <cx "lujvo,=
 and plausibility">  XE "lujvo: and plausibility"   And of course, we have =
to consider the listener. On hearing an unknown word, the listener will dec=
ompose it and get a tanru that makes no sense or the wrong sense for the co=
ntext. If the listener realizes that the grouping operators may have been d=
ropped out, he or she may try alternate groupings, or try inserting an abst=
raction operator if that seems plausible. (The grouping of tanru is explain=
ed in <a href=3Dchap5.html>Chapter 5</a>; abstraction is explained in <a hr=
ef=3Dchap11.html>Chapter 11</a>.)  Plausibility is the key to learning new =
ideas and to evaluating unfamiliar lujvo.
+<p>
+<a name=3Ds11><h3>The lujvo-making algorithm</h3>
+<p>
+<cx "lujvo, algorithm for">  XE "lujvo: algorithm for"   The following is =
the current algorithm for generating Lojban lujvo given a known tanru and a=
 complete list of gismu and their assigned rafsi. The algorithm was designe=
d by Bob LeChevalier and Dr. James Cooke Brown for computer program impleme=
ntation. It was modified in 1989 with the assistance of Nora LeChevalier, w=
ho detected a flaw in the original =93tosmabru test=94.
+<p>
+Given a tanru that is to be made into a lujvo:
+<p>
+<dl compact><dt>1)	<dd>Choose a 3-letter or 4-letter rafsi for each of the=
 gismu and cmavo in the tanru except the last.
+<p><dt>2)	<dd>Choose a 3-letter (CVV-form or CCV-form) or 5-letter rafsi f=
or the final gismu in the tanru.
+<p><dt>3)	<dd>Join the resulting string of rafsi, initially without hyphen=
s.
+</dl><cx "hyphens in lujvo, proscribed where not required">  XE "hyphens i=
n lujvo: proscribed where not required"   <dl compact><dt>4)	<dd>Add hyphen=
 letters where necessary. It is illegal to add a hyphen at a place that is =
not required by this algorithm. Right-to-left tests are recommended, for re=
asons discussed below.
+
+<p><dt>4a)	<dd>If there are more than two words in the tanru, put an =93r=
=94-hyphen (or an =93n=94-hyphen) after the first rafsi if it is CVV-form. =
If there are exactly two words, then put an =93r=94-hyphen (or an =93n=94-h=
yphen) between the two rafsi if the first rafsi is CVV-form, unless the sec=
ond rafsi is CCV-form (for example, =93saicli=94 requires no hyphen). Use a=
n =93r=94-hyphen unless the letter after the hyphen is =93r=94, in which ca=
se use an =93n=94-hyphen. Never use an =93n=94-hyphen unless it is required=
.
+<p><dt>4b)	<dd>Put a =93y=94-hyphen between the consonants of any impermis=
sible consonant pair. This will always appear between rafsi.
+<p><dt>4c)	<dd>Put a =93y=94-hyphen after any 4-letter rafsi form.
+</dl><cx "tosmabru test">  XE "tosmabru test"   <dl compact><dt>5)	<dd>Tes=
t all forms with one or more initial CVC-form rafsi =97 with the pattern =
=93CVC=A0<dots>=85</dots>CVC + X=94 =97 for =93tosmabru failure=94. X must =
either be a CVCCV long rafsi that happens to have a permissible initial pai=
r as the consonant cluster, or is something which has caused a =93y=94-hyph=
en to be installed between the previous CVC and itself by one of the above =
rules.
+		The test is as follows:
+<p><dt>5a)	<dd>Examine all the C/C consonant pairs that join the CVC rafsi=
, and also the pair between the last CVC and the X portion, ignoring any =
=93y=94-hyphen before the X.
+	These consonant pairs are called =93joints=94.
+<p><dt>5b)	<dd>If all of those joints are permissible initials, then the t=
rial word will break up into a cmavo and a shorter brivla. If not, the word=
 will not break up, and no further hyphens are needed.
+<p><dt>5c)	<dd>Install a =93y=94-hyphen at the first such joint.
+</dl><p>Note that the =93tosmabru test=94 implies that the algorithm will =
be more efficient if rafsi junctures are tested for required hyphens from r=
ight to left, instead of from left to right; when the test is required, it =
cannot be completed until hyphenation to the right has been determined.
+<p>
+<a name=3Ds12><h3>The lujvo scoring algorithm</h3>
+<p>
+<cx "lujvo, scoring of">  XE "lujvo: scoring of"   <cx "lujvo, selection o=
f best form of">  XE "lujvo: selection of best form of"   This algorithm wa=
s devised by Bob and Nora LeChevalier in 1989. It is not the only possible =
algorithm, but it usually gives a choice that people find preferable. The a=
lgorithm may be changed in the future. The lowest-scoring variant will usua=
lly be the dictionary form of the lujvo. (In previous versions, it was the =
highest-scoring variant.)
+<p>
+<dl compact><dt>1)	<dd>Count the total number of letters, including hyphen=
s and apostrophes; call it =93L=94.
+<p><dt>2)	<dd>Count the number of apostrophes; call it =93A=94.
+<p><dt>3)	<dd>Count the number of =93y=94-, =93r=94-, and =93n=94-hyphens;=
 call it =93H=94.
+
+<p><dt>4)	<dd>For each rafsi, find the value in the following table. Sum t=
his value over all rafsi; call it =93R=94:
+<p><dt>	<dd>CVC/CV (final)	(-sarji)	1
+	CVC/C		(-sarj-)	2
+	CCVCV (final)	(-zbasu)	3
+	CCVC		(-zbas-)	4
+	CVC		(-nun-)	5
+	CVV with an apostrophe	(-ta'u-)	6
+	CCV		(-zba-)	7
+	CVV with no apostrophe	(-sai-)	8
+<p><dt>5)	<dd>Count the number of vowels, not including =93y=94; call it =
=93V=94.
+		</dl><p>The score is then:
+<p>
+<dl compact><dt>		<dd>(1000 * L) - (500 * A) + (100 * H) - (10 * R) - V
+</dl><cx "hierarchy of priorities for selecting lujvo form">  XE "hierarch=
y of priorities for selecting lujvo form"   <cx "lujvo form, hierarchy of p=
riorities for selection of">  XE "lujvo form: hierarchy of priorities for s=
election of"   In case of ties, there is no preference. This should be rare=
. Note that the algorithm essentially encodes a hierarchy of priorities: sh=
ort words are preferred (counting apostrophes as half a letter), then words=
 with fewer hyphens, words with more pleasing rafsi (this judgment is subje=
ctive), and finally words with more vowels are chosen. Each decision princi=
ple is applied in turn if the ones before it have failed to choose; it is p=
ossible that a lower-ranked principle might dominate a higher-ranked one if=
 it is ten times better than the alternative.
+<p>
+<cx "lujvo, scored examples of">  XE "lujvo: scored examples of"   Here ar=
e some lujvo with their scores (not necessarily the lowest scoring forms fo=
r these lujvo, nor even necessarily sensible lujvo):
+<p>
+<pre><a name=3De12d1>12.1)	zbasai
+	zba + sai
+	(1000 * 6) - (500 * 0) + (100 * 0) - (10 * 15) - 3=20
+=3D 5847
+<a name=3De12d2>12.2)	nunynau
+	nun + y + nau
+	32500 - (1000 * 7) + (500 * 0) - (100 * 1) + (10 * 13) + 3=20
+=3D 6967
+<a name=3De12d3>12.3)	sairzbata'u
+	sai + r + zba + ta'u
+	32500 - (1000 * 11) + (500 * 1) - (100 * 1) + (10 * 21) + 5
+=3D 10385
+<a name=3De12d4>12.4)	zbazbasysarji
+	zba + zbas + y + sarji
+	32500 - (1000 * 13) + (500 * 0) - (100 * 1) + (10 * 12) + 4
+=3D 12976
+</pre><a name=3Ds13><h3>lujvo-making examples</h3>
+<p>
+<cx "lujvo, examples of making">  XE "lujvo: examples of making"   <ex "do=
ghouse">  XE "doghouse: example=93  This section contains examples of makin=
g and scoring lujvo. First, we will start with the tanru =93gerku zdani=94 =
(=93dog house=94) and construct a lujvo meaning =93doghouse=94, that is, a =
house where a dog lives. We will use a brute-force application of the algor=
ithm in <a href=3D#s12>Section 12</a>, using every possible rafsi.
+<p>
+The rafsi for =93gerku=94 are:
+<p>
+<dl compact><dt>		<dd>-ger-, -ge'u-, -gerk-, -gerku
+</dl><p>The rafsi for =93zdani=94 are:
+<p>
+<dl compact><dt>		<dd>-zda-, -zdan-, -zdani.
+</dl><p>Step 1 of the algorithm directs us to use =93-ger-=94, =93-ge'u-=
=94 and =93-gerk-=94 as possible rafsi for =93gerku=94; Step 2 directs us t=
o use =93-zda-=94 and =93-zdani=94 as possible rafsi for =93zdani=94. The s=
ix possible forms of the lujvo are then:
+<p>
+<dl compact><dt>	<dd>ger-zda
+	ger-zdani
+	ge'u-zda
+	ge'u-zdani
+	gerk-zda
+	gerk-zdani
+</dl><p>We must then insert appropriate hyphens in each case. The first tw=
o forms need no hyphenation: =93ge=94 cannot fall off the front, because th=
e following word would begin with =93rz=94, which is not a permissible init=
ial consonant pair. So the lujvo forms are =93gerzda=94 and =93gerzdani=94.
+<p>
+The third form, =93ge'u-zda=94, needs no hyphen, because even though the f=
irst rafsi is CVV, the second one is CCV, so there is a consonant cluster i=
n the first five letters. So =93ge'uzda=94 is this form of the lujvo.
+<p>
+The fourth form, =93ge'u-zdani=94, however, requires an =93r=94-hyphen; ot=
herwise, the =93ge'u-=94 part would fall off as a cmavo. So this form of th=
e lujvo is =93ge'urzdani=94.
+<p>
+The last two forms require =93y=94-hyphens, as all 4-letter rafsi do, and =
so are =93gerkyzda=94 and =93gerkyzdani=94 respectively.
+<p>
+The scoring algorithm is heavily weighted in favor of short lujvo, so we m=
ight expect that =93gerzda=94 would win. Its L score is 6, its A score is 0=
, its H score is 0, its R score is 12, and its V score is 3, for a final sc=
ore of 5878. The other forms have scores of 7917, 6367, 9506, 8008, and 100=
47 respectively. Consequently, this lujvo would probably appear in the dict=
ionary in the form =93gerzda=94.
+<p>
+<ex "boat class">  XE "boat class: example=93  For the next example, we wi=
ll use the tanru =93bloti klesi=94 (=93boat class=94) presumably referring =
to the category (rowboat, motorboat, cruise liner) into which a boat falls.=
 We will omit the long rafsi from the process, since lujvo containing long =
rafsi are almost never preferred by the scoring algorithm when there are sh=
ort rafsi available.
+<p>
+The rafsi for =93bloti=94 are =93-lot-=94, =93-blo-=94, and =93-lo'i-=94; =
for =93klesi=94 they are =93-kle-=94 and =93-lei-=94. Both these gismu are =
among the handful which have both CVV-form and CCV-form rafsi, so there is =
an unusual number of possibilities available for a two-part tanru:
+<p>
+<pre> 	lotkle	blokle	lo'ikle
+	lotlei	blolei	lo'irlei
+
+</pre><p>Only =93lo'irlei=94 requires hyphenation (to avoid confusion with=
 the cmavo sequence =93lo'i lei=94). All six forms are valid versions of th=
e lujvo, as are the six further forms using long rafsi; however, the scorin=
g algorithm produces the following results:
+<p>
+<pre> 	lotkle	5878	blokle	5858	lo'ikle	6367
+	lotlei	5867	blolei	5847	lo'irlei	7456
+</pre><p>So the form =93blolei=94 is preferred, but only by a tiny margin =
over =93blokle=94;  =93lotlei=94 and =93lotkle=94 are only slightly worse; =
=93lo'ikle=94 suffers because of its apostrophe, and =93lo'irlei=94 because=
 of having both apostrophe and hyphen.
+<p>
+<ex "Logical Language Group">  XE "Logical Language Group: example=93  Our=
 third example will result in forming both a lujvo and a name from the tanr=
u =93logji bangu girzu=94, or =93logical-language group=94 in English. (=93=
The Logical Language Group=94 is the name of the publisher of this book and=
 the organization for the promotion of Lojban.)
+The available rafsi are =93-loj-=94 and =93-logj-=94; =93-ban-=94, =93-bau=
-=94, and =93-bang-=94; and =93-gri-=94 and =93-girzu=94, and (for name pur=
poses only) =93-gir-=94 and =93-girz-=94. The resulting 12 lujvo possibilit=
ies are:
+<p>
+<pre> 	loj-ban-gri	loj-bau-gri	loj-bang-gri
+	logj-ban-gri	logj-bau-gri	logj-bang-gri
+	loj-ban-girzu	loj-bau-girzu	loj-bang-girzu
+	logj-ban-girzu	logj-bau-girzu	logj-bang-girzu
+</pre>and the 12 name possibilities are:
+<p>
+<pre> 	loj-ban-gir.	loj-bau-gir.	loj-bang-gir.
+	logj-ban-gir.	logj-bau-gir.	logj-bang-gir.
+	loj-ban-girz.	loj-bau-girz.	loj-bang-girz.
+	logj-ban-girz.	logj-bau-girz.	logj-bang-girz.
+</pre><p>After hyphenation, we have:
+<p>
+<pre> 	lojbangri	lojbaugri	lojbangygri
+	logjybangri	logjybaugri	logjybangygri
+	lojbangirzu	lojbaugirzu	lojbangygirzu
+	logjybangirzu	logjybaugirzu	logjybangygirzu
+	lojbangir.	lojbaugir.	lojbangygir.
+	logjybangir.	logjybaugir.	logjybangygir.
+	lojbangirz.	lojbaugirz.	lojbangygirz.
+	logjybangirz.	logjybaugirz.	logjybangygirz.
+</pre><p>The only fully reduced lujvo forms are =93lojbangri=94 and =93loj=
baugri=94, of which the latter has a slightly lower score: 8827 versus 8796=
, respectively. However, for the name of the organization, we chose to make=
 sure the name of the language was embedded in it, and to use the clearer l=
ong-form rafsi for =93girzu=94, producing =93lojbangirz.=94
+<p>
+<ex "male sexual teacher">  XE "male sexual teacher: example=93  <ex "sexu=
al teacher, male">  XE "sexual teacher: male, example=93  Finally, here is =
a four-part lujvo with a cmavo in it, based on the tanru =93nakni ke cinse =
ctuca=94 or =93male (sexual teacher)=94. The =93ke=94 cmavo ensures the int=
erpretation =93teacher of sexuality who is male=94, rather than =93teacher =
of male sexuality=94. Here are the possible forms of the lujvo, both before=
 and after hyphenation:
+
+<p>
+<pre> 	nak-kem-cin-ctu	nakykemcinctu
+	nak-kem-cin-ctuca	nakykemcinctuca
+	nak-kem-cins-ctu	nakykemcinsyctu
+	nak-kem-cins-ctuca	nakykemcinsyctuca
+	nakn-kem-cin-ctu	naknykemcinctu
+	nakn-kem-cin-ctuca	naknykemcinctuca
+	nakn-kem-cins-ctu	naknykemcinsyctu
+	nakn-kem-cins-ctuca	naknykemcinsyctuca
+</pre><p>Of these forms, =93nakykemcinctu=94 is the shortest and is prefer=
red by the scoring algorithm. On the whole, however, it might be better to =
just make a lujvo for =93cinse ctuca=94 (which would be =93cinctu=94) since=
 the sex of the teacher is rarely important. If there was a reason to speci=
fy =93male=94, then the simpler tanru =93nakni cinctu=94 (=93male sexual-te=
acher=94) would be appropriate. This tanru is actually shorter than the fou=
r-part lujvo, since the =93ke=94 required for grouping need not be expresse=
d.
+<p>
+<a name=3Ds14><h3>The gismu creation algorithm</h3>
+<p>
+<cx "gismu, algorithm for">  XE "gismu: algorithm for"   The gismu were cr=
eated through the following process:
+<p>
+<cx "source languages, use in creating gismu">  XE "source languages: use =
in creating gismu"   <dl compact><dt>1)	<dd>At least one word was found in =
each of the six source languages (Chinese, English, Hindi, Spanish, Russian=
, Arabic) corresponding to the proposed gismu. This word was rendered into =
Lojban phonetics rather liberally: consonant clusters consisting of a stop =
and the corresponding fricative were simplified to just the fricative (=93t=
c=94 became =93c=94, =93dj=94 became =93j=94) and non-Lojban vowels were ma=
pped onto Lojban ones. Furthermore, morphological endings were dropped. The=
 same mapping rules were applied to all six languages for the sake of consi=
stency.
+<p><dt>2)	<dd>All possible gismu forms were matched against the six source=
-language forms. The matches were scored as follows:
+</dl><cx "gismu creation, scoring rules">  XE =93gismu: creation, scoring =
rules"   <dl compact><dt>2a)	<dd>If three or more letters were the same in =
the proposed gismu and the source-language word, and appeared in the same o=
rder, the score was equal to the number of letters that were the same. Inte=
rvening letters, if any, did not matter.
+<p><dt>2b)	<dd>If exactly two letters were the same in the proposed gismu =
and the source-language word, and either the two letters were consecutive i=
n both words, or were separated by a single letter in both words, the score=
 was 2. Letters in reversed order got no score.=20
+<p><dt>2c)	<dd>Otherwise, the score was 0.
+<p><dt>3)	<dd>The scores were divided by the length of the source-language=
 word in its Lojbanized form, and then multiplied by a weighting value spec=
ific to each language, reflecting the proportional number of first-language=
 and second-language speakers of the language. (Second-language speakers we=
re reckoned at half their actual numbers.) The weights were chosen to sum t=
o 1.00. The sum of the weighted scores was the total score for the proposed=
 gismu form.
+</dl><cx "gismu creation, considerations for selection after scoring">  XE=
 =93gismu: creation, considerations for selection after scoring"   <dl comp=
act><dt>4)	<dd>Any gismu forms that conflicted with existing gismu were rem=
oved. Obviously, being identical with an existing gismu constitutes a confl=
ict. In addition, a proposed gismu that was identical to an existing gismu =
except for the final vowel was considered a conflict, since two such gismu =
would have identical 4-letter rafsi.
+		</dl><cx "gismu, too-similar">  XE "gismu: too-similar"   <cx "gismu cre=
ation, proscribed gismu pairs">  XE =93gismu: creation, proscribed gismu pa=
irs"   <dl compact><dt><dd>More subtly: If the proposed gismu was identical=
 to an existing gismu except for a single consonant, and the consonant was =
=93too similar=94 based on the following table, then the proposed gismu was=
 rejected.
+<p><dt>	<dd>proposed gismu		   existing gismu
+<p><dt>		<dd>b			p, v
+		c			j, s
+		d			t
+		f			p, v
+		g			k, x
+		j			c, z
+		k			g, x
+		l			r
+		m			n
+		n			m
+		p			b, f
+		r			l
+		s			c, z
+		t			d
+		v			b, f
+		x			g, k
+		z			j, s
+	</dl><p>	See <a href=3D#s4>Section 4 </a>for an example.
+<p>
+<cx "gismu creation, and transcription blunders">  XE =93gismu: creation, =
and transcription blunders"   <dl compact><dt>5)	<dd>The gismu form with th=
e highest score usually became the actual gismu. Sometimes a lower-scoring =
form was used to provide a better rafsi. A few gismu were changed in error =
as a result of transcription blunders (for example, the gismu =93gismu=94 s=
hould have been =93gicmu=94, but it's too late to fix it now).
+		</dl><cx "gismu, source-language weights for">  XE "gismu: source-langua=
ge weights for"   The language weights used to make most of the gismu were =
as follows:
+<p>
+<pre> 	Chinese	0.36
+	English	0.21
+	Hindi		0.16
+	Spanish	0.11
+	Russian	0.09
+	Arabic		0.07
+		</pre>reflecting 1985 number-of-speakers data. A few gismu were made muc=
h later <dl compact><dt>using updated weights:	<dd>
+<p><dt>	<dd>Chinese	0.347
+	Hindi		0.196
+	English	0.160
+	Spanish	0.123
+	Russian	0.089
+	Arabic		0.085
+		</dl>(English and Hindi switched places due to demographic changes.)
+<p>
+Note that the stressed vowel of the gismu was considered sufficiently dist=
inctive that two or more gismu may differ only in this vowel; as an extreme=
 example, =93bradi=94, =93bredi=94, =93bridi=94, and =93brodi=94 (but fortu=
nately not =93brudi=94) are all existing gismu.
+<p>
+<a name=3Ds15><h3>Cultural and other non-algorithmic gismu</h3>
+<p>
+<cx "gismu, exceptions to gismu creation by algorithm">  XE "gismu: except=
ions to gismu creation by algorithm"   <cx "gismu, coined">  XE "gismu: coi=
ned"   The following gismu were not made by the gismu creation algorithm. T=
hey are, in effect, coined words similar to fu'ivla. They are exceptions to=
 the otherwise mandatory gismu creation algorithm where there was sufficien=
t justification for such exceptions. Except for the small metric prefixes a=
nd the assignable predicates beginning with =93brod=94, they all end in the=
 letter =93o=94, which is otherwise a rare letter in Lojban gismu.
+<p>
+<cx "gismu, Lojban-specific">  XE "gismu: Lojban-specific"   The following=
 gismu represent concepts that are sufficiently unique to Lojban that they =
were either coined from combining forms of other gismu, or else made up out=
 of whole cloth. These gismu are thus conceptually similar to lujvo even th=
ough they are only five letters long; however, unlike lujvo, they have rafs=
i assigned to them for use in building more complex lujvo. Assigning gismu =
to these concepts helps to keep the resulting lujvo reasonably short.
+<p>
+<pre> 	broda	1st assignable predicate
+	brode	2nd assignable predicate
+	brodi	3rd assignable predicate
+	brodo	4th assignable predicate
+	brodu	5th assignable predicate
+	cmavo	structure word (from =93cmalu valsi=94)
+	lojbo	Lojbanic (from =93logji bangu=94)
+	lujvo	compound word (from =93pluja valsi=94)
+	mekso	Mathematical EXpression
+</pre><p>It is important to understand that even though =93cmavo=94, =93lo=
jbo=94, and =93lujvo=94 were made up from parts of other gismu, they are no=
w full-fledged gismu used in exactly the same way as all other gismu, both =
in grammar and in word formation.
+<p>
+<cx "gismu, scientific-mathematical">  XE "gismu: scientific-mathematical"=
   The following three groups of gismu represent concepts drawn from the in=
ternational language of science and mathematics. They are used for concepts=
 that are represented in most languages by a root which is recognized inter=
nationally.
+<p>
+Small metric prefixes (values less than 1):
+<p>
+<pre> 	decti	.1/deci
+	centi	.01/centi
+	milti	.001/milli
+	mikri	1E-6/micro
+	nanvi	1E-9/nano
+	picti	1E-12/pico
+	femti	1E-15/femto
+	xatsi	1E-18/atto
+	zepti	1E-21/zepto
+	gocti	1E-24/yocto
+</pre><p>Large metric prefixes (values greater than 1):
+<p>
+<pre> 	dekto	10/deka
+	xecto	100/hecto
+	kilto	1000/kilo
+	megdo	1E6/mega
+	gigdo	1E9/giga
+	terto	1E12/tera
+	petso	1E15/peta
+	xexso	1E18/exa
+	zetro	1E21/zetta
+	gotro	1E24/yotta
+</pre><p>Other scientific or mathematical terms:
+<p>
+<pre> 	delno	candela
+	kelvo	kelvin
+	molro	mole
+	radno	radian
+	sinso	sine
+	stero	steradian
+	tanjo	tangent
+	xampo	ampere
+</pre><p>The gismu =93sinso=94 and =93tanjo=94 were only made non-algorith=
mically because they were identical (having been borrowed from a common sou=
rce) in all the dictionaries that had translations. The other terms in this=
 group are units in the international metric system; some metric units, how=
ever, were made by the ordinary process (usually because they are different=
 in Chinese).
+<p>
+<cx "gismu, cultural">  XE "gismu: cultural"   Finally, there are the cult=
ural gismu, which are also borrowed, but by modifying a word from one parti=
cular language, instead of using the multi-lingual gismu creation algorithm=
. Cultural gismu are used for words that have local importance to a particu=
lar culture; other cultures or languages may have no word for the concept a=
t all, or may borrow the word from its home culture, just as Lojban does. I=
n such a case, the gismu algorithm, which uses weighted averages, doesn't a=
ccurately represent the frequency of usage of the individual concept. Cultu=
ral gismu are not even required to be based on the six major languages.
+<p>
+<cx "gismu, for languages">  XE "gismu: for languages"   <cx "gismu for Lo=
jban source languages">  XE =93gismu: for Lojban source languages"   The si=
x Lojban source languages:
+<p>
+<pre>	jungo	Chinese (from =93Zhong<sup>1</sup>guo<sup>2</sup>=94)
+	glico	English
+	xindo	Hindi
+	spano	Spanish
+	rusko	Russian
+	xrabo	Arabic
+</pre><p>Seven other widely spoken languages that were on the list of cand=
idates for gismu-making, but weren't used:
+<p>
+<pre> 	bengo	Bengali
+	porto	Portuguese
+	baxso	Bahasa Melayu/Bahasa Indonesia
+	ponjo	Japanese (from =93Nippon=94)
+	dotco	German (from =93Deutsch=94)
+	fraso	French (from =93Fran=E7&ccedil;ais=94)
+	xurdo	Urdu
+</pre>(Urdu and Hindi began as the same language with different writing sy=
stems, but have now become somewhat different, principally in borrowed voca=
bulary. Urdu-speakers were counted along with Hindi-speakers when weights w=
ere assigned for gismu-making purposes.)
+<p>
+<cx "gismu, for countries">  XE "gismu: for countries"   Countries with a =
large number of speakers of any of the above languages (where the meaning o=
f =93large=94 is dependent on the specific language):
+<p>
+<dl compact><dt>	<dd>English:
+	merko	American
+	brito	British
+	skoto	Scottish
+	sralo	Australian
+	kadno	Canadian
+<p><dt>	<dd>Spanish:
+	gento	Argentinian
+	mexno	Mexican
+<p><dt>	<dd>Russian:
+	softo	Soviet/USSR
+	vukro	Ukrainian
+<p><dt>	<dd>Arabic:
+	filso	Palestinian
+	jerxo	Algerian
+	jordo	Jordanian
+	libjo	Libyan
+	lubno	Lebanese
+	misro	Egyptian (from =93Mizraim=94)
+	morko	Moroccan
+	rakso	Iraqi
+	sadjo	Saudi
+	sirxo	Syrian
+<p><dt>	<dd>Bahasa Melayu/Bahasa Indonesia:
+	bindo	Indonesian
+	meljo	Malaysian
+<p><dt>	<dd>Portuguese:
+	brazo	Brazilian
+<p><dt>	<dd>Urdu:
+	kisto	Pakistani
+</dl><cx "gismu, geographical">  XE "gismu: geographical"   <cx "continent=
s, gismu for">  XE "continents: gismu for"   The continents (and oceanic re=
gions) of the Earth:
+<p>
+<pre> 	bemro	North American (from =93berti merko=94)
+	dzipo	Antarctican (from =93cadzu cipni=94)
+	ketco	South American (from =93Quechua=94)
+	friko	African
+	polno	Polynesian/Oceanic
+	ropno	European
+	xazdo	Asiatic
+</pre><cx "gismu, ethnic">  XE "gismu: ethnic"   A few smaller but histori=
cally important cultures:
+<p>
+<pre> 	latmo	Latin/Roman
+	srito	Sanskrit
+	xebro	Hebrew/Israeli/Jewish
+	xelso	Greek (from =93Hellas=94)
+
+</pre><cx "gismu, religious">  XE "gismu: religious"   Major world religio=
ns:
+<p>
+<pre> 	budjo	Buddhist
+	dadjo	Taoist
+	muslo	Islamic/Moslem
+	xriso	Christian
+</pre><p>A few terms that cover multiple groups of the above:
+<p>
+<pre> 	jegvo	Jehovist (Judeo-Christian-Moslem)
+	semto	Semitic
+	slovo	Slavic
+	xispo	Hispanic (New World Spanish)
+</pre><a name=3Ds16><h3>rafsi fu'ivla: a proposal</h3>
+<p>
+<cx "rafsi fu'ivla">  XE "rafsi fu'ivla proposal"   <cx "cultural words, r=
afsi fu'ivla proposal for">  XE "cultural words: rafsi fu'ivla proposal for=
"   The list of cultures represented by gismu, given in <a href=3D#s15>Sect=
ion 15</a>, is unavoidably controversial. Much time has been spent debating=
 whether this or that culture =93deserves a gismu=94 or =93must languish in=
 fu'ivla space=94. To help defuse this argument, a last-minute proposal was=
 made when this book was already substantially complete. I have added it he=
re with experimental status: it is not yet a standard part of Lojban, since=
 all its implications have not been tested in open debate, and it affects a=
 part of the language (lujvo-making) that has long been stable, but is know=
n to be fragile in the face of small changes. (Many attempts were made to a=
dd general mechanisms for making lujvo that contained fu'ivla, but all fail=
ed on obvious or obscure counterexamples; finally the general =93zei=94 mec=
hanism was devised instead.)
+<p>
+<cx "CCVVCV fu'ivla, and rafsi fu'ivla proposal">  XE "CCVVCV fu'ivla: and=
 rafsi fu'ivla proposal"   <cx "fu'ivla, form for rafsi fu'ivla proposal"> =
 XE "fu'ivla: form for rafsi fu'ivla proposal"   The first part of the prop=
osal is uncontroversial and involves no change to the language mechanisms. =
All valid Type 4 fu'ivla of the form CCVVCV would be reserved for cultural =
brivla analogous to those described in <a href=3D#s15>Section 15</a>. For e=
xample,
+<p>
+<pre><a name=3De16d1>16.1)	tci'ile
+	Chilean
+is of the appropriate form, and passes all tests required of a Stage 4 fu'=
ivla. No two fu'ivla of this form would be allowed to coexist if they diffe=
red only in the final vowel; this rule was applied to gismu, but does not a=
pply to other fu'ivla or to lujvo.
+<p>
+The second, and fully experimental, part of the proposal is to allow rafsi=
 to be formed from these cultural fu'ivla by removing the final vowel and t=
reating the result as a 4-letter rafsi (although it would contain five lett=
ers, not four). These rafsi could then be used on a par with all other rafs=
i in forming lujvo. The tanru
+<p>
+<ex "Chilean desert">  XE "Chilean desert: example=93  <pre><a name=3De16d=
2>16.2)	tci'ile ke canre tutra
+	Chilean type-of (sand territory)
+	Chilean desert
+</pre>could be represented by the lujvo
+<p>
+<pre><a name=3De16d3>16.3)	tci'ilykemcantutra
+</pre>which is an illegal word in standard Lojban, but a valid lujvo under=
 this proposal. There would be no short rafsi or 5-letter rafsi assigned to=
 any fu'ivla, so no fu'ivla could appear as the last element of a lujvo.
+<p>
+The cultural fu'ivla introduced under this proposal are called =93rafsi fu=
'ivla=94, since they are distinguished from other Type 4 fu'ivla by the pro=
perty of having rafsi. If this proposal is workable and introduces no probl=
ems into Lojban morphology, it might become standard for all Type 4 fu'ivla=
, including those made for plants, animals, foodstuffs, and other things.
+
+<p>
+<body><center><p align=3Dcenter>
+<img src=3Dchap5.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D4=
05>
+
+<h2>Chapter 5
+<br>
+=93Pretty Little Girls' School=94: The Structure Of Lojban selbri</h2>
+<h6>$Revision: 4.0 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>Lojban content words: brivla</h3>
+<p>
+<cx "bridi, definition">  XE "bridi: definition"   <cx "selbri, definition=
">  XE "selbri: definition"   <cx "selbri, relation to bridi">  XE "selbri:=
 relation to bridi"   <cx "bridi, relation to selbri">  XE "bridi: relation=
 to selbri"   At the center, logically and often physically, of every Lojba=
n bridi is one or more words which constitute the selbri. A bridi expresses=
 a relationship between things: the selbri specifies which relationship is =
referred to. The difference between:
+<p>
+<pre><a name=3De1d1>1.1)	do mamta mi
+	You are-a-mother-of me.
+	You are my mother.
+</pre>and
+<p>
+<pre><a name=3De1d2>1.2)	do patfu mi
+	You are-a-father-of me.
+	You are my father.
+</pre>lies in the different selbri.
+<p>
+<cx "selbri, brivla as">  XE "selbri: brivla as"   <cx "brivla as selbri">=
  XE "brivla as selbri"   <cx "brivla, types">  XE "brivla: types"   The si=
mplest kind of selbri is a single Lojban content word: a brivla. There are =
three different varieties of brivla: those which are built into the languag=
e (the gismu), those which are derived from combinations of the gismu (the =
lujvo), and those which are taken (usually in a modified form) from other l=
anguages (the fu'ivla). In addition, there are a few cmavo that can act lik=
e brivla; these are mentioned in <a href=3D#s9>Section 9</a>, and discussed=
 in full in <a href=3Dchap7.html>Chapter 7</a>.
+<p>
+For the purposes of this chapter, however, all brivla are alike. For examp=
le,
+<p>
+<pre><a name=3De1d3>1.3)	ta bloti
+	That is-a-boat.
+	That is a boat.
+<a name=3De1d4>1.4)	ta brablo
+	That is-a-large-boat.
+	That is a ship.
+</pre<ex "schooner">  XE "schooner: example=93  <pre><a name=3De1d5>1.5)	t=
a blotrskunri
+	That is-a-(boat)-schooner.
+	That is a schooner.
+</pre>illustrate the three types of brivla (gismu, lujvo, and fu'ivla resp=
ectively), but in each case the selbri is composed of a single word whose m=
eaning can be learned independent of its origins.
+<p>
+The remainder of this chapter will mostly use gismu as example brivla, bec=
ause they are short. However, it is important to keep in mind that wherever=
 a gismu appears, it could be replaced by any other kind of brivla.
+<p>
+<a name=3Ds2><h3>Simple tanru</h3>
+<p>
+<cx "tanru, simple">  XE "tanru: simple"   <cx "tanru, definition">  XE "t=
anru: definition"   Beyond the single brivla, a selbri may consist of two b=
rivla placed together. When a selbri is built in this way from more than on=
e brivla, it is called a tanru, a word with no single English equivalent. T=
he nearest analogue to tanru in English are combinations of two nouns such =
as =93lemon tree=94. There is no way to tell just by looking at the phrase =
=93lemon tree=94 exactly what it refers to, even if you know the meanings o=
f =93lemon=94 and =93tree=94 by themselves. As English-speakers, we must si=
mply know that it refers to =93a tree which bears lemons as fruits=94. A pe=
rson who didn't know English very well might think of it as analogous to =
=93brown tree=94 and wonder, =93What kind of tree is lemon-colored?=94
+<p>
+<cx "adverb-verb combination, with tanru">  XE "adverb-verb combination: w=
ith tanru"   <cx "adjective-noun combination, with tanru">  XE "adjective-n=
oun combination: with tanru"   In Lojban, tanru are also used for the same =
purposes as English adjective-noun combinations like =93big boy=94 and adve=
rb-verb combinations like =93quickly run=94. This is a consequence of Lojba=
n not having any such categories as =93noun=94, =93verb=94, =93adjective=94=
, or =93adverb=94. English words belonging to any of these categories are t=
ranslated by simple brivla in Lojban. Here are some examples of tanru:
+<p>
+<ex "lemon tree">  XE "lemon tree: example=93  <pre><a name=3De2d1>2.1)	tu=
 pelnimre tricu
+	That-yonder is-a-(lemon tree).
+	That is a lemon tree.
+<a name=3De2d2>2.2)	la djan. barda nanla
+	John is-a-big boy.
+	John is a big boy.
+<a name=3De2d3>2.3)	mi sutra bajra
+	I quick run.
+	I quickly run/I run quickly.
+</pre><p>Note that =93pelnimre=94 is a lujvo for =93lemon=94; it is derive=
d from the gismu =93pelxu=94, yellow, and =93nimre=94, citrus. Note also th=
at =93sutra=94 can mean =93fast/quick=94 or =93quickly=94 depending on its =
use:
+<p>
+<pre><a name=3De2d4>2.4)	mi sutra
+	I am-fast/quick.
+</pre><ex "quick runner">  XE "quick runner: example=93  shows =93sutra=94=
 used to translate an adjective, whereas in <a href=3D#e2d3>Example 2.3 </a=
>it is translating an adverb. (Another correct translation of <a href=3D#e2=
d3>Example 2.3</a>, however, would be =93I am a quick runner=94.)
+<p>
+<cx "seltau, definition of">  XE "seltau: definition of"   <cx "tertau, de=
finition of">  XE "tertau: definition of"   There are special Lojban terms =
for the two components of a tanru, derived from the place structure of the =
word =93tanru=94. The first component is called the =93seltau=94, and the s=
econd component is called the =93tertau=94.
+<p>
+<cx "tanru, primary meaning of">  XE "tanru: primary meaning of"   <cx "te=
rtau, effect on meaning of tanru">  XE "tertau: effect on meaning of tanru"=
   The most important rule for use in interpreting tanru is that the tertau=
 carries the primary meaning. A =93pelnimre tricu=94 is primarily a tree, a=
nd only secondarily is it connected with lemons in some way. For this reaso=
n, an alternative translation of <a href=3D#e2d1>Example 2.1 </a>would be:
+<p>
+<pre><a name=3De2d5>2.5)	That is a lemon type of tree.
+</pre><p>This =93type of=94 relationship between the components of a tanru=
 is fundamental to the tanru concept.
+<p>
+<cx "seltau, effect on meaning of tanru">  XE "seltau: effect on meaning o=
f tanru"   <cx "modifier, seltau as">  XE "modifier: seltau as"   We may al=
so say that the seltau modifies the meaning of the tertau:
+<p>
+<pre><a name=3De2d6>2.6)	That is a tree which is lemon-ish 	(in the way ap=
propriate to trees)
+</pre>would be another possible translation of <a href=3D#e2d1>Example 2.1=
</a>. In the same way, a more explicit translation of <a href=3D#e2d2>Examp=
le 2.2 </a>might be:
+<p>
+<pre><a name=3De2d7>2.7)	John is a boy who is big in the way that boys are=
 big.
+</pre><p>This =93way that boys are big=94 would be quite different from th=
e way in which elephants are big; big-for-a-boy is small-for-an-elephant.
+<p>
+<cx "tanru, ambiguity of">  XE "tanru: ambiguity of"   <cx "ambiguity of t=
anru">  XE "ambiguity of tanru"   All tanru are ambiguous semantically. Pos=
sible translations of:
+<p>
+<pre><a name=3De2d8>2.8)	ta klama jubme
+	That is-a-goer type-of-table.
+</pre>include:
+<p>
+<ex "goer table">  XE "goer table: example=93  <dl compact><dt>	<dd>That i=
s a table which goes (a wheeled table, perhaps).
+	That is a table owned by one who goes.
+	That is a table used by those who go (a sports doctor's table?).
+	That is a table when it goes (otherwise it is a chair?).
+</dl><cx "tanru, as ambiguous">  XE "tanru: as ambiguous"   In each case t=
he object referred to is a =93goer type of table=94, but the ambiguous =93t=
ype of=94 relationship can mean one of many things. A speaker who uses tanr=
u (and pragmatically all speakers must) takes the risk of being misundersto=
od. Using tanru is convenient because they are short and expressive; the ci=
rcumlocution required to squeeze out all ambiguity can require too much eff=
ort.
+<p>
+<cx "tanru, meaning of">  XE "tanru: meaning of"   No general theory cover=
ing the meaning of all possible tanru exists; probably no such theory can e=
xist. However, some regularities obviously do exist:
+<p>
+<pre><a name=3De2d9>2.9)	do barda prenu
+	You are-a-large person.
+<a name=3De2d10>2.10)	do cmalu prenu
+	You are-a-small person.
+</pre>are parallel tanru, in the sense that the relationship between =93ba=
rda=94 and =93prenu=94 is the same as that between =93cmalu=94 and =93prenu=
=94. <a href=3D#s14>Section 14 </a>and <a href=3D#s15>Section 15 </a>contai=
n a partial listing of some types of tanru, with examples.
+<p>
+<a name=3Ds3><h3>Three-part tanru grouping with =93bo=94</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	bo	BO	closest scope grouping
+</pre><ex "girls' school, little">  XE "girls' school: little, example=93 =
 <cx "tanru grouping, three-part">  XE "tanru grouping: three-part"   Consi=
der the English sentence:
+<p>
+<pre><a name=3De3d1>3.1)	That's a little girls' school.
+</pre><p>What does it mean? Two possible readings are:
+<p>
+<pre><a name=3De3d2>3.2)	That's a little school for girls.
+<a name=3De3d3>3.3)	That's a school for little girls.
+ </pre><cx "speech rhythm, for grouping in English">  XE "speech rhythm: f=
or grouping in English"   This ambiguity is quite different from the simple=
 tanru ambiguity described in <a href=3D#s2>Section 2</a>. We understand th=
at =93girls' school=94 means =93a school where girls are the students=94, a=
nd not =93a school where girls are the teachers=94 or =93a school which is =
a girl=94 (!). Likewise, we understand that =93little girl=94 means =93girl=
 who is small=94. This is an ambiguity of grouping. Is =93girls' school=94 =
to be taken as a unit, with =93little=94 specifying the type of girls' scho=
ol? Or is =93little girl=94 to be taken as a unit, specifying the type of s=
chool? In English speech, different tones of voice, or exaggerated speech r=
hythm showing the grouping, are used to make the distinction; English writi=
ng usually leaves it unrepresented.
+<p>
+<lx "bo">  XE "bo"   <lx "BO">  XE "BO selma=92o"   Lojban makes no use of=
 tones of voice for any purpose; explicit words are used to do the work. Th=
e cmavo =93bo=94 (which belongs to selma'o BO) may be placed between the tw=
o brivla which are most closely associated. Therefore, a Lojban translation=
 of <a href=3D#e3d2>Example 3.2 </a>would be:
+<p>
+<pre><a name=3De3d4>3.4)	ta cmalu nixli bo ckule
+	That is-a-small girl =96 school.
+</pre><a href=3D#e3d3>Example 3.3 </a>might be translated:
+<p>
+<pre><a name=3De3d5>3.5)	ta cmalu bo nixli ckule
+	That is-a-small =96 girl school.
+</pre><p>The =93bo=94 is represented in the literal translation by a hyphe=
n because in written English a hyphen is sometimes used for the same purpos=
e: =93a big dog-catcher=94 would be quite different from a =93big-dog catch=
er=94 (presumably someone who catches only big dogs).
+<p>
+<cx "tanru nested within tanru">  XE "tanru nested within tanru"   Analysi=
s of <a href=3D#e3d4>Example 3.4 </a>and <a href=3D#e3d5>Example 3.5 </a>re=
veals a tanru nested within a tanru. In <a href=3D#e3d4>Example 3.4</a>, th=
e main tanru has a seltau of =93cmalu=94 and a tertau of =93nixli bo ckule=
=94; the tertau is itself a tanru with =93nixli=94 as the seltau and =93cku=
le=94 as the tertau. In <a href=3D#e3d5>Example 3.5</a>, on the other hand,=
 the seltau is =93cmalu bo nixli=94 (itself a tanru), whereas the tertau is=
 =93ckule=94. This structure of tanru nested within tanru forms the basis f=
or all the more complex types of selbri that will be explained below.
+<p>
+What about <a href=3D#e3d6>Example 3.6</a>? What does it mean?
+<p>
+<pre><a name=3De3d6>3.6)	ta cmalu nixli ckule
+	That is-a-small girl school.
+</pre><cx "left-grouping rule, definition of">  XE "left-grouping rule: de=
finition of"   <cx "tanru, default left-grouping of">  XE "tanru: default l=
eft-grouping of"   The rules of Lojban do not leave this sentence ambiguous=
, as the rules of English do with <a href=3D#e3d1>Example 3.1</a>. The choi=
ce made by the language designers is to say that <a href=3D#e3d6>Example 3.=
6 </a>means the same as <a href=3D#e3d5>Example 3.5</a>. This is true no ma=
tter what three brivla are used: the leftmost two are always grouped togeth=
er. This rule is called the =93left-grouping rule=94. Left-grouping in seem=
ingly ambiguous structures is quite common =97 though not universal =97 in =
other contexts in Lojban.
+<p>
+Another way to express the English meaning of <a href=3D#e3d4>Example 3.4 =
</a>and <a href=3D#e3d5>Example 3.5</a>, using parentheses to mark grouping=
, is:
+<p>
+<pre><a name=3De3d7>3.7)	ta cmalu nixli bo ckule
+	That is-a-small type-of (girl type-of school).
+<a name=3De3d8>3.8)	ta cmalu bo nixli ckule
+	That is-a-(small type-of girl) type-of school.
+</pre><p>Because =93type-of=94 is implicit in the Lojban tanru form, it ha=
s no Lojban equivalent.
+<p>
+Note: It is perfectly legal, though pointless, to insert =93bo=94 into a s=
imple tanru:
+<p>
+<pre><a name=3De3d9>3.9)	ta klama bo jubme
+	That is-a goer =96 table
+</pre>is a legal Lojban bridi that means exactly the same thing as <a href=
=3D#e2d8>Example 2.8</a>, and is ambiguous in exactly the same ways. The cm=
avo =93bo=94 serves only to resolve grouping ambiguity: it says nothing abo=
ut the more basic ambiguity present in all tanru.
+
+<p>
+<a name=3Ds4><h3>Complex tanru grouping</h3>
+<p>
+<cx "tanru grouping, complex">  XE "tanru grouping: complex"   If one elem=
ent of a tanru can be another tanru, why not both elements?
+<p>
+<pre><a name=3De4d1>4.1)	do mutce bo barda gerku bo kavbu
+	You are-a-(very type-of large) (dog type-of capturer).
+	You are a very large dog-catcher.
+</pre><p>In <a href=3D#e4d1>Example 4.1</a>, the selbri is a tanru with se=
ltau =93mutce bo barda=94 and tertau =93gerku bo kavbu=94. It is worth emph=
asizing once again that this tanru has the same fundamental ambiguity as al=
l other Lojban tanru: the sense in which the =93dog type-of capturer=94 is =
said to be =93very type-of large=94 is not precisely specified. Presumably =
it is his body which is large, but theoretically it could be one of his oth=
er properties.
+<p>
+<cx "pretty, English ambiguity of">  XE "pretty: English ambiguity of"   W=
e will now justify the title of this chapter by exploring the ramifications=
 of the phrase =93pretty little girls' school=94, an expansion of the tanru=
 used in <a href=3D#s3>Section 3 </a>to four brivla. (Although this example=
 has been used in the Loglan Project almost since the beginning =97 it firs=
t appeared in Quine's book <cite>Word and Object </cite>(1960) =97 it is ac=
tually a mediocre example because of the ambiguity of English =93pretty=94;=
 it can mean =93beautiful=94, the sense intended here, or it can mean =93ve=
ry=94. Lojban =93melbi=94 is not subject to this ambiguity: it means only =
=93beautiful=94.)
+Here are four ways to group this phrase:
+<p>
+<pre><a name=3De4d2>4.2)	ta melbi cmalu nixli ckule
+	That is-a-((pretty type-of little) type-of girl) type-of school.
+	That is a school for girls who are beautifully small.
+<a name=3De4d3>4.3)	ta melbi cmalu nixli bo ckule
+	That is-a-(pretty type-of little) (girl type-of school).
+	That is a girls' school which is beautifully small.
+<a name=3De4d4>4.4)	ta melbi cmalu bo nixli ckule
+	That is-a-(pretty type-of (little type-of girl)) type-of school.
+	That is a school for small girls who are beautiful.
+<a name=3De4d5>4.5)	ta melbi cmalu bo nixli bo ckule
+	That is-a-pretty type-of (little type-of (girl type-of school)).
+	That is a small school for girls which is beautiful.
+</pre><cx "right-grouping rule, definition of">  XE "right-grouping rule: =
definition of"   <cx "right-grouping in tanru, with bo">  XE "right-groupin=
g in tanru: with bo"   <cx "tanru grouping, with bo">  XE "tanru grouping: =
with bo"   <cx "bo, for right-grouping in tanru">  XE "bo: for right-groupi=
ng in tanru"   <a href=3D#e4d5>Example 4.5 </a>uses a construction which ha=
s not been seen before: =93cmalu bo nixli bo ckule=94, with two consecutive=
 uses of =93bo=94 between brivla. The rule for multiple =93bo=94 constructi=
ons is the opposite of the rule when no =93bo=94 is present at all: the las=
t two are grouped together. Not surprisingly, this is called the =93right-g=
rouping rule=94, and it is associated with every use of =93bo=94 in the lan=
guage. Therefore,
+<p>
+<pre><a name=3De4d6>4.6)	ta cmalu bo nixli bo ckule
+	That is-a-little type-of (girl type-of school).
+</pre>means the same as <a href=3D#e3d4>Example 3.4</a>, not <a href=3D#e3=
d5>Example 3.5</a>. This rule may seem peculiar at first, but one of its co=
nsequences is that =93bo=94 is never necessary between the first two elemen=
ts of any of the complex tanru presented so far: all of <a href=3D#e4d{2}>E=
xamples 4.2 </a>through <a href=3D#e4d5>4.5 </a>could have =93bo=94 inserte=
d between =93melbi=94 and =93cmalu=94 with no change in meaning.
+<p>
+<a name=3Ds5><h3>Complex tanru with =93ke=94 and =93ke'e=94</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ke	KE	start grouping
+	ke'e	KEhE	end grouping
+</pre><cx "grouping parentheses">  XE "grouping parentheses"   <lx "ke">  =
XE "ke"   <lx "ke'e">  XE "ke'e"   <lx "KE">  XE "KE selma=92o"   <lx "KEhE=
">  XE "KEhE selma=92o"   <cx "tanru grouping, with ke">  XE "tanru groupin=
g: with ke"   There is, in fact, a fifth grouping of =93pretty little girls=
' school=94 that cannot be expressed with the resources explained so far. T=
o handle it, we must introduce the grouping parentheses cmavo, =93ke=94 and=
 =93ke'e=94 (belonging to selma'o KE and KEhE respectively). Any portion of=
 a selbri sandwiched between these two cmavo is taken to be a single tanru =
component, independently of what is adjacent to it. Thus, <a href=3D#e4d2>E=
xample 4.2 </a>can be rewritten in any of the following ways:
+<p>
+<pre><a name=3De5d1>5.1)	ta ke melbi cmalu ke'e nixli ckule
+	That is-a-( pretty little ) girl school.
+<a name=3De5d2>5.2)	ta ke ke melbi cmalu ke'e nixli ke'e ckule
+	That is-a-( ( pretty little ) girl ) school.
+<a name=3De5d3>5.3)	ta ke ke ke melbi cmalu ke'e nixli ke'e ckule ke'e
+	That is-a-( ( ( pretty little ) girl ) school ).
+</pre><p>Even more versions could be created simply by placing any number =
of =93ke=94 cmavo at the beginning of the selbri, and a like number of =93k=
e'e=94 cmavo at its end. Obviously, all of these are a waste of breath once=
 the left-grouping rule has been grasped. However, the following is equival=
ent to <a href=3D#e4d4>Example 4.4 </a>and may be easier to understand:
+<p>
+<pre><a name=3De5d4>5.4)	ta melbi ke cmalu nixli ke'e ckule
+	That is-a-(pretty type-of (little type-of girl) ) type-of school.
+</pre><p>Likewise, a =93ke=94 and =93ke'e=94 version of <a href=3D#e4d3>Ex=
ample 4.3 </a>would be:
+<p>
+<pre><a name=3De5d5>5.5)	ta melbi cmalu ke nixli ckule [ke'e]
+	That is-a-(pretty type-of little) (girl type-of school).
+</pre><p>The final =93ke'e=94 is given in square brackets here to indicate=
 that it can be elided. It is always possible to elide =93ke'e=94 at the en=
d of the selbri, making <a href=3D#e5d5>Example 5.5 </a>as terse as <a href=
=3D#e4d3>Example 4.3</a>.
+<p>
+Now how about that fifth grouping? It is
+<p>
+<pre><a name=3De5d6>5.6)	ta melbi ke cmalu nixli ckule [ke'e]
+	That is-a-pretty type-of ( (little type-of girl) type-of school )
+	That is a beautiful school for small girls.
+</pre><a href=3D#e5d6>Example 5.6 </a>is distinctly different in meaning f=
rom any of Examples 4.2 through 4.5. Note that within the =93ke=A0<dots>=85=
</dots>ke'e=94 parentheses, the left-grouping rule is applied to =93cmalu n=
ixli ckule=94.
+<p>
+<cx "tanru grouping, with ke and bo">  XE "tanru grouping: with ke and bo"=
   It is perfectly all right to mix =93bo=94 and =93ke=A0<dots>=85</dots>ke=
'e=94 in a single selbri. For instance, <a href=3D#e4d5>Example 4.5</a>, wh=
ich in pure =93ke=A0<dots>=85</dots>ke'e=94 form is
+<p>
+<pre><a name=3De5d7>5.7)	ta melbi ke cmalu ke nixli ckule [ke'e] [ke'e]
+	That is-a-pretty type-of ( little type-of (girl type-of school) ).
+
+</pre>can equivalently be expressed as:
+<p>
+<pre><a name=3De5d8>5.8)	ta melbi ke cmalu nixli bo ckule [ke'e]
+	That is-a-pretty type-of ( little type-of (girl type-of school) ).
+</pre>and in many other different forms as well.
+<p>
+<a name=3Ds6><h3>Logical connection within tanru</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	je	JA	tanru logical =93and=94
+	ja	JA	tanru logical =93or=94
+	joi	JOI	mixed mass =93and=94
+	gu'e	GUhA	tanru forethought logical =93and=94
+	gi	GI	forethought connection separator
+</pre><ex "big red dog">  XE "big red dog: example=93  Consider the Englis=
h phrase =93big red dog=94. How shall this be rendered as a Lojban tanru? T=
he naive attempt:
+<p>
+<pre><a name=3De6d1>6.1)	barda xunre gerku
+	(big type-of red) type-of dog
+</pre>will not do, as it means a dog whose redness is big, in whatever way=
 redness might be described as =93big=94. Nor is
+<p>
+<pre><a name=3De6d2>6.2)	barda xunre bo gerku
+	big type-of (red type-of dog)
+</pre><cx "adjective ordering">  XE "adjective ordering"   much better. Af=
ter all, the straightforward understanding of the English phrase is that th=
e dog is big as compared with other dogs, not merely as compared with other=
 red dogs. In fact, the bigness and redness are independent properties of t=
he dog, and only obscure rules of English adjective ordering prevent us fro=
m saying =93red big dog=94.
+<p>
+<cx "logical connectives, in tanru">  XE "logical connectives: in tanru"  =
 <lx "je">  XE "je"   The Lojban approach to this problem is to introduce t=
he cmavo =93je=94, which is one of the many equivalents of English =93and=
=94. A big red dog is one that is both big and red, and we can say:
+<p>
+<pre><a name=3De6d3>6.3)	barda je xunre gerku
+	(big and red) type-of dog
+</pre><p>Of course,
+<p>
+<pre><a name=3De6d4>6.4)	xunre je barda gerku
+	(red and big) type-of dog
+</pre><cx "logical connectives in tanru, effect on tanru grouping">  XE "l=
ogical connectives in tanru: effect on tanru grouping"   is equally satisfa=
ctory and means the same thing. As these examples indicate, joining two bri=
vla with =93je=94 makes them a unit for tanru purposes. However, explicit g=
rouping with =93bo=94 or =93ke=A0<dots>=85</dots>ke'e=94 associates brivla =
more closely than =93je=94 does:
+<p>
+<pre><a name=3De6d5>6.5)	barda je pelxu bo xunre gerku
+	barda je ke pelxu xunre ke'e gerku
+	(big and (yellow type-of red)) dog
+	big yellowish-red dog
+
+</pre><p>With no grouping indicators, we get:
+<p>
+<pre><a name=3De6d6>6.6)	barda je pelxu xunre gerku
+	((big and yellow) type-of red) type-of dog
+	biggish- and yellowish-red dog
+</pre>which again raises the question of <a href=3D#e6d1>Example 6.1</a>: =
what is does =93biggish-red=94 mean?
+<p>
+<cx "logical connectives in tanru, usefulness of">  XE "logical connective=
s in tanru: usefulness of"   Unlike =93bo=94 and =93ke=A0<dots>=85</dots>ke=
'e=94, =93je=94 is useful as well as merely legal within simple tanru. It m=
ay be used to partly resolve the ambiguity of simple tanru:
+<p>
+<pre><a name=3De6d7>6.7)	ta blanu je zdani
+	that is-blue and is-a-house
+</pre>definitely refers to something which is both blue and is a house, an=
d not to any of the other possible interpretations of simple =93blanu zdani=
=94. Furthermore, =93blanu zdani=94 refers to something which is blue in th=
e way that houses are blue; =93blanu je zdani=94 has no such implication =
=97 the blueness of a =93blanu je zdani=94 is independent of its houseness.
+<p>
+With the addition of =93je=94, many more versions of =93pretty little girl=
s' school=94 are made possible: see <a href=3D#s16>Section 16 </a>for a com=
plete list.
+<p>
+A subtle point in the semantics of tanru like <a href=3D#e6d3>Example 6.3 =
</a>needs special elucidation. There are at least two possible interpretati=
ons of:
+<p>
+<pre><a name=3De6d8>6.8)	ta melbi je nixli ckule
+	That is-a-(beautiful and girl) type-of school.
+</pre><p>It can be understood as:
+<p>
+<pre><a name=3De6d9>6.9)	That is a girls' school and a beautiful school.
+</pre>or as:
+<p>
+<pre><a name=3De6d10>6.10)	That is a school for things which are both girl=
s and beautiful.
+</pre><cx "logical connectives in tanru, ambiguity of">  XE "logical conne=
ctives in tanru: ambiguity of"   The interpretation specified by <a href=3D=
#e6d9>Example 6.9 </a>treats the tanru as a sort of abbreviation for:
+<p>
+<pre><a name=3De6d11>6.11)	ta ke melbi ckule ke'e je ke nixli ckule [ke'e]
+	That is-a-( beautiful type-of school ) and (girl type-of school)
+</pre>whereas the interpretation specified by <a href=3D#e6d10>Example 6.1=
0 </a>does not. This is a kind of semantic ambiguity for which Lojban does =
not compel a firm resolution. The way in which the school is said to be of =
type =93beautiful and girl=94 may entail that it is separately a beautiful =
school and a girls' school; but the alternative interpretation, that the me=
mbers of the school are beautiful and girls, is also possible. Still anothe=
r interpretation is:
+<p>
+<pre><a name=3De6d12>6.12)	That is a school for beautiful things and also =
for girls.
+</pre>so while the logical connectives help to resolve the meaning of tanr=
u, they by no means compel a single meaning in and of themselves.
+<p>
+<cx "logical connectives in tanru, effect on formal logical manipulations"=
>  XE "logical connectives in tanru: effect on formal logical manipulations=
"   In general, logical connectives within tanru cannot undergo the formal =
manipulations that are possible with the related logical connectives that e=
xist outside tanru; see <a href=3Dchap14.html>Chapter 14 </a>for further de=
tails.
+<p>
+<lx "JA">  XE "JA selma=92o"   The logical connective =93je=94 is only one=
 of the fourteen logical connectives that Lojban provides. Here are a few e=
xamples of some of the others:
+<p>
+<pre><a name=3De6d13>6.13)	le bajra cu jinga ja te jinga
+	The runner(s) is/are winner(s) or loser(s).
+<a name=3De6d14>6.14)	blanu naja lenku skapi
+	(blue only-if cold) skin
+	skin which is blue only if it is cold
+<a name=3De6d15>6.15)	xamgu jo cortu nuntavla
+	(good if-and-only-if short) speech
+	speech which is good if (and only if) it is short
+<a name=3De6d16>6.16)	vajni ju selpluka nuntavla
+	(important whether-or-not pleasing) event-of-talking
+	speech which is important, whether or not it is pleasing
+</pre><p>In <a href=3D#e6d13>Example 6.13</a>, =93ja=94 is grammatically e=
quivalent to =93je=94 but means =93or=94 (more precisely, =93and/or=94). Li=
kewise, =93naja=94 means =93only if=94 in <a href=3D#e6d14>Example 6.14</a>=
, =93jo=94 means =93if and only if=94 in <a href=3D#e6d15>Example 6.15</a>,=
 and =93ju=94 means =93whether or not=94 in <a href=3D#e6d16>Example 6.16</=
a>.
+<p>
+<cx "multiple logical connectives, within tanru">  XE "multiple logical co=
nnectives: within tanru"   Now consider the following example:
+<p>
+<pre><a name=3De6d17>6.17)	ricfu je blanu jabo crino
+	rich and (blue or green)
+</pre><lx "jabo">  XE "jabo"   <cx "tanru grouping with JA+BO, effect on t=
anru grouping">  XE "tanru grouping with JA+BO: effect on tanru grouping"  =
 which illustrates a new grammatical feature: the use of both =93ja=94 and =
=93bo=94 between tanru components. The two cmavo combine to form a compound=
 whose meaning is that of =93ja=94 but which groups more closely; =93jabo=
=94 is to =93ja=94 as plain =93bo=94 is to no cmavo at all. However, both =
=93ja=94 and =93jabo=94 group less closely than =93bo=94 does:
+<p>
+<pre><a name=3De6d18>6.18)	ricfu je blanu jabo crino bo blanu
+	rich and (blue or green =96 blue)
+	rich and (blue or greenish-blue)
+</pre><p>An alternative form of <a href=3D#e6d17>Example 6.17 </a>is:
+<p>
+<pre><a name=3De6d19>6.19)	ricfu je ke blanu ja crino [ke'e]
+	rich and (blue or green)
+</pre><cx "non-logical connectives, within tanru">  XE "non-logical connec=
tives: within tanru"   In addition to the logical connectives, there are al=
so a variety of non-logical connectives, grammatically equivalent to the lo=
gical ones. The only one with a well-understood meaning in tanru contexts i=
s =93joi=94, which is the kind of =93and=94 that denotes a mixture:
+<p>
+<pre><a name=3De6d20>6.20)	ti blanu joi xunre bolci
+	This is-a-(blue and red) ball.
+</pre><p>The ball described is neither solely red nor solely blue, but pro=
bably striped or in some other way exhibiting a combination of the two colo=
rs. <a href=3D#e6d20>Example 6.20 </a>is distinct from:
+<p>
+<pre><a name=3De6d21>6.21)	ti blanu xunre bolci
+	This is a bluish-red ball
+</pre>which would be a ball whose color is some sort of purple tending tow=
ard red, since =93xunre=94 is the more important of the two components. On =
the other hand,
+<p>
+<pre><a name=3De6d22>6.22)	ti blanu je xunre bolci
+	This is a (blue and red) ball
+</pre>is probably self-contradictory, seeming to claim that the ball is in=
dependently both blue and red at the same time, although some sensible inte=
rpretation may exist.
+<p>
+<cx "forethought logical connectives, within tanru">  XE "forethought logi=
cal connectives: within tanru"   <lx "gu'e">  XE "gu'e"   <lx "gi">  XE "gi=
"   Finally, just as English =93and=94 has the variant form =93both=A0<dots=
>=85</dots>and=94, so =93je=94 between tanru components has the variant for=
m =93gu'e=A0<dots>=85</dots>gi=94, where =93gu'e=94 is placed before the co=
mponents and =93gi=94 between them:
+<p>
+<pre><a name=3De6d23>6.23)	gu'e barda gi xunre gerku
+	(both big and red) type-of dog
+</pre>is equivalent in meaning to <a href=3D#e6d3>Example 6.3</a>. For eac=
h logical connective related to =93je=94, there is a corresponding connecti=
ve related to =93gu'e=A0<dots>=85</dots>gi=94 in a systematic way.
+<p>
+<cx "forethought logical connectives in tanru, effect on tanru grouping"> =
 XE "forethought logical connectives in tanru: effect on tanru grouping"   =
The portion of a =93gu'e=A0<dots>=85</dots>gi=94 construction before the =
=93gi=94 is a full selbri, and may use any of the selbri resources includin=
g =93je=94 logical connections. After the =93gi=94, logical connections are=
 taken to be wider in scope than the =93gu'e=A0<dots>=85</dots>gi=94, which=
 has in effect the same scope as =93bo=94:
+<p>
+<pre><a name=3De6d24>6.24)	gu'e barda je xunre gi gerku ja mlatu
+	(both (big and red) and dog) or cat
+	something which is either big, red, and a dog, or else a cat
+</pre>leaves =93mlatu=94 outside the =93gu'e =85gi=94 construction. The sc=
ope of the =93gi=94 arm extends only to a single brivla or to two or more b=
rivla connected with =93bo=94 or =93ke =85ke'e=94.
+<p>
+<a name=3Ds7><h3>Linked sumti: =93be=96bei=96be'o=94</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	be	BE	linked sumti marker
+	bei	BEI	linked sumti separator
+	be'o	BEhO	linked sumti terminator
+</pre><p>The question of the place structures of selbri has been glossed o=
ver so far. This chapter does not attempt to treat place structure issues i=
n detail; they are discussed in <a href=3Dchap9.html>Chapter 9</a>. One gra=
mmatical structure related to places belongs here, however. In simple sente=
nces such as <a href=3D#e1d1>Example 1.1</a>, the place structure of the se=
lbri is simply the defined place structure of the gismu =93mamta=94. What a=
bout more complex selbri?
+<p>
+<cx "tanru, place structures of">  XE "tanru: place structures of"   For t=
anru, the place structure rule is simple: the place structure of a tanru is=
 always the place structure of its tertau. Thus, the place structure of =93=
blanu zdani=94 is that of =93zdani=94: the x1 place is a house or nest, and=
 the x2 place is its occupants.
+<p>
+What about the places of =93blanu=94? Is there any way to get them into th=
e act? In fact, =93blanu=94 has only one place, and this is merged, as it w=
ere, with the x1 place of =93zdani=94. It is whatever is in the x1 place th=
at is being characterized as blue-for-a-house. But if we replace =93blanu=
=94 with =93xamgu=94, we get:
+<p>
+<ex "good house">  XE "good house: example=93  <pre><a name=3De7d1>7.1)	ti=
 xamgu zdani
+	This is-a-good house.
+	This is a good (for someone, by some standard) house.
+</pre><p>Since =93xamgu=94 has three places (x1, the good thing; x2, the p=
erson for whom it is good; and x3, the standard of goodness), <a href=3D#e7=
d1>Example 7.1 </a>necessarily omits information about the last two: there =
is no room for them. Room can be made, however!
+<p>
+<pre><a name=3De7d2>7.2)	ti xamgu be do bei mi [be'o] zdani
+	This is-a-good (for you by-standard me) house.
+	This is a house that is good for you by my standards.
+</pre><lx "be">  XE "be"   <lx "BE">  XE "BE selma=92o"   <lx "bei">  XE "=
bei"   <lx "BEI">  XE "BEI selma=92o"   <lx "be'o">  XE "be'o"   <lx "BEhO"=
>  XE "BEhO selma=92o"   <cx "seltau, filling sumti places in">  XE "seltau=
: filling sumti places in"   <cx "linked sumti, in tanru">  XE "linked sumt=
i: in tanru"   Here, the gismu =93xamgu=94 has been followed by the cmavo =
=93be=94 (of selma'o BE), which signals that one or more sumti follows. The=
se sumti are not part of the overall bridi place structure, but fill the pl=
aces of the brivla they are attached to, starting with x2. If there is more=
 than one sumti, they are separated by the cmavo =93bei=94 (of selma'o BEI)=
, and the list of sumti is terminated by the elidable terminator =93be'o=94=
 (of selma'o BEhO).
+<p>
+<cx "linked sumti, definition">  XE "linked sumti: definition"   Grammatic=
ally, a brivla with sumti linked to it in this fashion plays the same role =
in tanru as a simple brivla. To illustrate, here is a fully fleshed-out ver=
sion of <a href=3D#e3d4>Example 3.4</a>, with all places filled in:
+<p>
+<ex "Brooklyn">  XE "Brooklyn: example=93  <pre><a name=3De7d3>7.3)	ti cma=
lu be le ka canlu
+			bei lo'e ckule be'o
+		nixli be li mu bei lo merko be'o bo
+			ckule la bryklyn. loi pemci
+				le mela nu,IORK. prenu
+le jecta
+	This is a small (in-dimension the property-of volume
+			by-standard the-typical school)
+		(girl (of-years the-number five by-standard some American-thing)
+			school) in-Brooklyn with-subject poems=20
+				for-audience New-York persons=20
+with-operator the state.
+	This is a school, small in volume compared to the typical school, pertain=
ing to five-year-old girls (by American standards), in Brooklyn, teaching p=
oetry to the New York community and operated by the state.
+</pre><p>Here the three places of =93cmalu=94, the three of =93nixli=94, a=
nd the four of =93ckule=94 are fully specified. Since the places of =93ckul=
e=94 are the places of the bridi as a whole, it was not necessary to link t=
he sumti which follow =93ckule=94. It would have been legal to do so, howev=
er:
+<p>
+<pre><a name=3De7d4>7.4)	mi klama be le zarci bei le zdani [be'o]
+	I go (to-the market from-the house).
+</pre>means the same as
+<p>
+<pre><a name=3De7d5>7.5)	mi klama le zarci le zdani
+	I go to-the market from-the house.
+</pre><cx "tanru, place structures of">  XE "tanru: place structures of"  =
 No matter how complex a tanru gets, the last brivla always dictates the pl=
ace structure: the place structure of
+<p>
+<pre><a name=3De7d6>7.6)	melbi je cmalu nixli bo ckule
+	a (pretty and little) (girl school)
+	a school for girls which is both beautiful and small
+</pre>is simply that of =93ckule=94. (The sole exception to this rule is d=
iscussed in <a href=3D#s8>Section 8</a>.)
+<p>
+<lx "FA">  XE "FA selma=92o"   <cx "linked sumti and FA tags">  XE "linked=
 sumti and FA tags"   <cx "FA tags and linked sumti">  XE "FA tags and link=
ed sumti"   It is possible to precede linked sumti by the place structure o=
rdering tags =93fe=94, =93fi=94, =93fo=94, and =93fu=94 (of selma'o FA, dis=
cussed further in <a href=3Dchap9.html>Chapter 9</a>), which serve to expli=
citly specify the x2, x3, x4, and x5 places respectively. Normally, the pla=
ce following the =93be=94 is the x2 place and the other places follow in or=
der. If it seems convenient to change the order, however, it can be accompl=
ished as follows:
+<p>
+<pre><a name=3De7d7>7.7)	ti xamgu be fi mi bei fe do [be'o] zdani
+	This is-a-good (by-standard me for you) house
+</pre>which is equivalent in meaning to <a href=3D#e7d2>Example 7.2</a>. N=
ote that the order of =93be=94, =93bei=94, and =93be'o=94 does not change; =
only the inserted =93fi=94 tells us that =93mi=94 is the x3 place (and corr=
espondingly, the inserted =93fe=94 tells us that =93do=94 is the x2 place).=
 Changing the order of sumti is often done to match the order of another la=
nguage, or for emphasis or rhythm.
+<p>
+Of course, using FA cmavo makes it easy to specify one place while omittin=
g a previous place:
+<p>
+<pre><a name=3De7d8>7.8)	ti xamgu be fi mi [be'o] zdani
+	This is-a-good (by-standard me) house
+	This is a good house by my standards.
+</pre><cx "linked sumti and sumti tcita">  XE "linked sumti and sumti tcit=
a"   <cx "sumti tcita and linked sumti">  XE "sumti tcita and linked sumti"=
   <cx "modal tags and sumti tcita">  XE "modal tag: and sumti tcita"   <cx=
 "sumti tcita and modal tags">  XE "sumti tcita and modal tags"   <cx "tens=
e tags and sumti tcita">  XE "tense tags and sumti tcita"   <cx "sumti tcit=
a and tense tags">  XE "sumti tcita and tense tags"   Similarly, sumti labe=
led by modal or tense tags can be inserted into strings of linked sumti jus=
t as they can into bridi:
+<p>
+<pre><a name=3De7d9>7.9)	ta blanu be ga'a mi [be'o] zdani
+	That is-a-blue (to-observer me) house.
+	That is a blue, as I see it, house.
+</pre><p>The meaning of <a href=3D#e7d9>Example 7.9 </a>is slightly differ=
ent from:
+<p>
+<pre><a name=3De7d10>7.10)	ta blanu zdani ga'a mi
+	That is-a-blue house to-observer me.
+	That is a blue house, as I see it.
+</pre><p>See discussions in <a href=3Dchap9.html>Chapter 9 </a>of modals a=
nd in <a href=3Dchap10.html>Chapter 10 </a>of tenses for more explanations.
+<p>
+<cx "elidability of =93be'o=94">  XE "elidability of be'o"   <cx "be'o, el=
idability of">  XE "be'o: elidability of"   <cx "relative clauses, effect o=
n elidability of be'o">  XE =93relative clause: effect on elidability of be=
'o"   <cx "be'o, effect of relative clauses on elidability of">  XE "be'o: =
effect of relative clauses on elidability of"   The terminator =93be'o=94 i=
s almost always elidable: however, if the selbri belongs to a description, =
then a relative clause following it will attach to the last linked sumti un=
less =93be'o=94 is used, in which case it will attach to the outer descript=
ion:
+<p>
+<pre><a name=3De7d11>7.11)	le xamgu be do noi barda cu zdani
+	The good-thing for you (who are-large) is-a-house.
+<a name=3De7d12>7.12)	le xamgu be do be'o noi barda cu zdani
+	The (good-thing for you) (which is-large) is-a-house
+</pre>(Relative clauses are explained in <a href=3Dchap8.html>Chapter 8</a=
>.)
+<p>
+<cx "ku, effect on elidability of be'o"><cx "be'o, effect of ku on elidabi=
lity of">  XE "be'o: effect of ku on elidability of"   In other cases, howe=
ver, =93be'o=94 cannot be elided if =93ku=94 has also been elided:
+<p>
+<pre><a name=3De7d13>7.13)	le xamgu be le ctuca [ku] be'o zdani
+	the good (for the teacher ) house
+</pre>requires either =93ku=94 or =93be'o=94, and since there is only one =
occurrence of =93be=94, the =93be'o=94 must match it, whereas it may be con=
fusing which occurrence of =93le=94 the =93ku=94 terminates (in fact the se=
cond one is correct).
+
+<p>
+<a name=3Ds8><h3>Inversion of tanru: =93co=94</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	co	CO	tanru inversion marker
+</pre><cx "tanru inversion">  XE "tanru inversion"   <lx "co">  XE "co"   =
The standard order of Lojban tanru, whereby the modifier precedes what it m=
odifies, is very natural to English-speakers: we talk of =93blue houses=94,=
 not of =93houses blue=94. In other languages, however, such matters are di=
fferently arranged, and Lojban supports this reverse order (tertau before s=
eltau) by inserting the particle =93co=94. <a href=3D#e8d1>Example 8.1 </a>=
and <a href=3D#e8d2>Example 8.2 </a>mean exactly the same thing:
+<p>
+<pre><a name=3De8d1>8.1)	ta blanu zdani
+	That is-a-blue type-of-house.
+	That is a blue house.
+<a name=3De8d2>8.2)	ta zdani co blanu
+	That is-a-house of-type blue.
+	That is a blue house.
+</pre><cx "tanru inversion, definition">  XE "tanru inversion: definition"=
   <cx "tertau, definition">  XE "tertau: definition"   <cx "seltau, defini=
tion">  XE "seltau: definition"   This change is called =93tanru inversion=
=94. In tanru inversion, the element before =93co=94 (=93zdani=94 in <a hre=
f=3D#e8d2>Example 8.2</a>) is the tertau, and the element following =93co=
=94 (=93blanu=94) in <a href=3D#e8d2>Example 8.2</a>) is the seltau.
+<p>
+<cx "tanru inversion and place structure">  XE "tanru inversion and place =
structure"   <cx "place structure and tanru inversion">  XE "place structur=
e and tanru inversion"   The meaning, and more specifically, the place stru=
cture, of a tanru is not affected by inversion: the place structure of =93z=
dani co blanu=94 is still that of =93zdani=94. However, the existence of in=
version in a selbri has a very special effect on any sumti which follow tha=
t selbri. Instead of being interpreted as filling places of the selbri, the=
y actually fill the places (starting with x2) of the seltau. In <a href=3D#=
s7>Section 7</a>, we saw how to fill interior places with =93be=A0<dots>=85=
</dots>bei=A0<dots>=85</dots>be'o=94, and in fact <a href=3D#e8d3>Example 8=
.3 </a>and <a href=3D#e8d4>Example 8.4 </a>have the same meaning:
+<p>
+<pre><a name=3De8d3>8.3)	mi klama be le zarci bei le zdani be'o troci
+	I am-a-(goer to the market from the house) type-of trier.
+	I try to go to the market from the house.
+<a name=3De8d4>8.4)	mi troci co klama le zarci le zdani
+	I am-a-trier of-type (goer to-the market from-the house).
+	I try to go to the market from the house.
+</pre><a href=3D#e8d4>Example 8.4 </a>is a less deeply nested construction=
, requiring fewer cmavo. As a result it is probably easier to understand.
+<p>
+<ex "try to go">  XE "try to go: example=93  Note that in Lojban =93trying=
 to go=94 is expressed using =93troci=94 as the tertau. The reason is that =
=93trying to go=94 is a =93going type of trying=94, not a =93trying type of=
 going=94. The trying is more fundamental than the going =97 if the trying =
fails, we may not have a going at all.
+<p>
+<cx "unfilled places of inverted tanru">  XE "unfilled places of inverted =
tanru"   <cx "inverted tanru, effect on sumti before the selbri">  XE "inve=
rted tanru: effect on sumti before the selbri"   <cx "inverted tanru, effec=
t on sumti after the selbri">  XE "inverted tanru: effect on sumti after th=
e selbri"   Any sumti which precede a selbri with an inverted tanru fill th=
e places of the selbri (i.e., the places of the tertau) in the ordinary way=
. In <a href=3D#e8d4>Example 8.4</a>, =93mi=94 fills the x1 place of =93tro=
ci co klama=94, which is the x1 place of =93troci=94. The other places of t=
he selbri remain unfilled. The trailing sumti =93le zarci=94 and =93le zdan=
i=94 do not occupy selbri places, despite appearances.
+<p>
+As a result, the regular mechanisms (involving selma'o VOhA and GOhI, expl=
ained in <a href=3Dchap7.html>Chapter 7</a>) for referring to individual su=
mti of a bridi cannot refer to any of the trailing places of <a href=3D#e8d=
4>Example 8.4</a>, because they are not really =93sumti of the bridi=94 at =
all.
+<p>
+<cx "tanru inversion, in complex tanru">  XE "tanru inversion: in complex =
tanru"   <cx "tanru inversion, where allowed">  XE "tanru inversion: where =
allowed"   When inverting a more complex tanru, it is possible to invert it=
 only at the most general modifier-modified pair. The only possible inversi=
on of <a href=3D#e3d4>Example 3.4</a>, for instance, is:
+<p>
+<pre><a name=3De8d5>8.5)	ta nixli [bo] ckule co cmalu
+	That (is-a-girl type-of school) of-type little.
+	That's a girls' school which is small.
+</pre><cx "tanru inversion, effect on tanru grouping">  XE "tanru inversio=
n: effect on tanru grouping"   <cx "tanru grouping, effect of tanru inversi=
on on">  XE "tanru grouping: effect of tanru inversion on"   Note that the =
=93bo=94 of <a href=3D#e3d4>Example 3.4 </a>is optional in <a href=3D#e8d5>=
Example 8.5</a>, because =93co=94 groups more loosely than any other cmavo =
used in tanru, including none at all. Not even =93ke=A0<dots>=85</dots>ke'e=
=94 parentheses can encompass a =93co=94:
+<p>
+<pre><a name=3De8d6>8.6)	ta cmalu ke nixli ckule [ke'e] co melbi
+	That is-a-(little type-of (girl type-of school)) of-type pretty.
+	That's a small school for girls which is beautiful.
+</pre><cx "tanru inversion, rule for removing">  XE "tanru inversion: rule=
 for removing"   In <a href=3D#e8d6>Example 8.6</a>, the =93ke'e=94 is auto=
matically inserted before the =93co=94 rather than at its usual place at th=
e end of the selbri. As a result, there is a simple and mechanical rule for=
 removing =93co=94 from any selbri: change =93A co B=94 to =93ke B ke'e A=
=94. (At the same time, any sumti following the selbri must be transformed =
into =93be=A0<dots>=85</dots>bei=A0<dots>=85</dots>be'o=94 form and attache=
d following B.) Therefore,
+<p>
+<pre><a name=3De8d7>8.7)	ckule co melbi nixli
+	school of-type pretty girl
+	school for beautiful girls
+</pre>means the same as:
+<p>
+<pre><a name=3De8d8>8.8)	ke melbi nixli ke'e ckule
+	(pretty girl) school
+</pre><cx "tanru inversion, multiple">  XE "tanru inversion: multiple"   <=
cx "multiple tanru inversion, effect on grouping">  XE "multiple tanru inve=
rsion: effect on grouping"   Multiple =93co=94 cmavo can appear within a se=
lbri, indicating multiple inversions: a right-grouping rule is employed, as=
 for =93bo=94. The above rule can be applied to interpret such selbri, but =
all =93co=94 cmavo must be removed simultaneously:
+<p>
+<pre><a name=3De8d9>8.9)	ckule co nixli co cmalu
+	school of-type (girl of-type little)
+</pre>becomes formally
+<p>
+<pre><a name=3De8d10>8.10)	ke ke cmalu ke'e nixli ke'e ckule
+	( (little) girl ) school
+</pre>which by the left-grouping rule is simply
+<p>
+<pre><a name=3De8d11>8.11)	cmalu nixli ckule
+	little girl school
+	school for little girls
+</pre><p>As stated above, the selbri places, other than the first, of
+<p>
+<pre><a name=3De8d12>8.12)	mi klama co sutra
+	I am-a-goer of-type quick
+	I go quickly
+</pre>cannot be filled by placing sumti after the selbri, because any sumt=
i in that position fill the places of =93sutra=94, the seltau. However, the=
 tertau places (which means in effect the selbri places) can be filled with=
 =93be=94:
+<p>
+<pre><a name=3De8d13>8.13)	mi klama be le zarci co sutra
+	I am-a-goer (to the store) of-type quick.
+	I go to the store quickly.
+</pre><a name=3Ds9><h3>Other kinds of simple selbri</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	go'i	GOhA	repeats the previous bridi
+	du	GOhA	equality
+	nu'a	NUhA	math operator to selbri
+	moi	MOI	changes number to ordinal selbri
+	mei	MOI	changes number to cardinal selbri
+	nu	NU	event abstraction
+	kei	KEI	terminator for =93nu=94
+</pre><cx "brivla equivalents">  XE "brivla equivalents"   <cx "equivalent=
s to brivla">  XE "equivalents to brivla"   So far we have only discussed b=
rivla and tanru built up from brivla as possible selbri. In fact, there are=
 a few other constructions in Lojban which are grammatically equivalent to =
brivla: they can be used either directly as selbri, or as components in tan=
ru. Some of these types of simple selbri are discussed at length in <a href=
=3Dchap7.html>Chapter 7</a>, <a href=3Dchap11.html>Chapter 11</a>, and <a h=
ref=3Dchap18.html>Chapter 18</a>; but for completeness these types are ment=
ioned here with a brief explanation and an example of their use in selbri.
+<p>
+<lx "GOhA">  XE "GOhA selma=92o"   <cx "GOhA, as selbri">  XE "GOhA selma=
=92o: as selbri"   <cx "GOhA, as component in tanru">  XE "GOhA selma=92o: =
as component in tanru"   <cx "selbri, with GOhA">  XE "selbri: with GOhA"  =
 <cx "tanru, with GOhA">  XE "tanru: with GOhA"   The cmavo of selma'o GOhA=
 (with one exception) serve as pro-bridi, providing a reference to the cont=
ent of other bridi; none of them has a fixed meaning. The most commonly use=
d member of GOhA is probably =93go'i=94, which amounts to a repetition of t=
he previous bridi, or part of it. If I say:
+<p>
+<lx "go'i">  XE "go'i"   <pre><a name=3De9d1>9.1)	la djan. klama le zarci
+	John goes-to the market.
+</pre>you may retort:
+<p>
+<pre><a name=3De9d2>9.2)	la djan. go'i troci
+	John [repeat last] are-a-tryer
+	John tries to.
+</pre><a href=3D#e9d2>Example 9.2 </a>is short for:
+<p>
+<pre><a name=3De9d3>9.3)	la djan. klama be le zarci be'o troci
+	John is-a-goer (to the market) type-of trier.
+</pre>because the whole bridi of <a href=3D#e9d1>Example 9.1 </a>has been =
packaged up into the single word =93go'i=94 and inserted into <a href=3D#e9=
d2>Example 9.2</a>.
+<p>
+<cx "du, as an exception in GOhA">  XE "du: as an exception within GOhA se=
lma=92o"   The exceptional member of GOhA is =93du=94, which represents the=
 relation of identity. Its place structure is:
+<p>
+<dl compact><dt>du:  <dd>x1 is identical with x2, x3,=A0<dots>=85</dots>
+</dl>for as many places as are given. More information on selma'o GOhA is =
available in <a href=3Dchap7.html>Chapter 7</a>.
+<p>
+<cx "mathematical expressions in tanru">  XE "mathematical expressions in =
tanru"   <lx "nu'a">  XE "nu'a"   <lx "su'i">  XE "su'i"   <cx "tanru, cont=
aining mathematical expressions">  XE "tanru: containing mathematical expre=
ssions"   Lojban mathematical expressions (mekso) can be incorporated into =
selbri in two different ways. Mathematical operators such as =93su'i=94, me=
aning =93plus=94, can be transformed into selbri by prefixing them with =93=
nu'a=94 (of selma'o NUhA). The resulting place structure is:
+<p>
+<dl compact><dt><dd>x1 is the result of applying (the operator) to argumen=
ts x2, x3, etc.
+</dl>for as many arguments as are required. (The result goes in the x1 pla=
ce because the number of following places may be indefinite.) For example:
+<p>
+<pre><a name=3De9d4>9.4)	li vo nu'a su'i li re li re
+	The-number 4 is-the-sum-of the-number 2 and-the-number 2.
+</pre><p>A possible tanru example might be:
+<p>
+<ex "addition problems">  XE "addition problems: example=93  <pre><a name=
=3De9d5>9.5)	mi jimpe tu=92a loi nu'a su'i nabmi
+	I understand something about the-mass-of is-the-sum-of problems.
+	I understand addition problems.
+</pre><lx "MOI">  XE "MOI selma=92o"   More usefully, it is possible to co=
mbine a mathematical expression with a cmavo of selma'o MOI to create one o=
f various numerical selbri. Details are available in <a href=3Dchap18.html>=
Chapter 18</a>. Here are a few tanru:
+<p>
+<ex "Preem Palver">  XE "Preem Palver: example=93  <pre><a name=3De9d6>9.6=
)	la prim. palvr. pamoi cusku
+	Preem Palver is-the-1-th speaker.
+	Preem Palver is the first speaker.
+</pre><ex "two brothers">  XE "two brothers: example=93  <pre><a name=3De9=
d7>9.7)	la an,iis. joi la .asun. bruna remei
+	Anyi massed-with Asun are-a-brother type-of-twosome.
+	Anyi and Asun are two brothers.
+</pre><lx "NU">  XE "NU selma=92o"   <lx "KEI">  XE "KEI selma=92o"   Fina=
lly, an important type of simple selbri which is not a brivla is the abstra=
ction. Grammatically, abstractions are simple: a cmavo of selma'o NU, follo=
wed by a bridi, followed by the elidable terminator =93kei=94 of selma'o KE=
I. Semantically, abstractions are an extremely subtle and powerful feature =
of Lojban whose full ramifications are documented in <a href=3Dchap11.html>=
Chapter 11</a>. For example:
+<p>
+<cx "abstraction bridi, contrasted with component non-abstraction bridi in=
 meaning">  XE "abstraction bridi: contrasted with component non-abstractio=
n bridi in meaning"   <pre><a name=3De9d8>9.8)	ti nu zdile kei kumfa
+	This is-an-event-of amusement room.
+	This is an amusement room.
+</pre><a href=3D#e9d8>Example 9.8 </a>is quite distinct in meaning from:
+<p>
+<pre><a name=3De9d9>9.9)	ti zdile kumfa
+	This is-an-amuser room.
+</pre>which suggests the meaning =93a room that amuses someone=94.
+<p>
+<a name=3Ds10><h3>selbri based on sumti: =93me=94</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	me	ME	changes sumti to simple selbri
+	me'u	MEhU	terminator for =93me=94
+</pre><lx "me">  XE "me"   <lx "ME">  XE "ME selma=92o"   <lx "me'u">  XE =
"me'u"   <lx "MEhU">  XE "MEhU selma=92o"   <cx "selbri from sumti">  XE "s=
elbri from sumti"   <cx "sumti into selbri">  XE "sumti into selbri"   <cx =
"conversion, sumti into selbri">  XE "conversion of sumti into selbri"   <c=
x "me, place structure of">  XE "me: place structure of"   A sumti can be m=
ade into a simple selbri by preceding it with =93me=94 (of selma'o ME) and =
following it with the elidable terminator =93me'u=94 (of selma'o MEhU). Thi=
s makes a selbri with the place structure
+<p>
+<dl compact><dt><dd>x1 is one of the referents of =93[the sumti]=94
+</dl>which is true of the thing, or things, that are the referents of the =
sumti, and not of anything else. For example, consider the sumti
+<p>
+<pre><a name=3De10d1>10.1)	le ci nolraitru
+	the three noblest-governors
+	the three kings
+</pre><ex "Three Kings">  XE "Three Kings: example=93  If these are unders=
tood to be the Three Kings of Christian tradition, who arrive every year on=
 January 6, then we may say:
+<p>
+<pre><a name=3De10d2>10.2)	la BALtazar. cu me le ci nolraitru
+	Balthazar is one-of-the-referents-of =93the three kings=94.
+	Balthazar is one of the three kings.
+</pre>and likewise
+<p>
+<pre><a name=3De10d3>10.3)	la kaspar. cu me le ci nolraitru
+	Caspar is one of the three kings.
+</pre>and
+<p>
+<pre><a name=3De10d4>10.4)	la melxi,or. cu me le ci nolraitru
+	Melchior is one of the three kings.
+</pre><cx "=93me=94/=93du=94 equivalence">  XE "me/du equivalence"   <cx "=
me, compared with du in effect">  XE "me: compared with du in effect"   <cx=
 "du, compared with me in effect">  XE "du: compared with me in effect"   I=
f the sumti refers to a single object, then the effect of =93me=94 is much =
like that of =93du=94:
+<p>
+<pre><a name=3De10d5>10.5)	do du la djan.
+	You are-identical-with the-one-called =93John=94.
+	You are John.
+</pre>means the same as
+<p>
+<pre><a name=3De10d6>10.6)	do me la djan.
+	You are-the-referent-of =93the-one-called =91John=92=94.
+	You are John.
+</pre><cx "=93me=94 and names">  XE "me: used with names"   It is common t=
o use =93me=94 selbri, especially those based on name sumti using =93la=94,=
 as seltau. For example:
+<p>
+<ex "Chrysler">  XE "Chrysler: example=93  <pre><a name=3De10d7>10.7)	ta m=
e lai kraislr. [me'u] karce
+	That (is-a-referent of =93the-mass-called =91Chrysler=92=94) car.
+	That is a Chrysler car.
+</pre><cx "elidability of =93me'u=94">  XE "elidability of me'u"   <cx "me=
'u, relative precedence with logical connectives">  XE "me'u: relative prec=
edence with logical connectives"   <cx "logical connectives, relative prece=
dence with me'u">  XE "logical connectives: relative precedence with me'u" =
  The elidable terminator =93me'u=94 can usually be omitted. It is absolute=
ly required only if the =93me=94 selbri is being used in an indefinite desc=
ription (a type of sumti explained in <a href=3Dchap6.html>Chapter 6</a>), =
and if the indefinite description is followed by a relative clause (explain=
ed in <a href=3Dchap8.html>Chapter 8</a>) or a sumti logical connective (ex=
plained in <a href=3Dchap14.html>Chapter 14</a>). Without a =93me'u=94, the=
 relative clause or logical connective would appear to belong to the sumti =
embedded in the =93me=94 expression. Here is a contrasting pair of sentence=
s:
+<p>
+<pre><a name=3De10d8>10.8)	re me le ci nolraitru .e la djan. [me'u] cu bla=
bi
+	Two of the group =93the three kings and John=94 are white.
+<a name=3De10d9>10.9)	re me le ci nolraitru me'u .e la djan. cu blabi
+	Two of the three kings, and John, are white.
+</pre><p>In <a href=3D#e10d8>Example 10.8 </a>the =93me=94 selbri covers t=
he three kings plus John, and the indefinite description picks out two of t=
hem that are said to be white: we cannot say which two. In <a href=3D#e10d9=
>Example 10.9</a>, though, the =93me=94 selbri covers only the three kings:=
 two of them are said to be white, and so is John.
+<p>
+Finally, here is another example requiring =93me'u=94:
+<p>
+<pre><a name=3De10d10>10.10)	ta me la'e le se cusku be do me'u cukta
+	That is-a-(what-you-said) type of book.
+	That is the kind of book you were talking about.
+</pre><p>There are other sentences where either =93me'u=94 or some other e=
lidable terminator must be expressed:
+<p>
+<pre><a name=3De10d11>10.11)	le me le ci nolraitru [ku] me'u nunsalci
+	the (the three kings) type-of-event-of-celebrating
+	the Three Kings celebration
+</pre>requires either =93ku=94 or =93me'u=94 to be explicit, and (as with =
=93be'o=94 in <a href=3D#s7>Section 7</a>) the =93me'u=94 leaves no doubt w=
hich cmavo it is paired with.
+<p>
+<a name=3Ds11><h3>Conversion of simple selbri</h3>
+<p>
+<lx "SE">  XE "SE selma=92o"   <lx "se">  XE "se"   <lx "te">  XE "te"   <=
lx "ve">  XE "ve"   <lx "xe">  XE "xe"   <cx "conversion and tanru">  XE "c=
onversion and tanru"   <cx "tanru and conversion">  XE "tanru and conversio=
n"   <cx "conversion, definition">  XE "conversion: definition"   <cx "plac=
e structure, re-ordering by conversion">  XE "place structure: re-ordering =
by conversion"   Conversion is the process of changing a selbri so that its=
 places appear in a different order. This is not the same as labeling the s=
umti with the cmavo of FA, as mentioned in <a href=3D#s7>Section 7</a>, and=
 then rearranging the order in which the sumti are spoken or written. Conve=
rsion transforms the selbri into a distinct, though closely related, selbri=
 with renumbered places.
+<p>
+In Lojban, conversion is accomplished by placing a cmavo of selma'o SE bef=
ore the selbri:
+<p>
+<pre><a name=3De11d1>11.1)	mi prami do
+	I love you.
+</pre>is equivalent in meaning to:
+<p>
+<pre><a name=3De11d2>11.2)	do se prami mi
+	You [swap x1 and x2] love me.
+	You are loved by me.
+</pre><p>Conversion is fully explained in <a href=3Dchap9.html>Chapter 9</=
a>. For the purposes of this chapter, the important point about conversion =
is that it applies only to the following simple selbri. When trying to conv=
ert a tanru, therefore, it is necessary to be careful! Consider <a href=3D#=
e11d3>Example 11.3</a>:
+<p>
+<ex "walk to market">  XE "walk to market: example=93  <pre><a name=3De11d=
3>11.3)	la .alis. cu cadzu klama le zarci
+	Alice is-a-walker type-of goer to-the market.
+	Alice walkingly goes to the market.
+	Alice walks to the market.
+</pre><p>To convert this sentence so that =93le zarci=94 is in the x1 plac=
e, one correct way is:
+<p>
+<pre><a name=3De11d4>11.4)	le zarci cu se ke cadzu klama [ke'e] la .alis.
+	The market is-a-[swap x1/x2] (walker type-of goer) Alice.
+	The market is-walkingly gone-to by-Alice.
+</pre><cx "conversion with `ke'">  XE "conversion with `ke'"   <cx "ke, fo=
r conversion of tanru">  XE "ke: for conversion of tanru"   The =93ke=A0<do=
ts>=85</dots>ke'e=94 brackets cause the entire tanru to be converted by the=
 =93se=94, which would otherwise convert only =93cadzu=94, leading to:
+<p>
+<pre><a name=3De11d5>11.5)	le zarci cu se cadzu klama la .alis.
+	The market (is-a-[swap x1/x2] walker) type-of goer to Alice.
+	The market is-a-walking-surface type-of goer to Alice.
+</pre>whatever that might mean. An alternative approach, since the place s=
tructure of =93cadzu klama=94 is that of =93klama=94 alone, is to convert o=
nly the latter:
+<p>
+<pre><a name=3De11d6>11.6)	le zarci cu cadzu se klama la .alis.
+	The market walkingly is-gone-to by-Alice.
+</pre><p>But the tanru in <a href=3D#e11d6>Example 11.6 </a>may or may not=
 have the same meaning as that in <a href=3D#e11d3>Example 11.3</a>; in par=
ticular, because =93cadzu=94 is not converted, there is a suggestion that a=
lthough Alice is the goer, the market is the walker. With a different sumti=
 as x1, this seemingly odd interpretation might make considerable sense:
+<p>
+<pre><a name=3De11d7>11.7)	la djan. cu cadzu se klama la .alis
+	John walkingly is-gone-to by Alice
+</pre>suggests that Alice is going to John, who is a moving target.
+<p>
+<cx "jai, for modal conversion">  XE "jai: conversion using"   <cx "tense =
conversion, with jai">  XE "conversion: using jai"   There is an alternativ=
e type of conversion, using the cmavo =93jai=94 of selma'o JAI optionally f=
ollowed by a modal or tense construction. Grammatically, such a combination=
 behaves exactly like conversion using SE. More details can be found in <a =
href=3Dchap9.html>Chapter 9</a>.
+<p>
+<a name=3Ds12><h3>Scalar negation of selbri</h3>
+<p>
+<lx "NAhE">  XE "NAhE selma=92o"   <cx "scalar negation, effect on selbri"=
>  XE "scalar negation: effect on selbri"   <cx "selbri, scalar negation of=
">  XE "selbri: scalar negation of"   Negation is too large and complex a t=
opic to explain fully in this chapter; see <a href=3Dchap15.html>Chapter 15=
</a>. In brief, there are two main types of negation in Lojban. This sectio=
n is concerned with so-called =93scalar negation=94, which is used to state=
 that a true relation between the sumti is something other than what the se=
lbri specifies. Scalar negation is expressed by cmavo of selma'o NAhE:
+<p>
+<cx "ke, for expanding scope of scalar negation">  XE "ke: for expanding s=
cope of scalar negation"   <pre><a name=3De12d1>12.1)	la .alis. cu na'e ke =
cadzu klama [ke'e] le zarci
+	Alice non- (walkingly goes) to-the market.
+	Alice other-than (walkingly goes) to-the market.
+	Alice doesn't walk to the market.
+</pre>meaning that Alice's relationship to the market is something other t=
han that of walking there. But if the =93ke=94 were omitted, the result wou=
ld be:
+<p>
+<pre><a name=3De12d2>12.2)	la .alis. cu na'e cadzu klama le zarci
+	Alice non- walkingly goes to-the market.
+	Alice doesn't walk to the market.
+</pre>meaning that Alice does go there in some way (=93klama=94 is not neg=
ated), but by a means other than that of walking. <a href=3D#e12d1>Example =
12.1 </a>negates both =93cadzu=94 and =93klama=94, suggesting that Alice's =
relation to the market is something different from walkingly-going; it migh=
t be walking without going, or going without walking, or neither.
+<p>
+Of course, any of the simple selbri types explained in <a href=3D#s9>Secti=
on 9 </a>may be used in place of brivla in any of these examples:
+
+<p>
+<pre><a name=3De12d3>12.3)	la djonz. cu na'e pamoi cusku
+	Jones is non-1st speaker
+	Jones is not the first speaker.
+</pre><p>Since only =93pamoi=94 is negated, an appropriate inference is th=
at he is some other kind of speaker.
+<p>
+<cx "complex negation, examples">  XE "complex negation: examples"   <cx "=
negation, complex examples">  XE "negation: complex examples"   Here is an =
assortment of more complex examples showing the interaction of scalar negat=
ion with =93bo=94 grouping, =93ke=94 and =93ke'e=94 grouping, logical conne=
ction, and sumti linked with =93be=94 and =93bei=94:
+<p>
+<cx "na'e, contrasted with na'e ke">  XE "na'e: contrasted with na'e ke"  =
 <pre><a name=3De12d4>12.4)	mi na'e sutra cadzu be fi le birka be'o klama l=
e zarci
+	I ( (non-quickly) (walking using the arms) ) go-to the market.
+	I go to the market, walking using my arms other than quickly.
+</pre><p>In <a href=3D#e12d4>Example 12.4</a>, =93na'e=94 negates only =93=
sutra=94. Contrast <a href=3D#e12d5>Example 12.5</a>:
+<p>
+<pre><a name=3De12d5>12.5)	mi na'e ke sutra cadzu be fi le birka [be'o] ke=
'e klama le zarci
+	I non-( quickly (walking using the arms) ) go-to the market.
+	I go to the market, other than by walking quickly on my arms.
+</pre><p>Now consider <a href=3D#e12d6>Example 12.6 </a>and <a href=3D#e12=
d7>Example 12.7</a>, which are equivalent in meaning, but use =93ke=94 grou=
ping and =93bo=94 grouping respectively:
+<p>
+<pre><a name=3De12d6>12.6)	mi sutra cadzu be fi le birka be'o je masno kla=
ma le zarci
+	I (quickly =96 (walking using the arms) and slowly) go-to the market.
+	I go to the market, both quickly walking using my arms and slowly.
+<a name=3De12d7>12.7)	mi ke sutra cadzu be fi le birka [be'o] ke'e je masn=
o klama le zarci
+	I ((quickly (walking using the arms)) and slowly) go-to the market.
+	I go to the market, both quickly walking using my arms and slowly.
+</pre><p>However, if we place a =93na'e=94 at the beginning of the selbri =
in both <a href=3D#e12d6>Example 12.6 </a>and <a href=3D#e12d7>Example 12.7=
</a>, we get different results:
+<p>
+<pre><a name=3De12d8>12.8)	mi na'e sutra cadzu be fi le birka be'o je masn=
o klama le zarci
+	I ((non- quickly) =96 (walking using the arms) and slowly) go-to the mark=
et.
+	I go to the market, both walking using my arms other than quickly, and=20
+also slowly.
+<a name=3De12d9>12.9)	mi na'e ke sutra cadzu be fi le birka [be'o] ke'e je=
 masno klama le zarci
+	I (non-(quickly (walking using the arms)) and slowly) go-to the market.
+	I go to the market, both other than quickly walking using my arms, and=20
+also slowly.
+</pre><p>The difference arises because the =93na'e=94 in <a href=3D#e12d9>=
Example 12.9 </a>negates the whole construction from =93ke=94 to =93ke'e=94=
, whereas in <a href=3D#e12d8>Example 12.8 </a>it negates =93sutra=94 alone=
.
+<p>
+<cx "omitting terminators, perils of">  XE "omitting terminators: perils o=
f"   <cx "perils of omitting terminators">  XE "perils of omitting terminat=
ors"   Beware of omitting terminators in these complex examples! If the exp=
licit =93ke'e=94 is left out in <a href=3D#e12d9>Example 12.9</a>, it is tr=
ansformed into:
+<p>
+<pre><a name=3De12d10>12.10)	mi na'e ke sutra cadzu be fi le birka be'o je=
 masno klama [ke'e] le zarci
+	I non-(quickly ((walking using the arms)) and slowly) go-to) the market.
+	I do something other than quickly both going to the market walking=20
+using my arms and slowly going to the market.
+
+</pre><p>And if both =93ke'e=94 and =93be'o=94 are omitted, the results ar=
e even sillier:
+<p>
+<pre><a name=3De12d11>12.11)	mi na'e ke sutra cadzu be fi le birka je masn=
o klama [be=92o] [ke'e] le zarci
+	I non-(quickly walk on my (arm-type and slow) goers) on the market.
+	I do something other than quickly walking using the goers, both arm-type=
=20
+and slow, relative-to the market.
+</pre><p>In <a href=3D#e12d11>Example 12.11</a>, everything after =93be=94=
 is a linked sumti, so the place structure is that of =93cadzu=94, whose x2=
 place is the surface walked upon. It is less than clear what an =93arm-typ=
e goer=94 might be. Furthermore, since the x3 place has been occupied by th=
e linked sumti, the =93le zarci=94 following the selbri falls into the none=
xistent x4 place of =93cadzu=94. As a result, the whole example, though gra=
mmatical, is complete nonsense. (The bracketed Lojban words appear where a =
fluent Lojbanist would understand them to be implied.)
+<p>
+<cx "na'e, before gu'e">  XE "na'e: before gu'e"   Finally, it is also pos=
sible to place =93na'e=94 before a =93gu'e=A0<dots>=85</dots>gi=94 logicall=
y connected tanru construction. The meaning of this usage has not yet been =
firmly established.
+<p>
+<a name=3Ds13><h3>Tenses and bridi negation</h3>
+<p>
+A bridi can have cmavo associated with it which specify the time, place, o=
r mode of action. For example, in
+<p>
+<pre><a name=3De13d1>13.1)	mi pu klama le zarci
+	I [past] go to-the market.
+	I went to the market.
+</pre>the cmavo =93pu=94 specifies that the action of the speaker going to=
 the market takes place in the past. Tenses are explained in full detail in=
 <a href=3Dchap10.html>Chapter 10</a>. Tense is semantically a property of =
the entire bridi; however, the usual syntax for tenses attaches them at the=
 front of the selbri, as in <a href=3D#e13d1>Example 13.1</a>. There are al=
ternative ways of expressing tense information as well. Modals, which are e=
xplained in <a href=3Dchap9.html>Chapter 9</a>, behave in the same way as t=
enses.
+<p>
+Similarly, a bridi may have the particle =93na=94 (of selma'o NA) attached=
 to the beginning of the selbri to negate the bridi. A negated bridi expres=
ses what is false without saying anything about what is true. Do not confus=
e this usage with the scalar negation of <a href=3D#s12>Section 12</a>. For=
 example:
+<p>
+<pre><a name=3De13d2>13.2)	la djonz. na pamoi cusku
+	Jones (Not!) is-the-first speaker
+	It is not true that Jones is the first speaker.
+	Jones isn't the first speaker.
+</pre><p>Jones may be the second speaker, or not a speaker at all; <a href=
=3D#e13d2>Example 13.2 </a>doesn't say. There are other ways of expressing =
bridi negation as well; the topic is explained fully in <a href=3Dchap15.ht=
ml>Chapter 15</a>.
+<p>
+<cx "tense, relative order with bridi negation">  XE "tense: relative orde=
r with bridi negation"   <cx "bridi negation, relative order with tense">  =
XE "bridi negation: relative order with tense"   Various combinations of te=
nse and bridi negation cmavo are permitted. If both are expressed, either o=
rder is permissible with no change in meaning:
+<p>
+<pre><a name=3De13d3>13.3)	mi na pu klama le zarci
+	mi pu na klama le zarci
+	It is false that I went to the market.
+	I didn't go to the market.
+
+</pre><cx "bridi negation, multiple">  XE "bridi negation: multiple"   <lx=
 "na">  XE "na"   It is also possible to have more than one =93na=94, in wh=
ich case pairs of =93na=94 cmavo cancel out:
+<p>
+<pre><a name=3De13d4>13.4)	mi na na klama le zarci
+	It is false that it is false that I go to the market.
+	I go to the market.
+</pre><cx "na and tense, multiple">  XE "na and tense: multiple"   <cx "te=
nse and na, multiple">  XE "tense and na: multiple"   It is even possible, =
though somewhat pointless, to have multiple =93na=94 cmavo and tense cmavo =
mixed together, subject to the limitation that two adjacent tense cmavo wil=
l be understood as a compound tense, and must fit the grammar of tenses as =
explained in <a href=3Dchap10.html>Chapter 10</a>.
+<p>
+<pre><a name=3De13d5>13.5)	mi na pu na ca klama le zarci
+	I [not] [past] [not] [present] go to-the market
+	It is not the case that in the past it was not the case that in the prese=
nt I
+went to the market.
+	I didn't not go to the market.
+	I went to the market.
+</pre><cx "tense cmavo, position relative to selbri">  XE "tense cmavo: po=
sition relative to selbri"   <cx "modal cmavo, position relative to selbri"=
>  XE "modal cmavo: position relative to selbri"   <cx "negation cmavo, pos=
ition relative to selbri">  XE "negation cmavo: position relative to selbri=
"   Tense, modal, and negation cmavo can appear only at the beginning of th=
e selbri. They cannot be embedded within it.
+<p>
+<a name=3Ds14><h3>Some types of asymmetrical tanru</h3>
+<p>
+<cx "asymmetrical tanru">  XE "asymmetrical tanru"   <cx "tanru, asymmetri=
cal">  XE "tanru: asymmetrical"   This section and <a href=3D#s15>Section 1=
5 </a>contain some example tanru classified into groups based on the type o=
f relationship between the modifying seltau and the modified tertau. All th=
e examples are paralleled by compounds actually observed in various natural=
 languages. In the tables which follow, each group is preceded by a brief e=
xplanation of the relationship. The tables themselves contain a tanru, a li=
teral gloss, an indication of the languages which exhibit a compound analog=
ous to this tanru, and (for those tanru with no English parallel) a transla=
tion.
+<p>
+<cx "languages, abbreviations for">  XE "languages: abbreviations for"   H=
ere are the 3-letter abbreviations used for the various languages (it is pr=
esumed to be obvious whether a compound is found in English or not, so Engl=
ish is not explicitly noted):
+<p>
+<dl compact><dt><dd>	Aba =3D Abazin		Kaz =3D Kazakh
+	Chi =3D Chinese		Kor =3D Korean
+	Ewe =3D Ewe		Mon =3D Mongolian
+	Fin =3D Finnish		Qab =3D Qabardian
+	Geo =3D Georgian		Que =3D Quechua
+	Gua =3D Guarani		Rus =3D Russian
+	Hop =3D Hopi		Skt =3D Sanskrit
+	Hun =3D Hungarian	Swe =3D Swedish
+	Imb =3D Imbabura Quechua	Tur =3D Turkish
+	Kar =3D Karaitic		Udm =3D Udmurt=09
+</dl><p>Any lujvo or fu'ivla used in a group are glossed at the end of tha=
t group.
+<p>
+<cx "asymmetrical tanru, definition">  XE "asymmetrical tanru: definition"=
   The tanru discussed in this section are asymmetrical tanru; that is, one=
s in which the order of the terms is fundamental to the meaning of the tanr=
u. For example, =93junla dadylsi=94, or =93clock pendulum=94, is the kind o=
f pendulum used in a clock, whereas =93dadysli junla=94, or =93pendulum clo=
ck=94, is the kind of clock that employs a pendulum. Most tanru are asymmet=
rical in this sense. Symmetrical tanru are discussed in <a href=3D#s15>Sect=
ion 15</a>.
+<p>
+<cx "asymmetrical tanru types, object-of-action + action">  XE "asymmetric=
al tanru types: object-of-action + action"   The tertau represents an actio=
n, and the seltau then represents the object of that action:
+<p>
+<dl compact><dt>	pinsi nunkilbra		<dd>pencil sharpener (Hun)
+	zgike nunctu		music instruction (Hun)=20
+	mirli nunkalte		deer hunting (Hun)=20
+	finpe nunkalte		fish hunting (Tur,Kor,Udm,Aba =3D fishing)=20
+	smacu terkavbu		mousetrap (Tur,Kor,Hun,Udm,Aba)=20
+	zdani turni		house ruler (Kar =3D host)=20
+	zerle'a nunte'a		thief fear (Skt =3D fear of thieves)=20
+	cevni zekri		god crime (Skt =3D offense against the gods)
+	nunkilbra =3D sharpness-apparatus
+	nunctu =3D event-of-teaching
+	nunkalte =3D event-of-hunting
+	terkavbu =3D trap
+	zerle'a =3D crime-taker
+	nunte'a =3D event-of-fearing
+</dl><cx "asymmetrical tanru types, elements-in-set + set">  XE "asymmetri=
cal tanru types: elements-in-set + set"   The tertau represents a set, and =
the seltau the type of the elements contained in that set:
+<p>
+<dl compact><dt>	zdani lijgri		<dd>house row=20
+	selci lamgri		cell block=20
+	karda mulgri		card pack (Swe)=20
+	rokci derxi		stone heap (Swe)=20
+	tadni girzu		student group (Hun)=20
+	remna girzu		human-being group (Qab =3D group of people)=20
+	cpumi'i lijgri		tractor column (Qab)=20
+	cevni jenmi		god army (Skt)=20
+	cevni prenu		god folk (Skt)
+	lijgri =3D line-group
+	lamgri =3D adjacent-group
+	mulgri =3D complete-group
+	cpami'i =3D pull-machine
+</dl><cx "asymmetrical tanru types, set + element-of-set">  XE "asymmetric=
al tanru types: set + element-of-set"   Conversely: the tertau is an elemen=
t, and the seltau represents a set in which that element is contained. Impl=
icitly, the meaning of the tertau is restricted from its usual general mean=
ing to the specific meaning appropriate for elements in the given set. Note=
 the opposition between =93zdani linji=94 in the previous group, and =93lin=
ji zdani=94 in this one, which shows why this kind of tanru is called =93as=
ymmetrical=94.
+<p>
+<dl compact><dt>	carvi dirgo		<dd>raindrop (Tur,Kor,Hun,Udm,Aba)=20
+	linji zdani		row house
+
+</dl><cx "asymmetrical tanru types, object + component/detail">  XE "asymm=
etrical tanru types: object + component/detail"   The seltau specifies an o=
bject and the tertau a component or detail of that object; the tanru as a w=
hole refers to the detail, specifying that it is a detail of that whole and=
 not some other.
+<p>
+<dl compact><dt>	junla dadysli		<dd>clock pendulum (Hun)=20
+	purdi vorme		garden door (Qab)=20
+	purdi bitmu		garden wall (Que)=20
+	moklu skapi		mouth skin (Imb =3D lips)=20
+	nazbi kevna		nose hole (Imb =3D nostril)=20
+	karce xislu		automobile wheel (Chi)=20
+	jipci pimlu		chicken feather (Chi)=20
+	vinji rebla		airplane tail (Chi)
+dadysli =3D hang-oscillator
+</dl><cx "asymmetrical tanru types, characteristic/detail + object">  XE "=
asymmetrical tanru types: characteristic/detail + object"   Conversely: the=
 seltau specifies a characteristic or important detail of the object descri=
bed by the tertau; objects described by the tanru as a whole are differenti=
ated from other similar objects by this detail.
+<p>
+<dl compact><dt>	pixra cukta		<dd>picture book=20
+	kerfa silka		hair silk (Kar =3D velvet)=20
+	plise tapla		apple cake (Tur)=20
+	dadysli junla		pendulum clock (Hun)
+	dadysli =3D hang-oscillator
+</dl><cx "asymmetrical tanru types, general-class + sub-class">  XE "asymm=
etrical tanru types: general-class + sub-class"   The tertau specifies a ge=
neral class of object (a genus), and the seltau specifies a sub-class of th=
at class (a species):
+<p>
+<dl compact><dt>	ckunu tricu		<dd>pine tree (Hun,Tur,Hop)
+</dl><cx "asymmetrical tanru types, possessor + object">  XE "asymmetrical=
 tanru types: possessor + object"   The tertau specifies an object of posse=
ssion, and the seltau may specify the possessor (the possession may be intr=
insic or otherwise). In English, these compounds have an explicit possessiv=
e element in them: =93lion's mane=94, =93child's foot=94, =93noble's cow=94=
.
+<p>
+<dl compact><dt>	cinfo kerfa		<dd>lion mane (Kor,Tur,Hun,Udm,Qab)=20
+	verba jamfu		child foot (Swe)=20
+	nixli tuple		girl leg (Swe)=20
+	cinfo jamfu		lion foot (Que)=20
+	danlu skapi		animal skin (Ewe)=20
+	ralju zdani		chief house (Ewe)=20
+	jmive munje		living world (Skt)=20
+	nobli bakni		noble cow (Skt)=20
+	nolraitru ralju		king chief (Skt =3D emperor)
+	nolraitru =3D nobly-superlative-ruler
+</dl><cx "asymmetrical tanru types, inhabitant + habitat">  XE "asymmetric=
al tanru types: inhabitant + habitat"   The tertau specifies a habitat, and=
 the seltau specifies the inhabitant:
+<p>
+<dl compact><dt>	lanzu tumla		<dd>family land
+
+</dl><cx "asymmetrical tanru types, effect + causative agent">  XE "asymme=
trical tanru types: effect + causative agent"   The tertau specifies a caus=
ative agent, and the seltau specifies the effect of that cause:
+<p>
+<dl compact><dt>	kalselvi'i gapci		<dd>tear gas (Hun)=20
+	terbi'a jurme		disease germ (Tur)=20
+	fenki litki		crazy liquid (Hop =3D whisky)=20
+	pinca litki		urine liquid (Hop =3D beer)
+	kalselvi'i =3D eye-excreted-thing
+	terbi'a =3D disease
+</dl><cx "asymmetrical tanru types, cause + effect">  XE "asymmetrical tan=
ru types: cause + effect"   Conversely: the tertau specifies an effect, and=
 the seltau specifies its cause.
+<p>
+<dl compact><dt>	djacu barna		<dd>water mark (Chi)
+</dl><cx "asymmetrical tanru types, purpose-of-instrument + instrument">  =
XE "asymmetrical tanru types: purpose-of-instrument + instrument"   The ter=
tau specifies an instrument, and the seltau specifies the purpose of that i=
nstrument:
+<p>
+<dl compact><dt>	taxfu dadgreku		<dd>garment rack (Chi)=20
+	tergu'i ti'otci		lamp shade (Chi)=20
+	xirma zdani		horse house (Chi =3D stall)=20
+	nuzba tanbo		news board (Chi =3D bulletin board)
+	dadgreku =3D hang-frame
+	tergu'i =3D source of illumination
+	ti'otci =3D shadow-tool
+</dl><cx "asymmetrical tanru types, object-of-purpose-of-instrument + inst=
rument">  XE "asymmetrical tanru types: object-of-purpose-of-instrument + i=
nstrument"   More vaguely: the tertau specifies an instrument, and the selt=
au specifies the object of the purpose for which that instrument is used:
+<p>
+<dl compact><dt>	cpina rokci		<dd>pepper stone (Que =3D stone for grinding=
 pepper)=20
+	jamfu djacu		foot water (Skt =3D water for washing the feet)=20
+	grana mudri		post wood (Skt =3D wood for making a post)=20
+	moklu djacu		mouth water (Hun =3D water for washing the
+					mouth)=20
+	lanme gerku		sheep dog (dog for working sheep)
+</dl><cx "asymmetrical tanru types, source + product">  XE "asymmetrical t=
anru types: source + product"   The tertau specifies a product from some so=
urce, and the seltau specifies the source of the product:
+<p>
+<dl compact><dt>	moklu djacu		<dd>mouth water (Aba,Qab =3D saliva)=20
+	ractu mapku		rabbit hat (Rus)=20
+	jipci sovda		chicken egg (Chi)=20
+	sikcurnu silka		silkworm silk (Chi)=20
+	mlatu kalci		cat feces (Chi)=20
+	bifce lakse		bee wax (Chi =3D beeswax)=20
+	cribe rectu		bear meat (Tur,Kor,Hun,Udm,Aba)=20
+	solxrula grasu		sunflower oil (Tur,Kor,Hun,Udm,Aba)=20
+	bifce jisra		bee juice (Hop =3D honey)=20
+	tatru litki		breast liquid (Hop =3D milk)=20
+	kanla djacu		eye water (Kor =3D tear)
+	sikcurnu =3D silk-worm
+	solxrula =3D solar-flower
+</dl><cx "asymmetrical tanru types, product + source">  XE "asymmetrical t=
anru types: product + source"   Conversely: the tertau specifies the source=
 of a product, and the seltau specifies the product:
+<p>
+<dl compact><dt>	silna jinto		<dd>salt well (Chi)=20
+	kolme terkakpa		coal mine (Chi)=20
+	ctile jinto		oil well (Chi)
+	terkakpa =3D source of digging
+</dl><cx "asymmetrical tanru types, source-material + object">  XE "asymme=
trical tanru types: source-material + object"   The tertau specifies an obj=
ect, and the seltau specifies the material from which the object is made. T=
his case is especially interesting, because the referent of the tertau may =
normally be made from just one kind of material, which is then overridden i=
n the tanru.
+<p>
+<dl compact><dt>	rokci cinfo		<dd>stone lion=20
+	snime nanmu		snow man (Hun)=20
+	kliti cipni		clay bird=20
+	blaci kanla		glass eye (Hun)=20
+	blaci kanla		glass eye (Que =3D spectacles)=20
+	solji sicni		gold coin (Tur)=20
+	solji junla		gold watch (Tur,Kor,Hun)=20
+	solji djine		gold ring (Udm,Aba,Que)=20
+	rokci zdani		stone house (Imb)=20
+	mudri zdani		wood house (Ewe =3D wooden house)=20
+	rokci bitmu		stone wall (Ewe)=20
+	solji carce		gold chariot (Skt)=20
+	mudri xarci		wood weapon (Skt =3D wooden weapon)=20
+	cmaro'i dargu		pebble road (Chi)=20
+	sudysrasu cutci		straw shoe (Chi)
+	cmaro'i =3D small-rock
+	sudysrasu =3D dry-grass
+</dl><p>Note: the two senses of =93blaci kanla=94 can be discriminated as:
+<p>
+<dl compact><dt>	blaci kanla bo tarmi	<dd>glass (eye shape) =3D glass eye=
=20
+	blaci kanla bo sidju	glass (eye helper) =3D spectacles
+</dl><cx "asymmetrical tanru types, object-measured + standard-object">  X=
E "asymmetrical tanru types: object-measured + standard-object"   The terta=
u specifies a typical object used to measure a quantity and the seltau spec=
ifies something measured. The tanru as a whole refers to a given quantity o=
f the thing being measured. English does not have compounds of this form, a=
s a rule.
+<p>
+<dl compact><dt>	tumla spisa		<dd>land piece (Tur =3D piece of land)=20
+	tcati kabri		tea cup (Kor,Aba =3D cup of tea)=20
+	nanba spisa		bread piece (Kor =3D piece of bread)=20
+	bukpu spisa		cloth piece (Udm,Aba =3D piece of cloth)=20
+	djacu calkyguzme	water calabash (Ewe =3D calabash of water)
+	calkyguzme =3D shell-fruit, calabash
+</dl><cx "asymmetrical tanru types, overriding-property + object-with-impl=
icit-properties">  XE "asymmetrical tanru types: overriding-property + obje=
ct-with-implicit-properties"   The tertau specifies an object with certain =
implicit properties, and the seltau overrides one of those implicit propert=
ies:
+<p>
+<dl compact><dt>	kensa bloti		<dd>spaceship=20
+	bakni verba		cattle child (Ewe =3D calf)
+</dl><cx "asymmetrical tanru types, whole + part">  XE "asymmetrical tanru=
 types: whole + part"   The seltau specifies a whole, and the tertau specif=
ies a part which normally is associated with a different whole. The tanru t=
hen refers to a part of the seltau which stands in the same relationship to=
 the whole seltau as the tertau stands to its typical whole.
+<p>
+<dl compact><dt>	kosta degji		<dd>coat finger (Hun =3D coat sleeve)=20
+	denci genja		tooth root (Imb)=20
+	tricu stedu		tree head (Imb =3D treetop)
+</dl><cx "asymmetrical tanru types, product + producer">  XE "asymmetrical=
 tanru types: product + producer"   The tertau specifies the producer of a =
certain product, and the seltau specifies the product. In this way, the tan=
ru as a whole distinguishes its referents from other referents of the terta=
u which do not produce the product.
+<p>
+<dl compact><dt>	silka curnu		<dd>silkworm (Tur,Hun,Aba)
+</dl><cx "asymmetrical tanru types, object-giving-characteristic + other-o=
bject">  XE "asymmetrical tanru types: object-giving-characteristic + other=
-object"   The tertau specifies an object, and the seltau specifies another=
 object which has a characteristic property. The tanru as a whole refers to=
 those referents of the tertau which possess the property.
+<p>
+<dl compact><dt>	sonci manti		<dd>soldier ant=20
+	ninmu bakni		woman cattle (Imb =3D cow)=20
+	mamta degji		mother finger (Imb =3D thumb)=20
+	cifnu degji		baby finger (Imb =3D pinky)=20
+	pacraistu zdani		hell house (Skt)=20
+	fagri dapma		fire curse (Skt =3D curse destructive as fire)
+	pacraistu =3D evil-superlative-site
+</dl><cx "asymmetrical tanru types, similar-appearance-object + object">  =
XE "asymmetrical tanru types: similar-appearance-object + object"   As a pa=
rticular case (when the property is that of resemblance): the seltau specif=
ies an object which the referent of the tanru resembles.
+<p>
+<dl compact><dt>	grutrceraso jbama	<dd>cherry bomb=20
+	solji kerfa		gold hair (Hun =3D golden hair)=20
+	kanla djacu		eye water (Kar =3D spring)=20
+	bakni rokci		bull stone (Mon =3D boulder)
+	grutrceraso =3D fu'ivla for =93cherry=94 based on Linnean name
+</dl><cx "asymmetrical tanru types, typical-place + object">  XE "asymmetr=
ical tanru types: typical-place + object"   The seltau specifies a place, a=
nd the tertau an object characteristically located in or at that place.
+<p>
+<dl compact><dt>	ckana boxfo		<dd>bed sheet (Chi)=20
+	rostu mojysu'a		tomb monument (Chi =3D tombstone)=20
+	jubme tergusni		table lamp (Chi)=20
+	foldi smacu		field mouse (Chi)=20
+	briju ci'ajbu		office desk (Chi)=20
+	rirxe xirma		river horse (Chi =3D hippopotamus)=20
+	xamsi gerku		sea dog (Chi =3D seal)=20
+	cagyce'u zdani		village house (Skt)
+	mrostu =3D dead-site
+	mojysu'a =3D remember-structure
+	ci'ajbu =3D write-table
+	cagyce'u =3D farm-community
+</dl><cx "asymmetrical tanru types, object + place-sold">  XE "asymmetrica=
l tanru types: object + place-sold"   Specifically: the tertau is a place w=
here the seltau is sold or made available to the public.
+<p>
+<dl compact><dt>	cidja barja		<dd>food bar (Chi =3D restaurant)=20
+	cukta barja		book bar (Chi =3D library)
+</dl><cx "asymmetrical tanru types, locus-of-application + object">  XE "a=
symmetrical tanru types: locus-of-application + object"   The seltau specif=
ies the locus of application of the tertau.
+<p>
+<dl compact><dt>	kanla velmikce		<dd>eye medicine (Chi)=20
+	jgalu grasu		nail oil (Chi =3D nail polish)=20
+	denci pesxu		tooth paste (Chi)
+	velmikce =3D treatment used by doctor
+</dl><cx "asymmetrical tanru types, activity + implement-used">  XE "asymm=
etrical tanru types: activity + implement-used"   The tertau specifies an i=
mplement used in the activity denoted by the seltau.
+<p>
+<dl compact><dt>	me la pinpan. bolci	<dd>Ping-Pong ball (Chi)
+</dl><cx "asymmetrical tanru types, undesired-object + protection-object">=
  XE "asymmetrical tanru types: undesired-object + protection-object"   The=
 tertau specifies a protective device against the undesirable features of t=
he referent of the seltau.
+<p>
+<dl compact><dt>	carvi mapku		<dd>rain cap (Chi)=20
+	carvi taxfu		rain garment (Chi =3D raincoat)=20
+	vindu firgai		poison mask (Chi =3D gas mask)
+	firgai =3D face-cover
+</dl><cx "asymmetrical tanru types, object + usual-container">  XE "asymme=
trical tanru types: object + usual-container"   The tertau specifies a cont=
ainer characteristically used to hold the referent of the seltau.
+<p>
+<dl compact><dt>	cukta vasru		<dd>book vessel (Chi =3D satchel)=20
+	vanju kabri		wine cup (Chi)=20
+	spatrkoka lanka		coca basket (Que)=20
+	rismi dakli		rice bag (Ewe,Chi)=20
+	tcati kabri		tea cup (Chi)=20
+	ladru botpi		milk bottle (Chi)=20
+	rismi patxu		rice pot (Chi)=20
+	festi lante		trash can (Chi)=20
+	bifce zdani		bee house (Kor =3D beehive)=20
+	cladakyxa'i zdani	sword house (Kor =3D sheath)=20
+	manti zdani		ant nest (Gua =3D anthill)
+	spatrkoka =3D fu'ivla for =93coca=94
+	cladakyxa'i =3D (long-knife)-weapon
+</dl><cx "asymmetrical tanru types, characteristic-time + event">  XE "asy=
mmetrical tanru types: characteristic-time + event"   The seltau specifies =
the characteristic time of the event specified by the tertau.
+<p>
+<dl compact><dt>	vensa djedi		<dd>spring day (Chi)=20
+	crisa citsi		summer season (Chi)=20
+	cerni bumru		morning fog (Chi)=20
+	critu lunra		autumn moon (Chi)=20
+	dunra nicte		winter night (Chi)=20
+	nicte ckule		night school (Chi)
+</dl><cx "asymmetrical tanru types, energy-source + powered">  XE "asymmet=
rical tanru types: energy-source + powered"   The seltau specifies a source=
 of energy for the referent of the tertau.
+<p>
+<dl compact><dt>	dikca tergusni		<dd>electric lamp (Chi)=20
+	ratni nejni		atom energy (Chi)=20
+	brife molki		windmill (Tur,Kor,Hun,Udm,Aba)
+	tergusni =3D illumination-source
+</dl><cx "asymmetrical tanru types, miscellaneous">  XE "asymmetrical tanr=
u types: miscellaneous"   Finally, some tanru which don't fall into any of =
the above categories.
+<p>
+<dl compact><dt>	ladru denci		<dd>milk tooth (Tur,Hun,Udm,Qab)=20
+	kanla denci		eye tooth
+</dl><p>It is clear that =93tooth=94 is being specified, and that =93milk=
=94 and =93eye=94 act as modifiers. However, the relationship between =93la=
dru=94 and =93denci=94 is something like =93tooth which one has when one is=
 drinking milk from one's mother=94, a relationship certainly present nowhe=
re except in this particular concept. As for =93kanla denci=94, the relatio=
nship is not only not present on the surface, it is hardly possible to form=
ulate it at all.
+<p>
+<a name=3Ds15><h3>Some types of symmetrical tanru</h3>
+<p>
+<cx "symmetrical tanru">  XE "symmetrical tanru"   This section deals with=
 symmetrical tanru, where order is not important. Many of these tanru can b=
e expressed with a logical or non-logical connective between the components=
.
+<p>
+<cx "symmetrical tanru types, both separately true">  XE "symmetrical tanr=
u types: both separately true"   The tanru may refer to things which are co=
rrectly specified by both tanru components. Some of these instances may als=
o be seen as asymmetrical tanru where the seltau specifies a material. The =
connective =93je=94 is appropriate:
+<p>
+<dl compact><dt>	cipnrstrigi pacru'i	<dd>owl demon (Skt)=20
+	nolraitru prije		royal sage (Skt)=20
+	remna nakni		human-being male (Qab =3D man)=20
+	remna fetsi		human-being female (Qab =3D woman)=20
+	sonci tolvri		soldier coward (Que)=20
+	panzi nanmu		offspring man (Ewe =3D son)=20
+	panzi ninmu		offspring woman (Ewe =3D daughter)=20
+	solji sicni		gold coin (Tur)=20
+	solji junla		gold watch (Tur,Kor,Hun)=20
+	solji djine		gold ring (Udm,Aba,Que)=20
+	rokci zdani		stone house (Imb)=20
+	mudri zdani		wooden house (Ewe)=20
+	rokci bitmu		stone wall (Ewe)=20
+	solji carce		gold chariot (Skt)=20
+	mudri xarci		wooden weapon (Skt)=20
+	zdani tcadu		home town (Chi)
+	cipnrstrigi =3D fu'ivla for =93owl=94 based on Linnean name
+	pacru'i =3D evil-spirit
+	tolvri =3D opposite-of-brave
+
+</dl><cx "symmetrical tanru types, one or other true">  XE "symmetrical ta=
nru types: one or other true"   The tanru may refer to all things which are=
 specified by either of the tanru components. The connective =93ja=94 is ap=
propriate:
+<p>
+<dl compact><dt>	nunji'a nunterji'a	<dd>victory defeat (Skt =3D victory or=
 defeat)=20
+	donri nicte		day night (Skt =3D day and night)=20
+	lunra tarci		moon stars (Skt =3D moon and stars)=20
+	patfu mamta		father mother (Imb,Kaz,Chi =3D parents)=20
+	tuple birka		leg arm (Kaz =3D extremity)=20
+	nuncti nunpinxe	eating drinking (Udm =3D cuisine)=20
+	bersa tixnu		son daughter (Chi =3D children)
+	nunji'a =3D event-of-winning
+	nunterji'a =3D event-of-losing
+	nuncti =3D event-of-eating
+	nunpinxe =3D event-of-drinking
+</dl><cx "symmetrical tanru types, using more inclusive class">  XE "symme=
trical tanru types: using more inclusive class"   Alternatively, the tanru =
may refer to things which are specified by either of the tanru components o=
r by some more inclusive class of things which the components typify:
+<p>
+<dl compact><dt>	curnu jalra		<dd>worm beetle (Mon =3D insect)=20
+	jalra curnu		beetle worm (Mon =3D insect)=20
+	kabri palta		cup plate (Kaz =3D crockery)=20
+	jipci gunse		hen goose (Qab =3D housefowl)=20
+	xrula tricu		flower tree (Chi =3D vegetation)
+</dl><cx "symmetrical tanru types, using crucial/typical parts">  XE "symm=
etrical tanru types: using crucial/typical parts"   The tanru components sp=
ecify crucial or typical parts of the referent of the tanru as a whole:
+<p>
+<dl compact><dt>	tumla vacri		<dd>land air (Fin =3D world)=20
+	moklu stedu		mouth head (Aba =3D face)=20
+	sudysrasu cunmi	hay millet (Qab =3D agriculture)=20
+	gugde ciste		state system (Mon =3D politics)=20
+	prenu so'imei		people multitude (Mon =3D masses)=20
+	djacu dertu		water earth (Chi =3D climate)
+	sudysrasu =3D dry-grass
+	so'imei =3D manysome
+ =93Pretty little girls' school=94: forty ways to say it</h3>
+<p></dl><cx "pretty little girls' school, forty ways">  XE "pretty little =
girls' school: forty ways, examples"  =20
+The following examples show every possible grouping arrangement of =93melb=
i cmalu nixli ckule=94 using =93bo=94 or =93ke=A0<dots>=85</dots>ke'e=94 fo=
r grouping and =93je=94 or =93jebo=94 for logical connection. Most of these=
 are definitely not plausible interpretations of the English phrase =93pret=
ty little girls' school=94, especially those which describe something which=
 is both a girl and a school.
+<p>
+Examples 4.2, 4.3, 4.4, 4.5, and 5.6 are repeated here as Examples 16.1, 1=
6.9, 16.17, 16.25, and 16.33 respectively. The seven examples following eac=
h of these share the same grouping pattern, but differ in the presence or a=
bsence of =93je=94 at each possible site. Some of the examples have more th=
an one Lojban version. In that case, they differ only in grouping mechanism=
, and are always equivalent in meaning.
+<p>
+The logical connective =93je=94 is associative: that is, =93A and (B and C=
)=94 is the same as =93(A and B) and C=94. Therefore, some of the examples =
have the same meaning as others. In particular, 16.8, 16.16, 16.24, 16.32, =
and 16.40 all have the same meaning because all four brivla are logically c=
onnected and the grouping is simply irrelevant. Other equivalent forms are =
noted in the examples themselves. However, if =93je=94 were replaced by =93=
naja=94 or =93jo=94 or most of the other logical connectives, the meanings =
would become distinct.
+<p>
+It must be emphasized that, because of the ambiguity of all tanru, the Eng=
lish translations are by no means definitive =97 they represent only one po=
ssible interpretation of the corresponding Lojban sentence.
+<p>
+<pre><a name=3De16d1>16.1)	melbi cmalu nixli ckule
+	((pretty type-of little) type-of girl) type-of school
+	school for girls who are beautifully small
+<a name=3De16d2>16.2)	melbi je cmalu nixli ckule
+	((pretty and little) type-of girl) type-of school
+	school for girls who are beautiful and small
+<a name=3De16d3>16.3)	melbi bo cmalu je nixli ckule
+	((pretty type-of little) and girl) type-of school
+	school for girls and for beautifully small things
+<a name=3De16d4>16.4)	ke melbi cmalu nixli ke'e je ckule
+	((pretty type-of little) type-of girl) and school
+	thing which is a school and a beautifully small girl
+<a name=3De16d5>16.5)	melbi je cmalu je nixli ckule
+	((pretty and little) and girl) type-of school
+	school for things which are beautiful, small, and girls
+	Note:  same as 16.21
+<a name=3De16d6>16.6)	melbi bo cmalu je nixli je ckule
+	((pretty type-of little) and girl) and school
+	thing which is beautifully small, a school, and a girl
+	Note:  same as 16.14
+<a name=3De16d7>16.7)	ke melbi je cmalu nixli ke'e je ckule
+	((pretty and little) type-of girl) and school
+	thing which is a school and a girl who is both beautiful and small
+<a name=3De16d8>16.8)	melbi je cmalu je nixli je ckule
+	((pretty and little) and girl) and school
+	thing which is beautiful, small, a girl, and a school
+<a name=3De16d9>16.9)	melbi cmalu nixli bo ckule
+	(pretty type-of little) type-of (girl type-of school)
+	girls' school which is beautifully small
+<a name=3De16d10>16.10)	melbi je cmalu nixli bo ckule
+	(pretty and little) type-of (girl type-of school)
+	girls' school which is beautiful and small
+<a name=3De16d11>16.11)	melbi cmalu nixli je ckule
+	(pretty type-of little) type-of (girl and school)
+	something which is a girl and a school which is beautifully small
+
+<a name=3De16d12>16.12)	melbi bo cmalu je nixli bo ckule
+	(pretty type-of little) and (girl type-of school)
+	something which is beautifully small and a girls' school
+<a name=3De16d13>16.13)	melbi je cmalu nixli je ckule
+	(pretty and little) type-of (girl and school)
+	a pretty and little type of thing which is both a girl and a school
+<a name=3De16d14>16.14)	melbi bo cmalu je nixli jebo ckule
+	(pretty type-of little) and (girl and school)
+	thing which is beautifully small, a school, and a girl
+	Note:  same as 16.6
+<a name=3De16d15>16.15)	melbi jebo cmalu je nixli bo ckule
+	(pretty and little) and (girl type-of school)
+	thing which is beautiful and small and a girl's school
+	Note:  same as 16.30
+<a name=3De16d16>16.16)	melbi jebo cmalu je nixli jebo ckule
+	(pretty and little) and (girl and school)
+	thing which is beautiful, small, a girl, and a school
+<a name=3De16d17>16.17)	melbi cmalu bo nixli ckule
+	(pretty type-of (little type-of girl)) type-of school
+	school for beautiful girls who are small
+<a name=3De16d18>16.18)	melbi cmalu je nixli ckule
+	(pretty type-of (little and girl)) type-of school
+	school for beautiful things which are small and are girls
+<a name=3De16d19>16.19)	melbi je cmalu bo nixli ckule
+	(pretty and (little type-of girl)) type-of school
+	school for things which are beautiful and are small girls
+<a name=3De16d20>16.20)	ke melbi cmalu bo nixli ke'e je ckule
+	melbi bo cmalu bo nixli je ckule
+	(pretty type-of (little type-of girl)) and school
+	thing which is a school and a small girl who is beautiful
+<a name=3De16d21>16.21)	melbi je cmalu jebo nixli ckule
+	(pretty and (little and girl)) type-of school
+	school for things which are beautiful, small, and girls
+	Note:  same as 16.5
+<a name=3De16d22>16.22)	melbi je cmalu bo nixli je ckule
+	(pretty and (little type-of girl)) and school
+	thing which is beautiful, a small girl, and a school
+	Note:  same as 16.38
+<a name=3De16d23>16.23)	ke melbi cmalu je nixli ke'e je ckule
+	(pretty type-of (little and girl)) and school
+	thing which is beautifully small, a beautiful girl, and a school
+
+<a name=3De16d24>16.24)	melbi je cmalu jebo nixli je ckule
+	(pretty and (little and girl)) and school
+	thing which is beautiful, small, a girl, and a school
+<a name=3De16d25>16.25)	melbi cmalu bo nixli bo ckule
+	melbi ke cmalu ke nixli ckule [ke'e] [ke'e]
+	pretty type-of (little type-of (girl type-of school))
+	small school for girls which is beautiful
+<a name=3De16d26>16.26)	melbi ke cmalu nixli je ckule [ke'e]
+	pretty type-of (little type-of (girl and school))
+	small thing, both a girl and a school, which is beautiful
+<a name=3De16d27>16.27)	melbi cmalu je nixli bo ckule
+	pretty type-of (little and (girl type-of school))
+	thing which is beautifully small and a girls' school that is beautiful
+<a name=3De16d28>16.28)	melbi je cmalu bo nixli bo ckule
+	melbi je ke cmalu nixli bo ckule [ke'e]
+	melbi je ke cmalu ke nixli ckule [ke'e] [ke'e]
+	pretty and (little type-of (girl type-of school))
+	thing which is beautiful and a small type of girls' school
+<a name=3De16d29>16.29)	melbi cmalu je nixli jebo ckule
+	melbi cmalu je ke nixli je ckule [ke'e]
+	pretty type-of (little and (girl and school))
+	thing which is beautifully small, a beautiful girl, and a beautiful schoo=
l
+	Note:  same as 16.37
+<a name=3De16d30>16.30)	melbi je cmalu jebo nixli bo ckule
+	melbi je ke cmalu je nixli bo ckule [ke'e]
+	pretty and (little and (girl type-of school))
+	thing which is beautiful, small and a girls' school
+	Note:  same as 16.15
+<a name=3De16d31>16.31)	melbi je ke cmalu nixli je ckule [ke'e]
+	pretty and (little type-of (girl and school))
+	beautiful thing which is a small girl and a small school
+<a name=3De16d32>16.32)	melbi jebo cmalu jebo nixli jebo ckule
+	pretty and (little and (girl and school))
+	thing which is beautiful, small, a girl, and a school
+<a name=3De16d33>16.33)	melbi ke cmalu nixli ckule [ke'e]
+	pretty type-of ((little type-of girl) type-of school)
+	beautiful school for small girls
+<a name=3De16d34>16.34)	melbi ke cmalu je nixli ckule [ke'e]
+	pretty type-of ((little and girl) type-of school
+	beautiful school for things which are small and are girls
+<a name=3De16d35>16.35)	melbi ke cmalu bo nixli je ckule [ke'e]
+	pretty type-of ((little type-of girl) and school)
+	beautiful thing which is a small girl and a school
+<a name=3De16d36>16.36)	melbi je ke cmalu nixli ckule [ke'e]
+	pretty and ((little type-of girl) type-of school)
+	thing which is beautiful and a school for small girls
+<a name=3De16d37>16.37)	melbi cmalu je nixli je ckule
+	pretty type-of ((little and girl) and school)
+	thing which is beautifully small, a beautiful girl, and a beautiful schoo=
l
+	Note:  same as 16.29
+<a name=3De16d38>16.38)	melbi je ke cmalu bo nixli je ckule [ke'e]
+	pretty and ((little type-of girl) and school)
+	thing which is beautiful, a small girl and a school
+	Note:  same as 16.22
+<a name=3De16d39>16.39)	melbi je ke cmalu je nixli ckule [ke'e]
+	pretty and ((little and girl) type-of school)
+	thing which is beautiful and is a small school and a girls' school
+<a name=3De16d40>16.40)	melbi je ke cmalu je nixli je ckule [ke'e]
+	pretty and ((little and girl) and school)
+	thing which is beautiful, small, a girl, and a school
+
+<center><p align=3Dcenter>
+<img src=3Dchap6.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D4=
05>
+
+<h2>Chapter 6
+<br>
+To Speak Of Many Things: The Lojban sumti</h2>
+<h6>$Revision: 4.1 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>The five kinds of simple sumti</h3>
+<p>
+<cx "simple sumti">  XE "simple sumti"   <cx "sumti, definition">  XE "sum=
ti: definition"   If you understand anything about Lojban, you know what a =
sumti is by now, right? An argument, one of those things that fills the pla=
ces of simple Lojban sentences like:
+<p>
+<pre><a name=3De1d1>1.1)	mi klama le zarci
+	I go-to the market
+</pre><cx "pro-sumti, contrasted with description">  XE "pro-sumti: contra=
sted with description"   In <a href=3D#e1d1>Example 1.1</a>, =93mi=94 and =
=93le zarci=94 are the sumti. It is easy to see that these two sumti are no=
t of the same kind: =93mi=94 is a pro-sumti (the Lojban analogue of a prono=
un) referring to the speaker, whereas =93le zarci=94 is a description which=
 refers to something described as being a market.
+<p>
+<cx "sumti, kinds of">  XE "sumti: kinds of"   There are five kinds of sim=
ple sumti provided by Lojban:
+<p>
+<cx "sumti, descriptions as">  XE "sumti: descriptions as"   <cx "gadri, d=
efinition">  XE "gadri: definition"   <lx "le">  XE "le"   <dl compact><dt>=
1)	<dd>descriptions like =93le zarci=94, which usually begin with a descrip=
tor (called a =93gadri=94 in Lojban) such as =93le=94;
+</dl><cx "sumti, pro-sumti as">  XE "sumti: pro-sumti as"   <lx "mi">  XE =
"mi"   <dl compact><dt>2)	<dd>pro-sumti, such as =93mi=94;
+</dl><cx "sumti, names as">  XE "sumti: names as"   <lx "la">  XE "la"   <=
dl compact><dt>3)	<dd>names, such as =93la lojban.=94, which usually begin =
with =93la=94;
+</dl><cx "sumti, quotations as">  XE "sumti: quotations as"   <lx "lu">  X=
E "lu"   <lx "le'u">  XE "le'u"   <lx "zo">  XE "zo"   <lx "zoi">  XE "zoi"=
   <dl compact><dt>4)	<dd>quotations, which begin with =93lu=94, =93le'u=94=
, =93zo=94, or =93zoi=94;
+</dl><cx "sumti, numbers as">  XE "sumti: numbers as"   <lx "li">  XE "li"=
   <dl compact><dt>5)	<dd>pure numbers, which usually begin with =93li=94.
+</dl><p>Here are a few examples of each kind of sumti:
+<p>
+<lx "ko">  XE "ko"   <pre><a name=3De1d2>1.2)	e'osai ko sarji la lojban.
+	Please support Lojban!
+</pre><a href=3D#e1d2>Example 1.2 </a>exhibits =93ko=94, a pro-sumti; and =
=93la lojban.=94, a name.
+<p>
+<lx "mi">  XE "mi"   <lx "lu">  XE "lu"   <lx "li'u">  XE "li'u"   <lx "le=
">  XE "le"   <pre><a name=3De1d3>1.3)	mi cusku lu e'osai li'u le tcidu
+	I express =93Please!=94 to-the reader.
+</pre><a href=3D#e1d3>Example 1.3 </a>exhibits =93mi=94, a pro-sumti; =93l=
u e'osai li'u=94, a quotation; and =93le tcidu=94, a description.
+<p>
+<lx "ti">  XE "ti"   <lx "li">  XE "li"   <pre><a name=3De1d4>1.4)	ti mitr=
e li ci
+	This measures-in-meters the-number three.
+	This is three meters long.
+</pre><a href=3D#e1d4>Example 1.4 </a>exhibits =93ti=94, a pro-sumti; and =
=93li ci=94, a number.
+<p>
+Most of this chapter is about descriptions, as they have the most complica=
ted syntax and usage. Some attention is also given to names, which are clos=
ely interwoven with descriptions. Pro-sumti, numbers, and quotations are de=
scribed in more detail in <a href=3Dchap7.html>Chapter 7</a>, <a href=3Dcha=
p18.html>Chapter 18</a>, and <a href=3Dchap19.html>Chapter 19 </a>respectiv=
ely, so this chapter only gives summaries of their forms and uses. See <a h=
ref=3D#s13>Section 13 </a>through <a href=3D#s15>Section 15 </a>for these s=
ummaries.
+
+<p>
+<a name=3Ds2><h3>The three basic description types</h3>
+<p>
+<cx "descriptions, types of">  XE "descriptions: types of"   The following=
 cmavo are discussed in this section:
+<p>
+<pre>	le	LE	the, the one(s) described as
+	lo	LE	some, some of those which really are
+	la	LA	the one(s) named
+	ku	KU	elidable terminator for LE, LA
+</pre><cx "descriptions, types of">  XE "descriptions: types of"   <cx "de=
scriptions, components of">  XE "descriptions: components of"   <cx "descri=
ptor, as part of description">  XE "descriptor: as part of description"   <=
cx "selbri, as part of description">  XE "selbri: as part of description"  =
 <lx "LE">  XE "LE selma=92o"   <lx "LA">  XE "LA selma=92o"   The syntax o=
f descriptions is fairly complex, and not all of it can be explained within=
 the confines of this chapter: relative clauses, in particular, are discuss=
ed in <a href=3Dchap8.html>Chapter 8</a>. However, most descriptions have j=
ust two components: a descriptor belonging to selma'o LE or LA, and a selbr=
i. (The difference between selma'o LE and selma'o LA is not important until=
 <a href=3D#s12>Section 12</a>.) Furthermore, the selbri is often just a si=
ngle brivla. Here is an elementary example:
+<p>
+<pre><a name=3De2d1>2.1)	le zarci
+	one-or-more-specific-things-each-of-which-I-describe-as being-a-market
+	the market
+</pre><cx "le, compared with English =93the=94">  XE "le: compared with En=
glish \=94the\=94"   The long gloss for =93le=94 is of course far too long =
to use most of the time, and in fact =93le=94 is quite close in meaning to =
English =93the=94. It has particular implications, however, which =93the=94=
 does not have.
+<p>
+<cx "descriptors, purpose of">  XE "descriptors: purpose of"   <cx "descri=
ptions, importance of selbri first place in">  XE "descriptions: importance=
 of selbri first place in"   The general purpose of all descriptors is to c=
reate a sumti which might occur in the x1 place of the selbri belonging to =
the description. Thus =93le zarci=94 conveys something which might be found=
 in the x1 place of =93zarci=94, namely a market.
+<p>
+<cx "le, implications of">  XE "le: implications of"   <cx "le, and specif=
icity">  XE "le: and specificity"   <cx "le, and truth of selbri">  XE "le:=
 and truth of selbri"   The specific purpose of =93le=94 is twofold. First,=
 it indicates that the speaker has one or more specific markets in mind (wh=
ether or not the listener knows which ones they are). Second, it also indic=
ates that the speaker is merely describing the things he or she has in mind=
 as markets, without being committed to the truth of that description.
+<p>
+<pre><a name=3De2d2>2.2)	le zarci cu barda
+	One-or-more-specific-things-which-I-describe as =93markets=94 is/are-big.
+	The market is big.
+	The markets are big.
+</pre><cx "plurals, Lojban contrasted with English in necessity of marking=
">  XE "plural: Lojban contrasted with English in necessity of marking"   N=
ote that English-speakers must state whether a reference to markets is to j=
ust one (=93the market=94) or to more than one (=93the markets=94). Lojban =
requires no such forced choice, so both colloquial translations of <a href=
=3D#e2d2>Example 2.2 </a>are valid. Only the context can specify which is m=
eant. (This rule does not mean that Lojban has no way of specifying the num=
ber of markets in such a case: that mechanism is explained in <a href=3D#s7=
>Section 7</a>.)
+<p>
+<ex "The men are women">  XE "The men are women: example=93  Now consider =
the following strange-looking example:
+<p>
+<pre><a name=3De2d3>2.3)	le nanmu cu ninmu
+	One-or-more-specific-things-which-I-describe as =93men=94 are women
+	The man is a woman.
+	The men are women.
+</pre><cx "le, in false-to-fact descriptions">  XE "le: in false-to-fact d=
escriptions"   <a href=3D#e2d3>Example 2.3 </a>is not self-contradictory in=
 Lojban, because =93le nanmu=94 merely means something or other which, for =
my present purposes, I choose to describe as a man, whether or not it reall=
y is a man. A plausible instance would be: someone we had assumed to be a m=
an at a distance turned out to be actually a woman on closer observation. <=
a href=3D#e2d3>Example 2.3 </a>is what I would say to point out my observat=
ion to you.
+<p>
+<cx "specific descriptions">  XE "specific descriptions"   <cx "descriptio=
ns, specific">  XE "descriptions: specific"   In all descriptions with =93l=
e=94, the listener is presumed to either know what I have in mind or else n=
ot to be concerned at present (perhaps I will give more identifying details=
 later). In particular, I might be pointing at the supposed man or men: <a =
href=3D#e2d3>Example 2.3 </a>would then be perfectly intelligible, since =
=93le nanmu=94 merely clarifies that I am pointing at the supposed man, not=
 at a landscape, or a nose, which happens to lie in the same direction.
+<p>
+<cx "non-specific descriptions">  XE "non-specific descriptions"   <cx "de=
scriptions, non-specific">  XE "descriptions: non-specific"   <lx "lo">  XE=
 "lo"   <cx "lo, implications of">  XE "lo: implications of"   The second d=
escriptor dealt with in this section is =93lo=94. Unlike =93le=94, =93lo=94=
 is nonspecific:
+<p>
+<pre><a name=3De2d4>2.4)	lo zarci
+	one-or-more-of-all-the-things-which-really are-markets
+	a market
+	some markets
+</pre><cx "lo, contrasted with le in specificity">  XE "lo: contrasted wit=
h le in specificity"   <cx "le, contrasted with lo in specificity">  XE "le=
: contrasted with lo in specificity"   <cx "le, contrasted with lo in truth=
 requirement">  XE "le: contrasted with lo in truth requirement"   <cx "lo,=
 contrasted with le in truth requirement">  XE "lo: contrasted with le in t=
ruth requirement"   <cx "lo, and truth of selbri">  XE "lo: and truth of se=
lbri"   Again, there are two colloquial English translations. The effect of=
 using =93lo=94 in <a href=3D#e2d4>Example 2.4 </a>is to refer generally to=
 one or more markets, without being specific about which. Unlike =93le zarc=
i=94, =93lo zarci=94 must refer to something which actually is a market (th=
at is, which can appear in the x1 place of a truthful bridi whose selbri is=
 =93zarci=94). Thus
+<p>
+<dl compact><dt>2.5)	<dd>lo nanmu cu ninmu
+	Some man is a woman.
+	Some men are women.
+</dl>must be false in Lojban, given that there are no objects in the real =
world which are both men and women. Pointing at some specific men or women =
would not make <a href=3D#e2d5>Example 2.5 </a>true, because those specific=
 individuals are no more both-men-and-women than any others. In general, =
=93lo=94 refers to whatever individuals meet its description.
+<p>
+<lx "la">  XE "la"   <cx "la, implications of">  XE "la: implications of" =
  <cx "la, compared with le in specificity">  XE "la: compared with le in s=
pecificity"   <cx "le, compared with la in specificity">  XE "le: compared =
with la in specificity"   <cx "la, use with descriptions contrasted with us=
e before Lojbanized names">  XE "la: use with descriptions contrasted with =
use before Lojbanized names"   The last descriptor of this section is =93la=
=94, which indicates that the selbri which follows it has been dissociated =
from its normal meaning and is being used as a name. Like =93le=94 descript=
ions, =93la=94 descriptions are implicitly restricted to those I have in mi=
nd. (Do not confuse this use of =93la=94 with its use before regular Lojban=
ized names, which is discussed in <a href=3D#s12>Section 12</a>.) For examp=
le:
+<p>
+<ex "bear wrote story">  XE "bear wrote story: example=93  <pre><a name=3D=
e2d6>2.6)	la cribe pu finti le lisri
+	The-one-named =93bear=94 [past] creates the story.
+	Bear wrote the story.
+</pre><cx "naming predicate">  XE "naming predicate"   In <a href=3D#e2d6>=
Example 2.6</a>, =93la cribe=94 refers to someone whose naming predicate is=
 =93cribe=94, i.e. =93Bear=94. In English, most names don't mean anything, =
or at least not anything obvious. The name =93Frank=94 coincides with the E=
nglish word =93frank=94, meaning =93honest=94, and so one way of translatin=
g =93Frank ate some cheese=94 into Lojban would be:
+<p>
+<pre><a name=3De2d7>2.7)	la stace pu citka lo cirla
+	The-one-called =93Honest/Frank=94 [past] eats some cheese.
+</pre><p>English-speakers typically would not do this, as we tend to be mo=
re attached to the sound of our names than their meaning, even if the meani=
ng (etymological or current) is known. Speakers of other languages may feel=
 differently. (In point of fact, =93Frank=94 originally meant =93the free o=
ne=94 rather than =93the honest one=94.)
+
+<p>
+<cx "la, contrasted with le in implications">  XE "la: contrasted with le =
in implications"   It is important to note the differences between <a href=
=3D#e2d6>Example 2.6 </a>and the following:
+<p>
+<pre><a name=3De2d8>2.8)	le cribe pu finti le lisri
+	One-or-more-specific-things-which-I-describe-as a-bear
+[past] creates the story.
+	The bear(s) wrote the story.
+</pre><cx "la, contrasted with lo in implications">  XE "la: contrasted wi=
th lo in implications"   <pre><a name=3De2d9>2.9)	lo cribe pu finti le lisr=
i
+	One-or-more-of-the-things-which-really are-bears
+		[past] creates the story.
+	A bear wrote the story.
+	Some bears wrote the story.
+</pre><a href=3D#e2d8>Example 2.8 </a>is about a specific bear or bearlike=
 thing(s), or thing(s) which the speaker (perhaps whimsically or metaphoric=
ally) describes as a bear (or more than one); <a href=3D#e2d9>Example 2.9 <=
/a>is about one or more of the really existing, objectively defined bears. =
In either case, though, each of them must have contributed to the writing o=
f the story, if more than one bear (or =93bear=94) is meant.
+<p>
+<cx "descriptions with lo, teddy bear contrasted with real bear">  XE "des=
criptions with lo: teddy bear contrasted with real bear"   (The notion of a=
 =93really existing, objectively defined bear=94 raises certain difficultie=
s. Is a panda bear a =93real bear=94? How about a teddy bear? In general, t=
he answer is =93yes=94. Lojban gismu are defined as broadly as possible, al=
lowing tanru and lujvo to narrow down the definition. There probably are no=
 necessary and sufficient conditions for defining what is and what is not a=
 bear that can be pinned down with complete precision: the real world is fu=
zzy. In borderline cases, =93le=94 may communicate better than =93lo=94.)
+<p>
+So while <a href=3D#e2d6>Example 2.6 </a>could easily be true (there is a =
real writer named =93Greg Bear=94), and <a href=3D#e2d8>Example 2.8 </a>cou=
ld be true if the speaker is sufficiently peculiar in what he or she descri=
bes as a bear, <a href=3D#e2d9>Example 2.9 </a>is certainly false.
+<p>
+Similarly, compare the following two examples, which are analogous to <a h=
ref=3D#e2d8>Example 2.8 </a>and <a href=3D#e2d9>Example 2.9 </a>respectivel=
y:
+<p>
+<cx "le, contrasted with lo in implications">  XE "le: contrasted with lo =
in implications"   <cx "lo, contrasted with le in implications">  XE "lo: c=
ontrasted with le in implications"   <pre><a name=3De2d10>2.10)	le remna pu=
 finti le lisri
+	The human being(s) wrote the story.
+<a name=3De2d11>2.11)	lo remna pu finti le lisri
+	A human being wrote the story.
+	Some human beings wrote the story.
+</pre><a href=3D#e2d10>Example 2.10 </a>says who the author of the story i=
s: one or more particular human beings that the speaker has in mind. If the=
 topic of conversation is the story, then <a href=3D#e2d10>Example 2.10 </a=
>identifies the author as someone who can be pointed out or who has been pr=
eviously mentioned; whereas if the topic is a person, then =93le remna=94 i=
s in effect a shorthand reference to that person. <a href=3D#e2d11>Example =
2.11 </a>merely says that the author is human.
+<p>
+<lx "ku">  XE "ku"   <cx "ku, as elidable terminator for descriptions">  X=
E "ku: as elidable terminator for descriptions"   <cx "ku, effect of follow=
ing selbri on elidability of">  XE "ku: effect of following selbri on elida=
bility of"   <cx "cu, effect on elidability of ku">  XE "cu: effect on elid=
ability of ku"   <cx "ku, uses of">  XE "ku: uses of"   The elidable termin=
ator for all descriptions is =93ku=94. It can almost always be omitted with=
 no danger of ambiguity. The main exceptions are in certain uses of relativ=
e clauses, which are discussed in <a href=3Dchap8.html>Chapter 8</a>, and i=
n the case of a description immediately preceding the selbri. In this latte=
r case, using an explicit =93cu=94 before the selbri makes the =93ku=94 unn=
ecessary. There are also a few other uses of =93ku=94: in the compound nega=
tor =93naku=94 (discussed in <a href=3Dchap15.html>Chapter 15</a>) and to t=
erminate place-structure, tense, and modal tags that do not have associated=
 sumti (discussed in <a href=3Dchap9.html>Chapter 9 </a>and <a href=3Dchap1=
0.html>Chapter 10</a>).
+<p>
+<a name=3Ds3><h3>Individuals and masses</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	lei	LE	the mass I describe as
+	loi	LE	part of the mass of those which really are
+	lai	LA	the mass of those named
+</pre><cx "sumti, classified by types of objects referred to">  XE "sumti:=
 classified by types of objects referred to"   <cx "sumti, for individual o=
bjects">  XE "sumti: for individual objects"   <cx "sumti, for mass objects=
">  XE "sumti: for mass objects"   <cx "sumti, for set objects">  XE "sumti=
: for set objects"   All Lojban sumti are classified by whether they refer =
to one of three types of objects, known as =93individuals=94, =93masses=94,=
 and =93sets=94. The term =93individual=94 is misleading when used to refer=
 to more than one object, but no less-confusing term has as yet been found.=
 All the descriptions in <a href=3D#s1>Sections 1 </a>and <a href=3D#s2>2 <=
/a>refer to individuals, whether one or more than one. Consider the followi=
ng example:
+<p>
+<pre><a name=3De3d1>3.1)	le prenu cu bevri le pipno
+	One-or-more-of-those-I-describe-as persons carry the piano.
+	The person(s) carry the piano.
+</pre>(Of course the second =93le=94 should really get the same translatio=
n as the first, but I am putting the focus of this discussion on the first =
=93le=94, the one preceding =93prenu=94. I will assume that there is only o=
ne piano under discussion.)
+<p>
+<cx "plurals with le, meaning of">  XE "plural: meaning of le with"   <cx =
"le, meaning of in the plural">  XE "le: meaning of in the plural"   <cx "m=
ultiple individual objects, meaning of">  XE "multiple individual objects: =
meaning of"   <cx "individual objects, multiple">  XE "individual objects: =
multiple"   Suppose the context of <a href=3D#e3d1>Example 3.1 </a>is such =
that you can determine that I am talking about three persons. What am I cla=
iming? I am claiming that each of the three persons carried the piano. This=
 claim can be true if the persons carried the piano one at a time, or by tu=
rns, or in a variety of other ways. But in order for <a href=3D#e3d1>Exampl=
e 3.1 </a>to be true, I must be willing to assert that person 1 carried the=
 piano, and that person 2 carried the piano, and that person 3 carried the =
piano.
+<p>
+But suppose I am not willing to claim that. For in fact pianos are heavy, =
and very few persons can carry a piano all by themselves. The most likely f=
actual situation is that person 1 carried one end of the piano, and person =
2 the other end, while person 3 either held up the middle or else supervise=
d the whole operation without actually lifting anything. The correct way of=
 expressing such a situation in Lojban is:
+<p>
+<lx "lei">  XE "lei"   <pre><a name=3De3d2>3.2)	lei prenu cu bevri le pipn=
o
+	The-mass-of-one-or-more-of-those-I-describe-as persons carry the piano.
+</pre><ex "piano-moving">  XE "piano-moving: example=93  <cx "mass object,=
 contrasted with multiple individual objects">  XE "mass object: contrasted=
 with multiple individual objects"   <cx "multiple individual objects, cont=
rasted with mass object">  XE "multiple individual objects: contrasted with=
 mass object"   <cx "mass object, properties of">  XE "mass object: propert=
ies of"   <cx "mass object, and logical reasoning">  XE "mass object: and l=
ogical reasoning"   Here the same three persons are treated not as individu=
als, but as a so-called =93mass entity=94, or just =93mass=94. A mass has t=
he properties of each individual which composes it, and may have other prop=
erties of its own as well. This can lead to apparent contradictions. Thus s=
uppose in the piano-moving example above that person 1 has fair skin, where=
as person 2 has dark skin. Then it is correct to say that the person-mass h=
as both fair skin and dark skin. Using the mass descriptor =93lei=94 signal=
s that ordinary logical reasoning is not applicable: contradictions can be =
maintained, and all sorts of other peculiarities may exist. However, we can=
 safely say that a mass inherits only the component properties that are rel=
evant to it; it would be ludicrous to say that a mass of two persons is of =
molecular dimensions, simply because some of the parts (namely, the molecul=
es) of the persons are that small.
+<p>
+<lx "loi">  XE "loi"   <lx "lai">  XE "lai"   <cx "loi, as mass counterpar=
t of lo">  XE "loi: as mass counterpart of lo"   <cx "lai, as mass counterp=
art of lai">  XE "lai: as mass counterpart of lai"   The descriptors =93loi=
=94 and =93lai=94 are analogous to =93lo=94 and =93la=94 respectively, but =
refer to masses either by property (=93loi=94) or by name (=93lai=94). A cl=
assic example of =93loi=94 use is:
+
+<p>
+<ex "lions in Africa">  XE "lions in Africa: example=93  <pre><a name=3De3=
d3>3.3)	loi cinfo cu xabju le fi'ortu'a
+	Part-of-the-mass-of-those-which-really are-lions dwell in-the African-lan=
d.
+	The lion dwells in Africa.
+	Lions dwell in Africa.
+</pre><cx "lei, contrasted with loi in specificity">  XE "lei: contrasted =
with loi in specificity"   <cx "loi, contrasted with lei in specificity">  =
XE "loi: contrasted with lei in specificity"   The difference between =93le=
i=94 and =93loi=94 is that =93lei cinfo=94 refers to a mass of specific ind=
ividuals which the speaker calls lions, whereas =93loi cinfo=94 refers to s=
ome part of the mass of all those individuals which actually are lions. The=
 restriction to =93some part of the mass=94 allows statements like <a href=
=3D#e3d3>Example 3.3 </a>to be true even though some lions do not dwell in =
Africa =97 they live in various zoos around the world. On the other hand, <=
a href=3D#e3d3>Example 3.3 </a>doesn't actually say that most lions live in=
 Africa: equally true is
+<p>
+<ex "Englishman in Africa">  XE "Englishman in Africa: example=93  <pre><a=
 name=3De3d4>3.4)	loi glipre=20
+cu xabju le fi'ortu'a
+	Part-of-the-mass-of-those-which-really are-English-persons=20
+		dwell in-the African-land.
+	The English dwell in Africa.
+</pre>since there is at least one English person living there. <a href=3D#=
s4>Section 4 </a>explains another method of saying what is usually meant by=
 =93The lion lives in Africa=94 which does imply that living in Africa is n=
ormal, not exceptional, for lions.
+<p>
+<cx "mass objects, peculiarities of English translation of">  XE "mass obj=
ects: peculiarities of English translation of"   Note that the Lojban mass =
articles are sometimes translated by English plurals (the most usual case),=
 sometimes by English singulars (when the singular is used to express typic=
alness or abstraction), and sometimes by singulars with no article:
+<p>
+<ex "butter is soft">  XE "butter is soft: example=93  <lx "matne">  XE "m=
atne"   <pre><a name=3De3d5>3.5)	loi matne cu ranti
+	Part-of-the-mass-of-that-which-really is-a-quantity-of-butter is-soft.
+	Butter is soft.
+</pre><cx "mass object, as dependent on intention">  XE "mass object: as d=
ependent on intention"   Of course, some butter is hard (for example, if it=
 is frozen butter), so the =93part-of=94 implication of =93loi=94 becomes o=
nce again useful. The reason this mechanism works is that the English words=
 like =93butter=94, which are seen as already describing masses, are transl=
ated in Lojban by non-mass forms. The place structure of =93matne=94 is =93=
x1 is a quantity of butter from source x2=94, so the single English word =
=93butter=94 is translated as something like =93a part of the mass formed f=
rom all the quantities of butter that exist=94. (Note that the operation of=
 forming a mass entity does not imply, in Lojban, that the components of th=
e mass are necessarily close to one another or even related in any way othe=
r than conceptually. Masses are formed by the speaker's intention to form a=
 mass, and can in principle contain anything.)
+<p>
+<cx "mass name, use of">  XE "mass name: use of"   The mass name descripto=
r =93lai=94 is used in circumstances where we wish to talk about a mass of =
things identified by a name which is common to all of them. It is not used =
to identify a mass by a single name peculiar to it. Thus the mass version o=
f <a href=3D#e2d5>Example 2.5</a>,
+<p>
+<ex "Bears wrote book">  XE "Bears wrote book: example=93  <pre><a name=3D=
e3d6>3.6)	lai cribe pu finti le vi cukta
+	The-mass-of-those-named =93bear=94 [past] creates the nearby book.
+	The Bears wrote this book.
+</pre><cx "la, contrasted with lai in implications">  XE "la: contrasted w=
ith lai in implications"   <cx "lai, contrasted with la in implications">  =
XE "lai: contrasted with la in implications"   in a context where =93la cri=
be=94 would be understood as plural, would mean that either Tom Bear or Fre=
d Bear (to make up some names) might have written the book, or that Tom and=
 Fred might have written it as collaborators. Using =93la=94 instead of =93=
lai=94 in <a href=3D#e3d6>Example 3.6 </a>would give the implication that e=
ach of Tom and Fred, considered individually, had written it.
+<p>
+<a name=3Ds4><h3>Masses and sets</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	le'i	LE	the set described as
+	lo'i	LE	the set of those which really are
+	la'i	LA	the set of those named
+</pre><cx "set, contrasted with mass in attribution of component propertie=
s">  XE "set: contrasted with mass in attribution of component properties" =
  <cx "mass, contrasted with set in attribution of component properties">  =
XE "mass: contrasted with set in attribution of component properties"   <cx=
 "set, compared with mass as abstract of multiple individuals">  XE "set: c=
ompared with mass as abstract of multiple individuals"   <cx "mass, compare=
d with set as abstract of multiple individuals">  XE "mass: compared with s=
et as abstract of multiple individuals"   Having said so much about masses,=
 let us turn to sets. Sets are easier to understand than masses, but are mo=
re rarely used. Like a mass, a set is an abstract object formed from a numb=
er of individuals; however, the properties of a set are not derived from an=
y of the properties of the individuals that compose it.
+<p>
+<cx "inclusion, property of sets">  XE "inclusion: property of sets"   <cx=
 "membership, property of sets">  XE "membership: property of sets"   <cx "=
cardinality, property of sets">  XE "cardinality: property of sets"   <cx "=
cardinality, definition">  XE "cardinality: definition"   <cx "sets, proper=
ties of">  XE "sets: properties of"   <lx "le'i">  XE "le'i"   <lx "lo'i"> =
 XE "lo'i"   <lx "la'i">  XE "la'i"   <cx "le'i, as set counterpart of lei"=
>  XE "le'i: as set counterpart of lei"   <cx "lo'i, as set counterpart of =
loi">  XE "lo'i: as set counterpart of loi"   <cx "la'i, as set counterpart=
 of lai">  XE "la'i: as set counterpart of lai"   Sets have properties like=
 cardinality (how many elements in the set), membership (the relationship b=
etween a set and its elements), and set inclusion (the relationship between=
 two sets, one of which =97 the superset =96 contains all the elements of t=
he other =97 the subset). The set descriptors =93le'i=94, =93lo'i=94 and =
=93la'i=94 correspond exactly to the mass descriptors =93lei=94, =93loi=94,=
 and =93lai=94 except that normally we talk of the whole of a set, not just=
 part of it. Here are some examples contrasting =93lo=94, =93loi=94, and =
=93lo'i=94:
+<p>
+<cx "lo, contrasted with loi and lo'i">  XE "lo: contrasted with loi and l=
o'i"   <cx "loi, contrasted with lo and lo'i">  XE "loi: contrasted with lo=
 and lo'i"   <cx "lo'i, contrasted with lo and loi">  XE "lo'i: contrasted =
with lo and loi"   <cx "rats are brown">  XE "rats are brown: example"   <p=
re><a name=3De4d1>4.1)	lo ratcu cu bunre
+	One-or-more-of-those-which-really-are rats are-brown.
+	Some rats are brown.
+<a name=3De4d2>4.2)	loi ratcu cu cmalu
+	Part-of-the-mass-of-those-which-really-are rats are-small.
+	Rats are small.
+<a name=3De4d3>4.3)	lo'i ratcu cu barda
+	The-set-of rats is-large.
+	There are a lot of rats.
+</pre><p>The mass of rats is small because at least one rat is small; the =
mass of rats is also large; the set of rats, though, is unquestionably larg=
e =96 it has billions of members. The mass of rats is also brown, since som=
e of its components are; but it would be incorrect to call the set of rats =
brown =97 brown-ness is not the sort of property that sets possess.
+<p>
+<cx "sets, use in Lojban place structure">  XE "sets: use in Lojban place =
structure"   Lojban speakers should generally think twice before employing =
the set descriptors. However, certain predicates have places that require s=
et sumti to fill them. For example, the place structure of =93fadni=94 is:
+fadni:  <p>
+<dl compact><dt><dd>x1 is ordinary/common/typical/usual in property x2=20
+among the members of set x3
+</dl><p>Why is it necessary for the x3 place of =93fadni=94 to be a set? B=
ecause it makes no sense for an individual to be typical of another individ=
ual: an individual is typical of a group. In order to make sure that the br=
idi containing =93fadni=94 is about an entire group, its x3 place must be f=
illed with a set:
+<p>
+<ex "typical Lojban user">  XE "typical Lojban user: example=93  <pre><a n=
ame=3De4d4>4.4)	mi fadni zo=92e lo'i lobypli
+	I am-ordinary among the-set-of Lojban-users.
+	I am a typical Lojban user.
+
+</pre><p>Note that the x2 place has been omitted; I am not specifying in e=
xactly which way I am typical =97 whether in language knowledge, or age, or=
 interests, or something else. If =93lo'i=94 were changed to =93lo=94 in <a=
 href=3D#e4d4>Example 4.4</a>, the meaning would be something like =93I am =
typical of some Lojban user=94, which is nonsense.
+<p>
+<a name=3Ds5><h3>Descriptors for typical objects</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	lo'e	LE	the typical
+	le'e	LE	the stereotypical
+</pre><lx "lo'e">  XE "lo'e"   As promised in <a href=3D#s3>Section 3</a>,=
 Lojban has a method for discriminating between =93the lion=94 who lives in=
 Africa and =93the Englishman=94 who, generally speaking, doesn't live in A=
frica even though some Englishmen do. The descriptor =93lo'e=94 means =93th=
e typical=94, as in
+<p>
+<ex "lion in Africa">  XE "lion in Africa: example=93  <pre><a name=3De5d1=
>5.1)	lo'e cinfo cu xabju le fi'ortu'a
+	The-typical lion dwells-in the African-land.
+	Typically, lions dwell in Africa.
+	The lion dwells in Africa.
+</pre><cx "typical objects, determining characteristics of">  XE "typical =
objects: determining characteristics of"   <cx "typical objects, and instan=
tiation">  XE "typical objects: and instantiation"   What is this =93typica=
l lion=94? Surely it is not any particular lion, because no lion has all of=
 the =93typical=94 characteristics, and (worse yet) some characteristics th=
at all real lions have can't be viewed as typical. For example, all real li=
ons are either male or female, but it would be bizarre to suppose that the =
typical lion is either one. So the typical lion has no particular sex, but =
does have a color (golden brown), a residence (Africa), a diet (game), and =
so on. Likewise we can say that
+<p>
+<ex "typical Englishman">  XE "typical Englishman: example=93  <pre><a nam=
e=3De5d2>5.2)	lo'e glipre cu xabju le fi'ortu'a na.e
+		le gligugde
+	The-typical English-person dwells-in the African-land (Not!) and
+		the English-country.
+	The typical English person dwells not in Africa but in England.
+</pre><cx "lo'e, relationship to lo'i">  XE "lo'e: relationship to lo'i"  =
 <cx "lo'i, relationship to lo'e">  XE "lo'i: relationship to lo'e"   <cx "=
le'e, relationship to le'i">  XE "le'e: relationship to le'i"   <cx "le'i, =
relationship to le'e">  XE "le'i: relationship to le'e"   The relationship =
between =93lo'e cinfo=94 and =93lo'i cinfo=94 may be explained thus: the ty=
pical lion is an imaginary lion-abstraction which best exemplifies the set =
of lions. There is a similar relationship between =93le'e=94 and =93le'i=94=
:
+<p>
+<lx "le'e">  XE "le'e"   <cx "stereotypical objects">  XE "stereotypical o=
bjects"   <cx "Greek-Americans own restaurants">  XE "Greek-Americans own r=
estaurants"   <pre><a name=3De5d3>5.3)	le'e xelso merko cu gusta ponse
+	The-stereotypical Greek-type-of American is-a-restaurant-type-of owner.
+	Lots of Greek-Americans own restaurants.
+</pre><cx "stereotypical, as not derogatory in Lojban">  XE "stereotypical=
: as not derogatory in Lojban"   <cx "typical, compared with stereotypical"=
>  XE "typical: compared with stereotypical"   <cx "stereotypical, compared=
 with typical">  XE "stereotypical: compared with typical"   Here we are co=
ncerned not with the actual set of Greek-Americans, but with the set of tho=
se the speaker has in mind, which is typified by one (real or imaginary) wh=
o owns a restaurant. The word =93stereotypical=94 is often derogatory in En=
glish, but =93le'e=94 need not be derogatory in Lojban: it simply suggests =
that the example is typical in the speaker's imagination rather than in som=
e objectively agreed-upon way. Of course, different speakers may disagree a=
bout what the features of =93the typical lion=94 are (some would include ha=
ving a short intestine, whereas others would know nothing of lions' intesti=
nes), so the distinction between =93lo'e cinfo=94 and =93le'e cinfo=94 may =
be very fine.
+
+<p>
+Furthermore,
+<p>
+<ex "Hollywood">  XE "Hollywood: example=93  <pre><a name=3De5d4>5.4)	le'e=
 skina cu se finti ne'i la xali,uyd.
+	The-stereotypical movie is-invented in Hollywood.
+</pre>is probably true to an American, but might be false (not the stereot=
ype) to someone living in India or Russia.
+<p>
+<ex "typical Smith">  XE "typical Smith: example=93  <cx "name equivalent =
for =93typical=94, rationale for lack of">  XE "name equivalent for \=93typ=
ical\=94: rationale for lack of"   Note that there is no naming equivalent =
of =93lo'e=94 and =93le'e=94, because there is no need, as a rule, for a =
=93typical George=94 or a =93typical Smith=94. People or things who share a=
 common name do not, in general, have any other common attributes worth men=
tioning.
+<p>
+<a name=3Ds6><h3>Quantified sumti</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ro	PA	all of/each of
+	su'o	PA	at least (one of)
+</pre><cx "quantifiers, with sumti">  XE =93quantifier: with sumti"   <cx =
"sumti, as having implicit quantifiers">  XE "sumti: as having implicit qua=
ntifiers"   Quantifiers tell us how many: in the case of quantifiers with s=
umti, how many things we are talking about. In Lojban, quantifiers are expr=
essed by numbers and mathematical expressions: a large topic discussed in s=
ome detail in <a href=3Dchap18.html>Chapter 18</a>. For the purposes of thi=
s chapter, a simplified treatment will suffice. Our examples will employ ei=
ther the simple Lojban numbers =93pa=94, =93re=94, =93ci=94, =93vo=94, and =
=93mu=94, meaning =93one=94, =93two=94, =93three=94, =93four=94, =93five=94=
 respectively, or else one of four special quantifiers, two of which are di=
scussed in this section and listed above. These four quantifiers are import=
ant because every Lojban sumti has either one or two of them implicitly pre=
sent in it =97 which one or two depends on the particular kind of sumti. Th=
ere is more explanation of implicit quantifiers later in this section. (The=
 other two quantifiers, =93piro=94 and =93pisu'o=94, are explained in <a hr=
ef=3D#s7>Section 7</a>.)
+<p>
+<cx "sumti, with explicit quantifiers">  XE "sumti: with explicit quantifi=
ers"   <cx "quantifiers, explicit on sumti">  XE =93quantifier: explicit on=
 sumti"   Every Lojban sumti may optionally be preceded by an explicit quan=
tifier. The purpose of this quantifier is to specify how many of the things=
 referred to by the sumti are being talked about. Here are some simple exam=
ples contrasting sumti with and without explicit quantifiers:
+<p>
+<cx "sumti with explicit quantifier, contrasted with sumti without explici=
t quantifier">  XE "sumti with explicit quantifier: contrasted with sumti w=
ithout explicit quantifier"   <pre><a name=3De6d1>6.1)	do cadzu le bisli
+	You walk-on the ice.
+<a name=3De6d2>6.2)	re do cadzu le bisli
+	Two-of you walk-on the ice.
+</pre><cx "number, meaning of when used as a quantifier">  XE "numbers: me=
aning when used as quantifiers"   <cx "quantifiers on sumti, numbers as an =
exact amounts">  XE "quantifier: on sumti, indicating exact number"   The d=
ifference between <a href=3D#e6d1>Example 6.1 </a>and <a href=3D#e6d2>Examp=
le 6.2 </a>is the presence of the explicit quantifier =93re=94 in the latte=
r example. Although =93re=94 by itself means =93two=94, when used as a quan=
tifier it means =93two-of=94. Out of the group of listeners (the number of =
which isn't stated), two (we are not told which ones) are asserted to be =
=93walkers on the ice=94. Implicitly, the others (if any) are not walkers o=
n the ice. In Lojban, you cannot say =93I own three shoes=94 if in fact you=
 own four shoes. Numbers need never be specified, but if they are specified=
 they must be correct.
+<p>
+<cx "quantifiers on sumti, expressing inexact amounts with">  XE "quantifi=
er: on sumti, expressing inexact amount with"   (This rule does not mean th=
at there is no way to specify a number which is vague. The sentence
+<p>
+<pre><a name=3De6d3>6.3)	mi ponse su'o ci cutci
+	I possess at-least three shoes.
+</pre><cx "vague numbers">  XE "vague numbers"   is true if you own three =
shoes, or four, or indeed any larger number. More details on vague numbers =
appear in the discussion of mathematical expressions in <a href=3Dchap18.ht=
ml>Chapter 18</a>.)
+<p>
+Now consider <a href=3D#e6d1>Example 6.1 </a>again. How many of the listen=
ers are claimed to walk on the ice? The answer turns out to be: all of them=
, however many that is. So <a href=3D#e6d1>Example 6.1 </a>and <a href=3D#e=
6d4>Example 6.4</a>:
+<p>
+<pre><a name=3De6d4>6.4)	ro do cadzu le bisli
+	All-of you walk-on the ice.
+</pre><cx "implicit quantifier, on personal pro-sumti">  XE "implicit quan=
tifier: on personal pro-sumti"   <lx "ro">  XE "ro"   <cx "ro, as implicit =
quantifier on personal pro-sumti">  XE "ro: as implicit quantifier on perso=
nal pro-sumti"   <cx "personal pro-sumti, implicit quantifier for">  XE "pe=
rsonal pro-sumti: implicit quantifier for"   <cx "implicit quantifier, defi=
nition">  XE "implicit quantifier: definition"   turn out to mean exactly t=
he same thing. This is a safe strategy, because if one of my listeners does=
n't turn out to be walking on the ice, I can safely claim that I didn't int=
end that person to be a listener! And in fact, all of the personal pro-sumt=
i such as =93mi=94 and =93mi'o=94 and =93ko=94 obey the same rule. We say t=
hat personal pro-sumti have a so-called =93implicit quantifier=94 of =93ro=
=94 (all). This just means that if no quantifier is given explicitly, the m=
eaning is the same as if the implicit quantifier had been used.
+<p>
+<cx "implicit quantifier, on quotations, discussion of">  XE "implicit qua=
ntifier: on quotations, discussion of"   Not all sumti have =93ro=94 as the=
 implicit quantifier, however. Consider the quotation in:
+<p>
+<pre><a name=3De6d5>6.5)	mi cusku lu do cadzu le bisli li'u
+	I express [quote] you walk-on the ice [unquote].
+	I say, =93You walk on the ice.=94
+</pre><p>What is the implicit quantifier of the quotation =93lu do cadzu l=
e bisli li'u=94? Surely not =93ro=94. If =93ro=94 were supplied explicitly,=
 thus:
+<p>
+<pre><a name=3De6d6>6.6)	mi cusku ro lu do cadzu le bisli li'u
+	I express all-of [quote] you walk-on the ice [unquote].
+</pre>the meaning would be something like =93I say every occurrence of the=
 sentence 'You walk on the ice=92=94. Of course I don't say every occurrenc=
e of it, only some occurrences. One might suppose that <a href=3D#e6d5>Exam=
ple 6.5 </a>means that I express exactly one occurrence, but it is more Loj=
banic to leave the number unspecified, as with other sumti. We can say defi=
nitely, however, that I say it at least once.
+<p>
+<lx "su'o">  XE "su'o"   The Lojban cmavo meaning =93at least=94 is =93su'=
o=94, and if no ordinary number follows, =93su'o=94 means =93at least once=
=94. (See <a href=3D#e6d3>Example 6.3 </a>for the use of =93su'o=94 with an=
 ordinary number). Therefore, the explicitly quantified version of <a href=
=3D#e6d5>Example 6.5 </a>is
+<p>
+<cx "su'o, as implicit quantifier for quotations">  XE "su'o: as implicit =
quantifier for quotations"   <cx "quotations, implicit quantifier for">  XE=
 =93quotation: implicit quantifier for"   <cx "implicit quantifier, for quo=
tations">  XE "implicit quantifier: for quotations"   <pre><a name=3De6d7>6=
.7)	mi cusku su'o lu do cadzu le bisli li'u
+	I express at-least-one-of [quote] you walk-on the ice [unquote].
+	I say one or more instances of =93You walk on the ice=94.
+	I say =93You walk on the ice=94.
+</pre><p>If an explicit ordinary number such as =93re=94 were to appear, i=
t would have to convey an exact expression, so
+<p>
+<pre><a name=3De6d8>6.8)	mi cusku re lu do cadzu le bisli li'u
+	I express two-of [quote] you walk-on the ice [unquote].
+</pre>means that I say the sentence exactly twice, neither more nor less.
+
+<p>
+<a name=3Ds7><h3>Quantified descriptions</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	piro	PA	the whole of
+	pisu'o	PA	a part of
+</pre><cx "quantification, before description sumti compared with before n=
on-description sumti">  XE "quantification: before description sumti compar=
ed with before non-description sumti"   Like other sumti, descriptions can =
be quantified. When a quantifier appears before a description, it has the s=
ame meaning as one appearing before a non-description sumti: it specifies h=
ow many things, of all those referred to by the description, are being talk=
ed about in this particular bridi. Suppose that context tells us that =93le=
 gerku=94 refers to three dogs. Then we can say that exactly two of them ar=
e white as follows:
+<p>
+<ex "two dogs are white">  XE "two dogs are white: example=93  <pre><a nam=
e=3De7d1>7.1)	re le gerku cu blabi
+	Two-of the dogs are-white.
+	Two of the dogs are white.
+</pre><cx "outer quantifier, definition">  XE "outer quantifier: definitio=
n"   <cx "inner quantifier, definition">  XE "inner quantifier: definition"=
   <cx "outer quantifier, contrasted with inner quantifier">  XE "outer qua=
ntifier: contrasted with inner quantifier"   <cx "inner quantifier, contras=
ted with outer quantifier">  XE "inner quantifier: contrasted with outer qu=
antifier"   <cx "outer quantifier, effect of on meaning">  XE "outer quanti=
fier: effect of on meaning"   <cx "inner quantifier, effect of on meaning">=
  XE "inner quantifier: effect of on meaning"   When discussing description=
s, this ordinary quantifier is called an =93outer quantifier=94, since it a=
ppears outside the description. But there is another possible location for =
a quantifier: between the descriptor and the selbri. This quantifier is cal=
led an =93inner quantifier=94, and its meaning is quite different: it tells=
 the listener how many objects the description selbri characterizes.
+<p>
+For example, the context of <a href=3D#e7d1>Example 7.1 </a>supposedly tol=
d us that =93le gerku=94 referred to some three specific dogs. This assumpt=
ion can be made certain with the use of an explicit inner quantifier:
+<p>
+<pre><a name=3De7d2>7.2)	re le ci gerku cu blabi
+	Two-of the three dogs are-white.
+	Two of the three dogs are white.
+</pre>(As explained in the discussion of <a href=3D#e6d3>Example 6.3</a>, =
simple numbers like those in <a href=3D#e7d2>Example 7.2 </a>must be exact:=
 it therefore follows that the third dog cannot be white.)
+<p>
+<cx "inner quantifier, explicit">  XE "inner quantifier: explicit"   You m=
ay also specify an explicit inner quantifier and leave the outer quantifier=
 implicit:
+<p>
+<pre><a name=3De7d3>7.3)	le ci gerku cu blabi
+	The three dogs are-white.
+	The three dogs are white.
+</pre><cx "descriptors, implicit quantifiers for">  XE "descriptors: impli=
cit quantifiers for"   <cx "inner quantifier, implicit on descriptors">  XE=
 "inner quantifier: implicit on descriptors"   <cx "outer quantifier, impli=
cit on descriptors">  XE "outer quantifier: implicit on descriptors"   Ther=
e are rules for each of the 11 descriptors specifying what the implicit val=
ues for the inner and outer quantifiers are. They are meant to provide sens=
ible default values when context is absent, not necessarily to prescribe ha=
rd and fast rules. The following table lists the implicit values:
+<p>
+<lx "le">  XE "le"   <lx "ro">  XE "ro"   <lx "su'o">  XE "su'o"   <dl com=
pact><dt>le:	<dd>ro le su'o
+		all of the at-least-one described as
+</dl><lx "lo">  XE "lo"   <dl compact><dt>lo:	<dd>su'o lo ro
+		at least one of all of those which really are
+</dl><lx "la">  XE "la"   <dl compact><dt>la:	<dd>ro la su'o
+		all of the at least one named
+
+</dl><lx "lei">  XE "lei"   <dl compact><dt>lei:	<dd>pisu'o lei su'o
+		some part of the mass of the at-least-one described as
+</dl><lx "loi">  XE "loi"   <dl compact><dt>loi:	<dd>pisu'o loi ro
+		some part of the mass of all those that really are
+</dl><lx "lai">  XE "lai"   <dl compact><dt>lai:	<dd>pisu'o lai su'o
+		some part of the mass of the at-least-one named
+</dl><lx "le'i">  XE "le'i"   <dl compact><dt>le'i:	<dd>piro le'i su'o
+		the whole of the set of the at-least-one described as
+</dl><lx "lo'i">  XE "lo'i"   <dl compact><dt>lo'i:	<dd>piro lo'i ro
+		the whole of the set of all those that really are
+</dl><lx "la'i">  XE "la'i"   <dl compact><dt>la'i:	<dd>piro la'i su'o
+		the whole of the set of the at-least-one named
+</dl><lx "le'e">  XE "le'e"   <dl compact><dt>le'e:	<dd>ro le'e su'o
+		all the stereotypes of the at-least-one described as
+</dl><lx "lo'e">  XE "lo'e"   <dl compact><dt>lo'e:	<dd>su'o lo'e ro
+		at least one of the types of all those that really are
+</dl><cx "le-series descriptors, compared with la-series in implicit quant=
ification">  XE "le-series descriptors: compared with la-series in implicit=
 quantification"   <cx "la-series descriptors, compared with le-series in i=
mplicit quantification">  XE "la-series descriptors: compared with le-serie=
s in implicit quantification"   <cx "le-series cmavo, definition">  XE "le-=
series cmavo: definition"   <cx "le-series cmavo, as encompassing le-series=
 and la-series descriptors for quantification discussion">  XE "le-series c=
mavo: as encompassing le-series and la-series descriptors for quantificatio=
n discussion"   When examined for the first time, this table looks dreadful=
ly arbitrary. In fact, there are quite a few regularities in it. First of a=
ll, the la-series (that is, the descriptors =93la=94, =93lai=94, and =93la'=
i=94) and the le-series (that is, the descriptors =93le=94, =93lei=94, =93l=
e'i=94, and =93le'e=94) always have corresponding implicit quantifiers, so =
we may subsume the la-series under the le-series for the rest of this discu=
ssion: =93le-series cmavo=94 will refer to both the le-series proper and to=
 the la-series.
+<p>
+<cx "lo-series cmavo, rule for implicit inner quantifier">  XE "lo-series =
cmavo: rule for implicit inner quantifier"   <cx "le-series cmavo, rule for=
 implicit inner quantifier">  XE "le-series cmavo: rule for implicit inner =
quantifier"   The rule for the inner quantifier is very simple: the lo-seri=
es cmavo (namely, =93lo=94, =93loi=94, =93lo'i=94, and =93lo'e=94) all have=
 an implicit inner quantifier of =93ro=94, whereas the le-series cmavo all =
have an implicit inner quantifier of =93su'o=94.
+<p>
+<cx "lo-series cmavo, rationale for implicit inner quantifier">  XE "lo-se=
ries cmavo: rationale for implicit inner quantifier"   <cx "le-series cmavo=
, rationale for implicit inner quantifier">  XE "le-series cmavo: rationale=
 for implicit inner quantifier"   Why? Because lo-series descriptors always=
 refer to all of the things which really fit into the x1 place of the selbr=
i. They are not restricted by the speaker's intention. Descriptors of the l=
e-series, however, are so restricted, and therefore talk about some number,=
 definite or indefinite, of objects the speaker has in mind =97 but never l=
ess than one.
+<p>
+<cx "sets, rule for implicit outer quantifier">  XE "sets: rule for implic=
it outer quantifier"   <cx "masses, rule for implicit outer quantifier">  X=
E "mass: rule for implicit outer quantifier"   <lx "piro">  XE "piro"   Und=
erstanding the implicit outer quantifier requires rules of greater subtlety=
. In the case of mass and set descriptors, a single rule suffices for each:=
 reference to a mass is implicitly a reference to some part of the mass; re=
ference to a set is implicitly a reference to the whole set. Masses and set=
s are inherently singular objects: it makes no sense to talk about two dist=
inct masses with the same components, or two distinct sets with the same me=
mbers. Therefore, the largest possible outer quantifier for either a set de=
scription or a mass description is =93piro=94, the whole of it.
+<p>
+<cx "plural masses, possible use for">  XE "plural masses: possible use fo=
r"   (Pedantically, it is possible that the mass of water molecules composi=
ng an ice cube might be thought of as different from the same mass of water=
 molecules in liquid form, in which case we might talk about =93re lei djac=
u=94, two masses of the water-bits I have in mind.)
+<p>
+<lx "pi">  XE "pi"   <cx "piro, explanation of meaning">  XE "piro: explan=
ation of meaning"   <cx "pisu'o, explanation of meaning">  XE "pisu'o: expl=
anation of meaning"   Why =93pi-=94? It is the Lojban cmavo for the decimal=
 point. Just as =93pimu=94 means =93.5=94, and when used as a quantifier sp=
ecifies a portion consisting of five tenths of a thing, =93piro=94 means a =
portion consisting of the all-ness =96 the entirety =97 of a thing. Similar=
ly, =93pisu'o=94 specifies a portion consisting of at least one part of a t=
hing, i.e. some of it.
+<p>
+<cx "subsets, expressing with outer quantifiers">  XE "subsets: expressing=
 with outer quantifiers"   <cx "outer quantifiers, for expressing subsets">=
  XE "outer quantifier: for expressing subset"   <cx "portion, on set contr=
asted with on individual">  XE "portion: on set contrasted with on individu=
al"   Smaller quantifiers are possible for sets, and refer to subsets. Thus=
 =93pimu le'i nanmu=94 is a subset of the set of men I have in mind; we don=
't know precisely which elements make up this subset, but it must have half=
 the size of the full set. This is the best way to say =93half of the men=
=94; saying =93pimu le nanmu=94 would give us a half-portion of one of them=
 instead! Of course, the result of =93pimu le'i nanmu=94 is still a set; if=
 you need to refer to the individuals of the subset, you must say so (see =
=93lu'a=94 in <a href=3D#s10>Section 10</a>).
+<p>
+<cx "outer quantifiers on individual descriptors, rationale for difference=
s in implicit quantifier">  XE "outer quantifier: rationale for differences=
 in implicit quantifier on descriptors"   <cx "individual descriptors, diff=
erent implicit outer quantifiers among">  XE "descriptors: different implic=
it outer quantifiers among"   <cx "le, implicit outer quantifier for">  XE =
"le: implicit outer quantifier for"   <cx "lo, implicit outer quantifier fo=
r">  XE "lo: implicit outer quantifier for"   The case of outer quantifiers=
 for individual descriptors (including =93le=94, =93lo=94, =93la=94, and th=
e typical descriptors =93le'e=94 and =93lo'e=94) is special. When we refer =
to specific individuals with =93le=94, we mean to refer to all of those we =
have in mind, so =93ro=94 is appropriate as the implicit quantifier, just a=
s it is appropriate for =93do=94. Reference to non-specific individuals wit=
h =93lo=94, however, is typically to only some of the objects which can be =
correctly described, and so =93su'o=94 is the appropriate implicit quantifi=
er, just as for quotations.
+<p>
+<cx "le, contrasted with lo in implicit quantification">  XE "le: contrast=
ed with lo in implicit quantification"   <cx "lo, contrasted with le in imp=
licit quantification">  XE "lo: contrasted with le in implicit quantificati=
on"   From the English-speaking point of view, the difference in structure =
between the following example using =93le=94:
+<p>
+<pre><a name=3De7d4>7.4)	[ro] le ci gerku cu blabi
+	[All-of] those-described-as three dogs are-white.
+	The three dogs are white.
+</pre>and the corresponding form with =93lo=94:
+<p>
+<pre><a name=3De7d5>7.5)	ci lo [ro] gerku cu blabi
+	Three-of those-which-are [all] dogs are-white
+	Three dogs are white.
+</pre>looks very peculiar. Why is the number =93ci=94 found as an inner qu=
antifier in <a href=3D#e7d4>Example 7.4 </a>and as an outer quantifier in <=
a href=3D#e7d5>Example 7.5</a>? The number of dogs is the same in either ca=
se. The answer is that the =93ci=94 in <a href=3D#e7d4>Example 7.4 </a>is p=
art of the specification: it tells us the actual number of dogs in the grou=
p that the speaker has in mind. In <a href=3D#e7d5>Example 7.5</a>, however=
, the dogs referred to by =93<dots>=85 </dots>lo gerku=94 are all the dogs =
that exist: the outer quantifier then restricts the number to three; which =
three, we cannot tell. The implicit quantifiers are chosen to avoid claimin=
g too much or too little: in the case of =93le=94, the implicit outer quant=
ifier =93ro=94 says that each of the dogs in the restricted group is white;=
 in the case of =93lo=94, the implicit inner quantifier simply says that th=
ree dogs, chosen from the group of all the dogs there are, are white.
+<p>
+<cx "lo-series description, caution on exact numbers as inner quantifiers =
on">  XE "lo-series description: caution on exact numbers as inner quantifi=
ers on"   Using exact numbers as inner quantifiers in lo-series description=
s is dangerous, because you are stating that exactly that many things exist=
 which really fit the description. So examples like
+<p>
+<pre><a name=3De7d67>7.6)	[su'o] lo ci gerku cu blabi
+	[Some-of] those-which-really-are three dogs are-white
+</pre>are semantically anomalous; <a href=3D#e7d7>Example 7.6 </a>claims t=
hat some dog (or dogs) is white, but also that there are just three dogs in=
 the universe!
+<p>
+Nevertheless, inner quantifiers are permitted on =93lo=94 descriptors for =
consistency's sake, and may occasionally be useful.
+<p>
+Note that the inner quantifier of =93le=94, even when exact, need not be t=
ruthful: =93le ci nanmu=94 means =93what I describe as three men=94, not =
=93three of what I describe as men=94. This follows from the rule that what=
 is described by a =93le=94 description represents the speaker's viewpoint =
rather than the objective way things are.
+<p>
+<a name=3Ds8><h3>Indefinite descriptions</h3>
+<p>
+<cx "lo, omission of">  XE "lo: omission of"   <cx "descriptors, omission =
of">  XE "descriptors: omission of"   By a quirk of Lojban syntax, it is po=
ssible to omit the descriptor =93lo=94, but never any other descriptor, fro=
m a description like that of <a href=3D#e7d5>Example 7.5</a>; namely, one w=
hich has an explicit outer quantifier but no explicit inner quantifier. The=
 following example:
+<p>
+<pre><a name=3De8d1>8.1)	ci gerku [ku] cu blabi
+	Three dogs are white.
+</pre><cx "ku, effect on of omitting descriptor">  XE "ku: effect on of om=
itting descriptor"   <cx "omission of descriptor, effect on ku">  XE "omiss=
ion of descriptor: effect on ku"   <cx "indefinite description, definition"=
>  XE "indefinite description: definition"   is equivalent in meaning to <a=
 href=3D#e7d5>Example 7.5</a>. Even though the descriptor is not present, t=
he elidable terminator =93ku=94 may still be used. The name =93indefinite d=
escription=94 for this syntactic form is historically based: of course, it =
is no more and no less indefinite than its counterpart with an explicit =93=
lo=94. Indefinite descriptions were introduced into the language in order t=
o imitate the syntax of English and other natural languages.
+<p>
+<cx "indefinite description, as needing explicit outer quantifier">  XE "i=
ndefinite description: as needing explicit outer quantifier"   <cx "indefin=
ite description, as prohibiting explicit inner quantifier">  XE "indefinite=
 description: as prohibiting explicit inner quantifier"   <cx "outer quanti=
fier, in indefinite description">  XE "outer quantifier: in indefinite desc=
ription"   <cx "inner quantifier, in indefinite description">  XE "inner qu=
antifier: in indefinite description"   Indefinite descriptions must fit thi=
s mold exactly: there is no way to make one which does not have an explicit=
 outer quantifier (thus =93*gerku cu blabi=94 is ungrammatical), or which h=
as an explicit inner quantifier (thus =93*reboi ci gerku cu blabi=94 is als=
o ungrammatical =97 =93re ci gerku cu blabi=94 is fine, but means =9323 dog=
s are white=94).
+<p>
+Note: <a href=3D#e6d3>Example 6.3 </a>also contains an indefinite descript=
ion, namely =93su'o ci cutci=94; another version of that example using an e=
xplicit =93lo=94 would be:
+<p>
+<pre><a name=3De8d2>8.2)	mi ponse su'o ci lo cutci
+	I possess at-least three things-which-really-are shoes
+	I own three (or more) shoes.
+</pre><a name=3Ds9><h3>sumti-based descriptions</h3>
+<p>
+As stated in <a href=3D#s2>Section 2</a>, most descriptions consist of jus=
t a descriptor and a selbri. (In this chapter, the selbri have always been =
single gismu, but of course any selbri, however complex, can be employed in=
 a description. The syntax and semantics of selbri are explained in <a href=
=3Dchap5.html>Chapter 5</a>.) In the intervening sections, inner and outer =
quantifiers have been added to the syntax. Now it is time to discuss a desc=
ription of a radically different kind: the sumti-based description.
+<p>
+<cx "sumti-based description, definition">  XE "sumti-based description: d=
efinition"   <cx "sumti-based description, inner quantifier on">  XE "sumti=
-based description: inner quantifier on"   <cx "sumti-based description, ou=
ter quantifier on">  XE "sumti-based description: outer quantifier on"   A =
sumti-based description has a sumti where the selbri would normally be, and=
 the inner quantifier is required =97 it cannot be implicit. An outer quant=
ifier is permitted but not required.
+<p>
+A full theory of sumti-based descriptions has yet to be worked out. One co=
mmon case, however, is well understood. Compare the following:
+<p>
+<pre><a name=3De9d1>9.1)	re do cu nanmu
+	Two-of you are-men.
+</pre><ex "the two of you">  XE "the two of you: example=93  <pre><a name=
=3De9d2>9.2)	le re do cu nanmu
+	The two-of you are men.
+</pre><a href=3D#e9d1>Example 9.1 </a>simply specifies that of the group o=
f listeners, size unknown, two are men. <a href=3D#e9d2>Example 9.2</a>, wh=
ich has the sumti-based description =93le re do=94, says that of the two li=
steners, all (the implicit outer quantifier =93ro=94) are men. So in effect=
 the inner quantifier =93re=94 gives the number of individuals which the in=
ner sumti =93do=94 refers to.
+
+<p>
+Here is another group of examples:
+<p>
+<ex "three bears">  XE "three bears: example=93  <pre><a name=3De9d3>9.3)	=
re le ci cribe cu bunre
+	Two-of the three bears are-brown.
+<a name=3De9d4>9.4)	le re le ci cribe cu bunre
+The two-of the three bears are-brown.
+<a name=3De9d5>9.5)	pa le re le ci cribe cu bunre
+	One-of the two-of the three bears are-brown.
+</pre><cx "sumti-based descriptions with le, as increasing restricting to =
in-mind">  XE "sumti-based descriptions with le: as increasing restricting =
to in-mind"   In each case, =93le ci cribe=94 restricts the bears (or alleg=
ed bears) being talked of to some group of three which the speaker has in m=
ind. <a href=3D#e9d3>Example 9.3 </a>says that two of them (which two is no=
t stated) are brown. <a href=3D#e9d4>Example 9.4 </a>says that a specific p=
air of them are brown. <a href=3D#e9d5>Example 9.5 </a>says that of a speci=
fic pair chosen from the original three, one or the other of that pair is b=
rown.
+<p>
+<a name=3Ds10><h3>sumti qualifiers</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<cx "sumti qualifiers, list of">  XE "sumti qualifiers: list of"   <pre>	l=
a'e		LAhE	something referred to by
+	lu'e		LAhE	a reference to
+	tu'a	LAhE	an abstraction involving
+	lu'a		LAhE	an individual/member/component of
+	lu'i		LAhE	a set formed from
+	lu'o		LAhE	a mass formed from
+	vu'i		LAhE	a sequence formed from
+	na'ebo	NAhE+BO	something other than
+	to'ebo	NAhE+BO	the opposite of
+	no'ebo	NAhE+BO	the neutral form of
+	je'abo	NAhE+BO	that which indeed is
+	lu'u		LUhU	elidable terminator for LAhE and NAhE+BO
+</pre><p>Well, that's quite a list of cmavo. What are they all about?
+<p>
+<cx "LAhE">  XE "LAhE selma=92o"   <cx "NAhE">  XE "NAhE selma=92o"   <lx =
"BO">  XE "BO selma=92o"   <cx "sumti qualifiers, internal syntax of">  XE =
"sumti qualifiers: internal syntax of"   <cx "sumti qualifiers, external sy=
ntax of">  XE "sumti qualifiers: external syntax of"   <cx "sumti qualifier=
s, elidable terminator for qualified sumti">  XE "sumti qualifiers: elidabl=
e terminator for qualified sumti"   <lx "lu'u">  XE "lu'u"   <cx "lu'u, as =
elidable terminator for qualified sumti">  XE "lu'u: as elidable terminator=
 for qualified sumti"   The above cmavo and compound cmavo are called the =
=93sumti qualifiers=94. All of them are either single cmavo of selma'o LAhE=
, or else compound cmavo involving a scalar negation cmavo of selma'o NAhE =
immediately followed by =93bo=94 of selma'o BO. Syntactically, you can pref=
ix a sumti qualifier to any sumti and produce another simple sumti. (You ma=
y need to add the elidable terminator =93lu'u=94 to show where the qualifie=
d sumti ends.)
+<p>
+<cx "sumti qualifiers, as short forms for common special cases">  XE "sumt=
i qualifiers: as short forms for common special cases"   Semantically, sumt=
i qualifiers represent short forms of certain common special cases. Suppose=
 you want to say =93I see 'The Red Pony=92=94, where =93The Red Pony=94 is =
the title of a book. How about:
+<p>
+<cx "qualified sumti, contrasted with unqualified sumti">  XE "qualified s=
umti: contrasted with unqualified sumti"   <cx "unqualified sumti, contrast=
ed with qualified sumti">  XE "unqualified sumti: contrasted with qualified=
 sumti"   <ex "Red Pony">  XE "Red Pony: example=93  <pre><a name=3De10d1>1=
0.1)	mi viska lu le xunre cmaxirma li'u
+	I see [quote] the red small-horse [unquote]
+</pre><p>But <a href=3D#e10d1>Example 10.1 </a>doesn't work: it says that =
you see a piece of text =93The Red Pony=94. That might be all right if you =
were looking at the cover of the book, where the words =93The Red Pony=94 a=
re presumably written. (More precisely, where the words =93le xunre cmaxirm=
a=94 are written =96 but we may suppose the book has been translated into L=
ojban.)
+<p>
+What you really want to say is:
+<p>
+<pre><a name=3De10d2>10.2)	mi viska le selsinxa be lu le xunre cmaxirma li=
'u
+	I see the thing-represented-by [quote] the red small-horse [unquote]
+</pre><p>The x2 place of =93selsinxa=94 (the x1 place of =93sinxa=94) is a=
 sign or symbol, and the x1 place of =93selsinxa=94 (the x2 place of =93sin=
xa=94) is the thing represented by the sign. <a href=3D#e10d2>Example 10.2 =
</a>allows us to use a symbol (namely the title of a book) to represent the=
 thing it is a symbol of (namely the book itself).
+<p>
+This operation turns out to be needed often enough that it's useful to be =
able to say:
+<p>
+<cx "la'e, as short for =93le selsinxa be=94">  XE "la'e: as short for \=
=93le selsinxa be\=94"   <pre><a name=3De10d3>10.3)	mi viska la'e lu le xun=
re cmaxirma li'u [lu'u]
+	I see the-referent-of [quote] the red small-horse [unquote].
+</pre><cx "dereferencing a pointer, with la'e">  XE "dereferencing a point=
er: with la'e"   <cx "la'e, effect of on meaning">  XE "la'e: effect of on =
meaning"   <cx "referent, referring to with la'e">  XE "referent: referring=
 to with la'e"   <lx "la'e">  XE "la'e"   So when =93la'e=94 is prefixed to=
 a sumti referring to a symbol, it produces a sumti referring to the refere=
nt of that symbol. (In computer jargon, =93la'e=94 dereferences a pointer.)
+<p>
+By introducing a sumti qualifier, we correct a false sentence (<a href=3D#=
e10d1>Example 10.1</a>), which too closely resembles its literal English eq=
uivalent, into a true sentence (<a href=3D#e10d3>Example 10.3</a>), without=
 having to change it overmuch; in particular, the structure remains the sam=
e. Most of the uses of sumti qualifiers are of this general kind.
+<p>
+<lx "lu'e">  XE "lu'e"   <cx "lu'e, effect of on meaning">  XE "lu'e: effe=
ct of on meaning"   <cx "symbol, referring to with lu'e">  XE "symbol: refe=
rring to with lu'e"   The sumti qualifier =93lu'e=94 provides the converse =
operation: it can be prefixed to a sumti referring to some thing to produce=
 a sumti referring to a sign or symbol for the thing. For example,
+<p>
+<cx "lu'e, as short for =93le sinxa be=94">  XE "lu'e: as short for \=93le=
 sinxa be\=94"   <ex "title of book">  XE "title of book: example=93  <pre>=
<a name=3De10d4>10.4)	mi pu cusku lu'e le vi cukta
+	I [past] express a-symbol-for the nearby book.
+	I said the title of this book.
+</pre><p>The equivalent form not using a sumti qualifier would be:
+<p>
+<pre><a name=3De10d5>10.5)	mi pu cusku le sinxa be le vi cukta
+	I [past] express the symbol-for the nearby book.
+</pre>which is equivalent to <a href=3D#e10d4>Example 10.4</a>, but longer=
.
+<p>
+<lx "tu'a">  XE "tu'a"   <lx "lu'a">  XE "lu'a"   <lx "lu'i">  XE "lu'i"  =
 <lx "lu'o">  XE "lu'o"   <lx "vu'i">  XE "vu'i"   <cx "tu'a, effect of on =
meaning">  XE "tu'a: effect of on meaning"   <cx "lu'a, effect of on meanin=
g">  XE "lu'a: effect of on meaning"   <cx "lu'i, effect of on meaning">  X=
E "lu'i: effect of on meaning"   <cx "lu'o, effect of on meaning">  XE "lu'=
o: effect of on meaning"   <cx "vu'i, effect of on meaning">  XE "vu'i: eff=
ect of on meaning"   <cx "tu'a, use for forming abstractions">  XE "tu'a: u=
se for forming abstractions"   <cx "vu'i, use for creating sequence">  XE "=
vu'i: use for creating sequence"   <cx "sequence, contrasted with set">  XE=
 "sequence: contrasted with set"   The other sumti qualifiers follow the sa=
me rules. The cmavo =93tu'a=94 is used in forming abstractions, and is expl=
ained more fully in <a href=3Dchap11.html>Chapter 11</a>. The triplet =93lu=
'a=94, =93lu'i=94, and =93lu'o=94 convert between individuals, masses, and =
sets; =93vu'i=94 belongs to this group as well, but creates a sequence, whi=
ch is similar to a set but has a definite order. (The set of John and Charl=
es is the same as the set of Charles and John, but the sequences are differ=
ent.) Here are some examples:
+<p>
+<pre><a name=3De10d6>10.6)	mi troci tu'a le vorme
+	I try some-abstraction-about the door.
+	I try (to open) the door.
+</pre><cx "tu'a, as being deliberately vague">  XE "tu'a: as being deliber=
ately vague"   <a href=3D#e10d6>Example 10.6 </a>might mean that I try to d=
o something else involving the door; the form is deliberately vague.
+<p>
+Most of the following examples make use of the cmavo =93ri=94, belonging t=
o selma'o KOhA. This cmavo means =93the thing last mentioned=94; it is equi=
valent to repeating the immediately previous sumti (but in its original con=
text). It is explained in more detail in <a href=3Dchap7.html>Chapter 7</a>=
.
+
+<p>
+<ex "set of rats">  XE "set of rats: example=93  <pre><a name=3De10d7>10.7=
)	lo'i ratcu cu barda  .iku'i lu'a ri cmalu
+	The-set-of rats is-large.  But some-members-of it-last-mentioned is-small=
.
+	The set of rats is large, but some of its members are small.
+<a name=3De10d8>10.8)	lo ratcu cu cmalu  .iku'i lu'i ri barda
+	Some rats are-small.  But the-set-of them-last-mentioned is-large.
+	Some rats are small, but the set of rats is large.
+<a name=3De10d9>10.9)	mi ce do girzu
+		.i lu'o ri gunma
+		.i vu'i ri porsi
+	I in-a-set-with you are-a-set.
+		The-mass-of it-last-mentioned is-a-mass.
+		The-sequence-of it-last-mentioned is-a-sequence
+	The set of you and me is a set.
+		The mass of you and me is a mass.
+		The sequence of you and me is a sequence.
+</pre>(Yes, I know these examples are a bit silly. This set was introduced=
 for completeness, and practical examples are as yet hard to come by.)
+<p>
+<cx "sumti qualifiers, for negation">  XE "sumti qualifiers: for negation"=
   <cx "negation sumti qualifiers, meanings of">  XE "negation sumti qualif=
iers: meanings of"   Finally, the four sumti qualifiers formed from a cmavo=
 of NAhE and =93bo=94 are all concerned with negation, which is discussed i=
n detail in <a href=3Dchap15.html>Chapter 15</a>. Here are a few examples o=
f negation sumti qualifiers:
+<p>
+<lx "na'ebo">  XE "na'ebo"   <pre><a name=3De10d10>10.10)	mi viska na'ebo =
le gerku
+	I see something-other-than the dog.
+</pre><p>This compound, =93na'ebo=94, is the most common of the four negat=
ion sumti qualifiers. The others usually only make sense in the context of =
repeating, with modifications, something already referred to:
+<p>
+<ex "lukewarm food">  XE "lukewarm food: example=93  <pre><a name=3De10d11=
>10.11)	mi nelci loi glare cidja
+		.ije do nelci to'ebo ri
+		.ije la djein. nelci no'ebo ra
+	I like part-of-the-mass-of hot-type-of food.
+		And you like the-opposite-of the-last-mentioned.
+		And Jane likes the-neutral-value-of something-mentioned.
+	I like hot food, and you like cold food, and Jane likes lukewarm food.
+</pre>(In <a href=3D#e10d11>Example 10.11</a>, the sumti =93ra=94 refers t=
o some previously mentioned sumti other than that referred to by =93ri=94. =
We cannot use =93ri=94 here, because it would signify =93la djein.=94, that=
 being the most recent sumti available to =93ri=94. See more detailed expla=
nations in <a href=3Dchap7.html>Chapter 7</a>.)
+<p>
+<a name=3Ds11><h3>The syntax of vocative phrases</h3>
+<p>
+<cx "vocative phrases, as a =93free modifier=94">  XE "vocative phrase: as=
 a free modifier"   Vocative phrases are not sumti, but are explained in th=
is chapter because their syntax is very similar to that of sumti. Grammatic=
ally, a vocative phrase is one of the so-called =93free modifiers=94 of Loj=
ban, along with subscripts, parentheses, and various other constructs expla=
ined in <a href=3Dchap19.html>Chapter 19</a>. They can be placed after many=
, but not all, constructions of the grammar: in general, after any elidable=
 terminator (which, however, must not then be elided!), at the beginnings a=
nd ends of sentences, and in many other places.
+<p>
+<cx "vocative phrase, purpose of">  XE "vocative phrase: purpose of"   <lx=
 "COI">  XE "COI selma=92o"   <lx "DOI">  XE "DOI selma=92o"   The purpose =
of a vocative phrase is to indicate who is being addressed, or to indicate =
to that person that he or she ought to be listening. A vocative phrase begi=
ns with a cmavo of selma'o COI or DOI, all of which are explained in more d=
etail in <a href=3Dchap13.html>Chapter 13</a>. Sometimes that is all there =
is to the phrase:
+<p>
+<pre><a name=3De11d1>11.1)	coi
+	[greetings]
+	Hello.
+<a name=3De11d2>11.2)	je'e
+	[acknowledgement]
+	Uh-huh.
+	Roger!
+</pre><cx "vocative word, phrase following">In these cases, the person bei=
ng addressed is obvious from the context. However, a vocative word (more pr=
ecisely, one or more cmavo of COI, possibly followed by =93doi=94, or else =
just =93doi=94 by itself) can be followed by one of several kinds of phrase=
s, all of which are intended to indicate the addressee. The most common cas=
e is a name:
+<p>
+<lx "coi">  XE "coi"   <pre><a name=3De11d2>11.2)	coi. djan.
+	Hello, John.
+</pre><p>A pause is required (for morphological reasons) between a member =
of COI and a name. You can use =93doi=94 instead of a pause:
+<p>
+<lx "doi">  XE "doi"   <pre><a name=3De11d3>11.3)	coi doi djan.
+	Hello, John.
+</pre>means exactly the same thing and does not require a pause. Using =93=
doi=94 by itself is like just saying someone's name to attract his or her a=
ttention:
+<p>
+<pre><a name=3De11d4>11.4)	doi djan.
+	John!
+</pre><cx "vocative phrase, forms of">  XE "vocative phrase: forms of"   <=
cx "vocative phrase, with sumti without descriptor">  XE "vocative phrase: =
with sumti without descriptor"   <cx "vocative phrase, implicit descriptor =
on">  XE "vocative phrase: implicit descriptor on"   In place of a name, a =
description may appear, lacking its descriptor, which is understood to be =
=93le=94:
+<p>
+<pre><a name=3De11d5>11.5)	coi xunre pastu nixli
+	Hello, (red-type-of dress)-type-of girl.
+	Hello, girl with the red dress!
+</pre><cx "vocative phrase, implicit quantifiers on">  XE "vocative phrase=
: implicit quantifiers on"   <cx "vocative phrase, explicit quantifiers pro=
hibited on">  XE "vocative phrase: explicit quantifiers prohibited on"   Th=
e listener need not really be a =93xunre pastu nixli=94, as long as she und=
erstands herself correctly from the description. (Actually, only a bare sel=
bri can appear; explicit quantifiers are forbidden in this form of vocative=
, so the implicit quantifiers =93su'o le ro=94 are in effect.)
+<p>
+Finally, a complete sumti may be used, the most general case.
+<p>
+<cx "vocative phrase, with complete sumti">  XE "vocative phrase: with com=
plete sumti"   <pre><a name=3De11d6>11.6)	co'o la bab. .e la noras.
+	Goodbye, Bob and Nora.
+</pre><a href=3D#e11d5>Example 11.5 </a>is thus the same as:
+<p>
+<pre><a name=3De11d7>11.7)	coi le xunre pastu nixli
+	Hello, the-one-described-as red-dress girl!
+</pre>and <a href=3D#e11d4>Example 11.4 </a>is the same as:
+<p>
+<pre><a name=3De11d8>11.8)	doi la djan.
+	The-one-named John!
+</pre><cx "vocative phrase, elidable terminator for">  XE "vocative phrase=
: elidable terminator for"   <lx "do'u">  XE "do'u"   <lx "DOhU">  XE "DOhU=
 selma=92o"   <cx "vocative phrase terminator, elidability of">  XE "vocati=
ve phrase terminator: elidability of"   Finally, the elidable terminator fo=
r vocative phrases is =93do'u=94 (of selma'o DOhU), which is rarely needed =
except when a simple vocative word is being placed somewhere within a bridi=
. It may also be required when a vocative is placed between a sumti and its=
 relative clause, or when there are a sequence of so-called =93free modifie=
rs=94 (vocatives, subscripts, utterance ordinals =97 see <a href=3Dchap18.h=
tml>Chapter 18 </a>=97 metalinguistic comments =97 see <a href=3Dchap19.htm=
l>Chapter 19 </a>=97 or reciprocals =97 see <a href=3Dchap19.html>Chapter 1=
9</a>) which must be properly separated.
+<p>
+<cx "vocative phrase, effect of position on meaning">  XE "vocative phrase=
: effect of position on meaning"   The meaning of a vocative phrase that is=
 within a sentence is not affected by its position in the sentence: thus <a=
 href=3D#e11d9>Example 11.9 </a>and <a href=3D#e11d10>Example 11.10 </a>mea=
n the same thing:
+<p>
+<pre><a name=3De11d9>11.9)	doi djan. ko klama mi
+	John, come to me!
+<a name=3De11d10>11.10)	ko klama mi doi djan.
+	Come to me, John!
+</pre><p>As usual for this chapter, the full syntax of vocative phrases ha=
s not been explained: relative clauses, discussed in <a href=3Dchap8.html>C=
hapter 8</a>, make for more possibilities.
+<p>
+<a name=3Ds12><h3>Lojban names</h3>
+<p>
+Names have been used freely as sumti throughout this chapter without too m=
uch explanation. The time for the explanation has now come.
+<p>
+<cx "names, two kinds of">  XE "names: two kinds of"   <cx "name words, re=
cognition of">  XE "name words: recognition of"   First of all, there are t=
wo different kinds of things usually called =93names=94 when talking about =
Lojban. The naming predicates of <a href=3D#s2>Section 2 </a>are just ordin=
ary predicates which are being used in a special sense. In addition, though=
, there is a class of Lojban words which are used only to name things: thes=
e can be recognized by the fact that they end in a consonant followed by a =
pause. Some examples:
+<p>
+<pre><a name=3De12d1>12.1)	djan.	meris.	djein.	.alis.
+	John.	Mary.	Jane.	Alice.
+</pre>(Note that =93.alis.=94 begins as well as ends with a pause, because=
 all Lojban words beginning with a vowel must be preceded by a pause. See <=
a href=3Dchap4.html>Chapter 4 </a>for more information.)
+<p>
+<cx "names, uses of">  XE "names: uses of"   <cx "names, in vocative phras=
e">  XE "names: in vocative phrase"   <cx "names, with LA descriptor">  XE =
"names: with LA descriptor"   Names of this kind have two basic uses in Loj=
ban: when used in a vocative phrase (see <a href=3D#s11>Section 11</a>) the=
y indicate who the listener is or should be. When used with a descriptor of=
 selma'o LA, namely =93la=94, =93lai=94, or =93la'i=94, they form sumti whi=
ch refer to the persons or things known by the name.
+<p>
+<lx "la">  XE "la"   <pre><a name=3De12d2>12.2)	la djonz. klama le zarci
+	Jones goes to-the store.
+	The Joneses go to-the store.
+</pre><lx "lai">  XE "lai"   <pre><a name=3De12d3>12.3)	lai djonz. klama l=
e zarci
+	The-mass-of Joneses go to-the store.
+	The Joneses go to the store.
+</pre><p>In <a href=3D#e12d2>Example 12.2</a>, the significance is that al=
l the persons (perhaps only one) I mean to refer to by the name =93djonz.=
=94 are going to the store. In <a href=3D#e12d3>Example 12.3</a>, the Jones=
es are massified, and only some part of them needs to be going. Of course, =
by =93djonz.=94 I can mean whomever I want: that person need not use the na=
me =93djonz.=94 at all.
+<p>
+<cx "LE, contrasted with LA in use of name-words">  XE "LE selma=92o: cont=
rasted with LA in use of name-words"   <cx "LA, contrasted with LE in use o=
f name-words">  XE "LA selma=92o: contrasted with LE in use of name-words" =
  The sumti in <a href=3D#e12d2>Example 12.2 </a>and <a href=3D#e12d3>Examp=
le 12.3 </a>operate exactly like the similar uses of =93la=94 and =93lai=94=
 in <a href=3D#e2d{5}>Examples 2.5 </a>and <a href=3D#e3d6>3.6 </a>respecti=
vely. The only difference is that these descriptors are followed by Lojban =
name-words. And in fact, the only difference between descriptors of selma'o=
 LA (these three) and of selma'o LE (all the other descriptors) is that the=
 former can be followed by name-words, whereas the latter cannot.
+<p>
+<cx "name-words, limitations on">  XE "name-words: limitations on"   <cx "=
name-words, pause requirements before">  XE "name-words: pause requirements=
 before"   <cx "LA, effect on necessity for pause before name-word">  XE "L=
A selma=92o: effect on necessity for pause before name-word"   <cx "doi, ef=
fect on necessity for pause before name-word">  XE "doi: effect on necessit=
y for pause before name-word"   <lx "doi">  XE "doi"   <lx "LA">  XE "LA se=
lma=92o"   <lx "la">  XE "la"   <lx "lai">  XE "lai"   <lx "la'i">  XE "la'=
i"   There are certain limitations on the form of name-words in Lojban. In =
particular, they cannot contain the letter-sequences (or sound-sequences) =
=93la=94, =93lai=94, or =93doi=94 unless a consonant immediately precedes w=
ithin the name. Reciprocally, every name not preceded by =93la=94, =93lai=
=94, =93la'i=94, or =93doi=94 must be preceded by a pause instead:
+<p>
+<pre><a name=3De12d4>12.4)	coi .djan.
+	Hello, John.
+<a name=3De12d5>12.5)	zo .djan. cmene mi
+	The-word =93John=94 is-the-name-of me.
+	My name is John.
+</pre><p>In <a href=3D#e12d4>Example 12.4 </a>and <a href=3D#e12d5>Example=
 12.5</a>, =93.djan.=94 appears with a pause before it as well as after it,=
 because the preceding word is not one of the four special cases. These rul=
es force names to always be separable from the general word-stream.
+<p>
+<cx "names, multiple">  XE "names: multiple"   Unless some other rule prev=
ents it (such as the rule that =93zo=94 is always followed by a single word=
, which is quoted), multiple names may appear wherever one name is permitte=
d, each with its terminating pause:
+<p>
+<ex "Newport News">  XE "Newport News: example=93  <ex "John Paul Jones"> =
 XE "John Paul Jones: example=93  <pre><a name=3De12d6>12.6)	doi djan. pol.=
 djonz. le bloti cu klama fi la niuport. niuz.
+	John Paul Jones, the boat comes (to somewhere) from Newport News.
+</pre><cx "name-words, permissible consonant combinations">  XE "name-word=
s: permissible consonant combinations"   A name may not contain any consona=
nt combination that is illegal in Lojban words generally: the =93impermissi=
ble consonant clusters=94 of Lojban morphology (explained in <a href=3Dchap=
3.html>Chapter 3</a>). Thus =93djeimz.=94 is not a valid version of =93Jame=
s=94 (because =93mz=94 is invalid): =93djeimyz=94 will suffice. Similarly, =
=93la=94 may be replaced by =93ly=94, =93lai=94 by =93ly'i=94, =93doi=94 by=
 =93do'i=94 or =93dai=94. Here are a few examples:
+<p>
+<pre>    English name	invalid Lojban name	valid Lojban names
+</pre><ex "Lottie">  XE "Lottie: example=93  <ex "Lyra">  XE "Lyra: exampl=
e=93  <ex "Doyle">  XE "Doyle: example=93  <pre> 	Doyle		       *doi,l		   =
do'il or dai,l
+	Lyra		       *lairas		   ly'iras
+	Lottie 		       *latis		   LYtis. or lotis.
+(American pronunciation)
+</pre><cx "names, using rafsi">  XE "names: using rafsi"   Names may be bo=
rrowed from other languages or created arbitrarily. Another common practice=
 is to use one or more rafsi, arranged to end with a consonant, to form a n=
ame: thus the rafsi =93loj-=94 for =93logji=94 (logical) and =93ban-=94 for=
 =93bangu=94 (language) unite to form the name of this language:
+<p>
+<pre><a name=3De12d7>12.7)	lojban.
+	Lojban
+</pre><cx "names, borrowing from other languages">  XE "names: borrowing f=
rom other languages"   <cx "names from vowel-final base, commonly used cons=
onant endings">  XE "names from vowel-final base: commonly used consonant e=
ndings"   When borrowing names from another language which end in a vowel, =
or when turning a Lojban brivla (all of which end in vowels) into a name, t=
he vowel may be removed or an arbitrary consonant added. It is common (but =
not required) to use the consonants =93s=94 or =93n=94 when borrowing vowel=
-final names from English; speakers of other languages may wish to use othe=
r consonant endings.
+<p>
+<cx "names with la, implicit quantifier for">  XE "names with la: implicit=
 quantifier for"   The implicit quantifier for name sumti of the form =93la=
=94 followed by a name is =93su'o=94, just as for =93la=94 followed by a se=
lbri.
+<p>
+<a name=3Ds13><h3>Pro-sumti summary</h3>
+<p>
+<cx "pro-sumti, classes of">  XE "pro-sumti: classes of"   The Lojban pro-=
sumti are the cmavo of selma'o KOhA. They fall into several classes: person=
al, definable, quantificational, reflexive, back-counting, indefinite, demo=
nstrative, metalinguistic, relative, question. More details are given in <a=
 href=3Dchap7.html>Chapter 7</a>; this section mostly duplicates informatio=
n found there, but adds material on the implicit quantifier of each pro-sum=
ti.
+<p>
+<cx "pro-sumti, implicit quantifier for">  XE "pro-sumti: implicit quantif=
ier for"   <lx "ro">  XE "ro"   The following examples illustrate each of t=
he classes. Unless otherwise noted below, the implicit quantification for p=
ro-sumti is =93ro=94 (all). In the case of pro-sumti which refer to other s=
umti, the =93ro=94 signifies =93all of those referred to by the other sumti=
=94: thus it is possible to restrict, but not to extend, the quantification=
 of the other sumti.
+<p>
+<cx "personal pro-sumti">  XE "personal pro-sumti"   Personal pro-sumti (=
=93mi=94, =93do=94, =93mi'o=94, =93mi'a=94, =93ma'a=94, =93do'o=94, =93ko=
=94) refer to the speaker or the listener or both, with or without third pa=
rties:
+<p>
+<pre><a name=3De13d1>13.1)	mi prami do
+	I love you.
+</pre><cx "personal pro-sumti, implicit quantifier for">  XE "personal pro=
-sumti: implicit quantifier for"   The personal pro-sumti may be interprete=
d in context as either representing individuals or masses, so the implicit =
quantifier may be =93pisu'o=94 rather than =93ro=94: in particular, =93mi'o=
=94, =93mi'a=94, =93ma'a=94, and =93do'o=94 specifically represent mass com=
binations of the individuals (you and I, I and others, you and I and others=
, you and others) that make them up.
+<p>
+<cx "definable pro-sumti">  XE "definable pro-sumti"   Definable pro-sumti=
 (=93ko'a=94, =93ko'e=94, =93ko'i=94, =93ko'o=94, =93ko'u=94, =93fo'a=94, =
=93fo'e=94, =93fo'i=94, =93fo'o=94, =93fo'u=94) refer to whatever the speak=
er has explicitly made them refer to. This reference is accomplished with =
=93goi=94 (of selma'o GOI), which means =93defined-as=94.
+<p>
+<pre><a name=3De13d2>13.2)	le cribe goi ko'a cu xekri  .i ko'a citka le sm=
acu
+	The bear defined-as it-1 is-black.  It-1 eats the mouse.
+</pre><cx "quantificational pro-sumti">  XE "quantificational pro-sumti"  =
 Quantificational pro-sumti (=93da=94, =93de=94, =93di=94) are used as vari=
ables in bridi involving predicate logic:
+<p>
+<pre><a name=3De13d3>13.3)	ro da poi prenu cu prami pa de poi finpe
+	All somethings-1 which-are persons love one something-2 which-is a-fish.
+	All persons love a fish (each his/her own).
+</pre><cx "quantificational pro-sumti, implicit quantification rules">  XE=
 "quantificational pro-sumti: implicit quantification rules"   (This is not=
 the same as =93All persons love a certain fish=94; the difference between =
the two is one of quantifier order.) The implicit quantification rules for =
quantificational pro-sumti are particular to them, and are discussed in det=
ail in <a href=3Dchap16.html>Chapter 16</a>. Roughly speaking, the quantifi=
er is =93su'o=94 (at least one) when the pro-sumti is first used, and =93ro=
=94 (all) thereafter.
+<p>
+<cx "reflexive pro-sumti">  XE "reflexive pro-sumti"   Reflexive pro-sumti=
 (=93vo'a=94, =93vo'e=94, =93vo'i=94, =93vo'o=94, =93vo'u=94) refer to the =
same referents as sumti filling other places in the same bridi, with the ef=
fect that the same thing is referred to twice:
+<p>
+<pre><a name=3De13d4>13.4)	le cribe cu batci vo'a
+	The bear bites what-is-in-the-x1-place.
+	The bear bites itself.
+</pre><cx "back-counting pro-sumti">  XE "back-counting pro-sumti"   Back-=
counting pro-sumti (=93ri=94, =93ra=94, =93ru=94) refer to the referents of=
 previous sumti counted backwards from the pro-sumti:
+<p>
+<pre><a name=3De13d5>13.5)	mi klama la frankfurt. ri
+	I go to-Frankfurt from-the-referent-of-the-last-sumti
+	I go from Frankfurt to Frankfurt (by some unstated route).
+</pre><cx "indefinite pro-sumti">  XE "indefinite pro-sumti"   Indefinite =
pro-sumti (=93zo'e=94, =93zu'i=94, =93zi'o=94) refer to something which is =
unspecified:
+<p>
+<pre><a name=3De13d6>13.6)	mi klama la frankfurt. zo'e zo'e zo'e
+	I go to-Frankfurt from-unspecified via-unspecified by-means-unspecified.
+</pre><cx "indefinite pro-sumti, implicit quantifier for">  XE "indefinite=
 pro-sumti: implicit quantifier for"   The implicit quantifier for indefini=
te pro-sumti is, well, indefinite. It might be =93ro=94 (all) or =93su'o=94=
 (at least one) or conceivably even =93no=94 (none), though =93no=94 would =
require a very odd context indeed.
+<p>
+<cx "demonstrative pro-sumti">  XE "demonstrative pro-sumti"   Demonstrati=
ve pro-sumti (=93ti=94, =93ta=94, =93tu=94) refer to things pointed at by t=
he speaker, or when pointing is not possible, to things near or far from th=
e speaker:
+<p>
+<pre><a name=3De13d7>13.7)	ko muvgau=20
+ti ta tu
+	You [imperative] move=20
+this-thing from-that-nearby-place to-that-further-away-place.
+	Move this from there to over there!
+</pre><cx "metalinguistic pro-sumti">  XE "metalinguistic pro-sumti"   Met=
alinguistic pro-sumti (=93di'u=94, =93de'u=94, =93da'u=94, =93di'e=94, =93d=
e'e=94, =93da'e=94, =93dei=94, =93do'i=94) refer to spoken or written utter=
ances, either preceding, following, or the same as the current utterance.
+<p>
+<pre><a name=3De13d8>13.8)	li re su'i re du li vo
+		.i la=92edi'u jetnu
+	The-number two plus two equals the-number four.
+		The-previous-utterance is-true.
+</pre><cx "metalinguistic pro-sumti, implicit quantifier for">  XE "metali=
nguistic pro-sumti: implicit quantifier for"   The implicit quantifier for =
metalinguistic pro-sumti is =93su'o=94 (at least one), because they are con=
sidered analogous to =93lo=94 descriptions: they refer to things which real=
ly are previous, current, or following utterances.
+<p>
+<cx "relative pro-sumti">  XE "relative pro-sumti"   The relative pro-sumt=
i (=93ke'a=94) is used within relative clauses (see <a href=3Dchap8.html>Ch=
apter 8 </a>for a discussion of relative clauses) to refer to whatever sumt=
i the relative clause is attached to.
+<p>
+<pre><a name=3De13d9>13.9)	mi viska le mlatu ku poi zo'e zbasu ke'a
+		loi slasi
+	I see the cat(s) such-that something-unspecified makes it/them (the cats)
+		from-a-mass-of plastic.
+	I see the cat(s) made of plastic.
+</pre><cx "question pro-sumti">  XE "question pro-sumti"   The question pr=
o-sumti (=93ma=94) is used to ask questions which request the listener to s=
upply a sumti which will make the question into a truth:
+<p>
+<pre><a name=3De13d10>13.10)	do klama ma
+	You go to-what-sumti?
+	Where are you going?
+</pre><cx "question pro-sumti, implicit quantifier for">  XE "question pro=
-sumti: implicit quantifier for"   The implicit quantifier for the question=
 pro-sumti is =93su'o=94 (at least one), because the listener is only being=
 asked to supply a single answer, not all correct answers.
+<p>
+<cx "definable pro-sumti, sequences of lerfu words as">  XE "definable pro=
-sumti: sequences of lerfu words as"   In addition, sequences of lerfu word=
s (of selma'o BY and related selma'o) can also be used as definable pro-sum=
ti.
+<p>
+<a name=3Ds14><h3>Quotation summary</h3>
+<p>
+<cx "quotation, four kinds">  XE "quotation: four kinds"   There are four =
kinds of quotation in Lojban: text quotation, words quotation, single-word =
quotation, non-Lojban quotation. More information is provided in <a href=3D=
chap19.html>Chapter 19</a>.
+<p>
+<cx "text quotation, syntax of">  XE "text quotation: syntax of"   <cx "te=
xt quotation, as internally grammatical">  XE "text quotation: as internall=
y grammatical"   <lx "lu">  XE "lu"   <lx "li'u">  XE "li'u"   Text quotati=
ons are preceded by =93lu=94 and followed by =93li'u=94, and are an essenti=
al part of the surrounding text: they must be grammatical Lojban texts.
+<p>
+<pre><a name=3De14d1>14.1)	mi cusku lu mi'e djan. li'u
+	I say the-text [quote] I-am John [unquote].
+	I say =93I'm John=94.
+</pre><cx "word quotation, as morphologically valid">  XE "word quotation:=
 as morphologically valid"   <cx "word quotation, internal grammar of">  XE=
 "word quotation: internal grammar of"   Words quotations are quotations of=
 one or more Lojban words. The words need not mean anything, but they must =
be morphologically valid so that the end of the quotation can be discerned.
+<p>
+<lx "lo'u">  XE "lo'u"   <lx "le'u">  XE "le'u"   <pre><a name=3De14d2>14.=
2)	mi cusku lo'u li mi le'u
+	I say the-words [quote] =93li mi=94 [unquote].
+	I say =93li mi=94.
+</pre><p>Note that the translation of <a href=3D#e14d2>Example 14.2 </a>do=
es not translate the Lojban words, because they are not presumed to have an=
y meaning (in fact, they are ungrammatical).
+<p>
+<cx "single-word quotation">  XE "single-word quotation"   Single-word quo=
tation quotes a single Lojban word. Compound cmavo are not allowed.
+<p>
+<pre><a name=3De14d3>14.3)	mi cusku zo .ai
+	I say the-word =93.ai=94.
+</pre><cx "non-Lojban quotation">  XE "non-Lojban quotation"   Non-Lojban =
quotation can quote anything, Lojban or not, even non-speech such as drum t=
alk, whistle words, music, or belching. A Lojban word which does not appear=
 within the quotation is used before and after it to set it off from the su=
rrounding Lojban text.
+<p>
+<pre><a name=3De14d4>14.4)	mi cusku zoi kuot. I'm John .kuot
+	I say =93I'm John=94.
+</pre><cx "quotation, implicit quantifier for">  XE "quotation: implicit q=
uantifier for"   The implicit quantifier for all types of quotation is =93s=
u'o=94 (at least one), because quotations are analogous to =93lo=94 descrip=
tions: they refer to things which actually are words or sequences of words.
+<p>
+<a name=3Ds15><h3>Number summary</h3>
+<p>
+<cx "number sumti, syntax of">  XE "number sumti: syntax of"   <cx "number=
 sumti, with li">  XE "number sumti: with li"   <lx "li">  XE "li"   The su=
mti which refer to numbers consist of the cmavo =93li=94 (of selma'o LI) fo=
llowed by an arbitrary Lojban mekso, or mathematical expression. This can b=
e anything from a simple number up to the most complicated combination of n=
umbers, variables, operators, and so on. Much more information on numbers i=
s given in <a href=3Dchap18.html>Chapter 18</a>. Here are a few examples of=
 increasing complexity:
+<p>
+<pre><a name=3De15d1>15.1)	li vo
+	the-number four
+	4
+<a name=3De15d2>15.2)	li re su'i re
+	the-number two plus two
+	<math>2 + 2</math>
+<a name=3De15d3>15.3)	li .abu bopi'i xy. bote'a re su'i by. bopi'i xy. su'=
i cy.
+	the-number a times x to-power 2 plus b times x plus c
+	<math>ax<sup>2</sup>+ bx + c</math>
+</pre><cx "number sumti, with me'o">  XE "number sumti: with me'o"   <cx "=
me'o">  XE "me'o"   <cx "LI">  XE "LI selma=92o"   <cx "number sumti, with =
me'o contrasted with li">  XE "number sumti: with me'o contrasted with li" =
  <cx "number sumti, with li contrasted with me'o">  XE "number sumti: with=
 li contrasted with me'o"   An alternative to =93li=94 is =93me'o=94, also =
of selma'o LI. Number expressions beginning with =93me'o=94 refer to the ac=
tual expression, rather than its value. Thus <a href=3D#e15d1>Example 15.1 =
</a>and <a href=3D#e15d2>Example 15.2 </a>above have the same meaning, the =
number four, whereas
+<p>
+<pre><a name=3De15d4>15.4)	me'o vo
+	the-expression four
+	=934=94
+</pre>and
+<p>
+<pre><a name=3De15d5>15.5)	me'o re su'i re
+	the-expression two plus two
+	=932+2=94
+</pre>refer to different pieces of text.
+<p>
+<cx "numbers, implicit quantifier for">  XE "numbers: implicit quantifier =
for"   <cx "mathematical expressions, implicit quantifier for">  XE "mathem=
atical expressions: implicit quantifier for"   The implicit quantifier for =
numbers and mathematical expressions is =93su'o=94, because these sumti are=
 analogous to =93lo=94 descriptions: they refer to things which actually ar=
e numbers or pieces of text. In the case of numbers (with =93li=94), this i=
s a distinction without a difference, as there is only one number which is =
4; but there are many texts =934=94, as many as there are documents in whic=
h that numeral appears.
+
+<p>
+</body></html>
+<img src=3Dchap7.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D4=
05>
+
+<h2>Chapter 7
+<br>
+Brevity Is The Soul Of Language: Pro-sumti And Pro-bridi</h2>
+<h6>$Revision: 4.0 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>What are pro-sumti and pro-bridi? What are they for?</h3>
+<p>
+<cx "pronouns in English, as noun abbreviations">  XE "pronouns in English=
: as noun abbreviations"   Speakers of Lojban, like speakers of other langu=
ages, require mechanisms of abbreviation. If every time we referred to some=
thing, we had to express a complete description of it, life would be too sh=
ort to say what we have to say. In English, we have words called =93pronoun=
s=94 which allow us to replace nouns or noun phrases with shorter terms. An=
 English with no pronouns might look something like this:
+<p>
+<pre><a name=3De1d1>1.1)	Speakers of Lojban, like speakers of other langua=
ges, require mechanisms of
+abbreviation. If every time speakers of Lojban referred to a thing to whic=
h
+	speakers of Lojban refer, speakers of Lojban had to express a complete
+description of what speakers of Lojban referred to, life would be too shor=
t to
+say what speakers of Lojban have to say.
+</pre><cx "pronouns in English, as independent of abbreviations">  XE "pro=
nouns in English: as independent of abbreviations"   Speakers of this kind =
of English would get mightily sick of talking. Furthermore, there are uses =
of pronouns in English which are independent of abbreviation. There is all =
the difference in the world between:
+<p>
+<ex "shook stick">  XE "shook stick: example=93  <pre><a name=3De1d2>1.2)	=
John picked up a stick and shook it.
+</pre>and
+<p>
+<pre><a name=3De1d3>1.3)	John picked up a stick and shook a stick.
+</pre><a href=3D#e1d3>Example 1.3 </a>does not imply that the two sticks a=
re necessarily the same, whereas <a href=3D#e1d2>Example 1.2 </a>requires t=
hat they are.
+<p>
+<cx "pronouns, compared to pro-sumti in usage as abbreviations">  XE "pron=
ouns: compared to pro-sumti in usage as abbreviations"   <cx "pro-sumti, co=
mpared to pronouns in usage as abbreviations">  XE "pro-sumti: compared to =
pronouns in usage as abbreviations"   <cx "pro-sumti, definition">  XE "pro=
-sumti: definition"   <cx "pro-bridi, definition">  XE "pro-bridi: definiti=
on"   <lx "KOhA">  XE "KOhA selma=92o"   <lx "GOhA">  XE "GOhA selma=92o"  =
 <cx "pro-bridi, compared to pro-sumti as means of abbreviation">  XE "pro-=
bridi: compared to pro-sumti as means of abbreviation"   <cx "pro-sumti, co=
mpared to pro-bridi as means of abbreviation">  XE "pro-sumti: compared to =
pro-bridi as means of abbreviation"   <cx "pro-sumti, series">  XE "pro-sum=
ti: series"   In Lojban, we have sumti rather than nouns, so our equivalent=
 of pronouns are called by the hybrid term =93pro-sumti=94. A purely Lojban=
 term would be =93sumti cmavo=94: all of the pro-sumti are cmavo belonging =
to selma'o KOhA. In exactly the same way, Lojban has a group of cmavo (belo=
nging to selma'o GOhA) which serve as selbri or full bridi. These may be ca=
lled =93pro-bridi=94 or =93bridi cmavo=94. This chapter explains the uses o=
f all the members of selma'o KOhA and GOhA. They fall into a number of grou=
ps, known as series: thus, in selma'o KOhA, we have among others the mi-ser=
ies, the ko'a-series, the da-series, and so on. In each section, a series o=
f pro-sumti is explained, and if there is a corresponding series of pro-bri=
di, it is explained and contrasted. Many pro-sumti series don't have pro-br=
idi analogues, however.
+<p>
+<cx "referent of pro-sumti, definition">  XE "referent of pro-sumti: defin=
ition"   <cx "antecedent of pro-sumti, definition">  XE "antecedent of pro-=
sumti: definition"   <cx "referent of pro-bridi, definition">  XE "referent=
 of pro-bridi: definition"   <cx "antecedent of pro-bridi, definition">  XE=
 "antecedent of pro-bridi: definition"   A few technical terms: The term =
=93referent=94 means the thing to which a pro-sumti (by extension, a pro-br=
idi) refers. If the speaker of a sentence is James, then the referent of th=
e word =93I=94 is James. On the other hand, the term =93antecedent=94 refer=
s to a piece of language which a pro-sumti (or pro-bridi) implicitly repeat=
s. In
+<p>
+<pre><a name=3De1d4>1.4)	John loves himself
+</pre>the antecedent of =93himself=94 is =93John=94; not the person, but a=
 piece of text (a name, in this case). John, the person, would be the refer=
ent of =93himself=94. Not all pro-sumti or pro-bridi have antecedents, but =
all of them have referents.
+<p>
+<a name=3Ds2><h3>Personal pro-sumti: the mi-series</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	mi	KOhA	mi-series	I, me
+	do	KOhA	mi-series	you
+	mi'o	KOhA	mi-series	you and I
+	mi'a	KOhA	mi-series	I and others,
+					we but not you
+	ma'a	KOhA	mi-series	you and I and others
+	do'o	KOhA	mi-series	you and others
+	ko	KOhA	mi-series	you-imperative
+</pre><ex "foreman of a jury">  XE "foreman of a jury: example=93  <cx "mi=
-series, of pro-sumti">  XE "mi-series: of pro-sumti"   <cx "pro-sumti, mi-=
series">  XE "pro-sumti: mi-series"   <lx "mi">  XE "mi"   <lx "do">  XE "d=
o"   <cx "pro-sumti, for speaker(s)">  XE "pro-sumti: for speaker(s)"   <cx=
 "pro-sumti, for listener(s)">  XE "pro-sumti: for listener(s)"   <cx "pers=
onal pronouns, with mi-series for I/you">  XE "personal pronouns: with mi-s=
eries for I/you"   The mi-series of pro-sumti refer to the speaker, the lis=
tener, and others in various combinations. =93mi=94 refers to the speaker a=
nd perhaps others for whom the speaker speaks; it may be a Lojbanic mass. =
=93do=94 refers to the listener or listeners. Neither =93mi=94 nor =93do=94=
 is specific about the number of persons referred to; for example, the fore=
man of a jury may refer to the members of the jury as =93mi=94, since in sp=
eaking officially he represents all of them.
+<p>
+<cx "COI, effect on referent of =93mi=94">  XE "COI selma=92o: effect on r=
eferent of mi"   <cx "COI, effect on referent of =93do=94">  XE "COI selma=
=92o: effect on referent of do"   <lx "mi'e">  XE "mi'e"   <lx "COI">  XE "=
COI selma=92o"   The referents of =93mi=94 and =93do=94 are usually obvious=
 from the context, but may be assigned by the vocative words of selma'o COI=
, explained in <a href=3Dchap13.html>Chapter 13</a>. The vocative =93mi'e=
=94 assigns =93mi=94, whereas all of the other vocatives assign =93do=94.
+<p>
+<pre><a name=3De2d1>2.1)	mi'e djan. doi frank. mi cusku lu mi bajra li'u d=
o
+	I-am John, O Frank, I express [quote] I run [unquote] to-you
+	I am John, Frank; I tell you =93I run=94.
+</pre><lx "mi'o">  XE "mi'o"   <lx "mi'a">  XE "mi'a"   <lx "ma'a">  XE "m=
a'a"   <lx "do'o">  XE "do'o"   <cx "pro-sumti, for listeners and/or speake=
rs and/or others">  XE "pro-sumti: for listeners and/or speakers and/or oth=
ers"   The cmavo =93mi'o=94, =93mi'a=94, =93ma'a=94, and =93do'o=94 express=
 various combinations of the speaker and/or the listener and/or other peopl=
e:
+<p>
+<dl compact><dt><dd>=93mi'o=94 includes only the speaker and the listener =
but no one else;
+=93mi'a=94 includes the speaker and others but excludes the listener;
+<p><dt><dd>=93do'o=94 includes the listener and others but excludes the sp=
eaker;
+<p><dt><dd>=93ma'a=94 includes all three:  speaker, listener, others.
+</dl><cx "pro-sumti for speaker/listener/others, relation to joi">  XE "pr=
o-sumti for speaker/listener/others: relation to joi"   <cx "pro-sumti for =
speaker/listener/others, as masses">  XE "pro-sumti for speaker/listener/ot=
hers: as masses"   All of these pro-sumti represent masses. For example, =
=93mi'o=94 is the same as =93mi joi do=94, the mass of me and you considere=
d jointly.
+<p>
+<cx "English =93we=94, contrasted with Lojban pro-sumti for =93we=94">  XE=
 "English we: contrasted with Lojban pro-sumti for we"   <cx "pro-sumti for=
 =93we=94, contrasted with English =93we=94">  XE "pro-sumti for we: contra=
sted with English we"   In English, =93we=94 can mean =93mi=94 or =93mi'o=
=94 or =93mi'a=94 or even =93ma'a=94, and English-speakers often suffer bec=
ause they cannot easily distinguish  =93mi'o=94 from =93mi'a=94:
+<p>
+<pre><a name=3De2d2>2.2)	We're going to the store.
+</pre><p>Does this include the listener or not? There's no way to be sure.
+<p>
+<lx "ko">  XE "ko"   <cx "commands, with ko">  XE "commands: with ko"   <c=
x "imperatives, with ko">  XE "imperatives: with ko"   <cx "ko, use for imp=
eratives">  XE "ko: use for imperatives"   <cx "ko, use for commands">  XE =
"ko: use for commands"   Finally, the cmavo =93ko=94 is logically equivalen=
t to =93do=94; its referent is the listener. However, its use alters an ass=
ertion about the listener into a command to the listener to make the assert=
ion true:
+<p>
+<pre><a name=3De2d3>2.3)	do klama le zarci
+	You go to-the store.
+</pre>becomes:
+
+<p>
+<pre><a name=3De2d4>2.4)	ko klama le zarci
+	You [imperative] go to-the store.
+	Make =93you go to the store=94 true!
+	Go to the store!
+</pre><cx "imperatives, English contrasted with Lojban in presence of subj=
ect of command">  XE "imperatives: English contrasted with Lojban in presen=
ce of subject of command"   <cx "ko, in later selbri place in imperative"> =
 XE "ko: in later selbri place in imperative"   In English, the subject of =
a command is omitted, but in Lojban, the word =93ko=94 must be used. Howeve=
r, =93ko=94 does not have to appear in the x1 place:
+<p>
+<pre><a name=3De2d5>2.5)	mi viska ko
+	I see you [imperative]
+	Make =93I see you=94 true!
+	Be seen by me!
+</pre><cx "ko, in sub-clause of main bridi">  XE "ko: in sub-clause of mai=
n bridi"   In <a href=3D#e2d5>Example 2.5</a>, it is necessary to make the =
verb passive in English in order to convey the effect of =93ko=94 in the x2=
 place. Indeed, =93ko=94 does not even have to be a sumti of the main bridi=
:
+<p>
+<pre><a name=3De2d6>2.6)	mi viska le prenu poi prami ko
+	I see the person that loves you [imperative]
+	Make =93I see the person that loves you=94 true!
+	Be such that the person who loves you is seen by me!
+	Show me the person who loves you!
+</pre><cx "mi-series pro-sumti, lack of pro-bridi equivalent">  XE "mi-ser=
ies pro-sumti: lack of pro-bridi equivalent"   As mentioned in <a href=3D#s=
1>Section 1</a>, some pro-sumti series have corresponding pro-bridi series.=
 However, there is no equivalent of the mi-series among pro-bridi, since a =
person isn't a relationship.
+<p>
+<a name=3Ds3><h3>Demonstrative pro-sumti: the ti-series</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ti	KOhA	ti-series	this here, a nearby object
+	ta	KOhA	ti-series	that there, a medium-distant object
+	tu	KOhA	ti-series	that yonder, a far-distant object
+</pre><cx "pointing, reference by">  XE "pointing: reference by"   <cx "de=
monstrative pro-sumti">  XE "demonstrative pro-sumti"   <cx "ti-series pro-=
sumti, compared with English this/that">  XE "ti-series pro-sumti: compared=
 with English this/that"   <cx "this/that in English, compared with ti-seri=
es pro-sumti">  XE "this/that in English: compared with ti-series pro-sumti=
"   <cx "ti-series pro-sumti, as pointing referents only">  XE "ti-series p=
ro-sumti: as pointing referents only"   <cx "pro-sumti, ti-series">  XE "pr=
o-sumti: ti-series"   <lx "ti">  XE "ti"   <lx "ta">  XE "ta"   <lx "tu">  =
XE "tu"   It is often useful to refer to things by pointing to them or by s=
ome related non-linguistic mechanism. In English, the words =93this=94 and =
=93that=94 serve this function among others: =93this=94 refers to something=
 pointed at that is near the speaker, and =93that=94 refers to something fu=
rther away. The Lojban pro-sumti of the ti-series serve the same functions,=
 but more narrowly. The cmavo =93ti=94, =93ta=94, and =93tu=94 provide only=
 the pointing function of =93this=94 and =93that=94; they are not used to r=
efer to things that cannot be pointed at.
+<p>
+<cx "yon, as archaic English equivalent of tu">  XE "yon: as archaic Engli=
sh equivalent of tu"   <cx "tu, archaic English yon as equivalent of">  XE =
"tu: archaic English yon as equivalent of"   <cx "ti-series pro-sumti, 3 de=
grees of distance with">  XE "ti-series pro-sumti: 3 degrees of distance wi=
th"   There are three pro-sumti of the ti-series rather than just two becau=
se it is often useful to distinguish between objects that are at more than =
two different distances. Japanese, among other languages, regularly does th=
is. Until the 16th century, English did too; the pronoun =93that=94 referre=
d to something at a medium distance from the speaker, and the now-archaic p=
ronoun =93yon=94 to something far away.
+<p>
+<cx "ti-series pro-sumti, conversational convention for">  XE "ti-series p=
ro-sumti: conversational convention for"   <cx "ti-series pro-sumti, proble=
ms in written text">  XE "ti-series pro-sumti: problems in written text"   =
In conversation, there is a special rule about =93ta=94 and =93tu=94 that i=
s often helpful in interpreting them. When used contrastingly, =93ta=94 ref=
ers to something that is near the listener, whereas =93tu=94 refers to some=
thing far from both speaker and listener. This makes for a parallelism betw=
een =93ti=94 and =93mi=94, and =93ta=94 and =93do=94, that is convenient wh=
en pointing is not possible; for example, when talking by telephone. In wri=
tten text, on the other hand, the meaning of the ti-series is inherently va=
gue; is the writer to be taken as pointing to something, and if so, to what=
? In all cases, what counts as =93near=94 and =93far away=94 is relative to=
 the current situation.
+<p>
+<cx "this, pronoun usage contrasted with adjective usage">  XE "this: pron=
oun usage contrasted with adjective usage"   <cx "this, adjective usage con=
trasted with pronoun usage">  XE "this: adjective usage contrasted with pro=
noun usage"   <cx "ti, as pronoun usage for =93this=94">  XE "ti: as pronou=
n expression for English this"   <cx "this, pronoun usage with =93ti=94">  =
XE "this: English, pronoun expression with ti"   It is important to disting=
uish between the English pronoun =93this=94 and the English adjective =93th=
is=94 as in =93this boat=94. The latter is not represented in Lojban by =93=
ti=94:
+<p>
+<ex "this boat">  XE "this boat: example=93  <pre><a name=3De3d1>3.1)	le t=
i bloti
+	the this boat
+</pre><cx "vi, as adjective usage for =93this=94">  XE "vi: as adjective e=
xpression for English this"   <cx "this, adjective usage with =93vi=94">  X=
E "this: English, adjective expression with vi"   does not mean =93this boa=
t=94 but rather =93this one's boat=94, =93the boat associated with this thi=
ng=94, as explained in <a href=3Dchap8.html>Chapter 8</a>. A correct Lojban=
 translation of <a href=3D#e3d1>Example 3.1 </a>is
+<p>
+<pre><a name=3De3d2>3.2)	le vi bloti
+	the here boat
+	the nearby boat
+</pre>using a spatial tense before the selbri =93bloti=94 to express that =
the boat is near the speaker. (Tenses are explained in full in <a href=3Dch=
ap11.html>Chapter 11</a>.) Another correct translation would be:
+<p>
+<cx "ti noi, as adjective usage for =93this=94">  XE "ti noi: as adjective=
 expression for this"   <cx "this, adjective usage with =93ti noi=94">  XE =
"this: adjective expression with ti noi"   <pre><a name=3De3d3>3.3)	ti noi =
bloti
+	this-thing which-incidentally is-a-boat
+</pre><cx "ti-series pro-sumti, lack of pro-bridi equivalent">  XE "ti-ser=
ies pro-sumti: lack of pro-bridi equivalent"   There are no demonstrative p=
ro-bridi to correspond to the ti-series: you can't point to a relationship.
+<p>
+<a name=3Ds4><h3>Utterance pro-sumti: the di'u-series</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	di'u	KOhA	di'u-series	the previous utterance
+	de'u	KOhA	di'u-series	an earlier utterance
+	da'u	KOhA	di'u-series	a much earlier utterance
+	di'e	KOhA	di'u-series	the next utterance
+	de'e	KOhA	di'u-series	a later utterance
+	da'e	KOhA	di'u-series	a much later utterance
+	dei	KOhA	di'u-series	this very utterance
+	do'i	KOhA	di'u-series	some utterance
+</pre><cx "pro-sumti for utterances">  XE "pro-sumti for utterances"   <cx=
 "utterance pro-sumti (see also di'u-series pro-sumti)">  XE "utterance pro=
-sumti (see also di'u-series pro-sumti)"   <cx "di'u-series pro-sumti">  XE=
 "di'u-series pro-sumti /r dihuseries   <cx "this, as utterance reference i=
n English">  XE "this: as utterance reference in English"   <cx "pro-sumti,=
 di'u-series">  XE "pro-sumti: di'u-series" /r dihuseries   The cmavo of th=
e di'u-series enable us to talk about things that have been, are being, or =
will be said. In English, it is normal to use =93this=94 and =93that=94 for=
 this (indeed, the immediately preceding =93this=94 is an example of such a=
 usage):
+<p>
+<pre><a name=3De4d1>4.1)	You don't like cats.
+	That is untrue.
+</pre><p>Here =93that=94 does not refer to something that can be pointed t=
o, but to the preceding sentence =93You don't like cats=94. In Lojban, ther=
efore, <a href=3D#e4d1>Example 4.1 </a>is rendered:
+<p>
+<pre><a name=3De4d2>4.2)	do na nelci loi mlatu  .i di'u jitfa jufra
+	You (Not!) like the-mass-of cats.  The-previous-utterance is-a-false-sent=
ence.
+</pre><cx "ta, contrasted with di'u">  XE "ta: contrasted with di'u"   <cx=
 "di'u, contrasted with ta">  XE "di'u: contrasted with ta"   <cx "ti-serie=
s pro-sumti, contrasted with di'u-series pro-sumti">  XE "ti-series pro-sum=
ti: contrasted with di'u-series pro-sumti"   Using =93ta=94 instead of =93d=
i'u=94 would cause the listener to look around to see what the speaker of t=
he second sentence was physically pointing to.
+<p>
+<lx "di'u">  XE "di'u"   <lx "de'u">  XE "de'u"   <lx "da'u">  XE "da'u"  =
 As with =93ti=94, =93ta=94, and =93tu=94, the cmavo of the di'u-series com=
e in threes: a close utterance, a medium-distance utterance, and a distant =
utterance, either in the past or in the future. It turned out to be impossi=
ble to use the =93i=94/=93a=94/=93u=94 vowel convention of the demonstrativ=
es in <a href=3D#s3>Section 3 </a>without causing collisions with other cma=
vo, and so the di'u-series has a unique =93i=94/=93e=94/=93a=94 convention =
in the first vowel of the cmavo.
+<p>
+<lx "di'e">  XE "di'e"   <lx "de'e">  XE "de'e"   <lx "da'e">  XE "da'e"  =
 Most references in speech are to the past (what has already been said), so=
 =93di'e=94, =93de'e=94, and =93da'e=94 are not very useful when speaking. =
In writing, they are frequently handy:
+<p>
+<ex "Simon says">  XE "Simon says: example=93  <pre><a name=3De4d3>4.3)	la=
 saimn. cusku di'e
+	Simon expresses the-following-utterance.
+	Simon says:
+</pre><a href=3D#e4d3>Example 4.3 </a>would typically be followed by a quo=
tation. Note that although presumably the quotation is of something Simon h=
as said in the past, the quotation utterance itself would appear after <a h=
ref=3D#e4d3>Example 4.3</a>, and so =93di'e=94 is appropriate.
+<p>
+<lx "dei">  XE "dei"   <lx "do'i">  XE "do'i"   The remaining two cmavo, =
=93dei=94 and =93do'i=94, refer respectively to the very utterance that the=
 speaker is uttering, and to some vague or unspecified utterance uttered by=
 someone at some time:
+<p>
+<pre><a name=3De4d4>4.4)	dei jetnu jufra
+	This-utterance is-a-true-sentence.
+	What I am saying (at this moment) is true.
+<a name=3De4d5>4.5)	do'i jetnu jufra
+	Some-utterance is-a-true-sentence.
+	That's true (where =93that=94 is not necessarily what was just said).
+The cmavo of the di'u-series have a meaning that is relative to the contex=
t. The referent of =93dei=94 in the current utterance is the same as the re=
ferent of =93di'u=94 in the next utterance. The term =93utterance=94 is use=
d rather than =93sentence=94 because the amount of speech or written text r=
eferred to by any of these words is vague. Often, a single bridi is intende=
d, but longer utterances may be thus referred to.
+<p>
+<lx "la'e">  XE "la'e"   <lx "LAhE">  XE "LAhE selma=92o"   <lx "la'edi'u"=
>  XE "la'edi'u"   Note one very common construction with =93di'u=94 and th=
e cmavo =93la'e=94 (of selma'o LAhE; see <a href=3Dchap6.html>Chapter 6</a>=
) which precedes a sumti and means =93the thing referred to by (the sumti)=
=94:
+<p>
+<pre><a name=3De4d6>4.6)	mi prami la djein.  .i mi nelci la'e di'u
+	I love Jane. And I like the-referent-of the-last-utterance.
+	I love Jane, and I like that.
+<cx "di'u, contrasted with la'edi'u">  XE "di'u: contrasted with la'edi'u"=
   <cx "la'edi'u, contrasted with di'u">  XE "la'edi'u: contrasted with di'=
u"   The effect of =93la'e di'u=94 in <a href=3D#e4d6>Example 4.6 </a>is th=
at the speaker likes, not the previous sentence, but rather the state of af=
fairs referred to by the previous sentence, namely his loving Jane. This cm=
avo compound is often written as a single word: =93la'edi'u=94. It is impor=
tant not to mix up =93di'u=94 and =93la'edi'u=94, or the wrong meaning will=
 generally result:
+<p>
+<pre><a name=3De4d7>4.7)	mi prami la djein.  .i mi nelci di'u
+	I love Jane. And I like the-last-utterance.
+</pre>says that the speaker likes one of his own sentences.
+<p>
+There are no pro-bridi corresponding to the di'u-series.
+<p>
+<a name=3Ds5><h3>Assignable pro-sumti and pro-bridi: the ko'a-series and t=
he broda-series</h3>
+<p>
+The following cmavo and gismu are discussed in this section:
+<p>
+<pre> 	ko'a	KOhA	ko'a-series	it-1
+	ko'e	KOhA	ko'a-series	it-2
+	ko'i		KOhA	ko'a-series	it-3
+	ko'o	KOhA	ko'a-series	it-4
+	ko'u	KOhA	ko'a-series	it-5
+	fo'a	KOhA	ko'a-series	it-6
+	fo'e	KOhA	ko'a-series	it-7
+	fo'i		KOhA	ko'a-series	it-8
+	fo'o	KOhA	ko'a-series	it-9
+	fo'u	KOhA	ko'a-series	it-10
+	broda	BRIVLA	broda-series	is-thing-1
+	brode	BRIVLA	broda-series	is-thing-2
+	brodi	BRIVLA	broda-series	is-thing-3
+	brodo	BRIVLA	broda-series	is-thing-4
+	brodu	BRIVLA	broda-series	is-thing-5
+	goi		GOI			pro-sumti assignment
+	cei		CEI			pro-bridi assignment
+</pre><cx "personal pronouns, with ko'a-series for he/she/it/they">  XE "p=
ersonal pronouns: with ko'a-series for he/she/it/they"   <cx "pro-sumti, ko=
'a-series">  XE "pro-sumti: ko'a-series"   <cx "ko'a-series pro-sumti">  XE=
 "ko'a-series pro-sumti"   The discussion of personal pro-sumti in <a href=
=3D#s2>Section 2 </a>may have seemed incomplete. In English, the personal p=
ronouns include not only =93I=94 and =93you=94 but also =93he=94, =93she=94=
, =93it=94, and =93they=94. Lojban does have equivalents of this latter gro=
up: in fact, it has more of them than English does. However, they are organ=
ized and used very differently.
+<p>
+<cx "personal pronouns for he/she/it/they, English contrasted with Lojban =
in organization">  XE "personal pronouns for he/she/it/they: English contra=
sted with Lojban in organization"   <cx "ko'a-series pro-sumti, as assignab=
le">  XE "ko'a-series pro-sumti: as assignable"   There are ten cmavo in th=
e ko'a-series, and they may be assigned freely to any sumti whatsoever. The=
 English word =93he=94 can refer only to males, =93she=94 only to females (=
and ships and a few other things), =93it=94 only to inanimate things, and =
=93they=94 only to plurals; the cmavo of the ko'a-series have no restrictio=
ns at all. Therefore, it is almost impossible to guess from the context wha=
t ko'a-series cmavo might refer to if they are just used freely:
+<p>
+<lx "ko'a">  XE "ko'a"   <pre><a name=3De5d1>5.1)	la .alis. klama le zarci=
  .i ko'a blanu
+	Alice goes-to the store.  It-1 is-blue.
+</pre><p>The English gloss =93it-1=94, plus knowledge about the real world=
, would tend to make English-speakers believe that =93ko'a=94 refers to the=
 store; in other words, that its antecedent is =93le zarci=94. To a Lojbani=
st, however, =93la .alis.=94 is just as likely an antecedent, in which case=
 <a href=3D#e5d1>Example 5.1 </a>means that Alice, not the store, is blue.
+<p>
+<lx "goi">  XE "goi"   <cx "ko'a-series pro-sumti, assigning with goi">  X=
E "ko'a-series pro-sumti: assigning with goi"   To avoid this pitfall, Lojb=
an employs special syntax, using the cmavo =93goi=94:
+<p>
+<pre><a name=3De5d2>5.2)	la .alis. klama le zarci  .i ko'a goi la .alis. c=
u blanu
+	Alice goes-to the store.  It-1, also-known-as Alice, is-blue.
+</pre><cx "ko'a-series pro-sumti, assignment with goi as symmetrical">  XE=
 "ko'a-series pro-sumti: assignment with goi as symmetrical"   Syntacticall=
y, =93goi la .alis.=94 is a relative phrase (relative phrases are explained=
 in <a href=3Dchap8.html>Chapter 8</a>). Semantically, it says that =93ko'a=
=94 and =93la .alis.=94 refer to the same thing, and furthermore that this =
is true because =93ko'a=94 is being defined as meaning =93la .alis.=94. It =
is equally correct to say:
+
+<p>
+<pre><a name=3De5d3>5.3)	la .alis. klama le zarci  .i la .alis. goi ko'a c=
u blanu
+	Alice goes-to the store.  Alice, also-known-as it-1, is-blue.
+</pre><lx "ge'u">  XE "ge'u"   in other words, =93goi=94 is symmetrical. T=
here is a terminator, =93ge'u=94 (of selma'o GEhU), which is almost always =
elidable. The details are in <a href=3Dchap8.html>Chapter 8</a>.
+<p>
+<cx "goi assignment of ko'a-series pro-sumti, use in speech contrasted wit=
h writing">  XE "goi assignment of ko'a-series pro-sumti: use in speech con=
trasted with writing"   The afterthought form of =93goi=94 shown in <a href=
=3D#e5d2>Example 5.2 </a>and <a href=3D#e5d3>Example 5.3 </a>is probably mo=
st common in speech, where we do not know until part way through our uttera=
nce that we will want to refer to Alice again. In writing, though, =93ko'a=
=94 may be assigned at the point where Alice is first mentioned. An example=
 of this forethought form of =93goi=94 is:
+<p>
+<pre><a name=3De5d4>5.4)	la .alis. goi ko'a klama le zarci  .i ko'a cu bla=
nu
+	Alice, also-known-as it-1, goes-to the store.  It-1 is-blue.
+</pre><ex "legal jargon">  XE "legal jargon: example=93  <ex "hereafter kn=
own as">  XE "hereafter known as: example=93  Again, =93ko'a goi la .alis.=
=94 would have been entirely acceptable in <a href=3D#e5d4>Example 5.4</a>.=
 This last form is reminiscent of legal jargon: =93The party of the first p=
art, hereafter known as Buyer,=A0<dots>=85</dots>=94.
+<p>
+<cx "broda-series pro-bridi">  XE "broda-series pro-bridi"   <cx "pro-brid=
i, broda-series">  XE "pro-bridi: broda-series"   <cx "ko'a-series for pro-=
sumti, compared with broda-series for pro-bridi">  XE "ko'a-series for pro-=
sumti: compared with broda-series for pro-bridi"   <cx "broda-series for pr=
o-bridi, compared with ko'a-series for pro-sumti">  XE "broda-series for pr=
o-bridi: compared with ko'a-series for pro-sumti"   <cx "pro-bridi, as abbr=
eviation for bridi">  XE "pro-bridi: as abbreviation for bridi"   Just as t=
he ko'a-series of pro-sumti allows a substitute for a sumti which is long o=
r complex, or which for some other reason we do not want to repeat, so the =
broda-series of pro-bridi allows a substitute for a selbri or even a whole =
bridi:
+<p>
+<ex "thingy">  XE "thingy: example=93  <lx "broda">  XE "broda"   <pre><a =
name=3De5d5>5.5)	ti slasi je mlatu bo cidja lante gacri cei broda
+		.i le crino broda cu barda  .i le xunre broda cu cmalu
+	These are plastic cat-food can covers or thingies.
+		The green thingy is large.  The red thingy is small.
+</pre><cx "antecedent, for pro-bridi">  XE "antecedent: for pro-bridi"   <=
lx "cei">  XE "cei"   <cx "cei for broda-series assignment, compared with g=
oi for ko'a-series assignment">  XE "cei for broda-series assignment: compa=
red with goi for ko'a-series assignment"   <cx "goi for ko'a-series assignm=
ent, compared with cei for broda-series assignment">  XE "goi for ko'a-seri=
es assignment: compared with cei for broda-series assignment"   <cx "broda-=
series pro-bridi, assigning with cei">  XE "broda-series pro-bridi: assigni=
ng with cei"   <cx "cei, for broda-series pro-bridi assignment">  XE "cei: =
for broda-series pro-bridi assignment"   <cx "broda-series pro-bridi, word-=
form rationale">  XE "broda-series pro-bridi: word-form rationale"   The pr=
o-bridi =93broda=94 has as its antecedent the selbri =93slasi je mlatu bo c=
idja lante gacri=94. The cmavo =93cei=94 performs the role of =93goi=94 in =
assigning =93broda=94 to this long phrase, and =93broda=94 can then be used=
 just like any other brivla. (In fact, =93broda=94 and its relatives actual=
ly <i>are </i>brivla: they are gismu in morphology, although they behave ex=
actly like the members of selma'o GOhA. The reasons for using gismu rather =
than cmavo are buried in the Loglan Project's history.)
+<p>
+<cx "antecedent, for pro-bridi as full bridi">  XE "antecedent: for pro-br=
idi as full bridi"   Note that pro-bridi are so called because, even though=
 they have the grammar of selbri, their antecedents are whole bridi. In the=
 following rather contrived example, the antecedent of =93brode=94 is the w=
hole bridi =93mi klama le zarci=94:
+<p>
+<pre><a name=3De5d6>5.6)	mi klama cei brode le zarci  .i do brode
+	I go-to (which-is claim-1) the store. You claim-1
+	I go to the store. You, too.
+</pre><cx "pro-bridi, overriding sumti of antecedent bridi for">  XE "pro-=
bridi: overriding sumti of antecedent bridi for"   In the second bridi, =93=
do brode=94 means =93do klama le zarci=94, because =93brode=94 carries the =
x2 sumti of =93mi klama le zarci=94 along with it. It also potentially carr=
ies the x1 sumti as well, but the explicit x1 sumti =93do=94 overrides the =
=93mi=94 of the antecedent bridi. Similarly, any tense or negation that is =
present in the antecedent is also carried, and can be overridden by explici=
t tense or negation cmavo on the pro-bridi. These rules hold for all pro-br=
idi that have antecedents.
+<p>
+<cx "broda-series pro-bridi, use as sample gismu">  XE "broda-series pro-b=
ridi: use as sample gismu"   <cx "broda-series pro-bridi, use as abstract p=
attern">  XE "broda-series pro-bridi: use as abstract pattern"   <cx "broda=
-series pro-bridi, with no assignment">  XE "broda-series pro-bridi: with n=
o assignment"   Another use of =93broda=94 and its relatives, without assig=
nment, is as =93sample gismu=94:
+<p>
+<pre><a name=3De5d7>5.7)	broda ke brode brodi
+	a thing-1 type of (thing-2 type-of thing-3)
+</pre>represents an abstract pattern, a certain kind of tanru. (Historical=
ly, this use was the original one.)
+<p>
+<cx "lerfu, as assignable pro-sumti">  XE "lerfu: as assignable pro-sumti"=
   <cx "pro-sumti, lerfu as">  XE "pro-sumti: lerfu as"   <cx "lerfu as pro=
-sumti, contrasted with ko'a-series in explicit assignment of">  XE "lerfu =
as pro-sumti: contrasted with ko'a-series in explicit assignment of"   <cx =
"ko'a-series pro-sumti, contrasted with lerfu as pro-sumti in explicit assi=
gnment of">  XE "ko'a-series pro-sumti: contrasted with lerfu as pro-sumti =
in explicit assignment of"   <cx "lerfu as pro-sumti, implicit assignment o=
f antecedent">  XE "lerfu as pro-sumti: implicit assignment of antecedent" =
  As is explained in <a href=3Dchap17.html>Chapter 17</a>, the words for Lo=
jban letters, belonging to selma'o BY and certain related selma'o, are also=
 usable as assignable pro-sumti. The main difference between letter pro-sum=
ti and ko'a-series pro-sumti is that, in the absence of an explicit assignm=
ent, letters are taken to refer to the most recent name or description sumt=
i beginning with the same letter:
+<p>
+<pre><a name=3De5d8>5.8)	mi viska le gerku  .i gy. cusku zo arf.
+	I see the dog.  D expresses the-word =93Arf!=94.
+</pre><cx "lerfu as pro-sumti, explicit assignment of antecedent">  XE "le=
rfu as pro-sumti: explicit assignment of antecedent"   <cx "goi, use in ass=
igning lerfu as pro-sumti">  XE "goi: use in assigning lerfu as pro-sumti" =
  The Lojban word =93gerku=94 begins with =93g=94, so the antecedent of =93=
gy.=94, the cmavo for the letter =93g=94, must be =93le gerku=94. In the En=
glish translation, we use the same principle to refer to the dog as =93D=94=
. Of course, in case of ambiguity, =93goi=94 can be used to make an explici=
t assignment.
+<p>
+<cx "goi, use in assigning name">  XE "goi: use in assigning name"   <cx "=
name, assigning with goi">  XE "names: assigning with goi"   Furthermore, =
=93goi=94 can even be used to assign a name:
+<p>
+<pre><a name=3De5d9>5.9)	le ninmu goi la sam. cu klama le zarci
+	The woman also-known-as Sam goes to-the store.
+	The woman, whom I'll call Sam, goes to the store.
+</pre><p>This usage does not imply that the woman's name is Sam, or even t=
hat the speaker usually calls the woman =93Sam=94. =93Sam=94 is simply a na=
me chosen, as if at random, for use in the current context only.
+<p>
+<a name=3Ds6><h3>Anaphoric pro-sumti and pro-bridi: the ri-series and the =
go'i-series</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ri	KOhA	ri-series		(repeats last sumti)
+	ra	KOhA	ri-series		(repeats previous sumti)
+	ru	KOhA	ri-series		(repeats long-ago sumti)
+	go'i	GOhA	go'i-series	(repeats last bridi)
+	go'a	GOhA	go'i-series	(repeats previous bridi)
+	go'u	GOhA	go'i-series	(repeats long-ago bridi)
+	go'e	GOhA	go'i-series	(repeats last-but-one bridi)
+	go'o	GOhA	go'i-series	(repeats future bridi)
+	nei	GOhA	go'i-series	(repeats current bridi)
+	no'a	GOhA	go'i-series	(repeats outer bridi)
+	ra'o	RAhO			pro-cmavo update
+</pre><cx "anaphora, definition">  XE "anaphora: definition"   <cx "pronou=
ns, as anaphora">  XE "pronouns: as anaphora"   <cx "anaphora, pro-sumti ri=
-series as">  XE "anaphora: pro-sumti ri-series as"   <cx "anaphora, pro-br=
idi go'i-series as">  XE "anaphora: pro-bridi go'i-series as"   <cx "ri-ser=
ies pro-sumti">  XE "ri-series pro-sumti"   <cx "go'i-series pro-bridi">  X=
E "go'i-series pro-bridi"   <cx "ri-series pro-sumti, compared with ti-seri=
es in word formation">  XE "ri-series pro-sumti: compared with ti-series in=
 word formation"   <cx "go'i-series pro-bridi, compared with ri-series in w=
ord formation">  XE "go'i-series pro-bridi: compared with ri-series in word=
 formation"   The term =93anaphora=94 literally means =93repetition=94, but=
 is used in linguistics to refer to pronouns whose significance is the repe=
tition of earlier words, namely their antecedents. Lojban provides three pr=
o-sumti anaphora, =93ri=94, =93ra=94, and =93ru=94; and three corresponding=
 pro-bridi anaphora, =93go'i=94, =93go'a=94, and =93go'u=94. These cmavo re=
veal the same vowel pattern as the ti-series, but the =93distances=94 refer=
red to are not physical distances, but distances from the anaphoric cmavo t=
o its antecedent.
+<p>
+<lx "ri">  XE "ri"   <cx "ri, referent of">  XE "ri: referent of"   The cm=
avo =93ri=94 is the simplest of these; it has the same referent as the last=
 complete sumti appearing before the =93ri=94:
+<p>
+<pre><a name=3De6d1>6.1)	la .alis. sipna le ri kumfa
+	Alice sleeps-in the of-[repeat last sumti] room.
+	Alice sleeps in her room.
+</pre><cx "ri-series pro-sumti, effect of use on meaning">  XE "ri-series =
pro-sumti: effect of use on meaning"   The =93ri=94 in <a href=3D#e6d1>Exam=
ple 6.1 </a>is equivalent to repeating the last sumti, which is =93la .alis=
.=94, so <a href=3D#e6d1>Example 6.1 </a>is equivalent to:
+<p>
+<pre><a name=3De6d2>6.2)	la .alis. sipna le la .alis. kumfa
+	Alice sleeps-in the of-Alice room.
+	Alice sleeps in Alice's room.
+</pre><cx "ri, non-self-reference of">  XE "ri: non-self-reference of"   <=
cx "ri-series pro-sumti, and order of possible referents">  XE "ri-series p=
ro-sumti: and order of possible referents"   Note that =93ri=94 does not re=
peat =93le ri kumfa=94, because that sumti is not yet complete when =93ri=
=94 appears. This prevents =93ri=94 from getting entangled in paradoxes of =
self-reference. (There are plenty of other ways to do that!) Note also that=
 sumti within other sumti, as in quotations, abstractions, and the like, ar=
e counted in the order of their beginnings; thus a lower level sumti like =
=93la alis.=94 in <a href=3D#e6d2>Example 6.2 </a>is considered to be more =
recent than a higher level sumti that contains it.
+<p>
+<cx "ri-series pro-sumti, possible referents of">  XE "ri-series pro-sumti=
: possible referents of"   <cx "ri-series pro-sumti, non-allowable referent=
s of">  XE "ri-series pro-sumti: non-allowable referents of"   <cx "ri-seri=
es pro-sumti, effect of ko'a-series pro-sumti on">  XE "ri-series pro-sumti=
: effect of ko'a-series pro-sumti on"   <cx "ko'a-series pro-sumti, effect =
on ri-series pro-sumti">  XE "ko'a-series pro-sumti: effect on ri-series pr=
o-sumti"   <cx "ri-series pro-sumti, effect of lerfu pro-sumti on">  XE "ri=
-series pro-sumti: effect of lerfu pro-sumti on"   <cx "lerfu pro-sumti, ef=
fect on ri-series pro-sumti">  XE "lerfu pro-sumti: effect on ri-series pro=
-sumti"   Certain sumti are ignored by =93ri=94; specifically, most of the =
other cmavo of KOhA, and the almost-grammatically-equivalent lerfu words of=
 selma'o BY. It is simpler just to repeat these directly:
+<p>
+<pre><a name=3De6d3>6.3)	mi prami mi
+	I love me.
+	I love myself.
+</pre><cx "ri-series pro-sumti, effect of ti-series pro-sumti on">  XE "ri=
-series pro-sumti: effect of ti-series pro-sumti on"   <cx "ti-series pro-s=
umti, effect on ri-series pro-sumti">  XE "ti-series pro-sumti: effect on r=
i-series pro-sumti"   <cx "ri-series pro-sumti, effect of other ri-series p=
ro-sumti on">  XE "ri-series pro-sumti: effect of other ri-series pro-sumti=
 on"   <cx "ri-series pro-sumti, effect on other ri-series pro-sumti">  XE =
"ri-series pro-sumti: effect on other ri-series pro-sumti"   However, the c=
mavo of the ti-series can be picked up by =93ri=94, because you might have =
changed what you are pointing at, so repeating =93ti=94 may not be effectiv=
e. Likewise, =93ri=94 itself (or rather its antecedent) can be repeated by =
a later =93ri=94; in fact, a string of =93ri=94 cmavo with no other interve=
ning sumti always all repeat the same sumti:
+<p>
+<pre><a name=3De6d4>6.4)	la djan. viska le tricu  .i ri se jadni le ri jim=
ca
+	John sees the tree.  [repeat last] is-adorned-by the of-[repeat last] bra=
nch
+	John sees the tree.  It is adorned by its branches.
+</pre><p>Here the second =93ri=94 has as antecedent the first =93ri=94, wh=
ich has as antecedent =93le tricu=94. All three refer to the same thing: a =
tree.
+<p>
+<cx "ri, subscripting for referring further back">  XE "ri: subscripting f=
or referring further back"   <cx "subscripts, on ri">  XE "subscripts: on r=
i"   To refer to the next-to-last sumti, the third-from-last sumti, and so =
on, =93ri=94 may be subscripted (subscripts are explained in <a href=3Dchap=
19.html>Chapter 19</a>):
+<p>
+<pre><a name=3De6d5>6.5)	lo smuci .i lo forca .i la rik. pilno rixire
+		.i la .alis. pilno riximu
+	A spoon. A fork. Rick uses [repeat next-to-last].
+		Alice uses [repeat fifth-from-last].
+</pre><p>Here =93rixire=94, or =93ri-sub-2=94, skips =93la rik.=94 to reac=
h =93lo forca=94. In the same way, =93riximu=94, or =93ri-sub-5=94, skips =
=93la .alis.=94, =93rixire=94, =93la rik.=94, and =93lo forca=94 to reach =
=93lo smuci=94. As can clearly be seen, this procedure is barely practicabl=
e in writing, and would break down totally in speech.
+<p>
+<lx "ra">  XE "ra"   <lx "ru">  XE "ru"   <cx "ra, practical referent conv=
entions">  XE "ra: practical referent conventions"   <cx "ru, practical ref=
erent conventions">  XE "ru: practical referent conventions"   Therefore, t=
he vaguer =93ra=94 and =93ru=94 are also provided. The cmavo =93ra=94 repea=
ts a recently used sumti, and =93ru=94 one that was further back in the spe=
ech or text. The use of =93ra=94 and =93ru=94 forces the listener to guess =
at the referent, but makes life easier for the speaker. Can =93ra=94 refer =
to the last sumti, like =93ri=94? The answer is no if =93ri=94 has also bee=
n used. If =93ri=94 has not been used, then =93ra=94 might be the last sumt=
i. Likewise, if =93ra=94 has been used, then any use of =93ru=94 would repe=
at a sumti earlier than the one =93ra=94 is repeating. A more reasonable ve=
rsion of Example 6.5, but one that depends more on context, is:
+
+<p>
+<pre><a name=3De6d6>6.6)	lo smuci .i lo forca .i la rik. pilno ra
+		.i la .alis. pilno ru
+	A spoon. A fork. Rick uses [some previous thing].
+		Alice uses [some more remote thing].
+</pre><p>In <a href=3D#e6d6>Example 6.6</a>, the use of =93ra=94 tells us =
that something other than =93la rik.=94 is the antecedent; =93lo forca=94 i=
s the nearest sumti, so it is probably the antecedent. Similarly, the antec=
edent of =93ru=94 must be something even further back in the utterance than=
 =93lo forca=94, and =93lo smuci=94 is the obvious candidate.
+<p>
+<cx "ri-series pro-sumti, assigning for permanent reference">  XE "ri-seri=
es pro-sumti: assigning for permanent reference"   <lx "goi">  XE "goi"   T=
he meaning of =93ri=94 must be determined every time it is used. Since =93r=
a=94 and =93ru=94 are more vaguely defined, they may well retain the same m=
eaning for a while, but the listener cannot count on this behavior. To make=
 a permanent reference to something repeated by =93ri=94, =93ra=94, or =93r=
u=94, use =93goi=94 and a ko'a-series cmavo:
+<p>
+<pre><a name=3De6d7>6.7)	la .alis. klama le zarci  .i ri goi ko'a blanu
+	Alice goes-to the store.  It-last-mentioned also-known-as it-1 is-blue.
+</pre>allows the store to be referred to henceforth as =93ko'a=94 without =
ambiguity. <a href=3D#e6d7>Example 6.7 </a>is equivalent to <a href=3D#e5d1=
>Example 5.1 </a>and eliminates any possibility of =93ko'a=94 being interpr=
eted by the listener as referring to Alice.
+<p>
+<lx "go'i">  XE "go'i"   <lx "go'a">  XE "go'a"   <lx "go'u">  XE "go'u"  =
 <cx "go'i-series pro-bridi, compared with ri-series pro-sumti in rules of =
reference">  XE "go'i-series pro-bridi: compared with ri-series pro-sumti i=
n rules of reference"   <cx "go'i-series pro-bridi, referent of">  XE "go'i=
-series pro-bridi: referent of"   <cx "go'i-series pro-bridi, effect of sub=
-clauses on">  XE "go'i-series pro-bridi: effect of sub-clauses on"   <cx "=
go'i-series pro-bridi, as main-bridi anaphora only">  XE "go'i-series pro-b=
ridi: as main-bridi anaphora only"   <cx "go'i-series pro-bridi, effect of =
sumti of referent bridi on">  XE "go'i-series pro-bridi: effect of sumti of=
 referent bridi on"   <cx "go'i, as affirmative answer to yes/no question">=
  XE "go'i: as affirmative answer to yes/no question"   <cx "question, answ=
ering with go'i">  XE =93questions: answering with go'i"   <cx "answer, go'=
i for yes/no questions">  XE "answers: go'i for yes/no questions"   The cma=
vo =93go'i=94, =93go'a=94, and =93go'u=94 follow exactly the same rules as =
=93ri=94, =93ra=94, and =93ru=94, except that they are pro-bridi, and there=
fore repeat bridi, not sumti =97 specifically, main sentence bridi. Any bri=
di that are embedded within other bridi, such as relative clauses or abstra=
ctions, are not counted. Like the cmavo of the broda-series, the cmavo of t=
he go'i-series copy all sumti with them. This makes =93go'i=94 by itself co=
nvenient for answering a question affirmatively, or for repeating the last =
bridi, possibly with new sumti:
+<p>
+<pre><a name=3De6d8>6.8)	xu zo djan. cmene do  .i go'i
+	[True-false?] The-word =93John=94 is-the-name of you?  [repeat last bridi=
].
+	Is John your name? Yes.
+<a name=3De6d9>6.9)	mi klama le zarci .i do go'i
+	I go-to the store. You [repeat last bridi].
+	I go to the store. You, too.
+</pre><cx "go'i-series pro-sumti, assigning for permanent reference">  XE =
"go=92i-series pro-bridi : assigning for permanent reference"   <lx "cei"> =
 XE "cei"   Note that <a href=3D#e6d9>Example 6.9 </a>means the same as <a =
href=3D#e5d6>Example 5.6</a>, but without the bother of assigning an actual=
 broda-series word to the first bridi. For long-term reference, use =93go'i=
 cei broda=94 or the like, analogously to =93ri goi ko'a=94 in <a href=3D#e=
6d7>Example 6.7</a>.
+<p>
+<lx "go'ixire">  XE "go'ixire"   <lx "go'e">  XE "go'e"   The remaining fo=
ur cmavo of the go'i-series are provided for convenience or for achieving s=
pecial effects. The cmavo =93go'e=94 means the same as =93go'ixire=94: it r=
epeats the last bridi but one. This is useful in conversation:
+<p>
+<pre><a name=3De6d10>6.10)	A:  mi ba klama le zarci
+	B:  mi nelci le si'o mi go'i
+	A:  do go'e
+	A:  I [future] go-to the store.
+	B:  I like the concept-of I [repeat last bridi].
+	A:  You [repeat last bridi but one].
+	A:  I am going to the store.
+	B:  I like the idea of my going.
+	A:  You'll go, too.
+</pre><p>Here B's sentence repeats A's within an abstraction (explained in=
 Chapter 11): =93le si'o mi go'i=94 means =93le si'o mi klama le zarci=94. =
Why must B use the word =93mi=94 explicitly to replace the x1 of =93mi klam=
a le zarci=94, even though it looks like =93mi=94 is replacing =93mi=94? Be=
cause B's =93mi=94 refers to B, whereas A's =93mi=94 refers to A. If B said=
:
+<p>
+<pre><a name=3De6d11>6.11)	mi nelci le si'o go'i
+</pre>that would mean:
+<p>
+<dl compact><dt>	<dd>I like the idea of your going to the store.
+</dl><cx "go'i-series pro-bridi, as repeating referent concept">  XE "go'i=
-series pro-bridi: as repeating referent concept"   The repetition signalle=
d by =93go'i=94 is not literally of words, but of concepts. Finally, A repe=
ats her own sentence, but with the x1 changed to =93do=94, meaning B. Note =
that in <a href=3D#e6d10>Example 6.10</a>, the tense =93ba=94 (future time)=
 is carried along by both =93go'i=94 and =93go'e=94.
+<p>
+<cx "go'i-series pro-bridi, as basis for description">  XE "go'i-series pr=
o-bridi: as basis for description"   <cx "descriptions, based on go'i-serie=
s pro-bridi">  XE "descriptions: based on go'i-series pro-bridi"   Descript=
ions based on go'i-series cmavo can be very useful for repeating specific s=
umti of previous bridi:
+<p>
+<ex "black cat">  XE "black cat: example=93  <pre><a name=3De6d12>6.12)	le=
 xekri mlatu cu klama le zarci
+		.i le go'i cu cadzu le bisli
+	The black cat goes-to the store.
+		That-described-as-the-x1-place-of [repeat last bridi] walks-on the ice.
+	The black cat goes to the store.  It walks on the ice.
+</pre><p>Here the =93go'i=94 repeats =93le xekri mlatu cu klama le zarci=
=94, and since =93le=94 makes the x1 place into a description, and the x1 p=
lace of this bridi is =93le xekri mlatu=94, =93le go'i=94 means =93le xekri=
 mlatu=94.
+<p>
+<lx "go'o">  XE "go'o"   <lx "nei">  XE "nei"   <lx "no'a">  XE "no'a"   <=
cx "no'a, contrasted with other members of go'i-series in possible referent=
s">  XE "no'a: contrasted with other members of go'i-series in possible ref=
erents"   <cx "go'i-series pro-bridi, no'a as exception to only main-bridi =
anaphora">  XE "go'i-series pro-bridi: no'a as exception to only main-bridi=
 anaphora"   <cx "go'i-series pro-bridi, as main-bridi anaphora only, excep=
tion">  XE "go'i-series pro-bridi: as main-bridi anaphora only, exception" =
  The cmavo =93go'o=94, =93nei=94, and =93no'a=94 have been little used so =
far. They repeat respectively some future bridi, the current bridi, and the=
 bridi that encloses the current bridi (=93no'a=94, unlike the other member=
s of the go'i- series, can repeat non-sentence bridi). Here are a few examp=
les:
+<p>
+<pre><a name=3De6d13>6.13)	mi nupre le nu mi go'o
+		.i ba dunda le djini le bersa
+		.i ba dunda le zdani le tixnu
+	I promise the event-of I [repeat future bridi]
+		[Future] give the money to-the son
+		[Future] give the house to-the daughter
+	I promise to do the following:
+		Give the money to my son.
+		Give the house to my daughter.
+</pre>(Note: The Lojban does not contain an equivalent of the =93my=94 in =
the colloquial English; it leaves the fact that it is the speaker's son and=
 daughter that are referred to implicit. To make the fact explicit, use =93=
le bersa/tixnu be mi=94.)
+<p>
+For good examples of =93nei=94 and =93no'a=94, we need nested bridi contex=
ts:
+<p>
+<pre><a name=3De6d14>6.14)	mi se pluka le nu do pensi
+		le nu nei kei pu le nu do zukte
+	I am-pleased-by the event-of (you think-about
+		(the event-of [main bridi]) before the-event of (your acting).
+	I am pleased that you thought about whether I
+		would be pleased (about=A0<dots>=85</dots>) before you acted.
+
+<a name=3De6d15>6.15)	mi ba klama ca le nu do no'a
+	I [future] go [present] the event-of you [repeats outer bridi]
+	I will go when you do.
+</pre><lx "ra'o">  XE "ra'o"   <cx "go'i-series pro-bridi, reinterpreting =
sumti references with ra'o">  XE "go'i-series pro-bridi: reinterpreting sum=
ti references with ra'o"   <cx "ra'o, for reinterpreting go'i-series pro-br=
idi sumti references">  XE "ra'o: for reinterpreting go'i-series pro-bridi =
sumti references"   <cx "go'i, contrasted with go'i ra'o">  XE "go'i: contr=
asted with go'i ra'o"   <cx "go'i ra'o, contrasted with go'i">  XE "go'i ra=
'o: contrasted with go'i"   Finally, =93ra'o=94 is a cmavo that can be appe=
nded to any go'i-series cmavo, or indeed any cmavo of selma'o GOhA, to sign=
al that pro-sumti or pro-bridi cmavo in the antecedent are to be repeated l=
iterally and reinterpreted in their new context. Normally, any pro-sumti us=
ed within the antecedent of the pro-bridi keep their meanings intact. In th=
e presence of =93ra'o=94, however, their meanings must be reinterpreted wit=
h reference to the new environment. If someone says to you:
+<p>
+<pre><a name=3De6d16>6.16)	mi ba lumci lemi karca
+	I will wash my car.
+</pre>you might reply either:
+<p>
+<pre><a name=3De6d17>6.17)	mi go'i
+	I will wash your car.
+</pre>or:
+<p>
+<pre><a name=3De6d18>6.18)	mi go'i ra'o
+	I will wash my car.
+</pre><p>The =93ra'o=94 forces the second =93mi=94 from the original bridi=
 to mean the new speaker rather than the former speaker. This means that =
=93go'e ra'o=94 would be an acceptable alternative to =93do go'e=94 in B's =
statement in <a href=3D#e6d10>Example 6.10</a>.
+<p>
+<cx "ri-series pro-sumti, in quotations">  XE "ri-series pro-sumti: in quo=
tations"   <cx "go'i-series pro-sumti, in quotations">  XE "go=92i-series p=
ro-bridi : in quotations"   The anaphoric pro-sumti of this section can be =
used in quotations, but never refer to any of the supporting text outside t=
he quotation, since speakers presumably do not know that they may be quoted=
 by someone else.
+<p>
+<cx "ri-series pro-sumti, in quotation series">  XE "ri-series pro-sumti: =
in quotation series"   <cx "go'i-series pro-sumti, in quotation series">  X=
E "go=92i-series pro-bridi : in quotation series"   However, a =93ri=94-ser=
ies or =93go'a=94-series reference within a quotation can refer to somethin=
g mentioned in an earlier quotation if the two quotations are closely relat=
ed in time and context. This allows a quotation to be broken up by narrativ=
e material without interfering with the pro-sumti within it. Here's an exam=
ple:
+<p>
+<pre><a name=3De6d19>6.19)	la djan. cusku lu mi klama le zarci li'u
+		.i la .alis. cusku lu mi go'i li'u
+	John says [quote] I go-to the store [unquote].
+		Alice says [quote] I [repeat] [unquote].
+	John says, =93I am going to the store.=94  Alice says, =93Me too.=94
+</pre><cx "ri-series pro-sumti, in narrative about quotation">  XE "ri-ser=
ies pro-sumti: in narrative about quotation"   <cx "go'i-series pro-sumti, =
in narrative about quotation">  XE "go=92i-series pro-bridi : in narrative =
about quotation"   Of course, there is no problem with narrative material r=
eferring to something within a quotation: people who quote, unlike people w=
ho are quoted, are aware of it.
+<p>
+<a name=3Ds7><h3>Indefinite pro-sumti and pro-bridi: the zo'e-series and t=
he co'e-series</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	zo'e	KOhA	zo'e-series	the obvious value
+	zu'i	KOhA	zo'e-series	the typical value
+	zi'o	KOhA	zo'e-series	the nonexistent value
+	co'e	GOhA	co'e-series	has the obvious relationship
+</pre><cx "zo'e-series pro-sumti">  XE "zo'e-series pro-sumti"   <cx "co'e=
-series pro-bridi">  XE "co'e-series pro-bridi"   <cx "indefinite pro-sumti=
">  XE "indefinite pro-sumti"   <cx "elliptical pro-sumti">  XE "elliptical=
 pro-sumti"   <cx "pro-sumti, unspecified">  XE "pro-sumti: unspecified"   =
<cx "indefinite pro-bridi">  XE "indefinite pro-bridi"   <cx "elliptical pr=
o-bridi">  XE "elliptical pro-bridi"   <lx "zo'e">  XE "zo'e"   <cx "zo'e, =
as place-holder for sumti">  XE "zo'e: as place-holder for sumti"   <cx "el=
liptical sumti">  XE "elliptical sumti"   <cx "elliptical value, contrasted=
 with typical value for sumti">  XE "elliptical value: contrasted with typi=
cal value for sumti"   <cx "typical value, contrasted with elliptical value=
 for sumti">  XE "typical value: contrasted with elliptical value for sumti=
"   The cmavo of the zo'e-series represent indefinite, unspecified sumti. T=
he cmavo =93zo'e=94 represents an elliptical value for this sumti place; it=
 is the optional spoken place holder when a sumti is skipped without being =
specified. Note that the elliptical value is not always the typical value. =
The properties of ellipsis lead to an elliptical sumti being defined as =93=
whatever I want it to mean but haven't bothered to figure out, or figure ou=
t how to express=94.
+<p>
+<lx "zu'i">  XE "zu'i"   <cx "pro-sumti, typical">  XE "pro-sumti: typical=
"   <cx "typical sumti">  XE "typical sumti"   The cmavo =93zu'i=94, on the=
 other hand, represents the typical value for this place of this bridi:
+<p>
+<pre><a name=3De7d1>7.1)	mi klama le bartu be le zdani le nenri be le zdan=
i
+		zu'i zu'i
+	I go to-the outside of the house from-the inside of the house
+		[by-typical-route] [by-typical-means]
+</pre><p>In <a href=3D#e7d1>Example 7.1</a>, the first =93zu'i=94 probably=
 means something like =93by the door=94, and the second =93zu'i=94 probably=
 means something like =93on foot=94, those being the typical route and mean=
s for leaving a house. On the other hand, if you are at the top of a high r=
ise during a fire, neither =93zu'i=94 is appropriate. It's also common to u=
se =93zu'i=94 in =93by standard=94 places.
+<p>
+<cx "zi'o">  XE "zi'o"   <cx "sumti, irrelevant to relationship">  XE "sum=
ti: irrelevant to relationship"   <cx "irrelevant, specifying of sumti plac=
e">  XE "irrelevant: specifying of sumti place"   Finally, the cmavo =93zi'=
o=94 represents a value which does not even exist. When a bridi fills one o=
f its places with =93zi'o=94, what is really meant is that the selbri has a=
 place which is irrelevant to the true relationship the speaker wishes to e=
xpress. For example, the place structure of =93zbasu=94 is
+<p>
+<dl compact><dt>zbasu:  <dd>actor x1 makes x2 from materials x3
+</dl><ex "living things">  XE "living things: example=93  Consider the sen=
tence
+<p>
+<dl compact><dt>	<dd>Living things are made from cells.
+</dl><p>This cannot be correctly expressed as:
+<p>
+<pre><a name=3De7d2>7.2)	loi jmive cu se zbasu [zo'e] fi loi selci
+	The-mass-of living-things is-made [by-something] from the-mass-of cells
+</pre>because the =93zo'e=94, expressed or understood, in <a href=3D#e7d2>=
Example 7.2 </a>indicates that there is still a =93maker=94 in this relatio=
nship. We do not generally suppose, however, that someone =93makes=94 livin=
g things from cells. The best answer is probably to find a different selbri=
, one which does not imply a =93maker=94: however, an alternative strategy =
is to use =93zi'o=94 to eliminate the maker place:
+<p>
+<pre><a name=3De7d3>7.3)	loi jmive cu se zbasu zi'o loi selci
+	The-mass-of living-things is-made [without-maker] from the-mass-of cells.
+</pre><cx "zi'o, as creating new selbri">  XE "zi'o: as creating new selbr=
i"   Note: The use of =93zi'o=94 to block up, as it were, one place of a se=
lbri actually creates a new selbri with a different place structure. Consid=
er the following examples:
+<p>
+<pre><a name=3De7d4>7.4)	mi zbasu le dinju loi mudri
+	I make the building from-some-of-the-mass-of wood.
+	I make the building out of wood.
+<a name=3De7d5>7.5)	zi'o zbasu le dinju loi mudri
+	[without-maker] makes the building from-some-of-the-mass-of wood.
+	The building is made out of wood.
+
+<a name=3De7d6>7.6)	mi zbasu zi'o loi mudri
+	I make [without-thing-made] from-some-of-the-mass-of wood.
+	I build using wood.
+<a name=3De7d7>7.7)	mi zbasu loi mudri zi'o
+	I make the building [without-material].
+	I make the building.
+</pre><p>If <a href=3D#e7d4>Example 7.4 </a>is true, then <a href=3D#e7d{5=
}>Examples 7.5 </a>through <a href=3D#e7d7>7.7 </a>must be true also. Howev=
er, <a href=3D#e7d3>Example 7.3 </a>does not correspond to any sentence wit=
h three regular (non-=93zi'o=94) sumti.
+<p>
+<lx "co'e">  XE "co'e"   <cx "selbri, omitting with co'e">  XE "selbri: om=
itting with co'e"   <cx "co'e, as selbri place-holder">  XE "co'e: as selbr=
i place-holder"   The pro-bridi =93co'e=94 (which by itself constitutes the=
 co'e-series of selma'o GOhA) represents the elliptical selbri. Lojban gram=
mar does not allow the speaker to merely omit a selbri from a bridi, althou=
gh any or all sumti may be freely omitted. Being vague about a relationship=
 requires the use of =93co'e=94 as a selbri place-holder:
+<p>
+<pre><a name=3De7d8>7.8)	mi troci le nu mi co'e le vorme
+	I try the event-of my [doing-the-obvious-action] to-the door.
+	I try the door.
+</pre><p>The English version means, and the Lojban version probably means,=
 that I try to open the door, but the relationship of opening is not actual=
ly specified; the Lojbanic listener must guess it from context. Lojban, unl=
ike English, makes it clear that there is an implicit action that is not be=
ing expressed.
+<p>
+<cx "co'e, rationale for word form">  XE "co'e: rationale for word form"  =
 The form of =93co'e=94 was chosen to resemble =93zo'e=94; the cmavo =93do'=
e=94 of selma'o BAI (see <a href=3Dchap9.html>Chapter 9</a>) also belongs t=
o the same group of cmavo.
+<p>
+<cx "do'i, compared with zo'e-series as indefinite pro-sumti">  XE "do'i: =
compared with zo'e-series as indefinite pro-sumti"   <cx "zo'e-series, comp=
ared with do'i as indefinite pro-sumti">  XE "zo'e-series: compared with do=
'i as indefinite pro-sumti"   Note that =93do'i=94, of the di'u-series, is =
also a kind of indefinite pro-sumti: it is indefinite in referent, but is r=
estricted to referring only to an utterance.
+<p>
+<a name=3Ds8><h3>Reflexive and reciprocal pro-sumti: the vo'a-series</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	vo'a	KOhA	vo'a-series	x1 of this bridi
+	vo'e	KOhA	vo'a-series	x2 of this bridi
+	vo'i	KOhA	vo'a-series	x3 of this bridi
+	vo'o	KOhA	vo'a-series	x4 of this bridi
+	vo'u	KOhA	vo'a-series	x5 of this bridi
+	soi	SOI			reciprocity
+	se'u	SEhU			soi terminator
+</pre><cx "anaphora, pro-sumti vo'a-series as">  XE "anaphora: pro-sumti v=
o'a-series as"   <cx "pro-sumti, vo'a-series">  XE "pro-sumti: vo'a-series"=
   <cx "pro-sumti, referring to place of same bridi with vo'a-series">  XE =
"pro-sumti: referring to place of same bridi with vo'a-series"   <cx "refle=
xive pro-sumti">  XE "reflexive pro-sumti"   <cx "reciprocal pro-sumti">  X=
E "reciprocal pro-sumti"   The cmavo of the vo'a-series are pro-sumti anaph=
ora, like those of the ri-series, but have a specific function. These cmavo=
 refer to the other places of the same bridi; the five of them represent up=
 to five places. The same vo'a-series cmavo mean different things in differ=
ent bridi. Some examples:
+<p>
+<lx "vo'a">  XE "vo'a"   <ex "wash self">  XE "wash self: example=93  <pre=
><a name=3De8d1>8.1)	mi lumci vo'a
+	I wash myself
+</pre><lx "vo'e">  XE "vo'e"   <pre><a name=3De8d2>8.2)	mi klama le zarci =
vo'e
+	I go to the store from itself [by some route unspecified].
+</pre><cx "pro-sumti, referring to place of different bridi with go'i-seri=
es">  XE "pro-sumti: referring to place of different bridi with go'i-series=
"   To refer to places of neighboring bridi, constructions like =93le se go=
'i ku=94 do the job: this refers to the 2nd place of the previous main brid=
i, as explained in <a href=3D#s6>Section 6</a>.
+<p>
+<lx "soi">  XE "soi"   <lx "SOI">  XE "SOI selma=92o"   <cx "reciprocity, =
expressing with vo'a-series pro-sumti and soi">  XE "reciprocity: expressin=
g with vo'a-series pro-sumti and soi"   <cx "soi, use in expressing recipro=
city with vo'a-series pro-sumti">  XE "soi: use in expressing reciprocity w=
ith vo'a-series pro-sumti"   <cx "vo'a-series pro-sumti, use in expressing =
reciprocity with soi">  XE "vo'a-series pro-sumti: use in expressing recipr=
ocity with soi"   <cx "=93vice versa=94, expressing with vo'a-series pro-su=
mti and soi">  XE "vice versa: English, expressing with vo'a-series pro-sum=
ti and soi"   The cmavo of the vo'a-series are also used with =93soi=94 (of=
 selma'o SOI) to precisely express reciprocity, which in English is impreci=
sely expressed with a discursive phrase like =93vice versa=94:
+<p>
+<ex "vice versa">  XE "vice versa: example=93  <pre><a name=3De8d3>8.3)	mi=
 prami do soi vo'a vo'e
+	I love you [reciprocity] [x1 of this bridi] [x2 of this bridi].
+	I love you and vice versa (swapping =93I=94 and =93you=94).
+</pre><cx "soi with one following sumti, convention">  XE "soi with one fo=
llowing sumti: convention"   The significance of =93soi vo'a vo'e=94 is tha=
t the bridi is still true even if the x1 (specified by =93vo'a=94) and the =
x2 (specified by =93vo'e=94) places are interchanged. If only a single sumt=
i follows =93soi=94, then the sumti immediately preceding =93soi=94 is unde=
rstood to be one of those involved:
+<p>
+<pre><a name=3De8d4>8.4)	mi prami do soi vo'a
+	I love you [reciprocity] [x1 of this bridi].
+</pre>again involves the x1 and x2 places.
+<p>
+<cx "reciprocity, expressing with soi">  XE "reciprocity: expressing with =
soi"   <cx "soi, use in expressing reciprocity">  XE "soi: use in expressin=
g reciprocity"   Of course, other places can be involved, and other sumti m=
ay be used in place of vo'a-series cmavo, provided those other sumti can be=
 reasonably understood as referring to the same things mentioned in the bri=
di proper. Here are several examples that mean the same thing:
+<p>
+<pre><a name=3De8d5>8.5)	mi bajykla ti ta soi vo'e
+	mi bajykla ti ta soi vo'e vo'i
+	soi vo'e vo'i mi bajra ti ta
+	I runningly-go to this from that and vice versa (to that from this).
+</pre><lx "se'u">  XE "se'u"   <lx "SEhU">  XE "SEhU selma=92o"   <cx "se'=
u, as elidable terminator for soi">  XE "se'u: as elidable terminator for s=
oi"   <cx "se'u, elidability considerations">  XE "se'u: elidability consid=
erations"   The elidable terminator for =93soi=94 is =93se'u=94 (selma'o SE=
hU), which is normally needed only if there is just one sumti after the =93=
soi=94, and the =93soi=94 construction is not at the end of the bridi. Cons=
tructions using =93soi=94 are free modifiers, and as such can go almost any=
where. Here is an example where =93se'u=94 is required:
+<p>
+<pre><a name=3De8d6>8.6)	mi bajykla ti soi vo'i se'u ta
+	I runningly-go to-this [reciprocity] [x3 of this bridi] from-that
+	I runningly-go to this from that and vice versa.
+</pre><a name=3Ds9><h3>sumti and bridi questions: =93ma=94 and =93mo=94</h=
3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ma	KOhA	sumti question
+	mo	GOhA	bridi question
+</pre><lx "ma">  XE "ma"   <cx "ma, as sumti question">  XE "ma: as sumti =
question"   <cx "questions, sumti">  XE "questions: sumti"   Lojban questio=
ns are more fully explained in <a href=3Dchap19.html>Chapter 19 </a>, but =
=93ma=94 and =93mo=94 are listed in this chapter for completeness. The cmav=
o =93ma=94 asks for a sumti to make the bridi true:
+<p>
+<pre><a name=3De9d1>9.1)	do klama ma
+	You go to-what-destination?
+	Where are you going?
+</pre><lx "mo">  XE "mo"   <cx "mo, as selbri question">  XE "mo: as selbr=
i question"   <cx "questions, selbri">  XE "questions: selbri"   <cx "mo, c=
ompared with go'i in overriding of arguments">  XE "mo: compared with go'i =
in overriding of arguments"   <cx "go'i, compared with mo in overriding of =
arguments">  XE "go'i: compared with mo in overriding of arguments"   The c=
mavo =93mo=94, on the other hand, asks for a selbri which makes the questio=
n bridi true. If the answer is a full bridi, then the arguments of the answ=
er override the arguments in the question, in the same manner as the go'i-s=
eries cmavo. A simple example is:
+<p>
+<pre><a name=3De9d2>9.2)	do mo
+	What predicate is true as applied to you?
+	How are you?
+	What are you doing?
+	What are you?
+</pre><a href=3D#e9d2>Example 9.2 </a>is a truly pregnant question that wi=
ll have several meanings depending on context.
+<p>
+(One thing it probably does not mean is =93Who are you?=94 in the sense =
=93What is your name/identity?=94, which is better expressed by:
+<p>
+<ex "what is your name">  XE "what is your name: example=93  <pre><a name=
=3De9d3>9.3)	ma cmene do
+	What sumti is-the-name-of you?
+	What is your name?
+</pre>or even
+<p>
+<pre><a name=3De9d4>9.4)	doi ma
+	O [what sumti?]
+</pre>which uses the vocative =93doi=94 to address someone, and simultaneo=
usly asks who the someone is.)
+<p>
+A further example of =93mo=94:
+<p>
+<pre><a name=3De9d5>9.5)	lo mo prenu cu darxi do  .i barda
+	A [what selbri?] type-of person hit you?  (Observative:) A big thing.
+	Which person hit you? The big one.
+</pre><cx "multiple questions in one bridi, expressing">  XE "multiple que=
stions in one bridi: expressing"   <cx "multiple ma, as multiple questions"=
>  XE "multiple ma: as multiple questions"   <cx "multiple mo, as multiple =
questions">  XE "multiple mo: as multiple questions"   When =93ma=94 or =93=
mo=94 is repeated, multiple questions are being asked simultaneously:
+<p>
+<ex "who knows what">  XE "who knows what: example=93  <pre><a name=3De9d6=
>9.6)	ma djuno ma
+	[What sumti] knows [what sumti]?
+	Who knows what?
+</pre><a name=3Ds10><h3>Relativized pro-sumti: =93ke'a=94</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ke'a	KOhA	relativized sumti
+</pre><lx "ke'a">  XE "ke'a"   <cx "pro-sumti, for relativized sumti in re=
lative clauses">  XE "pro-sumti: for relativized sumti in relative clauses"=
   <cx "ke'a, for relativized sumti in relative clauses">  XE "ke'a: for re=
lativized sumti in relative clauses"     XE "ke'a: for relativized sumti in=
 relative clauses"   <cx "relative clauses, use of ke'a for referral to rel=
ativized sumti in">  XE =93relative clause: use of ke'a for referral to rel=
ativized sumti in"   This pro-sumti is used in relative clauses (explained =
in <a href=3Dchap8.html>Chapter 8</a>) to indicate how the sumti being rela=
tivized fits within the clause. For example:
+<p>
+<ex "cat of plastic">  XE "cat of plastic: example=93  <pre><a name=3De10d=
1>10.1)	mi catlu lo mlatu poi [zo'e] zbasu
+ke'a lei slasi
+	I see a cat such-that something-unspecified makes
+		the-thing-being-relativized [the cat] from-some-mass-of plastic.
+	I see a cat made of plastic.
+
+</pre><cx "ke'a, ambiguity when omitted">  XE "ke'a: ambiguity when omitte=
d"   If =93ke'a=94 were omitted from <a href=3D#e10d1>Example 10.1</a>, it =
might be confused with:
+<p>
+<pre><a name=3De10d2>10.2)	mi catlu lo mlatu poi [ke'a]=20
+		zbasu lei slasi
+	I see a cat such-that the-thing-being-relativized
+		[the cat] makes a-mass-of plastic
+	I see a cat that makes plastic.
+</pre><cx "ke'a, contrasted with ri in relative clauses">  XE "ke'a: contr=
asted with ri in relative clauses"   <cx "ri, contrasted with ke'a in relat=
ive clauses">  XE "ri: contrasted with ke'a in relative clauses"   The anap=
hora cmavo =93ri=94 cannot be used in place of =93ke'a=94 in <a href=3D#e10=
d1>Example 10.1 </a>and <a href=3D#e10d2>Example 10.2</a>, because the rela=
tivized sumti is not yet complete when the =93ke'a=94 appears.
+<p>
+<cx "ke'a, and abstract descriptions">  XE "ke'a: and abstract description=
s"   <cx "ke'a, subscripting for nested relative clauses">  XE "ke'a: subsc=
ripting for nested relative clauses"   <cx "subscripts, on ke'a for nested =
relative clauses">  XE "subscripts: on ke'a for nested relative clauses"   =
Note that =93ke'a=94 is used only with relative clauses, and not with other=
 embedded bridi such as abstract descriptions. In the case of relative clau=
ses within relative clauses, =93ke'a=94 may be subscripted to make the diff=
erence clear (see <a href=3Dchap8.html>Chapter 8</a>).
+<p>
+<a name=3Ds11><h3>Abstraction focus pro-sumti: =93ce'u=94</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	ce'u	KOhA	abstraction focus
+</pre><lx "ce'u">  XE "ce'u"   The cmavo =93ce'u=94 is used within abstrac=
tion bridi, particularly property abstractions introduced by the cmavo =93k=
a=94. Abstractions, including the uses of =93ce'u=94, are discussed in full=
 in <a href=3Dchap11.html>Chapter 11</a>.
+<p>
+<cx "property abstraction, specifying sumti place of property with ce'u"> =
 XE "property abstraction(s): specifying sumti place of property with ce'u"=
   <cx "ce'u, use in specifying sumti place of property in abstraction">  X=
E "ce'u: use in specifying sumti place of property in abstraction"   In bri=
ef: Every property abstraction specifies a property of one of the sumti in =
it; that sumti place is filled by using =93ce'u=94. This convention enables=
 us to distinguish clearly between:
+<p>
+<ex "happiness">  XE "happiness: example=93  <pre><a name=3De11d1>11.1)	le=
 ka ce'u gleki
+	the property-of (X being-happy)
+	the property of being happy
+	happiness
+</pre>and
+<p>
+<pre><a name=3De11d2>11.2)	le ka gleki ce'u
+	the property-of (being-happy about-X)
+	the property of being that which someone is happy about
+</pre><a name=3Ds12><h3>Bound variable pro-sumti and pro-bridi: the da-ser=
ies and the bu'a-series</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	da	KOhA	da-series		something-1
+	de	KOhA	da-series		something-2
+	di	KOhA	da-series		something-3
+	bu'a	GOhA	bu'a-series	some-predicate-1
+	bu'e	GOhA	bu'a-series	some-predicate-2
+	bu'i	GOhA	bu'a-series	some-predicate-3
+</pre><cx "da-series pro-sumti, for bound variables">  XE "da-series pro-s=
umti: for bound variables"   <cx "bu'a-series pro-sumti, for bound variable=
s">  XE "bu'a-series pro-sumti: for bound variables"   Bound variables belo=
ng to the predicate-logic part of Lojban, and are listed here for completen=
ess only. Their semantics is explained in <a href=3Dchap16.html>Chapter 16<=
/a>. It is worth mentioning that the Lojban translation of <a href=3D#e1d2>=
Example 1.2 </a>is:
+<p>
+<ex "shook stick">  XE "shook stick: example=93  <lx "da">  XE "da"   <pre=
><a name=3De12d1>12.1)	la djan. cu lafti da poi grana ku'o gi'e desygau da
+	John raised something-1 which is-a-stick and shake-did something-1
+	John picked up a stick and shook it.
+</pre><a name=3Ds13><h3>Pro-sumti and pro-bridi cancelling</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	da'o	DAhO	cancel all pro-sumti/pro-bridi
+</pre><cx "pro-sumti assignment, stability of">  XE "pro-sumti assignment:=
 stability of"   <cx "pro-bridi assignment, stability of">  XE "pro-bridi a=
ssignment: stability of"   How long does a pro-sumti or pro-bridi remain st=
able? In other words, once we know the referent of a pro-sumti or pro-bridi=
, how long can we be sure that future uses of the same cmavo have the same =
referent? The answer to this question depends on which series the cmavo bel=
ongs to.
+<p>
+<cx "personal pro-sumti, implicit cancellation of by change of speaker/lis=
tener">  XE "personal pro-sumti: implicit cancellation of by change of spea=
ker/listener"   <cx "personal pro-sumti, stability of">  XE "personal pro-s=
umti: stability of"   <cx "assignable pro-sumti, stability of">  XE "assign=
able pro-sumti: stability of"   <cx "assignable pro-sumti, explicit cancell=
ation of by rebinding">  XE "assignable pro-sumti: explicit cancellation of=
 by rebinding"   <lx "goi">  XE "goi"   <lx "cei">  XE "cei"   <cx "bound v=
ariable pro-sumti, stability of">  XE "bound variable pro-sumti: stability =
of"   Personal pro-sumti are stable until there is a change of speaker or l=
istener, possibly signaled by a vocative. Assignable pro-sumti and pro-brid=
i last indefinitely or until rebound with =93goi=94 or =93cei=94. Bound var=
iable pro-sumti and pro-bridi also generally last until re-bound; details a=
re available in <a href=3Dchap16.html>Chapter 16</a>.
+<p>
+<cx "utterance pro-sumti, stability of">  XE "utterance pro-sumti: stabili=
ty of"   <cx "reflexive pro-sumti, stability of">  XE "reflexive pro-sumti:=
 stability of"   <cx "ke'a, stability of">  XE "ke'a: stability of"   <cx "=
anaphoric pro-sumti, stability of">  XE "anaphoric pro-sumti: stability of"=
   <cx "anaphoric pro-bridi, stability of">  XE "anaphoric pro-bridi: stabi=
lity of"   Utterance pro-sumti are stable only within the utterance in whic=
h they appear; similarly, reflexive pro-sumti are stable only within the br=
idi in which they appear; and =93ke'a=94 is stable only within its relative=
 clause. Anaphoric pro-sumti and pro-bridi are stable only within narrow li=
mits depending on the rules for the particular cmavo.
+<p>
+<cx "demonstrative pro-sumti, stability of">  XE "demonstrative pro-sumti:=
 stability of"   <cx "indefinite pro-sumti, stability of">  XE "indefinite =
pro-sumti: stability of"   <cx "indefinite pro-bridi, stability of">  XE "i=
ndefinite pro-bridi: stability of"   Demonstrative pro-sumti, indefinite pr=
o-sumti and pro-bridi, and sumti and bridi questions potentially change ref=
erents every time they are used.
+<p>
+<cx "pro-sumti assignment, explicit cancellation of with da'o">  XE "pro-s=
umti assignment: explicit cancellation of with da'o"   <cx "pro-bridi assig=
nment, explicit cancellation of with da'o">  XE "pro-bridi assignment: expl=
icit cancellation of with da'o"   <lx "da'o">  XE "da'o"   <lx "DAhO">  XE =
"DAhO selma=92o"   <cx "cancellation of pro-sumti/pro-bridi assignment, wit=
h da'o">  XE "cancellation of pro-sumti/pro-bridi assignment: with da'o"   =
<cx "da'o, for cancellation of pro-sumti/pro-bridi assignment">  XE "da'o: =
for cancellation of pro-sumti/pro-bridi assignment"   <cx "da'o, syntax of"=
>  XE "da'o: syntax of"   However, there are ways to cancel all pro-sumti a=
nd pro-bridi, so that none of them have known referents. (Some, such as =93=
mi=94, will acquire the same referent as soon as they are used again after =
the cancellation.) The simplest way to cancel everything is with the cmavo =
=93da'o=94 of selma'o DAhO, which is used solely for this purpose; it may a=
ppear anywhere, and has no effect on the grammar of texts containing it. On=
e use of =93da'o=94 is when entering a conversation, to indicate that one's=
 pro-sumti assignments have nothing to do with any assignments already made=
 by other participants in the conversation.
+<p>
+<cx "ni'o, effect on pro-sumti/pro-bridi assignments">  XE "ni'o: effect o=
n pro-sumti/pro-bridi assignments"   <cx "no'i, effect on pro-sumti/pro-bri=
di assignments">  XE "no'i: effect on pro-sumti/pro-bridi assignments"   <c=
x "pro-sumti assignments, no'i effect on">  XE "pro-sumti assignment: no'i =
effect on"   <cx "pro-bridi assignments, no'i effect on">  XE "pro-bridi as=
signment: no'i effect on"   In addition, the cmavo =93ni'o=94 and =93no'i=
=94 of selma'o NIhO, which are used primarily to indicate shifts in topic, =
may also have the effect of canceling pro-sumti and pro-bridi assignments, =
or of reinstating ones formerly in effect. More explanations of NIhO can be=
 found in <a href=3Dchap19.html>Chapter 19</a>.
+<p>
+<a name=3Ds14><h3>The identity predicate: du</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	du	GOhA	identity
+</pre><cx "identity predicate">  XE "identity predicate"   <lx "du">  XE "=
du"   The cmavo =93du=94 has the place structure:
+<p>
+<dl compact><dt>du:  <dd>x1 is identical with x2, x3,=A0<dots>=85</dots>
+</dl><cx "du, rationale for selection of selma'o for">  XE "du: rationale =
for selection of selma'o for"   and appears in selma'o GOhA for reasons of =
convenience: it is not a pro-bridi. =93du=94 serves as mathematical =93=3D=
=94, and outside mathematical contexts is used for defining or identifying.=
 Mathematical examples may be found in <a href=3Dchap18.html>Chapter 18</a>=
.
+
+<p>
+<cx "du, contrasted with mintu">  XE "du: contrasted with mintu"   <cx "mi=
ntu, contrasted with du">  XE "mintu: contrasted with du"   The main differ=
ence between
+<p>
+<pre><a name=3De14d1>14.1)	ko'a du le nanmu
+	It-1 is-identical-to the man
+</pre>and
+<p>
+<pre><a name=3De14d2>14.2)	ko'a mintu le nanmu
+	It-1 is-the-same-as the man
+</pre><cx "du, meaning of">  XE "du: meaning of"   is this defining nature=
. <a href=3D#e14d1>Example 14.1 </a>presumes that the speaker is responding=
 to a request for information about what =93ko'a=94 refers to, or that the =
speaker in some way feels the need to define =93ko'a=94 for later reference=
. A bridi with =93du=94 is an identity sentence, somewhat metalinguisticall=
y saying that all attached sumti are representations for the same referent.=
 There may be any number of sumti associated with =93du=94, and all are sai=
d to be identical.
+<p>
+<a href=3D#e14d2>Example 14.2</a>, however, predicates; it is used to make=
 a claim about the identity of =93ko'a=94, which presumably has been define=
d previously.
+<p>
+<cx "du, derivation of">  XE "du: derivation of"   <cx "du, contrasted wit=
h dunli">  XE "du: contrasted with dunli"   <cx "dunli, contrasted with du"=
>  XE "dunli: contrasted with du"   Note: =93du=94 historically is derived =
from =93dunli=94, but =93dunli=94 has a third place which =93du=94 lacks: t=
he standard of equality.
+<p>
+<a name=3Ds15><h3>lujvo based on pro-sumti</h3>
+<p>
+<cx "rafsi, based on pro-sumti">  XE "rafsi: based on pro-sumti"   <cx "pr=
o-sumti, rafsi for">  XE "pro-sumti: rafsi for"   There exist rafsi allocat=
ed to a few cmavo of selma'o KOhA, but they are rarely used. (See <a href=
=3D#s16>Section 16 </a>for a complete list.) The obvious way to use them is=
 as internal sumti, filling in an appropriate place of the gismu or lujvo t=
o which they are attached; as such, they usually stand as the first rafsi i=
n their lujvo.
+<p>
+<ex "you-talk">  XE "you-talk: example=93  <cx "lujvo, pro-sumti rafsi eff=
ect on place structure of">  XE "lujvo: pro-sumti rafsi effect on place str=
ucture of"   <cx "pro-sumti rafsi, effect of on place structure of lujvo"> =
 XE "pro-sumti rafsi: effect of on place structure of lujvo"   Thus =93dont=
a'a=94, meaning =93you-talk=94, would be interpreted as =93tavla be do=94, =
and would have the place structure
+<p>
+<pre><a name=3De15d1>15.1)	t1 talks to you about subject t3 in language t4
+</pre>since t2 (the addressee) is already known to be =93do=94.
+<p>
+<ex "you-cmavo">  XE "you-cmavo: example=93  On the other hand, the lujvo =
=93donma'o=94, literally =93you-cmavo=94, which means =93a second person pe=
rsonal pronoun=94, would be interpreted as =93cmavo be zo do=94, and have t=
he place structure:
+<p>
+<pre><a name=3De15d2>15.2)	c1 is a second person pronoun in language c4
+</pre>since both the c2 place (the grammatical class) and the c3 place (th=
e meaning) are obvious from the context =93do=94.
+<p>
+<cx "pro-sumti rafsi, anticipated use of for abbreviating inconvenient for=
ms">  XE "pro-sumti rafsi: anticipated use of for abbreviating inconvenient=
 forms"   <lx "fo'a">  XE "fo'a"   An anticipated use of rafsi for cmavo in=
 the =93fo'a=94 series is to express lujvo which can't be expressed in a co=
nvenient rafsi form, because they are too long to express, or are formally =
inconvenient (fu'ivla, cmene, and so forth.) An example would be:
+<p>
+<pre><a name=3De15d3>15.3)	fo'a goi le kulnrsu,omi  .i lo fo'arselsanga
+	x6 stands for Finnish-culture.  An x6-song.
+</pre><ex "beverage">  XE "beverage: example=93  <lx "zi'o">  XE "zi'o"   =
<cx "lujvo, zi'o rafsi effect on place structure of">  XE "lujvo: zi'o rafs=
i effect on place structure of"   <cx "zi'o rafsi, effect of on place struc=
ture of lujvo">  XE "zi'o rafsi: effect of on place structure of lujvo"   F=
inally, lujvo involving =93zi'o=94 are also possible, and are fully discuss=
ed in <a href=3Dchap12.html>Chapter 12</a>. In brief, the convention is to =
use the rafsi for =93zi'o=94 as a prefix immediately followed by the rafsi =
for the number of the place to be deleted. Thus, if we consider a beverage =
(something drunk without considering who, if anyone, drinks it) as a =93se =
pinxe be zi'o=94, the lujvo corresponding to this is =93zilrelselpinxe=94 (=
deleting the second place of =93se pinxe=94). Deleting the x1 place in this=
 fashion would move all remaining places up by one. This would mean that =
=93zilpavypinxe=94 has the same place structure as =93zilrelselpinxe=94, an=
d =93lo zilpavypinxe=94, like =93lo zilrelselpinxe=94, refers to a beverage=
, and not to a non-existent drinker.
+<p>
+<cx "pro-bridi rafsi, as producing context-dependent meanings">  XE "pro-b=
ridi rafsi: as producing context-dependent meanings"   <lx "co'e">  XE "co'=
e"   <lx "du">  XE "du"   <lx "bu'a">  XE "bu'a"   The pro-bridi =93co'e=94=
, =93du=94, and =93bu'a=94 also have rafsi, which can be used just as if th=
ey were gismu. The resulting lujvo have (except for =93du=94-based lujvo) h=
ighly context-dependent meanings.
+<p>
+<a name=3Ds16><h3>KOhA cmavo by series</h3>
+<p>
+<cx "pro-sumti, list by series">  XE "pro-sumti: list by series"   <cx "pr=
o-sumti, mi-series list">  XE "pro-sumti: mi-series list"   <dl compact><dt=
>		<dd>mi-series:
+	mi		I (rafsi:  =93mib=94)
+	do		you (rafsi:  =93don=94 and =93doi=94)
+	mi'o		you and I
+	mi'a		I and others, we but not you
+	ma'a		you and I and others
+	do'o		you and others
+	ko		you-imperative
+</dl><cx "pro-sumti, ti-series list">  XE "pro-sumti: ti-series list"   <d=
l compact><dt>		<dd>ti-series:
+	ti		this here; something nearby (rafsi:  =93tif=94)
+	ta		that there; something distant (rafsi:  =93taz=94)
+	tu		that yonder; something far distant (rafsi:  =93tuf=94)
+</dl><cx "pro-sumti, di'u-series list">  XE "pro-sumti: di'u-series list" =
  <dl compact><dt>		<dd>di'u-series:
+	di'u		the previous utterance
+	de'u		an earlier utterance
+	da'u		a much earlier utterance
+	di'e		the next utterance
+	de'e		a later utterance
+	da'e		a much later utterance
+	dei		this very utterance
+	do'i		some utterance
+	</dl><cx "pro-sumti, ko'a-series list">  XE "pro-sumti: ko'a-series list"=
   <dl compact><dt>	<dd>ko'a-series:
+	ko'a		it-1; 1st assignable pro-sumti
+	ko'e		it-2; 2nd assignable pro-sumti
+	ko'i		it-3; 3rd assignable pro-sumti
+	ko'o		it-4; 4th assignable pro-sumti
+	ko'u		it-5; 5th assignable pro-sumti
+	fo'a		it-6; 6th assignable pro-sumti (rafsi:  =93fo'a=94)
+	fo'e		it-7; 7th assignable pro-sumti (rafsi:  =93fo'e=94)
+	fo'i		it-8; 8th assignable pro-sumti (rafsi:  =93fo'i=94)
+	fo'o		it-9; 9th assignable pro-sumti
+	fo'u		it-10; 10th assignable pro-sumti
+</dl><cx "pro-sumti, ri-series list">  XE "pro-sumti: ri-series list"   <d=
l compact><dt>		<dd>ri-series:
+	ri		(repeats the last sumti)
+	ra		(repeats a previous sumti)
+	ru		(repeats a long-ago sumti)
+</dl><cx "pro-sumti, zo'e-series list">  XE "pro-sumti: zo'e-series list" =
  <dl compact><dt>	=09
+<dd>zo'e-series:
+	zo'e		the obvious value
+	zu'i		the typical value
+	zi'o		the nonexistent value (rafsi:  =93zil=94)
+</dl><cx "pro-sumti, vo'a-series list">  XE "pro-sumti: vo'a-series list" =
  <dl compact><dt>		<dd>vo'a-series:
+	vo'a		x1 of this bridi
+	vo'e		x2 of this bridi
+	vo'i		x3 of this bridi
+	vo'o		x4 of this bridi
+	vo'u		x5 of this bridi
+</dl><cx "pro-sumti, da-series list">  XE "pro-sumti: da-series list"   <d=
l compact><dt>		<dd>da-series:
+	da		something-1 (rafsi:  =93dav=94/=93dza=94)
+	de		something-2
+	di		something-3
+</dl><cx "pro-sumti, miscellaneous list">  XE "pro-sumti: miscellaneous li=
st"   <dl compact><dt>		<dd>others:
+	ke'a		relativized sumti
+	ma		sumti question
+	ce'u		abstraction focus
+</dl><a name=3Ds17><h3>GOhA and other pro-bridi by series</h3>
+<p>
+<cx "pro-bridi, list by series">  XE "pro-bridi: list by series"   <cx "pr=
o-bridi, broda-series list">  XE "pro-bridi: broda-series list"   <dl compa=
ct><dt>		<dd>broda-series (not GOhA):
+	broda	is-1; 1st assignable pro-bridi
+	brode	is-2; 2nd assignable pro-bridi
+	brodi		is-3; 3rd assignable pro-bridi
+	brodo	is-4; 4th assignable pro-bridi
+	brodu	is-5; 5th assignable pro-bridi
+</dl><cx "pro-bridi, go'i-series list">  XE "pro-bridi: go'i-series list" =
  <dl compact><dt>		<dd>go'i-series:
+	go'i		(repeats the last bridi)
+	go'a		(repeats a previous bridi)
+	go'u		(repeats a long-ago bridi)
+	go'e		(repeats the last-but-one bridi)
+	go'o		(repeats a future bridi)
+	nei		(repeats the current bridi)
+	no'a		(repeats the next outer bridi)
+</dl><cx "pro-bridi, bu'a-series list">  XE "pro-bridi: bu'a-series list" =
  <dl compact><dt>		<dd>bu'a-series:
+	bu'a		some-predicate-1 (rafsi:  =93bul=94)
+	bu'e		some-predicate-2
+	bu'i		some-predicate-3
+</dl><cx "pro-bridi, miscellaneous list">  XE "pro-bridi: miscellaneous li=
st"   <dl compact><dt>		<dd>others:
+	co'e		has the obvious relationship  (rafsi:  =93com=94/=93co'e=94)
+	mo		bridi question
+	du		identity:  x1 is identical to x2, x3=A0<dots>=85</dots> (rafsi:  =93d=
ub=94/=93du'o=94)
+</dl><a name=3Ds18><h3>Other cmavo discussed in this chapter</h3>
+<p>
+<cx "pro-bridi, list of miscellaneous cmavo used with">  XE "pro-bridi: li=
st of miscellaneous cmavo used with"   <cx "pro-sumti, list of miscellaneou=
s cmavo used with">  XE "pro-sumti: list of miscellaneous cmavo used with" =
  <pre> 	goi	GOI	pro-sumti assignment (ko'a-series)
+	cei	CEI	pro-bridi assignment (broda-series)
+	ra'o	RAhO	pro-sumti/pro-bridi update
+	soi	SOI	reciprocity
+	se'u	SEhU	soi terminator
+	da'o	DAhO	cancel all pro-sumti/pro-bridi
+
+</body></html>
+<img src=3Dchap8.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D4=
05>
+
+<h2>Chapter 8
+<br>
+Relative Clauses, Which Make sumti Even More Complicated</h2>
+<h6>$Revision: 4.0 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>What are you pointing at?</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	poi	NOI	restrictive relative clause introducer
+	ke'a	GOhA	relative pro-sumti
+	ku'o	KUhO	relative clause terminator
+</pre><lx "ti">  XE "ti"   <lx "ta">  XE "ta"   <lx "tu">  XE "tu"   Let u=
s think about the problem of communicating what it is that we are pointing =
at when we are pointing at something. In Lojban, we can refer to what we ar=
e pointing at by using the pro-sumti =93ti=94 if it is nearby, or =93ta=94 =
if it is somewhat further away, or =93tu=94 if it is distant. (Pro-sumti ar=
e explained in full in <a href=3Dchap7.html>Chapter 7</a>.)
+<p>
+<cx "reference, ambiguity of ti/ta/tu">  XE "reference: ambiguity of ti/ta=
/tu"   However, even with the assistance of a pointing finger, or pointing =
lips, or whatever may be appropriate in the local culture, it is often hard=
 for a listener to tell just what is being pointed at. Suppose one is point=
ing at a person (in particular, in the direction of his or her face), and s=
ays:
+<p>
+<pre><a name=3De1d1>1.1)	ti cu barda
+	This-one is-big.
+</pre><p>What is the referent of =93ti=94? Is it the person? Or perhaps it=
 is the person's nose? Or even (for =93ti=94 can be plural as well as singu=
lar, and mean =93these ones=94 as well as =93this one=94) the pores on the =
person's nose?
+<p>
+<cx "relative clause, use for reference">  XE "relative clause: use for re=
ference"   <cx "reference, use of relative clause for">  XE "reference: use=
 of relative clause for"   <lx "NOI">  XE "NOI selma=92o"   <lx "ku'o">  XE=
 "ku'o"   <lx "KUhO">  XE "KUhO selma=92o"   <lx "poi">  XE "poi"   To help=
 solve this problem, Lojban uses a construction called a =93relative clause=
=94. Relative clauses are usually attached to the end of sumti, but there a=
re other places where they can go as well, as explained later in this chapt=
er. A relative clause begins with a word of selma'o NOI, and ends with the =
elidable terminator =93ku'o=94 (of selma'o KUhO). As you might suppose, =93=
noi=94 is a cmavo of selma'o NOI; however, first we will discuss the cmavo =
=93poi=94, which also belongs to selma'o NOI.
+<p>
+<cx "poi, syntax of">  XE "poi: syntax of"   <lx "ke'a">  XE "ke'a"   <cx =
"relativized sumti, definition">  XE "relativized sumti: definition"   <cx =
"reference, to relativized sumti with ke'a">  XE "reference: to relativized=
 sumti with ke'a"   <cx "ke'a, as referent for relativized sumti">  XE "ke'=
a: as referent for relativized sumti"   In between the =93poi=94 and the =
=93ku'o=94 appears a full bridi, with the same syntax as any other bridi. A=
nywhere within the bridi of a relative clause, the pro-sumti =93ke'a=94 (of=
 selma'o KOhA) may be used, and it stands for the sumti to which the relati=
ve clause is attached (called the =93relativized sumti=94). Here are some e=
xamples before we go any further:
+<p>
+<ex "big person">  XE "big person: example=93  <pre><a name=3De1d2>1.2)	ti=
 poi ke'a prenu ku'o cu barda
+	This-thing such-that-(IT is-a-person) is-large.
+	This thing which is a person is big.
+	This person is big.
+</pre><ex "big nose">  XE "big nose: example=93  <pre><a name=3De1d3>1.3)	=
ti poi ke'a nazbi ku'o cu barda
+	This-thing such-that-(IT is-a-nose) is-large.
+	This thing which is a nose is big.
+	This nose is big.
+
+</pre><ex "big nose-pores">  XE "big nose-pores: example=93  <pre><a name=
=3De1d4>1.4)	ti poi ke'a nazbi kapkevna ku'o cu barda
+	This-thing such-that-(IT is-a-nose-type-of skin-hole) is-big.
+	These things which are nose-pores are big.
+	These nose-pores are big.
+</pre><cx "IT, as notation convention in relative clause chapter">  XE "IT=
: as notation convention in relative clause chapter"   In the literal trans=
lations throughout this chapter, the word =93IT=94, capitalized, is used to=
 represent the cmavo =93ke'a=94. In each case, it serves to represent the s=
umti (in <a href=3D#e1d{2}>Examples 1.2 </a>through <a href=3D#e1d4>1.4</a>=
, the cmavo =93ti=94) to which the relative clause is attached.
+<p>
+<cx "ke'a, non-initial place use in relative clause">  XE "ke'a: non-initi=
al place use in relative clause"   Of course, there is no reason why =93ke'=
a=94 needs to appear in the x1 place of a relative clause bridi; it can app=
ear in any place, or indeed even in a sub-bridi within the relative clause =
bridi. Here are two more examples:
+<p>
+<pre><a name=3De1d5>1.5)	tu poi le mlatu pu lacpu ke'a ku'o cu ratcu
+	That-distant-thing such-that (the cat [past] drags IT) is-a-rat.
+	That thing which the cat dragged is a rat.
+	What the cat dragged is a rat.
+<a name=3De1d6>1.6)	ta poi mi djica le nu mi ponse ke'a [kei] ku'o cu blot=
i
+	That-thing such-that (I desire the event-of (I own IT) ) is-a-boat.
+	That thing that I want to own is a boat.
+</pre><p>In <a href=3D#e1d6>Example 1.6</a>, =93ke'a=94 appears in an abst=
raction clause (abstractions are explained in <a href=3Dchap11.html>Chapter=
 11</a>) within a relative clause.
+<p>
+<cx "ke'a, effect of omission of">  XE "ke'a: effect of omission of"   <cx=
 "relative clause, effect of omission of =93ke'a=94 on">  XE "relative clau=
se: effect of omission of ke'a on"   Like any sumti, =93ke'a=94 can be omit=
ted. The usual presumption in that case is that it then falls into the x1 p=
lace:
+<p>
+<pre><a name=3De1d7>1.7)	ti poi nazbi cu barda
+	This-thing which is-a-nose is-big.
+</pre>almost certainly means the same thing as <a href=3D#e1d3>Example 1.3=
</a>. However, =93ke'a=94 can be omitted if it is clear to the listener tha=
t it belongs in some place other than x1:
+<p>
+<pre><a name=3De1d8>1.8)	tu poi le mlatu pu lacpu cu ratcu
+	That-distant-thing which the cat [past] drags is-a-rat
+</pre>is equivalent to <a href=3D#e1d4>Example 1.4</a>.
+<p>
+<cx "ku'o, elidability for relative clauses">  XE "ku'o: elidability for r=
elative clauses"   As stated before, =93ku'o=94 is an elidable terminator, =
and in fact it is almost always elidable. Throughout the rest of this chapt=
er, =93ku'o=94 will not be written in any of the examples unless it is abso=
lutely required: thus, <a href=3D#e1d2>Example 1.2 </a>can be written:
+<p>
+<pre><a name=3De1d9>1.9)	ti poi prenu cu barda
+	That which is-a-person is-big.
+	That person is big.
+</pre><cx "poi, discussion of translation">  XE "poi: discussion of transl=
ation"   without any change in meaning. Note that =93poi=94 is translated =
=93which=94 rather than =93such-that=94 when =93ke'a=94 has been omitted fr=
om the x1 place of the relative clause bridi. The word =93which=94 is used =
in English to introduce English relative clauses: other words that can be u=
sed are =93who=94 and =93that=94, as in:
+<p>
+<pre><a name=3De1d10>1.10)	I saw a man who was going to the store.
+</pre>and
+<p>
+<pre><a name=3De1d11>1.11)	The building that the school was located in is =
large.
+</pre><p>In <a href=3D#e1d10>Example 1.10 </a>the relative clause is =93wh=
o was going to the store=94, and in <a href=3D#e1d11>Example 1.11 </a>it is=
 =93that the school was located in=94. Sometimes =93who=94, =93which=94, an=
d =93that=94 are used in literal translations in this chapter in order to m=
ake them read more smoothly.
+<p>
+<a name=3Ds2><h3>Incidental relative clauses</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	noi	NOI	incidental relative clause introducer
+</pre><cx "relative clauses, kinds of">  XE =93relative clause: kinds of" =
  <cx "relative clause, restrictive (see also restrictive relative clause)"=
>  XE "relative clause: restrictive (see also restrictive relative clause)"=
   <cx "restrictive relative clause, definition">  XE "restrictive relative=
 clause: definition"   <cx "incidental relative clause, definition">  XE "i=
ncidental relative clause: definition"   <cx "non-restrictive relative clau=
se, definition (see also incidental relative clause)">  XE "non-restrictive=
 relative clause: definition (see also incidental relative clause)"   <cx "=
relative clauses, restricted contrasted with incidental">  XE =93relative c=
lause: restricted contrasted with incidental"   <lx "noi">  XE "noi"   Ther=
e are two basic kinds of relative clauses: restrictive relative clauses int=
roduced by =93poi=94, and incidental (sometimes called simply =93non-restri=
ctive=94) relative clauses introduced by =93noi=94. The difference between =
restrictive and incidental relative clauses is that restrictive clauses pro=
vide information that is essential to identifying the referent of the sumti=
 to which they are attached, whereas incidental relative clauses provide ad=
ditional information which is helpful to the listener but is not essential =
for identifying the referent of the sumti. All of the examples in <a href=
=3D#s1>Section 1 </a>are restrictive relative clauses: the information in t=
he relative clause is essential to identification. (The title of this chapt=
er, though, uses an incidental relative clause.)
+<p>
+Consider the following examples:
+<p>
+<pre><a name=3De2d1>2.1)	le gerku poi blanu cu barda
+	The dog which is-blue is-large.
+	The dog which is blue is large.
+<a name=3De2d2>2.2)	le gerku noi blanu cu barda
+	The dog incidentally-which is-blue is-large.
+	The dog, which is blue, is large.
+</pre><p>In <a href=3D#e2d1>Example 2.1</a>, the information conveyed by =
=93poi blanu=94 is essential to identifying the dog in question: it restric=
ts the possible referents from dogs in general to dogs that are blue. This =
is why =93poi=94 relative clauses are called restrictive. In <a href=3D#e2d=
2>Example 2.2</a>, on the other hand, the dog which is referred to has pres=
umably already been identified clearly, and the relative clause =93noi blan=
u=94 just provides additional information about it. (If in fact the dog has=
n't been identified clearly, then the relative clause does not help identif=
y it further.)
+<p>
+<cx "relative clauses, restricted contrasted with incidental in English ex=
pression">  XE =93relative clause: restricted contrasted with incidental in=
 English expression"   <cx "relative clauses, effect of commas in English">=
  XE =93relative clause: effect of commas in English"   <cx "comma, effect =
on relative clause in English">  XE "comma: effect on relative clause in En=
glish"   <cx "incidental relative clause, as a parenthetical device">  XE "=
incidental relative clause: as a parenthetical device"   In English, the di=
stinction between restrictive and incidental relative clauses is expressed =
in writing by surrounding incidental, but not restrictive, clauses with com=
mas. These commas are functioning as parentheses, because incidental relati=
ve clauses are essentially parenthetical. This distinction in punctuation i=
s represented in speech by a difference in tone of voice. In addition, Engl=
ish restrictive relative clauses can be introduced by =93that=94 as well as=
 =93which=94 and =93who=94, whereas incidental relative clauses cannot begi=
n with =93that=94. Lojban, however, always uses the cmavo =93poi=94 and =93=
noi=94 rather than punctuation or intonation to make the distinction.
+<p>
+Here are more examples of incidental relative clauses:
+<p>
+<pre><a name=3De2d3>2.3)	mi noi jdice cu zvati
+	I who-incidentally am-a-judge am-at [some-place].
+	I, a judge, am present.
+</pre><p>In this example, =93mi=94 is already sufficiently restricted, and=
 the additional information that I am a judge is being provided solely for =
the listener's edification.
+
+<p>
+<pre><a name=3De2d4>2.4)	xu do viska le mi karce noi blabi
+	[True?] You see my car incidentally-which is-white.
+	Do you see my car, which is white?
+</pre><p>In <a href=3D#e2d4>Example 2.4</a>, the speaker is presumed to ha=
ve only one car, and is providing incidental information that it is white. =
(Alternatively, he or she might have more than one car, since =93le karce=
=94 can be plural, in which case the incidental information is that each of=
 them is white.) Contrast <a href=3D#e2d5>Example 2.5 </a>with a restrictiv=
e relative clause:
+<p>
+<pre><a name=3De2d5>2.5)	xu do viska le mi karce poi blabi
+	[True?] You see my car which is-white.
+	Do you see my car that is white?
+	Do you see my white car?
+</pre><cx "relative clause, compared with tanru">  XE "relative clause: co=
mpared with tanru"   Here the speaker probably has several cars, and is res=
tricting the referent of the sumti =93le mi karce=94 (and thereby the liste=
ner's attention) to the white one only. <a href=3D#e2d5>Example 2.5 </a>mea=
ns much the same as <a href=3D#e2d6>Example 2.6</a>, which does not use a r=
elative clause:
+<p>
+<pre><a name=3De2d6>2.6)	xu do viska le mi blabi karce
+	[True?] You see my white car.
+	Do you see my car, the white one?
+</pre><cx "relative clause, contrasted with tanru">  XE "relative clause: =
contrasted with tanru"   So a restrictive relative clause attached to a des=
cription can often mean the same as a description involving a tanru. Howeve=
r, =93blabi karce=94, like all tanru, is somewhat vague: in principle, it m=
ight refer to a car which carries white things, or even express some more c=
omplicated concept involving whiteness and car-ness; the restrictive relati=
ve clause of <a href=3D#e2d5>Example 2.5 </a>can only refer to a car which =
is white, not to any more complex or extended concept.
+<p>
+<a name=3Ds3><h3>Relative phrases</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	pe	GOI	restrictive association
+	po	GOI	restrictive possession
+	po'e	GOI	restrictive intrinsic possession
+	po'u	GOI	restrictive identification
+	ne	GOI	incidental association
+	no'u	GOI	incidental identification
+	ge'u	GEhU	relative phrase terminator
+</pre><cx "relative phrase, rationale for">  XE "relative phrase: rational=
e for"   <lx "GOI">  XE "GOI selma=92o"   <cx "relative phrase, syntax of">=
  XE "relative phrase: syntax of"   <cx "relative phrase, as an abbreviatio=
n of a common relative clause">  XE "relative phrase: as an abbreviation of=
 a common relative clause"   There are types of relative clauses (those whi=
ch have a certain selbri) which are frequently wanted in Lojban, and can be=
 expressed using a shortcut called a relative phrase. Relative phrases are =
introduced by cmavo of selma'o GOI, and consist of a GOI cmavo followed by =
a single sumti.
+<p>
+<lx "pe">  XE "pe"   <cx "pe, compared with =93poi ke'a srana=94">  XE "pe=
: compared with poi ke'a srana"   <cx "pe, as loose association">  XE "pe: =
as loose association"   <cx "loose association, expressing with pe">  XE "l=
oose association: expressing with pe"   Here is an example of =93pe=94, plu=
s an equivalent sentence using a relative clause:
+<p>
+<pre><a name=3De3d1>3.1)	le stizu pe mi cu blanu
+	The chair associated-with me is-blue.
+	My chair is blue.
+
+<a name=3De3d2>3.2)	le stizu poi ke'a srana mi cu blanu
+	The chair such-that (IT is-associated-with me) is-blue.
+</pre><p>In <a href=3D#e3d1>Example 3.1 </a>and <a href=3D#e3d2>Example 3.=
2</a>, the link between the chair and the speaker is of the loosest kind.
+<p>
+<lx "po">  XE "po"   <cx "po, compared with =93poi ke'a se steci srana=94"=
>  XE "po: compared with poi ke'a se steci srana"   <cx "po, as restrictive=
 possession">  XE "po: as restrictive possession"   <cx "possession, expres=
sing with po">  XE "possession: expressing with po"   <cx "specificity, exp=
ressing with po">  XE "specificity: expressing with po"   Here is an exampl=
e of =93po=94:
+<p>
+<pre><a name=3De3d3>3.3)	le stizu po mi cu xunre
+	The chair specific-to me is red.
+<a name=3De3d4>3.4)	le stizu poi ke'a se steci srana mi cu xunre
+	The chair such-that (IT is-specifically associated-with me) is-red.
+</pre><cx "pe, contrasted with po">  XE "pe: contrasted with po"   <cx "po=
, contrasted with pe">  XE "po: contrasted with pe"   <a href=3D#e3d3>Examp=
le 3.3 </a>and <a href=3D#e3d4>Example 3.4 </a>contrast with <a href=3D#e3d=
1>Example 3.1 </a>and <a href=3D#e3d2>Example 3.2</a>: the chair is more pe=
rmanently connected with the speaker. A plausible (though not the only poss=
ible) contrast between <a href=3D#e3d1>Example 3.1 </a>and <a href=3D#e3d3>=
Example 3.3 </a>is that =93pe mi=94 would be appropriate for a chair the sp=
eaker is currently sitting on (whether or not the speaker owned that chair)=
, and =93po mi=94 for a chair owned by the speaker (whether or not he or sh=
e was currently occupying it).
+<p>
+<cx "po, contrasted with English =93possession=94">  XE "po: contrasted wi=
th English possession"   As a result, the relationship expressed between tw=
o sumti by =93po=94 is usually called =93possession=94, although it does no=
t necessarily imply ownership, legal or otherwise. The central concept is t=
hat of specificity (=93steci=94 in Lojban).
+<p>
+<lx "po'e">  XE "po'e"   <cx "po'e, compared with =93poi ke'a jinzi ke se =
steci srana=94">  XE "po'e: compared with poi ke'a jinzi ke se steci srana"=
   <cx "po'e, as intrinsic possession">  XE "po'e: as intrinsic possession"=
   <cx "possession, intrinsic, expressing with po'e">  XE "possession: intr=
insic, expressing with po'e"   <cx "intrinsic possession, expressing with p=
o'e">  XE "intrinsic possession: expressing with po'e"   <cx "inalienable p=
ossession, expressing with po'e">  XE "inalienable possession: expressing w=
ith po'e"   Here is an example of =93po'e=94, as well as another example of=
 =93po=94:
+<p>
+<ex "person's arm">  XE "person's arm: example=93  <pre><a name=3De3d5>3.5=
)	le birka po'e mi cu spofu
+	The arm intrinsically-possessed-by me is-broken
+<a name=3De3d6>3.6)	le birka poi jinzi ke se steci srana mi cu spofu
+	The arm which is-intrinsically (specifically associated-with) me is-broke=
n.
+<a name=3De3d7>3.7)	le botpi po mi cu spofu
+	The bottle specific-to me is-broken
+</pre><cx "po'e, contrasted with po">  XE "po'e: contrasted with po"   <cx=
 "po, contrasted with po'e">  XE "po: contrasted with po'e"   <cx "intrinsi=
c possession, definition">  XE "intrinsic possession: definition"   <cx "in=
alienable possession, definition">  XE "inalienable possession: definition"=
   <cx "extrinsic possession, definition">  XE "extrinsic possession: defin=
ition"   <cx "alienable possession, definition">  XE "alienable possession:=
 definition"   <a href=3D#e3d5>Example 3.5 </a>and <a href=3D#e3d6>Example =
3.6 </a>on the one hand, and <a href=3D#e3d7>Example 3.7 </a>on the other, =
illustrate the contrast between two types of possession called =93intrinsic=
=94 and =93extrinsic=94, or sometimes =93inalienable=94 and =93alienable=94=
, respectively. Something is intrinsically (or inalienably) possessed by so=
meone if the possession is part of the possessor, and cannot be changed wit=
hout changing the possessor. In the case of <a href=3D#e3d5>Example 3.5</a>=
, people are usually taken to intrinsically possess their arms: even if an =
arm is cut off, it remains the arm of that person. (If the arm is transplan=
ted to another person, however, it becomes intrinsically possessed by the n=
ew user, though, so intrinsic possession is a matter of degree.)
+<p>
+<cx "inalienable, distinguishing from alienable">  XE "inalienable: distin=
guishing from alienable"   By contrast, the bottle of <a href=3D#e3d7>Examp=
le 3.7 </a>can be given away, or thrown away, or lost, or stolen, so it is =
possessed extrinsically (alienably). The exact line between intrinsic and e=
xtrinsic possession is culturally dependent. The U.S. Declaration of Indepe=
ndence speaks of the =93inalienable rights=94 of men, but just what those r=
ights are, and even whether the concept makes sense at all, varies from cul=
ture to culture.
+<p>
+Note that <a href=3D#e3d5>Example 3.5 </a>can also be expressed without a =
relative clause:
+<p>
+<pre><a name=3De3d8>3.8)	le birka be mi cu spofu
+	The arm of-body me is broken
+</pre><cx "intrinsic possession, expressing by using place in some selbri"=
>  XE "intrinsic possession: expressing by using place in some selbri"   re=
flecting the fact that the gismu =93birka=94 has an x2 place representing t=
he body to which the arm belongs. Many, but not all, cases of intrinsic pos=
session can be thus covered without using =93po'e=94 by placing the possess=
or into the appropriate place of the description selbri.
+<p>
+<lx "po'u">  XE "po'u"   Here is an example of =93po'u=94:
+<p>
+<pre><a name=3De3d9>3.9)	le gerku po'u le mi pendo cu cinba mi
+	The dog which-is my friend kisses me.
+<a name=3De3d10>3.10)	le gerku poi du le mi pendo cu cinba mi
+	The dog which =3D my friend kisses me.
+</pre><cx "po'u, compared with =93poi ke'a du=94">  XE "po'u: compared wit=
h poi ke'a du"   <cx "po'u, as identity">  XE "po'u: as identity"   <cx "id=
entity, expressing with po'u">  XE "identity: expressing with po'u"   The c=
mavo =93po'u=94 does not represent possession at all, but rather identity. =
(Note that it means =93poi du=94 and its form was chosen to suggest the rel=
ationship.)
+<p>
+In <a href=3D#e3d9>Example 3.9 </a>, the use of =93po'u=94 tells us that =
=93le gerku=94 and =93le mi pendo=94 represent the same thing. Consider the=
 contrast between <a href=3D#e3d9>Example 3.9 </a>and:
+<p>
+<pre><a name=3De3d11>3.11)	le mi pendo po'u le gerku cu cinba mi
+	My friend which-is the dog kisses me.
+</pre><cx "po'u, relative phrase of contrasted with relativized sumti of">=
  XE "po'u: relative phrase of contrasted with relativized sumti of"   The =
facts of the case are the same, but the listener's knowledge about the situ=
ation may not be. In <a href=3D#e3d9>Example 3.9</a>, the listener is presu=
med not to understand which dog is meant by =93le gerku=94, so the speaker =
adds a relative phrase clarifying that it is the particular dog which is th=
e speaker's friend.
+<p>
+<a href=3D#e3d11>Example 3.11</a>, however, assumes that the listener does=
 not know which of the speaker's friends is referred to, and specifies that=
 it is the friend that is the dog (which dog is taken to be obvious). Here =
is another example of the same contrast:
+<p>
+<ex "New York city">  XE "New York city: example=93  <ex "New York state">=
  XE "New York state: example=93  <pre><a name=3De3d12>3.12)	le tcadu po'u =
la nu,iork
+	The city of New York [not another city].
+<a name=3De3d13>3.13)	la nu,iork po'u le tcadu
+	New York the city (not the state or some other New York)
+</pre><cx "possessor in relative phrases, compared with possessed">  XE "p=
ossessor in relative phrases: compared with possessed"   <cx "possessed in =
relative phrases, compared with possessor">  XE "possessed in relative phra=
ses: compared with possessor"   The principle that the possessor and the po=
ssessed may change places applies to all the GOI cmavo, and allows for the =
possibility of odd effects:
+<p>
+<ex "friend's cup">  XE "friend's cup: example=93  <pre><a name=3De3d14>3.=
14)	le kabri pe le mi pendo cu cmalu
+	The cup associated-with my friend is small.
+	My friend's cup is small
+</pre><ex "cup's friend">  XE "cup's friend: example=93  <pre><a name=3De3=
d15>3.15)	le mi pendo pe le kabri cu cmalu
+	My friend associated-with the cup is small.
+	My friend, the one with the cup, is small.
+</pre><a href=3D#e3d14>Example 3.14 </a>is useful in a context which is ab=
out my friend, and states that his or her cup is small, whereas <a href=3D#=
e3d15>Example 3.15 </a>is useful in a context that is primarily about a cer=
tain cup, and makes a claim about =93my friend of the cup=94, as opposed to=
 some other friend of mine. Here the cup appears to =93possess=94 the perso=
n! English can't even express this relationship with a possessive =97 =93th=
e cup's friend of mine=94 looks like nonsense =97 but Lojban has no trouble=
 doing so.
+<p>
+<lx "ne">  XE "ne"   <lx "no'u">  XE "no'u"   <cx "ne, compared with pe"> =
 XE "ne: compared with pe"   <cx "no'u, compared with po'u">  XE "no'u: com=
pared with po'u"   <cx "pe, compared with ne">  XE "pe: compared with ne"  =
 <cx "po'u, compared with no'u">  XE "po'u: compared with no'u"   <cx "inci=
dental association, expressing with ne">  XE "incidental association: expre=
ssing with ne"   <cx "incidental identification, expressing with no'u">  XE=
 "incidental identification: expressing with no'u"   Finally, the cmavo =93=
ne=94 and =93no'u=94 stand to =93pe=94 and =93po'u=94, respectively, as =93=
noi=94 does to =93poi=94 =97 they provide incidental information:
+<p>
+<pre><a name=3De3d16>3.16)	le blabi gerku ne mi cu batci do
+	The white dog, incidentally-associated-with me, bites you.
+	The white dog, which is mine, bites you.
+</pre><p>In <a href=3D#e3d16>Example 3.16</a>, the white dog is already fu=
lly identified (after all, presumably the listener know which dog bit him o=
r her!). The fact that it is yours is merely incidental to the main bridi c=
laim.
+<p>
+<cx "no'u, contrasted with po'u">  XE "no'u: contrasted with po'u"   <cx "=
po'u, contrasted with no'u">  XE "po'u: contrasted with no'u"   Distinguish=
ing between =93po'u=94 and =93no'u=94 can be a little tricky. Consider a ro=
om with several men in it, one of whom is named Jim. If you don't know thei=
r names, I might say:
+<p>
+<pre><a name=3De3d17>3.17)	le nanmu no'u la djim. cu terpemci
+	The man, incidentally-who-is Jim, is-a-poet.
+	The man, Jim, is a poet.
+</pre><p>Here I am saying that one of the men is a poet, and incidentally =
telling you that he is Jim. But if you do know the names, then
+<p>
+<pre><a name=3De3d18>3.18)	le nanmu po'u la djim. cu terpemci
+	The man who-is Jim is-a-poet.
+	The man named Jim is a poet.
+</pre>is appropriate. Now I am using the fact that the man I am speaking o=
f is Jim in order to pick out which man I mean.
+<p>
+<cx "possession, Lojban usage contrasted with English in omission/inclusio=
n">  XE "possession: Lojban usage contrasted with English in omission/inclu=
sion"   <cx "possession, Lojban usage compared with French and German in om=
ission/inclusion">  XE "possession: Lojban usage compared with French and G=
erman in omission/inclusion"   It is worth mentioning that English sometime=
s over-specifies possession from the Lojban point of view (and the point of=
 view of many other languages, including ones closely related to English). =
The idiomatic English sentence
+<p>
+<pre><a name=3De3d19>3.19)	The man put his hands in his pockets.
+</pre>seems strange to a French- or German-speaking person: whose pockets =
would he put his hands into? and even odder, whose hands would he put into =
his pockets? In Lojban, the sentence
+<p>
+<ex "hands in pockets">  XE "hands in pockets: example=93  <pre><a name=3D=
e3d20>3.20)	le nanmu cu punji le xanci le daski
+	The man puts the hand at-locus-the pocket.
+</pre>is very natural. Of course, if the man is in fact putting his hands =
into another's pockets, or another's hands into his pockets, the fact can b=
e specified.
+<p>
+<lx "ge'u">  XE "ge'u"   <lx "GEhU">  XE "GEhU selma=92o"   <cx "ge'u, eli=
dability of from relative phrases">  XE "ge'u: elidability of from relative=
 phrases"   <cx "logical connective, effect on elidability of ge'u from pre=
ceding relative phrase">  XE "logical connectives: effect on elidability of=
 ge'u from preceding relative phrase"   <cx "ge'u, effect of following logi=
cal connective on elidability">  XE "ge'u: effect of following logical conn=
ective on elidability"   <cx "goi, rationale for non-inclusion in relative =
clause chapter">  XE "goi: rationale for non-inclusion in relative clause c=
hapter"   Finally, the elidable terminator for GOI cmavo is =93ge'u=94 of s=
elma'o GEhU; it is almost never required. However, if a logical connective =
immediately follows a sumti modified by a relative phrase, then an explicit=
 =93ge'u=94 is needed to allow the connective to affect the relativized sum=
ti rather than the sumti of the relative phrase. (What about the cmavo afte=
r which selma'o GOI is named? It is discussed in <a href=3Dchap7.html>Chapt=
er 7</a>, as it is not semantically akin to the other kinds of relative phr=
ases, although the syntax is the same.)
+<p>
+a name=3Ds4><h3>Multiple relative clauses: =93zi'e=94</h3>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	zi'e	ZIhE	relative clause joiner
+</pre><cx "multiple relative clauses, attaching with zi'e">  XE "multiple =
relative clauses: attaching with zi'e"   <lx "zi'e">  XE "zi'e"   <lx "ZIhE=
">  XE "ZIhE selma=92o"   Sometimes it is necessary or useful to attach mor=
e than one relative clause to a sumti. This is made possible in Lojban by t=
he cmavo =93zi'e=94 (of selma'o ZIhE), which is used to join one or more re=
lative clauses together into a single unit, thus making them apply to the s=
ame sumti. For example:
+<p>
+<pre><a name=3De4d1>4.1)	le gerku poi blabi zi'e poi batci le nanmu cu kla=
ma
+	The dog which is white and which bites the man goes.
+</pre><cx "zi'e, contrasted with logical connectives">  XE "zi'e: contrast=
ed with logical connectives"   <cx "zi'e, compared with =93and=94">  XE "zi=
'e: compared with English and"   The most usual translation of =93zi'e=94 i=
n English is =93and=94, but =93zi'e=94 is not really a logical connective: =
unlike most of the true logical connectives (which are explained in <a href=
=3Dchap14.html>Chapter 14</a>), it cannot be converted into a logical conne=
ction between sentences.
+<p>
+<cx "multiple relative clauses, connecting different kinds with zi'e">  XE=
 "multiple relative clauses: connecting different kinds with zi'e"   <cx "r=
elative clause, connecting to relative phrase with zi'e">  XE "relative cla=
use: connecting to relative phrase with zi'e"   <cx "relative phrase, conne=
cting to relative clause with zi'e">  XE "relative phrase: connecting to re=
lative clause with zi'e"   It is perfectly correct to use =93zi'e=94 to con=
nect relative clauses of different kinds:
+<p>
+<pre><a name=3De4d2>4.2)	le gerku poi blabi zi'e noi le mi pendo cu ponse =
ke'a cu klama
+	The dog that-is (white) and incidentally-such-that (my friend owns IT) go=
es.
+	The dog that is white, which my friend owns, is going.
+</pre><p>In <a href=3D#e4d2>Example 4.2</a>, the restrictive clause =93poi=
 blabi=94 specifies which dog is referred to, but the incidental clause =93=
noi le mi pendo cu ponse=94 is mere incidental information: the listener is=
 supposed to already have identified the dog from the =93poi blabi=94. Of c=
ourse, the meaning (though not necessarily the emphasis) is the same if the=
 incidental clause appears first.
+<p>
+<cx "zi'e, use in connecting relative phrase/clause to relative phrase/cla=
use">  XE "zi'e: use in connecting relative phrase/clause to relative phras=
e/clause"   It is also possible to connect relative phrases with =93zi'e=94=
, or a relative phrase with a relative clause:
+<p>
+<pre><a name=3De4d3>4.3)	le botpi po mi zi'e poi blanu cu spofu
+	The bottle specific-to me and which-is blue is-broken.
+	My blue bottle is broken.
+</pre><p>Note that if the colloquial translation of <a href=3D#e4d3>Exampl=
e 4.3 </a>were =93My bottle, which is blue, is broken=94, then =93noi=94 ra=
ther than =93poi=94 would have been correct in the Lojban version, since th=
at version of the English implies that you do not need to know the bottle i=
s blue. As written, <a href=3D#e4d3>Example 4.3 </a>suggests that I probabl=
y have more than one bottle, and the one in question needs to be picked out=
 as the blue one.
+<p>
+<ex "my chair">  XE "my chair: example=93  <pre><a name=3De4d4>4.4)	mi ba =
zutse le stizu pe mi zi'e po do
+		zi'e poi xunre
+	I [future] sit-in the chair associated-with me and specific-to you
+		and which-is red.
+	I will sit in my chair (really yours), the red one.
+</pre><a href=3D#e4d4>Example 4.4 </a>illustrates that more than two relat=
ive phrases or clauses can be connected with =93zi'e=94. It almost defies c=
olloquial translation because of the very un-English contrast between =93pe=
 mi=94, implying that the chair is temporarily connected with me, and =93po=
 do=94, implying that the chair has a more permanent association with you. =
(Perhaps I am a guest in your house, in which case the chair would naturall=
y be your property.)
+<p>
+Here is another example, mixing a relative phrase and two relative clauses=
, a restrictive one and a non-restrictive one:
+<p>
+<ex "beans">  XE "beans: example=93  <pre><a name=3De4d5>4.5)	mi ba citka =
le dembi pe mi
+		zi'e poi cpana le mi palta
+		zi'e noi do dunda ke'a mi
+	I [future] eat the beans associated-with me
+		and which are-upon my plate
+		and which-incidentally you gave IT to-me.
+	I'll eat my beans that are on my plate, the ones you gave me.
+
+</pre><a name=3Ds5><h3>Non-veridical relative clauses: =93voi=94</h3>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	voi	NOI	non-veridical relative clause introducer
+</pre><lx "voi">  XE "voi"   <cx "voi, contrasted with poi in veridicality=
">  XE "voi: contrasted with poi in veridicality"   <cx "poi, contrasted wi=
th voi in veridicality">  XE "poi: contrasted with voi in veridicality"   <=
cx "restrictive relative clauses, non-veridical using voi">  XE "restrictiv=
e relative clauses: non-veridical using voi"   <cx "restrictive relative cl=
auses, veridical using poi">  XE "restrictive relative clauses: veridical u=
sing poi"   There is another member of selma'o NOI which serves to introduc=
e a third kind of relative clause: =93voi=94. Relative clauses introduced b=
y =93voi=94 are restrictive, like those introduced by =93poi=94. However, t=
here is a fundamental difference between =93poi=94 and =93voi=94 relative c=
lauses. A =93poi=94 relative clause is said to be veridical, in the same se=
nse that a description using =93lo=94 or =93loi=94 is: it is essential to t=
he interpretation that the bridi actually be true. For example:
+<p>
+<pre><a name=3De5d1>5.1)	le gerku poi blabi cu klama
+	The dog which is-white goes.
+</pre>it must actually be true that the dog is white, or the sentence cons=
titutes a miscommunication. If there is a white dog and a brown dog, and th=
e speaker uses =93le gerku poi blabi=94 to refer to the brown dog, then the=
 listener will not understand correctly. However,
+<p>
+<pre><a name=3De5d2>5.2)	le gerku voi blabi cu klama
+	The dog which-I-describe-as white goes
+</pre>puts the listener on notice that the dog in question may not actuall=
y meet objective standards (whatever they are) for being white: only the sp=
eaker can say exactly what is meant by the term. In this way, =93voi=94 is =
like =93le=94; the speaker's intention determines the meaning.
+<p>
+As a result, the following two sentences
+<p>
+<ex "man is woman">  XE "man is woman: example=93  <pre><a name=3De5d3>5.3=
)	le nanmu cu ninmu
+	That-which-I-describe-as a-man is-a-woman.
+	The =93guy=94 is actually a gal.
+<a name=3De5d4>5.4)	ti voi nanmu cu ninmu
+	This-thing which-I-describe-as a-man is-a-woman.
+</pre>mean essentially the same thing (except that <a href=3D#e5d5>Example=
 5.5 </a>involves pointing thanks to the use of =93ti=94, whereas <a href=
=3D#e5d4>Example 5.4 </a>doesn't), and neither one is self-contradictory: i=
t is perfectly all right to describe something as a man (although perhaps c=
onfusing to the listener) even if it actually is a woman.
+<p>
+<a name=3Ds6><h3>Relative clauses and descriptors</h3>
+<p>
+<cx "relative clause after sumti, as common placement in sentence">  XE "r=
elative clause after sumti: as common placement in sentence"   So far, this=
 chapter has described the various kinds of relative clauses (including rel=
ative phrases). The list is now complete, and the rest of the chapter will =
be concerned with the syntax of sumti that include relative clauses. So far=
, all relative clauses have appeared directly after the sumti to which they=
 are attached. This is the most common position (and originally the only on=
e), but a variety of other placements are also possible which produce a var=
iety of semantic effects.
+<p>
+<cx "relative clauses, placement in sentence">  XE =93relative clause: pla=
cement in sentence"   <cx "relative clauses, effect of elided =93ku=94 of r=
elativized sumti">  XE =93relative clause: effect of elided ku of relativiz=
ed sumti"   <lx "ku">  XE "ku"   <lx "le">  XE "le"   There are actually th=
ree places where a relative clause can be attached to a description sumti: =
after the descriptor (=93le=94, =93lo=94, or whatever), after the embedded =
selbri but before the elidable terminator (which is =93ku=94), and after th=
e =93ku=94. The relative clauses attached to descriptors that we have seen =
have occupied the second position. Thus <a href=3D#e5d1>Example 5.1</a>, if=
 written out with all elidable terminators, would appear as:
+<p>
+<pre><a name=3De6d1>6.1)	le gerku poi blabi ku'o ku cu klama vau
+	The (dog which (is-white) ) goes.
+	The dog which is white is going.
+</pre><lx "ku'o">  XE "ku'o"   <lx "vau">  XE "vau"   Here =93ku'o=94 is t=
he terminator paired with =93poi=94 and =93ku=94 with =93le=94, and =93vau=
=94 is the terminator of the whole bridi.
+<p>
+<cx "relative clause placement, on sumti with simple descriptor">  XE "rel=
ative clause placement: on sumti with simple descriptor"   When a simple de=
scriptor using =93le=94, like =93le gerku=94, has a relative clause attache=
d, it is purely a matter of style and emphasis where the relative clause sh=
ould go. Therefore, the following examples are all equivalent in meaning to=
 <a href=3D#e6d1>Example 6.1</a>:
+<p>
+<pre><a name=3De6d2>6.2)	le poi blabi ku'o gerku cu klama
+	The such-that (it-is-white) dog goes.
+<a name=3De6d3>6.3)	le gerku ku poi blabi cu klama
+	The (dog) which is-white goes.
+</pre><cx "relative clause placement, English contrasted with Chinese and =
Finnish">  XE "relative clause placement: English contrasted with Chinese a=
nd Finnish"   <cx "relative clause after descriptor, effect on elidability =
of ku'o">  XE "relative clause after descriptor: effect on elidability of k=
u'o"   <cx "ku'o, effect of relative clause after descriptor on elidability=
">  XE "ku'o: effect of relative clause after descriptor on elidability"   =
<a href=3D#e6d1>Example 6.1 </a>will seem most natural to speakers of langu=
ages like English, which always puts relative clauses after the noun phrase=
s they are attached to; <a href=3D#e6d2>Example 6.2</a>, on the other hand,=
 may seem more natural to Finnish or Chinese speakers, who put the relative=
 clause first. Note that in <a href=3D#e6d2>Example 6.2</a>, the elidable t=
erminator =93ku'o=94 must appear, or the selbri of the relative clause (=93=
blabi=94) will merge with the selbri of the description (=93gerku=94), resu=
lting in an ungrammatical sentence. The purpose of the form appearing in <a=
 href=3D#e6d3>Example 6.3 </a>will be apparent shortly.
+<p>
+<cx "inner quantifier of sumti, meaning of">  XE "inner quantifier of sumt=
i: meaning of"   <cx "outer quantifier of sumti, meaning of">  XE "outer qu=
antifier of sumti: meaning of"   As is explained in detail in <a href=3Dcha=
p6.html>Chapter 6</a>, two different numbers (known as the =93inner quantif=
ier=94 and the =93outer quantifier=94) can be attached to a description. Th=
e inner quantifier specifies how many things the descriptor refers to: it a=
ppears between the descriptor and the description selbri. The outer quantif=
ier appears before the descriptor, and specifies how many of the things ref=
erred to by the descriptor are involved in this particular bridi. In the fo=
llowing example,
+<p>
+<ex "five people">  XE "five people: example=93  <pre><a name=3De6d4>6.4)	=
re le mu prenu cu klama le zarci
+	Two-of the five persons go to-the market.
+	Two of the five people [that I have in mind] are going to the market.
+</pre><cx "relative clauses, effect of relativized sumti quantifiers on"> =
 XE =93relative clause: effect of relativized sumti quantifiers on"   <cx "=
quantifiers on sumti, effect on relative clauses for sumti">  XE "quantifie=
r: on sumti, effect on relative clause"   <cx "relative clause placement, e=
ffect on scope">  XE "relative clause placement: effect on scope"   =93=93m=
u=94 is the inner quantifier and =93re=94 is the outer quantifier. Now what=
 is meant by attaching a relative clause to the sumti =93re le mu prenu=94?=
 Suppose the relative clause is =93poi ninmu=94 (meaning =93who are women=
=94). Now the three possible attachment points discussed previously take on=
 significance.
+<p>
+<cx "relative clause before inner quantifier, meaning">  XE "relative clau=
se before inner quantifier: meaning"   <pre><a name=3De6d5>6.5)	re le poi n=
inmu ku'o mu prenu cu klama le zarci
+	Two of the such-that ([they] are-women) five persons go to-the market.
+	Two women out of the five persons go to the market.
+</pre><cx "relative clause before relativized sumti ku, meaning">  XE "rel=
ative clause before relativized sumti ku: meaning"   <ex "five women">  XE =
"five women: example=93  <pre><a name=3De6d6>6.6)	re le mu prenu poi ninmu =
[ku] cu klama le zarci
+	Two of the (five persons which are-women) go to-the market.
+	Two of the five women go to the market.
+</pre><cx "relative clause after relativized sumti ku, meaning">  XE "rela=
tive clause after relativized sumti ku: meaning"   <pre><a name=3De6d7>6.7)=
	re le mu prenu ku poi ninmu cu klama le zarci
+	(Two of the five persons ) which are-women go to-the market.
+	Two women out of the five persons go to the market.
+</pre><cx "relative clause scope, with quantified relativized sumti">  XE =
"relative clause scope: with quantified relativized sumti"   <cx "quantifie=
d sumti, relative clause scope with">  XE "quantified sumti: relative claus=
e scope with"   <cx "relative clauses and quantified sumti, placement consi=
derations">  XE =93relative clause and quantified sumti: placement consider=
ations"   As the parentheses show, <a href=3D#e6d6>Example 6.6 </a>means th=
at all five of the persons are women, whereas <a href=3D#e6d7>Example 6.7 <=
/a>means that the two who are going to the market are women. How do we reme=
mber which is which? If the relative clause comes after the explicit =93ku=
=94, as in <a href=3D#e6d7>Example 6.7</a>, then the sumti as a whole is qu=
alified by the relative clause. If there is no =93ku=94, or if the relative=
 clause comes before an explicit =93ku=94, then the relative clause is unde=
rstood to apply to everything which the underlying selbri applies to.
+<p>
+What about <a href=3D#e6d5>Example 6.5</a>? By convention, it means the sa=
me as <a href=3D#e6d7>Example 6.7</a>, and it requires no =93ku=94, but it =
does typically require a =93ku'o=94 instead. Note that the relative clause =
comes before the inner quantifier.
+<p>
+<cx "relative clause placement considerations,  for simple descriptors con=
trasted with for quantified sumti">  XE "relative clause placement: conside=
rations for simple descriptors contrasted with for quantified sumti"   <cx =
"relative clause placement considerations, for =93lo=94 sumti contrasted wi=
th =93le=94 sumti">  XE "relative clause placement: considerations for lo-s=
umti contrasted with le-sumti"   <cx "lo contrasted with le, for relative c=
lause placement considerations">  XE "lo contrasted with le: for relative c=
lause placement considerations"   <cx "le contrasted with lo, for relative =
clause placement considerations">  XE "le contrasted with lo: for relative =
clause placement considerations"   <cx "relative clauses, impact of le on p=
lacement">  XE =93relative clause: impact of le on placement"   <cx "relati=
ve clauses, impact of lo on placement">  XE =93relative clause: impact of l=
o on placement"   <cx "relative clauses and =93le=94 sumti, placement consi=
derations">  XE =93relative clause and le-sumti: placement considerations" =
  <cx "relative clauses and =93lo=94 sumti, placement considerations">  XE =
=93relative clause and lo-sumti: placement considerations"   When =93le=94 =
is the descriptor being used, and the sumti has no explicit outer quantifie=
r, then the outer quantifier is understood to be =93ro=94 (meaning =93all=
=94), as is explained in <a href=3Dchap6.html>Chapter 6</a>. Thus =93le ger=
ku=94 is taken to mean =93all of the things I refer to as dogs=94, possibly=
 all one of them. In that case, there is no difference between a relative c=
lause after the =93ku=94 or before it. However, if the descriptor is =93lo=
=94, the difference is quite important:
+<p>
+<pre><a name=3De6d8>6.8)	lo prenu ku noi blabi cu klama le zarci
+	(Some persons) incidentally-which are-white go to-the market.
+	Some people, who are white, go to the market.
+<a name=3De6d9>6.9)	lo prenu noi blabi [ku] cu klama le zarci
+	Some (persons incidentally-which are-white) go to-the market.
+	Some of the people, who by the way are white, go to the market.
+</pre><p>Both <a href=3D#e6d{8}>Examples 6.8 </a>and <a href=3D#e6d9>6.9 <=
/a>tell us that one or more persons are going to the market. However, they =
make very different incidental claims. Now, what does =93lo prenu noi blabi=
=94 mean? Well, the default inner quantifier is =93ro=94 (meaning =93all=94=
), and the default outer quantifier is =93su'o=94 (meaning =93at least one=
=94). Therefore, we must first take all persons, then choose at least one o=
f them. That one or more people will be going.
+<p>
+In <a href=3D#e6d8>Example 6.8</a>, the relative clause described the sumt=
i once the outer quantifier was applied: one or more people, who are white,=
 are going. But in <a href=3D#e6d9>Example 6.9</a>, the relative clause act=
ually describes the sumti before the outer quantification is applied, so th=
at it ends up meaning =93First take all persons =97 by the way, they're all=
 white=94. But not all people are white, so the incidental claim being made=
 here is false.
+<p>
+<cx "relative clauses on lo, syntax suggestion">  XE =93relative clause on=
 lo: syntax suggestion"   The safe strategy, therefore, is to always use =
=93ku=94 when attaching a =93noi=94 relative clause to a =93lo=94 descripto=
r. Otherwise we may end up claiming far too much.
+<p>
+<cx "relative clauses, impact of la on placement">  XE =93relative clause:=
 impact of la on placement"   <cx "relative clauses, as part of name">  XE =
=93relative clause: as part of name"   <cx "relative clauses, on names">  X=
E =93relative clause: on names"   <cx "relative clauses and names, placemen=
t considerations">  XE =93relative clause and names: placement consideratio=
ns"   When the descriptor is =93la=94, indicating that what follows is a se=
lbri used for naming, then the positioning of relative clauses has a differ=
ent significance. A relative clause inside the =93ku=94, whether before or =
after the selbri, is reckoned part of the name; a relative clause outside t=
he =93ku=94 is not. Therefore,
+<p>
+<ex "afraid of horse">  XE "afraid of horse: example=93  <pre><a name=3De6=
d10>6.10)	mi viska la nanmu poi terpa le ke'a xirma [ku]
+	I see that-named (=93man which fears the of-IT horse=94).
+	I see Man Afraid Of His Horse.
+</pre>says that the speaker sees a person with a particular name, who does=
 not necessarily fear any horses, whereas
+<p>
+<pre><a name=3De6d11>6.11)	mi viska la nanmu ku poi terpa le ke'a xirma.
+	I see that-named(=93Man=94 ) which fears the of-IT horse.
+	I see the person named =93Man=94 who is afraid of his horse.
+</pre>refers to one (or more) of those named =93Man=94, namely the one(s) =
who are afraid of their horses.
+<p>
+<cx "relative clauses, impact of indefinite sumti on placement">  XE =93re=
lative clause: impact of indefinite sumti on placement"   <cx "relative cla=
uses and indefinite sumti, placement considerations">  XE =93relative claus=
e and indefinite sumti: placement considerations"   Finally, so-called inde=
finite sumti like =93re karce=94, which means almost the same as =93re lo k=
arce=94 (which in turn means the same as =93re lo ro karce=94), can have re=
lative clauses attached; these are taken to be of the outside-the-=93ku=94 =
variety. Here is an example:
+<p>
+<pre><a name=3De6d12>6.12)	mi ponse re karce [ku] poi xekri
+	I possess two cars which-are black.
+</pre><cx "relative clauses, syntax with indefinite sumti">  XE =93relativ=
e clause: syntax with indefinite sumti"   <cx "relative clauses on indefini=
te sumti, syntax considerations">  XE =93relative clause on indefinite sumt=
i: syntax considerations"   The restrictive relative clause only affects th=
e two cars being affected by the main bridi, not all cars that exist. It is=
 ungrammatical to try to place a relative clause within an indefinite sumti=
 (that is, before an explicitly expressed terminating =93ku=94.) Use an exp=
licit =93lo=94 instead.
+<p>
+<a name=3Ds7><h3>Possessive sumti</h3>
+<p>
+<cx "sumti, between descriptor and description selbri">  XE "sumti: betwee=
n descriptor and description selbri"   <lx "pe">  XE "pe"   <cx "relative p=
hrase, compared with possessor sumti">  XE "relative phrase: compared with =
possessive sumti"   <cx "possessor sumti, compared with relative phrase">  =
XE "possessive sumti: compared with relative phrase"   In <a href=3D#e2d{4}=
>Examples 2.4 </a>through <a href=3D#e2d6>2.6</a>, the sumti =93le mi karce=
=94 appears, glossed as =93my car=94. Although it might not seem so, this s=
umti actually contains a relative phrase. When a sumti appears between a de=
scriptor and its description selbri, it is actually a =93pe=94 relative phr=
ase. So
+<p>
+<ex "my">  XE "my: example=93  <pre><a name=3De7d1>7.1)	le mi karce cu xun=
re
+	My car is-red.
+</pre>and
+<p>
+<pre><a name=3De7d2>7.2)	le pe mi karce cu xunre
+	The (associated-with me) car is-red.
+</pre>mean exactly the same thing. Furthermore, since there are no special=
 considerations of quantifiers here,
+<p>
+<pre><a name=3De7d3>7.3)	le karce pe mi cu xunre
+	The car associated-with me is-red.
+</pre><cx "possessive sumti, definition">  XE "possessive sumti: definitio=
n"   <cx "possessor sumti, definition">  XE "possessive sumti: definition" =
  means the same thing as well. A sumti like the one in <a href=3D#e7d1>Exa=
mple 7.1 </a>is called a =93possessive sumti=94. Of course, it does not rea=
lly indicate possession in the sense of ownership, but like =93pe=94 relati=
ve phrases, indicates only weak association; you can say =93le mi karce=94 =
even if you've only borrowed it for the night. (In English, =93my car=94 us=
ually means =93le karce po mi=94, but we do not have the same sense of poss=
ession in =93my seat on the bus=94; Lojban simply makes the weaker sense th=
e standard one.) The inner sumti, =93mi=94 in <a href=3D#e7d1>Example 7.1</=
a>, is correspondingly called the =93possessor sumti=94.
+<p>
+<cx "possessive sumti and relative clauses, development history">  XE "pos=
sessive sumti and relative clauses: development history"   <cx "relative cl=
auses and possessive sumti, development history">  XE =93relative clause an=
d possessive sumti: development history"   Historically, possessive sumti e=
xisted before any other kind of relative phrase or clause, and were retaine=
d when the machinery of relative phrases and clauses as detailed in this ch=
apter so far was slowly built up. When preposed relative clauses of the <a =
href=3D#e7d2>Example 7.2 </a>type were devised, possessive sumti were most =
easily viewed as a special case of them.
+<p>
+<cx "possessor sumti, syntax allowed">  XE "possessive sumti: syntax allow=
ed"   <cx "possessor sumti, contrasted with relative phrases in complexity =
allowed">  XE "possessive sumti: contrasted with relative phrases in comple=
xity allowed"   <cx "relative phrases, contrasted with possessor sumti in c=
omplexity allowed">  XE =93relative phrase: contrasted with possessive sumt=
i in complexity allowed"   <cx "pro-sumti, as possessor sumti">  XE "pro-su=
mti: as possessive sumti"   <cx "names, as possessor sumti">  XE "names: as=
 possessive sumti"   <cx "descriptions, as possessor sumti">  XE "descripti=
ons: as possessive sumti"   <cx "quotations, as possessor sumti">  XE =93qu=
otation: as possessive sumti"   <cx "numbers, as possessor sumti">  XE "num=
bers: as possessive sumti"   Although any sumti, however complex, can appea=
r in a full-fledged relative phrase, only simple sumti can appear as posses=
sor sumti, without a =93pe=94. Roughly speaking, the legal possessor sumti =
are: pro-sumti, quotations, names and descriptions, and numbers. In additio=
n, the possessor sumti may not be preceded by a quantifier, as such a form =
would be interpreted as the unusual =93descriptor + quantifier + sumti=94 t=
ype of description. All these sumti forms are explained in full in <a href=
=3Dchap6.html>Chapter 6</a>.
+
+<p>
+Here is an example of a description used in a possessive sumti:
+<p>
+<pre><a name=3De7d4>7.4)	le le nanmu ku karce cu blanu
+	The (associated-with-the man) car is blue.
+	The man's car is blue.
+<cx "ku, effect of possessor sumti on elidability of">  XE "ku: effect of =
possessive sumti on elidability of"   <cx "possessor sumti, effect on elida=
bility of ku">  XE "possessive sumti: effect on elidability of ku"   Note t=
he explicit =93ku=94 at the end of the possessor sumti, which prevents the =
selbri of the possessor sumti from merging with the selbri of the main desc=
ription sumti. Because of the need for this =93ku=94, the most common kind =
of possessor sumti are pro-sumti, especially personal pro-sumti, which requ=
ire no elidable terminator. Descriptions are more likely to be attached wit=
h relative phrases.
+</pre><p>And here is a number used as a possessor sumti:
+<p>
+<ex "juror 5">  XE "juror 5: example=93  <pre><a name=3De7d5>7.5)	le li mu=
 jdice se bende
+	The of-the-number-five judging team-member
+	Juror number 5
+</pre>which is not quite the same as =93the fifth juror=94; it simply indi=
cates a weak association between the particular juror and the number 5.
+<p>
+<cx "relative clauses, on possessor sumti">  XE =93relative clause: on pos=
sessive sumti"   <cx "possessor sumti, relative clauses on">  XE "possessiv=
e sumti: relative clauses on"   <cx "possessive sumti, with relative clause=
s on possessor sumti">  XE "possessive sumti: with relative clauses on poss=
essive sumti"   <cx "relative clauses with possessive sumti, effect of plac=
ement">  XE =93relative clause with possessive sumti: effect of placement" =
  <cx "possessive sumti with relative clauses, effect of placement">  XE "p=
ossessive sumti with relative clauses: effect of placement"   A possessive =
sumti may also have regular relative clauses attached to it. This would nee=
d no comment if it were not for the following special rule: a relative clau=
se immediately following the possessor sumti is understood to affect the po=
ssessor sumti, not the possessive. For example:
+<p>
+<pre><a name=3De7d6>7.6)	le mi noi sipna vau karce cu na klama
+	The of-me incidentally-which-(is-sleeping) car isn't going.
+</pre>means that my car isn't going; the incidental claim of =93noi sipna=
=94 applies to me, not my car, however. If I wanted to say that the car is =
sleeping (whatever that might mean) I would need:
+<p>
+<pre><a name=3De7d7>7.7)	le mi karce poi sipna cu na klama
+	The of-me car which sleeps isn't going.
+</pre><cx "vau, effect on elidability ku'o">  XE "vau: effect on elidabili=
ty ku'o"   <cx "ku'o, effect of vau on elidability">  XE "ku'o: effect of v=
au on elidability"   Note that <a href=3D#e7d6>Example 7.6 </a>uses =93vau=
=94 rather than =93ku'o=94 at the end of the relative clause: this terminat=
or ends every simple bridi and is almost always elidable; in this case, tho=
ugh, it is a syllable shorter than the equally valid alternative, =93ku'o=
=94.
+<p>
+<a name=3Ds8><h3>Relative clauses and complex sumti: =93vu'o=94</h3>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	vu'o	VUhO	relative clause attacher
+</pre><p>Normally, relative clauses attach only to simple sumti or parts o=
f sumti: pro-sumti, names and descriptions, pure numbers, and quotations. A=
n example of a relative clause attached to a pure number is:
+<p>
+<cx "relative clause, on number">  XE "relative clause: on number"   <ex "=
irrational number">  XE "irrational number: example=93  <pre><a name=3De8d1=
>8.1)	li pai noi na'e frinu namcu
+	The-number pi, incidentally-which is-a-non- fraction number
+	The irrational number pi
+
+</pre><cx "relative clause, on quotation">  XE "relative clause: on quotat=
ion"   And here is an incidental relative clause attached to a quotation:
+<p>
+<pre><a name=3De8d2>8.2)	lu mi klama le zarci li'u
+		noi mi cusku ke'a cu jufra
+	[quote] I go to-the market [unquote]
+		incidentally-which-(I express IT) is-a-sentence.
+	=93I'm going to the market=94, which I'd said, is a sentence.
+</pre>which may serve to identify the author of the quotation or some othe=
r relevant, but subsidiary, fact about it. All such relative clauses appear=
 only after the simple sumti, never before it.
+<p>
+<lx "LAhE">  XE "LAhE selma=92o"   <lx "NAhE">  XE "NAhE selma=92o"   <cx =
"relative clauses, impact of LAhE on placement">  XE =93relative clause: im=
pact of LAhE on placement"   <cx "relative clauses and LAhE, placement cons=
iderations">  XE =93relative clause and LAhE: placement considerations"   <=
cx "relative clauses, impact of NAhE on placement">  XE =93relative clause:=
 impact of NAhE on placement"   <cx "relative clauses and NAhE, placement c=
onsiderations">  XE =93relative clause and NAhE: placement considerations" =
  <cx "LAhE, effect of relative clause placement with">  XE "LAhE selma=92o=
: effect of relative clause placement with"   <cx "NAhE, effect of relative=
 clause placement with">  XE "NAhE selma=92o: effect of relative clause pla=
cement with"   In addition, sumti with attached sumti qualifiers of selma'o=
 LAhE or NAhE+BO (which are explained in detail in <a href=3Dchap6.html>Cha=
pter 6</a>) can have a relative clause appearing after the qualifier and be=
fore the qualified sumti, as in:
+<p>
+<ex "red pony">  XE "red pony: example=93  <lx "la'e">  XE "la'e"   <pre><=
a name=3De8d3>8.3)	la'e poi tolcitno vau lu le xunre cmaxirma li'u
+		cu zvati le vu kumfa
+	A-referent-of (which is-old) [quote] The Red Small-horse [unquote]
+		is-at the [far distance] room.
+	An old =93The Red Pony=94 is in the far room.
+</pre><a href=3D#e8d3>Example 8.3 </a>is a bit complex, and may need some =
picking apart. The quotation =93lu le xunre cmaxirma li'u=94 means the stri=
ng of words =93The Red Pony=94. If the =93la'e=94 at the beginning of the s=
entence were omitted, <a href=3D#e8d3>Example 8.3 </a>would claim that a ce=
rtain string of words is in a room distant from the speaker. But obviously =
a string of words can't be in a room! The effect of the =93la'e=94 is to mo=
dify the sumti so that it refers not to the words themselves, but to the re=
ferent of those words, a novel by John Steinbeck (presumably in Lojban tran=
slation). The particular copy of =93The Red Pony=94 is identified by the re=
strictive relative clause. <a href=3D#e8d3>Example 8.3 </a>means exactly th=
e same as:
+<p>
+<pre><a name=3De8d4>8.4)	la'e lu le xunre cmaxirma li'u lu'u
+		poi to'ercitno cu zvati le vu kumfa
+	A-referent-of ([quote] The Red Small-horse [unquote])
+		which is-old is-at the [far distance] room.
+</pre>and the two sentences can be considered stylistic variants. Note the=
 required =93lu'u=94 terminator, which prevents the relative clause from at=
taching to the quotation itself: we do not wish to refer to an old quotatio=
n!
+<p>
+<cx "relative clauses, on connected sumti">  XE =93relative clause: on con=
nected sumti"   Sometimes, however, it is important to make a relative clau=
se apply to the whole of a more complex sumti, one which involves logical o=
r non-logical connection (explained in <a href=3Dchap14.html>Chapter 14</a>=
). For example,
+<p>
+<pre><a name=3De8d5>8.5)	la frank. .e la djordj. noi nanmu cu klama le zda=
ni
+	Frank and George incidentally-who is-a-man go to-the house.
+	Frank and George, who is a man, go to the house.
+</pre><p>The incidental claim in <a href=3D#e8d5>Example 8.5 </a>is not th=
at Frank and George are men, but only that George is a man, because the inc=
idental relative clause attaches only to =93la djordj=94, the immediately p=
receding simple sumti.
+<p>
+<lx "vu'o">  XE "vu'o"   <lx "VUhO">  XE "VUhO selma=92o"   <cx "relative =
clause scope, extending to preceding sumti with vu'o">  XE "relative clause=
 scope: extending to preceding sumti with vu'o"   To make a relative clause=
 attach to both parts of the logically connected sumti in <a href=3D#e8d5>E=
xample 8.5</a>, a new cmavo is needed, =93vu'o=94 (of selma'o VUhO). It is =
placed between the sumti and the relative clause, and extends the sphere of=
 influence of that relative clause to the entire preceding sumti, including=
 however many logical or non-logical connectives there may be.
+<p>
+<pre><a name=3De8d6>8.6)	la frank. .e la djordj. vu'o noi nanmu cu klama l=
e zdani
+	Frank and George incidentally-who are-men go to-the house.
+	Frank and George, who are men, go to the house.
+</pre><p>The presence of =93vu'o=94 here means that the relative clause =
=93noi nanmu=94 extends to the entire logically connected sumti =93la frank=
. .e la djordj.=94; in other words, both Frank and George are claimed to be=
 men, as the colloquial translation shows.
+<p>
+<cx "relative clauses on complex sumti, Lojban contrasted with English">  =
XE =93relative clause on complex sumti: Lojban contrasted with English"   E=
nglish is able to resolve the distinction correctly in the case of <a href=
=3D#e8d5>Example 8.5 </a>and <a href=3D#e8d6>Example 8.6 </a>by making use =
of number: =93who is=94 rather than =93who are=94. Lojban doesn't distingui=
sh between singular and plural verbs: =93nanmu=94 can mean =93is a man=94 o=
r =93are men=94, so another means is required. Furthermore, Lojban's mechan=
ism works correctly in general: if =93nanmu=94 (meaning =93is-a-man=94) wer=
e replaced with =93pu bajra=94 (=93ran=94), English would have to make the =
distinction some other way:
+<p>
+<pre><a name=3De8d7>8.7)	la frank. .e la djordj. noi pu bajra cu klama le =
zdani
+	Frank and (George who [past] runs) go to-the house.
+	Frank and George, who ran, go to the house.
+<a name=3De8d8>8.8)	la frank. .e la djordj. vu'o noi pu bajra cu klama le =
zdani
+	(Frank and George) who [past] runs go to-the house.
+	Frank and George, who ran, go to the house.
+</pre><p>In spoken English, tone of voice would serve; in written English,=
 one or both sentences would need rewriting.
+<p>
+<a name=3Ds9><h3>Relative clauses in vocative phrases</h3>
+<p>
+<lx "COI">  XE "COI selma=92o"   <lx "DOI">  XE "DOI selma=92o"   Vocative=
 phrases are explained in more detail in <a href=3Dchap6.html>Chapter 6</a>=
. Briefly, they are a method of indicating who a sentence or discourse is a=
ddressed to: of identifying the intended listener. They take three general =
forms, all beginning with cmavo from selma'o COI or DOI (called =93vocative=
 words=94; there can be one or many), followed by either a name, a selbri, =
or a sumti. Here are three examples:
+<p>
+<pre><a name=3De9d1>9.1)	coi. frank.
+	Hello, Frank.
+<a name=3De9d2>9.2)	co'o xirma
+	Goodbye, horse.
+<a name=3De9d3>9.3)	fi'i la frank. .e la djordj.
+	Welcome, Frank and George!
+</pre><p>Note that <a href=3D#e9d2>Example 9.2 </a>says farewell to someth=
ing which doesn't really have to be a horse, something that the speaker sim=
ply thinks of as being a horse, or even might be something (a person, for e=
xample) who is named =93Horse=94. In a sense, <a href=3D#e9d2>Example 9.2 <=
/a>is ambiguous between =93co'o le xirma=94 and =93co'o la xirma=94, a rela=
tively safe semantic ambiguity, since names are ambiguous in general: sayin=
g =93George=94 doesn't distinguish between the possible Georges.
+
+<p>
+Similarly, <a href=3D#e9d1>Example 9.1 </a>can be thought of as an abbrevi=
ation of:
+<p>
+<pre><a name=3De9d4>9.4)	coi la frank.
+	Hello, the-one-named =93Frank=94.
+</pre><p>Syntactically, vocative phrases are a kind of free modifier, and =
can appear in many places in Lojban text, generally at the beginning or end=
 of some complete construct; or, as in <a href=3D#e9d{1}>Examples 9.1 </a>t=
o <a href=3D#e9d3>9.3</a>, as sentences by themselves.
+<p>
+<cx "relative clauses, on vocative phrases">  XE =93relative clause: on vo=
cative phrases"   <cx "vocative phrases, relative clauses on">  XE "vocativ=
e phrase: relative clauses on"   <cx "relative clauses, placement with voca=
tive phrases">  XE =93relative clause: placement with vocative phrases"   <=
cx "vocative phrase with name, placement of relative clause on">  XE "vocat=
ive phrase with name: placement of relative clause on"   As can be seen, th=
e form of vocative phrases is similar to that of sumti, and as you might ex=
pect, vocative phrases allow relative clauses in various places. In vocativ=
e phrases which are simple names (after the vocative words), any relative c=
lauses must come just after the names:
+<p>
+<pre><a name=3De9d5>9.5)	coi. frank. poi xunre se bende
+	Hello, Frank who is-a-red team-member
+	Hello, Frank from the Red Team!
+</pre><p>The restrictive relative clause in <a href=3D#e9d5>Example 9.5 </=
a>suggests that there is some other Frank (perhaps on the Green Team) from =
whom this Frank, the one the speaker is greeting, must be distinguished.
+<p>
+<cx "vocative phrase with selbri, placement of relative clause on">  XE "v=
ocative phrase with selbri: placement of relative clause on"   A vocative p=
hrase containing a selbri can have relative clauses either before or after =
the selbri; both forms have the same meaning. Here are some examples:
+<p>
+<pre><a name=3De9d6>9.6)	co'o poi mi zvati ke'a ku'o xirma
+	Goodbye, such-that-(I am-at IT) horse
+	Goodbye, horse where I am!
+<a name=3De9d7>9.7)	co'o xirma poi mi zvati
+	Goodbye, horse such-that-(I am-at-it).
+</pre><a href=3D#e9d6>Example 9.6 </a>and <a href=3D#e9d7>Example 9.7 </a>=
mean the same thing. In fact, relative clauses can appear in both places.
+<p>
+<a name=3Ds10><h3>Relative clauses within relative clauses</h3>
+<p>
+<cx "relative clauses, relative clauses within">  XE =93relative clause: r=
elative clauses within"   For the most part, these are straightforward and =
uncomplicated: a sumti that is part of a relative clause bridi may itself b=
e modified by a relative clause:
+<p>
+<pre><a name=3De10d1>10.1)	le prenu poi zvati le kumfa poi blanu cu masno
+	The person who is-in the room which is-blue is-slow.
+</pre><cx "ke'a, meaning in relative clause inside relative clause">  XE "=
ke'a: meaning in relative clause inside relative clause"   <cx "relativized=
 sumti, in relative clauses within relative clauses">  XE "relativized sumt=
i: in relative clauses within relative clauses"   <cx "outer sumti, referri=
ng to from within relative clause within relative clause">  XE "outer sumti=
: referring to from within relative clause within relative clause"   <cx "i=
nner sumti, referring to from within relative clause within relative clause=
">  XE "inner sumti: referring to from within relative clause within relati=
ve clause"   <cx "subscripts, use with ke'a for outer sumti reference">  XE=
 "subscripts: use with ke'a for outer sumti reference"   <cx "ke'a with sub=
script, use for outer sumti reference">  XE "ke'a with subscript: use for o=
uter sumti reference"   However, an ambiguity can exist if =93ke'a=94 is us=
ed in a relative clause within a relative clause: does it refer to the oute=
rmost sumti, or to the sumti within the outer relative clause to which the =
inner relative clause is attached? The latter. To refer to the former, use =
a subscript on =93ke'a=94:
+<p>
+<ex "room which he built">  XE "room which he built: example=93  <pre><a n=
ame=3De10d2>10.2)	le prenu poi zvati le kumfa poi ke'axire zbasu ke'a cu ma=
sno
+	The person who is-in the room which IT-sub-2 built IT is-slow.
+	The person who is in the room which he built is slow.
+</pre><lx "ke'axipa">  XE "ke'axipa"   Here, the meaning of =93IT-sub-2=94=
 is that sumti attached to the second relative clause, counting from the in=
nermost, is used. Therefore, =93ke'axipa=94 (IT-sub-1) means the same as pl=
ain =93ke'a=94.
+<p>
+<cx "outer sumti, prenex for referring to from within relative clause with=
in relative clause">  XE "outer sumti: prenex for referring to from within =
relative clause within relative clause"   <cx "prenex, use for outer sumti =
reference">  XE "prenex: use for outer sumti reference"   Alternatively, yo=
u can use a prenex (explained in full in <a href=3Dchap16.html>Chapter 16</=
a>), which is syntactically a series of sumti followed by the special cmavo=
 =93zo'u=94, prefixed to the relative clause bridi:
+<p>
+<pre><a name=3De10d3>10.3)	le prenu poi ke'a goi ko'a zo'u ko'a zvati le k=
umfa
+		poi ke'a goi ko'e zo'u ko'a zbasu ke'a cu masno
+	The man who (IT =3D it1 :  it1 is-in the room
+		which (IT =3D it2 :  it1 built it2) is-slow.
+</pre><a href=3D#e10d3>Example 10.3 </a>is more verbose than <a href=3D#e1=
0d2>Example 10.2</a>, but may be clearer, since it explicitly spells out th=
e two =93ke'a=94 cmavo, each on its own level, and assigns them to the assi=
gnable cmavo =93ko'a=94 and =93ko'e=94 (explained in Chapter 6).
+<p>
+<a name=3Ds11><h3>Index of relative clause cmavo</h3>
+<p>
+<cx "relative clauses, list of cmavo for">  XE =93relative clause: list of=
 cmavo for"   Relative clause introducers (selma'o NOI):
+ 	<pre>noi		incidental clauses
+	poi		restrictive clauses
+	voi		restrictive clauses (non-veridical)
+</pre><p>Relative phrase introducers (selma'o GOI):
+ <pre>	goi		pro-sumti assignment
+	pe		restrictive association
+	ne		incidental association
+	po		extrinsic (alienable) possession
+	po'e		intrinsic (inalienable) possession
+	po'u		restrictive identification
+	no'u		incidental identification
+</pre><p>Relativizing pro-sumti (selma'o KOhA): <pre>
+	ke'a		pro-sumti for relativized sumti
+</pre><p>Relative clause joiner (selma'o ZIhE): <pre>
+	zi'e		joins relative clauses applying to a single sumti
+</pre><p>Relative clause associator (selma'o VUhO): <pre>
+	vu'o		causes relative clauses to apply to all of a complex sumti
+</pre><p>Elidable terminators (each its own selma'o): <pre>
+	ku'o		relative clause elidable terminator
+	ge'u		relative phrase elidable terminator
+
+</body></html>
+<img src=3Dchap9.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D4=
05>
+
+<h2>Chapter 9
+<br>
+To Boston Via The Road Go I, With An Excursion Into The Land Of Modals</h2=
>
+<h6>$Revision: 4.0 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>Introductory</h3>
+<p>
+<cx "sumti, definition">  XE "sumti: definition"   <cx "selbri, definition=
">  XE "selbri: definition"   <cx "bridi", definition">  XE "bridi: definit=
ion"   <cx "sentence, basic Lojban">  XE "sentence: basic Lojban"   <cx "re=
lationship, as basis of sentence">  XE "relationship: as basis of sentence"=
   <cx "relationship, objects of">  XE "relationship: objects of"   The bas=
ic type of Lojban sentence is the bridi: a claim by the speaker that certai=
n objects are related in a certain way. The objects are expressed by Lojban=
 grammatical forms called =93sumti=94; the relationship is expressed by the=
 Lojban grammatical form called a =93selbri=94.
+<p>
+<cx "selbri, place structure of">  XE "selbri: place structure of"   <cx "=
place structure, definition">  XE "place structure: definition"   <cx "plac=
e structure, empty slots in">  XE "place structure: empty slots in"   <cx "=
sumti, as objects in place structure slots">  XE "sumti: as objects in plac=
e structure slots"   The sumti are not randomly associated with the selbri,=
 but according to a systematic pattern known as the =93place structure=94 o=
f the selbri. This chapter describes the various ways in which the place st=
ructure of Lojban bridi is expressed and by which it can be manipulated. Th=
e place structure of a selbri is a sequence of empty slots into which the s=
umti associated with that selbri are placed. The sumti are said to occupy t=
he places of the selbri.
+<p>
+<cx "place structure of selbri, determining">  XE "place structure of selb=
ri: determining"   For our present purposes, every selbri is assumed to hav=
e a well-known place structure. If the selbri is a brivla, the place struct=
ure can be looked up in a dictionary (or, if the brivla is a lujvo not in a=
ny dictionary, inferred from the principles of lujvo construction as explai=
ned in <a href=3Dchap12.html>Chapter 12</a>); if the selbri is a tanru, the=
 place structure is the same as that of the final component in the tanru.
+<p>
+<lx "klama">  XE "klama"   <cx "klama, place structure of">  XE "klama: pl=
ace structure of"   <cx "place structure, notation conventions">  XE "place=
 structure: notation conventions"   <cx "x1, in place structure notation"> =
 XE "x1: in place structure notation"   <ex "go">  XE "go: example=93  The =
stock example of a place structure is that of the gismu =93klama=94:
+klama:  <p>
+<dl compact><dt><dd>x1 comes/goes to destination x2 from origin x3 via rou=
te x4
+employing means of transport x5
+</dl><p>The =93x1=A0<dots>=85</dots>x5=94 indicates that =93klama=94 is a =
five-place predicate, and show the natural order (as assigned by the langua=
ge engineers) of those places: agent, destination, origin, route, means.
+<p>
+<cx "place structure, instability of">  XE "place structure: instability o=
f"   The place structures of brivla are not absolutely stable aspects of th=
e language. The work done so far has attempted to establish a basic place s=
tructure on which all users can, at first, agree. In the light of actual ex=
perience with the individual selbri of the language, there will inevitably =
be some degree of change to the brivla place structures.
+<p>
+<a name=3Ds2><h3>Standard bridi form: =93cu=94</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	cu	CU	prefixed selbri separator
+</pre><cx "bridi, building from selbri and sumti">  XE "bridi: building fr=
om selbri and sumti"   The most usual way of constructing a bridi from a se=
lbri such as =93klama=94 and an appropriate number of sumti is to place the=
 sumti intended for the x1 place before the selbri, and all the other sumti=
 in order after the selbri, thus:
+<p>
+<ex "Boston from Atlanta">  XE "Boston from Atlanta: example=93  <ex "go t=
o Boston from Atlanta">  XE "go to Boston from Atlanta: example=93  <pre><a=
 name=3De2d1>2.1)	mi cu klama la bastn. la .atlantas. le dargu le karce
+	I go to-Boston from-Atlanta via-the road using-the car.
+</pre><p>Here the sumti are assigned to the places as follows:
+
+<p>
+<pre> 		x1	agent	mi
+		x2	destination	la bastn.
+		x3	origin	la .atlantas.
+		x4	route	le dargu
+		x5	means	le karce
+</pre>(Note: Many of the examples in the rest of this chapter will turn ou=
t to have the same meaning as <a href=3D#e2d1>Example 2.1</a>; this fact wi=
ll not be reiterated.)
+<p>
+<cx "bridi, standard form of">  XE "bridi: standard form of"   <cx "standa=
rd bridi form, definition">  XE "standard bridi form: definition"   <cx "br=
idi, non-standard form">  XE "bridi: non-standard form"   This ordering, wi=
th the x1 place before the selbri and all other places in natural order aft=
er the selbri, is called =93standard bridi form=94, and is found in the bul=
k of Lojban bridi, whether used in main sentences or in subordinate clauses=
. However, many other forms are possible, such as:
+<p>
+<pre><a name=3De2d2>2.2)	mi la bastn. la .atlantas. le dargu le karce cu k=
lama
+	I, to-Boston from-Atlanta via-the road using-the car, go.
+</pre><cx "bridi, effect of alternate form on sumti order">  XE "bridi: ef=
fect of alternate form on sumti order"   Here the selbri is at the end; all=
 the sumti are placed before it. However, the same order is maintained.
+<p>
+Similarly, we may split up the sumti, putting some before the selbri and o=
thers after it:
+<p>
+<pre><a name=3De2d3>2.3)	mi la bastn. cu klama la .atlantas. le dargu le k=
arce
+	I to-Boston go from-Atlanta via-the road using-the car.
+</pre><cx "bridi, effect of using non-standard form">  XE "bridi: effect o=
f using non-standard form"   <cx "emphasis, changing by using non-standard =
form of bridi">  XE "emphasis: changing by using non-standard form of bridi=
"   All of the variant forms in this section and following sections can be =
used to place emphasis on the part or parts which have been moved out of th=
eir standard places. Thus, <a href=3D#e2d2>Example 2.2 </a>places emphasis =
on the selbri (because it is at the end); <a href=3D#e2d3>Example 2.3 </a>e=
mphasizes =93la bastn.=94, because it has been moved before the selbri. Mov=
ing more than one component may dilute this emphasis. It is permitted, but =
no stylistic significance has yet been established for drastic reordering.
+<p>
+<lx "cu">  XE "cu"   <lx "CU">  XE "CU selma=92o"   <cx "cu, as selbri sep=
arator">  XE "cu: as selbri separator"   <cx "cu, necessity of">  XE "cu: n=
ecessity of"   <cx "cu, usefulness of">  XE "cu: usefulness of"   <cx "cu, =
effect on elidable terminators">  XE "cu: effect on elidable terminators"  =
 In all these examples, the cmavo =93cu=94 (belonging to selma'o CU) is use=
d to separate the selbri from any preceding sumti. It is never absolutely n=
ecessary to use =93cu=94. However, providing it helps the reader or listene=
r to locate the selbri quickly, and may make it possible to place a complex=
 sumti just before the selbri, allowing the speaker to omit elidable termin=
ators, possibly a whole stream of them, that would otherwise be necessary.
+<p>
+<cx "sumti, order in selbri">  XE "sumti: order in selbri"   <cx "sumti, o=
rder in selbri-first bridi">  XE "sumti: order in selbri-first bridi"   <cx=
 "sumti, omitted first place in selbri-first bridi">  XE "sumti: omitted fi=
rst place in selbri-first bridi"   <cx "selbri-first bridi, effect on sumti=
 places">  XE "selbri-first bridi: effect on sumti places"   <cx "bridi, se=
lbri-first as exceptional">  XE "bridi: selbri-first as exceptional"   The =
general rule, then, is that the selbri may occur anywhere in the bridi as l=
ong as the sumti maintain their order. The only exception (and it is an imp=
ortant one) is that if the selbri appears first, the x1 sumti is taken to h=
ave been omitted:
+<p>
+<pre><a name=3De2d4>2.4)	klama la bastn. la .atlantas. le dargu le karce
+	A-goer to-Boston from-Atlanta via-the road using-the car.
+	Goes to-Boston from-Atlanta via-the road using-the car.
+	Look:  a goer to Boston from Atlanta via the road using the car!
+</pre><cx "bridi, omitting the first sumti place">  XE "bridi: omitting th=
e first sumti place"   <cx "bridi, exception to sumti place structure in"> =
 XE "bridi: exception to sumti place structure in"   <cx "observative, defi=
nition">  XE "observative: definition"   <cx "observative form, contrasted =
with command">  XE "observative form: contrasted with command"   <cx "comma=
nd, contrasted with observative form">  XE "command: contrasted with observ=
ative form"   Here the x1 place is empty: the listener must guess from cont=
ext who is going to Boston. In <a href=3D#e2d4>Example 2.4</a>, =93klama=94=
 is glossed =93a goer=94 rather than =93go=94 because =93Go=94 at the begin=
ning of an English sentence would suggest a command: =93Go to Boston!=94. <=
a href=3D#e2d4>Example 2.4 </a>is not a command, simply a normal statement =
with the x1 place unspecified, causing the emphasis to fall on the selbri =
=93klama=94. Such a bridi, with empty x1, is called an =93observative=94, b=
ecause it usually calls on the listener to observe something in the environ=
ment which would belong in the x1 place. The third translation above shows =
this observative nature. Sometimes it is the relationship itself which the =
listener is asked to observe.
+<p>
+(There is a way to both provide a sumti for the x1 place and put the selbr=
i first in the bridi: see <a href=3D#e3d7>Example 3.7</a>.)
+<p>
+<cx "unspecified trailing sumti, dropping">  XE "unspecified trailing sumt=
i: dropping"   <cx "sumti, dropping trailing unspecified">  XE "sumti: drop=
ping trailing unspecified"   <cx "bridi, leaving end sumti places unspecifi=
ed in">  XE "bridi: leaving end sumti places unspecified in"   Suppose the =
speaker desires to omit a place other than the x1 place? (Presumably it is =
obvious or, for one reason or another, not worth saying.) Places at the end=
 may simply be dropped:
+<p>
+<ex "unspecified route">  XE "unspecified route: example=93  <pre><a name=
=3De2d5>2.5)	mi klama la bastn. la .atlantas.
+	I go to-Boston from-Atlanta (via an unspecified route, using an unspecifi=
ed
+means).
+</pre><cx "unspecified sumti, non-trailing">  XE "unspecified sumti: non-t=
railing"   <a href=3D#e2d5>Example 2.5 </a>has empty x4 and x5 places: the =
speaker does not specify the route or the means of transport. However, simp=
le omission will not work for a place when the places around it are to be s=
pecified: in
+<p>
+<pre><a name=3De2d6>2.6)	mi klama la bastn. la .atlantas. le karce
+	I go to-Boston from-Atlanta via-the car.
+</pre>=93le karce=94 occupies the x4 place, and therefore <a href=3D#e2d6>=
Example 2.6 </a>means:
+<p>
+<dl compact><dt>	<dd>I go to Boston from Atlanta, using the car as a route=
.
+</dl><p>This is nonsense, since a car cannot be a route. What the speaker =
presumably meant is expressed by:
+<p>
+<pre><a name=3De2d7>2.7)	mi klama la bastn. la .atlantas. zo'e le karce
+	I go to-Boston from-Atlanta via-something-unspecified using-the car.
+</pre><lx "zo'e">  XE "zo'e"   <cx "unspecified sumti, using zo'e as place=
-holder for">  XE "unspecified sumti: using zo'e as place-holder for"   <cx=
 "zo'e, as place-holder for unspecified sumti">  XE "zo'e: as place-holder =
for unspecified sumti"   <cx "bridi, leaving a sumti place unspecified in w=
ith zo'e">  XE "bridi: leaving a sumti place unspecified in with zo'e"   <c=
x "place structure, leaving a sumti place unspecified in with zo'e">  XE "p=
lace structure: leaving a sumti place unspecified in with zo'e"   <cx "plac=
e structures, omitting places with zo'e">  XE "place structure: omitting pl=
aces with zo'e"   Here the sumti cmavo =93zo'e=94 is used to explicitly fil=
l the x4 place; =93zo'e=94 means =93the unspecified thing=94 and has the sa=
me meaning as leaving the place empty: the listener must infer the correct =
meaning from context.
+<p>
+<a name=3Ds3><h3>Tagging places: FA</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	fa	FA	tags x1 place
+	fe	FA	tags x2 place
+	fi	FA	tags x3 place
+	fo	FA	tags x4 place
+	fu	FA	tags x5 place
+	fi'a	FA	place structure question
+</pre><lx "FA">  XE "FA selma=92o"   <lx "fa">  XE "fa"   <lx "fe">  XE "f=
e"   <lx "fi">  XE "fi"   <lx "fo">  XE "fo"   <lx "fu">  XE "fu"   <cx "FA=
, syntax of">  XE "FA selma=92o: syntax of"   In sentences like <a href=3D#=
e2d1>Example 2.1</a>, it is easy to get lost and forget which sumti falls i=
n which place, especially if the sumti are more complicated than simple nam=
es or descriptions. The place structure tags of selma'o FA may be used to h=
elp clarify place structures. The five cmavo =93fa=94, =93fe=94, =93fi=94, =
=93fo=94, and =93fu=94 may be inserted just before the sumti in the x1 to x=
5 places respectively:
+<p>
+<pre><a name=3De3d1>3.1)	fa mi cu klama fe la bastn. fi la .atlantas. fo l=
e dargu fu le karce
+	x1=3D I  go  x2=3D Boston  x3=3D Atlanta  x4=3D the road  x5=3D the car.
+	I go to Boston from Atlanta via the road using the car.
+</pre><cx "FA, effect on place structure">  XE "FA selma=92o: effect on pl=
ace structure"   <cx "place structure, effect of FA on">  XE "place structu=
re: effect of FA on"   <cx "place structure, explicitly mapping sumti to pl=
ace with FA">  XE "place structure: explicitly mapping sumti to place with =
FA"   <cx "sumti, explicitly mapping into place structure with FA">  XE "su=
mti: explicitly mapping into place structure with FA"   <cx "FA, for access=
ing a selbri place explicitly by relative number">  XE "FA selma=92o: for a=
ccessing a selbri place explicitly by relative number"   In <a href=3D#e3d1=
>Example 3.1</a>, the tag =93fu=94 before =93le karce=94 clarifies that =93=
le karce=94 occupies the x5 place of =93klama=94. The use of =93fu=94 tells=
 us nothing about the purpose or meaning of the x5 place; it simply says th=
at =93le karce=94 occupies it.
+<p>
+In <a href=3D#e3d1>Example 3.1</a>, the tags are overkill; they serve only=
 to make <a href=3D#e2d1>Example 2.1 </a>even longer than it is. Here is a =
better illustration of the use of FA tags for clarification:
+<p>
+<pre><a name=3De3d2>3.2)	fa mi klama fe le zdani be mi be'o poi nurma vau =
fi la nu,IORK.
+	x1=3D I  go   x2=3D (the house of me) which is-rural   x3=3D New York.
+</pre><p>In <a href=3D#e3d2>Example 3.2</a>, the place structure of =93kla=
ma=94 is as follows:
+<p>
+<pre>	  x1	agent	mi
+	  x2	destination	le zdani be mi be'o poi nurma vau
+	  x3	origin	la nu,IORK.
+	  x4	route	(empty)
+	  x5	means	(empty)
+</pre><cx "FA, as a reminder of place in place structure">  XE "FA selma=
=92o: as a reminder of place in place structure"   The =93fi=94 tag serves =
to remind the hearer that what follows is in the x3 place of =93klama=94; a=
fter listening to the complex sumti occupying the x2 place, it's easy to ge=
t lost.
+<p>
+<cx "FA, effect on place structure order">  XE "FA selma=92o: effect on pl=
ace structure order"   <cx "place structure order, effect of FA on">  XE "p=
lace structure order: effect of FA on"   <cx "selbri-first bridi, specifyin=
g first sumti place in with fa">  XE "selbri-first bridi: specifying first =
sumti place in with fa"   <cx "sumti, re-ordering with FA">  XE "sumti: re-=
ordering with FA"   Of course, once the sumti have been tagged, the order i=
n which they are specified no longer carries the burden of distinguishing t=
he places. Therefore, it is perfectly all right to scramble them into any o=
rder desired, and to move the selbri to anywhere in the bridi, even the beg=
inning:
+<p>
+<pre><a name=3De3d3>3.3)	klama fa mi fi la .atlantas. fu le karce fe la ba=
stn. fo le dargu
+	go  x1=3D I  x3=3D Atlanta  x5=3D the car  x2=3D Boston  x4=3D the road.
+	Go I from Atlanta using the car to Boston via the road.
+</pre><lx "cu">  XE "cu"   <cx "cu, effect of selbri-first bridi on">  XE =
"cu: effect of selbri-first bridi on"   <cx "selbri-first bridi, effect on =
use of cu">  XE "selbri-first bridi: effect on use of cu"   Note that no =
=93cu=94 is permitted before the selbri in <a href=3D#e3d3>Example 3.3</a>,=
 because =93cu=94 separates the selbri from any preceding sumti, and <a hre=
f=3D#e3d3>Example 3.3 </a>has no such sumti.
+<p>
+<pre><a name=3De3d4>3.4)	fu le karce fo le dargu fi la .atlantas. fe la ba=
stn. cu klama fa mi
+	x5=3D the car  x4=3D the road  x3=3D Atlanta  x2=3D Boston  go  x1=3DI
+	Using the car, via the road, from Atlanta to Boston go I.
+</pre><a href=3D#e3d4>Example 3.4 </a>exhibits the reverse of the standard=
 bridi form seen in <a href=3D#e2d{1}>Examples 2.1 </a>and <a href=3D#e3d1>=
3.1</a>, but still means exactly the same thing. If the FA tags were left o=
ut, however, producing:
+<p>
+<pre><a name=3De3d5>3.5)	le karce le dargu la .atlantas. la bastn. cu klam=
a mi
+	The car to-the road from-Atlanta via-Boston goes using-me.
+	The car goes to the road from Atlanta, with Boston as the route, using me
+as a means of transport.
+</pre>the meaning would be wholly changed, and in fact nonsensical.
+<p>
+<cx "place structures, omitting places with FA">  XE "place structure: omi=
tting places with FA"   <cx "FA, compared with zo'e for omitting places">  =
XE "FA selma=92o: compared with zo'e for omitting places"   <cx "zo'e, comp=
ared with FA for omitting places">  XE "zo'e: compared with FA for omitting=
 places"   Tagging places with FA cmavo makes it easy not only to reorder t=
he places but also to omit undesirable ones, without any need for =93zo'e=
=94 or special rules about the x1 place:
+<p>
+<pre><a name=3De3d6>3.6)	klama fi la .atlantas. fe la bastn. fu le karce
+	A-goer  x3=3D Atlanta  x2=3D Boston  x5 =3D the car.
+	A goer from Atlanta to Boston using the car.
+</pre><p>Here the x1 and x4 places are empty, and so no sumti are tagged w=
ith =93fa=94 or =93fo=94; in addition, the x2 and x3 places appear in rever=
se order.
+<p>
+<cx "FA, effect on subsequent non-tagged places">  XE "FA selma=92o: effec=
t on subsequent non-tagged places"   What if some sumti have FA tags and ot=
hers do not? The rule is that after a FA-tagged sumti, any sumti following =
it occupy the places numerically succeeding it, subject to the proviso that=
 an already-filled place is skipped:
+<p>
+<pre><a name=3De3d7>3.7)	klama fa mi la bastn. la .atlantas. le dargu le k=
arce
+	Go  x1=3D I  x2=3D Boston  x3=3D Atlanta  x4=3D the road  x5=3D the car.
+	Go I to Boston from Atlanta via the road using the car.
+</pre><p>In <a href=3D#e3d7>Example 3.7</a>, the =93fa=94 causes =93mi=94 =
to occupy the x1 place, and then the following untagged sumti occupy in ord=
er the x2 through x5 places. This is the mechanism by which Lojban allows p=
lacing the selbri first while specifying a sumti for the x1 place.
+<p>
+Here is a more complex (and more confusing) example:
+<p>
+<pre><a name=3De3d8>3.8)	mi klama fi la .atlantas. le dargu fe la bastn. l=
e karce
+	I  go  x3=3D Atlanta, the road  x2=3D Boston, the car.
+	I go from Atlanta via the road to Boston using the car.
+</pre><p>In <a href=3D#e3d8>Example 3.8</a>, =93mi=94 occupies the x1 plac=
e because it is the first sumti in the sentence (and is before the selbri).=
 The second sumti, =93la .atlantas.=94, occupies the x3 place by virtue of =
the tag =93fi=94, and =93le dargu=94 occupies the x4 place as a result of f=
ollowing =93la .atlantas.=94. Finally, =93la bastn.=94 occupies the x2 plac=
e because of its tag =93fe=94, and =93le karce=94 skips over the already-oc=
cupied x3 and x4 places to land in the x5 place.
+<p>
+<cx "FA, avoidance of complex usage of">  XE "FA selma=92o: avoidance of c=
omplex usage of"   Such a convoluted use of tags should probably be avoided=
 except when trying for a literal translation of some English (or other nat=
ural-language) sentence; the rules stated here are merely given so that som=
e standard interpretation is possible.
+<p>
+<cx "FA, for putting more than one sumti in a single place">  XE "FA selma=
=92o: for putting more than one sumti in a single place"   <cx "sumti, mult=
iple in one place with FA">  XE "sumti: multiple in one place with FA"   <c=
x "multiple sumti in one place, meaning">  XE "multiple sumti in one place:=
 meaning"   It is grammatically permitted to tag more than one sumti with t=
he same FA cmavo. The effect is that of making more than one claim:
+<p>
+<ex "to movie, house, office">  XE "to movie: house, office, example=93  <=
pre><a name=3De3d9>3.9)	[fa] la rik. fa la djein. klama [fe] le skina fe le=
 zdani fe le zarci
+	[x1=3D] Rick  x1=3D Jane  goes-to  x2=3D the movie  x2=3D the house  x2=
=3D the office
+</pre><cx "multiple sumti in one place, avoiding">  XE "multiple sumti in =
one place: avoiding"   may be taken to say that both Rick and Jane go to th=
e movie, the house, and the office, merging six claims into one. More likel=
y, however, it will simply confuse the listener. There are better ways, inv=
olving logical connectives (explained in <a href=3Dchap14.html>Chapter 14</=
a>), to say such things in Lojban. In fact, putting more than one sumti int=
o a place is odd enough that it can only be done by explicit FA usage: this=
 is the motivation for the proviso above, that already-occupied places are =
skipped. In this way, no sumti can be forced into a place already occupied =
unless it has an explicit FA cmavo tagging it.
+<p>
+<lx "fi'a">  XE "fi'a"   <cx "place structure questions">  XE "place struc=
ture questions"   <cx "questions, place structure position">  XE "questions=
: place structure position"   The cmavo =93fi'a=94 also belongs to selma'o =
FA, and allows Lojban users to ask questions about place structures. A brid=
i containing =93fi'a=94 is a question, asking the listener to supply the ap=
propriate other member of FA which will make the bridi a true statement:
+<p>
+<ex "give or receive?">  XE "give or receive?: example=93   <pre><a name=
=3De3d10>3.10)	fi'a do dunda [fe] le vi rozgu
+	[what place]? you give x2=3D the nearby rose
+	In what way are you involved in the giving of this rose?
+	Are you the giver or the receiver of this rose?
+</pre><p>In <a href=3D#e3d10>Example 3.10</a>, the speaker uses the selbri=
 =93dunda=94, whose place structure is:
+<p>
+<dl compact><dt>dunda:  <dd>x1 gives x2 to x3
+</dl><cx "answer, to place structure question">  XE "answers: to place str=
ucture questions"   The tagged sumti =93fi'a do=94 indicates that the speak=
er wishes to know whether the sumti =93do=94 falls in the x1 or the x3 plac=
e (the x2 place is already occupied by =93le rozgu=94). The listener can re=
ply with a sentence consisting solely of a FA cmavo: =93fa=94 if the listen=
er is the giver, =93fi=94 if he/she is the receiver.
+<p>
+<cx "fi'a, effect on subsequent untagged sumti">  XE "fi'a: effect on subs=
equent untagged sumti"   I have inserted the tag =93fe=94 in brackets into =
<a href=3D#e3d10>Example 3.10</a>, but it is actually not necessary, becaus=
e =93fi'a=94 does not count as a numeric tag; therefore, =93le vi rozgu=94 =
would necessarily be in the x2 place even if no tag were present, because i=
t immediately follows the selbri.
+<p>
+There is also another member of FA, namely =93fai=94, which is discussed i=
n <a href=3D#s12>Section 12</a>.
+<p>
+<a name=3Ds4><h3>Conversion: SE</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre>	se	SE	2nd place conversion
+	te	SE	3rd place conversion
+	ve	SE	4th place conversion
+	xe	SE	5th place conversion
+</pre><cx "converted selbri, definition">  XE "converted selbri: definitio=
n"   <cx "converted selbri, forming with SE">  XE "converted selbri: formin=
g with SE"   <cx "converted selbri, place structure of">  XE "converted sel=
bri: place structure of"   <cx "converted selbri, as different selbri from =
unconverted">  XE "converted selbri: as different selbri from unconverted" =
  <lx "SE">  XE "SE selma=92o"   <cx "SE, effect on selbri place structure"=
>  XE "SE selma=92o: effect on selbri place structure"   <lx "se">  XE "se"=
   <cx "SE, for converting place structure">  XE "SE selma=92o: for convert=
ing place structure"   So far we have seen ways to move sumti around within=
 a bridi, but the actual place structure of the selbri has always remained =
untouched. The conversion cmavo of selma'o SE are incorporated within the s=
elbri itself, and produce a new selbri (called a converted selbri) with a d=
ifferent place structure. In particular, after the application of any SE cm=
avo, the number and purposes of the places remain the same, but two of them=
 have been exchanged, the x1 place and another. Which place has been exchan=
ged with x1 depends on the cmavo chosen. Thus, for example, when =93se=94 i=
s used, the x1 place is swapped with the x2 place.
+<p>
+<cx "SE, rationale for no 1st place conversion">  XE "SE selma=92o: ration=
ale for no 1st place conversion"   <cx "SE, word formation of cmavo in">  X=
E "SE selma=92o: word formation of cmavo in"   Note that the cmavo of SE be=
gin with consecutive consonants in alphabetical order. There is no =931st p=
lace conversion=94 cmavo, because exchanging the x1 place with itself is a =
pointless maneuver.
+<p>
+<cx "se klama, place structure of">  XE "se klama: place structure of"   H=
ere are the place structures of =93se klama=94:
+<p>
+<dl compact><dt><dd>x1 is the destination of x2's going from x3 via x4 usi=
ng x5
+</dl><lx "te">  XE "te"   and =93te klama=94:
+<p>
+<dl compact><dt><dd>x1 is the origin and x2 the destination of x3 going vi=
a x4 using x5
+</dl><lx "ve">  XE "ve"   and =93ve klama=94:
+<p>
+<dl compact><dt><dd>x1 is the route to x2 from x3 used by x4 going via x5
+</dl><lx "xe">  XE "xe"   and =93xe klama=94:
+<p>
+<dl compact><dt><dd>x1 is the means in going to x2 from x3 via x4 employed=
 by x5
+</dl><cx "SE, effect on place structure numbering">  XE "SE selma=92o: eff=
ect on place structure numbering"   Note that the place structure numbers i=
n each case continue to be listed in the usual order, x1 to x5.
+<p>
+Consider the following pair of examples:
+<p>
+<pre><a name=3De4d1>4.1)	la bastn. cu se klama mi
+	Boston is-the-destination of-me.
+	Boston is my destination.
+	Boston is gone to by me.
+
+<a name=3De4d2>4.2)	fe la bastn. cu klama fa mi
+	x2=3D Boston  go  x1=3DI.
+	To Boston go I.
+</pre><cx "converted selbri, contrasted with selbri with FA in structure">=
  XE "converted selbri: contrasted with selbri with FA in structure"   <cx =
"FA in selbri, contrasted converted selbri with in structure">  XE "FA in s=
elbri: contrasted converted selbri with in structure"   <cx "converted selb=
ri, compared with selbri with FA in meaning">  XE "converted selbri: compar=
ed with selbri with FA in meaning"   <cx "FA in selbri, compared with conve=
rted selbri in meaning">  XE "FA in selbri: compared with converted selbri =
in meaning"   <cx "converted selbri, as resetting standard order">  XE "con=
verted selbri: as resetting standard order"   <a href=3D#e4d1>Example 4.1 <=
/a>and <a href=3D#e4d2>Example 4.2 </a>mean the same thing, in the sense th=
at there is a relationship of going with the speaker as the agent and Bosto=
n as the destination (and with unspecified origin, route, and means). Struc=
turally, however, they are quite different. <a href=3D#e4d1>Example 4.1 </a=
>has =93la bastn.=94 in the x1 place and =93mi=94 in the x2 place of the se=
lbri =93se klama=94, and uses standard bridi order; <a href=3D#e4d2>Example=
 4.2 </a>has =93mi=94 in the x1 place and =93la bastn.=94 in the x2 place o=
f the selbri =93klama=94, and uses a non-standard order.
+<p>
+<cx "converted selbri, in descriptions">  XE "converted selbri: in descrip=
tions"   <lx "LA">  XE "LA selma=92o"   <lx "LE">  XE "LE selma=92o"   The =
most important use of conversion is in the construction of descriptions. A =
description is a sumti which begins with a cmavo of selma'o LA or LE, calle=
d the descriptor, and contains (in the simplest case) a selbri. We have alr=
eady seen the descriptions =93le dargu=94 and =93le karce=94. To this we co=
uld add:
+<p>
+<ex "the go-er">  XE "the go-er: example=93  <pre><a name=3De4d3>4.3)	le k=
lama
+	the go-er, the one who goes
+</pre><cx "descriptions, as based on first place of following selbri">  XE=
 "descriptions: as based on first place of following selbri"   <cx "descrip=
tions, use of SE in">  XE "descriptions: use of SE in"   <cx "converted sel=
bri, to access non-first place in description">  XE "converted selbri: to a=
ccess non-first place in description"   In every case, the description is a=
bout something which fits into the x1 place of the selbri. In order to get =
a description of a destination (that is, something fitting the x2 place of =
=93klama=94), we must convert the selbri to =93se klama=94, whose x1 place =
is a destination. The result is
+<p>
+<ex "the destination">  XE "the destination: example=93  <pre><a name=3De4=
d4>4.4)	le se klama
+	the destination gone to by someone
+</pre><p>Likewise, we can create three more converted descriptions:
+<p>
+<pre><a name=3De4d5>4.5)	le te klama
+	the origin of someone's going
+<a name=3De4d6>4.6)	le ve klama
+	the route of someone's going
+<a name=3De4d7>4.7)	le xe klama
+	the means by which someone goes
+</pre><ex "Mars road">  XE "Mars road: example=93  <cx "converted selbri, =
contrasted with other similar selbri">  XE "converted selbri: contrasted wi=
th other similar selbri"   <cx "converted selbri, retention of basic meanin=
g in">  XE "converted selbri: retention of basic meaning in"   <cx "ve klam=
a, contrasted with pluta">  XE "ve klama: contrasted with pluta"   <cx "plu=
ta, contrasted with ve klama">  XE "pluta: contrasted with ve klama"   <lx =
"ve klama">  XE "ve klama"   <lx "pluta">  XE "pluta"   <a href=3D#e4d6>Exa=
mple 4.6 </a>does not mean =93the route=94 plain and simple: that is =93le =
pluta=94, using a different selbri. It means a route that is used by someon=
e for an act of =93klama=94; that is, a journey with origin and destination=
. A =93road=94 on Mars, on which no one has traveled or is ever likely to, =
may be called =93le pluta=94, but it cannot be =93le ve klama=94, since the=
re exists no one for whom it is =93le ve klama be fo da=94 (the route taken=
 in an actual journey by someone [da]).
+<p>
+<cx "SE, scope of">  XE "SE selma=92o: scope of"   <cx "SE, extending scop=
e of">  XE "SE selma=92o: extending scope of"   <cx "conversion, scope of">=
  XE "conversion: scope of"   <cx "conversion, extending scope of">  XE "co=
nversion: extending scope of"   <lx "ke">  XE "ke"   <lx "ke'e">  XE "ke'e"=
   When converting selbri that are more complex than a single brivla, it is=
 important to realize that the scope of a SE cmavo is only the following br=
ivla (or equivalent unit). In order to convert an entire tanru, it is neces=
sary to enclose the tanru in =93ke<dots>=85</dots>ke'e=94 brackets:
+<p>
+<ex "blue house">  XE "blue house: example=93  <pre><a name=3De4d8>4.8)	mi=
 se ke blanu zdani [ke'e] ti
+	I [2nd conversion] blue house this-thing
+</pre><p>The place structure of =93blanu zdani=94 (blue house) is the same=
 as that of =93zdani=94, by the rule given in <a href=3D#s1>Section 1</a>. =
The place structure of =93zdani=94 is:
+zdani:  <p>
+<dl compact><dt><dd>x1 is a house/nest/lair/den for inhabitant x2
+</dl><p>The place structure of =93se ke blanu zdani [ke'e]=94 is therefore=
:
+<p>
+<dl compact><dt><dd>x1 is the inhabitant of the blue house (etc.) x2
+</dl><p>Consequently, <a href=3D#e4d8>Example 4.8 </a>means:
+<p>
+<dl compact><dt><dd>I am the inhabitant of the blue house which is this th=
ing.
+</dl><p>Conversion applied to only part of a tanru has subtler effects whi=
ch are explained in <a href=3Dchap5.html>Chapter 5</a>.
+<p>
+<cx "SE, effect of multiple on a selbri">  XE "SE selma=92o: effect of mul=
tiple on a selbri"   <cx "multiple SE, effect of ordering">  XE "multiple S=
E: effect of ordering"   <cx "conversion, effect of multiple on a selbri"> =
 XE "conversion: effect of multiple on a selbri"   <cx "multiple conversion=
, effect of ordering">  XE "multiple conversion: effect of ordering"   It i=
s grammatical to convert a selbri more than once with SE; later (inner) con=
versions are applied before earlier (outer) ones. For example, the place st=
ructure of =93se te klama=94 is achieved by exchanging the x1 and x2 place =
of =93te klama=94, producing:
+<p>
+<lx "se te">  XE "se te"   <dl compact><dt><dd>x1 is the destination and x=
2 is the origin of x3 going via x4 using x5
+</dl><p>On the other hand, =93te se klama=94 has a place structure derived=
 from swapping the x1 and x3 places of =93se klama=94:
+<p>
+<dl compact><dt><dd>x1 is the origin of x2's going to x3 via x4 using x5
+</dl><cx "multiple conversion, avoiding">  XE "multiple conversion: avoidi=
ng"   which is quite different. However, multiple conversions like this are=
 never necessary. Arbitrary scrambling of places can be achieved more easil=
y and far more intelligibly with FA tags, and only a single conversion is e=
ver needed in a description.
+<p>
+<cx "conversion, swapping non-first places">  XE "conversion: swapping non=
-first places"   <lx "setese">  XE "setese"   (Although no one has made any=
 real use of it, it is perhaps worth noting that compound conversions of th=
e form =93setese=94, where the first and third cmavo are the same, effectiv=
ely swap the two given places while leaving the others, including x1, alone=
: =93setese=94 (or equivalently =93tesete=94) swap the x2 and x3 places, wh=
ereas =93texete=94 (or =93xetexe=94) swap the x3 and x5 places.)
+<p>
+<a name=3Ds5><h3>Modal places: FIhO, FEhU</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	fi'o	FIhO	modal place prefix
+	fe'u	FEhU	modal terminator
+</pre><p>Sometimes the place structures engineered into Lojban are inadequ=
ate to meet the needs of actual speech. Consider the gismu =93viska=94, who=
se place structure is:
+<p>
+<lx "viska">  XE "viska"   <dl compact><dt><dd>viska:  x1 sees x2 under co=
nditions x3
+</dl><p>Seeing is a threefold relationship, involving an agent (le viska),=
 an object of sight (le se viska), and an environment that makes seeing pos=
sible (le te viska). Seeing is done with one or more eyes, of course; in ge=
neral, the eyes belong to the entity in the x1 place.
+<p>
+<ex "see with left eye">  XE "see with left eye: example=93  <cx "fi'o, as=
 modal tag">  XE "fi'o: as modal tag"   <cx "place structure, adding new pl=
aces to with modal sumti">  XE "place structure: adding new places to with =
modal sumti"   <cx "fi'o, use in adding places to place structure">  XE "fi=
'o: use in adding places to place structure"   <lx "fi'o">  XE "fi'o"   <lx=
 "FIhO">  XE "FIhO selma=92o"   <lx "fe'u">  XE "fe'u"   <lx "FEhU">  XE "F=
EhU selma=92o"   Suppose, however, that you are blind in one eye and are ta=
lking to someone who doesn't know that. You might want to say, =93I see you=
 with the left eye.=94 There is no place in the place structure of =93viska=
=94 such as =93with eye x4=94 or the like. Lojban allows you to solve the p=
roblem by adding a new place, changing the relationship:
+<p>
+<pre><a name=3De5d1>5.1)	mi viska do fi'o kanla [fe'u] le zunle
+	I see you [modal] eye:  the left-thing
+	I see you with the left eye.
+</pre><cx "fi'o, effect on following selbri">  XE "fi'o: effect on followi=
ng selbri"   <cx "modal tag, fi'o with selbri as">  XE "modal tag: fi'o wit=
h selbri as"   <cx "fi'o with selbri, meaning of">  XE "fi'o with selbri: m=
eaning of"   The three-place relation =93viska=94 has now acquired a fourth=
 place specifying the eye used for seeing. The combination of the cmavo =93=
fi'o=94 (of selma'o FIhO) followed by a selbri, in this case the gismu =93k=
anla=94, forms a tag which is prefixed to the sumti filling the new place, =
namely =93le zunle=94. The semantics of =93fi'o kanla le zunle=94 is that =
=93le zunle=94 fills the x1 place of =93kanla=94, whose place structure is
+<p>
+<dl compact><dt><dd>kanla:  x1 is an/the eye of body x2
+</dl><cx "modal place, relation of to selbri">  XE "modal place: relation =
of to selbri"   <cx "fi'o tag, relation of modal sumti following to selbri"=
>  XE "fi'o tag: relation of modal sumti following to selbri"   <cx "modal =
place relation, importance of first place in">  XE "modal place relation: i=
mportance of first place in"   <cx "modal sumti, as first place of modal ta=
g selbri">  XE "modal sumti: as first place of modal tag selbri"   Thus =93=
le zunle=94 is an eye. The x2 place of =93kanla=94 is unspecified and must =
be inferred from the context. It is important to remember that even though =
=93le zunle=94 is placed following =93fi'o kanla=94, semantically it belong=
s in the x1 place of =93kanla=94. The selbri may be terminated with =93fe'u=
=94 (of selma'o FEhU), an elidable terminator which is rarely required unle=
ss a non-logical connective follows the tag (omitting =93fe'u=94 in that ca=
se would make the connective affect the selbri).
+<p>
+<cx "modal place, definition">  XE "modal place: definition"   <cx "modal =
place, rationale for term name">  XE "modal place: rationale for term name"=
   <cx "modal tag, definition (see also sumti tcita)">  XE "modal tag: defi=
nition (see also sumti tcita)"   <cx "modal sumti, definition (see also sel=
tcita sumti)">  XE "modal sumti: definition (see also seltcita sumti)"   <c=
x "sumti tcita, definition (see also modal tag)">  XE "sumti tcita: definit=
ion (see also modal tag)"   <cx "seltcita sumti, definition (see also modal=
 sumti)">  XE "seltcita sumti: definition (see also modal sumti)"   <cx "mo=
dal sumti, position in bridi">  XE "modal sumti: position in bridi"   <cx "=
modal sumti, effect on place structure">  XE "modal sumti: effect on place =
structure"   <cx "modal sumti, and FA marking">  XE "modal sumti: and FA ma=
rking"   The term for such an added place is a =93modal place=94, as distin=
guished from the regular numbered places. (This use of the word =93modal=94=
 is specific to the Loglan Project, and does not agree with the standard us=
es in either logic or linguistics, but is now too entrenched to change easi=
ly.) The =93fi'o=94 construction marking a modal place is called a =93modal=
 tag=94, and the sumti which follows it a =93modal sumti=94; the purely Loj=
ban terms =93sumti tcita=94 and =93seltcita sumti=94, respectively, are als=
o commonly used. Modal sumti may be placed anywhere within the bridi, in an=
y order; they have no effect whatever on the rules for assigning unmarked b=
ridi to numbered places, and they may not be marked with FA cmavo.
+<p>
+Consider <a href=3D#e5d1>Example 5.1 </a>again. Another way to view the si=
tuation is to consider the speaker's left eye as a tool, a tool for seeing.=
 The relevant selbri then becomes =93pilno=94, whose place structure is
+<p>
+<dl compact><dt><dd>pilno:  x1 uses x2 as a tool for purpose x3
+</dl>and we can rewrite <a href=3D#e5d1>Example 5.1 </a>as
+<p>
+<pre><a name=3De5d2>5.2)	mi viska do fi'o se pilno le zunle kanla
+	I see you [modal] [conversion] use:  the left eye
+	I see you using my left eye.
+</pre><p>Here the selbri belonging to the modal is =93se pilno=94. The con=
version of =93pilno=94 is necessary in order to get the =93tool=94 place in=
to x1, since only x1 can be the modal sumti. The =93tool user=94 place is t=
he x2 of =93se pilno=94 (because it is the x1 of =93pilno=94) and remains u=
nspecified. The tag =93fi'o pilno=94 would mean =93with tool user=94, leavi=
ng the tool unspecified.
+<p>
+<a name=3Ds6><h3>Modal tags: BAI</h3>
+<p>
+There are certain selbri which seem particularly useful in constructing mo=
dal tags. In particular, =93pilno=94 is one of them. The place structure of=
 =93pilno=94 is:
+<p>
+pilno:  x1 uses x2 as a tool for purpose x3
+</dl><lx "sepi'o">  XE "sepi'o"   <cx "BAI modal tags, rationale for">  XE=
 "BAI modal tags: rationale for"   and almost any selbri which represents a=
n action may need to specify a tool. Having to say =93fi'o se pilno=94 freq=
uently would make many Lojban sentences unnecessarily verbose and clunky, s=
o an abbreviation is provided in the language design: the compound cmavo =
=93sepi'o=94.
+<p>
+<lx "pi'o">  XE "pi'o"   <lx "BAI">  XE "BAI selma=92o"   <cx "BAI, as sho=
rt forms for fi'o constructs">  XE "BAI selma=92o: as short forms for fi'o =
constructs"   <cx "fi'o constructs, short forms as BAI cmavo">  XE "fi'o co=
nstructs: short forms as BAI"   <cx "modal tags, short forms as BAI cmavo">=
  XE "modal tag: short form as BAI"   <cx "conversion, of BAI cmavo">  XE "=
conversion: of BAI"   <lx "SE">  XE "SE selma=92o"   Here =93se=94 is used =
before a cmavo, namely =93pi'o=94, rather than before a brivla. The meaning=
 of this cmavo, which belongs to selma'o BAI, is exactly the same as that o=
f =93fi'o pilno fe'u=94. Since what we want is a tag based on =93se pilno=
=94 rather than =93pilno=94 =97 the tool, not the tool user =97 the grammar=
 allows a BAI cmavo to be converted using a SE cmavo. <a href=3D#e5d2>Examp=
le 5.2 </a>may therefore be rewritten as:
+<p>
+<pre><a name=3De6d1>6.1)	mi viska do sepi'o le zunle kanla
+	I see you with-tool:  the left eye
+	I see you using my left eye.
+</pre><p>The compound cmavo =93sepi'o=94 is much shorter than =93fi'o se p=
ilno [fe'u]=94 and can be thought of as a single word meaning =93with-tool=
=94. The modal tag =93pi'o=94, with no =93se=94, similarly means =93with-to=
ol-user=94, probably a less useful concept. Nevertheless, the parallelism w=
ith the place structure of =93pilno=94 makes the additional syllable worthw=
hile.
+<p>
+<lx "ka'a">  XE "ka'a"   <cx "BAI, effect of conversion on">  XE "BAI selm=
a=92o: effect of conversion on"   <cx "conversion, effect on BAI">  XE "con=
version: effect on BAI"   Some BAI cmavo make sense with as well as without=
 a SE cmavo; for example, =93ka'a=94, the BAI corresponding to the gismu =
=93klama=94, has five usable forms corresponding to the five places of =93k=
lama=94 respectively:
+<p>
+<pre> 	ka'a	with-goer
+	seka'a	with-destination
+	teka'a	with-origin
+	veka'a	with-route
+	xeka'a	with-means-of-transport
+</pre><p>Any of these tags may be used to provide modal places for bridi, =
as in the following examples:
+<p>
+<ex "traveling salesperson">  XE "traveling salesperson: example=93  <ex "=
Avon">  XE "Avon: example=93  <pre><a name=3De6d2>6.2)	la .eivn. cu vecnu l=
oi flira cinta ka'a mi
+	Avon sells a-mass-of face paint with-goer me.
+	I am a traveling cosmetics salesperson for Avon.
+</pre>(<a href=3D#e6d2>Example 6.2 </a>may seem a bit strained, but it ill=
ustrates the way in which an existing selbri, =93vecnu=94 in this case, may=
 have a place added to it which might otherwise seem utterly unrelated.)
+<p>
+<pre><a name=3De6d3>6.3)	mi cadzu seka'a la bratfyd.
+	I walk with-destination Bradford.
+	I am walking to Bradford.
+<a name=3De6d4>6.4)	bloti teka'a la nu,IORK.
+	[Observative:] is-a-boat with-origin New York
+	A boat from New York!
+<a name=3De6d5>6.5)	do bajra veka'a lo djine
+	You run with-route a circle.
+	You are running in circles.
+<a name=3De6d6>6.6)	mi citka xeka'a le vinji
+	I eat with-means-of-transport the airplane.
+	I eat in the airplane.
+</pre><cx "BAI cmavo, rationale for selection">  XE "BAI selma=92o: ration=
ale for selection"   <cx "modal tags, contrasted with English prepositions =
in preciseness">  XE "modal tag: contrasted with English preposition in pre=
ciseness"   <cx "English prepositions, contrasted with modal tags in precis=
eness">  XE "English prepositions: contrasted with modal tags in precisenes=
s"   There are sixty-odd cmavo of selma'o BAI, based on selected gismu that=
 seemed useful in a variety of settings. The list is somewhat biased toward=
 English, because many of the cmavo were selected on the basis of correspon=
ding English prepositions and preposition compounds such as =93with=94, =93=
without=94, and =93by means of=94. The BAI cmavo, however, are far more pre=
cise than English prepositions, because their meanings are fixed by the pla=
ce structures of the corresponding gismu.
+<p>
+<cx "BAI, form of cmavo in">  XE "BAI selma=92o: form of cmavo in"   All B=
AI cmavo have the form CV'V or CVV. Most of them are CV'V, where the C is t=
he first consonant of the corresponding gismu and the two Vs are the two vo=
wels of the gismu. The table in <a href=3D#s16>Section 16 </a>shows the exc=
eptions.
+<p>
+<cx "modal tag, for vague relationship">  XE "modal tag: for vague relatio=
nship"   <cx "vague relationship, modal tag for">  XE "vague relationship: =
modal tag for"   <lx "do'e">  XE "do'e"   There is one additional BAI cmavo=
 that is not derived from a gismu: =93do'e=94. This cmavo is used when an e=
xtra place is needed, but it seems useful to be vague about the semantic im=
plications of the extra place:
+<p>
+<pre><a name=3De6d7>6.7)	lo nanmu be do'e le berti cu klama le tcadu
+	Some man [related to] the north came to-the city.
+	A man of the north came to the city.
+</pre><cx "do'e, compared with English =93of=94">  XE "do'e: compared with=
 English of"   <cx "=93of=94 in English, compared with do'e">  XE "of: in E=
nglish, compared with do'e"   <cx "modal place, on description selbri">  XE=
 "modal place: on description selbri"   <lx "be">  XE "be"   Here =93le ber=
ti=94 is provided as a modal place of the selbri =93nanmu=94, but its exact=
 significance is vague, and is paralleled in the colloquial translation by =
the vague English preposition =93of=94. <a href=3D#e6d7>Example 6.7 </a>als=
o illustrates a modal place bound into a selbri with =93be=94. This constru=
ction is useful when the selbri of a description requires a modal place; th=
is and other uses of =93be=94 are more fully explained in <a href=3Dchap5.h=
tml>Chapter 5</a>.
+<p>
+<a name=3Ds7><h3>Modal sentence connection: the causals</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<lx "ri'a">  XE "ri'a"   <lx "ki'u">  XE "ki'u"   <lx "mu'i">  XE "mu'i"  =
 <lx "ni'i">  XE "ni'i"   <pre> 	ri'a	BAI	rinka modal:  physical cause
+	ki'u	BAI	krinu modal:  justification
+	mu'i	BAI	mukti modal:  motivation
+	ni'i	BAI	nibli modal:  logical entailment
+</pre><p>This section has two purposes. On the one hand, it explains the g=
rammatical construct called =93modal sentence connection=94. On the other, =
it exemplifies some of the more useful BAI cmavo: the causals. (There are o=
ther BAI cmavo which have causal implications: =93ja'e=94 means =93with res=
ult=94, and so =93seja'e=94 means =93with cause of unspecified nature=94; l=
ikewise, =93gau=94 means =93with agent=94 and =93tezu'e=94 means =93with pu=
rpose=94. These other modal cmavo will not be further discussed here, as my=
 purpose is to explain modal sentence connection rather than Lojbanic views=
 of causation.)
+<p>
+<cx "causals, gismu">  XE "causals: gismu"   There are four causal gismu i=
n Lojban, distinguishing different versions of the relationships lumped in =
English as =93causal=94:
+<p>
+<dl compact><dt><dd>rinka:  event x1 physically causes event x2
+krinu:  event x1 is the justification for event x2
+mukti:  event x1 is the (human) motive for event x2
+nibli:  event x1 logically entails event x2
+</dl><cx "modals, for causal gismu">  XE "modals: for causal gismu"   <cx =
"causals, modal">  XE "causals: modal"   <cx "modal causals, implication di=
fferences">  XE "modal causals: implication differences"   Each of these gi=
smu has a related modal: =93ri'a=94, =93ki'u=94, =93mu'i=94, and =93ni'i=94=
 respectively. Using these gismu and these modals, we can create various ca=
usal sentences with different implications:
+<p>
+<ex "plant grows">  XE "plant grows: example=93  <pre><a name=3De7d1>7.1)	=
le spati cu banro ri'a le nu do djacu dunda=20
+fi le spati
+	The plant grows with-physical-cause the event-of you water give=20
+to the plant.
+	The plant grows because you water it.
+
+<a name=3De7d2>7.2)	la djan. cpacu le pamoi se jinga ki'u le nu la djan. j=
inga
+	John gets the first prize with-justification the event-of John wins.
+	John got the first prize because he won.
+<a name=3De7d3>7.3)	mi lebna le cukta mu'i le nu mi viska le cukta
+	I took the book with-motivation the event-of I saw the book.
+	I took the book because I saw it.
+</pre><ex "Socrates">  XE "Socrates: example=93  <pre><a name=3De7d4>7.4)	=
la sokrates. morsi binxo ni'i le nu la sokrates. remna
+	Socrates dead-became with-logical-justification Socrates is-human.
+	Socrates died because Socrates is human.
+</pre><cx "=93because=94, four varieties of">  XE "because: English word, =
four varieties of"   In <a href=3D#e7d{1}>Examples 7.1 </a>through <a href=
=3D#e7d4>7.4</a>, the same English word =93because=94 is used to translate =
all four modals, but the types of cause being expressed are quite different=
. Let us now focus on <a href=3D#e7d1>Example 7.1</a>, and explore some var=
iations on it.
+<p>
+<cx "abstraction bridi, effect on claim of bridi">  XE "abstraction bridi:=
 effect on claim of bridi"   <cx "causals, claiming the relation contrasted=
 with claiming cause and/or effect and/or relation">  XE "causals: claiming=
 the relation contrasted with claiming cause and/or effect and/or relation"=
   As written, <a href=3D#e7d1>Example 7.1 </a>claims that the plant grows,=
 but only refers to the event of watering it in an abstraction bridi (abstr=
actions are explained in <a href=3Dchap11.html>Chapter 11</a>) without actu=
ally making a claim. If I express <a href=3D#e7d1>Example 7.1</a>, I have s=
aid that the plant in fact grows, but I have not said that you actually wat=
er it, merely that there is a causal relationship between watering and grow=
ing. This is semantically asymmetrical. Suppose I wanted to claim that the =
plant was being watered, and only mention its growth as ancillary informati=
on? Then we could reverse the main bridi and the abstraction bridi, saying:
+<p>
+<pre><a name=3De7d5>7.5)	do djacu dunda fi le spati seri'a le nu ri banro
+	You water-give to the plant with-physical-effect it grows.
+	You water the plant; therefore, it grows.
+</pre>with the =93ri'a=94 changed to =93seri'a=94. In addition, there are =
also symmetrical forms:
+<p>
+<pre><a name=3De7d6>7.6)	le nu do djacu dunda fi le spati cu rinka
+		le nu le spati cu banro
+	The event-of (you water-give to the plant) causes
+		the event-of (the plant grows).
+	Your watering the plant causes its growth.
+	If you water the plant, then it grows.
+</pre>does not claim either event, but asserts only the causal relationshi=
p between them. So in <a href=3D#e7d6>Example 7.6</a>, I am not saying that=
 the plant grows nor that you have in fact watered it. The second colloquia=
l translation shows a form of =93if-then=94 in English quite distinct from =
the logical connective =93if-then=94 explained in <a href=3Dchap14.html>Cha=
pter 14</a>.
+<p>
+Suppose we wish to claim both events as well as their causal relationship?=
 We can use one of two methods:
+<p>
+<lx ".iri'abo">  XE "iri'abo"   <lx ".i">  XE "i"   <lx "bo">  XE "bo"   <=
cx "modal sentence connection">  XE "modal sentence connection"   <pre><a n=
ame=3De7d7>7.7)	le spati cu banro  .iri'abo do djacu dunda fi le spati
+	The plant grows. Because you water-give to the plant.
+	The plant grows because you water it.
+<a name=3De7d8>7.8)	do djacu dunda fi le spati  .iseri'abo le spati cu ban=
ro
+	You water-give to the plant.  Therefore it grows.
+	You water the plant; therefore, it grows.
+</pre><p>The compound cmavo =93.iri'abo=94 and =93.iseri'abo=94 serve to c=
onnect two bridi, as the initial =93.i=94 indicates. The final =93bo=94 is =
necessary to prevent the modal from =93taking over=94 the following sumti. =
If the =93bo=94 were omitted from <a href=3D#e7d7>Example 7.7 </a>we would =
have:
+<p>
+<pre><a name=3De7d9>7.9)	le spati cu banro  .i ri'a do djacu dunda fi le s=
pati
+	The plant grows.  Because of you, [something] water-gives to the plant.
+	The plant grows. Because of you, water is given to the plant.
+</pre><p>Because =93ri'a do=94 is a modal sumti in <a href=3D#e7d9>Example=
 7.9</a>, there is no longer an explicit sumti in the x1 place of =93djacu =
dunda=94, and the translation must be changed.
+<p>
+<cx "modal sentence connection, effect on modal">  XE "modal sentence conn=
ection: effect on modal"   <cx "modal sentence connection, relation to moda=
l of first sentence in">  XE "modal sentence connection: relation to modal =
of first sentence in"   <cx "modal sentence connection, relation to modal o=
f second sentence in">  XE "modal sentence connection: relation to modal of=
 second sentence in"   The effect of sentences like <a href=3D#e7d7>Example=
 7.7 </a>and <a href=3D#e7d8>Example 7.8 </a>is that the modal, =93ri'a=94 =
in this example, no longer modifies an explicit sumti. Instead, the sumti i=
s implicit, the event given by a full bridi. Furthermore, there is a second=
 implication: that the first bridi fills the x2 place of the gismu =93rinka=
=94; it specifies an event which is the effect. I am therefore claiming thr=
ee things: that the plant grows, that you have watered it, and that there i=
s a cause-and-effect relationship between the two.
+<p>
+<cx "modal sentence connection, with other than causals">  XE "modal sente=
nce connection: with other than causals"   In principle, any modal tag can =
appear in a sentence connective of the type exemplified by <a href=3D#e7d7>=
Example 7.7 </a>and <a href=3D#e7d8>Example 7.8</a>. However, it makes litt=
le sense to use any modals which do not expect events or other abstractions=
 to fill the places of the corresponding gismu. The sentence connective =93=
.ibaubo=94 is perfectly grammatical, but it is hard to imagine any two sent=
ences which could be connected by an =93in-language=94 modal. This is becau=
se a sentence describes an event, and an event can be a cause or an effect,=
 but not a language.
+<p>
+<a name=3Ds8><h3>Other modal connections</h3>
+<p>
+Like many Lojban grammatical constructions, sentence modal connection has =
both forethought and afterthought forms. (See <a href=3Dchap14.html>Chapter=
 14 </a>for a more detailed discussion of Lojban connectives.) <a href=3D#s=
7>Section 7 </a>exemplifies only afterthought modal connection, illustrated=
 here by:
+<p>
+<ex "grasp water">  XE "grasp water: example=93  <pre><a name=3De8d1>8.1)	=
mi jgari lei djacu .iri'abo mi jgari le kabri
+	I grasp the-mass-of water with-physical-cause I grasp the cup.
+	Causing the mass of water to be grasped by me, I grasped the cup.
+	I grasp the water because I grasp the cup.
+</pre><cx "afterthought connection, definition">  XE "afterthought connect=
ion: definition"   <cx "forethought connection, definition">  XE "forethoug=
ht connection: definition"   An afterthought connection is one that is sign=
aled only by a cmavo (or a compound cmavo, in this case) between the two co=
nstructs being connected. Forethought connection uses a signal both before =
the first construct and between the two: the use of =93both=94 and =93and=
=94 in the first half of this sentence represents a forethought connection =
(though not a modal one).
+<p>
+<cx "forethought modal sentence connection">  XE "forethought modal senten=
ce connection"   <cx "modal sentence connection, forethought">  XE "modal s=
entence connection: forethought"   <lx "gi">  XE "gi"   To make forethought=
 modal sentence connections in Lojban, place the modal plus =93gi=94 before=
 the first bridi, and =93gi=94 between the two. No =93.i=94 is used within =
the construct. The forethought equivalent of <a href=3D#e8d1>Example 8.1 </=
a>is:
+<p>
+<pre><a name=3De8d2>8.2)	ri'agi mi jgari le kabri gi mi jgari lei djacu
+	With-physical-cause I grasp the cup, I grasp the-mass-of water.
+	Because I grasp the cup, I grasp the water.
+</pre><cx "forethought modal sentence connection for causals, order of cau=
se and effect">  XE "forethought modal sentence connection for causals: ord=
er of cause and effect"   <cx "forethought modal sentence connection, relat=
ion to modal of first bridi in">  XE "forethought modal sentence connection=
: relation to modal of first bridi in"   <cx "forethought modal sentence co=
nnection, relation to modal of second bridi in">  XE "forethought modal sen=
tence connection: relation to modal of second bridi in"   Note that the cau=
se, the x1 of =93rinka=94 is now placed first. To keep the two bridi in the=
 original order of <a href=3D#e8d1>Example 8.1</a>, we could say:
+<p>
+<pre><a name=3De8d3>8.3)	seri'agi mi jgari lei djacu gi mi jgari le kabri
+	With-physical-effect I grasp the-mass-of water, I grasp the cup.
+</pre><p>In English, the sentence =93*Therefore I grasp the water, I grasp=
 the cup=94 is ungrammatical, because =93therefore=94 is not grammatically =
equivalent to =93because=94. In Lojban, =93seri'agi=94 can be used just lik=
e =93ri'agi=94.
+<p>
+<cx "modal sentence connection, condensing">  XE "modal sentence connectio=
n: condensing"   When the two bridi joined by a modal connection have one o=
r more elements (selbri or sumti or both) in common, there are various cond=
ensed forms that can be used in place of full modal sentence connection wit=
h both bridi completely stated.
+<p>
+<cx "modal sumti connection">  XE "modal sumti connection"   <cx "sumti mo=
dal connection">  XE "sumti modal connection"   When the bridi are the same=
 except for a single sumti, as in Examples 8.1 through 8.3, then a sumti mo=
dal connection may be employed:
+<p>
+<pre><a name=3De8d4>8.4)	mi jgari ri'agi le kabri gi lei djacu
+	I grasp because the cup, the-mass-of water.
+</pre><a href=3D#e8d4>Example 8.4 </a>means exactly the same as <a href=3D=
#e8d{1}>Examples 8.1 </a>through <a href=3D#e8d3>8.3</a>, but there is no i=
diomatic English translation that will distinguish it from them.
+<p>
+<cx "modal termset connection">  XE "modals: termset connection"   <cx "te=
rmset modal connection">  XE "termset modal connection"   If the two connec=
ted bridi are different in more than one sumti, then a termset may be emplo=
yed. Termsets are explained more fully in <a href=3Dchap14.html>Chapter 14<=
/a>, but are essentially a mechanism for creating connections between multi=
ple sumti simultaneously.
+<p>
+<pre><a name=3De8d5>8.5)	mi dunda le cukta la djan.  .imu'ibo la djan. dun=
da lei jdini mi
+	I gave the book to John.  Motivated-by John gave the-mass-of money to-me.
+	I gave the book to John, because John gave money to me.
+</pre>means the same as:
+<p>
+<lx "nu'i">  XE "nu'i"   <lx "nu'u">  XE "nu'u"   <pre><a name=3De8d6>8.6)=
	nu'i mu'igi mi le cukta la djan. gi la djan. lei jdini mi nu'u dunda
+	[start] because I, the book, John; John, the-mass-of money, me [end] give=
s.
+</pre><p>Here there are three sumti in each half of the termset, because t=
he two bridi share only their selbri.
+<p>
+<cx "modal connection of selbri, using bridi-tail modal connection">  XE "=
modal connection of selbri: using bridi-tail modal connection"   <cx "modal=
 bridi-tail connection">  XE "modal bridi-tail connection"   <cx "bridi-tai=
l modal connection">  XE "bridi-tail modal connection"   There is no modal =
connection between selbri as such: bridi which differ only in the selbri ca=
n be modally connected using bridi-tail modal connection. The bridi-tail co=
nstruct is more fully explained in <a href=3Dchap14.html>Chapter 14</a>, bu=
t essentially it consists of a selbri with optional sumti following it. <a =
href=3D#e7d3>Example 7.3 </a>is suitable for bridi-tail connection, and cou=
ld be shortened to:
+<p>
+<pre><a name=3De8d7>8.7)	mi mu'igi viska le cukta gi lebna le cukta
+	I, because saw the book, took the book.
+</pre><p>Again, no straightforward English translation exists. It is even =
possible to shorten <a href=3D#e8d7>Example 8.7 </a>further to:
+<p>
+<lx "vau">  XE "vau"   <pre><a name=3De8d8>8.8)	mi mu'igi viska gi lebna v=
au le cukta
+	I because saw, therefore took, the book.
+</pre>where =93le cukta=94 is set off by the non-elidable =93vau=94 and is=
 made to belong to both bridi-tails =97 see <a href=3Dchap14.html>Chapter 1=
4 </a>for more explanations.
+<p>
+<cx "vau for shared bridi-tail sumti, avoiding">  XE "vau for shared bridi=
-tail sumti: avoiding"   <cx "shared bridi-tail sumti, avoiding">  XE "shar=
ed bridi-tail sumti: avoiding"   Since this is a chapter on rearranging sum=
ti, it is worth pointing out that <a href=3D#e8d8>Example 8.8 </a>can be fu=
rther rearranged to:
+<p>
+<pre><a name=3De8d9>8.9)	mi le cukta mu'igi viska gi lebna
+	I, the book, because saw, therefore took
+</pre>which doesn't require the extra =93vau=94; all sumti before a conjun=
ction of bridi-tails are shared.
+
+<p>
+<cx "modal operand connection">  XE "modal operand connection"   <cx "oper=
and modal connection">  XE "operand modal connection"   Finally, mathematic=
al operands can be modally connected.
+<p>
+<pre><a name=3De8d10>8.10)	li ny. du li vo =20
+.ini'ibo li ny. du li re su'i re
+	The number n =3D the-number 4. =20
+Entailed-by the-number n =3D the-number 2 + 2.
+	n =3D 4 because n =3D 2 + 2.
+</pre>can be reduced to:
+<p>
+<pre><a name=3De8d11>8.11)	li ny. du li ni'igi vei re su'i re [ve'o] gi vo
+	The-number n =3D the-number because ( 2 + 2 ) therefore 4.
+	n is 2 + 2, and is thus 4.
+</pre><lx "vei">  XE "vei"   <lx "ve'o">  XE "ve'o"   The cmavo =93vei=94 =
and =93ve'o=94 represent mathematical parentheses, and are required so that=
 =93ni'igi=94 affects more than just the immediately following operand, nam=
ely the first =93re=94. (The right parenthesis, =93ve'o=94, is an elidable =
terminator.) As usual, no English translation does <a href=3D#e8d11>Example=
 8.11 </a>justice.
+<p>
+<cx "fi'o, restriction on use">  XE "fi'o: restriction on use"   <cx "moda=
l connectives, fi'o prohibited in">  XE "modal connectives: fi'o prohibited=
 in"   Note: Due to restrictions on the Lojban parsing algorithm, it is not=
 possible to form modal connectives using the =93fi'o=94-plus-selbri form o=
f modal. Only the predefined modals of selma'o BAI can be compounded as sho=
wn in <a href=3D#s7>Sections 7 </a>and <a href=3D#s8>8</a>.
+<p>
+<a name=3Ds9><h3>Modal selbri</h3>
+<p>
+Consider the example:
+<p>
+<ex "under compulsion">  XE "under compulsion: example=93  <lx "bai">  XE =
"bai"   <pre><a name=3De9d1>9.1)	mi tavla bau la lojban. bai tu'a la frank.
+	I speak in-language Lojban with-compeller some-act-by Frank.
+	I speak in Lojban, under compulsion by Frank.
+</pre><cx "modal sumti, leaving vague">  XE "modal sumti: leaving vague"  =
 <cx "modal sumti, unspecified">  XE "modal sumti: unspecified"   <a href=
=3D#e9d1>Example 9.1 </a>has two modal sumti, using the modals =93bau=94 an=
d =93bai=94. Suppose we wanted to specify the language explicitly but be va=
gue about who's doing the compelling. We can simplify <a href=3D#e9d1>Examp=
le 9.1 </a>to:
+<p>
+<lx "ku">  XE "ku"   <pre><a name=3De9d2>9.2)	mi tavla bau la lojban. bai =
[ku].
+	I speak in-language Lojban under-compulsion.
+</pre><p>In <a href=3D#e9d2>Example 9.2</a>, the elidable terminator =93ku=
=94 has taken the place of the sumti which would normally follow =93bai=94.=
 Alternatively, we could specify the one who compels but keep the language =
vague:
+<p>
+<pre><a name=3De9d3>9.3)	mi tavla bau [ku] bai tu'a la frank.
+	I speak in-some-language under-compulsion-by some-act-by Frank.
+</pre><p>We are also free to move the modal-plus-=93ku=94 around the bridi=
:
+<p>
+<pre><a name=3De9d4>9.4)	bau [ku] bai ku mi tavla
+	In-some-language under-compulsion I speak.
+</pre><cx "modal followed by selbri, effect on eliding cu">  XE "modal fol=
lowed by selbri: effect on eliding cu"   An alternative to using =93ku=94 i=
s to place the modal cmavo right before the selbri, following the =93cu=94 =
which often appears there. When a modal is present, the =93cu=94 is almost =
never necessary.
+<p>
+<pre><a name=3De9d5>9.5)	mi bai tavla bau la lojban.
+	I compelledly speak in-language Lojban.
+</pre><cx "modal followed by selbri, compared with tanru modification in m=
eaning">  XE "modal followed by selbri: compared with tanru modification in=
 meaning"   <cx "modal followed by selbri, contrasted with tanru modificati=
on in grammar">  XE "modal followed by selbri: contrasted with tanru modifi=
cation in grammar"   In this use, the modal is like a tanru modifier semant=
ically, although grammatically it is quite distinct. <a href=3D#e9d5>Exampl=
e 9.5 </a>is very similar in meaning to:
+<p>
+<pre><a name=3De9d6>9.6)	mi se bapli tavla bau la lojban.
+	I compelledly-speak in-language Lojban.
+</pre><p>The =93se=94 conversion is needed because =93bapli tavla=94 would=
 be a =93compeller type of speaker=94 rather than a =93compelled [by someon=
e] type of speaker=94, which is what a =93bai tavla=94 is.
+<p>
+<cx "fi'o modal followed by selbri, effect on eliding fe'u">  XE "fi'o mod=
al followed by selbri: effect on eliding fe'u"   If the modal preceding a s=
elbri is constructed using =93fi'o=94, then =93fe'u=94 is required to preve=
nt the main selbri and the modal selbri from colliding:
+<p>
+<ex "see with eye">  XE "see with eye: example=93  <pre><a name=3De9d7>9.7=
)	mi fi'o kanla fe'u viska do
+	I with-eye see you.
+	I see you with my eye(s).
+</pre><cx "modal, expanding scope over logical connection with ke=85ke'e">=
  XE "modals: expanding scope over logical connection with ke=85ke'e"   <cx=
 "modal, expanding scope over non-logical connection">  XE "modals: expandi=
ng scope over non-logical connection"   <cx "modal, expanding scope over in=
ner modal connection">  XE "modals: expanding scope over inner modal connec=
tion"   <lx "bai ke">  XE "bai ke"   There are two other uses of modals. A =
modal can be attached to a pair of bridi-tails that have already been conne=
cted by a logical, non-logical, or modal connection (see <a href=3Dchap14.h=
tml>Chapter 14 </a>for more on logical and non-logical connections):
+<p>
+<pre><a name=3De9d8>9.8)	mi bai ke ge klama le zarci gi cadzu le bisli [ke=
'e]
+	I under-compulsion ( both go to-the market and walk on-the ice ).
+	Under compulsion, I both go to the market and walk on the ice.
+</pre><p>Here the =93bai=94 is spread over both =93klama le zarci=94 and =
=93cadzu le bisli=94, and the =93ge=A0<dots>=85</dots>gi=94 represents the =
logical connection =93both-and=94 between the two.
+<p>
+<cx "modal, expanding scope over multiple sentences with tu'e=85tu'u">  XE=
 "modals: expanding scope over multiple sentences with tu'e=85tu'u"   <lx "=
tu'e">  XE "tu'e"   <lx "tu'u">  XE "tu'u"   Similarly, a modal can be atta=
ched to multiple sentences that have been combined with =93tu'e=94 and =93t=
u'u=94, which are explained in more detail in <a href=3Dchap19.html>Chapter=
 19</a>:
+<p>
+<pre><a name=3De9d9>9.9)	bai tu'e mi klama le zarci  .i mi cadzu le bisli =
[tu'u]
+	Under-compulsion [start] I go to-the market.  I walk on-the ice [end].
+</pre>means the same thing as <a href=3D#e9d8>Example 9.8</a>.
+<p>
+Note: Either BAI modals or =93fi'o=94-plus-selbri modals may correctly be =
used in any of the constructions discussed in this section.
+<p>
+<a name=3Ds10><h3>Modal relative phrases; Comparison</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	pe	GOI	restrictive relative phrase
+	ne	GOI	incidental relative phrase
+	mau	BAI	zmadu modal
+	me'a	BAI	mleca modal
+</pre><p>Relative phrases and clauses are explained in much more detail in=
 <a href=3Dchap8.html>Chapter 8</a>. However, there is a construction which=
 combines a modal with a relative phrase which is relevant to this chapter.=
 Consider the following examples of relative clauses:
+<p>
+<ex "Appassionata">  XE "Appassionata: example=93  <ex "Artur Rubenstein">=
  XE "Artur Rubenstein: example=93  <ex "Beethoven">  XE "Beethoven: exampl=
e=93  <pre><a name=3De10d1>10.1)	la .apasionatas. poi se cusku la .artr. ru=
bnstain. cu se nelci mi
+	The Appassionata which is-expressed-by Artur Rubenstein is-liked-by me.
+<a name=3De10d2>10.2)	la .apasionatas. noi se finti la betovn. cu se nelci=
 mi
+	The Appassionata, which is-created-by Beethoven, is-liked-by me.
+</pre><lx "poi">  XE "poi"   In <a href=3D#e10d1>Example 10.1</a>, =93la .=
apasionatas.=94 refers to a particular performance of the sonata, namely th=
e one performed by Rubenstein. Therefore, the relative clause =93poi se cus=
ku=94 uses the cmavo =93poi=94 (of selma'o NOI) to restrict the meaning of =
=93la .apasionatas=94 to the performance in question.
+<p>
+<lx "noi">  XE "noi"   In <a href=3D#e10d2>Example 10.2</a>, however, =93l=
a .apasionatas.=94 refers to the sonata as a whole, and the information tha=
t it was composed by Beethoven is merely incidental. The cmavo =93noi=94 (a=
lso of selma'o NOI) expresses the incidental nature of this relationship.
+<p>
+<lx "pe">  XE "pe"   <lx "ne">  XE "ne"   The cmavo =93pe=94 and =93ne=94 =
(of selma'o GOI) are roughly equivalent to =93poi=94 and =93noi=94 respecti=
vely, but are followed by sumti rather than full bridi. We can abbreviate <=
a href=3D#e10d1>Example 10.1 </a>and <a href=3D#e10d2>Example 10.2 </a>to:
+<p>
+<pre><a name=3De10d3>10.3)	la .apasionatas pe la .artr. rubnstain. se nelc=
i mi
+	The Appassionata of Artur Rubenstein is-liked-by me.
+<a name=3De10d4>10.4)	la .apasionatas ne la betovn. se nelci mi
+	The Appassionata, which is of Beethoven, is-liked-by me.
+</pre><cx "relative phrases, contrasted with relative clauses in precisene=
ss">  XE =93relative phrase: contrasted with relative clause in preciseness=
"   Here the precise selbri of the relative clauses is lost: all we can tel=
l is that the Appassionata is connected in some way with Rubenstein (in <a =
href=3D#e10d3>Example 10.3</a>) and Beethoven (in <a href=3D#e10d4>Example =
10.4</a>), and that the relationships are respectively restrictive and inci=
dental.
+<p>
+<lx "cu'u">  XE "cu'u"   <lx "fi'e">  XE "fi'e"   It happens that both =93=
cusku=94 and =93finti=94 have BAI cmavo, namely =93cu'u=94 and =93fi'e=94. =
We can recast <a href=3D#e10d3>Example 10.3 </a>and <a href=3D#e10d4>Exampl=
e 10.4 </a>as:
+<p>
+<pre><a name=3De10d5>10.5)	la .apasionatas pe cu'u la .artr. rubnstain. cu=
 se nelci mi
+	The Appassionata expressed-by Artur Rubenstein is-liked-by me.
+<a name=3De10d6>10.6)	la .apasionatas ne fi'e la betovn. cu se nelci mi
+	The Appassionata, invented-by Beethoven, is-liked-by me.
+</pre><cx "relative phrases with modals, compared to relative clauses in p=
reciseness">  XE "relative phrases with modals: compared to relative clause=
s in preciseness"   <cx "relative phrases, improving preciseness with modal=
s">  XE =93relative phrase: improving preciseness with modals"   <cx "modal=
s, improving relative phrase preciseness with ">  XE "modals: improving rel=
ative phrase preciseness with "   <a href=3D#e10d5>Example 10.5 </a>and <a =
href=3D#e10d6>Example 10.6 </a>have the full semantic content of <a href=3D=
#e10d1>Example 10.1 </a>and <a href=3D#e10d2>Example 10.2 </a>respectively.
+<p>
+<lx "mau">  XE "mau"   <lx "me'a">  XE "me'a"   <cx "=93more=94, expressin=
g with relative phrases">  XE "more: English word, expressing with relative=
 phrases"   <cx "=93less=94, expressing with relative phrases">  XE "less: =
English word, expressing with relative phrases"   Modal relative phrases ar=
e often used with the BAI cmavo =93mau=94 and =93me'a=94, which are based o=
n the comparative gismu =93zmadu=94 (more than) and =93mleca=94 (less than)=
 respectively. The place structures are:
+<p>
+<pre>zmadu:  x1 is more than x2 in property/quantity x3 by amount x4
+mleca:  x1 is less than x2 in property/quantity x3 by amount x4
+</pre><p>Here are some examples:
+<p>
+<ex "likes more than">  XE "likes more than: example=93  <pre><a name=3De1=
0d7>10.7)	la frank. nelci la betis. ne semau la meiris.
+	Frank likes Betty, which-is more-than Mary.
+	Frank likes Betty more than (he likes) Mary.
+</pre><a href=3D#e10d7>Example 10.7 </a>requires that Frank likes Betty, b=
ut adds the information that his liking for Betty exceeds his liking for Ma=
ry. The modal appears in the form =93semau=94 because the x2 place of =93zm=
adu=94 is the basis for comparison: in this case, Frank's liking for Mary.
+<p>
+<pre><a name=3De10d8>10.8)	la frank. nelci la meiris. ne seme'a la betis.
+	Frank likes Mary, which-is less-than Betty.
+	Frank likes Mary less than (he likes) Betty.
+</pre><cx "mau, avoiding in favor of seme'a">  XE "mau: avoiding in favor =
of seme'a"   <cx "me'a, avoiding in favor of semau">  XE "me'a: avoiding in=
 favor of semau"   Here we are told that Frank likes Mary less than he like=
s Betty; the information about the comparison is the same. It would be poss=
ible to rephrase <a href=3D#e10d7>Example 10.7 </a>using =93me'a=94 rather =
than =93semau=94, and <a href=3D#e10d8>Example 10.8 </a>using =93mau=94 rat=
her than =93seme'a=94, but such usage would be unnecessarily confusing. Lik=
e many BAI cmavo, =93mau=94 and =93me'a=94 are more useful when converted w=
ith =93se=94.
+<p>
+<cx "=93more=94, importance of relative phrase to">  XE "more: English wor=
d, importance of relative phrase to"   <cx "=93less=94, importance of relat=
ive phrase to">  XE "less: English word, importance of relative phrase to" =
  If the =93ne=94 were omitted in <a href=3D#e10d7>Example 10.7 </a>and <a =
href=3D#e10d8>Example 10.8</a>, the modal sumti (=93la meiris.=94 and =93la=
 betis.=94 respectively) would become attached to the bridi as a whole, pro=
ducing a very different translation. <a href=3D#e10d8>Example 10.8 </a>woul=
d become:
+<p>
+<pre><a name=3De10d9>10.9)	la frank. nelci la meiris. seme'a la betis.
+	Frank likes Mary is-less-than Betty.
+	Frank's liking Mary is less than Betty.
+</pre>which compares a liking with a person, and is therefore nonsense.
+<p>
+<cx "comparison, claims related to based on form">  XE "comparison: claims=
 related to based on form"   <cx "comparison with relative phrase, contrast=
ed with bridi-based comparison, in claims about parts">  XE "comparison wit=
h relative phrase: contrasted with bridi-based comparison, in claims about =
parts"   <cx "bridi-based comparison, contrasted with comparison with relat=
ive phrase, in claims about parts">  XE "bridi-based comparison: contrasted=
 with comparison with relative phrase, in claims about parts"   Pure compar=
ison, which states only the comparative information but says nothing about =
whether Frank actually likes either Mary or Betty (he may like neither, but=
 dislike Betty less), would be expressed differently, as:
+<p>
+<pre><a name=3De10d10>10.10)	le ni la frank. nelci la betis. cu zmadu
+		le ni la frank. nelci la meiris.
+	The quantity-of Frank's liking Betty is-more-than
+		the quantity-of Frank's liking Mary.
+</pre><cx "modals often attached with relative phrases, list">  XE "modals=
 often attached with relative phrases: list"   <lx "semau">  XE "semau"   <=
lx "seme'a">  XE "seme'a"   <lx "seba'i">  XE "seba'i"   <lx "ci'u">  XE "c=
i'u"   <lx "de'i">  XE "de'i"   <lx "du'i">  XE "du'i"   The mechanisms exp=
lained in this section are appropriate to many modals other than =93semau=
=94 and =93seme'a=94. Some other modals that are often associated with rela=
tive phrases are: =93seba'i=94 (=93instead of=94), =93ci'u=94 (=93on scale=
=94), =93de'i=94 (=93dated=94), =93du'i=94 (=93as much as=94). Some BAI tag=
s can be used equally well in relative phrases or attached to bridi; others=
 seem useful only attached to bridi. But it is also possible that the usefu=
lness of particular BAI modals is an English-speaker bias, and that speaker=
s of other languages may find other BAIs useful in divergent ways.
+<p>
+<cx "fi'o modals, usage in relative phrases">  XE "fi'o modals: usage in r=
elative phrases"   Note: The uses of modals discussed in this section are a=
pplicable both to BAI modals and to =93fi'o=94-plus-selbri modals.
+<p>
+<a name=3Ds11><h3>Mixed modal connection</h3>
+<p>
+It is possible to mix logical connection (explained in <a href=3Dchap14.ht=
ml>Chapter 14</a>) with modal connection, in a way that simultaneously asse=
rts the logical connection and the modal relationship. Consider the sentenc=
es:
+<p>
+<pre><a name=3De11d1>11.1)	mi nelci do .ije mi nelci la djein.
+	I like you.  And I like Jane
+</pre>which is a logical connection, and
+<p>
+<pre><a name=3De11d2>11.2)	mi nelci do  .iki'ubo mi nelci la djein.
+	I like you.  Justified-by I like Jane.
+</pre><cx "connection, simultaneously modal and logical">  XE "connection:=
 simultaneously modal and logical"   <cx "modal connection, simultaneous wi=
th logical">  XE "modal connection: simultaneous with logical"   <cx "mixed=
 modal connection, of sentences">  XE "mixed modal connection: of sentences=
"   The meanings of <a href=3D#e11d1>Example 11.1 </a>and <a href=3D#e11d2>=
Example 11.2 </a>can be simultaneously expressed by combining the two compo=
und cmavo, thus:
+<p>
+<pre><a name=3De11d3>11.3)	mi nelci do  .ijeki'ubo mi nelci la djein.
+	I like you.  And justified-by I like Jane.
+</pre><cx "mixed modal connection, definition">  XE "mixed modal connectio=
n: definition"   Here the two sentences =93mi nelci do=94 and =93mi nelci l=
a djein.=94 are simultaneously asserted, their logical connection is assert=
ed, and their causal relationship is asserted. The logical connective =93je=
=94 comes before the modal =93ki'u=94 in all such mixed connections.
+<p>
+Since =93mi nelci do=94 and =93mi nelci la djein.=94 differ only in the fi=
nal sumti, we can transform <a href=3D#e11d3>Example 11.3 </a>into a mixed =
sumti connection:
+<p>
+<cx "mixed modal connection, of sumti">  XE "mixed modal connection: of su=
mti"   <pre><a name=3De11d4>11.4)	mi nelci do .eki'ubo la djein.
+	I like you and/because Jane.
+</pre><cx "mixed modal connection, afterthought">  XE "mixed modal connect=
ion: afterthought"   <cx "mixed modal connection, as proscribed in forethou=
ght">  XE "mixed modal connection: as proscribed in forethought"   Note tha=
t this connection is an afterthought one. Mixed connectives are always afte=
rthought; forethought connectives must be either logical or modal.
+<p>
+<cx "mixed modal connection, of bridi-tails">  XE "mixed modal connection:=
 of bridi-tails"   There are numerous other afterthought logical and non-lo=
gical connectives that can have modal information planted within them. For =
example, a bridi-tail connected version of <a href=3D#e11d4>Example 11.4 </=
a>would be:
+<p>
+<pre><a name=3De11d5>11.5)	mi nelci do gi'eki'ubo nelci la djein.
+	I like you and/because like Jane.
+</pre><p>The following three complex examples all mean the same thing.
+<p>
+<ex "carry sack">  XE "carry sack: example=93  <pre><a name=3De11d6>11.6)	=
mi bevri le dakli
+		.ijeseri'abo tu'e mi bevri le gerku  .ijadu'ibo mi bevri le mlatu [tu'u]
+	I carry the sack.
+		And [effect] (I carry the dog.  And/or [equal] I carry the cat.)
+	I carry the sack.  As a result I carry the dog or I carry the cat, equall=
y.
+<a name=3De11d7>11.7)	mi bevri le dakli
+		gi'eseri'ake bevri le gerku gi'adu'ibo bevri le mlatu [ke'e]
+	I carry the sack
+		and [effect] (carry the dog and/or [equal] carry the cat ).
+	I carry the sack and as a result carry the dog or carry the cat equally.
+<a name=3De11d8>11.8)	mi bevri le dakli .eseri'ake le gerku .adu'ibo le ml=
atu [ke'e]
+	I carry the sack and [effect] (the cat and/or [equal] the dog).
+	I carry the sack, and as a result the cat or the dog equally.
+</pre><lx "tu'e">  XE "tu'e"   <lx "tu'u">  XE "tu'u"   <lx "ke">  XE "ke"=
   <lx "ke'e">  XE "ke'e"   In <a href=3D#e11d6>Example 11.6</a>, the =93tu=
'e=A0<dots>=85</dots>tu'u=94 brackets are the equivalent of the =93ke=A0<do=
ts>=85</dots>ke'e=94 brackets in <a href=3D#e11d7>Example 11.7 </a>and <a h=
ref=3D#e11d8>Example 11.8</a>, because =93ke=A0<dots>=85</dots>ke'e=94 cann=
ot extend across more than one sentence. It would also be possible to chang=
e the =93.ijeseri'abo=94 to =93.ije seri'a=94, which would show that the =
=93tu'e=A0<dots>=85</dots>tu'u=94 portion was an effect, but would not pin =
down the =93mi bevri le dakli=94 portion as the cause. It is legal for a mo=
dal (or a tense; see <a href=3Dchap10.html>Chapter 10</a>) to modify the wh=
ole of a =93tu'e=A0<dots>=85</dots>tu'u=94 construct.
+<p>
+<cx "fi'o, mixed modal connection with">  XE "fi'o: mixed modal connection=
 with"   Note: The uses of modals discussed in this section are applicable =
both to BAI modals and to =93fi'o=94-plus-selbri modals.
+<p>
+<a name=3Ds12><h3>Modal conversion: JAI</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	jai	JAI	modal conversion
+	fai	FA	modal place structure tag
+</pre><cx "conversion, swapping with modal place">  XE "conversion: swappi=
ng with modal place"   <lx "SE">  XE "SE selma=92o"   <lx "BAI">  XE "BAI s=
elma=92o"   So far, conversion of numbered bridi places with SE and the add=
ition of modal places with BAI have been two entirely separate operations. =
However, it is possible to convert a selbri in such a way that, rather than=
 exchanging two numbered places, a modal place is made into a numbered plac=
e. For example,
+<p>
+<pre><a name=3De12d1>12.1)	mi cusku bau la lojban.
+	I express [something] in-language Lojban.
+</pre><lx "jai">  XE "jai"   <lx "JAI">  XE "JAI selma=92o"   <cx "modal c=
onversion, grammar of">  XE "modal conversion: grammar of"   <cx "modal con=
version, place structure of">  XE "modal conversion: place structure of"   =
<cx "conversion, modal">  XE "conversion: modal"   <cx "place structure, ef=
fect of modal conversion on">  XE "place structure: effect of modal convers=
ion on"   <lx "fai">  XE "fai"   <lx "FA">  XE "FA selma=92o"   has an expl=
icit x1 place occupied by =93mi=94 and an explicit =93bau=94 place occupied=
 by =93la lojban.=94 To exchange these two, we use a modal conversion opera=
tor consisting of =93jai=94 (of selma'o JAI) followed by the modal cmavo. T=
hus, the modal conversion of <a href=3D#e12d1>Example 12.1 </a>is:
+<p>
+<pre><a name=3De12d2>12.2)	la lojban. jai bau cusku fai mi
+	Lojban is-the-language-of-expression used-by me.
+</pre><cx "modal conversion, access to original first place with fai">  XE=
 "modal conversion: access to original first place with fai"   <cx "fai, as=
 allowing access to original first place in modal conversion">  XE "fai: as=
 allowing access to original first place in modal conversion"   In <a href=
=3D#e12d2>Example 12.2</a>, the modal place =93la lojban.=94 has become the=
 x1 place of the new selbri =93jai bau cusku=94. What has happened to the o=
ld x1 place? There is no numbered place for it to move to, so it moves to a=
 special =93unnumbered place=94 marked by the tag =93fai=94 of selma'o FA.
+<p>
+<cx "fai, effect on numbering of place structure places">  XE "fai: effect=
 on numbering of place structure places"   Note: For the purposes of place =
numbering, =93fai=94 behaves like =93fi'a=94; it does not affect the number=
ing of the other places around it.
+<p>
+<cx "modal conversions, in descriptions">  XE "modal conversions: in descr=
iptions"   Like SE conversions, JAI conversions are especially convenient i=
n descriptions. We may refer to =93the language of an expression=94 as =93l=
e jai bau cusku=94, for example.
+<p>
+<cx "jai without modal, meaning">  XE "jai without modal: meaning"   <cx "=
modal conversion, with no modal specified">  XE "modal conversion: with no =
modal specified"   In addition, it is grammatical to use =93jai=94 without =
a following modal. This usage is not related to modals, but is explained he=
re for completeness. The effect of =93jai=94 by itself is to send the x1 pl=
ace, which should be an abstraction, into the =93fai=94 position, and to ra=
ise one of the sumti from the abstract sub-bridi into the x1 place of the m=
ain bridi. This feature is discussed in more detail in <a href=3Dchap11.htm=
l>Chapter 11</a>. The following two examples mean the same thing:
+<p>
+<pre><a name=3De12d3>12.3)	le nu mi lebna le cukta cu se krinu le nu mi vi=
ska le cukta
+	The event-of (I take the book) is-justified-by the event-of (I see the bo=
ok).
+	My taking the book is justified by my seeing it.
+<a name=3De12d4>12.4)	mi jai se krinu le nu mi viska le cukta kei
+		[fai le nu mi lebna le cukta]
+	I am-justified by the event-of (I see the book)=20
+		[namely, the event-of (I take the book)].
+	I am justified in taking the book by seeing the book.
+</pre><cx "modal conversion without modal, as vague">  XE "modal conversio=
n without modal: as vague"   <a href=3D#e12d4>Example 12.4</a>, with the br=
acketed part omitted, allows us to say that =93I am justified=94 whereas in=
 fact it is my action that is justified. This construction is vague, but us=
eful in representing natural-language methods of expression.
+<p>
+<cx "modal conversion with fi'o">  XE "modal conversion with fi'o"   <cx "=
fi'o, and modal conversion">  XE "fi'o: and modal conversion"   Note: The u=
ses of modals discussed in this section are applicable both to BAI modals a=
nd to =93fi'o=94-plus-selbri modals.
+<p>
+<a name=3Ds13><h3>Modal negation</h3>
+<p>
+<cx "negation of modals">  XE "negation of modals"   <cx "modals, negation=
 of">  XE "modals: negation of"   Negation is explained in detail in <a hre=
f=3Dchap15.html>Chapter 15</a>. There are two forms of negation in Lojban: =
contradictory and scalar negation. Contradictory negation expresses what is=
 false, whereas scalar negation says that some alternative to what has been=
 stated is true. A simple example is the difference between =93John didn't =
go to Paris=94 (contradictory negation) and =93John went to (somewhere) oth=
er than Paris=94 (scalar negation).
+<p>
+<cx "negation of modals, contradictory">  XE "negation of modals: contradi=
ctory"   <lx "nai">  XE "nai"   <cx "modals, contradictory negation of">  X=
E "modals: contradictory negation of"   Contradictory negation involving BA=
I cmavo is performed by appending =93-nai=94 (of selma'o NAI) to the BAI. A=
 common use of modals with =93-nai=94 is to deny a causal relationship:
+<p>
+<pre><a name=3De13d1>13.1)	mi nelci do mu'inai le nu do nelci mi
+	I like you, but not because you like me.
+</pre><cx "contradictory negation of modals, explanation of meaning">  XE =
"contradictory negation of modals: explanation of meaning"   <a href=3D#e13=
d1>Example 13.1 </a>denies that the relationship between my liking you (whi=
ch is asserted) and your liking me (which is not asserted) is one of motiva=
tion. Nothing is said about whether you like me or not, merely that that hy=
pothetical liking is not the motivation for my liking you.
+<p>
+<cx "negation of modals, scalar">  XE "negation of modals: scalar"   <lx "=
na'e">  XE "na'e"   <cx "modals, scalar negation of">  XE "modals: scalar n=
egation of"   Scalar negation is achieved by prefixing =93na'e=94 (of selma=
'o NAhE), or any of the other cmavo of NAhE, to the BAI cmavo.
+<p>
+<ex "plant grows">  XE "plant grows: example=93  <pre><a name=3De13d2>13.2=
)	le spati cu banro na'emu'i le nu
+		do djacu dunda fi le spati
+	The plant grows other-than-motivated-by the event-of
+		you water-give to the plant.
+</pre><cx "scalar negation of modals, explanation of meaning">  XE "scalar=
 negation of modals: explanation of meaning"   <a href=3D#e13d2>Example 13.=
2 </a>says that the relationship between the plant's growth and your wateri=
ng it is not one of motivation: the plant is not motivated to grow, as plan=
ts are not something which can have motivation as a rule. Implicitly, some =
other relationship between watering and growth exists, but <a href=3D#e13d2=
>Example 13.2 </a>doesn't say what it is (presumably =93ri'a=94).
+<p>
+<cx "negation of fi'o modals, by negating selbri">  XE "negation of fi'o m=
odals: by negating selbri"   <cx "fi'o modals, negation of by negating selb=
ri">  XE "fi'o modals: negation of by negating selbri"   Note: Modals made =
with =93fi'o=94 plus a selbri cannot be negated directly. The selbri can it=
self be negated either with contradictory or with scalar negation, however.
+<p>
+<a name=3Ds14><h3>Sticky modals</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre>	ki	KI	stickiness flag
+</pre><cx "sticky modals, definition">  XE "sticky modals: definition"   <=
cx "modals, making long-scope">  XE "modals: making long-scope"   <lx "ki">=
  XE "ki"   <cx "modals, making sticky">  XE "modals: making sticky"   Like=
 tenses, modals can be made persistent from the bridi in which they appear =
to all following bridi. The effect of this =93stickiness=94 is to make the =
modal, along with its following sumti, act as if it appeared in every succe=
ssive bridi. Stickiness is put into effect by following the modal (but not =
any following sumti) with the cmavo =93ki=94 of selma'o KI. For example,
+<p>
+<pre><a name=3De14d1>14.1)	mi tavla bau la lojban. bai ki tu'a la frank.
+		.ibabo mi tavla bau la gliban.
+	I speak in-language Lojban compelled-by some-property-of Frank.
+		Afterward, I speak in-language English.
+</pre>means the same as:
+<p>
+<pre><a name=3De14d2>14.2)	mi tavla bau la lojban. bai tu'a la frank.
+		.ibabo mi tavla bau la gliban.=20
+bai tu'a la frank.
+	I speak in-language Lojban compelled-by some-property-of Frank.
+		Afterward, I speak in-language English
+compelled-by some-property-of Frank.
+</pre><p>In <a href=3D#e14d1>Example 14.1</a>, =93bai=94 is made sticky, a=
nd so Frank's compelling is made applicable to every following bridi. =93ba=
u=94 is not sticky, and so the language may vary from bridi to bridi, and i=
f not specified in a particular bridi, no assumption can safely be made abo=
ut its value.
+<p>
+<cx "sticky modal, canceling">  XE "sticky modals: canceling"   To cancel =
stickiness, use the form =93BAI ki ku=94, which stops any modal value for t=
he specified BAI from being passed to the next bridi. To cancel stickiness =
for all modals simultaneously, and also for any sticky tenses that exist (=
=93ki=94 is used for both modals and tenses), use =93ki=94 by itself, eithe=
r before the selbri or (in the form =93ki ku=94) anywhere in the bridi:
+<p>
+<pre><a name=3De14d3>14.3)	mi ki tavla
+	I speak (no implication about language or compulsion).
+</pre><cx "sticky modals, fi'o proscribed from">  XE "sticky modals: fi'o =
proscribed from"   <cx "fi'o, proscribed for sticky modals">  XE "fi'o: pro=
scribed for sticky modals"   Note: Modals made with =93fi'o=94-plus-selbri =
cannot be made sticky. This is an unfortunate, but unavoidable, restriction=
.
+<p>
+<a name=3Ds15><h3>Logical and non-logical connection of modals</h3>
+<p>
+<cx "logical connection, of modals">  XE "logical connection: of modals"  =
 <cx "non-logical connection, of modals">  XE "non-logical connection: of m=
odals"   Logical and non-logical connectives are explained in detail in <a =
href=3Dchap14.html>Chapter 14</a>. For the purposes of this chapter, it suf=
fices to point out that a logical (or non-logical) connection between two b=
ridi which differ only in a modal can be reduced to a single bridi with a c=
onnective between the modals. As a result, <a href=3D#e15d1>Example 15.1 </=
a>and <a href=3D#e15d2>Example 15.2 </a>mean the same thing:
+<p>
+<pre><a name=3De15d1>15.1)	la frank. bajra seka'a le zdani  .ije la frank.=
 bajra=20
+teka'a le zdani
+	Frank runs with-destination the house.  And Frank runs=20
+with-origin the house.
+	Frank runs to the house, and Frank runs from the house.
+</pre><lx "je">  XE "je"   <pre><a name=3De15d2>15.2)	la frank. bajra seka=
'a je teka'a le zdani
+	Frank runs with-destination and with-origin the house.
+	Frank runs to and from the house.
+</pre><lx "ce'e">  XE "ce'e"   Neither example implies whether a single ac=
t, or two acts, of running is referred to. To compel the sentence to refer =
to a single act of running, you can use the form:
+<p>
+<pre><a name=3De15d3>15.3)	la frank. bajra seka'a le zdani ce'e teka'a le =
zdani
+	Frank runs with-destination the house [joined-to] with-origin the-house.
+</pre><p>The cmavo =93ce'e=94 creates a termset containing two terms (term=
sets are explained in <a href=3Dchap14.html>Chapter 14 </a>and <a href=3Dch=
ap16.html>Chapter 16</a>). When a termset contains more than one modal tag =
derived from a single BAI, the convention is that the two tags are derived =
from a common event.
+<p>
+<a name=3Ds16><h3>CV'V cmavo of selma'o BAI with irregular forms</h3>
+<p>
+<cx "modal cmavo, regular form for derivation">  XE "modal cmavo: regular =
form for derivation"   There are 65 cmavo of selma'o BAI, of which all but =
one (=93do'e=94, discussed in <a href=3D#s6>Section 6</a>), are derived dir=
ectly from selected gismu. Of these 64 cmavo, 36 are entirely regular and h=
ave the form CV'V, where C is the first consonant of the corresponding gism=
u, and the Vs are the two vowels of the gismu. The remaining BAI cmavo, whi=
ch are irregular in one way or another, are listed in the table below. The =
table is divided into sub-tables according to the nature of the exception; =
some cmavo appear in more than one sub-table, and are so noted.
+
+<p>
+<cx "modal cmavo, list of irregular derivation">  XE "modal cmavo: list of=
 irregular derivation"   <pre>	cmavo	gismu	comments
+		Monosyllables of the form CVV:
+	bai	bapli
+	bau	bangu
+	cau	claxu
+	fau	fasnu
+	gau	gasnu
+	kai	ckaji	uses 2nd consonant of gismu
+	mau	zmadu	uses 2nd consonant of gismu
+	koi	korbi
+	rai	traji	uses 2nd consonant of gismu
+	sau	sarcu
+	tai	tamsmi	based on lujvo, not gismu
+	zau	zanru
+		Second consonant of the gismu as the C:
+			(the gismu is always of the form CCVCV)
+	ga'a	zgana
+	kai	ckaji	has CVV form (monosyllable)
+	ki'i	ckini
+	la'u	klani	has irregular 2nd V
+	le'a	klesi	has irregular 2nd V
+	mau	zmadu	has CVV form (monosyllable)
+	me'e	cmene
+	ra'a	srana
+	ra'i	krasi
+	rai	traji	has CVV form (monosyllable)
+	ti'i	stidi
+	tu'i	stuzi
+		Irregular 2nd V:
+	fi'e	finti
+	la'u	klani	uses 2nd consonant of gismu
+	le'a	klesi	uses 2nd consonant of gismu
+	ma'e	marji
+	mu'u	mupli
+	ti'u	tcika
+	va'o	vanbi
+		Special cases:
+	ri'i	lifri	uses 3rd consonant of gismu
+	tai	tamsmi	based on lujvo, not gismu
+	va'u	xamgu	CV'V cmavo can't begin with =93x=94
+
+</pre><a name=3Ds17><h3>Complete table of BAI cmavo with rough English equ=
ivalents</h3>
+<p>
+<cx "modal cmavo, table with English equivalents">  XE "modal cmavo: table=
 with English equivalents"   <cx "modal cmavo table, format of">  XE "modal=
 cmavo table: format of"   The following table shows all the cmavo belongin=
g to selma'o BAI, and has five columns. The first column is the cmavo itsel=
f; the second column is the gismu linked to it. The third column gives an E=
nglish phrase which indicates the meaning of the cmavo; and the fourth colu=
mn indicates its meaning when preceded by =93se=94.
+<p>
+For those cmavo with meaningful =93te=94, =93ve=94, and even =93xe=94 conv=
ersions (depending on the number of places of the underlying gismu), the me=
anings of these are shown on one or two extra rows following the primary ro=
w for that cmavo.
+<p>
+<cx "modal cmavo, basis in gismu place structure">  XE "modal cmavo: basis=
 in gismu place structure"   It should be emphasized that the place structu=
res of the gismu control the meanings of the BAI cmavo. The English phrases=
 shown here are only suggestive, and are often too broad or too narrow to c=
orrectly specify what the acceptable range of uses for the modal tag are.
+<p>
+<pre> 	ba'i	basti	replaced by	instead of
+	bai	bapli	compelled by	compelling
+	bau	bangu	in language	in language of
+	be'i	benji	sent by	transmitting
+			te=3Dsent to	ve=3Dwith transmit origin
+			xe=3Dtransmitted via
+	ca'i	catni	by authority of	with authority over
+	cau	claxu	lacked by	without
+	ci'e	ciste	in system	with system function
+			te=3Dof system components
+	ci'o	cinmo	felt by	 feeling emotion
+	ci'u	ckilu	on the scale	on scale measuring
+	cu'u	cusku	as said by	expressing
+			te=3Das told to	ve=3Dexpressed in medium
+	de'i	detri	dated	on the same date as
+	di'o	diklo	at the locus of 	at specific locus
+	do'e	=97	vaguely related to
+	du'i	dunli	as much as	equal to
+	du'o	djuno	according to	knowing facts
+			te=3Dknowing about
+			ve=3Dunder epistemology
+	fa'e	fatne	reverse of	in reversal of
+	fau	fasnu	in the event of
+	fi'e	finti	created by	creating work
+			te=3Dcreated for purpose
+	ga'a	zgana	to observer	observing
+			te=3Dobserved by means
+			ve=3Dobserved under conditions
+	gau	gasnu	with agent	as agent in doing
+	ja'e	jalge	resulting in	results because of
+	ja'i	javni	by rule	by rule prescribing
+	ji'e	jimte	up to limit	as a limit of
+	ji'o	jitro	under direction	controlling
+	ji'u	jicmu	based on	supporting
+
+	ka'a	klama	gone to by	with destination
+			te=3Dwith origin	ve=3Dvia route
+			xe=3Dby transport mode
+	ka'i	krati	represented by	on behalf of
+	kai	ckaji	characterizing	with property
+	ki'i	ckini	as relation of	related to
+			te=3Dwith relation
+	ki'u	krinu	justified by	with justified result
+	koi	korbi	bounded by	as boundary of
+			te=3Dbordering
+	ku'u	kulnu	in culture	in culture of
+	la'u	klani	as quantity of	in quantity
+	le'a	klesi	in category	as category of
+			te=3Ddefined by quality
+	li'e	lidne	led by	 leading
+	ma'e	marji	of material	made from material
+			te=3Din material form of
+	ma'i	manri	in ref. frame	as a standard for
+	mau	zmadu	exceeded by	more than
+	me'a	mleca	undercut by	less than
+	me'e	cmene	with name	as a name for
+			te=3Das a name to
+	mu'i	mukti	motivated by	motive therefore
+	mu'u	mupli	exemplified by	as an example of
+	ni'i	nibli	entailed by	entails
+	pa'a	panra	in addition to	similar to
+			te=3Dsimilar in pattern
+			ve=3Dsimilar by standard
+	pa'u	pagbu	with component	as a part of
+	pi'o	pilno	used by	using tool
+	po'i	porsi	in the sequence	sequenced by rule
+	pu'a	pluka	pleased by	in order to please
+	pu'e	pruce	by process	processing from
+			te=3Dprocessing into
+			ve=3Dpassing through stages
+	ra'a	srana	pertained to by	concerning
+	ra'i	krasi	from source	as an origin of
+	rai	traji	with superl.	superlative in
+			te=3Dat extreme	ve=3Dsuperlative among
+	ri'a	rinka	caused by	causing
+	ri'i	lifri	experienced by	experiencing
+	sau	sarcu	requiring	necessarily for
+			te=3Dnecessarily under conditions
+	si'u	sidju	aided by	assisting in
+	ta'i	tadji	by method	as a method for
+	tai	tamsmi	as a form of	in form
+			te=3Din form similar to
+	ti'i	stidi	suggested by	suggesting
+			te=3Dsuggested to
+	ti'u	tcika	with time	at the time of
+	tu'i	stuzi	with site	as location of
+	va'o	vanbi	under conditions	as conditions for
+	va'u	xamgu	benefiting from	with beneficiary
+	zau	zanru	approved by	approving
+	zu'e	zukte	with actor	with means to goal
+			te=3Dwith goal
+</pre><p>The lujvo =93tamsmi=94 on which =93tai=94 is based is derived fro=
m the tanru =93tarmi simsa=94 and has the place structure:
+<p>
+<dl compact><dt><dd>tamsmi:  x1 has form x2, similar in form to x3 in prop=
erty/quality x4
+</dl><p>This lujvo is employed because =93tarmi=94 does not have a place s=
tructure useful for the modal's purpose.
+<p>
+</body></html>
+<img src=3Dchap10.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+
+<h2>Chapter 10
+<br>
+Imaginary Journeys: The Lojban Space/Time Tense System</h2>
+<h6>$Revision: 4.0 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>Introductory</h3>
+<p>
+<cx "tense, explanation of presentation method">  XE "tense: explanation o=
f presentation method"   This chapter attempts to document and explain the =
space/time tense system of Lojban. It does not attempt to answer all questi=
ons of the form =93How do I say such-and-such (an English tense) in Lojban?=
=94 Instead, it explores the Lojban tense system from the inside, attemptin=
g to educate the reader into a Lojbanic viewpoint. Once the overall system =
is understood and the resources that it makes available are familiar, the r=
eader should have some hope of using appropriate tense constructs and being=
 correctly understood.
+<p>
+<cx "tense, Lojban contrasted with native languages">  XE "tense: Lojban c=
ontrasted with native languages"   The system of Lojban tenses presented he=
re may seem really complex because of all the pieces and all the options; i=
ndeed, this chapter is the longest one in this book. But tense is in fact c=
omplex in every language. In your native language, the subtleties of tense =
are intuitive. In foreign languages, you are seldom taught the entire syste=
m until you have reached an advanced level. Lojban tenses are extremely sys=
tematic and productive, allowing you to express subtleties based on what th=
ey mean rather than on how they act similarly to English tenses. This chapt=
er concentrates on presenting an intuitive approach to the meaning of Lojba=
n tense words and how they may be creatively and productively combined.
+<p>
+<cx "temporal tense, historical definition">  XE "temporal tense: historic=
al definition"   <cx "temporal tense, real relationship to time in English"=
>  XE "temporal tense: real relationship to time in English"   <cx "tempora=
l tense, as mandatory in English">  XE "temporal tense: as mandatory in Eng=
lish"   <cx "temporal tense, Lojban contrasted with English in necessity"> =
 XE "temporal tense: Lojban contrasted with English in necessity"   What is=
 =93tense=94? Historically, =93tense=94 is the attribute of verbs in Englis=
h and related languages that expresses the time of the action. In English, =
three tenses are traditionally recognized, conventionally called the past, =
the present, and the future. There are also a variety of compound tenses us=
ed in English. However, there is no simple relationship between the form of=
 an English tense and the time actually expressed:
+<p>
+<dl compact><dt>	<dd>I go to London tomorrow.
+	I will go to London tomorrow.
+	I am going to London tomorrow.
+</dl>all mean the same thing, even though the first sentence uses the pres=
ent tense; the second, the future tense; and the third, a compound tense us=
ually called =93present progressive=94. Likewise, a newspaper headline says=
 =93JONES DIES=94, although it is obvious that the time referred to must be=
 in the past. Tense is a mandatory category of English: every sentence must=
 be marked for tense, even if in a way contrary to logic, because every mai=
n verb has a tense marker built into to it. By contrast, Lojban brivla have=
 no implicit tense marker attached to them.
+<p>
+<cx "tense, selbri types applicable to">  XE "tense: selbri types applicab=
le to"   <cx "tense system, and space location">  XE "tense system: and spa=
ce location"   <cx "tense system, and space location">  XE "tense system: a=
nd space location"   <cx "time, as part of tense system (see also tense, te=
mporal tense)">  XE "time: as part of tense system (see also tense, tempora=
l tense)"   <cx "space location, as part of tense system (see also tense, s=
patial tense)">  XE "space location: as part of tense system (see also tens=
e, spatial tense)"   <cx "elided tense, meaning of">  XE "elided tense: mea=
ning of"   In Lojban, the concept of tense extends to every selbri, not mer=
ely the verb-like ones. In addition, tense structures provide information a=
bout location in space as well as in time. All tense information is optiona=
l in Lojban: a sentence like:
+<p>
+<ex "go to market">  XE "go to market: example=93  <pre><a name=3De1d1>1.1=
)	mi klama le zarci
+	I go-to the market.
+</pre>can be understood as:
+
+<p>
+<dl compact><dt>	<dd>I went to the market.
+	I am going to the market.
+	I have gone to the market.
+	I will go to the market.
+	I continually go to the market.
+</dl>as well as many other possibilities: context resolves which is correc=
t.
+<p>
+<cx "tense, position of in sentence">  XE "tense: position of in sentence"=
   <lx "cu">  XE "cu"   <cx "tense specification, effect on =93cu=94">  XE =
"tense specification: effect on cu"   <cx "tense specification, effect on e=
lidability of terminators">  XE "tense specification: effect on elidability=
 of terminators"   <cx "cu, effect of tense specification">  XE "cu: effect=
 of tense specification"   The placement of a tense construct within a Lojb=
an bridi is easy: right before the selbri. It goes immediately after the =
=93cu=94, and can in fact always replace the =93cu=94 (although in very com=
plex sentences the rules for eliding terminators may be changed as a result=
). In the following examples, =93pu=94 is the tense marker for =93past time=
=94:
+<p>
+<pre><a name=3De1d2>1.2)	mi cu pu klama le zarci
+	mi pu klama le zarci
+	I in-the-past go-to the market.
+	I went to the market.
+</pre><cx "tense, position in sentence alternative">  XE "tense: position =
in sentence alternative"   <cx "tense, with ku">  XE "tense: with ku"   <lx=
 "ku">  XE "ku"   <cx "ku, with tense">  XE "ku: with tense"   It is also p=
ossible to put the tense somewhere else in the bridi by adding =93ku=94 aft=
er it. This =93ku=94 is an elidable terminator, but it's almost never possi=
ble to actually elide it except at the end of the bridi:
+<p>
+<pre><a name=3De1d3>1.3)	puku mi klama le zarci
+	In-the-past I go-to the market.
+	Earlier, I went to the market.
+<a name=3De1d4>1.4)	mi klama puku le zarci
+	I go-to in-the-past the market.
+	I went earlier to the market.
+<a name=3De1d5>1.5)	mi klama le zarci pu [ku]
+	I go-to the market in-the-past.
+	I went to the market earlier.
+</pre><cx "tense, effect of different position in sentence">  XE "tense: e=
ffect of different position in sentence"   <cx "tense, emphasizing by posit=
ion in sentence">  XE "tense: emphasizing by position in sentence"   <a hre=
f=3D#e1d{2}>Examples 1.2 </a>through <a href=3D#e1d5>1.5 </a>are different =
only in emphasis. Abnormal order, such as <a href=3D#e1d{3}>Examples 1.3 </=
a>through <a href=3D#e1d5>1.5 </a>exhibit, adds emphasis to the words that =
have been moved; in this case, the tense cmavo =93pu=94. Words at either en=
d of the sentence tend to be more noticeable.
+<p>
+<a name=3Ds2><h3>Spatial tenses: FAhA and VA</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	vi	VA	short distance
+	va	VA	medium distance
+	vu	VA	long distance
+	zu'a	FAhA	left
+	ri'u	FAhA	right
+	ga'u	FAhA	up
+	ni'a	FAhA	down
+	ca'u	FAhA	front
+	ne'i	FAhA	within
+	be'a	FAhA	north of
+</pre>(The complete list of FAhA cmavo can be found in <a href=3D#s28>Sect=
ion 28</a>.)
+<p>
+Why is this section about spatial tenses rather than the more familiar tim=
e tenses of <a href=3D#s1>Section 1</a>, asks the reader? Because the model=
 to be used in explaining both will be easier to grasp for space than for t=
ime. The explanation of time tenses will resume in <a href=3D#s4>Section 4<=
/a>.
+<p>
+<cx "spatial tense, as optional in English">  XE "spatial tense: as option=
al in English"   <cx "spatial tense, compared with temporal tense in elidab=
ility">  XE "spatial tense: compared with temporal tense in elidability"   =
<cx "temporal tense, compared with spatial tense in elidability">  XE "temp=
oral tense: compared with spatial tense in elidability"   <cx "temporal ten=
se elision, compared with spatial tense elision in meaning">  XE "temporal =
tense elision: compared with spatial tense elision in meaning"   English do=
esn't have mandatory spatial tenses. Although there are plenty of ways in E=
nglish of showing where an event happens, there is absolutely no need to do=
 so. Considering this fact may give the reader a feel for what the optional=
 Lojban time tenses are like. From the Lojban point of view, space and time=
 are interchangeable, although they are not treated identically.
+<p>
+<cx "spatial tense, definition">  XE "spatial tense: definition"   <cx "im=
aginary journey, and spatial tense">  XE "imaginary journey: and spatial te=
nse"   <cx "spatial tense, as an imaginary journey">  XE "spatial tense: as=
 an imaginary journey"   <cx "VA, and distance">  XE "VA selma=92o: and dis=
tance"   <cx "distance, specification with VA">  XE "distance: specificatio=
n with VA"   <lx "FAhA">  XE "FAhA selma=92o"   <lx "VA">  XE "VA selma=92o=
"   <cx "spatial tense, distance">  XE "spatial tense: distance"   Lojban s=
pecifies the spatial tense of a bridi (the place at which it occurs) by usi=
ng words from selma'o FAhA and VA to describe an imaginary journey from the=
 speaker to the place referred to. FAhA cmavo specify the direction taken i=
n the journey, whereas VA cmavo specify the distance gone. For example:
+<p>
+<cx "man biting dog">  XE "man biting dog"   <pre><a name=3De2d1>2.1)	le n=
anmu va batci le gerku
+	The man [medium distance] bites the dog.
+	Over there the man is biting the dog.
+</pre><cx "imaginary journey, starting point">  XE "imaginary journey: sta=
rting point"   <cx "imaginary journey, ending point">  XE "imaginary journe=
y: ending point"   <cx "spatial tense, referent of">  XE "spatial tense: re=
ferent of"   <cx "spatial tense, reference frame">  XE "spatial tense: refe=
rence frame"   <lx "va">  XE "va"   What is at a medium distance? The event=
 referred to by the bridi: the man biting the dog. What is this event at a =
medium distance from? The speaker's location. We can understand the =93va=
=94 as saying: =93If you want to get from the speaker's location to the loc=
ation of the bridi, journey for a medium distance (in some direction unspec=
ified).=94 This =93imaginary journey=94 can be used to understand not only =
<a href=3D#e2d1>Example 2.1</a>, but also every other spatial tense constru=
ct.
+<p>
+<cx "spatial tense, direction">  XE "spatial tense: direction"   <cx "FAhA=
, and direction">  XE "FAhA selma=92o: and direction"   <cx "direction, spe=
cification with FAhA">  XE "direction: specification with FAhA"   Suppose y=
ou specify a direction with a FAhA cmavo, rather than a distance with a VA =
cmavo:
+<p>
+<lx "zu'a">  XE "zu'a"   <pre><a name=3De2d2>2.2)	le nanmu zu'a batci le g=
erku
+	The man [left] bites the dog.
+</pre><p>Here the imaginary journey is again from the speaker's location t=
o the location of the bridi, but it is now performed by going to the left (=
in the speaker's reference frame) for an unspecified distance. So a reasona=
ble translation is:
+<p>
+<dl compact><dt>	<dd>To my left, the man bites the dog.
+</dl><p>The =93my=94 does not have an explicit equivalent in the Lojban, b=
ecause the speaker's location is understood as the starting point.
+<p>
+<cx "zu'a, derivation of word">  XE "zu'a: derivation of word"   <cx "VA, =
relation of words to ti, ta, tu">  XE "VA selma=92o: relation of words to t=
i, ta, tu"   <lx "zu'a">  XE "zu'a"   <lx "VA">  XE "VA selma=92o"   <lx "v=
i">  XE "vi"   <lx "va">  XE "va"   <lx "vu">  XE "vu"   (Etymologically, b=
y the way, =93zu'a=94 is derived from =93zunle=94, the gismu for =93left=94=
, whereas =93vi=94, =93va=94, and =93vu=94 are intended to be reminiscent o=
f =93ti=94, =93ta=94, and =93tu=94, the demonstrative pronouns =93this-here=
=94, =93that-there=94, and =93that-yonder=94.)
+<p>
+<cx "tense, order of direction specification in">  XE "tense: order of dir=
ection specification in"   <cx "tense, order of distance specification in">=
  XE "tense: order of distance specification in"   <cx "spatial tenses, ord=
er of direction and distance specifications">  XE "spatial tense: order of =
direction and distance specification"   <cx "direction, order of relative t=
o distance in spatial tenses">  XE "direction: order of relative to distanc=
e in spatial tenses"   <cx "distance, order of relative to direction in spa=
tial tenses">  XE "distance: order of relative to direction in spatial tens=
es"   What about specifying both a direction and a distance? The rule here =
is that the direction must come before the distance:
+<p>
+<pre><a name=3De2d3>2.3)	le nanmu zu'avi batci le gerku
+	The man [left] [short distance] bites the dog.
+	Slightly to my left, the man bites the dog.
+</pre><p>As explained in <a href=3D#s1>Section 1</a>, it would be perfectl=
y correct to use =93ku=94 to move this tense to the beginning or the end of=
 the sentence to emphasize it:
+<p>
+<pre><a name=3De2d4>2.4)	zu'aviku le nanmu cu batci le gerku
+	[Left] [short distance] the man bites the dog.
+	Slightly to my left, the man bites the dog.
+</pre><a name=3Ds3><h3>Compound spatial tenses</h3>
+<p>
+<cx "compound tense, definition">  XE "compound tense: definition"   Humph=
, says the reader: this talk of =93imaginary journeys=94 is all very well, =
but what's the point of it? =97 =93zu'a=94 means =93on the left=94 and =93v=
i=94 means =93nearby=94, and there's no more to be said. The imaginary-jour=
ney model becomes more useful when so-called compound tenses are involved. =
A compound tense is exactly like a simple tense, but has several FAhAs run =
together:
+<p>
+<ex "manhole">  XE "manhole: example=93  <pre><a name=3De3d1>3.1)	le nanmu=
 ga'u zu'a batci le gerku
+	The man [up] [left] bites the dog.
+</pre><cx "compound spatial tense, explanation of">  XE "compound spatial =
tense: explanation of"   <cx "imaginary journey, stages of in compound tens=
es">  XE "imaginary journey: stages of in compound tenses"   The proper int=
erpretation of <a href=3D#e3d1>Example 3.1 </a>is that the imaginary journe=
y has two stages: first move from the speaker's location upward, and then t=
o the left. A translation might read:
+<p>
+<dl compact><dt>	<dd>Left of a place above me, the man bites the dog.
+</dl>(Perhaps the speaker is at the bottom of a manhole, and the dog-bitin=
g is going on at the edge of the street.)
+<p>
+<cx "compound tense, Lojban contrasted with English in order of specificat=
ion">  XE "compound tense: Lojban contrasted with English in order of speci=
fication"   <cx "compound tense ordering, Lojban contrasted with English"> =
 XE "compound tense ordering: Lojban contrasted with English"   In the Engl=
ish translation, the keywords =93left=94 and =93above=94 occur in reverse o=
rder to the Lojban order. This effect is typical of what happens when we =
=93unfold=94 Lojban compound tenses into their English equivalents, and sho=
ws why it is not very useful to try to memorize a list of Lojban tense cons=
tructs and their colloquial English equivalents.
+<p>
+The opposite order also makes sense:
+<p>
+<pre><a name=3De3d2>3.2)	le nanmu zu'a ga'u batci le gerku
+	The man [left] [up] bites the dog.
+	Above a place to the left of me, the man bites the dog.
+</pre><cx "compound spatial tense, effect of different ordering">  XE "com=
pound spatial tense: effect of different ordering"   In ordinary space, the=
 result of going up and then to the left is the same as that of going left =
and then up, but such a simple relationship does not apply in all environme=
nts or to all directions: going south, then east, then north may return one=
 to the starting point, if that point is the North Pole.
+<p>
+<cx "compound spatial tense, with direction and distance">  XE "compound s=
patial tense: with direction and distance"   Each direction can have a dist=
ance following:
+<p>
+<pre><a name=3De3d3>3.3)	le nanmu zu'avi ga'uvu batci le gerku
+	The man [left] [short distance] [up] [long distance] bites the dog.
+	Far above a place slightly to the left of me, the man bites the dog.
+</pre><cx "compound spatial tense, beginning with distance only">  XE "com=
pound spatial tense: beginning with distance only"   A distance can also co=
me at the beginning of the tense construct, without any specified direction=
. (<a href=3D#e2d1>Example 2.1</a>, with VA alone, is really a special case=
 of this rule when no directions at all follow.)
+<p>
+<pre><a name=3De3d4>3.4)	le nanmu vi zu'a batci le gerku
+	The man [short distance] [left] bites the dog.
+	Left of a place near me, the man bites the dog.
+</pre><cx "compound spatial tense, as direction with-or-without distance">=
  XE "compound spatial tense: as direction with-or-without distance"   Any =
number of directions may be used in a compound tense, with or without speci=
fied distances for each:
+
+<p>
+<pre><a name=3De3d5>3.5)	le nanmu ca'uvi ni'ava ri'uvu ne'i
+		batci le gerku
+	The man [front] [short] [down] [medium] [right] [long] [within]
+		bites the dog.
+	Within a place a long distance to the right of a place which is a medium
+distance downward from a place a short distance in front of me,=20
+the man bites the dog.
+</pre><p>Whew! It's a good thing tense constructs are optional: having to =
say all that could certainly be painful. Note, however, how much shorter th=
e Lojban version of <a href=3D#e3d5>Example 3.5 </a>is than the English ver=
sion.
+<p>
+<a name=3Ds4><h3>Temporal tenses: PU and ZI</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	pu	PU	past
+	ca	PU	present
+	ba	PU	future
+	zi	ZI	short time distance
+	za	ZI	medium time distance
+	zu	ZI	long time distance
+</pre><cx "temporal tenses, compared with spatial tenses">  XE "temporal t=
enses: compared with spatial tenses"   <cx "PU, compared with FAhA">  XE "P=
U selma=92o: compared with FAhA"   <cx "ZI, compared with VA">  XE "ZI selm=
a=92o: compared with VA"   <lx "PU">  XE "PU selma=92o"   <lx "ZI">  XE "ZI=
 selma=92o"   <cx "tense, order of temporal and spatial in">  XE "tense: or=
der of temporal and spatial in"   <cx "tense, rationale for relative order =
of temporal and spatial in">  XE "tense: rationale for relative order of te=
mporal and spatial in"   <cx "spatial tense, order relative to temporal">  =
XE "spatial tense: order relative to temporal"   <cx "temporal tense, order=
 relative to spatial">  XE "temporal tense: order relative to spatial"   No=
w that the reader understands spatial tenses, there are only two main facts=
 to understand about temporal tenses: they work exactly like the spatial te=
nses, with selma'o PU and ZI standing in for FAhA and VA; and when both spa=
tial and temporal tense cmavo are given in a single tense construct, the te=
mporal tense is expressed first. (If space could be expressed before or aft=
er time, then certain constructions would be ambiguous.)
+<p>
+<pre><a name=3De4d1>4.1)	le nanmu pu batci le gerku
+	The man [past] bites the dog.
+	The man bit the dog.
+</pre>means that to reach the dog-biting, you must take an imaginary journ=
ey through time, moving towards the past an unspecified distance. (Of cours=
e, this journey is even more imaginary than the ones talked about in the pr=
evious sections, since time-travel is not an available option.)
+<p>
+<lx "pu">  XE "pu"   <lx "ca">  XE "ca"   <lx "ba">  XE "ba"   <cx "time, =
contrasted with space in number of directions">  XE "time: contrasted with =
space in number of directions"   <cx "space, contrasted with time in number=
 of directions">  XE "space: contrasted with time in number of directions" =
  Lojban recognizes three temporal directions: =93pu=94 for the past, =93ca=
=94 for the present, and =93ba=94 for the future. (Etymologically, these de=
rive from the corresponding gismu =93purci=94, =93cabna=94, and =93balvi=94=
. See <a href=3D#s23>Section 23 </a>for an explanation of the exact relatio=
nship between the cmavo and the gismu.) There are many more spatial directi=
ons, since there are FAhA cmavo for both absolute and relative directions a=
s well as =93direction-like relationships=94 like =93surrounding=94, =93wit=
hin=94, =93touching=94, etc. (See <a href=3D#s28>Section 28 </a>for a compl=
ete list.) But there are really only two directions in time: forward and ba=
ckward, toward the future and toward the past. Why, then, are there three c=
mavo of selma'o PU?
+<p>
+<cx "ca, rational for">  XE "ca: rational for"   <cx "ca, compared with bu=
'u">  XE "ca: compared with bu'u"   <cx "bu'u, compared with ca">  XE "bu'u=
: compared with ca"   <cx "tense, as subjective perception">  XE "tense: as=
 subjective perception"   <lx "ca">  XE "ca"   <lx "bu'u">  XE "bu'u"   The=
 reason is that tense is subjective: human beings perceive space and time i=
n a way that does not necessarily agree with objective measurements. We hav=
e a sense of =93now=94 which includes part of the objective past and part o=
f the objective future, and so we naturally segment the time line into thre=
e parts. The Lojban design recognizes this human reality by providing a sep=
arate time-direction cmavo for the =93zero direction=94, Similarly, there i=
s a FAhA cmavo for the zero space direction: =93bu'u=94, which means someth=
ing like =93coinciding=94.
+<p>
+<cx "relativity theory, relation to Lojban tense system">  XE "relativity =
theory: relation to Lojban tense system"   <cx "tense, as observer-based"> =
 XE "tense: as observer-based"   (Technical note for readers conversant wit=
h relativity theory: The Lojban time tenses reflect time as seen by the spe=
aker, who is assumed to be a =93point-like observer=94 in the relativistic =
sense: they do not say anything about physical relationships of relativisti=
c interval, still less about implicit causality. The nature of tense is not=
 only subjective but also observer-based.)
+<p>
+Here are some examples of temporal tenses:
+<p>
+<pre><a name=3De4d2>4.2)	le nanmu puzi batci le gerku
+	The man [past] [short distance] bites the dog.
+	A short time ago, the man bit the dog.
+<a name=3De4d3>4.3)	le nanmu pu pu batci le gerku
+	The man [past] [past] bites the dog.
+	Earlier than an earlier time than now, the man bit the dog.
+	The man had bitten the dog.
+	The man had been biting the dog.
+<a name=3De4d4>4.4)	le nanmu ba puzi batci le gerku
+	The man [future] [past] [short] bites the dog.
+	Shortly earlier than some time later than now, the man will bite the dog.
+	Soon before then, the man will have bitten the dog.
+	The man will have just bitten the dog.
+	The man will just have been biting the dog.
+</pre><cx "compound temporal tense, beginning with distance only">  XE "co=
mpound temporal tense: beginning with distance only"   What about the analo=
gue of an initial VA without a direction? Lojban does allow an initial ZI w=
ith or without following PUs:
+<p>
+<pre><a name=3De4d5>4.5)	le nanmu zi pu batci le gerku
+	The man [short] [past] bites the dog.
+	Before a short time from or before now, the man bit or will bite the dog.
+<a name=3De4d6>4.6)	le nanmu zu batci le gerku
+	The man [long] bites the dog.
+	A long time from or before now, the man will bite or bit the dog.
+</pre><cx "unspecified direction, temporal contrasted with in spatial">  X=
E "unspecified direction: temporal contrasted with in spatial"   <a href=3D=
#e4d5>Example 4.5 </a>and <a href=3D#e4d6>Example 4.6 </a>are perfectly leg=
itimate, but may not be very much used: =93zi=94 by itself signals an event=
 that happens at a time close to the present, but without saying whether it=
 is in the past or the future. A rough translation might be =93about now, b=
ut not exactly now=94.
+<p>
+<ex "nearby in time">  XE "nearby in time: example=93  Because we can move=
 in any direction in space, we are comfortable with the idea of events happ=
ening in an unspecified space direction (=93nearby=94 or =93far away=94), b=
ut we live only from past to future, and the idea of an event which happens=
 =93nearby in time=94 is a peculiar one. Lojban provides lots of such possi=
bilities that don't seem all that useful to English-speakers, even though y=
ou can put them together productively; this fact may be a limitation of Eng=
lish.
+<p>
+<cx "tense, with both temporal and spatial">  XE "tense: with both tempora=
l and spatial"   Finally, here are examples which combine temporal and spat=
ial tense:
+<p>
+<ex "long ago and far away">  XE "long ago and far away: example=93  <pre>=
<a name=3De4d7>4.7)	le nanmu puzu vu batci le gerku
+	The man [past] [long time] [long space] bites the dog.
+	Long ago and far away, the man bit the dog.
+</pre><p>Alternatively,
+<p>
+<pre><a name=3De4d8>4.8)	le nanmu batci le gerku puzuvuku
+	The man bites the dog [past] [long time] [long space].
+	The man bit the dog long ago and far away.
+</pre><a name=3Ds5><h3>Interval sizes: VEhA and ZEhA</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ve'i	VEhA	short space interval
+	ve'a	VEhA	medium space interval
+	ve'u	VEhA	long space interval
+	ze'i	ZEhA	short time interval
+	ze'a	ZEhA	medium time interval
+	ze'u	ZEhA	long time interval
+</pre><cx "tense, point contrasted with interval">  XE "tense: point contr=
asted with interval"   <cx "tense, interval contrasted with point">  XE "te=
nse: interval contrasted with point"   <lx "VEhA">  XE "VEhA selma=92o"   <=
lx "ZEhA">  XE "ZEhA selma=92o"   So far, we have considered only events th=
at are usually thought of as happening at a particular point in space and t=
ime: a man biting a dog at a specified place and time. But Lojbanic events =
may be much more =93spread out=94 than that: =93mi vasxu=94 (I breathe) is =
something which is true during the whole of my life from birth to death, an=
d over the entire part of the earth where I spend my life. The cmavo of VEh=
A (for space) and ZEhA (for time) can be added to any of the tense construc=
ts we have already studied to specify the size of the space or length of th=
e time over which the bridi is claimed to be true.
+<p>
+<ex "child on ice">  XE "child on ice: example=93  <pre><a name=3De5d1>5.1=
)	le verba ve'i cadzu le bisli
+	The child [small space interval] walks-on the ice.
+	In a small space, the child walks on the ice.
+	The child walks about a small area of the ice.
+</pre>means that her walking was done in a small area. Like the distances,=
 the interval sizes are classified only roughly as =93small, medium, large=
=94, and are relative to the context: a small part of a room might be a lar=
ge part of a table in that room.
+<p>
+Here is an example using a time interval:
+<p>
+<pre><a name=3De5d2>5.2)	le verba ze'a cadzu le bisli
+	The child [medium time interval] walks-on the ice.
+	For a medium time, the child walks/walked/will walk on the ice.
+</pre><cx "tense, order of direction, distance and interval in">  XE "tens=
e: order of direction, distance and interval in"   <cx "interval, relative =
order with direction and distance in tense">  XE "interval: relative order =
with direction and distance in tense"   Note that with no time direction wo=
rd, <a href=3D#e5d2>Example 5.2 </a>does not say when the walking happened:=
 that would be determined by context. It is possible to specify both direct=
ions or distances and an interval, in which case the interval always comes =
afterward:
+<p>
+<pre><a name=3De5d3>5.3)	le verba pu ze'a cadzu le bisli
+	The child [past] [medium time interval] walks-on the ice.
+	For a medium time, the child walked on the ice.
+	The child walked on the ice for a while.
+</pre><cx "interval, relation to point specified by direction and distance=
">  XE "interval: relation to point specified by direction and distance"   =
<cx "tense, relation of interval to point specified by direction and distan=
ce">  XE "tense: relation of interval to point specified by direction and d=
istance"   <cx "tense, relation of point specified by direction and distanc=
e to interval">  XE "tense: relation of point specified by direction and di=
stance to interval"   <cx "interval, specifying relation to point specified=
 by direction and distance">  XE "interval: specifying relation to point sp=
ecified by direction and distance"   <cx "tense, specifying relation of int=
erval to point specified by direction and distance">  XE "tense: specifying=
 relation of interval to point specified by direction and distance"   <cx "=
interval, followed by direction in tense construct">  XE "interval: followe=
d by direction in tense construct"   <cx "direction, following interval in =
tense construct">  XE "direction: following interval in tense construct"   =
<cx "ca, meaning when following interval specification">  XE "ca: meaning w=
hen following interval specification"   <cx "interval direction, specifying=
">  XE "interval direction: specifying"   In <a href=3D#e5d3>Example 5.3</a=
>, the relationship of the interval to the specified point in time or space=
 is indeterminate. Does the interval start at the point, end at the point, =
or is it centered on the point? By adding an additional direction cmavo aft=
er the interval, this question can be conclusively answered:
+<p>
+<pre><a name=3De5d4>5.4)	mi ca ze'ica cusku dei
+	I [present] [short time interval =96 present] express this-utterance.
+	I am now saying this sentence.
+</pre><cx "interval size, as context-dependent">  XE "interval size: as co=
ntext-dependent"   means that for an interval starting a short time in the =
past and extending to a short time in the future, I am expressing the utter=
ance which is <a href=3D#e5d4>Example 5.4</a>. Of course, =93short=94 is re=
lative, as always in tenses. Even a long sentence takes up only a short par=
t of a whole day; in a geological context, the era of Homo sapiens would on=
ly be a =93ze'i=94 interval.
+<p>
+By contrast,
+<p>
+<cx "pu, meaning when following interval specification">  XE "pu: meaning =
when following interval specification"   <pre><a name=3De5d5>5.5)	mi ca ze'=
ipu cusku dei
+	I [present] [short time interval =96 past] express this-utterance.
+	I have just been saying this sentence.
+</pre><cx "imaginary journey, with interval direction">  XE "imaginary jou=
rney: with interval direction"   means that for a short time interval exten=
ding from the past to the present I have been expressing <a href=3D#e5d5>Ex=
ample 5.5</a>. Here the imaginary journey starts at the present, lays down =
one end point of the interval, moves into the past, and lays down the other=
 endpoint. Another example:
+<p>
+<pre><a name=3De5d6>5.6)	mi pu ze'aba citka le mi sanmi
+	I [past] [medium time interval - future] eat my meal.
+	For a medium time afterward, I ate my meal.
+	I ate my meal for a while.
+</pre><p>With =93ca=94 instead of =93ba=94, <a href=3D#e5d6>Example 5.6 </=
a>becomes <a href=3D#e5d7>Example 5.7</a>,
+<p>
+<pre><a name=3De5d7>5.7)	mi pu ze'aca citka le mi sanmi
+	I [past] [medium time interval - present] eat my meal
+	For a medium time before and afterward, I ate my meal.
+	I ate my meal for a while.
+</pre>because the interval would then be centered on the past moment rathe=
r than oriented toward the future of that moment. The colloquial English tr=
anslations are the same =97 English is not well-suited to representing this=
 distinction.
+<p>
+Here are some examples of the use of space intervals with and without spec=
ified directions:
+<p>
+<ex "fish on right">  XE "fish on right: example=93  <pre><a name=3De5d8>5=
.8)	ta ri'u ve'i finpe
+	That-there [right] [short space interval] is-a-fish.
+	That thing on my right is a fish.
+</pre><p>In <a href=3D#e5d8>Example 5.8</a>, there is no equivalent in the=
 colloquial English translation of the =93small interval=94 which the fish =
occupies. Neither the Lojban nor the English expresses the orientation of t=
he fish. Compare <a href=3D#e5d9>Example 5.9</a>:
+<p>
+<pre><a name=3De5d9>5.9)	ta ri'u ve'ica'u finpe
+	That-there [right] [short space interval - front] is-a-fish.
+	That thing on my right extending forwards is a fish.
+</pre><p>Here the space interval occupied by the fish extends from a point=
 on my right to another point in front of the first point.
+<p>
+<a name=3Ds6><h3>Vague intervals and non-specific tenses</h3>
+<p>
+<cx "interval size, unspecified">  XE "interval size: unspecified"   <cx "=
interval size, vague">  XE "interval size: vague"   What is the significanc=
e of failing to specify an interval size of the type discussed in <a href=
=3D#s5>Section 5</a>? The Lojban rule is that if no interval size is given,=
 the size of the space or time interval is left vague by the speaker. For e=
xample:
+<p>
+<pre><a name=3De6d1>6.1)	mi pu klama le zarci
+	I [past] go-to the market.
+</pre>really means:
+<p>
+<dl compact><dt>	<dd>At a moment in the past, and possibly other moments a=
s well, the event
+	=93I went to the market=94 was in progress.
+</dl><cx "past event, possible extension into present">  XE "past event: p=
ossible extension into present"   <cx "tense direction, implications on sco=
pe of event">  XE "tense direction: implications on scope of event"   The v=
ague or unspecified interval contains an instant in the speaker's past. How=
ever, there is no indication whether or not the whole interval is in the sp=
eaker's past! It is entirely possible that the interval during which the go=
ing-to-the-market is happening stretches into the speaker's present or even=
 future.
+<p>
+<cx "tense, Lojban contrasted with English in implications of completeness=
">  XE "tense: Lojban contrasted with English in implications of completene=
ss"   <a href=3D#e6d1>Example 6.1 </a>points up a fundamental difference be=
tween Lojban tenses and English tenses. An English past-tense sentence like=
 =93I went to the market=94 generally signifies that the going-to-the-marke=
t is entirely in the past; that is, that the event is complete at the time =
of speaking. Lojban =93pu=94 has no such implication.
+<p>
+<cx "Classical Greek aorist tense, compared with Lojban tense">  XE "Class=
ical Greek aorist tense: compared with Lojban tense"   <cx "aorist, definit=
ion">  XE "aorist: definition"   <cx "tense, aorist">  XE "tense: aorist"  =
 This property of a past tense is sometimes called =93aorist=94, in referen=
ce to a similar concept in the tense system of Classical Greek. All of the =
Lojban tenses have the same property, however:
+<p>
+<cx "future event, possible extension into present">  XE "future event: po=
ssible extension into present"   <pre><a name=3De6d2>6.2)	le tricu ba crino
+	The tree [future] is-green.
+	The tree will be green.
+</pre>does not imply (as the colloquial English translation does) that the=
 tree is not green now. The vague interval throughout which the tree is, in=
 fact, green may have already started.
+<p>
+This general principle does not mean that Lojban has no way of indicating =
that a tree will be green but is not yet green. Indeed, there are several w=
ays of expressing that concept: see <a href=3D#s10>Section 10 </a>(event co=
ntours) and <a href=3D#s20>Section 20 </a>(logical connection between tense=
s).
+<p>
+<a name=3Ds7><h3>Dimensionality: VIhA</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	vi'i	VIhA	on a line
+	vi'a	VIhA	in an area
+	vi'u	VIhA	through a volume
+	vi'e	VIhA	throughout a space/time interval
+</pre><cx "spatial tense, contrasted with temporal in dimensionality">  XE=
 "spatial tense: contrasted with temporal in dimensionality"   <cx "spatial=
 tense, linear">  XE "spatial tense: linear"   <cx "spatial tense, planar">=
  XE "spatial tense: planar"   <cx "spatial tense, one-dimensional">  XE "s=
patial tense: one-dimensional"   <cx "spatial tense, two-dimensional">  XE =
"spatial tense: two-dimensional"   <cx "spatial tense, three-dimensional"> =
 XE "spatial tense: three-dimensional"   The cmavo of ZEhA are sufficient t=
o express time intervals. One fundamental difference between space and time=
, however, is that space is multi-dimensional. Sometimes we want to say not=
 only that something moves over a small interval, but also perhaps that it =
moves in a line. Lojban allows for this. I can specify that a motion =93in =
a small space=94 is more specifically =93in a short line=94, =93in a small =
area=94, or =93through a small volume=94.
+<p>
+<lx "VIhA">  XE "VIhA selma=92o"   What about the child walking on the ice=
 in <a href=3D#e5d{1}>Examples 5.1 </a>through <a href=3D#e5d3>5.3</a>? Giv=
en the nature of ice, probably the area interpretation is most sensible. I =
can make this assumption explicit with the appropriate member of selma'o VI=
hA:
+<p>
+<lx "vi'a">  XE "vi'a"   <pre><a name=3De7d1>7.1)	le verba ve'a vi'a cadzu=
 le bisli
+	The child [medium space interval] [2-dimensional] walks-on the ice.
+	In a medium-sized area, the child walks on the ice.
+</pre><cx "spatial tense intervals, order of VEhA and VIhA in">  XE "spati=
al tense interval: order of VEhA and VIhA in"   <cx "spatial tense interval=
s, order of size and dimensionality in">  XE "spatial tense interval: order=
 of size and dimensionality in"   <cx "dimensionality, order with size in s=
patial tense intervals">  XE "dimensionality: order with size in spatial te=
nse intervals"   <cx "size, order with dimensionality in spatial tense inte=
rvals">  XE "size: order with dimensionality in spatial tense intervals"   =
Space intervals can contain either VEhA or VIhA or both, but if both, VEhA =
must come first, as <a href=3D#e7d1>Example 7.1 </a>shows.
+<p>
+<cx "dimensionality of interval, as subjective">  XE "dimensionality of in=
terval: as subjective"   <cx "dimensionality, of walking">  XE "dimensional=
ity: of walking"   The reader may wish to raise a philosophical point here.=
 (Readers who don't wish to, should skip this paragraph.) The ice may be tw=
o-dimensional, or more accurately its surface may be, but since the child i=
s three-dimensional, her walking must also be. The subjective nature of Loj=
ban tense comes to the rescue here: the action is essentially planar, and t=
he third dimension of height is simply irrelevant to walking. Even walking =
on a mountain could be called =93vi'a=94, because relatively speaking the m=
ountain is associated with an essentially two-dimensional surface. Motion w=
hich is not confined to such a surface (e.g., flying, or walking through a =
three-dimensional network of tunnels, or climbing among mountains rather th=
an on a single mountain) would be properly described with =93vi'u=94. So th=
e cognitive, rather than the physical, dimensionality controls the choice o=
f VIhA cmavo.
+<p>
+<cx "Einsteinian, space-time intervals with 4 dimensions">  XE "Einsteinia=
n: space-time intervals with 4 dimensions"   <lx "vi'e">  XE "vi'e"   <cx "=
spatial tense, four-dimensional">  XE "spatial tense: four-dimensional"   <=
cx "tense, space-time dimension for intervals">  XE "tense: space-time dime=
nsion for intervals"   <cx "pastward, as a spatial tense">  XE "pastward: a=
s a spatial tense"   <cx "futureward, as a spatial tense">  XE "futureward:=
 as a spatial tense"   <cx "temporal tense, interaction with 4-dimensional =
spatial tense">  XE "temporal tense: interaction with 4-dimensional spatial=
 tense"   <cx "spatial tense, 4-dimensional interaction with temporal tense=
">  XE "spatial tense: 4-dimensional interaction with temporal tense"   VIh=
A has a member =93vi'e=94 which indicates a 4-dimensional interval, one tha=
t involves both space and time. This allows the spatial tenses to invade, t=
o some degree, the temporal tenses; it is possible to make statements about=
 space-time considered as an Einsteinian whole. (There are presently no cma=
vo of FAhA assigned to =93pastward=94 and =93futureward=94 considered as sp=
ace rather than time directions =97 they could be added, though, if Lojbani=
sts find space-time expression useful.) If a temporal tense cmavo is used i=
n the same tense construct with a =93vi'e=94 interval, the resulting tense =
may be self-contradictory.
+<p>
+<a name=3Ds8><h3>Movement in space: MOhI</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	mo'i	MOhI	movement flag
+</pre><cx "tense, static contrasted with moving">  XE "tense: static contr=
asted with moving"   <lx "mo'i">  XE "mo'i"   <lx "MOhI">  XE "MOhI selma=
=92o"   <cx "tense, expressing movement in">  XE "tense: expressing movemen=
t in"   <cx "direction, interaction with movement specification in tenses">=
  XE "direction: interaction with movement specification in tenses"   <cx "=
movement specification, interaction with direction in tenses">  XE "movemen=
t specification: interaction with direction in tenses"   All the informatio=
n carried by the tense constructs so far presented has been presumed to be =
static: the bridi is occurring somewhere or other in space and time, more o=
r less remote from the speaker. Suppose the truth of the bridi itself depen=
ds on the result of a movement, or represents an action being done while th=
e speaker is moving? This too can be represented by the tense system, using=
 the cmavo =93mo'i=94 (of selma'o MOhI) plus a spatial direction and option=
al distance; the direction now refers to a direction of motion rather than =
a static direction from the speaker.
+<p>
+<cx "toward right, contrasted with on right">  XE "toward right: contraste=
d with on right"   <cx "on right, contrasted with toward right">  XE "on ri=
ght: contrasted with toward right"   <ex "toward my right">  XE "toward my =
right: example=93  <pre><a name=3De8d1>8.1)	le verba mo'i ri'u cadzu le bis=
li
+	The child [movement] [right] walks-on the ice.
+	The child walks toward my right on the ice.
+</pre><p>This is quite different from:
+<p>
+<pre><a name=3De8d2>8.2)	le verba ri'u cadzu le bisli
+	The child [right] walks-on the ice.
+	To the right of me, the child walks on the ice.
+</pre><cx "reference frame for directions in tenses">  XE "reference frame=
 for directions in tenses"   <cx "direction, reference frame for">  XE "dir=
ection: reference frame for"   <cx "reference frame, specifying for directi=
on tenses">  XE "reference frame: specifying for direction tenses"   <lx "m=
a'i">  XE "ma'i"   <lx "BAI">  XE "BAI selma=92o"   In either case, however=
, the reference frame for defining =93right=94 and =93left=94 is the speake=
r's, not the child's. This can be changed thus:
+<p>
+<ex "toward her right">  XE "toward her right: example=93  <pre><a name=3D=
e8d3>8.3)	le verba mo'i ri'u cadzu le bisli
+		ma'i vo'a
+	The child [movement] [right] walks on the ice
+		in-reference-frame the-x1-place.
+	The child walks toward her right on the ice.
+</pre><a href=3D#e8d3>Example 8.3 </a>is analogous to <a href=3D#e8d1>Exam=
ple 8.1</a>. The cmavo =93ma'i=94 belongs to selma'o BAI (explained in <a h=
ref=3Dchap9.html>Chapter 9</a>), and allows specifying a reference frame.
+<p>
+<cx "movement, order in tense constructs">  XE "movement: order in tense c=
onstructs"   <cx "tense, order of movement specification in">  XE "tense: o=
rder of movement specification in"   Both a regular and a =93mo'i=94-flagge=
d spatial tense can be combined, with the =93mo'i=94 construct coming last:
+<p>
+<pre><a name=3De8d4>8.4)	le verba zu'avu mo'i ri'uvi cadzu le bisli
+	The child [left] [long] [movement] [right] [short] walks-on the ice.
+	Far to the left of me, the child walks a short distance toward my right=
=20
+on the ice.
+</pre><cx "complex movements, expressing">  XE "complex movements: express=
ing"   <cx "directions, multiple with movement">  XE "directions: multiple =
with movement"   <cx "movement, with multiple directions">  XE "movement: w=
ith multiple directions"   It is not grammatical to use multiple directions=
 like =93zu'a ca'u=94 after =93mo'i=94, but complex movements can be expres=
sed in a separate bridi.
+<p>
+Here is an example of a movement tense on a bridi not inherently involving=
 movement:
+<p>
+<ex "eat in airplane">  XE "eat in airplane: example=93  <pre><a name=3De8=
d5>8.5)	mi mo'i ca'uvu citka le mi sanmi
+	I [movement] [front] [long] eat my meal.
+	While moving a long way forward, I eat my meal.
+</pre>(Perhaps I am eating in an airplane.)
+<p>
+<cx "movement, time">  XE "movement: time"   <cx "time travel">  XE "time =
travel"   There is no parallel facility in Lojban at present for expressing=
 movement in time =97 time travel =97 but one could be added easily if it e=
ver becomes useful.
+<p>
+<a name=3Ds9><h3>Interval properties: TAhE and =93roi=94</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	di'i		TAhE	regularly
+	na'o		TAhE	typically
+	ru'i		TAhE	continuously
+	ta'e		TAhE	habitually
+	di'inai		TAhE	irregularly
+	na'onai	TAhE	atypically
+	ru'inai		TAhE	intermittently
+	ta'enai		TAhE	contrary to habit
+	roi		ROI	=93n=94 times
+	roinai		ROI	other than =93n=94 times
+	ze'e		ZEhA	whole time interval
+	ve'e		VEhA	whole space interval
+</pre><cx "continuous, of tense intervals">  XE "continuous: of tense inte=
rvals"   <cx "discrete, of tense intervals">  XE "discrete: of tense interv=
als"   <lx "TAhE">  XE "TAhE selma=92o"   <cx "intervals, spread of actions=
 over">  XE "interval: spread of actions over"   Consider Lojban bridi whic=
h express events taking place in time. Whether a very short interval (a poi=
nt) or a long interval of time is involved, the event may not be spread con=
sistently throughout that interval. Lojban can use the cmavo of selma'o TAh=
E to express the idea of continuous or non-continuous actions.
+<p>
+<ex "attend school">  XE "attend school: example=93  <pre><a name=3De9d1>9=
.1)	mi puzu ze'u velckule
+	I [past] [long distance] [long interval] am-a-school-attendee (pupil).
+	Long ago I attended school for a long time.
+</pre>probably does not mean that I attended school continuously throughou=
t the whole of that long-ago interval. Actually, I attended school every da=
y, except for school holidays. More explicitly,
+<p>
+<ex "regularly">  XE "regularly: example=93  <pre><a name=3De9d2>9.2)	mi p=
uzu ze'u di'i velckule
+	I [past] [long distance] [long interval] [regularly] am-a-pupil.
+	Long ago I regularly attended school for a long time.
+</pre><lx "ru'i">  XE "ru'i"   <lx "di'i">  XE "di'i"   <lx "na'o">  XE "n=
a'o"   <lx "ta'e">  XE "ta'e"   <cx "interval spread, mutually contrasted">=
  XE "interval spread: mutually contrasted"   The four TAhE cmavo are diffe=
rentiated as follows: =93ru'i=94 covers the entirety of the interval, =93di=
'i=94 covers the parts of the interval which are systematically spaced subi=
ntervals; =93na'o=94 covers part of the interval, but exactly which part is=
 determined by context; =93ta'e=94 covers part of the interval, selected wi=
th reference to the behavior of the actor (who often, but not always, appea=
rs in the x1 place of the bridi).
+<p>
+<cx "interval spread, with unspecified interval">  XE "interval spread: wi=
th unspecified interval"   Using TAhE does not require being so specific. E=
ither the time direction or the time interval or both may be omitted (in wh=
ich case they are vague). For example:
+<p>
+<pre><a name=3De9d3>9.3)	mi ba ta'e klama le zarci
+	I [future] [habitually] go-to the market.
+	I will habitually go to the market.
+	I will make a habit of going to the market.
+</pre>specifies the future, but the duration of the interval is indefinite=
. Similarly,
+<p>
+<pre><a name=3De9d4>9.4)	mi na'o klama le zarci
+	I [typically] go-to the market
+	I typically go/went/will go to the market
+</pre>illustrates an interval property in isolation. There are no distance=
 or direction cmavo, so the point of time is vague; likewise, there is no i=
nterval cmavo, so the length of the interval during which these goings-to-t=
he-market take place is also vague. As always, context will determine these=
 vague values.
+<p>
+<ex "intermittently">  XE "intermittently: example=93  <cx "interval sprea=
d, expressing =93intermittently=94">  XE "interval spread: expressing Engli=
sh intermittently"   <lx "nai">  XE "nai"   =93=93Intermittently=94 is the =
polar opposite notion to =93continuously=94, and is expressed not with its =
own cmavo, but by adding the negation suffix =93-nai=94 (which belongs to s=
elma'o NAI) to =93ru'i=94. For example:
+<p>
+<pre><a name=3De9d5>9.5)	le verba ru'inai cadzu le bisli
+	The child [continuously-not] walks-on the ice.
+	The child intermittently walks on the ice.
+</pre><cx "interval spread, negation with nai">  XE "interval spread: nega=
tion with nai"   As shown in the cmavo table above, all the cmavo of TAhE m=
ay be negated with =93-nai=94; =93ru'inai=94 and =93di'inai=94 are probably=
 the most useful.
+<p>
+<cx "quantified temporal tense, definition">  XE "quantified temporal tens=
e: definition"   <cx "tense, quantified">  XE "tense: quantified"   <ex "on=
ce">  XE "once: example=93  <lx "roi">  XE "roi"   <lx "ROI">  XE "ROI selm=
a=92o"   An intermittent event can also be specified by counting the number=
 of times during the interval that it takes place. The cmavo =93roi=94 (whi=
ch belongs to selma'o ROI) can be appended to a number to make a quantified=
 tense. Quantified tenses are common in English, but not so commonly named:=
 they are exemplified by the adverbs =93never=94, =93once=94, =93twice=94, =
=93thrice=94,=A0<dots>=85</dots>=93always=94, and by the related phrases =
=93many times=94, =93a few times=94, =93too many times=94, and so on. All o=
f these are handled in Lojban by a number plus =93-roi=94:
+
+<p>
+<pre><a name=3De9d6>9.6)	mi paroi klama le zarci
+	I [one time] go-to the market.
+	I go to the market once.
+<a name=3De9d7>9.7)	mi du'eroi klama le zarci
+	I [too-many times] go-to the market.
+	I go to the market too often.
+</pre><cx "temporal tense, quantified with direction">  XE "temporal tense=
: quantified with direction"   With the quantified tense alone, we don't kn=
ow whether the past, the present, or the future is intended, but of course =
the quantified tense need not stand alone:
+<p>
+<pre><a name=3De9d8>9.8)	mi pu reroi klama le zarci
+	I [past] [two times] go-to the market.
+	I went to the market twice.
+</pre><cx "quantified temporal tense with direction, Lojban contrasted wit=
h English in implications">  XE "quantified temporal tense with direction: =
Lojban contrasted with English in implications"   The English is slightly o=
ver-specific here: it entails that both goings-to- the-market were in the p=
ast, which may or may not be true in the Lojban sentence, since the implied=
 interval is vague. Therefore, the interval may start in the past but exten=
d into the present or even the future.
+<p>
+<cx "quantified temporal tense, negating with nai">  XE "quantified tempor=
al tense: negating with nai"   Adding =93-nai=94 to =93roi=94 is also permi=
tted, and has the meaning =93other than (the number specified)=94:
+<p>
+<ex "rat eats cheese">  XE "rat eats cheese: example=93  <pre><a name=3De9=
d9>9.9)	le ratcu reroinai citka le cirli
+	The rat [twice-not] eats the cheese.
+	The rat eats the cheese other than twice
+</pre><p>This may mean that the rat eats the cheese fewer times, or more t=
imes, or not at all.
+<p>
+<ex "only once">  XE "only once: example=93  <cx "quantified temporal tens=
es, caveat on implication of">  XE "quantified temporal tenses: caveat on i=
mplication of"   <cx "quantified temporal tenses, =93once=94 contrasted wit=
h =93only once=94">  XE "quantified temporal tenses: \=93once\=94 contraste=
d with \=93only once\=94"   <lx "ze'e">  XE "ze'e"   <cx "whole time interv=
al, expressing">  XE "whole time interval: expressing"   It is necessary to=
 be careful with sentences like <a href=3D#e9d6>Example 9.6 </a>and <a href=
=3D#e9d8>Example 9.8</a>, where a quantified tense appears without an inter=
val. What <a href=3D#e9d8>Example 9.8 </a>really says is that during an int=
erval of unspecified size, at least part of which was set in the past, the =
event of my going to the market happened twice. The example says nothing ab=
out what happened outside that vague time interval. This is often less than=
 we mean. If we want to nail down that I went to the market once and only o=
nce, we can use the cmavo =93ze'e=94 which represents the =93whole time int=
erval=94: conceptually, an interval which stretches from time's beginning t=
o its end:
+<p>
+<pre><a name=3De9d10>9.10)	mi ze'e paroi klama le zarci
+	I [whole interval] [once] go-to the market.
+</pre><p>Since specifying no ZEhA leaves the interval vague, <a href=3D#e9=
d8>Example 9.8 </a>might in appropriate context mean the same as <a href=3D=
#e9d10>Example 9.10 </a>after all =97 but <a href=3D#e9d10>Example 9.10 </a=
>allows us to be specific when specificity is necessary.
+<p>
+<lx "PU">  XE "PU selma=92o"   <cx "temporal direction, exception in meani=
ng when following ze'e">  XE "temporal direction: exception in meaning when=
 following ze'e"   <cx "ze'e, effect on following PU direction">  XE "ze'e:=
 effect on following PU direction"   <lx "ze'epu">  XE "ze'epu"   <lx "ze'e=
ba">  XE "ze'eba"   <lx "ze'eca">  XE "ze'eca"   <ex "have never">  XE "hav=
e never: example=93  <cx "ze'epu, meaning of">  XE "ze'epu: meaning of"   <=
cx "ze'eba, meaning of">  XE "ze'eba: meaning of"   <cx "ze'eca, meaning of=
">  XE "ze'eca: meaning of"   A PU cmavo following =93ze'e=94 has a slightl=
y different meaning from one that follows another ZEhA cmavo. The compound =
cmavo =93ze'epu=94 signifies the interval stretching from the infinite past=
 to the reference point (wherever the imaginary journey has taken you); =93=
ze'eba=94 is the interval stretching from the reference point to the infini=
te future. The remaining form, =93ze'eca=94, makes specific the =93whole of=
 time=94 interpretation just given. These compound forms make it possible t=
o assert that something has never happened without asserting that it never =
will.
+
+<p>
+<pre><a name=3De9d11>9.11)	mi ze'epu noroi klama le zarci
+	I [whole interval] [past] [never] go-to the market.
+	I have never gone to the market.
+</pre>says nothing about whether I might go in future.
+<p>
+<cx "quantified space">  XE "quantified space"   <lx "ve'e">  XE "ve'e"   =
The space equivalent of =93ze'e=94 is =93ve'e=94, and it can be used in the=
 same way with a quantified space tense: see <a href=3D#s11>Section 11 </a>=
for an explanation of space interval modifiers.
+<p>
+<a name=3Ds10><h3>Event contours: ZAhO and =93re'u=94</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<cx "inchoative event contour">  XE "inchoative event contour"   <cx "cont=
inuitive event contour">  XE "continuitive event contour"   <cx "perfective=
 event contour">  XE "perfective event contour"   <cx "initiative event con=
tour">  XE "initiative event contour"   <cx "cessitive event contour">  XE =
"cessitive event contour"   <cx "completitive event contour">  XE "completi=
tive event contour"   <cx "superfective event contour">  XE "superfective e=
vent contour"   <cx "achievative event contour">  XE "achievative event con=
tour"   <cx "pausative event contour">  XE "pausative event contour"   <cx =
"resumptive event contour">  XE "resumptive event contour"   <pre> 	pu'o	ZA=
hO	inchoative
+	ca'o	ZAhO	continuitive
+	ba'o	ZAhO	perfective
+	co'a	ZAhO	initiative
+	co'u	ZAhO	cessitive
+	mo'u	ZAhO	completitive
+	za'o	ZAhO	superfective
+	co'i	ZAhO	achievative
+	de'a	ZAhO	pausative
+	di'a	ZAhO	resumptive
+	re'u	ROI	ordinal tense
+</pre><cx "aspect, expressing">  XE "aspect: expressing"   <lx "ZAhO">  XE=
 "ZAhO selma=92o"   <cx "aspect, natural languages compared with respect to=
">  XE "aspect: natural languages compared with respect to"   The cmavo of =
selma'o ZAhO express the Lojban version of what is traditionally called =93=
aspect=94. This is not a notion well expressed by English tenses, but many =
languages (including Chinese and Russian among Lojban's six source language=
s) consider it more important than the specification of mere position in ti=
me.
+<p>
+<cx "event contours, definition">  XE "event contours: definition"   <cx "=
events, considered as a process">  XE "events: considered as a process"   <=
cx "event contours, syntax of">  XE "event contours: syntax of"   <cx "even=
t contours, order with respect to TAhE and ROI">  XE "event contours: order=
 with respect to TAhE and ROI"   The =93event contours=94 of selma'o ZAhO, =
with their bizarre keywords, represent the natural portions of an event con=
sidered as a process, an occurrence with an internal structure including a =
beginning, a middle, and an end. Since the keywords are scarcely self-expla=
natory, each ZAhO will be explained in detail here. Note that from the view=
point of Lojban syntax, ZAhOs are interval modifiers like TAhEs or ROI comp=
ounds; if both are found in a single tense, the TAhE/ROI comes first and th=
e ZAhO afterward. The imaginary journey described by other tense cmavo move=
s us to the portion of the event-as-process which the ZAhO specifies.
+<p>
+<cx "event contours, as timeless in perspective">  XE "event contours: as =
timeless in perspective"   <cx "PU tenses, contrasted with ZAhO tenses in v=
iewpoint">  XE "PU tenses: contrasted with ZAhO tenses in viewpoint"   <cx =
"tenses, viewpoint of PU contrasted with viewpoint of ZAhO">  XE =93tense: =
viewpoint of PU contrasted with viewpoint of ZAhO"   <cx "event-relative vi=
ewpoint, contrasted with speaker-relative viewpoint">  XE "event-relative v=
iewpoint: contrasted with speaker-relative viewpoint"   <cx "speaker-relati=
ve viewpoint, contrasted with event-relative viewpoint">  XE "speaker-relat=
ive viewpoint: contrasted with event-relative viewpoint"   <cx "event conto=
urs, as characteristic portions of events">  XE "event contours: as charact=
eristic portions of events"   It is important to understand that ZAhO cmavo=
, unlike the other tense cmavo, specify characteristic portions of the even=
t, and are seen from an essentially timeless perspective. The =93beginning=
=94 of an event is the same whether the event is in the speaker's present, =
past, or future. It is especially important not to confuse the speaker-rela=
tive viewpoint of the PU tenses with the event-relative viewpoint of the ZA=
hO tenses.
+<p>
+<lx "pu'o">  XE "pu'o"   <lx "ca'o">  XE "ca'o"   <lx "ba'o">  XE "ba'o"  =
 <cx "pu'o, derivation of word">  XE "pu'o: derivation of word"   <cx "ba'o=
, derivation of word">  XE "ba'o: derivation of word"   <cx "ca'o, derivati=
on of word">  XE "ca'o: derivation of word"   The cmavo =93pu'o=94, =93ca'o=
=94, and =93ba'o=94 (etymologically derived from the PU cmavo) refer to an =
event that has not yet begun, that is in progress, or that has ended, respe=
ctively:
+<p>
+<cx "event contour, inchoative">  XE "event contours: inchoative"   <ex "o=
n verge">  XE "on verge: example=93  <pre><a name=3De10d1>10.1)	mi pu'o dam=
ba
+	I [inchoative] fight.
+	I'm on the verge of fighting.
+</pre><ex "continues">  XE "continues: example=93  <cx "event contour, con=
tinuitive">  XE "event contours: continuitive"   <pre><a name=3De10d2>10.2)=
	la stiv. ca'o bacru
+	Steve [continuitive] utters.
+	Steve continues to talk.
+</pre><ex "finished">  XE "finished: example=93  <cx "event contour, perfe=
ctive">  XE "event contours: perfective"   <pre><a name=3De10d3>10.3)	le ve=
rba ba'o cadzu le bisli
+	The child [perfective] walks-on the ice.
+	The child is finished walking on the ice.
+</pre><cx "event contours, implications on scope of event">  XE "event con=
tours: implications on scope of event"   <cx "event contours, contrasted wi=
th tense direction in implication of extent">  XE "event contours: contrast=
ed with tense direction in implication of extent"   <cx "tense direction, c=
ontrasted with event contours in implication of extent">  XE "tense directi=
on: contrasted with event contours in implication of extent"   As discussed=
 in <a href=3D#s6>Section 6</a>, the simple PU cmavo make no assumptions ab=
out whether the scope of a past, present, or future event extends into one =
of the other tenses as well. <a href=3D#e10d{1}>Examples 10.1 </a>through <=
a href=3D#e10d3>10.3 </a>illustrate that these ZAhO cmavo do make such assu=
mptions possible: the event in 10.1 has not yet begun, definitively; likewi=
se, the event in 10.3 is definitely over.
+<p>
+<cx "pu'o, explanation of derivation">  XE "pu'o: explanation of derivatio=
n"   <cx "ba'o, explanation of derivation">  XE "ba'o: explanation of deriv=
ation"   <cx "pu'o, as pastward of event">  XE "pu'o: as pastward of event"=
   <cx "ba'o, as futureward of event">  XE "ba'o: as futureward of event"  =
 Note that in <a href=3D#e10d1>Example 10.1 </a>and <a href=3D#e10d3>Exampl=
e 10.3</a>, =93pu'o=94 and =93ba'o=94 may appear to be reversed: =93pu'o=94=
, although etymologically connected with =93pu=94, is referring to a future=
 event; whereas =93ba'o=94, connected with =93ba=94, is referring to a past=
 event. This is the natural result of the event-centered view of ZAhO cmavo=
. The inchoative, or =93pu'o=94, part of an event, is in the =93pastward=94=
 portion of that event, when seen from the perspective of the event itself.=
 It is only by inference that we suppose that <a href=3D#e10d1>Example 10.1=
 </a>refers to the speaker's future: in fact, no PU tense is given, so the =
inchoative part of the event need not be coincident with the speaker's pres=
ent: =93pu'o=94 is not necessarily, though in fact often is, the same as =
=93ca pu'o=94.
+<p>
+<cx "event contours, points associated with">  XE "event contours: points =
associated with"   <cx "event contours, division of the event into">  XE "e=
vent contours: division of the event into"   <cx "event contour, initiative=
">  XE "event contours: initiative"   <cx "event contour, cessative">  XE "=
event contours: cessative"   The cmavo in <a href=3D#e10d{1}>Examples 10.1 =
</a>through <a href=3D#e10d3>10.3 </a>refer to spans of time. There are als=
o two points of time that can be usefully associated with an event: the beg=
inning, marked by =93co'a=94, and the end, marked by =93co'u=94. Specifical=
ly, =93co'a=94 marks the boundary between the =93pu'o=94 and =93ca'o=94 par=
ts of an event, and =93co'u=94 marks the boundary between the =93ca'o=94 an=
d =93ba'o=94 parts:
+<p>
+<pre><a name=3De10d4>10.4)	mi ba co'a citka le mi sanmi
+	I [future] [initiative] eat my meal.
+	I will begin to eat my meal.
+<a name=3De10d5>10.5)	mi pu co'u citka le mi sanmi
+	I [past] [cessitive] eat my meal.
+	I ceased eating my meal.
+</pre><p>Compare <a href=3D#e10d4>Example 10.4 </a>with:
+<p>
+<pre><a name=3De10d6>10.6)	mi ba di'i co'a bajra
+	I [future] [regularly] [initiative] run.
+	I will regularly begin to run.
+</pre>which illustrates the combination of a TAhE with a ZAhO.
+<p>
+<cx "natural end, contrasted with actual stop">  XE "natural end: contrast=
ed with actual stop"   <cx "actual stop, contrasted with natural end">  XE =
"actual stop: contrasted with natural end"   <cx "finish, contrasted with s=
top">  XE "finish: contrasted with stop"   <cx "stop, contrasted with finis=
h">  XE "stop: contrasted with finish"   <cx "event contour, completitive">=
  XE "event contours: completitive"   A process can have two end points, on=
e reflecting the =93natural end=94 (when the process is complete) and the o=
ther reflecting the =93actual stopping point=94 (whether complete or not). =
<a href=3D#e10d5>Example 10.5 </a>may be contrasted with:
+<p>
+<pre><a name=3De10d7>10.7)	mi pu mo'u citka le mi sanmi
+	I [past] [completitive] eat my meal.
+	I finished eating my meal.
+</pre><p>In <a href=3D#e10d7>Example 10.7</a>, the meal has reached its na=
tural end; in <a href=3D#e10d5>Example 10.5</a>, the meal has merely ceased=
, without necessarily reaching its natural end.
+<p>
+<cx "stop, contrasted with pause">  XE "stop: contrasted with pause"   <cx=
 "pause, contrasted with stop">  XE "pause: contrasted with stop"   <cx "re=
sume, contrasted with begin">  XE "resume: contrasted with begin"   <cx "be=
gin, contrasted with resume">  XE "begin: contrasted with resume"   <cx "ev=
ent contours, interruption">  XE "event contours: interruption"   <cx "even=
t contours, resumption">  XE "event contours: resumption"   <lx "de'a">  XE=
 "de'a"   <lx "di'a">  XE "di'a"   <cx "event contour, pausative">  XE "eve=
nt contours: pausative"   <cx "event contour, resumptive">  XE "event conto=
urs: resumptive"   A process such as eating a meal does not necessarily pro=
ceed uninterrupted. If it is interrupted, there are two more relevant point=
 events: the point just before the interruption, marked by =93de'a=94, and =
the point just after the interruption, marked by =93di'a=94. Some examples:
+<p>
+<pre><a name=3De10d8>10.8)	mi pu de'a citka le mi sanmi
+	I [past] [pausative] eat my meal.
+	I stopped eating my meal (with the intention of resuming).
+<a name=3De10d9>10.9)	mi ba di'a citka le mi sanmi
+	I [future] [resumptive] eat my meal.
+	I will resume eating my meal.
+</pre><cx "natural end, continuing beyond">  XE "natural end: continuing b=
eyond"   <lx "za'o">  XE "za'o"   <cx "event contour, superfective">  XE "e=
vent contours: superfective"   In addition, it is possible for a process to=
 continue beyond its natural end. The span of time between the natural and =
the actual end points is represented by =93za'o=94:
+<p>
+<ex "keeps on">  XE "kept on too long: example=93  XE "too long:Example"  =
  <pre><a name=3De10d10>10.10)	le ctuca pu za'o ciksi le cmaci seldanfu=20
+le tadgri
+	The teacher [past] [superfective] explained the mathematics problem=20
+to the student-group.
+The teacher kept on explaining the mathematics problem to the class
+too long.
+That is, the teacher went on explaining after the class already understood=
 the problem.</pre>
+<p>
+<cx "event contour, achievative">  XE "event contours: achievative"   <cx =
"point, event considered as">  XE "point: event considered as"   <lx "co'i"=
>  XE "co'i"   An entire event can be treated as a single moment using the =
cmavo =93co'i=94:
+<p>
+<pre><a name=3De10d11>10.11)	la djan. pu co'i catra la djim
+	John [past] [achievative] kills Jim.
+	John was at the point in time where he killed Jim.
+</pre><cx "ordinal tense">  XE "ordinal tense"   <lx "re'u">  XE "re'u"   =
<lx "ROI">  XE "ROI selma=92o"   <cx "cycles">  XE "cycles"   Finally, sinc=
e an activity is cyclical, an individual cycle can be referred to using a n=
umber followed by =93re'u=94, which is the other cmavo of selma'o ROI:
+<p>
+<pre><a name=3De10d12>10.12)	mi pare'u klama le zarci
+	I [first time] go-to the store.
+	I go to the store for the first time (within a vague interval).
+</pre><p>Note the difference between:
+<p>
+<pre><a name=3De10d13>10.13)	mi pare'u paroi klama le zarci
+	I [first time] [one time] go-to the store.
+	For the first time, I go to the store once.
+</pre>and
+<p>
+<pre><a name=3De10d14>10.14)	mi paroi pare'u klama le zarci
+	I [one time] [first time] go-to the store.
+	There is one occasion on which I go to the store for the first time.
+</pre><a name=3Ds11><h3>Space interval modifiers: FEhE</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre>	fe'e	FEhE	space interval modifier flag
+</pre><cx "space intervals, compared with time intervals in continuity">  =
XE "space interval: compared with time intervals in continuity"   <lx "fe'e=
">  XE "fe'e"   <lx "FEhE">  XE "FEhE selma=92o"   <cx "spatial intervals, =
expressing degree of continuity over">  XE "spatial interval: expressing de=
gree of continuity over"   <cx "spatial interval modifiers, order in tense"=
>  XE "spatial interval modifiers: order in tense"   <cx "tense, order of s=
patial interval modifiers in ">  XE "tense: order of spatial interval modif=
iers in "   Like time intervals, space intervals can also be continuous, di=
scontinuous, or repetitive. Rather than having a whole separate set of selm=
a'o for space interval properties, we instead prefix the flag =93fe'e=94 to=
 the cmavo used for time interval properties. A space interval property wou=
ld be placed just after the space interval size and/or dimensionality cmavo=
:
+<p>
+<ex "sow grain">  XE "sow grain: example=93  <pre><a name=3De11d1>11.1)	ko=
 vi'i fe'e di'i sombo le gurni
+	You-imperative [1-dimensional] [space:] [regularly] sow the grain.
+	Sow the grain in a line and evenly!
+</pre><ex "salad ingredients">  XE "salad ingredients: example=93  <pre><a=
 name=3De11d2>11.2)	mi fe'e ciroi tervecnu lo selsalta
+	I [space:] [three places] buy those-which-are salad-ingredients.
+	I buy salad ingredients in three locations.
+</pre><ex "always and everywhere">  XE "always and everywhere: example=93 =
 <pre><a name=3De11d3>11.3)	ze'e roroi ve'e fe'e roroi ku
+		li re su'i re du li vo
+	[whole time] [all times] [whole space] [space:] [all places]
+		The-number 2 + 2 =3D the-number 4.
+	Always and everywhere, two plus two is four.
+</pre><p>As shown in <a href=3D#e11d3>Example 11.3</a>, when a tense comes=
 first in a bridi, rather than in its normal position before the selbri (in=
 this case =93du=94), it is emphasized.
+<p>
+<lx "fe'e">  XE "fe'e"   <lx "ZAhO">  XE "ZAhO selma=92o"   <cx "spatial c=
ontours, expressing">  XE "spatial contours: expressing"   <lx "be'a">  XE =
"be'a"   The =93fe'e=94 marker can also be used for the same purpose before=
 members of ZAhO. (The cmavo =93be'a=94 belongs to selma'o FAhA; it is the =
space direction meaning =93north of=94.)
+<p>
+<ex "south face">  XE "south face: example=93  <ex "rock face">  XE "rock =
face: example=93  <pre><a name=3De11d4>11.4)	tu ve'abe'a fe'e co'a rokci
+	That-yonder [medium space interval - north] [space] [initiative] is-a-roc=
k.
+	That is the beginning of a rock extending to my north.
+	That is the south face of a rock.
+</pre><cx "beginning point, spatial">  XE "beginning point: spatial"   <cx=
 "spatial contours, contrasted with temporal event contours">  XE "spatial =
contours: contrasted with temporal event contours"   <cx "event contours, t=
emporal contrasted with spatial">  XE "event contours: temporal contrasted =
with spatial"   Here the notion of a =93beginning point=94 represented by t=
he cmavo =93co'a=94 is transferred from =93beginning in time=94 to =93begin=
ning in space=94 under the influence of the =93fe'e=94 flag. Space is not i=
nherently oriented, unlike time, which flows from past to future: therefore=
, some indication of orientation is necessary, and the =93ve'abe'a=94 provi=
des an orientation in which the south face is the =93beginning=94 and the n=
orth face is the =93end=94, since the rock extends from south (near me) to =
north (away from me).
+<p>
+<cx "time, as space-based metaphor">  XE "time: as space-based metaphor"  =
 <cx "space, as time-based metaphor">  XE "space: as time-based metaphor"  =
 <cx "space/time metaphor, expressing direction mapping for">  XE "space/ti=
me metaphor: expressing direction mapping for"   <cx "FAhA, use in specifyi=
ng space/time mapping direction">  XE "FAhA selma=92o: use in specifying sp=
ace/time mapping direction"   Many natural languages represent time by a sp=
ace-based metaphor: in English, what is past is said to be =93behind us=94.=
 In other languages, the metaphor is reversed. Here, Lojban is representing=
 space (or space interval modifiers) by a time-based metaphor: the choice o=
f a FAhA cmavo following a VEhA cmavo indicates which direction is mapped o=
nto the future. (The choice of future rather than past is arbitrary, but co=
nvenient for English-speakers.)
+<p>
+<cx "TAhE, effect of ZAhO on fe'e flag">  XE "TAhE selma=92o: effect of ZA=
hO on fe'e flag"   <cx "ROI, effect of ZAhO on fe'e flag">  XE "ROI selma=
=92o: effect of ZAhO on fe'e flag"   <cx "ZAhO, effect on fe'e flag for TAh=
E and ROI">  XE "ZAhO selma=92o: effect on fe'e flag for TAhE and ROI"   <c=
x "fe'e, effect of TAhE/ROI with ZAhO on">  XE "fe'e: effect of TAhE/ROI wi=
th ZAhO on"   If both a TAhE (or ROI) and a ZAhO are present as space inter=
val modifiers, the =93fe'e=94 flag must be prefixed to each.
+<p>
+<a name=3Ds12><h3>Tenses as sumti tcita</h3>
+<p>
+<cx "tenses, use as sumti tcita">  XE =93tense: use as sumti tcita"   <cx =
"spatial information, adding to a sentence with tense sumti tcita">  XE "sp=
atial information: adding to a sentence with tense sumti tcita"   <cx "temp=
oral information, adding to a sentence with tense sumti tcita">  XE "tempor=
al information: adding to a sentence with tense sumti tcita"   <cx "sumti t=
cita, based on tenses">  XE "sumti tcita: based on tenses"   <cx "argument =
tags, based on tenses (see also sumti tcita)">  XE "argument tags: based on=
 tenses (see also sumti tcita)"   So far, we have seen tenses only just bef=
ore the selbri, or (equivalently in meaning) floating about the bridi with =
=93ku=94. There is another major use for tenses in Lojban: as sumti tcita, =
or argument tags. A tense may be used to add spatial or temporal informatio=
n to a bridi as, in effect, an additional place:
+
+<p>
+<pre><a name=3De12d1>12.1)	mi klama le zarci ca le nu do klama le zdani
+	I go-to the market [present] the event-of you go-to the house.
+	I go to the market when you go to the house.
+</pre><lx "ca">  XE "ca"   <cx "ca, meaning as a sumti tcita">  XE "ca: me=
aning as a sumti tcita"   <cx "tense direction, as sumti tcita">  XE "tense=
 direction: as sumti tcita"   <cx "sumti tcita, based on tense direction"> =
 XE "sumti tcita: based on tense direction"   Here =93ca=94 does not appear=
 before the selbri, nor with =93ku=94; instead, it governs the following su=
mti, the =93le nu=94 construct. What <a href=3D#e12d1>Example 12.1 </a>asse=
rts is that the action of the main bridi is happening at the same time as t=
he event mentioned by that sumti. So =93ca=94, which means =93now=94 when u=
sed with a selbri, means =93simultaneously-with=94 when used with a sumti. =
Consider another example:
+<p>
+<lx "pu">  XE "pu"   <cx "pu, meaning as a sumti tcita">  XE "pu: meaning =
as a sumti tcita"   <pre><a name=3De12d2>12.2)	mi klama le zarci pu le nu d=
o pu klama le zdani
+	I go-to the market [past] the event-of you [past] go-to the house.
+</pre><p>The second =93pu=94 is simply the past tense marker for the event=
 of your going to the house, and says that this event is in the speaker's p=
ast. How are we to understand the first =93pu=94, the sumti tcita?
+<p>
+<cx "imaginary journey, starting point">  XE "imaginary journey: starting =
point"   <cx "imaginary journey, starting at a different point">  XE "imagi=
nary journey: starting at a different point"   All of our imaginary journey=
s so far have started at the speaker's location in space and time. Now we a=
re specifying an imaginary journey that starts at a different location, nam=
ely at the event of your going to the house. <a href=3D#e12d2>Example 12.2 =
</a>then says that my going to the market is in the past, relative not to t=
he speaker's present moment, but instead relative to the moment when you we=
nt to the house. <a href=3D#e12d2>Example 12.2 </a>can therefore be transla=
ted:
+<p>
+<dl compact><dt>	<dd>I had gone to the market before you went to the house=
.
+</dl><cx "spatial tenses, as sumti tcita">  XE "spatial tense: as sumti tc=
ita"   <cx "tense distance, as sumti tcita">  XE "tense distance: as sumti =
tcita"   <cx "sumti tcita, based on tense distance">  XE "sumti tcita: base=
d on tense distance"   (Other translations are possible, depending on the e=
ver-present context.) Spatial direction and distance sumti tcita are exactl=
y analogous:
+<p>
+<ex "near the park">  XE "near the park: example=93  <ex "rat eats cheese"=
>  XE "rat eats cheese: example=93  <pre><a name=3De12d3>12.3)	le ratcu cu =
citka le cirla vi le panka
+	The rat eats the cheese [short distance] the park.
+	The rat eats the cheese near the park.
+<a name=3De12d4>12.4)	le ratcu cu citka le cirla vi le vu panka
+	The rat eats the cheese [short distance] the [long distance] park
+	The rat eats the cheese near the faraway park.
+<a name=3De12d5>12.5)	le ratcu cu citka le cirla vu le vi panka
+	The rat eats the cheese [long distance] the [short distance] park
+	The rat eats the cheese far away from the nearby park.
+</pre><cx "event contours, as sumti tcita">  XE "event contours: as sumti =
tcita"   <cx "sumti tcita, based on event contours">  XE "sumti tcita: base=
d on event contours"   <cx "spatial contours, as sumti tcita">  XE "spatial=
 contours: as sumti tcita"   <cx "sumti tcita, based on spatial contours"> =
 XE "sumti tcita: based on spatial contours"   <lx "fe'e">  XE "fe'e"   <cx=
 "ZAhO">  XE "ZAhO selma=92o"   <cx "event contours as sumti tcita, contras=
ted with direction and distance">  XE "event contours as sumti tcita: contr=
asted with direction and distance"   <cx "tense direction/distance as sumti=
 tcita, contrasted with event contours">  XE "tense direction/distance as s=
umti tcita: contrasted with event contours"   <cx "sumti tcita, event conto=
urs contrasted with direction/distance as basis for">  XE "sumti tcita: eve=
nt contours contrasted with direction/distance as basis for"   <cx "sumti t=
cita based on event contours, relation of main bridi to sumti process in"> =
 XE "sumti tcita based on event contours: relation of main bridi to sumti p=
rocess in"   The event contours of selma'o ZAhO (and their space equivalent=
s, prefixed with =93fe'e=94) are also useful as sumti tcita. The interpreta=
tion of ZAhO tcita differs from that of FAhA, VA, PU, and ZI tcita, however=
. The event described in the sumti is viewed as a process, and the action o=
f the main bridi occurs at the phase of the process which the ZAhO specifie=
s, or at least some part of that phase. The action of the main bridi itself=
 is seen as a point event, so that there is no issue about which phase of t=
he main bridi is intended. For example:
+<p>
+<ex "die after living">  XE "die after living: example=93  <ex "in the aft=
ermath">  XE "in the aftermath: example=93  <pre><a name=3De12d6>12.6)	mi m=
orsi ba'o le nu mi jmive
+	I am-dead [perfective] the event-of I live.
+	I die in the aftermath of my living.
+</pre><p>Here the (point-)event of my being dead is the portion of my livi=
ng-process which occurs after the process is complete. Contrast <a href=3D#=
e12d6>Example 12.6 </a>with:
+
+<p>
+<pre><a name=3De12d7>12.7)	mi morsi ba le nu mi jmive
+	I am-dead [future] the event-of I live.
+</pre><p>As explained in <a href=3D#s6>Section 6</a>, <a href=3D#e12d7>Exa=
mple 12.7 </a>does not exclude the possibility that I died before I ceased =
to live!
+<p>
+Likewise, we might say:
+<p>
+<pre><a name=3De12d8>12.8)	mi klama le zarci pu'o le nu mi citka
+	I go-to the store [inchoative] the event-of I eat
+</pre>which indicates that before my eating begins, I go to the store, whe=
reas
+<p>
+<pre><a name=3De12d9>12.9)	mi klama le zarci ba'o le nu mi citka
+	I go-to the store [perfective] the event-of I eat
+</pre>would indicate that I go to the store after I am finished eating.
+<p>
+Here is an example which mixes temporal ZAhO (as a tense) and spatial ZAhO=
 (as a sumti tcita):
+<p>
+<ex "boat sailed">  XE "boat sailed: example=93  <ex "too long">  XE "too =
long: example=93  <pre><a name=3De12d10>12.10)	le bloti pu za'o xelklama=20
+fe'e ba'o le lalxu
+	The boat [past] [superfective] is-a-transport-mechanism
+		[space] [perfective] the lake.
+	The boat sailed for too long and beyond the lake.
+</pre><p>Probably it sailed up onto the dock. One point of clarification: =
although =93xelklama=94 appears to mean simply =93is-a-mode-of-transport=94=
, it does not =96 the bridi of <a href=3D#e12d10>Example 12.10 </a>has four=
 omitted arguments, and thus has the (physical) journey which goes on too l=
ong as part of its meaning.
+<p>
+<cx "interval size, meaning as sumti tcita">  XE "interval size: meaning a=
s sumti tcita"   <cx "sumti tcita based on interval size">  XE "sumti tcita=
 based on interval size"   <cx "dimension, meaning as sumti tcita">  XE "di=
mension: meaning as sumti tcita"   <cx "sumti tcita based on dimension">  X=
E "sumti tcita based on dimension"   <cx "interval continuousness, meaning =
as sumti tcita">  XE "interval continuousness: meaning as sumti tcita"   <c=
x "sumti tcita based on interval continuousness">  XE "sumti tcita based on=
 interval continuousness"   <cx "interval properties, meaning as sumti tcit=
a">  XE "interval properties: meaning as sumti tcita"   <cx "sumti tcita ba=
sed on interval properties">  XE "sumti tcita based on interval properties"=
   <cx "quantified tenses, as sumti tcita">  XE "quantified tenses: as sumt=
i tcita"   <cx "sumti tcita based on quantified tenses">  XE "sumti tcita b=
ased on quantified tenses"   The remaining tense cmavo, which have to do wi=
th interval size, dimension, and continuousness (or lack thereof) are inter=
preted to let the sumti specify the particular interval over which the main=
 bridi operates:
+<p>
+<ex "twice today">  XE "twice today: example=93  <pre><a name=3De12d11>12.=
11)	mi klama le zarci reroi le ca djedi
+	I go-to the market [twice] the [present] day
+	I go/went/will go to the market twice today.
+</pre><cx "tense as sumti tcita, contrasted with tense inside sumti">  XE =
"tense as sumti tcita: contrasted with tense inside sumti"   <cx "tense ins=
ide sumti, contrasted with tense as sumti tcita">  XE "tense inside sumti: =
contrasted with tense as sumti tcita"   Be careful not to confuse a tense u=
sed as a sumti tcita with a tense used within a seltcita sumti:
+<p>
+<ex "snow falls">  XE "snow falls: example=93  <pre><a name=3De12d12>12.12=
)	loi snime cu carvi ze'u le ca dunra
+	Some-of-the-mass-of snow rains [long time interval] the [present] winter.
+	Snow falls during this winter.
+</pre>claims that the interval specified by =93this winter=94 is long, as =
events of snowfall go, whereas
+<p>
+<pre><a name=3De12d13>12.13)	loi snime cu carvi ca le ze'u dunra
+	Some-of-the-mass-of snow rains [present] the [long time] winter.
+	Snow falls in the long winter.
+</pre>claims that during some part of the winter, which is long as winters=
 go, snow falls.
+
+<p>
+<a name=3Ds13><h3>Sticky and multiple tenses: KI</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	ki	KI	sticky tense set/reset
+</pre><cx "tense, scope of">  XE "tense: scope of"   <cx "sticky tenses, d=
efinition">  XE "sticky tenses: definition"   <cx "sticky tense, effect on =
future tense meaning">  XE "sticky tenses: effect on future tense meaning" =
  <cx "imaginary journey origin, with sticky tenses">  XE "imaginary journe=
y origin: with sticky tenses"   So far we have only considered tenses in is=
olated bridi. Lojban provides several ways for a tense to continue in effec=
t over more than a single bridi. This property is known as =93stickiness=94=
: the tense gets =93stuck=94 and remains in effect until explicitly =93unst=
uck=94. In the metaphor of the imaginary journey, the place and time set by=
 a sticky tense may be thought of as a campsite or way-station: it provides=
 a permanent origin with respect to which other tenses are understood. Late=
r imaginary journeys start from that point rather than from the speaker.
+<p>
+<lx "ki">  XE "ki"   <lx "KI">  XE "KI selma=92o"   To make a tense sticky=
, suffix =93ki=94 to it:
+<p>
+<pre><a name=3De13d1>13.1)	mi puki klama le zarci  .i le nanmu cu batci le=
 gerku
+	I [past] [sticky] go-to the market.  The man bites the dog.
+	I went to the market. The man bit the dog.
+</pre><p>Here the use of =93puki=94 rather than just =93pu=94 ensures that=
 the tense will affect the next sentence as well. Otherwise, since the seco=
nd sentence is tenseless, there would be no way of determining its tense; t=
he event of the second sentence might happen before, after, or simultaneous=
ly with that of the first sentence.
+<p>
+(The last statement does not apply when the two sentences form part of a n=
arrative. See <a href=3D#s14>Section 14 </a>for an explanation of =93story =
time=94, which employs a different set of conventions.)
+<p>
+What if the second sentence has a tense anyway?
+<p>
+<cx "tense, effect of sticky tense on">  XE "tense: effect of sticky tense=
 on"   <ex "had earlier">  XE "had earlier: example=93  <pre><a name=3De13d=
2>13.2)	mi puki klama le zarci  .i le nanmu pu batci le gerku
+	I [past] sticky go-to the market.  The man [past] bites the dog.
+</pre><p>Here the second =93pu=94 does not replace the sticky tense, but a=
dds to it, in the sense that the starting point of its imaginary journey is=
 taken to be the previously set sticky time. So the translation of <a href=
=3D#e13d2>Example 13.2 </a>is:
+<p>
+<pre><a name=3De13d3>13.3)	I went to the market.  The man had earlier bitt=
en the dog.
+</pre><cx "tense in scope of sticky tense, compared with compound tense"> =
 XE "tense in scope of sticky tense: compared with compound tense"   <cx "c=
ompound tense, compared with tense in scope of sticky tense">  XE "compound=
 tense: compared with tense in scope of sticky tense"   and it is equivalen=
t in meaning (when considered in isolation from any other sentences) to:
+<p>
+<pre><a name=3De13d4>13.4)	mi pu klama le zarci  .i le nanmu pupu batci le=
 gerku
+	I [past] go-to the market.  The man [past] [past] bites the dog.
+</pre><cx "tenses, multiple in sentence">  XE =93tense: multiple in senten=
ce"   <cx "tenses, multiple in sentence compared with compound tense">  XE =
=93tense: multiple in sentence compared with compound tense"   <cx "compoun=
d tense, compared with multiple tenses in sentence">  XE "compound tense: c=
ompared with multiple tenses in sentence"   The point has not been discusse=
d so far, but it is perfectly grammatical to have more than one tense const=
ruct in a sentence:
+<p>
+<pre><a name=3De13d5>13.5)	puku mi ba klama le zarci
+	[past] I [future] go-to the market.
+	Earlier, I was going to go to the market.
+</pre><p>Here there are two tenses in the same bridi, the first floating f=
ree and specified by =93puku=94, the second in the usual place and specifie=
d by =93ba=94. They are considered cumulative in the same way as the two te=
nses in separate sentences of <a href=3D#e13d4>Example 13.4</a>. <a href=3D=
#e13d5>Example 13.5 </a>is therefore equivalent in meaning, except for emph=
asis, to:
+<p>
+<pre><a name=3De13d6>13.6)	mi puba klama le zarci
+	I [past] [future] go-to the market.
+	I was going to go to the market.
+</pre><cx "multiple tenses, effect of order in sentence">  XE "multiple te=
nses: effect of order in sentence"   Compare <a href=3D#e13d7>Example 13.7 =
</a>and <a href=3D#e13d8>Example 13.8</a>, which have a different meaning f=
rom <a href=3D#e13d5>Example 13.5 </a>and <a href=3D#e13d6>Example 13.6</a>=
:
+<p>
+<pre><a name=3De13d7>13.7)	mi ba klama le zarci puku
+	I [future] go-to the market [past].
+	I will have gone to the market earlier.
+<a name=3De13d8>13.8)	mi bapu klama le zarci
+	I [future] [past] go-to the market.
+	I will have gone to the market.
+</pre><p>So when multiple tense constructs in a single bridi are involved,=
 order counts =97 the tenses cannot be shifted around as freely as if there=
 were only one tense to worry about.
+<p>
+<cx "sticky tenses, from part of a multiple tense">  XE "sticky tenses: fr=
om part of a multiple tense"   But why bother to allow multiple tense const=
ructs at all? They specify separate portions of the imaginary journey, and =
can be useful in order to make part of a tense sticky. Consider <a href=3D#=
e13d9>Example 13.9</a>, which adds a second bridi and a =93ki=94 to <a href=
=3D#e13d5>Example 13.5</a>:
+<p>
+<dl compact><dt>13.9)	<dd>pukiku mi ba klama le zarci  .i le nanmu cu batc=
i le gerku
+	[past] [sticky] I [future] go-to the market.  The man bites the dog.
+</dl><p>What is the implied tense of the second sentence? Not =93puba=94, =
but only =93pu=94, since only =93pu=94 was made sticky with =93ki=94. So th=
e translation is:
+<p>
+<dl compact><dt>	<dd>I was going to go to the market.  The man bit the dog=
.
+</dl><cx "tense, on embedded bridi">  XE "tense: on embedded bridi"   <cx =
"tense on main bridi, effect on embedded bridi tenses">  XE "tense on main =
bridi: effect on embedded bridi tenses"   <cx "tense on main bridi, effect =
on embedded sumti with tenses">  XE "tense on main bridi: effect on embedde=
d sumti with tenses"   <cx "embedded bridi tenses, effect of main bridi ten=
se on">  XE "embedded bridi tenses: effect of main bridi tense on"   <cx "s=
umti with tense, effect of main bridi tense on">  XE "sumti with tense: eff=
ect of main bridi tense on"   Lojban has several ways of embedding a bridi =
within another bridi: descriptions, abstractors, relative clauses. (Technic=
ally, descriptions contain selbri rather than bridi.) Any of the selbri of =
these subordinate bridi may have tenses attached. These tenses are interpre=
ted relative to the tense of the main bridi:
+<p>
+<ex "former market">  XE "former market: example=93  <pre><a name=3De13d10=
>13.10)	mi pu klama le ba'o zarci
+	I [past] go-to the [perfective] market
+	I went to the former market.
+</pre><p>The significance of the =93ba'o=94 in <a href=3D#e13d10>Example 1=
3.10 </a>is that the speaker's destination is described as being =93in the =
aftermath of being a market=94; that is, it is a market no longer. In parti=
cular, the time at which it was no longer a market is in the speaker's past=
, because the =93ba'o=94 is interpreted relative to the =93pu=94 tense of t=
he main bridi.
+<p>
+Here is an example involving an abstraction bridi:
+<p>
+<pre><a name=3De13d11>13.11)	mi ca jinvi le du'u mi ba morsi
+	I now opine the fact-that I will-be dead.
+	I now believe that I will be dead.
+</pre><p>Here the event of being dead is said to be in the future with res=
pect to the opinion, which is in the present.
+<p>
+<cx "sticky tense, canceling">  XE "sticky tenses: canceling"   <cx "ki, w=
ith no tense">  XE "ki: with no tense"   =93ki=94 may also be used as a ten=
se by itself. This cancels all stickiness and returns the bridi and all fol=
lowing bridi to the speaker's location in both space and time.
+
+<p>
+</pre><cx "subscript, for sticky tense">  XE =93subscripts: for sticky ten=
se"   <cx "tense, subscripting">  XE "tense: subscripting"   <cx "tense, ha=
ndling multiple episodes">  XE "tense: handling multiple episodes"   In com=
plex descriptions, multiple tenses may be saved and then used by adding a s=
ubscript to =93ki=94. A time made sticky with =93kixipa=94 (ki-sub-1) can b=
e returned to by specifying =93kixipa=94 as a tense by itself. In the case =
of written expression, the writer's here-and-now is often different from th=
e reader's, and a pair of subscripted =93ki=94 tenses could be used to dist=
inguish the two.
+<p>
+<a name=3Ds14><h3>Story time</h3>
+<p>
+<cx "story time, definition">  XE "story time: definition"   <cx "story ti=
me, rationale for">  XE "story time: rationale for"   <cx "stories, flow of=
 time in">  XE "stories: flow of time in"   Making strict use of the conven=
tions explained in <a href=3D#s13>Section 13 </a>would be intolerably awkwa=
rd when a story is being told. The time at which a story is told by the nar=
rator is usually unimportant to the story. What matters is the flow of time=
 within the story itself. The term =93story=94 in this section refers to an=
y series of statements related in more-or-less time-sequential order, not j=
ust a fictional one.
+<p>
+<cx "story time, as a convention for inferring tense">  XE "story time: as=
 a convention for inferring tense"   <cx "tenseless sentences in story time=
">  XE "tenseless sentences in story time"   <cx "story time, tenseless sen=
tences in">  XE "story time: tenseless sentences in"   Lojban speakers use =
a different set of conventions, commonly called =93story time=94, for infer=
ring tense within a story. It is presumed that the event described by each =
sentence takes place some time more or less after the previous ones. Theref=
ore, tenseless sentences are implicitly tensed as =93what happens next=94. =
In particular, any sticky time setting is advanced by each sentence.
+<p>
+The following mini-story illustrates the important features of story time.=
 A sentence-by-sentence explication follows:
+<p>
+<ex "cave">  XE "cave: example=93  <pre><a name=3De14d1>14.1)	puzuki ku ne=
'iki le kevna
+		le ninmu goi ko'a zutse le rokci
+	[past] [long] [sticky] [,] [inside] [sticky] the cave,
+		the woman defined-as she-1 sat-on the rock
+	Long ago, in a cave, a woman sat on a rock.
+<a name=3De14d2>14.2)	.i ko'a citka loi kanba rectu
+	She-1 [tenseless] eat some-of-the-mass-of goat flesh.
+	She was eating goat's meat.
+<a name=3De14d3>14.3)	.i ko'a pu jukpa ri le mudyfagri
+	She [past] cook the-last-mentioned by-method the wood-fire.
+	She had cooked the meat over a wood fire.
+<a name=3De14d4>14.4)	.i lei rectu cu zanglare
+	The-mass-of flesh is-(favorable)-warm.
+	The meat was pleasantly warm.
+<a name=3De14d5>14.5)	.i le labno goi ko'e bazaki nenri klama le kevna
+	The wolf defined-as it-2 [future] [medium] [sticky] within-came to-the ca=
ve.
+	A while later, a wolf came into the cave.
+<a name=3De14d6>14.6)	.i ko'e lebna lei rectu ko'a
+	It-2 [tenseless] takes the-mass-of flesh from-her-1.
+	It took the meat from her.
+<a name=3De14d7>14.7)	.i ko'e bartu klama
+It-2 out ran
+	It ran out.
+</pre><cx "story tense, Lojban convention contrasted with English conventi=
on">  XE "story tense: Lojban convention contrasted with English convention=
"   <a href=3D#e14d1>Example 14.1 </a>sets both the time (long ago) and the=
 place (in a cave) using =93ki=94, just like the sentence sequences in <a h=
ref=3D#s13>Section 13</a>. No further space cmavo are used in the rest of t=
he story, so the place is assumed to remain unchanged. The English translat=
ion of <a href=3D#e14d1>Example 14.1 </a>is marked for past tense also, as =
the conventions of English storytelling require: consequently, all other En=
glish translation sentences are also in the past tense. (We don't notice ho=
w strange this is; even stories about the future are written in past tense!=
) This conventional use of past tense is not used in Lojban narratives.
+<p>
+<a href=3D#e14d2>Example 14.2 </a>is tenseless. Outside story time, it wou=
ld be assumed that its event happens simultaneously with that of <a href=3D=
#e14d1>Example 14.1</a>, since a sticky tense is in effect; the rules of st=
ory time, however, imply that the event occurs afterwards, and that the sto=
ry time has advanced (changing the sticky time set in <a href=3D#e14d1>Exam=
ple 14.1</a>).
+<p>
+<a href=3D#e14d3>Example 14.3 </a>has an explicit tense. This is taken rel=
ative to the latest setting of the sticky time; therefore, the event of <a =
href=3D#e14d3>Example 14.3 </a>happens before that of <a href=3D#e14d2>Exam=
ple 14.2</a>. It cannot be determined if <a href=3D#e14d3>Example 14.3 </a>=
happens before or after <a href=3D#e14d1>Example 14.1</a>.
+<p>
+<ex "flashbacks in story time">  XE "flashbacks in story time: example=93 =
 <a href=3D#e14d4>Example 14.4 </a>is again tenseless. Story time was not c=
hanged by the flashback in <a href=3D#e14d3>Example 14.3</a>, so <a href=3D=
#e14d4>Example 14.4 </a>happens after <a href=3D#e14d2>Example 14.2</a>.
+<p>
+<a href=3D#e14d5>Example 14.5 </a>specifies the future (relative to <a hre=
f=3D#e14d4>Example 14.4</a>) and makes it sticky. So all further events hap=
pen after <a href=3D#e14d5>Example 14.5</a>.
+<p>
+<a href=3D#e14d6>Example 14.6 </a>and <a href=3D#e14d7>Example 14.7 </a>ar=
e again tenseless, and so happen after <a href=3D#e14d5>Example 14.5</a>. (=
Story time is changed.)
+<p>
+So the overall order is 14.1 - 14.3 - 14.2 - 14.4 - (medium interval) - 14=
.5 - 14.6 - 14.7. It is also possible that 14.3 happens before 14.1.
+<p>
+<cx "story time, with no initial sticky time">  XE "story time: with no in=
itial sticky time"   If no sticky time (or space) is set initially, the sto=
ry is set at an unspecified time (or space): the effect is like that of cho=
osing an arbitrary reference point and making it sticky. This style is comm=
on in stories that are jokes. The same convention may be used if the contex=
t specifies the sticky time sufficiently.
+<p>
+<p>
+<a name=3Ds15><h3>Tenses in subordinate bridi</h3>
+<p>
+<cx "subordinate clauses, tense usage rules in English">  XE "subordinate =
clauses: tense usage rules in English"   English has a set of rules, formal=
ly known as =93sequence of tense rules=94, for determining what tense shoul=
d be used in a subordinate clause, depending on the tense used in the main =
sentence. Here are some examples:
+<p>
+<pre><a name=3De15d1>15.1)	John says that George is going to the market.
+<a name=3De15d2>15.2)	John says that George went to the market.
+<a name=3De15d3>15.3)	John said that George went to the market.
+<a name=3De15d4>15.4)	John said that George had gone to the market.
+</pre><p>In <a href=3D#e15d1>Example 15.1 </a>and <a href=3D#e15d2>Example=
 15.2</a>, the tense of the main sentence is the present: =93says=94. If Ge=
orge goes when John speaks, we get the present tense =93is going=94 (=93goe=
s=94 would be unidiomatic); if George goes before John speaks, we get the p=
ast tense =93went=94. But if the tense of the main sentence is the past, wi=
th =93said=94, then the tense required in the subordinate clause is differe=
nt. If George goes when John speaks, we get the past tense =93went=94; if G=
eorge goes before John speaks, we get the past-perfect tense =93had gone=94=
.
+<p>
+The rule of English, therefore, is that both the tense of the main sentenc=
e and the tense of the subordinate clause are understood relative to the sp=
eaker of the main sentence (not John, but the person who speaks <a href=3D#=
e15d{1}>Examples 15.1 </a>through <a href=3D#e15d4>15.4</a>).
+<p>
+<cx "sequence of tense rules, Lojban contrasted with English">  XE "sequen=
ce of tense rules: Lojban contrasted with English"   <cx "subordinate claus=
e tense, Lojban contrasted with English">  XE "subordinate clause tense: Lo=
jban contrasted with English"   <cx "subordinate clause tense, Lojban compa=
red with Russian">  XE "subordinate clause tense: Lojban compared with Russ=
ian"   <cx "subordinate clause tense, Lojban compared with Esperanto">  XE =
"subordinate clause tense: Lojban compared with Esperanto"   <cx "subordina=
te clause tense, effect of main bridi tense on">  XE "subordinate clause te=
nse: effect of main bridi tense on"   Lojban, like Russian and Esperanto, u=
ses a different convention. A tense in a subordinate bridi is understood to=
 be relative to the tense already set in the main bridi. Thus <a href=3D#e1=
5d{1}>Examples 15.1 </a>through <a href=3D#e15d4>15.4 </a>can be expressed =
in Lojban respectively thus:
+<p>
+<ex "John says that George goes to market">  XE "John says that George goe=
s to market: example=93  <pre><a name=3De15d5>15.5)	la djan. ca cusku le se=
 du'u la djordj. ca klama le zarci
+	John [present] says the statement-that George [present] goes-to the marke=
t.
+<a name=3De15d6>15.6)	la djan. ca cusku le se du'u la djordj. pu klama le =
zarci
+	John [present] says the statement-that George [past] goes-to the market.
+<a name=3De15d7>15.7)	la djan. pu cusku le se du'u la djordj. ca klama le =
zarci
+	John [past] says the statement-that George [present] goes-to the market.
+<a name=3De15d8>15.8)	la djan. pu cusku le se du'u la djordj. pu klama le =
zarci
+	John [past] says the statement-that George [past] goes-to the market.
+</pre><p>Probably the most counterintuitive of the Lojban examples is <a h=
ref=3D#e15d7>Example 15.7</a>. The =93ca=94 looks quite odd, as if George w=
ere going to the market right now, rather than back when John spoke. But th=
is =93ca=94 is really a =93ca=94 with respect to a reference point specifie=
d by the outer =93pu=94. This behavior is the same as the additive behavior=
 of multiple tenses in the same bridi, as explained in <a href=3D#s13>Secti=
on 13</a>.
+<p>
+<lx "nau">  XE "nau"   <lx "CUhE">  XE "CUhE selma=92o"   <cx "tense, spea=
ker's current">  XE "tense: speaker's current"   <cx "tense, overriding to =
speaker's current">  XE "tense: overriding to speaker's current"   <cx "nau=
, syntax">  XE "nau: syntax"   There is a special cmavo =93nau=94 (of selma=
'o CUhE) which can be used to override these rules and get to the speaker's=
 current reference point. (Yes, it sounds like English =93now=94.) It is no=
t grammatical to combine =93nau=94 with any other cmavo in a tense, except =
by way of a logical or non-logical connection (see <a href=3D#s20>Section 2=
0</a>). Here is a convoluted sentence with several nested bridi which uses =
=93nau=94 at the lowest level:
+<p>
+<pre><a name=3De15d9>15.9)	la djan. pu cusku le se du'u la .alis pu cusku =
le se du'u
+		la djordj. pu cusku le se du'u la maris. nau klama le zarci
+	John [past] says the statement-that Alice [past] says the statement-that
+		George [past] says the statement that Mary [now] goes-to the market.
+	John said that Alice had said that George had earlier said that Mary is n=
ow
+going to the market.
+</pre><cx "nau, effect on sticky tenses">  XE "nau: effect on sticky tense=
s"   <cx "sticky tenses, effect of nau on">  XE "sticky tenses: effect of n=
au on"   The use of =93nau=94 does not affect sticky tenses.
+<p>
+<a name=3Ds16><h3>Tense relations between sentences</h3>
+<p>
+<cx "tense with sumti tcita, asymmetry of">  XE "tense with sumti tcita: a=
symmetry of"   The sumti tcita method, explained in <a href=3D#s12>Section =
12</a>, of asserting a tense relationship between two events suffers from a=
symmetry. Specifically,
+<p>
+<pre><a name=3De16d1>16.1)	le verba cu cadzu le bisli zu'a le nu le nanmu =
cu batci le gerku
+	The child walks-on the ice [left] the event-of the man bites the dog.
+	The child walks on the ice to the left of where the man bites the dog.
+</pre><lx ".i">  XE "i"   <lx "bo">  XE "bo"   <cx "tense, connecting sent=
ences in with">  XE "tense: connecting sentences in with"   <cx "sentences,=
 connecting with tense">  XE "sentences: connecting with tense"   which spe=
cifies an imaginary journey leftward from the man biting the dog to the chi=
ld walking on the ice, claims only that the child walks on the ice. By the =
nature of =93le nu=94, the man's biting the dog is merely referred to witho=
ut being claimed. If it seems desirable to claim both, each event can be ex=
pressed as a main sentence bridi, with a special form of =93.i=94 connectin=
g them:
+<p>
+<pre><a name=3De16d2>16.2)	le nanmu cu batci le gerku  .izu'abo le verba c=
u cadzu le bisli
+	The man bites the dog.  [Left] the child walks-on the ice.
+	The man bites the dog.  To the left, the child walks on the ice.
+</pre>=93.izu'abo=94 is a compound cmavo: the =93.i=94 separates the sente=
nces and the =93zu'a=94 is the tense. The =93bo=94 is required to prevent t=
he =93zu'a=94 from gobbling up the following sumti, namely =93le verba=94.
+<p>
+<cx "tense connection of sentences, order of">  XE "tense connection of se=
ntences: order of"   <cx "imaginary journey, origin of in tense-connected s=
entences">  XE "imaginary journey: origin of in tense-connected sentences" =
  <cx "tense connection of sentences, contrasted with sumti tcita form">  X=
E "tense connection of sentences: contrasted with sumti tcita form"   <cx "=
tense, sumti tcita form contrasted with connected sentences">  XE "tense: s=
umti tcita form contrasted with connected sentences"   Note that the bridi =
in <a href=3D#e16d2>Example 16.2 </a>appear in the reverse order from their=
 appearance in <a href=3D#e16d1>Example 16.1</a>. With =93.izu'abo=94 (and =
all other afterthought tense connectives) the sentence specifying the origi=
n of the journey comes first. This is a natural order for sentences, but re=
quires some care when converting between this form and the sumti tcita form=
.
+<p>
+<a href=3D#e16d2>Example 16.2 </a>means the same thing as:
+<p>
+<pre><a name=3De16d3>16.3)	le nanmu cu batci le gerku =20
+		.i zu'a la'edi'u le verba cu cadzu le bisli
+	The man bites the dog.
+		[Left] the-referent-of-the-last-sentence the child walks-on the ice.
+	The man bites the dog. Left of what I just mentioned, the child walks on=
=20
+the ice.
+</pre><cx "tense connected sentences, importance of =93bo=94 in">  XE "ten=
se connected sentences: importance of \=93bo\=94 in"   If the =93bo=94 in E=
xample 16.2 is omitted, the meaning changes:
+<p>
+<pre><a name=3De16d4>16.4)	le nanmu cu batci le gerku  .i zu'a le verba cu=
 cadzu le bisli
+	The man bites the dog.  [Left] the child [something] walks-on the ice.
+	The man bites the dog. To the left of the child, something walks on the i=
ce.
+</pre><p>Here the first place of the second sentence is unspecified, becau=
se =93zu'a=94 has absorbed the sumti =93le verba=94.
+<p>
+Do not confuse either <a href=3D#e16d2>Example 16.2 </a>or <a href=3D#e16d=
4>Example 16.4 </a>with the following:
+<p>
+<cx "tense connected sentences, contrasted with separately tensed sentence=
s">  XE "tense connected sentences: contrasted with separately tensed sente=
nces"   <cx "separately tensed sentences, contrasted with tense connected s=
entences">  XE "separately tensed sentences: contrasted with tense connecte=
d sentences"   <pre><a name=3De16d5>16.5)	le nanmu cu batci le gerku  .i zu=
'aku le verba cu cadzu le bisli
+	The man bites the dog.  [Left] the child walks-on the ice.
+	The man bites the dog. Left of me, the child walks on the ice.
+</pre><p>In <a href=3D#e16d5>Example 16.5</a>, the origin point is the spe=
aker, as is usual with =93zu'aku=94. <a href=3D#e16d2>Example 16.2 </a>make=
s the origin point of the tense the event described by the first sentence.
+<p>
+<cx "tense connected sentences, forethought mode">  XE "tense connected se=
ntences: forethought mode"   Two sentences may also be connected in foretho=
ught by a tense relationship. Just like afterthought tense connection, fore=
thought tense connection claims both sentences, and in addition claims that=
 the time or space relationship specified by the tense holds between the ev=
ents the two sentences describe.
+<p>
+<cx "forethought tense connection of sentences, order of">  XE "forethough=
t tense connection of sentences: order of"   <cx "sentences, forethought te=
nse connection of">  XE "sentences: forethought tense connection of"   <cx =
"imaginary journey, origin in tense forethought sentence connection">  XE "=
imaginary journey: origin in tense forethought sentence connection"   <lx "=
gi">  XE "gi"   The origin sentence is placed first, preceded by a tense pl=
us =93gi=94. Another =93gi=94 is used to separate the sentences:
+<p>
+<pre><a name=3De16d6>16.6)	pugi mi klama le zarci gi mi klama le zdani
+	[past] I go-to the market [,] I go-to the house.
+	Before I go to the market, I go to the house.
+</pre><p>A parallel construction can be used to express a tense relationsh=
ip between sumti:
+<p>
+<cx "forethought tense connection of sumti, order of">  XE "forethought te=
nse connection of sumti: order of"   <cx "imaginary journey, origin in tens=
e forethought sumti connection">  XE "imaginary journey: origin in tense fo=
rethought sumti connection"   <cx "sumti, forethought tense connection of">=
  XE "sumti: forethought tense connection of"   <lx "gi">  XE "gi"   <pre><=
a name=3De16d7>16.7)	mi klama pugi le zarci gi le zdani
+	I go-to [past] the market [,] the house.
+</pre><p>Because English does not have any direct way of expressing a tens=
e-like relationship between nouns, <a href=3D#e16d7>Example 16.7 </a>cannot=
 be expressed in English without paraphrasing it either into <a href=3D#e16=
d6>Example 16.6 </a>or else into =93I go to the house before the market=94,=
 which is ambiguous =97 is the market going?
+<p>
+<cx "forethought tense connection of bridi-tails, order of">  XE "forethou=
ght tense connection of bridi-tails: order of"   <cx "imaginary journey, or=
igin in tense forethought bridi-tail connection">  XE "imaginary journey: o=
rigin in tense forethought bridi-tail connection"   <cx "bridi-tails, foret=
hought tense connection of">  XE "bridi-tails: forethought tense connection=
 of"   <lx "gi">  XE "gi"   Finally, a third forethought construction expre=
sses a tense relationship between bridi-tails rather than whole bridi. (The=
 construct known as a =93bridi-tail=94 is explained fully in <a href=3Dchap=
14.html>Chapter 14</a>; roughly speaking, it is a selbri, possibly with fol=
lowing sumti.) <a href=3D#e16d8>Example 16.8 </a>is equivalent in meaning t=
o <a href=3D#e16d6>Example 16.6 </a>and <a href=3D#e16d7>Example 16.7</a>:
+<p>
+<pre><a name=3De16d8>16.8)	mi pugi klama le zarci gi klama le zdani
+	I [past] go-to the market [,] go-to the house.
+	I, before going to the market, go to the house.
+</pre><cx "tense connection of sumti, meaning of">  XE "tense connection o=
f sumti: meaning of"   <cx "tense connection of bridi-tails, meaning of">  =
XE "tense connection of bridi-tails: meaning of"   In both <a href=3D#e16d7=
>Example 16.7 </a>and <a href=3D#e16d8>Example 16.8</a>, the underlying sen=
tences =93mi klama le zarci=94 and =93mi klama le zdani=94 are not claimed;=
 only the relationship in time between them is claimed.
+<p>
+<cx "tense forethought connection forms, selma'o allowed">  XE "tense fore=
thought connection forms: selma'o allowed"   <cx "tense afterthought connec=
tion forms, selma'o allowed">  XE "tense afterthought connection forms: sel=
ma'o allowed"   Both the forethought and the afterthought forms are appropr=
iate with PU, ZI, FAhA, VA, and ZAhO tenses. In all cases, the equivalent f=
orms are (where X and Y stand for sentences, and TENSE for a tense cmavo):
+<p>
+<cx "tense connection, equivalent meanings">  XE "tense connection: equiva=
lent meanings"   <cx "tense connection, expansions of">  XE "tense connecti=
on: expansions of"   <dl compact><dt>	<dd>subordinate:  X TENSE le nu Y
+	afterthought coordinate:  Y  .i+TENSE+bo X
+	forethought coordinate:  TENSE+gi X gi Y
+</dl><a name=3Ds17><h3>Tensed logical connectives</h3>
+<p>
+<cx "logical connectives, tensed">  XE "logical connectives: tensed"   <cx=
 "tensed logical connectives">  XE "tensed logical connective"   The Lojban=
 tense system interacts with the Lojban logical connective system. That sys=
tem is a separate topic, explained in <a href=3Dchap14.html>Chapter 14 </a>=
and touched on only in summary here. By the rules of the logical connective=
 system, <a href=3D#e17d1>Example 17.1 </a>through 17.3 are equivalent in m=
eaning:
+<p>
+<pre><a name=3De17d1>17.1)	la teris. satre le mlatu  .ije la teris. satre =
le ractu
+	Terry strokes the cat. And Terry strokes the rabbit.
+<a name=3De17d2>17.2)	la teris. satre le mlatu gi'e satre le ractu
+	Terry strokes the cat and strokes the rabbit.
+<a name=3De17d3>17.3)	la teris. satre le mlatu .e le ractu
+	Terry strokes the cat and the rabbit.
+</pre><lx "bo">  XE "bo"   <ex "and then">  XE "and then: example=93  <ex =
"stoke cat then rabbit">  XE "stoke cat then rabbit: example=93  Suppose we=
 wish to add a tense relationship to the logical connective =93and=94? To s=
ay that Terry strokes the cat and later strokes the rabbit, we can combine =
a logical connective with a tense connective by placing the logical connect=
ive first, then the tense, and then the cmavo =93bo=94, thus:
+<p>
+<cx "tensed logically connected sentences">  XE "tensed logically connecte=
d sentences"   <pre><a name=3De17d4>17.4)	la teris. satre le mlatu  .ijebab=
o la teris. satre le ractu
+	Terry strokes the cat.  And then Terry strokes the rabbit.
+</pre><cx "tensed logically connected bridi-tails">  XE "tensed logically =
connected bridi-tails"   <pre><a name=3De17d5>17.5)	la teris. satre le mlat=
u gi'ebabo satre le ractu
+	Terry strokes the cat, and then strokes the rabbit.
+</pre><cx "tensed logically connected sumti">  XE "tensed logically connec=
ted sumti"   <pre><a name=3De17d6>17.6)	la teris. satre le mlatu .ebabo le =
ractu
+	Terry strokes the cat and then the rabbit.
+</pre><a href=3D#e17d4>Example 17.4 </a>through 17.6 are equivalent in mea=
ning. They are also analogous to <a href=3D#e17d{1}>Examples 17.1 </a>throu=
gh <a href=3D#e17d3>17.3 </a>respectively. The =93bo=94 is required for the=
 same reason as in <a href=3D#e16d2>Example 16.2</a>: to prevent the =93ba=
=94 from functioning as a sumti tcita for the following sumti (or, in <a hr=
ef=3D#e17d5>Example 17.5</a>, from being attached to the following selbri).
+<p><cx "tensed logical connectives, with ke=85ke'e">  XE "tensed logical c=
onnective: with ke=85ke'e"   <cx "tensed logical connectives, with tu'e=85t=
u'u">  XE "tensed logical connective: with tu'e=85tu'u"   In addition to th=
e =93bo=94 construction of <a href=3D#e17d{4}>Examples 17.4 </a>through <a =
href=3D#e17d6>17.6</a>, there is also a form of tensed logical connective w=
ith =93ke=85ke'e=94 (=93tu'e=85tu'u=94 for sentences). The logical connecti=
ve system makes <a href=3D#e17d{7}>Examples 17.7 </a>through <a href=3D#e17=
d9>17.9 </a>equivalent in meaning:
+<p>
+<ex "carry sack and dog">  XE "carry sack and dog: example=93  <pre><a nam=
e=3De17d7>17.7)	mi bevri le dakli  .ije tu'e mi bevri le gerku  .ija mi bev=
ri le mlatu tu'u
+	I carry the sack.  And (I carry the dog.  And/or I carry the cat).
+	I carry the sack.  And I carry the dog, or I carry the cat, or I carry bo=
th.
+<a name=3De17d8>17.8)	mi bevri le dakli gi'eke bevri le gerku gi'a bevri l=
e mlatu
+	I carry the sack and (carry the dog and/or carry the cat).
+	I carry the sack, and also carry the dog or carry the cat or carry both.
+<a name=3De17d9>17.9)	mi bevri le dakli .eke le gerku .a le mlatu
+	I carry the sack and (the dog or the cat)
+	I carry the sack and also the dog or the cat or both.
+</pre><p>Note the uniformity of the Lojban, as contrasted with the variety=
 of ways in which the English provides for the correct grouping. In all cas=
es, the meaning is that I carry the sack in any case, and either the cat or=
 the dog or both.
+<p>
+To express that I carry the sack first (earlier in time), and then the dog=
 or the cat or both simultaneously, I can insert tenses to form <a href=3D#=
e17d{10}>Examples 17.10 </a>through <a href=3D#e17d12>17.12</a>:
+<p>
+<cx "tensed logically connected sentences, with grouping">  XE "tensed log=
ically connected sentences: with grouping"   <pre><a name=3De17d10>17.10)	m=
i bevri le dakli  .ije ba tu'e mi bevri le gerku =20
+.ijacabo mi bevri le mlatu tu'u
+	I carry the sack.  And [future] (I carry the dog. =20
+And/or [present] I carry the cat.)
+	I carry the sack.  And then I will carry the dog or I will carry the cat =
or=20
+I will carry both at once.
+</pre><cx "tensed logically connected bridi-tails, with grouping">  XE "te=
nsed logically connected bridi-tails: with grouping"   <pre><a name=3De17d1=
1>17.11)	mi bevri le dakli gi'ebake bevri le gerku gi'acabo bevri le mlatu
+	I carry the sack and [future] (carry the dog and/or [present] carry the c=
at).
+	I carry the sack and then will carry the dog or carry the cat or carry bo=
th=20
+at once.
+</pre><cx "tensed logically connected sumti, with grouping">  XE "tensed l=
ogically connected sumti: with grouping"   <pre><a name=3De17d12>17.12)	mi =
bevri le dakli .ebake bevri le gerku .acabo le mlatu
+	I carry the sack and [future] (the cat and/or [present] the dog).
+	I carry the sack, and then the cat or the dog or both at once.
+</pre><a href=3D#e17d{10}>Examples 17.10 </a>through <a href=3D#e17d12>17.=
12 </a>are equivalent in meaning to each other, and correspond to the tense=
less <a href=3D#e17d{7}>Examples 17.7 </a>through <a href=3D#e17d9>17.9 </a=
>respectively.
+<p>
+<a name=3Ds18><h3>Tense negation</h3>
+<p>
+<cx "tenses, negating">  XE =93tense: negating"   <cx "negation, of tenses=
">  XE "negation: of tenses"   <lx "nai">  XE "nai"   <cx "PU, contradictor=
y negation of">  XE "PU selma=92o: contradictory negation of"   <cx "FAhA, =
contradictory negation of">  XE "FAhA selma=92o: contradictory negation of"=
   <cx "ZAhO, contradictory negation of">  XE "ZAhO selma=92o: contradictor=
y negation of"   Any bridi which involves tenses of selma'o PU, FAhA, or ZA=
hO can be contradicted by a =93-nai=94 suffixed to the tense cmavo. Some ex=
amples:
+<p>
+<pre><a name=3De18d1>18.1)	mi punai klama le zarci
+	I [past] [not] go-to the market.
+	I didn't go to the market.
+</pre><cx "tenses, contradictory negation of with nai">  XE =93tense: cont=
radictory negation of with nai"   <cx "negation of tenses, meaning of">  XE=
 "negation of tenses: meaning of"   <lx "nai">  XE "nai"   As a contradicto=
ry negation, <a href=3D#e18d1>Example 18.1 </a>implies that the bridi as a =
whole is false without saying anything about what is true. When the negated=
 tense is a sumti tcita, =93nai=94 negation indicates that the stated relat=
ionship does not hold:
+
+<p>
+<pre><a name=3De18d2>18.2)	mi klama le zarci canai le nu do klama le zdani
+	I go-to the market [present] [not] the event-of you go-to the house.
+	It is not true that I went to the market at the same time that you went t=
o
+the house.
+<a name=3De18d3>18.3)	le nanmu batci le gerku ne'inai le kumfa
+	The man bites the dog [within] [not] the room.
+	The man didn't bite the dog inside the room.
+<a name=3De18d4>18.4)	mi morsi ca'onai le nu mi jmive
+	I am-dead [continuitive - negated] the event-of I live.
+	It is false that I am dead during my life.
+</pre><cx "tense, scalar negation of with NAhE">  XE "tense: scalar negati=
on of with NAhE"   <lx "NAhE">  XE "NAhE selma=92o"   <cx "tense, scalar ne=
gation contrasted with contradictory negation of">  XE "tense: scalar negat=
ion contrasted with contradictory negation of"   <cx "tense, contradictory =
negation contrasted with scalar negation of">  XE "tense: contradictory neg=
ation contrasted with scalar negation of"   It is also possible to perform =
scalar negation of whole tense constructs by placing a member of NAhE befor=
e them. Unlike contradictory negation, scalar negation asserts a truth: tha=
t the bridi is true with some tense other than that specified. The followin=
g examples are scalar negation analogues of <a href=3D#e18d{1}>Examples 18.=
1 </a>to <a href=3D#e18d3>18.3</a>:
+<p>
+<pre><a name=3De18d5>18.5)	mi na'e pu klama le zarci
+	I [non-] [past] go-to the market.
+	I go to the market other than in the past.
+<a name=3De18d6>18.6)	le nanmu batci le gerku to'e ne'i le kumfa
+	The man bites the dog [opposite-of] [within] the room.
+	The man bites the dog outside the room.
+<a name=3De18d7>18.7)	mi klama le zarci na'e ca le nu do klama le zdani
+	I go-to the market [non-] [present] the event-of you go-to the house.
+	I went to the market at a time other than the time at which you went to=
=20
+the house.
+<a name=3De18d8>18.8)	mi morsi na'e ca'o le nu mi jmive
+	I am-dead [non-] [continuitive] the event-of I live.
+	I am dead other than during my life.
+</pre><cx "contradictory negation of tenses, selma'o allowed with">  XE "c=
ontradictory negation of tenses: selma'o allowed with"   <cx "scalar negati=
on of tenses, selma'o allowed with">  XE "scalar negation of tenses: selma'=
o allowed with"   <lx "PU">  XE "PU selma=92o"   <lx "FAhA">  XE "FAhA selm=
a=92o"   Unlike =93-nai=94 contradictory negation, scalar negation of tense=
s is not limited to PU and FAhA:
+<p>
+<pre><a name=3De18d9>18.9)	le verba na'e ri'u cadzu le bisli
+	The child [non-] [right] walks-on the ice
+	The child walks on the ice other than to my right.
+</pre><cx "TAhE, scalar negation of">  XE "TAhE selma=92o: scalar negation=
 of"   <cx "ROI, scalar negation of">  XE "ROI selma=92o: scalar negation o=
f"   The use of =93-nai=94 on cmavo of TAhE and ROI has already been discus=
sed in <a href=3D#s9>Section 9</a>; this use is also a scalar negation.
+
+<p>
+<a name=3Ds19><h3>Actuality, potentiality, capability: CAhA</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ca'a	CAhA	actually is
+	ka'e	CAhA	is innately capable of
+	nu'o	CAhA	can but has not
+	pu'i	CAhA	can and has
+</pre><cx "potential events, expressing implicitly">  XE "potential events=
: expressing implicitly"   <cx "innate capabilities, expressing implicitly"=
>  XE "innate capability: expressing implicitly"   Lojban bridi without ten=
se markers may not necessarily refer to actual events: they may also refer =
to capabilities or potential events. For example:
+<p>
+<ex "ducks swim">  XE "ducks swim: example=93  <pre><a name=3De19d1>19.1)	=
ro datka cu flulimna
+	All ducks are-float-swimmers
+	All ducks swim by floating.
+</pre><cx "tense, Lojban contrasted with English in implying actuality">  =
XE "tense: Lojban contrasted with English in implying actuality"   <cx "act=
uality, Lojban contrasted with English in implying">  XE "actuality: Lojban=
 contrasted with English in implying"   is a Lojban truth, even though the =
colloquial English translation is false or at best ambiguous. This is becau=
se the tenseless Lojban bridi doesn't necessarily claim that every duck is =
swimming or floating now or even at a specific time or place. Even if we ad=
d a tense marker to <a href=3D#e19d1>Example 19.1</a>,
+<p>
+<pre><a name=3De19d2>19.2)	ro datka ca flulimna
+	All ducks [present] are-float-swimmers.
+	All ducks are now swimming by floating.
+</pre>the resulting <a href=3D#e19d2>Example 19.2 </a>might still be consi=
dered a truth, even though the colloquial English seems even more likely to=
 be false. All ducks have the potential of swimming even if they are not ex=
ercising that potential at present. To get the full flavor of =93All ducks =
are now swimming=94, we must append a marker from selma'o CAhA to the tense=
, and say:
+<p>
+<cx "actual events, explicitly expressing">  XE "actual events: explicitly=
 expressing"   <lx "ca'a">  XE "ca'a"   <pre><a name=3De19d3>19.3)	ro datka=
 ca ca'a flulimna
+	All ducks [present] [actual] are-float-swimmers
+	All ducks are now actually swimming by floating.
+</pre><cx "CAhA, order in tense construct">  XE "CAhA selma=92o: order in =
tense construct"   <lx "CAhA">  XE "CAhA selma=92o"   <lx "ki">  XE "ki"   =
<cx "CAhA, making sticky">  XE "CAhA selma=92o: making sticky"   <cx "stick=
y tenses, and CAhA">  XE "sticky tenses: and CAhA"   A CAhA cmavo is always=
 placed after any other tense cmavo, whether for time or for space. However=
, a CAhA cmavo comes before =93ki=94, so that a CAhA condition can be made =
sticky.
+<p>
+<a href=3D#e19d3>Example 19.3 </a>is false in both Lojban and English, sin=
ce it claims that the swimming is an actual, present fact, true of every du=
ck that exists, whereas in fact there is at least one duck that is not swim=
ming now.
+<p>
+<cx "innate capability, expressing explicitly">  XE "innate capability: ex=
pressing explicitly"   <lx "ka'e">  XE "ka'e"   Furthermore, some ducks are=
 dead (and therefore sink); some ducks have just hatched (and do not know h=
ow to swim yet), and some ducks have been eaten by predators (and have ceas=
ed to exist as separate objects at all). Nevertheless, all these ducks have=
 the innate capability of swimming =97 it is part of the nature of duckhood=
. The cmavo =93ka'e=94 expresses this notion of innate capability:
+<p>
+<pre><a name=3De19d4>19.4)	ro datka ka'e flulimna
+	All ducks [capable] are-float-swimmers.
+	All ducks are innately capable of swimming.
+</pre><cx "innate properties, extension of from mass to individuals">  XE =
"innate property: extension of from mass to individuals"   Under some epist=
emologies, innate capability can be extended in order to apply the innate p=
roperties of a mass to which certain individuals belong to the individuals =
themselves, even if those individuals are themselves not capable of fulfill=
ing the claim of the bridi. For example:
+<p>
+<ex "can see">  XE "can see: example=93  <pre><a name=3De19d5>19.5)	la dja=
n. ka'e viska
+	John [capable] sees.
+	John is innately capable of seeing.
+	John can see.
+</pre><cx "innate properties, extension to individuals not actually capabl=
e">  XE "innate property: extension to individuals not actually capable"   =
might be true about a human being named John, even though he has been blind=
 since birth, because the ability to see is innately built into his nature =
as a human being. It is theoretically possible that conditions might occur =
that would enable John to see (an operation, for example). On the other han=
d,
+<p>
+<pre><a name=3De19d6>19.6)	le cukta ka'e viska
+	The book [capable] sees.
+	The book can see.
+</pre>is not true in most epistemologies, since the ability to see is not =
part of the innate nature of a book.
+<p>
+<cx "undemonstrated potential, expressing">  XE "undemonstrated potential:=
 expressing"   Consider once again the newly hatched ducks mentioned earlie=
r. They have the potential of swimming, but have not yet demonstrated that =
potential. This may be expressed using =93nu'o=94, the cmavo of CAhA for un=
demonstrated potential:
+<p>
+<ex "infant ducks">  XE "infant ducks: example=93  <lx "nu'o">  XE "nu'o" =
  <pre><a name=3De19d7>19.7)	ro cifydatka nu'o flulimna
+	All infant-ducks [can but has not] are-float-swimmers.
+	All infant ducks have an undemonstrated potential for swimming by floatin=
g.
+	Baby ducks can swim but haven't yet.
+</pre><cx "demonstrated potential, expressing">  XE "demonstrated potentia=
l: expressing"   Contrariwise, if Frank is not blind from birth, then =93pu=
'i=94 is appropriate:
+<p>
+<lx "pu'i">  XE "pu'i"   <pre><a name=3De19d8>19.8)	la frank. pu'i viska
+	Frank [can and has] sees.
+	Frank has demonstrated a potential for seeing.
+	Frank can see and has seen.
+</pre><cx "potential, expressing in past/future">  XE "potential: expressi=
ng in past/future"   <cx "actuality, expressing in past/future">  XE "actua=
lity: expressing in past/future"   Note that the glosses given at the begin=
ning of this section for =93ca'a=94, =93nu'o=94, and =93pu'i=94 incorporate=
 =93ca=94 into their meaning, and are really correct for =93ca ca'a=94, =93=
ca nu'o=94, and =93ca pu'i=94. However, the CAhA cmavo are perfectly meanin=
gful with other tenses than the present:
+<p>
+<pre><a name=3De19d9>19.9)	mi pu ca'a klama le zarci
+	I [past] [actual] go-to the store.
+	I actually went to the store.
+<a name=3De19d10>19.10)	la frank. ba nu'o klama le zdani
+	Frank [future] [can but has not] goes-to the store.
+	Frank could have, but will not have, gone to the store (at some understoo=
d
+moment in the future).
+</pre><cx "tenses with elided CAhA, meaning">  XE =93tense with elided CAh=
A: meaning"   As always in Lojban tenses, a missing CAhA can have an indete=
rminate meaning, or the context can be enough to disambiguate it. Saying
+<p>
+<pre><a name=3De19d11>19.11)	ta jelca
+	That burns/is-burning/might-burn/will-burn.
+</pre><ex "inflammable">  XE "inflammable: example=93  with no CAhA specif=
ied can translate the two very different English sentences =93That is on fi=
re=94 and =93That is inflammable.=94 The first demands immediate action (us=
ually), whereas the second merely demands caution. The two cases can be dis=
ambiguated with:
+<p>
+<pre><a name=3De19d12>19.12)	ta ca ca'a jelca
+	That [present] [actual] burns.
+	That is on fire.
+</pre>and
+<p>
+<pre><a name=3De19d13>19.13)	ta ka'e jelca
+	That [capable] burns.
+	That is capable of burning.
+	That is inflammable.
+</pre><cx "observative with elided CAhA, convention">  XE "observative wit=
h elided CAhA: convention"   When no indication is given, as in the simple =
observative
+<p>
+<pre><a name=3De19d14>19.14)	jelca
+	It burns!
+</pre>the prudent Lojbanist will assume the meaning =93Fire!=94
+<p>
+<a name=3Ds20><h3>Logical and non-logical connections between tenses</h3>
+<p>
+Like many things in Lojban, tenses may be logically connected; logical con=
nection is explained in more detail in <a href=3Dchap14.html>Chapter 14</a>=
. Some of the terminology in this section will be clear only if you already=
 understand logical connectives.
+<p>
+<lx "JA">  XE "JA selma=92o"   <cx "logically connected tenses, with JA"> =
 XE "logically connected tenses: with JA"   <cx "tenses, logically connecte=
d with JA">  XE =93tense: logically connected with JA"   <cx "logically con=
nected tenses, expansion to sentences">  XE "logically connected tenses: ex=
pansion to sentences"   The appropriate logical connectives belong to selma=
'o JA. A logical connective between tenses can always be expanded to one be=
tween sentences:
+<p>
+<pre><a name=3De20d1>20.1)	mi pu je ba klama le zarci
+	I [past] and [future] go-to the market.
+	I went and will go to the market.
+</pre>means the same as:
+<p>
+<pre><a name=3De20d2>20.2)	mi pu klama le zarci  .ije mi ba klama le zarci
+	I [past] go-to the market.  And I [future] go-to the market.
+	I went to the market, and I will go to the market.
+</pre><cx "tenses, connected, with negation">  XE =93tense: connected, wit=
h negation"   Tense connection and tense negation are combined in:
+<p>
+<pre><a name=3De20d3>20.3)	mi punai je canai je ba klama le zarci
+	I [past] [not] and [present] [not] and [future] go-to the market.
+	I haven't yet gone to the market, but I will in future.
+</pre><a href=3D#e20d3>Example 20.3 </a>is far more specific than
+<p>
+<pre><a name=3De20d4>20.4)	mi ba klama le zarci
+	I [future] go-to the market.
+</pre>which only says that I will go, without claiming anything about my p=
ast or present. =93ba=94 does not imply =93punai=94 or =93canai=94; to comp=
el that interpretation, either a logical connection or a ZAhO is needed.
+<p>
+<cx "connected tenses, negation of compared with negation in connective"> =
 XE "connected tenses: negation of compared with negation in connective"   =
Tense negation can often be removed in favor of negation in the logical con=
nective itself. The following examples are equivalent in meaning:
+<p>
+<pre><a name=3De20d5>20.5)	mi mo'izu'anai je mo'iri'u cadzu
+	I [motion] [left-not] and [motion] [right] walk.
+	I walk not leftward but rightward.
+<a name=3De20d6>20.6)	mi mo'izu'a naje mo'iri'u cadzu
+	I [motion] [left] not-and [motion] [right] walk.
+	I walk not leftward but rightward.
+</pre><cx "tenses, forethought logical connections">  XE =93tense: foretho=
ught logical connections"   <cx "tenses, possible groupings of">  XE =93ten=
se: possible groupings of"   There are no forethought logical connections b=
etween tenses allowed by the grammar, to keep tenses simpler. Nor is there =
any way to override simple left-grouping of the connectives, the Lojban def=
ault.
+<p>
+<cx "tenses, non-logical connection of">  XE =93tense: non-logical connect=
ion of"   <lx "JOI">  XE "JOI selma=92o"   <lx "BIhI">  XE "BIhI selma=92o"=
   <lx "GAhO">  XE "GAhO selma=92o"   <cx "intervals, expressing by endpoin=
ts with bi'o">  XE "interval: expressing by endpoints with bi'o"   The non-=
logical connectives of selma'o JOI, BIhI, and GAhO are also permitted betwe=
en tenses. One application is to specify intervals not by size, but by thei=
r end-points (=93bi'o=94 belongs to selma'o BIhI, and connects the end-poin=
ts of an ordered interval, like English =93from=A0<dots>=85</dots>to=94):
+<p>
+<pre><a name=3De20d7>20.7)	mi puza bi'o bazu vasxu
+	I [past] [medium] from=A0<dots>=85</dots>to [future] [long] breathe.
+	I breathe from a medium time ago till a long time to come.
+</pre>(It is to be hoped that I have a long life ahead of me.)
+<p>
+One additional use of non-logical connectives within tenses is discussed i=
n <a href=3D#s21>Section 21</a>. Other uses will probably be identified in =
future.
+<p>
+<a name=3Ds21><h3>Sub-events</h3>
+<p>
+<ex "six-shooter">  XE "six-shooter: example=93  <cx "tenses, non-logical =
connection of for sub-events">  XE =93tense: non-logical connection of for =
sub-events"   Another application of non-logical tense connection is to tal=
k about sub-events of events. Consider a six-shooter: a gun which can fire =
six bullets in succession before reloading. If I fire off the entire magazi=
ne twice, I can express the fact in Lojban thus:
+<p>
+<ex "on two occasions">  XE "on two occasions: example=93  <pre><a name=3D=
e21d1>21.1)	mi reroi pi'u xaroi cecla le seldanti
+	I [twice] [cross-product] [six times] shoot the projectile-launcher.
+	On two occasions, I fire the gun six times.
+</pre><lx "pi'u">  XE "pi'u"   <cx "pi'u, use in connecting tenses">  XE "=
pi'u: use in connecting tenses"   <cx "Cartesian product, with tenses">  XE=
 "Cartesian product: with tenses"   <cx "cross product, with tenses">  XE "=
cross product: with tenses"   It would be confusing, though grammatical, to=
 run the =93reroi=94 and the =93xaroi=94 directly together. However, the no=
n-logical connective =93pi'u=94 expresses a Cartesian product (also known a=
s a cross product) of two sets. In this case, there is a set of two firings=
 each of which is represented by a set of six shots, for twelve shots in al=
l (hence the name =93product=94: the product of 2 and 6 is 12). Its use spe=
cifies very precisely what occurs.
+<p>
+<cx "interval properties, strings of">  XE "interval properties: strings o=
f"   <cx "event contours, strings of">  XE "event contours: strings of"   I=
n fact, you can specify strings of interval properties and event contours w=
ithin a single tense without the use of a logical or non-logical connective=
 cmavo. This allows tenses of the type:
+<p>
+<pre><a name=3De21d2>21.2)	la djordj. ca'o co'a ciska
+	George [continuitive] [initiative] writes.
+	George continues to start to write.
+<a name=3De21d3>21.3)	mi reroi ca'o xaroi darxi le damri
+	I [twice] [continuitive] [six times] hit the drum.
+	On two occasions, I continue to beat the drum six times.
+</pre><a name=3Ds22><h3>Conversion of sumti tcita: JAI</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	jai	JAI	tense conversion
+	fai	FA	indefinite place
+</pre><lx "SE">  XE "SE selma=92o"   <cx "conversion, definition">  XE "co=
nversion: definition"   Conversion is the regular Lojban process of moving =
around the places of a place structure. The cmavo of selma'o SE serve this =
purpose, exchanging the first place with one of the others:
+<p>
+<pre><a name=3De22d1>22.1)	mi cu klama le zarci
+	I go-to the market.
+<a name=3De22d2>22.2)	le zarci cu se klama mi
+	The market is-gone-to by-me.
+</pre><cx "conversion, accessing tense of bridi with jai">  XE "conversion=
: accessing tense of bridi with jai"   <cx "tense conversion, accessing ten=
se of bridi with jai">  XE "tense conversion: accessing tense of bridi with=
 jai"   <lx "jai">  XE "jai"   <cx "jai with tense, as equivalent of SE in =
grammar">  XE "jai with tense: as equivalent of SE in grammar"   It is also=
 possible to bring a place that is specified by a sumti tcita (for the purp=
oses of this chapter, a tense sumti tcita) to the front, by using =93jai=94=
 plus the tense as the grammatical equivalent of SE:
+<p>
+<ex "rat eats cheese in park">  XE "rat eats cheese in park: example=93  <=
pre><a name=3De22d3>22.3)	le ratcu cu citka le cirla vi le panka
+	The rat eats the cheese [short distance] the park.
+	The rat eats the cheese in the park.
+<a name=3De22d4>22.4)	le panka cu jai vi citka le cirla fai le ratcu
+	The park is-the-place-of eating the cheese by-the rat.
+	The park is where the rat eats the cheese.
+</pre><cx "tense conversion, accessing original first place with fai">  XE=
 "tense conversion: accessing original first place with fai"   <lx "fai">  =
XE "fai"   <lx "FA">  XE "FA selma=92o"   In <a href=3D#e22d4>Example 22.4<=
/a>, the construction JAI+tense converts the location sumti into the first =
place. The previous first place has nowhere to go, since the location sumti=
 is not a numbered place; however, it can be inserted back into the bridi w=
ith =93fai=94, the indefinite member of selma'o FA.
+<p>
+(The other members of FA are used to mark the first, second, etc. places o=
f a bridi explicitly:
+<p>
+<pre><a name=3De22d5>22.5)	fa mi cu klama fe le zarci
+</pre>means the same as
+<p>
+<pre><a name=3De22d6>22.6)	fe le zarci cu klama fa mi
+</pre>as well as the simple
+<p>
+<pre><a name=3De22d7>22.7)	mi cu klama le zarci
+</pre>in which the place structure is determined by position.)
+<p>
+<cx "tense conversion, use in sumti descriptions">  XE "tense conversion: =
use in sumti descriptions"   <lx "LE">  XE "LE selma=92o"   Like SE convers=
ion, JAI+tense conversion is especially useful in descriptions with LE selm=
a'o:
+<p>
+<ex "place of eating">  XE "place of eating: example=93  <pre><a name=3De2=
2d8>22.8)	mi viska le jai vi citka be le cirla
+	I saw the place-of eating the cheese.
+</pre><p>Here the eater of the cheese is elided, so no =93fai=94 appears.
+
+<p>
+<cx "tense conversion, of temporal tenses">  XE "tense conversion: of temp=
oral tenses"   Of course, temporal tenses are also usable with JAI:
+<p>
+<ex "time of death">  XE "time of death: example=93  <pre><a name=3De22d9>=
22.9)	mi djuno fi le jai ca morsi be fai la djan.
+	I know about the [present] is-dead of-the-one-called =93John=94
+	I know the time of John's death.
+	I know when John died.
+</pre><a name=3Ds23><h3>Tenses versus modals</h3>
+<p>
+<cx "tenses, compared with modals in syntax">  XE =93tense: compared with =
modals in syntax"   <cx "modals, compared with tenses in syntax">  XE "moda=
ls: compared with tenses in syntax"   <cx "tenses, contrasted with modals i=
n semantics">  XE =93tense: contrasted with modals in semantics"   <cx "mod=
als, contrasted with tenses in semantics">  XE "modals: contrasted with ten=
ses in semantics"   Grammatically, every use of tenses seen so far is exact=
ly paralleled by some use of modals as explained in <a href=3Dchap9.html>Ch=
apter 9</a>. Modals and tenses alike can be followed by sumti, can appear b=
efore the selbri, can be used in pure and mixed connections, can participat=
e in JAI conversions. The parallelism is perfect. However, there is a deep =
difference in the semantics of tense constructs and modal constructs, groun=
ded in historical differences between the two forms. Originally, modals and=
 tenses were utterly different things in earlier versions of Loglan; only i=
n Lojban have they become grammatically interchangeable. And even now, diff=
erences in semantics continue to be maintained.
+<p>
+<cx "modals, importance of 1st sumti place for sumti tcita use">  XE "moda=
ls: importance of 1st sumti place for sumti tcita use"   The core distincti=
on is that whereas the modal bridi
+<p>
+<pre><a name=3De23d1>23.1)	mi nelci do mu'i le nu do nelci mi
+	I like you with-motivation the event-of you like me.
+	I like you because you like me.
+</pre>places the =93le nu=94 sumti in the x1 place of the gismu =93mukti=
=94 (which underlies the modal =93mu'i=94), namely the motivating event, th=
e tensed bridi
+<p>
+<cx "tenses, importance of 2nd sumti place for sumti tcita use">  XE =93te=
nse: importance of 2nd sumti place for sumti tcita use"   <pre><a name=3De2=
3d2>23.2)	mi nelci do ba le nu do nelci mi
+	I like you after the event-of you like me.
+	I like you after you like me.
+</pre>places the =93le nu=94 sumti in the x2 place of the gismu =93balvi=
=94 (which underlies the tense =93ba=94), namely the point of reference for=
 the future tense. Paraphrases of <a href=3D#e23d1>Example 23.1 </a>and <a =
href=3D#e23d2>Example 23.2</a>, employing the brivla =93mukti=94 and =93bal=
vi=94 explicitly, would be:
+<p>
+<pre><a name=3De23d3>23.3)	le nu do nelci mi cu mukti le nu mi nelci do
+	The event-of you like me motivates the event-of I like you
+	Your liking me is the motive for my liking you.
+</pre>and
+<p>
+<pre><a name=3De23d4>23.4)	le nu mi nelci do cu balvi le nu do nelci mi
+	The event-of I like you is after the event of you like me.
+	My liking you follows (in time) your liking me.
+</pre>(Note that the paraphrase is not perfect due to the difference in wh=
at is claimed; <a href=3D#e23d3>Example 23.3 </a>and <a href=3D#e23d4>Examp=
le 23.4 </a>claim only the causal and temporal relationships between the ev=
ents, not the existence of the events themselves.)
+<p>
+<cx "afterthought sentence connection, modal contrasted with tense">  XE "=
afterthought sentence connection: modal contrasted with tense"   As a resul=
t, the afterthought sentence-connective forms of <a href=3D#e23d1>Example 2=
3.1 </a>and <a href=3D#e23d2>Example 23.2 </a>are, respectively:
+<p>
+<pre><a name=3De23d5>23.5)	mi nelci do .imu'ibo do nelci mi
+	I like you. [That is] Because you like me.
+<a name=3De23d6>23.6)	do nelci mi .ibabo mi nelci do
+	You like me. Afterward, I like you.
+</pre><p>In <a href=3D#e23d5>Example 23.5</a>, the order of the two bridi =
=93mi nelci do=94 and =93do nelci mi=94 is the same as in <a href=3D#e23d1>=
Example 23.1</a>. In <a href=3D#e23d6>Example 23.6</a>, however, the order =
is reversed: the origin point =93do nelci mi=94 physically appears before t=
he future-time event =93mi nelci do=94. In both cases, the bridi characteri=
zing the event in the x2 place appears before the bridi characterizing the =
event in the x1 place of =93mukti=94 or =93balvi=94.
+<p>
+<cx "forethought connections, modal compared with tense in semantics">  XE=
 "forethought connections: modal compared with tense in semantics"   <cx "f=
orethought tense connection, contrasted with afterthought in likeness to mo=
dal connection">  XE "forethought tense connection: contrasted with afterth=
ought in likeness to modal connection"   <cx "afterthought tense connection=
, contrasted with forethought in likeness to modal connection">  XE "aftert=
hought tense connection: contrasted with forethought in likeness to modal c=
onnection"   In forethought connections, however, the asymmetry between mod=
als and tenses is not found. The forethought equivalents of <a href=3D#e23d=
5>Example 23.5 </a>and <a href=3D#e23d6>Example 23.6 </a>are
+<p>
+<pre><a name=3De23d7>23.7)	mu'igi do nelci mi gi mi nelci do
+	Because you like me, I like you.
+</pre>and
+<p>
+<pre><a name=3De23d8>23.8)	bagi do nelci mi gi mi nelci do
+	After you like me, I like you.
+</pre>respectively.
+<p>
+<cx "modal sentence connection, table of equivalent schemata">  XE "modal =
sentence connection: table of equivalent schemata"   The following modal se=
ntence schemata (where X and Y represent sentences) all have the same meani=
ng:
+<p>
+<dl compact><dt><dd>X  .i BAI bo Y
+BAI gi Y gi X
+X BAI le nu Y
+</dl><cx "tense sentence connection, table of equivalent schemata">  XE "t=
ense sentence connection: table of equivalent schemata"   whereas the follo=
wing tensed sentence schemata also have the same meaning:
+<p>
+<dl compact><dt><dd>X  .i TENSE bo Y
+TENSE gi X gi Y
+Y TENSE le nu X
+</dl>neglecting the question of what is claimed. In the modal sentence sch=
emata, the modal tag is always followed by Y, the sentence representing the=
 event in the x1 place of the gismu that underlies the BAI. In the tensed s=
entences, no such simple rule exists.
+<p>
+<a name=3Ds24><h3>Tense questions: =93cu'e=94</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	cu'e	CUhE	tense question
+</pre><lx "ma">  XE "ma"   <cx "tense questions, methods of asking">  XE "=
tense questions: methods of asking"   <cx "tense questions with ma">  XE "t=
ense questions with ma"   <cx "ma, for tense questions">  XE "ma: for tense=
 questions"   There are two main ways to ask questions about tense. The mai=
n English tense question words are =93When?=94 and =93Where?=94. These may =
be paraphrased respectively as =93At what time?=94 and =93At what place?=94=
 In these forms, their Lojban equivalents simply involve a tense plus =93ma=
=94, the Lojban sumti question:
+<p>
+<ex "when">  XE "when: example=93  <pre><a name=3De24d1>24.1)	do klama le =
zdani ca ma
+	You go-to the house [present] [what sumti?].
+	You go to the house at what time?
+	When do you go to the house?
+</pre><ex "where">  XE "where: example=93  <pre><a name=3De24d2>24.2)	le v=
erba vi ma pu cadzu le bisli
+	The child [short space] [what sumti?] [past] walks-on the ice.
+	The child at/near what place walked on the ice?
+	Where did the child walk on the ice?
+</pre><cx "tense-or-modal questions, with cu'e">  XE "tense-or-modal quest=
ions: with cu'e"   <cx "modal-or-tense question, with cu'e">  XE "modal-or-=
tense question: with cu'e"   There is also a non-specific tense and modal q=
uestion, =93cu'e=94, belonging to selma'o CUhE. This can be used wherever a=
 tense or modal construct can be used.
+<p>
+<ex "when/where/how">  XE "when/where/how: example=93  <pre><a name=3De24d=
3>24.3)	le nanmu cu'e batci le gerku
+	The man [what tense?] bites the dog.
+	When/Where/How does the man bite the dog?
+</pre><cx "answers, to tense-or-modal questions">  XE "answers: to tense-o=
r-modal questions"   Possible answers to <a href=3D#e24d3>Example 24.3 </a>=
might be:
+<p>
+<pre><a name=3De24d4>24.4)	va
+	[medium space].
+	Some ways from here.
+<a name=3De24d5>24.5)	puzu
+	[past] [long time].
+	A long time ago.
+<a name=3De24d6>24.6)	vi le lunra
+	[short space] The moon.
+	On the moon.
+<a name=3De24d7>24.7)	pu'o
+	[inchoative]
+	He hasn't yet done so.
+</pre>or even the modal reply (from selma'o BAI; see <a href=3Dchap9.html>=
Chapter 9</a>):
+<p>
+<pre><a name=3De24d8>24.8)	seka=92a le briju
+	With-destination the office..
+</pre><cx "cu'e, combining with other tense cmavo">  XE "cu'e: combining w=
ith other tense cmavo"   <cx "tense-or-modal questions, pre-specifying some=
 information"><cx "modal-or-tense questions, pre-specifying some informatio=
n">The only way to combine =93cu'e=94 with other tense cmavo is through log=
ical connection, which makes a question that pre-specifies some information=
:
+<p>
+<ex "sowed grain">  XE "sowed grain: example=93  <ex "when else">  XE "whe=
n else: example=93  <pre><a name=3De24d9>24.9)	do puzi je cu'e sombo le gur=
ni
+	You [past] [short] and [when?] sow the grain?
+	You sowed the grain a little while ago; when else do you sow it?
+</pre><p>Additionally, the logical connective itself can be replaced by a =
question word:
+<p>
+<cx "tense questions, by using logical connective question">  XE "tense qu=
estions: by using logical connective question"   <pre><a name=3De24d10>24.1=
0)	la .artr. pu je'i ba nolraitru
+	Arthur [past] [which?] [future] is-a-king
+	Was Arthur a king or will he be?
+</pre><p>Answers to <a href=3D#e24d10>Example 24.10 </a>would be logical c=
onnectives such as =93je=94, meaning =93both=94, =93naje=94 meaning =93the =
latter=94, or =93jenai=94 meaning =93the former=94.
+<p>
+Explicit magnitudes</h3>
+<p><cx "explicit magnitude">  XE "explicit magnitude"   <cx "tense magnitu=
de">  XE "tense magnitude"   <cx "magnitude, tense">  XE "magnitude: tense"=
   <a name=3Ds25><h3>
+It is a limitation of the VA and ZI system of specifying magnitudes that t=
hey can only prescribe vague magnitudes: small, medium, or large. In order =
to express both an origin point and an exact distance, the Lojban construct=
ion called a =93termset=94 is employed. (Termsets are explained further in =
<a href=3Dchap14.html>Chapter 14 </a>and <a href=3Dchap16.html>Chapter 16</=
a>.) It is grammatical for a termset to be placed after a tense or modal ta=
g rather than a sumti, which allows both the origin of the imaginary journe=
y and its distance to be specified. Here is an example:
+
+<p>
+<pre><a name=3De25d1>25.1)	la frank. sanli zu'a nu'i la djordj.
+		lu'a lo mitre be li mu [nu'u]
+	Frank stands [left] [start termset] George
+		[quantity] a thing-measuring-in-meters the-number 5 [end termset].
+	Frank is standing five meters to the left of George.
+</pre><p>Here the termset extends from the =93nu'i=94 to the implicit =93n=
u'u=94 at the end of the sentence, and includes the terms =93la djordj.=94,=
 which is the unmarked origin point, and the tagged sumti =93lo mitre be li=
 mu=94, which the cmavo =93la'u=94 (of selma'o BAI, and meaning =93with qua=
ntity=94; see <a href=3Dchap9.html>Chapter 9</a>) marks as a quantity. Both=
 terms are governed by the tag =93zu'a=94.
+<p>
+It is not necessary to have both an origin point and an explicit magnitude=
: a termset may have only a single term in it. A less precise version of <a=
 href=3D#e25d1>Example 25.1 </a>is:
+<p>
+<pre><a name=3De25d2>25.2)	la frank. sanli zu'a nu'i lu'a
+		lo mitre be li mu
+	Frank stands [left] [termset] [quantity]=20
+		a thing-measuring-in-meters the-number 5.
+	Frank stands five meters to the left.
+</pre><a name=3Ds26><h3>Finally (an exercise for the much-tried reader)</h=
3>
+<p>
+<pre><a name=3De26d1>26.1)	a'o do pu seju ba roroi ca'o fe'e su'oroi jimpe
+		fi le lojbo temci selsku ciste
+
+</pre><a name=3Ds27><h3>Summary of tense selma'o</h3>
+<p>
+<cx "tense selma'o, summary of">  XE "tense selma'o: summary of"   <dl com=
pact><dt>PU	<dd>temporal direction
+		pu =3D past, ca =3D present, ba =3D future=20
+ZI	temporal distance
+		zi =3D short, za =3D medium, zu =3D long=20
+ZEhA	temporal interval
+		ze'i =3D short, ze'a =3D medium, ze'u =3D long, ze'e =3D infinite=20
+ROI	objective quantified tense flag
+		noroi =3D never, paroi =3D once,=A0<dots>=85</dots>, roroi =3D always, e=
tc.=20
+		pare=92u =3D the first time, rere=92u =3D the second time, etc.
+TAhE	subjective quantified tense
+		di'i =3D regularly, na'o =3D typically, ru'i =3D continuously, ta'e =3D =
habitually=20
+ZAhO	event contours
+		see <a href=3D#s10>Section 10 </a>
+<p><dt>FAhA	<dd>spatial direction
+		see <a href=3D#s27>Section 28=20
+</a>VA	spatial distance
+		vi =3D short, va =3D medium, vu =3D long=20
+VEhA	spatial interval
+		ve'i =3D short, ve'a =3D medium, ve'u =3D long, ve'e =3D infinite=20
+VIhA	spatial dimensionality
+		vi'i =3D line, vi'a =3D plane, vi'u =3D space, vi'e =3D space-time=20
+FEhE	spatial interval modifier flag
+		fe'enoroi =3D nowhere, fe'eroroi =3D everywhere, fe'eba'o =3D beyond, et=
c.=20
+MOhI	spatial movement flag
+		mo'i =3D motion; see <a href=3D#s27>Section 27 </a>
+<p><dt>KI	<dd>set or reset sticky tense
+		tense+=93ki=94 =3D set, =93ki=94 alone =3D reset=20
+CUhE	tense question, reference point
+		cu'e =3D asks for a tense or aspect, nau =3D use speaker's reference poi=
nt=20
+JAI	tense conversion
+		jaica =3D the time of, jaivi =3D the place of, etc.
+
+</dl><a name=3Ds28><h3>List of spatial directions and direction-like relat=
ions</h3>
+<p>
+<cx "spatial directions, list of">  XE "spatial directions: list of"   The=
 following list of FAhA cmavo gives rough English glosses for the cmavo, fi=
rst when used without =93mo'i=94 to express a direction, and then when used=
 with =93mo'i=94 to express movement in the direction. When possible, the g=
ismu from which the cmavo is derived is also listed.
+<p>
+<pre>cmavo	gismu	without mo'i	with mo'i
+ca'u	crane	in front (of)	forward=20
+ti'a	trixe	behind	backward=20
+zu'a	zunle	on the left (of)	leftward=20
+ri'u	pritu	on the right (of)	rightward=20
+ga'u	gapru	above	upward(ly)=20
+ni'a	cnita	below	downward(ly)=20
+ne'i	nenri	within	into=20
+ru'u	sruri	surrounding	orbiting=20
+pa'o	pagre	transfixing	passing through=20
+ne'a		next to	moving while next to=20
+te'e		bordering	moving along the border (of)
+re'o		adjacent (to)	along=20
+fa'a	farna	towards	arriving at=20
+to'o		away from	departing from=20
+zo'i		inward (from)	approaching=20
+ze'o		outward (from)	receding from=20
+zo'a		tangential (to)	passing (by)=20
+bu'u		coincident (with)	moving to coincide with=20
+be'a	berti	north (of)	northward(ly)=20
+ne'u	snanu	south (of)	southward(ly)=20
+du'a	stuna	east (of)	eastward(ly)=20
+vu'a		west (of)	westward(ly)
+</pre><cx "fa'a, special note on direction orientation">  XE "fa'a: specia=
l note on direction orientation"   <cx "to'o, special note on direction ori=
entation">  XE "to'o: special note on direction orientation"   <cx "zo'i, s=
pecial note on direction orientation">  XE "zo'i: special note on direction=
 orientation"   <cx "ze'o, special note on direction orientation">  XE "ze'=
o: special note on direction orientation"   Special note on =93fa'a=94, =93=
to'o=94, =93zo'i=94, and =93ze'o=94:
+<p>
+=93zo'i=94 and =93ze'o=94 refer to direction towards or away from the spea=
ker's location, or whatever the origin is.
+<p>
+=93fa'a=94 and =93to'o=94 refer to direction towards or away from some oth=
er point.
+<p>
+</body></html>
+<img src=3Dchap11.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+=20
+<h2>Chapter 11
+<br>
+Events, Qualities, Quantities, And Other Vague Words: On Lojban Abstractio=
n</h2>
+<h6>$Revision: 4.1 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+The syntax of abstraction</h3>
+<p><cx "abstraction, syntax">  XE "abstraction(s): syntax"   <cx "NU, synt=
ax">  XE "NU selma=92o: syntax"   <a name=3Ds1><h3>
+<cx "clause, subordinate using abstraction">  XE "clause: subordinate usin=
g abstraction"   <cx "gerund, using abstraction">  XE "gerund: using abstra=
ction"   The purpose of the feature of Lojban known as =93abstraction=94 is=
 to provide a means for taking whole bridi and packaging them up, as it wer=
e, into simple selbri. Syntactically, abstractions are very simple and unif=
orm; semantically, they are rich and complex, with few features in common b=
etween one variety of abstraction and another. We will begin by discussing =
syntax without regard to semantics; as a result, the notion of abstraction =
may seem unmotivated at first. Bear with this difficulty until <a href=3D#s=
2>Section 2</a>.
+<p>
+<lx "NU">  XE "NU selma=92o"   <lx "kei">  XE "kei"   <lx "KEI">  XE "KEI =
selma=92o"   An abstraction selbri is formed by taking a full bridi and pre=
ceding it by any cmavo of selma'o NU. There are twelve such cmavo; they are=
 known as =93abstractors=94. The bridi is closed by the elidable terminator=
 =93kei=94, of selma'o KEI. Thus, to change the bridi
+<p>
+<pre><a name=3De1d1>1.1)	mi klama le zarci
+	I go-to the store.
+</pre>into an abstraction using =93nu=94, one of the members of selma'o NU=
, we change it into
+<p>
+<cx "observatives, and abstractions">  XE "observatives: and abstractions"=
   <pre><a name=3De1d2>1.2)	nu mi klama le zarci [kei]
+	an-event-of my going-to the store
+</pre><cx "KEI, eliding">  XE "KEI selma=92o: eliding"   The bridi may be =
a simple selbri, or it may have associated sumti, as here. It is important =
to beware of eliding =93kei=94 improperly, as many of the common uses of ab=
straction selbri involve following them with words that would appear to be =
part of the abstraction if =93kei=94 had been elided.
+<p>
+(Technically, =93kei=94 is never necessary, because the elidable terminato=
r =93vau=94 that closes every bridi can substitute for it; however, =93kei=
=94 is specific to abstractions, and using it is almost always clearer.)
+<p>
+<cx "abstraction, grammatical uses">  XE "abstraction(s): grammatical uses=
"   <cx "tanru, and abstractions">  XE "tanru: and abstractions"   The gram=
matical uses of an abstraction selbri are exactly the same as those of a si=
mple brivla. In particular, abstraction selbri may be used as observatives,=
 as in <a href=3D#e1d2>Example 1.2</a>, or used in tanru:
+<p>
+<ex "want to be a soldier">  XE "want to be a soldier: example=93  <pre><a=
 name=3De1d3>1.3)	la djan. cu nu sonci kei djica
+	John is-an-(event-of being-a-soldier) type-of desirer.
+	John wants to be a soldier.
+</pre><cx "descriptions, and abstractions">  XE "descriptions: and abstrac=
tions"   Abstraction selbri may also be used in descriptions, preceded by =
=93le=94 (or any other member of selma'o LE):
+<p>
+<pre><a name=3De1d4>1.4)	la djan. cu djica le nu sonci [kei]
+	John desires the event-of being-a-soldier.
+</pre><p>We will most often use descriptions containing abstraction either=
 at the end of a bridi, or just before the main selbri with its =93cu=94; i=
n either of these circumstances, =93kei=94 can normally be elided.
+<p>
+<cx "abstractions, place structure">  XE "abstraction(s): place structure"=
   The place structure of an abstraction selbri depends on the particular a=
bstractor, and will be explained individually in the following sections.
+<p>
+Note: In glosses of bridi within abstractions, the grammatical form used i=
n the English changes. Thus, in the gloss of <a href=3D#e1d2>Example 1.2 </=
a>we see =93my going-to the store=94 rather than =93I go-to the store=94; l=
ikewise, in the glosses of <a href=3D#e1d3>Example 1.3 </a>and <a href=3D#e=
1d4>Example 1.4 </a>we see =93being-a-soldier=94 rather than =93is-a-soldie=
r=94. This procedure reflects the desire for more understandable glosses, a=
nd does not indicate any change in the Lojban form. A bridi is a bridi, and=
 undergoes no change when it is used as part of an abstraction selbri.
+<p>
+Event abstraction</h3>
+<p><cx "event abstractions">  XE "event abstractions"   <cx "abstractions,=
 event">  XE "abstraction(s): event"   <a name=3Ds2><h3>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	nu	NU	event abstractor
+</pre><cx "common abstractor">  XE "common abstractor"   <lx "nu">  XE "nu=
"   <lx "NU">  XE "NU selma=92o"   <cx "nu, definition">  XE "nu: definitio=
n"   <cx "le nu, definition">  XE "le nu: definition"   The examples in <a =
href=3D#s1>Section 1 </a>made use of =93nu=94 as the abstractor, and it is =
certainly the most common abstractor in Lojban text. Its purpose is to capt=
ure the event or state of the bridi considered as a whole. Do not confuse t=
he =93le=94 description built on a =93nu=94 abstraction with ordinary descr=
iptions based on =93le=94 alone. The following sumti are quite distinct:
+<p>
+<pre><a name=3De2d1>2.1)	le klama
+	the comer, that which comes
+<a name=3De2d2>2.2)	le se klama
+	the destination
+<a name=3De2d3>2.3)	le te klama
+	the origin
+<a name=3De2d4>2.4)	le ve klama
+	the route
+<a name=3De2d5>2.5)	le xe klama
+	the means of transportation
+<a name=3De2d6>2.6)	le nu klama
+	the event of someone coming to somewhere from somewhere by some route
+using some means
+</pre><a href=3D#e2d{1}>Examples 2.1 </a>through <a href=3D#e2d5>2.5 </a>a=
re descriptions that isolate the five individual sumti places of the selbri=
 =93klama=94. <a href=3D#e2d6>Example 2.6 </a>describes something associate=
d with the bridi as a whole: the event of it.
+<p>
+<cx "events, duration">  XE "events: duration"   In Lojban, the term =93ev=
ent=94 is divorced from its ordinary English sense of something that happen=
s over a short period of time. The description:
+<p>
+<pre><a name=3De2d7>2.7)	le nu mi vasxu
+	the event-of my breathing
+</pre>is an event which lasts for the whole of my life (under normal circu=
mstances). On the other hand,
+<p>
+  XE "kissing Jane, example"  <pre><a name=3De2d8>2.8)	le nu la djan. cinb=
a la djein.
+	the event-of John kissing Jane
+</pre><cx "normal circumstances">  XE "normal circumstances"   is relative=
ly brief by comparison (again, under normal circumstances).
+<p>
+<cx "abstraction, sumti ellipsis in">  XE "abstraction(s): sumti ellipsis =
in"   We can see from <a href=3D#e2d{6}>Examples 2.6 </a>through <a href=3D=
#e2d8>2.8 </a>that ellipsis of sumti is valid in the bridi of abstraction s=
elbri, just as in the main bridi of a sentence. Any sumti may be ellipsized=
 if the listener will be able to figure out from context what the proper va=
lue of it is, or else to recognize that the proper value is unimportant. It=
 is extremely common for =93nu=94 abstractions in descriptions to have the =
x1 place ellipsized:
+<p>
+<pre><a name=3De2d9>2.9)	mi nelci le nu limna
+	I like the event-of swimming.
+	I like swimming.
+</pre>is elliptical, and most probably means:
+<p>
+<pre><a name=3De2d10>2.10)	mi nelci le nu mi limna
+	I like the event-of I swim.
+</pre><p>In the proper context, of course, <a href=3D#e2d9>Example 2.9 </a=
>could refer to the event of somebody else swimming. Its English equivalent=
, =93I like swimming=94, can't be interpreted as =93I like Frank's swimming=
=94; this is a fundamental distinction between English and Lojban. In Lojba=
n, an omitted sumti can mean whatever the context indicates that it should =
mean.
+<p>
+<cx "abstraction, implicit in sumti">  XE "abstraction(s): implicit in sum=
ti"   Note that the lack of an explicit NU cmavo in a sumti can sometimes h=
ide an implicit abstraction. In the context of <a href=3D#e2d10>Example 2.1=
0</a>, the appearance of =93le se nelci=94 (=93that which is liked=94) is i=
n effect an abstraction:
+<p>
+<pre><a name=3De2d11>2.11)	le se nelci cu cafne
+	The liked-thing is-frequent.
+	The thing which I like happens often.
+</pre>which in this context means
+<p>
+<dl compact><dt>	<dd>My swimming happens often.
+</dl><p>Event descriptions with =93le nu=94 are commonly used to fill the =
=93under conditions=A0<dots>=85</dots>=94 places, among others, of gismu an=
d lujvo place structures:
+<p>
+<ex "under conditions">  XE "under conditions: example=93  <pre><a name=3D=
e2d12>2.12)	la lojban. cu frili mi le nu mi tadni [kei]
+	Lojban is-easy for-me under-conditions-the event-of I study
+	Lojban is easy for me when I study.
+</pre>(The =93when=94 of the English would also be appropriate for a const=
ruction involving a Lojban tense, but the Lojban sentence says more than th=
at the studying is concurrent with the ease.)
+<p>
+<cx "events, place structure">  XE "events: place structure"   <cx "nu, pl=
ace structure">  XE "nu: place structure"   The place structure of a =93nu=
=94 abstraction selbri is simply:
+<p>
+<dl compact><dt><dd>nu:  x1 is an event of (the bridi)
+Types of event abstractions</h3>
+<p></dl><cx "event abstractions, types">  XE "event abstraction(s): types"=
   <a name=3Ds3><h3>
+The following cmavo are discussed in this section:
+<p>
+<lx "mu'e">  XE "mu'e"   <lx "pu'u">  XE "pu'u"   <lx "zu'o">  XE "zu'o"  =
 <lx "za'i">  XE "za'i"   <lx "NU">  XE "NU selma=92o"   <pre> 	mu'e	NU	poi=
nt-event abstractor
+	pu'u	NU	process abstractor
+	zu'o	NU	activity abstractor
+	za'i	NU	state abstractor
+</pre><p>Event abstractions with =93nu=94 suffice to express all kinds of =
events, whether long, short, unique, repetitive, or whatever. Lojban also h=
as more finely discriminating machinery for talking about events, however. =
There are four other abstractors of selma'o NU for talking about four speci=
fic types of events, or four ways of looking at the same event.
+<p>
+<cx "point-event abstractor">  XE "point-event abstractor"   <cx "triumph"=
>  XE "triumph"   <cx "abstraction, point-event">  XE "abstraction(s): poin=
t-event"   <cx "abstraction, achievement">  XE "abstraction(s): achievement=
"   <cx "point-event abstraction, definition">  XE "point-event abstraction=
(s): definition"   <cx "achievement abstraction, definition">  XE "achievem=
ent abstraction(s): definition"   <lx "mu'e">  XE "mu'e"   An event conside=
red as a point in time is called a =93point-event=94, or sometimes an =93ac=
hievement=94. (This latter word should be divorced, in this context, from a=
ll connotations of success or triumph.) A point-event can be extended in du=
ration, but it is still a point-event if it is thought of as unitary, havin=
g no internal structure. The abstractor =93mu'e=94 means =93point-event-of=
=94:
+<p>
+<cx "killing Jim">  XE "killing Jim"   <ex "killing Jim">  XE "killing Jim=
: example=93  <pre><a name=3De3d1>3.1)	le mu'e la djan. catra la djim. cu z=
ekri
+	The point-event-of (John kills Jim) is-a-crime
+	John's killing Jim (considered as a point in time) is a crime.
+</pre><cx "process abstractor">  XE "process abstractor"   <cx "process ab=
straction, definition">  XE "process abstraction(s): definition"   <cx "abs=
traction, process">  XE "abstraction(s): process"   <lx "pu'u">  XE "pu'u" =
  An event considered as extended in time, and structured with a beginning,=
 a middle containing one or more stages, and an end, is called a =93process=
=94. The abstractor =93pu'u=94 means =93process-of=94:
+<p>
+<cx "Roman Empire">  XE "Roman Empire"   <ex "Roman Empire">  XE "Roman Em=
pire: example=93  <pre><a name=3De3d2>3.2)	ca'o le pu'u le latmo balje'a cu=
 porpi kei
+		so'i je'atru cu selcatra
+	[continuitive] The process-of-(the Latin great-state breaking-up)
+		many state-rulers were-killed.
+	During the fall of the Roman Empire, many Emperors were killed.
+</pre><cx "activity abstractor">  XE "activity abstractor"   <cx "activity=
 abstraction, definition">  XE "activity abstraction(s): definition"   <cx =
"abstraction, activity">  XE "abstraction(s): activity"   <lx "zu'o">  XE "=
zu'o"   An event considered as extended in time and cyclic or repetitive is=
 called an =93activity=94. The abstractor =93zu'o=94 means =93activity-of=
=94:
+<p>
+<pre><a name=3De3d3>3.3)	mi tatpi ri'a le zu'o mi plipe
+	I am-tired because-of the activity-of (I jump)
+	I am tired because I jump.
+</pre><cx "state abstractor">  XE "state abstractor"   <cx "state abstract=
ion, definition">  XE "state abstraction(s): definition"   <cx "abstraction=
, state">  XE "abstraction(s): state"   <lx "za'i">  XE "za'i"   An event c=
onsidered as something that is either happening or not happening, with shar=
p boundaries, is called a =93state=94. The abstractor =93za'i=94 means =93s=
tate-of=94:
+<p>
+<ex "being alive">  XE "being alive: example=93  <pre><a name=3De3d4>3.4)	=
le za'i mi jmive cu ckape do
+	The state-of (I am-alive) is-dangerous-to you
+	My being alive is dangerous to you.
+</pre><cx "event types, described">  XE "event types: described"   The abs=
tractors in <a href=3D#e3d{1}>Examples 3.1 </a>through <a href=3D#e3d4>3.4 =
</a>could all have been replaced by =93nu=94, with some loss of precision. =
Note that Lojban allows every sort of event to be viewed in any of these fo=
ur ways:
+<p>
+<cx "state event, described">  XE "state event: described"   <dl compact><=
dt><dd>the =93state of running=94 begins when the runner starts and ends wh=
en the runner stops;
+</dl><cx "activity event, described">  XE "activity event: described"   <d=
l compact><dt><dd>the =93activity of running=94 consists of the cycle =93li=
ft leg, step forward, drop leg, lift other leg=A0<dots>=85</dots>=94 (each =
such cycle is a process, but the activity consists in the repetition of the=
 cycle);
+</dl><cx "steady speed">  XE "steady speed"   <cx "slowdown">  XE "slowdow=
n"   <cx "process event, described">  XE "process event: described"   <dl c=
ompact><dt><dd>the =93process of running=94 puts emphasis on the initial sp=
rint, the steady speed, and the final slowdown;
+</dl><cx "Pheidippides">  XE "Pheidippides"   <cx "marathon">  XE "maratho=
n"   <cx "indivisible">  XE "indivisible"   <cx "Athens">  XE "Athens"   <c=
x "achievement event, described">  XE "achievement event: described"   <dl =
compact><dt><dd>the =93achievement of running=94 is most alien to English, =
but sees the event of running as a single indivisible thing, like =93Pheidi=
ppides' run from Marathon to Athens=94 (the original marathon).
+</dl><p>Further information on types of events can be found in <a href=3D#=
s12>Section 12</a>.
+
+<p>
+The four event type abstractors have the following place structures:
+<p>
+<cx "achievement, place structure">  XE "achievement abstraction: place st=
ructure"   <cx "point-event, place structure">  XE "point-event abstraction=
: place structure"   <cx "mu'e, place structure">  XE "mu'e: place structur=
e"   <dl compact><dt><dd>=93mu'e=94:  x1 is a point event of (the bridi)
+<cx "process, place structure">  XE "process abstraction: place structure"=
   <cx "pu'u, place structure">  XE "pu'u: place structure"   =93pu'u=94:  =
x1 is a process of (the bridi) with stages x2
+<cx "state, place structure">  XE "state abstraction: place structure"   <=
cx "za'i, place structure">  XE "za'i: place structure"   =93za'i=94:  x1 i=
s a continuous state of (the bridi) being true
+<cx "activity, place structure">  XE "activity abstraction: place structur=
e"   <cx "zu'o, place structure">  XE "zu'o: place structure"   =93zu'o=94:=
  x1 is an activity of (the bridi) consisting of repeated actions x2
+Property abstractions</h3>
+</dl><cx "property abstractions">  XE "property abstraction"   <cx "abstra=
ctions, property">  XE "abstraction(s): property"   <a name=3Ds4><h3><p>
+The following cmavo are discussed in this section:=20
+  XE "ka"    XE "NU"    XE "property abstractor"  <pre>	ka	NU	property abs=
tractor
+	ce'u	KOhA	abstraction focus
+</pre><p>The things described by =93le nu=94 descriptions (or, to put it a=
nother way, the things of which =93nu=94 selbri may correctly be predicated=
) are only moderately =93abstract=94. They are still closely tied to happen=
ings in space and time. Properties, however, are much more ethereal. What i=
s =93the property of being blue=94, or =93the property of being a go-er=94?=
 They are what logicians call =93intensions=94. If John has a heart, then =
=93the property of having a heart=94 is an abstract object which, when appl=
ied to John, is true. In fact,
+<p>
+<ex "has a heart">  XE "has a heart: example=93  <pre><a name=3De4d1>4.1)	=
la djan. cu se risna zo'e
+	John has-as-heart something-unspecified.
+	John has a heart.
+</pre>has the same truth conditions as
+<p>
+<pre><a name=3De4d2>4.2)	la djan. cu ckaji le ka se risna [zo'e] [kei]
+	John has-the-property the property-of having-as-heart something.
+	John has the property of having a heart.
+</pre><cx "having, of properties">  XE "having: of properties"   (The Engl=
ish word =93have=94 frequently appears in any discussion of Lojban properti=
es: things are said to =93have=94 properties, but this is not the same sens=
e of =93have=94 as in =93I have money=94, which is possession.)
+<p>
+Property descriptions, like event descriptions, are often wanted to fill p=
laces in brivla place structures:
+<p>
+<pre><a name=3De4d3>4.3)	do cnino mi le ka xunre [kei]
+	You are-new to-me in-the-quality-of-the property-of being-red.
+	You are new to me in redness.
+</pre><cx "-ness">  XE "-ness"   <cx "-ity">  XE "-ity"   (The English suf=
fix =93-ness=94 often signals a property abstraction, as does the suffix =
=93-ity=94.)
+<p>
+<cx "property description">  XE "property description"   We can also move =
the property description to the x1 place of <a href=3D#e4d3>Example 4.3</a>=
, producing:
+<p>
+<pre><a name=3De4d4>4.4)	le ka do xunre [kei] cu cnino mi
+	The property-of your being-red is-new to me.
+	Your redness is new to me.
+</pre><cx "sunburn">  XE "sunburn: example"   <cx "beach">  XE "beach: exa=
mple"   It would be suitable to use <a href=3D#e4d{3}>Examples 4.3 </a>and =
<a href=3D#e4d4>4.4 </a>to someone who has returned from the beach with a s=
unburn.
+<p>
+<cx "property abstraction, sumti ellipsis in">  XE "property abstraction(s=
): sumti ellipsis in"   <cx "property abstraction, specifying determining p=
lace by sumti ellipsis">  XE "property abstraction(s): specifying determini=
ng place by sumti ellipsis"   There are several different properties that c=
an be extracted from a bridi, depending on which place of the bridi is =93u=
nderstood=94 as being specified externally. Thus:
+
+<p>
+<ex "property of loving">  XE "property of loving: example=93  <pre><a nam=
e=3De4d5>4.5)	ka mi prami [zo'e] [kei]
+	a-property-of me loving something-unspecified
+</pre>is quite different from
+<p>
+<pre><a name=3De4d6>4.6)	ka [zo'e] prami mi [kei]
+	a-property-of something-unspecified loving me
+</pre><p>In particular, sentences like <a href=3D#e4d7>Example 4.7 </a>and=
 <a href=3D#e4d8>Example 4.8 </a>are quite different in meaning:
+<p>
+<ex "love more">  XE "love more: example=93  <pre><a name=3De4d7>4.7)	la d=
jan. cu zmadu la djordj. le ka mi prami
+	John exceeds George in-the property-of (I love X)
+	I love John more than I love George.
+<a name=3De4d8>4.8)	la djan. cu zmadu la djordj. le ka prami mi
+	John exceeds George in the property of (X loves me).
+	John loves me more than George loves me.
+</pre><lx "ce'u">  XE "ce'u"   <lx "KOhA">  XE "KOhA selma=92o"   <p>
+<cx "property abstraction, specifying determining place with ce'u">  XE "p=
roperty abstraction(s): specifying determining place with ce'u"   The =93X=
=94 used in the glosses of <a href=3D#e4d{7}>Examples 4.7 </a>through <a hr=
ef=3D#e4d8>4.8 </a>as a place-holder cannot be represented only by ellipsis=
 in Lojban, because ellipsis means that there must be a specific value that=
 can fill the ellipsis, as mentioned in <a href=3D#s2>Section 2</a>. Instea=
d, the cmavo =93ce'u=94 of selma'o KOhA is employed when an explicit sumti =
is wanted. (The form =93X=94 will be used in literal translations.)
+<p>
+Therefore, an explicit equivalent of <a href=3D#e4d7>Example 4.7</a>, with=
 no ellipsis, is:
+<p>
+<pre><a name=3De4d9>4.9)	la djan. cu zmadu la djordj. le ka mi prami ce'u
+	John exceeds George in-the property-of (I love X).
+</pre>and of <a href=3D#e4d8>Example 4.8 </a>is:
+<p>
+<pre><a name=3De4d10>4.10)	la djan. cu zmadu la djordj. le ka ce'u prami m=
i
+	John exceeds George in-the property-of (X loves me).
+</pre><p>This convention allows disambiguation of cases like:
+<p>
+<ex "giving the horse">  XE "giving the horse: example=93  <pre><a name=3D=
e4d11>4.11)	le ka [zo'e] dunda le xirma [zo'e] [kei]
+	the property-of giving the horse
+</pre>into
+<p>
+<pre><a name=3De4d12>4.12)	le ka ce'u dunda le xirma [zo'e] [kei]
+	the property-of (X is-a-giver of-the horse to someone-unspecified)
+	the property of being a giver of the horse
+</pre>which is the most natural interpretation of <a href=3D#e4d11>Example=
 4.11</a>, versus
+<p>
+<pre><a name=3De4d13>4.13)	le ka [zo'e] dunda le xirma ce'u [kei]
+	the property-of (someone-unspecified is-a-giver of-the horse to X)
+	the property of being one to whom the horse is given
+</pre>which is also a possible interpretation.
+<p>
+<cx "relationship abstraction">  XE "relationship abstraction"   <cx "prop=
erty abstractions for relationships">  XE "property abstraction: use of mul=
tiple ce=92u for relationship abstraction"   It is also possible to have mo=
re than one =93ce'u=94 in a =93ka=94 abstraction, which transforms it from =
a property abstraction into a relationship abstraction.  Relationship abstr=
actions =93package up=94 a complex relationship for future use; such an abs=
traction can be translated back into a selbri by placing it in the x2 place=
 of the selbri =93bridi=94, whose place structure is:
+=93bridi=94: x1 is a predicate relationship with relation x2 (abstraction)
+among arguments (sequence/set) x3
+<p>
+<cx "properties, place structure">  XE "properties: place structure"   The=
 place structure of =93ka=94 abstraction selbri is simply:
+<p>
+<dl compact><dt>ka:  <dd>x1 is a property of (the bridi)
+Amount abstractions</h3>
+<p></dl><cx "amount abstraction">  XE "amount abstraction"   <cx "abstract=
ion, amount">  XE "abstraction(s): amount"   <a name=3Ds5><h3>
+The following cmavo is discussed in this section:
+<p>
+  XE "ni"    XE "NU"    XE "amount abstraction"  <pre> 	ni	NU	amount abstr=
action
+</pre><p>Amount abstractions are far more limited than event or property a=
bstractions. They really make sense only if the selbri of the abstracted br=
idi is subject to measurement of some sort. Thus we can speak of:
+<p>
+<ex "blueness">  XE "blueness: example=93  <pre><a name=3De5d1>5.1)	le ni =
le pixra cu blanu [kei]
+	the amount-of (the picture being-blue)
+	the amount of blueness in the picture
+</pre><cx "colorimeter">  XE "colorimeter"   because =93blueness=94 could =
be measured with a colorimeter or a similar device. However,
+<p>
+<pre><a name=3De5d2>5.2)	le ni la djein. cu mamta [kei]
+	the amount-of (Jane being-a-mother)
+	the amount of Jane's mother-ness (?)
+	the amount of mother-ness in Jane (?)
+</pre><cx "measurement scale">  XE "measurement scale"   makes very little=
 sense in either Lojban or English. We simply do not have any sort of measu=
rement scale for being a mother.
+<p>
+<cx "amount abstraction and mathematics">  XE "amount abstraction and math=
ematics"   Semantically, a sumti with =93le ni=94 is a number; however, it =
cannot be treated grammatically as a quantifier in Lojban unless prefixed b=
y the mathematical cmavo =93mo'e=94:
+<p>
+<pre><a name=3De5d3>5.3)	li pa vu'u mo'e le ni le pixra cu blanu [kei]
+	The-number 1 minus the-operand the amount-of (the picture being-blue)
+	1 - B, where B =3D blueness of the picture
+</pre><p>Mathematical Lojban is beyond the scope of this chapter, and is e=
xplained more fully in <a href=3Dchap18.html>Chapter 18</a>.
+<p>
+<cx "amount abstraction, specifying determining place with ce'u">  XE "amo=
unt abstraction(s): specifying determining place with ce'u"   <lx "ce'u">  =
XE "ce'u"   <cx "abstraction, amount contrasted with property">  XE "abstra=
ction(s): amount contrasted with property"   <cx "property abstraction, con=
trasted with amount abstraction">  XE "property abstraction(s): contrasted =
with amount abstraction"   There are contexts where either property or amou=
nt abstractions make sense, and in such constructions, amount abstractions =
can make use of =93ce'u=94 just like property abstractors. Thus,
+<p>
+<ex "blueness varying">  XE "blueness varying: example=93  <pre><a name=3D=
e5d4>5.4)	le pixra cu cenba le ka ce'u blanu [kei]
+	The picture varies in-the property-of (X is blue)
+	The picture varies in being blue.
+	The picture varies in blueness.
+</pre>is not the same as
+<p>
+<pre><a name=3De5d5>5.5)	le pixra cu cenba le ni ce'u blanu [kei]
+	The picture varies in-the amount-of (X is blue)
+	The picture varies in how blue it is.
+	The picture varies in blueness.
+</pre><a href=3D#e5d4>Example 5.4 </a>conveys that the blueness comes and =
goes, whereas <a href=3D#e5d5>Example 5.5 </a>conveys that its quantity cha=
nges over time.
+<p>
+<cx "amount abstraction, scale">  XE "amount abstraction(s): scale"   <cx =
"amount abstraction, place structure">  XE "amount abstraction(s): place st=
ructure"   Whenever we talk of measurement of an amount, there is some sort=
 of scale, and so the place structure of =93ni=94 abstraction selbri is:
+<p>
+<dl compact><dt>ni:  <dd>x1 is the amount of (the bridi) on scale x2
+</dl><lx "ni">  XE "ni"   <lx "kei">  XE "kei"   <lx "be">  XE "be"   Note=
: the best way to express the x2 places of abstract sumti is to use somethi=
ng like =93le ni=A0<dots>=85</dots>kei be=94. See <a href=3D#e9d5>Example 9=
.5 </a>for the use of this construction.
+<p>
+Truth-value abstraction: =93jei=94</h3>
+<p><cx "value abstraction">  XE "value abstraction"   <cx "truth-value abs=
traction">  XE "truth-value abstraction"   <cx "abstraction, truth-value"> =
 XE "abstraction(s): truth-value"   <a name=3Ds6><h3>
+<lx "jei">  XE "jei"   <lx "NU">  XE "NU selma=92o"   <cx "abstraction, tr=
uth-value  contrasted with amount">  XE "abstraction(s): truth-value  contr=
asted with amount"   The =93blueness of the picture=94 discussed in <a href=
=3D#s5>Section 5 </a>refers to the measurable amount of blue pigment (or ot=
her source of blueness), not to the degree of truth of the claim that bluen=
ess is present. That abstraction is expressed in Lojban using =93jei=94, wh=
ich is closely related semantically to =93ni=94. In the simplest cases, =93=
le jei=94 produces not a number but a truth value:
+<p>
+<ex "2 + 2">  XE "2 + 2: example=93  <pre><a name=3De6d1>6.1)	le jei li re=
 su'i re du li vo [kei]
+	the truth-value-of the-number 2 + 2 =3D the-number 4
+	the truth of 2 + 2 being 4
+</pre>is equivalent to =93truth=94, and
+<p>
+<pre><a name=3De6d2>6.2)	le jei li re su'i re du li mu [kei]
+	the truth-value-of the-number 2 + 2 =3D the-number 5
+	the truth of 2 + 2 being 5
+</pre>is equivalent to =93falsehood=94.
+<p>
+However, not everything in life (or even in Lojban) is simply true or fals=
e. There are shades of gray even in truth value, and =93jei=94 is Lojban's =
mechanism for indicating the shade of grey intended:
+<p>
+<ex "whether criminal">  XE "whether criminal: example=93  <pre><a name=3D=
e6d3>6.3)	mi ba jdice le jei la djordj. cu zekri gasnu [kei]
+	I [future] decide the truth-value of (George being-a-(crime doer)).
+	I will decide whether George is a criminal.
+</pre><cx "legal system">  XE "legal system"   <cx "truth-value abstractio=
n, place structure">  XE "truth-value abstraction(s): place structure"   <c=
x "jei, place structure">  XE "jei: place structure"   <a href=3D#e6d3>Exam=
ple 6.3 </a>does not imply that George is, or is not, definitely a criminal=
. Depending on the legal system I am using, I may make some intermediate de=
cision. As a result, =93jei=94 requires an x2 place analogous to that of =
=93ni=94:
+<p>
+<dl compact><dt>jei:  <dd>x1 is the truth value of (the bridi) under epist=
emology x2
+</dl><cx "abstraction, truth-value and fuzzy logic">  XE "abstraction(s): =
truth-value and fuzzy logic"   <cx "fuzzy logic and truth-value abstraction=
">  XE "fuzzy logic and truth-value abstraction"   Abstractions using =93je=
i=94 are the mechanism for fuzzy logic in Lojban; the =93jei=94 abstraction=
 refers to a number between 0 and 1 inclusive (as distinct from =93ni=94 ab=
stractions, which are often on open-ended scales). The detailed conventions=
 for using =93jei=94 in fuzzy-logic contexts have not yet been established.
+<p>
+Predication/sentence abstraction</h3>
+<p><cx "predication abstraction">  XE "predication abstraction"   <cx "sen=
tence abstraction">  XE "sentence abstraction"   <cx "abstraction, predicat=
ion/sentence">  XE "abstraction(s): predication/sentence"   <a name=3Ds7><h=
3>
+The following cmavo is discussed in this section:
+<p>
+<pre>	du'u	NU	predication abstraction
+</pre><cx "propositional attitudes">  XE "propositional attitudes"   <cx "=
abstraction, with knowing, believing, etc.">  XE "abstraction(s): with know=
ing, believing, etc."   <cx "abstraction, mental activity">  XE "abstractio=
n(s): mental activity"   There are some selbri which demand an entire predi=
cation as a sumti; they make claims about some predication considered as a =
whole. Logicians call these the =93propositional attitudes=94, and they inc=
lude (in English) things like knowing, believing, learning, seeing, hearing=
, and the like. Consider the English sentence:
+<p>
+<ex "know">  XE "know: example=93  <ex "Frank is a fool">  XE "Frank is a =
fool: example=93  <pre><a name=3De7d1>7.1)	I know that Frank is a fool.
+</pre><p>How's that in Lojban? Let us try:
+<p>
+<pre><a name=3De7d2>7.2)	mi djuno le nu la frank. cu bebna [kei]
+	I know the event of Frank being a fool.
+</pre><cx "mental activity">  XE "mental activity"   Not quite right. Even=
ts are actually or potentially physical, and can't be contained inside one'=
s mind, except for events of thinking, feeling, and the like; <a href=3D#e7=
d2>Example 7.2 </a>comes close to claiming that Frank's being-a-fool is pur=
ely a mental activity on the part of the speaker. (In fact, <a href=3D#e7d2=
>Example 7.2 </a>is an instance of improperly marked =93sumti raising=94, a=
 concept discussed further in <a href=3D#s10>Section 10</a>).
+<p>
+Try again:
+<p>
+<pre><a name=3De7d3>7.3)	mi djuno le jei la frank. cu bebna [kei]
+	I know the truth-value of Frank being a fool.
+</pre><p>Closer. <a href=3D#e7d3>Example 7.3 </a>says that I know whether =
or not Frank is a fool, but doesn't say that he is one, as <a href=3D#e7d1>=
Example 7.1 </a>does. To catch that nuance, we must say:
+<p>
+<lx "du'u">  XE "du'u"   <lx "NU">  XE "NU selma=92o"   <pre><a name=3De7d=
4>7.4)	mi djuno le du'u la frank. cu bebna [kei]
+	I know the predication that Frank is a fool.
+</pre><p>Now we have it. Note that the implied assertion =93Frank is a foo=
l=94 is not a property of =93le du'u=94 abstraction, but of =93djuno=94; we=
 can only know what is in fact true. (As a result, =93djuno=94 like =93jei=
=94 has a place for epistemology, which specifies how we know.) <a href=3D#=
e7d5>Example 7.5 </a>has no such implied assertion:
+<p>
+<cx "curious">  XE "curious"   <ex "curious">  XE "curious: example=93  <e=
x "Frank is a fool">  XE "Frank is a fool: example=93  <pre><a name=3De7d5>=
7.5)	mi kucli le du'u la frank. cu bebna [kei]
+	I am curious about whether Frank is a fool.
+</pre>and here =93du'u=94 could probably be replaced by =93jei=94 without =
much change in meaning:
+<p>
+  XE "curious:example"  <pre><a name=3De7d6>7.6)	mi kucli le jei la frank.=
 cu bebna [kei]
+	I am curious about how true it is that Frank is a fool.
+</pre><cx "truth-value abstraction, place structure">  XE "truth-value abs=
traction(s): place structure"   As a matter of convenience rather than logi=
cal necessity, =93du'u=94 has been given an x2 place, which is a sentence (=
piece of language) expressing the bridi:
+<p>
+<dl compact><dt>du=92u:  <dd>x1 is the predication (the bridi) expressed i=
n sentence x2
+</dl><cx "linguistic behavior">  XE "linguistic behavior"   <cx "se du'u">=
  XE "se du'u"   <cx "abstraction, speaking, writing, etc.">  XE "abstracti=
on(s): speaking, writing, etc."   and =93le se du'u=A0<dots>=85</dots>=94 i=
s very useful in filling places of selbri which refer to speaking, writing,=
 or other linguistic behavior regarding bridi:
+<p>
+<cx "abstraction of sentences, contrasted with quotation">  XE "abstractio=
n(s): sentence, contrasted with quotation"   <cx "quotation, contrasted wit=
h sentence abstraction">  XE "quotation: contrasted with sentence abstracti=
on"   <pre><a name=3De7d7>7.7)	la djan. cusku le se du'u la djordj. klama l=
e zarci [kei]
+	John expresses the sentence-expressing-that George goes-to the store
+	John says that George goes to the store.
+</pre><a href=3D#e7d7>Example 7.7 </a>differs from
+<p>
+<pre><a name=3De7d8>7.8)	la djan cusku lu la djordj. klama le zarci li'u
+	John expresses, quote, George goes to the store, unquote.
+	John says =93George goes to the store=94.
+</pre>because <a href=3D#e7d8>Example 7.8 </a>claims that John actually sa=
id the quoted words, whereas <a href=3D#e7d7>Example 7.7 </a>claims only th=
at he said some words or other which were to the same purpose.
+<p>
+<lx "lu'e">  XE "lu'e"   =93le se du'u=94 is much the same as =93lu'e le d=
u'u=94, a symbol for the predication, but =93se du'u=94 can be used as a se=
lbri, whereas =93lu'e=94 is ungrammatical in a selbri. (See <a href=3Dchap5=
.html>Chapter 5 </a>for a discussion of =93lu'e=94.)
+<p>
+Indirect questions</h3>
+<p><cx "indirect questions">  XE "indirect questions"   <cx "questions, in=
direct">  XE "questions: indirect"   <a name=3Ds8><h3>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	kau	UI	indirect question marker
+</pre><lx "du'u">  XE "du'u"   There is an alternative type of sentence in=
volving =93du'u=94 and a selbri expressing a propositional attitude. In add=
ition to sentences like
+<p>
+<pre><a name=3De8d1>8.1)	I know that John went to the store.
+</pre>we can also say things like
+<p>
+<ex "know who">  XE "know who: example=93  <pre><a name=3De8d2>8.2)	I know=
 who went to the store.
+</pre><cx "abstraction, with wonder, doubt, etc.">  XE "abstraction(s): wi=
th wonder, doubt, etc."   This form is called an =93indirect question=94 in=
 English because the embedded English sentence is a question: =93Who went t=
o the store?=94 A person who says <a href=3D#e8d2>Example 8.2 </a>is claimi=
ng to know the answer to this question. Indirect questions can occur with m=
any other English verbs as well: I can wonder, or doubt, or see, or hear, a=
s well as know who went to the store.
+<p>
+<lx "kau">  XE "kau"   <lx "UI">  XE "UI selma=92o"   To express indirect =
questions in Lojban, we use a =93le du'u=94 abstraction, but rather than us=
ing a question word like =93who=94 (=93ma=94 in Lojban), we use any word th=
at will fit grammatically and mark it with the suffix particle =93kau=94. T=
his cmavo belongs to selma'o UI, so grammatically it can appear anywhere. T=
he simplest Lojban translation of <a href=3D#e8d2>Example 8.2 </a>is theref=
ore:
+<p>
+<pre><a name=3De8d3>8.3)	mi djuno le du'u makau pu klama le zarci
+	I know the predication-of X [indirect question] [past] going to the store=
.
+<cx "kau, =93ma kau=94 contrasted with =93la djan. kau=94">  XE "kau: ma k=
au, contrasted with la djan. kau"   <cx "Indirect questions, =93ma kau=94 c=
ontrasted with =93la djan. kau=94">  XE "indirect questions: \=93ma kau\=94=
 contrasted with \=93la djan. kau\=94"   <cx "know who, contrasted with kno=
w that">  XE "know who: contrasted with know that"   In <a href=3D#e8d3>Exa=
mple 8.3</a>, we have chosen to use =93ma=94 as the word marked by =93kau=
=94. In fact, any other sumti would have done as well: =93zo'e=94 or =93da=
=94 or even =93la djan.=94. Using =93la djan.=94 would suggest that it was =
John who I knew had gone to the store, however:
+<p>
+<pre><a name=3De8d4>8.4)	mi djuno le du'u la djan. kau pu klama=20
+le zarci
+	I know the predication-of/fact-that John [indirect question] [past] going=
=20
+to the store.
+	I know who went to the store, namely John.
+	I know that it was John who went to the store.
+</pre><p>Using one of the indefinite pro-sumti such as =93ma=94, =93zo'e=
=94, or =93da=94 does not suggest any particular value.
+<p>
+Why does Lojban require the =93kau=94 marker, rather than using =93ma=94 a=
s English and Chinese and many other languages do? Because =93ma=94 always =
signals a direct question, and so
+<p>
+<pre><a name=3De8d5>8.5)	mi djuno le du'u ma pu klama le zarci
+	I know the predication-of [what sumti?] [past] goes-to the store
+</pre>means
+<p>
+<pre><a name=3De8d6>8.6)	Who is it that I know goes to the store?
+</pre><cx "indirect question involving sumti">  XE "indirect question invo=
lving sumti"   <cx "indirect questions without =93kau=94">  XE "indirect qu=
estions without \=93kau\=94"   It is actually not necessary to use =93le du=
'u=94 and =93kau=94 at all if the indirect question involves a sumti; there=
 is generally a paraphrase of the type:
+<p>
+<pre><a name=3De8d7>8.7)	mi djuno fi le pu klama be le zarci
+	I know about the [past] goer to-the store.
+	I know something about the one who went to the store (namely, his
+identity).
+</pre>because the x3 place of =93djuno=94 is the subject of knowledge, as =
opposed to the fact that is known. But when the questioned point is not a s=
umti, but (say) a logical connection, then there is no good alternative to =
=93kau=94:
+<p>
+<pre><a name=3De8d8>8.8)	mi ba zgana le du'u la djan.=20
+jikau la djordj. cu zvati le panka
+	I [future] observe the predication-of/fact-that John=20
+[connective indirect question] George is-at the park.
+	I will see whether John or George (or both) is at the park.
+</pre><p>In addition, <a href=3D#e8d7>Example 8.7 </a>is only a loose para=
phrase of <a href=3D#e8d3>Example 8.3</a>, because it is left to the listen=
er's insight to realize that what is known about the goer-to-the-store is h=
is identity rather than some other of his attributes.
+<p>
+Minor abstraction types</h3>
+<p><cx "abstraction types">  XE "abstraction(s): types"   <a name=3Ds9><h3=
>
+The following cmavo are discussed in this section:
+<p>
+<pre>	li'i	NU	experience abstractor
+	si'o	NU	concept abstractor
+	su'u	NU	general abstractor
+</pre><cx "experience abstractor">  XE "experience abstractor"   <cx "expe=
rience abstraction">  XE "experience abstraction"   <cx "abstraction, exper=
ience">  XE "abstraction(s): experience"   <lx "li'i">  XE "li'i"   There a=
re three more abstractors in Lojban, all of them little used so far. The ab=
stractor =93li'i=94 expresses experience:
+<p>
+<pre><a name=3De9d1>9.1)	mi morji le li'i mi verba
+	I remember the experience-of (my being-a-child)
+</pre><cx "concept abstractor">  XE "concept abstractor"   <cx "idea abstr=
action">  XE "idea abstraction"   <cx "concept abstraction">  XE "concept a=
bstraction"   <cx "abstraction, idea">  XE "abstraction(s): idea"   <cx "ab=
straction, concept">  XE "abstraction(s): concept"   <lx "si'o">  XE "si'o"=
   The abstractor =93si'o=94 expresses a mental image, a concept, an idea:
+<p>
+<pre><a name=3De9d2>9.2)	mi nelci le si'o la lojban. cu mulno
+	I enjoy the concept-of Lojban being-complete.
+</pre><cx "vague abstractor">  XE "vague abstractor"   <cx "vague abstract=
ion">  XE "vague abstraction"   <cx "abstraction, vague">  XE "abstraction(=
s): vague"   <lx "su'u">  XE "su'u"   Finally, the abstractor =93su'u=94 is=
 a vague abstractor, whose meaning must be grasped from context:
+<p>
+  XE "mice: example"  <pre><a name=3De9d3>9.3)	ko zgana le su'u le ci smac=
u cu bajra
+	You [imperative] observe the abstract-nature-of the three mice running
+	See how the three mice run!
+</pre><cx "experience abstraction, place structure">  XE "experience abstr=
action(s): place structure"   All three of these abstractors have an x2 pla=
ce. An experience requires an experiencer, so the place structure of =93li'=
i=94 is:
+<p>
+<dl compact><dt>li=92i:  <dd>x1 is the experience of (the bridi) as experi=
enced by x2
+</dl><cx "concept abstraction, place structure">  XE "concept abstraction(=
s): place structure"   <cx "idea abstraction, place structure">  XE "idea a=
bstraction(s): place structure"   Similarly, an idea requires a mind to hol=
d it, so the place structure of =93si'o=94 is:
+<p>
+<dl compact><dt>si=92o:  <dd>x1 is the idea/concept of (the bridi) in the =
mind of x2
+</dl><cx "vague abstraction, place structure">  XE "vague abstraction(s): =
place structure"   Finally, there needs to be some way of specifying just w=
hat sort of abstraction =93su'u=94 is representing, so its place structure =
is:
+<p>
+<dl compact><dt>su=92u:  <dd>x1 is an abstract nature of (the bridi) of ty=
pe x2
+</dl><cx "template">  XE "template"   <cx "abstraction, creating new types=
">  XE "abstraction(s): creating new types"   The x2 place of =93su'u=94 al=
lows it to serve as a substitute for any of the other abstractors, or as a =
template for creating new ones. For example,
+<p>
+<pre><a name=3De9d4>9.4)	le nu mi klama
+	the event-of my going
+</pre>can be paraphrased as
+<p>
+<pre><a name=3De9d5>9.5)	le su'u mi klama kei be lo fasnu
+	the abstract-nature-of (my going) of-type an event
+</pre>and there is a book whose title might be rendered in Lojban as:
+<p>
+<cx "Jesus">  XE "Jesus"   <cx "intersect">  XE "intersect"   <ex "Jesus">=
  XE "Jesus: example=93  <ex "bicycle race">  XE "bicycle race: example=93 =
 <pre><a name=3De9d6>9.6)	le su'u la .iecuas. kuctra selcatra kei
+		be lo sao'rdzifa'a=20
+ke nalmatma'e sutyterjvi
+	The abstract-nature-of (Jesus is-an-intersect-shape type-of-killed-one)=
=20
+of-type a slope-low-direction=20
+type-of non-motor-vehicle speed-competition
+	The Crucifixion of Jesus Considered As A Downhill Bicycle Race
+</pre><p>Note the importance of using =93kei=94 after =93su'u=94 when the =
x2 of =93su'u=94 (or any other abstractor) is being specified; otherwise, t=
he =93be lo=94 ends up inside the abstraction bridi.
+<p>
+Lojban sumti raising</h3>
+<p><cx "sumti raising">  XE "sumti raising"   <cx "abstraction, simplifica=
tion to sumti">  XE "abstraction(s): simplification to sumti"   <a name=3Ds=
10><h3>
+The following cmavo are discussed in this section:
+<p>
+  XE "abstraction conversion"  <pre>	tu'a	LAhE	an abstraction involving
+	jai	JAI	abstraction conversion
+</pre><cx "abstract description">  XE "abstract description"  It is someti=
mes inconvenient, in a situation where an abstract description is logically=
 required, to express the abstraction. In English we can say:
+<p>
+<pre><a name=3De10d1>10.1)	I try to open the door.
+</pre>which in Lojban is:
+<p>
+<pre><a name=3De10d2>10.2)	mi troci le nu [mi] gasnu le nu le vorme cu kar=
bi'o
+	I try the event-of (I am-agent-in the event-of (the door open-becomes)).
+</pre>which has an abstract description within an abstract description, qu=
ite a complex structure. In English (but not in all other languages), we ma=
y also say:
+<p>
+<cx "abstraction, simplification to sumti with tu'a">  XE "abstraction(s):=
 simplification to sumti with tu'a"   <lx "tu'a">  XE "tu'a"   <lx "LAhE"> =
 XE "LAhE selma=92o"   <ex "try the door">  XE "try the door: example=93  <=
pre><a name=3De10d3>10.3)	I try the door.
+</pre>where it is understood that what I try is actually not the door itse=
lf, but the act of opening it. The same simplification can be done in Lojba=
n, but it must be marked explicitly using a cmavo. The relevant cmavo is =
=93tu'a=94, which belongs to selma'o LAhE. The Lojban equivalent of <a href=
=3D#e10d3>Example 10.3 </a>is:
+<p>
+<pre><a name=3De10d4>10.4)	mi troci tu'a le vorme
+	I try some-action-to-do-with the door.
+</pre><cx "intermediate abstraction">  XE "intermediate abstraction"   The=
 term =93sumti-raising=94, as in the title of this section, signifies that =
a sumti which logically belongs within an abstraction (or even within an ab=
straction which is itself inside an intermediate abstraction) is =93raised=
=94 to the main bridi level. This transformation from <a href=3D#e10d2>Exam=
ple 10.2 </a>to <a href=3D#e10d4>Example 10.4 </a>loses information: nothin=
g except convention tells us what the abstraction was.
+<p>
+Using =93tu'a=94 is a kind of laziness: it makes speaking easier at the po=
ssible expense of clarity for the listener. The speaker must be prepared fo=
r the listener to respond something like:
+<p>
+<lx "lu'u">  XE "lu'u"   <lx "LUhU">  XE "LUhU selma=92o"   <pre><a name=
=3De10d5>10.5)	tu'a le vorme lu'u ki'a
+	something-to-do-with the door [terminator] [confusion!]
+</pre>which indicates that =93tu'a le vorme=94 cannot be understood. (The =
terminator for =93tu'a=94 is =93lu'u=94, and is used in <a href=3D#e10d5>Ex=
ample 10.5 </a>to make clear just what is being questioned: the sumti-raisi=
ng, rather than the word =93vorme=94 as such.) An example of a confusing ra=
ised sumti might be:
+<p>
+<pre><a name=3De10d6>10.6)	tu'a la djan. cu cafne
+	Something-to-do-with John frequently-occurs
+</pre><p>This must mean that something which John does, or which happens t=
o John, occurs frequently: but without more context there is no way to figu=
re out what. Note that without the =93tu'a=94, <a href=3D#e10d6>Example 10.=
6 </a>would mean that John considered as an event frequently occurs =97 in =
other words, that John has some sort of on-and-off existence! Normally we d=
o not think of people as events in English, but the x1 place of =93cafne=94=
 is an event, and if something that does not seem to be an event is put the=
re, the Lojbanic listener will attempt to construe it as one. (Of course, t=
his analysis assumes that =93djan.=94 is the name of a person, and not the =
name of some event.)
+<p>
+<lx "jai">  XE "jai"   <lx "JAI">  XE "JAI selma=92o"   <cx "abstraction, =
making concrete">  XE "abstraction(s): making concrete"   <cx "abstraction,=
 simplification to sumti with jai">  XE "abstraction(s): simplification to =
sumti with jai"   Logically, a counterpart of some sort is needed to =93tu'=
a=94 which transposes an abstract sumti into a concrete one. This is achiev=
ed at the selbri level by the cmavo =93jai=94 (of selma'o JAI). This cmavo =
has more than one function, discussed in <a href=3Dchap9.html>Chapter 9 </a=
>and <a href=3Dchap11.html>Chapter 11</a>; for the purposes of this chapter=
, it operates as a conversion of selbri, similarly to the cmavo of selma'o =
SE. This conversion changes
+<p>
+<ex "cause death">  XE "cause death: example=93  <pre><a name=3De10d7>10.7=
)	tu'a mi rinka le nu do morsi
+	Something-to-do-with me causes the event-of you are-dead
+	My action causes your death.
+</pre>into
+<p>
+<pre><a name=3De10d8>10.8)	mi jai rinka le nu do morsi
+	I am-associated-with causing the event-of your death.
+	I cause your death.
+</pre><p>In English, the subject of =93cause=94 can either be the actual c=
ause (an event), or else the agent of the cause (a person, typically); not =
so in Lojban, where the x1 of =93rinka=94 is always an event. <a href=3D#e1=
0d7>Example 10.7 </a>and <a href=3D#e10d8>Example 10.8 </a>look equally con=
venient (or inconvenient), but in making descriptions, <a href=3D#e10d8>Exa=
mple 10.8 </a>can be altered to:
+<p>
+<pre><a name=3De10d9>10.9)	le jai rinka be le nu do morsi
+	that-which-is associated-with causing (the event-of your death)
+	the one who caused your death
+</pre>because =93jai=94 modifies the selbri and can be incorporated into t=
he description =97 not so for =93tu'a=94.
+<p>
+The weakness of =93jai=94 used in descriptions in this way is that it does=
 not specify which argument of the implicit abstraction is being raised int=
o the x1 place of the description selbri. One can be more specific by using=
 the modal form of =93jai=94 explained in <a href=3Dchap9.html>Chapter 9</a=
>:
+<p>
+<pre><a name=3De10d10>10.10)	le jai gau rinka be le nu do morsi
+	that-which-is agent-in causing (the event-of your death)
+Event-type abstractors and event contour tenses</h3>
+<p></pre><cx "abstraction contours, compared with contour tenses">  XE "ab=
straction contours: compared with contour tenses"   <cx "tense contours, co=
mpared with event abstraction contours">  XE "tense contours: compared with=
 event abstraction contours"   <cx "NU compared with ZAhO">  XE "NU compare=
d with ZAhO"   <cx "ZAhO compared with NU">  XE "ZAhO compared with NU"   <=
a name=3Ds11><h3>
+This section is a logical continuation of <a href=3D#s3>Section 3</a>.
+<p>
+<lx "NU">  XE "NU selma=92o"   <lx "ZAhO">  XE "ZAhO selma=92o"   There ex=
ists a relationship between the four types of events explained in <a href=
=3D#s3>Section 3 </a>and the event contour tense cmavo of selma'o ZAhO. The=
 specific cmavo of NU and of ZAhO are mutually interdefining; the ZAhO cont=
ours were chosen to fit the needs of the NU event types and vice versa. Eve=
nt contours are explained in full in <a href=3Dchap10.html>Chapter 10</a>, =
and only summarized here.
+<p>
+The purpose of ZAhO cmavo is to represent the natural portions of an event=
, such as the beginning, the middle, and the end. They fall into several gr=
oups:
+<p>
+<cx "process abstraction, related tense contours">  XE "process abstractio=
n(s): related tense contours"   <dl compact><dt><dd>The cmavo =93pu'o=94, =
=93ca'o=94, and =93ba'o=94 represent spans of time: before an event begins,=
 while it is going on, and after it is over, respectively.
+<p><dt><dd>The cmavo =93co'a=94, =93de'a=94, =93di'a=94, and =93co'u=94 re=
present points of time: the start of an event, the temporary stopping of an=
 event, the resumption of an event after a stop, and the end of an event, r=
espectively. Not all events can have breaks in them, in which case =93de'a=
=94 and =93di'a=94 do not apply.
+<p><dt><dd>The cmavo =93mo'u=94 and =93za'o=94 correspond to =93co'u=94 an=
d =93ba'o=94 respectively, in the case of those events which have a natural=
 ending point that may not be the same as the actual ending point: =93mo'u=
=94 refers to the natural ending point, and =93za'o=94 to the time between =
the natural ending point and the actual ending point (the =93excessive=94 o=
r =93superfective=94 part of the event).
+<p><dt><dd>The cmavo =93co'i=94 represents an entire event considered as a=
 point-event or achievement.
+</dl><lx "pu'u">  XE "pu'u"   All these cmavo are applicable to events see=
n as processes and abstracted with =93pu'u=94. Only processes have enough i=
nternal structure to make all these points and spans of time meaningful.
+<p>
+<cx "state abstraction, related tense contours">  XE "state abstraction(s)=
: related tense contours"   <lx "za'i">  XE "za'i"   For events seen as sta=
tes and abstracted with =93za'i=94, the meaningful event contours are the s=
pans =93pu'o=94, =93ca'o=94, and =93ba'o=94; the starting and ending points=
 =93co'a=94 and =93co'u=94, and the achievement contour =93co'i=94. States =
do not have natural endings distinct from their actual endings. (It is an o=
pen question whether states can be stopped and resumed.)
+<p>
+<cx "activity abstraction, related tense contours">  XE "activity abstract=
ion(s): related tense contours"   <lx "zu'o">  XE "zu'o"   For events seen =
as activities and abstracted with =93zu'o=94, the meaningful event contours=
 are the spans =93pu'o=94, =93ca'o=94, and =93ba'o=94, and the achievement =
contour =93co'i=94. Because activities are inherently cyclic and repetitive=
, the beginning and ending points are not well-defined: you do not know whe=
ther an activity has truly begun until it begins to repeat.
+<p>
+<cx "point-event abstraction, related tense contours">  XE "point-event ab=
straction(s): related tense contours"   <cx "achievement abstraction, relat=
ed tense contours">  XE "achievement abstraction(s): related tense contours=
"   For events seen as point-events and abstracted with =93mu'e=94, the mea=
ningful event contours are the spans =93pu'o=94 and =93ba'o=94 but not =93c=
a'o=94 (a point-event has no duration), and the achievement contour =93co'i=
=94.
+<p>
+Note that the parts of events are themselves events, and may be treated as=
 such. The points in time may be seen as =93mu'e=94 point-events; the spans=
 of time may constitute processes or activities. Therefore, Lojban allows u=
s to refer to processes within processes, activities within states, and man=
y other complicated abstract things.
+<p>
+Abstractor connection</h3>
+<p><cx "abstractions, connection">  XE "abstraction(s): connection"   <cx =
"connectives, for abstractions">  XE "connectives: for abstractions"   <a n=
ame=3Ds12><h3>
+An abstractor may be replaced by two or more abstractors joined by logical=
 or non-logical connectives. Connectives are explained in detail in <a href=
=3Dchap14.html>Chapter 14</a>. The connection can be expanded to one betwee=
n two bridi which differ only in abstraction marker. <a href=3D#e13d1>Examp=
le 13.1 </a>and <a href=3D#e13d2>Example 13.2 </a>are equivalent in meaning=
:
+<p>
+<pre><a name=3De13d1>13.1)	le ka la frank. ciska cu xlali
+		.ije le ni la frank. ciska cu xlali
+	The quality-of Frank's writing is bad,=20
+		and the quantity of Frank's writing is bad.
+<a name=3De13d2>13.2)	le ka je ni la frank. ciska cu xlali
+	The quality and quantity of Frank's writing is bad.
+</pre><p>This feature of Lojban has hardly ever been used, and nobody know=
s what uses it may eventually have.
+<p>
+Table of abstractors</h3>
+<p><cx "abstraction, table">  XE "abstraction(s): table"   <a name=3Ds13><=
h3>
+The following table gives each abstractor, an English gloss for it, a Lojb=
an gismu which is connected with it (more or less remotely: the association=
s between abstractors and gismu are meant more as memory hooks than for any=
 kind of inference), the rafsi associated with it, and (on the following li=
ne) its place structure.
+<p>
+<pre>nu 	event of		fasnu		nun
+		x1 is an event of (the bridi)=20
+ka	property of	ckaji		kam
+		x1 is a property of (the bridi)=20
+ni	amount of	klani		nil
+		x1 is an amount of (the bridi) measured on scale x2=20
+jei	truth-value of	jetnu		jez
+		x1 is a truth-value of (the bridi) under epistemology x2=20
+li'i	experience of	lifri		liz
+		x1 is an experience of (the bridi) to experiencer x2=20
+si'o	idea of		sidbo		siz
+		x1 is an idea/concept of (the bridi) in the mind of x2=20
+du'u	predication of	=97=96		dum
+		x1 is the bridi (the bridi) expressed by sentence x2=20
+
+su'u	abstraction of	sucta		sus
+		x1 is an abstract nature of (the bridi)=20
+za'i	state of		zasti		zam
+		x1 is a state of (the bridi)=20
+zu'o	activity of	zukte		zum
+		x1 is an activity of (the bridi)=20
+pu'u	process of	pruce		pup
+		x1 is a process of (the bridi)=20
+mu'e	point-event of	mulno		mub
+		x1 is a point-event/achievement of (the bridi)
+
+</body></html>
+<img src=3Dchap12.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+=20
+
+<h2>Chapter 12
+<br>
+Dog House And White House: Determining lujvo Place Structures</h2>
+<h6>$Revision: 4.0 $<br>
+mkhtml: 1.1</h6></p></center>
+Why have lujvo?</h3>
+<p><cx "lujvo, place structure of">  XE "lujvo: place structure of"   <a n=
ame=3Ds1><h3>
+<cx "world's languages">  XE "world's languages"   <cx "word lists">  XE "=
word lists"   <cx "semantic primitives">  XE "semantic primitives"   <cx "m=
inimal list">  XE "minimal list"   <cx "gismu, rationale for">  XE "gismu: =
rationale for"   <cx "lujvo, grammar of">  XE "lujvo: grammar of"   <cx "co=
mpound words">  XE "compound words"   The Lojban vocabulary is founded on i=
ts list of 1350-plus gismu, made up by combining word lists from various so=
urces. These gismu are not intended to be either a complete vocabulary for =
the language nor a minimal list of semantic primitives. Instead, the gismu =
list serves as a basis for the creation of compound words, or lujvo. The in=
tention is that (except in certain semantically broad but shallow fields su=
ch as cultures, nations, foods, plants, and animals) suitable lujvo can be =
devised to cover the ten million or so concepts expressible in all the worl=
d's languages taken together. Grammatically, lujvo behave just like gismu: =
they have place structures and function as selbri.
+<p>
+<cx "lujvo, compared with tanru">  XE "lujvo: compared with tanru"   There=
 is a close relationship between lujvo and tanru. In fact, lujvo are conden=
sed forms of tanru:
+<p>
+<pre><a name=3De1d1>1.1)	ti fagri festi
+	That is-fire waste.
+</pre>contains a tanru which can be reduced to the lujvo in:
+<p>
+<pre><a name=3De1d2>1.2)	ti fagyfesti
+	That is-fire-waste.
+	That is-ashes.
+</pre><p>Although the lujvo =93fagyfesti=94 is derived from the tanru =93f=
agri festi=94, it is not equivalent in meaning to it. In particular, =93fag=
yfesti=94 has a distinct place structure of its own, not the same as that o=
f =93festi=94. (In contrast, the tanru does have the same place structure a=
s =93festi=94.) The lujvo needs to take account of the places of =93fagri=
=94 as well. When a tanru is made into a lujvo, there is no equivalent of =
=93be=A0<dots>=85</dots>bei=A0<dots>=85</dots>be'o=94 (described in <a href=
=3Dchap5.html>Chapter 5</a>) to incorporate sumti into the middle of the lu=
jvo.
+<p>
+<cx "creative understanding">  XE "creative understanding"   <cx "lujvo, r=
ationale for">  XE "lujvo: rationale for"   So why have lujvo? Primarily to=
 reduce semantic ambiguity. On hearing a tanru, there is a burden on the li=
stener to figure out what the tanru might mean. Adding further terms to the=
 tanru reduces ambiguity in one sense, by providing more information; but i=
t increases ambiguity in another sense, because there are more and more tan=
ru joints, each with an ambiguous significance. Since lujvo, like other bri=
vla, have a fixed place structure and a single meaning, encapsulating a com=
monly-used tanru into a lujvo relieves the listener of the burden of creati=
ve understanding. In addition, lujvo are typically shorter than the corresp=
onding tanru.
+<p>
+<cx "alternative guidelines">  XE "alternative guidelines"   <cx "absolute=
 laws">  XE "absolute laws"   <cx "lujvo, guidelines for place structure"> =
 XE "lujvo: guidelines for place structure"   <cx "lujvo place structure, g=
uidelines">  XE "lujvo place structure: guidelines"   There are no absolute=
 laws fixing the place structure of a newly created lujvo. The maker must c=
onsider the place structures of all the components of the tanru and then de=
cide which are still relevant and which can be removed. What is said in thi=
s chapter represents guidelines, presented as one possible standard, not ne=
cessarily complete, and not the only possible standard. There may well be l=
ujvo that are built without regard for these guidelines, or in accordance w=
ith entirely different guidelines, should such alternative guidelines somed=
ay be developed. The reason for presenting any guidelines at all is so that=
 Lojbanists have a starting point for deciding on a likely place structure =
=97 one that others seeing the same word can also arrive at by similar cons=
ideration.
+<p>
+<cx "lujvo, cmavo incorporation">  XE "lujvo: cmavo incorporation"   If th=
e tanru includes connective cmavo such as =93bo=94, =93ke=94, =93ke'e=94, o=
r =93je=94, or conversion or abstraction cmavo such as =93se=94 or =93nu=94=
, there are ways of incorporating them into the lujvo as well. Sometimes th=
is makes the lujvo excessively long; if so, the cmavo may be dropped. This =
leads to the possibility that more than one tanru could produce the same lu=
jvo. Typically, however, only one of the possible tanru is useful enough to=
 justify making a lujvo for it.
+<p>
+The exact workings of the lujvo-making algorithm, which takes a tanru buil=
t from gismu (and possibly cmavo) and produces a lujvo from it, are describ=
ed in <a href=3Dchap4.html>Chapter 4</a>.
+<p>
+The meaning of tanru: a necessary detour</h3>
+<p><cx "necessary detour">  XE "necessary detour"   <a name=3Ds2><h3>
+<cx "lujvo meaning">  XE "lujvo meaning"   <cx "veljvo, definition">  XE "=
veljvo: definition"   <cx "seltau, definition">  XE "seltau: definition"   =
<cx "tertau, definition">  XE "tertau: definition"   <cx "modifier, of a ta=
nru">  XE "modifier: of a tanru"   <cx "modified, of a tanru">  XE "modifie=
d: of a tanru"   The meaning of a lujvo is controlled by =97 but is not the=
 same as =97 the meaning of the tanru from which the lujvo was constructed.=
 The tanru corresponding to a lujvo is called its =93veljvo=94 in Lojban, a=
nd since there is no concise English equivalent, that term will be used in =
this chapter. Furthermore, the left (modifier) part of a tanru will be call=
ed the =93seltau=94, and the right (modified) part the =93tertau=94, follow=
ing the usage of <a href=3Dchap5.html>Chapter 5</a>. For brevity, we will s=
peak of the seltau or tertau of a lujvo, meaning of course the seltau or te=
rtau of the veljvo of that lujvo. (If this terminology is confusing, substi=
tuting =93modifier=94 for =93seltau=94 and =93modified=94 for =93tertau=94 =
may help.)
+<p>
+<cx "tanru, place structure of">  XE "tanru: place structure of"   The pla=
ce structure of a tanru is always the same as the place structure of its te=
rtau. As a result, the meaning of the tanru is a modified version of the me=
aning of the tertau; the tanru will typically, but not always, refer to a s=
ubset of the things referred to by the tertau.
+<p>
+<cx "wine-dark sea">  XE "wine-dark sea"   <cx "tanru, purpose">  XE "tanr=
u: purpose"   The purpose of a tanru is to join concepts together without n=
ecessarily focusing on the exact meaning of the seltau. For example, in the=
 <cite>Iliad</cite>, the poet talks about =93the wine-dark sea=94, in which=
 =93wine=94 is a seltau relative to =93dark=94, and the pair of words is a =
seltau relative to =93sea=94. We're talking about the sea, not about wine o=
r color. The other words are there to paint a scene in the listener's mind,=
 in which the real action will occur, and to evoke relations to other sagas=
 of the time similarly describing the sea. Logical inferences about wine or=
 color will be rejected as irrelevant.
+<p>
+<ex "goer-house">  XE "goer-house: example=93  As a simple example, consid=
er the rather non-obvious tanru =93klama zdani=94, or =93goer-house=94. The=
 gismu =93zdani=94 has two places:
+<p>
+<pre><a name=3De2d1>2.1)	x1 is a nest/house/lair/den for inhabitant x2
+</pre>(but in this chapter we will use simply =93house=94, for brevity), a=
nd the gismu =93klama=94 has five:
+<p>
+<pre><a name=3De2d2>2.2)	x1 goes to destination x2 from origin point x3 vi=
a route x4 using means x5
+</pre><p>The tanru =93klama zdani=94 will also have two places, namely tho=
se of =93zdani=94. Since a =93klama zdani=94 is a type of =93zdani=94, we c=
an assume that all goer-houses =97 whatever they may be =97 are also houses=
.
+<p>
+<cx "fleas">  XE "fleas"   <ex "dog house">  XE "dog house: example=93  <c=
x "tanru, possible meanings of">  XE "tanru: possible meanings of"   But is=
 knowing the places of the tertau everything that is needed to understand t=
he meaning of a tanru? No. To see why, let us switch to a less unlikely tan=
ru: =93gerku zdani=94, literally =93dog house=94. A tanru expresses a very =
loose relation: a =93gerku zdani=94 is a house that has something to do wit=
h some dog or dogs. What the precise relation might be is left unstated. Th=
us, the meaning of =93lo gerku zdani=94 can include all of the following: h=
ouses occupied by dogs, houses shaped by dogs, dogs which are also houses (=
e.g. houses for fleas), houses named after dogs, and so on. All that is ess=
ential is that the place structure of =93zdani=94 continues to apply.
+<p>
+For something (call it z1) to qualify as a =93gerku zdani=94 in Lojban, it=
's got to be a house, first of all. For it to be a house, it's got to house=
 someone (call that z2). Furthermore, there's got to be a dog somewhere (ca=
lled g1). For g1 to count as a dog in Lojban, it's got to belong to some br=
eed as well (called g2). And finally, for z1 to be in the first place of =
=93gerku zdani=94, as opposed to just =93zdani=94, there's got to be some r=
elationship (called r) between some place of =93zdani=94 and some place of =
=93gerku=94. It doesn't matter which places, because if there's a relations=
hip between some place of =93zdani=94 and any place of =93gerku=94, then th=
at relationship can be compounded with the relationship between the places =
of =93gerku=94 =97 namely, =93gerku=94 itself =97 to reach any of the other=
 =93gerku=94 places. Thus, if the relationship turns out to be between z2 a=
nd g2, we can still state r in terms of z1 and g1: =93the relationship invo=
lves the dog g1, whose breed has to do with the occupant of the house z1=94=
.
+<p>
+<ex "Bill Clinton">  XE "Bill Clinton: example=93  Doubtless to the relief=
 of the reader, here's an illustration. We want to find out whether the Whi=
te House (the one in which the U. S. President lives, that is) counts as a =
=93gerku zdani=94. We go through the five variables. The White House is the=
 z1. It houses Bill Clinton as z2, as of this writing, so it counts as a =
=93zdani=94. Let's take a dog =97 say, Spot (g1). Spot has to have a breed;=
 let's say it's a Saint Bernard (g2). Now, the White House counts as a =93g=
erku zdani=94 if there is any relationship (r) at all between the White Hou=
se and Spot. (We'll choose the g1 and z1 places to relate by r; we could ha=
ve chosen any other pair of places, and simply gotten a different relations=
hip.)
+<p>
+<cx "Chelsea Clinton">  XE "Chelsea Clinton"   The sky is the limit for r;=
 it can be as complicated as =93The other day, g1 (Spot) chased Socks, who =
is owned by Chelsea Clinton, who is the daughter of Bill Clinton, who lives=
 in z1 (the White House)=94 or even worse. If no such r can be found, well,=
 you take another dog, and keep going until no more dogs can be found. Only=
 then can we say that the White House cannot fit into the first place of =
=93gerku zdani=94.
+<p>
+As we have seen, no less than five elements are involved in the definition=
 of =93gerku zdani=94: the house, the house dweller, the dog, the dog breed=
 (everywhere a dog goes in Lojban, a dog breed follows), and the relationsh=
ip between the house and the dog. Since tanru are explicitly ambiguous in L=
ojban, the relationship r cannot be expressed within a tanru (if it could, =
it wouldn't be a tanru any more!) All the other places, however, can be exp=
ressed =97 thus:
+<p>
+<pre><a name=3De2d3>2.3)	la blabi zdani cu gerku be fa la spot. bei la san=
kt. berNARD. be'o
+		zdani la bil. klinton.
+	The White House is-a-dog (namely Spot of-breed Saint Bernard)
+		type-of-house-for Bill Clinton.
+</pre><cx "derogatory terms">  XE "derogatory terms"   Not the most elegan=
t sentence ever written in either Lojban or English. Yet if there is any re=
lation at all between Spot and the White House, <a href=3D#e2d4>Example 2.4=
 </a>is arguably true. If we concentrate on just one type of relation in in=
terpreting the tanru =93gerku zdani=94, then the meaning of =93gerku zdani=
=94 changes. So if we understand =93gerku zdani=94 as having the same meani=
ng as the English word =93doghouse=94, the White House would no longer be a=
 =93gerku zdani=94 with respect to Spot, because as far as we know Spot doe=
s not actually live in the White House, and the White House is not a doghou=
se (derogatory terms for incumbents notwithstanding).
+<p>
+<a name=3Ds3><h3>The meaning of lujvo</h3>
+<p>
+This is a fairly long way to go to try and work out how to say =93doghouse=
=94! The reader can take heart; we're nearly there. Recall that one of the =
components involved in fixing the meaning of a tanru =97 the one left delib=
erately vague =97 is the precise relation between the tertau and the seltau=
. Indeed, fixing this relation is tantamount to giving an interpretation to=
 the ambiguous tanru.
+<p>
+<cx "disambiguated instance">  XE "disambiguated instance"   <cx "lujvo, i=
nterpreting">  XE "lujvo: interpreting"   <cx "lujvo, and seltau/tertau rel=
ationship">  XE "lujvo: and seltau/tertau relationship"   A lujvo is define=
d by a single disambiguated instance of a tanru. That is to say, when we tr=
y to design the place structure of a lujvo, we don't need to try to discove=
r the relation between the tertau and the seltau. We already know what kind=
 of relation we're looking for; it's given by the specific need we wish to =
express, and it determines the place structure of the lujvo itself.
+<p>
+Therefore, it is generally not appropriate to simply devise lujvo and deci=
de on place structures for them without considering one or more specific us=
ages for the coinage. If one does not consider specifics, one will be likel=
y to make erroneous generalizations on the relationship r.
+<p>
+<cx "lujvo, design consideration for relationship">  XE "lujvo: design con=
sideration for relationship"   The insight driving the rest of this chapter=
 is this: while the relation expressed by a tanru can be very distant (e.g.=
 Spot chasing Socks, above), the relationship singled out for disambiguatio=
n in a lujvo should be quite close. This is because lujvo-making, paralleli=
ng natural language compounding, picks out the most salient relationship r =
between a tertau place and a seltau place to be expressed in a single word.=
 The relationship =93dog chases cat owned by daughter of person living in h=
ouse=94 is too distant, and too incidental, to be likely to need expression=
 as a single short word; the relationship =93dog lives in house=94 is not. =
From all the various interpretations of =93gerku zdani=94, the person creat=
ing =93gerzda=94 should pick the most useful value of r. The most useful on=
e is usually going to be the most obvious one, and the most obvious one is =
usually the closest one.
+<p>
+In fact, the relationship will almost always be so close that the predicat=
e expressing r will be either the seltau or the tertau predicate itself. Th=
is should come as no surprise, given that a word like =93zdani=94 in Lojban=
 is a predicate. Predicates express relations; so when you're looking for a=
 relation to tie together =93le zdani=94 and =93le gerku=94, the most obvio=
us relation to pick is the very relation named by the tertau, =93zdani=94: =
the relation between a home and its dweller. As a result, the object which =
fills the first place of =93gerku=94 (the dog) also fills the second place =
of =93zdani=94 (the house-dweller).
+<p>
+<cx "lujvo place structure, dropping redundant places">  XE "lujvo place s=
tructure: dropping redundant places"   The seltau-tertau relationship in th=
e veljvo is expressed by the seltau or tertau predicate itself. Therefore, =
at least one of the seltau places is going to be equivalent to a tertau pla=
ce. This place is thus redundant, and can be dropped from the place structu=
re of the lujvo. As a corollary, the precise relationship between the veljv=
o components can be implicitly determined by finding one or more places to =
overlap in this way.
+<p>
+So what is the place structure of =93gerzda=94? We're left with three plac=
es, since the dweller, the =93se zdani=94, turned out to be identical to th=
e dog, the =93gerku=94. We can proceed as follows:
+<p>
+<cx "lujvo place structure, notation conventions">  XE "lujvo place struct=
ure: notation conventions"   (The notation introduced casually in <a href=
=3D#s2>Section 2 </a>will be useful in the rest of this chapter. Rather tha=
n using the regular x1, x2, etc. to represent places, we'll use the first l=
etter of the relevant gismu in place of the =93x=94, or more than one lette=
r where necessary to resolve ambiguities. Thus, z1 is the first place of =
=93zdani=94, and g2 is the second place of =93gerku=94.)
+<p>
+<cx "new notation">  XE "new notation"   <cx "lujvo place structure, expli=
cated walk-through">  XE "lujvo place structure: explicated walk-through"  =
 The place structure of =93zdani=94 is given as <a href=3D#e2d1>Example 2.1=
</a>, but is repeated here using the new notation:
+<p>
+<ex "doghouse">  XE "doghouse: example=93  <pre><a name=3De3d1>3.1)	z1 is =
a nest/house/lair/den of z2
+</pre><p>The place structure of =93gerku=94 is:
+<p>
+<pre><a name=3De3d2>3.2)	g1 is a dog of breed g2
+</pre><p>But z2 is the same as g1; therefore, the tentative place structur=
e for =93gerzda=94 now becomes:
+<p>
+<pre><a name=3De3d3>3.3)	z1 is a house for dweller z2 of breed g2
+</pre>which can also be written
+<p>
+<pre><a name=3De3d4>3.4)	z1 is a house for dog g1 of breed g2
+</pre>or more comprehensively
+<p>
+<pre><a name=3De3d5>3.5)	z1 is a house for dweller/dog z2=3Dg1 of breed g2
+</pre><p>Despite the apparently conclusive nature of <a href=3D#e3d5>Examp=
le 3.5</a>, our task is not yet done: we still need to decide whether any o=
f the remaining places should also be eliminated, and what order the lujvo =
places should appear in. These concerns will be addressed in the remainder =
of the chapter; but we are now equipped with the terminology needed for tho=
se discussions.
+<p>
+<a name=3Ds4><h3>Selecting places</h3>
+<p>
+<cx "lujvo place structure, basis of">  XE "lujvo place structure: basis o=
f"   The set of places of an ordinary lujvo are selected from the places of=
 its component gismu. More precisely, the places of such a lujvo are derive=
d from the set of places of the component gismu by eliminating unnecessary =
places, until just enough places remain to give an appropriate meaning to t=
he lujvo. In general, including a place makes the concept expressed by a lu=
jvo more general; excluding a place makes the concept more specific, becaus=
e omitting the place requires assuming a standard value or range of values =
for it.
+<p>
+<cx "lujvo place structure, rationale for standardization">  XE "lujvo pla=
ce structure: rationale for standardization"   It would be possible to desi=
gn the place structure of a lujvo from scratch, treating it as if it were a=
 gismu, and working out what arguments contribute to the notion to be expre=
ssed by the lujvo. There are two reasons arguing against doing so and in fa=
vor of the procedure detailed in this chapter.
+<p>
+The first is that it might be very difficult for a hearer or reader, who h=
as no preconceived idea of what concept the lujvo is intended to convey, to=
 work out what the place structure actually is. Instead, he or she would ha=
ve to make use of a lujvo dictionary every time a lujvo is encountered in o=
rder to work out what a =93se jbopli=94 or a =93te klagau=94 is. But this w=
ould mean that, rather than having to learn just the 1300-odd gismu place s=
tructures, a Lojbanist would also have to learn myriads of lujvo place stru=
ctures with little or no apparent pattern or regularity to them. The purpos=
e of the guidelines documented in this chapter is to apply regularity and t=
o make it conventional wherever possible.
+<p>
+The second reason is related to the first: if the veljvo of the lujvo has =
not been properly selected, and the places for the lujvo are formulated fro=
m scratch, then there is a risk that some of the places formulated may not =
correspond to any of the places of the gismu used in the veljvo of the lujv=
o. If that is the case =97 that is to say, if the lujvo places are not a su=
bset of the veljvo gismu places =97 then it will be very difficult for the =
hearer or reader to understand what a particular place means, and what it i=
s doing in that particular lujvo. This is a topic that will be further disc=
ussed in <a href=3D#s14>Section 14</a>.
+<p>
+However, second-guessing the place structure of the lujvo is useful in gui=
ding the process of subsequently eliminating places from the veljvo. If the=
 Lojbanist has an idea of what the final place structure should look like, =
he or she should be able to pick an appropriate veljvo to begin with, in or=
der to express the idea, and then to decide which places are relevant or no=
t relevant to expressing that idea.
+<p>
+<a name=3Ds5><h3>Symmetrical and asymmetrical lujvo</h3>
+<p>
+<cx "lujvo, symmetrical">  XE "lujvo: symmetrical"   <cx "veljvo, symmetri=
cal">  XE "veljvo: symmetrical"   <ex "great soldier">  XE "great soldier: =
example=93  <cx "lujvo place structure, when first places redundant">  XE "=
lujvo place structure: when first places redundant"   A common pattern, per=
haps the most common pattern, of lujvo-making creates what is called a =93s=
ymmetrical lujvo=94. A symmetrical lujvo is one based on a tanru interpreta=
tion such that the first place of the seltau is equivalent to the first pla=
ce of the tertau: each component of the tanru characterizes the same object=
. As an illustration of this, consider the lujvo =93balsoi=94: it is intend=
ed to mean =93both great and a soldier=94 =97 that is, =93great soldier=94,=
 which is the interpretation we would tend to give its veljvo, =93banli son=
ci=94. The underlying gismu place structures are:
+<p>
+<pre><a name=3De5d1>5.1)	=93banli=94:  b1 is great in property b2 by stand=
ard b3
+	=93sonci=94:  s1 is a soldier of army s2
+</pre><p>In this case the s1 place of =93sonci=94 is redundant, since it i=
s equivalent to the b1 place of =93banli=94. Therefore the place structure =
of =93balsoi=94 need not include places for both s1 and b1, as they refer t=
o the same thing. So the place structure of =93balsoi=94 is at most
+<p>
+<pre><a name=3De5d2>5.2)	b1=3Ds1 is a great soldier of army s2 in property=
 b2 by standard b3
+</pre><cx "symmetrical veljvo">  XE "symmetrical veljvo"   <ex "listen att=
entively">  XE "listen attentively: example=93  <cx "lujvo place structure,=
 when first places redundant plus others">  XE "lujvo place structure: when=
 first places redundant plus others"   Some symmetrical veljvo have further=
 equivalent places in addition to the respective first places. Consider the=
 lujvo =93tinju'i=94, =93to listen=94 (=93to hear attentively, to hear and =
pay attention=94). The place structures of the gismu =93tirna=94 and =93jun=
di=94 are:
+<p>
+<cx "background noise">  XE "background noise"   <pre><a name=3De5d3>5.3)	=
=93tirna=94:  t1 hears sound t2 against background noise t3
+	=93jundi=94:  j1 pays attention to j2
+</pre>and the place structure of the lujvo is:
+<p>
+<pre><a name=3De5d4>5.4)	j1=3Dt1 listens to j2=3Dt2 against background noi=
se t3
+</pre><p>Why so? Because not only is the j1 place (the one who pays attent=
ion) equivalent to the t1 place (the hearer), but the j2 place (the thing p=
aid attention to) is equivalent to the t2 place (the thing heard).
+<p>
+<cx "lujvo, asymmetrical">  XE "lujvo: asymmetrical"   <cx "lujvo place st=
ructure, when first place redundant with non-first">  XE "lujvo place struc=
ture: when first place redundant with non-first"   A substantial minority o=
f lujvo have the property that the first place of the seltau (=93gerku=94 i=
n this case) is equivalent to a place other than the first place of the ter=
tau; such lujvo are said to be =93asymmetrical=94. (There is a deliberate p=
arallel here with the terms =93asymmetrical tanru=94 and =93symmetrical tan=
ru=94 used in <a href=3Dchap5.html>Chapter 5</a>.)
+<p>
+<cx "lujvo place structure, effect of =93SE=94">  XE "lujvo place structur=
e: effect of SE"   In principle any asymmetrical lujvo could be expressed a=
s a symmetrical lujvo. Consider =93gerzda=94, discussed in <a href=3D#s3>Se=
ction 3</a>, where we learned that the g1 place was equivalent to the z2 pl=
ace. In order to get the places aligned, we could convert =93zdani=94 to =
=93se zdani=94 (or =93selzda=94 when expressed as a lujvo). The place struc=
ture of =93selzda=94 is
+<p>
+<pre><a name=3De5d5>5.5)	s1 is housed by nest s2
+</pre>and so the three-part lujvo =93gerselzda=94 would have the place str=
ucture
+<p>
+<pre><a name=3De5d6>5.6)	s1=3Dg1 is a dog housed in nest s2 of dog breed g=
2
+</pre><p>However, although =93gerselzda=94 is a valid lujvo, it doesn't tr=
anslate =93doghouse=94; its first place is the dog, not the doghouse. Furth=
ermore, it is more complicated than necessary; =93gerzda=94 is simpler than=
 =93gerselzda=94.
+<p>
+From the reader's or listener's point of view, it may not always be obviou=
s whether a newly met lujvo is symmetrical or asymmetrical, and if the latt=
er, what kind of asymmetrical lujvo. If the place structure of the lujvo is=
n't given in a dictionary or elsewhere, then plausibility must be applied, =
just as in interpreting tanru.
+<p>
+<ex "car goer">  XE "car goer: example=93  The lujvo =93karcykla=94, for e=
xample, is based on =93karce klama=94, or =93car goer=94. The place structu=
re of =93karce=94 is:
+<p>
+<pre><a name=3De5d7>5.7)	karce:  ka1 is a car carrying ka2 propelled by ka=
3
+</pre><p>An asymmetrical interpretation of =93karcykla=94 that is strictly=
 analogous to the place structure of =93gerzda=94, equating the kl2 (destin=
ation) and ka1 (car) places, would lead to the place structure
+<p>
+<pre><a name=3De5d8>5.8)	kl1 goes to car kl2=3Dka1 which carries ka2 prope=
lled by ka3 from origin kl3
+		via route kl4 by means of kl5
+</pre><p>But in general we go about in cars, rather than going to cars, so=
 a far more likely place structure treats the ka1 place as equivalent to th=
e kl5 place, leading to
+<p>
+<pre><a name=3De5d9>5.9)	kl1 goes to destination kl2 from origin kl3 via r=
oute kl4=20
+by means of car kl5=3Dka1 carrying ka2 propelled by ka3.
+</pre>instead.
+<p>
+<a name=3Ds6><h3>Dependent places</h3>
+<p>
+<cx "lujvo place structure, dependent places">  XE "lujvo place structure:=
 dependent places"   In order to understand which places, if any, should be=
 completely removed from a lujvo place structure, we need to understand the=
 concept of dependent places. One place of a brivla is said to be dependent=
 on another if its value can be predicted from the values of one or more of=
 the other places. For example, the g2 place of =93gerku=94 is dependent on=
 the g1 place. Why? Because when we know what fits in the g1 place (Spot, l=
et us say, a well-known dog), then we know what fits in the g2 place (=93St=
. Bernard=94, let us say). In other words, when the value of the g1 place h=
as been specified, the value of the g2 place is determined by it. Conversel=
y, since each dog has only one breed, but each breed contains many dogs, th=
e g1 place is not dependent on the g2 place; if we know only that some dog =
is a St. Bernard, we cannot tell by that fact alone which dog is meant.
+<p>
+For =93zdani=94, on the other hand, there is no dependency between the pla=
ces. When we know the identity of a house-dweller, we have not determined t=
he house, because a dweller may dwell in more than one house. By the same t=
oken, when we know the identity of a house, we do not know the identity of =
its dweller, for a house may contain more than one dweller.
+<p>
+<cx "lujvo place structure, dropping dependent seltau places">  XE "lujvo =
place structure: dropping dependent seltau places"   The rule for eliminati=
ng places from a lujvo is that dependent places provided by the seltau are =
eliminated. Therefore, in =93gerzda=94 the dependent g2 place is removed fr=
om the tentative place structure given in <a href=3D#e3d5>Example 3.5</a>, =
leaving the place structure:
+<p>
+<pre><a name=3De6d1>6.1)	z1 is the house dwelt in by dog z2=3Dg1
+</pre><p>Informally put, the reason this has happened =97 and it happens a=
 lot with seltau places =97 is that the third place was describing not the =
doghouse, but the dog who lives in it. The sentence
+<p>
+<ex "Mon Repos">  XE "Mon Repos: example=93  <pre><a name=3De6d2>6.2)	la m=
on. rePOS. gerzda la spat.
+	Mon Repos is a doghouse of Spot.
+</pre>really means
+<p>
+<pre><a name=3De6d3>6.3)	la mon. rePOS. zdani la spat. noi gerku
+	Mon Repos is a house of Spot, who is a dog.
+</pre>since that is the interpretation we have given =93gerzda=94. But tha=
t in turn means
+<p>
+  XE "unspecified breed:example"  <pre><a name=3De6d4>6.4)	la mon. rePOS. =
zdani la spat noi ke'a gerku zo'e
+	Mon Repos is a house of Spot, who is a dog of unspecified breed.
+</pre><p>Specifically,
+<p>
+<pre><a name=3De6d5>6.5)	la mon. rePOS. zdani la spat. noi ke'a gerku la s=
ankt. berNARD.
+	Mon Repos is a house of Spot, who is a dog of breed St. Bernard.
+</pre>and in that case, it makes little sense to say
+<p>
+<pre><a name=3De6d6>6.6)	la mon. rePOS. gerzda la spat.noi ke'a gerku la s=
ankt. berNARD. ku'o
+la sankt. berNARD.
+	Mon Repos is a doghouse of Spot, who is a dog of breed St. Bernard,
+of breed St. Bernard.
+</pre><cx "supplementary information">  XE "supplementary information"   e=
mploying the over-ample place structure of <a href=3D#e3d5>Example 3.5</a>.=
 The dog breed is redundantly given both in the main selbri and in the rela=
tive clause, and (intuitively speaking) is repeated in the wrong place, sin=
ce the dog breed is supplementary information about the dog, and not about =
the doghouse.
+<p>
+<ex "beetle">  XE "beetle: example=93  As a further example, take =93cakci=
nki=94, the lujvo for =93beetle=94, based on the tanru =93calku cinki=94, o=
r =93shell-insect=94. The gismu place structures are:
+<p>
+<cx "arthropod">  XE "arthropod"   <pre><a name=3De6d7>6.7)	=93calku=94:  =
ca1 is a shell/husk around ca2 made of ca3
+	=93cinki=94:  ci1 is an insect/arthropod of species ci2
+</pre><cx "cross-dependency">  XE "cross-dependency"   <cx "lujvo place st=
ructure, cross-dependent places">  XE "lujvo place structure: cross-depende=
nt places"   <cx "lujvo place structure, dropping cross-dependent places"> =
 XE "lujvo place structure: dropping cross-dependent places"   This example=
 illustrates a cross-dependency between a place of one gismu and a place of=
 the other. The ca3 place is dependent on ci1, because all insects (which f=
it into ci1) have shells made of chitin (which fits into ca3). Furthermore,=
 ca1 is dependent on ci1 as well, because each insect has only a single she=
ll. And since ca2 (the thing with the shell) is equivalent to ci1 (the inse=
ct), the place structure is
+<p>
+<pre><a name=3De6d8>6.8)	ci1=3Dca2 is a beetle of species ci2
+</pre>with not a single place of =93calku=94 surviving independently!
+<p>
+<cx "Coleoptera">  XE "Coleoptera"   <cx "beetles">  XE "beetles"   (Note =
that there is nothing in this explanation that tells us just why =93cakcink=
i=94 means =93beetle=94 (member of Coleoptera), since all insects in their =
adult forms have chitin shells of some sort. The answer, which is in no way=
 predictable, is that the shell is a prominent, highly noticeable feature o=
f beetles in particular.)
+<p>
+<cx "lujvo place structure, dropping dependent tertau places">  XE "lujvo =
place structure: dropping dependent tertau places"   What about the depende=
ncy of ci2 on ci1? After all, no beetle belongs to more than one species, s=
o it would seem that the ci2 place of =93cakcinki=94 could be eliminated on=
 the same reasoning that allowed us to eliminate the g2 place of =93gerzda=
=94 above. However, it is a rule that dependent places are not eliminated f=
rom a lujvo when they are derived from the tertau of its veljvo. This rule =
is imposed to keep the place structures of lujvo from drifting too far from=
 the tertau place structure; if a place is necessary in the tertau, it's tr=
eated as necessary in the lujvo as well.
+<p>
+<cx "wrong concept">  XE "wrong concept"   <cx "shoehorn">  XE "shoehorn" =
  <cx "lujvo place structure, selecting tertau">  XE "lujvo place structure=
: selecting tertau"   In general, the desire to remove places coming from t=
he tertau is a sign that the veljvo selected is simply wrong. Different pla=
ce structures imply different concepts, and the lujvo maker may be trying t=
o shoehorn the wrong concept into the place structure of his or her choosin=
g. This is obvious when someone tries to shoe-horn a =93klama=94 tertau int=
o a =93litru=94 or =93cliva=94 concept, for example: these gismu differ in =
their number of arguments, and suppressing places of =93klama=94 in a lujvo=
 doesn't make any sense if the resulting modified place structure is that o=
f =93litru=94 or =93cliva=94.
+<p>
+Sometimes the dependency is between a single place of the tertau and the w=
hole event described by the seltau. Such cases are discussed further in <a =
href=3D#s13>Section 13</a>.
+<p>
+<cx "recital rooms">  XE "recital rooms"   <cx "playgrounds">  XE "playgro=
unds"   <cx "elementary schools">  XE "elementary schools"   <cx "auditoriu=
ms">  XE "auditoriums"   <ex "school building">  XE "school building: examp=
le=93  <cx "lujvo place structure, dropping dependent places, caveat">  XE =
"lujvo place structure: dropping dependent places, caveat"   Unfortunately,=
 not all dependent places in the seltau can be safely removed: some of them=
 are necessary to interpreting the lujvo's meaning in context. It doesn't m=
atter much to a doghouse what breed of dog inhabits it, but it can make qui=
te a lot of difference to the construction of a school building what kind o=
f school is in it! Music schools need auditoriums and recital rooms, elemen=
tary schools need playgrounds, and so on: therefore, the place structure of=
 =93kuldi'u=94 (from =93ckule dinju=94, and meaning =93school building=94) =
needs to be
+<p>
+<pre><a name=3De6d9>6.9)	d1 is a building housing school c1 teaching subje=
ct c3 to audience c4
+</pre>even though c3 and c4 are plainly dependent on c1. The other places =
of =93ckule=94, the location (c2) and operators (c5), don't seem to be nece=
ssary to the concept =93school building=94, and are dependent on c1 to boot=
, so they are omitted. Again, the need for case-by-case consideration of pl=
ace structures is demonstrated.
+<p>
+<a name=3Ds7><h3>Ordering lujvo places</h3>
+<p>
+<cx "lujvo place order">  XE "lujvo place order"   So far, we have concent=
rated on selecting the places to go into the place structure of a lujvo. Ho=
wever, this is only half the story. In using selbri in Lojban, it is import=
ant to remember the right order of the sumti. With lujvo, the need to atten=
d to the order of sumti becomes critical: the set of places selected should=
 be ordered in such a way that a reader unfamiliar with the lujvo should be=
 able to tell which place is which.
+<p>
+<cx "lujvo place order, rationale for standardization">  XE "lujvo place o=
rder: rationale for standardization"   <ex "prayer">  XE "prayer: example=
=93  If we aim to make understandable lujvo, then, we should make the order=
 of places in the place structure follow some conventions. If this does not=
 occur, very real ambiguities can turn up. Take for example the lujvo =93jd=
aselsku=94, meaning =93prayer=94. In the sentence
+<p>
+  XE "Dong: example"  <pre><a name=3De7d1>7.1)	di'e jdaselsku la dong.
+	This-utterance is-a-prayer somehow-related-to-Dong.
+</pre>we must be able to know if Dong is the person making the prayer, giv=
ing the meaning
+<p>
+<pre><a name=3De7d2>7.2)	This is a prayer by Dong.
+</pre>or is the entity being prayed to, resulting in
+<p>
+<pre><a name=3De7d3>7.3)	This is a prayer to Dong.
+</pre><p>We could resolve such problems on a case-by-case basis for each l=
ujvo (<a href=3D#s14>Section 14 </a>discusses when this is actually necessa=
ry), but case-by-case resolution for run-of-the-mill lujvo makes the task o=
f learning lujvo place structures unmanageable. People need consistent patt=
erns to make sense of what they learn. Such patterns can be found across gi=
smu place structures (see <a href=3D#s16>Section 16</a>), and are even more=
 necessary in lujvo place structures. Case-by-case consideration is still n=
ecessary; lujvo creation is a subtle art, after all. But it is helpful to t=
ake advantage of any available regularities.
+<p>
+<cx "elimination process">  XE "elimination process"   <cx "lujvo place or=
der, symmetrical lujvo">  XE "lujvo place order: symmetrical lujvo"   <ex "=
great soldier">  XE "great soldier: example=93  We use two different orderi=
ng rules: one for symmetrical lujvo and one for asymmetrical ones. A symmet=
rical lujvo like =93balsoi=94 (from <a href=3D#s5>Section 5</a>) has the pl=
aces of its tertau followed by whatever places of the seltau survive the el=
imination process. For =93balsoi=94, the surviving places of =93banli=94 ar=
e b2 and b3, leading to the place structure:
+<p>
+<pre><a name=3De7d4>7.4)	b1=3Ds1 is a great soldier of army s2 in property=
 b2 by standard b3
+</pre>just what appears in <a href=3D#e5d1>Example 5.1</a>. In fact, all p=
lace structures shown until now have been in the correct order by the conve=
ntions of this section, though the fact has been left tacit until now.
+<p>
+The motivation for this rule is the parallelism between the lujvo bridi-sc=
hema
+<p>
+<pre><a name=3De7d5>7.5)	b1 bansoi s2 b2 b3
+	b1 is-a-great-soldier of-army-s2 in-property-b2 by-standard-b3
+</pre>and the more or less equivalent bridi-schema
+<p>
+<pre><a name=3De7d6>7.6)	b1 sonci s2 gi'e banli b2 b3
+	b1 is-a-soldier of-army-s2 and is-great in-property-b2 by-standard-b3
+</pre>where =93gi'e=94 is the Lojban word for =93and=94 when placed betwee=
n two partial bridi, as explained in <a href=3Dchap14.html>Chapter 14</a>.
+<p>
+<cx "animal doctor">  XE "animal doctor: example"   <cx "lujvo place order=
, asymmetrical lujvo">  XE "lujvo place order: asymmetrical lujvo"   <ex "v=
eterinarian">  XE "veterinarian: example=93  Asymmetrical lujvo like =93ger=
zda=94, on the other hand, employ a different rule. The seltau places are i=
nserted not at the end of the place structure, but rather immediately after=
 the tertau place which is equivalent to the first place of the seltau. Con=
sider =93dalmikce=94, meaning =93veterinarian=94: its veljvo is =93danlu mi=
kce=94, or =93animal doctor=94. The place structures for those gismu are:
+<p>
+<cx "ailment">  XE "ailment"   <pre><a name=3De7d7>7.7)	=93danlu=94:  d1 i=
s an animal of species d2
+	=93mikce=94:  m1 is a doctor to patient m2 for ailment m3 using treatment=
 m4
+</pre>and the lujvo place structure is:
+<p>
+<pre><a name=3De7d8>7.8)	m1 is a doctor for animal m2=3Dd1 of species d2 f=
or ailment m3
+ using treatment m4
+</pre><cx "animal patient">  XE "animal patient"   Since the shared place =
is m2=3Dd1, the animal patient, the remaining seltau place d2 is inserted i=
mmediately after the shared place; then the remaining tertau places form th=
e last two places of the lujvo.
+<p>
+<a name=3Ds8><h3>lujvo with more than two parts</h3>
+<p>
+<cx "lujvo place order, based on 3-or-more part veljvo">  XE "lujvo place =
order: based on 3-or-more part veljvo"   <ex "tomorrow">  XE "tomorrow: exa=
mple=93  The theory we have outlined so far is an account of lujvo with two=
 parts. But often lujvo are made containing more than two parts. An example=
 is =93bavlamdei=94, =93tomorrow=94: it is composed of the rafsi for =93fut=
ure=94, =93adjacent=94, and =93day=94. How does the account we have given a=
pply to lujvo like this?
+<p>
+The best way to approach such lujvo is to continue to classify them as bas=
ed on binary tanru, the only difference being that the seltau or the tertau=
 or both is itself a lujvo. So it is easiest to make sense of =93bavlamdei=
=94 as having two components: =93bavla'i=94, =93next=94, and =93djedi=94. I=
f we know or invent the lujvo place structure for the components, we can co=
mpose the new lujvo place structure in the usual way.
+<p>
+In this case, =93bavla'i=94 is given the place structure
+<p>
+<pre><a name=3De8d1>8.1)	b1=3Dl1 is next after b2=3Dl2
+</pre>making it a symmetrical lujvo. We combine this with =93djedi=94, whi=
ch has the place structure:
+<p>
+<pre><a name=3De8d2>8.2)	duration d1 is d2 days long (default 1) by standa=
rd d3
+</pre><cx "anomalous ordering of lujvo places">  XE "anomalous ordering of=
 lujvo places"   While symmetrical lujvo normally put any trailing tertau p=
laces before any seltau places, the day standard is a much less important c=
oncept than the day the tomorrow follows, in the definition of =93bavlamdei=
=94. This is an example of how the guidelines presented for selecting and o=
rdering lujvo places are just that, not laws that must be rigidly adhered t=
o. In this case, we choose to rank places in order of relative importance. =
The resulting place structure is:
+<p>
+<pre><a name=3De8d3>8.3)	d1=3Db1=3Dl1 is a day following b2=3Dl2, d2 days =
later (default 1) by standard d3
+</pre><cx "medieval weapon">  XE "medieval weapon"   <ex "long-sword">  XE=
 "long-sword: example=93  Here is another example of a multi-part lujvo: =
=93cladakyxa'i=94, meaning =93long-sword=94, a specific type of medieval we=
apon. The gismu place structures are:
+<p>
+<pre><a name=3De8d4>8.4)	=93clani=94:  c1 is long in direction c2 by stand=
ard c3
+	=93dakfu=94:  d1 is a knife for cutting d2 with blade made of d3
+	=93xarci=94:  xa1 is a weapon for use against xa2 by wielder xa3
+</pre><cx "sword blade">  XE "sword blade"   Since =93cladakyxa'i=94 is a =
symmetrical lujvo based on =93cladakfu xarci=94, and =93cladakfu=94 is itse=
lf a symmetrical lujvo, we can do the necessary analyses all at once. Plain=
ly c1 (the long thing), d1 (the knife), and xa1 (the weapon) are all the sa=
me. Likewise, the d2 place (the thing cut) is the same as the xa2 place (th=
e target of the weapon), given that swords are used to cut victims. Finally=
, the c2 place (direction of length) is always along the sword blade in a l=
ongsword, by definition, and so is dependent on c1=3Dd1=3Dxa1. Adding on th=
e places of the remaining gismu in right-to-left order we get:
+<p>
+<pre><a name=3De8d5>8.5)	xa1=3Dd1=3Dc1 is a long-sword for use against xa2=
=3Dd2 by wielder xa3,=20
+with a blade made of d3, length measured by standard c3.
+</pre><p>If the last place sounds unimportant to you, notice that what cou=
nts legally as a =93sword=94, rather than just a =93knife=94, depends on th=
e length of the blade (the legal limit varies in different jurisdictions). =
This fifth place of =93cladakyxa'i=94 may not often be explicitly filled, b=
ut it is still useful on occasion. Because it is so seldom important, it is=
 best that it be last.
+<p>
+<a name=3Ds9><h3>Eliding SE rafsi from seltau</h3>
+<p>
+<cx "lujvo, dropping SE rafsi">  XE "lujvo: dropping SE rafsi"   It is com=
mon to form lujvo that omit the rafsi based on cmavo of selma'o SE, as well=
 as other cmavo rafsi. Doing so makes lujvo construction for common or usef=
ul constructions shorter. Since it puts more strain on the listener who has=
 not heard the lujvo before, the shortness of the word should not necessari=
ly outweigh ease in understanding, especially if the lujvo refers to a rare=
 or unusual concept.
+<p>
+<cx "proposed law">  XE "proposed law"   Consider as an example the lujvo =
=93ti'ifla=94, from the veljvo =93stidi flalu=94, and meaning =93bill, prop=
osed law=94. The gismu place structures are:
+<p>
+<pre><a name=3De9d1>9.1)	=93stidi=94:  agent st1 suggests idea/action st2 =
to audience st3
+	=93flalu=94:  f1 is a law specifying f2 for community f3 under conditions=
 f4
+by lawgiver f5
+</pre><cx "lined up">  XE "lined up"   This lujvo does not fit any of our =
existing molds: it is the second seltau place, st2, that is equivalent to o=
ne of the tertau places, namely f1. However, if we understand =93ti'ifla=94=
 as an abbreviation for the lujvo =93selti'ifla=94, then we get the first p=
laces of seltau and tertau lined up. The place structure of =93selti'i=94 i=
s: <dl compact><dt>		<dd>
+<a name=3De9d2>9.2)	=93selti'i=94:  idea/action se1 is suggested by agent =
se2 to audience se3
+</dl><p>Here we can see that se1 (what is suggested) is equivalent to f1 (=
the law), and we get a normal symmetrical lujvo. The final place structure =
is:
+<p>
+<pre><a name=3De9d3>9.3)	f1=3Dse1 is a bill specifying f2 for community f3=
 under conditions f4
+by suggester se2 to audience/lawgivers f5=3Dse3
+</pre>or, relabeling the places,
+<p>
+<pre><a name=3De9d4>9.4)	f1=3Dst2 is a bill specifying f2 for community f3=
 under conditions f4
+by suggester st1 to audience/lawgivers f5=3Dst3
+</pre>where the last place (st3) is probably some sort of legislature.
+<p>
+<cx "lujvo, abbreviated">  XE "lujvo: abbreviated"   Abbreviated lujvo lik=
e =93ti'ifla=94 are more intuitive (for the lujvo-maker) than their more ex=
plicit counterparts like =93selti'ifla=94 (as well as shorter). They don't =
require the coiner to sit down and work out the precise relation between th=
e seltau and the tertau: he or she can just rattle off a rafsi pair. But sh=
ould the lujvo get to the stage where a place structure needs to be worked =
out, then the precise relation does need to be specified. And in that case,=
 such abbreviated lujvo form a trap in lujvo place ordering, since they obs=
cure the most straightforward relation between the seltau and tertau. To gi=
ve our lujvo-making guidelines as wide an application as possible, and to e=
ncourage analyzing the seltau-tertau relation in lujvo, lujvo like =93ti'if=
la=94 are given the place structure they would have with the appropriate SE=
 added to the seltau.
+<p>
+<cx "implausible">  XE "implausible"   <cx "abbreviated lujvo and plausibi=
lity">  XE "abbreviated lujvo and plausibility"   <cx "plausibility, in abb=
reviated lujvo">  XE "plausibility: in abbreviated lujvo"   Note that, with=
 these lujvo, an interpretation requiring SE insertion is safe only if the =
alternatives are either implausible or unlikely to be needed as a lujvo. Th=
is may not always be the case, and Lojbanists should be aware of the risk o=
f ambiguity.
+<p>
+<a name=3Ds10><h3>Eliding SE rafsi from tertau</h3>
+<p>
+<cx "lujvo place structure, effect of =93SE=94-dropping in tertau">  XE "l=
ujvo place structure: effect of SE-dropping in tertau"   Eliding SE rafsi f=
rom tertau gets us into much more trouble. To understand why, recall that l=
ujvo, following their veljvo, describe some type of whatever their tertau d=
escribe. Thus, =93posydji=94 describes a type of =93djica=94, =93gerzda=94 =
describes a type of =93zdani=94, and so on. What is certain is that =93gerz=
da=94 does not describe a =93se zdani=94 =97 it is not a word that could be=
 used to describe an inhabitant such as a dog.
+<p>
+<ex "blue-eyed">  XE "blue-eyed: example=93  Now consider how we would tra=
nslate the word =93blue-eyed=94. Let's tentatively translate this word as =
=93blakanla=94 (from =93blanu kanla=94, meaning =93blue eye=94). But immedi=
ately we are in trouble: we cannot say
+<p>
+<pre><a name=3De10d1>10.1)	la djak. cu blakanla
+	Jack is-a-blue-eye
+</pre>because Jack is not an eye, =93kanla=94, but someone with an eye, =
=93se kanla=94. At best we can say
+<p>
+<pre><a name=3De10d2>10.2)	la djak. cu se blakanla
+	Jack is-the-bearer-of-blue-eyes
+</pre><p>But look now at the place structure of =93blakanla=94: it is a sy=
mmetrical lujvo, so the place structure is:
+<p>
+<pre><a name=3De10d3>10.3)	bl1=3Dk1 is a blue eye of bl2=3Dk2
+</pre><p>We end up being most interested in talking about the second place=
, not the first (we talk much more of people than of their eyes), so =93se=
=94 would almost always be required.
+<p>
+What is happening here is that we are translating the tertau wrongly, unde=
r the influence of English. The English suffix =93-eyed=94 does not mean =
=93eye=94, but someone with an eye, which is =93selkanla=94.
+<p>
+Because we've got the wrong tertau (eliding a =93se=94 that really should =
be there), any attempt to accommodate the resulting lujvo into our guidelin=
es for place structure is fitting a square peg in a round hole. Since they =
can be so misleading, lujvo with SE rafsi elided from the tertau should be =
avoided in favor of their more explicit counterparts: in this case, =93blas=
elkanla=94.
+<p>
+<a name=3Ds11><h3>Eliding KE and KEhE rafsi from lujvo</h3>
+<p>
+<cx "lujvo place structure, dropping =93KE=94">  XE "lujvo place structure=
: dropping KE"   <cx "lujvo place structure, dropping =93KEhE=94">  XE "luj=
vo place structure: dropping KEhE"   People constructing lujvo usually want=
 them to be as short as possible. To that end, they will discard any cmavo =
they regard as niceties. The first such cmavo to get thrown out are usually=
 =93ke=94 and =93ke'e=94, the cmavo used to structure and group tanru. We c=
an usually get away with this, because the interpretation of the tertau wit=
h =93ke=94 and =93ke'e=94 missing is less plausible than that with the cmav=
o inserted, or because the distinction isn't really important.
+<p>
+<cx "beefsteak">  XE "beefsteak"   <ex "beefsteak">  XE "beefsteak: exampl=
e=93  For example, in =93bakrecpa'o=94, meaning =93beefsteak=94, the veljvo=
 is
+<p>
+  XE "bovine: example"  <pre><a name=3De11d1>11.1)	[ke] bakni rectu [ke'e]=
 panlo
+	( bovine meat ) slice
+</pre>because of the usual Lojban left-grouping rule. But there doesn't se=
em to be much difference between that veljvo and
+<p>
+  XE "meat slice: example"  <pre><a name=3De11d2>11.2)	bakni ke rectu panl=
o [ke'e]
+	bovine ( meat slice )
+</pre><ex "sneak in">  XE "sneak in: example=93  On the other hand, the lu=
jvo =93zernerkla=94, meaning =93to sneak in=94, almost certainly was formed=
 from the veljvo
+<p>
+<pre><a name=3De11d3>11.3)	zekri ke nenri klama [ke'e]
+	crime ( inside go )
+	to go within, criminally
+</pre>because the alternative,
+<p>
+<pre><a name=3De11d4>11.4)	[ke] zekri nenri [ke'e] klama
+	(crime inside) go
+</pre>doesn't make much sense. (To go to the inside of a crime? To go into=
 a place where it is criminal to be inside =97 an interpretation almost ide=
ntical with <a href=3D#e11d3>Example 11.3 </a>anyway?)
+<p>
+<cx "shellfish">  XE "shellfish"   <ex "shellfish">  XE "shellfish: exampl=
e=93  There are cases, however, where omitting a KE or KEhE rafsi can produ=
ce another lujvo, equally useful. For example, =93xaskemcakcurnu=94 means =
=93oceanic shellfish=94, and has the veljvo
+<p>
+  XE "shell worm: example"  <pre><a name=3De11d5>11.5)	xamsi ke calku curn=
u
+	ocean type-of (shell worm)
+</pre><cx "invertebrate">  XE "invertebrate"   (=93worm=94 in Lojban refer=
s to any invertebrate), but =93xascakcurnu=94 has the veljvo
+
+<p>
+  XE "ocean shell: example"  <pre><a name=3De11d6>11.6)	[ke] xamsi calku [=
ke'e] curnu
+	(ocean shell) type-of worm
+</pre><cx "parasitic worms">  XE "parasitic worms: example"   <cx "clamshe=
lls">  XE "clamshells"   and might refer to the parasitic worms that infest=
 clamshells.
+<p>
+<cx "misinterpretation">  XE "misinterpretation"   <cx "lujvo creation, in=
teraction of KE with SE">  XE "lujvo creation: interaction of KE with SE"  =
 <cx "lujvo creation, interaction of KE with NAhE">  XE "lujvo creation: in=
teraction of KE with NAhE"   Such misinterpretation is more likely than not=
 in a lujvo starting with =93sel-=94 (from =93se=94), =93nal-=94 (from =93n=
a'e=94) or =93tol-=94 (from =93to'e=94): the scope of the rafsi will likeli=
est be presumed to be as narrow as possible, since all of these cmavo norma=
lly bind only to the following brivla or =93ke=A0<dots>=85</dots>ke'e=94 gr=
oup. For that reason, if we want to modify an entire lujvo by putting =93se=
=94, =93na'e=94 or =93to'e=94 before it, it's better to leave the result as=
 two words, or else to insert =93ke=94, than to just stick the SE or NAhE r=
afsi on.
+<p>
+It is all right to replace the phrase =93se klama=94 with =93selkla=94, an=
d the places of =93selkla=94 are exactly those of =93se klama=94. But consi=
der the related lujvo =93dzukla=94, meaning =93to walk to somewhere=94. It =
is a symmmetrical lujvo, derived from the veljvo =93cadzu klama=94 as follo=
ws:
+<p>
+<pre><a name=3De11d7>	=93cadzu=94:  c1 walks on surface c2 using limbs c3
+	=93klama=94:  k1 goes to k2 from k3 via route k4 using k5
+	=93dzukla=94:  c1=3Dk1 walks to k2 from k3 via route k4 using limbs k5=3D=
c3
+		on surface c2
+</pre><p>We can swap the k1 and k2 places using =93se dzukla=94, but we ca=
nnot directly make =93se dzukla=94 into =93seldzukla=94, which would repres=
ent the veljvo =93selcadzu klama=94 and plausibly mean something like =93to=
 go to a walking surface=94. Instead, we would need =93selkemdzukla=94, wit=
h an explicit rafsi for =93ke=94. Similarly, =93nalbrablo=94 (from =93na'e =
barda bloti=94) means =93non-big boat=94, whereas =93na'e brablo=94 means =
=93other than a big boat=94.
+<p>
+<cx "lujvo creation, use of multiple SE in">  XE "lujvo creation: use of m=
ultiple SE in"   If the lujvo we want to modify with SE has a seltau alread=
y starting with a SE rafsi, we can take a shortcut. For instance, =93gekmau=
=94 means =93happier than=94, while =93selgekmau=94 means =93making people =
happier than, more enjoyable than, more of a 'se gleki' than=94. If somethi=
ng is less enjoyable than something else, we can say it is =93se selgekmau=
=94.
+<p>
+But we can also say it is =93selselgekmau=94. Two =93se=94 cmavo in a row =
cancel each other (=93se se gleki=94 means the same as just =93gleki=94), s=
o there would be no good reason to have =93selsel=94 in a lujvo with that m=
eaning. Instead, we can feel free to interpret =93selsel-=94 as =93selkemse=
l-=94. The rafsi combinations =93terter-=94, =93velvel-=94 and =93xelxel-=
=94 work in the same way.
+<p>
+Other SE combinations like =93selter-=94, although they might conceivably =
mean =93se te=94, more than likely should be interpreted in the same way, n=
amely as =93se ke te=94, since there is no need to re-order places in the w=
ay that =93se te=94 provides. (See <a href=3Dchap9.html>Chapter 9</a>.)
+<p>
+<a name=3Ds12><h3>Abstract lujvo</h3>
+<p>
+<cx "lujvo, abstract">  XE "lujvo: abstract"   <cx "abstract lujvo">  XE "=
abstract lujvo"   <cx "lujvo place structure, =93nu=94 lujvo">  XE "lujvo p=
lace structure: \=93nu\=94 lujvo"   The cmavo of NU can participate in the =
construction of lujvo of a particularly simple and well-patterned kind. Con=
sider that old standard example, =93klama=94:
+<p>
+<pre><a name=3De12d1>12.1)	k1 comes/goes to k2 from k3 via route k4 by mea=
ns k5.
+</pre><p>The selbri =93nu klama [kei]=94 has only one place, the event-of-=
going, but the full five places exist implicitly between =93nu=94 and =93ke=
i=94, since a full bridi with all sumti may be placed there. In a lujvo, th=
ere is no room for such inside places, and consequently the lujvo =93nunkla=
=94 (=93nun-=94 is the rafsi for =93nu=94), needs to have six places:
+<p>
+<pre><a name=3De12d2>12.2)	nu1 is the event of k1's coming/going to k2 fro=
m k3 via route k4=20
+by means k5.
+</pre><p>Here the first place of =93nunklama=94 is the first and only plac=
e of =93nu=94, and the other five places have been pushed down by one to oc=
cupy the second through the sixth places. Full information on =93nu=94, as =
well as the other abstractors mentioned in this section, is given in <a hre=
f=3Dchap11.html>Chapter 11</a>.
+<p>
+<cx "lujvo place structure, =93ni=94 lujvo">  XE "lujvo place structure: \=
=93ni\=94 lujvo"   <cx "lujvo place structure, multi-place abstraction lujv=
o">  XE "lujvo place structure: multi-place abstraction lujvo"   For those =
abstractors which have a second place as well, the standard convention is t=
o place this place after, rather than before, the places of the brivla bein=
g abstracted. The place structure of =93nilkla=94, the lujvo derived from =
=93ni klama=94, is the imposing:
+<p>
+<pre><a name=3De12d3>12.3)	ni1 is the amount of k1's coming/going to k2 fr=
om k3 via route k4=20
+by means k5, measured on scale ni2.
+</pre><p>It is not uncommon for abstractors to participate in the making o=
f more complex lujvo as well. For example, =93nunsoidji=94, from the veljvo
+<p>
+<pre><a name=3De12d4>12.4)	nu sonci kei djica
+	event-of being-a-soldier desirer
+</pre>has the place structure
+<p>
+<pre><a name=3De12d5>12.5)	d1 desires the event of (s1 being a soldier of =
army s2) for purpose d3
+</pre><cx "lujvo place order, multi-part with =93NU=94">  XE "lujvo place =
order: multi-part with NU"   where the d2 place has disappeared altogether,=
 being replaced by the places of the seltau. As shown in <a href=3D#e12d5>E=
xample 12.5</a>, the ordering follows this idea of replacement: the seltau =
places are inserted at the point where the omitted abstraction place exists=
 in the tertau.
+<p>
+<cx "asymmetric lujvo">  XE "asymmetric lujvo"   The lujvo =93nunsoidji=94=
 is quite different from the ordinary asymmetric lujvo =93soidji=94, a =93s=
oldier desirer=94, whose place structure is just
+<p>
+<pre><a name=3De12d6>12.6)	d1 desires (a soldier of army s2) for purpose d=
3
+</pre><p>A =93nunsoidji=94 might be someone who is about to enlist, wherea=
s a =93soidji=94 might be a camp-follower.
+<p>
+<cx "abstract lujvo, contrasted with abstract bridi">  XE "abstract lujvo:=
 contrasted with abstract bridi"   One use of abstract lujvo is to eliminat=
e the need for explicit =93kei=94 in tanru: =93nunkalri gasnu=94 means much=
 the same as =93nu kalri kei gasnu=94, but is shorter. In addition, many En=
glish words ending in =93-hood=94 are represented with =93nun-=94 lujvo, an=
d other words ending in =93-ness=94 or =93-dom=94 are often representable w=
ith =93kam-=94 lujvo (=93kam-=94 is the rafsi for =93ka=94); =93kambla=94 i=
s =93blueness=94.
+<p>
+<cx "lujvo, scope abstraction in underlying veljvo">  XE "lujvo: scope abs=
traction in underlying veljvo"   Even though the cmavo of NU are long-scope=
 in nature, governing the whole following bridi, the NU rafsi should genera=
lly be used as short-scope modifiers, like the SE and NAhE rafsi discussed =
in <a href=3D#s9>Section 9</a>.
+<p>
+<cx "lujvo, with =93jai=94">  XE "lujvo: with \=93jai\=94"   There is also=
 a rafsi for the cmavo =93jai=94, namely =93jax=94, which allows sentences =
like
+<p>
+<ex "cause death">  XE "cause death: example=93  <pre><a name=3De12d7>12.7=
)	mi jai rinka le nu do morsi
+	I am-associated-with causing the event-of your death.
+	I cause your death.
+</pre>explained in <a href=3Dchap11.html>Chapter 11</a>, to be rendered wi=
th lujvo:
+<p>
+<pre><a name=3De12d8>12.8)	mi jaxri'a le nu do morsi
+	I am-part-of-the-cause-of the event-of your dying.
+</pre><cx "lujvo place structure, with =93jai=94 lujvo">  XE "lujvo place =
structure: with \=93jai\=94 lujvo"   <lx "fai">  XE "fai"   <lx "jai">  XE =
"jai"   In making a lujvo that contains =93jax-=94 for a selbri that contai=
ns =93jai=94, the rule is to leave the =93fai=94 place as a =93fai=94 place=
 of the lujvo; it does not participate in the regular lujvo place structure=
. (The use of =93fai=94 is also explained in <a href=3Dchap11.html>Chapter =
11</a>.)
+<p>
+<a name=3Ds13><h3>Implicit-abstraction lujvo</h3>
+<p>
+<cx "lujvo, dropping NU rafsi">  XE "lujvo: dropping NU rafsi"   Eliding N=
U rafsi involves the same restrictions as eliding SE rafsi, plus additional=
 ones. In general, NU rafsi should not be elided from the tertau, since tha=
t changes the kind of thing the lujvo is talking about from an abstraction =
to a concrete sumti. However, they may be elided from the seltau if no reas=
onable ambiguity would result.
+<p>
+<cx "lujvo, NU-dropping contrasted with SE-dropping">  XE "lujvo: NU-dropp=
ing contrasted with SE-dropping"   <cx "lujvo, implicit-abstraction">  XE "=
lujvo: implicit-abstraction"   A major difference, however, between SE elis=
ion and NU elision is that the former is a rather sparse process, providing=
 a few convenient shortenings. Eliding =93nu=94, however, is extremely impo=
rtant in producing a class of lujvo called =93implicit-abstraction lujvo=94=
.
+<p>
+<ex "feed">  XE "feed: example=93  Let us make a detailed analysis of the =
lujvo =93nunctikezgau=94, meaning =93to feed=94. (If you think this lujvo i=
s excessively longwinded, be patient.) The veljvo of =93nunctikezgau=94 is =
=93nu citka kei gasnu=94. The relevant place structures are:
+<p>
+<pre><a name=3De13d1>13.1)	=93nu=94:  n1 is an event
+	=93citka=94:  c1 eats c2
+	=93gasnu=94:  g1 does action/is the agent of event g2
+</pre><p>In accordance with the procedure for analyzing three-part lujvo g=
iven in <a href=3D#s8>Section 8</a>, we will first create an intermediate l=
ujvo, =93nuncti=94, whose veljvo is =93nu citka [kei]=94. By the rules give=
n in <a href=3D#s12>Section 12</a>, =93nuncti=94 has the place structure
+<p>
+<pre><a name=3De13d2>13.2)	n1 is the event of c1 eating c2
+</pre><cx "lujvo place structure, dropping first place of NU">  XE "lujvo =
place structure: dropping first place of NU"   Now we can transform the vel=
jvo of =93nunctikezgau=94 into =93nuncti gasnu=94. The g2 place (what is br=
ought about by the actor g1) obviously denotes the same thing as n1 (the ev=
ent of eating). So we can eliminate g2 as redundant, leaving us with a tent=
ative place structure of
+<p>
+<pre><a name=3De13d3>13.3)	g1 is the actor in the event n1=3Dg2 of c1 eati=
ng c2
+</pre><p>But it is also possible to omit the n1 place itself! The n1 place=
 describes the event brought about; an event in Lojban is described as a br=
idi, by a selbri and its sumti; the selbri is already known (it's the selta=
u), and the sumti are also already known (they're in the lujvo place struct=
ure). So n1 would not give us any information we didn't already know. In fa=
ct, the n1=3Dg2 place is dependent on c1 and c2 jointly =97 it does not dep=
end on either c1 or c2 by itself. Being dependent and derived from the selt=
au, it is omissible.
+So the final place structure of =93nunctikezgau=94 is:
+<p>
+<pre><a name=3De13d4>13.4)	g1 is the actor in the event of c1 eating c2
+</pre><cx "lujvo, asymmetric abstraction">  XE "lujvo: asymmetric abstract=
ion"   <cx "abstraction lujvo, asymmetric">  XE "abstraction lujvo: asymmet=
ric"   There is one further step that can be taken. As we have already seen=
 with =93balsoi=94 in <a href=3D#s5>Section 5</a>, the interpretation of lu=
jvo is constrained by the semantics of gismu and of their sumti places. Now=
, any asymmetrical lujvo with =93gasnu=94 as its tertau will involve an eve=
nt abstraction either implicitly or explicitly, since that is how the g2 pl=
ace of =93gasnu=94 is defined.
+<p>
+<cx "lujvo, dropping NU in implicit abstractions">  XE "lujvo: dropping NU=
 in implicit abstractions"   Therefore, if we assume that =93nu=94 is the t=
ype of abstraction one would expect to be a =93se gasnu=94, then the rafsi =
=93nun=94 and =93kez=94 in =93nunctikezgau=94 are only telling us what we w=
ould already have guessed =97 that the seltau of a =93gasnu=94 lujvo is an =
event. If we drop these rafsi out, and use instead the shorter lujvo =93cti=
gau=94, rejecting its symmetrical interpretation (=93someone who both does =
and eats=94; =93an eating doer=94), we can still deduce that the seltau ref=
ers to an event.
+<p>
+(You can't =93do an eater=94/=93gasnu lo citka=94, with the meaning of =93=
do=94 as =93bring about an event=94; so the seltau must refer to an event, =
=93nu citka=94. The English slang meanings of =93do someone=94, namely =93s=
ocialize with someone=94 and =93have sex with someone=94, are not relevant =
to =93gasnu=94.)
+<p>
+So we can simply use =93ctigau=94 with the same place structure as =93nunc=
tikezgau=94:
+<p>
+<pre><a name=3De13d5>13.5)	agent g1 causes c1 to eat c2
+	g1 feeds c2 to c1.
+</pre><cx "implicit-abstraction lujvo, definition">  XE "implicit-abstract=
ion lujvo: definition"   This particular kind of asymmetrical lujvo, in whi=
ch the seltau serves as the selbri of an abstraction which is a place of th=
e tertau, is called an implicit-abstraction lujvo, because one deduces the =
presence of an abstraction which is unexpressed (implicit).
+<p>
+<ex "replace">  XE "replace: example=93  To give another example: the gism=
u =93basti=94, whose place structure is
+<p>
+<pre><a name=3De13d6>13.6)	b1 replaces b2 in circumstances b3
+</pre>can form the lujvo =93basygau=94, with the place structure:
+<p>
+<pre><a name=3De13d7>13.7)	g1 (agent) replaces b1 with b2 in circumstances=
 b3
+</pre>where both =93basti=94 and =93basygau=94 are translated =93replace=
=94 in English, but represent different relations: =93basti=94 may be used =
with no mention of any agent doing the replacing.
+<p>
+In addition, =93gasnu=94-based lujvo can be built from what we would consi=
der nouns or adjectives in English. In Lojban, everything is a predicate, s=
o adjectives, nouns and verbs are all treated in the same way. This is cons=
istent with the use of similar causative affixes in other languages. For ex=
ample, the gismu =93litki=94, meaning =93liquid=94, with the place structur=
e
+<p>
+<ex "liquefy">  XE "liquefy: example=93  <pre><a name=3De13d8>13.8)	l1 is =
a quantity of liquid of composition l2 under conditions l3
+</pre>can give =93likygau=94, meaning =93to liquefy=94:
+<p>
+<pre><a name=3De13d9>13.9)	g1 (agent) causes l1 to be a quantity of liquid=
 of composition l2
+under conditions l3.
+</pre><p>While =93likygau=94 correctly represents =93causes to be a liquid=
=94, a different lujvo based on =93galfi=94 (meaning =93modify=94) may be m=
ore appropriate for =93causes to become a liquid=94. On the other hand, =93=
fetsygau=94 is potentially confusing, because it could mean =93agent in the=
 event of something becoming female=94 (the implicit-abstraction interpreta=
tion) or simply =93female agent=94 (the parallel interpretation), so using =
implicit-abstraction lujvo is always accompanied with some risk of being mi=
sunderstood.
+<p>
+<ex "sun liquefies">  XE "sun liquefies: example=93  Many other Lojban gis=
mu have places for event abstractions, and therefore are good candidates fo=
r the tertau of an implicit-abstraction lujvo. For example, lujvo based on =
=93rinka=94, with its place structure
+<p>
+<pre><a name=3De13d10>13.10)	event r1 causes event r2 to occur
+</pre>are closely related to those based on =93gasnu=94. However, =93rinka=
=94 is less generally useful than =93gasnu=94, because its r1 place is anot=
her event rather than a person: =93lo rinka=94 is a cause, not a causer. Th=
us the place structure of =93likyri'a=94, a lujvo analogous to =93likygau=
=94, is
+<p>
+<pre><a name=3De13d11>13.11)	event r1 causes l1 to be a quantity of liquid=
 of composition l2
+under conditions l3
+</pre>and would be useful in translating sentences like =93The heat of the=
 sun liquefied the block of ice.=94
+<p>
+<cx "expressive power">  XE "expressive power"   Implicit-abstraction lujv=
o are a powerful means in the language of rendering quite verbose bridi int=
o succinct and manageable concepts, and increasing the expressive power of =
the language.
+<p>
+<a name=3Ds14><h3>Anomalous lujvo</h3>
+<p>
+<cx "unusual position">  XE "unusual position"   <cx "lujvo, anomalous">  =
XE "lujvo: anomalous"   Some lujvo that have been coined and actually emplo=
yed in Lojban writing do not follow the guidelines expressed above, either =
because the places that are equivalent in the seltau and the tertau are in =
an unusual position, or because the seltau and tertau are related in a comp=
lex way, or both. An example of the first kind is =93jdaselsku=94, meaning =
=93prayer=94, which was mentioned in <a href=3D#s7>Section 7</a>. The gismu=
 places are:
+<p>
+<cx "lujvo place order, redundant non-first places">  XE "lujvo place orde=
r: redundant non-first places"   <ex "prayer">  XE "prayer: example=93  <pr=
e><a name=3De14d1>14.1)	=93lijda=94:  l1 is a religion with believers l2 an=
d beliefs l3
+	=93cusku=94:  c1 expresses text c2 to audience c3 in medium c4
+</pre>and =93selsku=94, the tertau of =93jdaselsku=94, has the place struc=
ture
+<p>
+<pre><a name=3De14d2>14.2)	s1 is a text expressed by s2 to audience s3 in =
medium s4
+Now it is easy to see that the l2 and s2 places are equivalent: the believ=
er in the religion (l2) is the one who expresses the prayer (s2). This is n=
ot one of the cases for which a place ordering rule has been given in <a hr=
ef=3D#s7>Section 7 </a>or <a href=3D#s13>Section 13</a>; therefore, for lac=
k of a better rule, we put the tertau places first and the remaining seltau=
 places after them, leading to the place structure:
+<a name=3De14d3>14.3)	s1 is a prayer expressed by s2=3Dl2 to audience s3 i=
n medium s4=20
+pertaining to religion l1
+</pre><p>The l3 place (the beliefs of the religion) is dependent on the l1=
 place (the religion) and so is omitted.
+<p>
+We could make this lujvo less messy by replacing it with =93se seljdasku=
=94, where =93seljdasku=94 is a normal symmetrical lujvo with place structu=
re:
+<p>
+<pre><a name=3De14d4>14.4)	c1=3Dl2 religiously expresses prayer c2 to audi=
ence c3 in medium s4=20
+pertaining to religion l1
+</pre>which, according to the rule expressed in <a href=3D#s9>Section 9</a=
>, can be further expressed as =93selseljdasku=94. However, there is no nee=
d for the ugly =93selsel-=94 prefix just to get the rules right: =93jdasels=
ku=94 is a reasonable, if anomalous, lujvo.
+<p>
+However, there is a further problem with =93jdaselsku=94, not resolvable b=
y using =93seljdasku=94. No veljvo involving just the two gismu =93lijda=94=
 and =93cusku=94 can fully express the relationship implicit in prayer. A p=
rayer is not just anything said by the adherents of a religion; nor is it e=
ven anything said by them acting as adherents of that religion. Rather, it =
is what they say under the authority of that religion, or using the religio=
n as a medium, or following the rules associated with the religion, or some=
thing of the kind. So the veljvo is somewhat elliptical.
+<p>
+As a result, both =93seljdasku=94 and =93jdaselsku=94 belong to the second=
 class of anomalous lujvo: the veljvo doesn't really supply all that the lu=
jvo requires.
+<p>
+<cx "sheepdog">  XE "sheepdog"   <ex "sheepdog">  XE "sheepdog: example=93=
  <cx "sheep breed">  XE "sheep breed"   <cx "lujvo place order, non-overla=
pping place structures">  XE "lujvo place order: non-overlapping place stru=
ctures"   <cx "lujvo place order, complex relation">  XE "lujvo place order=
: complex relation"   Another example of this kind of anomalous lujvo, draw=
n from the tanru lists in <a href=3Dchap5.html>Chapter 5</a>, is =93lange'u=
=94, meaning =93sheepdog=94. Clearly a sheepdog is not a dog which is a she=
ep (the symmetrical interpretation is wrong), nor a dog of the sheep breed =
(the asymmetrical interpretation is wrong). Indeed, there is simply no over=
lap in the places of =93lanme=94 and =93gerku=94 at all. Rather, the lujvo =
refers to a dog which controls sheep flocks, a =93terlanme jitro gerku=94, =
the lujvo from which is =93terlantroge'u=94 with place structure:
+<p>
+  XE "sheep flock: example"  <pre><a name=3De14d5>14.5)	g1=3Dj1 is a dog t=
hat controls sheep flock l3=3Dj2 made up of sheep l1=20
+in activity j3 of dog breed g2
+</pre>based on the gismu place structures
+<p>
+<pre><a name=3De14d6>14.6)	=93lanme=94:  l1 is a sheep of breed l2 belongi=
ng to flock l3
+	=93gerku=94:  g1 is a dog of breed g2
+	=93jitro=94:  j1 controls j2 in activity j3
+</pre><p>Note that this lujvo is symmetrical between =93lantro=94 (sheep-c=
ontroller) and =93gerku=94, but =93lantro=94 is itself a asymmetrical lujvo=
. The l2 place, the breed of sheep, is removed as dependent on l1. However,=
 the lujvo =93lange'u=94 is both shorter than =93terlantroge'u=94 and suffi=
ciently clear to warrant its use: its place structure, however, should be t=
he same as that of the longer lujvo, for which =93lange'u=94 can be underst=
ood as an abbreviation.
+<p>
+<ex "beckon">  XE "beckon: example=93  <cx "lujvo place order, elliptical =
lujvo">  XE "lujvo place order: elliptical lujvo"   Another example is =93x=
anmi'e=94, =93to command by hand, to beckon=94. The component place structu=
res are:
+<p>
+<pre><a name=3De14d7>14.7)	=93xance=94:  xa1 is the hand of xa2
+	=93minde=94:  m1 gives commands to m2 to cause m3 to happen
+</pre><p>The relation between the seltau and tertau is close enough for th=
ere to be an overlap: xa2 (the person with the hand) is the same as m1 (the=
 one who commands). But interpreting =93xanmi'e=94 as a symmetrical lujvo w=
ith an elided =93sel-=94 in the seltau, as if from =93se xance mindu=94, mi=
sses the point: the real relation expressed by the lujvo is not just =93one=
 who commands and has a hand=94, but =93to command using the hand=94. The c=
oncept of =93using=94 suggests the gismu =93pilno=94, with place structure
+<p>
+<pre><a name=3De14d8>14.8)	p1 uses tool p2 for purpose p3
+</pre><p>Some possible three-part veljvo are (depending on how strictly yo=
u want to constrain the veljvo)
+<p>
+<pre><a name=3De14d9>14.9)	[ke] xance pilno [ke'e] minde
+	(hand user) type-of commander
+<a name=3De14d10>14.10)	[ke] minde xance [ke'e] pilno
+	(commander hand) type-of user
+</pre>or even
+<p>
+<pre><a name=3De14d11>14.11)	minde ke xance pilno [ke'e]
+	commander type-of (hand user)
+</pre>which lead to the three different lujvo =93xanplimi'e=94, =93mi'erxa=
npli=94, and =93minkemxanpli=94 respectively.
+<p>
+<cx "latent component">  XE "latent component"   Does this make =93xanmi'e=
=94 wrong? By no means. But it does mean that there is a latent component t=
o the meaning of =93xanmi'e=94, the gismu =93pilno=94, which is not explici=
t in the veljvo. And it also means that, for a place structure derivation t=
hat actually makes sense, rather than being ad-hoc, the Lojbanist should pr=
obably go through a derivation for =93xancypliminde=94 or one of the other =
possibilities that is analogous to the analysis of =93terlantroge'u=94 abov=
e, even if he or she decides to stick with a shorter, more convenient form =
like =93xanmi'e=94. In addition, of course, the possibilities of elliptical=
 lujvo increase their potential ambiguity enormously =97 an unavoidable fac=
t which should be borne in mind.
+<p>
+<a name=3Ds15><h3>Comparatives and superlatives</h3>
+<p>
+<cx "lujvo, comparatives">  XE "lujvo: comparatives"   <cx "lujvo, superla=
tives">  XE "lujvo: superlatives"   English has the concepts of =93comparat=
ive adjectives=94 and =93superlative adjectives=94 which can be formed from=
 other adjectives, either by adding the suffixes =93-er=94 and =93-est=94 o=
r by using the words =93more=94 and =93most=94, respectively. The Lojbanic =
equivalents, which can be made from any brivla, are lujvo with the tertau =
=93zmadu=94, =93mleca=94, =93zenba=94, =93jdika=94, and =93traji=94. In ord=
er to make these lujvo regular and easy to make, certain special guidelines=
 are imposed.
+<p>
+We will begin with lujvo based on =93zmadu=94 and =93mleca=94, whose place=
 structures are:
+<p>
+<pre><a name=3De15d1>15.1)	=93zmadu=94:  z1 is more than z2 in property z3=
 in quantity z4
+	=93mleca=94:  m1 is less than m2 in property m3 in quantity m4
+</pre><p>For example, the concept =93young=94 is expressed by the gismu =
=93citno=94, with place structure
+<p>
+<pre><a name=3De15d2>15.2)	=93citno=94:  c1 is young
+</pre><cx "lujvo place order, comparatives">  XE "lujvo place order: compa=
ratives"   <ex "younger">  XE "younger: example=93  The comparative concept=
 =93younger=94 can be expressed by the lujvo =93citmau=94 (based on the vel=
jvo =93citno zmadu=94, meaning =93young more-than=94).
+<p>
+<pre><a name=3De15d3>15.3)	mi citmau do lo nanca be li xa
+	I am-younger-than you by-years the-number six.
+	I am six years younger than you.
+</pre><p>The place structure for =93citmau=94 is
+<p>
+<pre><a name=3De15d4>15.4)	z1=3Dc1 is younger than z2=3Dc1 by amount z4
+</pre><p>Similarly, in Lojban you can say:
+<p>
+<pre><a name=3De15d5>15.5)	do citme'a mi lo nanca be li xa
+	You are-less-young-than me by-years the-number six.
+	You are six years less young than me.
+</pre><p>In English, =93more=94 comparatives are easier to make and use th=
an =93less=94 comparatives, but in Lojban the two forms are equally easy.
+<p>
+Because of their much simpler place structure, lujvo ending in =93mau=94 a=
nd =93me'a=94 are in fact used much more frequently than =93zmadu=94 and =
=93mleca=94 themselves as selbri. It is highly unlikely for such lujvo to b=
e construed as anything other than implicit-abstraction lujvo. But there is=
 another type of ambiguity relevant to these lujvo, and which has to do wit=
h what is being compared.
+<p>
+<cx "comparative lujvo, potential ambiguity in">  XE "comparative lujvo: p=
otential ambiguity in"   For example, does =93nelcymau=94 mean =93X likes Y=
 more than X likes Z=94, or =93X likes Y more than Z likes Y=94? Does =93kl=
amau=94 mean: =93X goes to Y more than to Z=94, =93X goes to Y more than Z =
does=94, =93X goes to Y from Z more than from W=94, or what?
+<p>
+<cx "comparative lujvo, standardized meanings">  XE "comparative lujvo: st=
andardized meanings"   <cx "lujvo place structure, comparative lujvo">  XE =
"lujvo place structure: comparative lujvo"   We answer this concern by putt=
ing regularity above any considerations of concept usefulness: by conventio=
n, the two things being compared always fit into the first place of the sel=
tau. In that way, each of the different possible interpretations can be exp=
ressed by SE-converting the seltau, and making the required place the new f=
irst place. As a result, we get the following comparative lujvo place struc=
tures:
+<p>
+<pre><a name=3De15d6>15.6)	=93nelcymau=94:  z1, more than z2, likes n2 by =
amount z4
+	=93selnelcymau=94:  z1, more than z2, is liked by n1 in amount z4
+	=93klamau=94:  z1, more than z2, goes to k2 from k3 via k4 by means of k5
+	=93selklamau=94:  z1, more than z2, is gone to by k1 from k3 via k4=20
+by means of k5
+	=93terklamau=94:  z1, more than z2, is an origin point from destination k=
2=20
+for k1's going via k4 by means of k5
+</pre>(See <a href=3Dchap11.html>Chapter 11 </a>for the way in which this =
problem is resolved when lujvo aren't used.)
+<p>
+The ordering rule places the things being compared first, and the other se=
ltau places following. Unfortunately the z4 place, which expresses by how m=
uch one entity exceeds the other, is displaced into a lujvo place whose num=
ber is different for each lujvo. For example, while =93nelcymau=94 has z4 a=
s its fourth place, =93klamau=94 has it as its sixth place. In any sentence=
 where a difficulty arises, this amount-place can be redundantly tagged wit=
h =93vemau=94 (for =93zmadu=94) or =93veme'a=94 (for =93mleca=94) to help m=
ake the speaker's intention clear.
+<p>
+<cx "nonagenarian">  XE "nonagenarian"   <cx "octogenarian">  XE "octogena=
rian"   <cx "comparative lujvo, and seltau presupposition">  XE "comparativ=
e lujvo: and seltau presupposition"   It is important to realize that such =
comparative lujvo do not presuppose their seltau. Just as in English, sayin=
g someone is younger than someone else doesn't imply that they're young in =
the first place: an octogenarian, after all, is still younger than a nonage=
narian. Rather, the 80-year-old has a greater =93ni citno=94 than the 90-ye=
ar-old. Similarly, a 5-year-old is older than a 1-year-old, but is not cons=
idered =93old=94 by most standards.
+<p>
+<cx "former state">  XE "former state"   <cx "comparative lujvo, against f=
ormer state">  XE "comparative lujvo: against former state"   There are som=
e comparative concepts which are in which the =93se zmadu=94 is difficult t=
o specify. Typically, these involve comparisons implicitly made with a form=
er state of affairs, where stating a z2 place explicitly would be problemat=
ic.
+<p>
+In such cases, it is best not to use =93zmadu=94 and leave the comparison =
hanging, but to use instead the gismu =93zenba=94, meaning =93increase=94 (=
and =93jdika=94, meaning =93decrease=94, in place of =93mleca=94). The gism=
u =93zenba=94 was included in the language precisely in order to capture th=
ose notions of increase which =93zmadu=94 can't quite cope with; in additio=
n, we don't have to waste a place in lujvo or tanru on something that we'd =
never fill in with a value anyway. So we can translate =93I'm stronger now=
=94 not as
+<p>
+<pre><a name=3De15d7>15.7)	mi ca tsamau
+	I now am-stronger.
+</pre>which implies that I'm currently stronger than somebody else (the el=
ided occupant of the second or z2 place), but as
+<p>
+<pre><a name=3De15d8>15.8)	mi ca tsaze'a
+	I increase in strength.
+</pre><p>Finally, lujvo with a tertau of =93traji=94 are used to build sup=
erlatives. The place structure of =93traji=94 is
+<p>
+<pre><a name=3De15d9>15.9)	t1 is superlative in property t2, being the t3 =
extremum (largest by default)=20
+of set t4
+</pre><p>Consider the gismu =93xamgu=94, whose place structure is:
+<p>
+<pre><a name=3De15d10>15.10)	xa1 is good for xa2 by standard xa3
+</pre><ex "better">  XE "better: example=93  The comparative form is =93xa=
gmau=94, corresponding to English =93better=94, with a place structure (by =
the rules given above) of
+<p>
+<pre><a name=3De15d11>15.11)	z1 is better than z2 for xa2 by standard xa3 =
in amount z4
+</pre><cx "lujvo place structure, superlatives">  XE "lujvo place structur=
e: superlatives"   <cx "lujvo place order, superlatives">  XE "lujvo place =
order: superlatives"   We would expect the place structure of =93xagrai=94,=
 the superlative form, to somehow mirror that, given that comparatives and =
superlatives are comparable concepts, resulting in:
+<p>
+<pre><a name=3De15d12>15.12)	xa1=3Dt1 is the best of the set t4 for xa2 by=
 standard xa3.
+</pre><p>The t2 place in =93traji=94, normally filled by a property abstra=
ction, is replaced by the seltau places, and the t3 place specifying the ex=
tremum of =93traji=94 (whether the most or the least, that is) is presumed =
by default to be =93the most=94.
+<p>
+<cx "lujvo place order, superlatives as exceptions">  XE "lujvo place orde=
r: superlatives as exceptions"   But the set against which the t1 place of =
=93traji=94 is compared is not the t2 place (which would make the place str=
ucture of =93traji=94 fully parallel to that of =93zmadu=94), but rather th=
e t4 place. Nevertheless, by a special exception to the rules of place orde=
ring, the t4 place of =93traji=94-based lujvo becomes the second place of t=
he lujvo. Some examples:
+<p>
+  XE "Judy: example"  <pre><a name=3De15d12>15.13)	la djudis. cu citrai lo=
'i lobypli
+	Judy is the youngest of all Lojbanists.
+  XE "Einstein: example"  <pre><a name=3De15d13>15.14)	la ainctain. cu bal=
rai lo'i skegunka
+	Einstein was the greatest of all scientists.
+</pre><a name=3Ds16><h3>Notes on gismu place structures</h3>
+<p>
+<cx "gismu place structures">  XE =93gismu: place structures"   <cx "place=
 structure, gismu">  XE "place structure: gismu"   Unlike the place structu=
res of lujvo, the place structures of gismu were assigned in a far less sys=
tematic way through a detailed case-by-case analysis and repeated reviews w=
ith associated changes. (The gismu list is now baselined, so no further cha=
nges are contemplated.) Nevertheless, certain regularities were imposed bot=
h in the choice of places and in the ordering of places which may be helpfu=
l to the learner and the lujvo-maker, and which are therefore discussed her=
e.
+<p>
+<cx "gismu place structures, rationale">  XE =93gismu: place structures, r=
ationale"   The choice of gismu places results from the varying outcome of =
four different pressures: brevity, convenience, metaphysical necessity, and=
 regularity. (These are also to some extent the underlying factors in the l=
ujvo place structures generated by the methods of this chapter.) The implic=
ations of each are roughly as follows:
+<p>
+<dl compact><dt><dd>Brevity tends to remove places: the fewer places a gis=
mu has, the easier it is to learn, and the less specific it is. As mentione=
d in <a href=3D#s4>Section 4</a>, a brivla with fewer place structures is l=
ess specific, and generality is a virtue in gismu, because they must thorou=
ghly blanket all of semantic space.
+<p><dt><dd>Convenience tends to increase the number of places: if a concep=
t can be expressed as a place of some existing gismu, there is no need to m=
ake another gismu, a lujvo or a fu'ivla for it.
+<p><dt><dd>Metaphysical necessity can either increase or decrease places: =
it is a pressure tending to provide the =93right number=94 of places. If so=
mething is part of the essential nature of a concept, then a place must be =
made for it; on the other hand, if instances of the concept need not have s=
ome property, then this pressure will tend to remove the place.
+<p><dt><dd>Regularity is a pressure which can also either increase or decr=
ease places. If a gismu has a given place, then gismu which are semanticall=
y related to it are likely to have the place also.
+</dl><p>Here are some examples of gismu place structures, with a discussio=
n of the pressures operating on them:
+<p>
+<pre><a name=3De16d1>16.1)	=93xekri=94:  xe1 is black
+</pre><cx "color standards">  XE "color standards"   Brevity was the most =
important goal here, reinforced by one interpretation of metaphysical neces=
sity. There is no mention of color standards here, as many people have poin=
ted out; like all color gismu, =93xekri=94 is explicitly subjective. Object=
ive color standards can be brought in by an appropriate BAI tag such as =93=
ci'u=94 (=93in system=94; see <a href=3Dchap9.html>Chapter 9</a>) or by mak=
ing a lujvo.
+<p>
+<pre><a name=3De16d2>16.2)	=93jbena=94:  j1 is born to j2 at time j3 and l=
ocation j4
+</pre><p>The gismu =93jbena=94 contains places for time and location, whic=
h few other gismu have: normally, the time and place at which something is =
done is supplied by a tense tag (see <a href=3Dchap10.html>Chapter 10</a>).=
 However, providing these places makes =93le te jbena=94 a simple term for =
=93birthday=94 and =93le ve jbena=94 for =93birthplace=94, so these places =
were provided despite their lack of metaphysical necessity.
+<p>
+<pre><a name=3De16d3>16.3)	=93rinka=94:  event r1 is the cause of event r2
+</pre><cx "melting">  XE "melting"   The place structure of =93rinka=94 do=
es not have a place for the agent, the one who causes, as a result of the p=
ressure toward metaphysical necessity. A cause-effect relationship does not=
 have to include an agent: an event (such as snow melting in the mountains)=
 may cause another event (such as the flooding of the Nile) without any hum=
an intervention or even knowledge.
+<p>
+<cx "lujvo, as suppliers of agent place">  XE "lujvo: as suppliers of agen=
t place"   Indeed, there is a general tendency to omit agent places from mo=
st gismu except for a few such as =93gasnu=94 and =93zukte=94 which are the=
n used as tertau in order to restore the agent place when needed: see <a hr=
ef=3D#s13>Section 13</a>.
+<p>
+<pre><a name=3De16d4>16.4)	=93cinfo=94:  c1 is a lion of species/breed c2
+</pre><cx "general terms">  XE "general terms"   <cx "diversified species"=
>  XE "diversified species"   The c2 place of =93cinfo=94 is provided as a =
result of the pressure toward regularity. All animal and plant gismu have s=
uch an x2 place; although there is in fact only one species of lion, and br=
eeds of lion, though they exist, aren't all that important in talking about=
 lions. The species/breed place must exist for such diversified species as =
dogs, and for general terms like =93cinki=94 (insect), and are provided for=
 all other animals and plants as a matter of regularity.
+<p>
+<cx "gismu place order, rationale">  XE =93gismu: place order, rationale" =
  Less can be said about gismu place structure ordering, but some regularit=
ies are apparent. The places tend to appear in decreasing order of psycholo=
gical saliency or importance. There is an implication within the place stru=
cture of =93klama=94, for example, that =93lo klama=94 (the one going) will=
 be talked about more often, and is thus more important, than =93lo se klam=
a=94 (the destination), which is in turn more important than =93lo xe klama=
=94 (the means of transport).
+<p>
+Some specific tendencies (not really rules) can also be observed. For exam=
ple, when there is an agent place, it tends to be the first place. Similarl=
y, when a destination and an origin point are mentioned, the destination is=
 always placed just before the origin point. Places such as =93under condit=
ions=94 and =93by standard=94, which often go unfilled, are moved to near t=
he end of the place structure.
+
+<p>
+</body></html>
+
+<img src=3Dchap13.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+<h2>Chapter 13
+<br>
+Oooh! Arrgh! Ugh! Yecch! Attitudinal and Emotional Indicators</h2>
+<h6>$Revision: 4.1 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>What are attitudinal indicators?</h3>
+<cx "attitude">  XE "attitude"   <cx "punctuation">  XE "punctuation"   <c=
x "tone of voice">  XE "tone of voice"   <cx "feeling">  XE "feeling"   <cx=
 "emotion">  XE "emotion"   This chapter explains the various words that Lo=
jban provides for expressing attitude and related notions. In natural langu=
ages, attitudes are usually expressed by the tone of voice when speaking, a=
nd (very imperfectly) by punctuation when writing. For example, the bare wo=
rds
+<p>
+<pre><a name=3De1d1>1.1)	John is coming.
+</pre>can be made, through tone of voice, to express the speaker's feeling=
 of happiness, pity, hope, surprise, or disbelief. These fine points of ton=
e cannot be expressed in writing. Attitudes are also expressed with various=
 sounds which show up in print as oddly spelled words, such as the =93Oooh!=
=94, =93Arrgh!=94, =93Ugh!=94, and =93Yecch!=94 in the title. These are par=
t of the English language; people born to other languages use a different s=
et; yet you won't find any of these words in a dictionary.
+<p>
+<cx "attitudinal indicators">  XE "attitudinal indicators"   In Lojban, ev=
erything that can be spoken can also be written. Therefore, these tones of =
voice must be represented by explicit words known as =93attitudinal indicat=
ors=94, or just =93attitudinals=94. This rule seems awkward and clunky to E=
nglish-speakers at first, but is an essential part of the Lojbanic way of d=
oing things.
+<p>
+<cx "attitidinals, prevailing attitude">  XE "attitudinals: prevailing att=
itude"   <cx "attitidinals, placement for prevailing attitude">  XE "attitu=
dinals: placement for prevailing attitude"   The simplest way to use attitu=
dinal indicators is to place them at the beginning of a text. In that case,=
 they express the speaker's prevailing attitude. Here are some examples, co=
rrelated with the attitudes mentioned following <a href=3D#e1d1>Example 1.1=
</a>:
+<p>
+<ex "John is coming">  XE "John is coming: example=93  <lx ".ui">  XE "ui"=
   <pre><a name=3De1d2>1.2)	.ui la djan klama
+	[Whee!] John is coming!
+</pre><lx ".uu">  XE "uu"   <pre><a name=3De1d3>1.3)	.uu la djan klama
+	[Alas!] John is coming.
+</pre><lx ".a'o">  XE "a'o"   <pre><a name=3De1d4>1.4)	.a'o la djan klama
+	[Hopefully] John is coming.
+</pre><lx ".ue">  XE "ue"   <pre><a name=3De1d5>1.5)	.ue la djan klama
+	[Wow!] John is coming!
+</pre><lx ".ianai">  XE "ianai"   <pre><a name=3De1d6>1.6)	.ianai la djan =
klama
+	[Nonsense!] John is coming.
+</pre><cx "attitudinals, word-form for primary">  XE "attitudinals: word-f=
orm for primary"   <lx "UI">  XE "UI selma=92o"   The primary Lojban attitu=
dinals are all the cmavo of the form VV or V'V: one of the few cases where =
cmavo have been classified solely by their form. There are 39 of these cmav=
o: all 25 possible vowel pairs of the form V'V, the four standard diphthong=
s (=93.ai=94, =93.au=94, =93.ei=94, and =93.oi=94), and the ten more diphth=
ongs that are permitted only in these attitudinal indicators and in names a=
nd borrowings (=93.ia=94, =93.ie=94, =93.ii=94, =93.io=94, =93.iu=94, =93.u=
a=94, =93.ue=94, =93.ui=94, =93.uo=94, and =93.uu=94). Note that each of th=
ese cmavo has a period before it, marking the pause that is mandatory befor=
e every word beginning with a vowel. Attitudinals, like most of the other k=
inds of indicators described in this chapter, belong to selma'o UI.
+<p>
+<cx "attitudinals, compound">  XE "attitudinals: compound"   Attitudinals =
can also be compound cmavo, of the types explained in Sections 4-8; <a href=
=3D#e1d6>Example 1.6 </a>illustrates one such possibility, the compound att=
itudinal =93.ianai=94. In attitudinals, =93-nai=94 indicates polar negation=
: the opposite of the simple attitudinal without the =93-nai=94. Thus, as y=
ou might suppose, =93.ia=94 expresses belief, since =93.ianai=94 expresses =
disbelief.
+<p>
+<cx "indicators">  XE "indicators"   <cx "indicators, types of">  XE "indi=
cators: types of"   In addition to the attitudinals, there are other classe=
s of indicators: intensity markers, emotion categories, attitudinal modifie=
rs, observationals, and discursives. All of them are grammatically equivale=
nt, which is why they are treated together in this chapter.
+<p>
+<cx "indicators, placement of">  XE "indicators: placement of"   Every ind=
icator behaves in more or less the same way with respect to the grammar of =
the rest of the language. In general, one or more indicators can be inserte=
d at the beginning of an utterance or after any word. Indicators at the beg=
inning apply to the whole utterance; otherwise, they apply to the word that=
 they follow. More details can be found in <a href=3D#s9>Section 9</a>.
+<p>
+<cx "indicator tables, format convention">  XE "indicator tables: format c=
onvention"   Throughout this chapter, tables of indicators will be written =
in four columns. The first column is the cmavo itself. The second column is=
 a corresponding English word, not necessarily a literal translation. The f=
ourth column represents the opposite of the second column, and shows the ap=
proximate meaning of the attitudinal when suffixed with =93-nai=94. The thi=
rd column, which is sometimes omitted, indicates a neutral point between th=
e second and fourth columns, and shows the approximate meaning of the attit=
udinal when it is suffixed with =93-cu'i=94. The cmavo =93cu'i=94 belongs t=
o selma'o CAI, and is explained more fully in <a href=3D#s4>Section 4</a>.
+<p>
+<cx "feelings, expression of contrasted with talking about">  XE "feelings=
: expression of contrasted with talking about"   <cx "claims, contrasted wi=
th expression of feelings">  XE "claims: contrasted with expression of feel=
ings"   One flaw that the English glosses are particularly subject to is th=
at in English it is often difficult to distinguish between expressing your =
feelings and talking about them, particularly with the limited resource of =
the written word. So the gloss for =93.ui=94 should not really be =93happin=
ess=94 but some sound or tone that expresses happiness. However, there aren=
't nearly enough of those that have unambiguous or obvious meanings in Engl=
ish to go around for all the many, many different emotions Lojban speakers =
can readily express.
+<p>
+<cx "indicators, derived from gismu">  XE "indicators: derived from gismu"=
   <cx "indicators derived from gismu, notation convention">  XE "indicator=
s derived from gismu: notation convention"   Many indicators of CV'V form a=
re loosely derived from specific gismu. The gismu should be thought of as a=
 memory hook, not an equivalent of the cmavo. Such gismu are shown in this =
chapter between square brackets, thus: [gismu].
+<p>
+<a name=3Ds2><h3>Pure emotion indicators</h3>
+<p>
+<cx "attitudinals and claims">  XE "attitudinals and claims"   <cx "attitu=
dinals and truth value">  XE "attitudinals and truth value"   Attitudinals =
make no claim: they are expressions of attitude, not of facts or alleged fa=
cts. As a result, attitudinals themselves have no truth value, nor do they =
directly affect the truth value of a bridi that they modify. However, since=
 emotional attitudes are carried in your mind, they reflect reactions to th=
at version of the world that the mind is thinking about; this is seldom ide=
ntical with the real world. At times, we are thinking about our idealized v=
ersion of the real world; at other times we are thinking about a potential =
world that might or might not ever exist.
+<p>
+<cx "attitudinals, pure emotion">  XE "attitudinals: pure emotion"   <cx "=
attitudinals, u- series">  XE "attitudinals: u- series"   <cx "attitudinals=
, o- series">  XE "attitudinals: o- series"   <cx "attitudinals, i- series"=
>  XE "attitudinals: i- series"   Therefore, there are two groups of attitu=
dinals in Lojban. The =93pure emotion indicators=94 express the way the spe=
aker is feeling, without direct reference to what else is said. These indic=
ators comprise the attitudinals which begin with =93u=94 or =93o=94 and man=
y of those beginning with =93i=94.
+<p>
+<cx "attitudinals for emotional reaction">  XE "attitudinals for emotional=
 reaction"   The cmavo beginning with =93u=94 are simple emotions, which re=
present the speaker's reaction to the world as it is, or as it is perceived=
 to be.
+<p>
+<pre> 	.ua	discovery		confusion
+	.u'a	gain		loss
+	.ue	surprise	no surprise	expectation
+	.u'e	wonder		commonplace
+	.ui	happiness		unhappiness
+	.u'i	amusement		weariness
+	.uo	completion		incompleteness
+	.u'o	courage	 timidity	cowardice
+	.uu	pity		cruelty
+	.u'u	repentance	lack of regret	innocence
+</pre><p>Here are some typical uses of the =93u=94 attitudinals:
+<p>
+<ex "Eureka!">  XE "Eureka!: example=93   <pre><a name=3De2d1>2.1)	.ua mi =
facki fi le mi mapku
+	[Eureka!] I found my hat!  [emphasizes the discovery of the hat]
+<a name=3De2d2>2.2)	.u'a mi facki fi le mi mapku
+	[Gain!] I found my hat!  [emphasizes the obtaining of the hat]
+</pre><ex "Yay!">  XE "Yay!: example=93  <ex "Hooray!">  XE "Hooray!: exam=
ple=93  <pre><a name=3De2d3>2.3)	.ui mi facki fi le mi mapku
+	[Yay!] I found my hat!  [emphasizes the feeling of happiness]
+<a name=3De2d4>2.4)	.uo mi facki fi le mi mapku
+	[At last!] I found my hat!  [emphasizes that the finding is complete]
+</pre><ex "sympathy">  XE "sympathy: example=93  <pre><a name=3De2d5>2.5)	=
.uu do cortu
+	[Pity!] You feel-pain.  [expresses speaker's sympathy]
+<a name=3De2d6>2.6)	.u'u do cortu
+	[Repentance!] You feel-pain.  [expresses that speaker feels guilty]
+</pre><cx "attitudinal phrase">  XE "attitudinal phrase"   In <a href=3D#e=
2d4>Example 2.4</a>, note that the attitudinal =93.uo=94 is translated by a=
n English non-attitudinal phrase: =93At last!=94 It is common for the Engli=
sh equivalents of Lojban attitudinals to be short phrases of this sort, wit=
h more or less normal grammar, but actually expressions of emotion.
+<p>
+<lx "uu">  XE "uu"   <lx "u'u">  XE "u'u"   <cx "uu, contrasted with u'u">=
  XE "uu: contrasted with u'u"   <cx "u'u, contrasted with uu">  XE "u'u: c=
ontrasted with uu"   In particular, both =93.uu=94 and =93.u'u=94 can be tr=
anslated into English as =93I'm sorry=94; the difference between these two =
attitudes frequently causes confusion among English-speakers who use this p=
hrase, leading to responses like =93Why are you sorry? It's not your fault!=
=94
+<p>
+<cx "attitudinals and irony">  XE "attitudinals and irony"   It is importa=
nt to realize that =93.uu=94, and indeed all attitudinals, are meant to be =
used sincerely, not ironically. In English, the exclamation =93Pity!=94 is =
just as likely to be ironically intended, but this usage does not extend to=
 Lojban. Lying with attitudinals is (normally) as inappropriate to Lojban d=
iscourse as any other kind of lying: perhaps worse, because misunderstood e=
motions can cause even greater problems than misunderstood statements.
+<p>
+<lx "nai">  XE "nai"   <lx "cu'i">  XE "cu'i"   <cx "attitudinals, effect =
of nai">  XE "attitudinals: effect of nai"   <cx "attitudinals, effect of c=
u'i">  XE "attitudinals: effect of cu'i"   The following examples display t=
he effects of =93nai=94 and =93cu'i=94 when suffixed to an attitudinal:
+
+<p>
+<pre><a name=3De2d7>2.7)	.ue la djan. klama
+	[Surprise!] John comes.
+<a name=3De2d8>2.8)	.uecu'i la djan. klama
+	[Ho hum.] John comes.
+<a name=3De2d9>2.9)	.uenai la djan. klama
+	[Expected!] John comes.
+</pre><p>In <a href=3D#e2d9>Example 2.9</a>, John's coming has been antici=
pated by the speaker. In <a href=3D#e2d7>Example 2.7 </a>and <a href=3D#e2d=
8>Example 2.8</a>, no such anticipation has been made, but in <a href=3D#e2=
d7>Example 2.7 </a>the lack-of-anticipation goes no further =97 in <a href=
=3D#e2d8>Example 2.8</a>, it amounts to actual surprise.
+<p>
+<cx "attitudinals, complex emotion words">  XE "attitudinals: complex emot=
ion words"   <cx "attitudinals, ambivalent emotion words">  XE "attitudinal=
s: ambivalent emotion words"   <cx "attitudinals, difficult emotion words">=
  XE "attitudinals: difficult emotion words"   It is not possible to firmly=
 distinguish the pure emotion words beginning with =93o=94 or =93i=94 from =
those beginning with =93u=94, but in general they represent more complex, m=
ore ambivalent, or more difficult emotions.
+<p>
+<pre> 	.o'a	pride	modesty	shame
+	.o'e	closeness	detachment	distance
+	.oi	complaint/pain	doing OK	pleasure
+	.o'i	caution	boldness	rashness
+	.o'o	patience	mere tolerance	anger
+	.o'u	relaxation	composure	stress
+</pre><p>Here are some examples:
+<p>
+<ex ".oi">  XE "oi: example=93  <pre><a name=3De2d10>2.10)	.oi la djan. kl=
ama
+	[Complaint!] John is coming.
+</pre><p>Here the speaker is distressed or discomfited over John's coming.=
 The word =93.oi=94 is derived from the Yiddish word =93oy=94 of similar me=
aning. It is the only cmavo with a Yiddish origin.
+<p>
+<ex ".o'onai">  XE "o'onai: example=93  <pre><a name=3De2d11>2.11)	.o'onai=
 la djan. klama
+	[Anger!] John is coming!
+</pre><p>Here the speaker feels anger over John's coming.
+<p>
+<ex ".o'i">  XE "o'i: example=93  <pre><a name=3De2d12>2.12)	.o'i la djan.=
 klama
+	[Beware!] John is coming.
+</pre><p>Here there is a sense of danger in John's arrival.
+<p>
+<ex ".o'ecu'i">  XE "o'ecu'i: example=93  <pre><a name=3De2d13>2.13)	.o'ec=
u'i la djan. klama
+	[Detachment!] John is coming.
+</pre><ex ".o'u">  XE "o'u: example=93  <pre><a name=3De2d14>2.14)	.o'u la=
 djan. klama
+	[Phew!] John is coming.
+</pre><p>In <a href=3D#e2d13>Example 2.13 </a>and <a href=3D#e2d14>Example=
 2.14</a>, John's arrival is no problem: in the former example, the speaker=
 feels emotional distance from the situation; in the latter example, John's=
 coming is actually a relief of some kind.
+<p>
+The pure emotion indicators beginning with =93i=94 are those which could n=
ot be fitted into the =93u=94 or =93o=94 groups because there was a lack of=
 room, so they are a mixed lot. =93.ia=94, =93.i'a=94, =93.ie=94, and =93.i=
'e=94 do not appear here, as they belong in <a href=3D#s3>Section 3 </a>ins=
tead.
+
+<p>
+<pre> 	.ii	fear	nervousness	security
+	.i'i	togetherness		privacy
+	.io	respect		disrespect
+	.i'o	appreciation		envy
+	.iu	love	no love lost	hatred
+	.i'u	familiarity		mystery
+</pre><p>Here are some examples:
+<lx ".ii">  XE "ii"   <ex "Eek!">  XE "Eek!: example=93  <pre><a name=3De2=
d15>2.15)	.ii smacu
+	[Fear!] [Observative:] a-mouse
+	Eek! A mouse!
+</pre><lx ".iu">  XE "iu"   <pre><a name=3De2d16>2.16)	la djan. .iu klama
+	John [love!] is coming.
+</pre><lx ".ionai">  XE "ionai"   <pre><a name=3De2d17>2.17)	la djan. .ion=
ai klama
+	John [disrespect!] is coming.
+</pre><a href=3D#e2d15>Example 2.15 </a>shows an attitude-colored observat=
ive; the attitudinal modifies the situation described by the observative, n=
amely the mouse that is causing the emotion. Lojban-speaking toddlers, if t=
here ever are any, will probably use sentences like <a href=3D#e2d15>Exampl=
e 2.15 </a>a lot.
+<p>
+<cx "attitudinals, comparison of meaning based on position">  XE "attitudi=
nals: comparison of meaning based on position"   <a href=3D#e2d16>Example 2=
.16 </a>and <a href=3D#e2d17>Example 2.17 </a>use attitudinals that follow =
=93la djan.=94 rather than being at the beginning of the sentence. This for=
m means that the attitude is attached to John rather than the event of his =
coming; the speaker loves or disrespects John specifically. Compare:
+<p>
+<pre><a name=3De2d18>2.18)	la djan. klama .iu
+	John is-coming [love!]
+</pre>where it is specifically the coming of John that inspires the feelin=
g.
+<p>
+<a href=3D#e2d17>Example 2.17 </a>is a compact way of swearing at John: yo=
u could translate it as =93That good-for-nothing John is coming.=94
+<p>
+<a name=3Ds3><h3>Propositional attitude indicators</h3>
+<p>
+<cx "attitudinals, propositional indicators">  XE "attitudinals: propositi=
onal indicators"   <cx "propositional, of attitudinals">  XE "propositional=
: of attitudinals"   <cx "internal world">  XE "internal world"   <cx "hypo=
thetical world">  XE "hypothetical world"   As mentioned at the beginning o=
f <a href=3D#s2>Section 2</a>, attitudinals may be divided into two groups,=
 the pure emotion indicators explained in that section, and a contrasting g=
roup which may be called the =93propositional attitude indicators=94. These=
 indicators establish an internal, hypothetical world which the speaker is =
reacting to, distinct from the world as it really is. Thus we may be expres=
sing our attitude towards =93what the world would be like if=A0<dots>=85</d=
ots>=94, or more directly stating our attitude towards making the potential=
 world a reality.
+<p>
+<cx "attitudinals, emotional contrasted with propositional">  XE "attitudi=
nals: emotional contrasted with propositional"   <cx "attitudinals, proposi=
tional contrasted with emotional">  XE "attitudinals: propositional contras=
ted with emotional"   <cx "attitudinals, propositional effect on claim">  X=
E "attitudinals: propositional effect on claim"   In general, the bridi par=
aphrases of pure emotions look (in English) something like =93I'm going to =
the market, and I'm happy about it=94. The emotion is present with the subj=
ect of the primary claim, but is logically independent of it. Propositional=
 attitudes, though, look more like =93I intend to go to the market=94, wher=
e the main claim is logically subordinate to the intention: I am not claimi=
ng that I am actually going to the market, but merely that I intend to.
+<p>
+<cx "attitudinals, a- series">  XE "attitudinals: a- series"   <cx "attitu=
dinals, e- series">  XE "attitudinals: e- series"   <cx "attitudinals, i- s=
eries">  XE "attitudinals: i- series"   There is no sharp distinction betwe=
en attitudinals beginning with =93a=94 and those beginning with =93e=94; ho=
wever, the original intent (not entirely realized due to the need to cram t=
oo many attitudes into too little space) was to make the members of the =93=
a=94-series the purer, more attitudinal realizers of a potential world, whi=
le the members of the =93e=94-series were more ambivalent or complex about =
the speaker's intention with regard to the predication. The relationship be=
tween the =93a=94-series and the =93e=94-series is similar to that between =
the =93u=94-series and the =93o=94-series, respectively. A few propositiona=
l attitude indicators overflowed into the =93i=94-series as well.
+<p>
+<cx "attitudinals, emotional/propositional caveat">  XE "attitudinals: emo=
tional/propositional caveat"   <cx "attitudinals, propositional/emotional c=
aveat">  XE "attitudinals: propositional/emotional caveat"   <cx "attitudin=
als, logical language and">  XE "attitudinals: logical language and"   In f=
act, the entire distinction between pure emotions and propositional attitud=
es is itself a bit shaky: =93.u'u=94 can be seen as a propositional attitud=
e indicator meaning =93I regret that=A0<dots>=85</dots>=94, and =93a'e=94 (=
discussed below) can be seen as a pure emotion meaning =93I'm awake/aware=
=94. The division of the attitudinals into pure-emotion and propositional-a=
ttitude classes in this chapter is mostly by way of explanation; it is not =
intended to permit firm rulings on specific points. Attitudinals are the pa=
rt of Lojban most distant from the =93logical language=94 aspect.
+<p>
+Here is the list of propositional attitude indicators grouped by initial l=
etter, starting with those beginning with =93a=94:
+<p>
+<pre> 	.a'a	attentive	inattentive	avoiding
+	.a'e	alertness		exhaustion
+	.ai	intent	indecision	refusal
+	.a'i	effort	no real effort	repose
+	.a'o	hope		despair
+	.au	desire	indifference	reluctance
+	.a'u	interest	no interest	repulsion
+</pre><p>Some examples (of a parental kind):
+<p>
+<lx ".a'a">  XE "a'a"   <pre><a name=3De3d1>3.1)	.a'a do zgana le veltivni
+	[attentive] you observe the television-receiver.
+	I'm noticing that you are watching the TV.
+</pre><lx ".a'enai">  XE "a'enai"   <pre><a name=3De3d2>3.2)	.a'enai do ra=
nji bacru
+	[exhaustion] you continuously utter.
+	I'm worn out by your continuous talking.
+</pre><lx ".ai">  XE "ai"   <pre><a name=3De3d3>3.3)	.ai mi benji do le ck=
ana
+	[intent] I transfer you to-the bed.
+	I'm putting you to bed.
+</pre><lx ".a'i">  XE "a'i"   <pre><a name=3De3d4>3.4)	.a'i mi ba gasnu le=
 nu do cikna binxo
+	[effort] I [future] am-the-actor-in the event-of you awake-ly become.
+	It'll be hard for me to wake you up.
+</pre><lx ".a'o">  XE "a'o"   <pre><a name=3De3d5>3.5)	.a'o mi kanryze'a c=
a le bavlamdei
+	[hope] I am-health-increased at-time the future-adjacent-day.
+	I hope I feel better tomorrow!
+</pre><lx ".au">  XE "au"   <pre><a name=3De3d6>3.6)	.au mi sipna
+	[desire] I sleep.
+	I want to sleep.
+</pre><lx ".a'ucu'i">  XE "a'ucu'i"   <pre><a name=3De3d7>3.7)	a'ucu'i do =
pante
+	[no interest] you complain
+	I have no interest in your complaints.
+</pre>(In a real-life situation, Examples 3.1-3.7 would also be decorated =
by various pure emotion indicators, certainly including =93.oicai=94, but p=
robably also =93.iucai=94.)
+<p>
+<cx "attitudinals, contrasted with bridi">  XE "attitudinals: contrasted w=
ith bridi"   <cx "attitudinals, rationale for">  XE "attitudinals: rational=
e for"   Splitting off the attitude into an indicator allows the regular br=
idi grammar to do what it does best: express the relationships between conc=
epts that are intended, desired, hoped for, or whatever. Rephrasing these e=
xamples to express the attitude as the main selbri would make for unaccepta=
bly heavyweight grammar.
+<p>
+Here are the propositional attitude indicators beginning with =93e=94, whi=
ch stand roughly in the relation to those beginning with =93a=94 as the pur=
e-emotion indicators beginning with =93o=94 do to those beginning with =93u=
=94 =97 they are more complex or difficult:
+<p>
+<pre> 	.e'a	permission		prohibition
+	.e'e	competence		incompetence
+	.ei	obligation		freedom
+	.e'i	constraint	independence	resistance to constraint
+	.e'o	request		negative request
+	.e'u	suggestion	no suggestion	warning
+</pre><ex "after sleep">  XE "after sleep: example=93  More examples (afte=
r a good night's sleep):
+<p>
+<lx ".e'a">  XE "e'a"   <pre><a name=3De3d8>3.8)	.e'a do sazri le karce
+	[permission] You drive the car.
+	Sure, you can drive the car.
+</pre><lx ".e'e">  XE "e'e"   <pre><a name=3De3d9>3.9)	e'e mi lifri tu'a d=
o
+	[competence] I experience something-related-to you
+	I feel up to dealing with you.
+</pre><lx ".ei">  XE "ei"   <pre><a name=3De3d10>3.10)	.ei mi tisna le kar=
ce ctilyvau
+	[obligation] I fill the car-type-of petroleum-container.
+	I should fill the car's gas tank.
+</pre><lx ".e'o">  XE "e'o"   <pre><a name=3De3d11>3.11)	.e'o ko ko kurji
+	[request] You-imperative of-you-imperative take-care.
+	Please take care of yourself!
+</pre><lx ".e'u">  XE "e'u"   <pre><a name=3De3d12>3.12)	.e'u do klama le =
panka
+	[suggestion] You go to-the park.
+	I suggest going to the park.
+</pre><p>Finally, the propositional attitude indicators beginning with =93=
i=94, which are the overflow from the other sets:
+<p>
+<pre>	.ia	belief	skepticism	disbelief
+	.i'a	acceptance		blame
+	.ie	agreement		disagreement
+	.i'e	approval	non-approval	disapproval
+</pre><p>Still more examples (much, much later):
+<p>
+<lx ".ianai">  XE "ianai"   <pre><a name=3De3d13>3.13)	.ianai do pu pensi =
le nu tcica mi
+	[disbelief] You [past] think the event-of deceiving me.
+	I can't believe you thought you could fool me.
+
+</pre><lx ".i'anai">  XE "i'anai"   <pre><a name=3De3d14>3.14)	do .i'anai =
na xruti do le zdani
+	You [blame] did-not return you to-the house
+	I blame you for not coming home.
+</pre><lx ".ie">  XE "ie"   <pre><a name=3De3d15>3.15)	.ie mi na cusku lu'=
e le tcika
+		be le nu xruti
+	[agreement] I did-not express a-symbol-for the time-of-day=20
+		of the event-of (you return)
+	It's true I didn't tell you when to come back.
+</pre><lx ".i'enai">  XE "i'enai"   <lx ".i'e">  XE "i'e"   <pre><a name=
=3De3d16>3.16)	.i'enai do .i'e zukte
+	[disapproval] you [approval] act
+	I don't approve of what you did, but I approve of you.
+</pre><a href=3D#e3d16>Example 3.16 </a>illustrates the use of a propositi=
onal attitude indicator, =93i'e=94, in both the usual sense (at the beginni=
ng of the bridi) and as a pure emotion (attached to =93do=94). The event ex=
pressed by the main bridi is disapproved of by the speaker, but the referen=
t of the sumti in the x1 place (namely the listener) is approved of.
+<p>
+<cx "attitudinal, signaling as non-propositional">  XE "attitudinal: signa=
ling as non-propositional"   To indicate that an attitudinal discussed in t=
his section is not meant to indicate a propositional attitude, the simplest=
 expedient is to split the attitudinal off into a separate sentence. Thus, =
a version of <a href=3D#e3d8>Example 3.8 </a>which actually claimed that th=
e listener was or would be driving the car might be:
+<p>
+<pre><a name=3De3d17>3.17)	do sazri le karce .i e'a
+	You drive the car. [Permission].
+	You're driving (or will drive) the car, and that's fine.
+</pre><a name=3Ds4><h3>Attitudes as scales</h3>
+<p>
+<cx "attitudinals, scale of">  XE "attitudinals: scale of"   <cx "attitudi=
nals, positive">  XE "attitudinals: positive"   <cx "attitudinals, negative=
">  XE "attitudinals: negative"   <cx "attitudinals, neutral">  XE "attitud=
inals: neutral"   In Lojban, all emotions and attitudes are scales. These s=
cales run from some extreme value (which we'll call =93positive=94) to an o=
pposite extreme (which we'll call =93negative=94). In the tables above, we =
have seen three points on the scale: =93positive=94, neutral, and =93negati=
ve=94. The terms =93positive=94 and =93negative=94 are put into quotation m=
arks because they are loaded words when applied to emotions, and the attitu=
dinal system reflects this loading, which is a known cultural bias. Only tw=
o of the =93positive=94 words, namely =93.ii=94 (fear) and =93.oi=94 (pain/=
complaint), represent emotions commonly thought of as less =93virtuous=94 i=
n most cases than their negative counterparts. But these two were felt to b=
e instinctive, distinct, and very powerful emotions that needed to be expre=
ssible in a monosyllable when necessary, while their counterparts are less =
commonly expressed.
+<p>
+<cx "attitudinal scales, rationale for assignment">  XE "attitudinal scale=
s: rationale for assignment"   (Why the overt bias? Because there are a lot=
 of attitudinals and they will be difficult to learn as an entire set. By a=
ligning our scales arbitrarily, we give the monosyllable =93nai=94 a useful=
 meaning and make it easier for a novice to recognize at least the positive=
 or negative alignment of an indicator, if not the specific word. Other cho=
ices considered were =93random=94 orientation, which would have unknown bia=
ses and be difficult to learn, and orientation based on our guesses as to w=
hich scale orientations made the most frequent usages shorter, which would =
be biased in favor of American perceptions of =93usefulness=94. If bias mus=
t exist in our indicator set, it might as well be a known bias that eases l=
earning, and in addition might as well favor a harmonious and positive worl=
d-view.)
+<p>
+<cx "emotional scale">  XE "emotional scale"   <cx "attitudinal scale, sev=
en-position">  XE "attitudinal scale: seven-position"   <lx "cai">  XE "cai=
"   <lx "sai">  XE "sai"   <lx "ru'e">  XE "ru'e"   <lx "cu'i">  XE "cu'i" =
  <lx "nairu'e">  XE "nairu'e"   <lx "naisai">  XE "naisai"   <lx "naicai">=
  XE "naicai"   <lx "CAI">  XE "CAI selma=92o"   In fact, though, each emot=
ional scale has seven positions defined, three =93positive=94 ones (shown b=
elow on the left), three =93negative=94 ones (shown below on the right), an=
d a neutral one indicating that no particular attitude on this scale is fel=
t. The following chart indicates the seven positions of the scale and the a=
ssociated cmavo. All of these cmavo, except =93nai=94, are in selma'o CAI.
+<p>
+<pre>cai	sai	ru'e	cu'i	nairu'e	naisai	naicai
+[carmi]	[tsali]	[ruble]	[cumki]
+</pre><cx "scalar attitude">  XE "scalar attitude"   <cx "attitude, scalar=
">  XE "attitude: scalar"   A scalar attitude is expressed by using the att=
itudinal word, and then following it by the desired scalar intensity. The b=
ias creeps in because the =93negative=94 emotions take the extra syllable =
=93nai=94 to indicate their negative position on the axis, and thus require=
 a bit more effort to express.
+<p>
+<cx "attitudinal scale, usage">  XE "attitudinal scale: usage"   Much of t=
his system is optional. You can express an attitude without a scale indicat=
or, if you don't want to stop and think about how strongly you feel. Indeed=
, for most attitudinals, we've found that either no scalar value is used, o=
r =93cai=94 is used to indicate especially high intensity. Less often, =93r=
u'e=94 is used for a recognizably weak intensity, and =93cu'i=94 is used in=
 response to the attitudinal question =93pei=94 (see <a href=3D#s10>Section=
 10</a>) to indicate that the emotion is not felt.
+<p>
+<cx "attitudinal, example of scale effect">  XE "attitudinal: example of s=
cale effect"   The following shows the variations resulting from intensity =
variation:
+<p>
+<lx ".ei">  XE "ei"   <pre><a name=3De4d1>4.1)	.ei
+	I ought to
+	(a non-specific obligation)
+</pre><lx ".eicai">  XE "eicai"   <pre><a name=3De4d2>4.2)	.eicai
+	I shall/must
+	(an intense obligation or requirement, possibly a formal one)
+</pre><lx ".eisai">  XE "eisai"   <ex "formal requirement">  XE "formal re=
quirement: example=93  <pre><a name=3De4d3>4.3)	.eisai
+	I should
+	(a strong obligation or necessity, possibly an implied but not formal
+requirement)
+</pre><lx ".eiru'e">  XE "eiru'e"   <pre><a name=3De4d4>4.4)	.eiru'e
+	I might
+	(a weak obligation =97 in English often mixed with permission and desire)
+</pre><lx ".eicu'i">  XE "eicu'i"   <pre><a name=3De4d5>4.5)	.eicu'i
+	No matter
+	(no particular obligation)
+</pre><lx ".einai">  XE "einai"   <pre><a name=3De4d6>4.6)	.einai
+	I need not
+	(a non-obligation)
+</pre><cx "attitudinal scale, stand-alone usage">  XE "attitudinal scale: =
stand-alone usage"   You can also utter a scale indicator without a specifi=
c emotion. This is often used in the language: in order to emphasize a poin=
t about which you feel strongly, you mark what you are saying with the scal=
e indicator =93cai=94. You could also indicate that you don't care using =
=93cu'i=94 by itself.
+<p>
+<a name=3Ds5><h3>The space of emotions</h3>
+<p>
+<cx "attitudinal scale, as axis in emotion-space">  XE "attitudinal scale:=
 as axis in emotion-space"   Each of the attitude scales constitutes an axi=
s in a multi-dimensional space. In effect, given our total so far of 39 sca=
les, we have a 39-dimensional space. At any given time, our emotions and at=
titudes are represented by a point in this 39-dimensional space, with the i=
ntensity indicators serving as coordinates along each dimension. A complete=
 attitudinal inventory, should one decide to express it, would consist of r=
eading off each of the scale values for each of the emotions, with the vect=
or sum serving as a distinct single point, which is our attitude.
+<p>
+<cx "compound emotions">  XE "compound emotions"   <cx "emotions, compound=
">  XE "emotions: compound"   <cx "emotion, insights">  XE "emotions: insig=
hts"   <cx "emotion, when expressed">  XE "emotions: when expressed"   Now =
no one is going to ever utter a string of 100-odd attitudinals to express t=
heir emotions. If asked, we normally do not recognize more than one or two =
emotions at a time =97 usually the ones that are strongest or which most re=
cently changed in some significant way. But the scale system provides some =
useful insights into a possible theory of emotion (which might be testable =
using Lojban), and incidentally explains how Lojbanists express compound em=
otions when they do recognize them.
+<p>
+<cx "attitudinal scale, neutral compared with positive + negative">  XE "a=
ttitudinal scale: neutral compared with positive + negative"   The existenc=
e of 39 scales highlights the complexity of emotion. We also aren't bound t=
o the 39. There are modifiers described in <a href=3D#s6>Section 6 </a>that=
 multiply the set of scales by an order of magnitude. You can also have mix=
ed feelings on a scale, which might be expressed by =93cu'i=94, but could a=
lso be expressed by using both the =93positive=94 and =93negative=94 scale =
emotions at once. One expression of =93fortitude=94 might be =93.ii.iinai=
=94 =97 fear coupled with security.
+<p>
+<cx "attitudinals, order of">  XE "attitudinals: order of"   <cx "attitudi=
nals, contrasted with rationalizations of emotion">  XE "attitudinals: cont=
rasted with rationalizations of emotion"   Uttering one or more attitudinal=
s to express an emotion reflects several things. We will tend to utter emot=
ions in their immediate order of importance to us. We feel several emotions=
 at once, and our expression reflects these emotions simultaneously, althou=
gh their order of importance to us is also revealing =97 of our attitude to=
wards our attitude, so to speak. There is little analysis necessary; for th=
ose emotions you feel, you express them; the =93vector sum=94 naturally exp=
resses the result. This is vital to their nature as attitudinals =97 if you=
 had to stop and think about them, or to worry about grammar, they wouldn't=
 be emotions but rationalizations.
+<p>
+<cx "attitudinals, contrasted with bridi">  XE "attitudinals: contrasted w=
ith bridi"   People have proposed that attitudinals be expressed as bridi j=
ust like everything else; but emotions aren't logical or analytical =97 say=
ing =93I'm awed=94 is not the same as saying =93Wow!!!=94. The Lojban syste=
m is intended to give the effects of an analytical system without the thoug=
ht involved. Thus, you can simply feel in Lojban.
+<p>
+<cx "attitudinals, design benefit">  XE "attitudinals: design benefit"   A=
 nice feature of this design is that you can be simple or complex, and the =
system works the same way. The most immediate benefit is in learning. You o=
nly need to learn a couple of the scale words and a couple of attitude word=
s, and you're ready to express your emotions Lojbanically. As you learn mor=
e, you can express your emotions more thoroughly and more precisely, but ev=
en a limited vocabulary offers a broad range of expression.
+<p>
+<a name=3Ds6><h3>Emotional categories</h3>
+<p>
+<cx "emotional categories">  XE "emotional categories"   <cx "attitudinal =
categories">  XE "attitudinal categories"   <cx "attitudinal categories, ra=
tionale">  XE "attitudinal categories: rationale"   The Lojban attitudinal =
system was designed by starting with a long list of English emotion words, =
far too many to fit into the 39 available VV-form cmavo. To keep the number=
 of cmavo limited, the emotion words in the list were grouped together by c=
ommon features: each group was then assigned a separate cmavo. This was lik=
e making tanru in reverse, and the result is a collection of indicators tha=
t can be combined, like tanru, to express very complex emotions. Some examp=
les in a moment.
+<p>
+The most significant =93common feature=94 we identified was that the emoti=
onal words on the list could easily be broken down into six major groups, e=
ach of which was assigned its own cmavo:
+<p>
+<lx "ro'a">  XE "ro'a"   <lx "ro'e">  XE "ro'e"   <lx "ro'i">  XE "ro'i"  =
 <lx "ro'o">  XE "ro'o"   <lx "ro'u">  XE "ro'u"   <lx "re'e">  XE "re'e"  =
 <pre>	ro'a	social	asocial	antisocial
+	ro'e	mental		mindless
+	ro'i	emotional		denying emotion
+	ro'o	physical		denying physical
+	ro'u	sexual		sexual abstinence
+	re'e	spiritual	secular	sacrilegious
+</pre><p>Using these, we were able to assign =93o'u=94 to mark a scale of =
what we might call =93generalized comfort=94. When you are comfortable, rel=
axed, satisfied, you express comfort with =93o'u=94, possibly followed by a=
 scale indicator to indicate how comfortable you are. The six cmavo given a=
bove allow you to turn this scale into six separate ones, should you wish.
+<p>
+<ex "sexual discomfort">  XE "sexual discomfort: example=93  <ex "spiritua=
l discomfort">  XE "spiritual discomfort: example=93  <ex "embarrassment"> =
 XE "embarrassment: example=93  <ex "stress">  XE "stress: example=93  <ex =
"physical distress">  XE "physical distress: example=93  <ex "mental discom=
fort">  XE "mental discomfort: example=93  <cx "attitudinal categories, exa=
mple of effect">  XE "attitudinal categories: example of effect"   For exam=
ple, embarrassment is a social discomfort, expressible as =93.o'unairo'a=94=
. Some emotions that we label =93stress=94 in English are expressed in Lojb=
an with =93.o'unairo'i=94. Physical distress can be expressed with =93.o'un=
airo'o=94, which makes a nice groan if you say it with feeling. Mental disc=
omfort might be what you feel when you don't know the answer to the test qu=
estion, but feel that you should. Most adults can recall some instance wher=
e we felt sexual discomfort, =93o'unairo'u=94. Spiritual discomfort, =93o'u=
naire'e=94, might be felt by a church-goer who has wandered into the wrong =
kind of religious building.
+<p>
+Most of the time when expressing an emotion, you won't categorize it with =
these words. Emotional expressions should be quickly expressible without ha=
ving to think about them. However, we sometimes have mixed emotions within =
this set, as for example emotional discomfort coupled with physical comfort=
 or vice versa.
+<p>
+<lx ".eiro'u">  XE "eiro'u"   Coupling these six words with our 39 attitud=
e scales, each of which has a positive and negative side, already gives you=
 far more emotional expression words than we have emotional labels in Engli=
sh. Thus, you'll never see a Lojban-English emotional dictionary that cover=
s all the Lojban possibilities. Some may be useless, but others convey emot=
ions that probably never had a word for them before, though many have felt =
them (=93.eiro'u=94, for example =97 look it up).
+<p>
+<cx "attitudinals, categories with scale markers">  XE "attitudinals: cate=
gories with scale markers"   <cx "attitudinals, categories with nai">  XE "=
attitudinals: categories with nai"   <cx "attitudinals, stand-alone categor=
ies">  XE "attitudinals: stand-alone categories"   <ex "ro'anai">  XE "ro'a=
nai: example=93  You can use scale markers and =93nai=94 on these six categ=
ory words, and you can also use category words without specifying the emoti=
on. Thus, =93I'm trying to concentrate=94 could be expressed simply as =93r=
o'e=94, and if you are feeling anti-social in some non-specific way, =93ro'=
anai=94 will express it.
+<p>
+<cx "attitudinal categories, mnemonic for">  XE "attitudinal categories: m=
nemonic for"   There is a mnemonic device for the six emotion categories, b=
ased on moving your arms about. In the following table, your hands begin ab=
ove your head and move down your body in sequence.
+<p>
+<pre> 	ro'a	hands above head	social
+	ro'e	hands on head	intellectual
+	ro'i	hands on heart	emotional
+	ro'o	hands on belly	physical
+	ro'u	hands on groin	sexual
+	re'e	hands moving around	spiritual
+</pre><p>The implicit metaphors =93heart=94 for emotional and =93belly=94 =
for physical are not really Lojbanic, but they work fine for English-speake=
rs.
+<p>
+<a name=3Ds7><h3>Attitudinal modifiers</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<ex "self-orientation">  XE "self-orientation: example=93  <pre>ga'i	[galt=
u]	hauteur	equal rank	meekness
+		rank		lack of rank
+le'o		aggressive	passive	defensive
+vu'e	[vrude]	virtue (zabna)		sin (mabla)
+se'i	[sevzi]	self-orientation		other-orientation
+ri'e	[zifre]	release	restraint	control
+fu'i	[frili]	with help	without help	with opposition
+		easily		with difficulty
+be'u		lack	presence	satiation
+		need	satisfaction=09
+se'a	[sevzi]	self-sufficiency		dependency
+		=09
+</pre><cx "attitudinal modifiers">  XE "attitudinal modifiers"   It turned=
 out that, once we had devised the six emotion categories, we also recogniz=
ed some other commonalities among emotions. These tended to fit nicely on s=
cales of their own, but generally tend not to be thought of as separate emo=
tions. Some of these are self-explanatory, some need to be placed in contex=
t. Some of these tend to go well with only a few of the attitudinals, other=
s go with nearly all of them. To really understand these modifiers, try to =
use them in combination with one or two of the attitudinals found in <a hre=
f=3D#s2>Sections 2 </a>and <a href=3D#s3>3</a>, and see what emotional pict=
ures you can build:
+<p>
+<ex "inferior">  XE "inferior: example=93  <ex "deference">  XE "deference=
: example=93  <ex "condescension">  XE "condescension: example=93  <lx "ga'=
i">  XE "ga'i"   <lx "ga'inai">  XE "ga'inai"   The cmavo =93ga'i=94 expres=
ses the scale used to indicate condescension or polite deference; it is not=
 respect in general, which is =93.io=94. Whatever it is attached to is mark=
ed as being below (for =93ga'i=94) or above (for =93ga'inai=94) the speaker=
's rank or social position. Note that it is always the referent, not the sp=
eaker or listener, who is so marked: in order to mark the listener, the lis=
tener must appear in the sentence, as with =93doi ga'inai=94, which can be =
appended to a statement addressed to a social superior.
+<p>
+<pre><a name=3De7d1>7.1)	ko ga'inai nenri klama le mi zdani
+	You-imperative [low-rank!] enter-type-of come-to my house.
+	I would be honored if you would enter my residence.
+</pre><cx "imperatives, attitude">  XE "imperatives: attitude"   Note that=
 imperatives in Lojban need not be imperious! Corresponding examples with =
=93ga'icu'i=94 and =93ga'inai=94:
+<p>
+<pre><a name=3De7d2>7.2)	ko ga'icu'i nenri klama le mi zdani
+	You-imperative [equal-rank!] enter-type-of come-to my house.
+	Come on in to my place.
+<a name=3De7d3>7.3)	ko ga'i nenri klama le mi zdani
+	You-imperative [high-rank!] enter-type-of come-to my house.
+	You! Get inside!
+</pre><p>Since =93ga'i=94 expresses the relative rank of the speaker and t=
he referent, it does not make much sense to attach it to =93mi=94, unless t=
he speaker is using =93mi=94 to refer to a group (as in English =93we=94), =
or a past or future version of himself with a different rank.
+<p>
+It is also possible to attach =93ga'i=94 to a whole bridi, in which case i=
t expresses the speaker's superiority to the event the bridi refers to:
+<p>
+<pre><a name=3De7d4>7.4)	ga'i le xarju pu citka
+	[High-rank!] The pig [past] eats.
+	The pig ate (which is an event beneath my notice).
+</pre><lx "ga'icai">  XE "ga'icai"   When used without being attached to a=
ny bridi, =93ga'i=94 expresses the speaker's superiority to things in gener=
al, which may represent an absolute social rank: =93ga'icai=94 is an approp=
riate opening word for an emperor's address from the throne.
+<p>
+<lx "le'o">  XE "le'o"   The cmavo =93le'o=94 represents the scale of aggr=
essiveness. We seldom overtly recognize that we are feeling aggressive or d=
efensive, but perhaps in counseling sessions, a psychologist might encourag=
e someone to express these feelings on this scale. And football teams could=
 be urged on by their coach using =93ro'ole'o=94. =93le'o=94 is also useful=
 in threats as an alternative to =93o'onai=94, which expresses anger.
+<p>
+<ex "righteous indignation">  XE "righteous indignation: example=93  <lx "=
vu'e">  XE "vu'e"   <ex "virtue">  XE "virtue: example=93  <ex "sinful">  X=
E "sinful: example=93  The cmavo =93vu'e=94 represents ethical virtue or it=
s absence. An excess of almost any emotion is usually somewhat =93sinful=94=
 in the eyes of most ethical systems. On the other hand, we often feel virt=
uous about our feelings =97 what we call righteous indignation might be =93=
o'onaivu'e=94. Note that this is distinct from lack of guilt: =93.u'unai=94=
.
+<p>
+<lx "se'i">  XE "se'i"   The cmavo =93se'i=94 expresses the difference bet=
ween selfishness and generosity, for example (in combination with =93.au=94=
):
+<p>
+<pre><a name=3De7d5>7.5)	ause'i
+	[desire] [self]
+	I want it!
+<a name=3De7d6>7.6)	ause'inai
+	[desire] [other]
+	I want you to have it!
+</pre><p>In both cases, the English =93it=94 is vague, reflecting the abse=
nce of a bridi. <a href=3D#e7d5>Example 7.5 </a>and <a href=3D#e7d6>Example=
 7.6 </a>are pure expressions of attitude. Analogously, =93.uuse'i=94 is se=
lf-pity, whereas =93.uuse'inai=94 is pity for someone else.
+<p>
+<lx "ri'e">  XE "ri'e"   The modifier =93ri'e=94 indicates emotional relea=
se versus emotional control. =93I will not let him know how angry I am=94, =
you say to yourself before entering the room. The Lojban is much shorter:
+<p>
+<pre><a name=3De7d7>7.7)	.o'onai ri'enai
+	[anger] [control]
+</pre><p>On the other hand, =93ri'e=94 can be used by itself to signal an =
emotional outburst.
+<p>
+<lx "fu'i">  XE "fu'i"   The cmavo =93fu'i=94 may express a reason for fee=
ling the way we do, as opposed to a feeling in itself; but it is a reason t=
hat is more emotionally determined than most. For example, it could show th=
e difference between the mental discomfort mentioned in <a href=3D#s6>Secti=
on 6 </a>when it is felt on an easy test, as opposed to on a hard test. Whe=
n someone gives you a back massage, you could use =93.o'ufu'i=94 to show ap=
preciation for the assistance in your comfort.
+<p>
+<lx "be'u">  XE "be'u"   The cmavo =93be'u=94 expresses, roughly speaking,=
 whether the emotion it modifies is in response to something you don't have=
 enough of, something you have enough of, or something you have too much of=
. It is more or less the attitudinal equivalent of the subjective quantifie=
r cmavo =93mo'a=94, =93rau=94, and =93du'e=94 (these belong to selma'o PA, =
and are discussed in <a href=3Dchap18.html>Chapter 18</a>). For example,
+<p>
+<pre><a name=3De7d8>7.8)	.uiro'obe'unai
+	[Yay!] [physical] [Enough!]
+</pre><ex "large meal">  XE "large meal: example=93  might be something yo=
u say after a large meal which you enjoyed.
+<p>
+Like all modifiers, =93be'u=94 can be used alone:
+<p>
+<pre><a name=3De7d9>7.9)	le cukta be'u cu zvati ma
+	The book [Needed!] is at-location [what sumti?]
+	Where's the book? =97 I need it!
+</pre><lx "se'a">  XE "se'a"   Lastly, the modifier =93se'a=94 shows wheth=
er the feeling is associated with self-sufficiency or with dependence on ot=
hers.
+<p>
+<pre><a name=3De7d10>7.10)	.e'ese'a
+	[I can!] [self-sufficient!]
+	I can do it all by myself!
+</pre>is something a Lojban-speaking child might say. On the other hand,
+<p>
+<pre><a name=3De7d11>7.11)	.e'ese'anai
+	[I can!] [dependent]
+	I can do it if you help me.
+</pre>from the same child would indicate a (hopefully temporary) loss of s=
elf- confidence. It is also possible to negate the =93.e'e=94 in <a href=3D=
#e7d10>Example 7.10 </a>and <a href=3D#e7d11>Example 7.11</a>, leading to:
+<p>
+<pre><a name=3De7d12>7.12)	.e'enaise'a
+	[I can't!] [self-sufficient]
+	I can't do it if you insist on =93helping=94 me!
+</pre>and
+<p>
+<pre><a name=3De7d13>7.13)	.e'enaise'anai
+	[I can't!] [dependent]
+	I can't do it by myself!
+</pre><cx "attitudinals, complexity">  XE "attitudinals: complexity"   Som=
e of the emotional expressions may seem too complicated to use. They might =
be for most circumstances. It is likely that most combinations will never g=
et used. But if one person uses one of these expressions, another person ca=
n understand (as unambiguously as the expresser intends) what emotion is be=
ing expressed. Most probably as the system becomes well-known and internali=
zed by Lojban-speakers, particular attitudinal combinations will come to be=
 standard expressions (if not cliches) of emotion.
+<p>
+<a name=3Ds8><h3>Compound indicators</h3>
+<p>
+<cx "indicators, grammar for compounding">  XE "indicators: grammar for co=
mpounding"   <cx "indicators, meaning when compounded">  XE "indicators: me=
aning when compounded"   The grammar of indicators is quite simple; almost =
all facets are optional. You can combine indicators in any order, and they =
are still grammatical. The presumed denotation is additive; thus the whole =
is the sum of the parts regardless of the order expressed, although the fir=
st expressed is presumed most important to the speaker. Every possible stri=
ng of UI cmavo has some meaning.
+<p>
+<cx "unspecified emotion">  XE "unspecified emotion"   <cx "attitudinal in=
dicators, conventions of interpretation">  XE "attitudinal indicators: conv=
entions of interpretation"   <lx "ge'e">  XE "ge'e"   <cx "attitudinal indi=
cator, unspecified">  XE "attitudinal indicator: unspecified"   Within a st=
ring of indicators, there will be conventions of interpretation which amoun=
t to a kind of second-order grammar. Each of the modifier words is presumed=
 to modify an indicator to the left, if there is one. (There is an =93unspe=
cified emotion=94 word, =93ge'e=94, reserved to ensure that if you want to =
express a modifier without a root emotion, it doesn't attach to and modify =
a previous but distinct emotional expression.)
+<p>
+<cx "unstated emotion">  XE "unstated emotion"   <cx "unspecified level of=
 emotion">  XE "unspecified level of emotion"   For example, =93.ieru'e=94 =
expresses a weak positive value on the scale of agreement: the speaker agre=
es (presumably with the listener or with something else just stated), but w=
ith the least possible degree of intensity. But =93.ie ge'eru'e=94 expresse=
s agreement (at an unspecified level), followed by some other unstated emot=
ion which is felt at a weak level. A rough English equivalent of =93.ie ge'=
eru'e=94 might be =93I agree, but=A0<dots>=85</dots>=94 where the =93but=94=
 is left hanging. (Again, attitudes aren't always expressed in English by E=
nglish attitudinals.)
+<p>
+<cx "attitudinal indicators, placement of scale in">  XE "attitudinal indi=
cators: placement of scale in"   A scale variable similarly modifies the pr=
evious emotion word. You put the scale word for a root emotion word before =
a modifier, since the latter can have its own scale word. This merely maxim=
izes the amount of information expressible. For example, =93.oinaicu'i ro'u=
cai=94 expresses a feeling midway between pain (=93.oi=94) and pleasure (=
=93.oinai=94) which is intensely sexual (=93ro'u=94) in nature.
+<p>
+<cx "attitudinal indicators, placement of =93nai=94 in">  XE "attitudinal =
indicators: placement of \=93nai\=94 in"   <cx "attitudinals, placement in =
sentences with =93nai=94">  XE "attitudinals: placement in sentences with \=
=93nai\=94"   The cmavo =93nai=94 is the most tightly bound modifier in the=
 language: it always negates exactly one word =97 the preceding one. Of all=
 the words used in indicator constructs, =93nai=94 is the only one with any=
 meaning outside the indicator system. If you try to put an indicator betwe=
en a non-indicator cmavo and its =93nai=94 negator, the =93nai=94 will end =
up negating the last word of the indicator. The result, though unambiguous,=
 is not what you want. For example,
+<p>
+<pre><a name=3De8d1>8.1)	mi .e .ui nai do
+	I and [Yay!] [Not!] you
+</pre>means =93I and (unfortunately) you=94, whereas
+<p>
+<pre><a name=3De8d2>8.2)	mi .e nai .ui do
+	I and [Not!] [Yay!] you
+</pre>means =93I but (fortunately) not you=94. Attitudinal =93nai=94 expre=
sses a =93scalar negation=94, a concept explained in <a href=3Dchap15.html>=
Chapter 15</a>; since every attitudinal word implies exactly one scale, the=
 effect of =93nai=94 on each should be obvious.
+<p>
+<cx "attitudinals, internal grammar, complete">  XE "attitudinals: interna=
l grammar, complete"   <cx "attitudinals, grammar of internal compounding">=
  XE "attitudinals: grammar of internal compounding"   Thus, the complete i=
nternal grammar of UI is as follows, with each listed part optionally prese=
nt or absent without affecting grammaticality, though it obviously would af=
fect meaning.
+<p>
+<dl compact><dt><dd>attitudinal  =93nai=94  intensity-word  =93nai=94  mod=
ifier  =93nai=94  intensity-word  =93nai=94
+		(possibly repeated)
+</dl>=93ge'e=94, the non-specific emotion word, functions as an attitudina=
l. If multiple attitudes are being expressed at once, then in the 2nd or gr=
eater position, either =93ge'e=94 or a VV word must be used to prevent any =
modifiers from modifying the previous attitudinal.
+
+<p>
+<a name=3Ds9><h3>The uses of indicators</h3>
+<p>
+<cx "attitudinals, grammar of placement in bridi">  XE "attitudinals: gram=
mar of placement in bridi"   <cx "attitudinals, external grammar">  XE "att=
itudinals: external grammar"   <lx "si">  XE "si"   <lx "sa">  XE "sa"   <l=
x "su">  XE "su"   The behavior of indicators in the =93outside grammar=94 =
is nearly as simple as their internal structure. Indicator groupings are id=
entified immediately after the metalinguistic erasers =93si=94, =93sa=94, a=
nd =93su=94 and some, though not all, kinds of quotations. The details of s=
uch interactions are discussed in <a href=3Dchap19.html>Chapter 19</a>.
+<p>
+<lx "zo">  XE "zo"   A group of indicators may appear anywhere that a sing=
le indicator may, except in those few situations (as in =93zo=94 quotation,=
 explained in <a href=3Dchap19.html>Chapter 19</a>) where compound cmavo ma=
y not be used.
+<p>
+<cx "attitudinals, at beginning of text">  XE "attitudinals: at beginning =
of text"   At the beginning of a text, indicators modify everything followi=
ng them indefinitely: such a usage is taken as a raw emotional expression, =
and we normally don't turn off our emotions when we start and stop sentence=
s. In every other place in an utterance, the indicator (or group) attaches =
to the word immediately to its left, and indicates that the attitude is bei=
ng expressed concerning the object or concept to which the word refers.
+<p>
+<cx "attitudinals, affecting whole grammatical structures">  XE "attitudin=
als: affecting whole grammatical structures"   If the word that an indicato=
r (or group) attaches to is itself a cmavo which governs a grammatical stru=
cture, then the indicator construct pertains to the referent of the entire =
structure. There is also a mechanism, discussed in <a href=3Dchap19.html>Ch=
apter 19</a>, for explicitly marking the range of words to which an indicat=
or applies.
+<p>
+<cx "attitudinals, referent uncertainty">  XE "attitudinals: referent unce=
rtainty"   More details about the uses of indicators, and the way they inte=
ract with other specialized cmavo, are given in <a href=3Dchap19.html>Chapt=
er 19</a>. It is worth mentioning that real-world interpretation is not nec=
essarily consistent with the formal scope rules. People generally express e=
motions when they feel them, with only a minimum of grammatical constraint =
on that expression; complexities of emotional expression are seldom logical=
ly analyzable. Lojban attempts to provide a systematic reference that could=
 possibly be ingrained to an instinctive level. However, it should always b=
e assumed that the referent of an indicator has some uncertainty.
+<p>
+<cx "multiple indicators">  XE "multiple indicators"   For example, in cas=
es of multiple indicators expressed together, the combined form has some am=
biguity of interpretation. It is possible to interpret the second indicator=
 as expressing an attitude about the first, or to interpret both as express=
ing attitudes about the common referent. For example, in
+<p>
+<pre><a name=3De9d1>9.1)	mi pu tavla do .o'onai .oi
+	I [past] talk-to you [Grrr!] [Oy!]
+</pre>can be interpreted as expressing complaint about the anger, in which=
 case it means =93Damn, I snapped at you=94; or as expressing both anger an=
d complaint about the listener, in which case it means =93I told you, you p=
est!=94
+<p>
+Similarly, an indicator after the final brivla of a tanru may be taken to =
express an attitude about the particular brivla placed there =97 as the rul=
es have it =97 or about the entire bridi which hinges on that brivla. Remem=
bering that indicators are supposedly direct expressions of emotion, this a=
mbiguity is acceptable.
+<p>
+<cx "attitudinals, benefit in written expression">  XE "attitudinals: bene=
fit in written expression"   Even if the scope rules given for indicators t=
urn out to be impractical or unintuitive for use in conversation, they are =
still useful in written expression. There, where you can go back and put in=
 markers or move words around, the scope rules can be used in lieu of elabo=
rate nuances of body language and intonation to convey the writer's intent.
+
+<p>
+<a name=3Ds10><h3>Attitude questions; empathy; attitude contours</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	pei	attitude question
+	dai	empathy
+	bu'o	start emotion	continue emotion	end emotion
+</pre><p>You can ask someone how they are feeling with a normal bridi sent=
ence, but you will get a normal bridi answer in response, one which may be =
true or false. Since the response to a question about emotions is no more l=
ogical than the emotion itself, this isn't appropriate.
+<p>
+<lx "pei">  XE "pei"   <cx "attitudinal questions">  XE "attitudinal quest=
ions"   The word =93pei=94 is therefore reserved for attitude questions. As=
ked by itself, it captures all of the denotation of English =93How are you?=
=94 coupled with =93How do you feel?=94 (which has a slightly different ran=
ge of usage).
+<p>
+<cx "attitudinal answers, plausibility">  XE "attitudinal answers: plausib=
ility"   When asked in the context of discourse, =93pei=94 acts like other =
Lojban question words =97 it requests the respondent to =93fill in the blan=
k=94, in this case with an appropriate attitudinal describing the responden=
t's feeling about the referent expression. As with other questions, plausib=
ility is polite; if you answer with an irrelevant UI cmavo, such as a discu=
rsive, you are probably making fun of the questioner. (A =93ge'e=94, howeve=
r, is always in order =97 you are not required to answer emotionally. This =
is not the same as =93.i'inai=94, which is privacy as the reverse of conviv=
iality.)
+<p>
+<cx "attitudinal questions, asking intensity">  XE "attitudinal questions:=
 asking intensity"   Most often, however, the asker will use =93pei=94 as a=
 place holder for an intensity marker. (As a result, =93pei=94 is placed in=
 selma'o CAI, although selma'o UI would have been almost as appropriate. Gr=
ammatically, there is no difference between UI and CAI.) Such usage corresp=
onds to a whole range of idiomatic usages in natural languages:
+<p>
+<pre><a name=3De10d1>10.1)	.iepei
+	[agreement] [question]
+	Do you agree?
+<a name=3De10d2>10.2)	.iare'epei
+	[belief] [spiritual] [question]
+	Are you a Believer?
+<a name=3De10d3>10.3)	.aipei
+	[intention] [question]
+	Are you going to do it?
+</pre><a href=3D#e10d3>Example 10.3 </a>might appear at the end of a comma=
nd, to which the response
+<p>
+<pre><a name=3De10d4>10.4)	.aicai
+	[intention] [maximal]
+</pre>corresponds to =93Aye! Aye!=94 (hence the choice of cmavo).
+<p>
+<pre><a name=3De10d5>10.5)	.e'apei
+	[permission] [question]
+	Please, Mommy! Can I??
+</pre><cx "attitudinal questions, asking about specific attitude">  XE "at=
titudinal questions: asking about specific attitude"   Additionally, when =
=93pei=94 is used at the beginning of an indicator construct, it asks speci=
fically if that construct reflects the attitude of the respondent, as in (a=
sked of someone who has been ill or in pain):
+<p>
+<pre><a name=3De10d6>10.6)	pei.o'u
+	[question] [comfort]
+	Are you comfortable?
+<a name=3De10d7>10.7)	pei.o'ucu'i
+	[question] [comfort] [neutral]
+	Are you no longer in pain?
+<a name=3De10d8>10.8)	pei.o'usai
+	[question] [comfort] [strong]
+	Are you again healthy?
+</pre><ex "empathy">  XE "empathy: example=93  <ex "sympathy">  XE "sympat=
hy: example=93  <cx "attitudinals, attributing emotion to others">  XE "att=
itudinals: attributing emotion to others"   <cx "attitudes, empathy contras=
ted with sympathy">  XE "attitudes: empathy contrasted with sympathy"   <lx=
 "dai">  XE "dai"   <lx ".uuse'inai">  XE "uuse'inai"   Empathy, which is n=
ot really an emotion, is expressed by the indicator =93dai=94. (Don't confu=
se empathy with sympathy, which is =93.uuse'inai=94.) Sometimes, as when te=
lling a story, you want to attribute emotion to someone else. You can of co=
urse make a bridi claim that so-and-so felt such-and-such an emotion, but y=
ou can also make use of the attitudinal system by adding the indicator =93d=
ai=94, which attributes the preceding attitudinal to someone else =97 exact=
ly whom, must be determined from context. You can also use =93dai=94 conver=
sationally when you empathize, or feel someone else's emotion as if it were=
 your own:
+<p>
+<pre><a name=3De10d9>10.9)	.oiro'odai
+	[Pain!] [physical] [empathy]
+	Ouch, that must have hurt!
+</pre><p>It is even possible to =93empathize=94 with a non-living object:
+<p>
+<ex "ship sank">  XE "ship sank: example=93  <pre><a name=3De10d10>10.10)	=
le bloti .iidai .uu pu klama le xasloi
+	The ship [fear!] [empathy] [pity!] [past] goes-to the ocean-floor
+	Fearfully the ship, poor thing, sank.
+</pre>suggesting that the ship felt fear at its impending destruction, and=
 simultaneously reporting the speaker's pity for it.
+<p>
+<cx "attitudinals, exceptions">  XE "attitudinals: exceptions"   <cx "atti=
tudinals, non-speaker attitudes">  XE "attitudinals: non-speaker attitudes"=
   Both =93pei=94 and =93dai=94 represent exceptions to the normal rule tha=
t attitudinals reflect the speaker's attitude.
+<p>
+<cx "attitudes, expressing changes in">  XE "attitudes: expressing changes=
 in"   <cx "attitudinals, contours">  XE "attitudinals: contours"   <cx "at=
titudes, continuing">  XE "attitudes: continuing"   <cx "attitudes, beginni=
ng">  XE "attitudes: beginning"   <cx "attitudes, ceasing">  XE "attitudes:=
 ceasing"   <lx "bu'o">  XE "bu'o"   <lx "bu'onai">  XE "bu'onai"   Finally=
, we often want to report how our attitudes are changing. If our attitude h=
as not changed, we can just repeat the attitudinal. (Therefore, =93.ui .ui =
.ui=94 is not the same as =93.uicai=94, but simply means that we are contin=
uing to be happy.) If we want to report that we are beginning to feel, cont=
inuing to feel, or ceasing to feel an emotion, we can use the attitudinal c=
ontour cmavo =93bu'o=94.
+<p>
+When attached to an attitudinal, =93bu'o=94 means that you are starting to=
 have that attitude, =93bu'ocu'i=94 that you are continuing to have it, and=
 =93bu'onai=94 that you are ceasing to have it. Some examples:
+<p>
+<pre><a name=3De10d11>10.11)	o'onai bu'o
+	[Anger!] [start emotion]
+	I'm getting angry!
+<a name=3De10d12>10.12)	.iu bu'onai .uinai
+	[Love!] [end emotion] [unhappiness!]
+	I don't love you any more; I'm sad.
+
+</pre><p>Note the difference in effect between <a href=3D#e10d12>Example 1=
0.12 </a>and:
+<p>
+<pre><a name=3De10d13>10.13)	mi ca ba'o prami do ja'e le nu mi badri
+	I [present] [cessitive] love you with-result the event-of (I am-sad).
+	I no longer love you; therefore, I am sad.
+</pre>which is a straightforward bridi claim. <a href=3D#e10d13>Example 10=
.13 </a>states that you have (or have had) certain emotions; <a href=3D#e10=
d12>Example 10.12 </a>expresses those emotions directly.
+<p>
+<a name=3Ds11><h3>Evidentials</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre>ja'o	[jalge]	I conclude
+ca'e		I define
+ba'a	[balvi]	I expect	I experience	I remember
+su'a	[sucta]	I generalize		I particularize
+ti'e	[tirna]	I hear (hearsay)
+ka'u	[kulnu]	I know by cultural means
+se'o	[senva]	I know by internal experience
+za'a	[zgana]	I observe
+pe'i	[pensi]	I opine
+ru'a	[sruma]	I postulate
+ju'a	[jufra]	I state
+</pre><cx "Elgin, Suzette Haden and evidentials">  XE "Elgin: Suzette Hade=
n and evidentials"   <cx "American Indian languages and evidentials">  XE "=
American Indian languages and evidentials"   <cx "L=E1&aacute;adan evidenti=
als"><cx "evidentials, definition">  XE "evidentials: definition"   <cx "ev=
identials, inspiration for">  XE "evidentials: inspiration for"   Now we pr=
oceed from the attitudinal indicators and their relatives to the other, sem=
antically unrelated, categories of indicators. The indicators known as =93e=
videntials=94 show how the speaker came to say the utterance; i.e. the sour=
ce of the information or the idea. Lojban's list of evidentials was derived=
 from lists describing several American Indian languages. Evidentials are a=
lso essential to the constructed language L=E1&aacute;adan, designed by the=
 linguist and novelist Suzette Haden Elgin. L=E1&aacute;adan's set of indic=
ators was drawn on extensively in developing the Lojban indicator system.
+<p>
+<cx "evidentials in English">  XE "evidentials: in English"   It is import=
ant to realize, however, that evidentials are not some odd system used by s=
ome strange people who live at the other end of nowhere: although their Eng=
lish equivalents aren't single words, English-speakers have vivid notions o=
f what constitutes evidence, and of the different kinds of evidence.
+<p>
+<cx "evidentials, grammar">  XE "evidentials: grammar"   <cx "evidentials,=
 scales">  XE "evidentials: scales"   Like the attitudinal indicators, the =
evidentials belong to selma'o UI, and may be treated identically for gramma=
tical purposes. Most of them are not usually considered scalar in nature, b=
ut a few have associated scales.
+<p>
+<cx "indisputable bridi">  XE "indisputable bridi"   <cx "evidential, indi=
sputable bridi">  XE "evidentials: indisputable bridi"   <cx "evidential, r=
hetorical flavor">  XE "evidentials: rhetorical flavor"   A bridi with an e=
vidential in it becomes =93indisputable=94, in the sense that the speaker i=
s saying =93how it is with him or her=94, which is beyond argument. Claims =
about one's own mental states may be true or false, but are hardly subject =
to other people's examination. If you say that you think, or perceive, or p=
ostulate such-and-such a predication, who can contradict you? Discourse tha=
t uses evidentials has therefore a different rhetorical flavor than discour=
se that does not; arguments tend to become what can be called dialogues or =
alternating monologues, depending on your prejudices.
+<p>
+<cx "evidentials, placement in bridi">  XE "evidentials: placement in brid=
i"   Evidentials are most often placed at the beginning of sentences, and a=
re often attached to the =93.i=94 that separates sentences in connected dis=
course. It is in the nature of an evidential to affect the entire bridi in =
which it is placed: like the propositional attitude indicators, they strong=
ly affect the claim made by the main bridi.
+<p>
+<lx "ja'o">  XE "ja'o"   <ex "thus">  XE "thus: example=93  <ex "deduction=
">  XE "deduction: example=93  A bridi marked by =93ja'o=94 is a conclusion=
 by the speaker based on other (stated or unstated) information or ideas. R=
ough English equivalents of =93ja'o=94 are =93thus=94 and =93therefore=94.
+<p>
+<ex "husband and wife">  XE "husband and wife: example=93  <lx "ca'e">  XE=
 "ca'e"   <ex "pronouncement">  XE "pronouncement: example=93  A bridi mark=
ed by =93ca'e=94 is true because the speaker says so. In addition to defini=
tions of words, =93ca'e=94 is also appropriate in what are called performat=
ives, where the very act of speaking the words makes them true. An English =
example is =93I now pronounce you husband and wife=94, where the very act o=
f uttering the words makes the listeners into husband and wife. A Lojban tr=
anslation might be:
+<p>
+<pre><a name=3De11d1>11.1)	ca'e le re do cu simxu speni
+	[I define!] The two of-you are-mutual spouses.
+</pre><lx "ba'a">  XE "ba'a"   <lx "ba'acu'i">  XE "ba'acu'i"   <lx "ba'an=
ai">  XE "ba'anai"   <cx "evidentials, ba'a scale">  XE "evidentials: ba'a =
scale"   <ex "anticipated">  XE "anticipated: example=93  <ex "remembered">=
  XE "remembered: example=93  <ex "experienced">  XE "experienced: example=
=93  The three scale positions of =93ba'a=94, when attached to a bridi, ind=
icate that it is based on the speaker's view of the real world. Thus =93ba'=
a=94 means that the statement represents a future event as anticipated by t=
he speaker; =93ba'acu'i=94, a present event as experienced by the speaker; =
=93ba'anai=94, a past event as remembered by the speaker. It is accidental =
that this scale runs from future to past instead of past to future.
+<p>
+<pre><a name=3De11d2>11.2)	ba'acu'i le tuple be mi cu se cortu
+	[I experience!] The leg of me is-the-locus-of-pain.
+	My leg hurts.
+</pre><ex "induction">  XE "induction: example=93  <lx "su'a">  XE "su'a" =
  <cx "evidentials, su'a contrasted with ja'o">  XE "evidentials: su'a cont=
rasted with ja'o"   <cx "evidentials, ja'o contrasted with su'a">  XE "evid=
entials: ja'o contrasted with su'a"   A bridi marked by =93su'a=94 is a gen=
eralization by the speaker based on other (stated or unstated) information =
or ideas. The difference between =93su'a=94 and =93ja'o=94 is that =93ja'o=
=94 suggests some sort of reasoning or deduction (not necessarily rigorous)=
, whereas =93su'a=94 suggests some sort of induction or pattern recognition=
 from existing examples (not necessarily rigorous).
+<p>
+<ex "abduction">  XE "abduction: example=93  <lx "su'anai">  XE "su'anai" =
  The opposite point of the scale, =93su'anai=94, indicates abduction, or d=
rawing specific conclusions from general premises or patterns.
+<p>
+<cx "discursive, su'a as">  XE "discursives: su'a"   This cmavo can also f=
unction as a discursive (see <a href=3D#s12>Section 12</a>), in which case =
=93su'a=94 means =93abstractly=94 or =93in general=94, and =93su'anai=94 me=
ans =93concretely=94 or =93in particular=94.
+<p>
+<lx "ti'e">  XE "ti'e"   <ex "hearsay">  XE "hearsay: example=93  A bridi =
marked by =93ti'e=94 is relayed information from some source other than the=
 speaker. There is no necessary implication that the information was relaye=
d via the speaker's ears; what we read in a newspaper is an equally good ex=
ample of =93ti'e=94, unless we have personal knowledge of the content.
+<p>
+<pre><a name=3De11d3>11.3)	ti'e la .uengas cu zergau
+	[I hear!] Wenga is-a-criminal-doer.
+	I hear that Wenga is a crook.
+</pre><ex "myth">  XE "myth: example=93  <lx "ka'u">  XE "ka'u"   <ex "cul=
tural knowledge">  XE "cultural knowledge: example=93  A bridi marked by =
=93ka'u=94 is one held to be true in the speaker's cultural context, as a m=
atter of myth or custom, for example. Such statements should be agreed on b=
y a community of people =97 you cannot just make up your own cultural conte=
xt =97 although =93objectivity=94 in the sense of actual correspondence wit=
h the facts is certainly not required.
+<p>
+<ex "revelation">  XE "revelation: example=93  <ex "dream">  XE "dream: ex=
ample=93  <lx "se'o">  XE "se'o"   <lx "ka'u">  XE "ka'u"   <cx "evidential=
s, se'o contrasted with ka'u">  XE "evidentials: se'o contrasted with ka'u"=
   <cx "evidentials, ka'u contrasted with se'o">  XE "evidentials: ka'u con=
trasted with se'o"   On the other hand, =93se'o=94 marks a bridi whose trut=
h is asserted by the speaker as a result of an internal experience not dire=
ctly available to others, such as a dream, vision, or personal revelation. =
In some cultures, the line between =93ka'u=94 and =93se'o=94 is fuzzy or ev=
en nonexistent.
+<p>
+<lx "za'a">  XE "za'a"   <ex "observation">  XE "observation: example=93  =
<cx "observative, contrasted with observation evidential">  XE "observative=
: contrasted with observation evidential"   <cx "observation evidential, co=
ntrasted with observative ">  XE "observation evidential: contrasted with o=
bservative "   A bridi marked by =93za'a=94 is based on perception or direc=
t observation by the speaker. This use of =93observe=94 is not connected wi=
th the Lojban =93observative=94, or bridi with the first sumti omitted. The=
 latter has no explicit aspect, and could be a direct observation, a conclu=
sion, an opinion, or other aspectual point of view.
+<p>
+<pre><a name=3De11d4>11.4)	za'a do tatpi
+	[I observe!] You are-tired.
+	I see you are tired.
+</pre><lx "pe'i">  XE "pe'i"   <ex "opinion">  XE "opinion: example=93  <l=
x "pe'ipei">  XE "pe'ipei"   A bridi marked by =93pe'i=94 is the opinion of=
 the speaker. The form =93pe'ipei=94 is common, meaning =93Is this your opi=
nion?=94. (Strictly, this should be =93peipe'i=94, in accordance with the d=
istinction explained in Examples 10.6-10.8, but since =93pe'i=94 is not rea=
lly a scale, there is no real difference between the two orders.)
+<p>
+<ex "Carthage destroyed">  XE "Carthage destroyed: example=93  <pre><a nam=
e=3De11d5>11.5)	pe'i la kartagos. .ei se daspo
+	[I opine!] Carthage [obligation] is-destroyed.
+	In my opinion, Carthage should be destroyed.
+</pre><lx "ru'a">  XE "ru'a"   <cx "ru'a, compared with e'u">  XE "ru'a: c=
ompared with e'u"   <cx "e'u, compared with ru'a">  XE "e'u: compared with =
ru'a"   <lx "e'u">  XE "e'u"   <ex "assumption">  XE "assumption: example=
=93  A bridi marked by =93ru'a=94 is an assumption made by the speaker. Thi=
s is similar to one possible use of =93.e'u=94.
+<p>
+<ex "Livingston">  XE "Livingston: example=93  <pre><a name=3De11d6>11.6)	=
ru'a doi livinston.
+	Dr. Livingstone, I presume? =20
+		(A rhetorical question:  Stanley knew who he was.)
+</pre><lx "ju'a">  XE "ju'a"   <ex "basis">  XE "basis: example=93  <lx "j=
u'apei">  XE "ju'apei"   Finally, the evidential =93ju'a=94 is used to avoi=
d stating a specific basis for a statement. It can also be used when the ba=
sis for the speaker's statement is not covered by any other evidential. For=
 the most part, using =93ju'a=94 is equivalent to using no evidential at al=
l, but in question form it can be useful: =93ju'apei=94 means =93What is th=
e basis for your statement?=94 and serves as an evidential, as distinct fro=
m emotional, question.
+<p>
+<a name=3Ds12><h3>Discursives</h3>
+<p>
+<cx "discursive, definition">  XE "discursives: definition"   <cx "utteran=
ce, expressing relation to discourse">  XE "utterance: expressing relation =
to discourse"   <cx "discourse, expressing utterance relation to">  XE "dis=
course: expressing utterance relation to"   The term =93discursive=94 is us=
ed for those members of selma'o UI that provide structure to the discourse,=
 and which show how a given word or utterance relates to the whole discours=
e. To express these concepts in regular bridi would involve extra layers of=
 nesting: rather than asserting that =93I also came=94, we would have to sa=
y =93I came; furthermore, the event of my coming is an additional instance =
of the relationship expressed by the previous sentence=94, which is intoler=
ably clumsy. Typical English equivalents of discursives are words or phrase=
s like =93however=94, =93summarizing=94, =93in conclusion=94, and =93for ex=
ample=94.
+<p>
+<cx "discursives, contrasted with attitudinals">  XE "discursives: contras=
ted with attitudinals"   <cx "attitudinals, contrasted with discursives">  =
XE "attitudinals: contrasted with discursives"   <cx "discursives, as metal=
inguistic claims">  XE "discursives: as metalinguistic claims"   Discursive=
s are not attitudinals: they express no particular emotion. Rather, they ar=
e abbreviations for metalinguistic claims that reference the sentence or te=
xt they are found in.
+<p>
+<cx "discursives, placement in sentence">  XE "discursives: placement in s=
entence"   Discursives are most often used at the beginning of sentences, o=
ften attached to the =93.i=94 that separates sentences in running discourse=
, but can (like all other indicators) be attached to single words when it s=
eems necessary or useful.
+<p>
+<cx "discursives for consecutive discourse">  XE "discursives for consecut=
ive discourse"   The discursives discussed in this section are given in gro=
ups, roughly organized by function. First, the =93consecutive discourse=94 =
group:
+<p>
+<lx "ku'i">  XE "ku'i"   <lx "ji'a">  XE "ji'a"   <lx "si'a">  XE "si'a"  =
 <lx "mi'u">  XE "mi'u"   <lx "po'o">  XE "po'o"   <pre>ku'i	[karbi]	howeve=
r/but/in contrast=09
+ji'a	[jmina]	additionally	=09
+si'a	[simsa]	similarly	=09
+mi'u	[mintu]	ditto	=09
+po'o		the only relevant case
+</pre><cx "discursives for consecutive discourse, contrasted">  XE "discur=
sives for consecutive discourse: contrasted"   <cx "mi'u, contrasted with g=
o'i">  XE "mi'u: contrasted with go'i"   <cx "go'i, contrasted with mi'u"> =
 XE "go'i: contrasted with mi'u"   <lx "go'i">  XE "go'i"   <ex "ditto">  X=
E "ditto: example=93  <ex "but">  XE "but: example=93  <ex "too">  XE "too:=
 example=93  These five discursives are mutually exclusive, and therefore t=
hey are not usually considered as scales. The first four are used in consec=
utive discourse. The first, =93ku'i=94, makes an exception to the previous =
argument. The second, =93ji'a=94, adds weight to the previous argument. The=
 third, =93si'a=94, adds quantity to the previous argument, enumerating an =
additional example. The fourth, =93mi'u=94, adds a parallel case to the pre=
vious argument, and can also be used in tables or the like to show that som=
ething is being repeated from the previous column. It is distinct from =93g=
o'i=94 (of selma'o GOhA, discussed in <a href=3Dchap7.html>Chapter 7</a>), =
which is a non-discursive version of =93ditto=94 that explicitly repeats th=
e claim of the previous bridi.
+<p>
+<ex "only">  XE "only: example=93  Lastly, =93po'o=94 is used when there i=
s no other comparable case, and thus corresponds to some of the uses of =93=
only=94, a word difficult to express in pure bridi form:
+<p>
+<ex "hit cousin">  XE "hit cousin: example=93  <ex "hit nose">  XE "hit no=
se: example=93  <pre><a name=3De12d1>12.1)	mi po'o darxi le mi tamne fo le =
nazbi
+	I [only] hit my cousin at-locus the nose.
+	Only I (nobody else) hit my cousin on his nose.
+<a name=3De12d2>12.2)	mi darxi po'o le mi tamne fo le nazbi
+	I hit [only] my cousin at-locus the nose.
+	I only hit my cousin on his nose (I did nothing else to him).
+<a name=3De12d3>12.3)	mi darxi le mi tamne po'o fo le nazbi
+	I hit my cousin [only] at-locus the nose.
+	I hit only my cousin on his nose (no one else).
+<a name=3De12d4>12.4)	mi darxi le mi tamne fo le nazbi po'o
+	I hit my cousin at-locus the nose [only].
+	I hit my cousin only on his nose (nowhere else).
+</pre><cx "po'o, placement in sentence">  XE "po'o: placement in sentence"=
   Note that =93only=94 can go before or after what it modifies in English,=
 but =93po'o=94, as an indicator, always comes afterward.
+<p>
+Next, the =93commentary on words=94 group:
+<p>
+<lx "va'i">  XE "va'i"   <lx "ta'u">  XE "ta'u"   <pre>va'i	[valsi]	in oth=
er words		in the same words
+ta'u	[tanru]	expanding a tanru	making a tanru
+</pre><cx "discursives, word-level">  XE "discursives: word-level"   <cx "=
discursives, expressing how things are said">  XE "discursives: expressing =
how things are said"   The discursives =93va'i=94 and =93ta'u=94 operate at=
 the level of words, rather than discourse proper, or if you like, they dea=
l with how things are said. An alternative English expression for =93va'i=
=94 is =93rephrasing=94; for =93va'inai=94, =93repeating=94. Also compare =
=93va'i=94 with =93ke'u=94, discussed below.
+<p>
+<cx "tanru, expanding">  XE "tanru: expanding"   <cx "tanru, explicating">=
  XE "tanru: explicating"   <cx "tanru, explicitly defining">  XE "tanru: e=
xplicitly defining"   <lx "ta'unai">  XE "ta'unai"   The cmavo =93ta'u=94 i=
s a discursive unique to Lojban; it expresses the particularly Lojbanic dev=
ice of tanru. Since tanru are semantically ambiguous, they are subject to m=
isunderstanding. This ambiguity can be removed by expanding the tanru into =
some semantically unambiguous structure, often involving relative clauses o=
r the introduction of additional brivla. The discursive =93ta'u=94 marks th=
e transition from the use of a brief but possibly confusing tanru to its fu=
ller, clearer expansion; the discursive =93ta'unai=94 marks a transition in=
 the reverse direction.
+<p>
+Next, the =93commentary on discourse=94 group:
+<p>
+<lx "li'a">  XE "li'a"   <lx "ba'u">  XE "ba'u"   <lx "zo'o">  XE "zo'o"  =
 <lx "sa'e">  XE "sa'e"   <lx "to'u">  XE "to'u"   <lx "do'a">  XE "do'a"  =
 <lx "sa'u">  XE "sa'u"   <lx "pa'e">  XE "pa'e"   <lx "ge'u">  XE "ge'u"  =
 <pre> li'a	[klina]	clearly		obscurely
+		obviously=09
+ba'u	[banli]	exaggeration	accuracy	understatement
+zo'o		humorously	dully	seriously
+sa'e	[satci]	precisely speaking	loosely speaking
+to'u	[tordu]	in brief		in detail
+do'a	[dunda]	generously		parsimoniously
+sa'u	[sampu]	simply		elaborating
+pa'e	[pajni]	justice		prejudice
+je'u	[jetnu]	truly		falsely
+</pre><lx "zo'o">  XE "zo'o"   <cx "discursives, discourse commentary">  X=
E "discursives: discourse commentary"   <cx "discourse, commentary on">  XE=
 "discourse: commentary on"   <cx "discourse, tone of voice markers">  XE "=
discourse: tone of voice markers"   <cx "discourse, gesture markers">  XE "=
discourse: gesture markers"   This group is used by the speaker to characte=
rize the nature of the discourse, so as to prevent misunderstanding. It is =
well-known that listeners often fail to recognize a humorous statement and =
take it seriously, or miss an exaggeration, or try to read more into a stat=
ement than the speaker intends to put there. In speech, the tone of voice o=
ften provides the necessary cue, but the reader of ironic or understated or=
 imprecise discourse is often simply clueless. As with the attitudinals, th=
e use of these cmavo may seem fussy to new Lojbanists, but it is important =
to remember that =93zo'o=94, for example, is the equivalent of smiling whil=
e you speak, not the equivalent of a flat declaration like =93What I'm abou=
t to say is supposed to be funny.=94
+<p>
+<lx "sa'enai">  XE "sa'enai"   A few additional English equivalents: for =
=93sa'enai=94, =93roughly speaking=94 or =93approximately speaking=94; for =
=93sa'unai=94, =93furthermore=94; for =93to'u=94, =93in short=94 or =93skip=
ping details=94; for =93do'a=94, =93broadly construed=94; for =93do'anai=94=
 (as you might expect), =93narrowly construed=94.
+<p>
+<lx "pa'e">  XE "pa'e"   <lx "pa'enai">  XE "pa'enai"   The cmavo =93pa'e=
=94 is used to claim (truly or falsely) that one is being fair or just to a=
ll parties mentioned, whereas =93pa'enai=94 admits (or proclaims) a bias in=
 favor of one party.
+<p>
+<ex "sarcasm">  XE "sarcasm: example=93  <ex "irony">  XE "irony: example=
=93  <cx "sarcasm, expressing">  XE "sarcasm: expressing"   <cx "irony, exp=
ressing">  XE "irony: expressing"   <lx "zo'o">  XE "zo'o"   <lx ".ianai"> =
 XE "ianai"   The scale of =93je'u=94 and =93je'unai=94 is a little differe=
nt from the others in the group. By default, we assume that people speak th=
e truth =97 or at least, that if they are lying, they will do their best to=
 conceal it from us. So under what circumstances would =93je'unai=94 be use=
d, or =93je'u=94 be useful? For one thing, =93je'u=94 can be used to mark a=
 tautology: a sentence that is a truth of logic, like =93All cats are cats.=
=94 Its counterpart =93je'unai=94 then serves to mark a logical contradicti=
on. In addition, =93je'unai=94 can be used to express one kind of sarcasm o=
r irony, where the speaker pretends to believe what he/she says, but actual=
ly wishes the listener to infer a contrary opinion. Other forms of irony ca=
n be marked with =93zo'o=94 (humor) or =93.ianai=94 (disbelief).
+<p>
+<lx "su'a">  XE "su'a"   When used as a discursive, =93su'a=94 (see <a hre=
f=3D#s11>Section 11</a>) belongs to this group.
+<p>
+Next, the =93knowledge=94 group:
+<p>
+<cx "knowledge discursives">  XE "knowledge discursives"   <cx "discursive=
s, knowledge">  XE "discursives: knowledge"   <lx "ju'o">  XE "ju'o"   <lx =
"la'a">  XE "la'a"   <pre>ju'o	[djuno]	certainly	uncertain	certainly not
+la'a	[lakne]	probably		improbably
+</pre><cx "speaker's state of knowledge">  XE "speaker's state of knowledg=
e"   <cx "propositional attitudes, compared with knowledge discursives">  X=
E "propositional attitudes: compared with knowledge discursives"   <cx "kno=
wledge discursives, compared with propositional attitudes">  XE "knowledge =
discursives: compared with propositional attitudes"   These two discursives=
 describe the speaker's state of knowledge about the claim of the associate=
d bridi. They are similar to the propositional attitudes of <a href=3D#s3>S=
ection 3</a>, as they create a hypothetical world. We may be quite certain =
that something is true, and label our bridi with =93ju'o=94; but it may be =
false all the same.
+<p>
+Next, the =93discourse management=94 group:
+<p>
+<lx "ta'o">  XE "ta'o"   <lx "ra'u">  XE "ra'u"   <lx "mu'a">  XE "mu'a"  =
 <lx "zu'u">  XE "zu'u"   <lx "ke'u">  XE "ke'u"   <lx "da'i">  XE "da'i"  =
 <pre>ta'o	[tanjo]	by the way		returning to point
+ra'u	[ralju]	chiefly	equally	incidentally
+mu'a	[mupli]	for example	omitting	end examples
+			  examples
+zu'u		on the one hand		on the other hand		  =20
+ke'u	[krefu]	repeating		continuing
+da'i		supposing		in fact
+</pre><cx "discursives, discourse management">  XE "discursives: discourse=
 management"   <cx "discursives for managing discourse flow">  XE "discursi=
ves for managing discourse flow"   <cx "flow of discourse, managing with di=
scursives">  XE "flow of discourse: managing with discursives"   This final=
 group is used to perform what may be called =93managing the discourse=94: =
providing reference points to help the listener understand the flow from on=
e sentence to the next.
+<p>
+<lx "ta'onao">  XE "ta'onao"   Other English equivalents of =93ta'onai=94 =
are =93anyway=94, =93anyhow=94, =93in any case=94, =93in any event=94, =93a=
s I was saying=94, and =93continuing=94.
+<p>
+<cx "importance of point, scale with ra'u">  XE "importance of point: scal=
e with ra'u"   <cx "ra'u, scale of importance">  XE "ra'u: scale of importa=
nce"   The scale of =93ra'u=94 has to do with the importance of the point b=
eing, or about to be, expressed: =93ra'u=94 is the most important point, =
=93ra'ucu'i=94 is a point of equal importance, and =93ra'unai=94 is a lesse=
r point. Other English equivalents of =93ra'u=94 are =93above all=94 and =
=93primarily=94.
+<p>
+<cx "va'i, contrasted with ke'u">  XE "va'i: contrasted with ke'u"   <cx "=
ke'u, contrasted with va'i">  XE "ke'u: contrasted with va'i"   <lx "va'i">=
  XE "va'i"   <lx "va'inai">  XE "va'inai"   <lx "ke'unai">  XE "ke'unai"  =
 The cmavo =93ke'u=94 is very similar to =93va'i=94, although =93ke'unai=94=
 and =93va'inai=94 are quite different. Both =93ke'u=94 and =93va'i=94 indi=
cate that the same idea is going to be expressed using different words, but=
 the two cmavo differ in emphasis. Using =93ke'u=94 emphasizes that the con=
tent is the same; using =93va'i=94 emphasizes that the words are different.=
 Therefore, =93ke'unai=94 shows that the content is new (and therefore the =
words are also); =93va'inai=94 shows that the words are the same (and there=
fore so is the content). One English equivalent of =93ke'unai=94 is =93furt=
hermore=94.
+<p>
+<cx "real world point of view">  XE "real world point of view"   <cx "hypo=
thetical world point of view">  XE "hypothetical world point of view"   <cx=
 "if, expressing hypothetical world">  XE =93if, expressing hypothetical wo=
rld"   <cx "if, expressing real world">  XE =93if: expressing real world"  =
 The discursive =93da'i=94 marks the discourse as possibly taking a non-rea=
l-world viewpoint (=93Supposing that=94, =93By hypothesis=94), whereas =93d=
a'inai=94 insists on the real-world point of view (=93In fact=94, =93In tru=
th=94, =93According to the facts=94). A common use of =93da'i=94 is to dist=
inguish between:
+<p>
+<ex "sister pregnant">  XE "sister pregnant: example=93  <ex "pregnant sis=
ter">  XE "pregnant sister: example=93  <cx "real world, contrasted with hy=
pothetical world, example">  XE "real world: contrasted with hypothetical w=
orld, example"   <cx "hypothetical world, contrasted with real world, examp=
le">  XE "hypothetical world: contrasted with real world, example"   <pre><=
a name=3De12d5>12.5)	ganai da'i do viska le mi citno mensi gi ju'o do djuno=
=20
+le du'u ri pazvau
+	If you [hypothetical] see my young sister, then [certain] you know that
+she is-pregnant.
+	If you were to see my younger sister, you would certainly know she=20
+is pregnant.
+</pre>and:
+<p>
+<pre><a name=3De12d6>12.6)	ganai da'inai do viska le mi citno mensi gi ju'=
o do djuno=20
+le du'u ri pazvau
+	If you [factual] see my young sister, then [certainty] you know that=20
+she is-pregnant.
+	If you saw my younger sister, you would certainly know she is pregnant.
+</pre><p>It is also perfectly correct to omit the discursive altogether, a=
nd leave the context to indicate which significance is meant. (Chinese alwa=
ys leaves this distinction to the context: the Chinese sentence
+<p>
+<pre><a name=3De12d7>12.7)	ru2<sup>2</sup>guo3 ni3 kan4<sup>4</sup>dao4 wo=
3 mei4<sup>4</sup>mei, ni3 yi2<sup>2</sup>ding4 zhi1<sup>1</sup>dao4 ta1 hu=
ai2<sup>2</sup>yun4 le
+	if you see-arrive my younger-sister, you certainly know she pregnant
+</pre>is the equivalent of either <a href=3D#e12d5>Example 12.5 </a>or <a =
href=3D#e12d6>Example 12.6</a>.)
+<p>
+<a name=3Ds13><h3>Miscellaneous indicators</h3>
+<p>
+Some indicators do not fall neatly into the categories of attitudinal, evi=
dential, or discursive. This section discusses the following miscellaneous =
indicators:
+<p>
+<pre> 	ki'a	metalinguistic confusion
+	na'i	metalinguistic negator
+	jo'a	metalinguistic affirmer
+	li'o	omitted text (quoted material)
+	sa'a	material inserted by editor/narrator
+	xu	true-false question
+	pau	question premarker			rhetorical question
+	pe'a	figurative language			literal language
+	bi'u	new information			old information
+	ge'e	non-specific indicator
+</pre><lx "ki'a">  XE "ki'a"   <ex "huh?">  XE "huh?: example=93   <cx "co=
nfusion, metalinguistic">  XE "confusion: metalinguistic"   <cx "confusion =
about what was said">  XE "confusion about what was said"   The cmavo =93ki=
'a=94 is one of the most common of the miscellaneous indicators. It express=
es metalinguistic confusion; i.e. confusion about what has been said, as op=
posed to confusion not tied to the discourse (which is =93.uanai=94). The c=
onfusion may be about the meaning of a word or of a grammatical construct, =
or about the referent of a sumti. One of the uses of English =93which=94 co=
rresponds to =93ki'a=94:
+<p>
+<pre><a name=3De13d1>13.1)	mi nelci le ctuca
+.i le ki'a ctuca
+	I like the teacher
+	Which teacher?
+</pre><p>Here, the second speaker does not understand the referent of the =
sumti =93le ctuca=94, and so echoes back the sumti with the confusion marke=
r.
+<p>
+<lx "na'i">  XE "na'i"   <lx "jo'a">  XE "jo'a"   <cx "error marking, meta=
linguistic">  XE "error marking: metalinguistic"   <cx "invalid speech, mar=
king as error with na'i">  XE "invalid speech: marking as error with na'i" =
  The metalinguistic negation cmavo =93na'i=94 and its opposite =93jo'a=94 =
are explained in full in <a href=3Dchap15.html>Chapter 15</a>. In general, =
=93na'i=94 indicates that there is something wrong with a piece of discours=
e: either an error, or a false underlying assumption, or something else of =
the sort. The discourse is invalid or inappropriate due to the marked word =
or construct.
+<p>
+<cx "valid speech, marking as error with jo'a">  XE "valid speech: marking=
 as error with jo'a"   Similarly, =93jo'a=94 marks something which looks wr=
ong but is in fact correct. These two cmavo constitute a scale, but are kep=
t apart for two reasons: =93na'inai=94 means the same as =93jo'a=94, but wo=
uld be too confusing as an affirmation; =93jo'anai=94 means the same as =93=
na'i=94, but is too long to serve as a convenient metalinguistic negator.
+<p>
+<cx "fragmentary text">  XE "fragmentary text"   <lx "li'o">  XE "li'o"   =
<cx "partial quotation">  XE "partial quotation"   The next two cmavo are u=
sed to assist in quoting texts written or spoken by others. It is often the=
 case that we wish to quote only part of a text, or to supply additional ma=
terial either by way of commentary or to make a fragmentary text grammatica=
l. The cmavo =93li'o=94 serves the former function. It indicates that words=
 were omitted from the quotation. What remains of the quotation must be gra=
mmatical, however, as =93li'o=94 does not serve any grammatical function. I=
t cannot, for example, take the place of a missing selbri in a bridi, or su=
pply the missing tail of a description sumti: =93le li'o=94 in isolation is=
 not grammatical.
+<p>
+<lx "sa'a">  XE "sa'a"   <cx "editorial insertion, with "sa'a">  XE "edito=
rial insertion: with sa'a"   <lx "sei">  XE "sei"   <lx "to'i">  XE "to'i" =
  <cx "sa'a, interaction with li'o">  XE "sa'a: interaction with li'o"   <c=
x "sa'a, interaction with sei">  XE "sa'a: interaction with sei"   <cx "sa'=
a, interaction with to'i">  XE "sa'a: interaction with to'i"   The cmavo =
=93sa'a=94 indicates in a quotation that the marked word or construct was n=
ot actually expressed, but is inserted for editorial, narrative, or grammat=
ical purposes. Strictly, even a =93li'o=94 should appear in the form =93li'=
osa'a=94, since the =93li'o=94 was not part of the original quotation. In p=
ractice, this and other forms which are already associated with metalinguis=
tic expressions, such as =93sei=94 (of selma'o SEI) or =93to'i=94 (of selma=
'o TO) need not be marked except where confusion might result.
+<p>
+<cx "editorial insertion, of text already containing sa'a">  XE "editorial=
 insertion: of text already containing sa'a"   <cx "sa'a, editorial inserti=
on of text already containing sa'a">  XE "sa'a: editorial insertion of text=
 already containing sa'a"   In the rare case that the quoted material alrea=
dy contains one or more instances of =93sa'a=94, they can be changed to =93=
sa'asa'a=94.
+<p>
+<lx "xu">  XE "xu"   <cx "yes/no questions">  XE "yes/no questions"   <cx =
"truth questions">  XE "truth questions"   <cx "questions, with "xu">  XE "=
questions: with \"xu"   The cmavo =93xu=94 marks truth questions, which are=
 discussed in detail in <a href=3Dchap15.html>Chapter 15</a>. In general, =
=93xu=94 may be translated =93Is it true that=A0<dots>=85</dots>?=94 and qu=
estions whether the attached bridi is true. When =93xu=94 is attached to a =
specific word or construct, it directs the focus of the question to that wo=
rd or construct.
+<p>
+<lx "pau">  XE "pau"   <cx "questions, marking in advance">  XE "questions=
: marking in advance"   Lojban question words, unlike those of English, fre=
quently do not stand at the beginning of the question. Placing the cmavo =
=93pau=94 at the beginning of a bridi helps the listener realize that the b=
ridi is a question, like the symbol at the beginning of written Spanish que=
stions that looks like an upside-down question mark. The listener is then w=
arned to watch for the actual question word.
+<p>
+<cx "pau, placement in sentence">  XE "pau: placement in sentence"   <cx "=
rhetorical question">  XE "rhetorical question"   <cx "question, rhetorical=
">  XE =93questions: rhetorical"   <lx "paunai">  XE "paunai"   Although =
=93pau=94 is grammatical in any location (like all indicators), it is not r=
eally useful except at or near the beginning of a bridi. Its scalar opposit=
e, =93paunai=94, signals that a bridi is not really a question despite its =
form. This is what we call in English a rhetorical question: an example app=
ears in the English text near the beginning of <a href=3D#s11>Section 11</a=
>.
+<p>
+<lx "pe'a">  XE "pe'a"   <cx "figurative speech">  XE "figurative speech" =
  <ex "blue, as sad">  XE "blue: as sad, example=93  The cmavo =93pe'a=94 i=
s the indicator of figurative speech, indicating that the previous word sho=
uld be taken figuratively rather than literally:
+<p>
+<pre><a name=3De13d2>13.2)	mi viska le blanu pe'a zdani
+	I see the blue [figurative] house.
+	I see the =93blue=94 house.
+</pre><p>Here the house is not blue in the sense of color, but in some oth=
er sense, whose meaning is entirely culturally dependent. The use of =93pe'=
a=94 unambiguously marks a cultural reference: =93blanu=94 in <a href=3D#e1=
3d2>Example 13.2 </a>could mean =93sad=94 (as in English) or something comp=
letely different.
+<p>
+<lx "pe'anai">  XE "pe'anai"   <cx "literally">  XE "literally"   The nega=
ted form, =93pe'anai=94, indicates that what has been said is to be interpr=
eted literally, in the usual way for Lojban; natural-language intuition is =
to be ignored.
+<p>
+<cx "lujvo, place structure of figurative lujvo">  XE "lujvo: place struct=
ure of figurative lujvo"   <cx "figurative lujvo">  XE "figurative lujvo"  =
 <cx "figurative lujvo, place structure">  XE "figurative lujvo: place stru=
cture"   <cx "culturally dependent lujvo">  XE "culturally dependent lujvo"=
   <ex "heartburn">  XE "heartburn: example=93  Alone among the cmavo of se=
lma'o UI, =93pe'a=94 has a rafsi, namely =93pev=94. This rafsi is used in f=
orming figurative (culturally dependent) lujvo, whose place structure need =
have nothing to do with the place structure of the components. Thus =93risn=
yjelca=94 (heart burn) might have a place structure like:
+<p>
+<dl compact><dt><dd>x1 is the heart of x2, burning in atmosphere x3 at tem=
perature x4
+</dl>whereas =93pevrisnyjelca=94, explicitly marked as figurative, might h=
ave the place structure:
+<p>
+<dl compact><dt><dd>x1 is indigestion/heartburn suffered by x2
+</dl>which obviously has nothing to do with the places of either =93risna=
=94 or =93jelca=94.
+<p>
+<lx "bi'u">  XE "bi'u"   <lx "bi'unai">  XE "bi'unai"   <ex "the">  XE "th=
e: example=93  <ex "a">  XE "a: example=93  <ex "an">  XE "an: example=93  =
<cx "the, contrasted with a/an">  XE "the: contrasted with a/an"   <cx "a/a=
n, contrasted with the">  XE "a/an: contrasted with the"   The uses of =93b=
i'u=94 and =93bi'unai=94 correspond to one of the uses of the English artic=
les =93the=94 and =93a/an=94. An English-speaker telling a story may begin =
with =93I saw a man who=A0<dots>=85</dots>=94. Later in the story, the same=
 man will be referred to with the phrase =93the man=94. Lojban does not use=
 its articles in the same way: both =93a man=94 and =93the man=94 would be =
translated =93le nanmu=94, since the speaker has in mind a specific man. Ho=
wever, the first use might be marked =93le bi'u nanmu=94, to indicate that =
this is a new man, not mentioned before. Later uses could correspondingly b=
e tagged =93le bi'unai nanmu=94.
+<p>
+Most of the time, the distinction between =93bi'u=94 and =93bi'unai=94 nee=
d not be made, as the listener can infer the right referent. However, if a =
different man were referred to still later in the story, =93le bi'u nanmu=
=94 would clearly show that this man was different from the previous one.
+<p>
+<lx "ge'e">  XE "ge'e"   <cx "attitude, avoidance of expression">  XE "att=
itude: avoidance of expression"   Finally, the indicator =93ge'e=94 has bee=
n discussed in <a href=3D#s8>Sections 8 </a>and <a href=3D#s10>10</a>. It i=
s used to express an attitude which is not covered by the existing set, or =
to avoid expressing any attitude.
+<p>
+Another use for =93ge'e=94 is to explicitly avoid expressing one's feeling=
 on a given scale; in this use, it functions like a member of selma'o CAI: =
=93.iige'e=94 means roughly =93I'm not telling whether I'm afraid or not.=
=94
+<p>
+<lx "kau">  XE "kau"   <cx "indirect question">  XE "indirect question"   =
<pre>	kau		indirect question
+</pre><p>This cmavo is explained in detail in <a href=3Dchap11.html>Chapte=
r 11</a>. It marks the word it is attached to as the focus of an indirect q=
uestion:
+<p>
+<pre><a name=3De13d3>13.3)	mi djuno le du'u dakau klama le zarci
+	I know the statement-that somebody [indirect ?] goes to-the store.
+	I know who goes to the store.
+</pre><a name=3Ds14><h3>Vocative scales</h3>
+<p>
+<cx "vocatives, definition">  XE "vocatives: definition"   <cx "vocatives,=
 contrasted with =93la=94">  XE "vocatives: contrasted with \=93la\=94"   <=
cx "=93la=94, contrasted with vocatives">  XE "la: contrasted with vocative=
s"   <lx "COI">  XE "COI selma=92o"   <cx "direct address">  XE "direct add=
ress"   =93=93Vocatives=94 are words used to address someone directly; they=
 precede and mark a name used in direct address, just as =93la=94 (and the =
other members of selma'o LA) mark a name used to refer to someone. The voca=
tives actually are indicators =97 in fact, discursives =97 but the need to =
tie them to names and other descriptions of listeners requires them to be s=
eparated from selma'o UI. But like the cmavo of UI, the members of selma'o =
COI can be =93negated=94 with =93nai=94 to get the opposite part of the sca=
le.
+<p>
+<cx "redundancy, effect on vocative design">  XE "redundancy: effect on vo=
cative design"   <cx "vocatives, rationale for redundancy">  XE "vocatives:=
 rationale for redundancy"   Because of the need for redundancy in noisy en=
vironments, the Lojban design does not compress the vocatives into a minimu=
m number of scales. Doing so would make a non-redundant =93nai=94 too often=
 vital to interpretation of a protocol signal, as explained later in this s=
ection.
+<p>
+<lx "DOhU">  XE "DOhU selma=92o"   <lx "do'u">  XE "do'u"   <cx "vocatives=
, grammar overview">  XE "vocatives: grammar overview"   The grammar of voc=
atives is explained in <a href=3Dchap6.html>Chapter 6</a>; but in brief, a =
vocative may be followed by a name (without =93la=94), a description (witho=
ut =93le=94 or its relatives), a complete sumti, or nothing at all (if the =
addressee is obvious from the context). There is an elidable terminator, =
=93do'u=94 (of selma'o DOhU) which is almost never required unless no name =
(or other indication of the addressee) follows the vocative.
+<p>
+<cx "you, defining">  XE "you: defining"   <cx "vocatives, and definition =
of =93you=94">  XE "vocatives: and definition of \=93you\=94"   Using any v=
ocative except =93mi'e=94 (explained below) implicitly defines the meaning =
of the pro-sumti =93do=94, as the whole point of vocatives is to specify th=
e listener, or at any rate the desired listener =97 even if the desired lis=
tener isn't listening! We will use the terms =93speaker=94 and =93listener=
=94 for clarity, although in written Lojban the appropriate terms would be =
=93writer=94 and =93reader=94.
+<p>
+<cx "vocatives, notation convention symbol =93X=94">  XE "vocatives: notat=
ion convention symbol \=93X\=94"   In the following list of vocatives, the =
translations include the symbol X. This represents the name (or identifying=
 description, or whatever) of the listener.
+<p>
+<lx "doi">  XE "doi"   <lx "DOI">  XE "DOI selma=92o"   <cx "pause before =
name, effect of doi">  XE "pause before name: effect of doi"   <cx "doi, ef=
fect on pause before name">  XE "doi: effect on pause before name"   The cm=
avo =93doi=94 is the general-purpose vocative. Unlike the cmavo of selma'o =
COI, explained below, =93doi=94 can precede a name directly without an inte=
rvening pause. It is not considered a scale, and =93doinai=94 is not gramma=
tical. In general, =93doi=94 needs no translation in English (we just use n=
ames by themselves without any preceding word, although in poetic styles we=
 sometimes say =93Oh X=94, which is equivalent to =93doi=94). One may attac=
h an attitudinal to =93doi=94 to express various English vocatives. For exa=
mple, =93doi .io=94 means =93Sir/Madam!=94, whereas =93doi .ionai=94 means =
=93You there!=94.
+<p>
+<cx "pause before name, effect of vocatives of COI">  XE "pause before nam=
e: effect of vocatives of COI"   <cx "COI, effect on pause before name">  X=
E "COI selma=92o: effect on pause before name"   All members of selma'o COI=
 require a pause when used immediately before a name, in order to prevent t=
he name from absorbing the COI word. This is unlike selma'o DOI and LA, whi=
ch do not require pauses because the syllables of these cmavo are not permi=
tted to be embedded in a Lojban name. When calling out to someone, this is =
fairly natural, anyway. =93Hey! John!=94 is thus a better translation of =
=93ju'i .djan.=94 than =93Hey John!=94. No pause is needed if the vocative =
reference is something other than a name, as in the title of the Lojban jou=
rnal, =93ju'i lobypli=94.
+<p>
+(Alternatively, =93doi=94 can be inserted between the COI cmavo and the na=
me, making a pause unnecessary: =93coi doi djan.=94)
+<p>
+<lx "coi">  XE "coi"   <pre>coi		greetings
+</pre>=93Hello, X=94; =93Greetings, X=94; indicates a greeting to the list=
ener.
+<p>
+<lx "co'o">  XE "co'o"   <pre>co'o		partings
+</pre>=93Good-bye, X=94; indicates parting from immediate company by eithe=
r the speaker or the listener. =93coico'o=94 means =93greeting in passing=
=94.
+<p>
+<lx "ju'i">  XE "ju'i"   <pre>ju'i	[jundi]	attention	at ease	ignore me/us
+</pre>=93Attention/Lo/Hark/Behold/Hey!/Listen, X=94; indicates an importan=
t communication that the listener should listen to.
+<p>
+<lx "nu'e">  XE "nu'e"   <pre>nu'e	[nupre]	promise	release promise		non-pr=
omise
+</pre>=93I promise, X=94; indicates a promise to the listener. In some con=
texts, =93nu'e=94 may be prefixed to an oath or other formal declaration.
+<p>
+<lx "ta'e">  XE "ta'e"   <pre>ta'a	[tavla]	interruption
+</pre>=93I interrupt, X=94, =93I desire the floor, X=94; a vocative expres=
sion to (possibly) interrupt and claim the floor to make a statement or exp=
ression. This can be used for both rude and polite interruptions, although =
rude interruptions will probably tend not to use a vocative at all. An appr=
opriate response to an interruption might be =93re'i=94 (or =93re'inai=94 t=
o ignore the interruption).
+<p>
+<lx "pe'u">  XE "pe'u"   <pre>pe'u	[cpedu]	request
+</pre><cx "pe'u, contrasted with e'o">  XE "pe'u: contrasted with e'o"   <=
cx "e'o, contrasted with pe'u">  XE "e'o: contrasted with pe'u"   =93=93Ple=
ase, X=94; indicates a request to the listener. It is a formal, non-attitud=
inal, equivalent of =93.e'o=94 with a specific recipient being addressed. O=
n the other hand, =93.e'o=94 may be used when there is no specific listener=
, but merely a =93sense of petition floating in the air=94, as it were.
+<p>
+<lx "ki'e">  XE "ki'e"   <pre>ki'e	[ckire]	appreciation		disappreciation
+		gratitude		ingratitude
+</pre><ex "thank you">  XE "thank you: example=93  <cx "politeness, thank =
you and you're welcome">  XE "politeness: thank you and you're welcome"   <=
cx "politeness, you're welcome">  XE "politeness: you're welcome"   <lx "je=
'e">  XE "je'e"   <lx "fi'i">  XE "fi'i"   =93=93Thank you, X=94; indicates=
 appreciation or gratitude toward the listener. The usual response is =93je=
'e=94, but =93fi'i=94 is appropriate on rare occasions: see the explanation=
 of =93fi'i=94.
+<p>
+<pre>fi'i	[friti]	welcome,		unwelcome,
+		offering		inhospitality
+</pre><lx "fi'i">  XE "fi'i"   <cx "you're welcome, fi'i contrasted with j=
e'e">  XE "you're welcome: fi'i contrasted with je'e"   <cx "you're welcome=
, je'e contrasted with fi'i">  XE "you're welcome: je'e contrasted with fi'=
i"   <ex "hospitality">  XE "hospitality: example=93  =93=93At your service=
, X=94; =93Make yourself at home, X=94; offers hospitality (possibly in res=
ponse to thanks, but not necessarily) to the listener. Note that =93fi'i=94=
 is not the equivalent of American English =93You're welcome=94 as a mechan=
ical response to =93Thank you=94; that is =93je'e=94, as noted below.
+<p>
+<pre>be'e	[benji]	request to send
+</pre><lx "be'e">  XE "be'e"   <cx "telephone conversation, hello">  XE "t=
elephone conversation: hello"   =93=93Request to send to X=94; indicates th=
at the speaker wishes to express something, and wishes to ensure that the l=
istener is listening. In a telephone conversation, can be used to request t=
he desired conversant(s). A more colloquial equivalent is =93Hello? Can I s=
peak to X?=94.
+<p>
+<pre>re'i	[bredi]	ready to receive		not ready
+</pre><lx "re'i">  XE "re'i"   =93=93Ready to receive, X=94; indicates tha=
t the speaker is attentive and awaiting communication from the listener. It=
 can be used instead of =93mi'e=94 to respond when called to the telephone.=
 The negative form can be used to prevent the listener from continuing to t=
alk when the speaker is unable to pay attention: it can be translated =93Ho=
ld on!=94 or =93Just a minute=94.
+<p>
+<pre>mu'o	[mulno]	completion of utterance	more to follow
+</pre><lx "mu'o">  XE "mu'o"   =93=93Over, X=94; indicates that the speake=
r has completed the current utterance and is ready to hear a response from =
the listener. The negative form signals that the pause or non-linguistic so=
und which follows does not represent the end of the current utterance: more=
 colloquially, =93I'm not done talking!=94
+<p>
+<pre>je'e	[jimpe]	successful receipt		unsuccessful receipt
+</pre><lx "je'e">  XE "je'e"   <ex "roger">  XE "roger: example=93  <cx "p=
oliteness, you're welcome">  XE "politeness: you're welcome"   =93=93Roger,=
 X!=94, =93I understand=94; acknowledges the successful receipt of a commun=
ication from the listener. The negative form indicates failure to receive c=
orrectly, and is usually followed by =93ke'o=94. The colloquial English equ=
ivalents of =93je'e=94 and =93je'enai=94 are the grunt typically written =
=93uh-huh=94 and =93What?/Excuse me?=94: =93je'e=94 is also used to mean =
=93You're welcome=94 when that is a response to =93Thank you=94.
+<p>
+<pre>vi'o		will comply		will not comply
+</pre><lx "vi'o">  XE "vi'o"   <cx "je'e, contrasted with vi'o">  XE "je'e=
: contrasted with vi'o"   <cx "vi'o, contrasted with je'e">  XE "vi'o: cont=
rasted with je'e"   =93=93Wilco, X=94, =93I understand and will comply=94. =
Similar to =93je'e=94 but signals an intention (similar to =93.ai=94) to co=
mply with the other speaker's request. This cmavo is the main way of saying=
 =93OK=94 in Lojban, in the usual sense of =93Agreed!=94, although =93.ie=
=94 carries some of the same meaning. The negative form indicates that the =
message was received but that you will not comply: a very colloquial versio=
n is =93No way!=94.
+<p>
+<pre>ke'o	[krefu]	please repeat		no repeat needed
+</pre><lx "ke'o">  XE "ke'o"   <cx "ke'o, compared to ki'a">  XE "ke'o: co=
mpared to ki'a"   <cx "ki'a, compared to ke=92o">  XE "ki'a: compared to ke=
=92o"   =93=93What did you say, X?=94; a request for repetition or clarific=
ation due to unsuccessful receipt or understanding. This is the vocative eq=
uivalent of =93ki'a=94, and is related to =93je'enai=94. The negative form =
may be rendered =93Okay, already; I get the point!=94
+<p>
+<pre>fe'o	[fanmo]	end of communication	not done
+</pre><lx "fe'o">  XE "fe'o"   =93=93Over and out, X=94; indicates complet=
ion of statement(s) and communication directed at the identified person(s).=
 Used to terminate a letter if a signature is not required because the send=
er has already been identified (as in memos). The negative form means =93Wa=
it, hold it, we're not done!=94 and differs from =93mu'onai=94 in that it m=
eans more exchanges are to follow, rather than that the current exchange is=
 incomplete.
+<p>
+<cx "fe'o, contrasted with fa'o">  XE "fe'o: contrasted with fa'o"   <cx "=
fa'o, contrasted with fe'o">  XE "fa'o: contrasted with fe'o"   Do not conf=
use =93fe'o=94 with =93fa'o=94 (selma'o FAhO) which is a mechanical, extra-=
grammatical signal that a text is complete. One may say =93fe'o=94 to one p=
articipant of a multi-way conversation and then go on speaking to the other=
s.
+<p>
+<lx "mi'e">  XE "mi'e"   <pre>mi'e	[cmavo: mi]	self-identification	non-ide=
ntification
+</pre><cx "introduce oneself">  XE "introduce oneself"   <cx "mi'e, contra=
sted with other members of COI">  XE "mi'e: contrasted with other members o=
f COI"   <cx "me, explicitly specifying">  XE "me: explicitly specifying"  =
 =93=93And I am X=94; a generalized self-vocative. Although grammatically j=
ust like the other members of selma'o COI, =93mi'e=94 is quite different se=
mantically. In particular, rather than specifying the listener, the person =
whose name (or description) follows =93mi'e=94 is taken to be the speaker. =
Therefore, using =93mi'e=94 specifies the meaning of the pro-sumti =93mi=94=
. It can be used to introduce oneself, to close letters, or to identify one=
self on the telephone.
+<p>
+<cx "closings, letter">  XE "closings: letter"   <lx "fe=92omi'e">  XE "fe=
=92omi'e"   <lx "re'imi'e">  XE "re'imi'e"   <cx "COI, ordering multiple wi=
th mi'e">  XE "COI selma=92o: ordering multiple with mi'e"   <cx "mi'e, eff=
ect of ordering multiple COI">  XE "mi'e: effect of ordering multiple COI" =
  This cmavo is often combined with other members of COI: =93fe'omi'e=94 wo=
uld be an appropriate closing at the end of a letter; =93re'imi'e=94 would =
be a self-vocative used in delayed responses, as when called to the phone, =
or possibly in a roll-call. As long as the =93mi'e=94 comes last, the follo=
wing name is that of the speaker; if another COI cmavo is last, the followi=
ng name is that of the listener. It is not possible to name both speaker an=
d listener in a single vocative expression, but this fact is of no importan=
ce, because wherever one vocative expression is grammatical, any number of =
consecutive ones may appear.
+<p>
+<lx "mi'enai">  XE "mi'enai"   The negative form denies an identity which =
someone else has attributed to you; =93mi'enai .djan.=94 means that you are=
 saying you are not John.
+<p>
+<cx "protocol, parliamentary using COI">  XE "protocol: parliamentary usin=
g COI"   <cx "protocol, computer communications using COI">  XE "protocol: =
computer communications using COI"   <cx "protocol, using vocatives">  XE "=
protocol: using vocatives"   <lx "ta'apei">  XE "ta'apei"   Many of the voc=
atives been listed with translations which are drawn from radio use: =93rog=
er=94, =93wilco=94, =93over and out=94. This form of translation does not m=
ean that Lojban is a language of CB enthusiasts, but rather that in most na=
tural languages these forms are so well handled by the context that only in=
 specific domains (like speaking on the radio) do they need special words. =
In Lojban, dependence on the context can be dangerous, as speaker and liste=
ner may not share the right context, and so the vocatives provide a formal =
protocol for use when it is appropriate. Other appropriate contexts include=
 computer communications and parliamentary procedure: in the latter context=
, the protocol question =93ta'apei=94 would mean =93Will the speaker yield?=
=94
+<p>
+<a name=3Ds15><h3>A sample dialogue</h3>
+<p>
+The following dialogue in Lojban illustrates the uses of attitudinals and =
protocol vocatives in conversation. The phrases enclosed in =93sei=A0<dots>=
=85</dots>se'u=94 indicate the speaker of each sentence.
+<p>
+<pre><a name=3De15d1>15.1)	la rik. .e la .alis. nerkla le kafybarja
+	Rick and Alice in-go to-the coffee-bar.
+	Rick and Alice go into the coffee bar.
+<a name=3De15d2>15.2)	.i sei la rik. cusku se'u ta'a ro zvati be ti
+		mi baza speni ti .iu
+	[Comment] Rick says, [end-comment] [Interrupt] all at this-place,=20
+		I [future] [medium] am-spouse-to this-one [love].
+	Rick said, =93Sorry to break in, everybody. Pretty soon I'm getting marri=
ed=20
+to my love here.=94
+<a name=3De15d3>15.3)	.i sei la djordj. cusku se'u
+		a'o ko gleki doi ma
+	[Comment] George says, [end-comment]
+		[Hope] [You-imperative] are-happy, O [who?].
+	George said, =93I hope you'll be happy, um,=A0<dots>=85</dots>?=94
+<a name=3De15d4>15.4)	.i sei la pam. cusku se'u pe'u .alis.
+		xu mi ba terfriti le nunspenybi'o
+	[Comment] Pam says, [Please] Alice, [end-comment]
+		[Is it true?] I [future] receive-offer-of the event-of-spouse-becoming?
+	Pam said, =93Please, Alice, am I going to be invited to the wedding?=94
+
+<a name=3De15d5>15.5)	.i sei la mark. cusku se'u
+		coi baza speni
+		a'o le re do lifri le ka xamgu
+	[Comment] Mark says, [end-comment]=20
+		[Greetings] [future] [medium] spouse(s),
+		[Hope] the two of-you experience the-property-of being-happy
+	Mark said, =93Hello, spouses-to-be. I hope both of you will be very happy=
.=94
+<a name=3De15d6>15.6)	.i sei la rik. cusku se'u mi'e .rik. doi terpreti
+	[Comment] Rick says, [end-comment] [I am] Rick, O questioners.
+	Rick said, =93My name is Rick, for those of you who want to know.=94
+<a name=3De15d7>15.7)	.i sei la .alis. cusku se'u
+		nu'e .pam. .o'ero'i do ba zvati
+	[Comment] Alice says, [end-comment]
+		[Promise-to] Pam, [closeness] [emotional] you [future] are-at.
+	Alice said, =93I promise you'll be there, Pam honey.=94
+<a name=3De15d8>15.8)	.i sei la fred. cusku se'u .uinaicairo'i
+		mi ji'a prami la .alis. fe'o .rik.
+	[Comment] Fred says, [end-comment] [Happy] [not] [emphatic] [emotional]
+		I [additionally] love Alice. [Over and out to] Rick.
+	=93I love Alice too,=94 said Fred miserably. =93Have a nice life, Rick.=
=94
+<a name=3De15d9>15.9)	.i la fred. cliva
+	Fred leaves.
+	And he left.
+<a name=3De15d10>15.10)	.i sei la rik. cusku se'u
+		fi'i ro zvati
+		ko pinxe pa ckafi fi'o pleji mi
+	[Comment] Rick says, [end-comment]
+		[Welcome-to] all at-place,
+		[You-imperative] drink one coffee with-payer me.
+	Rick said, raising his voice, =93A cup of coffee for the house,
+on me.=94
+<a name=3De15d11>15.11)	.i sei la pam. cusku se'u be'e selfu
+	[Comment] Pam says, [end-comment] [Request to speak to] server.
+	Pam said, =93Waiter!=94
+<a name=3De15d12>15.12)	.i sei le selfu cu cusku se'u re'i [end-comment]
+	[Comment] The server says, [Ready to receive].
+	The waiter replied, =93May I help you?=94
+<a name=3De15d13>15.13)	.i sei la pam. cusku se'u
+		e'o ko selfu le traji xamgu ckafi
+		le baza speni fi'o selpleji mi
+	[Comment] Pam says, [end-comment]
+		[Petition] [You-imperative] serve the (superlatively good) coffee
+		to-the [future] [medium] spouse with-payment me.
+	Pam said, =93One Jamaica Blue for the lovebirds here, on my tab.=94
+<a name=3De15d14>15.14)	.i sei le selfu cu cusku se'u vi'o
+	[Comment] The server says, [end-comment] [Will comply].
+	=93Gotcha=94, said the waiter.
+<a name=3De15d15>15.15)	.i sei la rik. cusku se'u ki'e .pam.
+	[Comment] Rick says, [end-comment] [Thanks O] Pam.
+	=93Thanks, Pam=94, said Rick.
+<a name=3De15d16>15.16)	.i sei la pam. cusku se'u je'e
+	[Comment] Pam says, [end-comment] [Acknowledge].
+	=93Sure=94, said Pam.
+<a name=3De15d17>15.17)	.i sei la djan. cusku se'u
+		.y. mi .y. mutce spopa .y.
+		le nu le speni si .y. ba speni .y. .y.
+		su .yyyyyy. mu'o
+	[Comment] John says, [end-comment]
+		[Uh] I [uh] very [nonexistent gismu] [uh]
+		the event-of the spouse [erase] [uh] [future] spouse [uh] [uh]
+		[erase all] [uh] [over]
+	John said, =93I, er, a lotta, uh, marriage, upcoming marriage,=A0<dots>=
=85</dots>. Oh, forget it.
+Er, later.=94
+<a name=3De15d18>15.18)	.i sei la djordj. cusku se'u ke'o .djan. zo'o
+	[Comment] George says, [end-comment] [Repeat O] John [humor].
+	=93How's that again, John?=94 said George.
+<a name=3De15d19>15.19)	.i sei la pam. cusku se'u
+		ju'i .djordj. .e'unai le kabri bazi farlu
+	[Comment] Pam says, [end-comment]
+		[Attention] George, [Warning] the cup [future] [short] falls
+	=93George, watch out!=94 said Pam. =93The cup's falling!=94
+<a name=3De15d20>15.20)	.i le kabri cu je'a farlu
+	The cup indeed falls.
+	The cup fell.
+<a name=3De15d21>15.21)	.i sei la djan. cusku se'u
+		e'o doi djordj. zo'o rapygau
+	[Comment] John says, [end-comment]
+		[Petition] O George [humor] repeat-cause.
+	John said, =93Try that again, George!=94
+<a name=3De15d22>15.22)	.i sei la djordj. cusku se'u
+		co'o ro zvati pe secau la djan. ga'i
+	[Comment] George says, [end-comment]
+		[Partings] all at-place without John [superiority]
+	=93Goodbye to all of you,=94 said George sneeringly, =93except John.=94
+<a name=3De15d23>15.23)	.i la djordj. cliva
+	George leaves.
+	George left.
+</pre><a name=3Ds16><h3>Tentative conclusion</h3>
+<p>
+<cx "Kzinti, communication with">  XE "Kzinti: communication with"   <cx "=
aliens, communication with">  XE "aliens: communication with"   <cx "indica=
tors, ramifications">  XE "indicators: ramifications"   The exact ramificat=
ions of the indicator system in actual usage are unknown. There has never b=
een anything like it in natural language before. The system provides great =
potential for emotional expression and transcription, from which significan=
t Sapir-Whorf effects can be anticipated. When communicating across cultura=
l boundaries, where different indicators are often used for the same emotio=
n, accidental offense can be avoided. If we ever ran into an alien race, a =
culturally neutral language of emotion could be vital. (A classic example, =
taken from the science fiction of Larry Niven, is to imagine speaking Lojba=
n to the carnivorous warriors called Kzinti, noting that a human smile bare=
s the teeth, and could be seen as an intent to attack.) And for communicati=
ng emotions to computers, when we cannot identify all of the signals involv=
ed in subliminal human communication (things like body language are also cu=
ltural), a system like this is needed.
+<p>
+<cx "indicators, rationale for selection">  XE "indicators: rationale for =
selection"   We have tried to err on the side of overkill. There are distin=
ctions possible in this system that no one may care to make in any culture.=
 But it was deemed more neutral to overspecify and let usage decide, than t=
o choose a limited set and constrain emotional expression. For circumstance=
s in which even the current indicator set is not enough, it is possible usi=
ng the cmavo =93sei=94, explained in <a href=3Dchap19.html>Chapter 19</a>, =
to create metalinguistic comments that act like indicators.
+<p>
+<cx "indicators, evolutionary development of">  XE "indicators: evolutiona=
ry development of"   We envision an evolutionary development. At this point=
, the system is little more than a mental toy. Many of you who read this wi=
ll try playing around with various combinations of indicators, trying to fi=
gure out what emotions they express and when the expressions might be usefu=
l. You may even find an expression for which there currently is no good Eng=
lish word and start using it. Why not, if it helps you express your feeling=
s?
+<p>
+There will be a couple dozen of these used pretty much universally =96 mos=
tly just simple attitudinals with, at most, intensity markers. These are th=
e ones that will quickly be expressed at the subconscious level. But every =
Lojbanist who plays with the list will bring in a couple of new words. Poet=
s will paint emotional pictures, and people who identify with those picture=
s will use the words so created for their own experiences.
+<p>
+Just as a library of tanru is built up, so will a library of attitudes be =
built. Unlike the tanru, though, the emotional expressions are built on som=
e fairly nebulous root emotions =97 words that cannot be defined with the p=
recision of the gismu. The emotion words of Lojban will very quickly take o=
n a life of their own, and the outline given here will evolve into a true s=
ystem of emotions.
+<p>
+<cx "emotions, recording using indicators">  XE "emotions: recording using=
 indicators"   <cx "emotions, research using indicators">  XE "emotions: re=
search using indicators"   There are several theories as to the nature of e=
motion, and they change from year to year as we learn more about ourselves.=
 Whether or not Lojban's additive/scalar emotional model is an accurate mod=
el for human emotions, it does support the linguistic needs for expressing =
those emotions. Researchers may learn more about the nature of human emotio=
ns by exploring the use of the system by Lojban speakers. They also may be =
able to use the Lojban system as a means for more clearly recording emotion=
s.
+<p>
+<cx "emotions, cultural bias of expression">  XE "emotions: cultural bias =
of expression"   The full list of scales and attitudes will probably not be=
 used until someone speaks the language from birth. Until then, people will=
 use the attitudes that are important to them. In this way, we counter cult=
ural bias =97 if a culture is prone to recognizing and/or expressing certai=
n emotions more than others, its members will use only those out of the eno=
rmous set available. If a culture hides certain emotions, its members simpl=
y won't express them.
+<p>
+<cx "Sapir-Whorf effects, and emotional indicators">  XE "Sapir-Whorf effe=
cts: and emotional indicators"   Perhaps native Lojban speakers will be mor=
e expressively clear about their emotions than others. Perhaps they will fe=
el some emotions more strongly than others in ways that can be correlated w=
ith the word choices; any difference from the norms of other cultures could=
 be significant. Psychologists have devised elaborate tests for measuring a=
ttitudes and personality; this may be the easiest area in which to detect a=
ny systematic cultural effect of the type sought to confirm Sapir-Whorf, si=
mply because we already have tools in existence to test it. Because Lojban =
is unique among languages in having such extensive and expressive indicator=
s, it is likely that a Sapir-Whorf effect will occur and will be recognized=
.
+<p>
+It is unlikely that we will know the true potential of a system like this =
one until and unless we have children raised entirely in a multi-cultural L=
ojban-speaking environment. We learn too many cultural habits in the realm =
of emotional communication =93at our mother's knee=94. Such children will h=
ave a Lojban system that has stronger reinforcement than any typical cultur=
e system. The second generation of such children, then, could be said to be=
 the start of a true Lojbanic culture.
+<p>
+<cx "emotional indicators, noticeable effects of">  XE "emotional indicato=
rs: noticeable effects of"   We shouldn't need to wait that long to detect =
significant effects. Emotion is so basic to our lives that even a small cha=
nge or improvement in emotional communication would have immediately notice=
able effects. Perhaps it will be the case that the most important contribut=
ion of our =93logical language=94 will be in the non-logical realm of emoti=
on!
+
+<p>
+</body></html>
+<img src=3Dchap14.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+
+<h2>Chapter 14
+<br>
+If Wishes Were Horses: The Lojban Connective System</h2>
+<h6>$Revision: 4.1 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>Logical connection and truth tables</h3>
+<p>
+<cx "logical language, truth functions">  XE "logical language: truth func=
tions"   <cx "truth functions">  XE "truth functions"   Lojban is a logical=
 language: the name of the language itself means =93logical language=94. Th=
e fundamentals of ordinary logic (there are variant logics, which aren't ad=
dressed in this book) include the notions of a =93sentence=94 (sometimes ca=
lled a =93statement=94 or =93proposition=94), which asserts a truth or fals=
ehood, and a small set of =93truth functions=94, which combine two sentence=
s to create a new sentence. The truth functions have the special characteri=
stic that the truth value (that is, the truth or falsehood) of the results =
depends only on the truth value of the component sentences. For example,
+<p>
+<ex "man or woman">  XE "man or woman: example=93  <pre><a name=3De1d1>1.1=
)	John is a man or James is a woman.
+</pre>is true if =93John is a man=94 is true, or if =93James is a woman=94=
 is true. If we know whether John is a man, and we know whether James is a =
woman, we know whether =93John is a man or James is a woman=94 is true, pro=
vided we know the meaning of =93or=94. Here =93John is a man=94 and =93Jame=
s is a woman=94 are the component sentences.
+<p>
+<cx "negating a sentence, and truth value">  XE "negating a sentence: and =
truth value"   We will use the phrase =93negating a sentence=94 to mean cha=
nging its truth value. An English sentence may always be negated by prefixi=
ng =93It is false that=A0<dots>=85</dots>=94, or more idiomatically by inse=
rting =93not=94 at the right point, generally before the verb. =93James is =
not a woman=94 is the negation of =93James is a woman=94, and vice versa. R=
ecent slang can also negate a sentence by following it with the exclamation=
 =93Not!=94
+<p>
+<cx "logical connectives">  XE "logical connectives"   Words like =93or=94=
 are called =93logical connectives=94, and Lojban has many of them, as befi=
ts a logical language. This chapter is mostly concerned with explaining the=
 forms and uses of the Lojban logical connectives. There are a number of ot=
her logical connectives in English such as =93and=94, =93and/or=94, =93if=
=94, =93only if=94, =93whether or not=94, and others; however, not every us=
e of these English words corresponds to a logical connective. This point wi=
ll be made clear in particular cases as needed. The other English meanings =
are supported by different Lojban connective constructs.
+<p>
+The Lojban connectives form a system (as the title of this chapter suggest=
s), regular and predictable, whereas natural-language connectives are rathe=
r less systematic and therefore less predictable.
+<p>
+<cx "truth functions, 16 possible">  XE "truth functions: 16 possible"   <=
cx "truth table, explanation">  XE "truth table: explanation"   There exist=
 16 possible different truth functions. A truth table is a graphical device=
 for specifying a truth function, making it clear what the value of the tru=
th function is for every possible value of the component sentences. Here is=
 a truth table for =93or=94:
+<p>
+<pre> 	first	second	result
+	True	True	True
+	True	False	True
+	False	True	True
+	False	False	False
+</pre><p>This table means that if the first sentence stated is true, and t=
he second sentence stated is true, then the result of the truth function is=
 also true. The same is true for every other possible combination of truth =
values except the one where both the first and the second sentences are fal=
se, in which case the truth value of the result is also false.
+<p>
+Suppose that =93John is a man=94 is true (and =93John is not a man=94 is f=
alse), and that =93James is a woman=94 is false (and =93James is not a woma=
n=94 is true). Then the truth table tells us that
+<p>
+<dl compact><dt>	<dd>=93John is a man, or James is not a woman=94 	(true, =
true) is true;
+	=93John is a man, or James is a woman=94		(true, false) is true;
+	=93John is not a man, or James is not a woman=94	(false, true) is true;
+	=93John is not a man, or James is a woman=94		(false, false) is false.
+</dl><cx "or=94, =93and/or=94 contrasted with =93either =85 or =85 but not=
 both">  XE "or\=94: \=93and/or\=94 contrasted with \=93either =85 or =85 b=
ut not both"   Note that the kind of =93or=94 used in this example can also=
 be expressed (in formal English) with =93and/or=94. There is a different t=
ruth table for the kind of =93or=94 that means =93either <dots>=85</dots> o=
r <dots>=85</dots> but not both=94.
+<p>
+<cx "truth tables, abbreviated format">  XE "truth tables: abbreviated for=
mat"   <cx "truth tables, notation convention">  XE "truth tables: notation=
 convention"   To save space, we will write truth tables in a shorter forma=
t henceforth. Let the letters T and F stand for True and False. The rows wi=
ll always be given in the order shown above: TT, TF, FT, FF for the two sen=
tences. Then it is only necessary to give the four letters from the result =
column, which can be written TTTF, as can be seen by reading down the third=
 column of the table above. So TTTF is the abbreviated truth table for the =
=93or=94 truth function. Here are the 16 possible truth functions, with an =
English version of what it means to assert that each function is, in fact, =
true (=93first=94 refers to the first sentence, and =93second=94 to the sec=
ond sentence):
+<p>
+<cx "truth tables, list of 16 in abbreviated form">  XE "truth tables: lis=
t of 16 in abbreviated form"   <pre> 	TTTT	(always true)
+	TTTF	first is true and/or second is true.
+	TTFT	first is true if second is true.
+	TTFF	first is true whether or not second is true.
+	TFTT	first is true only if second is true.
+	TFTF	whether or not first is true, second is true.
+	TFFT	first is true if and only if second is true.
+	TFFF	first is true and second is true
+	FTTT	first and second are not both true.
+	FTTF	first or second is true, but not both.
+	FTFT	whether or not first is true, second is false.
+	FTFF	first is true, but second is false.
+	FFTT	first is false whether or not second is true.
+	FFTF	first is false, but second is true.
+	FFFT	neither first nor second is true.
+	FFFF	(always false)
+</pre><p>Skeptics may work out the detailed truth tables for themselves.
+<p>
+<a name=3Ds2><h3>The four basic vowels</h3>
+<p>
+<cx "truth functions, fundamental 4 in Lojban">  XE "truth functions: fund=
amental 4 in Lojban"   <cx "truth functions, relation to logical connective=
s">  XE "truth functions: relation to logical connectives"   <cx "logical c=
onnectives, relation to truth functions">  XE "logical connectives: relatio=
n to truth functions"   Lojban regards four of these 16 truth functions as =
fundamental, and assigns them the four letters vowels <font face=3D"Brush S=
cript MT, Script">A</font>, <font face=3D"Brush Script MT, Script">E</font>=
, <font face=3D"Brush Script MT, Script">O</font>, and <font face=3D"Brush =
Script MT, Script">U</font>. These letters do not represent actual cmavo or=
 selma'o, but rather a component vowel from which actual logical-connective=
 cmavo are built up, as explained in the next section. Here are the four vo=
wels, their truth tables, and rough English equivalents:
+
+<p>
+<cx "truth tables, for 4 fundamental Lojban truth functions">  XE "truth t=
ables: for 4 fundamental Lojban truth functions"   <cx "fancy A, notation c=
onvention">  XE "fancy A: notation convention"   <cx "fancy E, notation con=
vention">  XE "fancy E: notation convention"   <cx "fancy O, notation conve=
ntion">  XE "fancy O: notation convention"   <cx "fancy U, notation convent=
ion">  XE "fancy U: notation convention"   <pre>	<font face=3D"Brush Script=
 MT, Script">A</font> 	TTTF	or, and/or
+	<font face=3D"Brush Script MT, Script">E</font> 	TFFF	and
+	<font face=3D"Brush Script MT, Script">O</font> 	TFFT	if and only if
+	<font face=3D"Brush Script MT, Script">U</font> 	TTFF	whether or not
+</pre><p>More precisely:
+<p>
+<dl compact><dt><dd><font face=3D"Brush Script MT, Script">	A  </font>is t=
rue if either or both sentences are true;
+	<font face=3D"Brush Script MT, Script">E  </font>is true if both sentence=
s are true, but not otherwise;
+	<font face=3D"Brush Script MT, Script">O  </font>is true if the sentences=
 are both true or both false;
+	<font face=3D"Brush Script MT, Script">U  </font>is true if the first sen=
tence is true, regardless of the truth value of the=20
+		second sentence.
+</dl><cx "truth functions, creating all 16 with Lojban's basic set">  XE "=
truth functions: creating all 16 with Lojban's basic set"   With the four v=
owels, the ability to negate either sentence, and the ability to exchange t=
he sentences, as if their order had been reversed, we can create all of the=
 16 possible truth functions except TTTT and FFFF, which are fairly useless=
 anyway. The following table illustrates how to create each of the 14 remai=
ning truth functions:
+<p>
+<pre> 	TTTF	<font face=3D"Brush Script MT, Script">A</font>
+	TTFT	<font face=3D"Brush Script MT, Script">A  </font>with second sentenc=
e negated
+	TTFF	<font face=3D"Brush Script MT, Script">U</font>
+	TFTT	<font face=3D"Brush Script MT, Script">A  </font>with first sentence=
 negated
+	TFTF	<font face=3D"Brush Script MT, Script">U  </font>with sentences exch=
anged
+	TFFT	<font face=3D"Brush Script MT, Script">O</font>
+	TFFF	<font face=3D"Brush Script MT, Script">E</font>
+	FTTT	<font face=3D"Brush Script MT, Script">A  </font>with both sentences=
 negated
+	FTTF	<font face=3D"Brush Script MT, Script">O  </font>with either first o=
r second negated (not both)
+	FTFT	<font face=3D"Brush Script MT, Script">U  </font>with sentences exch=
anged and then second negated
+	FTFF	<font face=3D"Brush Script MT, Script">E  </font>with second sentenc=
e negated
+	FFTT	<font face=3D"Brush Script MT, Script">U  </font>with first sentence=
 negated
+	FFTF	<font face=3D"Brush Script MT, Script">E  </font>with first sentence=
 negated
+	FFFT	<font face=3D"Brush Script MT, Script">E  </font>with both sentences=
 negated
+</pre><cx "commutative truth functions">  XE "commutative truth functions"=
   <cx "truth functions, commutative">  XE "truth functions: commutative"  =
 Note that exchanging the sentences is only necessary with <font face=3D"Br=
ush Script MT, Script">U</font>. The three other basic truth functions are =
commutative; that is, they mean the same thing regardless of the order of t=
he component sentences. There are other ways of getting some of these truth=
 tables; these just happen to be the methods usually employed.
+<p>
+<a name=3Ds3><h3>The six types of logical connectives</h3>
+<p>
+<cx "logical connectives, rationale for multiple sets in grammar">  XE "lo=
gical connectives: rationale for multiple sets in grammar"   In order to re=
main unambiguous, Lojban cannot have only a single logical connective for e=
ach truth function. There are many places in the grammar of the language wh=
ere logical connection is permitted, and each must have its appropriate set=
 of connectives. If the connective suitable for sumti were used to connect =
selbri, ambiguity would result.=20
+
+Consider the English sentence:
+<p>
+<ex "window">  XE "window: example=93  <pre><a name=3De3d1>3.1)	Mary went =
to the window and=A0<dots>=85</dots>
+</pre>where the last word could be followed by =93the door=94, a noun phra=
se, or by =93saw the horses=94, a sentence with subject omitted, or by =93J=
ohn went to the door=94, a full sentence, or by one of a variety of other E=
nglish grammatical constructions. Lojban cannot tolerate such grammatical l=
ooseness.
+<p>
+<lx "A">  XE =93A selma=92o"   <lx "GA">  XE "GA selma=92o"   <lx "GIhA"> =
 XE "GIhA selma=92o"   <lx "JA">  XE "JA selma=92o"   <cx "logical connecti=
ve selma'o, enumerated">  XE =93logical connectives: selma'o, enumerated"  =
 Instead, there are a total of five different selma'o used for logical conn=
ection: A, GA, GIhA, GUhA, and JA. Each of these includes four cmavo, one b=
ased on each of the four vowels, which is always the last vowel in the cmav=
o. In selma'o A, the vowel is the entire cmavo.
+<p>
+<cx "logical connective cmavo, format for each selma'o">  XE =93logical co=
nnectives: cmavo, format for each selma'o"   Thus, in selma'o A, the cmavo =
for the function <font face=3D"Brush Script MT, Script">A </font>is =93a=94=
. (Do not confuse A, which is a selma'o, with <font face=3D"Brush Script MT=
, Script">A</font>, which is a truth function, or =93a=94, which is a cmavo=
.) Likewise, the cmavo for <font face=3D"Brush Script MT, Script">E </font>=
in selma'o GIhA is =93gi'e=94, and the cmavo for <font face=3D"Brush Script=
 MT, Script">U </font>in selma'o GA is =93gu=94. This systematic regularity=
 makes the cmavo easier to learn.
+<p>
+<cx "compound logical connectives, components">  XE "compound logical conn=
ectives: components"   Obviously, four cmavo are not enough to express the =
14 truth functions explained in <a href=3D#s1>Section 1</a>. Therefore, com=
pound cmavo must be used. These compound cmavo follow a systematic pattern:=
 each has one cmavo from the five logical connection selma'o at its heart, =
and may also contain one or more of the auxiliary cmavo =93se=94, =93na=94,=
 or =93nai=94. Which auxiliaries are used with which logical connection cma=
vo, and with what grammar and meaning, will be explained in the following s=
ections. The uses of each of these auxiliary cmavo relates to its other use=
s in other parts of Lojban grammar.
+<p>
+<cx "compound logical connectives, naming convention">  XE "compound logic=
al connectives: naming convention"   <cx "-ek, in name for logical connecti=
ves">  XE "-ek: in name for logical connectives"   <cx "ek, definition">  X=
E "ek: definition"   <cx "jek, definition">  XE "jek: definition"   <lx "A"=
>  XE =93A selma=92o"   <lx "JA">  XE "JA selma=92o"   For convenience, eac=
h of the types of compound cmavo used for logical connection is designated =
by a Lojban name. The name is derived by changing the final =93-A=94 of the=
 selma'o name to =93-ek=94; the reasons for using =93-ek=94 are buried deep=
 in the history of the Loglan Project. Thus, compound cmavo based on selma'=
o A are known as eks, and those based on selma'o JA are known as jeks. (Whe=
n writing in English, it is conventional to use =93eks=94 as the plural of =
=93ek=94.) When the term =93logical connective=94 is used in this chapter, =
it refers to one or more of these kinds of compound cmavo.
+<p>
+<cx "ijek, definition">  XE "ijek: definition"   <lx "JA">  XE "JA selma=
=92o"   <lx "I">  XE "I selma=92o"   Why does the title of this section ref=
er to =93six types=94 when there are only five selma'o? A jek may be preced=
ed by =93.i=94, the usual Lojban cmavo for connecting two sentences. The co=
mpound produced by =93.i=94 followed by a jek is known as an ijek. It is us=
eful to think of ijeks as a sixth kind of logical connective, parallel to e=
ks, jeks, geks, giheks, and guheks.
+<p>
+<cx "gik, as name for compound cmavo">  XE "gik: as name for compound cmav=
o"   <cx "joik, as name for compound cmavo">  XE "joik: as name for compoun=
d cmavo"   <cx "joigik, as name for compound cmavo">  XE "joigik: as name f=
or compound cmavo"   <cx "ijoik, as name for compound cmavo">  XE "ijoik: a=
s name for compound cmavo"   <lx "GI">  XE "GI selma=92o"   <lx "JOI">  XE =
"JOI selma=92o"   <lx "I">  XE "I selma=92o"   There also exist giks, joiks=
, ijoiks, and joigiks, which are not logical connectives, but are other kin=
ds of compound cmavo which will be introduced later.
+<p>
+<a name=3Ds4><h3>Logical connection of bridi</h3>
+<p>
+<lx "JA">  XE "JA selma=92o"   <lx "I">  XE "I selma=92o"   <cx "ijek logi=
cal connective, connecting bridi">  XE "ijek logical connective(s): connect=
ing bridi"   <cx "bridi, logical connective for">  XE "bridi: logical conne=
ctive for"   Now we are ready to express <a href=3D#e1d1>Example 1.1 </a>in=
 Lojban! The kind of logical connective which is placed between two Lojban =
bridi to connect them logically is an ijek:
+<p>
+<pre><a name=3De4d1>4.1)	la djan. nanmu  .ija la djeimyz. ninmu
+	John is-a-man or James is-a-woman.
+</pre><p>Here we have two separate Lojban bridi, =93la djan. nanmu=94 and =
=93la djeimyz. ninmu=94. These bridi are connected by =93.ija=94, the ijek =
for the truth function <font face=3D"Brush Script MT, Script">A</font>. The=
 =93.i=94 portion of the ijek tells us that we are dealing with separate se=
ntences here. Similarly, we can now say:
+
+<p>
+<pre><a name=3De4d2>4.2)	la djan. nanmu  .ije la djeimyz. ninmu
+	John is-a-man and James is-a-woman.
+<a name=3De4d3>4.3)	la djan. nanmu  .ijo la djeimyz. ninmu
+	John is-a-man if-and-only-if James is-a-woman.
+<a name=3De4d4>4.4)	la djan. nanmu  .iju la djeimyz. ninmu
+	John is-a-man whether-or-not James is-a-woman.
+</pre><cx "bridi, logical connection with negation">  XE "bridi: logical c=
onnection with negation"   To obtain the other truth tables listed in <a hr=
ef=3D#s2>Section 2</a>, we need to know how to negate the two bridi which r=
epresent the component sentences. We could negate them directly by insertin=
g =93na=94 before the selbri, but Lojban also allows us to place the negati=
on within the connective itself.
+<p>
+<lx "JA">  XE "JA selma=92o"   <lx "I">  XE "I selma=92o"   To negate the =
first or left-hand bridi, prefix =93na=94 to the JA cmavo but after the =93=
.i=94. To negate the second or right-hand bridi, suffix =93-nai=94 to the J=
A cmavo. In either case, the negating word is placed on the side of the con=
nective that is closest to the bridi being negated.
+<p>
+So to express the truth table FTTF, which requires <font face=3D"Brush Scr=
ipt MT, Script">O </font>with either of the two bridi negated (not both), w=
e can say either:
+<p>
+<pre><a name=3De4d5>4.5)	la djan. nanmu  .inajo la djeimyz. ninmu
+	John is-not-a-man if-and-only-if James is-a-woman.
+<a name=3De4d6>4.6)	la djan. nanmu  .ijonai la djeimyz. ninmu
+	John is a man if-and-only-if James is-not-a-woman
+</pre><p>The meaning of both <a href=3D#e4d5>Example 4.5 </a>and <a href=
=3D#e4d6>Example 4.6 </a>is the same as that of:
+<p>
+<pre><a name=3De4d7>4.7)	John is a man or James is a woman, but not both.
+</pre><p>Here is another example:
+<p>
+<pre><a name=3De4d8>4.8)	la djan. nanmu  .ijanai la djeimyz. ninmu
+	John is-a-man or James is-not-a-woman.
+	John is a man if James is a woman.
+</pre><cx "if, meaning in logical connections">  XE "if: meaning in logica=
l connections"   <cx "if, English usage contrasted with Lojban logical conn=
ective">  XE "if: English usage contrasted with Lojban logical connective" =
  How's that again? Are those two English sentences in <a href=3D#e4d8>Exam=
ple 4.8 </a>really equivalent? In English, no. The Lojban TTFT truth functi=
on can be glossed =93A if B=94, but the =93if=94 does not quite have its En=
glish sense. <a href=3D#e4d8>Example 4.8 </a>is true so long as John is a m=
an, even if James is not a woman; likewise, it is true just because James i=
s not a woman, regardless of John's gender. This kind of =93if-then=94 is t=
echnically known as a =93material conditional=94.
+<p>
+Since James is not a woman (by our assertions in <a href=3D#s1>Section 1</=
a>), the English sentence =93John is a man if James is a woman=94 seems to =
be neither true nor false, since it assumes something which is not true. It=
 turns out to be most convenient to treat this =93if=94 as TTFT, which on i=
nvestigation means that <a href=3D#e4d8>Example 4.8 </a>is true. <a href=3D=
#e4d9>Example 4.9</a>, however, is equally true:
+<p>
+<pre><a name=3De4d9>4.9)	la djan. ninmu  .ijanai la djeimyz. ninmu
+	John is a woman if James is a woman.
+</pre><cx "principle of consistency, of logical-if statements">  XE "princ=
iple of consistency: of logical-if statements"   <cx "false statement, impl=
ications of">  XE "false statement: implications of"   This can be thought =
of as a principle of consistency, and may be paraphrased as follows: =93If =
a false statement is true, any statement follows from it.=94 All uses of En=
glish =93if=94 must be considered very carefully when translating into Lojb=
an to see if they really fit this Lojban mold.
+<p>
+<cx "if =85 then, compared with only if">  XE "if =85 then: compared with =
only if"   <cx "only if, compared with if =85 then">  XE "only if: compared=
 with if =85 then"   <a href=3D#e4d10>Example 4.10</a>, which uses the TFTT=
 truth function, is subject to the same rules: the stated gloss of TFTT as =
=93only if=94 works naturally only when the right-hand bridi is true; if it=
 is false, the left-hand bridi may be either true or false. The last gloss =
of <a href=3D#e4d10>Example 4.10 </a>illustrates the use of =93if=A0<dots>=
=85</dots> then=94 as a more natural substitute for =93only if=94.
+<p>
+<pre><a name=3De4d10>4.10)	la djan. nanmu  .inaja la djeimyz. ninmu
+	John is-not-a-man or James is-a-woman.
+	John is a man only if James is a woman.
+	If John is a man, then James is a woman.
+</pre><lx "se">  XE "se"   <cx "se, in logical connective to exchange sent=
ences">  XE "se: in logical connective to exchange sentences"   The followi=
ng example illustrates the use of =93se=94 to, in effect, exchange the two =
sentences. The normal use of =93se=94 is to (in effect) transpose places of=
 a bridi, as explained in <a href=3Dchap5.html>Chapter 5</a>.
+<p>
+<pre><a name=3De4d11>4.11)	la djan. nanmu  .iseju la djeimyz. ninmu
+	Whether or not John is a man, James is a woman.
+</pre><lx "na">  XE "na"   <lx "nai">  XE "nai"   <lx "se">  XE "se"   <cx=
 "se, order in logical connectives with na">  XE "se: order in logical conn=
ectives with na"   <cx "na, order in logical connectives with se">  XE "na:=
 order in logical connectives with se"   If both =93na=94 and =93se=94 are =
present, which is legal but never necessary, =93na=94 would come before =93=
se=94.
+<p>
+<cx "ijeks, syntax of">  XE "ijeks: syntax of"   <lx "I">  XE "I selma=92o=
"   <lx "JA">  XE "JA selma=92o"   The full syntax of ijeks, therefore, is:
+<p>
+<dl compact><dt><dd>.i [na] [se] JA [nai]
+</dl>where the cmavo in brackets are optional.
+<p>
+<a name=3Ds5><h3>Forethought bridi connection</h3>
+<p>
+<cx "afterthought, contrasted with forethought for connectives">  XE "afte=
rthought connective(s): contrasted with forethought connective"   <cx "fore=
thought, contrasted with afterthought for connectives">  XE "forethought co=
nnective(s): contrasted with afterthought connective"   Many concepts in Lo=
jban are expressible in two different ways, generally referred to as =93aft=
erthought=94 and =93forethought=94. <a href=3D#s4>Section 4 </a>discussed w=
hat is called =93afterthought bridi logical connection=94. The word =93afte=
rthought=94 is used because the connective cmavo and the second bridi were =
added, as it were, afterwards and without changing the form of the first br=
idi. This form might be used by someone who makes a statement and then wish=
es to add or qualify that statement after it has been completed. Thus,
+<p>
+<pre><a name=3De5d1>5.1)	la djan. nanmu
+</pre>is a complete bridi, and adding an afterthought connection to make
+<p>
+<pre><a name=3De5d2>5.2)	la djan. nanmu  .ija la djeimyz. ninmu
+	John is a man or James is a woman (or both)
+</pre>provides additional information without requiring any change in the =
form of what has come before; changes may not be possible or practical, esp=
ecially in speaking. (The meaning, however, may be changed by the use of a =
negating connective.) Afterthought connectives make it possible to construc=
t all the important truth-functional relationships in a variety of ways.
+<p>
+<cx "gek, definition">  XE "gek: definition"   <cx "ijek bridi connectives=
, contrasted with geks">  XE "ijek bridi connectives: contrasted with geks"=
   <cx "gek bridi connectives, contrasted with ijeks">  XE "gek bridi conne=
ctives: contrasted with ijeks"   <cx "forethought bridi connectives, contra=
sted with afterthought bridi connectives">  XE "forethought bridi connectiv=
es: contrasted with afterthought bridi connectives"   <cx "afterthought bri=
di connectives, contrasted with forethought bridi connectives">  XE "aftert=
hought bridi connectives: contrasted with forethought bridi connectives"   =
<lx "GA">  XE "GA selma=92o"   <lx "JA">  XE "JA selma=92o"   <lx "I">  XE =
"I selma=92o"   In forethought style the speaker decides in advance, before=
 expressing the first bridi, that a logical connection will be expressed. F=
orethought and afterthought connectives are expressed with separate selma'o=
. The forethought logical connectives corresponding to afterthought ijeks a=
re geks:
+<p>
+<pre><a name=3De5d3>5.3)	ga la djan. nanmu gi la djeimyz. ninmu
+	Either John is a man or James is a woman (or both).
+</pre><lx "GA">  XE "GA selma=92o"   <lx "GI">  XE "GI selma=92o"   =93ga=
=94 is the cmavo which represents the <font face=3D"Brush Script MT, Script=
">A </font>truth function in selma'o GA. The word =93gi=94 does not belong =
to GA at all, but constitutes its own selma'o: it serves only to separate t=
he two bridi without having any content of its own. The English translation=
 of =93ga=A0<dots>=85</dots>gi=94 is =93either=A0<dots>=85</dots> or=94, bu=
t in the English form the truth function is specified both by the word =93e=
ither=94 and by the word =93or=94: not so in Lojban.
+<p>
+<cx "forethought bridi connection, as grammatically one sentence">  XE "fo=
rethought bridi connection: as grammatically one sentence"   <cx "i, regard=
ing forethought bridi connection">  XE "i: regarding forethought bridi conn=
ection"   <lx "I">  XE "I selma=92o"   Even though two bridi are being conn=
ected, geks and giks do not have any =93.i=94 in them. The forethought cons=
truct binds up the two bridi into a single sentence as far as the grammar i=
s concerned.
+<p>
+Some more examples of forethought bridi connection are:
+<p>
+<pre><a name=3De5d4>5.4)	ge la djan. nanmu gi la djeimyz. ninmu
+	(It is true that) both John is a man and James is a woman.
+</pre><lx "gu">  XE "gu"   <lx "GU">  XE "GU selma=92o"   <pre><a name=3De=
5d5>5.5)	gu la djan. nanmu gi la djeimyz. ninmu
+	It is true that John is a man, whether or not James is a woman.
+</pre><p>It is worth emphasizing that <a href=3D#e5d5>Example 5.5 </a>does=
 not assert that James is (or is not) a woman. The =93gu=94 which indicates=
 that =93la djeimyz. ninmu=94 may be true or false is unfortunately rather =
remote from the bridi thus affected.
+<p>
+Perhaps the most important of the truth functions commonly expressed in fo=
rethought is TFTT, which can be paraphrased as =93if=A0<dots>=85 </dots>the=
n=A0<dots>=85</dots>=94:
+<p>
+<lx "ganai">  XE "ganai"   <lx "GA">  XE "GA selma=92o"   <pre><a name=3De=
5d6>5.6)	ganai la djan. nanmu gi la djeimyz. ninmu
+	Either John is not a man, or James is a woman.
+	If John is a man, then James is a woman.
+</pre><cx "nai, placement in forethought bridi connection contrasted with =
afterthought">  XE "nai: placement in forethought bridi connection contrast=
ed with afterthought"   <cx "nai, placement in afterthought bridi connectio=
n contrasted with forethought">  XE "nai: placement in afterthought bridi c=
onnection contrasted with forethought"   Note the placement of the =93nai=
=94 in <a href=3D#e5d6>Example 5.6</a>. When added to afterthought selma'o =
such as JA, a following =93nai=94 negates the second bridi, to which it is =
adjacent. Since GA cmavo precede the first bridi, a following =93nai=94 neg=
ates the first bridi instead.
+<p>
+<cx "logical connectives, negated first sentence as a potential problem fo=
r understanding">  XE "logical connectives: negated first sentence as a pot=
ential problem for understanding"   Why does English insist on forethought =
in the translation of <a href=3D#e5d6>Example 5.6</a>? Possibly because it =
would be confusing to seemingly assert a sentence and then make it conditio=
nal (which, as the Lojban form shows, involves a negation). Truth functions=
 which involve negating the first sentence may be confusing, even to the Lo=
jbanic understanding, when expressed using afterthought.
+<p>
+<cx "if =85 then, logical connectives contrasted with other translations">=
  XE "if =85 then: logical connectives contrasted with other translations" =
  It must be reiterated here that not every use of English =93if <dots>=85<=
/dots> then=94 is properly translated by =93.inaja=94 or =93ganai <dots>=85=
</dots>gi=94; anything with implications of time needs a somewhat different=
 Lojban translation, which will be discussed in <a href=3D#s18>Section 18</=
a>. Causal sentences like =93If you feed the pig, then it will grow=94 are =
not logical connectives of any type, but rather need a translation using =
=93rinka=94 as the selbri joining two event abstractions, thus:
+<p>
+<pre><a name=3De5d7>5.7)	le nu do cidja dunda fi le xarju cu rinka le nu r=
i ba banro
+	The event-of (you food-give to the pig) causes the event-of (it will grow=
).
+</pre><p>Causality is discussed in far more detail in <a href=3Dchap9.html=
>Chapter 9</a>.
+<p>
+<a href=3D#e5d8>Example 5.8 </a>and <a href=3D#e5d9>Example 5.9 </a>illust=
rates a truth function, FTTF, which needs to negate either the first or the=
 second bridi. We already understand how to negate the first bridi:
+<p>
+<pre><a name=3De5d8>5.8)	gonai la djan. nanmu gi la djeimyz. ninmu
+	John is-not-a-man if-and-only-if James is-a-woman,
+	Either John is a man or James is a woman but not both.
+
+</pre><lx "ginai">  XE "ginai"   <lx "GI">  XE "GI selma=92o"   How can th=
e second bridi be negated? By adding =93-nai=94 to the =93gi=94.
+<p>
+<pre><a name=3De5d9>5.9)	go la djan. nanmu ginai la djeimyz. ninmu
+	John is-a-man if-and-only-if James is-not-a-woman.
+	Either John is a man or James is a woman but not both.
+</pre><cx "gik, definition">  XE "gik: definition"   <lx "GI">  XE "GI sel=
ma=92o"   A compound cmavo based on =93gi=94 is called a gik; the only giks=
 are =93gi=94 itself and =93ginai=94.
+<p>
+Further examples:
+<p>
+<pre><a name=3De5d10>5.10)	ge la djan. nanmu ginai la djeimyz. ninmu
+	John is-a-man and James is-not-a-woman.
+</pre><lx "ganai">  XE "ganai"   <lx "GA">  XE "GA selma=92o"   <pre><a na=
me=3De5d11>5.11)	ganai la djan. nanmu ginai la djeimyz. ninmu
+	John is-not-a-man or James is-not-a-woman.
+</pre><lx "se">  XE "se"   <lx "nai">  XE "nai"   <lx "GA">  XE "GA selma=
=92o"   <cx "geks, syntax of">  XE "geks: syntax of"   The syntax of geks i=
s:
+<p>
+<dl compact><dt><dd>[se] GA [nai]
+</dl><lx "gi">  XE "gi"   <lx "GI">  XE "GI selma=92o"   <lx "nai">  XE "n=
ai"   <cx "giks, syntax of">  XE "giks: syntax of"   and of giks (which are=
 not themselves connectives, but part of the machinery of forethought conne=
ction) is:
+<p>
+<dl compact><dt><dd>gi [nai]
+</dl><a name=3Ds6><h3>sumti connection</h3>
+<p>
+<cx "sumti logical connection">  XE "sumti logical connection"   <cx "sumt=
i logical connection, rationale for">  XE "sumti logical connection: ration=
ale for"   <cx "sumti logical connection, compared with bridi logical conne=
ctions">  XE "sumti logical connection: compared with bridi logical connect=
ions"   <cx "bridi logical connection, compared with sumti logical connecti=
ons">  XE "bridi logical connection: compared with sumti logical connection=
s"   Geks and ijeks are sufficient to state every possible logical connecti=
on between two bridi. However, it is often the case that two bridi to be lo=
gically connected have one or more portions in common:
+<p>
+<pre><a name=3De6d1>6.1)	la djan. klama le zarci  .ije la .alis. klama le =
zarci
+	John goes to the market, and Alice goes to the market.
+</pre><p>Here only a single sumti differs between the two bridi. Lojban do=
es not require that both bridi be expressed in full. Instead, a single brid=
i can be given which contains both of the different sumti and uses a logica=
l connective from a different selma'o to combine the two sumti:
+<p>
+<lx ".e">  XE "e"   <lx "A">  XE =93A selma=92o"   <pre><a name=3De6d2>6.2=
)	la djan .e la .alis. klama le zarci
+	John and Alice go-to the market.
+</pre><cx "logical connection, transformation between forms">  XE "logical=
 connection: transformation between forms"   <a href=3D#e6d2>Example 6.2 </=
a>means exactly the same thing as <a href=3D#e6d1>Example 6.1</a>: one may =
be rigorously transformed into the other without any change of logical mean=
ing. This rule is true in general for every different kind of logical conne=
ction in Lojban; all of them, with one exception (see <a href=3D#s12>Sectio=
n 12</a>), can always be transformed into a logical connection between sent=
ences that expresses the same truth function.
+<p>
+<cx "sumti connection, afterthought">  XE "sumti connection: afterthought"=
   <cx "eks, in sumti forethought logical connection">  XE "eks: in sumti f=
orethought logical connection"   <lx "A">  XE =93A selma=92o"   The afterth=
ought logical connectives between sumti are eks, which contain a connective=
 cmavo of selma'o A. If ijeks were used in <a href=3D#e6d2>Example 6.2</a>,=
 the meaning would be changed:
+<p>
+<pre><a name=3De6d3>6.3)	la djan.  .ije la .alis. klama le zarci
+	John [is/does something].  And Alices goes-to the market.
+</pre>leaving the reader uncertain why John is mentioned at all.
+
+<p>
+Any ek may be used between sumti, even if there is no direct English equiv=
alent:
+<p>
+<pre><a name=3De6d4>6.4)	la djan. .o la .alis. klama le zarci
+	John if-and-only-if Alice goes-to the market.
+	John goes to the market if, and only if, Alice does.
+</pre><p>The second line of <a href=3D#e6d3>Example 6.3 </a>is highly stil=
ted English, but the first line (of which it is a literal translation) is e=
xcellent Lojban.
+<p>
+<cx "sumti connection, forethought">  XE "sumti connection: forethought"  =
 <cx "geks, in forethought sumti connection">  XE "geks: in forethought sum=
ti connection"   <lx "GA">  XE "GA selma=92o"   What about forethought sumt=
i connection? As is the case for bridi connection, geks are appropriate. Th=
ey are not the only selma'o of forethought logical-connectives, but are the=
 most commonly used ones.
+<p>
+<pre><a name=3De6d5>6.5)	ga la djan. gi la .alis. klama le zarci
+	Either John or Alice (or both) goes-to the market.
+</pre><cx "na writing convention, in eks">  XE "na writing convention: in =
eks"   <cx "se writing convention, in eks">  XE "se writing convention: in =
eks"   <lx "A">  XE =93A selma=92o"   Of course, eks include all the same p=
atterns of compound cmavo that ijeks do. When =93na=94 or =93se=94 is part =
of an ek, a special writing convention is invoked, as in the following exam=
ple:
+<p>
+<lx "na.a">  XE "na.a"   <lx "A">  XE =93A selma=92o"   <pre><a name=3De6d=
6>6.6)	la djan. na.a la .alis. klama le zarci
+	John only if Alice goes-to the market.
+	John goes to the market only if Alice does.
+</pre><p>Note the period in =93na.a=94. The cmavo of A begin with vowels, =
and therefore must always be preceded by a pause. It is conventional to wri=
te all connective compounds as single words (with no spaces), but this paus=
e must still be marked in writing as in speech; otherwise, the =93na=94 and=
 =93a=94 would tend to run together.
+<p>
+<a name=3Ds7><h3>More than two propositions</h3>
+<p>
+<cx "logical connectives, more than 2 sentences">  XE "logical connectives=
: more than 2 sentences"   So far we have seen logical connectives used to =
connect exactly two sentences. How about connecting three or more? Is this =
possible in Lojban? The answer is yes, subject to some warnings and some re=
strictions.
+<p>
+<cx "logical connectives, associative">  XE "logical connectives: associat=
ive"   Of the four primitive truth functions <font face=3D"Brush Script MT,=
 Script">A</font>, <font face=3D"Brush Script MT, Script">E</font>, <font f=
ace=3D"Brush Script MT, Script">O</font>, and <font face=3D"Brush Script MT=
, Script">U</font>, all but <font face=3D"Brush Script MT, Script">O </font=
>have the same truth values no matter how their component sentences are ass=
ociated in pairs. Therefore,
+<p>
+<pre><a name=3De7d1>7.1)	mi dotco  .ije mi ricfu  .ije mi nanmu
+	I am-German.  And I am-rich.  And I am-a-man.
+</pre>means that all three component sentences are true. Likewise,
+<p>
+<pre><a name=3De7d2>7.2)	mi dotco .ija mi ricfu .ija mi nanmu
+	I am-German. Or I am-rich. Or I am-a-man.
+</pre>means that one or more of the component sentences is true.
+<p>
+<cx "logical connectives, non-associative"><cx "logical connectives, equiv=
alence relation on 3 sentences"><font face=3D"Brush Script MT, Script">O</f=
ont>, however, is different. Working out the truth table for
+<p>
+<pre><a name=3De7d3>7.3)	mi dotco  .ijo mi ricfu  .ijo mi nanmu
+	I am-German. If-and-only-if I am-rich.  If-and-only-if I am-a-man.
+</pre>shows that <a href=3D#e7d3>Example 7.3 </a>does not mean that either=
 I am all three of these things or none of them; instead, an accurate trans=
lation would be:
+<p>
+<dl compact><dt><dd>Of the three properties =97 German-ness, wealth, and m=
anhood =97 I possess either
+ exctly one or else all three.
+</dl><cx "logical connection of more than 2 sentences, things to avoid">  =
XE =93logical connection: of more than 2 sentences, things to avoid"   <cx =
"logical connection, negation in connecting more than 2 sentences">  XE "lo=
gical connection: negation in connecting more than 2 sentences"   Because o=
f the counterintuitiveness of this outcome, it is safest to avoid <font fac=
e=3D"Brush Script MT, Script">O </font>with more than two sentences. Likewi=
se, the connectives which involve negation also have unexpected truth value=
s when used with more than two sentences.
+<p>
+<cx "logical connection of more than 2 sentences, all or none">  XE =93log=
ical connection: of more than 2 sentences, all or none"   In fact, no combi=
nation of logical connectives can produce the =93all or none=94 interpretat=
ion intended (but not achieved) by <a href=3D#e7d3>Example 7.3 </a>without =
repeating one of the bridi. See <a href=3D#e8d11>Example 8.11</a>.
+<p>
+There is an additional difficulty with the use of more than two sentences.=
 What is the meaning of:
+<p>
+<cx "logical connection of more than 2 sentences, mixed =93and=94 and =93o=
r=94">  XE =93logical connection: of more than 2 sentences, mixed \=93and\=
=94 and \=93or\=94"   <pre><a name=3De7d4>7.4)	mi nelci la djan.  .ije mi n=
elci la martas.  .ija mi nelci la meris.
+	I like John.  And I like Martha.  Or I like Mary.
+</pre><p>Does this mean:
+<p>
+<pre><a name=3De7d5>7.5)	I like John, and I like either Martha or Mary or =
both.
+</pre><p>Or is the correct translation:
+<p>
+<pre><a name=3De7d6>7.6)	Either I like John and I like Martha, or I like M=
ary, or both.
+</pre><cx "logical connectives, pairing from left">  XE "logical connectiv=
es: pairing from left"   <cx "logical connection of more than 2 sentences, =
forethought">  XE =93logical connection: of more than 2 sentences, forethou=
ght"   <a href=3D#e7d6>Example 7.6 </a>is the correct translation of <a hre=
f=3D#e7d4>Example 7.4</a>. The reason is that Lojban logical connectives pa=
ir off from the left, like many constructs in the language. This rule, call=
ed the left-grouping rule, is easy to forget, especially when intuition pul=
ls the other way. Forethought connectives are not subject to this problem:
+<p>
+<pre><a name=3De7d7>7.7)	ga ge mi nelci la djan. gi mi nelci la martas. gi=
 mi nelci la meris.
+	Either (Both I like John and I like Martha) or I like Mary.
+</pre>is equivalent in meaning to <a href=3D#e7d4>Example 7.4</a>, whereas
+<p>
+<pre><a name=3De7d8>7.8)	ge mi nelci la djan. gi ga mi nelci la martas. gi=
 mi nelci la meris.
+	Both I like John and (Either I like Martha or I like Mary).
+</pre>is not equivalent to <a href=3D#e7d4>Example 7.4</a>, but is instead=
 a valid translation into Lojban, using forethought, of <a href=3D#e7d5>Exa=
mple 7.5</a>.
+<p>
+<a name=3Ds8><h3>Grouping of afterthought connectives</h3>
+<p>
+<lx "bo">  XE "bo"   <lx "BO">  XE "BO selma=92o"   <cx "logical connectiv=
es, grouping with bo">  XE "logical connectives: grouping with bo"   <cx "l=
ogical connection with bo, precedence">  XE =93logical connection: with bo,=
 precedence"   <cx "bo, in logical connectives">  XE "bo: in logical connec=
tives"   There are several ways in Lojban to render <a href=3D#e7d5>Example=
 7.5 </a>using afterthought only. The simplest method is to make use of the=
 cmavo =93bo=94 (of selma'o BO). This cmavo has several functions in Lojban=
, but is always associated with high precedence and short scope. In particu=
lar, if =93bo=94 is placed after an ijek, the result is a grammatically dis=
tinct kind of ijek which overrides the regular left-grouping rule. Connecti=
ons marked with =93bo=94 are interpreted before connections not so marked. =
<a href=3D#e8d1>Example 8.1 </a>is equivalent in meaning to <a href=3D#e7d8=
>Example 7.8</a>:
+<p>
+<pre><a name=3De8d1>8.1)	mi nelci la djan.  .ije mi nelci la martas.  .ija=
bo mi nelci la meris.
+	I like John, and I like Martha or I like Mary.
+</pre><p>The English translation feebly indicates with a comma what the Lo=
jban marks far more clearly: the =93I like Martha=94 and =93I like Mary=94 =
sentences are joined by =93.ija=94 first, before the result is joined to =
=93I like John=94 by =93.ije=94. <dl compact><dt><dd>Eks can have =93bo=94 =
attached in exactly the same way, so that <a href=3D#e8d2>Example 8.2 </a>i=
s equivalent in meaning to Example 8.l:
+<p><dt><a name=3De8d2>8.2)	<dd>mi nelci la djan. .e la martas. .abo la mer=
is.
+</dl><cx "forethought connectives and bo">  XE "forethought connectives an=
d bo"   <cx "bo and forethought connectives">  XE "bo and forethought conne=
ctives"   Forethought connectives, however, never can be suffixed with =93b=
o=94, for every use of forethought connectives clearly indicates the intend=
ed pattern of grouping.
+<p>
+What happens if =93bo=94 is used on both connectives, giving them the same=
 high precedence, as in <a href=3D#e8d3>Example 8.3</a>?
+<p>
+<pre><a name=3De8d3>8.3)	mi nelci la djan. .ebo la martas. .abo la meris.
+</pre><cx "bo, right-grouping">  XE "bo: right-grouping"   <cx "logical co=
nnectives, right-grouping with bo">  XE "logical connectives: right-groupin=
g with bo"   <lx "BO">  XE "BO selma=92o"   Does this wind up meaning the s=
ame as <a href=3D#e7d4>Example 7.4 </a>and <a href=3D#e7d6>Example 7.6</a>?=
 Not at all. A second rule relating to =93bo=94 is that where several =93bo=
=94-marked connectives are used in succession, the normal Lojban left-group=
ing rule is replaced by a right-grouping rule. As a result, <a href=3D#e8d3=
>Example 8.3 </a>in fact means the same as <a href=3D#e8d{1}>Examples 8.1 <=
/a>and <a href=3D#e8d2>8.2</a>. This rule may be occasionally exploited for=
 special effects, but is tricky to keep straight; in writing intended to be=
 easy to understand, multiple consecutive connectives marked with =93bo=94 =
should be avoided.
+<p>
+<cx "complex logical connection, grouping strategies contrasted">  XE "com=
plex logical connection: grouping strategies contrasted"   <cx "logical con=
nection, grouping strategies for complex cases contrasted">  XE "logical co=
nnection: grouping strategies for complex cases contrasted"   <lx "tu'e">  =
XE "tu'e"   <lx "tu'u">  XE "tu'u"   <lx "ke">  XE "ke"   <lx "ke'e">  XE "=
ke'e"   <lx "TUhE">  XE "TUhE selma=92o"   <lx "TUhU">  XE "TUhU selma=92o"=
   <lx "KE">  XE "KE selma=92o"   <lx "KEhE">  XE "KEhE selma=92o"   <cx "c=
omplex logical connectives, grouping with parentheses">  XE "complex logica=
l connectives: grouping with parenthesis"   The use of =93bo=94, therefore,=
 gets tricky in complex connections of more than three sentences. Look back=
 at the English translations of <a href=3D#e7d{7}>Examples 7.7 </a>and <a h=
ref=3D#e7d8>7.8</a>: parentheses were used to clarify the grouping. These p=
arentheses have their Lojban equivalents, two sets of them actually. =93tu'=
e=94 and =93tu'u=94 are used with ijeks, and =93ke=94 and =93ke'e=94 with e=
ks and other connectives to be discussed later. (=93ke=94 and =93ke'e=94 ar=
e also used in other roles in the language, but always as grouping markers)=
. Consider the English sentence:
+<p>
+<pre><a name=3De8d4>8.4)	I kiss you and you kiss me, if I love you and you=
 love me.
+</pre><cx "complex logical connectives, grouping with bo">  XE "complex lo=
gical connectives: grouping with bo"   <lx "BO">  XE "BO selma=92o"   <lx "=
bo">  XE "bo"   where the semantics tells us that the instances of =93and=
=94 are meant to have higher precedence than that of =93if=94. If we wish t=
o express <a href=3D#e8d4>Example 8.4 </a>in afterthought, we can say:
+<p>
+<pre><a name=3De8d5>8.5)	mi cinba do .ije[bo] do cinba mi  .ijanai mi pram=
i do .ijebo do prami mi
+	I kiss you and you kiss me, if I love you and you love me.
+</pre>marking two of the ijeks with =93bo=94 for high precedence. (The fir=
st =93bo=94 is not strictly necessary, because of the left-grouping rule, a=
nd is shown here in brackets.)
+<p>
+<cx "complex logical connectives, grouping with parentheses">  XE "complex=
 logical connectives: grouping with parenthesis"   But it may be clearer to=
 use explicit parenthesis words and say:
+<p>
+<pre><a name=3De8d6>8.6)	tu'e mi cinba do .ije do cinba mi tu'u =20
+		.ijanai tu'e mi prami do  .ije do prami mi [tu'u]
+	( I kiss you and you kiss me )
+		if ( I love you and you love me).
+</pre>where the =93tu'e=A0<dots>=85</dots>tu'u=94 pairs set off the struct=
ure. The cmavo =93tu'u=94 is an elidable terminator, and its second occurre=
nce in <a href=3D#e8d6>Example 8.6 </a>is bracketed, because all terminator=
s may be elided at the end of a text.
+<p>
+In addition, parentheses are a general solution: multiple parentheses may =
be nested inside one another, and additional afterthought material may be a=
dded without upsetting the existing structure. Neither of these two advanta=
ges apply to =93bo=94 grouping. In general, afterthought constructions trad=
e generality for simplicity.
+<p>
+Because of the left-grouping rule, the first set of =93tu'e=A0<dots>=85</d=
ots>tu'u=94 parentheses may actually be left off altogether, producing:
+<p>
+<pre><a name=3De8d7>8.7)	mi cinba do .ije do cinba mi
+		.ijanai tu'e mi prami do  .ije do prami mi [tu'u]
+	I kiss you and you kiss me
+		if ( I love you and you love me ).
+
+</pre><p>What about parenthesized sumti connection? Consider
+<p>
+<pre><a name=3De8d8>8.8)	I walk to either the market and the house, or the=
 school and the office.
+</pre><cx "logical connection of sumti, grouping with parentheses">  XE =
=93logical connection: of sumti, grouping with parenthesis"   <cx "logical =
connection of sumti, restriction on ke">  XE =93logical connection: of sumt=
i, restriction on ke"   <cx "sumti, beginning with "ke">  XE "sumti: beginn=
ing with \"ke"   <lx "ke">  XE "ke"   <lx "KE">  XE "KE selma=92o"   Two pa=
irs of parentheses, analogous to <a href=3D#e8d6>Example 8.6</a>, would see=
m to be the right approach. However, it is a rule of Lojban grammar that a =
sumti may not begin with =93ke=94, so the first set of parentheses must be =
omitted, producing <a href=3D#e8d10>Example 8.10</a>, which is instead para=
llel to <a href=3D#e8d7>Example 8.7</a>:
+<p>
+<pre><a name=3De8d9>8.9)	mi dzukla le zarci .e le zdani .a ke le ckule .e =
le briju [ke'e]
+	I walk-to the market and the house or ( the school and the office ).
+</pre><cx "ke in sumti grouping, where allowed">  XE "ke in sumti grouping=
: where allowed"   <lx "ke">  XE "ke"   <lx "KE">  XE "KE selma=92o"   If s=
umti were allowed to begin with =93ke=94, unavoidable ambiguities would res=
ult, so =93ke=94 grouping of sumti is allowed only just after a logical con=
nective. This rule does not apply to =93tu'e=94 grouping of bridi, as <a hr=
ef=3D#e8d6>Example 8.6 </a>shows.
+<p>
+<ex "German rich man">  XE "German rich man: example=93  Now we have enoug=
h facilities to handle the problem of <a href=3D#e7d3>Example 7.3</a>: =93I=
 am German, rich, and a man =97 or else none of these.=94 The following par=
aphrase has the correct meaning:
+<p>
+<pre><a name=3De8d10>8.10)	 [tu'e] mi dotco .ijo mi ricfu [tu'u]
+		.ije tu'e mi dotco  .ijo mi nanmu [tu'u]
+	( I am-German if-and-only-if I am-rich )
+		and ( I am-German if-and-only-if I am-a-man ).
+</pre><p>The truth table, when worked out, produces T if and only if all t=
hree component sentences are true or all three are false.
+<p>
+<a name=3Ds9><h3>Compound bridi</h3>
+<p>
+<cx "logical connection of selbri">  XE =93logical connection: of selbri" =
  So far we have seen how to handle two sentences that need have no similar=
ity at all (bridi connection) and sentences that are identical except for a=
 difference in one sumti (sumti connection). It would seem natural to ask h=
ow to logically connect sentences that are identical except for having diff=
erent selbri.
+<p>
+<cx "compound bridi, definition">  XE "compound bridi: definition"   Surpr=
ise! Lojban provides no logical connective that is designed to handle selbr=
i and nothing else. Instead, selbri connection is provided as part of a mor=
e general-purpose mechanism called =93compound bridi=94. Compound bridi res=
ult from logically connecting sentences that differ in their selbri and pos=
sibly some of their sumti.
+<p>
+<cx "compound bridi, one sumti in common">  XE "compound bridi: one sumti =
in common"   The simplest cases result when the x1 sumti is the only common=
 point:
+<p>
+<pre><a name=3De9d1>9.1)	mi klama le zarci  .ije mi nelci la djan.
+	I go to the market, and I like John.
+</pre>is equivalent in meaning to the compound bridi:
+<p>
+<pre><a name=3De9d2>9.2)	mi klama le zarci gi'e nelci la djan.
+	I go-to the market and like John.
+</pre><cx "gihek, definition">  XE "gihek: definition"   <lx "gi'e">  XE "=
gi'e"   <cx "bridi-tail, definition">  XE "bridi-tail: definition"   <cx "c=
ompound bridi, logical connection of">  XE "compound bridi: logical connect=
ion of"   <lx "GIhA">  XE "GIhA selma=92o"   As <a href=3D#e9d2>Example 9.2=
 </a>indicates, giheks are used in afterthought to create compound bridi; =
=93gi'e=94 is the gihek corresponding to =93and=94. The actual phrases =93k=
lama le zarci=94 and =93nelci la djan.=94 that the gihek connects are known=
 as =93bridi-tails=94, because they represent (in this use) the =93tail end=
=94 of a bridi, including the selbri and any following sumti, but excluding=
 any sumti that precede the selbri:
+
+<p>
+<pre><a name=3De9d3>9.3)	mi ricfu gi'e klama le zarci
+	I am-rich and go-to the market.
+</pre><p>In <a href=3D#e9d3>Example 9.3</a>, the first bridi-tail is =93ri=
cfu=94, a simple selbri, and the second bridi-tail is =93klama le zarci=94,=
 a selbri with one following sumti.
+<p>
+<cx "compound bridi, more than one sumti in common">  XE "compound bridi: =
more than one sumti in common"   Suppose that more than a single sumti is i=
dentical between the two sentences:
+<p>
+<pre><a name=3De9d4>9.4)	mi dunda le cukta do  .ije mi lebna lo rupnu do
+	I give the book to-you, and I take some currency-units from-you.
+</pre><cx "compound bridi with more than one sumti in common, with common =
sumti first">  XE "compound bridi with more than one sumti in common: with =
common sumti first"   In <a href=3D#e9d4>Example 9.4</a>, the first and las=
t sumti of each bridi are identical; the selbri and the second sumti are di=
fferent. By moving the final sumti to the beginning, a form analogous to <a=
 href=3D#e9d2>Example 9.2 </a>can be achieved:
+<p>
+<pre><a name=3De9d5>9.5)	fi do fa mi dunda le cukta gi'e lebna lo rupnu
+	To/from you I give the book and take some currency-units.
+</pre><cx "logical connectives, bridi-tail connection">  XE "logical conne=
ctives: bridi-tail connection"   <cx "tail-terms, definition">  XE "tail-te=
rms: definition"   where the =93fi=94 does not have an exact English transl=
ation because it merely places =93do=94 in the third place of both =93lebna=
=94 and =93dunda=94. However, a form that preserves natural sumti order als=
o exists in Lojban. Giheks connect two bridi-tails, but also allow sumti to=
 be added following the bridi-tail. These sumti are known as tail-terms, an=
d apply to both bridi. The straightforward gihek version of <a href=3D#e9d4=
>Example 9.4 </a>therefore is:
+<p>
+<lx "vau">  XE "vau"   <lx "VAU">  XE "VAU selma=92o"   <cx "compound brid=
i with more than one sumti in common, with vau">  XE "compound bridi with m=
ore than one sumti in common: with vau"   <pre><a name=3De9d6>9.6)	mi dunda=
 le cukta gi'e lebna lo rupnu vau do
+	I (give the book) and (take some currency-units) to/from you.
+</pre><cx "bridi-tails, eliding vau in">  XE "bridi-tails: eliding vau in"=
   The =93vau=94 (of selma'o VAU) serves to separate the bridi-tail from th=
e tail-terms. Every bridi-tail is terminated by an elidable =93vau=94, but =
only in connection with compound bridi is it ever necessary to express this=
 =93vau=94. Thus:
+<p>
+<pre><a name=3De9d7>9.7)	mi klama le zarci [vau]
+	I go-to the market.
+</pre>has a single elided =93vau=94, and <a href=3D#e9d2>Example 9.2 </a>i=
s equivalent to:
+<p>
+<pre><a name=3De9d8>9.8)	mi klama le zarci [vau] gi'e nelci la djan. [vau]=
 [vau]
+</pre>where the double =93vau=94 at the end of <a href=3D#e9d8>Example 9.8=
 </a>terminates both the right-hand bridi-tail and the unexpressed tail-ter=
ms.
+<p>
+<cx "logical connectives, observative sentence connection">  XE "logical c=
onnectives: observative sentence connection"   A final use of giheks is to =
combine bridi-tails used as complete sentences, the Lojban observative:
+<p>
+<pre><a name=3De9d9>9.9)	klama le zarci gi'e dzukla le briju
+	A goer to-the market and a walker to-the office.
+</pre><cx "logical connection of observatives, relation of first places"> =
 XE =93logical connection: of observatives, relation of first places"   Sin=
ce x1 is omitted in both of the bridi underlying <a href=3D#e9d9>Example 9.=
9</a>, this compound bridi does not necessarily imply that the goer and the=
 walker are the same. Only the presence of an explicit x1 (other than =93zo=
'e=94, which is equivalent to omission) can force the goer and the walker t=
o be identical.
+<p>
+<cx "relation of first places in logical connection of observatives, ratio=
nale">  XE "relation of first places in logical connection of observatives:=
 rationale"   A strong argument for this convention is provided by analysis=
 of the following example:
+<p>
+<pre><a name=3De9d10>9.10)	 klama la nu,IORK. la finyks. gi'e klama la nu,=
IORK. la rom.
+	A goer to-New York from-Phoenix and a goer to-New York from-Rome.
+</pre><p>If the rule were that the x1 places of the two underlying bridi w=
ere identified, then (since there is nothing special about x1), the unspeci=
fied x4 (route) and x5 (means) places would also have to be the same, leadi=
ng to the absurd result that the route from Phoenix to New York is the same=
 as the route from Rome to New York. Inserting =93da=94, meaning roughly =
=93something=94, into the x1 place cures the problem:
+<p>
+<pre><a name=3De9d11>9.11)	da klama la nu,IORK. la finyks.
+		gi'e klama la nu,IORK. la rom.
+	Something is-a-goer to-New York from-Phoenix
+		and is-a-goer to-New York from-Rome.
+</pre><cx "giheks, syntax of">  XE "giheks: syntax of"   <lx "na">  XE "na=
"   <lx "se">  XE "se"   <lx "nai">  XE "nai"   <lx "GIhA">  XE "GIhA selma=
=92o"   The syntax of giheks is:
+<p>
+<dl compact><dt><dd>[na] [se] GIhA [nai]
+</dl>which is exactly parallel to the syntax of eks.
+<p>
+<a name=3Ds10><h3>Multiple compound bridi</h3>
+<p>
+<cx "compound bridi, multiple with bo">  XE "compound bridi: multiple with=
 bo"   Giheks can be combined with =93bo=94 in the same way as eks:
+<p>
+<pre><a name=3De10d1>10.1)	mi nelci la djan. gi'e nelci la martas. gi'abo =
nelci la meris.
+	I like John and ( like Martha or like Mary ).
+</pre><cx "compound bridi, multiple with ke=85ke'e">  XE "compound bridi: =
multiple with ke=85ke'e"   is equivalent in meaning to <a href=3D#e8d1>Exam=
ple 8.1 </a>and <a href=3D#e8d2>Example 8.2</a>. Likewise, =93ke<dots>=85</=
dots>ke'e=94 grouping can be used after giheks:
+<p>
+<pre><a name=3De10d2>10.2)	mi dzukla le zarci gi'e dzukla le zdani
+		gi'a ke dzukla le ckule gi'e dzukla le briju [ke'e]
+	I walk-to the market and walk-to the house,
+		or walk-to the school and walk-to the office.
+</pre><cx "multiple compound bridi, restriction on ke">  XE "multiple comp=
ound bridi: restriction on ke"   <cx "logical connection of bridi-tails, re=
striction on ke">  XE =93logical connection: of bridi-tails, restriction on=
 ke"   <lx "ke">  XE "ke"   <lx "KEhE">  XE "KEhE selma=92o"   is the gihek=
 version of <a href=3D#e8d9>Example 8.9</a>. The same rule about using =93k=
e=A0<dots>=85</dots>ke'e=94 bracketing only just after a connective applies=
 to bridi-tails as to sumti, so the first two bridi-tails in <a href=3D#e10=
d2>Example 10.2 </a>cannot be explicitly grouped; implicit left-grouping su=
ffices to associate them.
+<p>
+<cx "compound bridi, separate tail-terms for bridi-tails">  XE "compound b=
ridi: separate tail-terms for bridi-tails"   Each of the pairs of bridi-tai=
ls joined by multiple giheks can have its own set of tail-terms:
+<p>
+<pre><a name=3De10d3>10.3)	mi dejni lo rupnu la djan.  .inaja mi dunda le =
cukta la djan.
+		.ijabo mi lebna le cukta la djan.
+	[If] I owe some currency-units to John, then I give the book to John
+		or I take the book from John.
+</pre>is equivalent in meaning to:
+<p>
+<ex "owe money">  XE "owe money: example=93  <pre><a name=3De10d4>10.4)	mi=
 dejni lo rupnu nagi'a dunda gi'abo lebna vau le cukta vau la djan.
+	[If] I owe some currency-units then (give or take) 	a book to/from John.
+</pre><p>The literal English translation in <a href=3D#e10d4>Example 10.4 =
</a>is almost unintelligible, but the Lojban is perfectly grammatical. =93m=
i=94 fills the x1 place of all three selbri; =93lo rupnu=94 is the x2 of =
=93dejni=94, whereas =93le cukta=94 is a tail-term shared between =93dunda=
=94 and =93lebna=94; =93la djan.=94 is a tail-term shared by =93dejni=94 an=
d by =93dunda gi'abo lebna=94. In this case, greater clarity is probably ac=
hieved by moving =93la djan.=94 to the beginning of the sentence, as in <a =
href=3D#e9d5>Example 10.5</a>:
+<p>
+<pre><a name=3De10d5>10.5)	fi la djan. fa mi dejni lo rupnu nagi'a dunda g=
i'abo lebna vau le cukta
+	To/from John, [if] I owe some currency-units then [I] give or take the bo=
ok.
+</pre><cx "logical connection of bridi-tails, forethought">  XE =93logical=
 connection: of bridi-tails, forethought"   Finally, what about forethought=
 logical connection of bridi-tails? There is no direct mechanism for the pu=
rpose. Instead, Lojban grammar allows a pair of forethought-connected sente=
nces to function as a single bridi-tail, and of course the sentences need n=
ot have terms before their selbri. For example:
+<p>
+<pre><a name=3De10d6>10.6)	mi ge klama le zarci gi nelci la djan.
+	I both go to the market and like John.
+</pre>is equivalent in meaning to <a href=3D#e9d2>Example 9.2</a>.
+<p>
+Of course, either of the connected sentences may contain giheks:
+<p>
+<pre><a name=3De10d7>10.7)	mi ge klama le zarci gi'e dzukla le zdani gi ne=
lci la djan.
+	I both ( go to the market and walk to the house ) and like John.
+</pre><cx "negating a forethought-connected sentence pair">  XE "negating =
a forethought-connected sentence pair"   <cx "negating a forethought-connec=
ted bridi-tail pair">  XE "negating a forethought-connected bridi-tail pair=
"   The entire gek-connected sentence pair may be negated as a whole by pre=
fixing =93na=94:
+<p>
+<pre><a name=3De10d8>10.8)	mi na ge klama le zarci gi dzukla le zdani
+	[False!] I both go to the market and walk to the house.
+</pre><cx "compound bridi, separate tail-terms for forethought-connected b=
ridi-tails">  XE "compound bridi: separate tail-terms for forethought-conne=
cted bridi-tails"   Since a pair of sentences joined by geks is the equival=
ent of a bridi-tail, it may be followed by tail terms. The forethought equi=
valent of <a href=3D#e9d6>Example 9.6 </a>is:
+<p>
+<pre><a name=3De10d9>10.9)	mi ge dunda le cukta  gi lebna lo rupnu vau do
+	I both ( give the book ) and ( take some currency-units ) to/from you.
+</pre><cx "forethought connection, observatives">  XE "forethought connect=
ion: observatives"   Here is a pair of gek-connected observatives, a foreth=
ought equivalent of <a href=3D#e9d9>Example 9.9</a>:
+<p>
+<pre><a name=3De10d10>10.10)	ge klama le zarci gi dzukla le briju
+	Both a goer to-the market and a walker to-the office.
+</pre><p>Finally, here is an example of gek-connected sentences with both =
shared and unshared terms before their selbri:
+<p>
+<ex "office or ice-dance">  XE "office or ice-dance: example=93  <pre><a n=
ame=3De10d11>10.11)	mi gonai le zarci cu klama gi le bisli cu dansu
+	I either-but-not-both to-the office go or on-the ice dance.
+	I either go to the office or dance on the ice (but not both).
+</pre><a name=3Ds11><h3>Termset logical connection</h3>
+<cx "termset logical connection, contrasted with bridi connection">  XE "t=
ermset logical connection: contrasted with bridi connection"   <cx "termset=
 logical connection, contrasted with sumti connection">  XE "termset logica=
l connection: contrasted with sumti connection"   <cx "termset logical conn=
ection, contrasted with bridi-tail connection">  XE "termset logical connec=
tion: contrasted with bridi-tail connection"   <cx "termset logical connect=
ion, when used">  XE "termset logical connection: when used"   So far we ha=
ve seen sentences that differ in all components, and require bridi connecti=
on; sentences that differ in one sumti only, and permit sumti connection; a=
nd sentences that differ in the selbri and possibly one or more sumti, and =
permit bridi-tail connection. Termset logical connectives are employed for =
sentences that differ in more than one sumti but not in the selbri, such as=
:
+<p>
+<pre><a name=3De11d1>11.1)	I go to the market from the office and to the h=
ouse from the school.
+</pre><lx "ce'e">  XE "ce'e"   <lx "CEhE">  XE "CEhE selma=92o"   <lx "pe'=
e">  XE "pe'e"   <lx "PEhE">  XE "PEhE selma=92o"   <cx "term, definition">=
  XE "term: definition"   <cx "termset, formation">  XE "termset: formation=
"   <cx "logical connection, termsets">  XE "logical connection: termsets" =
  The Lojban version of <a href=3D#e11d1>Example 11.1 </a>requires two term=
sets joined by a logical connective. A =93term=94 is either a sumti or a su=
mti preceded by a tense or modal tag such as =93pu=94 or =93bai=94. Afterth=
ought termsets are formed by linking terms together by inserting the cmavo =
=93ce'e=94 (of selma'o CEhE) between each of them. Furthermore, the logical=
 connective (which is a jek) must be prefixed by the cmavo =93pe'e=94 (of s=
elma'o PEhE). (We could refer to the combination of =93pe'e=94 and a jek as=
 a =93pehejek=94, I suppose.)
+<p>
+<cx "to the market from the office">  XE "to the market from the office"  =
 <pre><a name=3De11d2>11.2)	mi klama le zarci ce'e le briju pe'e je
+		le zdani ce'e le ckule
+	I go to-the market [plus] from-the office [joint] and
+		to-the house [plus] from-the school.
+</pre><p>The literal translation uses =93[plus]=94 to indicate the termset=
 connective, and =93[joint]=94 to indicate the position of the logical conn=
ective joint. As usual, there is an equivalent bridi-connection form:
+<p>
+<pre><a name=3De11d3>11.3)	mi klama le zarci le briju  .ije mi klama le zd=
ani le ckule
+	I go to-the market from-the office, and I go to-the house from-the school=
.
+</pre>which illustrates that the two bridi differ in the x2 and x3 places =
only.
+<p>
+<cx "termset logical connection, unequal length">  XE "termset logical con=
nection: unequal length"   What happens if the two joined sets of terms are=
 of unequal length? Expanding to bridi connection will always make clear wh=
ich term goes in which place of which bridi. It can happen that a sumti may=
 fall in the x2 place of one bridi and the x3 place of another:
+<p>
+<pre><a name=3De11d4>11.4)	mi pe'e ja do ce'e le zarci cu klama le briju
+	I [joint] or you to-the market [plus] go to/from-the office.
+</pre>can be clearly understood by expansion to:
+<p>
+<pre><a name=3De11d5>11.5)	mi klama le briju  .ija do le zarci cu klama le=
 briju
+	I go to-the office, or you to-the market go from-the office.
+</pre><cx "unequal termset connection, compared with compound bridi connec=
tion with unequal separate bridi-tails">  XE "unequal termset connection: c=
ompared with compound bridi connection with unequal separate bridi-tails"  =
 So =93le briju=94 is your origin but my destination, and thus falls in the=
 x2 and x3 places of =93klama=94 simultaneously! This is legal because even=
 though there is only one selbri, =93klama=94, there are two distinct bridi=
 expressed here. In addition, =93mi=94 in <a href=3D#e11d4>Example 11.4 </a=
>is serving as a termset containing only one term. An analogous paradox app=
lies to compound bridi with tail-terms and unequal numbers of sumti within =
the connected bridi-tails:
+<p>
+<pre><a name=3De11d6>11.6)	mi klama le zarci gi'e dzukla vau le briju
+	I ( go to-the market and walk ) to/from-the office.
+</pre>means that I go to the market from the office, and I walk to the off=
ice; =93le briju=94 is the x3 place of =93klama=94 and the x2 place of =93d=
zukla=94.
+<p>
+<lx "nu'i">  XE "nu'i"   <lx "nu'u">  XE "nu'u"   <lx "NUhI">  XE "NUhI se=
lma=92o"   <lx "NUhU">  XE "NUhU selma=92o"   <cx "logical connection of fo=
rethought termsets">  XE =93logical connection: of forethought termsets"   =
<cx "forethought termsets, logical connection of">  XE "forethought termset=
s: logical connection of"   Forethought termsets also exist, and use =93nu'=
i=94 of selma'o NUhI to signal the beginning and =93nu'u=94 of selma'o NUhU=
 (an elidable terminator) to signal the end. Nothing is inserted between th=
e individual terms: they simply sit side-by-side. To make a logical connect=
ion in a forethought termset, use a gek, with the gek just after the =93nu'=
i=94, and an extra =93nu'u=94 just before the gik:
+<p>
+<pre><a name=3De11d7>11.7)	mi klama nu'i ge le zarci le briju
+		nu'u gi le zdani le ckule [nu'u]
+	I go [start termset] both to-the market from-the office
+		[joint] and to-the house from-the school [end termset].
+</pre><p>Note that even though two termsets are being connected, only one =
=93nu'i=94 is used.
+<p>
+The grammatical uses of termsets that do not contain logical connectives a=
re explained in <a href=3Dchap12.html>Chapter 12 </a>and <a href=3Dchap16.h=
tml>Chapter 16</a>.
+<p>
+<a name=3Ds12><h3>Logical connection within tanru</h3>
+<p>
+<cx "logical connectives in tanru">  XE "logical connectives in tanru"   A=
s noted at the beginning of <a href=3D#s9>Section 9</a>, there is no logica=
l connective in Lojban that joins selbri and nothing but selbri. However, i=
t is possible to have logical connectives within a selbri, forming a kind o=
f tanru that involves a logical connection. Consider the simple tanru =93bl=
anu zdani=94, blue house. Now anything that is a blue ball, in the most ord=
inary understanding of the phrase at least, is both blue and a ball. And in=
deed, instead of =93blanu bolci=94, Lojbanists can say =93blanu je bolci=94=
, using a jek connective within the tanru. (We saw jeks used in <a href=3D#=
s11>Section 11 </a>also, but there they were always prefixed by =93pe'e=94;=
 in this section they are used alone.) Here is a pair of examples:
+<p>
+<ex "blue house">  XE "blue house: example=93  <pre><a name=3De12d1>12.1)	=
ti blanu zdani
+	This is-a-blue type-of house.
+<a name=3De12d2>12.2)	ti blanu je zdani
+	This is-blue and a-house.
+</pre><cx "logical connection in tanru, contrasted with unconnected versio=
n">  XE =93logical connection: in tanru, contrasted with unconnected versio=
n"   <cx "unconnected tanru, contrasted with logically connected version"> =
 XE "unconnected tanru: contrasted with logically connected version"   But =
of course <a href=3D#e12d1>Example 12.1 </a>and <a href=3D#e12d2>Example 12=
.2 </a>are not necessarily equivalent in meaning! It is the most elementary=
 point about Lojban tanru that <a href=3D#e12d1>Example 12.1 </a>might just=
 as well mean
+<p>
+<pre><a name=3De12d3>12.3)	This is a house for blue inhabitants.
+</pre>and <a href=3D#e12d2>Example 12.2 </a>certainly is not equivalent in=
 meaning to <a href=3D#e12d3>Example 12.3</a>.
+<p>
+<cx "logical connection in tanru, expandability of">  XE =93logical connec=
tion: in tanru, expandability of"   A full explanation of logical connectio=
n within tanru belongs rather to a discussion of selbri structure than to l=
ogical connectives in general. Why? Because although <a href=3D#e12d2>Examp=
le 12.2 </a>happens to mean the same as
+<p>
+<pre><a name=3De12d4>12.4)	ti blanu gi'e zdani
+</pre>and therefore as
+<p>
+<pre><a name=3De12d5>12.5)	ti blanu  .ije ti zdani
+</pre>the rule of expansion into separate bridi simply does not always wor=
k for tanru connection. Supposing Alice to be a person who lives in blue ho=
uses, then
+<p>
+<pre><a name=3De12d6>12.6)	la .alis. cu blanu je zdani prenu
+	Alice is-a ( blue and house ) type-of-person.
+</pre>would be true, because tanru grouping with a jek has higher preceden=
ce than unmarked tanru grouping, but:
+<p>
+<pre><a name=3De12d7>12.7)	la .alis. cu blanu prenu  .ije la .alis. cu zda=
ni prenu
+	Alice is-a blue person, and Alice is-a house person.
+</pre>is probably false, because the blueness is associated with the house=
, not with Alice, even leaving aside the question of what it means to say =
=93Alice is a blue person=94. (Perhaps she belongs to the Blue team, or is =
wearing blue clothes.) The semantic ambiguity of tanru make such logical ma=
nipulations impossible.
+<p>
+<lx "bo">  XE "bo"   <lx "BO">  XE "BO selma=92o"   <cx "logical connectio=
n in tanru, grouping with bo">  XE =93logical connection: in tanru, groupin=
g with bo"   It suffices to note here, then, a few purely grammatical point=
s about tanru logical connection. =93bo=94 may be appended to jeks as to ek=
s, with the same rules:
+<p>
+<pre><a name=3De12d8>12.8)	la teris. cu ricfu je nakni jabo fetsi
+	Terry is rich and ( male or female ).
+</pre><cx "logical connection in tanru, grouping with ke">  XE =93logical =
connection: in tanru, grouping with ke"   <lx "ke">  XE "ke"   <lx "KE">  X=
E "KE selma=92o"   The components of tanru may be grouped with =93ke=94 bot=
h before and after a logical connective:
+<p>
+<pre><a name=3De12d9>12.9)	la .teris. cu [ke] ricfu ja pindi [ke'e] je ke =
nakni ja fetsi [ke'e]
+	Terry is (rich or poor) and (male or female).
+</pre>where the first =93ke=A0<dots>=85</dots>ke'e=94 pair may be omitted =
altogether by the rule of left-grouping, but is optionally permitted. In an=
y case, the last instance of =93ke'e=94 may be elided.
+<p>
+<lx "na">  XE "na"   <lx "se">  XE "se"   <lx "nai">  XE "nai"   <lx "JA">=
  XE "JA selma=92o"   <cx "jeks, syntax of">  XE "jeks: syntax of"   The sy=
ntax of jeks is:
+<p>
+<dl compact><dt><dd>[na] [se] JA [nai]
+</dl>parallel to eks and giheks.
+<p>
+<cx "forethought tanru connection">  XE "forethought tanru connection"   <=
cx "guhek, definition">  XE "guhek: definition"   <lx "GUhA">  XE "GUhA sel=
ma=92o"   Forethought tanru connection does not use geks, but uses guheks i=
nstead. Guheks have exactly the same form as geks:
+<p>
+<lx "se">  XE "se"   <lx "nai">  XE "nai"   <lx "GUhA">  XE "GUhA selma=92=
o"   <cx "guheks, syntax of">  XE "guheks: syntax of"   <dl compact><dt><dd=
>[se] GUhA [nai]
+</dl><cx "guheks for tanru connection, rationale">  XE "guheks for tanru c=
onnection: rationale"   <cx "logical connection, of bridi-tail as opposed t=
o tanru">  XE "logical connection: of bridi-tail as opposed to tanru"   <cx=
 "logical connection, of tanru as opposed to bridi-tail">  XE "logical conn=
ection: of tanru as opposed to bridi-tail"   Using guheks in tanru connecti=
on (rather than geks) resolves what would otherwise be an unacceptable ambi=
guity between bridi-tail and tanru connection:
+<p>
+<pre><a name=3De12d10>12.10)	la .alis. gu'e ricfu gi fetsi
+	Alice is both rich and female.
+</pre><cx "tanru grouping, guheks compared with jeks">  XE "tanru grouping=
: guheks compared with jeks"   <cx "tanru connection grouping, guheks unmar=
ked tanru">  XE "tanru connection grouping: guheks unmarked tanru"   <lx "G=
UhA">  XE "GUhA selma=92o"   <lx "JA">  XE "JA selma=92o"   Note that giks =
are used with guheks in exactly the same way they are used with geks. Like =
jeks, guheks bind more closely than unmarked tanru grouping does:
+<p>
+<pre><a name=3De12d11>12.11)	la .alis. gu'e blanu gi zdani prenu
+	Alice is-a-(both blue and a-house) type-of-person.
+</pre>is the forethought version of <a href=3D#e12d6>Example 12.6</a>.
+<p>
+<cx "logical connection of tanru, caveat">  XE =93logical connection: of t=
anru, caveat"   <cx "tanru, reducing logically connected sumti to, caveat">=
  XE "tanru: reducing logically connected sumti to, caveat"   <cx "tanru lo=
gical connection, contrasted with sumti logical connection">  XE "tanru log=
ical connection: contrasted with sumti logical connection"   <cx "sumti log=
ical connection, contrasted with tanru logical connection">  XE "sumti logi=
cal connection: contrasted with tanru logical connection"   A word of cauti=
on about the use of logically connected tanru within descriptions. English-=
based intuition can lead the speaker astray. In correctly reducing
+<p>
+<pre><a name=3De12d12>12.12)	mi viska pa nanmu  .ije mi viska pa ninmu
+	I see a man, and I see a woman.
+</pre>to
+<p>
+<pre><a name=3De12d13>12.13)	mi viska pa nanmu .e pa ninmu
+	I see a man and a woman.
+</pre>there is a great temptation to reduce further to:
+<p>
+<ex "man-woman">  XE "man-woman: example=93  <pre><a name=3De12d14>12.14)	=
mi viska pa nanmu je ninmu
+	I see a man and woman.
+</pre><p>But <a href=3D#e12d14>Example 12.14 </a>means that you see one th=
ing which is both a man and a woman simultaneously! A =93nanmu je ninmu=94 =
is a manwoman, a presumably non-existent creature who is both a =93nanmu=94=
 and a =93ninmu=94.
+<p>
+<a name=3Ds13><h3>Truth questions and connective questions</h3>
+<p>
+So far we have addressed only sentences which are statements. Lojban, like=
 all human languages, needs also to deal with sentences which are questions=
. There are many ways of asking questions in Lojban, but some of these (lik=
e questions about quantity, tense, and emotion) are discussed in other chap=
ters.
+<p>
+<cx "truth questions, simple">  XE "truth questions: simple"   The simples=
t kind of question is of the type =93Is it true that=A0<dots>=85</dots>=94 =
where some statement follows. This type is called a =93truth question=94, a=
nd can be represented in English by <a href=3D#e13d1>Example 13.1</a>:
+<p>
+<pre><a name=3De13d1>13.1)	Is it true that Fido is a dog?
+	Is Fido a dog?
+</pre><lx "xu">  XE "xu"   <lx "UI">  XE "UI selma=92o"   Note the two for=
mulations. English truth questions can always be formed by prefixing =93Is =
is true that=94 to the beginning of a statement; there is also usually a mo=
re idiomatic way involving putting the verb before its subject. =93Is Alice=
 a dog?=94 is the truth question corresponding to =93Alice is a dog=94. In =
Lojban, the equivalent mechanism is to prefix the cmavo =93xu=94 (of selma'=
o UI) to the statement:
+<p>
+<pre><a name=3De13d2>13.2)	xu la faidon. gerku
+	is-it-true-that Fido is-a-dog?
+</pre><a href=3D#e13d1>Example 13.1 </a>and <a href=3D#e13d2>Example 13.2 =
</a>are equivalent in meaning.
+<p>
+<cx "truth questions, as yes-or-no questions">  XE "truth questions: as ye=
s-or-no questions"   <cx "truth questions, answering =93yes=94">  XE "truth=
 questions: answering \=93yes\=94"   <cx "truth questions, answering =93no=
=94">  XE "truth questions: answering \=93no\=94"   A truth question can be=
 answered =93yes=94 or =93no=94, depending on the truth or falsity, respect=
ively, of the underlying statement. The standard way of saying =93yes=94 in=
 Lojban is =93go'i=94 and of saying =93no=94 is =93nago'i=94. (The reasons =
for this rule are explained in <a href=3Dchap7.html>Chapter 7</a>.) In answ=
er to <a href=3D#e13d2>Example 13.2</a>, the possible answers are:
+<p>
+<pre><a name=3De13d3>13.3)	go'i
+	Fido is a dog.
+</pre>and
+<p>
+<pre><a name=3De13d4>13.4)	nago'i
+	Fido is not a dog.
+</pre><cx "truth questions, contrasted with connection questions">  XE "tr=
uth questions: contrasted with connection questions"   Some English questio=
ns seemingly have the same form as the truth questions so far discussed. Co=
nsider
+<p>
+<ex "dog or cat">  XE "dog or cat: example=93  <pre><a name=3De13d5>13.5)	=
Is Fido a dog or a cat?
+</pre><p>Superficially, <a href=3D#e13d5>Example 13.5 </a>seems like a tru=
th question with the underlying statement:
+<p>
+<pre><a name=3De13d6>13.6)	Fido is a dog or a cat.
+</pre><p>By translating <a href=3D#e13d6>Example 13.6 </a>into Lojban and =
prefixing =93xu=94 to signal a truth question, we get:
+<p>
+<pre><a name=3De13d7>13.7)	xu la faidon. gerku gi'onai mlatu
+	is-it-true-that Fido is-a-dog or is-a-cat (but not both)?
+</pre><p>Given that Fido really is either a dog or a cat, the appropriate =
answer would be =93go'i=94; if Fido were a fish, the appropriate answer wou=
ld be =93nago'i=94.
+<p>
+But that is not what an English-speaker who utters <a href=3D#e13d5>Exampl=
e 13.5 </a>is asking! The true significance of <a href=3D#e13d5>Example 13.=
5 </a>is that the speaker desires to know the truth value of either of the =
two underlying bridi (it is presupposed that only one is true).
+<p>
+<cx "questions, connection">  XE "questions: connection"   Lojban has an e=
legant mechanism for rendering this kind of question which is very unlike t=
hat used in English. Instead of asking about the truth value of the connect=
ed bridi, Lojban users ask about the truth function which connects them. Th=
is is done by using a special question cmavo: there is one of these for eac=
h of the logical connective selma'o, as shown by the following table:
+
+<p>
+<lx "ge'i">  XE "ge'i"   <lx "gi'i">  XE "gi'i"   <lx "gu'i">  XE "gu'i"  =
 <lx "je'i">  XE "je'i"   <lx "ji">  XE "ji"   <lx "GA">  XE "GA selma=92o"=
   <lx "GIhA">  XE "GIhA selma=92o"   <lx "GUhA">  XE "GUhA selma=92o"   <l=
x "JA">  XE "JA selma=92o"   <lx "A">  XE =93A selma=92o"   <pre> 	ge'i	GA	=
forethought connective question
+	gi'i	GIhA	bridi-tail connective question
+	gu'i	GUhA	tanru forethought connective question
+	je'i	JA	tanru connective question
+	ji	A	sumti connective question
+</pre><cx "connective question cmavo, departure from regularity of">  XE "=
connective question cmavo: departure from regularity of"   (This list unfor=
tunately departs from the pretty regularity of the other cmavo for logical =
connection. The two-syllable selma'o, GIhA and GUhA, make use of the cmavo =
ending in =93-i=94 which is not used for a truth function, but =93gi=94 and=
 =93.i=94 were not available, and different cmavo had to be chosen. This ta=
ble must simply be memorized, like most other non-connective cmavo assignme=
nts.)
+<p>
+<cx "connective questions, answering">  XE "connective questions: answerin=
g"   One correct translation of <a href=3D#e13d5>Example 13.5 </a>employs a=
 question gihek:
+<p>
+<pre><a name=3De13d8>13.8)	la .alis gerku gi'i mlatu
+	Alice is-a-dog [truth function?] is-a-cat?
+</pre><p>Here are some plausible answers:
+<p>
+<pre><a name=3De13d9>13.9)	nagi'e
+	Alice is not a dog and is a cat.
+<a name=3De13d10>13.10)	gi'enai
+	Alice is a dog and is not a cat.
+<a name=3De13d11>13.11)	nagi'enai
+	Alice is not a dog and is not a cat.
+<a name=3De13d12>13.12)	nagi'o
+	gi'onai
+	Alice is a dog or is a cat but not both (I'm not saying which).
+</pre><a href=3D#e13d12>Example 13.12 </a>is correct but uncooperative.
+<p>
+<cx "connective, as complete grammatical utterance">  XE "connective(s): a=
s complete grammatical utterance"   <cx "afterthought connective, as comple=
te grammatical utterance">  XE "afterthought connective(s): as complete gra=
mmatical utterance"   As usual, Lojban questions are answered by filling in=
 the blank left by the question. Here the blank is a logical connective, an=
d therefore it is grammatical in Lojban to utter a bare logical connective =
without anything for it to connect.
+<p>
+The answer =93gi'e=94, meaning that Alice is a dog and is a cat, is imposs=
ible in the real world, but for:
+<p>
+<ex "coffee or tea">  XE "coffee or tea: example=93  <pre><a name=3De13d13=
>13.13)	do djica tu'a loi ckafi
+		ji loi tcati
+	You desire something-about a-mass-of coffee=20
+		[truth function?] a-mass-of tea?
+	Do you want coffee or tea?
+</pre>the answer =93.e=94, meaning that I want both, is perfectly plausibl=
e, if not necessarily polite.
+<p>
+<cx "connective, as ungrammatical utterance">  XE "connective(s): as ungra=
mmatical utterance"   <cx "forethought connective, as ungrammatical utteran=
ce">  XE "forethought connective(s): as ungrammatical utterance"   <cx "for=
ethought connection, contrasted with afterthought for grammatical utterance=
s">  XE "forethought connection: contrasted with afterthought for grammatic=
al utterances"   <cx "afterthought connection, contrasted with forethought =
for grammatical utterances">  XE "afterthought connection: contrasted with =
forethought for grammatical utterances"   The forethought questions =93ge'i=
=94 and =93gu'i=94 are used like the others, but ambiguity forbids the use =
of isolated forethought connectives as answers =97 they sound like the star=
t of forethought-connected bridi. So although <a href=3D#e13d14>Example 13.=
14 </a>is the forethought version of <a href=3D#e13d13>Example 13.13</a>:
+
+<p>
+<pre><a name=3De13d14>13.14)	do djica tu'a
+		ge'i loi ckafi gi loi tcati
+	You desire something-about=20
+		[truth function?] a-mass-of coffee [or] a-mass-of tea?
+</pre>the answer must be in afterthought form.
+<p>
+<cx "connective questions, compared with other languages">  XE "connective=
 questions: compared with other languages"   There are natural languages, n=
otably Chinese, which employ the Lojbanic form of connective question. The =
Chinese sentence
+<p>
+<pre><a name=3De13d15>13.15)	ni<sup>3</sup> zou<sup>3</sup> hai<sup>2</sup=
>shi pao<sup>3</sup>
+	you walk [or?] run?
+</pre>means =93Do you walk or run?=94, and is exactly parallel to the Lojb=
an:
+<p>
+<pre><a name=3De13d16>13.16)	do cadzu gi'i bajra
+	you walk [or?] run?
+</pre><cx "connective question answers, contrasted with other languages"> =
 XE "connective question answers: contrasted with other languages"   Howeve=
r, Chinese does not use logical connectives in the reply to such a question=
, so the resemblance, though striking, is superficial.
+<p>
+<cx "bridi connection, use of truth questions in">  XE "bridi connection: =
use of truth questions in"   <cx "bridi connection, use of imperatives in">=
  XE "bridi connection: use of imperatives in"   <cx "truth, in imperative =
sentences">  XE "truth: in imperative sentences"   <cx "imperatives, and tr=
uth">  XE "imperatives: and truth"   <ex "Abraham Lincoln">  XE "Abraham Li=
ncoln: example=93  <ex "if coffee, bring tea">  XE "if coffee: bring tea, e=
xample=93  Truth questions may be used in bridi connection. This form of se=
ntence is perfectly legitimate, and can be interpreted by using the convent=
ion that a truth question is true if the answer is =93yes=94 and false if t=
he answer is =93no=94. Analogously, an imperative sentence (involving the s=
pecial pro-sumti =93ko=94, which means =93you=94 but marks the sentence as =
a command) is true if the command is obeyed, and false otherwise. A request=
 of Abraham Lincoln's may be translated thus:
+<p>
+<pre><a name=3De13d17>13.17)	ganai ti ckafi gi ko bevri loi tcati mi
+		.ije ganai ti tcati gi ko bevri loi ckafi mi
+	If this is-coffee then [you!] bring a-mass-of tea to-me,
+		and if this is-tea then [you!] bring a-mass-of coffee to-me.
+	If this is coffee, bring me tea; but if this is tea, bring me coffee.
+</pre><cx "but, compared with and">  XE "but: compared with and"   <cx "an=
d, compared with but">  XE "and: compared with but"   <lx "ku'i">  XE "ku'i=
"   <lx "UI">  XE "UI selma=92o"   In logical terms, however, =93but=94 is =
the same as =93and=94; the difference is that the sentence after a =93but=
=94 is felt to be in tension or opposition to the sentence before it. Lojba=
n represents this distinction by adding the discursive cmavo =93ku'i=94 (of=
 selma'o UI), which is explained in <a href=3Dchap13.html>Chapter 13</a>, t=
o the logical =93.ije=94.)
+<p>
+<a name=3Ds14><h3>Non-logical connectives</h3>
+<p>
+<cx "and, as non-logical connective">  XE "and: as non-logical connective"=
   Way back in <a href=3D#s1>Section 1</a>, the point was made that not eve=
ry use of English =93and=94, =93if=A0<dots>=85</dots>then=94, and so on rep=
resents a Lojban logical connective. In particular, consider the =93and=94 =
of:
+<p>
+<ex "carried piano">  XE "carried piano: example=93  <pre><a name=3De14d1>=
14.1)	John and Alice carried the piano.
+</pre><cx "mass, joining elements into a">  XE "mass: joining elements int=
o a"   Given the nature of pianos, this probably means that John carried on=
e end and Alice the other. So it is not true that:
+<p>
+<pre><a name=3De14d2>14.2)	John carried the piano, and Alice carried the p=
iano.
+</pre><lx "joi">  XE "joi"   <lx "JOI">  XE "JOI selma=92o"   which would =
mean that each of them carried the piano by himself/herself. Lojban deals w=
ith this particular linguistic phenomenon as a =93mass=94. John and Alice a=
re joined together into a mass, John-and-Alice, and it is this mass which c=
arried the piano, not either of them separately. The cmavo =93joi=94 (of se=
lma'o JOI) is used to join two or more components into a mass:
+<p>
+<pre><a name=3De14d3>14.3)	la djan. joi la .alis. cu bevri le pipno
+	John massed-with Alice carry the piano.
+</pre><cx "supervising, as a contribution to mass action">  XE "supervisin=
g: as a contribution to mass action"   <cx "mass contrasted with components=
, in properties of">  XE "mass contrasted with components: in properties of=
"   <cx "components contrasted with mass, in properties of">  XE "component=
s contrasted with mass: in properties of"   <a href=3D#e14d3>Example 14.3 <=
/a>covers the case mentioned, where John and Alice divide the labor; it als=
o could mean that John did all the hauling and Alice did the supervising. T=
his possibility arises because the properties of a mass are the properties =
of its components, which can lead to apparent contradictions: if John is sm=
all and Alice is large, then John-and-Alice is both small and large. Masses=
 are also discussed in <a href=3Dchap6.html>Chapter 6</a>.
+<p>
+<lx "JOI">  XE "JOI selma=92o"   <lx "A">  XE =93A selma=92o"   <lx "JA"> =
 XE "JA selma=92o"   <cx "non-logical connection, and elidability of termin=
ators">  XE "non-logical connection: and elidability of terminators"   <cx =
"joi grammar, contrasted with eks">  XE "joi grammar: contrasted with eks" =
  <cx "joi grammar, contrasted with jeks">  XE "joi grammar: contrasted wit=
h jeks"   <cx "non-logical connection of sumti, distinguishing from connect=
ion in tanru">  XE =93non-logical connection: of sumti, distinguishing from=
 connection in tanru"   <cx "non-logical connection in tanru, distinguishin=
g from connection of sumti">  XE =93non-logical connection: in tanru, disti=
nguishing from connection of sumti"   Grammatically, =93joi=94 can appear b=
etween two sumti (like an ek) or between two tanru components (like a jek).=
 This flexibility must be paid for in the form of occasional terminators th=
at cannot be elided:
+<p>
+<lx "ku">  XE "ku"   <lx "le">  XE "le"   <cx "terminators, eliding ku in =
non-logical connections">  XE "terminators: eliding ku in non-logical conne=
ctions"   <lx "KU">  XE "KU selma=92o"   <lx "LE">  XE "LE selma=92o"   <pr=
e><a name=3De14d4>14.4)	le nanmu ku joi le ninmu [ku] cu klama le zarci
+	The man massed-with the woman go-to the market.
+</pre><p>The cmavo =93ku=94 is the elidable terminator for =93le=94, which=
 can almost always be elided, but not in this case. If the first =93ku=94 w=
ere elided here, Lojban's parsing rules would see =93le nanmu joi=94 and as=
sume that another tanru component is to follow; since the second =93le=94 c=
annot be part of a tanru, a parsing error results. No such problem can occu=
r with logical connectives, because an ek signals a following sumti and a j=
ek a following tanru component unambiguously.
+<p>
+<cx "joik, definition">  XE "joik: definition"   <lx "JOI">  XE "JOI selma=
=92o"   Single or compound cmavo involving members of selma'o JOI are calle=
d joiks, by analogy with the names for logical connectives. It is not gramm=
atical to use joiks to connect bridi-tails.
+<p>
+<cx "tanru connection, connotation of non-logical">  XE "tanru connection:=
 connotation of non-logical"   <ex "mixed with">  XE "mixed with: example=
=93  <ex "blue and red">  XE "blue and red: example=93  In tanru, =93joi=94=
 has the connotation =93mixed with=94, as in the following example:
+<p>
+<pre><a name=3De14d5>14.5)	ti blanu joi xunre bolci
+	This is-a-(blue mixed-with red) ball.
+	This is a blue and red ball.
+</pre><p>Here the ball is neither wholly blue nor wholly red, but partly b=
lue and partly red. Its blue/redness is a mass property. (Just how blue som=
ething has to be to count as =93wholly blue=94 is an unsettled question, th=
ough. A =93blanu zdani=94 may be so even though not every part of it is blu=
e.)
+<p>
+There are several other cmavo in selma'o JOI which can be used in the same=
 grammatical constructions. Not all of them are well-defined as yet in all =
contexts. All have clear definitions as sumti connectives; those definition=
s are shown in the following table:
+<p>
+<lx "JOI">  XE "JOI selma=92o"   <lx "joi">  XE "joi"   <lx "ce">  XE "ce"=
   <lx "ce'o">  XE "ce'o"   <lx "sece'o">  XE "sece'o"   <lx "jo'u">  XE "j=
o'u"   <lx "fa'u">  XE "fa'u"   <lx "sefa'u">  XE "sefa'u"   <lx "jo'e">  X=
E "jo'e"   <lx "ku'a">  XE "ku'a"   <lx "pi'u">  XE "pi'u"   <lx "sepi'u"> =
 XE "sepi'u"   <lx "se">  XE "se"   <pre> 	A joi B	the mass with components=
 A and B
+	A ce B	the set with elements A and B
+	A ce'o B	the sequence with elements A and B in order
+	A sece'o B	the sequence with elements B and A in order
+	A jo'u B	A and B considered jointly
+	A fa'u B	A and B respectively
+	A sefa'u B	B and A respectively
+	A jo'e B	the union of sets A and B
+	A ku'a B	the intersection of sets A and B
+	A pi'u B	the cross product of sets A and B
+	A sepi'u B	the cross product of sets B and A
+</pre><cx "se, as grammatical in JOI compounds">  XE "se: as grammatical i=
n JOI compounds"   <cx "joiks, use of =93se=94 in">  XE "joiks: use of \=93=
se\=94 in"   <lx "JOI">  XE "JOI selma=92o"   The cmavo =93se=94 is grammat=
ical before any JOI cmavo, but only useful with those that have inherent or=
der. Here are some examples of joiks:
+<p>
+<ex "choose from">  XE "choose from: example=93  <pre><a name=3De14d6>14.6=
)	mi cuxna la .alis. la frank. ce la .alis. ce la djeimyz.
+	I choose Alice from Frank and-member Alice and-member James.
+	I choose Alice from among Frank, Alice, and James.
+</pre><cx "set, as specified by members">  XE "set: as specified by member=
s"   <cx "set, contrasted with mass in distribution of properties">  XE "se=
t: contrasted with mass in distribution of properties"   <cx "mass, contras=
ted with set in distribution of properties">  XE "mass: contrasted with set=
 in distribution of properties"   The x3 place of =93cuxna=94 is a set from=
 which the choice is being made. A set is an abstract object which is deter=
mined by specifying its members. Unlike those of a mass, the properties of =
a set are unrelated to its members' properties: the set of all rats is larg=
e (since many rats exist), but the rats themselves are small. This chapter =
does not attempt to explain set theory (the mathematical study of sets) in =
detail: explaining propositional logic is quite enough for one chapter!
+<p>
+<lx "ce">  XE "ce"   <cx "set, by listing members with ce">  XE "set: by l=
isting members with ce"   In <a href=3D#e14d6>Example 14.6 </a>we specify t=
hat set by listing the members with =93ce=94 joining them.
+<p>
+<ex "list">  XE "list: example=93  <pre><a name=3De14d7>14.7)	ti liste mi =
ce'o do ce'o la djan.
+	This is-a-list-of me and-sequence you and-sequence John.
+	This is a list of you, me, and John.
+</pre><cx "list, contrasted with sequence">  XE "list: contrasted with seq=
uence"   <cx "sequence, contrasted with list">  XE "sequence: contrasted wi=
th list"   <cx "list, as a physical object">  XE "list: as a physical objec=
t"   <cx "sequence, as an abstract list">  XE "sequence: as an abstract lis=
t"   <cx "ordered sequence, by listing members">  XE "ordered sequence: by =
listing members"   <cx "ordered sequence, contrasted with set">  XE "ordere=
d sequence: contrasted with set"   <cx "ordered sequence, contrasted with m=
ass">  XE "ordered sequence: contrasted with mass"   <cx "mass, contrasted =
with ordered sequence">  XE "mass: contrasted with ordered sequence"   <cx =
"set, contrasted with ordered sequence">  XE "set: contrasted with ordered =
sequence"   The x2 place of =93liste=94 is a sequence of the things which a=
re mentioned in the list. (It is worth pointing out that =93lo liste=94 mea=
ns a physical object such as a grocery list: a purely abstract list is =93l=
o porsi=94, a sequence.) Here the three sumti connected by =93ce'o=94 are i=
n a definite order, not just lumped together in a set or a mass.
+<p>
+<cx "non-logical connection, of individuals into mass">  XE "non-logical c=
onnection: of individuals into mass"   <cx "non-logical connection, of indi=
viduals into set">  XE "non-logical connection: of individuals into set"   =
<cx "individuals into mass, by non-logical connection">  XE "individuals in=
to mass: by non-logical connection"   <cx "individuals into set, by non-log=
ical connection">  XE "individuals into set: by non-logical connection"   <=
lx "jo'u">  XE "jo'u"   <cx "jo'u, contrasted with joi">  XE "jo'u: contras=
ted with joi"   <cx "jo'u, contrasted with ce">  XE "jo'u: contrasted with =
ce"   <cx "jo'u, contrasted with ce'o">  XE "jo'u: contrasted with ce'o"   =
<cx "jo'u, result of connection with">  XE "jo'u: result of connection with=
"   So =93joi=94, =93ce=94, and =93ce'o=94 are parallel, in that the sumti =
connected are taken to be individuals, and the result is something else: a =
mass, a set, or a sequence respectively. The cmavo =93jo'u=94 serves as a f=
ourth element in this pattern: the sumti connected are individuals, and the=
 result is still individuals =97 but inseparably so. The normal Lojban way =
of saying that James and George are brothers is:
+<p>
+<ex "brothers">  XE "brothers: example=93  <pre><a name=3De14d8>14.8)	la d=
jeimyz. bruna la djordj.
+	James is-the-brother-of George.
+</pre>possibly adding a discursive element meaning =93and vice versa=94. H=
owever, =93James and George are brothers=94 cannot be correctly translated =
as:
+<p>
+<pre><a name=3De14d9>14.9)	la djeimyz. .e la djordj. bruna
+	James and George is-a-brother.
+</pre>since that expands to two bridi and means that James is a brother an=
d so is George, but not necessarily of each other. If the =93.e=94 is chang=
ed to =93jo'u=94, however, the meaning of <a href=3D#e14d8>Example 14.8 </a=
>is preserved:
+<p>
+<pre><a name=3De14d10>14.10)	la djeimyz. jo'u la djordj. cu remei bruna
+	James in-common-with George are-a-twosome type-of-brothers.
+</pre><p>The tanru =93remei bruna=94 is not strictly necessary in this sen=
tence, but is used to make clear that we are not saying that James and Geor=
ge are both brothers of some third person not specified. Alternatively, we =
could turn the tanru around: the x1 place of =93remei=94 is a mass with two=
 components, leading to:
+<p>
+<lx "joi">  XE "joi"   <pre><a name=3De14d11>14.11)	la djeimyz. joi la djo=
rdj. cu bruna remei
+	James massed-with George are-a-brother type-of-twosome.
+</pre>where =93joi=94 is used to create the necessary mass.
+<p>
+<lx "fa'u">  XE "fa'u"   <cx "connection, non-distributed">  XE "connectio=
n: non-distributed"   <cx "respectively, specifying with fa'u">  XE "respec=
tively: specifying with fa'u"   <ex "respectively">  XE "respectively: exam=
ple=93  Likewise, =93fa'u=94 can be used to put two individuals together wh=
ere order matters. Typically, there will be another =93fa'u=94 somewhere el=
se in the same bridi:
+<p>
+<pre><a name=3De14d12>14.12)	la djeimyz. fa'u la djordj. prami la meris. f=
a'u la martas.
+	James jointly-in-order-with George loves Mary jointly-in-order-with Marth=
a.
+	James and George love Mary and Martha, respectively.
+</pre><cx "fa'u, contrasted with .e">  XE "fa'u: contrasted with .e"   <cx=
 ".e, contrasted with fa'u">  XE "e: contrasted with fa'u"   Here the infor=
mation carried by the English adverb =93respectively=94, namely that James =
loves Mary and George loves Martha, is divided between the two occurrences =
of =93fa'u=94. If both uses of =93fa'u=94 were to be changed to =93.e=94, w=
e would get:
+<p>
+<pre><a name=3De14d13>14.13)	la djeimyz. .e la djordj. prami la meris. .e =
la martas.
+	James and George love Mary and Martha.
+</pre>which can be transformed to four bridi:
+<p>
+<pre><a name=3De14d14>14.14)	la djeimyz. prami la meris.  .ije la djordj. =
prami la meris.
+		.ije la djeimyz. prami la martas.  .ije la djordj. prami la martas.
+	James loves Mary, and George loves Mary,
+		and James loves Martha, and George loves Martha.
+</pre>which represents quite a different state of affairs from <a href=3D#=
e14d12>Example 14.12</a>. The meaning of <a href=3D#e14d12>Example 14.12 </=
a>can also be conveyed by a termset:
+<p>
+<cx "fa'u, compared to termsets">  XE "fa'u: compared to termsets"   <cx "=
termsets, compared to fa'u">  XE "termsets: compared to fa'u"   <pre><a nam=
e=3De14d15>14.15)	la djeimyz. ce'e la meris. pe'e .e la djordj. ce'e la mar=
tas. prami
+	James [plus] Mary [joint] and George [plus] Martha loves.
+</pre>at the expense of re-ordering the list of names so as to make the pa=
irs explicit. This option is not available when one of the lists is only de=
scribed rather than enumerated:
+<p>
+<pre><a name=3De14d16>14.16)	la djeimyz. fa'u la djordj. prami re sorme
+	James and-respectively George love two sisters.
+</pre>which conveys that James loves one sister and George the other, thou=
gh we are not able to tell which of the sisters is which.
+<p>
+<a name=3Ds15><h3>More about non-logical connectives</h3>
+<p>
+<lx "jo'e">  XE "jo'e"   <lx "ku'a">  XE "ku'a"   <lx "pi'u">  XE "pi'u"  =
 <cx "set operations">  XE "set operations"   <cx "union, of sets">  XE "un=
ion: of sets"   <cx "intersection, of sets">  XE "intersection: of sets"   =
<cx "cross-product, of sets">  XE "cross-product: of sets"   The final thre=
e JOI cmavo, =93jo'e=94, =93ku'a=94, and =93pi'u=94, are probably only usef=
ul when talking explicitly about sets. They represent three standard set op=
erators usually called =93union=94, =93intersection=94, and =93cross produc=
t=94 (also known as =93Cartesian product=94). The union of two sets is a se=
t containing all the members that are in either set; the intersection of tw=
o sets is a set containing all the members that are in both sets. The cross=
 product of two sets is the set of all possible ordered pairs, where each o=
rdered pair contains a single element from the first set followed by a sing=
le element from the second. This may seem very abstract; hopefully, the fol=
lowing examples will help:
+<p>
+<ex "rich and German">  XE "rich and German: example=93  <pre><a name=3De1=
5d1>15.1)	lo'i ricfu ku jo'e lo'i dotco cu barda
+	The-set-of rich-things union the-set-of German-things is large.
+<a name=3De15d2>15.2)	lo'i ricfu ku ku'a lo'i dotco cu cmalu
+	The set-of rich-things intersection the-set-of German-things is small.
+
+</pre><cx "union of sets, compared with or">  XE "union of sets: compared =
with or"   <cx "intersection of sets, compared with and">  XE "intersection=
 of sets: compared with and"   There is a parallelism between logic and set=
 theory that makes <a href=3D#e15d1>Example 15.1 </a>and <a href=3D#e15d2>E=
xample 15.2 </a>equivalent respectively to:
+<p>
+<pre><a name=3De15d3>15.3)	lo'i ricfu ja dotco cu barda
+	The-set-of rich-or-German-things is large.
+</pre>and
+<p>
+<pre><a name=3De15d4>15.4)	lo'i ricfu je dotco cu cmalu
+	The-set-of rich-and-German-things is small.
+</pre><p>The following example uses =93se remei=94, which is a set (not a =
mass) of two elements:
+<p>
+<cx "pi'u, contrasted with .e">  XE "pi'u: contrasted with .e"   <cx "e, c=
ontrasted with pi'u">  XE "e: contrasted with pi'u"   <cx "cross-product, c=
ontrasted with and">  XE "cross-product: contrasted with and"   <cx "and, c=
ontrasted with cross-product">  XE "and: contrasted with cross-product"   <=
pre><a name=3De15d5>15.5)	la djeimyz. ce[bo] la djordj. pi'u la meris. cebo=
 la martas.
+		cu prami se remei
+	James and-set George cross-product Mary and-set Martha=20
+		are-lover type-of-pairs.
+</pre>means that each of the pairs James/Mary, George/Mary, James/Martha, =
and George/Martha love each other. Therefore it is similar in meaning to <a=
 href=3D#e14d13>Example 14.13</a>; however, that example speaks only of the=
 men loving the women, not vice versa.
+<p>
+<cx "joiks, grouping">  XE "joiks: grouping"   <lx "JOI">  XE "JOI selma=
=92o"   <cx "non-logical connectives, grouping">  XE "non-logical connectiv=
es: grouping"   Joiks may be combined with =93bo=94 or with =93ke=94 in the=
 same way as eks and jeks; this allows grouping of non-logical connections =
between sumti and tanru units, in complete parallelism with logical connect=
ions:
+<p>
+<pre><a name=3De15d6>15.6)	mi joibo do ce la djan. joibo la djein.
+		cu gunma se remei
+	(I massed-with you) and (John massed-with Jane)=20
+		are-a-mass type-of-two-set
+</pre>asserts that there is a set of two items each of which is a mass.
+<p>
+<cx "non-logical connection, of termsets">  XE "non-logical connection: of=
 termsets"   <cx "termsets, non-logical connection of">  XE "termsets: non-=
logical connection of"   Non-logical connection is permitted at the joint o=
f a termset; this is useful for associating more than one sumti or tagged s=
umti with each side of the non-logical connection. The place structure of =
=93casnu=94 is:
+<p>
+<dl compact><dt>casnu:  <dd>the mass x1 discusses/talks about x2
+</dl>so the x1 place must be occupied by a mass (for reasons not explained=
 here); however, different components of the mass may discuss in different =
languages. To associate each participant with his or her language, we can s=
ay:
+<p>
+<ex "discuss in language">  XE "discuss in language: example=93  <pre><a n=
ame=3De15d7>15.7)	 mi ce'e bau la lojban.
+		pe'e joi do ce'e bau la gliban. nu'u casnu
+	( I [plus] in-language Lojban=20
+		massed-with you [plus] in-language English ) discuss.
+</pre><p>Like all non-logical connectives, the usage shown in <a href=3D#e=
15d7>Example 15.7 </a>cannot be mechanically converted into a non-logical c=
onnective placed at another location in the bridi. The forethought equivale=
nt of <a href=3D#e15d7>Example 15.7 </a>is:
+<p>
+<pre><a name=3De15d8>15.8)	 nu'i joigi mi bau la lojban gi do bau la gliba=
n. nu'u casnu
+
+</pre><cx "non-logical forethought termsets, connecting tagged sumti">  XE=
 "non-logical forethought termsets: connecting tagged sumti"   <cx "tagged =
sumti termsets, connecting with non-logical forethought connectives">  XE "=
tagged sumti termsets: connecting with non-logical forethought connectives"=
   <cx "respectively, with different relationships">  XE "respectively: wit=
h different relationships"   Non-logical forethought termsets are also usef=
ul when the things to be non-logically connected are sumti preceded with te=
nse or modal (BAI) tags:
+<p>
+<pre><a name=3De15d9>15.9)	la djan. fa'u la frank. cusku nu'i bau la lojba=
n.
+		nu'u fa'u bai tu=92a la djordj. [nu'u]
+	John respectively-with Frank express [start termset] in-language Lojban
+		[joint] respectively-with under-compulsion-by George.
+	John and Frank speak in Lojban and under George's compulsion, respectivel=
y.
+</pre><a href=3D#e14d17>Example 14.17 </a>associates speaking in Lojban wi=
th John, and speaking under George's compulsion with Frank. We do not know =
what language Frank uses, or whether John speaks under anyone's compulsion.
+<p>
+<cx "ijoik, definition">  XE "ijoik: definition"   <lx "I">  XE "I selma=
=92o"   <lx "JOI">  XE "JOI selma=92o"   <cx "non-logical connectives, sent=
ence">  XE "non-logical connectives: sentence"   <cx "sentences, connecting=
 non-logically">  XE "sentences: connecting non-logically"   <cx "ice'o, co=
ntrasted with .ibabo">  XE "ice'o: contrasted with .ibabo"   <cx "sequence =
of events, expressing non-time-related sequences">  XE "sequence of events:=
 expressing non-time-related sequences"   Joiks may be prefixed with =93.i=
=94 to produce ijoiks, which serve to non-logically connect sentences. The =
ijoik =93.ice'o=94 indicates that the event of the second bridi follows tha=
t of the first bridi in some way other than a time relationship (which is h=
andled with a tense):
+<p>
+<ex "to-do list">  XE "to-do list: example=93  <ex "list of things to do">=
  XE "list of things to do: example=93  <pre><a name=3De15d10>15.10)	mi ba =
gasnu la'edi'e
+		.i tu'e kanji lo ni cteki
+		.ice'o lumci le karce
+		.ice'o dzukansa le gerku tu'u
+	I [future] do the-referent-of-the-following:
+		( Compute the quantity of taxes.
+		And-then wash the car.
+		And-then walkingly-accompany the dog. )
+	List of things to do:
+		Figure taxes.
+		Wash car.
+		Walk dog.
+</pre><lx "TUhE">  XE "TUhE selma=92o"   <lx "TUhU">  XE "TUhU selma=92o" =
  <lx "di'e">  XE "di'e"   <lx "tu'u">  XE "tu'u"   <lx "tu'e">  XE "tu'e" =
  <cx "lists, use of tu'e/tu'u in">  XE "lists: use of tu'e/tu'u in"   <cx =
"di'e, effect of tu'e/tu'u on">  XE "di'e: effect of tu'e/tu'u on"   <cx "t=
u'e, effect on di'e">  XE "tu'e: effect on di'e"   <cx "tu'e, use in lists"=
>  XE "tu'e: use in lists"   <a href=3D#e15d10>Example 15.10 </a>represents=
 a list of things to be done in priority order. The order is important, hen=
ce the need for a sequence connective, but does not necessarily represent a=
 time order (the dog may end up getting walked first). Note the use of =93t=
u'e=94 and =93tu'u=94 as general brackets around the whole list. This is re=
lated to, but distinct from, their use in <a href=3D#s8>Section 8</a>, beca=
use there is no logical connective between the introductory phrase =93mi ba=
 gasnu la'edi'e=94 and the rest. The brackets effectively show how large an=
 utterance the word =93di'e=94, which means =93the following utterance=94, =
refers to.
+Similarly, =93.ijoi=94 is used to connect sentences that represent the com=
ponents of a joint event such as a joint cause: the Lojban equivalent of =
=93Fran hit her head and fell out of the boat, so that she drowned=94 would=
 join the events =93Fran hit her head=94 and =93Fran fell out of the boat=
=94 with =93.ijoi=94.
+<p>
+<cx "nai, effect on joiks">  XE "nai: effect on joiks"   <cx "joiks, effec=
t of nai on">  XE "joiks: effect of nai on"   <cx "non-logical connectives,=
 effect of nai on">  XE "non-logical connectives: effect of nai on"   <lx "=
nai">  XE "nai"   <cx "scalar negation of non-logical connective">  XE "sca=
lar negation of non-logical connective"   The following =93nai=94, if prese=
nt, does not negate either of the things to be connected, but instead speci=
fies that some other connection (logical or non-logical) is applicable: it =
is a scalar negation:
+<p>
+<pre><a name=3De15d11>15.11)	mi jo'unai do cu remei
+	I in-common-with [not!] you are-a-twosome
+</pre><p>The result of =93mi jo'u do=94 would be two individuals, not a ma=
ss, therefore =93jo'u=94 is not applicable; =93joi=94 would be the correct =
connective.
+<p>
+<cx "connective questions, non-logical">  XE "connective questions: non-lo=
gical"   <cx "connective answers, non-logical">  XE "connective answers: no=
n-logical"   There is no joik question cmavo as such; however, joiks and ij=
oiks may be uttered in isolation in response to a logical connective questi=
on, as in the following exchange:
+<p>
+<ex "coffee mixed with tea">  XE "coffee mixed with tea: example=93  <pre>=
<a name=3De15d12>15.12)	do djica tu'a
+		loi ckafi ji loi tcati
+	You desire something-about
+		a-mass-of coffee [what connective?] a-mass-of tea?
+	Do you want coffee or tea?
+<a name=3De15d13>15.13)	joi
+	Mixed-mass-and.
+	Both as a mass (i.e, mixed together).
+</pre><ex "ugh">  XE "ugh: example=93  Ugh. (Or in Lojban: .a'unaisairo'o.=
)
+<p>
+<a name=3Ds16><h3>Interval connectives and forethought non-logical connect=
ion</h3>
+<p>
+<lx "BIhI">  XE "BIhI selma=92o"   <lx "bi'i">  XE "bi'i"   <lx "bi'o">  X=
E "bi'o"   <lx "mi'i">  XE "mi'i"   <cx "non-logical connectives, intervals=
">  XE "non-logical connectives: intervals"   <cx "intervals, expressed as =
endpoints">  XE "interval: expressed as endpoints"   In addition to the non=
-logical connectives of selma'o JOI explained in <a href=3D#s14>Sections 14=
 </a>and <a href=3D#s15>15</a>, there are three other connectives which can=
 appear in joiks: =93bi'i=94, =93bi'o=94, and =93mi'i=94, all of selma'o BI=
hI. The first two cmavo are used to specify intervals: abstract objects def=
ined by two endpoints. The cmavo =93bi'i=94 is correct if the endpoints are=
 independent of order, whereas =93bi'o=94 or =93sebi'o=94 are used when ord=
er matters.
+<p>
+An example of =93bi'i=94 in sumti connection:
+<p>
+<ex "between Dresden and Frankfurt">  XE "between Dresden and Frankfurt: e=
xample=93  <lx "bi'i">  XE "bi'i"   <pre><a name=3De16d1>16.1)	mi ca sanli =
la drezdn. bi'i la frankfurt.
+	I [present] stand-on-surface Dresden [interval] Frankfurt.
+	I am standing between Dresden and Frankfurt.
+</pre><cx "non-logical connectives, un-ordered intervals">  XE "non-logica=
l connectives: un-ordered intervals"   In <a href=3D#e16d1>Example 16.1</a>=
, it is all the same whether I am standing between Dresden and Frankfurt or=
 between Frankfurt and Dresden, so =93bi'i=94 is the appropriate interval c=
onnective. The sumti =93la drezdn. bi'i la frankfurt.=94 falls into the x2 =
place of =93sanli=94, which is the surface I stand on; the interval specifi=
es that surface by its limits. (Obviously, I am not standing on the whole o=
f the interval; the x2 place of =93sanli=94 specifies a surface which is ty=
pically larger in extent than just the size of the stander's feet.)
+<p>
+<lx "bi'o">  XE "bi'o"   <ex "from one to two o'clock">  XE "from one to t=
wo o'clock: example=93  <pre><a name=3De16d2>16.2)	mi cadzu ca la pacac. bi=
'o la recac.
+	I walk simultaneous-with First-hour [ordered-interval] Second-hour.
+	I walk from one o'clock to two o'clock.
+</pre><cx "non-logical connectives, ordered intervals">  XE "non-logical c=
onnectives: ordered intervals"   In <a href=3D#e16d2>Example 16.2</a>, on t=
he other hand, it is essential that =93la pacac.=94 comes before =93la reca=
c.=94; otherwise we have an 11-hour (or 23-hour) interval rather than a one=
-hour interval. In this use of an interval, the whole interval is probably =
intended, or at least most of it.
+<p>
+<a href=3D#e16d2>Example 16.2 </a>is equivalent to:
+<p>
+<lx "sebi'o">  XE "sebi'o"   <pre><a name=3De16d3>16.3)	mi cadzu ca la rec=
ac. sebi'o la pacac.
+	I walk simultaneous-with Second-hour [reverse] [ordered] First-hour.
+</pre><p>English cannot readily express =93sebi'o=94, but its meaning can =
be understood by reversing the two sumti.
+<p>
+<lx "mi'i">  XE "mi'i"   <cx "intervals, expressed as center and distance"=
>  XE "interval: expressed as center and distance"   The third cmavo of sel=
ma'o BIhI, namely =93mi'i=94, expresses an interval seen from a different v=
iewpoint: not a pair of endpoints, but a center point and a distance. For e=
xample:
+
+<p>
+<ex "bomb destroyed fifty miles">  XE "bomb destroyed fifty miles: example=
=93  <pre><a name=3De16d4>16.4)	le jbama pu daspo la .uacintyn. mi'i lo min=
li
+		be li muno
+	The bomb [past] destroys Washington [center] what-is measured-in-miles=20
+by 50.
+	The bomb destroyed Washington and fifty miles around.
+</pre><p>Here we have an interval whose center is Washington and whose dis=
tance, or radius, is fifty miles.
+<p>
+<cx "closed interval">  XE "closed interval"   <cx "open interval">  XE "o=
pen interval"   <cx "interval, closed">  XE "interval: closed"   <cx "inter=
val, open">  XE "interval: open"   <cx "interval, inclusion of endpoints"> =
 XE "interval: inclusion of endpoints"   <cx "endpoints, inclusion in inter=
val">  XE "endpoints: inclusion in interval"   <lx "ga'o">  XE "ga'o"   <lx=
 "ke'i">  XE "ke'i"   <lx "GAhO">  XE "GAhO selma=92o"   In <a href=3D#e16d=
1>Example 16.1</a>, is it possible that I am standing in Dresden (or Frankf=
urt) itself? Yes. The connectives of selma'o BIhI are ambiguous about wheth=
er the endpoints themselves are included in or excluded from the interval. =
Two auxiliary cmavo =93ga'o=94 and =93ke'i=94 (of cmavo GAhO) are used to i=
ndicate the status of the endpoints: =93ga'o=94 means that the endpoint is =
included, =93ke'i=94 that it is excluded:
+<p>
+<ex "between Dresden and Frankfurt">  XE "between Dresden and Frankfurt: e=
xample=93  <pre><a name=3De16d5>16.5)	mi ca sanli la drezdn. ga'o bi'i ga'o=
 la frankfurt.
+	I [present] stand Dresden [inclusive] [interval] [inclusive] Frankfurt.
+	I am standing between Dresden and Frankfurt, inclusive of both.
+<a name=3De16d6>16.6)	mi ca sanli la drezdn. ga'o bi'i ke'i la frankfurt.
+	I [present] stand Dresden [inclusive] [interval] [exclusive] Frankfurt.
+	I am standing between Dresden (inclusive) and Frankfurt (exclusive).
+<a name=3De16d7>16.7)	mi ca sanli la drezdn. ke'i bi'i ga'o la frankfurt.
+	I [present] stand Dresden [exclusive] [interval] [inclusive] Frankfurt.
+	I am standing between Dresden (exclusive) and Frankfurt (inclusive).
+<a name=3De16d8>16.8)	mi ca sanli la drezdn. ke'i bi'i ke'i la frankfurt.
+	I [present] stand Dresden [exclusive] [interval] [exclusive] Frankfurt.
+	I am standing between Dresden and Frankfurt, exclusive of both.
+</pre><cx "GAhO, grammar of">  XE "GAhO selma=92o: grammar of"   As these =
examples should make clear, the GAhO cmavo that applies to a given endpoint=
 is the one that stands physically adjacent to it: the left-hand endpoint i=
s referred to by the first GAhO, and the right-hand endpoint by the second =
GAhO. It is ungrammatical to have just one GAhO.
+<p>
+<cx "ga'o, etymology of">  XE "ga'o: etymology of"   <cx "ke'i, etymology =
of">  XE "ke'i: etymology of"   (Etymologically, =93ga'o=94 is derived from=
 =93ganlo=94, which means =93closed=94, and =93ke'i=94 from =93kalri=94, wh=
ich means =93open=94. In mathematics, inclusive intervals are referred to a=
s closed intervals, and exclusive intervals as open ones.)
+<p>
+<cx "BIhI, grammar of">  XE "BIhI selma=92o: grammar of"   BIhI joiks are =
grammatical anywhere that other joiks are, including in tanru connection an=
d (as ijoiks) between sentences. No meanings have been found for these uses=
.
+<p>
+<cx "negated intervals, meaning of">  XE "negated interval: meaning of"   =
<cx "nai, effect on intervals">  XE "nai: effect on intervals"   <cx "inter=
vals, effect of nai on">  XE "interval: effect of nai on"   Negated interva=
ls, marked with a =93-nai=94 following the BIhI cmavo, indicate an interval=
 that includes everything but what is between the endpoints (with respect t=
o some understood scale):
+<p>
+<ex "except from 10 to 12">  XE "except from 10 to 12: example=93  <pre><a=
 name=3De16d9>16.9)	do dicra .e'a mi ca la daucac. bi'onai la gaicac.
+	You disturb (allowed) me at 10 not-from=A0<dots>=85</dots>to 12
+	You can contact me except from 10 to 12.
+</pre><p>The complete syntax of joiks is:
+<p>
+<lx "se">  XE "se"   <lx "nai">  XE "nai"   <lx "JOI">  XE "JOI selma=92o"=
   <lx "BIhI">  XE "BIhI selma=92o"   <lx "GAhO">  XE "GAhO selma=92o"   <c=
x "joiks, syntax of">  XE "joiks: syntax of"   <dl compact><dt><dd>[se] JOI=
 [nai]
+[se] BIhI [nai]
+GAhO [se] BIhI [nai] GAhO
+</dl><cx "intervals, forethought">  XE "interval: forethought"   <lx "gi">=
  XE "gi"   <cx "joigik, definition">  XE "joigik: definition"   <lx "GI"> =
 XE "GI selma=92o"   <lx "JOI">  XE "JOI selma=92o"   Notice that the collo=
quial English translations of =93bi'i=94 and =93bi'o=94 have forethought fo=
rm: =93between=A0<dots>=85</dots>and=94 for =93bi'i=94, and =93from=A0<dots=
>=85</dots>to=94 for =93bi'o=94. In Lojban too, non-logical connectives can=
 be expressed in forethought. Rather than using a separate selma'o, the for=
ethought logical connectives are constructed from the afterthought ones by =
suffixing =93gi=94. Such a compound cmavo is not unnaturally called a =93jo=
igik=94; the syntax of joigiks is any of:
+<p>
+<cx "joigiks, syntax of">  XE "joigiks: syntax of"   <lx "se">  XE "se"   =
<lx "nai">  XE "nai"   <lx "GI">  XE "GI selma=92o"   <lx "BIhI">  XE "BIhI=
 selma=92o"   <lx "JOI">  XE "JOI selma=92o"   <lx "GAhO">  XE "GAhO selma=
=92o"   <dl compact><dt><dd>[se] JOI [nai] GI
+[se] BIhI [nai] GI
+GAhO [se] BIhI [nai] GAhO GI
+</dl><cx "joigiks, connection types">  XE "joigiks: connection types"   Jo=
igiks may be used to non-logically connect bridi, sumti, and bridi-tails; a=
nd also in termsets.
+<p>
+<a href=3D#e14d3>Example 14.3 </a>in forethought becomes:
+<p>
+<ex "carry the piano">  XE "carry the piano: example=93  <pre><a name=3De1=
6d10>16.10)	joigi la djan. gi la .alis. bevri le pipno
+	[Together] John and Alice carry the piano.
+</pre><p>The first =93gi=94 is part of the joigik; the second =93gi=94 is =
the regular gik that separates the two things being connected in all foreth=
ought forms.
+<p>
+<a href=3D#e16d6>Example 16.6 </a>can be expressed in forethought as:
+<p>
+<ex "between Dresden and Frankfurt">  XE "between Dresden and Frankfurt: e=
xample=93  <pre><a name=3De16d11>16.11)	mi ca sanli ke'i bi'i ga'o gi la dr=
ezdn. gi la frankfurt.
+	I [present] stand [exclusive] between [inclusive] Dresden and Frankfurt.
+	I am standing between Dresden (exclusive) and Frankfurt (inclusive).
+</pre><cx "forethought intervals, GAhO position">  XE "forethought interva=
l: GAhO position"   <cx "GAhO position in forethought intervals">  XE "GAhO=
 position in forethought intervals"   In forethought, unfortunately, the GA=
hOs become physically separated from the endpoints, but the same rule appli=
es: the first GAhO refers to the first endpoint.
+<p>
+<a name=3Ds17><h3>Logical and non-logical connectives within mekso</h3>
+<p>
+<cx "logical connection in mathematical expressions">  XE =93logical conne=
ction: in mathematical expressions"   <cx "non-logical connection in mathem=
atical expressions">  XE =93non-logical connection: in mathematical express=
ions"   <cx "mathematical expressions, connectives in">  XE "mathematical e=
xpressions: connectives in"   Lojban has a separate grammar embedded within=
 the main grammar for representing mathematical expressions (or mekso in Lo=
jban) such as =932 + 2=94. Mathematical expressions are explained fully in =
<a href=3Dchap18.html>Chapter 18</a>. The basic components of mekso are ope=
rands, like =932=94, and operators, like =93+=94. Both of these may be eith=
er logically or non-logically connected.
+<p>
+<lx "bo">  XE "bo"   <lx "BO">  XE "BO selma=92o"   <lx "A">  XE =93A selm=
a=92o"   <lx "GA">  XE "GA selma=92o"   <lx "JA">  XE "JA selma=92o"   <lx =
"GUhA">  XE "GUhA selma=92o"   <lx "JOI">  XE "JOI selma=92o"   <cx "eks, c=
onnecting operands">  XE "eks: connecting operands"   <cx "geks, connecting=
 operands">  XE "geks: connecting operands"   <cx "operands, connecting">  =
XE "operands: connecting"   <cx "operators, connecting">  XE "operators: co=
nnecting"   <cx "jeks, connecting operators">  XE "jeks: connecting operato=
rs"   <cx "guheks, connecting operators">  XE "guheks: connecting operators=
"   <cx "bo, in jeks for operators">  XE "bo: in jeks for operators"   <cx =
"bo, in joiks for operators">  XE "bo: in joiks for operators"   <cx "conne=
cting operators, with bo in connective">  XE "connecting operators: with bo=
 in connective"   Operands are connected in afterthought with eks and in fo=
rethought with geks, just like sumti. Operators, on the other hand, are con=
nected in afterthought with jeks and in forethought with guheks, just like =
tanru components. (However, jeks and joiks with =93bo=94 are not allowed fo=
r operators.) This parallelism is no accident.
+<p>
+<lx "KE">  XE "KE selma=92o"   <lx "BO">  XE "BO selma=92o"   <lx "bo">  X=
E "bo"   <cx "connecting operands, with bo in connective">  XE "connecting =
operands: with bo in connective"   <cx "connecting operands, with ke in con=
nective">  XE "connecting operands: with ke in connective"   <cx "connectin=
g operators, with ke in connective">  XE "connecting operators: with ke in =
connective"   <lx "ke">  XE "ke"   <lx "ke'e">  XE "ke'e"   <cx "operators,=
 analogue of tanru in">  XE "operators: analogue of tanru in"   In addition=
, eks with =93bo=94 and with =93ke=A0<dots>=85</dots>ke'e=94 are allowed fo=
r grouping logically connected operands, and =93ke=A0<dots>=85</dots>ke'e=
=94 is allowed for grouping logically connected operators, although there i=
s no analogue of tanru among the operators.
+<p>
+Only a few examples of each kind of mekso connection will be given. Despit=
e the large number of rules required to support this feature, it is of rela=
tively minor importance in either the mekso or the logical-connective schem=
e of things. These examples are drawn from <a href=3Dchap18.html>Chapter 18=
</a>, and contain many mekso features not explained in this chapter.
+<p>
+<a href=3D#e17d1>Example 17.1 </a>exhibits afterthought logical connection=
 between operands:
+<p>
+<ex "three of four people">  XE "three of four people: example=93  <pre><a=
 name=3De17d1>17.1)	vei ci .a vo [ve'o] prenu cu klama le zarci
+	( Three or four ) people go-to the market.
+
+</pre><a href=3D#e17d2>Example 17.2 </a>is equivalent in meaning, but uses=
 forethought connection:
+<p>
+<pre><a name=3De17d2>17.2)	vei ga ci gi vo [ve'o] prenu cu klama le zarci
+	( Either 3 or 4 ) people go-to the market.
+</pre><lx "ve'o">  XE "ve'o"   Note that the mekso in <a href=3D#e17d1>Exa=
mple 17.1 </a>and <a href=3D#e17d2>Example 17.2 </a>are being used as quant=
ifiers. Lojban requires that any mekso other than a simple number be enclos=
ed in =93vei=94 and =93ve'o=94 parentheses when used as a quantifier. The r=
ight parenthesis mark, =93ve'o=94, is an elidable terminator.
+<p>
+Simple examples of logical connection between operators are hard to come b=
y. A contrived example is:
+<p>
+<pre><a name=3De17d3>17.3)	li re su'i je pi'i re du li vo
+	The-number 2 plus and times 2 equals the-number 4.
+	2 + 2 =3D 4 and 2 x 2 =3D 4.
+</pre><p>The forethought form of <a href=3D#e17d3>Example 17.3 </a>is:
+<p>
+<pre><a name=3De17d4>17.4)	li re ge su'i gi pi'i re du li vo
+	The-number two both plus and times two equals the-number four.
+	Both 2 + 2 =3D 4 and 2 x 2 =3D 4.
+</pre><cx "mathematical intervals">  XE "mathematical intervals"   <lx "bi=
'i">  XE "bi'i"   <lx "ga'o">  XE "ga'o"   <lx "ke'i">  XE "ke'i"   Non-log=
ical connection with joiks or joigiks is also permitted between operands an=
d between operators. One use for this construct is to connect operands with=
 =93bi'i=94 to create mathematical intervals:
+<p>
+<ex "zero to one">  XE "zero to one: example=93  <pre><a name=3De17d5>17.5=
)	li no ga'o bi'i ke'i pa
+	the-number zero (inclusive) from-to (exclusive) one
+	[0,1)
+	the numbers from zero to one, including zero but not including one
+</pre><cx "compound subscript">  XE "compound subscript"   <lx "ce'o">  XE=
 "ce'o"   You can also combine two operands with =93ce'o=94, the sequence c=
onnective of selma'o JOI, to make a compound subscript:
+<p>
+<ex "x{b,d}">  XE "x{b:d}, example=93  <pre><a name=3De17d6>17.6)	xy. boi =
xi vei by. ce'o dy. [ve'o]
+	=93x=94 sub (=93b=94 sequence =93d=94 )
+	x<sub>b,d</sub>
+</pre><lx "boi">  XE "boi"   Note that the =93boi=94 in <a href=3D#e17d6>E=
xample 17.6 </a>is not elidable, because the =93xi=94 subscript needs somet=
hing to attach to.
+<p>
+<a name=3Ds18><h3>Tenses, modals, and logical connection</h3>
+<p>
+<lx "PU">  XE "PU selma=92o"   The tense and modal systems of Lojban inter=
act with the logical connective system. No one chapter can explain all of t=
hese simultaneously, so each chapter must present its own view of the area =
of interaction with emphasis on its own concepts and terminology. In the ex=
amples of this chapter, the many tenses of various selma'o as well as the m=
odals of selma'o BAI are represented by the simple time cmavo =93pu=94, =93=
ca=94, and =93ba=94 (of selma'o PU) representing the past, the present, and=
 the future respectively. Preceding a selbri, these cmavo state the time wh=
en the bridi was, is, or will be true (analogous to English verb tenses); p=
receding a sumti, they state that the event of the main bridi is before, si=
multaneous with, or after the event given by the sumti (which is generally =
a =93le nu=94 abstraction; see <a href=3Dchap11.html>Chapter 11</a>).
+<p>
+<cx "logical connection, interaction with tenses">  XE "logical connection=
: interaction with tenses"   <cx "logically connected tenses, definition"> =
 XE "logically connected tenses: definition"   The two types of interaction=
 between tenses and logical connectives are logically connected tenses and =
tensed logical connections. The former are fairly simple. Jeks may be used =
between tense cmavo to specify two connected bridi that differ only in tens=
e:
+<p>
+<ex "once and future king">  XE "once and future king: example=93  <pre><a=
 name=3De18d1>18.1)	la .artr. pu nolraitru
+		.ije la .artr. ba nolraitru
+	Arthur [past] is-a-noblest-governor.
+		And Arthur [future] is-a-noblest-governor.
+	Arthur was a king, and Arthur will be a king.
+</pre>can be reduced to:
+<p>
+<pre><a name=3De18d2>18.2)	la .artr. pu je ba nolraitru
+	Arthur [past] and [future] is-a-noblest-governor.
+	Arthur was and will be king.
+</pre><a href=3D#e18d1>Example 18.1 </a>and <a href=3D#e18d2>Example 18.2 =
</a>are equivalent in meaning; neither says anything about whether Arthur i=
s king now.
+<p>
+<cx "non-logically connected tenses">  XE "non-logically connected tenses"=
   Non-logical connection with joiks is also possible between tenses:
+<p>
+<ex "breathe">  XE "breathe: example=93  <pre><a name=3De18d3>18.3)	mi pu =
bi'o ba vasxu
+	I [past] from=A0<dots>=85</dots>to [future] breathe.
+	I breathe from a past time until a future time.
+</pre><p>The full tense system makes more interesting tense intervals expr=
essible, such as =93from a medium time ago until a long time from now=94.
+<p>
+<cx "forethought connection, in tenses">  XE "forethought connection: in t=
enses"   <cx "tenses, forethought connection in">  XE =93tense: forethought=
 connection in"   <cx "grouping, of connection in tenses">  XE "grouping: o=
f connection in tenses"   <cx "tenses, grouping of connectives in">  XE =93=
tense: grouping of connectives in"   No forethought connections between ten=
ses are permitted by the grammar, nor is there any way to override the defa=
ult left-grouping rule; these limitations are imposed to keep the tense gra=
mmar simpler. Whatever can be said with tenses or modals can be said with s=
ubordinate bridi stating the time, place, or mode explicitly, so it is reas=
onable to try to remove at least some complications.
+<p>
+<cx "tensed logical connection">  XE "tensed logical connection"   Tensed =
logical connections are both more complex and more important than logical c=
onnections between tenses. Consider the English sentence:
+<p>
+<ex "went and bought">  XE "went and bought: example=93  <pre><a name=3De1=
8d4>18.4)	I went to the market, and I bought food.
+</pre><p>The verbatim translation of <a href=3D#e18d4>Example 18.4</a>, na=
mely:
+<p>
+<pre><a name=3De18d5>18.5)	mi pu klama le zarci .ije mi pu tervecnu lo cid=
ja
+	I [past] go-to the market. And I [past] buy items-of food.
+</pre>fails to fully represent a feature of the English, namely that the b=
uying came after the going. (It also fails to represent that the buying was=
 a consequence of the going, which can be expressed by a modal that is disc=
ussed in <a href=3Dchap9.html>Chapter 9</a>.) However, the tense informatio=
n =97 that the event of my going to the market preceded the event of my buy=
ing food =97 can be added to the logical connective as follows. The =93.ije=
=94 is replaced by =93.ijebo=94, and the tense cmavo =93ba=94 is inserted b=
etween =93.ije=94 and =93bo=94:
+<p>
+<lx ".ijebabo">  XE "ijebabo"   <lx "ba">  XE "ba"   <pre><a name=3De18d6>=
18.6)	mi pu klama le zarci  .ijebabo mi pu tervecnu lo cidja
+	I [past] go-to the market.  And [later] I [past] buy items-of food.
+</pre><p>Here the =93pu=94 cmavo in the two bridi-tails express the time o=
f both actions with respect to the speaker: in the past. The =93ba=94 relat=
es the two items to one another: the second item is later than the first it=
em. The grammar does not permit omitting the =93bo=94; if it were omitted, =
the =93ba=94 and the second =93pu=94 would run together to form a compound =
tense =93bapu=94 applying to the second bridi-tail only.
+<p>
+<lx "A">  XE =93A selma=92o"   <lx "BO">  XE "BO selma=92o"   <cx "tensed =
logical connective, in ek=85bo">  XE "tensed logical connective(s): in ek=
=85bo"   Adding tense or modal information to a logical connective is permi=
tted only in the following situations:
+<p>
+Between an ek (or joik) and =93bo=94, as in:
+<p>
+<ex "simultaneously">  XE "simultaneously: example=93  <pre><a name=3De18d=
7>18.7)	la .djan .ecabo la .alis. klama le zarci
+	John and [simultaneous] Alice go-to the market.
+	John and Alice go to the market simultaneously.
+</pre><lx "A">  XE =93A selma=92o"   <lx "KE">  XE "KE selma=92o"   <cx "t=
ensed logical connective, in ek=85ke">  XE "tensed logical connective(s): i=
n ek=85ke"   <lx "JOI">  XE "JOI selma=92o"   <lx "KE">  XE "KE selma=92o" =
  <cx "tensed logical connective, in joik=85ke">  XE "tensed logical connec=
tive(s): in joik=85ke"   Between an ek (or joik) and =93ke=94, as in:
+<p>
+<ex "and earlier">  XE "and earlier: example=93  <pre><a name=3De18d8>18.8=
)	mi dzukla le zarci .epuke le zdani .a le ckule [ke'e]
+	I walk-to the market and [earlier] ( the house or the school ).
+	I walk to the market and, before that, to the house or the school.
+</pre><lx "GIhA">  XE "GIhA selma=92o"   <lx "BO">  XE "BO selma=92o"   <c=
x "tensed logical connective, in gihek=85bo">  XE "tensed logical connectiv=
e(s): in gihek=85bo"   Between a gihek and =93bo=94, as in:
+<p>
+<ex "and then"><pre><a name=3De18d9>18.9)	mi dunda le cukta gi'ebabo lebna=
 lo rupnu vau do
+	I give the book and [later] take some currency-units from/to you.
+	I give you the book and then take some dollars (pounds, yen) from you.
+</pre><lx "GIhA">  XE "GIhA selma=92o"   <lx "KE">  XE "KE selma=92o"   <c=
x "tensed logical connective, in gihek=85ke">  XE "tensed logical connectiv=
e(s): in gihek=85ke"   Between a gihek and =93ke=94, as in:
+<p>
+<ex "and simultaneously">  XE "and simultaneously: example=93  <pre><a nam=
e=3De18d10>18.10)	mi dzukla le zarci gi'ecake cusku zo'e la djan. [ke'e]
+	I walk-to the market and [simultaneous] express something to-John.
+	I walk to the market and at the same time talk to John.
+</pre><lx "JA">  XE "JA selma=92o"   <lx "I">  XE "I selma=92o"   <lx "BO"=
>  XE "BO selma=92o"   <cx "tensed logical connective, in ijek=85bo">  XE "=
tensed logical connective(s): in ijek=85bo"   <cx "tensed logical connectiv=
e, in ijoik=85bo">  XE "tensed logical connective(s): in ijoik=85bo"   Betw=
een an ijek (or ijoik) and =93bo=94, as in:
+<p>
+<ex "and then">  XE "and then: example=93  <pre><a name=3De18d11>18.11)	mi=
 viska pa nanmu  .ijebabo mi viska pa ninmu
+	I see a man.  And [later] I see a woman.
+	I see a man, and then I see a woman.
+</pre><lx "JA">  XE "JA selma=92o"   <lx "JOI">  XE "JOI selma=92o"   <lx =
"I">  XE "I selma=92o"   <lx "TUhE">  XE "TUhE selma=92o"   <cx "tensed log=
ical connective, in ijek=85tu'e">  XE "tensed logical connective(s): in ije=
k=85tu'e"   <cx "tensed logical connective, in ijoik=85tu'e">  XE "tensed l=
ogical connective(s): in ijoik=85tu'e"   Between an ijek (or ijoik) and =93=
tu'e=94, as in:
+<p>
+<ex "and then">  XE "and then: example=93  <pre><a name=3De18d12>18.12)	mi=
 viska pa nanmu  .ijebatu'e mi viska pa ninmu [tu'u]
+	I see a man.  And [later] I see a woman.
+	I see a man, and then I see a woman.
+</pre><lx "JOI">  XE "JOI selma=92o"   <lx "JA">  XE "JA selma=92o"   <lx =
"BO">  XE "BO selma=92o"   <cx "tensed logical connective, in jek=85bo">  X=
E "tensed logical connective(s): in jek=85bo"   <cx "tensed logical connect=
ive, in joik=85bo">  XE "tensed logical connective(s): in joik=85bo"   And =
finally, between a jek (or joik) and =93bo=94, as in:
+<p>
+<ex "doctor and then rich">  XE "doctor and then rich: example=93  <pre><a=
 name=3De18d13>18.13)	mi mikce jebabo ricfu
+	I-am-a doctor and [later] rich
+	I am a doctor and future rich person.
+</pre><lx "BO">  XE "BO selma=92o"   <lx "KE">  XE "KE selma=92o"   <lx "T=
UhE">  XE "TUhE selma=92o"   <cx "bo, contrasted with ke for tensed logical=
 connection">  XE "bo: contrasted with ke for tensed logical connection"   =
<cx "ke, contrasted with bo for tensed logical connection">  XE "ke: contra=
sted with bo for tensed logical connection"   <cx "bo, contrasted with tu'e=
 for tensed logical connection">  XE "bo: contrasted with tu'e for tensed l=
ogical connection"   <cx "tu'e, contrasted with bo for tensed logical conne=
ction">  XE "tu'e: contrasted with bo for tensed logical connection"   <lx =
"bo">  XE "bo"   <lx "ke">  XE "ke"   <lx "tu'e">  XE "tu'e"   As can be se=
en from <a href=3D#e18d11>Example 18.11 </a>and <a href=3D#e18d12>Example 1=
8.12</a>, the choice between =93bo=94 and =93ke=94 (or =93tu'e=94) is arbit=
rary when there are only two things to be connected. If there were no tense=
 information to include, of course neither would be required; it is only th=
e rule that tense information must always be sandwiched between the logical=
 connective and a following =93bo=94, =93ke=94, or =93tu'e=94 that requires=
 the use of one of these grouping cmavo in <a href=3D#e18d7>Example 18.7 </=
a>and <a href=3D#e18d{9}>Examples 18.9 </a>through <a href=3D#e18d13>18.13<=
/a>.
+<p>
+<cx "non-logical connectives, including tense">  XE "non-logical connectiv=
es: including tense"   <cx "tensed non-logical connectives">  XE "tensed no=
n-logical connectives"   <cx "forethought connectives, with tense">  XE "fo=
rethought connectives: with tense"   <cx "tensed logical connectives, foret=
hought">  XE "tensed logical connective: forethought"   <cx "tensed non-log=
ical connectives, forethought">  XE "tensed non-logical connectives: foreth=
ought"   <cx "tensed connectives, in mathematical expressions">  XE "tensed=
 connectives: in mathematical expressions"   <cx "mathematical expressions,=
 tensed connection in">  XE "mathematical expressions: tensed connection in=
"   Non-logical connectives with =93bo=94 and =93ke=94 can include tense in=
formation in exactly the same way as logical connectives. Forethought conne=
ctives, however (except as noted below) are unable to do so, as are termset=
s or tense connectives. Mathematical operands and operators can also includ=
e tense information in their logical connectives as a result of their close=
 parallelism with sumti and tanru components respectively:
+<p>
+<pre><a name=3De18d14>18.14)	vei ci .ebabo vo [ve'o] tadni cu zvati le kum=
fa
+	( 3 and [future] 4 ) students are-at the room.
+	Three and, later, four students were in the room.
+</pre>is a simple example. <cx "tense, in forethought bridi-tail connectio=
n, special rule">  XE "tense: in forethought bridi-tail connection, special=
 rule"   <cx "forethought bridi-tail connection, special rule for tense">  =
XE "forethought bridi-tail connection: special rule for tense"   There is a=
 special grammatical rule for use when a tense applies to both of the selbr=
i in a forethought bridi-tail connection: the entire forethought constructi=
on can just be preceded by a tense. For example:
+<p>
+<ex "went and bought">  XE "went and bought: example=93  <lx "pu ge">  XE =
"pu ge"   <pre><a name=3De18d15>18.15)	mi pu ge klama le zarci gi tervecnu =
lo cidja
+	I [past] both go-to the market and buy some food
+	I went to the market and bought some food.
+</pre><a href=3D#e18d15>Example 18.15 </a>is similar to <a href=3D#e18d5>E=
xample 18.5</a>. There is no time relationship specified between the going =
and the buying; both are simply set in the past.
+<p>
+<a name=3Ds19><h3>Abstractor connection and connection within abstractions=
</h3>
+<p>
+<cx "logical connection of abstractors"><cx "abstraction, logical connecti=
on of">  XE "abstraction(s): logical connection of"   <cx "jeks, connecting=
 abstractors">  XE "jeks: connecting abstractors"   <lx "NU">  XE "NU selma=
=92o"   <lx "JA">  XE "JA selma=92o"   Last and (as a matter of fact) least=
: a logical connective is allowed between abstraction markers of selma'o NU=
. As usual, the connection can be expanded to a bridi connection between tw=
o bridi which differ only in abstraction marker. Jeks are the appropriate c=
onnective. <a href=3D#e19d1>Example 19.1 </a>and <a href=3D#e19d2>Example 1=
9.2 </a>are equivalent in meaning:
+<p>
+<ex "quality and quantity">  XE "quality and quantity: example=93  <pre><a=
 name=3De19d1>19.1)	le ka la frank. ciska cu xlali  .ije le ni la frank. ci=
ska=20
+cu xlali
+	The quality-of Frank's writing is bad, and the quantity of Frank's writin=
g=20
+is bad.
+<a name=3De19d2>19.2)	le ka je ni la frank. ciska cu xlali
+	The quality and quantity of Frank's writing is bad.
+</pre><cx "forethought connection, in abstractions">  XE "forethought conn=
ection: in abstractions"   <cx "abstractions, forethought connection in">  =
XE "abstraction(s): forethought connection in"   <cx "grouping, of connecti=
on in abstractions">  XE "grouping: of connection in abstractions"   <cx "a=
bstractions, grouping of connectives in">  XE "abstraction(s): grouping of =
connectives in"   As with tenses and modals, there is no forethought and no=
 way to override the left-grouping rule.
+<p>
+<cx "logical connection in abstractions, inner bridi contrasted with outer=
 bridi">  XE =93logical connection: in abstraction(s), inner bridi contrast=
ed with outer bridi"   <cx "logical connection inside an abstraction, contr=
asted with outside">  XE =93logical connection: inside an abstraction(s), c=
ontrasted with outside"   Logical connectives and abstraction are related i=
n another way as well, though. Since an abstraction contains a bridi, the b=
ridi may have a logical connection inside it. Is it legitimate to split the=
 outer bridi into two, joined by the logical connection? Absolutely not. Fo=
r example:
+<p>
+<ex "Jupiter life">  XE "Jupiter life: example=93  <pre><a name=3De19d3>19=
.3)	mi jinvi le du'u loi jmive cu zvati gi'onai na zvati vau la .iupiter.
+	I opine the fact-that a-mass-of living-things (is-at or-else isn't-at) Ju=
piter.
+	I believe there either is or isn't life on Jupiter.
+</pre>is true, since the embedded sentence is a tautology, but:
+<p>
+<pre><a name=3De19d4>19.4)	mi jinvi le du'u loi jmive cu zvati la .iupiter=
.
+		.ijonai mi jinvi le du'u loi jmive cu zvati la .iupiter.
+	I opine the fact-that a-mass-of living-things is-at Jupiter
+		or-else I opine the fact-that a-mass-of living-things isn't-at Jupiter
+</pre>is false, since I have no evidence one way or the other (=93jinvi=94=
 requires some sort of evidence, real or fancied, unlike =93krici=94).
+<p>
+<a name=3Ds20><h3>Constructs and appropriate connectives</h3>
+<p>
+<cx "connectives, table by constructs connected">  XE "connectives: table =
by constructs connected"   The following table specifies, for each kind of =
construct that can be logically or non-logically connected in Lojban, what =
kind of connective is required for both afterthought and (when possible) fo=
rethought modes. An asterisk (*) indicates that tensed connection is permit=
ted.
+<p>
+A dash indicates that connection of the specified type is not possible.
+<p>
+<pre>construct	afterthought	forethought	afterthought	forethought
+	      	logical	logical	non-logical	non-logical
+bridi		ijek*	gek	ijoik*	joigik
+sumti		ek*	gek	joik*	joigik
+bridi-tails	gihek*	gek	=97	joigik
+termsets	ek*	gek	joik*	joigik
+tanru parts	jek	guhek	joik*	=97
+operands	ek*	gek	joik*	joigik
+operators	jek	guhek	joik	=97
+tenses/modals	jek	=97	joik	=97
+abstractors	jek	=97	joik	=97
+</pre><a name=3Ds21><h3>Truth functions and corresponding logical connecti=
ves</h3>
+<p>
+<cx "truth functions, table of logical connectives">  XE "truth functions:=
 table of logical connectives"   <cx "logical connectives, table by truth f=
unction value">  XE "logical connectives: table by truth function value"   =
The following table specifies, for each truth function, the most-often used=
 cmavo or compound cmavo which expresses it for each of the six types of lo=
gical connective. (Other compound cmavo are often possible: for example, =
=93se.a=94 means the same as =93a=94, and could be used instead.)
+<p>
+<pre>truth	ek	jek	gihek	gek=96gik	guhek=96gik
+TTTF	.a	ja	gi'a	ga=96gi	gu'a=96gi=20
+TTFT	.anai	janai	gi'anai	ga=96ginai	gu'a=96ginai=20
+TTFF	.u	ju	gi'u	gu=96gi	gu'u=96gi=20
+TFTT	na.a	naja	nagi'a	ganai=96gi	gu'anai=96gi=20
+TFTF	se.u	seju	segi'u	segu=96gi	segu'u=96gi=20
+TFFT	.o	jo	gi'o	go=96gi	gu'o=96gi=20
+TFFF	.e	je	gi'e	ge=96gi	gu'e=96gi=20
+FTTT	na.anai	najanai	nagi'anai	ganai=96ginai	gu'anai=96ginai=20
+FTTF	.onai	jonai	gi'onai	go=96ginai	gu'o=96ginai=20
+FTFT	se.unai	sejunai	segi'unai	segu=96ginai	segu'u=96ginai=20
+FTFF	.enai	jenai	gi'enai	ge=96ginai	gu'e=96ginai=20
+FFTT	na.u	naju	nagi'u	gunai=96gi	gu'unai=96gi=20
+FFTF	na.e	naje	nagi'e	genai=96gi	gu'enai=96gi=20
+FFFT	na.enai	najenai	nagi'enai	genai=96ginai	gu'enai=96ginai
+</pre><p>Note: Ijeks are exactly the same as the corresponding jeks, excep=
t for the prefixed =93.i=94.
+<p>
+<a name=3Ds22><h3>Rules for making logical and non-logical connectives</h3=
>
+<p>
+<cx "logical connectives, syntax rules summary">  XE "logical connectives:=
 syntax rules summary"   <cx "non-logical connectives, syntax rules summary=
">  XE "non-logical connectives: syntax rules summary"   The full set of ru=
les for inserting =93na=94, =93se=94, and =93nai=94 into any connective is:
+<p>
+Afterthought logical connectives (eks, jeks, giheks, ijeks):
+<p>
+Negate first construct:<dl compact><dt>
+		<dd>Place =93na=94 before the connective cmavo (but after the =93.i=94 o=
f an ijek).
+
+</dl><p>Negate second construct:<dl compact><dt>
+		<dd>Place =93nai=94 after the connective cmavo.
+</dl><p>Exchange constructs: <dl compact><dt>
+		<dd>Place =93se=94 before the connective cmavo (after =93na=94 if any).
+</dl><p>Forethought logical connectives (geks, guheks):
+<p>
+Negate first construct:
+		<dl compact><dt><dd>Place =93nai=94 after the connective cmavo.
+</dl><p>Negate second construct:
+		<dl compact><dt><dd>Place =93nai=94 after the =93gi=94.
+</dl><p>Exchange constructs:
+		<dl compact><dt><dd>Place =93se=94 before the connective cmavo.
+</dl><p>Non-logical connectives (joiks, joigiks):
+<p>
+Negate connection:
+		<dl compact><dt><dd>Place =93nai=94 after the connective cmavo (but befo=
re the =93gi=94 of a joigik).
+</dl><p>Exchange constructs:
+		<dl compact><dt><dd>Place =93se=94 before the connective cmavo.
+</dl><a name=3Ds23><h3>Locations of other tables</h3>
+<p>
+<a href=3D#s1>Section 1</a>:  a table explaining the meaning of each truth=
 function in English.
+<p>
+<a href=3D#s2>Section 2</a>:  a table relating the truth functions to the =
four basic vowels.
+<p>
+<a href=3D#s13>Section 13</a>:  a table of the connective question cmavo.
+<p>
+<a href=3D#s14>Section 14</a>:  a table of the meanings of JOI cmavo when =
used to connect sumti.
+
+<p>
+</body></html>
+<center><p align=3Dcenter>
+<img src=3Dchap15.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+
+<h2>Chapter 15
+<br>
+=93No=94 Problems: On Lojban Negation</h2>
+<h6>$Revision: 4.0 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>Introductory</h3>
+<p>
+The grammatical expression of negation is a critical part of Lojban's clai=
m to being logical. The problem of negation, simply put, is to come up with=
 a complete definition of the word =93not=94. For Lojban's unambiguous gram=
mar, this means further that meanings of =93not=94 with different grammatic=
al effect must be different words, and even different grammatical structure=
s.
+<p>
+Logical assertions are implicitly required in a logical language; thus, an=
 apparatus for expressing them is built into Lojban's logical connectives a=
nd other structures.
+<p>
+In natural languages, especially those of Indo-European grammar, we have s=
entences composed of two parts which are typically called =93subject=94 and=
 =93predicate=94. In the statement
+<p>
+<pre><a name=3De1d1>1.1)	John goes to the store
+</pre>=93John=94 is the subject, and =93goes to the store=94 is the predic=
ate. Negating <a href=3D#e1d1>Example 1.1 </a>to produce
+<p>
+<pre><a name=3De1d2>1.2)	John doesn't go to the store.
+</pre>has the effect of declaring that the predicate does not hold for the=
 subject. <a href=3D#e1d2>Example 1.2 </a>says nothing about whether John g=
oes somewhere else, or whether someone else besides John goes to the store.
+<p>
+We will call this kind of negation =93natural language negation=94. This k=
ind of negation is difficult to manipulate by the tools of logic, because i=
t doesn't always follow the rules of logic. Logical negation is bi-polar: e=
ither a statement is true, or it is false. If a statement is false, then it=
s negation must be true. Such negation is termed contradictory negation.
+<p>
+Let's look at some examples of how natural language negation can violate t=
he rules of contradictory negation.
+<p>
+<pre><a name=3De1d3>1.3)	Some animals are not white.
+<a name=3De1d4>1.4)	Some animals are white.
+ </pre><p>Both of these statements are true; yet one is apparently the neg=
ation of the other. Another example:
+<p>
+<pre><a name=3De1d5>1.5)	I mustn't go to the dance.
+<a name=3De1d6>1.6)	I must go to the dance.
+</pre><p>At first thought, <a href=3D#e1d5>Example 1.5 </a>negates <a href=
=3D#e1d6>Example 1.6</a>. Thinking further, we realize that there is an int=
ermediate state wherein I am permitted to go to the dance, but not obligate=
d to do so. Thus, it is possible that both statements are false.
+
+<p>
+Sometimes order is significant:
+<p>
+<pre><a name=3De1d7>1.7)	The falling rock didn't kill Sam.
+<a name=3De1d8>1.8)	Sam wasn't killed by the falling rock.
+</pre><p>Our minds play tricks on us with this one. Because <a href=3D#e1d=
7>Example 1.7 </a>is written in what is called the =93active voice=94, we i=
mmediately get confused about whether =93the falling rock=94 is a suitable =
subject for the predicate =93did kill Sam=94. =93Kill=94 implies volition t=
o us, and rocks do not have volition. This confusion is employed by opponen=
ts of gun control who use the argument =93Guns don't kill people; people ki=
ll people.=94
+<p>
+Somehow, we don't have the same problem with <a href=3D#e1d8>Example 1.8</=
a>. The subject is Sam, and we determine the truth or falsity of the statem=
ent by whether he was or wasn't killed by the falling rock.
+<p>
+<a href=3D#e1d8>Example 1.8 </a>also helps us focus on the fact that there=
 are at least two questionable facts implicit in this sentence: whether Sam=
 was killed, and if so, whether the falling rock killed him. If Sam wasn't =
killed, the question of what killed him is moot.
+<p>
+This type of problem becomes more evident when the subject of the sentence=
 turns out not to exist:
+<p>
+<pre><a name=3De1d9>1.9)	The King of Mexico didn't come to dinner.
+<a name=3De1d10>1.10)	The King of Mexico did come to dinner.
+</pre><p>In the natural languages, we would be inclined to say that both o=
f these statements are false, since there is no King of Mexico.
+<p>
+The rest of this chapter is designed to explain the Lojban model of negati=
on.
+<p>
+<a name=3Ds2><h3>bridi negation</h3>
+<p>
+In discussing Lojban negation, we will call the form of logical negation t=
hat simply denies the truth of a statement =93bridi negation=94. Using brid=
i negation, we can say the equivalent of =93I haven't stopped beating my wi=
fe=94 without implying that I ever started, nor even that I have a wife, me=
aning simply =93It isn't true that I have stopped beating my wife.=94 Since=
 Lojban uses bridi as smaller components of complex sentences, bridi negati=
on is permitted in these components as well at the sentence level.
+<p>
+For the bridi negation of a sentence to be true, the sentence being negate=
d must be false. A major use of bridi negation is in making a negative resp=
onse to a yes/no question; such responses are usually contradictory, denyin=
g the truth of the entire sentence. A negative answer to
+<p>
+<pre><a name=3De2d1>2.1)	Did you go to the store?
+</pre>is taken as a negation of the entire sentence, equivalent to
+<p>
+<pre><a name=3De2d2>2.2)	No, I didn't go to the store.
+</pre><p>The most important rule about bridi negation is that if a bridi i=
s true, its negation is false, and vice versa.
+<p>
+The simplest way to express a bridi negation is to use the cmavo =93na=94 =
of selma'o NA before the selbri of the affirmative form of the bridi (but a=
fter the =93cu=94, if there is one):
+<p>
+<pre><a name=3De2d3>2.3)	mi klama le zarci
+	I go-to the store.
+</pre>when negated becomes:
+<p>
+<pre><a name=3De2d4>2.4)	mi na klama le zarci
+	I [false] go-to the store.
+</pre><p>Note that we have used a special convention to show in the Englis=
h that a bridi negation is present. We would like to use the word =93not=94=
, because this highlights the naturalness of putting the negation marker ju=
st before the selbri, and makes the form easier to learn. But there is a ma=
jor difference between Lojban's bridi negation with =93na=94 and natural la=
nguage negation with =93not=94. In English, the word =93not=94 can apply to=
 a single word, to a phrase, to an English predicate, or to the entire sent=
ence. In addition, =93not=94 may indicate either contradictory negation or =
another form of negation, depending on the sentence. Lojban's internal brid=
i negation, on the other hand, always applies to an entire bridi, and is al=
ways a contradictory negation; that is, it contradicts the claim of the who=
le bridi.
+<p>
+Because of the ambiguity of English =93not=94, we will use =93[false]=94 i=
n the translation of Lojban examples to remind the reader that we are expre=
ssing a contradictory negation. Here are more examples of bridi negation:
+<p>
+<pre><a name=3De2d5>2.5)	mi [cu] na ca klama le zarci
+	I [false] now am-a-go-er to the market.
+	I am not going to the market now.
+<a name=3De2d6>2.6)	lo ca nolraitru be le fasygu'e=20
+cu na krecau
+	The-actual present noblest-governor of the French country=20
+[false] is-hair-without.
+	The current king of France isn't bald.
+<a name=3De2d7>2.7)	ti na barda prenu co melbi mi
+	This [false] is a big-person of-type (beautiful to me).
+	This isn't a big person who is beautiful to me.
+</pre><p>Although there is this fundamental difference between Lojban's in=
ternal bridi negation and English negation, we note that in many cases, esp=
ecially when there are no existential or quantified variables (the cmavo =
=93da=94, =93de=94, and =93di=94 of selma'o KOhA, explained in <a href=3Dch=
ap16.html>Chapter 16</a>) in the bridi, you can indeed translate Lojban =93=
na=94 as =93not=94 (or =93isn't=94 or =93doesn't=94, as appropriate).
+<p>
+The most important rule about bridi negation is that if a bridi is true, i=
ts negation is false, and vice versa.
+<p>
+In Lojban, there are several structures that implicitly contain bridi, so =
that Lojban sentences may contain more than one occurrence of =93na=94. For=
 example:
+<p>
+<pre><a name=3De2d8>2.8)	mi na gleki le nu
+		na klama le nu dansu
+	I [false] am-happy-about the event-of
+([false] going-to the event-of dancing).
+	It is not the case that I am happy about it not being
+		the case that I am going to the dance.
+	I am not happy about not going to the dance.
+In the previous example, we used internal negations in abstraction bridi; =
bridi negation may also be found in descriptions within sumti. For example:
+<a name=3De2d9>2.9)	mi nelci le na melbi
+	I am fond of the-one-described-as ([false] beautiful)
+	I am fond of the one who isn't beautiful.
+</pre><p>A more extreme (and more indefinite) example is:
+<p>
+<pre><a name=3De2d10>2.10)	mi nelci lo na ca nolraitru be le frasygu'e
+	I am-fond-of one-who-is ([false] the current king of the French-country).
+	I am fond of one who isn't the current king of France.
+</pre><p>The claim of <a href=3D#e2d10>Example 2.10 </a>could apply to any=
one except a person who is fond of no one at all, since the relation within=
 the description is false for everyone. You cannot readily express these si=
tuations in colloquial English.
+<p>
+Negation with =93na=94 applies to an entire bridi, and not to just part of=
 a selbri. Therefore, you won't likely have reason to put =93na=94 inside a=
 tanru. In fact, the grammar currently does not allow you to do so (except =
in a lujvo and in elaborate constructs involving GUhA, the forethought conn=
ector for selbri). Any situation where you might want to do so can be expre=
ssed in a less-compressed non-tanru form. This grammatical restriction help=
s ensure that bridi negation is kept separate from other forms of negation.
+<p>
+The grammar of =93na=94 allows multiple adjacent negations, which cancel o=
ut, as in normal logic:
+<p>
+<pre><a name=3De2d11>2.11)	ti na na barda prenu co melbi mi
+	This [false] [false] is-a-big person that is (beautiful to me).
+</pre>which is the same as:
+<p>
+<pre><a name=3De2d12>2.12)	ti barda prenu co melbi mi
+	This is a big-person that is (beautiful to me).
+</pre><p>When a selbri is tagged with a tense or a modal, negation with =
=93na=94 is permitted in two positions: before or after the tag. No semanti=
c difference between these forms has yet been defined, but this is not fina=
lly determined, since the interactions between tenses/modals and bridi nega=
tion have not been fully explored. In particular, it remains to be seen whe=
ther sentences using less familiar tenses, such as:
+<p>
+<pre><a name=3De2d13>2.13)	mi [cu] ta'e klama le zarci
+	I habitually go to the market.
+</pre>mean the same thing with =93na=94 before the =93ta'e=94, as when the=
 negation occurs afterwards; we'll let future, Lojban-speaking, logicians d=
ecide on how they relate to each other.
+<p>
+A final caution on translating English negations into Lojban: if you trans=
late the English literally, you'll get the wrong one. With English causal s=
tatements, and other statements with auxiliary clauses, this problem is mor=
e likely.
+<p>
+Thus, if you translate the English:
+<p>
+<pre><a name=3De2d14>2.14)	I do not go to the market because the car is br=
oken.
+</pre>as:
+<p>
+<pre><a name=3De2d15>2.15)	mi na klama le zarci ki'u lenu le karce cu spof=
u
+	I [false] go-to the market because the car is broken.
+	It is false that:  =93I go to the market because the car is broken.=94
+</pre>you end up negating too much.
+<p>
+Such mistranslations result from the ambiguity of English compounded by th=
e messiness of natural language negation. A correct translation of the norm=
al interpretation of <a href=3D#e2d14>Example 2.14 </a>is:
+<p>
+<pre><a name=3De2d16>2.16)	lenu mi na klama le zarci cu se krinu
+		lenu le karce cu spofu
+	The event-of (my [false] going-to the market) is justified by
+		the event-of (the car being broken).
+	My not going to the market is because the car is broken.
+</pre><p>In <a href=3D#e2d16>Example 2.16</a>, the negation is clearly con=
fined to the event abstraction in the x1 sumti, and does not extend to the =
whole sentence. The English could also have been expressed by two separate =
sentences joined by a causal connective (which we'll not go into here).
+<p>
+The problem is not confined to obvious causals. In the English:
+<p>
+<pre><a name=3De2d17>2.17)	I was not conscripted into the Army with the he=
lp of my uncle the Senator.
+</pre>we do not intend the uncle's help to be part of the negation. We mus=
t thus move the negation into an event clause or use two separate sentences=
. The event-clause version would look like:
+<p>
+<pre><a name=3De2d18>2.18)	The event-of (my [false] being-conscripted-into=
 the Army) was aided by my uncle the Senator.
+</pre><p>It is possible that someone will want to incorporate bridi negati=
ons into lujvo. For this reason, the rafsi =93-nar-=94 has been reserved fo=
r =93na=94. However, before using this rafsi, make sure that you intend the=
 contradictory bridi negation, and not the scalar negation described in <a =
href=3D#s3>Section 3</a>, which will be much more common in tanru and lujvo=
.
+<p>
+<a name=3Ds3><h3>Scalar negation</h3>
+<p>
+Let us now consider some other types of negation. For example, when we say=
:
+<p>
+<pre><a name=3De3d1>3.1)	The chair is not brown.
+</pre>we make a positive inference =97 that the chair is some other color.=
 Thus, it is legitimate to respond:
+<p>
+<pre><a name=3De3d2>3.2)	It is green.
+</pre><p>Whether we agree that the chair is brown or not, the fact that th=
e statement refers to color has significant effect on how we interpret some=
 responses. If we hear the following exchange:
+<p>
+<pre><a name=3De3d3>3.3)	The chair is not brown.
+	Correct. The chair is wooden.
+</pre>we immediately start to wonder about the unusual wood that isn't bro=
wn. If we hear the exchange:
+<p>
+<pre><a name=3De3d4>3.4)	Is the chair green?
+	No, it is in the kitchen.
+</pre>we are unsettled because the response seems to be a non-sequitur. Bu=
t since it might be true and it is a statement about the chair, one can't s=
ay it is entirely irrelevant!
+<p>
+What is going on in these statements is something called =93scalar negatio=
n=94. As the name suggests, scalar negation presumes an implied scale. A ne=
gation of this type not only states that one scalar value is false, but imp=
lies that another value on the scale must be true. This can easily lead to =
complications. The following exchange seems reasonably natural (a little su=
spension of disbelief in such inane conversation will help):
+<p>
+<pre><a name=3De3d5>3.5)	That isn't a blue house.
+	Right!  That is a green house.
+</pre><p>We have acknowledged a scalar negation by providing a correct val=
ue which is another color in the set of colors permissible for houses. Whil=
e a little less likely, the following exchange is also natural:
+<p>
+<pre><a name=3De3d6>3.6)	That isn't a blue house.
+	Right!  That is a blue car.
+</pre><p>Again, we have acknowledged a scalar negation, and substituted a =
different object in the universe of discourse of things that can be blue.
+<p>
+Now, if the following exchange occurs:
+<p>
+<pre><a name=3De3d7>3.7)	That isn't a blue house.
+	Right!  That is a green car.
+</pre>we find the result unsettling. This is because it seems that two cor=
rections have been applied when there is only one negation. Yet out of cont=
ext, =93blue house=94 and =93green car=94 seem to be reasonably equivalent =
units that should be mutually replaceable in a sentence. It's just that we =
don't have a clear way in English to say:
+<p>
+<pre><a name=3De3d8>3.8)	That isn't a =93blue-house=94.
+</pre>aloud so as to clearly imply that the scalar negation is affecting t=
he pair of words as a single unit.
+<p>
+Another even more confusing example of scalar negation is to the sentence:
+<p>
+<pre><a name=3De3d9>3.9)	John didn't go to Paris from Rome.
+</pre><p>Might <a href=3D#e3d9>Example 3.9 </a>imply that John went to Par=
is from somewhere else? Or did he go somewhere else from Rome? Or perhaps h=
e didn't go anywhere at all: maybe someone else did, or maybe there was no =
event of going whatsoever. One can devise circumstances where any one, two =
or all three of these statements might be inferred by a listener.
+<p>
+In English, we have a clear way of distinguishing scalar negation from pre=
dicate negation that can be used in many situations. We can use the partial=
 word =93non-=94 as a prefix. But this is not always considered good usage,=
 even though it would render many statements much clearer. For example, we =
can clearly distinguish
+<p>
+<pre><a name=3De3d10>3.10)	That is a non-blue house.
+</pre>from the related sentence
+<p>
+<pre><a name=3De3d11>3.11)	That is a blue non-house.
+</pre><a href=3D#e3d10>Example 3.10 </a>and <a href=3D#e3d11>Example 3.11 =
</a>have the advantage that, while they contain a negative indication, they=
 are in fact positive assertions. They say what is true by excluding the fa=
lse; they do not say what is false.
+<p>
+We can't always use =93non-=94 though, because of the peculiarities of Eng=
lish's grammar. It would sound strange to say:
+<p>
+<pre><a name=3De3d12>3.12)	John went to non-Paris from Rome.
+</pre>or
+<p>
+<pre><a name=3De3d13>3.13)	John went to Paris from non-Rome.
+</pre>although these would clarify the vague negation. Another circumlocut=
ion for English scalar negation is =93other than=94, which works where =93n=
on-=94 does not, but is wordier.
+<p>
+Finally, we have natural language negations that are called polar negation=
s, or opposites:
+<p>
+<pre><a name=3De3d14>3.14)	John is moral
+<a name=3De3d15>3.15)	John is immoral
+</pre><p>To be immoral is much more than to just be not moral: it implies =
the opposite condition. Statements like <a href=3D#e3d15>Example 3.15 </a>a=
re strong negations which not only deny the truth of a statement, but asser=
t its opposite. Since, =93opposite=94 implies a scale, polar negations are =
a special variety of scalar negations.
+<p>
+To examine this concept more closely, let us draw a linear scale, showing =
two examples of how the scale is used:
+Affirmations	(positive)
+Negations	(negative)
+
+
+
+
+
+
+
+
+
+
+
+All	Most	Some	Few	None
+Excellent	Good	Fair	Poor	Awful
+<p>
+</pre><p>Some scales are more binary than the examples we diagrammed. Thus=
 we have =93not necessary=94 or =93unnecessary=94 being the polar opposite =
of necessary. Another scale, especially relevant to Lojban, is interpreted =
based on situations modified by one's philosophy: =93not true=94 may be equ=
ated with =93false=94 in a bi-valued truth-functional logic, while in tri-v=
alued logic an intermediate between =93true=94 and =93false=94 is permitted=
, and in fuzzy logic a continuous scale exists from true to false. The mean=
ing of =93not true=94 requires a knowledge of which variety of truth scale =
is being considered.
+<p>
+We will define the most general form of scalar negation as indicating only=
 that the particular point or value in the scale or range is not valid and =
that some other (unspecified) point on the scale is correct. This is the in=
tent expressed in most contexts by =93not mild=94, for example.
+<p>
+Using this paradigm, contradictory negation is less restrictive than scala=
r negation =97 it says that the point or value stated is incorrect (false),=
 and makes no statement about the truth of any other point or value, whethe=
r or not on the scale.
+<p>
+In English, scalar negation semantically includes phrases such as =93other=
 than=94, =93reverse of=94, or =93opposite from=94 expressions and their eq=
uivalents. More commonly, scalar negation is expressed in English by the pr=
efixes =93non-=94, =93un-=94, =93il-=94, and =93im-=94. Just which form and=
 permissible values are implied by a scalar negation is dependent on the se=
mantics of the word or concept which is being negated, and on the context. =
Much confusion in English results from the uncontrolled variations in meani=
ng of these phrases and prefixes.
+<p>
+In the examples of <a href=3D#s4>Section 4</a>, we will translate the gene=
ral case of scalar negation using the general formula =93other than=94 when=
 a phrase is scalar-negated, and =93non-=94 when a single word is scalar-ne=
gated.
+<p>
+<a name=3Ds4><h3>selbri and tanru negation</h3>
+<p>
+All the scalar negations illustrated in <a href=3D#s3>Section 3 </a>are ex=
pressed in Lojban using the cmavo =93na'e=94 (of selma'o NAhE). The most co=
mmon use of =93na'e=94 is as a prefix to the selbri:
+<p>
+<pre><a name=3De4d1>4.1)	mi klama le zarci
+	I go to the market.
+<a name=3De4d2>4.2)	mi na'e klama le zarci
+	I non-go to the market.
+</pre><p>Comparing these two, we see that the negation operator being used=
 in <a href=3D#e4d2>Example 4.2 </a>is =93na'e=94. But what exactly does =
=93na'e=94 negate? Does the negation include only the gismu =93klama=94, wh=
ich is the entire selbri in this case, or does it include the =93le zarci=
=94 as well? In Lojban, the answer is unambiguously =93only the gismu=94. T=
he cmavo =93na'e=94 always applies only to what follows it.
+<p>
+<a href=3D#e4d2>Example 4.2 </a>looks as if it were parallel to:
+<p>
+<pre><a name=3De4d3>4.3)	mi na klama le zarci
+	I [false] go-to the market.
+</pre>but in fact there is no real parallelism at all. A negation using =
=93na=94 denies the truth of a relationship, but a selbri negation with =93=
na'e=94 asserts that a relationship exists other than that stated, one whic=
h specifically involves the sumti identified in the statement. The grammar =
allotted to =93na'e=94 allows us to unambiguously express scalar negations =
in terms of scope, scale, and range within the scale. Before we explain the=
 scalar aspects, let us show how the scope of =93na'e=94 is determined.
+<p>
+In tanru, we may wish to negate an individual element before combining it =
with another to form the tanru. We in effect need a shorter-than-selbri-sco=
pe negation, for which we can use =93na'e=94 as well. The positive sentence
+<p>
+<pre><a name=3De4d4>4.4)	mi cadzu klama le zarci
+	I walking-ly go to the market.
+</pre>can be subjected to selbri negation in several ways. Two are:
+<p>
+<pre><a name=3De4d5>4.5)	mi na'e cadzu klama le zarci
+	I (other-than-walkingly)-go-to the market.
+<a name=3De4d6>4.6)	mi cadzu na'e klama le zarci
+	I walkingly-(other-than-go-to) the market.
+</pre><p>These negations show the default scope of =93na'e=94 is close-bin=
ding on an individual brivla in a tanru. <a href=3D#e4d5>Example 4.5 </a>sa=
ys that I am going to the market, but in some kind of a non-walking manner.=
 (As with most tanru, there are a few other possible interpretations, but w=
e'll assume this one =97 see <a href=3Dchap5.html>Chapter 5 </a>for a discu=
ssion of tanru meaning).
+<p>
+In neither <a href=3D#e4d5>Example 4.5 </a>nor <a href=3D#e4d6>Example 4.6=
 </a>does the =93na'e=94 negate the entire selbri. While both sentences con=
tain negations that deny a particular relationship between the sumti, they =
also have a component which makes a positive claim about such a relationshi=
p. This is clearer in <a href=3D#e4d5>Example 4.5</a>, which says that I am=
 going, but in a non-walking manner. In <a href=3D#e4d6>Example 4.6</a>, we=
 have claimed that the relationship between me and the market in some way i=
nvolves walking, but is not one of =93going to=94 (perhaps we are walking a=
round the market, or walking-in-place while at the market).
+<p>
+The =93scale=94, or actually the =93set=94, implied in Lojban tanru negati=
ons is anything which plausibly can be substituted into the tanru. (Plausib=
ility here is interpreted in the same way that answers to a =93mo=94 questi=
on must be plausible =97 the result must not only have the right number of =
places and have sumti values appropriate to the place structure, it must al=
so be appropriate or relevant to the context.) This minimal condition allow=
s a speaker to be intentionally vague, while still communicating meaningful=
 information. The speaker who uses selbri negation is denying one relations=
hip, while minimally asserting a different relationship.
+<p>
+We also need a scalar negation form that has a scope longer than a single =
brivla. There exists such a longer-scope selbri negation form, as exemplifi=
ed by (each Lojban sentence in the next several examples is given twice, wi=
th parentheses in the second copy showing the scope of the =93na'e=94):
+<p>
+<pre><a name=3De4d7>4.7)	mi na'eke cadzu klama [ke'e] le zarci
+	mi na'e (ke cadzu klama [ke'e]) le zarci
+	I other-than-(walkingly-go-to) the market.
+</pre><p>This negation uses the same =93ke=94 and =93ke'e=94 delimiters (t=
he =93ke'e=94 is always elidable at the end of a selbri) that are used in t=
anru. The sentence clearly negates the entire selbri. The =93ke'e=94, wheth=
er elided or not, reminds us that the negation does not include the trailin=
g sumti. While the trailing-sumti place-structure is defined as that of the=
 final brivla, the trailing sumti themselves are not part of the selbri and=
 are thus not negated by =93na'e=94.
+<p>
+Negations of just part of the selbri are also permitted:
+<p>
+<pre><a name=3De4d8>4.8)	mi na'eke sutra cadzu ke'e klama le zarci
+	mi na'e (ke sutra cadzu ke'e) klama le zarci
+	I other-than-(quickly-walkingly) go-to the market.
+</pre><p>In <a href=3D#e4d8>Example 4.8</a>, only the =93sutra cadzu=94 ta=
nru is negated, so the speaker is indeed going to the market, but not by wa=
lking quickly.
+<p>
+Negations made with =93na'e=94 or =93na'eke=94 also include within their s=
cope any sumti attached to the brivla or tanru with =93be=94 or =93bei=94. =
Such attached sumti are considered part of the brivla or tanru:
+<p>
+<pre><a name=3De4d9>4.9)	mi na'e ke sutra cadzu be le mi birka ke'e klama =
le zarci
+	I other-than-(quickly walking-on-my-arms-ly) go-to the market.
+</pre><p>Note that <a href=3D#e4d10>Example 4.10 </a>and <a href=3D#e4d11>=
Example 4.11 </a>do not express the same thing:
+<p>
+<pre><a name=3De4d10>4.10)	mi na'eke sutra cadzu [ke'e] lemi birka
+	mi na'e (ke sutra cadzu [ke'e]) lemi birka
+	I other-than-(quickly-walk-on) my-arms.
+<a name=3De4d11>4.11)	mi na'eke sutra cadzu be lemi birka [ke'e]
+	mi na'e (ke sutra cadzu be lemi birka [ke'e])
+	I other-than-(quickly-walk-on my-arms).
+</pre><p>The translations show that the negation in <a href=3D#e4d10>Examp=
le 4.10 </a>is more restricted in scope; i.e. less of the sentence is negat=
ed with respect to x1 (=93mi=94).
+<p>
+Logical scope being an important factor in Lojban's claims to be unambiguo=
us, let us indicate the relative precedence of =93na'e=94 as an operator. G=
rouping with =93ke=94 and =93ke'e=94, of course, has an overt scope, which =
is its advantage. =93na'e=94 is very close binding to its brivla. Internal =
binding of tanru, with =93bo=94, is not as tightly bound as =93na'e=94. =93=
co=94, the tanru inversion operator has a scope that is longer than all oth=
er tanru constructs.
+<p>
+In short, =93na'e=94 and =93na'eke=94 define a type of negation, which is =
shorter in scope than bridi negation, and which affects all or part of a se=
lbri. The result of =93na'e=94 negation remains an assertion of some specif=
ic truth and not merely a denial of another claim.
+<p>
+The similarity becomes striking when it is noticed that the rafsi =93-nal-=
=94, representing =93na'e=94 when a tanru is condensed into a lujvo, forms =
an exact parallel to the English usage of =93non-=94. Turning a series of r=
elated negations into lujvo gives:
+<p>
+<pre><a name=3De4d12>4.12)	na'e klama becomes nalkla
+	na'e cadzu klama becomes naldzukla
+	na'e sutra cadzu klama becomes nalsu'adzukla
+	nake sutra cadzu ke'e klama becomes nalsu'adzuke'ekla
+</pre><p>Note: =93-kem-=94 is the rafsi for =93ke=94, but it is omitted in=
 the final lujvo as superfluous =97 =93ke'e=94 is its own rafsi, and its in=
clusion in the lujvo implies a =93ke=94 after the =93-nal-=94, since it nee=
ds to close something; only a =93ke=94 immediately after the negation would=
 make the =93ke'e=94 meaningful in the tanru expressed in this lujvo.
+<p>
+In a lujvo, it is probably clearest to translate =93-nal-=94 as =93non-=94=
, to match the English combining forms, except when the =93na'e=94 has sing=
le word scope and English uses =93un-=94 or =93im-=94 to negate that single=
 word. Translation style should determine the use of =93other than=94, =93n=
on-=94, or another negator for =93na'e=94 in tanru; the translator must ren=
der the Lojban into English so it is clear in context. Let's go back to our=
 simplest example:
+<p>
+<pre><a name=3De4d13>4.13)	mi na'e klama le zarci
+	I am other-than-(going-to) the market.
+	?I am not going-to the market.
+<a name=3De4d14>4.14)	mi nalkla le zarci
+	I am-a-non-go-er-to the market.
+</pre><p>Note that to compare with the English translation form using =93n=
on-=94, we've translated the Lojban as if the selbri were a noun. Since Loj=
ban =93klama=94 is indifferently a noun, verb, or adjective, the difference=
 is purely a translation change, not a true change in meaning. The English =
difference seems significant, though, due to the strongly different English=
 grammatical forms and the ambiguity of English negation.
+<p>
+Consider the following highly problematic sentence:
+<p>
+<pre><a name=3De4d15>4.15)	lo ca nolraitru be le fasygu'e cu krecau
+	An-actual currently noblest-governor of the French country 	is-hair-witho=
ut.
+	The current King of France is bald.
+</pre><p>The selbri =93krecau=94 negates with =93na'e=94 as:
+<p>
+<pre><a name=3De4d16>4.16)	lo ca nolraitru be le fasygu'e
+		cu na'e krecau
+	An-actual currently noblest-governor of the French country
+		is-other-than hair-without.
+	The current King of France is other-than-bald.
+
+</pre>or, as a lujvo:
+<p>
+<pre><a name=3De4d17>4.17)	lo ca nolraitru be le fasygu'e
+		cu nalkrecau
+	An-actual currently noblest-governor of the French country
+		is-non-hair-without.
+	The current King of France is a non-bald-one.
+</pre><a href=3D#e4d16>Example 4.16 </a>and <a href=3D#e4d17>Example 4.17 =
</a>express the predicate negation forms using a negation word (=93na'e=94)=
 or rafsi (=93-nal-=94); yet they make positive assertions about the curren=
t King of France; ie., that he is other-than-bald or non-bald. This follows=
 from the close binding of =93na'e=94 to the brivla. The lujvo form makes t=
his overt by absorbing the negative marker into the word.
+<p>
+Since there is no current King of France, it is false to say that he is ba=
ld, or non-bald, or to make any other affirmative claim about him. Any sent=
ence about the current King of France containing only a selbri negation is =
as false as the sentence without the negation. No amount of selbri negation=
s have any effect on the truth value of the sentence, which is invariably =
=93false=94, since no affirmative statement about the current King of Franc=
e can be true. On the other hand, bridi negation does produce a truth:
+<p>
+<pre><a name=3De4d18>4.18)	lo ca nolraitru be le fasygu'e
+		cu na krecau
+	An-actual current noblest-governor of the French Country=20
+		[false] is-hair-without.
+	It is false that the current King of France is bald.
+</pre><p>Note: =93lo=94 is used in these sentences because negation relate=
s to truth conditions. To meaningfully talk about truth conditions in sente=
nces carrying a description, it must be clear that the description actually=
 applies to the referent. A sentence using =93le=94 instead of =93lo=94 can=
 be true even if there is no current king of France, as long as the speaker=
 and the listener agree to describe something as the current king of France=
. (See the explanations of =93le=94 in <a href=3Dchap6.html>Chapter 6</a>.)
+<p>
+<a name=3Ds5><h3>Expressing scales in selbri negation</h3>
+<p>
+In expressing a scalar negation, we can provide some indication of the sca=
le, range, frame-of-reference, or universe of discourse that is being dealt=
 with in an assertion. As stated in <a href=3D#s4>Section 4</a>, the defaul=
t is the set of plausible alternatives. Thus if we say:
+<p>
+<pre><a name=3De5d1>5.1)	le stizu cu na'e xunre
+	The chair is a non-(red-thing).
+</pre>the pragmatic interpretation is that we mean a different color and n=
ot
+<p>
+<pre><a name=3De5d2>5.2)	le stizu cu dzukla be le zarci
+	The chair walkingly-goes-to-the-market.
+</pre><p>However, if we have reason to be more explicit (an obtuse or cont=
rary listener, or simply an overt logical analysis), we can clarify that we=
 are referring to a color by saying:
+<p>
+<pre><a name=3De5d3>5.3)	le stizu cu na'e xunre skari
+	The chair is of a non-(red)-color (as perceived by something under
+some conditions).
+</pre><p>We might also have reduced the pragmatic ambiguity by making the =
two trailing sumti values explicit (the =93as perceived by=94 and =93under =
conditions=94 places have been added to the place structure of =93xunre=94)=
. But assume we have a really stubborn listener (an artificially semi-intel=
ligent computer?) who will find a way to misinterpret <a href=3D#e5d3>Examp=
le 5.3 </a>even with three specific sumti provided.
+<p>
+In this case, we use a sumti tagged with the sumti tcita =93ci'u=94, which=
 translates roughly as =93on a scale of X=94, where =93X=94 is the sumti. F=
or maximal clarity, the tagged sumti can be bound into the negated selbri w=
ith =93be=94. To clarify <a href=3D#e5d3>Example 5.3</a>, we might say:
+<p>
+<pre><a name=3De5d4>5.4)	le stizu cu na'e xunre be ci'u loka skari
+	The chair is a non-(red on-a-scale-of-colorness)-thing.
+</pre><p>We can alternately use the sumti tcita =93teci'e=94, based on =93=
ciste=94, which translates roughly as =93of a system of components X=94, fo=
r universes of discourse; in this case, we would express <a href=3D#e5d3>Ex=
ample 5.3 </a>as:
+<p>
+<pre><a name=3De5d5>5.5)	le stizu cu na'e xunre be teci'e le skari
+	The chair is a non-(red of-a-system-with-components-the-colors)-thing.
+</pre><p>Other places of =93ciste=94 can be brought out using the grammar =
of selma'o BAI modals, allowing slightly different forms of expression, thu=
s:
+<p>
+<pre><a name=3De5d6>5.6)	le stizu cu na'e xunre be ci'e lo'i skari
+	The chair is a non-(red of-a-system-which-is-the-set-of-colors)-thing.
+</pre><p>The cmavo =93le'a=94, also in selma'o BAI, can be used to specify=
 a category:
+<p>
+<pre><a name=3De5d7>5.7)	le stizu cu na'e xunre be le'a lo'i skari
+	The chair is a non-(red of-a-category-which-is-the-set-of-colors)-thing.
+</pre>which is minimally different in meaning from <a href=3D#e5d6>Example=
 5.6</a>.
+<p>
+The cmavo =93na'e=94 is not the only member of selma'o NAhE. If we want to=
 express a scalar negation which is a polar opposite, we use the cmavo =93t=
o'e=94, which is grammatically equivalent to =93na'e=94:
+<p>
+<pre><a name=3De5d8>5.8)	le stizu cu to'e xunre be ci'u loka skari
+	The chair is a ( opposite-of red ) on-scale a-property-of color-ness.
+</pre><p>Likewise, the midpoint of a scale can be expressed with the cmavo=
 =93no'e=94, also grammatically equivalent to =93na'e=94. Here are some par=
allel examples of =93na'e=94, =93no'e=94, and =93to'e=94:
+<p>
+<pre><a name=3De5d9>5.9)	ta melbi
+	That is-beautiful.
+<a name=3De5d10>5.10)	ta na'e melbi
+	That is-non-beautiful.
+	That is other than beautiful.
+	That is ugly [in one sense].
+<a name=3De5d11>5.11)	ta no'e melbi
+	That is-neutrally beautiful.
+	That is plain/ordinary-looking (neither ugly nor beautiful).
+
+<a name=3De5d12>5.12)	ta to'e melbi
+	That is-opposite-of beautiful.
+	That is ugly/very ugly/repulsive.
+</pre><p>The cmavo =93to'e=94 has the assigned rafsi =93-tol-=94 and =93-t=
o'e-=94; the cmavo =93no'e=94 has the assigned rafsi =93-nor-=94 and =93-no=
'e-=94. The selbri in <a href=3D#e5d10>Example 5.10 </a>through <a href=3D#=
e5d12>Example 5.12 </a>could be replaced by the lujvo =93nalmle=94, =93norm=
le=94, and =93tolmle=94 respectively.
+<p>
+This large variety of scalar negations is provided because different scale=
s have different properties. Some scales are open-ended in both directions:=
 there is no =93ultimately ugly=94 or =93ultimately beautiful=94. Other sca=
les, like temperature, are open at one end and closed at the other: there i=
s a minimum temperature (so-called =93absolute zero=94) but no maximum temp=
erature. Still other scales are closed at both ends.
+<p>
+Correspondingly, some selbri have no obvious =93to'e=94 =97 what is the op=
posite of a dog? =97 while others have more than one, and need =93ci'u=94 t=
o specify which opposite is meant.
+<p>
+<a name=3Ds6><h3>sumti negation</h3>
+<p>
+There are two ways of negating sumti in Lojban. We have the choice of quan=
tifying the sumti with zero, or of applying the sumti-negator =93na'ebo=94 =
before the sumti. It turns out that a zero quantification serves for contra=
dictory negation. As the cmavo we use implies, =93na'ebo=94 forms a scalar =
negation.
+<p>
+Let us show examples of each.
+<p>
+<pre><a name=3De6d1>6.1)	no lo ca nolraitru be le fasygu'e
+		cu krecau
+	Zero of those who are currently noblest-governors of the French country=
=20
+are-hair-without.
+	No current king of France is bald.
+</pre><p>Is <a href=3D#e6d1>Example 6.1 </a>true? Yes, because it merely c=
laims that of the current Kings of France, however many there may be, none =
are bald, which is plainly true, since there are no such current Kings of F=
rance.
+<p>
+Now let us look at the same sentence using =93na'ebo=94 negation:
+<p>
+<pre><a name=3De6d2>6.2)	na'ebo lo ca nolraitru
+		be le fasygu'e cu krecau
+	[Something] other-than-(the-current-noblest-governor
+		of the French country) is-hair-without.
+	Something other than the current King of France is bald.
+</pre><a href=3D#e6d2>Example 6.2 </a>is true provided that something reas=
onably describable as =93other than a current King of France=94, such as th=
e King of Saudi Arabia, or a former King of France, is in fact bald.
+<p>
+In place of =93na'ebo=94, you may also use =93no'ebo=94 and =93to'ebo=94, =
to be more specific about the sumti which would be appropriate in place of =
the stated sumti. Good examples are hard to come by, but here's a valiant t=
ry:
+<p>
+<pre><a name=3De6d3>6.3)	mi klama to'ebo la bastn.
+	I go to the-opposite-of Boston.
+	I go to Perth.
+</pre>(Boston and Perth are nearly, but not quite, antipodal cities. In a =
purely United States context, San Francisco might be a better =93opposite=
=94.) Coming up with good examples is difficult, because attaching =93to'eb=
o=94 to a description sumti is usually the same as attaching =93to'e=94 to =
the selbri of the description.
+<p>
+It is not possible to transform sumti negations of either type into bridi =
negations or scalar selbri negations. Negations of sumti will be used in Lo=
jban conversation. The inability to manipulate these negations logically wi=
ll, it is hoped, prevent the logical errors that result when natural langua=
ges attempt corresponding manipulations.
+<p>
+<a name=3Ds7><h3>Negation of minor grammatical constructs</h3>
+<p>
+We have a few other constructs that can be negated, all of them based on n=
egating individual words. For such negation, we use the suffix-combining ne=
gator, which is =93nai=94. =93nai=94, by the way, is almost always written =
as a compound into the previous word that it is negating, although it is a =
regular separate-word cmavo and the sole member of selma'o NAI.
+<p>
+Most of these negation forms are straightforward, and should be discussed =
and interpreted in connection with an analysis of the particular construct =
being negated. Thus, we will not go into much detail here.
+<p>
+The following are places where =93nai=94 is used:
+<p>
+When attached to tenses and modals (see <a href=3Dchap10.html>Chapter 10</=
a>), the =93nai=94 suffix usually indicates a contradictory negation of the=
 tagged bridi. Thus =93punai=94 as a tense inflection means =93not-in-the-p=
ast=94, or =93not-previously=94, without making any implication about any o=
ther time period unless explicitly stated. As a result,
+<p>
+<pre><a name=3De7d1>7.1)	mi na pu klama le zarci
+	I [false] [past] go-to the store.
+	I didn't go to the store.
+</pre>and
+<p>
+<pre><a name=3De7d2>7.2)	mi punai klama le zarci
+	I [past-not] go-to the store.
+	I didn't go to the store.
+</pre>mean exactly the same thing, although there may be a difference of e=
mphasis.
+<p>
+Tenses and modals can be logically connected, with the logical connectives=
 containing contradictory negations; this allows negated tenses and modals =
to be expressed positively using logical connectives. Thus =93punai je ca=
=94 means the same thing as =93pu naje ca=94.
+<p>
+As a special case, a =93-nai=94 attached to the interval modifiers of selm=
a'o TAhE, ROI, or ZAhO (explained in <a href=3Dchap10.html>Chapter 10</a>) =
signals a scalar negation:
+<p>
+<pre><a name=3De7d3>7.3)	mi paroinai dansu le bisli
+	I [once] [not] dance-on the ice
+</pre>means that I dance on the ice either zero or else two or more times =
within the relevant time interval described by the bridi. <a href=3D#e7d3>E=
xample 7.3 </a>is very different from the English use of =93not once=94, wh=
ich is an emphatic way of saying =93never=94 =97 that is, exactly zero time=
s.
+<p>
+In indicators and attitudinals of selma'o UI or CAI, =93nai=94 denotes a p=
olar negation. As discussed in <a href=3Dchap13.html>Chapter 13</a>, most i=
ndicators have an implicit scale, and =93nai=94 changes the indicator to re=
fer to the opposite end of the scale. Thus =93.uinai=94 expresses unhappine=
ss, and =93.ienai=94 expresses disagreement (not ambivalence, which is expr=
essed with the neutral or undecided intensity as =93.iecu'i=94).
+<p>
+Vocative cmavo of selma'o COI are considered a kind of indicator, but one =
which identifies the listener. Semantically, we could dispense with about h=
alf of the COI selma'o words based on the scalar paradigm. For example, =93=
co'o=94 could be expressed as =93coinai=94. However, this is not generally =
done.
+<p>
+Most of the COI cmavo are used in what are commonly called protocol situat=
ions. These protocols are used, for example, in radio conversations, which =
often take place in a noisy environment. The negatives of protocol words te=
nd to convey diametrically opposite communications situations (as might be =
expected). Therefore, only one protocol vocative is dependent on =93nai=94:=
 negative acknowledgement, which is =93je'enai=94 (=93I didn't get that=94)=
.
+<p>
+Unlike the attitudinal indicators, which tend to be unimportant in noisy s=
ituations, the protocol vocatives become more important. So if, in a noisy =
environment, a protocol listener makes out only =93nai=94, he or she can pr=
esume it is a negative acknowledgement and repeat transmission or otherwise=
 respond accordingly. <a href=3Dchap13.html>Chapter 13 </a>provides more de=
tail on this topic.
+<p>
+The abstractors of selma'o NU follow the pattern of the tenses and modals.=
 NU allows negative abstractions, especially in compound abstractions conne=
cted by logical connectives: =93su'ujeninai=94, which corresponds to =93su'=
u jenai ni=94 just as =93punai je ca=94 corresponds to =93pu naje ca=94. It=
 is not clear how much use logically connected abstractors will be: see <a =
href=3Dchap13.html>Chapter 13</a>.
+<p>
+A =93nai=94 attached to a non-logical connective (of selma'o JOI or BIhI) =
is a scalar negation, and says that the bridi is false under the specified =
mixture, but that another connective is applicable. Non-logical connectives=
 are discussed in <a href=3Dchap14.html>Chapter 14</a>.
+<p>
+<a name=3Ds8><h3>Truth questions</h3>
+<p>
+One application of negation is in answer to truth questions (those which e=
xpect the answers =93Yes=94 or =93No=94). The truth question cmavo =93xu=94=
 is in selma'o UI; placed at the beginning of a sentence, it asks whether t=
he sentence as a whole is true or false.
+<p>
+<pre><a name=3De8d1>8.1)	xu la djan. pu klama la paris. .e la rom.
+	Is it true that:  ( John previously went-to [both] Paris and Rome. )
+</pre><p>You can now use each of the several kinds of negation we've discu=
ssed in answer to this (presuming the same question and context for each an=
swer).
+<p>
+The straightforward negative answer is grammatically equivalent to the exp=
anded sentence with the =93na=94 immediately after the =93cu=94 (and before=
 any tense/modal):
+<p>
+<pre><a name=3De8d2>8.2)	na go'i
+	[false] [repeat previous]
+	No.
+</pre>which means
+<p>
+<pre><a name=3De8d3>8.3)	la djan. [cu] na pu klama la paris. .e la rom.
+	John [false] previously went-to [both] Paris and Rome.
+	It's not true that John went to Paris and Rome.
+</pre><p>The respondent can change the tense, putting the =93na=94 in eith=
er before or after the new tense:
+<p>
+<pre><a name=3De8d4>8.4)	na ba go'i
+	[false] [future] [repeat previous]
+</pre>meaning
+<p>
+<pre><a name=3De8d5>8.5)	la djan. [cu] na ba klama la paris. .e la rom.
+	John [false] later-will-go-to [both] Paris and Rome.
+	It is false that John will go to Paris and Rome.
+</pre>or alternatively
+<p>
+<pre><a name=3De8d6>8.6)	ba na go'i
+	[false] [future] [repeat previous]
+</pre>meaning
+<p>
+<pre><a name=3De8d7>8.7)	la djan. [cu] ba na klama la paris. .e la rom.
+	John later-will [false] go-to [both] Paris and Rome.
+</pre><p>We stated in <a href=3D#s3>Section 3 </a>that sentences like <a h=
ref=3D#e8d5>Example 8.5 </a>and <a href=3D#e8d7>Example 8.7 </a>appear to b=
e semantically identical, but that subtle semantic distinctions may eventua=
lly be found.
+<p>
+You can also use a scalar negation with =93na'e=94, in which case, it is e=
quivalent to putting a =93na'eke=94 immediately after any tense:
+<p>
+<pre><a name=3De8d8>8.8)	na'e go'i
+	other-than [repeat previous]
+</pre>which means
+<p>
+<pre><a name=3De8d9>8.9)	la djan. [cu] pu na'eke klama [ke'e] la paris. .e=
 la rom.
+	John previously other-than(went-to) [both] Paris and Rome.
+</pre><p>He might have telephoned the two cities instead of going there. T=
he unnecessary =93ke=94 and =93ke'e=94 would have been essential if the sel=
bri had been a tanru.
+<p>
+<a name=3Ds9><h3>Affirmations</h3>
+<p>
+There is an explicit positive form for both selma'o NA (=93ja'a=94) and se=
lma'o NAhE (=93je'a=94), each of which would supplant the corresponding neg=
ator in the grammatical position used, allowing one to assert the positive =
in response to a negative question or statement without confusion. Assuming=
 the same context as in <a href=3D#s8>Section 8</a>:
+<p>
+<pre><a name=3De9d1>9.1)	xu na go'i
+	Is-it-true-that [false] [repeat previous]?
+</pre>or equivalently
+<p>
+<pre><a name=3De9d2>9.2)	xu la djan. [cu] na pu klama la paris. .e la rom.
+	Is it true that:  John [false] previously-went-to [both] Paris and Rome.]
+</pre><p>The obvious, but incorrect, positive response to this negative qu=
estion is:
+<p>
+<pre><a name=3De9d3>9.3)	go'i
+	[repeat previous]
+</pre><p>A plain =93go'i=94 does not mean =93Yes it is=94; it merely abbre=
viates repeating the previous statement unmodified, including any negators =
present; and <a href=3D#e9d3>Example 9.3 </a>actually states that it is fal=
se that John went to both Paris and Rome.
+<p>
+When considering:
+<p>
+<pre><a name=3De9d4>9.4)	na go'i
+	[false] [repeat previous]
+</pre>as a response to a negative question like <a href=3D#e9d2>Example 9.=
2</a>, Lojban designers had to choose between two equally plausible interpr=
etations with opposite effects. Does <a href=3D#e9d4>Example 9.4 </a>create=
 a double negative in the sentence by adding a new =93na=94 to the one alre=
ady there (forming a double negative and hence a positive statement), or do=
es the =93na=94 replace the previous one, leaving the sentence unchanged?
+<p>
+It was decided that substitution, the latter alternative, is the preferabl=
e choice, since it is then clear whether we intend a positive or a negative=
 sentence without performing any manipulations. This is the way English usu=
ally works, but not all languages work this way =97 Russian, Japanese, and =
Navajo all interpret a negative reply to a negative question as positive.
+<p>
+The positive assertion cmavo of selma'o NA, which is =93ja=92a=94, can als=
o replace the =93na=94 in the context, giving:
+<p>
+<pre><a name=3De9d5>9.5)	ja'a go'i
+	(John truly-(previously went-to) [both] Paris and Rome.)
+</pre>=93ja'a=94 can replace =93na=94 in a similar manner wherever the lat=
ter is used:
+<p>
+<pre><a name=3De9d6>9.6)	mi ja=92a klama le zarci
+	I indeed go to the store.
+</pre>=93je'a=94 can replace =93na'e=94 in exactly the same way, stating t=
hat scalar negation does not apply, and that the relation indeed holds as s=
tated. In the absence of a negation context, it emphasizes the positive:
+<p>
+<pre><a name=3De9d7>9.7)	ta je'a melbi
+	that is-indeed beautiful.
+</pre><a name=3Ds10><h3>Metalinguistic negation forms</h3>
+<p>
+The question of truth or falsity is not entirely synonymous with negation.=
 Consider the English sentence
+<p>
+<pre><a name=3De10d1>10.1)	I have not stopped beating my wife.
+</pre><p>If I never started such a heinous activity, then this sentence is=
 neither true nor false. Such a negation simply says that something is wron=
g with the non-negated statement. Generally, we then use either tone of voi=
ce or else a correction to express a preferred true claim: =93I never have =
beaten my wife.=94
+<p>
+Negations which follow such a pattern are called =93metalinguistic negatio=
ns=94. In natural languages, the mark of metalinguistic negation is that an=
 indication of a correct statement always, or almost always, follows the ne=
gation. Tone of voice or emphasis may be further used to clarify the error.
+<p>
+Negations of every sort must be expressible in Lojban; errors are inherent=
 to human thought, and are not excluded from the language. When such negati=
ons are metalinguistic, we must separate them from logical claims about the=
 truth or falsity of the statement, as well as from scalar negations which =
may not easily express (or imply) the preferred claim. Because Lojban allow=
s concepts to be so freely combined in tanru, limits on what is plausible o=
r not plausible tend to be harder to determine.
+<p>
+Mimicking the muddled nature of natural language negation would destroy th=
is separation. Since Lojban does not use tone of voice, we need other means=
 to metalinguistically indicate what is wrong with a statement. When the st=
atement is entirely inappropriate, we need to be able to express metalingui=
stic negation in a more non-specific fashion.
+<p>
+Here is a list of some different kinds of metalinguistic negation with Eng=
lish-language examples:
+<p>
+<pre><a name=3De10d2>10.2)	I have not <i>stopped</i> beating my wife
+	(I never started =97 failure of presupposition).
+<a name=3De10d3>10.3)	5 is not blue
+	(color does not apply to abstract concepts -- failure of category).
+<a name=3De10d4>10.4)	The current King of France is not bald.
+	(there is no current King of France --- existential failure)
+<a name=3De10d5>10.5)	I do not have THREE children.
+	(I have two --- simple undue quantity)
+<a name=3De10d6>10.6)	I have not held THREE jobs previously, but four.
+	(inaccurate quantity; the difference from the previous example is that
+someone who has held four jobs has also held three jobs)
+<a name=3De10d7>10.7)	It is not good, but bad.
+	(undue quantity negation indicating that the value on a scale for measuri=
ng
+the predicate is incorrect)
+<a name=3De10d8>10.8)	She is not PRETTY; she is beautiful
+	(undue quantity transferred to a non-numeric scale)
+<a name=3De10d9>10.9)	The house is not blue, but green.
+	(the scale/category being used is incorrect, but a related category appli=
es)
+<a name=3De10d10>10.10)	The house is not blue, but is colored.
+	(the scale/category being used is incorrect, but a broader category appli=
es)
+<a name=3De10d11>10.11)	The cat is not blue, but long-haired.
+	(the scale/category being used is incorrect, but an unrelated category
+applies)
+<a name=3De10d12>10.12)	A:  He ain't coming today.
+	B:  =93Ain't=94 ain't a word.
+	(solecism, or improper grammatical action)
+<a name=3De10d13>10.13)	I haven't STOOPED beating my wife; I've STOPPED.
+	(spelling or mispronunciation error)
+<a name=3De10d14>10.14)	Not only was it a sheep, it was a black sheep.
+	(non-contradictory correction)
+</pre><p>The set of possible metalinguistic errors is open-ended.
+<p>
+Many of these forms have a counterpart in the various examples that we've =
discussed under logical negation. Metalinguistic negation doesn't claim tha=
t the sentence is false or true, though. Rather, it claims that, due to som=
e error in the statement, =93true=94 and =93false=94 don't really apply.
+<p>
+Because one can metalinguistically negate a true statement intending a non=
-contradictory correction (say, a spelling error); we need a way (or ways) =
to metalinguistically negate a statement which is independent of our logica=
l negation schemes using =93na=94, =93na'e=94 and kin. The cmavo =93na'i=94=
 is assigned this function. If it is present in a statement, it indicates m=
etalinguistically that something in the statement is incorrect. This metali=
nguistic negation must override any evaluation of the logic of the statemen=
t. It is equally allowed in both positive and negative statements.
+<p>
+Since =93na'i=94 is not a logical operator, multiple occurrences of =93na'=
i=94 need not be assumed to cancel each other. Indeed, we can use the posit=
ion of =93na'i=94 to indicate metalinguistically what is incorrect, prepara=
tory to correcting it in a later sentence; for this reason, we give =93na'i=
=94 the grammar of UI. The inclusion of =93na'i=94 anywhere in a sentence m=
akes it a non-assertion, and suggests one or more pitfalls in assigning a t=
ruth value.
+<p>
+Let us briefly indicate how the above-mentioned metalinguistic errors can =
be identified. Other metalinguistic problems can then be marked by devising=
 analogies to these examples:
+<p>
+Existential failure can be marked by attaching =93na'i=94 to the descripto=
r =93lo=94 or the =93poi=94 in a =93da poi=94-form sumti. (See Chapter 6 an=
d <a href=3Dchap16.html>Chapter 16 </a>for details on these constructions.)=
 Remember that if a =93le=94 sumti seems to refer to a non-existent referen=
t, you may not understand what the speaker has in mind =97 the appropriate =
response is then =93ki'a=94, asking for clarification.
+<p>
+Presupposition failure can be marked directly if the presupposition is ove=
rt; if not, one can insert a =93mock presupposition=94 to question with the=
 sumti tcita (selma'o BAI) word =93ji'u=94; =93ji'uku=94 thus explicitly re=
fers to an unexpressed assumption, and =93ji'una'iku=94 metalinguistically =
says that something is wrong with that assumption. (See <a href=3Dchap10.ht=
ml>Chapter 10</a>.)
+<p>
+Scale errors and category errors can be similarly expressed with selma'o B=
AI. =93le'a=94 has meaning =93of category/class/type X=94, =93ci'u=94 has m=
eaning =93on scale X=94, and =93ci'e=94, based on =93ciste=94, can be used =
to talk about universes of discourse defined either as systems or sets of c=
omponents, as shown in <a href=3D#s8>Section 8</a>. =93kai=94 and =93la'u=
=94 also exist in BAI for discussing other quality and quantity errors.
+<p>
+We have to make particular note of potential problems in the areas of undu=
e quantity and incorrect scale/category. Assertions about the relationships=
 between gismu are among the basic substance of the language. It is thus in=
valid to logically require that if something is blue, that it is colored, o=
r if it is not-blue, then it is some other color. In Lojban, =93blanu=94 (=
=93blue=94) is not explicitly defined as a =93skari=94 (=93color=94). Simil=
arly, it is not implicit that the opposite of =93good=94 is =93bad=94.
+<p>
+This mutual independence of gismu is only an ideal. Pragmatically, people =
will categorize things based on their world-views. We will write dictionary=
 definitions that will relate gismu, unfortunately including some of these =
world-view assumptions. Lojbanists should try to minimize these assumptions=
, but this seems a likely area where logical rules will break down (or wher=
e Sapir-Whorf effects will be made evident). In terms of negation, however,=
 it is vital that we clearly preserve the capability of denying a presumabl=
y obvious scale or category assumption.
+<p>
+Solecisms, grammatical and spelling errors will be marked by marking the o=
ffending word or phrase with =93na'i=94 (in the manner of any selma'o UI cm=
avo). In this sense, =93na'i=94 becomes equivalent to the English metalingu=
istic marker =93[sic]=94. Purists may choose to use ZOI or LOhU/LEhU quotes=
 or =93sa'a=94-marked corrections to avoid repeating a truly unparsable pas=
sage, especially if a computer is to analyze the speech/text. See <a href=
=3Dchap19.html>Chapter 19 </a>for explanations of these usages.
+<p>
+In summary, metalinguistic negation will typically take the form of referr=
ing to a previous statement and marking it with one or more =93na'i=94 to i=
ndicate what metalinguistic errors have been made, and then repeating the s=
tatement with corrections. References to previous statements may be full re=
petitions, or may use members of selma'o GOhA. =93na'i=94 at the beginning =
of a statement merely says that something is inappropriate about the statem=
ent, without specificity.
+<p>
+In normal use, metalinguistic negation requires that a corrected statement=
 follow the negated statement. In Lojban, however, it is possible to comple=
tely and unambiguously specify metalinguistic errors without correcting the=
m. It will eventually be seen whether an uncorrected metalinguistic negatio=
n remains an acceptable form in Lojban. In such a statement, metalinguistic=
 expression would involve an ellipsis not unlike that of tenseless expressi=
on.
+<p>
+Note that metalinguistic negation gives us another kind of legitimate nega=
tive answer to a =93xu=94 question (see <a href=3D#s8>Section 8</a>). =93na=
'i=94 will be used when something about the questioned statement is inappro=
priate, such as in questions like =93Have you stopped beating your wife?=94=
:
+<p>
+<pre><a name=3De10d15>10.15)	xu do sisti lezu'o do rapydarxi ledo fetspe
+	Have you ceased the activity of repeat-hitting your female-spouse?
+</pre><p>Responses could include:
+<p>
+<pre><a name=3De10d16>10.16)	na'i go'i
+	The bridi as a whole is inappropriate in some way.
+<a name=3De10d17>10.17)	go'i na'i
+	The selbri (sisti) is inappropriate in some way.
+</pre><p>One can also specifically qualify the metalinguistic negation, by=
 explicitly repeating the erroneous portion of the bridi to be metalinguist=
ically negated, or adding on of the selma'o BAI qualifiers mentioned above:
+<p>
+<pre><a name=3De10d18>10.18)	go'i ji'una'iku
+	Some presupposition is wrong with the previous bridi.
+</pre><p>Finally, one may metalinguistically affirm a bridi with =93jo'a=
=94, another cmavo of selma'o UI. A common use for =93jo'a=94 might be to a=
ffirm that a particular construction, though unusual or counterintuitive, i=
s in fact correct; another usage would be to disagree with =97 by overridin=
g =97 a respondent's metalinguistic negation.
+<p>
+<a name=3Ds11><h3>Summary =97 Are All Possible Questions About Negation No=
w Answered?</h3>
+<p>
+<pre><a name=3De11d1>11.1)	na go'i  .ije na'e go'i  .ije na'i go'i
+
+</body></html>
+<center><p align=3Dcenter>
+<img src=3Dchap16.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+
+<h2>Chapter 16
+<br>
+=93Who Did You Pass On The Road? Nobody=94: Lojban And Logic</h2>
+<h6>$Revision: 4.0 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>What's wrong with this picture?</h3>
+<p>
+<cx "nobody, interpretation of">  XE "nobody: interpretation of"   The fol=
lowing brief dialogue is from <a href=3Dchap7.html>Chapter 7 </a>of <cite>T=
hrough The Looking Glass </cite>by Lewis Carroll.
+<p>
+<pre><a name=3De1d1>1.1)	=93Who did you pass on the road?=94 the King went=
 on, holding out his hand to the Messenger for some more hay.
+<a name=3De1d2>1.2)	=93Nobody,=94 said the Messenger.
+<a name=3De1d3>1.3)	=93Quite right,=94 said the King:  =93this young lady =
saw him too. So of course Nobody walks slower than you. =94
+<a name=3De1d4>1.4)	=93I do my best,=94 the Messenger said in a sulky tone=
. =93I'm sure nobody walks much faster than I do!=94
+<a name=3De1d5>1.5)	=93He can't do that,=94 said the King, =93or else he'd=
 have been here first.=94
+</pre><p>This nonsensical conversation results because the King insists on=
 treating the word =93nobody=94 as a name, a name of somebody. However, the=
 essential nature of the English word =93nobody=94 is that it doesn't refer=
 to somebody; or to put the matter another way, there isn't anybody to whic=
h it refers.
+<p>
+<cx "nobody, ambiguous interpretations of">  XE "nobody: ambiguous interpr=
etations of"   The central point of contradiction in the dialogue arises in=
 <a href=3D#e1d3>Example 1.3</a>, when the King says =93<dots>=85 </dots>No=
body walks slower than you=94. This claim would be plausible if =93Nobody=
=94 were really a name, since the Messenger could only pass someone who doe=
s walk more slowly than he. But the Messenger interprets the word =93nobody=
=94 in the ordinary English way, and says (in <a href=3D#e1d4>Example 1.4</=
a>) =93<dots>=85 </dots>nobody walks much faster than I do=94 (i.e., I walk=
 faster, or as fast as, almost everyone), which the King then again misunde=
rstands. Both the King and the Messenger are correct according to their res=
pective understandings of the ambiguous word =93nobody/Nobody=94.
+<p>
+<cx "nobody, Lojban contrasted with English">  XE "nobody: Lojban contrast=
ed with English"   There are Lojban words or phrases corresponding to the p=
roblematic English words =93somebody=94, =93nobody=94, =93anybody=94, =93ev=
erybody=94 (and their counterparts =93some/no/any/everyone=94 and =93some/n=
o/any/everything=94), but they obey rules which can often be surprising to =
English-speakers. The dialogue above simply cannot be translated into Lojba=
n without distortion: the name =93Nobody=94 would have to be represented by=
 a Lojban name, which would spoil the perfection of the wordplay. As a matt=
er of fact, this is the desired result: a logical language should not allow=
 two conversationalists to affirm =93Nobody walks slower than the Messenger=
=94 and =93Nobody walks faster than the Messenger=94 and both be telling th=
e truth. (Unless, of course, nobody but the Messenger walks at all, or ever=
yone walks at exactly the same speed.)
+<p>
+<cx "logic, resolving ambiguities of =93nobody=94">  XE "logic: resolving =
ambiguities of \=93nobody\=94"   This chapter will explore the Lojban mecha=
nisms that allow the correct and consistent construction of sentences like =
those in the dialogue. There are no new grammatical constructs explained in=
 this chapter; instead, it discusses the way in which existing facilities t=
hat allow Lojban-speakers to resolve problems like the above, using the con=
cepts of modern logic. However, we will not approach the matter from the vi=
ewpoint of logicians, although readers who know something of logic will dis=
cover familiar notions in Lojban guise.
+<p>
+<cx "logic, limits of">  XE "logic: limits of"   Although Lojban is called=
 a logical language, not every feature of it is =93logical=94. In particula=
r, the use of =93le=94 is incompatible with logical reasoning based on the =
description selbri, because that selbri may not truthfully apply: you canno=
t conclude from my statement that
+<p>
+<pre><a name=3De1d6>1.6)	mi viska le nanmu
+	I see the-one-I-refer-to-as-the man.
+	I see the man/men.
+</pre>that there really is a man; the only thing you can conclude is that =
there is one thing (or more) that I choose to refer to as a man. You cannot=
 even tell which man is meant for sure without asking me (although communic=
ation is served if you already know from the context).
+<p>
+<cx "logic, and attitudinals">  XE "logic: and attitudinals"   <cx "attitu=
dinals, and logic">  XE "attitudinals: and logic"   In addition, the use of=
 attitudinals (see <a href=3Dchap13.html>Chapter 13</a>) often reduces or r=
emoves the ability to make deductions about the bridi to which those attitu=
dinals are applied. From the fact that I hope George will win the election,=
 you can conclude nothing about George's actual victory or defeat.
+<p>
+<a name=3Ds2><h3>Existential claims, prenexes, and variables</h3>
+<p>
+Let us consider, to begin with, a sentence that is not in the dialogue:
+<p>
+<ex "something sees me">  XE "something sees me: example=93  <pre><a name=
=3De2d1>2.1)	Something sees me.
+</pre><p>There are two plausible Lojban translations of <a href=3D#e2d1>Ex=
ample 2.1</a>. The simpler one is:
+<p>
+<lx "zo'e">  XE "zo'e"   <pre><a name=3De2d2>2.2)	[zo'e] viska mi
+	Something-unspecified sees me.
+</pre><cx "zo'e, as a translation for =93something=94">  XE "zo'e: as a tr=
anslation for \=93something\=94"   <cx "something, unspecified definite wit=
h =93zo'e=94">  XE "something: unspecified definite with \=93zo'e\=94"   Th=
e cmavo =93zo'e=94 indicates that a sumti has been omitted (indeed, even =
=93zo'e=94 itself can be omitted in this case, as explained in <a href=3Dch=
ap7.html>Chapter 7</a>) and the listener must fill in the correct value fro=
m context. In other words, <a href=3D#e2d2>Example 2.2 </a>means =93=91You-=
know-what=92 sees me.=94
+<p>
+However, <a href=3D#e2d1>Example 2.1 </a>is just as likely to assert simpl=
y that there is someone who sees me, in which case a correct translation is=
:
+<p>
+<lx "da">  XE "da"   <cx "da, as a translation for =93something=94">  XE "=
da: as a translation for \=93something\=94"   <pre><a name=3De2d3>2.3)	da z=
o'u da viska mi
+	There-is-an-X such-that X sees me.
+</pre><cx "existential claims, definition">  XE "existential claims: defin=
ition"   <cx "da, contrasted with zo'e">  XE "da: contrasted with zo'e"   <=
cx "zo'e, contrasted with da">  XE "zo'e: contrasted with da"   <a href=3D#=
e2d3>Example 2.3 </a>does not presuppose that the listener knows who sees t=
he speaker, but simply tells the listener that there is someone who sees th=
e speaker. Statements of this kind are called =93existential claims=94. (Fo=
rmally, the one doing the seeing is not restricted to being a person; it co=
uld be an animal or =97 in principle =97 an inanimate object. We will see i=
n <a href=3D#s4>Section 4 </a>how to represent such restrictions.)
+<p>
+<lx "zo'u">  XE "zo'u"   <lx "ZOhU">  XE "ZOhU selma=92o"   <cx "prenex, e=
xplanation">  XE "prenex: explanation"   <cx "prenex, syntax of">  XE "pren=
ex: syntax of"   <cx "variables, logical">  XE "variables: logical"   <cx "=
logical variables, notation convention">  XE "logical variables: notation c=
onvention"   <a href=3D#e2d3>Example 2.3 </a>has a two-part structure: ther=
e is the part =93da zo'u=94, called the prenex, and the part =93da viska mi=
=94, the main bridi. Almost any Lojban bridi can be preceded by a prenex, w=
hich syntactically is any number of sumti followed by the cmavo =93zo'u=94 =
(of selma'o ZOhU). For the moment, the sumti will consist of one or more of=
 the cmavo =93da=94, =93de=94, and =93di=94 (of selma'o KOhA), glossed in t=
he literal translations as =93X=94, =93Y=94, and =93Z=94 respectively. By a=
nalogy to the terminology of symbolic logic, these cmavo are called =93vari=
ables=94.
+
+<p>
+Here is an example of a prenex with two variables:
+<p>
+<ex "somebody loves somebody">  XE "somebody loves somebody: example=93  <=
pre><a name=3De2d4>2.4)	da de zo'u da prami de
+	There-is-an-X there-is-a-Y such that X loves Y.
+	Somebody loves somebody.
+</pre><cx "somebody, contrasted with somebody else">  XE "somebody: contra=
sted with somebody else"   <cx "logical variables, effect of using multiple=
 different">  XE "logical variables: effect of using multiple different"   =
In <a href=3D#e2d4>Example 2.4</a>, the literal interpretation of the two v=
ariables =93da=94 and =93de=94 as =93there-is-an-X=94 and =93there-is-a-Y=
=94 tells us that there are two things which stand in the relationship that=
 one loves the other. It might be the case that the supposed two things are=
 really just a single thing that loves itself; nothing in the Lojban versio=
n of <a href=3D#e2d4>Example 2.4 </a>rules out that interpretation, which i=
s why the colloquial translation does not say =93Somebody loves somebody el=
se.=94 The things referred to by different variables may be different or th=
e same. (We use =93somebody=94 here rather than =93something=94 for natural=
ness; lovers and beloveds are usually persons, though the Lojban does not s=
ay so.)
+<p>
+<cx "logical variables, with multiple appearances in bridi">  XE "logical =
variables: with multiple appearances in bridi"   It is perfectly all right =
for the variables to appear more than once in the main bridi:
+<p>
+<ex "somebody loves self">  XE "somebody loves self: example=93  <pre><a n=
ame=3De2d5>2.5)	da zo'u da prami da
+	There-is-an-X such that X loves X
+	Somebody loves himself/herself.
+</pre><cx "da prami da, contrasted with da prami de">  XE "da prami da: co=
ntrasted with da prami de"   <cx "da prami de, contrasted with da prami da"=
>  XE "da prami de: contrasted with da prami da"   What <a href=3D#e2d5>Exa=
mple 2.5 </a>claims is fundamentally different from what <a href=3D#e2d4>Ex=
ample 2.4 </a>claims, because =93da prami da=94 is not structurally the sam=
e as =93da prami de=94. However,
+<p>
+<cx "logical variables, effect of global substitution">  XE "logical varia=
bles: effect of global substitution"   <pre><a name=3De2d6>2.6)	de zo'u de =
prami de
+	There-is-a-Y such that Y loves Y
+</pre>means exactly the same thing as <a href=3D#e2d5>Example 2.5</a>; it =
does not matter which variable is used as long as they are used consistentl=
y.
+<p>
+It is not necessary for a variable to be a sumti of the main bridi directl=
y:
+<p>
+<ex "somebody's dog">  XE "somebody's dog: example=93  <pre><a name=3De2d7=
>2.7)	da zo'u le da gerku cu viska mi
+	There-is-an-X such-that the of-X dog sees me
+	Somebody's dog sees me
+</pre>is perfectly correct even though the =93da=94 is used only in a poss=
essive construction. (Possessives are explained in <a href=3Dchap8.html>Cha=
pter 8</a>.)
+<p>
+<cx "logical variable, when not in main bridi">  XE "logical variables: wh=
en not in main bridi"   It is very peculiar, however, even if technically g=
rammatical, for the variable not to appear in the main bridi at all:
+<p>
+<ex "Ralph">  XE "Ralph: example=93  <pre><a name=3De2d8>2.8)	da zo'u la r=
alf. gerku
+	There is something such that Ralph is a dog.
+</pre>has a variable bound in a prenex whose relevance to the claim of the=
 following bridi is completely unspecified.
+<p>
+<a name=3Ds3><h3>Universal claims</h3>
+<p>
+What happens if we substitute =93everything=94 for =93something=94 in <a h=
ref=3D#e2d1>Example 2.1</a>? We get:
+<p>
+<pre><a name=3De3d1>3.1)	Everything sees me.
+</pre><cx "universal claims, explanation">  XE "universal claims: explanat=
ion"   Of course, this example is false, because there are many things whic=
h do not see the speaker. It is not easy to find simple truthful examples o=
f so-called universal claims (those which are about everything), so bear wi=
th us for a while. (Indeed, some Lojbanists tend to avoid universal claims =
even in other languages, since they are so rarely true in Lojban.)
+<p>
+The Lojban translation of <a href=3D#e3d1>Example 3.1 </a>is
+ <ex "everything sees me">  XE "everything sees me: example=93  <pre><a na=
me=3De3d2>3.2)	ro da zo'u da viska mi
+	For-every X :  X sees me.
+</pre><lx "ro">  XE "ro"   <lx "ro da">  XE "ro da"   <cx "everything, exp=
ressing with =93ro da=94">  XE "everything: expressing with \=93ro da\=94" =
  When the variable cmavo =93da=94 is preceded by =93ro=94, the combination=
 means =93For every X=94 rather than =93There is an X=94. Superficially, th=
ese English formulations look totally unrelated: <a href=3D#s6>Section 6 </=
a>will bring them within a common viewpoint. For the moment, accept the use=
 of =93ro da=94 for =93everything=94 on faith.
+<p>
+Here is a universal claim with two variables:
+<p>
+<ex "everything loves everything">  XE "everything loves everything: examp=
le=93  <pre><a name=3De3d3>3.3)	ro da ro de zo'u da prami de
+	For-every X, for-every Y :  X loves Y.
+	Everything loves everything.
+</pre><p>Again, X and Y can represent the same thing, so <a href=3D#e3d3>E=
xample 3.3 </a>does not mean =93Everything loves everything else.=94 Furthe=
rmore, because the claim is universal, it is about every thing, not merely =
every person, so we cannot use =93everyone=94 or =93everybody=94 in the tra=
nslation.
+<p>
+<cx "mixed claim, definition">  XE "mixed claim: definition"   <cx "existe=
ntial, mixed claim with universal">  XE "existential: mixed claim with univ=
ersal"   <cx "universal, mixed claim with existential">  XE "universal: mix=
ed claim with existential"   Note that =93ro=94 appears before both =93da=
=94 and =93de=94. If =93ro=94 is omitted before either variable, we get a m=
ixed claim, partly existential like those of <a href=3D#s2>Section 2</a>, p=
artly universal.
+<p>
+<ex "everything sees something">  XE "everything sees something: example=
=93  <pre><a name=3De3d4>3.4)	ro da de zo'u da viska de
+	For-every X, there-is-a-Y :  X sees Y.
+	Everything sees something.
+</pre><ex "something sees everything">  XE "something sees everything: exa=
mple=93  <pre><a name=3De3d5>3.5)	da ro de zo'u da viska de
+	There-is-an-X such-that-for-every-Y :  X sees Y.
+	Something sees everything.
+</pre><cx "logical variables, effect of order in prenex">  XE "logical var=
iables: effect of order in prenex"   <a href=3D#e3d{4}>Examples 3.4 </a>and=
 <a href=3D#e3d5>3.5 </a>mean completely different things. <a href=3D#e3d4>=
Example 3.4 </a>says that for everything, there is something which it sees,=
 not necessarily the same thing seen for every seer. <a href=3D#e3d5>Exampl=
e 3.5</a>, on the other hand, says that there is a particular thing which c=
an see everything that there is (including itself). Both of these are fairl=
y silly, but they are different kinds of silliness.
+<p>
+There are various possible translations of universal claims in English: so=
metimes we use =93anybody/anything=94 rather than =93everybody/everything=
=94. Often it makes no difference which of these is used: when it does make=
 a difference, it is a rather subtle one which is explained in <a href=3D#s=
8>Section 8</a>.
+<p>
+<a name=3Ds4><h3>Restricted claims: =93da poi=94</h3>
+<p>
+<cx "universal claims, restricting">  XE "universal claims: restricting"  =
 <cx "restricted claims, definition">  XE "restricted claims: definition"  =
 The universal claims of <a href=3D#s3>Section 3 </a>are not only false but=
 absurd: there is really very little to be said that is both true and non-t=
rivial about every object whatsoever. Furthermore, we have been glossing ov=
er the distinction between =93everything=94 and =93everybody=94 and the oth=
er pairs ending in =93-thing=94 and =93-body=94. It is time to bring up the=
 most useful feature of Lojban variables: the ability to restrict their ran=
ges.
+<p>
+<cx "relative clauses, use in restricting existential claims">  XE =93rela=
tive clause: use in restricting existential claims"   <cx "existential clai=
ms, restricting">  XE "existential claims: restricting"   <lx "poi">  XE "p=
oi"   <lx "da poi">  XE "da poi"   <lx "ku'o">  XE "ku'o"   <lx "vau">  XE =
"vau"   In Lojban, a variable =93da=94, =93de=94, or =93di=94 may be follow=
ed by a =93poi=94 relative clause in order to restrict the range of things =
that the variable describes. Relative clauses are described in detail in <a=
 href=3Dchap8.html>Chapter 8</a>, but the kind we will need at present cons=
ist of =93poi=94 followed by a bridi (often just a selbri) terminated with =
=93ku'o=94 or =93vau=94 (which can usually be elided). Consider the differe=
nce between
+<p>
+<pre><a name=3De4d1>4.1)	da zo'u da viska la djim.
+	There-is-an-X :  X sees Jim.
+	Something sees Jim.
+</pre>and
+<p>
+<pre><a name=3De4d2>4.2)	da poi prenu zo'u da viska la djim.
+	There-is-an-X which is-a-person :  X sees Jim.
+	Someone sees Jim.
+</pre><cx "something, contrasted with someone">  XE "something: contrasted=
 with someone"   In <a href=3D#e4d1>Example 4.1</a>, the variable =93da=94 =
can refer to any object whatever; there are no restrictions on it. In <a hr=
ef=3D#e4d2>Example 4.2</a>, =93da=94 is restricted by the =93poi prenu=94 r=
elative clause to persons only, and so =93da poi prenu=94 translates as =93=
someone.=94 (The difference between =93someone=94 and =93somebody=94 is a m=
atter of English style, with no real counterpart in Lojban.) If <a href=3D#=
e4d2>Example 4.2 </a>is true, then <a href=3D#e4d1>Example 4.1 </a>must be =
true, but not necessarily vice versa.
+<p>
+<cx "relative clauses, use in restricting universal claims">  XE =93relati=
ve clause: use in restricting universal claims"   <cx "universal claims, re=
stricting">  XE "universal claims: restricting"   Universal claims benefit =
even more from the existence of relative clauses. Consider
+<p>
+<ex "everything breathes">  XE "everything breathes: example=93  <pre><a n=
ame=3De4d3>4.3)	ro da zo'u da vasxu
+	For-every X :  X breathes
+	Everything breathes
+</pre>and
+<p>
+<cx "dog breathes">  XE "dog breathes"   <pre><a name=3De4d4>4.4)	ro da po=
i gerku zo'u da vasxu
+	For-every X which is-a-dog :  X breathes.
+	Every dog breathes.
+	Each dog breathes.
+	All dogs breathe.
+</pre><a href=3D#e4d3>Example 4.3 </a>is a silly falsehood, but <a href=3D=
#e4d4>Example 4.4 </a>is an important truth (at least if applied in a timel=
ess or potential sense: see <a href=3Dchap10.html>Chapter 10</a>). Note the=
 various colloquial translations =93every dog=94, =93each dog=94, and =93al=
l dogs=94. They all come to the same thing in Lojban, since what is true of=
 every dog is true of all dogs. =93All dogs=94 is treated as an English plu=
ral and the others as singular, but Lojban makes no distinction.
+<p>
+If we make an existential claim about dogs rather than a universal one, we=
 get:
+<p>
+<pre><a name=3De4d5>4.5)	da poi gerku zo'u da vasxu
+	There-is-an-X which is-a-dog :  X breathes.
+	Some dog breathes.
+</pre><a name=3Ds5><h3>Dropping the prenex</h3>
+<p>
+<cx "prenex, considerations for dropping">  XE "prenex: considerations for=
 dropping"   It isn't really necessary for every Lojban bridi involving var=
iables to have a prenex on the front. In fact, none of the examples we've s=
een so far required prenexes at all! The rule for dropping the prenex is si=
mple: if the variables appear in the same order within the bridi as they di=
d in the prenex, then the prenex is superfluous. However, any =93ro=94 or =
=93poi=94 appearing in the prenex must be transferred to the first occurren=
ce of the variable in the main part of the bridi. Thus, <a href=3D#e2d3>Exa=
mple 2.3 </a>becomes just:
+
+<p>
+<pre><a name=3De5d1>5.1)	da viska mi
+	There-is-an-X-which sees me.
+	Something sees me.
+</pre>and <a href=3D#e4d4>Example 4.4 </a>becomes:
+<p>
+<pre><a name=3De5d2>5.2)	ro da poi gerku cu vasxu
+	For-every X which is-a-dog, it-breathes.
+	Every dog breathes.
+</pre><cx "prenex, purpose of">  XE "prenex: purpose of"   <cx "prenex, ef=
fect of order of variables in">  XE "prenex: effect of order of variables i=
n"   You might well suppose, then, that the purpose of the prenex is to all=
ow the variables in it to appear in a different order than the bridi order,=
 and that would be correct. Consider
+<p>
+<cx "everyone bitten by dog">  XE "everyone bitten by dog"   <pre><a name=
=3De5d3>5.3)	ro da poi prenu ku'o de poi gerku ku'o zo'u de batci da
+	For-every X which is-a-person, there-is-a-Y which is-a-dog:  Y bites X.
+</pre><p>The prenex of <a href=3D#e5d3>Example 5.3 </a>is like that of <a =
href=3D#e3d4>Example 3.4 </a>(but with relative clauses): it notes that the=
 following bridi is true of every person with respect to some dog, not nece=
ssarily the same dog for each. But in the main bridi part, the =93de=94 app=
ears before the =93da=94. Therefore, the true translation is
+<p>
+<pre><a name=3De5d4>5.4)	Every person is bitten by some dog (or other).
+</pre><p>If we tried to omit the prenex and move the =93ro=94 and the rela=
tive clauses into the main bridi, we would get:
+<p>
+<pre><a name=3De5d5>5.5)	de poi gerku cu batci ro da poi prenu
+	There-is-a-Y which is-a-dog which-bites every Y which is-a-person
+	Some dog bites everyone.
+</pre><ex "Fido">  XE "Fido: example=93    XE  selma=92o"dog bites: exampl=
e"  which has the structure of <a href=3D#e3d5>Example 3.5</a>: it says tha=
t there is a dog (call him Fido) who bites, has bitten, or will bite every =
person that has ever existed! We can safely rule out Fido's existence, and =
say that <a href=3D#e5d5>Example 5.5 </a>is false, while agreeing to <a hre=
f=3D#e5d3>Example 5.3</a>.
+<p>
+<cx "universal claims, dangers of using">  XE "universal claims: dangers o=
f using"   Even so, <a href=3D#e5d3>Example 5.3 </a>is most probably false,=
 since some people never experience dogbite. Examples like 5.3 and 4.4 (mig=
ht there be some dogs which never have breathed, because they died as embry=
os?) indicate the danger in Lojban of universal claims even when restricted=
. In English we are prone to say that =93Everyone says=94 or that =93Everyb=
ody does=94 or that =93Everything is=94 when in fact there are obvious coun=
terexamples which we are ignoring for the sake of making a rhetorical point=
. Such statements are plain falsehoods in Lojban, unless saved by a context=
 (such as tense) which implicitly restricts them.
+<p>
+<cx "se, using to re-order logical variables">  XE "se: using to re-order =
logical variables"   <cx "re-ordering logical variables with se">  XE "re-o=
rdering logical variables with se"   <lx "se">  XE "se"   <lx "SE">  XE "SE=
 selma=92o"   How can we express <a href=3D#e5d3>Example 5.3 </a>in Lojban =
without a prenex? Since it is the order in which variables appear that matt=
ers, we can say:
+<p>
+<pre><a name=3De5d6>5.6)	ro da poi prenu cu se batci de poi gerku
+	Every-X which is-a-person is-bitten-by some-Y which is-a-dog.
+</pre>using the conversion operator =93se=94 (explained in <a href=3Dchap5=
.html>Chapter 5</a>) to change the selbri =93batci=94 (=93bites=94) into =
=93se batci=94 (=93is bitten by=94). The translation given in <a href=3D#e5=
d4>Example 5.4 </a>uses the corresponding strategy in English, since Englis=
h does not have prenexes (except in strained =93logician's English=94). Thi=
s implies that a sentence with both a universal and an existential variable=
 can't be freely converted with =93se=94; one must be careful to preserve t=
he order of the variables.
+<p>
+<cx "logical variable with ro, in multiple appearances">  XE "logical vari=
ables: with ro, in multiple appearances"   <cx "logical variable with poi, =
in multiple appearances">  XE "logical variables: with poi, in multiple app=
earances"   <cx "ro, dropping from multiple appearances on logical variable=
s">  XE "ro: dropping from multiple appearances on logical variables"   <cx=
 "poi, dropping from multiple appearances on logical variables">  XE "poi: =
dropping from multiple appearances on logical variables"   <lx "ro">  XE "r=
o"   <lx "poi">  XE "poi"   If a variable occurs more than once, then any =
=93ro=94 or =93poi=94 decorations are moved only to the first occurrence of=
 the variable when the prenex is dropped. For example,
+<p>
+<ex "weapon against self">  XE "weapon against self: example=93  <pre><a n=
ame=3De5d7>5.7)	di poi prenu zo'u=20
+ti xarci di di
+	There-is-a-Z which is-a-person :
+this-thing is-a-weapon for-use-against-Z by-Z
+	This is a weapon for someone to use against himself/herself.
+</pre>(in which =93di=94 is used rather than =93da=94 just for variety) lo=
ses its prenex as follows:
+<p>
+<pre><a name=3De5d8>5.8)	ti xarci di poi prenu ku'o di
+	This-thing is-a-weapon-for-use-against some-Z which is-a-person by-Z.
+</pre><cx "prenex, dropping for terseness">  XE "prenex: dropping for ters=
eness"   As the examples in this section show, dropping the prenex makes fo=
r terseness of expression often even greater than that of English (Lojban i=
s meant to be an unambiguous language, not necessarily a terse or verbose o=
ne), provided the rules are observed.
+<p>
+<a name=3Ds6><h3>Variables with generalized quantifiers</h3>
+<p>
+<lx "PA">  XE "PA selma=92o"   <cx "quantifiers, with logical variables"> =
 XE =93quantifier: with logical variables"   So far, we have seen variables=
 with either nothing in front, or with the cmavo =93ro=94 in front. Now =93=
ro=94 is a Lojban number, and means =93all=94; thus =93ro prenu=94 means =
=93all persons=94, just as =93re prenu=94 means =93two persons=94. In fact,=
 unadorned =93da=94 is also taken to have an implicit number in front of it=
, namely =93su'o=94, which means =93at least one=94. Why is this? Consider =
<a href=3D#e2d3>Example 2.3 </a>again, this time with an explicit =93su'o=
=94:
+<p>
+<cx "something, expressing using =93su'o=94">  XE "something: expressing u=
sing \=93su'o\=94"   <ex "something sees me">  XE "something sees me: examp=
le=93  <lx "su'o">  XE "su'o"   <pre><a name=3De6d1>6.1)	su'o da zo'u da vi=
ska mi
+	For-at-least-one X :  X sees me.
+	Something sees me.
+</pre><p>From this version of <a href=3D#e2d3>Example 2.3</a>, we understa=
nd the speaker's claim to be that of all the things that there are, at leas=
t one of them sees him or her. The corresponding universal claim, <a href=
=3D#e3d2>Example 3.2</a>, says that of all the things that exist, every one=
 of them can see the speaker.
+<p>
+<cx "numbers, on logical variables">  XE "numbers: on logical variables"  =
 Any other number can be used instead of =93ro=94 or =93su'o=94 to precede =
a variable. Then we get claims like:
+<p>
+<pre><a name=3De6d2>6.2)	re da zo'u da viska mi
+	For-two-Xes :  X sees me.
+	Two things see me.
+</pre><cx "numbers, Lojban contrasted with English on exactness">  XE "num=
bers: Lojban contrasted with English on exactness"   <cx "numbers, English =
contrasted with Lojban on exactness">  XE "numbers: English contrasted with=
 Lojban on exactness"   This means that exactly two things, no more or less=
, saw the speaker on the relevant occasion. In English, we might take =93Tw=
o things see me=94 to mean that at least two things see the speaker, but th=
ere might be more; in Lojban, though, that claim would have to be made as:
+<p>
+<pre><a name=3De6d3>6.3)	su'ore da zo'u da viska mi
+	For-at-least-two Xes :  X sees me.
+</pre>which would be false if nothing, or only one thing, saw the speaker,=
 but not otherwise. We note the =93su'o=94 here meaning =93at least=94; =93=
su'o=94 by itself is short for =93su'opa=94 where =93pa=94 means =93one=94,=
 as is explained in <a href=3Dchap18.html>Chapter 18</a>.
+<p>
+<cx "prenex, removing when numeric quantifiers present">  XE "prenex: remo=
ving when numeric quantifiers present"   The prenex may be removed from <a =
href=3D#e6d{2}>Examples 6.2 </a>and <a href=3D#e6d3>6.3 </a>as from the oth=
ers, leading to:
+<p>
+<pre><a name=3De6d4>6.4)	re da viska mi
+	Two Xes see me.
+</pre>and
+<p>
+<pre><a name=3De6d5>6.5)	su'ore da viska mi
+	At-least-two Xes see me.
+</pre>respectively, subject to the rules prescribed in <a href=3D#s5>Secti=
on 5</a>.
+<p>
+<lx "ro prenu">  XE "ro prenu"   <ex "all persons">  XE "all persons: exam=
ple=93  <cx "indefinite description, definition">  XE "indefinite descripti=
on: definition"   <cx "indefinite description, compared with restricted var=
iable">  XE "indefinite description: compared with restricted variable"   <=
cx "restricted variable, compared with indefinite description">  XE "restri=
cted variable: compared with indefinite description"   Now we can explain t=
he constructions =93ro prenu=94 for =93all persons=94 and =93re prenu=94 fo=
r =93two persons=94 which were casually mentioned at the beginning of this =
Section. In fact, =93ro prenu=94, a so-called =93indefinite description=94,=
 is shorthand for =93ro DA poi prenu=94, where =93DA=94 represents a fictit=
ious variable that hasn't been used yet and will not be used in future. (Ev=
en if all three of =93da=94, =93de=94, and =93di=94 have been used up, it d=
oes not matter, for there are ways of getting more variables, discussed in =
<a href=3D#s14>Section 14</a>.) So in fact
+<p>
+<pre><a name=3De6d6>6.6)	re prenu viska mi
+	Two persons see me.
+</pre>is short for
+<p>
+<pre><a name=3De6d7>6.7)	re da poi prenu cu viska mi
+	Two Xes which are-persons see me.
+</pre>which in turn is short for:
+<p>
+<pre><a name=3De6d8>6.8)	re da poi prenu zo'u da viska mi
+	For-two Xes which are-persons :  X sees me.
+</pre><cx "order of variables, in moving to prenex">  XE "order of variabl=
es: in moving to prenex"   Note that when we move more than one variable to=
 the prenex (along with its attached relative clause), we must make sure th=
at the variables are in the same order in the prenex as in the bridi proper=
.
+<p>
+<a name=3Ds7><h3>Grouping of quantifiers</h3>
+<p>
+<cx "indefinite sumti, multiple in sentence">  XE "indefinite sumti: multi=
ple in sentence"   <cx "distribution of quantified sumti">  XE "distributio=
n of quantified sumti"   Let us consider a sentence containing two quantifi=
er expressions neither of which is =93ro=94 or =93su'o=94 (remembering that=
 =93su'o=94 is implicit where no explicit quantifier is given):
+<p>
+<ex "three dogs bite two men">  XE  selma=92o"dogs bite: example"    XE "t=
hree dogs bite two men: example=93  <pre><a name=3De7d1>7.1)	ci gerku cu ba=
tci re nanmu
+	Three dogs bite two men.
+</pre><cx "indefinite sumti, meaning when multiple in sentence">  XE "inde=
finite sumti: meaning when multiple in sentence"   <cx "multiple indefinite=
 sumti, meaning">  XE "multiple indefinite sumti: meaning"   The question r=
aised by <a href=3D#e7d1>Example 7.1 </a>is, does each of the dogs bite the=
 same two men, or is it possible that there are two different men per dog, =
for six men altogether? If the former interpretation is taken, the number o=
f men involved is fixed at two; but if the latter, then the speaker has to =
be taken as saying that there might be any number of men between two and si=
x inclusive. Let us transform <a href=3D#e7d1>Example 7.1 </a>step by step =
as we did with <a href=3D#e6d6>Example 6.6</a>:
+<p>
+<pre><a name=3De7d2>7.2)	ci da poi gerku cu batci re de poi nanmu
+	Three Xes which are-dogs bite two Ys which are-men.
+</pre>(Note that we need separate variables =93da=94 and =93de=94, because=
 of the rule that says each indefinite description gets a variable never us=
ed before or since.)
+<p>
+<pre><a name=3De7d3>7.3)	ci da poi gerku ku'o re de poi nanmu zo'u da batc=
i de
+	For-three Xes which are-dogs, for-two Ys which are-men :  X bites Y.
+</pre><p>Here we see that indeed each of the dogs is said to bite two men,=
 and it might be different men each time; a total of six biting events alto=
gether.
+<p>
+<cx "multiple indefinite sumti, effect of re-ordering in sentence">  XE "m=
ultiple indefinite sumti: effect of re-ordering in sentence"   How then are=
 we to express the other interpretation, in which just two men are involved=
? We cannot just reverse the order of variables in the prenex to
+<p>
+<pre><a name=3De7d4>7.4)	re de poi nanmu ku'o ci da poi gerku zo'u da batc=
i de
+	For-two Ys which are-men, for-three Xes which are-dogs, X bites Y
+</pre>for although we have now limited the number of men to exactly two, w=
e end up with an indeterminate number of dogs, from three to six. The disti=
nction is called a =93scope distinction=94: in <a href=3D#e7d2>Example 7.2<=
/a>, =93ci gerku=94 is said to have wider scope than =93re nanmu=94, and th=
erefore precedes it in the prenex. In <a href=3D#e7d4>Example 7.4 </a>the r=
everse is true.
+<p>
+<cx "multiple indefinite sumti, expressing with equal scope">  XE "multipl=
e indefinite sumti: expressing with equal scope"   <lx "ce'e">  XE "ce'e"  =
 <lx "CEhE">  XE "CEhE selma=92o"   <lx "nu'i">  XE "nu'i"   <lx "nu'u">  X=
E "nu'u"   <lx "NUhI">  XE "NUhI selma=92o"   <lx "NUhU">  XE "NUhU selma=
=92o"   <cx "multiple indefinite sumti scope, in termset">  XE "multiple in=
definite sumti scope: in termset"   <cx "termset, effect on scope of multip=
le indefinite sumti">  XE "termset: effect on scope of multiple indefinite =
sumti"   The solution is to use a termset, which is a group of terms either=
 joined by =93ce'e=94 (of selma'o CEhE) between each term, or else surround=
ed by =93nu'i=94 (of selma'o NUhI) on the front and =93nu'u=94 (of selma'o =
NUhU) on the rear. Terms (which are either sumti or sumti prefixed by tense=
 or modal tags) that are grouped into a termset are understood to have equa=
l scope:
+<p>
+<pre><a name=3De7d5>7.5)	ci gerku ce'e re nanmu cu batci
+	nu'i ci gerku re nanmu [nu'u] cu batci
+	Three dogs [plus] two men, bite.
+</pre>which picks out two groups, one of three dogs and the other of two m=
en, and says that every one of the dogs bites each of the men. The second L=
ojban version uses forethought; note that =93nu'u=94 is an elidable termina=
tor, and in this case can be freely elided.
+<p>
+<cx "quantified sumti, different types contrasted for scope for distributi=
on">  XE "quantified sumti: different types contrasted for scope for distri=
bution"   <cx "sumti with lo, compared to indefinite sumti">  XE "sumti wit=
h lo: compared to indefinite sumti"   <cx "indefinite sumti, compared to su=
mti with lo">  XE "indefinite sumti: compared to sumti with lo"   <cx "ro, =
effect of order when multiple in sentence">  XE "ro: effect of order when m=
ultiple in sentence"   What about descriptors, like =93ci lo gerku=94, =93l=
e nanmu=94 or =93re le ci mlatu=94? They too can be grouped in termsets, bu=
t usually need not be, except for the =93lo=94 case which functions like th=
e case without a descriptor. Unless an actual quantifier precedes it, =93le=
 nanmu=94 means =93ro le nanmu=94, as is explained in <a href=3Dchap6.html>=
Chapter 6</a>. Two sumti with =93ro=94 quantifiers are independent of order=
, so:
+<p>
+<pre><a name=3De7d6>7.6)	[ro] le ci gerku cu batci [ro] le re nanmu
+	[All of] the three dogs bite [all of] the two men.
+</pre>means that each of the dogs specified bites each of the men specifie=
d, for six acts of biting altogether. However, if there is an explicit quan=
tifier before =93le=94 other than =93ro=94, the problems of this section re=
appear.
+<p>
+<a name=3Ds8><h3>The problem of =93any=94</h3>
+<p>
+Consider the English sentence
+<p>
+<ex "anyone who goes, walks">  XE "anyone who goes: walks, example=93  <cx=
 "any, as a translation problem">  XE "any: as a translation problem"   <pr=
e><a name=3De8d1>8.1)	Anyone who goes to the store, walks across the field.
+</pre><p>Using the facilities already discussed, a plausible translation m=
ight be
+<p>
+<pre><a name=3De8d2>8.2)	ro da poi klama le zarci cu cadzu le foldi
+	All X such-that-it goes-to the store walks-on the field.
+	Everyone who goes to the store walks across the field.
+</pre><cx "any, as a restricted universal claim">  XE "any: as a restricte=
d universal claim"   <cx "anyone, contrasted with everyone in assumption of=
 existence">  XE "anyone: contrasted with everyone in assumption of existen=
ce"   <cx "everyone, contrasted with anyone in assumption of existence">  X=
E "everyone: contrasted with anyone in assumption of existence"   But there=
 is a subtle difference between <a href=3D#e8d1>Example 8.1 </a>and <a href=
=3D#e8d2>Example 8.2</a>. <a href=3D#e8d2>Example 8.2 </a>tells us that, in=
 fact, there are people who go to the store, and that they walk across the =
field. A sumti of the type =93ro da poi klama=94 requires that there are th=
ings which =93klama=94: Lojban universal claims always imply the correspond=
ing existential claims as well. <a href=3D#e8d1>Example 8.1</a>, on the oth=
er hand, does not require that there are any people who go to the store: it=
 simply states, conditionally, that if there is anyone who goes to the stor=
e, he or she walks across the field as well. This conditional form mirrors =
the true Lojban translation of <a href=3D#e8d1>Example 8.1</a>:
+<p>
+<pre><a name=3De8d3>8.3)	ro da zo'u ganai da klama le zarci gi cadzu le fo=
ldi
+	For-every X:  if X is-a-goer-to the store then X is-a-walker-on the field=
.
+</pre><cx "any, as a universal claim, later restricted">  XE "any: as a un=
iversal claim, later restricted"   Although <a href=3D#e8d3>Example 8.3 </a=
>is a universal claim as well, its universality only implies that there are=
 objects of some sort or another in the universe of discourse. Because the =
claim is conditional, nothing is implied about the existence of goers-to-th=
e-store or of walkers-on-the-field, merely that any entity which is one is =
also the other.
+<p>
+<cx "any, as an existential claim">  XE "any: as an existential claim"   T=
here is another use of =93any=94 in English that is not universal but exist=
ential. Consider
+<p>
+<cx "need any box">  XE "need any box"   <cx "any box">  XE "any box"   <p=
re><a name=3De8d4>8.4)	I need any box that is bigger than this one.
+</pre><a href=3D#e8d4>Example 8.4 </a>does not at all mean that I need eve=
ry box bigger than this one, for indeed I do not; I require only one box. B=
ut the naive translation
+<p>
+<pre><a name=3De8d5>8.5)	mi nitcu da poi tanxe gi'e bramau ti
+	I need some-X which is-a-box and is-bigger-than this-one
+</pre>does not work either, because it asserts that there really is such a=
 box, as the prenex paraphrase demonstrates:
+<p>
+<pre><a name=3De8d6>8.6)	da poi tanxe gi'e bramau ti zo'u mi nitcu da
+	There-is-an-X which is-a-box and is-bigger-than this :  I need X.
+</pre><p>What to do? Well, the x2 place of =93nitcu=94 can be filled with =
an event as well as an object, and in fact <a href=3D#e8d5>Example 8.5 </a>=
can also be paraphrased as:
+<p>
+<pre><a name=3De8d7>8.7)	mi nitcu lo nu mi ponse 	lo tanxe poi bramau ti
+	I need an event-of I possess some box(es) which-are bigger-than this-one.
+</pre><p>Rewritten using variables, <a href=3D#e8d7>Example 8.7 </a>become=
s
+<p>
+<pre><a name=3De8d8>8.8)	mi nitcu lo nu da zo'u
+		da se ponse mi gi'e tanxe gi'e bramau ti
+	I need an event-of there-being an-X such-that
+		X is-possessed-by me and is-a-box and is-bigger-than this-thing.
+</pre><cx "prenex, internal to a bridi">  XE "prenex: internal to a bridi"=
   <cx "logical variable, implicit placement in smallest enclosing bridi pr=
enex">  XE "logical variables: implicit placement in smallest enclosing bri=
di prenex"   <cx "logical variable, explicitly placing in outer prenex">  X=
E "logical variables: explicitly placing in outer prenex"   So we see that =
a prenex can be attached to a bridi that is within a sentence. By default, =
a variable always behaves as if it is bound in the prenex which (notionally=
) is attached to the smallest enclosing bridi, and its scope does not exten=
d beyond that bridi. However, the variable may be placed in an outer prenex=
 explicitly:
+<p>
+<pre><a name=3De8d9>8.9)	da poi tanxe gi'e bramau ti zo'u
+		mi nitcu le nu mi ponse da
+	There-is-an-X which is-a-box and is-bigger-than this-one such-that :
+		I need the event-of my possessing X.
+</pre><cx "existential variable, in abstraction contrasted with in main br=
idi">  XE "existential variable: in abstraction contrasted with in main bri=
di"   <cx "existential variable, in main bridi contrasted with in abstracti=
on">  XE "existential variable: in main bridi contrasted with in abstractio=
n"   But what are the implications of <a href=3D#e8d7>Example 8.7 </a>and <=
a href=3D#e8d9>Example 8.9</a>? The main difference is that in <a href=3D#e=
8d9>Example 8.9</a>, the =93da=94 is said to exist in the real world of the=
 outer bridi; but in <a href=3D#e8d7>Example 8.7</a>, the existence is only=
 within the inner bridi, which is a mere event that need not necessarily co=
me to pass. So <a href=3D#e8d9>Example 8.9 </a>means
+<p>
+<pre><a name=3De8d10>8.10)	There's a box, bigger than this one, that I nee=
d.
+</pre><cx "any, expressing as existential by variable in subordinate bridi=
">  XE "any: expressing as existential by variable in subordinate bridi"   =
which is what <a href=3D#e8d6>Example 8.6 </a>says, whereas <a href=3D#e8d7=
>Example 8.7 </a>turns out to be an effective translation of our original <=
a href=3D#e8d1>Example 8.1</a>. So uses of =93any=94 that aren't universal =
end up being reflected by variables bound in the prenex of a subordinate br=
idi.
+<p>
+<a name=3Ds9><h3>Negation boundaries</h3>
+<p>
+<cx "=93there is a Y=94, notation convention">  XE "there is a Y: expressi=
on, notation convention"   This section, as well as <a href=3D#s10>Section =
10 </a>through <a href=3D#s12>Section 12</a>, are in effect a continuation =
of <a href=3Dchap15.html>Chapter 15</a>, introducing features of Lojban neg=
ation that require an understanding of prenexes and variables. In the examp=
les below, =93there is a Y=94 and the like must be understood as =93there i=
s at least one Y, possibly more=94.
+<p>
+<cx "bridi negation, two forms of">  XE "bridi negation: two forms of"   A=
s explained in <a href=3Dchap15.html>Chapter 15</a>, the negation of a brid=
i is usually accomplished by inserting =93na=94 at the beginning of the sel=
bri:
+<p>
+<lx "na">  XE "na"   <pre><a name=3De9d1>9.1)	mi na klama le zarci
+	I [false] go-to the store.
+	It is false that I go to the store.
+	I don't go to the store.
+</pre><lx "naku">  XE "naku"   The other form of bridi negation is express=
ed by using the compound cmavo =93naku=94 in the prenex, which is identifie=
d and compounded by the lexer before looking at the sentence grammar. In Lo=
jban grammar, =93naku=94 is then treated like a sumti. In a prenex, =93naku=
=94 means precisely the same thing as the logician's =93it is not the case =
that=94 in a similar English context. (Outside of a prenex, =93naku=94 is a=
lso grammatically treated as a single entity =97 the equivalent of a sumti =
=97 but does not have this exact meaning; we'll discuss these other situati=
ons in <a href=3D#s11>Section 11</a>.)
+<p>
+<cx "bridi negation, naku in prenex compared to na before selbri">  XE "br=
idi negation: naku in prenex compared to na before selbri"   <cx "bridi neg=
ation, na before selbri compared to naku in prenex">  XE "bridi negation: n=
a before selbri compared to naku in prenex"   <cx "external bridi negation,=
 definition">  XE "external bridi negation: definition"   <cx "internal bri=
di negation, definition">  XE "internal bridi negation: definition"   <cx "=
internal bridi negation, compared to external bridi negation">  XE "interna=
l bridi negation: compared to external bridi negation"   <cx "external brid=
i negation, compared to internal bridi negation">  XE "external bridi negat=
ion: compared to internal bridi negation"   To represent a bridi negation u=
sing a prenex, remove the =93na=94 from before the selbri and place =93naku=
=94 at the left end of the prenex. This form is called =93external bridi ne=
gation=94, as opposed to =93internal bridi negation=94 using =93na=94. The =
prenex version of <a href=3D#e9d1>Example 9.1 </a>is
+<p>
+<pre><a name=3De9d2>9.2)	naku zo'u la djan. klama
+	It is not the case that:  John comes.
+	It is false that:  John comes.
+</pre><cx "negation in prenex, effects of position">  XE "negation in pren=
ex: effects of position"   However, =93naku=94 can appear at other points i=
n the prenex as well. Compare
+<p>
+<ex "nothing sits">  XE "nothing sits: example=93  <pre><a name=3De9d3>9.3=
)	naku de zo'u de zutse
+	It is not the case that:  for some Y, Y sits.
+	It is false that:  for at least one Y, Y sits.
+	It is false that something sits.
+	Nothing sits.
+</pre>with
+<p>
+<pre><a name=3De9d4>9.4)	su'ode naku zo'u de zutse
+	For at least one Y, it is false that:  Y sits.
+	There is something that doesn't sit.
+</pre><p>The relative position of negation and quantification terms within=
 a prenex has a drastic effect on meaning. Starting without a negation, we =
can have:
+<p>
+<ex "everybody loves something">  XE "everybody loves something: example=
=93  <pre><a name=3De9d5>9.5)	roda su'ode zo'u da prami de
+	For every X, there is a Y, such that X loves Y.
+	Everybody loves at least one thing (each, not necessarily the same thing)=
.
+</pre>or:
+<p>
+<ex "something is loved by everybody">  XE "something is loved by everybod=
y: example=93  <pre><a name=3De9d6>9.6)	su'ode roda zo'u da prami de
+	There is a Y, such that for each X, X loves Y.
+	There is at least one particular thing that is loved by everybody.
+</pre><p>The simplest form of bridi negation to interpret is one where the=
 negation term is at the beginning of the prenex:
+<p>
+<pre><a name=3De9d7>9.7)	naku roda su'ode zo'u da prami de
+	It is false that:  for every X, there is a Y, such that :  X loves Y.
+	It is false that:  everybody loves at least one thing.
+	(At least) someone doesn't love anything.
+</pre>the negation of <a href=3D#e9d5>Example 9.5</a>, and
+<p>
+<pre><a name=3De9d8>9.8)	naku su'ode roda zo'u da prami de
+	It is false that:  there is a Y such that for each X, X loves Y.
+	It is false that:  there is at least one thing that is loved by everybody=
.
+	There isn't any one thing that everybody loves.
+</pre>the negation of <a href=3D#e9d6>Example 9.6</a>.
+<p>
+<cx "negation boundary, effect of moving">  XE "negation boundary: effect =
of moving"   <cx "inversion of quantifiers, in moving negation boundary">  =
XE "inversion of quantifiers: in moving negation boundary"   <cx "inversion=
 of quantifiers, definition">  XE "inversion of quantifiers: definition"   =
The rules of formal logic require that, to move a negation boundary within =
a prenex, you must =93invert any quantifier=94 that the negation boundary p=
asses across. Inverting a quantifier means that any =93ro=94 (all) is chang=
ed to =93su'o=94 (at least one) and vice versa. Thus, <a href=3D#e9d7>Examp=
le 9.7 </a>and <a href=3D#e9d8>Example 9.8 </a>can be restated as, respecti=
vely:
+<p>
+<pre><a name=3De9d9>9.9)	su'oda naku su'ode zo'u da prami de
+	For some X, it is false that:  there is a Y such that : X loves Y.
+	There is somebody who doesn't love anything.
+</pre>and:
+<p>
+<pre><a name=3De9d10>9.10)	rode naku roda zo'u da prami de
+	For every Y, it is false that:  for every X, X loves Y.
+	For each thing, it is not true that everybody loves it.
+</pre><p>Another movement of the negation boundary produces:
+<p>
+<pre><a name=3De9d11>9.11)	su'oda rode naku zo'u da prami de
+	There is an X such that, for every Y, it is false that X loves Y.
+	There is someone who, for each thing, doesn't love that thing.
+ and
+<a name=3De9d12>9.12)	rode su'oda naku zo'u da prami de
+	For every Y, there is an X, such that it is false that: X loves Y.
+	For each thing there is someone who doesn't love it.
+</pre><cx "inversion of quantifiers on passing negation boundary, rational=
e for">  XE "inversion of quantifiers on passing negation boundary: rationa=
le for"   Investigation will show that, indeed, each transformation preserv=
es the meanings of <a href=3D#e9d7>Example 9.7 </a>and <a href=3D#e9d8>Exam=
ple 9.8</a>.
+<p>
+<cx "zero, relation to negation boundary">  XE "zero: relation to negation=
 boundary"   <cx "negation boundary, and zero">  XE "negation boundary: and=
 zero"   The quantifier =93no=94 (meaning =93zero of=94) also involves a ne=
gation boundary. To transform a bridi containing a variable quantified with=
 =93no=94, we must first expand it. Consider
+
+<p>
+<pre><a name=3De9d13>9.13)	noda rode zo'u da prami de
+	There is no X, for every Y, such that X loves Y.
+	Nobody loves everything.
+</pre>which is negated by:
+<p>
+<pre><a name=3De9d14>9.14)	naku noda rode zo'u da prami de
+	It is false that:  there is no X that, for every Y, X loves Y.
+	It is false that there is nobody who loves everything.
+</pre><cx "expanding =93no=94 quantifier">  XE "expanding \=93no\=94 quant=
ifier"   <cx "=93no=94 quantifier, expanding">  XE "no quantifier: expandin=
g"   <cx "noda, expanding to naku su'oda">  XE "noda: expanding to naku su'=
oda"   <cx "naku su'oda, as expansion of noda">  XE "naku su'oda: as expans=
ion of noda"   We can simplify <a href=3D#e9d14>Example 9.14 </a>by transfo=
rming the prenex. To move the negation phrase within the prenex, we must fi=
rst expand the =93no=94 quantifier. Thus =93for no x=94 means the same thin=
g as =93it is false for some x=94, and the corresponding Lojban =93noda=94 =
can be replaced by =93naku su'oda=94. Making this substitution, we get:
+<p>
+<pre><a name=3De9d15>9.15)	naku naku su'oda rode zo'u da prami de
+	It is false that it is false that:  for an X, for every Y:  X loves Y.
+</pre><p>Adjacent negation boundaries in the prenex can be dropped, so thi=
s means the same as:
+<p>
+<pre><a name=3De9d16>9.16)	su'oda rode zo'u da prami de
+	There is an X such that, for every Y, X loves Y.
+	At least one person loves everything.
+</pre>which is clearly the desired contradiction of <a href=3D#e9d13>Examp=
le 9.13</a>.
+<p>
+<cx "interactions between quantifiers and negation, effect">  XE "interact=
ions between quantifiers and negation: effect"   <cx "double negatives, eff=
ect of interactions between quantifiers and negation on">  XE "double negat=
ives: effect of interactions between quantifiers and negation on"   The int=
eractions between quantifiers and negation mean that you cannot eliminate d=
ouble negatives that are not adjacent. You must first move the negation phr=
ases so that they are adjacent, inverting any quantifiers they cross, and t=
hen the double negative can be eliminated.
+<p>
+<a name=3Ds10><h3>bridi negation and logical connectives</h3>
+<p>
+<cx "negation and logical connectives, caveat for logic chapter discussion=
s">  XE "negation and logical connectives: caveat for logic chapter discuss=
ions"   <cx "logical connectives and negation, caveat for logic chapter dis=
cussions">  XE "logical connectives and negation: caveat for logic chapter =
discussions"   <cx "bridi negation and logical connectives">  XE "bridi neg=
ation and logical connectives"   <cx "logical connectives and bridi negatio=
n">  XE "logical connectives and bridi negation"   A complete discussion of=
 logical connectives appears in <a href=3Dchap14.html>Chapter 14</a>. What =
is said here is intentionally quite incomplete and makes several oversimpli=
fications.
+<p>
+A logical connective is a cmavo or compound cmavo. In this chapter, we wil=
l make use of the logical connectives =93and=94 and =93or=94 (where =93or=
=94 really means =93and/or=94, =93either or both=94). The following simplif=
ied recipes explain how to make some logical connectives:
+<p>
+<cx "logical connective recipes, simplified for logic chapter discussion">=
  XE =93logical connectives: recipes, simplified for logic chapter discussi=
on"   <dl compact><dt><dd>To logically connect two Lojban sumti with =93and=
=94, put them both in the bridi and separate them with the cmavo =93.e=94.
+<p><dt><dd>To logically connect two Lojban bridi with =93and=94, replace t=
he regular separator cmavo =93.i=94 with the compound cmavo =93.ije=94.
+<p><dt><dd>To logically connect two Lojban sumti with =93or=94, put them b=
oth in the bridi and separate them with the cmavo =93.a=94.
+<p><dt><dd>To logically connect two Lojban bridi with =93or=94, replace th=
e regular separator cmavo =93.i=94 with the compound cmavo =93.ija=94.
+</dl><p>More complex logical connectives also exist; in particular, one ma=
y place =93na=94 before =93.e=94 or =93.a=94, or between =93.i=94 and =93je=
=94 or =93ja=94; likewise, one may place =93nai=94 at the end of a connecti=
ve. Both =93na=94 and =93nai=94 have negative effects on the sumti or bridi=
 being connected. Specifically, =93na=94 negates the first or left-hand sum=
ti or bridi, and =93nai=94 negates the second or right-hand one.
+<p>
+Whenever a logical connective occurs in a sentence, that sentence can be e=
xpanded into two sentences by repeating the common terms and joining the se=
ntences by a logical connective beginning with =93.i=94. Thus the following=
 sentence:
+<p>
+<pre><a name=3De10d1>10.1)	mi .e do klama ti
+	I and you come here.
+</pre>can be expanded to:
+<p>
+<pre><a name=3De10d2>10.2)	mi klama ti  .ije do klama ti
+	I come here, and, you come here.
+</pre><p>The same type of expansion can be performed for any logical conne=
ctive, with any valid combination of =93na=94 or =93nai=94 attached. No cha=
nge in meaning occurs under such a transformation.
+<p>
+Clearly, if we know what negation means in the expanded sentence forms, th=
en we know what it means in all of the other forms. But what does negation =
mean between sentences?
+<p>
+<cx "negation between sentences, compared with bridi negation">  XE "negat=
ion between sentences: compared with bridi negation"   <cx "bridi negation,=
 compared with negation between sentences">  XE "bridi negation: compared w=
ith negation between sentences"   <cx "negation between sentences, meaning =
of">  XE "negation between sentences: meaning of"   The mystery is easily s=
olved. A negation in a logical expression is identical to the corresponding=
 bridi negation, with the negator placed at the beginning of the prenex. Th=
us:
+<p>
+<pre><a name=3De10d3>10.3)	mi .enai do prami roda
+	I, and not you, love everything.
+</pre>expands to:
+<p>
+<pre><a name=3De10d4>10.4)	mi prami roda  .ijenai do prami roda
+	I love everything, and-not, you love everything.
+</pre>and then into prenex form as:
+<p>
+<pre><a name=3De10d5>10.5)	roda zo'u mi prami da  .ije naku zo'u do prami =
da
+	For each thing:  I love it, and it is false that you love (the same) it.
+</pre><cx "quantifier scope, in multiple connected sentences">  XE "quanti=
fier scope: in multiple connected sentences"   By the rules of predicate lo=
gic, the =93ro=94 quantifier on =93da=94 has scope over both sentences. Tha=
t is, once you've picked a value for =93da=94 for the first sentence, it st=
ays the same for both sentences. (The =93da=94 continues with the same fixe=
d value until a new paragraph or a new prenex resets the meaning.)
+<p>
+Thus the following example has the indicated translation:
+<p>
+<pre><a name=3De10d6>10.6)	su'oda zo'u mi prami da
+		.ije naku zo'u do prami da
+	For at least one thing:  I love that thing.
+		And it is false that:  you love that (same) thing.
+	There is something that I love that you don't.
+
+</pre><cx "prenex manipulation, rules">  XE "prenex manipulation: rules"  =
 If you remember only two rules for prenex manipulation of negations, you w=
on't go wrong:
+<p>
+<cx "prenex manipulation, moving naku past bound variable">  XE "prenex ma=
nipulation: moving naku past bound variable"   <dl compact><dt><dd>Within a=
 prenex, whenever you move =93naku=94 past a bound variable (da, de, di, et=
c.), you must invert the quantifier.
+</dl><cx "prenex manipulation, importing na from selbri">  XE "prenex mani=
pulation: importing na from selbri"   <cx "prenex manipulation, exporting n=
a from left of prenex">  XE "prenex manipulation: exporting na from left of=
 prenex"   <dl compact><dt><dd>A =93na=94 before the selbri is always trans=
formed into a =93naku=94 at the left-hand end of the prenex, and vice versa=
.
+</dl><a name=3Ds11><h3>Using =93naku=94 outside a prenex</h3>
+<p>
+Let us consider the English sentence
+<p>
+<ex "some do not go to school">  XE "some do not go to school: example=93 =
 <pre><a name=3De11d1>11.1)	Some children do not go to school.
+</pre><p>We cannot express this directly with =93na=94; the apparently obv=
ious translation
+<p>
+<pre><a name=3De11d2>11.2)	su'oda poi verba na klama
+		su'ode poi ckule
+	At-least-one X which-are child(ren) [false] go-to=20
+at-least-one Y which-are school(s).
+</pre>when converted to the external negation form produces:
+<p>
+<pre><a name=3De11d3>11.3)	naku zo'u su'oda poi verba cu klama su'ode poi =
ckule
+	It is false that some which are children go-to some which are schools.
+	All children don't go to some school (not just some children).
+</pre><cx "negation, form for emulating natural language negation">  XE "n=
egation: form for emulating natural language negation"   <cx "naku, outside=
 of prenex">  XE "naku: outside of prenex"   <cx "contradictory negation, u=
sing naku before selbri">  XE "contradictory negation: using naku before se=
lbri"   Lojban provides a negation form which more closely emulates natural=
 language negation. This involves putting =93naku=94 before the selbri, ins=
tead of a =93na=94. =93naku=94 is clearly a contradictory negation, given i=
ts parallel with prenex bridi negation. Using =93naku=94, <a href=3D#e11d1>=
Example 11.1 </a>can be expressed as:
+<p>
+<pre><a name=3De11d4>11.4)	su'oda poi verba naku klama su'ode poi ckule
+	Some which-are children don't go-to some which-are schools.
+	Some children don't go to a school.
+</pre><cx "naku, compared with sumti in grammar">  XE "naku: compared with=
 sumti in grammar"   Although it is not technically a sumti, =93naku=94 can=
 be used in most of the places where a sumti may appear. We'll see what thi=
s means in a moment.
+<p>
+<cx "naku, as creating a negation boundary">  XE "naku: as creating a nega=
tion boundary"   <cx "naku, effect on moving quantifiers">  XE "naku: effec=
t on moving quantifiers"   <cx "quantifiers, effect of moving naku">  XE =
=93quantifier: effect of moving naku"   <cx "inverting quantifiers, with mo=
vement relative to naku">  XE "inverting quantifiers: with movement relativ=
e to naku"   When you use =93naku=94 within a bridi, you are explicitly cre=
ating a negation boundary. As explained in <a href=3D#s9>Section 9</a>, whe=
n a prenex negation boundary expressed by =93naku=94 moves past a quantifie=
r, the quantifier has to be inverted. The same is true for =93naku=94 in th=
e bridi proper. We can move =93naku=94 to any place in the sentence where a=
 sumti can go, inverting any quantifiers that the negation boundary crosses=
. Thus, the following are equivalent to <a href=3D#e11d4>Example 11.4 </a>(=
no good English translations exist):
+<p>
+<pre><a name=3De11d5>11.5)	su'oda poi verba cu klama rode poi ckule naku
+	For some children, for every school, they don't go to it.
+<a name=3De11d6>11.6)	su'oda poi verba cu klama naku su'ode poi ckule
+	Some children don't go to (some) school(s).
+<a name=3De11d7>11.7)	naku roda poi verba cu klama su'ode poi ckule
+	It is false that all children go to some school(s).
+</pre><p>In <a href=3D#e11d5>Example 11.5</a>, we moved the negation bound=
ary rightward across the quantifier of =93de=94, forcing us to invert it. I=
n <a href=3D#e11d7>Example 11.7 </a>we moved the negation boundary across t=
he quantifier of =93da=94, forcing us to invert it instead. <a href=3D#e11d=
6>Example 11.6 </a>merely switched the selbri and the negation boundary, wi=
th no effect on the quantifiers.
+<p>
+<cx "inverting quantifiers, with movement relative to fixed negation">  XE=
 "inverting quantifiers: with movement relative to fixed negation"   The sa=
me rules apply if you rearrange the sentence so that the quantifier crosses=
 an otherwise fixed negation. You can't just convert the selbri of <a href=
=3D#e11d4>Example 11.4 </a>and rearrange the sumti to produce
+<p>
+<pre><a name=3De11d8>11.8)	su'ode poi ckule ku'o naku se klama roda poi ve=
rba
+	Some schools aren't gone-to-by every child.
+</pre><cx "conversion with se, effect of naku negation boundary on">  XE "=
conversion with se: effect of naku negation boundary on"   <cx "naku negati=
on boundary, effect on conversion with se">  XE "naku negation boundary: ef=
fect on conversion with se"   <cx "naku negation, rationale for considering=
 an advanced technique">  XE "naku negation: rationale for considering an a=
dvanced technique"   or rather, <a href=3D#e11d8>Example 11.8 </a>means som=
ething completely different from <a href=3D#e11d4>Example 11.4</a>. Convers=
ion with =93se=94 under =93naku=94 negation is not symmetric; not all sumti=
 are treated identically, and some sumti are not invariant under conversion=
. Thus, internal negation with =93naku=94 is considered an advanced techniq=
ue, used to achieve stylistic compatibility with natural languages.
+<p>
+It isn't always easy to see which quantifiers have to be inverted in a sen=
tence. <a href=3D#e11d4>Example 11.4 </a>is identical in meaning to:
+<p>
+<pre><a name=3De11d9>11.9)	su'o verba naku klama su'o ckule
+	Some children don't go-to some school.
+</pre><cx "indefinite sumti, as implicit quantification">  XE "indefinite =
sumti: as implicit quantification"   but in <a href=3D#e11d9>Example 11.9</=
a>, the bound variables =93da=94 and =93de=94 have been hidden.
+<p>
+<cx "exporting negation to prenex, internal bridi negation contrasted with=
 =93naku=94">  XE "exporting negation to prenex: internal bridi negation co=
ntrasted with \=93naku\=94"   <cx "exporting negation to prenex, =93naku=94=
 contrasted with internal bridi negation">  XE "exporting negation to prene=
x: \=93naku\=94 contrasted with internal bridi negation"   It is trivial to=
 export an internal bridi negation expressed with =93na=94 to the prenex, a=
s we saw in <a href=3D#s9>Section 9</a>; you just move it to the left end o=
f the prenex. In comparison, it is non-trivial to export a =93naku=94 to th=
e prenex because of the quantifiers. The rules for exporting =93naku=94 req=
uire that you export all of the quantified variables (implicit or explicit)=
 along with =93naku=94, and you must export them from left to right, in the=
 same order that they appear in the sentence. Thus <a href=3D#e11d4>Example=
 11.4 </a>goes into prenex form as:
+<p>
+<pre><a name=3De11d10>11.10)	su'oda poi verba ku'o naku
+		su'ode poi ckule zo'u da klama de
+	For some X which is a child, it is not the case that
+		there is a Y which is a school such that:  X goes to Y.
+</pre><p>We can now move the =93naku=94 to the left end of the prenex, get=
ting a contradictory negation that can be expressed with =93na=94:
+<p>
+<pre><a name=3De11d11>11.11)	naku roda poi verba
+		su'ode poi ckule zo'u da klama de
+	It is not the case that for all X's which are children,
+		there is a Y which is a school such that:  X goes to Y.
+</pre>from which we can restore the quantified variables to the sentence, =
giving:
+<p>
+<pre><a name=3De11d12>11.12)	naku zo'u roda poi verba cu klama su'ode poi =
ckule
+	It is not the case that all children go to some school.
+</pre>or more briefly
+<p>
+<pre><a name=3De11d13>11.13)	ro verba cu na klama su'o ckule
+	All children [false] go-to some school(s).
+</pre><cx "conversion of sentence with quantified variables, technique">  =
XE "conversion of sentence with quantified variables: technique"   As noted=
 in <a href=3D#s5>Section 5</a>, a sentence with two different quantified v=
ariables, such as <a href=3D#e11d13>Example 11.13</a>, cannot always be con=
verted with =93se=94 without first exporting the quantified variables. When=
 the variables have been exported, the sentence proper can be converted, bu=
t the quantifier order in the prenex must remain unchanged:
+<p>
+<pre><a name=3De11d14>11.14)	roda poi verba
+		su'ode poi ckule zo'u de na se klama da
+	It is not the case that for all X's which are children,
+		there is a Y which is a school such that: Y is gone to by X.
+</pre><cx "selbri placement among sumti, effect of multiple quantification=
 on">  XE "selbri placement among sumti: effect of multiple quantification =
on"   <cx "multiple quantification, effect on selbri placement among sumti"=
>  XE "multiple quantification: effect on selbri placement among sumti"   W=
hile you can't freely convert with =93se=94 when you have two quantified va=
riables in a sentence, you can still freely move sumti to either side of th=
e selbri, as long as the order isn't changed. If you use =93na=94 negation =
in such a sentence, nothing special need be done. If you use =93naku=94 neg=
ation, then quantified variables that cross the negation boundary must be i=
nverted.
+<p>
+<cx "negation manipulation, =93naku=94 contrasted with =93na=94 in difficu=
lty of">  XE "negation manipulation: \=93naku\=94 contrasted with \=93na\=
=94 in difficulty of"   <cx "negation manipulation, =93na=94 contrasted wit=
h =93naku=94 in difficulty of">  XE "negation manipulation: \=93na\=94 cont=
rasted with \=93naku\=94 in difficulty of"   Clearly, if all of Lojban nega=
tion was built on =93naku=94 negation instead of =93na=94 negation, logical=
 manipulation in Lojban would be as difficult as in natural languages. In <=
a href=3D#s12>Section 12</a>, for example, we'll discuss DeMorgan's Law, wh=
ich must be used whenever a sumti with a logical connection is moved across=
 a negation boundary.
+<p>
+<cx "naku, in linked sumti places">  XE "naku: in linked sumti places"   S=
ince =93naku=94 has the grammar of a sumti, it can be placed almost anywher=
e a sumti can go, including =93be=94 and =93bei=94 clauses; it isn't clear =
what these mean, and we recommend avoiding such constructs.
+<p>
+<cx "naku, multiple in sentence">  XE "naku: multiple in sentence"   <cx "=
double negation, and naku">  XE "double negation: and naku"   You can put m=
ultiple =93naku=94s in a sentence, each forming a separate negation boundar=
y. Two adjacent =93naku=94s in a bridi are a double negative and cancel out=
:
+<p>
+<pre><a name=3De11d15>11.15)	mi naku naku le zarci cu klama
+</pre><p>Other expressions using two =93naku=94s may or may not cancel out=
. If there is no quantified variable between them, then the =93naku=94s can=
cel.
+<p>
+Negation with internal =93naku=94 is clumsy and non-intuitive for logical =
manipulations, but then, so are the natural language features it is emulati=
ng.
+<p>
+<a name=3Ds12><h3>Logical Connectives and DeMorgan's Law</h3>
+<p>
+<cx "logical connectives within negation, effects of expansion on">  XE "l=
ogical connectives within negation: effects of expansion on"   <cx "negatio=
ns with logical connectives, effects on expansion of sentence">  XE "negati=
ons with logical connectives: effects on expansion of sentence"   DeMorgan'=
s Law states that when a logical connective between terms falls within a ne=
gation, then expanding the negation requires a change in the connective. Th=
us (where =93p=94 and =93q=94 stand for terms or sentences) =93not (p or q)=
=94 is identical to =93not p and not q=94, and =93not (p and q)=94 is ident=
ical to =93not p or not q=94. The corresponding changes for the other two b=
asic Lojban connectives are: =93not (p equivalent to q)=94 is identical to =
=93not p exclusive-or not q=94, and =93not (p whether-or-not q)=94 is ident=
ical to both =93not p whether-or-not q=94 and =93not p whether-or-not not q=
=94. In any Lojban sentence having one of the basic connectives, you can su=
bstitute in either direction from these identities. (These basic connective=
s are explained in <a href=3Dchap14.html>Chapter 14</a>.)
+<p>
+The effects of DeMorgan's Law on the logical connectives made by modifying=
 the basic connectives with =93nai=94, =93na=94 and =93se=94 can be derived=
 directly from these rules; modify the basic connective for DeMorgan's Law =
by substituting from the above identities, and then, apply each =93nai=94, =
=93na=94 and =93se=94 modifier of the original connectives. Cancel any doub=
le negatives that result.
+<p>
+<cx "distributing a negation">  XE "distributing a negation"   <cx "DeMorg=
an's Law, and distributing a negation">  XE "DeMorgan's Law: and distributi=
ng a negation"   <cx "DeMorgan's Law, and moving a logical connective relat=
ive to =93naku=94">  XE "DeMorgan's Law: and moving a logical connective re=
lative to \=93naku\=94"   When do we apply DeMorgan's Law? Whenever we wish=
 to =93distribute=94 a negation over a logical connective; and, for interna=
l =93naku=94 negation, whenever a logical connective moves in to, or out of=
, the scope of a negation =97 when it crosses a negation boundary.
+<p>
+<cx "DeMorgan's Law, sample applications">  XE "DeMorgan's Law: sample app=
lications"   <lx "ga">  XE "ga"   <lx "ge">  XE "ge"   <lx "gi">  XE "gi"  =
 <lx "nai">  XE "nai"   Let us apply DeMorgan's Law to some sample sentence=
s. These sentences make use of forethought logical connectives, which are e=
xplained in <a href=3Dchap14.html>Chapter 14</a>. It suffices to know that =
=93ga=94 and =93gi=94, used before each of a pair of sumti or bridi, mean =
=93either=94 and =93or=94 respectively, and that =93ge=94 and =93gi=94 used=
 similarly mean =93both=94 and =93and=94. Furthermore, =93ga=94, =93ge=94, =
and =93gi=94 can all be suffixed with =93nai=94 to negate the bridi or sumt=
i that follows.
+<p>
+<lx "na">  XE "na"   <lx "naku zo'u">  XE "naku zo'u"   <cx "na, and negat=
ion boundary">  XE "na: and negation boundary"   <cx "naku zo'u, and negati=
on boundary">  XE "naku zo'u: and negation boundary"   <cx "bridi negation,=
 and negation boundary">  XE "bridi negation: and negation boundary"   <cx =
"bridi negation, and DeMorgan's Law">  XE "bridi negation: and DeMorgan's L=
aw"   We have defined =93na=94 and =93naku zo'u=94 as, respectively, intern=
al and external bridi negation. These forms being identical, the negation b=
oundary always remains at the left end of the prenex. Thus, exporting or im=
porting negation between external and internal bridi negation forms never r=
equires DeMorgan's Law to be applied. <a href=3D#e12d1>Example 12.1 </a>and=
 <a href=3D#e12d2>Example 12.2 </a>are exactly equivalent:
+<p>
+<ex "go to Paris or Rome">  XE "go to Paris or Rome: example=93  <pre><a n=
ame=3De12d1>12.1)	la djan. na klama ga la paris. gi la rom.
+	John [false] goes-to either Paris or Rome.
+<a name=3De12d2>12.2)	naku zo'u la djan. klama ga la paris. gi la rom.
+	It-is-false that:  John goes-to either Paris or Rome.
+</pre><cx "negator, movement from bridi to sumti">  XE "negator: movement =
from bridi to sumti"   <cx "logically connected sentences, and DeMorgan's L=
aw">  XE "logically connected sentences: and DeMorgan's Law"   <cx "DeMorga=
n's Law, and logically connected sentences">  XE "DeMorgan's Law: and logic=
ally connected sentences"   It is not an acceptable logical manipulation to=
 move a negator from the bridi level to one or more sumti. However, <a href=
=3D#e12d1>Example 12.1 </a>and related examples are not sumti negations, bu=
t rather expand to form two logically connected sentences. In such a situat=
ion, DeMorgan's Law must be applied. For instance, <a href=3D#e12d2>Example=
 12.2 </a>expands to:
+<p>
+<pre><a name=3De12d3>12.3)	ge la djan. la paris. na klama
+		gi la djan. la rom. na klama
+	[It is true that] both John, to-Paris, [false] goes,=20
+		and John, to-Rome, [false] goes.
+</pre><p>The =93ga=94 and =93gi=94, meaning =93either-or=94, have become =
=93ge=94 and =93gi=94, meaning =93both-and=94, as a consequence of moving t=
he negators into the individual bridi.
+<p>
+<cx "bridi-tail logical connection, and DeMorgan's Law">  XE "bridi-tail l=
ogical connection: and DeMorgan's Law"   <cx "DeMorgan's Law, and bridi-tai=
l logical connection">  XE "DeMorgan's Law: and bridi-tail logical connecti=
on"   Here is another example of DeMorgan's Law in action, involving bridi-=
tail logical connection (explained in <a href=3Dchap14.html>Chapter 14</a>)=
:
+<p>
+<pre><a name=3De12d4>12.4)	la djein. le zarci na ge dzukla gi bajrykla
+	Jane to-the market [false] both walks and runs.
+<a name=3De12d5>12.5)	la djein. le zarci ganai dzukla ginai bajrykla
+	Jane to-the market either [false] walks or [false] runs.
+	Jane to-the market if walks then ([false] runs).
+</pre>(Placing =93le zarci=94 before the selbri makes sure that it is prop=
erly associated with both parts of the logical connection. Otherwise, it is=
 easy to erroneously leave it off one of the two sentences.)
+<p>
+<cx "transformations with logical connectives, steps">  XE "transformation=
s with logical connectives: steps"   It is wise, before freely doing transf=
ormations such as the one from <a href=3D#e12d4>Example 12.4 </a>to <a href=
=3D#e12d5>Example 12.5</a>, that you become familiar with expanding logical=
 connectives to separate sentences, transforming the sentences, and then re=
condensing. Thus, you would prove the transformation correct by the followi=
ng steps. By moving its =93na=94 to the beginning of the prenex as a =93nak=
u=94, <a href=3D#e12d4>Example 12.4 </a>becomes:
+
+<p>
+<pre><a name=3De12d6>12.6)	naku zo'u la djein. le zarci ge dzukla gi bajry=
kla
+	It is false that :  Jane to-the market (both walks and runs).
+</pre><p>And by dividing the bridi with logically connected selbri into tw=
o bridi,
+<p>
+<pre><a name=3De12d7>12.7)	naku zo'u ge la djein. le zarci dzukla
+		gi la djein. le zarci bajrykla
+	It-is-false-that:  both (Jane to-the market walks)
+		and (Jane to-the market runs).
+</pre>is the result.
+<p>
+At this expanded level, we apply DeMorgan's Law to distribute the negation=
 in the prenex across both sentences, to get
+<p>
+<pre><a name=3De12d8>12.8)	ga la djein. le zarci na dzukla gi la djein. le=
 zarci na bajrykla
+	Either Jane to-the market [false] walks, or Jane to-the market [false] ru=
ns.
+</pre>which is the same as
+<p>
+<pre><a name=3De12d9>12.9)	ganai la djein. le zarci dzukla ginai la djein.=
 le zarci bajrykla
+	If Jane to-the market walks, then Jane to-the market [false] runs.
+	If Jane walks to the market, then she doesn't run.
+</pre>which then condenses down to <a href=3D#e12d5>Example 12.5</a>.
+<p>
+<cx "internal naku negations, and DeMorgan's Law">  XE "internal naku nega=
tions: and DeMorgan's Law"   <cx "DeMorgan's Law, and internal naku negatio=
ns">  XE "DeMorgan's Law: and internal naku negations"   DeMorgan's Law mus=
t also be applied to internal =93naku=94 negations:
+<p>
+<pre><a name=3De12d10>12.10)	ga la paris. gi la rom. naku se klama la djan=
.
+	(Either Paris or Rome) is-not gone-to-by John.
+<a name=3De12d11>12.11)	la djan. naku klama ge la paris. gi la rom.
+	John doesn't go-to both Paris and Rome.
+</pre><p>That <a href=3D#e12d10>Example 12.10 </a>and <a href=3D#e12d11>Ex=
ample 12.11 </a>mean the same should become evident by studying the English=
. It is a good exercise to work through the Lojban and prove that they are =
the same.
+<p>
+<a name=3Ds13><h3>selbri variables</h3>
+<p>
+<lx "GOhA">  XE "GOhA selma=92o"   <lx "bu'a">  XE "bu'a"   <lx "bu'e">  X=
E "bu'e"   <lx "bu'i">  XE "bu'i"   <cx "logical variables, for selbri">  X=
E "logical variables: for selbri"   <cx "selbri logical variables">  XE "se=
lbri logical variables"   In addition to the variables =93da=94, =93de=94, =
and =93di=94 that we have seen so far, which function as sumti and belong t=
o selma'o KOhA, there are three corresponding variables =93bu'a=94, =93bu'e=
=94, and =93bu'i=94 which function as selbri and belong to selma'o GOhA. Th=
ese new variables allow existential or universal claims which are about the=
 relationships between objects rather than the objects themselves. We will =
start with the usual silly examples; the literal translation will represent=
 =93bu'a=94, =93bu'e=94 and =93bu'i=94 with F, G, and H respectively.
+<p>
+<ex "some relationship">  XE "some relationship: example=93  <pre><a name=
=3De13d1>13.1)	su'o bu'a zo'u la djim. bu'a la djan.=20
+	For-at-least-one relationship-F :  Jim stands-in-relationship-F to-John.
+	There's some relationship between Jim and John.
+</pre><cx "selbri variables, prenex form as indefinite description">  XE "=
selbri variables: prenex form as indefinite description"   The translations=
 of <a href=3D#e13d1>Example 13.1 </a>show how unidiomatic selbri variables=
 are in English; Lojban sentences like <a href=3D#e13d1>Example 13.1 </a>ne=
ed to be totally reworded in English. Furthermore, when a selbri variable a=
ppears in the prenex, it is necessary to precede it with a quantifier such =
as =93su'o=94; it is ungrammatical to just say =93bu'a zo'u=94. This rule i=
s necessary because only sumti can appear in the prenex, and =93su'o bu'a=
=94 is technically a sumti =97 in fact, it is an indefinite description lik=
e =93re nanmu=94, since =93bu'a=94 is grammatically equivalent to a brivla =
like =93nanmu=94. However, indefinite descriptions involving the bu'a-serie=
s cannot be imported from the prenex.
+<p>
+<cx "selbri variables, form when not in prenex">  XE "selbri variables: fo=
rm when not in prenex"   When the prenex is omitted, the preceding number h=
as to be omitted too:
+<p>
+<pre><a name=3De13d2>13.2)	la djim. bu'a la djan.
+	Jim stands-in-at-least-one-relationship to-John.
+</pre><cx "selbri variables, quantified">  XE "selbri variables: quantifie=
d"   As a result, if the number before the variable is anything but =93su'o=
=94, the prenex is required:
+<p>
+<pre><a name=3De13d3>13.3)	ro bu'a zo'u la djim. bu'a la djan.
+	For-every relationship-F : Jim stands-in-relationship-F to-John.
+	Every relationship exists between Jim and John.
+</pre><a href=3D#e13d1>Example 13.1 </a>and <a href=3D#e13d2>Example 13.2 =
</a>are almost certainly true: Jim and John might be brothers, or might liv=
e in the same city, or at least have the property of being jointly human. <=
a href=3D#e13d3>Example 13.3 </a>is palpably false, however; if Jim and Joh=
n were related by every possible relationship, then they would have to be b=
oth brothers and father-and-son, which is impossible.
+<p>
+<a name=3Ds14><h3>A few notes on variables</h3>
+<p>
+<cx "quantifier, on previously quantified variable">  XE "quantifier: on p=
reviously quantified variable"   A variable may have a quantifier placed in=
 front of it even though it has already been quantified explicitly or impli=
citly by a previous appearance, as in:
+<p>
+<ex "three cats white, and two big">  XE "three cats white: and two big, e=
xample=93  <pre><a name=3De14d1>14.1)	ci da poi mlatu cu blaci .ije re da c=
u barda
+	Three Xs which-are cats are white, and two Xs are big.
+</pre><p>What does <a href=3D#e14d1>Example 14.1 </a>mean? The appearance =
of =93ci da=94 quantifies =93da=94 as referring to three things, which are =
restricted by the relative clause to be cats. When =93re da=94 appears late=
r, it refers to two of the those three things =97 there is no saying which =
ones. Further uses of =93da=94 alone, if there were any, would refer once m=
ore to the three cats, so the requantification of =93da=94 is purely local.
+<p>
+<cx "prenex scope, for sentences joined by ijeks">  XE "prenex scope: for =
sentences joined by ijeks"   <cx "prenex scope, for sentences joined by .i"=
>  XE "prenex scope: for sentences joined by .i"   <cx "prenex scope, infor=
mal">  XE "prenex scope: informal"   <cx "prenex scope, in embedded bridi">=
  XE "prenex scope: in embedded bridi"   <cx "prenex scope, in relative cla=
uses">  XE "prenex scope: in relative clauses"   <cx "prenex scope, in abst=
ractions">  XE "prenex scope: in abstractions"   In general, the scope of a=
 prenex that precedes a sentence extends to following sentences that are jo=
ined by ijeks (explained in <a href=3Dchap14.html>Chapter 14</a>) such as t=
he =93.ije=94 in <a href=3D#e14d1>Example 14.1</a>. Theoretically, a bare =
=93.i=94 terminates the scope of the prenex. Informally, however, variables=
 may persist for a while even after an =93.i=94, as if it were an =93.ije=
=94. Prenexes that precede embedded bridi such as relative clauses and abst=
ractions extend only to the end of the clause, as explained in <a href=3D#s=
8>Section 8</a>. A prenex preceding =93tu'e=A0<dots>=85</dots>tu'u=94 long-=
scope brackets persists until the =93tu'u=94, which may be many sentences o=
r even paragraphs later.
+<p>
+<cx "logical variables, creating more by subscripting">  XE "logical varia=
bles: creating more by subscripting"   <cx "subscripts, use with logical va=
riables">  XE "subscripts: use with logical variables"   If the variables =
=93da=94, =93de=94, and =93di=94 (or the selbri variables =93bu'a=94, =93bu=
'e=94, and =93bu'i=94) are insufficient in number for handling a particular=
 problem, the Lojban approach is to add a subscript to any of them. Each po=
ssible different combination of a subscript and a variable cmavo counts as =
a distinct variable in Lojban. Subscripts are explained in full in <a href=
=3Dchap19.html>Chapter 19</a>, but in general consist of the cmavo =93xi=94=
 (of selma'o XI) followed by a number, one or more lerfu words forming a si=
ngle string, or a general mathematical expression enclosed in parentheses.
+<p>
+A quantifier can be prefixed to a variable that has already been bound eit=
her in a prenex or earlier in the bridi, thus:
+
+<p>
+<pre><a name=3De14d2>14.2)	ci da poi prenu cu se ralju pa da
+	Three Xs which are-persons are-led-by one-of X
+	Three people are led by one of them.
+</pre><p>The =93pa da=94 in <a href=3D#e14d2>Example 14.2 </a>does not spe=
cify the number of things to which =93da=94 refers, as the preceding =93ci =
da=94 does. Instead, it selects one of them for use in this sumti only. The=
 number of referents of =93da=94 remains three, but a single one (there is =
no way of knowing which one) is selected to be the leader.
+<p>
+<a name=3Ds15><h3>Conclusion</h3>
+<p>
+<cx "logic and Lojban, more aspects">  XE "logic and Lojban: more aspects"=
   This chapter is incomplete. There are many more aspects of logic that I =
neither fully understand nor feel competent to explain, neither in abstract=
 nor in their Lojban realization. Lojban was designed to be a language that=
 makes predicate logic speakable, and achieving that goal completely will n=
eed to wait for someone who understands both logic and Lojban better than I=
 do. I can only hope to have pointed out the areas that are well-understood=
 (and by implication, those that are not).
+
+<p>
+</body></html>
+<center><p align=3Dcenter>
+<img src=3Dchap17.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+
+<h2>Chapter 17
+<br>
+As Easy As A-B-C? The Lojban Letteral System And Its Uses</h2>
+<h6>$Revision: 4.1 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>What's a letteral, anyway?</h3>
+<p>
+<cx "Brown, James Cooke, and =93letteral=94">  XE "Brown: James Cooke, and=
 \=93letteral\=94"   <cx "letteral, definition">  XE "letteral: definition"=
   <cx "letter, alphabet">  XE "letter: alphabet"   James Cooke Brown, the =
founder of the Loglan Project, coined the word =93letteral=94 (by analogy w=
ith =93numeral=94) to mean a letter of the alphabet, such as =93f=94 or =93=
z=94. A typical example of its use might be
+<p>
+<ex "fourteen =93e=94s">  XE "fourteen \=93e\=94s: example=93  <pre><a nam=
e=3De1d1>1.1)	There are fourteen occurrences of the letteral
+		=93e=94 in this sentence.
+</pre><cx "lerfu, definition">  XE "lerfu: definition"   (Don't forget the=
 one within quotation marks.) Using the word =93letteral=94 avoids confusio=
n with =93letter=94, the kind you write to someone. Not surprisingly, there=
 is a Lojban gismu for =93letteral=94, namely =93lerfu=94, and this word wi=
ll be used in the rest of this chapter.
+<p>
+<cx "Latin, alphabet of Lojban">  XE "Latin: alphabet of Lojban"   <cx "al=
phabet, Latin used for Lojban">  XE "alphabet: Latin used for Lojban"   Loj=
ban uses the Latin alphabet, just as English does, right? Then why is there=
 a need for a chapter like this? After all, everyone who can read it alread=
y knows the alphabet. The answer is twofold:
+<p>
+<cx "alphabet, words for letters in, rationale">  XE "alphabet: words for =
letters in, rationale"   First, in English there are a set of words that co=
rrespond to and represent the English lerfu. These words are rarely written=
 down in English and have no standard spellings, but if you pronounce the E=
nglish alphabet to yourself you will hear them: ay, bee, cee, dee=A0<dots>=
=85</dots>. They are used in spelling out words and in pronouncing most acr=
onyms. The Lojban equivalents of these words are standardized and must be d=
ocumented somehow.
+<p>
+<cx "alphabets, words for non-Lojban letters, rationale">  XE "alphabet: w=
ords for non-Lojban letters, rationale"   Second, English has names only fo=
r the lerfu used in writing English. (There are also English names for Gree=
k and Hebrew lerfu: English-speakers usually refer to the Greek lerfu conve=
ntionally spelled =93phi=94 as =93fye=94, whereas =93fee=94 would more near=
ly represent the name used by Greek-speakers. Still, not all English-speake=
rs know these English names.) Lojban, in order to be culturally neutral, ne=
eds a more comprehensive system that can handle, at least potentially, all =
of the world's alphabets and other writing systems.
+<p>
+Letterals have several uses in Lojban: in forming acronyms and abbreviatio=
ns, as mathematical symbols, and as pro-sumti =97 the equivalent of English=
 pronouns.
+<p>
+<cx "lerfu, contrasted with lerfu word">  XE "lerfu: contrasted with lerfu=
 word"   <cx "lerfu word, contrasted with lerfu">  XE "lerfu word: contrast=
ed with lerfu"   <cx "letter, contrasted with word for the letter">  XE "le=
tter: contrasted with word for the letter"   In earlier writings about Lojb=
an, there has been a tendency to use the word =93lerfu=94 for both the lett=
erals themselves and for the Lojban words which represent them. In this cha=
pter, that tendency will be ruthlessly suppressed, and the term =93lerfu wo=
rd=94 will invariably be used for the latter. The Lojban equivalent would b=
e =93lerfu valsi=94 or =93lervla=94.
+<p>
+<a name=3Ds2><h3>A to Z in Lojban, plus one</h3>
+<p>
+<cx "lerfu words, Lojban coverage requirement">  XE "lerfu words: Lojban c=
overage requirement"   The first requirement of a system of lerfu words for=
 any language is that they must represent the lerfu used to write the langu=
age. The lerfu words for English are a motley crew: the relationship betwee=
n =93doubleyou=94 and =93w=94 is strictly historical in nature; =93aitch=94=
 represents =93h=94 but has no clear relationship to it at all; and =93z=94=
 has two distinct lerfu words, =93zee=94 and =93zed=94, depending on the di=
alect of English in question.
+
+<p>
+<cx "lerfu words, formation rules">  XE "lerfu words: formation rules"   <=
cx "lerfu words, for vowels">  XE "lerfu words: for vowels"   <cx "lerfu wo=
rds, for consonants">  XE "lerfu words: for consonants"   <cx "lerfu word, =
for =93 ' =94">  XE "lerfu word: for \=93 ' \=94"   <lx "BY">  XE "BY selma=
=92o"   All of Lojban's basic lerfu words are made by one of three rules:
+<p>
+<dl compact><dt><dd>to get a lerfu word for a vowel, add =93bu=94;
+to get a lerfu word for a consonant, add =93y=94;
+the lerfu word for =93 ' =94is =93.y'y=94.
+</dl><cx "lerfu words, table of Lojban">  XE "lerfu words: table of Lojban=
"   Therefore, the following table represents the basic Lojban alphabet:
+<p>
+<pre> 	'	a	b	c	d	e
+	.y'y.	.abu	by.	cy.	dy.	.ebu
+	f	g	i	j	k	l
+	fy.	gy.	.ibu	jy.	ky.	ly.
+	m	n	o	p	r	s
+	my.	ny.	.obu	py.	ry.	sy.
+	t	u	v	x	y	z
+	ty.	.ubu	vy.	xy.	.ybu	zy.
+</pre><cx "lerfu words for vowels, pause requirement before">  XE "lerfu w=
ords for vowels: pause requirement before"   <cx "lerfu words, consonant wo=
rds contrasted with vowel words">  XE "lerfu words: consonant words contras=
ted with vowel words"   <cx "lerfu words, vowel words contrasted with conso=
nant words">  XE "lerfu words: vowel words contrasted with consonant words"=
   <cx "lerfu words, composed of single cmavo">  XE "lerfu words: composed =
of single cmavo"   <cx "lerfu words, composed of compound cmavo">  XE "lerf=
u words: composed of compound cmavo"   <lx "bu">  XE "bu"   <lx "BU">  XE "=
BU selma=92o"   <cx "bu, effect on preceding word">  XE "bu: effect on prec=
eding word"   There are several things to note about this table. The conson=
ant lerfu words are a single syllable, whereas the vowel and =93 ' =94 lerf=
u words are two syllables and must be preceded by pause (since they all beg=
in with a vowel). Another fact, not evident from the table but important no=
netheless, is that =93by=94 and its like are single cmavo of selma'o BY, as=
 is =93.y'y=94. The vowel lerfu words, on the other hand, are compound cmav=
o, made from a single vowel cmavo plus the cmavo =93bu=94 (which belongs to=
 its own selma'o, BU). All of the vowel cmavo have other meanings in Lojban=
 (logical connectives, sentence separator, hesitation noise), but those mea=
nings are irrelevant when =93bu=94 follows.
+<p>
+Here are some illustrations of common Lojban words spelled out using the a=
lphabet above:
+<p>
+<pre><a name=3De2d1>2.1)	ty. .abu ny. ry. .ubu
+	=93t=94 =93a=94 =93n=94 =93r=94 =93u=94
+<a name=3De2d2>2.2)	ky. .obu .y'y. .abu
+	=93k=94 =93o=94 =93 ' =94 =93a=94
+</pre><cx "spelling out words, Lojban contrasted with English in usefulnes=
s">  XE "spelling out words: Lojban contrasted with English"   <cx "lerfu w=
ords, effect of systematic formulation">  XE "lerfu words: effect of system=
atic formulation"   Spelling out words is less useful in Lojban than in Eng=
lish, for two reasons: Lojban spelling is phonemic, so there can be no real=
 dispute about how a word is spelled; and the Lojban lerfu words sound more=
 alike than the English ones do, since they are made up systematically. The=
 English words =93fail=94 and =93vale=94 sound similar, but just hearing th=
e first lerfu word of either, namely =93eff=94 or =93vee=94, is enough to d=
iscriminate easily between them =97 and even if the first lerfu word were s=
omehow confused, neither =93vail=94 nor =93fale=94 is a word of ordinary En=
glish, so the rest of the spelling determines which word is meant. Still, t=
he capability of spelling out words does exist in Lojban.
+<p>
+<cx "lerfu words ending with =93y=94, pause after, rationale">  XE "lerfu =
words ending with \=93y\=94: pause after, rationale"   Note that the lerfu =
words ending in =93y=94 were written (in <a href=3D#e2d1>Example 2.1 </a>an=
d <a href=3D#e2d2>Example 2.2</a>) with pauses after them. It is not strict=
ly necessary to pause after such lerfu words, but failure to do so can in s=
ome cases lead to ambiguities:
+<p>
+<pre><a name=3De2d3>2.3)	mi cy. claxu
+	I lerfu-=93c=94 without
+	I am without (whatever is referred to by) the letter =93c=94.
+</pre>without a pause after =93cy=94 would be interpreted as:
+<p>
+<pre><a name=3De2d4>2.4)	micyclaxu
+	(Observative:) doctor-without
+	Something unspecified is without a doctor.
+</pre><p>A safe guideline is to pause after any cmavo ending in =93y=94 un=
less the next word is also a cmavo ending in =93y=94. The safest and easies=
t guideline is to pause after all of them.
+<p>
+<a name=3Ds3><h3>Upper and lower cases</h3>
+<p>
+<cx "upper-case letters, Lojban usage contrasted with English">  XE "upper=
-case letters: Lojban usage contrasted with English"   <cx "upper-case lett=
ers, English usage contrasted with Lojban">  XE "upper-case letters: Englis=
h usage contrasted with Lojban"   <cx "lower-case letters, Lojban usage con=
trasted with English">  XE "lower-case letters: Lojban usage contrasted wit=
h English"   <cx "lower-case letters, English usage contrasted with Lojban"=
>  XE "lower-case letters: English usage contrasted with Lojban"   <cx "str=
ess, irregular marked with upper-case">  XE "stress: irregular marked with =
upper-case"   <cx "capital letters, use in Lojban">  XE "capital letters: u=
se in Lojban"   <cx "lower case letters, use in Lojban">  XE "lower case le=
tters: use in Lojban"   Lojban doesn't use lower-case (small) letters and u=
pper-case (capital) letters in the same way that English does; sentences do=
 not begin with an upper-case letter, nor do names. However, upper-case let=
ters are used in Lojban to mark irregular stress within names, thus:
+<p>
+<pre><a name=3De3d1>3.1)	.iVAN.
+	the name =93Ivan=94 in Russian/Slavic pronunciation.
+</pre><cx "upper-case, lerfu word for">  XE "upper-case: lerfu word for"  =
 <cx "lower-case, lerfu word for">  XE "lower-case: lerfu word for"   <lx "=
ga'e">  XE "ga'e"   <lx "to'a">  XE "to'a"   <cx "case, upper/lower specifi=
cation">  XE "case: upper/lower specification"   It would require far too m=
any cmavo to assign one for each upper-case and one for each lower-case ler=
fu, so instead we have two special cmavo =93ga'e=94 and =93to'a=94 represen=
ting upper case and lower case respectively. They belong to the same selma'=
o as the basic lerfu words, namely BY, and they may be freely interspersed =
with them.
+<p>
+<lx "ga'e">  XE "ga'e"   <cx "lower-case word, effect on following lerfu w=
ords">  XE "lower-case word: effect on following lerfu words"   The effect =
of =93ga'e=94 is to change the interpretation of all lerfu words following =
it to be the upper-case version of the lerfu. An occurrence of =93to'a=94 c=
auses the interpretation to revert to lower case. Thus, =93ga'e .abu=94 mea=
ns not =93a=94 but =93A=94, and Ivan's name may be spelled out thus:
+<p>
+<pre><a name=3De3d2>3.2)	.ibu ga'e vy. .abu ny. to'a
+	i [upper] V A N [lower]
+</pre>The cmavo and compound cmavo of this type will be called =93shift wo=
rds=94.
+<p>
+<cx "shift word, scope">  XE "shift word: scope"   How long does a shift w=
ord last? Theoretically, until the next shift word that contradicts it or u=
ntil the end of text. In practice, it is common to presume that a shift wor=
d is only in effect until the next word other than a lerfu word is found.
+<p>
+<cx "shift word, for single letter">  XE "shift word: for single letter"  =
 <lx "LAU">  XE "LAU selma=92o"   <cx "shift, single-letter, grammar of">  =
XE "shift: single-letter, grammar of"   It is often convenient to shift jus=
t a single letter to upper case. The cmavo =93tau=94, of selma'o LAU, is us=
eful for the purpose. A LAU cmavo must always be immediately followed by a =
BY cmavo or its equivalent: the combination is grammatically equivalent to =
a single BY. (See <a href=3D#s14>Section 14 </a>for details.)
+<p>
+<cx "chemical elements, use of single-letter shift for">  XE "chemical ele=
ments: use of single-letter shift for"   A likely use of =93tau=94 is in th=
e internationally standardized symbols for the chemical elements. Each elem=
ent is represented using either a single upper-case lerfu or one upper-case=
 lerfu followed by one lower-case lerfu:
+<p>
+<pre><a name=3De3d3>3.3)	tau sy.
+	[single shift] S
+	S (chemical symbol for sulfur)
+</pre><pre><a name=3De3d4>3.4)	tau sy. .ibu
+	[single shift] S i
+	Si (chemical symbol for silicon)
+</pre><cx "single-letter shift, as toggle">  XE "single-letter shift: as t=
oggle"   If a shift to upper-case is in effect when =93tau=94 appears, it s=
hifts the next lerfu word only to lower case, reversing its usual effect.
+<p>
+<a name=3Ds4><h3>The universal =93bu=94</h3>
+<p>
+<cx "bu, for extension of lerfu word set">  XE "bu: for extension of lerfu=
 word set"   <cx "lerfu word set extension, with bu">  XE "lerfu word set e=
xtension: with bu"   So far we have seen =93bu=94 only as a suffix to vowel=
 cmavo to produce vowel lerfu words. Originally, this was the only use of =
=93bu=94. In developing the lerfu word system, however, it proved to be use=
ful to allow =93bu=94 to be attached to any word whatsoever, in order to al=
low arbitrary extensions of the basic lerfu word set.
+<p>
+<cx "bu, grammar of">  XE "bu: grammar of"   <cx "bu, and compound cmavo">=
  XE "bu: and compound cmavo"   <cx "bu, interaction with ba'e">  XE "bu: i=
nteractions"   <cx "ba'e, interaction with bu">  XE "ba'e: interaction with=
 bu"   <cx "za'e, interaction with bu">  XE "za'e: interaction with bu"   <=
cx "zei, interaction with bu">  XE "zei: interaction with bu"   <cx "zo, in=
teraction with bu">  XE "zo: interaction with bu"   <cx "zoi, interaction w=
ith bu">  XE "zoi: interaction with bu"   <cx "la'o, interaction with bu"> =
 XE "la'o: interaction with bu"   <cx "lo'u, interaction with bu">  XE "lo'=
u: interaction with bu"   <cx "si, interaction with bu">  XE "si: interacti=
on with bu"   <cx "sa, interaction with bu">  XE "sa: interaction with bu" =
  <cx "su, interaction with bu">  XE "su: interaction with bu"   <cx "fa'o,=
 interaction with bu">  XE "fa'o: interaction with bu"   <lx "ba'e">  XE "b=
a'e"   <lx "za'e">  XE "za'e"   <lx "zei">  XE "zei"   <lx "zo">  XE "zo"  =
 <lx "zoi">  XE "zoi"   <lx "la'o">  XE "la'o"   <lx "lo'u">  XE "lo'u"   <=
lx "si">  XE "si"   <lx "sa">  XE "sa"   <lx "su">  XE "su"   <lx "fa'o">  =
XE "fa'o"   Formally, =93bu=94 may be attached to any single Lojban word. C=
ompound cmavo do not count as words for this purpose. The special cmavo =93=
ba'e=94, =93za'e=94, =93zei=94, =93zo=94, =93zoi=94, =93la'o=94, =93lo'u=94=
, =93si=94, =93sa=94, =93su=94, and =93fa'o=94 may not have =93bu=94 attach=
ed, because they are interpreted before =93bu=94 detection is done; in part=
icular,
+<p>
+<ex "word =93bu=94">  XE "word \=93bu\=94: example=93   <pre><a name=3De4d=
1>4.1)	zo bu
+	the word =93bu=94
+</pre><cx "bu, effect of multiple">  XE "bu: effect of multiple"   <lx "bu=
bu">  XE "bubu"   <cx "name, pause requirement in lerfu words">  XE "names:=
 pause requirement in lerfu words"   is needed when discussing =93bu=94 in =
Lojban. It is also illegal to attach =93bu=94 to itself, but more than one =
=93bu=94 may be attached to a word; thus =93.abubu=94 is legal, if ugly. (I=
ts meaning is not defined, but it is presumably different from =93.abu=94.)=
 It does not matter if the word is a cmavo, a cmene, or a brivla. All such =
words suffixed by =93bu=94 are treated grammatically as if they were cmavo =
belonging to selma'o BY. However, if the word is a cmene it is always neces=
sary to precede and follow it by a pause, because otherwise the cmene may a=
bsorb preceding or following words.
+<p>
+<cx "unusual characters, words for">  XE "unusual characters: words for"  =
 <cx "smiley face, word for">  XE "smiley face: word for"   <cx "logograms,=
 words for">  XE "logograms: words for"   <ex "smiley face">  XE "smiley fa=
ce: example=93  <ex "happy face">  XE "happy face: example=93  The ability =
to attach =93bu=94 to words has been used primarily to make names for vario=
us logograms and other unusual characters. For example, the Lojban name for=
 the =93happy face=94 is =93.uibu=94, based on the attitudinal =93.ui=94 th=
at means =93happiness=94. Likewise, the =93smiley face=94, written =93:-)=
=94 and used on computer networks to indicate humor, is called =93zo'obu=94=
 The existence of these names does not mean that you should insert =93.uibu=
=94 into running Lojban text to indicate that you are happy, or =93zo'obu=
=94 when something is funny; instead, use the appropriate attitudinal direc=
tly.
+<p>
+<cx "=93&=94, word for">  XE "&: word for"   <cx "ampersand character, wor=
d for">  XE "ampersand character: word for"   <ex "ampersand">  XE "ampersa=
nd: example=93  Likewise, =93joibu=94 represents the ampersand character, =
=93&&amp;=94, based on the cmavo =93joi=94 meaning =93mixed and=94. Many mo=
re such lerfu words will probably be invented in future.
+<p>
+<cx "pause, symbol for">  XE "pause: symbol for"   <cx "syllable break, sy=
mbol for">  XE "syllable break: symbol for"   <cx "pause, word for">  XE "p=
ause: word for"   <cx "syllable break, word for">  XE "syllable break: word=
 for"   <cx "=93.=94, word for">  XE ". : word for"   <cx "=93,=94 , word f=
or">  XE ": , word for"   The =93.=94and =93,=94characters used in Lojbanic=
 writing to represent pause and syllable break respectively have been assig=
ned the lerfu words =93denpa bu=94 (literally, =93pause bu=94) and =93slaka=
 bu=94 (literally, =93syllable bu=94). The written space is mandatory here,=
 because =93denpa=94 and =93slaka=94 are normal gismu with normal stress: =
=93denpabu=94 would be a fu'ivla (word borrowed from another language into =
Lojban) stressed =93denPAbu=94. No pause is required between =93denpa=94 (o=
r =93slaka=94) and =93bu=94, though.
+<p>
+<a name=3Ds5><h3>Alien alphabets</h3>
+<p>
+As stated in <a href=3D#s1>Section 1</a>, Lojban's goal of cultural neutra=
lity demands a standard set of lerfu words for the lerfu of as many other w=
riting systems as possible. When we meet these lerfu in written text (parti=
cularly, though not exclusively, mathematical text), we need a standard Loj=
banic way to pronounce them.
+<p>
+There are certainly hundreds of alphabets and other writing systems in use=
 around the world, and it is probably an unachievable goal to create a sing=
le system which can express all of them, but if perfection is not demanded,=
 a usable system can be created from the raw material which Lojban provides=
.
+<p>
+<cx "letters, non-Lojban, representation with names">  XE "letters: non-Lo=
jban, representation with names"   <ex "alpha">  XE "alpha: example=93  One=
 possibility would be to use the lerfu word associated with the language it=
self, Lojbanized and with =93bu=94 added. Indeed, an isolated Greek =93alph=
a=94 in running Lojban text is probably most easily handled by calling it =
=93.alfas. bu=94. Here the Greek lerfu word has been made into a Lojbanized=
 name by adding =93s=94 and then into a Lojban lerfu word by adding =93bu=
=94. Note that the pause after =93.alfas.=94 is still needed.
+<p>
+<cx "letters, non-Lojban, representation with consonant-word + bu">  XE "l=
etters: non-Lojban, representation with consonant-word + bu"   Likewise, th=
e easiest way to handle the Latin letters =93h=94, =93q=94, and =93w=94 tha=
t are not used in Lojban is by a consonant lerfu word with =93bu=94 attache=
d. The following assignments have been made:
+<p>
+<pre>	.y'y.bu	h
+	ky.bu	q
+	vy.bu	w
+</pre><ex "quack">  XE "quack: example=93  As an example, the English word=
 =93quack=94 would be spelled in Lojban thus:
+<p>
+<pre><a name=3De5d1>5.1)	ky.bu .ubu .abu cy. ky.
+	=93q=94 =93u=94 =93a=94 =93c=94 =93k=94
+</pre><cx "letters, sound contrasted with symbol for spelling">  XE "lette=
rs: sound contrasted with symbol for spelling"   <cx "letters, symbol contr=
asted with sound for spelling">  XE "letters: symbol contrasted with sound =
for spelling"   Note that the fact that the letter =93c=94 in this word has=
 nothing to do with the sound of the Lojban letter =93c=94 is irrelevant; w=
e are spelling an English word and English rules control the choice of lett=
ers, but we are speaking Lojban and Lojban rules control the pronunciations=
 of those letters.
+<p>
+A few more possibilities for Latin-alphabet letters used in languages othe=
r than English:
+<p>
+<pre>			ty.bu		=FE&thorn; (thorn)
+			dy.bu		=F0&edh; (edh)
+<cx "letters, non-Lojban, representation with consonant-word + bu, drawbac=
k">  XE "letters: non-Lojban, representation with consonant-word + bu, draw=
back"   However, this system is not ideal for all purposes. For one thing, =
it is verbose. The native lerfu words are often quite long, and with =93bu=
=94 added they become even longer: the worst-case Greek lerfu word would be=
 =93.Omikron. bu=94, with four syllables and two mandatory pauses. In addit=
ion, alphabets that are used by many languages have separate sets of lerfu =
words for each language, and which set is Lojban to choose?
+</pre><cx "language shift, rationale for">  XE "language shift: rationale =
for"   <cx "language shift, effect on following words">  XE "language shift=
: effect on following words"   <cx "language shift, choice of Lojban-lerfu-=
word counterpart">  XE "language shift: choice of Lojban-lerfu-word counter=
part"   <cx "letters, non-Lojban, representation with language-shift">  XE =
"letters: non-Lojban, representation with language-shift"   The alternative=
 plan, therefore, is to use a shift word similar to those introduced in <a =
href=3D#s3>Section 3</a>. After the appearance of such a shift word, the re=
gular lerfu words are re-interpreted to represent the lerfu of the alphabet=
 now in use. After a shift to the Greek alphabet, for example, the lerfu wo=
rd =93ty=94 would represent not Latin =93t=94 but Greek =93tau=94. Why =93t=
au=94? Because it is, in some sense, the closest counterpart of =93t=94 wit=
hin the Greek lerfu system. In principle it would be all right to map =93ty=
.=94 to =93phi=94 or even =93omega=94, but such an arbitrary relationship w=
ould be extremely hard to remember.
+<p>
+<cx "language shift, interaction with bu">  XE "language shift: interactio=
n with bu"   <cx "bu, interaction with language shift">  XE "bu: interactio=
ns"   Where no obvious closest counterpart exists, some more or less arbitr=
ary choice must be made. Some alien lerfu may simply not have any shifted e=
quivalent, forcing the speaker to fall back on a =93bu=94 form. Since a =93=
bu=94 form may mean different things in different alphabets, it is safest t=
o employ a shift word even when =93bu=94 forms are in use.
+<p>
+Shifts for several alphabets have been assigned cmavo of selma'o BY:
+<p>
+<cx "Latin alphabet, language shift word for">  XE "Latin alphabet: langua=
ge shift word for"   <cx "Greek alphabet, language shift word for">  XE "Gr=
eek alphabet: language shift word for"   <cx "Hebrew alphabet, language shi=
ft word for">  XE "Hebrew alphabet: language shift word for"   <cx "Arabic =
alphabet, language shift word for">  XE "Arabic alphabet: language shift wo=
rd for"   <cx "Cyrillic alphabet, language shift word for">  XE "Cyrillic a=
lphabet: language shift word for"   <pre> 	lo'a	Latin/Roman/Lojban alphabet
+	ge'o	Greek alphabet
+	je'o	Hebrew alphabet
+	jo'o	Arabic alphabet
+	ru'o	Cyrillic alphabet
+</pre><lx "zai">  XE "zai"   <lx "LAU">  XE "LAU selma=92o"   <cx "languag=
e shift, formation of shift alphabet name">  XE "language shift: formation =
of shift alphabet name"   <cx "language shift, compound">  XE "language shi=
ft: compound"   <cx "language shift, based on name + bu">  XE "language shi=
ft: based on name + bu"   The cmavo =93zai=94 (of selma'o LAU) is used to c=
reate shift words to still other alphabets. The BY word which must follow a=
ny LAU cmavo would typically be a name representing the alphabet with =93bu=
=94 suffixed:
+<p>
+<ex "Devanagari">  XE "Devanagari: example=93  <pre><a name=3De5d2>5.2)	za=
i .devanagar. bu
+	Devanagari (Hindi) alphabet
+</pre><ex "katakana">  XE "katakana: example=93  <ex "Japanese katakana"> =
 XE "Japanese katakana: example=93  <pre><a name=3De5d3>5.3)	zai .katakan. =
bu
+	Japanese katakana syllabary
+</pre><ex "hiragana">  XE "hiragana: example=93  <ex "Japanese hiragana"> =
 XE "Japanese hiragana: example=93  <pre><a name=3De5d4>5.4)	zai .xiragan. =
bu
+	Japanese hiragana syllabary
+</pre><cx "language shift, standardization of">  XE "language shift: stand=
ardization of"   Unlike the cmavo above, these shift words have not been st=
andardized and probably will not be until someone actually has a need for t=
hem. (Note the =93.=94characters marking leading and following pauses.)
+<p>
+<ex "italic">  XE "italic: example=93  <ex "bold">  XE "bold: example=93  =
<cx "font, specifying for letters">  XE "font: specifying for letters"   <c=
x "face, specifying for letters">  XE "face: specifying for letters"   <cx =
"shift words, for font">  XE =93shift word: for font"   <cx "shift words, f=
or face">  XE =93shift word: for face"   <lx "ce'a">  XE "ce'a"   <lx "LAU"=
>  XE "LAU selma=92o"   In addition, there may be multiple visible represen=
tations within a single alphabet for a given letter: roman vs. italics, han=
dwriting vs. print, Bodoni vs. Helvetica. These traditional =93font and fac=
e=94 distinctions are also represented by shift words, indicated with the c=
mavo =93ce'a=94 (of selma'o LAU) and a following BY word:
+<p>
+<ex "Helvetica font">  XE "Helvetica font: example=93  <pre><a name=3De5d5=
>5.5)	ce'a .xelveticas. bu
+	Helvetica font
+</pre><ex "handwriting">  XE "handwriting: example=93  <pre><a name=3De5d6=
>5.6)	ce'a .xancisk. bu
+	handwriting
+</pre><ex "12-point">  XE "12-point: example=93  <ex "font">  XE "font: ex=
ample=93  <pre><a name=3De5d7>5.7)	ce'a .pavrel. bu
+	12-point font size
+</pre><cx "shift words, canceling effect">  XE =93shift word: canceling ef=
fect"   <lx "na'a">  XE "na'a"   <lx "BY">  XE "BY selma=92o"   <cx "cancel=
ing letter shifts">  XE "canceling letter shifts"   <cx "na'a, contrasted w=
ith lo'a">  XE "na'a: contrasted with lo'a"   <cx "lo'a, contrasted with na=
'a">  XE "lo'a: contrasted with na'a"   The cmavo =93na'a=94 (of selma'o BY=
) is a universal shift-word cancel: it returns the interpretation of lerfu =
words to the default of lower-case Lojban with no specific font. It is more=
 general than =93lo'a=94, which changes the alphabet only, potentially leav=
ing font and case shifts in place.
+<p>
+Several sections at the end of this chapter contain tables of proposed ler=
fu word assignments for various languages.
+<p>
+<a name=3Ds6><h3>Accent marks and compound lerfu words</h3>
+<p>
+<cx "diacritical marks, as lerfu">  XE "diacritical marks: as lerfu"   <cx=
 "letters, non-Lojban, representation of diacritical marks on">  XE "letter=
s: non-Lojban, representation of diacritical marks on"   <ex "accent mark, =
a diacritical mark">  XE "accent mark: a diacritical mark"   <ex "umlaut, a=
 diacritical mark">  XE "umlaut: a diacritical mark"   <ex "circumflex, a d=
iacritical mark">  XE "circumflex: a diacritical mark"   <ex "cedilla, a di=
acritical mark">  XE "cedilla: a diacritical mark"   <ex "tilde, a diacriti=
cal mark">  XE "tilde: a diacritical mark"   Many languages that make use o=
f the Latin alphabet add special marks to some of the lerfu they use. Frenc=
h, for example, uses three accent marks above vowels, called (in English) =
=93acute=94, =93grave=94, and =93circumflex=94. Likewise, German uses a mar=
k called =93umlaut=94; a mark which looks the same is also used in French, =
but with a different name and meaning.
+<p>
+<cx "diacritical marks, problem of position">  XE "diacritical marks: prob=
lem of position"   These marks may be considered lerfu, and each has a corr=
esponding lerfu word in Lojban. So far, no problem. But the marks appear ov=
er lerfu, whereas the words must be spoken (or written) either before or af=
ter the lerfu word representing the basic lerfu. Typewriters (for mechanica=
l reasons) and the computer programs that emulate them usually require thei=
r users to type the accent mark before the basic lerfu, whereas in speech t=
he accent mark is often pronounced afterwards (for example, in German =93a =
umlaut=94 is preferred to =93umlaut a=94).
+<p>
+<cx "diacritical marks, specifying with tei=85foi">  XE "diacritical marks=
: specifying with tei=85foi"   <lx "tei">  XE "tei"   <lx "foi">  XE "foi" =
  <lx "TEI">  XE "TEI selma=92o"   <lx "FOI">  XE "FOI selma=92o"   Lojban =
cannot settle this question by fiat. Either it must be left up to default i=
nterpretation depending on the language in question, or the lerfu-word comp=
ounding cmavo =93tei=94 (of selma'o TEI) and =93foi=94 (of selma'o FOI) mus=
t be used. These cmavo are always used in pairs; any number of lerfu words =
may appear between them, and the whole is treated as a single compound lerf=
u word. The French word =93=E9&eacute;t=E9&eacute;=94, with acute accent ma=
rks on both =93e=94 lerfu, could be spelled as:
+<p>
+<ex "accent mark">  XE "accent mark: example=93  <ex "ete">  XE "=E9t=E9: =
example=93  <pre><a name=3De6d1>6.1)	tei .ebu .akut. bu foi ty. tei .akut. =
bu .ebu foi
+	(=93e=94 acute ) =93t=94 ( acute =93e=94 )
+</pre><cx "diacritical marks, order of specification within tei=85foi">  X=
E "diacritical marks: order of specification within tei=85foi"   and it doe=
s not matter whether =93akut. bu=94 appears before or after =93.ebu=94; the=
 =93tei<dots>=85</dots>foi=94 grouping guarantees that the acute accent is =
associated with the correct lerfu. Of course, the level of precision repres=
ented by <a href=3D#e6d1>Example 6.1 </a>would rarely be required: it might=
 be needed by a Lojban-speaker when spelling out a French word for exact tr=
anscription by another Lojban-speaker who did not know French.
+<p>
+<cx "diacritical marks, problem with multiple on one lerfu">  XE "diacriti=
cal marks: problem with multiple on one lerfu"   This system breaks down in=
 languages which use more than one accent mark on a single lerfu; some othe=
r convention must be used for showing which accent marks are written where =
in that case. The obvious convention is to represent the mark nearest the b=
asic lerfu by the lerfu word closest to the word representing the basic ler=
fu. Any remaining ambiguities must be resolved by further conventions not y=
et established.
+<p>
+<cx "diacritical marks, considered as forming distinct letters">  XE "diac=
ritical marks: considered as forming distinct letters"   <cx "accented lett=
ers, considered as distinct from unaccented">  XE "accented letters: consid=
ered as distinct from unaccented"   <cx "compound letters, native language,=
 representing as distinct letters">  XE "compound letters: native language,=
 representing as distinct letters"   <ex "Spanish ll">  XE "Spanish ll: exa=
mple=93  <ex "Spanish ch">  XE "Spanish ch: example=93  Some languages, lik=
e Swedish and Finnish, consider certain accented lerfu to be completely dis=
tinct from their unaccented equivalents, but Lojban does not make a formal =
distinction, since the printed characters look the same whether they are re=
ckoned as separate letters or not. In addition, some languages consider cer=
tain 2-letter combinations (like =93ll=94 and =93ch=94 in Spanish) to be le=
tters; this may be represented by enclosing the combination in =93tei<dots>=
=85</dots>foi=94.
+<p>
+<cx "lerfu words, forming new for non-Lojban letters using bu">  XE "lerfu=
 words: forming new for non-Lojban letters using bu"   In addition, when di=
scussing a specific language, it is permissible to make up new lerfu words,=
 as long as they are either explained locally or well understood from conte=
xt: thus Spanish =93ll=94 or Croatian =93lj=94 could be called =93libu=94, =
but that usage would not necessarily be universally understood.
+<p>
+<a href=3D#s19>Section 19 </a>contains a table of proposed lerfu words for=
 some common accent marks.
+<p>
+<a name=3Ds7><h3>Punctuation marks</h3>
+<p>
+<cx "punctuation lerfu words, mechanism for creating">  XE "punctuation le=
rfu words: mechanism for creating"   <lx "lau">  XE "lau"   <lx "LAU">  XE =
"LAU selma=92o"   <cx "lau, effect on following lerfu word">  XE "lau: effe=
ct on following lerfu word"   Lojban does not have punctuation marks as suc=
h: the denpa bu and the slaka bu are really a part of the alphabet. Other l=
anguages, however, use punctuation marks extensively. As yet, Lojban does n=
ot have any words for these punctuation marks, but a mechanism exists for d=
evising them: the cmavo =93lau=94 of selma'o LAU. =93lau=94 must always be =
followed by a BY word; the interpretation of the BY word is changed from a =
lerfu to a punctuation mark. Typically, this BY word would be a name or bri=
vla with a =93bu=94 suffix.
+<p>
+<cx "punctuation lerfu words, rationale for lau">  XE "punctuation lerfu w=
ords: rationale for lau"   Why is =93lau=94 necessary at all? Why not just =
use a =93bu=94-marked word and announce that it is always to be interpreted=
 as a punctuation mark? Primarily to avoid ambiguity. The =93bu=94 mechanis=
m is extremely open-ended, and it is easy for Lojban users to make up =93bu=
=94 words without bothering to explain what they mean. Using the =93lau=94 =
cmavo flags at least the most important of such nonce lerfu words as having=
 a special function: punctuation. (Exactly the same argument applies to the=
 use of =93zai=94 to signal an alphabet shift or =93ce'a=94 to signal a fon=
t shift.)
+<p>
+<cx "punctuation lerfu words, interaction with different alphabet systems"=
>  XE "punctuation lerfu words: interaction with different alphabet systems=
"   Since different alphabets require different punctuation marks, the inte=
rpretation of a =93lau=94-marked lerfu word is affected by the current alph=
abet shift and the current font shift.
+<p>
+<a name=3Ds8><h3>What about Chinese characters?</h3>
+<p>
+<cx "Chinese characters, contrasted with alphabets and syllabaries">  XE "=
Chinese characters: contrasted with alphabets and syllabaries"   <cx "kanji=
, contrasted with alphabets and syllabaries">  XE "kanji: contrasted with a=
lphabets and syllabaries"   <cx "hiragana, contrasted with kanji">  XE "hir=
agana: contrasted with kanji"   <cx "syllabaries, lerfu word representation=
">  XE "syllabaries: lerfu word representation"   <cx "Amharic writing">  X=
E "Amharic writing"   Chinese characters (=93han<sup>4</sup>zi<sup>4</sup>=
=94 in Chinese, =93kanji=94 in Japanese) represent an entirely different ap=
proach to writing from alphabets or syllabaries. (A syllabary, such as Japa=
nese hiragana or Amharic writing, has one lerfu for each syllable of the sp=
oken language.) Very roughly, Chinese characters represent single elements =
of meaning; also very roughly, they represent single syllables of spoken Ch=
inese. There is in principle no limit to the number of Chinese characters t=
hat can exist, and many thousands are in regular use.
+<p>
+It is hopeless for Lojban, with its limited lerfu and shift words, to crea=
te an alphabet which will match this diversity. However, there are various =
possible ways around the problem.
+<p>
+<cx "Chinese characters, representing based on pinyin spelling">  XE "Chin=
ese characters: representing based on pinyin spelling"   <cx "kanji, repres=
enting based on romaji spelling">  XE "kanji: representing based on romaji =
spelling"   <cx "pinyin, as a basis for Chinese characters in Lojban lerfu =
words">  XE "pinyin: as a basis for Chinese characters in Lojban lerfu word=
s"   <cx "romaji, as a basis for kanji characters in Lojban lerfu words">  =
XE "romaji: as a basis for kanji characters in Lojban lerfu words"   First,=
 both Chinese and Japanese have standard Latin-alphabet representations, kn=
own as =93pinyin=94 for Chinese and =93romaji=94 for Japanese, and these ca=
n be used. Thus, the word =93han<sup>4</sup>zi<sup>4</sup>=94 is convention=
ally written with two characters, but it may be spelled out as:
+<p>
+<ex "han^{4}zi^{4}">  XE "han4zi4: example=93  <pre><a name=3De8d1>8.1)	.y=
'y.bu .abu ny. vo zy. .ibu vo
+	=93h=94 =93a=94 =93n=94 4 =93z=94 =93i=94 4
+</pre><cx "numeric digits in lerfu words, grammar considerations">  XE "nu=
meric digits in lerfu words: grammar considerations"   <cx "lerfu words wit=
h numeric digits, grammar considerations">  XE "lerfu words with numeric di=
gits: grammar considerations"   The cmavo =93vo=94 is the Lojban digit =934=
=94. It is grammatical to intersperse digits (of selma'o PA) into a string =
of lerfu words; as long as the first cmavo is a lerfu word, the whole will =
be interpreted as a string of lerfu words. In Chinese, the digits can be us=
ed to represent tones. Pinyin is more usually written using accent marks, t=
he mechanism for which was explained in <a href=3D#s6>Section 6</a>.
+<p>
+The Japanese company named =93Mitsubishi=94 in English is spelled the same=
 way in romaji, and could be spelled out in Lojban thus:
+<p>
+<ex "Mitsubishi">  XE "Mitsubishi: example=93  <pre><a name=3De8d2>8.2)	my=
. .ibu ty. sy. .ubu by. .ibu sy. .y'y.bu .ibu
+	=93m=94 =93i=94 =93t=94 =93s=94 =93u=94 =93b=94 =93i=94 =93s=94 =93h=94 =
=93i=94
+</pre><cx "Chinese characters, representing based on strokes">  XE "Chines=
e characters: representing based on strokes"   <cx "kanji, representing bas=
ed on strokes">  XE "kanji: representing based on strokes"   Alternatively,=
 a really ambitious Lojbanist could assign lerfu words to the individual st=
rokes used to write Chinese characters (there are about seven or eight of t=
hem if you are a flexible human being, or about 40 if you are a rigid compu=
ter program), and then represent each character with a =93tei=94, the strok=
e lerfu words in the order of writing (which is standardized for each chara=
cter), and a =93foi=94. No one has as yet attempted this project.
+<p>
+<a name=3Ds9><h3>lerfu words as pro-sumti</h3>
+<p>
+<cx "lerfu string, definition">  XE "lerfu string: definition"   So far, l=
erfu words have only appeared in Lojban text when spelling out words. There=
 are several other grammatical uses of lerfu words within Lojban. In each c=
ase, a single lerfu word or more than one may be used. Therefore, the term =
=93lerfu string=94 is introduced: it is short for =93sequence of one or mor=
e lerfu words=94.
+
+<p>
+<cx "lerfu string, as pro-sumti">  XE "lerfu string: as pro-sumti"   A ler=
fu string may be used as a pro-sumti (a sumti which refers to some previous=
 sumti), just like the pro-sumti =93ko'a=94, =93ko'e=94, and so on:
+<p>
+<ex "A loves B">  XE "A loves B: example=93  <pre><a name=3De9d1>9.1)	.abu=
 prami by.
+	A loves B
+</pre><p>In <a href=3D#e9d1>Example 9.1</a>, =93.abu=94 and =93by.=94 repr=
esent specific sumti, but which sumti they represent must be inferred from =
context.
+<p>
+<cx "lerfu string, as pro-sumti assigned by goi">  XE "lerfu string: as pr=
o-sumti assigned by goi"   <lx "goi">  XE "goi"   Alternatively, lerfu stri=
ngs may be assigned by =93goi=94, the regular pro-sumti assignment cmavo:
+<p>
+<pre><a name=3De9d2>9.2)	le gerku goi gy. cu xekri  .i gy. klama le zdani
+	The dog, or G, is black. G goes to the house.
+</pre><cx "lerfu string as pro-sumti, assumption of reference">  XE "lerfu=
 string: as pro-sumti, assumption of reference"   There is a special rule t=
hat sometimes makes lerfu strings more advantageous than the regular pro-su=
mti cmavo. If no assignment can be found for a lerfu string (especially a s=
ingle lerfu word), it can be assumed to refer to the most recent sumti whos=
e name or description begins in Lojban with that lerfu. So <a href=3D#e9d2>=
Example 9.2 </a>can be rephrased:
+<p>
+<pre><a name=3De9d3>9.3)	le gerku cu xekri.  .i gy. klama le zdani
+	The dog is black.  G goes to the house.
+</pre>(A less literal English translation would use =93D=94 for =93dog=94 =
instead.)
+<p>
+Here is an example using two names and longer lerfu strings:
+<p>
+<ex "Steven Mark Jones">  XE "Steven Mark Jones: example=93  <ex "Alexande=
r Pavlovitch Kuznetsov">  XE "Alexander Pavlovitch Kuznetsov: example=93  <=
pre><a name=3De9d4>9.4)	la stivn. mark. djonz. merko  .i la .aleksandr. pal=
iitc. kuzNIETsyf. rusko
+		.i symyjy. tavla .abupyky. bau la lojban.
+	Steven Mark Jones is-American. Alexander Pavlovitch Kuznetsov is-Russian.
+		SMJ talks-to APK in Lojban.
+</pre><p>Perhaps Alexander's name should be given as =93ru'o.abupyky=94 in=
stead.
+<p>
+<cx "lerfu strings as pro-sumti, for multiple sumti separated by boi">  XE=
 "lerfu string: as pro-sumti, for multiple sumti separated by boi"   What a=
bout
+<p>
+<ex "A gives BC">  XE "A gives BC: example=93  <pre><a name=3De9d5>9.5)	.a=
bu dunda by. cy.
+	A gives B C
+</pre><lx "boi">  XE "boi"   <lx "BOI">  XE "BOI selma=92o"   <cx "boi, el=
iding from lerfu strings">  XE "boi: eliding from lerfu strings"   Does thi=
s mean that A gives B to C? No. =93by. cy.=94 is a single lerfu string, alt=
hough written as two words, and represents a single pro-sumti. The true int=
erpretation is that A gives BC to someone unspecified. To solve this proble=
m, we need to introduce the elidable terminator =93boi=94 (of selma'o BOI).=
 This cmavo is used to terminate lerfu strings and also strings of numerals=
; it is required when two of these appear in a row, as here. (The other rea=
son to use =93boi=94 is to attach a free modifier =97 subscript, parenthesi=
s, or what have you =97 to a lerfu string.)  The correct version is:
+<p>
+<ex "A gives B to C">  XE "A gives B to C: example=93  <pre><a name=3De9d6=
>9.6)	.abu [boi] dunda by. boi cy. [boi]
+	A gives B to C
+</pre>where the two occurrences of =93boi=94 in brackets are elidable, but=
 the remaining occurrence is not. Likewise:
+<p>
+<pre><a name=3De9d7>9.7)	xy. boi ro [boi] prenu cu prami
+	X all persons loves.
+	X loves everybody.
+</pre><cx "boi, required between pro-sumti lerfu string and quantifier">  =
XE "boi: required between pro-sumti lerfu string and quantifier"   <cx "pro=
-sumti lerfu strings, interaction with quantifiers and boi">  XE "pro-sumti=
: lerfu string, interaction with quantifier and boi"   requires the first =
=93boi=94 to separate the lerfu string =93xy.=94 from the digit string =93r=
o=94.
+<p>
+<a name=3Ds10><h3>References to lerfu</h3>
+<p>
+<cx "lerfu, reference to">  XE "lerfu: reference to"   <cx "pro-sumti lerf=
u string, effect on reference to lerfu itself">  XE "pro-sumti: lerfu strin=
g, effect on reference to lerfu itself"   The rules of <a href=3D#s9>Sectio=
n 9 </a>make it impossible to use unmarked lerfu words to refer to lerfu th=
emselves. In the sentence:
+<p>
+<pre><a name=3De10d1>10.1)	.abu. cu lerfu
+	A is-a-letteral.
+</pre><lx "me'o">  XE "me'o"   <lx "LI">  XE "LI selma=92o"   <cx "lerfu, =
referring to with me'o">  XE "lerfu: referring to with me'o"   the hearer w=
ould try to find what previous sumti =93.abu=94 refers to. The solution to =
this problem makes use of the cmavo =93me'o=94 of selma'o LI, which makes a=
 lerfu string into a sumti representing that very string of lerfu. This use=
 of =93me'o=94 is a special case of its mathematical use, which is to intro=
duce a mathematical expression used literally rather than for its value.
+<p>
+<ex "=93a=94 is letteral">  XE "a is letteral: example=93  <pre><a name=3D=
e10d2>10.2)	me'o .abu cu lerfu
+	the-expression =93a=94 is-a-letteral.
+</pre><p>Now we can translate <a href=3D#e1d1>Example 1.1 </a>into Lojban:
+<p>
+<ex "four =93e=94s">  XE "four \=93e\=94s: example=93  <pre><a name=3De10d=
3>10.3)	dei vasru vo lerfu
+		po'u me'o .ebu
+	this-sentence contains four letterals
+		which-are the-expression =93e=94.
+	This sentence contains four =93e=94s.
+</pre><p>Since the Lojban sentence has only four =93e=94 lerfu rather than=
 fourteen, the translation is not a literal one =97 but <a href=3D#e10d3>Ex=
ample 10.3 </a>is a Lojban truth just as <a href=3D#e1d1>Example 1.1 </a>is=
 an English truth. Coincidentally, the colloquial English translation of <a=
 href=3D#e10d3>Example 10.3 </a>is also true!
+<p>
+<cx "quotation, contrasted with me'o for representing lerfu">  XE "quotati=
on: contrasted with me'o for representing lerfu"   <cx "me'o, contrasted wi=
th quotation for representing lerfu">  XE "me'o: contrasted with quotation =
for representing lerfu"   <cx "me'o, contrasted with lu=85li'u for represen=
ting lerfu">  XE "me'o: contrasted with lu=85li'u for representing lerfu"  =
 <cx "lu, contrasted with me'o for representing lerfu">  XE "lu: contrasted=
 with me'o for representing lerfu"   <cx "representing lerfu, lu contrasted=
 with me'o">  XE "representing lerfu: lu contrasted with me'o"   <cx "me'o,=
 compared with la'e lu">  XE "me'o: compared with la'e lu"   <cx "la'e lu, =
compared with me'o">  XE "la'e lu: compared with me'o"   <lx "me'o">  XE "m=
e'o"   <lx "lu">  XE "lu"   <lx "la'e">  XE "la'e"   The reader might be te=
mpted to use quotation with =93lu=A0<dots>=85</dots>li'u=94 instead of =93m=
e'o=94, producing:
+<p>
+<pre><a name=3De10d4>10.4)	lu .abu li'u cu lerfu
+	[quote] .abu [unquote] is-a-letteral.
+</pre>(The single-word quote =93zo=94 cannot be used, because =93.abu=94 i=
s a compound cmavo.) But <a href=3D#e10d4>Example 10.4 </a>is false, becaus=
e it says:
+<p>
+<ex "word =93abu=94">  XE "word \=93abu\=94: example=93   <pre><a name=3De=
10d5>10.5)	The word =93.abu=94 is a letteral
+</pre>which is not the case; rather, the thing symbolized by the word =93.=
abu=94 is a letteral. In Lojban, that would be:
+<p>
+<pre><a name=3De10d6>10.6)	la'e lu .abu li'u cu lerfu
+	The-referent-of [quote] .abu [unquote] is-a-letteral.
+</pre>which is correct.
+<p>
+<a name=3Ds11><h3>Mathematical uses of lerfu strings</h3>
+<p>
+<cx "mathematics, use of lerfu strings in">  XE "mathematics: use of lerfu=
 strings in"   <cx "lerfu strings, uses in mathematics">  XE "lerfu string:=
 use in mathematics"   This chapter is not about Lojban mathematics, which =
is explained in <a href=3Dchap18.html>Chapter 18</a>, so the mathematical u=
ses of lerfu strings will be listed and exemplified but not explained.
+<p>
+<cx "lerfu string, as mathematical variable">  XE "lerfu string: as mathem=
atical variable"   <cx "mathematical variables, lerfu strings as">  XE "mat=
hematical variables: lerfu strings as"   A lerfu string as mathematical var=
iable:
+<p>
+<pre><a name=3De11d1>11.1)	li .abu du li by. su'i cy.
+	the-number a equals the-number b plus c
+	a =3D b + c
+</pre><cx "lerfu string, as function name">  XE "lerfu string: as function=
 name"   <cx "function name, lerfu string as">  XE "function name: lerfu st=
ring as"   A lerfu string as function name (preceded by =93ma'o=94 of selma=
'o MAhO):
+<p>
+<ex "function f of x">  XE "function f of x: example=93  <pre><a name=3De1=
1d2>11.2)	li .y.bu du li ma'o fy. boi xy.
+	the-number y equals the number the-function f of x
+	y =3D f(x)
+</pre><p>Note the =93boi=94 here to separate the lerfu strings =93fy=94 an=
d =93xy=94.
+<p>
+<cx "lerfu string, as selbri">  XE "lerfu string: as selbri"   <cx "selbri=
, lerfu string as">  XE "selbri: lerfu string as"   A lerfu string as selbr=
i (followed by a cmavo of selma'o MOI):
+<p>
+<ex "Nth rat">  XE "Nth rat: example=93  <pre><a name=3De11d3>11.3)	le vi =
ratcu ny.moi le'i mi ratcu
+	the here rat is-nth-of the-set-of my rats
+	This rat is my Nth rat.
+</pre><cx "lerfu string, as utterance ordinal">  XE "lerfu string: as utte=
rance ordinal"   <cx "utterance ordinal, lerfu string as">  XE "utterance o=
rdinal: lerfu string as"   A lerfu string as utterance ordinal (followed by=
 a cmavo of selma'o MAI):
+<p>
+<ex "Nthly">  XE "Nthly: example=93  <pre><a name=3De11d4>11.4)	ny.mai
+	Nthly
+</pre><cx "lerfu string, as subscript">  XE "lerfu string: as subscript"  =
 <cx "subscript, lerfu string as">  XE =93subscripts: lerfu string as"   A =
lerfu string as subscript (preceded by =93xi=94 of selma'o XI):
+<p>
+<ex "x sub k">  XE "x{k}: x sub k, example=93  <pre><a name=3De11d5>11.5)	=
xy. xi ky.
+	x sub k
+</pre><cx "lerfu string, as quantifier">  XE "lerfu string: as quantifier"=
   <cx "quantifier, lerfu string as">  XE "quantifier: lerfu string as"   A=
 lerfu string as quantifier (enclosed in =93vei=A0<dots>=85</dots>ve'o=94 p=
arentheses):
+<p>
+<ex "n people">  XE "n people: example=93  <lx "vei">  XE "vei"   <pre><a =
name=3De11d6>11.6)	vei ny. [ve'o] lo prenu
+	(=93n=94 ) persons
+</pre><cx "lerfu strings as quantifiers, avoiding interaction with sumti q=
uantified">  XE "lerfu string: as quantifier, avoiding interaction with sum=
ti quantifier"   The parentheses are required because =93ny. lo prenu=94 wo=
uld be two separate sumti, =93ny.=94 and =93lo prenu=94. In general, any ma=
thematical expression other than a simple number must be in parentheses whe=
n used as a quantifier; the right parenthesis mark, the cmavo =93ve'o=94, c=
an usually be elided.
+<p>
+<cx "lerfu string interpretation, contrasted with mathematical interpretat=
ion">  XE "lerfu string: interpretation, contrasted with mathematical inter=
pretation"   <cx "lerfu juxtaposition interpretation, contrasted with mathe=
matical interpretation">  XE "lerfu juxtaposition interpretation: contraste=
d with mathematical interpretation"   All the examples above have exhibited=
 single lerfu words rather than lerfu strings, in accordance with the conve=
ntions of ordinary mathematics. A longer lerfu string would still be treate=
d as a single variable or function name: in Lojban, =93.abu by. cy.=94 is n=
ot the multiplication =93a ( b ( c=94 but is the variable =93abc=94. (Of co=
urse, a local convention could be employed that made the value of a variabl=
e like =93abc=94, with a multi-lerfu-word name, equal to the values of the =
variables =93a=94, =93b=94, and =93c=94 multiplied together.)
+<p>
+<cx "mathematical texts, effect on lerfu shift scope">  XE "mathematical t=
exts: effect on lerfu shift scope"   <cx "lerfu shift scope, exception for =
mathematical texts">  XE "lerfu shift scope: exception for mathematical tex=
ts"   There is a special rule about shift words in mathematical text: shift=
s within mathematical expressions do not affect lerfu words appearing outsi=
de <dl compact><dt>mathematical expressions, and vice versa.	<dd>
+</dl><a name=3Ds12><h3>Acronyms</h3>
+<p>
+<cx "acronym, definition">  XE "acronym: definition"   An acronym is a nam=
e constructed of lerfu. English examples are =93DNA=94, =93NATO=94, =93CIA=
=94. In English, some of these are spelled out (like =93DNA=94 and =93CIA=
=94) and others are pronounced more or less as if they were ordinary Englis=
h words (like =93NATO=94). Some acronyms fluctuate between the two pronunci=
ations: =93SQL=94 may be =93ess cue ell=94 or =93sequel=94.
+<p>
+<cx "acronyms, using names based on lerfu words">  XE "acronyms: using nam=
es based on lerfu words"   <cx "lerfu words, as a basis for acronym names">=
  XE "lerfu words: as a basis for acronym names"   In Lojban, a name can be=
 almost any sequence of sounds that ends in a consonant and is followed by =
a pause. The easiest way to Lojbanize acronym names is to glue the lerfu wo=
rds together, using =93 ' =94wherever two vowels would come together (pause=
s are illegal in names) and adding a final consonant:
+<p>
+<ex "DNA">  XE "DNA: example=93  <ex "CIA">  XE "CIA: example=93  <ex "IBM=
">  XE "IBM: example=93  <ex "NATO">  XE "NATO: example=93  <ex "SQL">  XE =
"SQL: example=93  <ex "NYC">  XE "NYC: example=93  <pre><a name=3De12d1>12.=
1)	la dyny'abub.  .i la ny'abuty'obub.
+	.i la cy'ibu'abub.  .i la sykybulyl.
+	.i la .ibubymym.  .i la ny'ybucyc.
+	DNA.  NATO.
+	CIA.  SQL.
+	IBM.  NYC.
+</pre><cx "acronym names from lerfu words, assigning final consonant">  XE=
 "acronym names from lerfu words: assigning final consonant"   There is no =
fixed convention for assigning the final consonant. In <a href=3D#e12d1>Exa=
mple 12.1</a>, the last consonant of the lerfu string has been replicated i=
nto final position.
+<p>
+<cx "acronyms names based on lerfu words, omitting bu">  XE "acronyms name=
s based on lerfu words: omitting bu"   <cx "bu, omitting in acronyms names =
based on lerfu words">  XE "bu: omitting in acronyms names based on lerfu w=
ords"   Some compression can be done by leaving out =93bu=94 after vowel le=
rfu words (except for =93.y.bu=94, wherein the =93bu=94 cannot be omitted w=
ithout ambiguity). Compression is moderately important because it's hard to=
 say long names without introducing an involuntary (and illegal) pause:
+<p>
+<pre><a name=3De12d2>12.2)	la dyny'am.  .i la ny'aty'om.
+	.i la cy'i'am.  .i la sykybulym.
+	.i la .ibymym.  .i la ny'ybucym.
+	DNA.  NATO.
+	CIA.  SQL.
+	IBM.  NYC.
+</pre><p>In <a href=3D#e12d2>Example 12.2</a>, the final consonant =93m=94=
 stands for =93merko=94, indicating the source culture of these acronyms.
+<p>
+<cx "acronyms names based on lerfu words, using =93z=94 instead of =93 ' =
=94in">  XE "acronyms names based on lerfu words: using \=93z\=94 instead o=
f \=93 ' \=94in"   <cx "=93z=94 instead of =93 ' =94, in acronyms names bas=
ed on lerfu words">  XE "z instead of ' : in acronymic names based on lerfu=
 words"   Another approach, which some may find easier to say and which is =
compatible with older versions of the language that did not have a =93 ' =
=94character, is to use the consonant =93z=94 instead of =93 ' =94:
+<p>
+<pre><a name=3De12d3>12.3)	la dynyzaz.  .i la nyzatyzoz.
+	.i la cyzizaz.  .i la sykybulyz.
+	.i la .ibymyz.  .i la nyzybucyz.
+	DNA.  NATO.
+	CIA.  SQL.
+	IBM.  NYC.
+</pre><cx "lerfu strings, as acronyms using =93me=94">  XE "lerfu string: =
as acronym using \=93me\=94"   <cx "acronyms, as lerfu strings using =93me=
=94">  XE "acronyms: as lerfu strings using \=93me\=94"   <lx "me">  XE "me=
"   One more alternative to these lengthy names is to use the lerfu string =
itself prefixed with =93me=94, the cmavo that makes sumti into selbri:
+<p>
+<pre><a name=3De12d4>12.4)	la me dy ny. .abu
+	that-named what-pertains-to =93d=94 =93n=94 =93a=94
+</pre><p>This works because =93la=94, the cmavo that normally introduces n=
ames used as sumti, may also be used before a predicate to indicate that th=
e predicate is a (meaningful) name:
+<p>
+<pre><a name=3De12d5>12.5)	la cribe cu ciska
+	That-named =93Bear=94 writes.
+	Bear is a writer.
+</pre><a href=3D#e12d5>Example 12.5 </a>does not of course refer to a bear=
 (=93le cribe=94 or =93lo cribe=94) but to something else, probably a perso=
n, named =93Bear=94. Similarly, =93me dy ny. .abu=94 is a predicate which c=
an be used as a name, producing a kind of acronym which can have pauses bet=
ween the individual lerfu words.
+<p>
+<a name=3Ds13><h3>Computerized character codes</h3>
+<p>
+<cx "character codes, definition">  XE "character codes: definition"   <cx=
 "characters, definition">  XE "characters: definition"   <cx "lerfu words,=
 using computer encoding schemes with se'e">  XE "lerfu words: using comput=
er encoding schemes with se'e"   <cx "character encoding schemes, applicati=
on to lerfu words">  XE "character encoding schemes: application to lerfu w=
ords"   <cx "letter encoding schemes, application to lerfu words">  XE "let=
ter encoding schemes: application to lerfu words"   <lx "se'e">  XE "se'e" =
  <lx "BY">  XE "BY selma=92o"   Since the first application of computers t=
o non-numerical information, character sets have existed, mapping numbers (=
called =93character codes=94) into selected lerfu, digits, and punctuation =
marks (collectively called =93characters=94). Historically, these character=
 sets have only covered the English alphabet and a few selected punctuation=
 marks. International efforts have now  created Unicode, a unified characte=
r set that can represent essentially all the characters in essentially all =
the world's writing systems. Lojban can take advantage of these encoding sc=
hemes by using the cmavo =93se'e=94 (of selma'o BY). This cmavo is conventi=
onally followed by digit cmavo of selma'o PA representing the character cod=
e, and the whole string indicates a single character in some computerized c=
haracter set:
+<p>
+<ex "American dollars">  XE "American dollars: example=93  <ex "$&dollar;"=
><pre><a name=3De13d1>13.1)	me'o se'ecixa cu lerfu la .asycy'i'is.=20
+loi merko rupnu
+	The-expression [code] 36 is-a-letteral in-set ASCII
+		for-the-mass-of American currency-units.
+	The character code 36 in ASCII represents American dollars.
+	=93$&dollar;=94 represents American dollars.
+</pre><cx "ASCII, application to lerfu words">  XE "ASCII: application to =
lerfu words"   Understanding <a href=3D#e13d1>Example 13.1 </a>depends on k=
nowing the value in the ASCII character set (one of the simplest and oldest=
) of the =93$&dollar;=94 character. Therefore, the =93se'e=94 convention is=
 only intelligible to those who know the underlying character set. For prec=
isely specifying a particular character, however, it has the advantages of =
unambiguity and (relative) cultural neutrality, and therefore Lojban provid=
es a means for those with access to descriptions of such character sets to =
take advantage of them.
+<p>
+<cx "Unicode">  XE "Unicode"   <cx "peace symbol">  XE "peace symbol"   As=
 another example, the Unicode character set (also known as ISO 10646) repre=
sents the international symbol of peace, an inverted trident in a circle, u=
sing the base-16 value 262E. In a suitable context, a Lojbanist may say:
+<p>
+<pre><a name=3De13d2>13.2)	me'o se'erexarerei sinxa le ka panpi
+	The-expression [code] 262E is-a-sign-of the quality-of being-at-peace.
+</pre><cx "se'e, and number base convention">  XE "se'e: and number base c=
onvention"   When a =93se'e=94 string appears in running discourse, some me=
talinguistic convention must specify whether the number is base 10 or some =
other base, and which character set is in use.
+<p>
+<a name=3Ds14><h3>List of all auxiliary lerfu-word cmavo</h3>
+<p>
+<cx "lerfu word cmavo, list of auxiliary">  XE "lerfu word cmavo: list of =
auxiliary"   <pre> 	cmavo	selma'o	meaning
+	bu	BU	makes previous word into a lerfu word
+	ga'e	BY	upper case shift
+	to'a	BY	lower case shift
+	tau	LAU	case-shift next lerfu word only
+	lo'a	BY	Latin/Lojban alphabet shift
+	ge'o	BY	Greek alphabet shift
+	je'o	BY	Hebrew alphabet shift
+	jo'o	BY	Arabic alphabet shift
+	ru'o	BY	Cyrillic alphabet shift
+
+	se'e	BY	following digits are a character code
+	na'a	BY	cancel all shifts
+	zai	LAU	following lerfu word specifies alphabet
+	ce'a	LAU	following lerfu word specifies font
+	lau	LAU	following lerfu word is punctuation
+	tei	TEI	start compound lerfu word
+	foi	FOI	end compound lerfu word
+</pre><cx "LAU, grammar of following BY cmavo">  XE "LAU selma=92o: gramma=
r of following BY"   Note that LAU cmavo must be followed by a BY cmavo or =
the equivalent, where =93equivalent=94 means: either any Lojban word follow=
ed by =93bu=94, another LAU cmavo (and its required sequel), or a =93tei=A0=
<dots>=85</dots>foi=94 compound cmavo.
+<p>
+<a name=3Ds15><h3>Proposed lerfu words =97 introduction</h3>
+<p>
+<cx "lerfu words, list of proposed, notation convention">  XE "lerfu words=
: list of proposed, notation convention"   The following sections contain t=
ables of proposed lerfu words for some of the standard alphabets supported =
by the Lojban lerfu system. The first column of each list is the lerfu (act=
ually, a Latin-alphabet name sufficient to identify it). The second column =
is the proposed name-based lerfu word, and the third column is the proposed=
 lerfu word in the system based on using the cmavo of selma'o BY with a shi=
ft word.
+<p>
+<cx "proposed lerfu words, as working basis">  XE "proposed lerfu words: a=
s working basis"   These tables are not meant to be authoritative (several =
authorities within the Lojban community have niggled over them extensively,=
 disagreeing with each other and sometimes with themselves). They provide a=
 working basis until actual usage is available, rather than a final resolut=
ion of lerfu word problems. Probably the system presented here will evolve =
somewhat before settling down into a final, conventional form.
+<p>
+For Latin-alphabet lerfu words, see <a href=3D#s2>Section 2 </a>(for Lojba=
n) and <a href=3D#s5>Section 5 </a>(for non-Lojban Latin-alphabet lerfu).
+<p>
+<a name=3Ds16><h3>Proposed lerfu words for the Greek alphabet</h3>
+<p>
+<cx "lerfu words, proposed for Greek alphabet">  XE "lerfu words: proposed=
 for Greek alphabet"   <cx "Greek alphabet, proposed lerfu words for">  XE =
"Greek alphabet: proposed lerfu words for"   <pre> 	alpha	.alfas. bu	.abu
+	beta	.betas. bu	by
+	gamma	.gamas. bu	gy
+	delta	.deltas. bu	dy
+	epsilon	.Epsilon. bu	.ebu
+	zeta	.zetas. bu	zy
+	eta	.etas. bu	.e'ebu
+	theta	.tetas. bu	ty. bu
+	iota	.iotas. bu	.ibu
+	kappa	.kapas. bu	ky
+	lambda	.lymdas. bu	ly
+	mu	.mus. bu	my
+	nu	.nus. bu	ny
+	xi	.ksis. bu	ksis. bu
+	omicron	.Omikron. bu	.obu
+	pi	.pis. bu	py
+	rho	.ros. bu	ry
+	sigma	.sigmas. bu	sy
+	tau	.taus. bu	ty
+	upsilon	.Upsilon. bu	.ubu
+	phi	.fis. bu	py. bu
+	chi	.xis. bu	ky. bu
+	psi	.psis. bu	psis. bu
+	omega	.omegas. bu	.o'obu
+	rough	.dasei,as. bu	.y'y
+	smooth	.psiles. bu	xutla bu
+</pre><a name=3Ds17><h3>Proposed lerfu words for the Cyrillic alphabet</h3=
>
+<p>
+<cx "lerfu words, proposed for Cyrillic alphabet">  XE "lerfu words: propo=
sed for Cyrillic alphabet"   <cx "Cyrillic alphabet, proposed lerfu words f=
or">  XE "Cyrillic alphabet: proposed lerfu words for"   The second column =
in this listing is based on the historical names of the letters in Old Chur=
ch Slavonic. Only those letters used in Russian are shown; other languages =
require more letters which can be devised as needed.
+<p>
+<pre> 	a	.azys. bu	.abu
+	b	.bukys. bu	by
+	v	.vedis. bu	vy
+	g	.glagolis. bu	gy
+	d	.dobros. bu	dy
+	e	.iestys. bu	.ebu
+	zh	.jivet. bu	jy
+	z	.zemlias. bu	zy
+	i	.ije,is. bu	.ibu
+	short i	.itord. bu	.itord. bu
+	k	.kakos. bu	ky
+	l	.liudi,ies. bu	ly
+	m	.myslites. bu	my
+	n	.naciys. bu	ny
+	o	.onys. bu	.obu
+	p	.pokois. bu	py
+	r	.riytsis. bu	ry
+	s	.slovos. bu	sy
+	t	.tvriydos. bu	ty
+	u	.ukys. bu	.ubu
+	f	.friytys. bu	fy
+	kh	.xerys. bu	xy
+	ts	.tsis. bu	tsys. bu
+	ch	.tcriyviys. bu	tcys. bu
+	sh	.cas. bu	cy
+	shch	.ctas. bu	ctcys. bu
+	hard sign	.ier. bu	jdari bu
+	yeri	.ierys. bu	.y.bu
+	soft sign	.ieriys. bu	ranti bu
+	reversed e	.ecarn. bu	.ecarn. bu
+	yu	.ius. bu	.iubu
+	ya	.ias. bu	.iabu
+
+</pre><a name=3Ds18><h3>Proposed lerfu words for the Hebrew alphabet</h3>
+<p>
+<cx "lerfu words, proposed for Hebrew alphabet">  XE "lerfu words: propose=
d for Hebrew alphabet"   <cx "Hebrew alphabet, proposed lerfu words for">  =
XE "Hebrew alphabet: proposed lerfu words for"   <pre> 	aleph	.alef. bu	.al=
ef. bu
+	bet	.bet. bu	by
+	gimel	.gimel. bu	gy
+	daled	.daled. bu	dy
+	he	.xex. bu	.y'y
+	vav	.vav. bu	vy
+	zayin	.zai,in. bu	zy
+	khet	.xet. bu	xy. bu
+	tet	.tet. bu	ty. bu
+	yud	.iud. bu	.iud. bu
+	kaf	.kaf. bu	ky
+	lamed	.LYmed. bu	ly
+	mem	.mem. bu	my
+	nun	.nun. bu	ny
+	samekh	.samex. bu	samex. bu
+	ayin	.ai,in. bu	.ai,in bu
+	pe	.pex. bu	py
+	tzadi	.tsadik. bu	tsadik. bu
+	quf	.kuf. bu	ky. bu
+	resh	.rec. bu	ry
+	shin	.cin. bu	cy
+	sin	.sin. bu	sy
+	taf	.taf. bu	ty.
+	dagesh	.daGEC. bu	daGEC. bu
+	hiriq	.xirik. bu	.ibu
+	tzeirekh	.tseirex. bu	.eibu
+	segol	.seGOL. bu	.ebu
+	qubbutz	.kubuts. bu	.ubu
+	qamatz	.kamats. bu	.abu
+	patach	.patax. bu	.a'abu
+	sheva	.cyVAS. bu	.y.bu
+	kholem	.xolem. bu	.obu
+	shuruq	.curuk. bu	.u'ubu
+
+</pre><a name=3Ds19><h3>Proposed lerfu words for some accent marks and mul=
tiple letters</h3>
+<p>
+<cx "lerfu words, proposed for accent marks">  XE "lerfu words: proposed f=
or accent marks"   <cx "lerfu words, proposed for diacritic marks">  XE "le=
rfu words: proposed for diacritic marks"   <cx "lerfu words, proposed for m=
ultiple letters">  XE "lerfu words: proposed for multiple letters"   <cx "a=
ccent marks, proposed lerfu words for">  XE "accent marks: proposed lerfu w=
ords for"   <cx "diacritic marks, proposed lerfu words for">  XE "diacritic=
 marks: proposed lerfu words for"   <cx "multiple letters, proposed lerfu w=
ords for">  XE "multiple letters: proposed lerfu words for"   <p>
+This list is intended to be suggestive, not complete: there are lerfu such=
 as Polish =93dark=94 l and Maltese h-bar that do not yet have symbols.
+	<p>
+<cx "over-ring, proposed lerfu word for">  XE "over-ring: proposed lerfu w=
ord for"   <cx "over-dot, proposed lerfu word for">  XE "over-dot: proposed=
 lerfu word for"   <cx "macron, proposed lerfu word for">  XE "macron: prop=
osed lerfu word for"   <cx "ligatured fi, proposed lerfu word for">  XE "li=
gatured fi: proposed lerfu word for"   <cx "Dutch ij, proposed lerfu word f=
or">  XE "Dutch ij: proposed lerfu word for"   <cx "cedilla, proposed lerfu=
 word for">  XE "cedilla: proposed lerfu word for"   <cx "umlaut, proposed =
lerfu word for">  XE "umlaut: proposed lerfu word for"   <cx "tilde, propos=
ed lerfu word for">  XE "tilde: proposed lerfu word for"   <pre>acute		.aku=
t. bu
+		or 	.pritygal. bu	[pritu galtu]
+	grave		.grav. bu=20
+		or	.zulgal. bu	[zunle galtu]
+	circumflex		.cirkumfleks. bu=20
+		or	.midgal. bu	[midju galtu]
+	tilde		.tildes. bu
+	macron		.makron. bu
+	breve		.brevis. bu
+	over-dot		.garmoc. bu	[gapru mokca]
+	umlaut/trema		.relmoc. bu	[re mokca]
+	over-ring		.garjin. bu	[gapru djine]
+	cedilla		.seDIlys. bu
+	double-acute		.re'akut. bu	[re akut.]
+	ogonek		.ogoniek. bu
+	hacek		.xatcek. bu
+	ligatured fi		tei fy. ibu foi
+	Danish/Latin ae	tei .abu .ebu foi
+	Dutch ij		tei .ibu jy. foi
+	German es-zed		tei sy. zy. foi
+</pre><a name=3Ds20><h3>Proposed lerfu words for radio communication</h3>
+<p>
+<cx "lerfu words, proposed for noisy environments">  XE "lerfu words: prop=
osed for noisy environments"   <cx "lerfu words, proposed for radio communi=
cation">  XE "lerfu words: proposed for radio communication"   <cx "radio c=
ommunication, proposed lerfu words for">  XE "radio communication: proposed=
 lerfu words for"   <cx "noisy environments, proposed lerfu words for">  XE=
 "noisy environments: proposed lerfu words for"   <cx "ICAO Phonetic Alphab=
et, proposed lerfu words for">  XE "ICAO Phonetic Alphabet: proposed lerfu =
words for"   <cx "Phonetic Alphabet, proposed lerfu words for">  XE "Phonet=
ic Alphabet: proposed lerfu words for"   There is a set of English words wh=
ich are used, by international agreement, as lerfu words (for the English a=
lphabet) over the radio, or in noisy situations where the utmost clarity is=
 required. Formally they are known as the =93ICAO Phonetic Alphabet=94, and=
 are used even in non-English-speaking countries.
+<p>
+This table presents the standard English spellings and proposed Lojban ver=
sions. The Lojbanizations are not straightforward renderings of the English=
 sounds, but make some concessions both to the English spellings of the wor=
ds and to the Lojban pronunciations of the lerfu (thus =93carlis. bu=94, no=
t =93tcarlis. bu=94).
+<p>
+<pre> 	Alfa	.alfas. bu		November	.novembr. bu
+	Bravo	.bravos. bu		Oscar	.oskar. bu
+	Charlie	.carlis. bu		Papa	.paPAS. bu
+	Delta	.deltas. bu		Quebec	.keBEK. bu
+	Echo	.ekos. bu		Romeo	.romios. bu
+	Foxtrot	.fokstrot. bu		Sierra	.sieras. bu
+	Golf	.golf. bu		Tango	.tangos. bu
+	Hotel	.xoTEL. bu		Uniform	.Uniform. bu
+	India	.indias. bu		Victor	.viktas. bu
+	Juliet	.juliet. bu		Whiskey	.uiskis. bu
+	Kilo	.kilos. bu		X-ray	.eksreis. bu
+	Lima	.limas. bu		Yankee	.iankis. bu
+	Mike	.maik. bu		Zulu	.zulus. bu
+
+</body></html>
+<center><p align=3Dcenter>
+<img src=3Dchap18.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+
+<h2>Chapter 18
+<br>
+lojbau mekso: Mathematical Expressions in Lojban</h2>
+<h6>$Revision: 4.3 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<a name=3Ds1><h3>Introductory</h3>
+<cx "mekso, definition">  XE "mekso: definition"   <cx "mathematical expre=
ssion, abbreviation notation">  XE "mathematical expression: abbreviation n=
otation"   <cx "mathematical expression, definition (see also =93mekso=94)"=
>  XE "mathematical expression: definition (see also \=93mekso\=94)"   lojb=
au mekso (=93Lojbanic mathematical-expression=94) is the part of the Lojban=
 language that is tailored for expressing statements of a mathematical char=
acter, or for adding numerical information to non-mathematical statements. =
Its formal design goals include:
+<p>
+<cx "mekso, design goals">  XE "mekso: design goals"   <cx "mekso goal, fo=
r mathematical writing">  XE "mekso goal: for mathematical writing"   <dl c=
ompact><dt>1)	<dd>representing all the different forms of expression used b=
y mathematicians in their normal modes of writing, so that a reader can una=
mbiguously read off mathematical text as written with minimal effort and ex=
pect a listener to understand it;
+</dl><cx "mekso goal, expandable">  XE "mekso goal: expandable"   <cx "mek=
so goal, for common use">  XE "mekso goal: for common use"   <dl compact><d=
t>2)	<dd>providing a vocabulary of commonly used mathematical terms which c=
an readily be expanded to include newly coined words using the full resourc=
es of Lojban;
+</dl><cx "mekso goal, unambiguous">  XE "mekso goal: unambiguous"   <dl co=
mpact><dt>3)	<dd>permitting the formulation, both in writing and in speech,=
 of unambiguous mathematical text;
+</dl><cx "mekso goal, coverage">  XE "mekso goal: coverage"   <cx "mekso g=
oal, precision">  XE "mekso goal: precision"   <dl compact><dt>4)	<dd>encom=
passing all forms of quantified expression found in natural languages, as w=
ell as encouraging greater precision in ordinary language situations than n=
atural languages allow.
+</dl><cx "mekso goals, mathematical notation form">  XE "mekso goals: math=
ematical notation form"   <cx "Polish notation, and mekso goals">  XE "Poli=
sh notation: and mekso goals"   <cx "reverse Polish notation, and mekso goa=
ls">  XE "reverse Polish notation: and mekso goals"   Goal 1 requires that =
mekso not be constrained to a single notation such as Polish notation or re=
verse Polish notation, but make provision for all forms, with the most comm=
only used forms the most easily used.
+<p>
+<cx "mekso goals, and non-mathematical expression">  XE "mekso goals: and =
non-mathematical expression"   Goal 2 requires the provision of several con=
version mechanisms, so that the boundary between mekso and full Lojban can =
be crossed from either side at many points.
+<p>
+<cx "mekso goals, and ambiguity">  XE "mekso goals: and ambiguity"   <cx "=
mathematical notation, international uniqueness of">  XE "mathematical nota=
tion: international uniqueness of"   Goal 3 is the most subtle. Written mat=
hematical expression is culturally unambiguous, in the sense that mathemati=
cians in all parts of the world understand the same written texts to have t=
he same meanings. However, international mathematical notation does not pre=
scribe unique forms. For example, the expression
+<p>
+<pre><a name=3De1d1>1.1)	<math>3x + 2y</math>
+</pre><cx "mathematical notation, and omitted operators">  XE "mathematica=
l notation: and omitted operators"   contains omitted multiplication operat=
ors, but there are other possible interpretations for the strings =933x=94 =
and =932y=94 than as mathematical multiplication. Therefore, the Lojban ver=
bal (spoken and written) form of <a href=3D#e1d1>Example 1.1 </a>must not o=
mit the multiplication operators.
+<p>
+<cx "mekso chapter, table notation convention">  XE "mekso chapter: table =
notation convention"   <cx "mekso chapter, completeness">  XE "mekso chapte=
r: completeness"   The remainder of this chapter explains (in as much detai=
l as is currently possible) the mekso system. This chapter is by intention =
complete as regards mekso components, but only suggestive about uses of tho=
se components =97 as of now, there has been no really comprehensive use mad=
e of mekso facilities, and many matters must await the test of usage to be =
fully clarified.
+
+<p>
+<a name=3Ds2><h3>Lojban numbers</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<lx "PA">  XE "PA selma=92o"   <pre> 	pa	PA	1		xa	PA	6
+	re	PA	2	 	ze	PA	7
+	ci	PA	3		bi	PA	8
+	vo	PA	4		so	PA	9
+	mu	PA	5		no	PA	0
+</pre><cx "numbers, expressing simple">  XE "numbers: expressing simple"  =
 <cx "digits, cmavo for">  XE "digits: cmavo for"   <cx "numbers, as compou=
nd cmavo">  XE "numbers: as compound cmavo"   <cx "ten, expressing as numbe=
r">  XE "ten: expressing as number"   <cx "hundred, expressing as number"> =
 XE "hundred: expressing as number"   The simplest kind of mekso are number=
s, which are cmavo or compound cmavo. There are cmavo for each of the 10 de=
cimal digits, and numbers greater than 9 are made by stringing together the=
 cmavo. Some examples:
+<p>
+<ex "123">  XE "123: example=93  <pre><a name=3De2d1>2.1)	pa re ci
+	one two three
+	123
+	one hundred and twenty three
+<a name=3De2d2>2.2)	pa no
+	one zero
+	10
+	ten
+<a name=3De2d3>2.3)	pa re ci vo mu xa ze bi so no
+	one two three four five six seven eight nine zero
+	1234567890
+	one billion, two hundred and thirty-four million, five hundred and
+sixty-seven thousand, eight hundred and ninety.
+</pre><cx "numbers, greater than 9">  XE "numbers: greater than 9"   There=
fore, there are no separate cmavo for =93ten=94, =93hundred=94, etc.
+<p>
+<cx "number words, pattern in">  XE "number words: pattern in"   There is =
a pattern to the digit cmavo (except for =93no=94, 0) which is worth explai=
ning. The cmavo from 1 to 5 end in the vowels =93a=94, =93e=94, =93i=94, =
=93o=94, =93u=94 respectively; and the cmavo from 6 to 9 likewise end in th=
e vowels =93a=94, =93e=94, =93i=94, and =93o=94 respectively. None of the d=
igit cmavo begin with the same consonant, to make them easy to tell apart i=
n noisy environments.
+<p>
+<a name=3Ds3><h3>Signs and numerical punctuation</h3>
+<p>
+<lx "PA">  XE "PA selma=92o"   The following cmavo are discussed in this s=
ection:
+<p>
+<pre> 	ma'u	PA	positive sign
+	ni'u	PA	negative sign
+	pi	PA	decimal point
+	fi'u	PA	fraction slash
+	ra'e	PA	repeating decimal
+	ce'i	PA	percent sign
+	ki'o	PA	comma between digits
+</pre><cx "signed numbers, expressing">  XE "signed numbers: expressing"  =
 <lx "mau">  XE "mau"   <lx "ni'u">  XE "ni'u"   <cx "positive numbers, exp=
licit expression">  XE "positive numbers: explicit expression"   <cx "negat=
ive numbers, expressing">  XE "negative numbers: expressing"   A number can=
 be given an explicit sign by the use of =93ma'u=94 and =93ni'u=94, which a=
re the positive and negative signs as distinct from the addition, subtracti=
on, and negation operators. For example:
+<p>
+<ex "-1">  XE "-1: example=93  <lx "pa">  XE "pa"   <pre><a name=3De3d1>3.=
1)	ni'u pa
+	negative-sign 1
+	-1
+</pre><cx "signs on numbers, grammar">  XE "signs on numbers: grammar"   G=
rammatically, the signs are part of the number to which they are attached. =
It is also possible to use =93ma'u=94 and =93ni'u=94 by themselves as numbe=
rs; the meaning of these numbers is explained in <a href=3D#s8>Section 8</a=
>.
+<p>
+<cx "numerical punctuation">  XE "numerical punctuation"   <cx "punctuatio=
n, in numbers">  XE "punctuation: in numbers"   <cx "decimal point, as nume=
rical punctuation">  XE "decimal point: as numerical punctuation"   Various=
 numerical punctuation marks are likewise expressed by cmavo, as illustrate=
d in the following examples:
+<p>
+<ex "3.1415">  XE "3.1415: example=93  <lx "pi">  XE "pi"   <pre><a name=
=3De3d2>3.2)	ci pi pa vo pa mu
+	three point one four one five
+	3.1415
+</pre><cx "decimal point, effect of different notations">  XE "decimal poi=
nt: effect of different notations"   (In some cultures, a comma is used ins=
tead of a period in the symbolic version of <a href=3D#e3d2>Example 3.2</a>=
; =93pi=94 is still the Lojban representation for the decimal point.)
+<p>
+<ex "2/7">  XE "2/7: example=93  <lx "fi'u">  XE "fi'u"   <pre><a name=3De=
3d3>3.3)	re fi'u ze
+	two fraction seven
+	2/7
+</pre><cx "fractions, expressing with numerical punctuation">  XE "fractio=
ns: expressing with numerical punctuation"   <cx "fractions, numerator defa=
ult">  XE "fractions: numerator default"   <cx "reciprocal, expression of m=
athematical">  XE "reciprocal: expression of mathematical"   <a href=3D#e3d=
3>Example 3.3 </a>is the name of the number two-sevenths; it is not the sam=
e as =93the result of 2 divided by 7=94 in Lojban, although numerically the=
se two are equal. If the denominator of the fraction is present but the num=
erator is not, the numerator is taken to be 1, thus expressing the reciproc=
al of the following number:
+<p>
+<pre><a name=3De3d4>3.4)	fi'u ze
+	fraction seven
+	1/7
+</pre><cx "repeating decimals, expressing with numerical punctuation">  XE=
 "repeating decimals: expressing with numerical punctuation"   <lx "ra'e"> =
 XE "ra'e"   <pre><a name=3De3d5>3.5)	pi ci mu ra'e pa vo re bi mu ze
+	point three five repeating one four two eight five seven
+	<math>.35142857142857<dots>=85</dots></math>
+</pre><cx "repeating decimals, marking start of repeating portion">  XE "r=
epeating decimals: marking start of repeating portion"   Note that the =93r=
a'e=94 marks unambiguously where the repeating portion =93142857=94 begins.
+<p>
+<lx "ce'i">  XE "ce'i"   <cx "percent, as numerical punctuation">  XE "per=
cent: as numerical punctuation"   <pre><a name=3De3d6>3.6)	ci mu ce'i
+	three five percent
+	35%
+</pre><cx "commas in numbers, as numerical punctuation">  XE "commas in nu=
mbers: as numerical punctuation"   <lx "ki'o">  XE "ki'o"   <pre><a name=3D=
e3d7>3.7)	pa ki'o re ci vo ki'o mu xa ze
+	one comma two three four comma five six seven
+	1,234,567
+</pre><cx "commas in numbers, effect of other notation conventions">  XE "=
commas in numbers: effect of other notation conventions"   (In some culture=
s, spaces are used in the symbolic representation of <a href=3D#e3d7>Exampl=
e 3.7</a>; =93ki'o=94 is still the Lojban representation.)
+<p>
+<cx "commas in numbers, with elided digits">  XE "commas in numbers: with =
elided digits"   It is also possible to have less than three digits between=
 successive =93ki'o=94s, in which case zeros are assumed to have been elide=
d:
+<p>
+<pre><a name=3De3d8>3.8)	pa ki'o re ci ki'o vo
+	one comma two three comma four
+	1,023,004
+</pre><p>In the same way, =93ki'o=94 can be used after =93pi=94 to divide =
fractions into groups of three:
+
+<p>
+<pre><a name=3De3d9>3.9)	pi ki'o re re
+	point comma two two
+	.022
+<a name=3De3d10>3.10)	pi pa ki'o pa re ki'o pa
+	point one comma one two comma one
+	.001012001
+</pre><a name=3Ds4><h3>Special numbers</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<cx "numbers, special">  XE "numbers: special"   <pre> 	ci'i	PA	infinity
+	ka'o	PA	imaginary i, sqrt(-1)
+	pai	PA	(, pi (approx <math>3.14159<dots>=85</dots></math>)
+	te'o	PA	exponential e (approx <math>2.71828<dots>=85</dots></math>)
+	fi'u	PA	golden ratio, (, phi, (1 + sqrt(5))/2 (approx. <math>1.61803<dots=
>=85</dots></math>)
+</pre><cx "fraction, meaning with elided numerator and denominator">The la=
st cmavo is the same as the fraction sign cmavo: a fraction sign with neith=
er numerator nor denominator represents the golden ratio.
+<p>
+Numbers can have any of these digit, punctuation, and special-number cmavo=
 of Sections 2, 3, and 4 in any combination:
+<p>
+<ex "infinity">  XE "infinity: example=93  <lx "ci'i">  XE "ci'i"   <pre><=
a name=3De4d1>4.1)	ma'u ci'i
+	<math>+<font face=3D"Symbol">(&#165;</font><!=96infty=96></math>
+</pre><cx "complex numbers, expressing">  XE "complex numbers: expressing"=
   <lx "ka'o">  XE "ka'o"   <pre><a name=3De4d2>4.2)	ci ka'o re
+	<math>3i2 </math>(a complex number equivalent to =93<math>3 + 2i</math>=
=94)
+</pre><cx "ka'o, as special number compared with as numerical punctuation"=
>  XE "ka'o: as special number compared with as numerical punctuation"   No=
te that =93ka'o=94 is both a special number (meaning =93i=94) and a number =
punctuation mark (separating the real and the imaginary parts of a complex =
number).
+<p>
+<ex "transfinite cardinal">  XE "transfinite cardinal: example=93  <ex "al=
eph null">  XE "aleph null: example=93  <lx "ci'i">  XE "ci'i"   <pre><a na=
me=3De4d3>4.3)	ci'i no
+	infinity zero
+	<math><font face=3D"Symbol">(&#192;</font><!=96aleph=96><sub>0</sub></mat=
h> (a transfinite cardinal)
+</pre><p>The special numbers =93pai=94 and =93te'o=94 are mathematically i=
mportant, which is why they are given their own cmavo:
+<p>
+<pre><a name=3De4d4>4.4)	pai
+	pi, (
+<a name=3De4d5>4.5)	te'o
+	e
+</pre><cx "numerical punctuation, undefined">  XE "numerical punctuation: =
undefined"   However, many combinations are as yet undefined:
+<p>
+<pre><a name=3De4d6>4.6)	pa pi re pi ci
+	1.2.3
+<a name=3De4d7>4.7)	pa ni'u re
+	1 negative-sign 2
+</pre><cx "negative sign, contrasted with subtraction operator">  XE "nega=
tive sign: contrasted with subtraction operator"   <cx "subtraction operato=
r, contrasted with negative sign">  XE "subtraction operator: contrasted wi=
th negative sign"   <a href=3D#e4d7>Example 4.7 </a>is not =931 minus 2=94,=
 which is represented by a different cmavo sequence altogether. It is a sin=
gle number which has not been assigned a meaning. There are many such numbe=
rs which have no well-defined meaning; they may be used for experimental pu=
rposes or for future expansion of the Lojban number system.
+<p>
+It is possible, of course, that some of these =93oddities=94 do have a mea=
ningful use in some restricted area of mathematics. A mathematician appropr=
iating these structures for specialized use needs to consider whether some =
other branch of mathematics would use the structure differently.
+<p>
+More information on numbers may be found in <a href=3D#s8>Sections 8 </a>t=
o <a href=3D#s12>12</a>.
+<p>
+<a name=3Ds5><h3>Simple infix expressions and equations</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	du	GOhA	equals
+	su'i	VUhU	plus
+	vu'u	VUhU	minus
+	pi'i	VUhU	times
+	te'a	VUhU	raised to the power
+	ny.	BY	letter =93n=94
+	vei	VEI	left parenthesis
+	ve'o	VEhO	right parenthesis
+</pre><cx "mathematical notation, infix">  XE "mathematical notation: infi=
x"   <ex "1 + 1 =3D 2">  XE "1 + 1 =3D 2: example=93  <lx "su'i">  XE "su'i=
"   <lx "VUhU">  XE "VUhU selma=92o"   Let us begin at the beginning: one p=
lus one equals two. In Lojban, that sentence translates to:
+<p>
+<pre><a name=3De5d1>5.1)	li pa su'i pa du li re
+	The-number one plus one equals the-number two.
+	<math>1 + 1 =3D 2</math>
+</pre><cx "mathematical equality, expressing">  XE "mathematical equality:=
 expressing"   <cx "du, grammar of">  XE "du: grammar of"   <a href=3D#e5d1=
>Example 5.1</a>, a mekso sentence, is a regular Lojban bridi that exploits=
 mekso features. =93du=94 is the predicate meaning =93x1 is mathematically =
equal to x2=94. It is a cmavo for conciseness, but it has the same grammati=
cal uses as any brivla. Outside mathematical contexts, =93du=94 means =93x1=
 is identical with x2=94 or =93x1 is the same object as x2=94.
+<p>
+<cx "article, number">  XE "article: number"   <cx "the, for talking about=
 numbers themselves">  XE "the: for talking about numbers themselves"   <lx=
 "li">  XE "li"   <cx "number article, explanation of use">  XE "number art=
icle: explanation of use"   <cx "numbers, talking about contrasted with usi=
ng for quantification">  XE "numbers: talking about contrasted with using f=
or quantification"   <cx "numbers, using for quantification contrasted with=
 talking about">  XE "numbers: using for quantification contrasted with tal=
king about"   The cmavo =93li=94 is the number article. It is required when=
ever a sentence talks about numbers as numbers, as opposed to using numbers=
 to quantify things. For example:
+<p>
+<pre><a name=3De5d2>5.2)	le ci prenu
+	the three persons
+</pre>requires no =93li=94 article, because the =93ci=94 is being used to =
specify the number of =93prenu=94. However, the sentence
+<p>
+<ex "3 grams">  XE "3 grams: example=93  <pre><a name=3De5d3>5.3)	levi sfa=
ni cu grake li ci
+	This fly masses-in-grams the-number three.
+	This fly has a mass of 3 grams.
+</pre><cx "measurements, expressing">  XE "measurements: expressing"   <cx=
 "units of measurement, expressing">  XE "units of measurement: expressing"=
   requires =93li=94 because =93ci=94 is being used as a sumti. Note that t=
his is the way in which measurements are stated in Lojban: all the predicat=
es for units of length, mass, temperature, and so on have the measured obje=
ct as the first place and a number as the second place. Using =93li=94 for =
=93le=94 in <a href=3D#e5d2>Example 5.2 </a>would produce
+<p>
+<pre><a name=3De5d4>5.4)	li ci prenu
+	The-number 3 is-a-person.
+</pre>which is grammatical but nonsensical: numbers are not persons.
+<p>
+<lx "su'i">  XE "su'i"   <lx "VUhU">  XE "VUhU selma=92o"   <cx "mathemati=
cal operators">  XE "mathematical operators"   <cx "addition, a mathematica=
l operator">  XE "addition: a mathematical operator"   <cx "positive sign, =
contrasted with addition operator">  XE "positive sign: contrasted with add=
ition operator"   <cx "addition operator, contrasted with positive sign">  =
XE "addition operator: contrasted with positive sign"   The cmavo =93su'i=
=94 belongs to selma'o VUhU, which is composed of mathematical operators, a=
nd means =93addition=94. As mentioned before, it is distinct from =93ma'u=
=94 which means the positive sign as an indication of a positive number:
+<p>
+<ex "+1 + -1 =3D 0">  XE "+1 + -1 =3D 0: example=93  <pre><a name=3De5d5>5=
.5)	li ma'u pa su'i ni'u pa du=20
+li no
+	The-number positive-sign one plus negative-sign one equals=20
+the-number zero.
+	<math>+1 + -1 =3D 0</math>
+</pre><p>Of course, it is legal to have complex mekso on both sides of =93=
du=94:
+<p>
+<cx "du, with complex mekso on both sides">  XE "du: with complex mekso on=
 both sides"   <pre><a name=3De5d6>5.6)	li mu su'i pa du li ci su'i ci
+	The-number five plus one equals the-number three plus three.
+	<math>5 + 1 =3D 3 + 3</math>
+</pre><cx "operators of VUhU, grammar of operands">  XE "operators of VUhU=
: grammar of operands"   <cx "VUhU operands">  XE "VUhU operands"   <cx "op=
erands, contrasted with general sumti">  XE "operands: contrasted with gene=
ral sumti"   <cx "general sumti, contrasted with operands">  XE "general su=
mti: contrasted with operands"   <cx "li, as converter of mekso into sumti"=
>  XE "li: as converter of mekso into sumti"   <cx "conversion of mekso int=
o sumti">  XE "conversion of mekso into sumti"   <cx "conversion into sumti=
 from mekso">  XE "conversion into sumti from mekso"   Why don't we say =93=
li mu su'i li pa=94 rather than just =93li mu su'i pa=94? The answer is tha=
t VUhU operators connect mekso operands (numbers, in <a href=3D#e5d6>Exampl=
e 5.6</a>), not general sumti. =93li=94 is used to make the entire mekso in=
to a sumti, which then plays the roles applicable to other sumti: in <a hre=
f=3D#e5d6>Example 5.6</a>, filling the places of a bridi.
+<p>
+<cx "calculator mathematics, as default in Lojban">  XE "calculator mathem=
atics: as default in Lojban"   <cx "operator precedence, in Lojban default"=
>  XE "operator precedence: in Lojban default"   <cx "operator left-right g=
rouping, as Lojban default">  XE "operator left-right grouping: as Lojban d=
efault"   <cx "precedence, mathematical default">  XE "precedence: mathemat=
ical default"   By default, Lojban mathematics is like simple calculator ma=
thematics: there is no notion of =93operator precedence=94. Consider the fo=
llowing example, where =93pi'i=94 means =93times=94, the multiplication ope=
rator:
+<p>
+<lx "pi'i">  XE "pi'i"   <pre><a name=3De5d7>5.7)	li ci su'i vo pi'i mu du=
 li reci
+	The-number three plus four times five equals the-number two-three.
+	<math>3 + 4 &times;( 5 =3D 23</math>
+</pre><p>Is the Lojban version of <a href=3D#e5d7>Example 5.7 </a>true? No=
! =93<math>3 + 4 &times; ( 5</math>=94 is indeed 23, because the usual conv=
entions of mathematics state that multiplication takes precedence over addi=
tion; that is, the multiplication =93<math>4&times; ( 5</math>=94 is done f=
irst, giving 20, and only then the addition =93<math>3 + 20</math>=94. But =
VUhU operators by default are done left to right, like other Lojban groupin=
g, and so a truthful bridi would be:
+<p>
+<pre><a name=3De5d8>5.8)	li ci su'i vo pi'i mu du li cimu
+	The-number three plus four times five equals the-number three-five.
+	<math>3 + 4 &times;( 5 =3D 35</math>
+</pre><cx "default operator precedence, contrasted with mekso goal">  XE "=
default operator precedence: contrasted with mekso goal"   Here we calculat=
e <math>3 + 4</math> first, giving 7, and then calculate <math>7 &times;(( =
5</math> second, leading to the result 35. While possessing the advantage o=
f simplicity, this result violates the design goal of matching the standard=
s of mathematics. What can be done?
+<p>
+<cx "mathematical notation, and operator precedence">  XE "mathematical no=
tation: and operator precedence"   <cx "operator precedence, and mathematic=
al notation">  XE "operator precedence: and mathematical notation"   <cx "o=
perator precedence, effect of pragmatic convention">  XE "operator preceden=
ce: effect of pragmatic convention"   There are three solutions, all of whi=
ch will probably be used to some degree. The first solution is to ignore th=
e problem. People will say =93li ci su'i vo pi'i mu=94 and mean 23 by it, b=
ecause the notion that multiplication takes precedence over addition is too=
 deeply ingrained to be eradicated by Lojban parsing, which totally ignores=
 semantics. This convention essentially allows semantics to dominate syntax=
 in this one area.
+<p>
+<cx "operator precedence, rationale for default left-grouping">  XE "opera=
tor precedence: rationale for default left-grouping"   <cx "operator preced=
ence in other languages">  XE "operator precedence in other languages"   (W=
hy not hard-wire the precedences into the grammar, as is done in computer p=
rogramming languages? Essentially because there are too many operators, kno=
wn and unknown, with levels of precedence that vary according to usage. The=
 programming language 'C' has 13 levels of precedence, and its list of oper=
ators is not even extensible. For Lojban this approach is just not practica=
l. In addition, hard-wired precedence could not be overridden in mathematic=
al systems such as spreadsheets where the conventions are different.)
+<p>
+<cx "operator precedence, generalized explicit specification">  XE "operat=
or precedence: generalized explicit specification"   The second solution is=
 to use explicit means to specify the precedence of operators. This approac=
h is fully general, but clumsy, and will be explained in <a href=3D#s20>Sec=
tion 20</a>.
+<p>
+<cx "operator precedence, scope modification with bi'e">  XE "operator pre=
cedence: scope modification with bi'e"   <lx "bi'e">  XE "bi'e"   <lx "BIhE=
">  XE "BIhE selma=92o"   <cx "bi'e, effect on following operator">  XE "bi=
'e: effect on following operator"   The third solution is simple but not ve=
ry general. When an operator is prefixed with the cmavo =93bi'e=94 (of selm=
a'o BIhE), it becomes automatically of higher precedence than other operato=
rs not so prefixed. Thus,
+<p>
+<pre><a name=3De5d9>5.9)	li ci su'i vo bi'e pi'i mu du li reci
+	The-number three plus our-times-five equals the-number two-three.
+	<math>3 + 4 (&times; 5 =3D 23</math>
+</pre>is a truthful Lojban bridi. If more than one operator has a =93bi'e=
=94 prefix, grouping is from the right; multiple =93bi'e=94 prefixes on a s=
ingle operator are not allowed.
+<p>
+<cx "operator precedence, specifying by parenthesis">  XE "operator preced=
ence: specifying by parenthesis"   <cx "parenthesis, mathematical">  XE "pa=
renthesis: mathematical"   <lx "vei">  XE "vei"   <lx "ve'o">  XE "ve'o"   =
In addition, of course, Lojban has the mathematical parentheses =93vei=94 a=
nd =93ve'o=94, which can be used just like their written equivalents =93(=
=94 and =93)=94 to group expressions in any way desired:
+<p>
+<lx "te'a">  XE "te'a"   <lx "ny">  XE "ny"   <ex "(n + 1)(n + 1) =3D n^2 =
+ 2n + 1">  XE "(n + 1)(n + 1) =3D n^2 + 2n + 1: example=93  <pre><a name=
=3De5d10>5.10)	li vei ny. su'i pa ve'o pi'i vei ny. su'i pa [ve'o] du
+		li ny. [bi'e] te'a re su'i re bi'e pi'i ny. su'i pa
+	The-number (=93n=94 plus one ) times (=93n=94 plus one ) equals
+		the-number n-power-two plus two-times-=93n=94 plus 1.
+	<math>(n + 1)(n + 1) =3D n<sup>2</sup>+ 2n + 1</math>
+</pre><cx "lerfu strings, in mathematical expressions">  XE "lerfu string:=
 in mathematical expressions"   <cx "lerfu strings, interpretation of contr=
asted with normal mathematical interpretation">  XE "lerfu string: interpre=
tation of contrasted with normal mathematical interpretation"   There are s=
everal new usages in <a href=3D#e5d10>Example 5.10</a>: =93te'a=94 means =
=93raised to the power=94, and we also see the use of the lerfu word =93ny=
=94, representing the letter =93n=94. In mekso, letters stand for just what=
 they do in ordinary mathematics: variables. The parser will accept a strin=
g of lerfu words (called a =93lerfu string=94) as the equivalent of a singl=
e lerfu word, in agreement with computer-science conventions; =93abc=94 is =
a single variable, not the equivalent of =93<math>a (&times; b (&times; c</=
math>=94. (Of course, a local convention could state that the value of a va=
riable like =93<math>abc</math>=94, with a multi-lerfu name, was equal to t=
he values of the variables =93<math>a</math>=94, =93<math>b</math>=94, and =
=93<math>c</math>=94 multiplied together.)
+<p>
+<cx "times, explicit expression of">  XE "times: explicit expression of"  =
 <cx "times, implicit expression of">  XE "times: implicit expression of"  =
 <cx "multiplication, explicit expression of">  XE "multiplication: explici=
t expression of"   <cx "multiplication, implicit expression of">  XE "multi=
plication: implicit expression of"   The explicit operator =93pi'i=94 is re=
quired in the Lojban verbal form whereas multiplication is implicit in the =
symbolic form. Note that =93ve'o=94 (the right parenthesis) is an elidable =
terminator: the first use of it in <a href=3D#e5d10>Example 5.10 </a>is req=
uired, but the second use (marked by square brackets) could be elided. Addi=
tionally, the first =93bi'e=94 (also marked by square brackets) is not nece=
ssary to get the proper grouping, but it is included here for symmetry with=
 the other one.
+<p>
+<a name=3Ds6><h3>Forethought operators (Polish notation, functions)</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre>	boi	BOI	numeral/lerfu string terminator
+	va'a	VUhU	negation/additive inverse
+	pe'o	PEhO	forethought flag
+	ku'e	KUhE	forethought terminator
+	ma'o	MAhO	convert operand to operator
+
+	py.	BY	letter =93p=94
+	xy.	BY	letter =93x=94
+	zy.	BY	letter =93z=94
+	fy.	BY	letter =93f=94
+</pre><cx "infix mathematical notation, shortcomings of">  XE "infix mathe=
matical notation: shortcomings of"   <cx "mathematical notation, infix shor=
tcomings">  XE "mathematical notation: infix shortcomings"   The infix form=
 explained so far is reasonable for many purposes, but it is limited and ri=
gid. It works smoothly only where all operators have exactly two operands, =
and where precedences can either be assumed from context or are limited to =
just two levels, with some help from parentheses.
+<p>
+<cx "Polish notation, definition">  XE "Polish notation: definition"   <cx=
 "forethought mathematical notation (see also Polish)">  XE "forethought ma=
thematical notation (see also Polish)"   <cx "mathematical notation, foreth=
ought (see also Polish)">  XE "mathematical notation: forethought (see also=
 Polish)"   But there are many operators which do not have two operands, or=
 which have a variable number of operands. The preferred form of expression=
 in such cases is the use of =93forethought operators=94, also known as Pol=
ish notation. In this style of writing mathematics, the operator comes firs=
t and the operands afterwards:
+<p>
+<ex "sum of 1, 2, 3">  XE "sum of 1: 2, 3, example=93  <pre><a name=3De6d1=
>6.1)	li su'i paboi reboi ci[boi] du li xa
+	The-number the-sum-of one two three equals the-number six.
+	<math><r>sum</r>(1,2,3) =3D 6</math>
+</pre><cx "Polish notation, separating operands in">  XE "Polish notation:=
 separating operands in"   <cx "Polish notation, and use of boi">  XE "Poli=
sh notation: and use of boi"   <lx "boi">  XE "boi"   <cx "boi, in Polish n=
otation">  XE "boi: in Polish notation"   Note that the normally elidable n=
umber terminator =93boi=94 is required after =93pa=94 and =93re=94 because =
otherwise the reading would be =93pareci=94 =3D 123. It is not required aft=
er =93ci=94 but is inserted here in brackets for the sake of symmetry. The =
only time =93boi=94 is required is, as in <a href=3D#e6d1>Example 6.1</a>, =
when there are two consecutive numbers or lerfu strings.
+<p>
+<lx "ku'e">  XE "ku'e"   <cx "Polish notation, end-of-operands indicator">=
  XE "Polish notation: end-of-operands indicator"   Forethought mekso can u=
se any number of operands, in <a href=3D#e6d1>Example 6.1, </a>three. How d=
o we know how many operands there are in ambiguous circumstances? The usual=
 Lojban solution is employed: an elidable terminator, namely =93ku'e=94. He=
re is an example:
+<p>
+<lx "va'a">  XE "va'a"   <ex "plus negative of">  XE "plus negative of: ex=
ample=93  <pre><a name=3De6d2>6.2)	li py. su'i va'a ny. ku'e su'i zy du li =
xy.
+	The-number =93p=94 plus negative-of (=93n=94) plus =93z=94 equals the-num=
ber =93x=94.
+	<math>p + -n + z =3D x</math>
+</pre>where we know that =93va'a=94 is a forethought operator because ther=
e is no operand preceding it.
+<p>
+<cx "va'a, contrasted with vu'u and ni'u">  XE "va'a: contrasted with vu'u=
 and ni'u"   <cx "vu'u, contrasted with va'a and ni'u">  XE "vu'u: contrast=
ed with va'a and ni'u"   <cx "ni'u, contrasted with va'a and vu'u">  XE "ni=
'u: contrasted with va'a and vu'u"   <cx "negation operator, contrasted wit=
h subtraction operator">  XE "negation operator: contrasted with subtractio=
n operator"   <cx "subtraction operator, contrasted with negation operator"=
>  XE "subtraction operator: contrasted with negation operator"   <cx "nega=
tion operator, contrasted with negative sign">  XE "negation operator: cont=
rasted with negative sign"   <cx "negative sign, contrasted with negation o=
perator">  XE "negative sign: contrasted with negation operator"   <lx "vu'=
u">  XE "vu'u"   <lx "ni'u">  XE "ni'u"   =93=93va'a=94 is the numerical ne=
gation operator, of selma'o VUhU. In contrast, j =93vu'u=94 is not used for=
 numerical negation, but only for subtraction, as it always has two or more=
 operands. Do not confuse =93va'a=94 and =93vu'u=94, which are operators, w=
ith =93ni'u=94, which is part of a number.
+<p>
+<cx "Polish notation, operator =85ku'e compared with parenthesization">  X=
E "Polish notation: operator =85ku'e compared with parenthesization"   <cx =
"Polish notation, vei =85ve'o contrasted with operator =85ku'e">  XE "Polis=
h notation: vei =85ve'o contrasted with operator =85ku'e"   <cx "vei =85ve'=
o, contrasted with operator =85ku'e in Polish notation">  XE "vei =85ve'o: =
contrasted with operator =85ku'e in Polish notation"   <cx "operator =85ku'=
e in Polish notation, contrasted with vei =85ve'o">  XE "operator =85ku'e i=
n Polish notation: contrasted with vei =85ve'o"   In <a href=3D#e6d2>Exampl=
e 6.2</a>, the operator =93va'a=94 and the terminator =93ku'e=94 serve in e=
ffect as parentheses. (The regular parentheses =93vei=94 and =93ve'o=94 are=
 NOT used for this purpose.) If the =93ku'e=94 were omitted, the =93su'i zy=
=94 would be swallowed up by the =93va'a=94 forethought operator, which wou=
ld then appear to have two operands, =93ny=94 and =93su'i zy.=94, where the=
 latter is also a forethought expression.
+<p>
+<cx "functional notation, standard">  XE "functional notation: standard"  =
 Forethought mekso is also useful for matching standard functional notation=
. How do we represent =93z =3D f(x)=94? The answer is:
+<p>
+<ex "z =3D f(x)">  XE "z =3D f(x): example=93  <lx "ma'o">  XE "ma'o"   <p=
re><a name=3De6d3>6.3)	li zy du li ma'o fy.boi xy.
+	The-number z equals the-number the-operator f x.
+	<math>z =3D f(x)</math>
+</pre><cx "lerfu strings as functions, in mathematics">  XE "lerfu string:=
 as function, in mathematics"   <cx "lerfu strings as variables, in mathema=
tics">  XE "lerfu string: as variable, in mathematics"   Again, no parenthe=
ses are used. The construct =93ma'o fy.boi=94 is the equivalent of an opera=
tor, and appears in forethought here (although it could also be used as a r=
egular infix operator). In mathematics, letters sometimes mean functions an=
d sometimes mean variables, with only the context to tell which. Lojban cho=
oses to accept the variable interpretation as the default, and uses the spe=
cial flag =93ma'o=94 to mark a lerfu string as an operator. The cmavo =93xy=
.=94 and =93zy.=94 are variables, but =93fy.=94 is an operator (a function)=
 because =93ma'o=94 marks it as such. The =93boi=94 is required because oth=
erwise the =93xy.=94 would look like part of the operator name. (The use of=
 =93ma'o=94 can be generalized from lerfu strings to any mekso operand: see=
 <a href=3D#s21>Section 21</a>.)
+<p>
+<lx "pe'o">  XE "pe'o"   <cx "Polish notation, explicitly marking as">  XE=
 "Polish notation: explicitly marking as"   When using forethought mekso, t=
he optional marker =93pe'o=94 may be placed in front of the operator. This =
usage can help avoid confusion by providing clearly marked =93pe'o=94 and =
=93ku'e=94 pairs to delimit the operand list. <a href=3D#e6d{1}>Examples 6.=
1 </a>to <a href=3D#e6d3>6.3</a>, respectively, with explicit =93pe'o=94 an=
d =93ku'e=94:
+<p>
+<pre><a name=3De6d4>6.4)	li pe'o su'i paboi reboi ciboi ku'e du li xa
+<a name=3De6d5>6.5)	li py. su'i pe'o va'a ny. ku'e su'i zy du li xy.
+<a name=3De6d6>6.6)	li zy du li pe'o ma'o fy.boi xy. ku'e
+</pre><cx "Polish notation, operands with infix expressions">  XE "Polish =
notation: operands with infix expressions"   <cx "infix expressions, in ope=
rands being used in Polish notation">  XE "infix expressions: in operands b=
eing used in Polish notation"   Note: When using forethought mekso, be sure=
 that the operands really are operands: they cannot contain regular infix e=
xpressions unless parenthesized with =93vei=94 and =93ve'o=94. An earlier v=
ersion of the complex <a href=3D#e17d6>Example 17.6 </a>came to grief becau=
se I forgot this rule.
+<p>
+<a name=3Ds7><h3>Other useful selbri for mekso bridi</h3>
+<p>
+<cx "mathematical inequalities, expressing">  XE "mathematical inequalitie=
s: expressing"   So far our examples have been isolated mekso (it is legal =
to have a bare mekso as a sentence in Lojban) and equation bridi involving =
=93du=94. What about inequalities such as =93<math>x < 5</math>=94? The ans=
wer is to use a bridi with an appropriate selbri, thus:
+<p>
+<ex "x < 5">  XE "x < 5: example=93  <pre><a name=3De7d1>7.1)	li xy. mleca=
 li mu
+	The-number x is-less-than the-number 5.
+</pre><p>Here is a partial list of selbri useful in mathematical bridi:
+<p>
+<pre> 	du:  x1 is identical to x2, x3, x4,=A0<dots>=85</dots>
+	dunli:  x1 is equal/congruent to x2 in/on property/quality/dimension/quan=
tity x3
+	mleca:  x1 is less than x2
+	zmadu:  x1 is greater than x2
+	dubjavme'a:  x1 is less than or equal to x2  [du ja mleca, equal or less]
+	dubjavmau:  x1 is greater than or equal to x2  [du ja zmadu, equal or gre=
ater]
+	tamdu'i:  x1 is similar to x2  [tarmi dunli, shape-equal]
+	turdu'i:  x1 is isomorphic to x2  [stura dunli, structure-equal]
+	cmima:  x1 is a member of set x2
+	gripau:  x1 is a subset of set x2  [girzu pagbu, set-part]
+	na'ujbi:  x1 is approximately equal to x2  [namcu jibni, number-near]
+	terci'e:  x1 is a component with function x2 of system x3
+</pre><cx "dunli, contrasted with du">  XE "dunli: contrasted with du"   <=
cx "du, contrasted with dunli">  XE "du: contrasted with dunli"   Note the =
difference between =93dunli=94 and =93du=94; =93dunli=94 has a third place =
that specifies the kind of equality that is meant. =93du=94 refers to actua=
l identity, and can have any number of places:
+<p>
+<ex "p =3D x =3D z">  XE "p =3D x =3D z: example=93  <pre><a name=3De7d2>7=
.2)	py. du xy.boi zy.
+	=93p=94 is-identical-to =93x=94 =93z=94
+	<math>p =3D x =3D z</math>
+</pre><p>Lojban bridi can have only one predicate, so the =93du=94 is not =
repeated.
+<p>
+<cx "falsity of mathematical relation, expressing">  XE "falsity of mathem=
atical relation: expressing"   Any of these selbri may usefully be prefixed=
 with =93na=94, the contradictory negation cmavo, to indicate that the rela=
tion is false:
+<p>
+<pre><a name=3De7d3>7.3)	li re su'i re na du li mu
+	The-number 2 + 2 is-not equal-to the-number 5.
+	<math>2 + 2 ( <font face=3D"Symbol">&#189;</font><!=96notequal=96>5</math=
>
+</pre><p>As usual in Lojban, negated bridi say what is false, and do not s=
ay anything about what might be true.
+<p>
+<a name=3Ds8><h3>Indefinite numbers</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ro	PA	all
+	so'a	PA	almost all
+	so'e	PA	most
+	so'i	PA	many
+	so'o	PA	several
+	so'u	PA	a few
+	no'o	PA	the typical number of
+	da'a	PA	all but (one) of
+	piro	PA+PA	the whole of/all of
+	piso'a	PA+PA	almost the whole of
+	piso'e	PA+PA	most of
+	piso'i	PA+PA	much of
+	piso'o	PA+PA	a small part of
+	piso'u	PA+PA	a tiny part of
+	pino'o	PA+PA	the typical portion of
+	rau	PA	enough
+	du'e	PA	too many
+	mo'a	PA	too few
+	pirau	PA+PA	enough of
+	pidu'e	PA+PA	too much of
+	pimo'a	PA+PA	too little of
+</pre><lx "PA">  XE "PA selma=92o"   <cx "ro, compared with pa">  XE "ro: =
compared with pa"   <cx "indefinite numbers">  XE "indefinite numbers"   <c=
x "numbers, indefinite">  XE "numbers: indefinite"   <lx "ro">  XE "ro"   N=
ot all the cmavo of PA represent numbers in the usual mathematical sense. F=
or example, the cmavo =93ro=94 means =93all=94 or =93each=94. This number d=
oes not have a definite value in the abstract: =93li ro=94 is undefined. Bu=
t when used to count or quantify something, the parallel between =93ro=94 a=
nd =93pa=94 is clearer:
+<p>
+<pre><a name=3De8d1>8.1)	mi catlu pa prenu
+	I look-at one person
+<a name=3De8d2>8.2)	mi catlu ro prenu
+	I look-at all persons
+</pre><a href=3D#e8d1>Example 8.1 </a>might be true, whereas <a href=3D#e8=
d2>Example 8.2 </a>is almost certainly false.
+<p>
+<lx "so'a">  XE "so'a"   <lx "so'e">  XE "so'e"   <lx "so'i">  XE "so'i"  =
 <lx "so'o">  XE "so'o"   <lx "so'u">  XE "so'u"   The cmavo =93so'a=94, =
=93so'e=94, =93so'i=94, =93so'o=94, and =93so'u=94 represent a set of indef=
inite numbers less than =93ro=94. As you go down an alphabetical list, the =
magnitude decreases:
+
+<p>
+<pre><a name=3De8d3>8.3)	mi catlu so'a prenu
+	I look-at almost-all persons
+<a name=3De8d4>8.4)	mi catlu so'e prenu
+	I look-at most persons
+<a name=3De8d5>8.5)	mi catlu so'i prenu
+	I look-at many persons
+<a name=3De8d6>8.6)	mi catlu so'o prenu
+	I look-at several persons
+<a name=3De8d7>8.7)	mi catlu so'u prenu
+	I look-at a-few persons
+</pre><cx "so'e, meaning of">  XE "so'e: meaning of"   The English equival=
ents are only rough: the cmavo provide space for up to five indefinite numb=
ers between =93ro=94 and =93no=94, with a built-in ordering. In particular,=
 =93so'e=94 does not mean =93most=94 in the sense of =93a majority=94 or =
=93more than half=94.
+<p>
+<lx "pi">  XE "pi"   <cx "pi, effect on indefinite numbers">  XE "pi: effe=
ct on indefinite numbers"   <cx "portion of whole, expressing">  XE "portio=
n of whole: expressing"   <cx "indefinite numbers, effect of pi on">  XE "i=
ndefinite numbers: effect of pi on"   <ex "whole of">  XE "whole of: exampl=
e=93  Each of these numbers, plus =93ro=94, may be prefixed with =93pi=94 (=
the decimal point) in order to make a fractional form which represents part=
 of a whole rather than some elements of a totality. =93piro=94 therefore m=
eans =93the whole of=94:
+<p>
+<ex "eat bread">  XE "eat bread: example=93  <pre><a name=3De8d8>8.8)	mi c=
itka piro lei nanba
+	I eat the-whole-of the-mass-of bread
+</pre><cx "masses, expressing portions of">  XE "mass: expressing portions=
 of"   Similarly, =93piso'a=94 means =93almost the whole of=94; and so on d=
own to =93piso'u=94, =93a tiny part of=94. These numbers are particularly a=
ppropriate with masses, which are usually measured rather than counted, as =
<a href=3D#e8d8>Example 8.8 </a>shows.
+<p>
+<lx "no'o">  XE "no'o"   <cx "typical value, contrasted with mathematical =
average">  XE "typical value: contrasted with mathematical average"   In ad=
dition to these cmavo, there is =93no'o=94, meaning =93the typical value=94=
, and =93pino'o=94, meaning =93the typical portion=94: Sometimes =93no'o=94=
 can be translated =93the average value=94, but the average in question is =
not, in general, a mathematical mean, median, or mode; these would be more =
appropriately represented by operators.
+<p>
+<pre><a name=3De8d9>8.9)	mi catlu no'o prenu
+	I look-at a-typical-number-of persons
+<a name=3De8d10>8.10)	mi citka pino'o lei nanba
+	I eat a-typical-amount-of the-mass-of bread.
+</pre><lx "da'a">  XE "da'a"   =93=93da'a=94 is a related cmavo meaning =
=93all but=94:
+<p>
+<pre><a name=3De8d11>8.11)	mi catlu da'a re prenu
+	I look-at all-but two persons
+<a name=3De8d12>8.12)	mi catlu da'a so'u prenu
+	I look-at all-but a-few persons
+</pre><a href=3D#e8d12>Example 8.12 </a>is similar in meaning to <a href=
=3D#e8d3>Example 8.3</a>.
+<p>
+<cx "da'a, default number for">  XE "da'a: default number for"   If no num=
ber follows =93da'a=94, then =93pa=94 is assumed; =93da'a=94 by itself mean=
s =93all but one=94, or in ordinal contexts =93all but the last=94:
+<p>
+<pre><a name=3De8d13>8.13)	ro ratcu ka'e citka da'a ratcu
+	all rats can eat all-but-one rats.
+	All rats can eat all other rats.
+</pre><ex "eat themselves">  XE "eat themselves: example=93  (The use of =
=93da'a=94 means that <a href=3D#e8d13>Example 8.13 </a>does not require th=
at all rats can eat themselves, but does allow it. Each rat has one rat it =
cannot eat, but that one might be some rat other than itself. Context often=
 dictates that =93itself=94 is, indeed, the =93other=94 rat.)
+<p>
+<cx "ma'u, with elided number">  XE "ma'u: with elided number"   <cx "ni'u=
, with elided number">  XE "ni'u: with elided number"   <lx "ma'u">  XE "ma=
'u"   <lx "ni'u">  XE "ni'u"   As mentioned in <a href=3D#s3>Section 3</a>,=
 =93ma'u=94 and =93ni'u=94 are also legal numbers, and they mean =93some po=
sitive number=94 and =93some negative number=94 respectively.
+<p>
+<pre><a name=3De8d14>8.14)	li ci vu'u re du li ma'u
+	the-number 3 - 2 =3D some-positive-number
+<a name=3De8d15>8.15)	li ci vu'u vo du li ni'u
+	the-number 3 - 4 =3D some-negative-number
+<a name=3De8d16>8.16)	mi ponse ma'u rupnu
+	I possess a-positive-number-of currency-units.
+</pre><cx "subjective amounts, expressing">  XE "subjective amounts: expre=
ssing"   <cx "indefinite values, subjective">  XE "indefinite values: subje=
ctive"   <lx "rau">  XE "rau"   <lx "du'e">  XE "du'e"   <lx "mo'a">  XE "m=
o'a"   All of the numbers discussed so far are objective, even if indefinit=
e. If there are exactly six superpowers (=93rairgugde=94, =93superlative-st=
ates=94) in the world, then =93ro rairgugde=94 means the same as =93xa rair=
gugde=94. It is often useful, however, to express subjective indefinite val=
ues. The cmavo =93rau=94 (enough), =93du'e=94 (too many), and =93mo'a=94 (t=
oo few) are then appropriate:
+<p>
+<ex "enough currency">  XE "enough currency: example=93  <pre><a name=3De8=
d17>8.17)	mi ponse rau rupnu
+	I possess enough currency-units.
+</pre><lx "pi">  XE "pi"   <cx "subjective portions, expressing">  XE "sub=
jective portions: expressing"   <cx "indefinite portions, subjective">  XE =
"indefinite portions: subjective"   Like the =93so'a=94-series, =93rau=94, =
=93du'e=94, and =93mo'a=94 can be preceded by =93pi=94; for example, =93pir=
au=94 means =93a sufficient part of.=94
+<p>
+<cx "indefinite numbers, combined with definite">  XE "indefinite numbers:=
 combined with definite"   <cx "definite numbers, combined with indefinite"=
>  XE "definite numbers: combined with indefinite"   Another possibility is=
 that of combining definite and indefinite numbers into a single number. Th=
is usage implies that the two kinds of numbers have the same value in the g=
iven context:
+<p>
+<ex "both dogs">  XE "both dogs: example=93  <pre><a name=3De8d18>8.18)	mi=
 viska le rore gerku
+	I saw the all-of/two dogs.
+	I saw both dogs.
+<a name=3De8d19>8.19)	mi speni so'ici prenu
+	I am-married-to many/three persons.
+	I am married to three persons (which is =93many=94 in the circumstances).
+</pre><a href=3D#e8d19>Example 8.19 </a>assumes a mostly monogamous cultur=
e by stating that three is =93many=94.
+<p>
+<a name=3Ds9><h3>Approximation and inexact numbers</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ji'i	PA	approximately
+	su'e	PA	at most
+	su'o	PA	at least
+	me'i	PA	less than
+	za'u	PA	more than
+</pre><lx "ji'i">  XE "ji'i"   <cx "approximate numbers, expressing">  XE =
"approximate numbers: expressing"   <cx "ji'i, effect of placement">  XE "j=
i'i: effect of placement"   The cmavo =93ji'i=94 (of selma'o PA) is used in=
 several ways to indicate approximate or rounded numbers. If it appears at =
the beginning of a number, the whole number is approximate:
+
+<p>
+<ex "approximately 40">  XE "approximately 40: example=93  <pre><a name=3D=
e9d1>9.1)	ji'i vo no
+	approximation four zero
+	approximately 40
+</pre><cx "approximate numbers, expressing some exactness of">  XE "approx=
imate numbers: expressing some exactness of"   If =93ji'i=94 appears in the=
 middle of a number, all the digits following it are approximate:
+<p>
+<pre><a name=3De9d2>9.2)	vo no ji'i mu no
+	four zero approximation five zero
+	roughly 4050 (where the =93four thousand=94 is exact, 	but the =93fifty=
=94 is
+approximate)
+</pre><cx "truncation of number, expressing">  XE "truncation of number: e=
xpressing"   <cx "rounded numbers, expressing">  XE "rounded numbers: expre=
ssing"   If =93ji'i=94 appears at the end of a number, it indicates that th=
e number has been rounded. In addition, it can then be followed by a sign c=
mavo (=93ma'u=94 or =93ni'u=94), which indicate truncation towards positive=
 or negative infinity respectively.
+<p>
+<pre><a name=3De9d3>9.3)	re pi ze re ji'i
+	two point seven two approximation
+	2.72 (rounded)
+</pre><ex "rounded up">  XE "rounded up: example=93  <pre><a name=3De9d4>9=
.4)	re pi ze re ji'i ma'u
+	two point seven two approximation positive-sign
+	2.72 (rounded up)
+</pre><ex "rounded down">  XE "rounded down: example=93  <pre><a name=3De9=
d5>9.5)	re pi ze pa ji'i ni'u
+	two point seven one approximation negative-sign
+	2.71 (rounded down)
+</pre><cx "ji'i, with elided number">  XE "ji'i: with elided number"   <a =
href=3D#e9d{3}>Examples 9.3 </a>through <a href=3D#e9d5>9.5 </a>are all app=
roximations to =93te'o=94 (exponential e). =93ji'i=94 can also appear by it=
self, in which case it means =93approximately the typical value in this con=
text=94.
+<p>
+<lx "su'e">  XE "su'e"   <lx "su'o">  XE "su'o"   <lx "me'i">  XE "me'i"  =
 <lx "za'u">  XE "za'u"   <cx "inexact numbers with bounds">  XE "inexact n=
umbers with bounds"   The four cmavo =93su'e=94, =93su'o=94, =93me'i=94, an=
d =93za'u=94, also of selma'o PA, express inexact numbers with upper or low=
er bounds:
+<p>
+<ex "at most">  XE "at most: example=93  <pre><a name=3De9d6>9.6)	mi catlu=
 su'e re prenu
+	I look-at at-most two persons.
+</pre><ex "at least">  XE "at least: example=93  <pre><a name=3De9d7>9.7)	=
mi catlu su'o re prenu
+	I look-at at-least two persons.
+</pre><ex "less than">  XE "less than: example=93  <pre><a name=3De9d8>9.8=
)	mi catlu me'i re prenu
+	I look-at less-than two persons.
+</pre><ex "more than">  XE "more than: example=93  <pre><a name=3De9d9>9.9=
)	mi catlu za'u re prenu
+	I look-at more-than two persons.
+</pre><cx "less than, contrasted with more than, at least, at most">  XE "=
less than: contrasted with more than, at least, at most"   <cx "more than, =
contrasted with less than, at least, at most">  XE "more than: contrasted w=
ith less than, at least, at most"   <cx "at least, contrasted with more tha=
n, less than, at most">  XE "at least: contrasted with more than, less than=
, at most"   <cx "at most, contrasted with more than, at least, less than">=
  XE "at most: contrasted with more than, at least, less than"   <ex "exact=
ly two">  XE "exactly two: example=93  <cx "exact number, expressing">  XE =
"exact number: expressing"   <cx "plural, Lojban equivalent of">  XE "plura=
l: Lojban equivalent of"   Each of these is a subtly different claim: <a hr=
ef=3D#e9d7>Example 9.7 </a>is true of two or any greater number, whereas <a=
 href=3D#e9d9>Example 9.9 </a>requires three persons or more. Likewise, <a =
href=3D#e9d6>Example 9.6 </a>refers to zero, one, or two; <a href=3D#e9d8>E=
xample 9.8 </a>to zero or one. (Of course, when the context allows numbers =
other than non-negative integers, =93me'i re=94 can be any number less than=
 2, and likewise with the other cases.) The exact quantifier, =93exactly 2,=
 neither more nor less=94 is just =93re=94. Note that =93su'ore=94 is the e=
xact Lojban equivalent of English plurals.
+<p>
+<cx "su'e, with elided number">  XE "su'e: with elided number"   <cx "su'o=
, with elided number">  XE "su'o: with elided number"   <cx "me'i, with eli=
ded number">  XE "me'i: with elided number"   <cx "za'u, with elided number=
">  XE "za'u: with elided number"   If no number follows one of these cmavo=
, =93pa=94 is understood: therefore,
+<p>
+<pre><a name=3De9d10>9.10)	mi catlu su'o prenu
+	I look-at at-least [one] person.
+</pre>is a meaningful claim.
+<p>
+<lx "pi">  XE "pi"   <cx "inexact portions with bounds">  XE "inexact port=
ions with bounds"   Like the numbers in <a href=3D#s8>Section 8</a>, all of=
 these cmavo may be preceded by =93pi=94 to make the corresponding quantifi=
ers for part of a whole. For example, =93pisu'o=94 means =93at least some p=
art of=94. The quantifiers =93ro=94, =93su'o=94, =93piro=94, and =93pisu'o=
=94 are particularly important in Lojban, as they are implicitly used in th=
e descriptions introduced by the cmavo of selma'o LA and LE, as explained i=
n <a href=3Dchap6.html>Chapter 6</a>. Descriptions in general are outside t=
he scope of this chapter.
+<p>
+<a name=3Ds10><h3>Non-decimal and compound bases</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ju'u	VUhU	to the base
+	dau	PA	hex digit A =3D 10
+	fei	PA	hex digit B =3D 11
+	gai	PA	hex digit C =3D 12
+	jau	PA	hex digit D =3D 13
+	rei	PA	hex digit E =3D 14
+	vai	PA	hex digit F =3D 15
+	pi'e	PA	compound base point
+</pre><cx "radix, decimal (see also base)">  XE "radix: decimal (see also =
base)"   In normal contexts, Lojban assumes that all numbers are expressed =
in the decimal (base 10) system. However, other bases are possible, and may=
 be appropriate in particular circumstances.
+<p>
+<cx "base, specifying">  XE "base: specifying"   <cx "binary system, speci=
fying numbers in (see also base)">  XE "binary system: specifying numbers i=
n (see also base)"   <cx "octal system, specifying numbers in (see also bas=
e)">  XE "octal system: specifying numbers in (see also base)"   To specify=
 a number in a particular base, the VUhU operator =93ju'u=94 is suitable:
+<p>
+<pre><a name=3De10d1>10.1)	li pa no pa no ju'u re du li pa no
+	The-number 1010 base 2 equals the-number 10.
+</pre><cx "base, assumed">  XE "base: assumed"   <cx "base, changing perma=
nently">  XE "base: changing permanently"   Here, the final =93pa no=94 is =
assumed to be base 10, as usual; so is the base specification. (The base ma=
y also be changed permanently by a metalinguistic specification; no standar=
d way of doing so has as yet been worked out.)
+<p>
+<cx "digits, rationale for having 16">  XE "digits: rationale for having 1=
6"   <cx "hexadecimal system, specifying numbers in (see also base)">  XE "=
hexadecimal system: specifying numbers in (see also base)"   Lojban has dig=
its for representing bases up to 16, because 16 is a base often used in com=
puter applications. In English, it is customary to use the letters A-F as t=
he base 16 digits equivalent to the numbers ten through fifteen. In Lojban,=
 this ambiguity is avoided:
+<p>
+<ex "ABC base 16">  XE "ABC base 16: example=93  <pre><a name=3De10d2>10.2=
)	li daufeigai ju'u paxa du li rezevobi
+	The-number ABC base 16 equals the-number 2748.
+<a name=3De10d3>10.3)	li jaureivai ju'u paxa du li cimuxaze
+	The-number DEF base 16 equals the-number 3567.
+</pre><cx "digits beyond 9, word pattern">  XE "digits beyond 9: word patt=
ern"   Note the pattern in the cmavo: the diphthongs =93au=94, =93ei=94, =
=93ai=94 are used twice in the same order. The digits for A to D use conson=
ants different from those used in the decimal digit cmavo; E and F unfortun=
ately overlap 2 and 4 =97 there was simply not enough available cmavo space=
 to make a full differentiation possible. The cmavo are also in alphabetica=
l order.
+<p>
+<cx "base point, in bases other than 10">  XE "base point: in bases other =
than 10"   <cx "decimal point, in bases other than 10">  XE "decimal point:=
 in bases other than 10"   The base point =93pi=94 is used in non-decimal b=
ases just as in base 10:
+<p>
+<ex "F.8 base 16">  XE "F.8 base 16: example=93  <pre><a name=3De10d4>10.4=
)	li vai pi bi ju'u paxa du li pamu pi mu
+	The-number F.8 base 16 equals the-number 15.5.
+</pre><cx "ju'u, grammar of">  XE "ju'u: grammar of"   <lx "VUhU">  XE "VU=
hU selma=92o"   <cx "base, non-constant">  XE "base: non-constant"   Since =
=93ju'u=94 is an operator of selma'o VUhU, it is grammatical to use any ope=
rand as the left argument. Semantically, however, it is undefined to use an=
ything but a numeral string on the left. The reason for making =93ju'u=94 a=
n operator is to allow reference to a base which is not a constant.
+<p>
+<ex "hours:minutes:seconds">  XE "hours\:minutes\:seconds: example=93  <cx=
 "base varying for each digit, separator for">  XE "base varying for each d=
igit: separator for"   <lx "pi'e">  XE "pi'e"   <cx "compound base, separat=
or for">  XE "compound base: separator for"   <cx "compound base, definitio=
n">  XE "compound base: definition"   There are some numerical values that =
require a =93base=94 that varies from digit to digit. For example, times re=
presented in hours, minutes, and seconds have, in effect, three =93digits=
=94: the first is base 24, the second and third are base 60. To express suc=
h numbers, the compound base separator =93pi'e=94 is used:
+<p>
+<pre><a name=3De10d5>10.5)	ci pi'e rere pi'e vono
+	3:22:40
+</pre><cx "compound base, expressing digits in">  XE "compound base: expre=
ssing digits in"   Each digit sequence separated by instances of =93pi'e=94=
 is expressed in decimal notation, but the number as a whole is not decimal=
 and can only be added and subtracted by special rules:
+<p>
+<pre><a name=3De10d6>10.6)	li ci pi'e rere pi'e vono su'i pi'e ci pi'e cic=
i 	du li ci pi'e rexa pi'e paci
+	The-number 3:22:40 plus :3:33 equals the-number 3:26:13.
+	<math>3:22:40 + 0:3:33 =3D 3:26:13</math>
+</pre><p>Of course, only context tells you that the first part of the numb=
ers in <a href=3D#e10d5>Example 10.5 </a>and <a href=3D#e10d6>Example 10.6 =
</a>is hours, the second minutes, and the third seconds.
+<p>
+<cx "Mayan mathematics, as a system with base larger than 16">  XE "Mayan =
mathematics: as a system with base larger than 16"   <cx "base greater than=
 16, expressing numbers in">  XE "base greater than 16: expressing numbers =
in"   The same mechanism using =93pi'e=94 can be used to express numbers wh=
ich have a base larger than 16. For example, base-20 Mayan mathematics migh=
t use digits from =93no=94 to =93paso=94, each separated by =93pi'e=94:
+<p>
+<pre><a name=3De10d7>10.7)	li pa pi'e re pi'e ci ju'u reno du li vovoci
+	The-number 1;2;3 base 20 equals the-number 443.
+</pre><cx "base greater than 16, compound single-digits contrasted with tw=
o digits">  XE "base greater than 16: compound single-digits contrasted wit=
h two digits"   <cx "base greater than 16, two digits contrasted with compo=
und single-digits">  XE "base greater than 16: two digits contrasted with c=
ompound single-digits"   Carefully note the difference between:
+<p>
+<pre><a name=3De10d8>10.8)	pano ju'u reno
+	the-digit-10 base 20
+</pre>which is equal to ten, and:
+<p>
+<pre><a name=3De10d9>10.9)	pa pi'e no ju'u reno
+	1;0 base 20
+</pre>which is equal to twenty.
+<p>
+<cx "large-base decimal fraction, expressing">  XE "large-base decimal fra=
ction: expressing"   Both =93pi=94 and =93pi'e=94 can be used to express la=
rge-base fractions:
+<p>
+<pre><a name=3De10d10>10.10)	li pa pi'e vo pi ze ju'u reno du li re vo pi =
ci mu
+	The-number 1;4.7 base 20 equals the-number 24.35.
+=93</pre><cx "base, vague">  XE "base: vague"   =93pi'e=94 is also used wh=
ere the base of each digit is vague, as in the numbering of the examples in=
 this chapter:
+<p>
+<ex "sentence 10.11">  XE "sentence 10.11: example=93  <pre><a name=3De10d=
11>10.11)	dei jufra panopi'epapamoi
+	This-utterance is-a-sentence-type-of 10;11th-thing.
+	This is Sentence 10.11.
+
+</pre><a name=3Ds11><h3>Special mekso selbri</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	mei	MOI	cardinal selbri
+	moi	MOI	ordinal selbri
+	si'e	MOI	portion selbri
+	cu'o	MOI	probability selbri
+	va'e	MOI	scale selbri
+	me	ME	make sumti into selbri
+	me'u	MEhU	terminator for ME
+</pre><lx "MOI">  XE "MOI selma=92o"   <cx "numerical selbri, special">  X=
E "numerical selbri: special"   Lojban possesses a special category of selb=
ri which are based on mekso. The simplest kind of such selbri are made by s=
uffixing a member of selma'o MOI to a number. There are five members of MOI=
, each of which serves to create number-based selbri with specific place st=
ructures.
+<p>
+<lx "mei">  XE "mei"   <cx "cardinal selbri, place structure">  XE "cardin=
al selbri: place structure"   <cx "cardinal selbri, definition">  XE "cardi=
nal selbri: definition"   =93The cmavo =93mei=94 creates cardinal selbri. T=
he basic place structure is:
+<p>
+<dl compact><dt><dd>x1 is a mass formed from the set x2 of <math>n</math> =
members, one or more of which is/are x3
+</dl><cx "set, expressing relation with mass formed from set">  XE "set: e=
xpressing relation with mass formed from set"   <cx "set, expressing relati=
on with individuals forming set">  XE "set: expressing relation with indivi=
duals forming set"   <cx "individuals, expressing relation with set formed"=
>  XE "individuals: expressing relation with set formed"   <cx "individuals=
, expressing relation with mass formed">  XE "individuals: expressing relat=
ion with mass formed"   <cx "mass, expressing relation with set forming">  =
XE "mass: expressing relation with set forming"   <cx "mass, expressing rel=
ation with individuals forming">  XE "mass: expressing relation with indivi=
duals forming"   A cardinal selbri interrelates a set with a given number o=
f members, the mass formed from that set, and the individuals which make th=
e set up. The mass argument is placed first as a matter of convenience, not=
 logical necessity.
+<p>
+Some examples:
+<p>
+<ex "three rats">  XE "three rats: example=93  <pre><a name=3De11d1>11.1)	=
lei mi ratcu cu cimei
+	those-I-describe-as-the-mass-of my rats are-a-threesome.
+	My rats are three.
+	I have three rats.
+</pre><p>Here, the mass of my rats is said to have three components; that =
is, I have three rats.
+<p>
+Another example, with one element this time:
+<p>
+<ex "individual">  XE "individual: example=93  <ex "singular me">  XE "sin=
gular me: example=93  <pre><a name=3De11d2>11.2)	mi poi pamei cu cusku dei
+	I who am-an-individual express this-sentence.
+</pre><p>In <a href=3D#e11d2>Example 11.2</a>, =93mi=94 refers to a mass, =
=93the mass consisting of me=94. Personal pronouns are vague between masses=
, sets, and individuals.
+<p>
+However, when the number expressed before =93-mei=94 is an objective indef=
inite number of the kind explained in <a href=3D#s8>Section 8</a>, a slight=
ly different place structure is required:
+<p>
+<cx "mei, place structure formed for objective indefinites">  XE "mei: pla=
ce structure formed for objective indefinites"   <cx "mass, expressing meas=
urement standard for indefinites">  XE "mass: expressing measurement standa=
rd for indefinites"   <cx "set, expressing measurement standard for indefin=
ites">  XE "set: expressing measurement standard for indefinites"   <cx "in=
dividuals of set, expressing measurement standard for indefinites">  XE "in=
dividuals of set: expressing measurement standard for indefinites"   <dl co=
mpact><dt><dd>x1 is a mass formed from a set x2 of <math>n </math>members, =
one or more of which is/are x3,
+measured relative to the set x4.
+</dl><p>An example:
+<p>
+<ex "fewsome">  XE "fewsome: example=93  <ex "rats in park">  XE "rats in =
park: example=93  <pre><a name=3De11d3>11.3)	lei ratcu poi zvati le panka c=
u so'umei fo
+		lo'i ratcu
+	the-mass-of rats which are-in the park are a-fewsome with-respect-to=20
+the-set-of rats.
+	The rats in the park are a small number of all the rats there are.
+</pre><lx "lo'i">  XE "lo'i"   <ex "set of all rats">  XE "set of all rats=
: example=93  <cx "lo'i, with elided quantifiers">  XE "lo'i: with elided q=
uantifiers"   In <a href=3D#e11d3>Example 11.3</a>, the x2 and x3 places ar=
e vacant, and the x4 place is filled by =93lo'i ratcu=94, which (because no=
 quantifiers are explicitly given) means =93the whole of the set of all tho=
se things which are rats=94, or simply =93the set of all rats.=94
+<p>
+<ex "manysome">  XE "manysome: example=93  <pre><a name=3De11d4>11.4)	le'i=
 ratcu poi zvati le panka cu se so'imei
+	The-set-of rats which-are in the park is-a manysome.
+	There are many rats in the park.
+</pre><p>In <a href=3D#e11d4>Example 11.4</a>, the conversion cmavo =93se=
=94 swaps the x1 and the x2 places, so that the new x1 is the set. The x4 s=
et is unspecified, so the implication is that the rats are =93many=94 with =
respect to some unspecified comparison set.
+<p>
+More explanations about the interrelationship of sets, masses, and individ=
uals can be found in <a href=3Dchap6.html>Chapter 6</a>.
+<p>
+<lx "moi">  XE "moi"   <cx "ordinal selbri, place structure">  XE "ordinal=
 selbri: place structure"   <cx "ordinal selbri, definition">  XE "ordinal =
selbri: definition"   =93The cmavo =93moi=94 creates ordinal selbri. The pl=
ace structure is:
+<p>
+<dl compact><dt><dd>x1 is the (n)th member of set x2 when ordered by rule =
x3
+</dl><p>Some examples:
+<p>
+<ex "first rat">  XE "first rat: example=93  <pre><a name=3De11d5>11.5)	ti=
 pamoi le'i mi ratcu
+	This-one is the first-of the rats associated-with me.
+	This is my first rat.
+</pre><ex "all-th">  XE "all-th: example=93  <pre><a name=3De11d6>11.6)	ta=
 romoi le'i mi ratcu
+	That is-the-all-th-of the rats associated-with me.
+	That is my last rat.
+</pre><ex "enough-th">  XE "enough-th: example=93  <pre><a name=3De11d7>11=
.7)	mi raumoi le velskina porsi
+	I am-enough-th-in the movie-audience sequence
+	I am enough-th in the movie line.
+</pre><a href=3D#e11d7>Example 11.7 </a>means, in the appropriate context,=
 that my position in line is sufficiently far to the front that I will get =
a seat for the movie.
+<p>
+<lx "si'e">  XE "si'e"   <cx "portion selbri, definition">  XE "portion se=
lbri: definition"   <cx "portion selbri, place structure">  XE "portion sel=
bri: place structure"   =93The cmavo =93si'e=94 creates portion selbri. The=
 place structure is:
+<p>
+<dl compact><dt><dd>x1 is an (n)th portion of mass x2
+</dl><p>Some examples:
+<p>
+<cx "one-third of food">  XE "one-third of food"   <pre><a name=3De11d8>11=
.8)	levi sanmi cu fi'ucisi'e lei mi djedi cidja
+	This-here meal is-a-slash-three-portion-of my day-food.
+	This meal is one-third of my daily food.
+</pre><lx "cu'o">  XE "cu'o"   <cx "probability selbri, definition">  XE "=
probability selbri: definition"   <cx "probability selbri, place structure"=
>  XE "probability selbri: place structure"   =93The cmavo =93cu'o=94 creat=
es probability selbri. The place structure is:
+<p>
+<dl compact><dt><dd>event x1 has probability (n) of occurring under condit=
ions x2
+</dl><cx "probability selbri, values">  XE "probability selbri: values"   =
The number must be between 0 and 1 inclusive. For example:
+<p>
+<ex "probability .5">  XE "probability .5: example=93  <ex "coin heads">  =
XE "coin heads: example=93  <pre><a name=3De11d9>11.9)	le nu lo sicni cu se=
dja'o cu pimucu'o
+	The event of a coin being a head-displayer has probability .5.
+</pre><lx "va'e">  XE "va'e"   <cx "scale selbri, definition">  XE "scale =
selbri: definition"   <cx "scale selbri, place structure">  XE "scale selbr=
i: place structure"   The cmavo =93va'e=94 creates a scale selbri. The plac=
e structure is:
+<p>
+<dl compact><dt><dd>x1 is at scale position (n) on the scale x2
+</dl><cx "scale, granular contrasted with continuous">  XE "scale: granula=
r contrasted with continuous"   <cx "unreduced fractions, use in granular s=
cales">  XE "unreduced fractions: use in granular scales"   If the scale is=
 granular rather than continuous, a form like =93cifi'uxa=94 (3/6) may be u=
sed; in this case, 3/6 is not the same as 1/2, because the third position o=
n a scale of six positions is not the same as the first position on a scale=
 of two positions . Here is an example:
+<p>
+<ex "8 out of ten">  XE "8 out of ten: example=93  <ex "scale of redness">=
  XE "scale of redness: example=93  <pre><a name=3De11d10>11.10)	le vi rozg=
u cu sofi'upanova'e xunre
+	This rose is 8/10-scale red
+	This rose is 8 out of 10 on the scale of redness.
+	This rose is very red.
+</pre><cx "subjective numbers, effect on place structure for cardinal selb=
ri">  XE "subjective numbers: effect on place structure for cardinal selbri=
"   <cx "subjective numbers, effect on place structure for ordinal selbri">=
  XE "subjective numbers: effect on place structure for ordinal selbri"   <=
cx "subjective numbers, effect on place structure for portion selbri">  XE =
"subjective numbers: effect on place structure for portion selbri"   <cx "s=
ubjective numbers, effect on place structure for probability selbri">  XE "=
subjective numbers: effect on place structure for probability selbri"   <cx=
 "subjective numbers, effect on place structure for scale selbri">  XE "sub=
jective numbers: effect on place structure for scale selbri"   <cx "cardina=
l selbri, place structure effect from subjective numbers">  XE "cardinal se=
lbri: place structure effect from subjective numbers"   <cx "ordinal selbri=
, place structure effect from subjective numbers">  XE "ordinal selbri: pla=
ce structure effect from subjective numbers"   <cx "portion selbri, place s=
tructure effect from subjective numbers">  XE "portion selbri: place struct=
ure effect from subjective numbers"   <cx "probability selbri, place struct=
ure effect from subjective numbers">  XE "probability selbri: place structu=
re effect from subjective numbers"   <cx "scale selbri, place structure eff=
ect from subjective numbers">  XE "scale selbri: place structure effect fro=
m subjective numbers"   <lx "rau">  XE "rau"   <lx "du'e">  XE "du'e"   <lx=
 "mo'a">  XE "mo'a"   When the quantifier preceding any MOI cmavo includes =
the subjective numbers =93rau=94, =93du'e=94, or =93mo'a=94 (enough, too ma=
ny, too few) then an additional place is added for =93by standard=94. For e=
xample:
+<p>
+<cx "too many rats">  XE "too many rats: example"   <pre><a name=3De11d11>=
11.11)	lei ratcu poi zvati le panka cu du'emei fo mi
+	The-mass-of rats which-are in the park are too-many by-standard me.
+	There are too many rats in the park for me.
+</pre><cx "subjective numbers, rationale for effect on place structure">  =
XE "subjective numbers: rationale for effect on place structure"   The extr=
a place (which for =93-mei=94 is the x4 place labeled by =93fo=94) is provi=
ded rather than using a BAI tag such as =93ma'i=94 because a specification =
of the standard for judgment is essential to the meaning of subjective word=
s like =93enough=94.
+<p>
+<cx "standard for subjective numbers, specifying">  XE "standard for subje=
ctive numbers: specifying"   <cx "subjective numbers, specifying standard f=
or">  XE "subjective numbers: specifying standard for"   This place is not =
normally explicit when using one of the subjective numbers directly as a nu=
mber. Therefore, =93du'e ratcu=94 means =93too many rats=94 without specify=
ing any standard.
+<p>
+<cx "numerical selbri, special, with lerfu strings">  XE "numerical selbri=
: special, with lerfu strings"   <cx "lerfu strings, with numerical selbri"=
>  XE "lerfu string: with numerical selbri"   It is also grammatical to sub=
stitute a lerfu string for a number:
+<p>
+<ex "nth rat">  XE "nth rat: example=93  <pre><a name=3De11d12>11.12)	ta n=
y.moi le'i mi ratcu
+	That is-nth-of the-set-of my rats
+	That is my nth rat.
+</pre><cx "numerical selbri, grammar">  XE "numerical selbri: grammar"   <=
cx "numerical selbri, restriction on numbers used for">  XE "numerical selb=
ri: restriction on numbers used for"   More complex mekso cannot be placed =
directly in front of MOI, due to the resulting grammatical ambiguities. Ins=
tead, a somewhat artificial form of expression is required. <p>
+<lx "me">  XE "me"   <lx "ME">  XE "ME selma=92o"   The cmavo =93me=94 (of=
 selma'o ME) has the function of making a sumti into a selbri. </pre><cx "n=
umerical selbri, use of =93me=94 with">  XE "numerical selbri: use of \=93m=
e\=94 with"   <cx "numerical selbri, complex">  XE "numerical selbri: compl=
ex"   <cx "=93me=94, effect of MOI on">  XE "me: effect of MOI on"   <lx "m=
e'u">  XE "me'u"   <cx "numerical selbri, alternative to compensate for res=
triction on numbers">  XE "numerical selbri: alternative to compensate for =
restriction on numbers"   A whole =93me=94 construction can have a member o=
f MOI added to the end to create a complex mekso selbri:
+<p>
+<ex "(n+1)-th rat">  XE "(n+1)-th rat: example=93  <pre><a name=3De11d14>1=
1.13)	ta me li ny. su'i pa me'u moi le'i mi ratcu
+	That is the-number n plus one-th-of the-set-of my rats.
+	That is my (n+1)-th rat.
+</pre><p>Here the mekso =93ny. su'i pa=94 is made into a sumti (with =93li=
=94) and then changed into a mekso selbri with =93me=94 and =93me'u moi=94.=
 The elidable terminator =93me'u=94 is required here in order to keep the =
=93pa=94 and the =93moi=94 separate; otherwise, the parser will combine the=
m into the compound =93pamoi=94 and reject the sentence as ungrammatical.
+<p>
+<cx "numerical selbri, based on non-numerical sumti">  XE "numerical selbr=
i: based on non-numerical sumti"   It is perfectly possible to use non-nume=
rical sumti after =93me=94 and before a member of MOI, producing strange re=
sults indeed:
+<p>
+<ex "snowball's chance">  XE "snowball's chance: example=93  <pre><a name=
=3De11d15>11.14)	le nu mi nolraitru=20
+		cu me le'e snime bolci be vi la xel. cu'o
+	The event-of me being-a-nobly-superlative-ruler
+		has-the-stereotypical snow type-of-ball at Hell probability.
+	I have a snowball's chance in Hell of being king.
+</pre><cx "boi, exception before MOI">  XE "boi: exception before MOI"   <=
cx "PA, exception on use of boi with MOI">  XE "PA selma=92o: exception on =
use of boi with MOI"   <cx "MOI, use of boi before">  XE "MOI selma=92o: us=
e of boi before"   <cx "boi, effect on elidability of me'u">  XE "boi: effe=
ct on elidability of me'u"   <lx "me'u">  XE "me'u"   <lx "boi">  XE "boi" =
  <lx "MOI">  XE "MOI selma=92o"   <lx "PA">  XE "PA selma=92o"   Note: the=
 elidable terminator =93boi=94 is not used between a number and a member of=
 MOI. As a result, the =93me'u=94 in <a href=3D#e11d14>Example 11.13 </a>co=
uld also be replaced by a =93boi=94, which would serve the same function of=
 preventing the =93pa=94 and =93moi=94 from joining into a compound.
+<p>
+<a name=3Ds12><h3>Number questions</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	xo	PA	number question
+</pre><lx "xo">  XE "xo"   <cx "questions, number">  XE "questions: number=
"   <cx "number questions">  XE "number questions"   The cmavo =93xo=94, a =
member of selma'o PA, is used to ask questions whose answers are numbers. L=
ike most Lojban question words, it fills the blank where the answer should =
go. (See <a href=3Dchap19.html>Chapter 19 </a>for more on Lojban questions.=
)
+<p>
+<pre><a name=3De12d1>12.1)	li re su'i re du li xo
+	The-number 2 plus 2 equals the-number what?
+	What is <math>2 + 2</math>?
+</pre><cx "police lineup">  XE "police lineup"   <pre><a name=3De12d2>12.2=
)	le xomoi prenu cu darxi do
+	The what-number-th person hit you?
+	Which person [as in a police lineup] hit you?
+</pre><cx "digit questions">  XE "digit questions"   <cx "questions, digit=
">  XE "questions: digit"   =93=93xo=94 can also be combined with other dig=
its to ask questions whose answers are already partly specified. This abili=
ty could be very useful in writing tests of elementary arithmetical knowled=
ge:
+<p>
+<pre><a name=3De12d3>12.3)	li remu pi'i xa du li paxono
+	The-number 25 times 6 equals the-number 1?0
+</pre><cx "numbers, as grammatically complete utterances">  XE "numbers: a=
s grammatically complete utterances"   <cx "number questions, answers to"> =
 XE "number questions: answers to"   to which the correct reply would be =
=93mu=94, or 5. The ability to utter bare numbers as grammatical Lojban sen=
tences is primarily intended for giving answers to =93xo=94 questions. (Ano=
ther use, obviously, is for counting off physical objects one by one.)
+<p>
+<a name=3Ds13><h3>Subscripts</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	xi	XI	subscript
+</pre><cx =93subscripts, external grammar of=94>  XE =93subscripts: extern=
al grammar of"  Subscripting is a general Lojban feature, not used only in =
mekso; there are many things that can logically be subscripted, and grammat=
ically a subscript is a free modifier, usable almost anywhere. In particula=
r, of course, mekso variables (lerfu strings) can be subscripted:
+<p>
+  XE =93x-sub-3"  <pre><a name=3De13d1>13.1)	li xy.boixici du li xy.boixip=
a su=92i xy.boixire
+	The-number x-sub-3 equals the-number x-sub-1 plus x-sub-2.
+	<math>x<sub>3</sub>=3D x<sub>1</sub>+ x<sub>2</sub></math>
+</pre>  XE "xi"     XE "XI selma=92o"     XE "subscripts: internal grammar=
 of"  =20
+Subscripts always begin with the flag =93xi=94 (of selma'o XI). =93xi=94 m=
ay be followed by a number, a lerfu string, or a general mekso expression i=
n parentheses:
+<p>
+<pre><a name=3De13d2>13.2)	xy.boixino
+	<math>x<sub>0</sub></math>
+<a name=3De13d3>13.3)	xy.boixiny.
+	<math>x<sub>n</sub></math>
+<a name=3De13d4>13.4)	xy.boixi vei ny. su'i pa [ve'o]
+	<math>x<sub>n+1</math>
+</pre><cx "subscripts on lerfu words, effect on elidability of boi">  XE "=
subscripts on lerfu words: effect on elidability of boi"   <cx "subscripts,=
 effects on elidability of terminators">  XE "subscripts: effects on elidab=
ility of terminators"   <cx "free modifiers, effects on elidability of term=
inators">  XE "free modifiers: effects on elidability of terminators"   Not=
e that subscripts attached directly to lerfu words (variables) generally ne=
ed a =93boi=94 terminating the variable. Free modifiers, of which subscript=
s are one variety, generally require the explicit presence of an otherwise =
elidable terminator.
+<p>
+<cx "subscripts, before main expression">  XE "subscripts: before main exp=
ression"   <cx "superscripts">  XE "superscripts"   There is no standard wa=
y of handling superscripts (other than those used as exponents) or for subs=
cripts or superscripts that come before the main expression. If necessary, =
further cmavo could be assigned to selma'o XI for these purposes.
+<p>
+<cx "subscripts, terminator for">  XE "subscripts: terminator for"   <lx "=
boi">  XE "boi"   <cx "subscripts, multiple as sub-subscript">  XE "subscri=
pts: multiple as sub-subscript"   <cx "sub-subscripts">  XE "sub-subscripts=
"   The elidable terminator for a subscript is that for a general number or=
 lerfu string, namely =93boi=94. By convention, a subscript following anoth=
er subscript is taken to be a sub-subscript:
+<p>
+<pre><a name=3De13d5>13.5)	xy.boi xi by.boi xi vo
+	xb4<math></sub></sub></math>
+</pre><p>See <a href=3D#e17d10>Example 17.10 </a>for the standard method o=
f specifying multiple subscripts on a single object.
+<p>
+More information on the uses of subscripts may be found in <a href=3Dchap1=
9.html>Chapter 19</a>.
+<p>
+<a name=3Ds14><h3>Infix operators revisited</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	tu'o	PA	null operand
+	ge'a	VUhU	null operator
+	gei	VUhU	exponential notation
+</pre><lx "tu'o">  XE "tu'o"   <lx "ge'a">  XE "ge'a"   The infix operator=
s presented so far have always had exactly two operands, and for more or fe=
wer operands forethought notation has been required. However, it is possibl=
e to use an operator in infix style even though it has more or fewer than t=
wo operands, through the use of a pair of tricks: the null operand =93tu'o=
=94 and the null operator =93ge'a=94. The first is suitable when there are =
too few operands, the second when there are too many. For example, suppose =
we wanted to express the numerical negation operator =93va'a=94 in infix fo=
rm. We would use:
+<p>
+<pre><a name=3De14d1>14.1)	li tu'o va'a ny. du li no vu'u ny.
+	The-number (null) additive-inverse n equals the-number zero minus n.
+	-<math>-n =3D 0 - n</math>
+</pre><cx "tu'o, for infix operations with too few operands">  XE "tu'o: f=
or infix operations with too few operands"   <cx "null operand, for infix o=
perations with too few operands">  XE "null operand: for infix operations w=
ith too few operands"   <cx "operands, too few for infix operation">  XE "o=
perands: too few for infix operation"   The =93tu'o=94 fulfills the grammat=
ical requirement for a left operand for the infix use of =93va'a=94, even t=
hough semantically none is needed or wanted.
+<p>
+<lx "gei">  XE "gei"   <cx "gei, as a binary operator">  XE "gei: as a bin=
ary operator"   <cx "scientific notation, with gei">  XE "scientific notati=
on: with gei"   <cx "exponential notation, with gei">  XE "exponential nota=
tion: with gei"   Finding a suitable example of =93ge'a=94 requires exhibit=
ing a ternary operator, and ternary operators are not common. The operator =
=93gei=94, however, has both a binary and a ternary use. As a binary operat=
or, it provides a terse representation of scientific (also called =93expone=
ntial=94) notation. The first operand of =93gei=94 is the exponent, and the=
 second operand is the mantissa or fraction:
+
+<p>
+<ex "3 ( 10^8">  XE "3 ( 10^8: example=93  <pre><a name=3De14d2>14.2)	li c=
inonoki'oki'o du
+		li bi gei ci
+	The-number three-zero-zero-comma-comma equals=20
+the-number eight scientific three.
+	<math>300,000,000 =3D 3 ( &times; 10<sup>8</sup></math>
+</pre><cx "scientific notation, rationale for order of places">  XE "scien=
tific notation: rationale for order of places"   <cx "gei, rationale for or=
der of places">  XE "gei: rationale for order of places"   Why are the argu=
ments to =93gei=94 in reverse order from the conventional symbolic notation=
? So that =93gei=94 can be used in forethought to allow easy specification =
of a large (or small) imprecise number:
+<p>
+<ex "10^20">  XE "10^20: example=93  <pre><a name=3De14d3>14.3)	gei reno
+	(scientific) two-zero
+	<math>10<sup>20</sup></math>
+</pre><cx "ge'a, for infix operations with too many operands">  XE "ge'a: =
for infix operations with too many operands"   <cx "null operator, for infi=
x operations with too many operands">  XE "null operator: for infix operati=
ons with too many operands"   <cx "operands, too many for infix operation">=
  XE "operands: too many for infix operation"   <cx "gei, as a ternary oper=
ator">  XE "gei: as a ternary operator"   <cx "exponential notation, with b=
ase other than 10">  XE "exponential notation: with base other than 10"   <=
cx "floating point numbers, expressing">  XE "floating point numbers: expre=
ssing"   Note, however, that although 10 is far and away the most common ex=
ponent base, it is not the only possible one. The third operand of =93gei=
=94, therefore, is the base, with 10 as the default value. Most computers i=
nternally store so-called =93floating-point=94 numbers using 2 as the expon=
ent base. (This has nothing to do with the fact that computers also represe=
nt all integers in base 2; the IBM 360 series used an exponent base of 16 f=
or floating point, although each component of the number was expressed in b=
ase 2.)  Here is a computer floating-point number with a value of 40:
+<p>
+<ex ".1010_2 ( 2^{110_2}">  XE ".1010_2 ( 2^{110_2}: example=93  <pre><a n=
ame=3De14d4>14.4)	papano bi'eju'u re gei pipanopano bi'eju'u re
+		ge'a re
+	(one-one-zero base 2) scientific (point-one-zero-one-zero base 2)=20
+with-base 2
+	<math>.1010 <sub>2</sub>&times;( 2<sup>110<sub>2</sub></sup></math>
+</pre><a name=3Ds15><h3>Vectors and matrices</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	jo'i	JOhI	start vector
+	te'u	TEhU	end vector
+	pi'a	VUhU	matrix row combiner
+	sa'i	VUhU	matrix column combiner
+</pre><cx "vector, definition">  XE "vector: definition"   <cx "matrix, de=
finition">  XE "matrix: definition"   A mathematical vector is a list of nu=
mbers, and a mathematical matrix is a table of numbers. Lojban considers ma=
trices to be built up out of vectors, which are in turn built up out of ope=
rands.
+<p>
+<lx "jo'i">  XE "jo'i"   <lx "JOhI">  XE "JOhI selma=92o"   <cx "vector in=
dicator">  XE "vector indicator"   <cx "jo'i, precedence of">  XE "jo'i: pr=
ecedence of"   <cx "vector, components of">  XE "vector: components of"   <=
lx "te'u">  XE "te'u"   <cx "vector indicator, terminator for">  XE "vector=
 indicator: terminator for"   =93jo'i=94, the only cmavo of selma'o JOhI, i=
s the vector indicator: it has a syntax reminiscent of a forethought operat=
or, but has very high precedence. The components must be simple operands ra=
ther than full expressions (unless parenthesized). A vector can have any nu=
mber of components; =93te'u=94 is the elidable terminator. An example:
+<p>
+<pre><a name=3De15d1>15.1)	li jo'i paboi reboi te'u su'i jo'i ciboi voboi =
du
+		li jo'i voboi xaboi
+	The-number array (one, two) plus array (three, four) equals
+		the-number array (four, six).
+	<math>(1,2) + (3,4) =3D (4,6)</math>
+</pre><lx "pi'a">  XE "pi'a"   <lx "sa'i">  XE "sa'i"   <cx "matrix row op=
erator">  XE "matrix row operator"   <cx "matrix column operator">  XE "mat=
rix column operator"   <cx "matrix, as combination of vectors">  XE "matrix=
: as combination of vectors"   <lx "ge'a">  XE "ge'a"   <cx "matrix, with g=
e'a for more than 2 rows/columns">  XE "matrix: with ge'a for more than 2 r=
ows/columns"   Vectors can be combined into matrices using either =93pi'a=
=94, the matrix row operator, or =93sa'i=94, the matrix column operator. Th=
e first combines vectors representing rows of the matrix, and the second co=
mbines vectors representing columns of the matrix. Both of them allow any n=
umber of arguments: additional arguments are tacked on with the null operat=
or =93ge'a=94.
+<p>
+<ex "magic square">  XE "magic square: example=93  Therefore, the =93magic=
 square=94 matrix
+<p>
+<dl compact><dt><dd>8  1  6
+3  5  7
+4  9  2
+</dl>can be represented either as:
+<p>
+<pre><a name=3De15d2>15.2)	jo'i biboi paboi xa pi'a jo'i ciboi muboi ze ge=
'a jo'i voboi soboi re
+	the-vector (8 1 6) matrix-row the-vector (3 5 7), the-vector (4 9 2)
+</pre>or as
+<p>
+<pre><a name=3De15d3>15.3)	jo'i biboi ciboi vo sa'i jo'i paboi muboi so ge=
'a jo'i xaboi zeboi re
+	the-vector (8 3 4) matrix-column the-vector (1 5 9), the-vector (6 7 2)
+</pre><cx "outer product">  XE "outer product"   <cx "inner product">  XE =
"inner product"   <cx "vectors, use as operands">  XE "vector: use as opera=
nd"   <cx "matrices, use as operands">  XE "matrix: use as operand"   <cx "=
vectors, use of parentheses with">  XE "vector: use of parentheses with"   =
<cx "matrices, use of parentheses with">  XE "matrix: use of parentheses wi=
th"   The regular mekso operators can be applied to vectors and to matrices=
, since grammatically both of these are expressions. It is usually necessar=
y to parenthesize matrices when used with operators in order to avoid incor=
rect groupings. There are no VUhU operators for the matrix operators of inn=
er or outer products, but appropriate operators can be created using a suit=
able symbolic lerfu word or string prefixed by =93ma'o=94.
+<p>
+<cx "matrix, with more than 2 dimensions">  XE "matrix: with more than 2 d=
imensions"   <cx "subscript, to form matrices of more than 2 dimensions">  =
XE =93subscripts: to form matrices of more than 2 dimensions"   <lx "xi">  =
XE "xi"   Matrices of more than two dimensions can be built up using either=
 =93pi'a=94 or =93sa'i=94 with an appropriate subscript numbering the dimen=
sion. When subscripted, there is no difference between =93pi'a=94 and =93sa=
'i=94.
+<p>
+<a name=3Ds16><h3>Reverse Polish notation</h3>
+<p>
+The following cmavo is discussed in this section:
+<p>
+<pre> fu'a	FUhA	reverse Polish flag
+</pre><cx "RP, as abbreviation for reverse Polish notation">  XE "RP (see =
reverse Polish notation)"   So far, the Lojban notational conventions have =
mapped fairly familiar kinds of mathematical discourse. The use of forethou=
ght operators may have seemed odd when applied to =93+=94, but when applied=
 to =93f=94 they appear as the usual functional notation. Now comes a sharp=
 break. Reverse Polish (RP) notation represents something completely differ=
ent; even mathematicians don't use it much. (The only common uses of RP, in=
 fact, are in some kinds of calculators and in the implementation of some p=
rogramming languages.)
+<p>
+<cx "reverse Polish notation, definition">  XE "reverse Polish notation: d=
efinition"   <cx "reverse Polish notation, use of parentheses in">  XE "rev=
erse Polish notation: use of parentheses in"   <cx "reverse Polish notation=
, marker">  XE "reverse Polish notation: marker"   <cx "reverse Polish nota=
tion, terminator">  XE "reverse Polish notation: terminator"   <lx "fu'a"> =
 XE "fu'a"   In RP notation, the operator follows the operands. (Polish not=
ation, where the operator precedes its operands, is another name for foreth=
ought mekso of the kind explained in <a href=3D#s6>Section 6</a>.) The numb=
er of operands per operator is always fixed. No parentheses are required or=
 permitted. In Lojban, RP notation is always explicitly marked by a =93fu'a=
=94 at the beginning of the expression; there is no terminator. Here is a s=
imple example:
+<p>
+<pre><a name=3De16d1>16.1)	li fu'a reboi ci su'i du li mu
+	the-number (RP!) two, three, plus equals the-number five.
+</pre><p>The operands are =93re=94 and =93ci=94; the operator is =93su'i=
=94.
+
+<p>
+Here is a more complex example:
+<p>
+<pre><a name=3De16d2>16.2)	li fu'a reboi ci pi'i voboi mu pi'i su'i du
+		li rexa
+	the-number (RP!) (two, three, times), (four, five, times), plus equals		t=
he-number two-six
+</pre><p>Here the operands of the first =93pi'i=94 are =93re=94 and =93ci=
=94; the operands of the second =93pi'i=94 are =93vo=94 and =93mu=94 (with =
=93boi=94 inserted where needed), and the operands of the =93su'i=94 are =
=93reboi ci pi'i=94, or 6, and =93voboi mu pi'i=94, or 20. As you can see, =
it is easy to get lost in the world of reverse Polish notation; on the othe=
r hand, it is especially easy for a mechanical listener (who has a deep men=
tal stack and doesn't get lost) to comprehend.
+<p>
+<cx "reverse Polish notation, operands of">  XE "reverse Polish notation: =
operands of"   <cx "reverse Polish notation, parentheses in operands of">  =
XE "reverse Polish notation: parentheses in operands of"   The operands of =
an RP operator can be any legal mekso operand, including parenthesized meks=
o that can contain any valid syntax, whether more RP or something more conv=
entional.
+<p>
+<cx "reverse Polish notation, number of operands">  XE "reverse Polish not=
ation: number of operands"   <lx "tu'o">  XE "tu'o"   <lx "ge'a">  XE "ge'a=
"   <cx "reverse Polish notation, with too few operands">  XE "reverse Poli=
sh notation: with too few operands"   <cx "reverse Polish notation, with to=
o many operands">  XE "reverse Polish notation: with too many operands"   <=
lx "tu'o va'a">  XE "tu'o va'a"   <lx "ge'a gei">  XE "ge'a gei"   In Lojba=
n, RP operators are always parsed with exactly two operands. What about ope=
rators which require only one operand, or more than two operands? The null =
operand =93tu'o=94 and the null operator =93ge'a=94 provide a simple soluti=
on. A one-operand operator like =93va'a=94 always appears in a reverse Poli=
sh context as =93tu'o va'a=94. The =93tu'o=94 provides the second operand, =
which is semantically ignored but grammatically necessary. Likewise, the th=
ree-operand version of =93gei=94 appears in reverse Polish as =93ge'a gei=
=94, where the =93ge'a=94 effectively merges the 2nd and 3rd operands into =
a single operand. Here are some examples:
+<p>
+<pre><a name=3De16d3>16.3)	li fu'a ciboi muboi vu'u du
+		li fu'a reboi tu'o va'a
+	The-number (RP!) (three, five, minus) equals=20
+the-number (RP!) two, null, negative-of.
+	<math>3 - 5 =3D -2</math>
+<a name=3De16d4>16.4)	li cinoki'oki'o du
+		li fu'a biboi ciboi panoboi ge'a gei
+	The-number 30-comma-comma equals
+		the-number (RP!) 8, (3, 10, null-op), exponential-notation.
+	<math>30,000,000 =3D 3 ( &times; 10<sup>8</sup></math>
+</pre><a name=3Ds17><h3>Logical and non-logical connectives within mekso</=
h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	.abu	BY	letter =93a=94
+	by	BY	letter =93b=94
+	cy	BY	letter =93c=94
+	fe'a	VUhU	nth root of (default square root)
+	lo'o	LOhO	terminator for LI
+</pre><cx "operand connection, afterthought">  XE "operand connection: aft=
erthought"   <cx "operand connection, forethought">  XE "operand connection=
: forethought"   <cx "operator connection, afterthought">  XE "operator con=
nection: afterthought"   <cx "operator connection, forethought">  XE "opera=
tor connection: forethought"   <cx "forethought connection, of operands">  =
XE "forethought connection: of operands"   <cx "forethought connection, of =
operators">  XE "forethought connection: of operators"   <cx "afterthought =
connection, of operands">  XE "afterthought connection: of operands"   <cx =
"afterthought connection, of operators">  XE "afterthought connection: of o=
perators"   <lx "A">  XE =93A selma=92o"   <lx "GA">  XE "GA selma=92o"   A=
s befits a logical language, Lojban has extensive provision for logical con=
nectives within both operators and operands. Full details on logical and no=
n-logical connectives are provided in <a href=3Dchap14.html>Chapter 14</a>.=
 Operands are connected in afterthought with selma'o A and in forethought w=
ith selma'o GA, just like sumti. Operators are connected in afterthought wi=
th selma'o JA and in forethought with selma'o GUhA, just like tanru compone=
nts. This parallelism is no accident.
+<p>
+<cx "connection of operands, grouping">  XE "connection of operands: group=
ing"   <cx "connection of operators, grouping">  XE "connection of operator=
s: grouping"   <lx "BO">  XE "BO selma=92o"   <lx "KE">  XE "KE selma=92o" =
  In addition, A+BO and A+KE constructs are allowed for grouping logically =
connected operands, and =93ke=A0<dots>=85</dots>ke'e=94 is allowed for grou=
ping logically connected operators, although there are no analogues of tanr=
u among the operators.
+<p>
+Despite the large number of rules required to support this feature, it is =
of relatively minor importance in the mekso scheme of things. <a href=3D#e1=
7d1>Example 17.1 </a>exhibits afterthought logical connection between opera=
nds:
+<p>
+<ex "three or four people">  XE "three or four people: example=93  <pre><a=
 name=3De17d1>17.1)	vei ci .a vo ve'o prenu cu klama le zarci
+	( Three or four ) people go-to the market.
+</pre><a href=3D#e17d2>Example 17.2 </a>is equivalent in meaning, but uses=
 forethought connection:
+<p>
+<pre><a name=3De17d2>17.2)	vei ga ci gi vo ve'o prenu cu klama le zarci
+	( Either 3 or 4 ) people go-to the market.
+</pre><cx "mekso, complex used as quantifier">  XE "mekso: complex used as=
 quantifier"   <cx "parentheses, for complex mekso used as quantifier">  XE=
 "parentheses: for complex mekso used as quantifier"   Note that the mekso =
here are being used as quantifiers. Lojban requires that any mekso other th=
an a simple number be enclosed in parentheses when used as a quantifier. Th=
is rule prevents ambiguities that do not exist when using =93li=94.
+<p>
+<cx "li, terminator for">  XE "li: terminator for"   <lx "lo'o">  XE "lo'o=
"   <cx "logical connection, effect on elidability of lo'o">  XE "logical c=
onnection: effect on elidability of lo'o"   <cx "lo'o, effect of logical co=
nnective on elidability of">  XE "lo'o: effect of logical connective on eli=
dability of"   By the way, =93li=94 has an elidable terminator, =93lo'o=94,=
 which is needed when a =93li=94 sumti is followed by a logical connective =
that could seem to be within the mekso. For example:
+<p>
+<pre><a name=3De17d3>17.3)	li re su'i re du
+		li vo lo'o .onai lo nalseldjuno namcu
+	The-number two plus two equals=20
+		the-number four or else a non-known number.
+</pre><p>Omitting the =93lo'o=94 would cause the parser to assume that ano=
ther operand followed the =93.onai=94 and reject =93lo=94 as an invalid ope=
rand.
+<p>
+Simple examples of logical connection between operators are hard to come b=
y. A contrived example is:
+<p>
+<pre><a name=3De17d4>17.4)	li re su'i je pi'i re du li vo
+	The-number two plus and times two equals the-number four.
+	<math>2 + 2 =3D 4 </math> and <math>2 (&times; 2 =3D 4</math>.
+</pre><p>The forethought-connection form of <a href=3D#e17d4>Example 17.4 =
</a>is:
+<p>
+<pre><a name=3De17d5>17.5)	li re ge su'i gi pi'i re du li vo
+	The-number two both plus and times two equals the-number four.
+	Both <math>2 + 2 =3D 4 </math>and <math>2 (&times; 2 =3D 4</math>.
+</pre><p>Here is a classic example of operand logical connection:
+
+<p>
+<ex "quadratic formula">  XE "quadratic formula: example=93  <cx "Polish n=
otation mixed with infix">  XE "Polish notation mixed with infix"   <cx "in=
fix notation mixed with Polish">  XE "infix notation mixed with Polish"   <=
pre><a name=3De17d6>17.6)	go li .abu bi'epi'i vei xy. te'a re ve'o su'i by.=
 bi'epi'i xy.
+		su'i cy. du li no
+	gi li xy. du li vei va'a by. ku'e su'i ja vu'u
+		fe'a vei by. bi'ete'a re vu'u vo bi'epi'i .abu bi'epi'i cy. ve'o [ku'e] =
ve'o
+		fe'i re bi'epi'i .abu
+	If-and-only-if the-number =93a=94-times-( =93x=94 power two ) plus =93b=
=94-times-=93x=94
+		plus =93c=94 equals the-number zero
+	then the-number x equals the-number [ the-negation-of( b ) plus or minus
+		the-root-of (=93b=94-power-2 minus four-times-=93a=94-times-=93c=94 ) ]
+		divided-by two-times-=93a=94.
+	Iff <math>ax<sup>2</sup>+ bx + c =3D 0</math>,
+=20
+-b &plusmn; <font face=3D"Symbol">&#214;</font><!=96sqrt=96>(b<sup>2</sup>=
- 4ac)</math>
+			=97=97=97=97=97=97=97=96
+</pre><cx "Polish notation mixed with infix, example">  XE "Polish notatio=
n mixed with infix: example"   <cx "infix notation mixed with Polish, examp=
le">  XE "infix notation mixed with Polish: example"   Note the mixture of =
styles in <a href=3D#e17d6>Example 17.6</a>: the negation of b and the squa=
re root are represented by forethought and most of the operator precedence =
by prefixed =93bi'e=94, but explicit parentheses had to be added to group t=
he numerator properly. In addition, the square root parentheses cannot be r=
emoved here in favor of simple =93fe'a=94 and =93ku'e=94 bracketing, becaus=
e infix operators are present in the operand. Getting <a href=3D#e17d6>Exam=
ple 17.6 </a>to parse perfectly using the current parser took several tries=
: a more relaxed style would dispense with most of the =93bi'e=94 cmavo and=
 just let the standard precedence rules be understood.
+<p>
+<lx "JOI">  XE "JOI selma=92o"   <lx "BIhI">  XE "BIhI selma=92o"   <cx "n=
on-logical connection of operands">  XE =93non-logical connection: of opera=
nds"   <cx "non-logical connection of operators">  XE =93non-logical connec=
tion: of operators"   Non-logical connection with JOI and BIhI is also perm=
itted between operands and between operators. One use for this construct is=
 to connect operands with =93bi'o=94 to create intervals:
+<p>
+<pre><a name=3De17d7>17.7)	li no ga'o bi'o ke'i pa
+	the-number zero (inclusive) from-to (exclusive) one
+	<math>[0,1)</math>
+	the numbers from zero to one, including zero but not including one
+</pre><lx "mi'i">  XE "mi'i"   Intervals defined by a midpoint and range r=
ather than beginning and end points can be expressed by =93mi'i=94:
+<p>
+<cx "open interval, expressed with mi'i">  XE "open interval: expressed wi=
th mi'i"   <cx "closed interval, expressed with mi'i">  XE "closed interval=
: expressed with mi'i"   <pre><a name=3De17d8>17.8)	li pimu ga'o mi'i ke'i =
pimu
+	the-number 0.5 (plus-or-minus 0.5
+</pre>which expresses the same interval as <a href=3D#e17d7>Example 17.7</=
a>. Note that the =93ga'o=94 and =93ke'i=94 still refer to the endpoints, a=
lthough these are now implied rather than expressed. Another way of express=
ing the same thing:
+<p>
+<pre><a name=3De17d9>17.9)	li pimu su'i ni'upimu bi'o ma'upimu
+	the-number 0.5 plus [-0.5 from-to +0.5]
+</pre><cx "connection of operands, precedence over operator">  XE "connect=
ion of operands: precedence over operator"   Here we have the sum of a numb=
er and an interval, which produces another interval centered on the number.=
 As <a href=3D#e17d9>Example 17.9 </a>shows, non-logical (or logical) conne=
ction of operands has higher precedence than any mekso operator.
+<p>
+<cx "compound subscript">  XE "compound subscript"   <cx "subscript, multi=
ple for same base word">  XE =93subscripts: multiple for same base word"   =
You can also combine two operands with =93ce'o=94, the sequence connective =
of selma'o JOI, to make a compound subscript:
+<p>
+<ex "x{b,d}">  XE "x{b:d}, example=93  <pre><a name=3De17d10>17.10)	xy. xi=
 vei by. ce'o dy. [ve'o]
+	=93x=94 sub (=93b=94 sequence =93d=94 )
+	<math>x<sub>b,d</math>
+</pre><a name=3Ds18><h3>Using Lojban resources within mekso</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	na'u	NAhU	selbri to operator
+	ni'e	NIhE	selbri to operand
+	mo'e	MOhE	sumti to operand
+	te'u	TEhU	terminator for all three
+</pre><lx "te'u">  XE "te'u"   <cx "na'u, terminator for">  XE "na'u: term=
inator for"   <cx "ni'e, terminator for">  XE "ni'e: terminator for"   <cx =
"mo'e, terminator for">  XE "mo'e: terminator for"   One of the mekso desig=
n goals requires the ability to make use of Lojban's vocabulary resources w=
ithin mekso to extend the built-in cmavo for operands and operators. There =
are three relevant constructs: all three share the elidable terminator =93t=
e'u=94 (which is also used to terminate vectors marked with =93jo'i=94).
+<p>
+<lx "na'u">  XE "na'u"   <cx "selbri, converting into an operator">  XE "s=
elbri: converting into an operator"   <cx "operator, converting selbri into=
">  XE "operator: converting selbri into"   <cx "conversion, selbri into op=
erator">  XE "conversion of selbri into operator"   <cx "selbri place struc=
ture, effect on operator formed by">  XE "selbri place structure: effect on=
 operator formed by"   <cx "operator derived from selbri, effect of selbri =
place structure on">  XE "operator derived from selbri: effect of selbri pl=
ace structure on"   The cmavo =93na'u=94 makes a selbri into an operator. I=
n general, the first place of the selbri specifies the result of the operat=
or, and the other unfilled places specify the operands:
+<p>
+<ex "tan(pi/2) =3D infinity">  XE "tan(pi/2) =3D infinity: example=93  <pr=
e><a name=3De18d1>18.1)	li na'u tanjo te'u vei pai fe'i re [ve'o] du li ci'=
i
+	The-number the-operator tangent ( pi( / 2 ) =3D the-number infinity.
+	<math>tan((pi/2) =3D (<font face=3D"Symbol">&#165;</font><!=96infty=96></=
math>
+</pre>=93tanjo=94 is the gismu for =93x1 is the tangent of x2=94, and the =
=93na'u=94 here makes it into an operator which is then used in forethought=
.
+<p>
+<lx "ni'e">  XE "ni'e"   <cx "selbri, converting into an operand">  XE "se=
lbri: converting into an operand"   <cx "operand, converting selbri into"> =
 XE "operand: converting selbri into"   <cx "conversion, selbri into operan=
d">  XE "conversion of selbri into operand"   <cx "formulae, expressing bas=
ed on pure dimensions">  XE "formulae: expressing based on pure dimensions"=
   The cmavo =93ni'e=94 makes a selbri into an operand. The x1 place of the=
 selbri generally represents a number, and therefore is often a =93ni=94 ab=
straction, since =93ni=94 abstractions represent numbers. The =93ni'e=94 ma=
kes that number available as a mekso operand. A common application is to ma=
ke equations relating pure dimensions:
+<p>
+<ex "Length ( Width ( Depth =3D Volume">  XE "Length ( Width ( Depth =3D V=
olume: example=93  <pre><a name=3De18d2>18.2)	li ni'e ni clani [te'u] pi'i =
ni'e ni ganra [te'u] pi'i
+		ni'e ni condi te'u du li ni'e ni canlu
+	The-number quantity-of length times quantity-of width times
+		quantity-of depth equals the-number quantity-of volume.
+	Length (&times; Width (&times; Depth =3D Volume
+</pre><lx "mo'e">  XE "mo'e"   <cx "sumti, converting into an operand">  X=
E "sumti: converting into an operand"   <cx "operand, converting sumti into=
">  XE "operand: converting sumti into"   <cx "conversion, sumti into opera=
nd">  XE "conversion of sumti into operand"   <cx "dimensioned numbers, exp=
ressing">  XE "dimensioned numbers: expressing"   The cmavo =93mo'e=94 oper=
ates similarly to =93ni'e=94, but makes a sumti (rather than a selbri) into=
 an operand. This construction is useful in stating equations involving dim=
ensioned numbers:
+<p>
+<ex "2 rats + 2 rabbits =3D 4 animals">  XE "2 rats + 2 rabbits =3D 4 anim=
als: example=93  <pre><a name=3De18d3>18.3)	li mo'e re ratcu su'i mo'e re r=
actu du li mo'e vo danlu
+	The-number two rats plus two rabbits equals the-number four animals.
+	2 rats + 2 rabbits =3D 4 animals.
+</pre><cx "folk quantifiers, expressing">  XE "folk quantifiers: expressin=
g"   <ex "pride of lions">  XE "pride of lions: example=93  Another use is =
in constructing Lojbanic versions of so-called =93folk quantifiers=94, such=
 as =93a pride of lions=94:
+<p>
+<pre><a name=3De18d4>18.4)	mi viska vei mo'e lo'e lanzu ve'o cinfo
+	I see ( the-typical family )-number-of lions.
+	I see a pride of lions.
+
+</pre><a name=3Ds19><h3>Other uses of mekso</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	me'o	LI	the mekso
+	nu'a	NUhA	operator to selbri
+	mai	MAI	utterance ordinal
+	mo'o	MAI	higher order utterance ordinal
+	roi	ROI	quantified tense
+</pre><p>So far we have seen mekso used as sumti (with =93li=94), as quant=
ifiers (often parenthesized), and in MOI and ME-MOI selbri. There are a few=
 other minor uses of mekso within Lojban.
+<p>
+<lx "me'o">  XE "me'o"   <cx "me'o, contrasted with li">  XE "me'o: contra=
sted with li"   <cx "li, contrasted with me'o">  XE "li: contrasted with me=
'o"   <cx "mathematical expression, referring to">  XE "mathematical expres=
sion: referring to"   The cmavo =93me'o=94 has the same grammatical use as =
=93li=94 but slightly different semantics. =93li=94 means =93the number whi=
ch is the value of the mekso=A0<dots>=85</dots>=94, whereas =93me'o=94 just=
 means =93the mekso=A0<dots>=85</dots>=94 So it is true that:
+<p>
+<pre><a name=3De19d1>19.1)	li re su'i re du li vo
+	The-number two plus two equals the-number four.
+	<math>2 + 2 =3D 4</math>
+</pre>but false that:
+<p>
+<pre><a name=3De19d2>19.2)	me'o re su'i re du me'o vo
+	The-mekso two plus two equals the-mekso four.
+	=932 + 2=94 =3D =934=94
+</pre><cx "li, relation to me'o compared with la/zo relation">  XE "li: re=
lation to me'o compared with la/zo relation"   <cx "me'o, relation to li co=
mpared with la/zo relation">  XE "me'o: relation to li compared with la/zo =
relation"   since the expressions =932 + 2=94 and =934=94 are not the same.=
 The relationship between =93li=94 and =93me'o=94 is related to that betwee=
n =93la djan.=94, the person named John, and =93zo .djan.=94, the name =93J=
ohn=94.
+<p>
+<lx "nu'a">  XE "nu'a"   <cx "selbri, converting operator into">  XE "selb=
ri: converting operator into"   <cx "operator, converting into selbri">  XE=
 "operator: converting into selbri"   <cx "conversion, operator into selbri=
">  XE "conversion of operator into selbri"   <cx "selbri, place structure =
of converted operator">  XE "selbri: place structure of converted operator"=
   The cmavo =93nu'a=94 is the inverse of =93na'u=94, and allows a mekso op=
erator to be used as a normal selbri, with the place structure:
+<p>
+<dl compact><dt><dd>x1 is the result of applying (operator) to x2, x3,=A0<=
dots>=85</dots>
+</dl>for as many places as may be required. For example:
+<p>
+<pre><a name=3De19d3>19.3)	li ni'umu cu nu'a va'a li ma'umu
+	The-number <math>-5 </math>is-the-negation-of the-number <math>+5</math>.
+</pre>uses =93nu'a=94 to make the operator =93va'a=94 into a two-place bri=
di.
+<p>
+<cx "questions, operator">  XE "questions: operator"   <cx "answers, to op=
erator questions">  XE "answers: to operator questions"   <cx "na'u, use in=
 asking operator questions">  XE "na'u: use in asking operator questions"  =
 <cx "nu'a, use in answering operator questions">  XE "nu'a: use in answeri=
ng operator questions"   Used together, =93nu'a=94 and =93na'u=94 make it p=
ossible to ask questions about mekso operators, even though there is no spe=
cific cmavo for an operator question, nor is it grammatical to utter an ope=
rator in isolation. Consider <a href=3D#e19d4>Example 19.4</a>, to which <a=
 href=3D#e19d5>Example 19.5 </a>is one correct answer:
+<p>
+<pre><a name=3De19d4>19.4)	li re na'u mo re du li vo
+	The-number two what-operator? two equals the-number four
+	<math>2 ? 2 =3D 4</math>
+<a name=3De19d5>19.5)	nu'a su'i
+	plus
+</pre><p>In <a href=3D#e19d4>Example 19.4</a>, =93na'u mo=94 is an operato=
r question, because =93mo=94 is the selbri question cmavo and =93na'u=94 ma=
kes the selbri into an operator. <a href=3D#e19d5>Example 19.5 </a>makes th=
e true answer =93su'i=94 into a selbri (which is a legal utterance) with th=
e inverse cmavo =93nu'a=94. Mechanically speaking, inserting <a href=3D#e19=
d5>Example 19.5 </a>into <a href=3D#e19d4>Example 19.4 </a>produces:
+<p>
+<pre><a name=3De19d6>19.6)	li re na'u nu'a su'i re du li vo
+	The-number two (the-operator the-selbri plus) two equals the-number four.
+</pre>where the =93na'u nu'a=94 cancels out, leaving a truthful bridi.
+<p>
+<lx "mai">  XE "mai"   <lx "MAI">  XE "MAI selma=92o"   <ex "firstly">  XE=
 "firstly: example=93  <cx "text, sub-division numbering with -mai">  XE "t=
ext: sub-division numbering with -mai"   <cx "section numbering">  XE "sect=
ion numbering"   <cx "digit string, definition of">  XE "digit string: defi=
nition of"   Numerical free modifiers, corresponding to English =93firstly=
=94, =93secondly=94, and so on, can be created by suffixing a member of sel=
ma'o MAI to a digit string or a lerfu string. (Digit strings are compound c=
mavo beginning with a cmavo of selma'o PA, and containing only cmavo of PA =
or BY; lerfu strings begin with a cmavo of selma'o BY, and likewise contain=
 only PA or BY cmavo.)  Here are some examples:
+<p>
+<pre><a name=3De19d7>19.7)	pamai
+	firstly
+<a name=3De19d8>19.8)	remai
+	secondly
+<a name=3De19d9>19.9)	romai
+	all-ly
+	lastly
+19.10)	ny.mai
+	nth-ly
+<a name=3De19d10><a name=3De19d11>19.11)	pasomo'o
+	nineteenthly (higher order)
+	<a href=3D#s19>Section 19</a>
+</pre><lx "mo'o">  XE "mo'o"   <lx "MAI">  XE "MAI selma=92o"   <ex "first=
ly">  XE "firstly: example=93  <cx "text, division numbering with -mai">  X=
E "text: division numbering with -mai"   <cx "mai, contrasted with mo'o">  =
XE "mai: contrasted with mo'o"   <cx "mo'o, contrasted with mai">  XE "mo'o=
: contrasted with mai"   <cx "chapter numbering">  XE "chapter numbering"  =
 The difference between =93mai=94 and =93mo'o=94 is that =93mo'o=94 enumera=
tes larger subdivisions of a text. Each =93mo'o=94 subdivision can then be =
divided into pieces and internally numbered with =93mai=94. If this chapter=
 were translated into Lojban, each section would be numbered with =93mo'o=
=94.  (See Chapter 19 for more on these words.)
+<p>
+<cx "tense, numerical">  XE "tense: numerical"   <lx "roi">  XE "roi"   <e=
x "once">  XE "once: example=93  A numerical tense can be created by suffix=
ing a digit string with =93roi=94. This usage generates tenses correspondin=
g to English =93once=94, =93twice=94, and so on. This topic belongs to a de=
tailed discussion of Lojban tenses, and is explained further in <a href=3Dc=
hap10.html>Chapter 10</a>.
+<p>
+<lx "boi">  XE "boi"   <cx "ROI, exception on use of boi before">  XE "ROI=
 selma=92o: exception on use of boi before"   <cx "MAI, exception on use of=
 boi before">  XE "MAI selma=92o: exception on use of boi before"   <cx "bo=
i, exception before MAI">  XE "boi: exception before MAI"   <cx "boi, excep=
tion before ROI">  XE "boi: exception before ROI"   <cx "numerical tenses, =
effect on use of boi">  XE "numerical tenses: effect on use of boi"   Note:=
 the elidable terminator =93boi=94 is not used between a number and a membe=
r of MAI or ROI.
+<p>
+<a name=3Ds20><h3>Explicit operator precedence</h3>
+<p>
+As mentioned earlier, Lojban does provide a way for the precedences of ope=
rators to be explicitly declared, although current parsers do not understan=
d these declarations.
+<p><lx "ti'o">  XE "ti'o"   <lx "SEI">  XE "SEI selma=92o"   The declarati=
on is made in the form of a metalinguistic comment using =93ti'o=94, a memb=
er of selma'o SEI. =93sei=94, the other member of SEI, is used to insert me=
talinguistic comments on a bridi which give information about the discourse=
 which the bridi comprises. The format of a =93ti'o=94 declaration has not =
been formally established, but presumably would take the form of mentioning=
 a mekso operator and then giving it either an absolute numerical precedenc=
e on some pre-established scale, or else specifying relative precedences be=
tween new operators and existing operators.
+<p>
+<cx "operator precedence, plans for future">  XE "operator precedence: pla=
ns for future"   In future, we hope to create an improved machine parser th=
at can understand declarations of the precedences of simple operators belon=
ging to selma'o VUhU. Originally, all operators would have the same precede=
nce. Declarations would have the effect of raising the specified cmavo of V=
UhU to higher precedence levels. Complex operators formed with =93na'u=94, =
=93ni'e=94, or =93ma'o=94 would remain at the standard low precedence; decl=
arations with respect to them are for future implementation efforts. It is =
probable that such a parser would have a set of =93commonly assumed precede=
nces=94 built into it (selectable by a special =93ti'o=94 declaration) that=
 would match mathematical intuition: times higher than plus, and so on.
+<p>
+<a name=3Ds21><h3>Miscellany</h3>
+<p>
+A few other points:
+<p>
+<cx "conversion, of operator places">  XE "conversion: of operator places"=
   <cx "se, use with operators">  XE "se: use with operators"   <lx "se">  =
XE "se"   =93se=94 can be used to convert an operator as if it were a selbr=
i, so that its arguments are exchanged. For example:
+<p>
+<pre><a name=3De21d1>21.1)	li ci se vu'u vo du li pa
+	The-number three (inverse) minus four equals the-number one.
+	3 subtracted from 4 equals 1.
+</pre><lx "SE">  XE "SE selma=92o"   The other converters of selma'o SE ca=
n also be used on operators with more than two operands, and they can be co=
mpounded to create (probably unintelligible) operators as needed.
+<p>
+<lx "NAhE">  XE "NAhE selma=92o"   <cx "negation, of operator">  XE "negat=
ion: of operator"   Members of selma'o NAhE are also legal on an operator t=
o produce a scalar negation of it. The implication is that some other opera=
tor would apply to make the bridi true:
+<p>
+<pre><a name=3De21d2>21.2)	li ci na'e su'i vo du li pare
+	The-number 3 non-plus 4 equals the-number 12.
+</pre><ex "opposite-of-minus">  XE "opposite-of-minus: example=93  <pre><a=
 name=3De21d3>21.3)	li ci to'e vu'u re du li mu
+	The-number 3 opposite-of-minus 2 equals the-number 5.
+</pre><p>The sense in which =93plus=94 is the opposite of =93minus=94 is n=
ot a mathematical but rather a linguistic one; negated operators are define=
d only loosely.
+<p>
+<lx "la'e">  XE "la'e"   <lx "lu'e">  XE "lu'e"   <cx "referent, of operan=
d">  XE "referent: of operand"   <cx "symbol, for operand">  XE "symbol: fo=
r operand"   <lx "bo">  XE "bo"   <cx "negation, of operand">  XE "negation=
: of operand"   =93=93la'e=94 and =93lu'e=94 can be used on operands with t=
he usual semantics to get the referent of or a symbol for an operand. Likew=
ise, a member of selma'o NAhE followed by =93bo=94 serves to scalar-negate =
an operand, implying that some other operand would make the bridi true:
+<p>
+<pre><a name=3De21d4>21.4)	li re su'i re du li na'ebo mu
+	The-number 2 plus 2 equals the-number non-5.
+	<math>2 + 2 =3D </math> something other than 5.
+</pre><cx "digits, rafsi for">  XE "digits: rafsi for"   <cx "digits, name=
s from">  XE "digits: names from"   The digits 0-9 have rafsi, and therefor=
e can be used in making lujvo. Additionally, all the rafsi have CVC form an=
d can stand alone or together as names:
+<p>
+<pre><a name=3De21d5>21.5)	la zel. poi gunta la tebes. pu nanmu
+	Those-named =93Seven=94 who attack that-named =93Thebes=94 [past] are-men=
.
+	The Seven Against Thebes were men.
+</pre><p>Of course, there is no guarantee that the name =93zel.=94 is conn=
ected with the number rafsi: an alternative which cannot be misconstrued is=
:
+<p>
+<pre><a name=3De21d6>21.6)	la zemei poi gunta la tebes. pu nanmu
+	Those-named-the Sevensome who attack Thebes [past] are-men.
+</pre><cx "PA, members with rafsi">  XE "PA selma=92o: members with rafsi"=
   <cx "rafsi, conventional meaning for frinu">  XE "rafsi: conventional me=
aning for frinu"   Certain other members of PA also have assigned rafsi: =
=93so'a=94, =93so'e=94, =93so'i=94, =93so'o=94, =93so'u=94, =93da'a=94, =93=
ro=94, =93su'e=94, =93su'o=94, =93pi=94, and =93ce'i=94. Furthermore, altho=
ugh the cmavo =93fi'u=94 does not have a rafsi as such, it is closely relat=
ed to the gismu =93frinu=94, meaning =93fraction=94; therefore, in a contex=
t of numeric rafsi, you can use any of the rafsi for =93frinu=94 to indicat=
e a fraction slash.
+<p>
+<cx "rafsi, conventional meaning for cu'o">  XE "rafsi: conventional meani=
ng for cu'o"   A similar convention is used for the cmavo =93cu'o=94 of sel=
ma'o MOI, which is closely related to =93cunso=94 (probability); use a rafs=
i for =93cunso=94 in order to create lujvo based on =93cu'o=94. The cmavo =
=93mei=94 and =93moi=94 of MOI have their own rafsi, two each in fact: =93m=
em=94/=93mei=94 and =93mom=94/=93moi=94 respectively.
+<p>
+<cx "lambda calculus, operator and operand distinction in">  XE "lambda ca=
lculus: operator and operand distinction in"   <cx "algebra of functions, o=
perator and operand distinction in">  XE "algebra of functions: operator an=
d operand distinction in"   <lx "ni'enu'a">  XE "ni'enu'a"   <cx "conversio=
n, operator into operand">  XE "conversion of operator into operand"   <cx =
"operator, converting into operand">  XE "operator: converting into operand=
"   <cx "operand, converting from operator">  XE "operand: converting from =
operator"   The grammar of mekso as described so far imposes a rigid distin=
ction between operators and operands. Some flavors of mathematics (lambda c=
alculus, algebra of functions) blur this distinction, and Lojban must have =
a method of doing the same. An operator can be changed into an operand with=
 =93ni'enu'a=94, which transforms the operator into a matching selbri and t=
hen the selbri into an operand.
+<p>
+<cx "conversion, operand into operator">  XE "conversion of operand into o=
perator"   <cx "operand, converting into operator">  XE "operand: convertin=
g into operator"   <cx "operator, converting from operand">  XE "operator: =
converting from operand"   <lx "ma'o">  XE "ma'o"   <lx "te'u">  XE "te'u" =
  To change an operand into an operator, we use the cmavo =93ma'o=94, alrea=
dy introduced as a means of changing a lerfu string such as =93fy.=94 into =
an operator. In fact, =93ma'o=94 can be followed by any mekso operand, usin=
g the elidable terminator =93te'u=94 if necessary.
+<p>
+<cx "ma'o, potential ambiguity caveat">  XE "ma'o: potential ambiguity cav=
eat"   There is a potential semantic ambiguity in =93ma'o fy. [te'u]=94 if =
=93fy.=94 is already in use as a variable: it comes to mean =93the function=
 whose value is always 'f=92=94. However, mathematicians do not normally us=
e the same lerfu words or strings as both functions and variables, so this =
case should not arise in practice.
+<p>
+<a name=3Ds22><h3>Four score and seven: a mekso problem</h3>
+<p>
+<ex "Gettysburg Address">  XE "Gettysburg Address: example=93  <ex "Four s=
core and seven">  XE "Four score and seven: example=93  <cx "mekso, and lit=
erary translation">  XE "mekso: and literary translation"   Abraham Lincoln=
's Gettysburg Address begins with the words =93Four score and seven years a=
go=94. This section exhibits several different ways of saying the number =
=93four score and seven=94. (A =93score=94, for those not familiar with the=
 term, is 20; it is analogous to a =93dozen=94 for 12.) The trivial way:
+<p>
+<pre><a name=3De22d1>22.1)	bize
+	eight seven
+	87
+</pre><a href=3D#e22d1>Example 22.1 </a>is mathematically correct, but sac=
rifices the spirit of the English words, which are intended to be complex a=
nd formal.
+<p>
+<pre><a name=3De22d2>22.2)	vo pi'i reno su'i ze
+	four times twenty plus seven
+	<math>4 &times;( 20 + 7</math>
+</pre><cx "score, as 20-year span">  XE "score: as 20-year span"   <a href=
=3D#e22d2>Example 22.2 </a>is also mathematically correct, but still misses=
 something. =93Score=94 is not a word for 20 in the same way that =93ten=94=
 is a word for 10: it contains the implication of 20 objects. The original =
may be taken as short for =93Four score years and seven years ago=94. Think=
ing of a score as a twentysome rather than as 20 leads to:
+<p>
+<pre><a name=3De22d3>22.3)	mo'e voboi renomei su'i ze
+	the-number-of four twentysomes plus seven
+</pre><p>In <a href=3D#e22d3>Example 22.3</a>, =93voboi renomei=94 is a su=
mti signifying four things each of which are groups of twenty; the =93mo'e=
=94 and =93te'u=94 then make this sumti into a number in order to allow it =
to be the operand of =93su'i=94.
+<p>
+<cx "score, as alternate base for years">  XE "score: as alternate base fo=
r years"   <cx "base-20 arithmetic, remnants of">  XE "base-20 arithmetic: =
remnants of"   Another approach is to think of =93score=94 as setting a rep=
resentation base. There are remnants of base-20 arithmetic in some language=
s, notably French, in which 87 is =93quatre-vingt-sept=94, literally =93fou=
r-twenties-seven=94. (This fact makes the Gettysburg Address hard to transl=
ate into French!) If =93score=94 is the representation base, then we have:
+<p>
+<pre><a name=3De22d4>22.4)	vo pi'e ze ju'u reno
+	four ; seven base 20
+	<math>47<sub>20</sub></math>
+</pre><p>Overall, <a href=3D#e22d3>Example 22.3 </a>probably captures the =
flavor of the English best. <a href=3D#e22d1>Example 22.1 </a>and <a href=
=3D#e22d2>Example 22.2 </a>are too simple, and <a href=3D#e22d4>Example 22.=
4 </a>is too tricky. Nevertheless, all four examples are good Lojban. Pedag=
ogically, these examples illustrate the richness of lojbau mekso: anything =
that can be said at all, can probably be said in more than one way.
+<p>
+<p>
+<a name=3Ds23><h3>mekso selma'o summary</h3>
+<p>
+Except as noted, each selma'o has only one cmavo.
+<p>
+<cx "mekso, list of selma'o for">  XE "mekso: list of selma'o for"   <pre>=
 	BOI	elidable terminator for numerals and lerfu strings
+	BY	lerfu for variables and functions (see <a href=3Dchap17.html>Chapter 1=
7</a>)
+	FUhA	reverse-Polish flag
+	GOhA	includes =93du=94 (mathematical equality) and other non-mekso cmavo
+	JOhI	array flag
+	KUhE	elidable terminator for forethought mekso
+	LI	mekso articles (li and me'o)
+	MAhO	make operand into operator
+	MOI	creates mekso selbri (moi, mei, si'e, and cu'o, see <a href=3D#s11>Se=
ction 11</a>)
+	MOhE	make sumti into operand
+	NAhU	make selbri into operator
+	NIhE	make selbri into operand
+	NUhA	make operator into selbri
+	PA	numbers (see <a href=3D#s25>Section 25</a>)
+	PEhO	optional forethought mekso marker
+	TEhU	elidable terminator for NAhU, NIhE, MOhE, MAhO, and JOhI
+	VEI	left parenthesis
+	VEhO	right parenthesis
+	VUhU	operators (see <a href=3D#s24>Section 24</a>)
+	XI	subscript flag
+</pre><a name=3Ds24><h3>Complete table of VUhU cmavo, with operand structu=
res</h3>
+<p>
+The operand structures specify what various operands (labeled <math>a</mat=
h>, <math>b</math>, <math>c</math>,=A0<dots>=85</dots>) mean. The implied c=
ontext is forethought, since only forethought operators can have a variable=
 number of operands; however, the same rules apply to infix and RP uses of =
VUhU.
+<p>
+<cx "operators, list of simple">  XE "operators: list of simple"   <dl com=
pact><dt>su'i	<dd>plus			<math>(((a + b) + c) +=A0<dots>=85</dots>)</math>
+pi'i	times			<math>(((a (&times; b) (&times; c) (&times;=A0<dots>=85</dots=
>)</math>
+vu'u	minus			<math>(((a - b) - c) -=A0<dots>=85</dots>)</math>
+fe'i	divided by		<math>(((a / b) / c) /=A0<dots>=85</dots>)</math>
+ju'u	number base		numeral string =93a=94 interpreted in the base b=20
+pa'i	ratio			the ratio of a to b, <math>a:b</math>
+fa'i	reciprocal/multiplicative inverse	<math>1 / a</math>
+gei	scientific notation	<math>b (&times; (c </math>[default 10] to the <ma=
th>a </math>power<math>)</math>
+ge'a	null operator 		(no operands)=20
+de'o	logarithm		log <i>a </i>to base <i>b </i>(default 10 or <i>e</i> as
+					appropriate)=20
+te'a	to the power/exponential	a to the b power=20
+fe'a	nth root/inverse power	bth root of a (default square root:  b =3D 2)=
=20
+cu'a	absolute value/norm	<math>| a |</math>
+ne'o	factorial		<math>a!</math>
+pi'a	matrix row vector combiner	(all operands are row vectors)=20
+sa'i	matrix column vector combiner	(all operands are column vectors)=20
+ri'o	integral		integral of a with respect to b over range c=20
+sa'o	derivative		derivative of a with respect to b of degree c 						(defa=
ult 1)=20
+fu'u	non-specific operator	(variable)=20
+si'i	sigma (() summation	summation of a using variable b over range c=20
+va'a	negation of/additive inverse	<math>-a</math>
+re'a	matrix transpose/dual	<math>a*</math>
+</dl><a name=3Ds25><h3>Complete table of PA cmavo: digits, punctuation, an=
d other numbers.</h3>
+<p>
+<cx "digits, list of decimal">  XE "digits: list of decimal"   Decimal dig=
its:
+<p>
+<dl compact><dt>	<dd>no, pa, re, ci, vo, mu, xa, ze, bi, so
+	0, 1, 2, 3, 4, 5, 6, 7, 8, 9
+	rafsi:  non, pav, rel, cib, von, mum, xav, zel, biv, soz
+</dl><cx "digits, list of hexadecimal">  XE "digits: list of hexadecimal" =
  Hexadecimal digits:
+<p>
+<dl compact><dt>	<dd>dau, fei, gai, jau, rei, vai
+	A/10, B/11, C/12, D/13, E/14, F/15
+</dl><cx "numbers, list of special">  XE "numbers: list of special"   Spec=
ial numbers:
+<p>
+<dl compact><dt>	<dd>pai, ka'o, te'o, ci'i
+	(, imaginary i, exponential e, infinity (()
+</dl><cx "punctuation, list of numerical">  XE "punctuation: list of numer=
ical"   Number punctuation:
+<p>
+<dl compact><dt>	<dd>pi, ce'i, fi'u
+	decimal point, percentage, fraction (not division)
+	rafsi:  piz, cez, fi'u (from frinu; see <a href=3D#s20>Section 20</a>)
+<p><dt>	<dd>pi'e, ma'u, ni'u
+	mixed-base point, plus sign (not addition), minus sign (not subtraction)
+<p><dt>	<dd>ki'o, ra'e
+	thousands comma, repeating-decimal indicator
+<p><dt>	<dd>ji'i, ka'o
+	approximation sign, complex number separator
+
+</dl><cx "numbers, list of indefinite">  XE "numbers: list of indefinite" =
  Indefinite numbers:
+<p>
+<dl compact><dt>	<dd>ro, so'a, so'e, so'i, so'o, so'u, da'a
+	all, almost all, most, many, several, few, all but
+	rafsi:  rol, soj, sor or so'i, sos, sot, daz
+<p><dt>	<dd>su'e, su'o
+	at most, at least
+	rafsi:  su'e, su'o
+<p><dt>	<dd>me'i, za'u
+	less than, more than
+<p><dt>	<dd>no'o
+	the typical number
+</dl><p>Subjective numbers:
+<p>
+<dl compact><dt>	<dd>rau, du'e, mo'a
+	enough, too many, too few
+</dl><p>Miscellaneous:
+<p>
+<dl compact><dt>	<dd>xo, tu'o
+	number question, null operand
+</dl><a name=3Ds26><h3>Table of MOI cmavo, with associated rafsi and place=
 structures</h3>
+<p>
+<lx "MOI">  XE "MOI selma=92o"   <cx "MOI, list of cmavo in">  XE "MOI sel=
ma=92o: list of cmavo in"   <pre> 	mei	x1 is a mass formed from a set x2 of=
 <math>n </math>members, one or more of=20
+which is/are x3, [measured relative to the set x4/by standard x4]
+	rafsi:  mem, mei
+	moi	x1 is the (n)th member of set x2 when ordered by rule x3=20
+[by standard x4]
+	rafsi:  mom, moi
+	si'e	x1 is an (n)th portion of mass x2 [by standard x3]
+	rafsi:  none
+	cu'o	event x1 has probability (n) of occurring under conditions x2
+[by standard x3]
+	rafsi:  cu'o (borrowed from cunso; see <a href=3D#s20>Section 20</a>)
+	va=92e	x1 is at scale position (n) on the scale x2
+[by standard x3]
+	rafsi:  none
+
+</body></html>
+<center><p align=3Dcenter>
+<img src=3Dchap19.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+
+<h2>Chapter 19
+<br>
+Putting It All Together: Notes on the Structure of Lojban Texts</h2>
+<h6>$Revision: 4.1 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+Introductory</h3>
+<p><cx "text, structure of">  XE "text: structure of"   <a name=3Ds1><h3>
+This chapter is incurably miscellaneous. It describes the cmavo that speci=
fy the structure of Lojban texts, from the largest scale (paragraphs) to th=
e smallest (single words). There are fewer examples than are found in other=
 chapters of this book, since the linguistic mechanisms described are gener=
ally made use of in conversation or else in long documents.
+<p>
+This chapter is also not very self-contained. It makes passing reference t=
o a great many concepts which are explained in full only in other chapters.=
 The alternative would be a chapter on text structure which was as complex =
as all the other chapters put together. Lojban is a unified language, and i=
t is not possible to understand any part of it (in full) before understandi=
ng every part of it (to some degree).
+<p>
+Sentences: I</h3>
+<p><cx "sentences, joining">  XE "sentences: joining"   <a name=3Ds2><h3>
+The following cmavo is discussed in this section:
+<p>
+<lx ".i">  XE "i"   <lx "I">  XE "I selma=92o"   <pre>	.i	I	sentence separ=
ator
+</pre><cx "audio-visually isomorphic">  XE "audio-visually isomorphic"   S=
ince Lojban is audio-visually isomorphic, there needs to be a spoken and wr=
itten way of signaling the end of a sentence and the start of the following=
 one. In written English, a period serves this purpose; in spoken English, =
a tone contour (rising or falling) usually does the job, or sometimes a lon=
g pause. Lojban uses a single separator: the cmavo =93.i=94 (of selma'o I):
+<p>
+<pre><a name=3De2d1>2.1)	mi klama le zarci  .i do cadzu le bisli
+	I go to-the store.  You walk on-the ice.
+</pre><cx "sentences, separator for joining">  XE "sentences: separator fo=
r joining"   The word =93separator=94 should be noted. =93.i=94 is not norm=
ally used after the last sentence nor before the first one, although both p=
ositions are technically grammatical. =93.i=94 signals a new sentence on th=
e same topic, not necessarily by the same speaker. The relationship between=
 the sentences is left vague, except in stories, where the relationship usu=
ally is temporal, and the following sentence states something that happened=
 after the previous sentence.
+<p>
+Note that although the first letter of an English sentence is capitalized,=
 the cmavo =93.i=94 is never capitalized. In writing, it is appropriate to =
place extra space before =93.i=94 to make it stand out better for the reade=
r. In some styles of Lojban writing, every =93.i=94 is placed at the beginn=
ing of a line, possibly leaving space at the end of the previous line.
+<p>
+An =93.i=94 cmavo may or may not be used when the speaker of the following=
 sentence is different from the speaker of the preceding sentence, dependin=
g on whether the sentences are felt to be connected or not.
+<p>
+An =93.i=94 cmavo can be compounded with a logical or non-logical connecti=
ve (a jek or joik), a modal or tense connective, or both: these constructs =
are explained in <a href=3Dchap9.html>Chapter 9</a>, <a href=3Dchap10.html>=
Chapter 10</a>, and <a href=3Dchap14.html>Chapter 14</a>. In all cases, the=
 =93.i=94 comes first in the compound. Attitudinals can also be attached to=
 an =93.i=94 if they are meant to apply to the whole sentence: see <a href=
=3Dchap13.html>Chapter 13</a>.
+<p>
+<lx "bo">  XE "bo"   <lx "BO">  XE "BO selma=92o"   <cx "sentences, close =
grouping">  XE "sentences: close grouping"   There exist a pair of mechanis=
ms for binding a sequence of sentences closely together. If the =93.i=94 (w=
ith or without connectives) is followed by =93bo=94 (of selma'o BO), then t=
he two sentences being separated are understood to be more closely grouped =
than sentences connected by =93.i=94 alone.
+<p>
+<lx "tu'e">  XE "tu'e"   <lx "tu'u">  XE "tu'u"   <lx "TUhE">  XE "TUhE se=
lma=92o"   <lx "TUhU">  XE "TUhU selma=92o"   <cx "title, specifying with t=
u'e=85tu'u">  XE "title: specifying with tu'e=85tu'u"   Similarly, a group =
of sentences can be preceded by =93tu'e=94 (of selma'o TUhE) and followed b=
y =93tu'u=94 (of selma'o TUhU) to fuse them into a single unit. A common us=
e of =93tu'e <dots>=85</dots>tu'u=94 is to group the sentences which compos=
e a poem: the title sentence would precede the group, separated from it by =
=93.i=94. Another use might be a set of directions, where each numbered dir=
ection might be surrounded by =93tu'e=A0<dots>=85</dots>tu'u=94 and contain=
 one or more sentences separated by =93.i=94. Grouping with =93tu'e=94 and =
=93tu'u=94 is analogous to grouping with =93ke=94 and =93ke'e=94 to establi=
sh the scope of logical or non-logical connectives (see <a href=3Dchap14.ht=
ml>Chapter 14</a>).
+<p>
+Paragraphs: NIhO</h3>
+<p><cx "paragraphs, separating">  XE "paragraphs: separating"   <a name=3D=
s3><h3>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ni'o	NIhO	new topic
+	no'i	NIhO	old topic
+	da'o	DAhO	cancel cmavo assignments
+</pre><p>The paragraph is a concept used in writing systems for two purpos=
es: to indicate changes of topic, and to break up the hard-to-read appearan=
ce of large blocks of text on the page. The former function is represented =
in both spoken and written Lojban by the cmavo =93ni'o=94 and =93no'i=94, b=
oth of selma'o NIhO. Of these two, =93ni'o=94 is the more common. By conven=
tion, written Lojban is broken into paragraphs just before any =93ni'o=94 o=
r =93no'i=94, but a very long passage on a single topic might be paragraphe=
d before an =93.i=94. On the other hand, it is conventional in English to s=
tart a new paragraph in dialogue when a new speaker starts, but this conven=
tion is not commonly observed in Lojban dialogues. Of course, none of these=
 conventions affect meaning in any way.
+<p>
+<lx "ni'o">  XE "ni'o"   <lx "NIhO">  XE "NIhO selma=92o"   <cx "old topic=
">  XE "old topic"   <cx "paragraphs, separator">  XE "paragraphs: separato=
r"   A =93ni'o=94 can take the place of an =93.i=94 as a sentence separator=
, and in addition signals a new topic or paragraph. Grammatically, any numb=
er of =93ni'o=94 cmavo can appear consecutively and are equivalent to a sin=
gle one; semantically, a greater number of =93ni'o=94 cmavo indicates a lar=
ger-scale change of topic. This feature allows complexly structured text, w=
ith topics, subtopics, and sub-subtopics, to be represented clearly and una=
mbiguously in both spoken and written Lojban. However, some conventional di=
fferences do exist between =93ni'o=94 in writing and in conversation.
+<p>
+<cx "indicator scope">  XE "indicator scope"   <cx "discursive indicator">=
  XE "discursive indicator"   <lx "da'o">  XE "da'o"   <lx "DAhO">  XE "DAh=
O selma=92o"   <cx "paragraphs, effects on scope">  XE "paragraphs: effects=
 on scope"   <cx "pro-bridi, scope effect of new paragraph">  XE "pro-bridi=
: scope effect of new paragraph"   <cx "pro-sumti, scope effect of new para=
graph">  XE "pro-sumti: scope effect of new paragraph"   <cx "indicators, s=
cope effect of new paragraph">  XE "indicators: scope effect of new paragra=
ph"   <cx "tense, scope effect of new paragraph">  XE "tense: scope effect =
of new paragraph"   <cx "paragraph separation, written text">  XE "paragrap=
h separation: written text"   In written text, a single =93ni'o=94 is a mer=
e discursive indicator of a new subject, whereas =93ni'oni'o=94 marks a cha=
nge in the context. In this situation, =93ni'oni'o=94 implicitly cancels th=
e definitions of all pro-sumti of selma'o KOhA as well as pro-bridi of selm=
a'o GOhA. (Explicit cancelling is expressed by the cmavo =93da'o=94 of selm=
a'o DAhO, which has the free grammar of an indicator =96 it can appear almo=
st anywhere.) The use of =93ni'oni'o=94 does not affect indicators (of selm=
a'o UI) or tense references, but =93ni'oni'oni'o=94, indicating a drastic c=
hange of topic, would serve to reset both indicators and tenses. (See <a hr=
ef=3D#s8>Section 8 </a>for a discussion of indicator scope.)
+<p>
+<cx "Arabian Nights">  XE "Arabian Nights: as multi-level narrative"   <cx=
 "paragraph separation, spoken text">  XE "paragraph separation: spoken tex=
t"   In spoken text, which is inherently less structured, these levels are =
reduced by one, with =93ni'o=94 indicating a change in context sufficient t=
o cancel pro-sumti and pro-bridi assignment. On the other hand, in a book, =
or in stories within stories such as =93The Arabian Nights=94, further leve=
ls may be expressed by extending the =93ni'o=94 string as needed. Normally,=
 a written text will begin with the number of =93ni'o=94 cmavo needed to si=
gnal the largest scale division which the text contains. =93ni'o=94 strings=
 may be subscripted to label each context of discourse: see <a href=3D#s6>S=
ection 6</a>.
+<p>
+<cx "previous topic">  XE "previous topic"   <lx "no'i">  XE "no'i"   <lx =
"NIhO">  XE "NIhO selma=92o"   =93no'i=94 is similar in effect to =93ni'o=
=94, but indicates the resumption of a previous topic. In speech, it is ana=
logous to (but much shorter than) such English discursive phrases as =93But=
 getting back to the point=A0<dots>=85</dots>=94. By default, the topic res=
umed is that in effect before the last =93ni'o=94. When subtopics are neste=
d within topics, then =93no'i=94 would resume the previous subtopic and =93=
no'ino'i=94 the previous topic. Note that =93no'i=94 also resumes tense and=
 pro-sumti assignments dropped at the previous =93ni'o=94.
+<p>
+<cx "subscripted topics">  XE "subscripted topics"   If a =93ni'o=94 is su=
bscripted, then a =93no'i=94 with the same subscript is assumed to be a con=
tinuation of it. A =93no'i=94 may also have a negative subscript, which wou=
ld specify counting backwards a number of paragraphs and resuming the topic=
 found thereby.
+<p>
+Topic-comment sentences: ZOhU</h3>
+<p><cx "topic-comment sentences">  XE "topic-comment sentences"   <a name=
=3Ds4><h3>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	zo'u	ZOhU	topic/comment separator
+</pre><p>The normal Lojban sentence is just a bridi, parallel to the norma=
l English sentence which has a subject and a predicate:
+<p>
+<pre><a name=3De4d1>4.1)	mi klama le zarci
+	I went to the market.
+</pre><cx "topic-comment, description">  XE "topic-comment: description"  =
 In Chinese, the normal sentence form is different: a topic is stated, and =
a comment about it is made. (Japanese also has the concept of a topic, but =
indicates it by attaching a suffix; other languages also distinguish topics=
 in various ways.) The topic says what the sentence is about:
+<p>
+<ex "news">  XE "news: example=93  <pre><a name=3De4d2>4.2)	zhe<sup>4</sup=
> xiao<sup>1</sup>xi<sup>2	</sup>wo<sup>3</sup> zhi<sup>1</sup>dao le
+	this news	I know [perfective]
+	As for this news, I knew it.
+	I've heard this news already.
+</pre><lx "zo'u">  XE "zo'u"   <lx "ZOhU">  XE "ZOhU selma=92o"   The wide=
 space in the first two versions of <a href=3D#e4d2>Example 4.2 </a>separat=
e the topic (=93this news=94) from the comment (=93I know already=94).
+<p>
+Lojban uses the cmavo =93zo'u=94 (of selma'o ZOhU) to separate topic (a su=
mti) from comment (a bridi):
+<p>
+<pre><a name=3De4d3>4.3)	le nuzba zu'o mi ba'o djuno
+	The news :  I [perfective] know.
+</pre><a href=3D#e4d3>Example 4.3 </a>is the literal Lojban translation of=
 <a href=3D#e4d2>Example 4.2</a>. Of course, the topic-comment structure ca=
n be changed to a straightforward bridi structure:
+<p>
+<pre><a name=3De4d4>4.4)	mi ba'o djuno le nuzba
+	I [perfective] know the news.
+</pre><a href=3D#e4d4>Example 4.4 </a>means the same as <a href=3D#e4d3>Ex=
ample 4.3</a>, and it is simpler. However, often the position of the topic =
in the place structure of the selbri within the comment is vague:
+<p>
+<ex "fish eat">  XE "fish eat: example=93  <pre><a name=3De4d5>4.5)	le fin=
pe zo'u citka
+	the fish :  eat
+</pre><p>Is the fish eating or being eaten? The sentence doesn't say. The =
Chinese equivalent of <a href=3D#e4d5>Example 4.5 </a>is:
+<p>
+<pre><a name=3De4d6>4.6)	yu<sup>2	</sup>chi<sup>1</sup>
+	fish	eat
+</pre>which is vague in exactly the same way.
+<p>
+Grammatically, it is possible to have more than one sumti before =93zo'u=
=94. This is not normally useful in topic-comment sentences, but is necessa=
ry in the other use of =93zo'u=94: to separate a quantifying section from a=
 bridi containing quantified variables. This usage belongs to a discussion =
of quantifier logic in Lojban (see <a href=3Dchap16.html>Chapter 16</a>), b=
ut an example would be:
+<p>
+<pre><a name=3De4d7>4.7)	roda poi prenu ku'o su'ode zo'u de patfu da
+	For-all X which-are-persons, there-exists-a-Y such-that Y is the father o=
f X.
+	Every person has a father.
+</pre><p>The string of sumti before =93zo'u=94 (called the =93prenex=94: s=
ee <a href=3Dchap16.html>Chapter 16</a>) may contain both a topic and bound=
 variables:
+<p>
+<pre><a name=3De4d8>4.8)	loi patfu roda poi prenu ku'o=20
+su'ode zo'u de patfu da
+	For-the-mass-of fathers for-all X which-are-persons,=20
+there-exists-a-Y such-that Y is the father of X.
+	As for fathers, every person has one.
+</pre><cx "topic/comment, multiple sentence">  XE "topic/comment: multiple=
 sentence"   To specify a topic which affects more than one sentence, wrap =
the sentences in =93tu'e=A0<dots>=85</dots>tu'u=94 brackets and place the t=
opic and the =93zo'u=94 directly in front. This is the exception to the rul=
e that a topic attaches directly to a sentence:
+<p>
+<pre><a name=3De4d9>4.9)	loi jdini zo'u tu'e do ponse  .inaja do djica [tu=
'u]
+	The-mass-of money :  ( [if] you possess, then you want )
+	Money:  if you have it, you want it.
+</pre><p>Note: In Lojban, you do not =93want money=94; you =93want to have=
 money=94 or something of the sort, as the x2 place of =93djica=94 demands =
an event. As a result, the straightforward rendering of <a href=3D#e4d8>Exa=
mple 4.8 </a>without a topic is not:
+<p>
+<pre><a name=3De4d10>4.10)	do ponse loi jdini  .inaja do djica ri
+	You possess money only-if you desire its-mere-existence.
+</pre>where =93ri=94 means =93loi jdini=94 and is interpreted as =93the me=
re existence of money=94, but rather:
+<p>
+<pre><a name=3De4d11>4.11)	do ponse loi jdini  .inaja do djica tu'a ri
+	You possess money only-if you desire something-about it.
+</pre>namely, the possession of money. But topic-comment sentences like <a=
 href=3D#e4d9>Example 4.9 </a>are inherently vague, and this difference bet=
ween =93ponse=94 (which expects a physical object in x2) and =93djica=94 is=
 ignored. See <a href=3D#e9d3>Example 9.3 </a>for another topic/comment sen=
tence.
+<p>
+The subject of an English sentence is often the topic as well, but in Lojb=
an the sumti in the x1 place is not necessarily the topic, especially if it=
 is the normal (unconverted) x1 for the selbri. Thus Lojban sentences don't=
 necessarily have a =93subject=94 in the English sense.
+<p>
+Questions and answers</h3>
+<p><cx "questions">  XE "questions"   <cx "answers">  XE "answers"   <a na=
me=3Ds5><h3>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	xu	UI	truth question
+	ma	KOhA	sumti question
+	mo	GOhA	bridi question
+	xo	PA	number question
+	ji	A	sumti connective question
+	ge'i	GA	forethought connective question
+	gi'i	GIhA	bridi-tail connective question
+	gu'i	GUhA	tanru forethought connective question
+	je'i	JA	tanru connective question
+	pei	UI	attitude question
+	fi'a	FA	place structure question
+	cu'e	CUhE	tense/modal question
+	pau	UI	question premarker
+</pre><cx "questions, truth">  XE "questions: truth"   <lx "xu">  XE "xu" =
  <lx "UI">  XE "UI selma=92o"   Lojban questions are not at all like Engli=
sh questions. There are two basic types: truth questions, of the form =93Is=
 it true that=A0<dots>=85</dots>=94, and fill-in-the-blank questions. Truth=
 questions are marked by preceding the bridi, or following any part of it s=
pecifically questioned, with the cmavo =93xu=94 (of selma'o UI):
+<p>
+<pre><a name=3De5d1>5.1)	xu do klama le zarci
+	[True or false?] You go to the store
+	Are you going to the store/Did you go to the store?
+</pre>(Since the Lojban is tenseless, either colloquial translation might =
be correct.) Truth questions are further discussed in <a href=3Dchap15.html=
>Chapter 15</a>.
+<p>
+<cx "questions, fill-in-the-blank">  XE "questions: fill-in-the-blank"   F=
ill-in-the-blank questions have a cmavo representing some Lojban word or ph=
rase which is not known to the questioner, and which the answerer is to sup=
ply. There are a variety of cmavo belonging to different selma'o which prov=
ide different kinds of blanks.
+<p>
+<lx "ma">  XE "ma"   <lx "KOhA">  XE "KOhA selma=92o"   <cx "questions, su=
mti">  XE "questions: sumti"   Where a sumti is not known, a question may b=
e formed with =93ma=94 (of selma'o KOhA), which is a kind of pro-sumti:
+<p>
+<pre><a name=3De5d2>5.2)	ma klama le zarci
+	[What sumti?] goes-to the store
+	Who is going to the store?
+</pre><p>Of course, the =93ma=94 need not be in the x1 place:
+<p>
+<pre><a name=3De5d3>5.3)	do klama ma
+	You go-to [what sumti?]
+	Where are you going?
+
+</pre><p>The answer is a simple sumti:
+<p>
+<pre><a name=3De5d4>5.4)	le zarci
+	The store.
+</pre><p>A sumti, then, is a legal utterance, although it does not by itse=
lf constitute a bridi =96 it does not claim anything, but merely completes =
the open-ended claim of the previous bridi.
+<p>
+<cx "questions, multiple">  XE "questions: multiple"   There can be two =
=93ma=94 cmavo in a single question:
+<p>
+<pre><a name=3De5d5>5.5)	ma klama ma
+	Who goes where?
+</pre>and the answer would be two sumti, which are meant to fill in the tw=
o =93ma=94 cmavo in order:
+<p>
+<pre><a name=3De5d6>5.6)	mi le zarci
+	I, to the store.
+</pre><lx "fa'u">  XE "fa'u"   <lx "JOI">  XE "JOI selma=92o"   An even mo=
re complex example, depending on the non-logical connective =93fa'u=94 (of =
selma'o JOI), which is like the English =93and=A0<dots>=85</dots>respective=
ly=94:
+<p>
+<pre><a name=3De5d7>5.7)	ma fa'u ma klama ma fa'u ma
+	Who and who goes where and where, respectively?
+</pre><p>An answer might be
+<p>
+  XE "Marsha: example"  <pre><a name=3De5d8>5.8)	la djan. la marcas. le za=
rci le briju
+	John, Marsha, the store, the office.
+	John and Marsha go to the store and the office,
+		respectively.
+</pre>(Note: A mechanical substitution of <a href=3D#e5d8>Example 5.8 </a>=
into <a href=3D#e5d7>Example 5.7 </a>produces an ungrammatical result, beca=
use =93*=A0<dots>=85</dots>le zarci fa'u le briju=94 is ungrammatical Lojba=
n: the first =93le zarci=94 has to be closed with its proper terminator =93=
ku=94, for reasons explained in <a href=3Dchap14.html>Chapter 14</a>. This =
effect is not important: Lojban behaves as if all elided terminators have b=
een supplied in both question and answer before inserting the latter into t=
he former. The exchange is grammatical if question and answer are each sepa=
rately grammatical.)
+<p>
+<lx "mo">  XE "mo"   <lx "GOhA">  XE "GOhA selma=92o"   <cx "questions, se=
lbri">  XE "questions: selbri"   Questions to be answered with a selbri are=
 expressed with =93mo=94 of selma'o GOhA, which is a kind of pro-bridi:
+<p>
+<pre><a name=3De5d9>5.9)	la lojban. mo
+	Lojban [what selbri?]
+	What is Lojban?
+</pre><cx "predicate answers">  XE "predicate answers"   Here the answerer=
 is to supply some predicate which is true of Lojban. Such questions are ex=
tremely open-ended, due to the enormous range of possible predicate answers=
. The answer might be just a selbri, or might be a full bridi, in which cas=
e the sumti in the answer override those provided by the questioner. To lim=
it the range of a =93mo=94 question, make it part of a tanru.
+<p>
+<cx "questions, number">  XE "questions: number"   Questions about numbers=
 are expressed with =93xo=94 of selma'o PA:
+<p>
+<pre><a name=3De5d10>5.10)	do viska xo prenu
+	You saw [what number?] persons.
+	How many people did you see?
+
+</pre><p>The answer would be a simple number, another kind of non-bridi ut=
terance:
+<p>
+<pre><a name=3De5d11>5.11)	vomu
+	Forty-five.
+</pre><p>Fill-in-the-blank questions may also be asked about: logical conn=
ectives (using cmavo =93ji=94 of A, =93ge'i=94 of GA, =93gi'i=94 of GIhA, =
=93gu'i=94 of GUhA, or =93je'i=94 of JA, and receiving an ek, gihek, ijek, =
or ijoik as an answer) =97 see <a href=3Dchap14.html>Chapter 14</a>; attitu=
des (using =93pei=94 of UI, and receiving an attitudinal as an answer) =97 =
see <a href=3Dchap13.html>Chapter 13</a>; place structures (using =93fi'a=
=94 of FA, and receiving a cmavo of FA as an answer) =97 see <a href=3Dchap=
9.html>Chapter 9</a>; tenses and modals (using =93cu'e=94 of CUhE, and rece=
iving any tense or BAI cmavo as an answer) =97 see <a href=3Dchap9.html>Cha=
pter 9 </a>and <a href=3Dchap10.html>Chapter 10</a>.
+<p>
+Questions can be marked by placing =93pau=94 (of selma'o UI) before the qu=
estion bridi. See <a href=3Dchap13.html>Chapter 13 </a>for details.
+<p>
+The full list of non-bridi utterances suitable as answers to questions is:
+<p>
+<cx "utterances, non-bridi">  XE "utterances: non-bridi"   <cx "linked arg=
uments">  XE "linked arguments"   <dl compact><dt><dd>any number of sumti (=
with elidable terminator =93vau=94, see <a href=3Dchap6.html>Chapter 6</a>)
+an ek or gihek (logical connectives, see <a href=3Dchap14.html>Chapter 14<=
/a>)
+a number, or any mathematical expression placed in parentheses (see <a hre=
f=3Dchap18.html>Chapter 18</a>)
+a bare =93na=94 negator (to negate some previously expressed bridi), or co=
rresponding =93ja'a=94 affirmer (see <a href=3Dchap15.html>Chapter 15</a>)
+a relative clause (to modify some previously expressed sumti, see <a href=
=3Dchap8.html>Chapter 8</a>)
+a prenex/topic (to modify some previously expressed bridi, see <a href=3Dc=
hap16.html>Chapter 16</a>)
+linked arguments (beginning with =93be=94 or =93bei=94 and attached to som=
e previously expressed selbri, often in a description, see <a href=3Dchap5.=
html>Chapter 5</a>)
+</dl><p>At the beginning of a text, the following non-bridi are also permi=
tted:
+<p>
+<dl compact><dt><dd>one or more names (to indicate direct address without =
=93doi=94, see <a href=3Dchap6.html>Chapter 6</a>)
+indicators (to express a prevailing attitude, see <a href=3Dchap13.html>Ch=
apter 13</a>)
+=93nai=94 (to vaguely negate something or other, see <a href=3Dchap15.html=
>Chapter 15</a>)
+</dl><p>Where not needed for the expression of answers, most of these are =
made grammatical for pragmatic reasons: people will say them in conversatio=
n, and there is no reason to rule them out as ungrammatical merely because =
most of them are vague.
+<p>
+Subscripts: XI</h3>
+<p><cx "subscripts">  XE "subscripts"   <a name=3Ds6><h3>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	xi	XI	subscript
+</pre><lx "xi">  XE "xi"   <lx "XI">  XE "XI selma=92o"   <cx "subscriptin=
g">  XE "subscripts: marker"   The cmavo =93xi=94 (of selma'o XI) indicates=
 that a subscript (a number, a lerfu string, or a parenthesized mekso) foll=
ows. Subscripts can be attached to almost any construction and are placed f=
ollowing the construction (or its terminator word, which is generally requi=
red). They are useful either to extend the finite cmavo list to infinite le=
ngth, or to make more refined distinctions than the standard cmavo list per=
mits. The remainder of this section mentions some places where subscripts m=
ight naturally be used.
+<p>
+Lojban gismu have at most five places:
+<p>
+<pre><a name=3De6d1>6.1)	mi cu klama le zarci le zdani le dargu le karce
+	I go to-the market from-the house via-the road using-the car.
+</pre><p>Consequently, selma'o SE (which operates on a selbri to change th=
e order of its places) and selma'o FA (which provides place number tags for=
 individual sumti) have only enough members to handle up to five places. Co=
nversion of <a href=3D#e6d1>Example 6.1</a>, using =93xe=94 to swap the x1 =
and x5 places, would produce:
+<p>
+<pre><a name=3De6d2>6.2)	le karce cu xe klama le zarci le zdani le dargu
+mi
+	The car is-a-transportation-means to-the market from-the house via-the ro=
ad
+for-me.
+</pre><p>And reordering of the place structures might produce:
+<p>
+<pre><a name=3De6d3>6.3)	fo le dargu fi le zdani fa mi fe le zarci fu le k=
arce cu klama
+	Via the road, from the house, I, to the market, using-the car, go.
+</pre><a href=3D#e6d{1}>Examples 6.1 </a>to <a href=3D#e6d3>6.3 </a>all me=
an the same thing. But consider the lujvo =93nunkla=94, formed by applying =
the abstraction operator =93nu=94 to =93klama=94:
+<p>
+<pre><a name=3De6d4>6.4)	la'edi'u cu nunkla
+		mi le zarci le zdani le dargu le karce
+	The-referent-of-the-previous-sentence is-an-event-of-going
+		by-me to-the market from-the house via-the road using-the car.
+</pre><cx "subscripting, and sumti re-ordering">  XE =93subscripts: and su=
mti re-ordering"   <cx "SE, after 5th place">  XE "SE selma=92o: after 5th =
place"   <cx "FA, after 5th place">  XE "FA selma=92o: after 5th place"   <=
lx "se">  XE "se"   <lx "SE">  XE "SE selma=92o"   <a href=3D#e6d4>Example =
6.4 </a>shows that =93nunkla=94 has six places: the five places of =93klama=
=94 plus a new one (placed first) for the event itself. Performing transfor=
mations similar to that of <a href=3D#e6d2>Example 6.2 </a>requires an addi=
tional conversion cmavo that exchanges the x1 and x6 places. The solution i=
s to use any cmavo of SE with a subscript =936=94 (see Chapter 19):
+<p>
+<pre><a name=3De6d5>6.5)	le karce cu sexixa nunkla mi
+		le zarci le zdani le dargu la'edi'u
+	The car is-a-transportation-means-in-the-event-of-going by-me
+		to-the market via-the road which-is-referred-to-by-the-last-sentence.
+</pre><p>Likewise, a sixth place tag can be created by using any cmavo of =
FA with a subscript:
+<p>
+<pre><a name=3De6d6>6.6)	 fu le dargu fo le zdani fe mi fa la'edi'u
+		fi le zarci faxixa le karce cu klama
+	Via the road, from the house, by me, the-referent-of-the-last-sentence,
+		to the market, using the car, is-an-event-of-going.
+</pre><a href=3D#e6d{4}>Examples 6.4 </a>to <a href=3D#e6d6>6.6 </a>also a=
ll mean the same thing, and each is derived straightforwardly from any of t=
he others, despite the tortured nature of the English glosses. In addition,=
 any other member of SE or FA could be substituted into =93sexixa=94 and =
=93faxixa=94 without change of meaning: =93vexixa=94 means the same thing a=
s =93sexixa=94.
+<p>
+<cx "subscripting, and pro-sumti">  XE =93subscripts: and pro-sumti"   <cx=
 "da-series, after third">  XE "da-series: after third"   <cx "ko'a-series,=
 after tenth">  XE "ko'a-series: after tenth"   Lojban provides two groups =
of pro-sumti, both belonging to selma'o KOhA. The ko'a-series cmavo are use=
d to refer to explicitly specified sumti to which they have been bound usin=
g =93goi=94. The da-series, on the other hand, are existentially or univers=
ally quantified variables. (These concepts are explained more fully in <a h=
ref=3Dchap16.html>Chapter 16</a>.)  There are ten ko'a-series cmavo and 3 d=
a-series cmavo available.
+<p>
+<lx "DA">  XE "DA selma=92o"   <lx "da">  XE "da"   If more are required, =
any cmavo of the ko'a-series or the da-series can be subscripted:
+<p>
+<pre><a name=3De6d7>6.7)	daxivo
+	X sub 4
+</pre>is the 4th bound variable of the 1st sequence of the da-series, and
+<p>
+<pre><a name=3De6d8>6.8)	ko'ixipaso
+	something-3 sub 18
+</pre>is the 18th free variable of the 3rd sequence of the ko'a-series. Th=
is convention allows 10 sequences of ko'a-type pro-sumti and 3 sequences of=
 da-type pro-sumti, each with as many members as needed. Note that =93daxiv=
o=94 and =93dexivo=94 are considered to be distinct pro-sumti, unlike the s=
ituation with =93sexixa=94 and =93vexixa=94 above. Exactly similar treatmen=
t can be given to the bu'a-series of selma'o GOhA and to the gismu pro-brid=
i =93broda=94, =93brode=94, =93brodi=94, =93brodo=94, and =93brodu=94.
+<p>
+<p>
+<cx "subscripting, mathematical">  XE =93subscripts: mathematical"   Subsc=
ripts on lerfu words are used in the standard mathematical way to extend th=
e number of variables:
+<p>
+<pre><a name=3De6d9>6.9)	li xy.boixipa du li xy.boixire su'i xy.boixici
+	The-number x-sub-1 equals the-number x-sub-2 plus x-sub-3
+	$x<sub>1</sub>  =3D x<sub>2</sub> + x<sub>3</sub>$
+</pre>and can be used to extend the number of pro-sumti as well, since ler=
fu strings outside mathematical contexts are grammatically and semantically=
 equivalent to pro-sumti of the ko'a-series. (In <a href=3D#e6d9>Example 6.=
9</a>, note the required terminator =93boi=94 after each =93xy.=94 cmavo; t=
his terminator allows the subscript to be attached without ambiguity.)
+<p>
+<p>
+<cx "subscripting, and names">  XE =93subscripts: and names"   Names, whic=
h are similar to pro-sumti, can also be subscripted to distinguish two indi=
viduals with the same name:
+<p>
+<pre><a name=3De6d10>6.10)	la djan. xipa cusku lu mi'enai do li'u la djan.=
 xire
+	John<sub>1</sub> expresses =93I-am-not you=94 to John<sub>2</sub>.
+<cx "subscripting, and tense">  XE =93subscripts: and tense"   Subscripts =
on tenses allow talking about more than one time or place that is described=
 by the same general cmavo. For example, =93puxipa=94 could refer to one po=
int in the past, and =93puxire=94 a second point (earlier or later).
+<p>
+<p>
+<cx "subscripting, and fuzzy truths">  XE =93subscripts: and fuzzy truths"=
   You can place a subscript on the word =93ja'a=94, the bridi affirmative =
of selma'o NA, to express so-called fuzzy truths. The usual machinery for f=
uzzy logic (statements whose truth value is not merely =93true=94 or =93fal=
se=94, but is expressed by a number in the range 0 to 1) in Lojban is the a=
bstractor =93jei=94:
+<p>
+<pre><a name=3De6d11>6.11)	li pimu jei mi ganra
+	The-number .5 is-the-truth-value-of my being-broad
+</pre><p>However, by convention we can attach a subscript to =93ja'a=94 to=
 indicate fuzzy truth (or to =93na=94 if we change the amount):
+<p>
+<pre><a name=3De6d12>6.12)	mi ja'a xipimu ganra
+	I truly-sub-.5 am-broad
+</pre><cx "subscripting, and paragraph separators">  XE =93subscripts: and=
 paragraph separators"   Finally, as mentioned in <a href=3D#s2>Section 2</=
a>, =93ni'o=94 and =93no'i=94 cmavo with matching subscripts mark the start=
 and the continuation of a given topic respectively. Different topics can b=
e assigned to different subscripts.
+<p>
+Other uses of subscripts will doubtless be devised in future.
+<p>
+Utterance ordinals: MAI</h3>
+<p><cx "ordinals, utterance">  XE "ordinals: utterance"   <a name=3Ds7><h3=
>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	mai	MAI	utterance ordinal, -thly
+	mo'o	MAI	higher order utterance ordinal
+</pre><lx "mai">  XE "mai"   <lx "mo'o">  XE "mo'o"   <lx "MAI">  XE "MAI =
selma=92o"   Numerical free modifiers, corresponding to English =93firstly=
=94, =93secondly=94, and so on, can be created by suffixing =93mai=94 or =
=93mo'o=94 of selma'o MAI to a number or a lerfu string. Here are some exam=
ples:
+<p>
+<pre><a name=3De7d1>7.1)	mi klama pamai le zarci .e remai le zdani
+	I go-to (firstly) the store and (secondly) the market.
+</pre><p>This does not imply that I go to the store before I go to the mar=
ket: that meaning requires a tense. The sumti are simply numbered for conve=
nience of reference. Like other free modifiers, the utterance ordinals can =
be inserted almost anywhere in a sentence without affecting its grammar or =
its meaning.
+<p>
+Any of the Lojban numbers can be used with MAI: =93romai=94, for example, =
means =93all-thly=94 or =93lastly=94. Likewise, if you are enumerating a lo=
ng list and have forgotten which number is wanted next, you can say =93ny.m=
ai=94, or =93Nthly=94.
+<p>
+The difference between =93mai=94 and =93mo'o=94 is that =93mo'o=94 enumera=
tes larger subdivisions of a text; =93mai=94 was designed for lists of numb=
ered items, whereas =93mo'o=94 was intended to subdivide structured works. =
If this chapter were translated into Lojban, it might number each section w=
ith =93mo'o=94: this section would then be introduced with =93zemo'o=94, or=
 =93<a href=3D#s7>Section 7</a>.=94
+<p>
+<a name=3Ds8><h3>Attitude scope markers: FUhE/FUhO</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	fu'e	FUhE	open attitudinal scope
+	fu'o	FUhO	close attitudinal scope
+</pre><lx "UI">  XE "UI selma=92o"   Lojban has a complex system of =93att=
itudinals=94, words which indicate the speaker's attitude to what is being =
said. The attitudinals include indicators of emotion, intensity markers, di=
scursives (which show the structure of discourse), and evidentials (which i=
ndicate =93how the speaker knows=94). Most of these words belong to selma'o=
 UI; the intensity markers belong to selma'o CAI for historical reasons, bu=
t the two selma'o are grammatically identical. The individual cmavo of UI a=
nd CAI are discussed in <a href=3Dchap13.html>Chapter 13</a>; only the rule=
s for applying them in discourse are presented here.
+<p>
+<cx "attitudinal, scope">  XE "attitudinal: scope"   Normally, an attitudi=
nal applies to the preceding word only. However, if the preceding word is a=
 structural cmavo which begins or ends a whole construction, then that whol=
e construction is affected by the attitudinal:
+
+<p>
+<pre><a name=3De8d1>8.1)	mi viska le blanu .ia zdani [ku]
+	I see the blue [belief] house.
+	I see the house, which I believe to be blue.
+<a name=3De8d2>8.2)	mi viska le blanu zdani .ia [ku]
+	I see the blue house [belief].
+	I see the blue thing, which I believe to be a house.
+<a name=3De8d3>8.3)	mi viska le .ia blanu zdani [ku]
+	I see the [belief] blue house.
+	I see what I believe to be a blue house.
+<a name=3De8d4>8.4)	mi viska le blanu zdani ku .ia
+	I see (the blue house ) [belief].
+	I see what I believe to be a blue house.
+</pre><p>An attitudinal meant to cover a whole sentence can be attached to=
 the preceding =93.i=94, expressed or understood:
+<p>
+<pre><a name=3De8d5>8.5)	[.i] .ia mi viska le blanu zdani
+	[belief] I see the blue house.
+	I believe I see a blue house.
+</pre>or to an explicit =93vau=94 placed at the end of a bridi.
+<p>
+Likewise, an attitudinal meant to cover a whole paragraph can be attached =
to =93ni'o=94 or =93no'i=94. An attitudinal at the beginning of a text appl=
ies to the whole text.
+<p>
+<lx "fu'e">  XE "fu'e"   <lx "FUhE">  XE "FUhE selma=92o"   <lx "fu'o">  X=
E "fu'o"   <lx "FUhO">  XE "FUhO selma=92o"   However, sometimes it is nece=
ssary to be more specific about the range of one or more attitudinals, part=
icularly if the range crosses the boundaries of standard Lojban syntactic c=
onstructions. The cmavo =93fu'e=94 (of selma'o FUhE) and =93fu'o=94 (of sel=
ma'o FUhO) provide explicit scope markers. Placing =93fu'e=94 in front of a=
n attitudinal disconnects it from what precedes it, and instead says that i=
t applies to all following words until further notice. The notice is given =
by =93fu'o=94, which can appear anywhere and cancels all in-force attitudin=
als. For example:
+<p>
+<pre><a name=3De8d6>8.6)	mi viska le fu'e .ia blanu zdani fu'o ponse
+	I see the [start] [belief] blue house [end] possessor.
+	I see the owner of what I believe to be a blue house.
+</pre><p>Here, only the =93blanu zdani=94 portion of the three-part tanru =
=93blanu zdani ponse=94 is marked as a belief of the speaker. Naturally, th=
e attitudinal scope markers do not affect the rules for interpreting multi-=
part tanru: =93blanu zdani=94 groups first because tanru group from left to=
 right unless overridden with =93ke=94 or =93bo=94.
+<p>
+<cx "global attitudinals">  XE "global attitudinals"   Other attitudinals =
of more local scope can appear after attitudinals marked by FUhE; these att=
itudinals are added to the globally active attitudinals rather than superse=
ding them.
+<p>
+Quotations: LU, LIhU, LOhU, LEhU</h3>
+<p><cx "quotation">  XE "quotation"   <a name=3Ds9><h3>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	lu	LU	begin quotation
+	li'u	LIhU	end quotation
+	lo'u	LOhU	begin error quotation
+	le'u	LEhU	end error quotation
+</pre><p>Grammatically, quotations are very simple in Lojban: all of them =
are sumti, and they all mean something like =93the piece of text here quote=
d=94:
+<p>
+<pre><a name=3De9d1>9.1)	mi pu cusku lu mi'e djan [li'u]
+	I [past] express [quote] I-am John [unquote]
+	I said, =93I'm John=94.
+</pre><p>But in fact there are four different flavors of quotation in the =
language, involving six cmavo of six different selma'o. This being the case=
, quotation deserves some elaboration.
+<p>
+<cx "opening quotation">  XE "opening quotation"   <lx "lu">  XE "lu"   <l=
x "LU">  XE "LU selma=92o"   <lx "li'u">  XE "li'u"   <lx "LIhU">  XE "LIhU=
 selma=92o"   <cx "quotation, of grammatical Lojban text">  XE "quotation: =
of grammatical Lojban text"   The simplest kind of quotation, exhibited in =
<a href=3D#e9d1>Example 9.1</a>, uses the cmavo =93lu=94 (of selma'o LU) as=
 the opening quotation mark, and the cmavo =93li'u=94 (of selma'o LIhU) as =
the closing quotation mark. The text between =93lu=94 and =93li'u=94 must b=
e a valid, parseable Lojban text. If the quotation is ungrammatical, so is =
the surrounding expression. The cmavo =93li'u=94 is technically an elidable=
 terminator, but it's almost never possible to elide it except at the end o=
f text.
+<p>
+<lx "lo'u">  XE "lo'u"   <lx "LOhU">  XE "LOhU selma=92o"   <lx "le'u">  X=
E "le'u"   <lx "LEhU">  XE "LEhU selma=92o"   <cx "quotation, of parseable =
Lojban text">  XE "quotation: of parseable Lojban text"   <cx "quotation, o=
f ungrammatical Lojban text">  XE "quotation: of ungrammatical Lojban text"=
   The cmavo =93lo'u=94 (of selma'o LOhU) and =93le'u=94 (of selma'o LEhU) =
are used to surround a quotation that is not necessarily grammatical Lojban=
. However, the text must consist of morphologically correct Lojban words (a=
s defined in <a href=3Dchap4.html>Chapter 4</a>), so that the =93le'u=94 ca=
n be picked out reliably. The words need not be meaningful, but they must b=
e recognizable as cmavo, brivla, or cmene. Quotation with =93lo'u=94 is ess=
ential to quoting ungrammatical Lojban for teaching in the language, the eq=
uivalent of the * that is used in English to mark such errors:
+<p>
+<pre><a name=3De9d2>9.2)	lo'u mi du do du la djan. le'u na tergerna=20
+la lojban.
+	[quote] mi du do du la djan. [unquote] is-not a-grammatical-structure=20
+in Lojban.
+</pre><cx "embedded quotation">  XE "embedded quotation"   <a href=3D#e9d2=
>Example 9.2 </a>is grammatical even though the embedded quotation is not. =
Similarly, =93lo'u=94 quotation can quote fragments of a text which themsel=
ves do not constitute grammatical utterances:
+<p>
+<pre><a name=3De9d3>9.3)	lu le mlatu cu viska le finpe li'u zo'u lo'u visk=
a le le'u
+		cu selbasti .ei lo'u viska lo le'u
+	[quote] le mlatu cu viska le finpe [unquote] : [quote] viska le [unquote]
+is-replaced-by [obligation!] [quote] viska lo [unquote].
+	In the sentence =93le mlatu viska le finpe=94, =93viska le=94 should be r=
eplaced by
+=93viska lo=94.
+</pre><p>Note the topic-comment formulation (<a href=3D#s4>Section 4</a>) =
and the indicator applying to the selbri only (<a href=3D#s8>Section 8</a>)=
. Neither =93viska le=94 nor =93viska lo=94 is a valid Lojban utterance, an=
d both require =93lo'u=94 quotation.
+<p>
+<cx "pro-sumti, quotation of">  XE "pro-sumti: quotation of"   <cx "pro-br=
idi, quotation of">  XE "pro-bridi: quotation of"   Additionally, pro-sumti=
 or pro-bridi in the quoting sentence can refer to words appearing in the q=
uoted sentence when =93lu=A0<dots>=85</dots>li'u=94 is used, but not when =
=93lo'u=A0<dots>=85</dots>le'u=94 is used:
+<p>
+<pre><a name=3De9d4>9.4)	la tcarlis. cusku lu le ninmu cu morsi li'u
+		.iku'i ri jmive
+	Charlie says [quote] the woman is-dead [unquote].=20
+		However, the-last-mentioned is-alive.
+	Charlie says =93The woman is dead=94, but she is alive.
+</pre><p>In <a href=3D#e9d4>Example 9.4</a>, =93ri=94 is a pro-sumti which=
 refers to the most recent previous sumti, namely =93le ninmu=94. Compare:
+<p>
+<pre><a name=3De9d5>9.5)	la tcarlis. cusku lo'u le ninmu cu morsi le'u
+		.iku'i ri jmive
+	Charlie says [quote] le ninmu cu morsi [unquote].
+		However, the-last-mentioned is-alive.
+	Charlie says =93le ninmu cu morsi=94, but he is alive.
+</pre><cx "uninterpreted sequence">  XE "uninterpreted sequence"   In <a h=
ref=3D#e9d5>Example 9.5</a>, =93ri=94 cannot refer to the referent of the a=
lleged sumti =93le ninmu=94, because =93le ninmu cu morsi=94 is a mere unin=
terpreted sequence of Lojban words. Instead, =93ri=94 ends up referring to =
the referent of the sumti =93la tcarlis.=94, and so it is Charlie who is al=
ive.
+<p>
+<lx "si">  XE "si"   <lx "sa">  XE "sa"   <lx "su">  XE "su"   <cx "metali=
nguistic erasers, within ungrammatical-Lojban quotation">  XE "metalinguist=
ic erasers: within ungrammatical-Lojban quotation"   <cx "quotation, ungram=
matical Lojban containing lo'u">  XE "quotation: ungrammatical Lojban conta=
ining lo'u"   <cx "quotation, ungrammatical Lojban containing le'u">  XE "q=
uotation: ungrammatical Lojban containing le'u"   The metalinguistic eraser=
s =93si=94, =93sa=94, and =93su=94, discussed in <a href=3D#s13>Section 13<=
/a>, do not operate in text between =93lo'u=94 and =93le'u=94. Since the fi=
rst =93le'u=94 terminates a =93lo'u=94 quotation, it is not directly possib=
le to have a =93lo'u=94 quotation within another =93lo'u=94 quotation. Howe=
ver, it is possible for a =93le'u=94 to occur within a =93lo'u=A0<dots>=85<=
/dots>le'u=94 quotation by preceding it with the cmavo =93zo=94, discussed =
in <a href=3D#s10>Section 10</a>. Note that =93le'u=94 is not an elidable t=
erminator; it is required.
+<p>
+<a name=3Ds10><h3>More on quotations: ZO, ZOI</h3>
+<p>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	zo	ZO	quote single word
+	zoi	ZOI	non-Lojban quotation
+	la'o	ZOI	non-Lojban name
+</pre><cx "strong quotation">  XE "strong quotation"   <lx "zo">  XE "zo" =
  <lx "ZO">  XE "ZO selma=92o"   <cx "quotation, of single word">  XE "quot=
ation: of single word"   The cmavo =93zo=94 (of selma'o ZO) is a strong quo=
tation mark for the single following word, which can be any Lojban word wha=
tsoever. Among other uses, =93zo=94 allows a metalinguistic word to be refe=
renced without having it act on the surrounding text. The word must be a mo=
rphologically legal (but not necessarily meaningful) single Lojban word; co=
mpound cmavo are not permitted. For example:
+<p>
+<pre><a name=3De10d1>10.1)	zo si cu lojbo valsi
+	=93si=94 is a Lojbanic word.
+</pre><p>Since =93zo=94 acts on a single word only, there is no correspond=
ing terminator. Brevity, then, is a great advantage of =93zo=94, since the =
terminators for other kinds of quotation are rarely or never elidable.
+<p>
+<cx "phoneme stream">  XE "phoneme stream"   <lx "zoi">  XE "zoi"   <lx "Z=
OI">  XE "ZOI selma=92o"   <cx "quotation, of non-Lojban">  XE "quotation: =
of non-Lojban"   The cmavo =93zoi=94 (of selma'o ZOI) is a quotation mark f=
or quoting non-Lojban text. Its syntax is =93zoi X. text .X=94, where X is =
a Lojban word (called the delimiting word) which is separated from the quot=
ed text by pauses, and which is not found in the written text or spoken pho=
neme stream. It is common, but not required, to use the lerfu word (of selm=
a'o BY) which corresponds to the Lojban name of the language being quoted:
+<p>
+<pre><a name=3De10d2>10.2)	zoi gy. John is a man .gy. cu glico jufra
+	=93John is a man=94 is an English sentence.
+</pre>where =93gy=94 stands for =93glico=94. Other popular choices of deli=
miting words are =93.kuot.=94, a Lojban name which sounds like the English =
word =93quote=94, and the word =93zoi=94 itself. Another possibility is a L=
ojban word suggesting the topic of the quotation.
+<p>
+Within written text, the Lojban written word used as a delimiting word may=
 not appear, whereas within spoken text, the sound of the delimiting word m=
ay not be uttered. This leads to occasional breakdowns of audio-visual isom=
orphism: <a href=3D#e10d3>Example 10.3 </a>is fine in speech but ungrammati=
cal as written, whereas <a href=3D#e10d4>Example 10.4 </a>is correct when w=
ritten but ungrammatical in speech.
+<p>
+<pre><a name=3De10d3>10.3)	?mi djuno fi le valsi po'u zoi gy. gyrations .g=
y.
+	I know about the word which-is =93gyrations=94.
+<a name=3De10d4>10.4)	?mi djuno fi le valsi po'u zoi jai. gyrations .jai
+	I know about the word which-is =93gyrations=94.
+</pre><p>The text =93gy=94 appears in the written word =93gyrations=94, wh=
ereas the sound represented in Lojban by =93jai=94 appears in the spoken wo=
rd =93gyrations=94. Such borderline cases should be avoided as a matter of =
good style.
+<p>
+<cx "quotation, of rafsi">  XE "quotation: of rafsi"   <cx "rafsi, quotati=
on of">  XE "rafsi: quotation of"   It should be noted particularly that =
=93zoi=94 quotation is the only way to quote rafsi, specifically CCV rafsi,=
 because they are not Lojban words, and =93zoi=94 quotation is the only way=
 to quote things which are not Lojban words. (CVC and CVV rafsi look like n=
ames and cmavo respectively, and so can be quoted using other methods.) For=
 example:
+<p>
+<pre><a name=3De10d5>10.5)	zoi ry. sku .ry. cu rafsi zo cusku
+	=93sku=94 is a rafsi of =93cusku=94.
+</pre><cx "lo'u, interaction with zoi">  XE "lo'u: interaction with zoi"  =
 <cx "le'u, interaction with zoi">  XE "le'u: interaction with zoi"   <cx "=
zoi, interaction with lo'u/le'u">  XE "zoi: interaction with lo'u/le'u"   (=
A minor note on interaction between =93lo'u=A0<dots>=85</dots>le'u=94 and =
=93zoi=94: The text between =93lo'u=94 and =93le'u=94 should consist of Loj=
ban words only. In fact, non-Lojban material in the form of a =93zoi=94 quo=
tation may also appear. However, if the word =93le'u=94 is used either as t=
he delimiting word for the =93zoi=94 quotation, or within the quotation its=
elf, the outer =93lo'u=94 quotation will be prematurely terminated. Therefo=
re, =93le'u=94 should be avoided as the delimiting word in any =93zoi=94 qu=
otation.)
+<p>
+<cx "names, with zo versus la">  XE "names: with zo versus la"   <cx "zo, =
contrasted with la for names">  XE "zo: contrasted with la for names"   <cx=
 "la, contrasted with zo">  XE "la: contrasted with zo"   <cx "quotation, r=
eferent versus symbol">  XE "quotation: referent versus symbol"   <cx "refe=
rent, contrasted with symbol">  XE "referent: contrasted with symbol"   <cx=
 "symbol, contrasted with referent">  XE "symbol: contrasted with referent"=
   Lojban strictly avoids any confusion between things and the names of thi=
ngs:
+<p>
+<pre><a name=3De10d6>10.6)	zo .bab. cmene la bab.
+	The-word =93Bob=94 is-the-name-of the-one-named Bob.
+</pre><lx "la'e">  XE "la'e"   <lx "lu'e">  XE "lu'e"   <lx "LAhE">  XE "L=
AhE selma=92o"   In <a href=3D#e10d6>Example 10.6</a>, =93zo .bab.=94 is th=
e word, whereas =93la bab.=94 is the thing named by the word. The cmavo =93=
la'e=94 and =93lu'e=94 (of selma'o LAhE) convert back and forth between ref=
erences and their referents:
+<p>
+<pre><a name=3De10d7>10.7)	zo .bab. cmene la'e zo .bab.
+	The-word =93Bob=94 is-the-name-of the-referent-of the-word =93Bob=94.
+<a name=3De10d8>10.8)	lu'e la bab. cmene la bab.
+	A-symbol-for Bob is-the-name-of Bob.
+</pre><a href=3D#e10d{6}>Examples 10.6 </a>through <a href=3D#e10d8>10.8 <=
/a>all mean approximately the same thing, except for differences in emphasi=
s. <a href=3D#e10d9>Example 10.9 </a>is different:
+<p>
+<pre><a name=3De10d9>10.9)	la bab. cmene la bab.
+	Bob is the name of Bob.
+</pre>and says that Bob is both the name and the thing named, an unlikely =
situation. People are not names.
+<p>
+(In <a href=3D#e10d{6}>Examples 10.6 </a>through <a href=3D#e10d7>10.7</a>=
, the name =93bab.=94 was separated from a preceding =93zo=94 by a pause, t=
hus: =93zo .bab.=94. The reason for this extra pause is that all Lojban nam=
es must be separated by pause from any preceding word other than =93la=94, =
=93lai=94, =93la'i=94 (all of selma'o LA) and =93doi=94 (of selma'o DOI). T=
here are numerous other cmavo that may precede a name: of these, =93zo=94 i=
s one of the most common.)
+<p>
+<cx "Linnaean">  XE "Linnaean"   <cx "Goethe">  XE "Goethe"   <lx "la'o"> =
 XE "la'o"   <lx "ZOI">  XE "ZOI selma=92o"   <cx "names, non-Lojban">  XE =
"names: non-Lojban"   <cx "Linnaean binomials">  XE "Linnaean binomials"   =
The cmavo =93la'o=94 also belongs to selma'o ZOI, and is mentioned here for=
 completeness, although it does not signal the beginning of a quotation. In=
stead, =93la'o=94 serves to mark non-Lojban names, especially the Linnaean =
binomial names (such as =93Homo sapiens=94) which are the internationally s=
tandardized names for species of animals and plants. Internationally known =
names which can more easily be recognized by spelling rather than pronuncia=
tion, such as =93Goethe=94, can also appear in Lojban text with =93la'o=94:
+<p>
+<pre><a name=3De10d10>10.10)	la'o dy. Goethe .dy. cu me la'o ly. Homo sapi=
ens .ly.
+	Goethe is a Homo sapiens.
+</pre><cx "cumbersome text">  XE "cumbersome text"   Using =93la'o=94 for =
all names rather than Lojbanizing, however, makes for very cumbersome text.=
 A rough equivalent of =93la'o=94 might be =93la me zoi=94.
+<p>
+Contrastive emphasis: BAhE</h3>
+<p><cx "emphasis">  XE "emphasis"   <a name=3Ds11><h3>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	ba'e	BAhE	emphasize next word
+	za'e	BAhE	next word is nonce
+</pre><p>English often uses strong stress on a word to single it out for c=
ontrastive emphasis, thus
+<p>
+<pre><a name=3De11d1>11.1)	I saw George.
+</pre>is quite different from
+<p>
+<pre><a name=3De11d2>11.2)	I saw <i>George</i>.
+</pre><p>The heavy stress on =93<i>George</i>=94 (represented in writing b=
y <i>italics</i>) indicates that I saw George rather than someone else. Loj=
ban does not use stress in this way: stress is used only to help separate w=
ords (because every brivla is stressed on the penultimate syllable) and in =
names to match other languages' stress patterns. Note that many other langu=
ages do not use stress in this way either; typically word order is rearrang=
ed, producing something like
+<p>
+<pre><a name=3De11d3>11.3)	It was George whom I saw.
+</pre><lx "ba'e">  XE "ba'e"   <lx "BAhE">  XE "BAhE selma=92o"   In Lojba=
n, the cmavo =93ba'e=94 (of selma'o BAhE) precedes a single word which is t=
o be emphasized:
+<p>
+<pre><a name=3De11d4>11.4)	mi viska la ba'e .djordj.
+	I saw the-one-named [emphasis] =93George=94.
+	I saw <i>George</i>.
+</pre><p>Note the pause before the name =93djordj.=94, which serves to sep=
arate it unambiguously from the =93ba'e=94. Alternatively, the =93ba'e=94 c=
an be moved to a position before the =93la=94, which in effect emphasizes t=
he whole construct =93la djordj.=94:
+<p>
+<pre><a name=3De11d5>11.5)	mi viska ba'e la djordj.
+	I saw [emphasis] the-one-named =93George=94.
+	I saw <i>George</i>.
+</pre><cx "emphasis marking">  XE "emphasis marking"   Marking a word with=
 a cmavo of BAhE does not change the word's grammar in any way. Any word in=
 a bridi can receive contrastive emphasis marking:
+
+<p>
+<pre><a name=3De11d6>11.6)	ba'e mi viska la djordj.
+	I, no one else, saw George.
+<a name=3De11d7>11.7)	mi ba'e viska la djordj.
+	I saw (not heard or smelled) George.
+</pre><p>Emphasis on one of the structural components of a Lojban bridi ca=
n also be achieved by rearranging it into an order that is not the speaker'=
s or writer's usual order. Any sumti moved out of place, or the selbri when=
 moved out of place, is emphatic to some degree.
+<p>
+<lx "za'e">  XE "za'e"   <lx "BAhE">  XE "BAhE selma=92o"   <cx "non-stand=
ard words, marking">  XE "non-standard words: marking"   <cx "words, markin=
g non-standard">  XE "words: marking non-standard"   For completeness, the =
cmavo =93za'e=94 should be mentioned, also of selma'o BAhE. It marks a word=
 as possibly irregular, non-standard, or nonce (created for the occasion):
+<p>
+  XE "Albania: example"  <pre><a name=3De11d8>11.8)	mi klama la za'e. .alb=
einias
+	I go-to so-called Albania
+</pre><cx "appropriate standard">  XE "appropriate standard"   marks a Loj=
banization of an English name, where a more appropriate standard form might=
 be something like =93la ctiipyris.=94, reflecting the country's name in Al=
banian.
+<p>
+<cx "unabridged dictionary">  XE "unabridged dictionary"   Before a lujvo =
or fu'ivla, =93za'e=94 indicates that the word has been made up on the spot=
 and may be used in a sense that is not found in the unabridged dictionary =
(when we have an unabridged dictionary!).
+<p>
+Parenthesis and metalinguistic commentary: TO, TOI, SEI</h3>
+<p><cx "parenthesis, textual">  XE "parenthesis: textual"   <cx "metalingu=
istic commentary">  XE "metalinguistic commentary"   <a name=3Ds12><h3>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	to	TO	open parenthesis
+	to'i	TO	open editorial parenthesis
+	toi	TOI	close parenthesis
+	sei	SEI	metalinguistic bridi marker
+</pre><lx "to">  XE "to"   <lx "TO">  XE "TO selma=92o"   <lx "toi">  XE "=
toi"   <lx "TOI">  XE "TOI selma=92o"   The cmavo =93to=94 and =93toi=94 ar=
e discursive (non-mathematical) parentheses, for inserting parenthetical re=
marks. Any text whatsoever can go within the parentheses, and it is complet=
ely invisible to its context. It can, however, refer to the context by the =
use of pro-sumti and pro-bridi: any that have been assigned in the context =
are still assigned in the parenthetical remarks, but the reverse is not tru=
e.
+<p>
+<pre><a name=3De12d1>12.1)	doi lisas. mi djica le nu to doi frank. ko sist=
i toi do viska le mlatu
+	O Lisa, I desire the event-of (O Frank, [imperative] stop!) you see the c=
at.
+	Lisa, I want you to (Frank! Stop!) see the cat.
+</pre><a href=3D#e12d1>Example 12.1 </a>implicitly redefines =93do=94 with=
in the parentheses: the listener is changed by =93doi frank.=94 When the co=
ntext sentence resumes, however, the old listener, Lisa, is automatically r=
estored.
+<p>
+<lx "to'i">  XE "to'i"   <lx "TO">  XE "TO selma=92o"   <cx "editorial com=
mentary">  XE "editorial commentary"   There is another cmavo of selma'o TO=
: =93to'i=94. The difference between =93to=94 and =93to'i=94 is the differe=
nce between parentheses and square brackets in English prose. Remarks withi=
n =93to=A0<dots>=85</dots>toi=94 cmavo are implicitly by the same speaker, =
whereas remarks within =93to'i=A0<dots>=85</dots>toi=94 are implicitly by s=
omeone else, perhaps an editor:
+<p>
+<pre><a name=3De12d2>12.2)	la frank. cusku lu mi prami do to'isa'a do du l=
a djein. toi li'u
+	Frank expresses =93I love you [you =3D Jane]=94
+</pre><cx "bracketed remark">  XE "bracketed remark"   <cx "actual confusi=
on=94><lx "sa'a">  XE "sa'a"   <lx "UI">  XE "UI selma=92o"   <cx "editoria=
l insertion">  XE "editorial insertion"   The =93sa'a=94 suffix is a discur=
sive cmavo (of selma'o UI) meaning =93editorial insertion=94, and indicatin=
g that the marked word or construct (in this case, the entire bracketed rem=
ark) is not part of the quotation. It is required whenever the =93to'i=A0<d=
ots>=85</dots>toi=94 remark is physically within quotation marks, at least =
when speaking to literal-minded listeners; the convention may be relaxed if=
 no actual confusion results.
+<p>
+Note: The parser believes that parentheses are attached to the previous wo=
rd or construct, because it treats them as syntactic equivalents of subscri=
pts and other such so-called =93free modifiers=94. Semantically, however, p=
arenthetical remarks are not necessarily attached either to what precedes t=
hem or what follows them.
+<p>
+<lx "sei">  XE "sei"   <lx "SEI">  XE "SEI selma=92o"   <cx "embedded disc=
ursive">  XE "embedded discursive"   <cx "discursive, embedded">  XE "discu=
rsives: embedded"   <cx "metalinguistic comment, with embedded discursive">=
  XE "metalinguistic comment: with embedded discursive"   The cmavo =93sei=
=94 (of selma'o SEI) begins an embedded discursive bridi. Comments added wi=
th =93sei=94 are called =93metalinguistic=94, because they are comments abo=
ut the discourse itself rather than about the subject matter of the discour=
se. This sense of the term =93metalinguistic=94 is used throughout this cha=
pter, and is not to be confused with the sense =93language for expressing o=
ther languages=94.
+<p>
+When marked with =93sei=94, a metalinguistic utterance can be embedded in =
another utterance as a discursive. In this way, discursives which do not ha=
ve cmavo assigned in selma'o UI can be expressed:
+<p>
+<pre><a name=3De12d3>12.3)	la frank. prami sei la frank. gleki la djein.
+	Frank loves (Frank is happy) Jane.
+</pre><p>Using the happiness attitudinal, =93.ui=94, would imply that the =
speaker was happy. Instead, the speaker attributes happiness to Frank. It w=
ould probably be safe to elide the one who is happy, and say:
+<p>
+<pre><a name=3De12d4>12.4)	la frank. prami sei gleki la djein.
+	Frank loves (he is happy) Jane.
+</pre><p>The grammar of the bridi following =93sei=94 has an unusual limit=
ation: the sumti must either precede the selbri, or must be glued into the =
selbri with =93be=94 and =93bei=94:
+<p>
+  XE "Susan: example"  <pre><a name=3De12d5>12.5)	la frank. prami sei glek=
i be fa la suzn. la djein.
+	Frank loves (Susan is happy) Jane.
+</pre><p>This restriction allows the terminator cmavo =93se'u=94 to almost=
 always be elided.
+<p>
+<cx "metalinguistic levels">  XE "metalinguistic levels"   <cx "metalingui=
stic levels or reference">  XE "metalinguistic levels or reference"   <cx "=
pro-sumti, and discursive utterances">  XE "pro-sumti: and discursive utter=
ances"   <cx "reference, and discursive utterances">  XE "reference: and di=
scursive utterances"   Since a discursive utterance is working at a =93high=
er=94 level of abstraction than a non-discursive utterance, a non-discursiv=
e utterance cannot refer to a discursive utterance. Specifically, the vario=
us back-counting, reciprocal, and reflexive constructs in selma'o KOhA igno=
re the utterances at =93higher=94 metalinguistic levels in determining thei=
r referent. It is possible, and sometimes necessary, to refer to lower meta=
linguistic levels. For example, the English =93he said=94 in a conversation=
 is metalinguistic. For this purpose, quotations are considered to be at a =
lower metalinguistic level than the surrounding context (a quoted text cann=
ot refer to the statements of the one who quotes it), whereas parenthetical=
 remarks are considered to be at a higher level than the context.
+<p>
+Lojban works differently from English in that the =93he said=94 can be mar=
ked instead of the quotation. In Lojban, you can say:
+<p>
+<ex "said John">  XE "said John: example=93  <pre><a name=3De12d6>12.6)	la=
 djan. cusku lu mi klama le zarci li'u
+	John expresses =93I go to-the store=94.
+</pre>which literally claims that John uttered the quoted text. If the cen=
tral claim is that John made the utterance, as is likely in conversation, t=
his style is the most sensible. However, in written text which quotes a con=
versation, you don't want the =93he said=94 or =93she said=94 to be conside=
red part of the conversation. If unmarked, it could mess up the anaphora co=
unting. Instead, you can use:
+<p>
+<pre><a name=3De12d7>12.7)	lu mi klama le zarci seisa'a la djan. cusku be =
dei li'u
+	=93I go to-the store (John expresses this-sentence)=94.
+	=93I go to the store=94, said John.
+</pre><p>And of course other orders are possible:
+<p>
+<pre><a name=3De12d8>12.8)	lu seisa'a la djan. cusku be dei mi klama le za=
rci
+	John said, =93I go to the store=94.
+<a name=3De12d9>12.9)	lu mi klama seisa'a la djan cusku le zarci
+	=93I go=94, John said, =93to the store=94.
+</pre><p>Note the =93sa'a=94 following each =93sei=94, marking the =93sei=
=94 and its attached bridi as an editorial insert, not part of the quotatio=
n. In a more relaxed style, these =93sa'a=94 cmavo would probably be droppe=
d.
+<p>
+<lx "se'u">  XE "se'u"   <lx "SEhU">  XE "SEhU selma=92o"   The elidable t=
erminator for =93sei=94 is =93se'u=94 (of selma'o SEhU); it is rarely neede=
d, except to separate a selbri within the =93sei=94 comment from an immedia=
tely following selbri (or component) outside the comment.
+<p>
+Erasure: SI, SA, SU</h3>
+<p><cx "erasure">  XE "erasure"   <a name=3Ds13><h3>
+The following cmavo are discussed in this section:
+<p>
+<pre> 	si	SI	erase word
+	sa	SA	erase phrase
+	su	SU	erase discourse
+</pre><lx "si">  XE "si"   <lx "SI">  XE "SI selma=92o"   <cx "erasure, wo=
rd">  XE "erasure: word"   The cmavo =93si=94 (of selma'o SI) is a metaling=
uistic operator that erases the preceding word, as if it had never been spo=
ken:
+<p>
+<pre><a name=3De13d1>13.1)	ti gerku si mlatu
+	This is-a-dog, er, is-a-cat.
+</pre>means the same thing as =93ti mlatu=94. Multiple =93si=94 cmavo in s=
uccession erase the appropriate number of words:
+<p>
+<pre><a name=3De13d2>13.2)	ta blanu zdani si si xekri zdani
+	That is-a-blue house, er, er, is-a-black house.
+</pre><cx "erasure, zo">  XE "erasure: zo"   In order to erase the word =
=93zo=94, it is necessary to use three =93si=94 cmavo in a row:
+<p>
+<pre><a name=3De13d3>13.3)	zo .bab. se cmene zo si si si la bab.
+	The-word =93Bob=94 is-the-name-of the word =93si=94, er, er, Bob.
+</pre><p>The first use of =93si=94 does not erase anything, but completes =
the =93zo=94 quotation. Two more =93si=94 cmavo are then necessary to erase=
 the first =93si=94 and the =93zo=94.
+<p>
+Incorrect names can likewise cause trouble with =93si=94:
+<p>
+<cx "erasure, names">  XE "erasure: names"   <pre><a name=3De13d4>13.4)	mi=
 tavla fo la .esperanto si si .esperanton.
+	I talk in-language that-named =93and=94 =93speranto=94, er, er, Esperanto=
.
+</pre><p>The Lojbanized spelling =93.esperanto=94 breaks up, as a conseque=
nce of the Lojban morphology rules (see <a href=3Dchap4.html>Chapter 4</a>)=
 into two Lojban words, the cmavo =93.e=94 and the undefined fu'ivla =93spe=
ranto=94. Therefore, two =93si=94 cmavo are needed to erase them. Of course=
, =93.e speranto=94 is not grammatical after =93la=94, but recognition of =
=93si=94 is done before grammatical analysis.
+<p>
+<cx "erasure, quotes">  XE "erasure: quotes"   Even more messy is the resu=
lt of an incorrect =93zoi=94:
+<p>
+<pre><a name=3De13d5>13.5)	mi cusku zoi fy. gy. .fy. si si si si zo .djan
+	I express [foreign] [quote] =93gy=94 [unquote], er, er, er, er, =93John=
=94.
+</pre><p>In <a href=3D#e13d5>Example 13.5</a>, the first =93fy=94 is taken=
 to be the delimiting word. The next word must be different from the delimi=
ting word, and =93gy.=94, the Lojban name for the letter =93g=94, was chose=
n arbitrarily. Then the delimiting word must be repeated. For purposes of =
=93si=94 erasure, the entire quoted text is taken to be a word, so four wor=
ds have been uttered, and four more =93si=94 cmavo are needed to erase them=
 altogether. Similarly, a stray =93lo'u=94 quotation mark must be erased wi=
th =93fy. le'u si si si=94, by completing the quotation and then erasing it=
 all with three =93si=94 cmavo.
+<p>
+What if less than the entire =93zo=94 or =93zoi=94 construct is erased? Th=
e result is something which has a loose =93zo=94 or =93zoi=94 in it, withou=
t its expected sequels, and which is incurably ungrammatical. Thus, to eras=
e just the word quoted by =93zo=94, it turns out to be necessary to erase t=
he =93zo=94 as well:
+<p>
+<pre><a name=3De13d6>13.6)	mi se cmene zo .djan. si si zo .djordj.
+	I am-named-by the-word =93John=94, er, er, the-word =93George=94.
+</pre><p>The parser will reject =93zo .djan. si .djordj.=94, because in th=
at context =93djordj.=94 is a name (of selma'o CMENE) rather than a quoted =
word.
+<p>
+Note: The current machine parser does not implement =93si=94 erasure.
+<p>
+<cx "starting marker">  XE "starting marker"   <cx "precise erasures">  XE=
 "precise erasures"   <lx "sa">  XE "sa"   <lx "SA">  XE "SA selma=92o"   <=
cx "erasure, multiple word">  XE "erasure: multiple word"   As the above ex=
amples plainly show, precise erasures with =93si=94 can be extremely hard t=
o get right. Therefore, the cmavo =93sa=94 (of selma'o SA) is provided for =
erasing more than one word. The cmavo following =93sa=94 should be the star=
ting marker of some grammatical construct. The effect of the =93sa=94 is to=
 erase back to and including the last starting marker of the same kind. For=
 example:
+<p>
+<pre><a name=3De13d7>13.7)	mi viska le sa  .i mi cusku zo .djan.
+	I see the=A0<dots>=85 </dots>I say the-word =93John=94.
+</pre><p>Since the word following =93sa=94 is =93.i=94, the sentence separ=
ator, its effect is to erase the preceding sentence. So <a href=3D#e13d7>Ex=
ample 13.7 </a>is equivalent to:
+<p>
+<pre><a name=3De13d8>13.8)	mi cusku zo .djan.
+</pre><p>Another example, erasing a partial description rather than a part=
ial sentence:
+<p>
+<pre><a name=3De13d9>13.9)	mi viska le blanu zdan. sa le xekri zdani
+	I see the blue hou=A0<dots>=85</dots>the black house.
+</pre><p>In <a href=3D#e13d9>Example 13.9</a>, =93le blanu zdan.=94 is ung=
rammatical, but clearly reflects the speaker's original intention to say =
=93le blanu zdani=94. However, the =93zdani=94 was cut off before the end a=
nd changed into a name. The entire ungrammatical =93le=94 construct is eras=
ed and replaced by =93le xekri zdani=94.
+
+<p>
+Note: The current machine parser does not implement =93sa=94 erasure. Gett=
ing =93sa=94 right is even more difficult (for a computer) than getting =93=
si=94 right, as the behavior of =93si=94 is defined in terms of words rathe=
r than in terms of grammatical constructs (possibly incorrect ones) and wor=
ds are conceptually simpler entities. On the other hand, =93sa=94 is genera=
lly easier for human beings, because the rules for using it correctly are l=
ess finicky.
+<p>
+<cx "multiple speakers">  XE "multiple speakers"   <lx "su">  XE "su"   <l=
x "SU">  XE "SU selma=92o"   <cx "erasure, total">  XE "erasure: total"   T=
he cmavo =93su=94 (of selma'o SU) is yet another metalinguistic operator th=
at erases the entire text. However, if the text involves multiple speakers,=
 then =93su=94 will only erase the remarks made by the one who said it, unl=
ess that speaker has said nothing. Therefore =93susu=94 is needed to eradic=
ate a whole discussion in conversation.
+<p>
+Note: The current machine parser does not implement either =93su=94 or =93=
susu=94 erasure.
+<p>
+Hesitation: Y</h3>
+<p><cx "hesitation">  XE "hesitation"   <a name=3Ds14><h3>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	.y.	Y	hesitation noise
+</pre><lx ".y">  XE "y"   <lx "Y">  XE "Y selma=92o"   Speakers often need=
 to hesitate to think of what to say next or for some extra-linguistic reas=
on. There are two ways to hesitate in Lojban: to pause between words (that =
is, to say nothing) or to use the cmavo =93.y.=94 (of selma'o Y). This rese=
mbles in sound the English hesitation noise written =93uh=94 (or =93er=94),=
 but differs from it in the requirement for pauses before and after. Unlike=
 a long pause, it cannot be mistaken for having nothing more to say: it hol=
ds the floor for the speaker. Since vowel length is not significant in Lojb=
an, the =93y=94 sound can be dragged out for as long as necessary. Furtherm=
ore, the sound can be repeated, provided the required pauses are respected.
+<p>
+<cx "hesitation sound">  XE "hesitation sound"   Since the hesitation soun=
d in English is outside the formal language, English-speakers may question =
the need for a formal cmavo. Speakers of other languages, however, often he=
sitate by saying (or, if necessary, repeating) a word (=93este=94 in some d=
ialects of Spanish, roughly meaning =93that is=94), and Lojban's audio-visu=
al isomorphism requires a written representation of all meaningful spoken b=
ehavior. Of course, =93.y.=94 has no grammatical significance: it can appea=
r anywhere at all in a Lojban sentence except in the middle of a word.
+<p>
+No more to say: FAhO</h3>
+<p><cx "text, end-marker">  XE "text: end-marker"   <a name=3Ds15><h3>
+The following cmavo is discussed in this section:
+<p>
+<pre> 	fa'o	FAhO	end of text
+</pre><lx "fa'o">  XE "fa'o"   <lx "FAhO">  XE "FAhO selma=92o"   <cx "unc=
onditional signal">  XE "unconditional signal"   <cx "computer interaction"=
>  XE "computer interaction"   <cx "end of file">  XE "end of file"   <cx "=
yielding the floor">  XE "yielding the floor"   The cmavo =93fa'o=94 (of se=
lma'o FAhO) is the usually omitted marker for the end of a text; it can be =
used in computer interaction to indicate the end of input or output, or for=
 explicitly giving up the floor during a discussion. It is outside the regu=
lar grammar, and the machine parser takes it as an unconditional signal to =
stop parsing unless it is quoted with =93zo=94 or with =93lo'u=A0<dots>=85<=
/dots>le'u=94. In particular, it is not used at the end of subordinate text=
s quoted with =93lu=A0<dots>=85</dots>li'u=94 or parenthesized with =93to=
=A0<dots>=85</dots>toi=94.
+
+<p>
+List of cmavo interactions</h3>
+<p><cx "cmavo, interaction list">  XE "cmavo: interaction list"   <cx "int=
eraction list, cmavo">  XE "interaction list: cmavo"   <a name=3Ds16><h3>
+The following list gives the cmavo and selma'o that are recognized by the =
earliest stages of the parser, and specifies exactly which of them interact=
 with which others. All of the cmavo are at least mentioned in this chapter=
. The cmavo are written in lower case, and the selma'o in UPPER CASE.
+<p>
+=93zo=94 quotes the following word, no matter what it is.
+<p>
+=93si=94 erases the preceding word unless it is a =93zo=94.
+<p>
+=93sa=94 erases the preceding word and other words, unless the preceding w=
ord is a =93zo=94.
+<p>
+=93su=94 is the same as =93sa=94, but erases more words.
+<p>
+=93lo'u=94 quotes all following words up to a =93le'u=94 (but not a =93zo =
le'u=94).
+<p>
+=93le'u=94 is ungrammatical except at the end of a =93lo'u quotation.
+<p>
+ZOI cmavo use the following word as a delimiting word, no matter what it i=
s, but using =93le'u=94 may create difficulties.
+<p>
+=93zei=94 combines the preceding and the following word into a lujvo, but =
does not affect =93zo=94, =93si=94, =93sa=94, =93su=94, =93lo'u=94, ZOI cma=
vo, =93fa'o=94, and =93zei=94.
+<p>
+BAhE cmavo mark the following word, unless it is =93si=94, =93sa=94, or =
=93su=94, or unless it is preceded by =93zo=94. Multiple BAhE cmavo may be =
used in succession.
+<p>
+=93bu=94 makes the preceding word into a lerfu word, except for =93zo=94, =
=93si=94, =93sa=94, =93su=94, =93lo'u=94, ZOI cmavo, =93fa'o=94, =93zei=94,=
 BAhE cmavo, and =93bu=94. Multiple =93bu=94 cmavo may be used in successio=
n.
+<p>
+UI and CAI cmavo mark the previous word, except for =93zo=94, =93si=94, =
=93sa=94, =93su=94, =93lo'u=94, ZOI, =93fa'o=94, =93zei=94, BAhE cmavo, and=
 =93bu=94. Multiple UI cmavo may be used in succession. A following =93nai=
=94 is made part of the UI.
+<p>
+=93.y.=94, =93da'o=94, =93fu'e=94, and =93fu'o=94 are the same as UI, but =
do not absorb a following =93nai=94.
+
+<p>
+List of Elidable Terminators</h3>
+<p><cx "elidable terminators, list">  XE "elidable terminators: list"   <a=
 name=3Ds17><h3>
+The following list shows all the elidable terminators of Lojban. The first=
 column is the terminator, the second column is the selma'o that starts the=
 corresponding construction, and the third column states what kinds of gram=
matical constructs are terminated. Each terminator is the only cmavo of its=
 selma'o, which naturally has the same name as the cmavo.
+<p>
+<pre> 	be'o	BE	sumti attached to a tanru unit
+	boi	PA/BY	number or lerfu string
+	do'u	COI/DOI	vocative phrases
+	fe'u	FIhO	ad-hoc modal tags
+	ge'u	GOI	relative phrases
+	kei	NU	abstraction bridi
+	ke'e	KE	groups of various kinds
+	ku	LE/LA	description sumti
+	ku'e	PEhO	forethought mekso
+	ku'o	NOI	relative clauses
+	li'u	LU	quotations
+	lo'o	LI	number sumti
+	lu'u	LAhE/NAhE+BO	sumti qualifiers
+	me'u	ME	tanru units formed from sumti
+	nu'u	NUhI	forethought termsets
+	se'u	SEI/SOI	metalinguistic insertions
+	te'u	various	mekso conversion constructs
+	toi	TO	parenthetical remarks
+	tu'u	TUhE	multiple sentences or paragraphs
+	vau	(none)	simple bridi or bridi-tails
+	ve'o	VEI	mekso parentheses
+
+</body></html>
+
+<h2>Chapter 20<br>
+A Catalogue of selma'o</h2>
+<h6>$Revision: 4.1 $<br>
+mkhtml: 1.1</h6></p></center>
+<p>
+<cx "selma'o, cross-reference list of, selma'o catalog">  XE "selma'o: cro=
ss-reference list of"   The following paragraphs list all the selma'o of Lo=
jban, with a brief explanation of what each one is about, and reference to =
the chapter number where each is explained more fully. As usual, all selma'=
o names are given in capital letters (with =93h=94 serving as the capital o=
f =93 ' =94) and are the names of a representative cmavo, often the most im=
portant or the first in alphabetical order. One example is given of each se=
lma'o: for selma'o which have several uses, the most common use is shown.
+<p>
+<lx "A, selma'o catalog">  XE =93A selma=92o"   <cx "connection, of sumti,=
 selma'o catalog">  XE "connection: of sumti"   selma'o A (<a href=3Dchap14=
.html>Chapter 14</a>)
+<p>
+Specifies a logical connection (e.g. =93and=94, =93or=94, =93if=94), usual=
ly between sumti.
+<p>
+<dl compact><dt>	<dd>la djan. .a la djein. klama le zarci
+	John and/or Jane goes to the store.
+</dl><p>Also used to create vowel lerfu words when followed with =93bu=94.
+<p>
+<lx "BAI, selma'o catalog">  XE "BAI selma=92o"   <cx "sumti place, additi=
onal, selma'o catalog">  XE "sumti place: additional"   selma'o BAI (<a hre=
f=3Dchap9.html>Chapter 9</a>)
+<p>
+May be prefixed to a sumti to specify an additional place, not otherwise p=
resent in the place structure of the selbri, and derived from a single plac=
e of some other selbri.
+<p>
+<dl compact><dt>	<dd>mi tavla bau la lojban.
+	I speak in-language Lojban.
+</dl><lx "BAhE, selma'o catalog">  XE "BAhE selma=92o"   <cx "emphasis, ma=
rking, selma'o catalog">  XE "emphasis: marking"   <cx "nonce word, marking=
, selma'o catalog">  XE "nonce word: marking"   selma'o BAhE (<a href=3Dcha=
p19.html>Chapter 19</a>)
+<p>
+Emphasizes the next single word, or marks it as a nonce word (one invented=
 for the occasion.
+<p>
+<dl compact><dt>	<dd>la ba'e .djordj. klama le zarci
+	<i>George </i>goes to the store.
+	It is George who goes to the store.
+</dl><lx "BE, selma'o catalog">  XE "BE selma=92o"   <cx "tanru unit, fill=
ing in places of, selma'o catalog">  XE "tanru unit: filling in places of" =
  selma'o BE (<a href=3Dchap5.html>Chapter 5</a>)
+<p>
+Attaches sumti which fill the place structure of a single unit making up a=
 tanru. Unless otherwise indicated, the sumti fill the x2, x3, and successi=
ve places in that order. BE is most useful in descriptions formed with LE. =
See BEI, BEhO.
+<p>
+<dl compact><dt>	<dd>mi klama be ta troci
+	I am-a-(goer to-that) type-of-trier.
+	I try to go to that place.
+</dl><cx "tanru unit, filling in places of, selma'o catalog">  XE "tanru u=
nit: filling in places of"   <lx "BEI, selma'o catalog">  XE "BEI selma=92o=
"   selma'o BEI (<a href=3Dchap5.html>Chapter 5</a>)
+<p>
+Separates multiple sumti attached by BE to a tanru unit.
+<p>
+<dl compact><dt>	<dd>mi klama be le zarci bei le zdani be'o troci
+	I am-a-(goer to-the store from-the home) type-of-trier.
+	I try to go from the home to the market.
+</dl><lx "BEhO, selma'o catalog">  XE "BEhO selma=92o"   <cx "BE, terminat=
or for, selma'o catalog">  XE "BE selma=92o: terminator for"   selma'o BEhO=
 (<a href=3Dchap5.html>Chapter 5</a>)
+<p>
+Elidable terminator for BE. Terminates sumti that are attached to a tanru =
unit.
+<p>
+<dl compact><dt>	<dd>mi klama be le zarci be'o troci
+	I am-a-(goer to-the market) type-of-trier.
+	I try to go to the market.
+</dl><lx "BIhE, selma'o catalog">  XE "BIhE selma=92o"   <cx "operator pri=
ority, selma'o catalog">  XE "operator priority"   selma'o BIhE (<a href=3D=
chap18.html>Chapter 18</a>)
+<p>
+Prefixed to a mathematical operator to mark it as higher priority than oth=
er mathematical operators, binding its operands more closely.
+<p>
+<dl compact><dt>	<dd>li ci bi'e pi'u vo su'i mu du li paze
+	The-number 3 [priority] times 4 plus 5 equals the-number 17.
+	3 ( 4 + 5 =3D 17
+</dl><lx "BIhI, selma'o catalog">  XE "BIhI selma=92o"   <cx "intervals, f=
orming, selma'o catalog">  XE "interval: forming"   selma'o BIhI (<a href=
=3Dchap14.html>Chapter 14</a>)
+<p>
+Joins sumti or tanru units (as well as some other things) to form interval=
s. See GAhO.
+<p>
+<dl compact><dt>	<dd>mi ca sanli la drezdn. bi'i la frankfurt.
+	I [present] stand-on-surface Dresden [interval] Frankfurt.
+	I am standing between Dresden and Frankfurt.
+</dl><lx "BO, selma'o catalog">  XE "BO selma=92o"   <cx "close-binding, s=
elma'o catalog">  XE "close-binding"   selma'o BO (<a href=3Dchap5.html>Cha=
pter 5</a>, <a href=3Dchap15.html>Chapter 15</a>, <a href=3Dchap18.html>Cha=
pter 18</a>)
+<p>
+Joins tanru units, binding them together closely. Also used to bind logica=
lly or non-logically connected phrases, sentences, etc. BO is always high p=
recedence and right-grouping.
+<p>
+<dl compact><dt>	<dd>ta cmalu nixli bo ckule
+	That is-a-small type-of (girl type-of school).
+	That is a small school for girls.
+</dl><lx "BOI, selma'o catalog">  XE "BOI selma=92o"   <cx "PA, terminator=
 for, selma'o catalog">  XE "PA selma=92o: terminator for"   <cx "BY, termi=
nator for, selma'o catalog">  XE "BY selma=92o: terminator for"   selma'o B=
OI (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Elidable terminator for PA or BY. Used to terminate a number (string of nu=
meric cmavo) or lerfu string (string of letter words) when another string i=
mmediately follows.
+<p>
+<dl compact><dt>	<dd>li re du li vu'u voboi re
+	The-number two equals the-number the-difference-of four-and two.
+</dl><lx "BU, selma'o catalog">  XE "BU selma=92o"   <cx "letter, making a=
 word into, selma'o catalog">  XE "letter: making a word into"   selma'o BU=
 (<a href=3Dchap17.html>Chapter 17</a>)
+<p>
+A suffix which can be attached to any word, typically a word representing =
a letter of the alphabet or else a name, to make a word for a symbol or a d=
ifferent letter of the alphabet. In particular, attached to single-vowel cm=
avo to make words for vowel letters.
+<p>
+<dl compact><dt>	<dd>.abu .ebu .ibu .obu .ubu .ybu
+	a, e, i, o, u, y.
+
+</dl><lx "BY, selma'o catalog">  XE "BY selma=92o"   <cx "letters, selma'o=
 catalog">  XE "letters"   selma'o BY (<a href=3Dchap17.html>Chapter 17</a>=
)
+<p>
+Words representing the letters of the Lojban alphabet, plus various shift =
words which alter the interpretation of other letter words. Terminated by B=
OI.
+<p>
+<dl compact><dt>	<dd>.abu tavla .by le la .ibymym. skami
+	A talks-to B about-the of-IBM computers.
+	A talks to B about IBM computers.
+</dl><lx "CAI, selma'o catalog">  XE "CAI selma=92o"   <cx "emotion, marki=
ng intensity of, selma'o catalog">  XE "emotions: marking intensity of"   s=
elma'o CAI (<a href=3Dchap13.html>Chapter 13</a>)
+<p>
+Indicates the intensity of an emotion: maximum, strong, weak, or not at al=
l. Typically follows another particle which specifies the emotion.
+<p>
+<dl compact><dt>	<dd>.ei cai mi klama le zarci
+	[Obligation!] [Intense!] I go-to the market.
+	I must go to the market.
+</dl><lx "CAhA, selma'o catalog">  XE "CAhA selma=92o"   <cx "actuality, m=
arking, selma'o catalog">  XE "actuality: marking"   selma'o CAhA (<a href=
=3Dchap10.html>Chapter 10</a>)
+<p>
+Specifies whether a bridi refers to an actual fact, a potential (achieved =
or not), or merely an innate capability.
+<p>
+<dl compact><dt>	<dd>ro datka ka'e flulimna
+	all ducks [capability] are-float-swimmers
+	All ducks have the capability of swimming by floating.
+</dl><lx "CEI, selma'o catalog">  XE "CEI selma=92o"   <cx "selbri assignm=
ent, selma'o catalog">  XE "selbri assignment"   selma'o CEI (<a href=3Dcha=
p7.html>Chapter 7</a>)
+<p>
+Assigns a selbri definition to one of the five pro-bridi gismu: =93broda=
=94, =93brode=94, =93brodi=94, =93brodo=94, or =93brodu=94, for later use.
+<p>
+<dl compact><dt>	<dd>ti slasi je mlatu bo cidja lante gacri cei broda
+		.i le crino broda cu barda  .i le xunre broda cu cmalu
+	This is a plastic cat-food can cover, or thingy.
+		The green thingy is large.  The red thingy is small.
+</dl><lx "CEhE, selma'o catalog">  XE "CEhE selma=92o"   <cx "termsets, se=
lma'o catalog">  XE "termsets"   selma'o CEhE (<a href=3Dchap14.html>Chapte=
r 14</a>, <a href=3Dchap16.html>Chapter 16</a>)
+<p>
+Joins multiple terms into a termset. Termsets are used to associate severa=
l terms for logical connectives, for equal quantifier scope, or for special=
 constructs in tenses.
+<p>
+<dl compact><dt>	<dd>mi ce'e do pe'e je la djan. ce'e la djeimyz. cu pendo
+	I [,] you [joint] and John [,] James are-friends-of.
+	I am a friend of you, and John are a friend of James.
+</dl><lx "CO, selma'o catalog">  XE "CO selma=92o"   <cx "inverting a tanr=
u, selma'o catalog">  XE "inverting a tanru"   selma'o CO (<a href=3Dchap5.=
html>Chapter 5</a>)
+<p>
+When inserted between the components of a tanru, inverts it, so that the f=
ollowing tanru unit modifies the previous one.
+<p>
+<dl compact><dt>	<dd>mi troci co klama le zarci le zdani
+	I am-a-trier of-type (goer to-the market from-the house).
+	I try to go to the market from the house.
+</dl><lx "COI, selma'o catalog">  XE "COI selma=92o"   <cx "vocative, selm=
a'o catalog">  XE "vocative"   selma'o COI (<a href=3Dchap6.html>Chapter 6<=
/a>, <a href=3Dchap13.html>Chapter 13</a>)
+<p>
+When prefixed to a name, description, or sumti, produces a vocative: a phr=
ase which indicates who is being spoken to (or who is speaking). Vocatives =
are used in conversational protocols, including greeting, farewell, and rad=
io communication. Terminated by DOhU. See DOI.
+<p>
+<dl compact><dt>	<dd>coi .djan.
+	Greetings, John.
+</dl><lx "CU, selma'o catalog">  XE "CU selma=92o"   <cx "selbri separator=
, selma'o catalog">  XE "selbri separator"   selma'o CU (<a href=3Dchap9.ht=
ml>Chapter 9</a>)
+<p>
+Separates the selbri of a bridi from any sumti which precede it. Never str=
ictly necessary, but often useful to eliminate various elidable terminators=
.
+<p>
+<dl compact><dt>	<dd>le gerku cu klama le zarci
+	The dog goes to-the store.
+</dl><lx "CUhE, selma'o catalog">  XE "CUhE selma=92o"   <cx "question, mo=
dal, selma'o catalog">  XE =93questions: modal"   selma'o CUhE (<a href=3Dc=
hap10.html>Chapter 10</a>)
+<p>
+Forms a question which asks when, where, or in what mode the rest of the b=
ridi is true. See PU, CAhA, TAhE, and BAI.
+<p>
+<dl compact><dt>	<dd>do cu'e klama le zarci
+	You [When/Where?] go to-the store?
+	When are you going to the store?
+</dl><lx "DAhO, selma'o catalog">  XE "DAhO selma=92o"   <cx "cancellation=
 of sumti assignment, selma'o catalog">  XE "cancellation of sumti assignme=
nt"   <cx "sumti assignment, cancellation of, selma'o catalog">  XE "sumti =
assignment: cancellation of"   selma'o DAhO (<a href=3Dchap7.html>Chapter 7=
</a>)
+<p>
+Cancels the assigned significance of all sumti cmavo (of selma'o KOhA) and=
 bridi cmavo (of selma'o GOhA).
+<p>
+<lx "DOI, selma'o catalog">  XE "DOI selma=92o"   <cx "vocative indicator,=
 selma'o catalog">  XE "vocative indicator"   selma'o DOI (<a href=3Dchap13=
.html>Chapter 13</a>)
+<p>
+The non-specific vocative indicator. May be used with or without COI. No p=
ause is required between =93doi=94 and a following name. See DOhU.
+<p>
+<dl compact><dt>	<dd>doi frank. mi tavla do
+	O Frank, I speak-to you.
+	Frank, I'm talking to you.
+</dl><lx "DOhU, selma'o catalog">  XE "DOhU selma=92o"   <cx "COI, termina=
tor for, selma'o catalog">  XE "COI selma=92o: terminator for"   <cx "DOI, =
terminator for, selma'o catalog">  XE "DOI selma=92o: terminator for"   sel=
ma'o DOhU (<a href=3Dchap13.html>Chapter 13</a>)
+<p>
+Elidable terminator for COI or DOI. Signals the end of a vocative.
+<p>
+<dl compact><dt>	<dd>coi do'u
+	Greetings [terminator]
+	Greetings, O unspecified one!
+
+</dl><lx "FA, selma'o catalog">  XE "FA selma=92o"   <cx "place number, sp=
ecifying, selma'o catalog">  XE "place number: specifying"   selma'o FA (<a=
 href=3Dchap9.html>Chapter 9</a>)
+<p>
+Prefix for a sumti, indicating which numbered place in the place structure=
 the sumti belongs in; overrides word order.
+<p>
+<dl compact><dt>	<dd>fa mi cu klama fi la .atlantas. fe la bastn. fo le da=
rgu fu le karce
+	x1=3D I   go   x3=3D Atlanta   x2=3D Boston   x4=3D the road   x5=3D the =
car.
+	I go from Atlanta to Boston via the road using the car.
+</dl><lx "FAhA, selma'o catalog">  XE "FAhA selma=92o"   <cx "direction, s=
patial, selma'o catalog">  XE "direction: spatial"   selma'o FAhA (<a href=
=3Dchap10.html>Chapter 10</a>)
+<p>
+Specifies the direction in which, or toward which (when marked with MOhI) =
or along which (when prefixed by VEhA or VIhA) the action of the bridi take=
s place.
+<p>
+<dl compact><dt>	<dd>le nanmu zu'a batci le gerku
+	The man [left] bites the dog.
+	To my left, the man bites the dog.
+</dl><lx "FAhO, selma'o catalog">  XE "FAhO selma=92o"   <cx "end of text =
indicator, selma'o catalog">  XE "end of text indicator"   selma'o FAhO (<a=
 href=3Dchap19.html>Chapter 19</a>)
+<p>
+A mechanical signal, outside the grammar, indicating that there is no more=
 text. Useful in talking to computers.
+<p>
+<lx "FEhE, selma'o catalog">  XE "FEhE selma=92o"   <cx "space indicator f=
or interval modifiers, selma'o catalog">  XE "space indicator for interval =
modifiers"   selma'o FEhE (<a href=3Dchap10.html>Chapter 10</a>)
+<p>
+Indicates that the following interval modifier (using TAhE, ROI, or ZAhO) =
refers to space rather than time.
+<p>
+<dl compact><dt>	<dd>ko vi'i fe'e di'i sombo le gurni
+	You-imperative [1-dimensional] [space] [regularly] sow the grain.
+	Sow the grain in a line and evenly!
+</dl><lx "FEhU, selma'o catalog">  XE "FEhU selma=92o"   <cx "FIhO, termin=
ator for, selma'o catalog">  XE "FIhO selma=92o: terminator for"   selma'o =
FEhU (<a href=3Dchap9.html>Chapter 9</a>)
+<p>
+Elidable terminator for FIhO. Indicates the end of an ad hoc modal tag: th=
e tagged sumti immediately follows.
+<p>
+<dl compact><dt>	<dd>mi viska do fi'o kanla [fe'u] le zunle
+	I see you [modal] eye:  the left-thing
+	I see you with my left eye.
+</dl><lx "FIhO, selma'o catalog">  XE "FIhO selma=92o"   <cx "selbri to mo=
dal converter, selma'o catalog">  XE "selbri to modal converter"   <cx "con=
version, selbri to modal, selma'o catalog">  XE "conversion of selbri to mo=
dal"   selma'o FIhO (<a href=3Dchap9.html>Chapter 9</a>)
+<p>
+When placed before a selbri, transforms the selbri into a modal tag, gramm=
atically and semantically equivalent to a member of selma'o BAI. Terminated=
 by FEhU.
+<p>
+<dl compact><dt>	<dd>mi viska do fi'o kanla le zunle
+	I see you with eye the left-thing.
+	I see you with my left eye.
+
+</dl><lx "FOI, selma'o catalog">  XE "FOI selma=92o"   <cx "TEI, terminato=
r for, selma'o catalog">  XE "TEI selma=92o: terminator for"   selma'o FOI =
(<a href=3Dchap17.html>Chapter 17</a>)
+<p>
+Signals the end of a compound alphabet letter word that begins with TEI. N=
ot an elidable terminator.
+<p>
+<dl compact><dt>	<dd>tei .ebu .akut. bu foi
+	(=93e=94 =93acute=94 )
+	the letter =93e=94 with an acute accent
+</dl><lx "FUhA, selma'o catalog">  XE "FUhA selma=92o"   <cx "reverse Poli=
sh indicator, selma'o catalog">  XE "reverse Polish notation: indicator"   =
selma'o FUhA (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Indicates that the following mathematical expression is to be interpreted =
as reverse Polish (RP), a mode in which mathematical operators follow their=
 operands.
+<p>
+<dl compact><dt>	<dd>li fu'a reboi re[boi] su'i du li vo
+	The-number [RP!] two, two, plus equals the-number four.
+	2 + 2 =3D 4
+</dl><lx "FUhE, selma'o catalog">  XE "FUhE selma=92o"   <cx "UI, extendin=
g the scope of, selma'o catalog">  XE "UI selma=92o: extending the scope of=
"   selma'o FUhE (<a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Indicates that the following indicator(s) of selma'o UI affect not the pre=
ceding word, as usual, but rather all following words until a FUhO.
+<p>
+<dl compact><dt>	<dd>mi viska le fu'e .ia blanu zdani fu'o ponse
+	I see the [start] [belief] blue house [end] possessor.
+	I see the owner of a blue house, or what I believe to be one.
+</dl><lx "FUhO, selma'o catalog">  XE "FUhO selma=92o"   <cx "cancellation=
 of indicators, selma'o catalog">  XE "cancellation of indicators"   <cx "i=
ndicators, cancellation of, selma'o catalog">  XE "indicators: cancellation=
 of"   selma'o FUhO (<a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Cancels all indicators of selma'o UI which are in effect.
+<p>
+<dl compact><dt>	<dd>mi viska le fu'e .ia blanu zdani fu'o ponse
+	I see the [start] [belief] blue house [end] possessor.
+	I see the owner of what I believe to be a blue house.
+</dl><lx "GA, selma'o catalog">  XE "GA selma=92o"   <cx "connection, fore=
thought, selma'o catalog">  XE "connection: forethought"   selma'o GA (<a h=
ref=3Dchap14.html>Chapter 14</a>)
+<p>
+Indicates the beginning of two logically connected sumti, bridi-tails, or =
various other things. Logical connections include =93both=A0<dots>=85</dots=
>and=94, =93either=A0<dots>=85</dots>or=94, =93if=A0<dots>=85</dots>then=94=
, and so on. See GI.
+<p>
+<dl compact><dt>	<dd>ga la djan. nanmu gi la djeimyz. ninmu
+	Either John is a man or James is a woman (or both).
+</dl><lx "GAhO, selma'o catalog">  XE "GAhO selma=92o"   <cx "interval, op=
en/closed specification, selma'o catalog">  XE "interval: open/closed speci=
fication"   selma'o GAhO (<a href=3Dchap14.html>Chapter 14</a>)
+<p>
+Specifies whether an interval specified by BIhI includes or excludes its e=
ndpoints. Used in pairs before and after the BIhI cmavo, to specify the nat=
ure of both the left- and the right-hand endpoints.
+<p>
+<dl compact><dt>	<dd>mi ca sanli la drezdn. ga'o bi'i ga'o la frankfurt.
+	I [present] stand Dresden [inclusive] [interval] [inclusive] Frankfurt.
+	I am standing between Dresden and Frankfurt, inclusive of both.
+</dl><lx "GEhU, selma'o catalog">  XE "GEhU selma=92o"   <cx "GOI, termina=
tor for, selma'o catalog">  XE "GOI selma=92o: terminator for"   selma'o GE=
hU (<a href=3Dchap8.html>Chapter 8</a>)
+<p>
+Elidable terminator for GOI. Marks the end of a relative phrase. See KUhO.
+<p>
+<dl compact><dt>	<dd>la djan. goi ko'a ge'u blanu
+	John (referred to as it-1) is-blue.
+</dl><lx "GI, selma'o catalog">  XE "GI selma=92o"   <cx "connection, fore=
thought separator, selma'o catalog">  XE "connection: forethought separator=
"   selma'o GI (<a href=3Dchap14.html>Chapter 14</a>)
+<p>
+Separates two logically or non-logically connected sumti, tanru units, bri=
di-tails, or other things, when the prefix is a forethought connective invo=
lving GA, GUhA, or JOI.
+<p>
+<dl compact><dt>	<dd>ge la djan. nanmu gi la djeimyz. ninmu
+	(It is true that) both John is a man and James is a woman.
+</dl><lx "GIhA, selma'o catalog">  XE "GIhA selma=92o"   <cx "connection, =
of bridi-tails, selma'o catalog">  XE "connection: of bridi-tails"   <cx "b=
ridi-tail, definition, selma'o catalog">  XE "bridi-tail: definition"   sel=
ma'o GIhA (<a href=3Dchap14.html>Chapter 14</a>)
+<p>
+Specifies a logical connective (e.g. =93and=94, =93or=94, =93if=94) betwee=
n two bridi-tails: a bridi-tail is a selbri with any associated following s=
umti, but not including any preceding sumti.
+<p>
+<dl compact><dt>	<dd>mi klama le zarci gi'e nelci la djan.
+	I go-to the market and like John.
+</dl><lx "GOI, selma'o catalog">  XE "GOI selma=92o"   <cx "relative phras=
e marker, selma'o catalog">  XE "relative phrase marker"   selma'o GOI (<a =
href=3Dchap8.html>Chapter 8</a>)
+<p>
+Specifies the beginning of a relative phrase, which associates a subordina=
te sumti (following) to another sumti (preceding). Terminated by GEhU. See =
NOI.
+<p>
+<dl compact><dt>	<dd>la djan. goi ko'a cu blanu
+	John (referred to as #1) is blue.
+</dl><lx "GOhA, selma'o catalog">  XE "GOhA selma=92o"   <cx "cmavo as bri=
vla, selma'o catalog">  XE "cmavo as brivla"   selma'o GOhA (<a href=3Dchap=
7.html>Chapter 7</a>)
+<p>
+A general selma'o for all cmavo which can take the place of brivla. There =
are several groups of these.
+<p>
+<dl compact><dt>	<dd>A: mi klama le zarci
+	B:  mi go'i
+	A:  I'm going to the market.
+	B:  Me, too.
+</dl><lx "GUhA, selma'o catalog">  XE "GUhA selma=92o"   <cx "connection, =
of tanru units, selma'o catalog">  XE "connection: of tanru units"   selma'=
o GUhA (<a href=3Dchap14.html>Chapter 14</a>)
+<p>
+Indicates the beginning of two logically connected tanru units. Takes the =
place of GA when forming logically-connected tanru. See GI.
+<p>
+<dl compact><dt>	<dd>la .alis. gu'e ricfu gi blanu
+	Alice is both rich and blue.
+</dl><lx "I, selma'o catalog">  XE =93I selma=92o"   <cx "sentence separat=
or, selma'o catalog">  XE "sentence separator"   selma'o I (<a href=3Dchap1=
9.html>Chapter 19</a>)
+<p>
+Separates two sentences from each other.
+<p>
+<dl compact><dt>	<dd>mi klama le zarci  .i mi klama le zdani
+	I go-to the market.  I go-to the office.
+</dl><lx "JA, selma'o catalog">  XE "JA selma=92o"   <cx "connection, of t=
anru units, selma'o catalog">  XE "connection: of tanru units"   selma'o JA=
 (<a href=3Dchap14.html>Chapter 14</a>)
+<p>
+Specifies a logical connection (e.g. =93and=94, =93or=94, =93if=94) betwee=
n two tanru units, mathematical operands, tenses, or abstractions.
+<p>
+<dl compact><dt>	<dd>ti blanu je zdani
+	This is-blue and a-house.
+</dl><lx "JAI, selma'o catalog">  XE "JAI selma=92o"   <cx "conversion, mo=
dal, selma'o catalog">  XE "conversion: modal"   selma'o JAI (<a href=3Dcha=
p9.html>Chapter 9</a>)
+<p>
+When followed by a tense or modal, creates a conversion operator attachabl=
e to a selbri which exchanges the modal place with the x1 place of the selb=
ri. When alone, is a conversion operator exchanging the x1 place of the sel=
bri (which should be an abstract sumti) with one of the places of the abstr=
acted-over bridi.
+<p>
+<dl compact><dt>	<dd>mi jai gau galfi le bitmu skari
+	I am-the-actor-in modifying the wall color.
+	I act so as to modify the wall color.
+	I change the color of the wall.
+</dl><lx "JOI, selma'o catalog">  XE "JOI selma=92o"   <cx "connection, of=
 sumti, selma'o catalog">  XE "connection: of sumti"   selma'o JOI (<a href=
=3Dchap14.html>Chapter 14</a>)
+<p>
+Specifies a non-logical connection (e.g. together-with-as-mass, -set, or -=
sequence) between two sumti, tanru units, or various other things. When imm=
ediately followed by GI, provides forethought non-logical connection analog=
ous to GA.
+<p>
+<dl compact><dt>	<dd>la djan. joi la .alis. cu bevri le pipno
+	John massed-with Alice carry the piano.
+</dl><lx "JOhI, selma'o catalog">  XE "JOhI selma=92o"   <cx "vector, form=
ing, selma'o catalog">  XE "vector: forming"   selma'o JOhI (<a href=3Dchap=
18.html>Chapter 18</a>)
+<p>
+Indicates that the following mathematical operands (a list terminated by T=
EhU) form a mathematical vector (one-dimensional array).
+<p>
+<dl compact><dt>	<dd>li jo'i paboi reboi te'u su'i jo'i ciboi voboi du
+		li jo'i voboi xaboi
+	The-number array (one, two) plus array (three, four) equals
+		the-number array (four, six).
+	(1,2) + (3,4) =3D (4,6)
+</dl><lx "KE, selma'o catalog">  XE "KE selma=92o"   <cx "grouping, indica=
tor for, selma'o catalog">  XE "grouping: indicator for"   selma'o KE (<a h=
ref=3Dchap5.html>Chapter 5</a>)
+<p>
+Groups everything between itself and a following KEhE for purposes of logi=
cal connection, tanru construction, or other purposes. KE and KEhE are not =
used for mathematical (see VEI and VEhO) or discursive (see TO and TOI) pur=
poses.
+<p>
+<dl compact><dt>	<dd>ta ke melbi cmalu ke'e nixli ckule
+	That is-a-( pretty little ) girl school.
+	That is a school for girls who are pretty in their littleness.
+
+</dl><lx "KEI, selma'o catalog">  XE "KEI selma=92o"   <cx "NU, terminator=
 for, selma'o catalog">  XE "NU selma=92o: terminator for"   selma'o KEI (<=
a href=3Dchap11.html>Chapter 11</a>)
+<p>
+Elidable terminator for NU. Marks the end of an abstraction bridi.
+<p>
+<dl compact><dt>	<dd>la djan. cu nu sonci kei djica
+	John is-an-(event-of being-a-soldier) type-of desirer.
+	John wants to be a soldier.
+</dl><lx "KEhE, selma'o catalog">  XE "KEhE selma=92o"   <cx "KE, terminat=
or for, selma'o catalog">  XE "KE selma=92o: terminator for"   selma'o KEhE=
 (<a href=3Dchap5.html>Chapter 5</a>)
+<p>
+Elidable terminator for KE. Marks the end of a grouping.
+<p>
+<dl compact><dt>	<dd>ta ke melbi cmalu ke'e nixli ckule
+	That is-a-( pretty little ) girl school.
+	That is a school for girls who are pretty in their littleness.
+</dl><lx "KI, selma'o catalog">  XE "KI selma=92o"   <cx "tense, making st=
icky, selma'o catalog">  XE "tense: making sticky"   selma'o KI (<a href=3D=
chap10.html>Chapter 10</a>)
+<p>
+When preceded by a tense or modal, makes it =93sticky=94, so that it appli=
es to all further bridi until reset by another appearance of KI. When alone=
, eliminates all sticky tenses.
+<p>
+<lx "KOhA, selma'o catalog">  XE "KOhA selma=92o"   <cx "cmavo as sumti, s=
elma'o catalog">  XE "cmavo as sumti"   selma'o KOhA (<a href=3Dchap7.html>=
Chapter 7</a>)
+<p>
+A general selma'o which contains all cmavo which can substitute for sumti.=
 These cmavo are divided into several groups.
+<p>
+<dl compact><dt>	<dd>le blanu zdani goi ko'a cu barda  .i ko'a na cmamau t=
i
+	The blue house (referred to as #1) is big.  #1 is-not smaller-than this-t=
hing.
+</dl><lx "KU, selma'o catalog">  XE "KU selma=92o"   <cx "LE, terminator f=
or, selma'o catalog">  XE "LE selma=92o: terminator for"   <cx "LA, termina=
tor for, selma'o catalog">  XE "LA selma=92o: terminator for"   <cx "tense =
with no sumti, indicator for, selma'o catalog">  XE "tense with no sumti: i=
ndicator for"   <cx "modal with no sumti, indicator for, selma'o catalog"> =
 XE "modal with no sumti: indicator for"   <cx "negation boundary, forming,=
 selma'o catalog">  XE "negation boundary: forming"   selma'o KU (<a href=
=3Dchap6.html>Chapter 6</a>, <a href=3Dchap10.html>Chapter 10</a>)
+<p>
+Elidable terminator for LE and some uses of LA. Indicates the end of a des=
cription sumti. Also used after a tense or modal to indicate that no sumti =
follows, and in the compound NA+KU to indicate natural language-style negat=
ion.
+<p>
+<dl compact><dt>	<dd>le prenu ku le zdani ku klama
+	The person, to-the house, goes.
+	The person goes to the house.
+</dl><lx "KUhE, selma'o catalog">  XE "KUhE selma=92o"   <cx "PEhO, termin=
ator for, selma'o catalog">  XE "PEhO selma=92o: terminator for"   selma'o =
KUhE (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Elidable terminator for PEhO: indicates the end of a forethought mathemati=
cal expression (one in which the operator precedes the operands).
+ <dl compact><dt>	<dd>li pe'o su'i reboi reboi re[boi] ku'e du
+		li xa
+	The number [forethought] the-sum-of two two two [end] equals=20
+the-number six.
+
+</dl><lx "KUhO, selma'o catalog">  XE "KUhO selma=92o"   <cx "NOI, termina=
tor for, selma'o catalog">  XE "NOI selma=92o: terminator for"   selma'o KU=
hO (<a href=3Dchap8.html>Chapter 8</a>)
+<p>
+Elidable terminator for NOI. Indicates the end of a relative clause.
+<dl compact><dt>	<dd>le zdani poi blanu ku'o barda
+	The house which is-blue is-big.
+</dl><lx "LA, selma'o catalog">  XE "LA selma=92o"   <cx "name descriptor,=
 selma'o catalog">  XE "name descriptor"   selma'o LA (<a href=3Dchap5.html=
>Chapter 5</a>)
+<p>
+Descriptors which change name words (or selbri) into sumti which identify =
people or things by name. Similar to LE. May be terminated with KU if follo=
wed by a description selbri.
+<p>
+<dl compact><dt>	<dd>la kikeros. du la tulis.
+	Cicero is Tully.
+</dl><lx "LAU, selma'o catalog">  XE "LAU selma=92o"   <cx "letter shift, =
selma'o catalog">  XE "letter shift"   selma'o LAU (<a href=3Dchap17.html>C=
hapter 17</a>)
+<p>
+Combines with the following alphabetic letter to represent a single marker=
: change from lower to upper case, change of font, punctuation, etc.)
+<p>
+<dl compact><dt>	<dd>tau sy. .ibu
+	[single-shift] =93s=94 =93i=94
+	Si (chemical symbol for silicon)
+</dl><lx "LAhE, selma'o catalog">  XE "LAhE selma=92o"   <cx "converting s=
umti to related meaning, selma'o catalog">  XE "converting sumti to related=
 meaning"   selma'o LAhE (<a href=3Dchap6.html>Chapter 6</a>)
+<p>
+Qualifiers which, when prefixed to a sumti, change it into another sumti w=
ith related meaning. Qualifiers can also consist of a cmavo from selma'o NA=
hE plus BO. Terminated by LUhU.
+<p>
+<dl compact><dt>	<dd>mi viska la'e zoi kuot. A Tale of Two Cities .kuot
+	I see that-represented-by the-text =93A Tale of Two Cities=94.
+	I see the book =93A Tale of Two Cities=94.
+</dl><lx "LE, selma'o catalog">  XE "LE selma=92o"   <cx "descriptor, selm=
a'o catalog">  XE "descriptor"   <cx "article, selma'o catalog">  XE "artic=
le"   selma'o LE (<a href=3Dchap6.html>Chapter 6</a>)
+<p>
+Descriptors which make selbri into sumti which describe or specify things =
that fit into the x1 place of the selbri. Termionated by KU. See LA.
+<p>
+<dl compact><dt>	<dd>le gerku klama le zdani
+	The dog goes-to the house.
+</dl><lx "LEhU, selma'o catalog">  XE "LEhU selma=92o"   <cx "LOhU, termin=
ator for, selma'o catalog">  XE "LOhU selma=92o: terminator for"   selma'o =
LEhU (<a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Indicates the end of a quotation begun with LOhU. Not an elidable terminat=
or.
+<p>
+<dl compact><dt>	<dd>lo'u mi du do du mi le'u cu na lojbo drani
+	[quote] mi du do du mi [unquote] is-not Lojbanically correct.
+	=93mi du do du mi=94 is not correct Lojban.
+
+</dl><lx "LI, selma'o catalog">  XE "LI selma=92o"   <cx "number, descript=
or for, selma'o catalog">  XE "numbers: descriptor for"   <cx "descriptor f=
or numbers, selma'o catalog">  XE "descriptor for numbers"   selma'o LI (<a=
 href=3Dchap18.html>Chapter 18</a>)
+<p>
+Descriptors which change numbers or other mathematical expressions into su=
mti which specify numbers or numerical expressions. Terminated by LOhO.
+<p>
+<dl compact><dt>	<dd>li re su'u re na du li vo su'i vo
+	The-number 2 minus 2 not equals the-number 4 plus 4.
+	$2 - 2<>( ( 4 + 4$
+</dl><lx "LIhU, selma'o catalog">  XE "LIhU selma=92o"   <cx "LU, terminat=
or for, selma'o catalog">  XE "LU selma=92o: terminator for"   selma'o LIhU=
 (<a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Elidable terminator for LU. Indicates the end of a text quotation.
+<p>
+<dl compact><dt>	<dd>mi cusku lu mi klama le zarci li'u
+	I express [quote] I go-to the market [end quote].
+</dl><lx "LOhO, selma'o catalog">  XE "LOhO selma=92o"   <cx "LI, terminat=
or for, selma'o catalog">  XE "LI selma=92o: terminator for"   selma'o LOhO=
 (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Elidable terminator for LI. Indicates the end of a mathematical expression=
 used in a LI description.
+<p>
+<dl compact><dt>	<dd>li vo lo'o li ci lo'o cu zmadu
+	The-number 4 [end number], the number 3 [end number], is greater.
+	$4 > 3$
+</dl><lx "LOhU, selma'o catalog">  XE "LOhU selma=92o"   <cx "quotation, o=
f Lojban words, selma'o catalog">  XE "quotation: of Lojban words"   selma'=
o LOhU (<a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Indicates the beginning of a quotation (a sumti) which is grammatical as l=
ong as the quoted material consists of Lojban words, whether they form a te=
xt or not. Terminated by LEhU.
+<p>
+<dl compact><dt>	<dd>do cusku lo'u mi du do du ko'a le'u
+	You express [quote] mi du do du ko'a [end quote].
+	You said, =93mi du do du ko'a=94.
+</dl><lx "LU, selma'o catalog">  XE "LU selma=92o"   <cx "quotation, gramm=
atical, selma'o catalog">  XE "quotation: grammatical"   selma'o LU (<a hre=
f=3Dchap19.html>Chapter 19</a>)
+<p>
+Indicates the beginning of a quotation (a sumti) which is grammatical only=
 if the quoted material also forms a grammatical Lojban text. Terminated by=
 LIhU.
+<p>
+<dl compact><dt>	<dd>mi cusku lu mi klama le zarci li'u
+	I express [quote] I go-to the market [end quote].
+</dl><lx "LUhU, selma'o catalog">  XE "LUhU selma=92o"   <cx "LAhE, termin=
ator for, selma'o catalog">  XE "LAhE selma=92o: terminator for"   <cx "NAh=
E+BO, terminator for, selma'o catalog">  XE "NAhE+BO: terminator for"   sel=
ma'o LUhU (<a href=3Dchap6.html>Chapter 6</a>)
+<p>
+Elidable terminator for LAhE and NAhE+BO. Indicates the end of a qualified=
 sumti.
+<p>
+<dl compact><dt>	<dd>mi viska la'e lu barda gerku li'u lu'u
+	I see the-referent-of [quote] big dog [end quote] [end ref]
+	I saw =93Big Dog=94 [not the words, but a book or movie].
+
+</dl><lx "MAI, selma'o catalog">  XE "MAI selma=92o"   <cx "index numberin=
g, selma'o catalog">  XE "index numbering"   selma'o MAI (<a href=3Dchap18.=
html>Chapter 18</a>, <a href=3Dchap19.html>Chapter 19</a>)
+<p>
+When suffixed to a number or string of letter words, produces a free modif=
ier which serves as an index number within a text.
+<p>
+<dl compact><dt>	<dd>pamai mi pu klama le zarci
+	1-thly, I [past] go to-the market.
+	First, I went to the market.
+</dl><lx "MAhO, selma'o catalog">  XE "MAhO selma=92o"   <cx "converting, =
operand to operator, selma'o catalog">  XE "converting: operand to operator=
"   selma'o MAhO (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Produces a mathematical operator from a letter or other operand. Terminate=
d by TEhU. See VUhU.
+<p>
+<dl compact><dt>	<dd>ma'o fy. boi xy.
+	[operator] f x
+	f(x)
+</dl><lx "ME, selma'o catalog">  XE "ME selma=92o"   <cx "converting, sumt=
i to tanru unit, selma'o catalog">  XE "converting: sumti to tanru unit"   =
selma'o ME (<a href=3Dchap5.html>Chapter 5</a>, <a href=3Dchap18.html>Chapt=
er 18</a>)
+<p>
+Produces a tanru unit from a sumti, which is applicable to the things refe=
renced by the sumti. Terminated by MEhU.
+<p>
+<dl compact><dt>	<dd>ta me la ford. karce
+	That is-a-Ford-type car.
+	That's a Ford car.
+</dl><lx "MEhU, selma'o catalog">  XE "MEhU selma=92o"   <cx "ME, terminat=
or for, selma'o catalog">  XE "ME selma=92o: terminator for"   selma'o MEhU=
 (<a href=3Dchap5.html>Chapter 5</a>)
+<p>
+The elidable terminator for ME. Indicates the end of a sumti converted to =
a tanru unit.
+<p>
+<dl compact><dt>	<dd>ta me mi me'u zdani
+	That's a me type of house.
+</dl><lx "MOI, selma'o catalog">  XE "MOI selma=92o"   <cx "converting, qu=
antifier to selbri, selma'o catalog">  XE "converting: quantifier to selbri=
"   selma'o MOI (<a href=3Dchap5.html>Chapter 5, Chapter 18</a>)
+<p>
+Suffixes added to numbers or other quantifiers to make various numerically=
-based selbri.
+<p>
+<dl compact><dt>	<dd>la djan. joi la frank. cu bruna remei
+	John in-a-mass-with Frank are-a-brother-type-of twosome.
+	John and Frank are two brothers.
+</dl><lx "MOhE, selma'o catalog">  XE "MOhE selma=92o"   <cx "converting, =
sumti to operand, selma'o catalog">  XE "converting: sumti to operand"   se=
lma'o MOhE (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Produces a mathematical operand from a sumti; used to make dimensioned uni=
ts. Terminated by TEhU.
+<p>
+<dl compact><dt>	<dd>li mo'e re ratcu su'i mo'e re ractu du li mo'e vo dan=
lu
+The-number two rats plus two rabbits equals the-number four animals.
+	2 rats + 2 rabbits =3D 4 animals.
+
+</dl><lx "MOhI, selma'o catalog">  XE "MOhI selma=92o"   <cx "space moveme=
nt indicator, selma'o catalog">  XE "space movement indicator"   selma'o MO=
hI (<a href=3Dchap10.html>Chapter 10</a>)
+<p>
+A tense flag indicating movement in space, in a direction specified by a f=
ollowing FAhA cmavo.
+<p>
+<dl compact><dt>	<dd>le verba mo'i ri'u cadzu le bisli
+	The child [movement] [right] walks-on the ice.
+	The child walks toward my right on the ice.
+</dl><lx "NA, selma'o catalog">  XE "NA selma=92o"   <cx "negator, contrad=
ictory, selma'o catalog">  XE "negator: contradictory"   selma'o NA (<a hre=
f=3Dchap14.html>Chapter 14</a>, <a href=3Dchap15.html>Chapter 15</a>)
+<p>
+Contradictory negators, asserting that a whole bridi is false (or true).
+<p>
+<dl compact><dt>	<dd>mi na klama le zarci
+	It is not true that I go to the market.
+</dl><p>Also used to construct logical connective compound cmavo.
+<p>
+<lx "NAI, selma'o catalog">  XE "NAI selma=92o"   <cx "negator, single-wor=
d, selma'o catalog">  XE "negator: single-word"   selma'o NAI (<a href=3Dch=
ap14.html>Chapter 14</a>, <a href=3Dchap15.html>Chapter 15</a>)
+<p>
+Negates the previous word, but can only be used with certain selma'o as sp=
ecified by the grammar.
+<p>
+<lx "NAhE, selma'o catalog">  XE "NAhE selma=92o"   <cx "negator, scalar, =
selma'o catalog">  XE "negator: scalar"   selma'o NAhE (<a href=3Dchap15.ht=
ml>Chapter 15</a>)
+<p>
+Scalar negators, modifying a selbri or a sumti to a value other than the o=
ne stated, the opposite of the one stated, etc. Also used with following BO=
 to construct a sumti qualifier; see LAhE.
+<p>
+<dl compact><dt>	<dd>ta na'e blanu zdani
+	That is-a-non- blue house.
+	That is a house which is other than blue.
+</dl><lx "NAhU, selma'o catalog">  XE "NAhU selma=92o"   <cx "converting, =
selbri to operator, selma'o catalog">  XE "converting: selbri to operator" =
  selma'o NAhU (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Creates a mathematical operator from a selbri. Terminated by TEhU. See VUh=
U.
+<p>
+<dl compact><dt>	<dd>li na'u tanjo te'u vei pai fe'i re [ve'o] du li ci'i
+	The-number the-operator tangent ((pi / 2 ) =3D the-number infinity.
+	tan((pi/2) =3D infinity(
+</dl><lx "NIhE, selma'o catalog">  XE "NIhE selma=92o"   <cx "converting, =
selbri to operand, selma'o catalog">  XE "converting: selbri to operand"   =
selma'o NIhE (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Creates a mathematical operand from a selbri, usually a =93ni=94 abstracti=
on. Terminated by TEhU.=20
+<p>
+<dl compact><dt>	<dd>li ni'e ni clani [te'u] pi'i ni'e ni ganra [te'u] pi'=
i
+		ni'e ni condi te'u du li ni'e ni canlu
+	The-number quantity-of length times quantity-of width times
+		quantity-of depth equals the-number quantity-of volume.
+	Length ( Width ( Depth =3D Volume
+
+</dl><lx "NIhO, selma'o catalog">  XE "NIhO selma=92o"   <cx "paragraph ma=
rker, selma'o catalog">  XE "paragraph marker"   selma'o NIhO (<a href=3Dch=
ap19.html>Chapter 19</a>)
+<p>
+Marks the beginning of a new paragraph, and indicates whether it contains =
old or new subject matter.
+<p>
+<lx "NOI, selma'o catalog">  XE "NOI selma=92o"   <cx "relative clause mar=
ker, selma'o catalog">  XE "relative clause marker"   selma'o NOI (<a href=
=3Dchap8.html>Chapter 8</a>)
+Introduces r<p>
+elative clauses. The following bridi modifies the preceding sumti. Termina=
ted by KUhO. See GOI.
+ <dl compact><dt>	<dd>le zdani poi blanu cu cmalu
+	The house which is blue is small.
+</dl><lx "NU, selma'o catalog">  XE "NU selma=92o"   <cx "abstractors, sel=
ma'o catalog">  XE "abstractors"   selma'o NU (<a href=3Dchap11.html>Chapte=
r 11</a>)
+<p>
+Abstractors which, when prefixed to a bridi, create abstraction selbri. Te=
rminated by KEI.
+<p>
+<dl compact><dt>	<dd>la djan. cu djica le nu sonci [kei]
+	John desires the event-of being-a-soldier.
+</dl><lx "NUhA, selma'o catalog">  XE "NUhA selma=92o"   <cx "converting, =
operator to selbri, selma'o catalog">  XE "converting: operator to selbri" =
  selma'o NUhA (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Creates a selbri from a mathematical operator. See VUhU.
+<p>
+<dl compact><dt>	<dd>li ni'umu cu nu'a va'a li ma'umu
+	The-number -5 is-the-negation-of the-number +5.
+</dl><lx "NUhI, selma'o catalog">  XE "NUhI selma=92o"   <cx "termset mark=
er, selma'o catalog">  XE "termset marker"   selma'o NUhI (<a href=3Dchap14=
.html>Chapter 14</a>, <a href=3Dchap16.html>Chapter 16</a>)
+<p>
+Marks the beginning of a termset, which is used to make simultaneous claim=
s involving two or more different places of a selbri. Terminated by NUhU.
+<p>
+<dl compact><dt>	<dd>mi klama nu'i ge le zarci le briju nu=92u gi le zdani
+		le ckule [nu'u]
+	I go [start] to-the market from-the office [joint] and to-the house
+		from-the school.
+</dl><lx "NUhU, selma'o catalog">  XE "NUhU selma=92o"   <cx "NUhU, termin=
ator for, selma'o catalog">  XE "NUhU selma=92o: terminator for"   selma'o =
NUhU (<a href=3Dchap14.html>Chapter 14</a>)
+<p>
+Elidable terminator for NUhI. Marks the end of a termset.
+<p>
+<dl compact><dt>	<dd> mi klama nu'i ge le zarci le briju nu=92u gi le zdan=
i
+		le ckule [nu'u]
+	I go [start] to-the market from-the office [joint] and to-the house
+		from-the school.
+</dl><lx "PA, selma'o catalog">  XE "PA selma=92o"   <cx "quantifiers, sel=
ma'o catalog">  XE =93quantifier"   <cx "digits, selma'o catalog">  XE "dig=
its"   selma'o PA (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Digits and related quantifiers (some, all, many, etc.). Terminated by BOI.
+<p>
+<dl compact><dt>	<dd>mi speni re ninmu
+	I am-married-to two women.
+
+</dl><lx "PEhE, selma'o catalog">  XE "PEhE selma=92o"   <cx "termset conn=
ectives, selma'o catalog">  XE "termset connectives"   selma'o PEhE (<a hre=
f=3Dchap14.html>Chapter 14</a>)
+<p>
+Precedes a logical or non-logical connective that joins two termsets. Term=
sets (see CEhE) are used to associate several terms for logical connectives=
, for equal quantifier scope, or for special constructs in tenses.
+<p>
+<dl compact><dt>	<dd>mi ce'e do pe'e je la djan. ce'e la djeimyz. cu pendo
+	I [,] you [joint] and John [,] James are-friends-of.
+	I am a friend of you, and John is a friend of James.
+</dl><lx "PEhO, selma'o catalog">  XE "PEhO selma=92o"   <cx "operator, fo=
rethought marker, selma'o catalog">  XE "operator: forethought marker"   se=
lma'o PEhO (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+An optional signal of forethought mathematical operators, which precede th=
eir operands. Terminated by KUhE.
+<p>
+<dl compact><dt>	<dd>li vo du li pe'o su'i reboi re
+	The-number four equals the-number [forethought] sum-of two two.
+</dl><lx "PU, selma'o catalog">  XE "PU selma=92o"   <cx "direction, time,=
 selma'o catalog">  XE "direction: time"   selma'o PU (<a href=3Dchap10.htm=
l>Chapter 10</a>)
+<p>
+Specifies simple time directions (future, past, or neither).
+<p>
+<dl compact><dt>	<dd>mi pu klama le zarci
+	I [past] go-to the market.
+	I went to the market.
+</dl><lx "RAhO, selma'o catalog">  XE "RAhO selma=92o"   <cx "pro-bridi up=
date, flag for, selma'o catalog">  XE "pro-bridi update: flag for"   <cx "r=
e-evaluation of referents, flag for, selma'o catalog">  XE "re-evaluation o=
f referents: flag for"   selma'o RAhO (<a href=3Dchap7.html>Chapter 7</a>)
+<p>
+The pro-bridi update flag: changes the meaning of sumti implicitly attache=
d to a pro-bridi (see GOhA) to fit the current context rather than the orig=
inal context.
+<p>
+<dl compact><dt>	<dd>A: mi ba lumci le mi karce
+	B:  mi go'i
+	A:  I [future] wash my car.
+	B:  I do-the-same-thing (i.e. wash A's car).
+	A:  mi ba lumci le mi karce
+	B:  mi go'i ra'o
+	A:  I [future] wash my car.
+	B:  I do-the-corresponding-thing (i.e. wash B's car).
+</dl><lx "ROI, selma'o catalog">  XE "ROI selma=92o"   <cx "tense, extensi=
onal, selma'o catalog">  XE "tense: extensional"   selma'o ROI (<a href=3Dc=
hap10.html>Chapter 10</a>)
+<p>
+When suffixed to a number, makes an extensional tense (e.g. once, twice, m=
any times).
+<p>
+<dl compact><dt>	<dd>mi reroi klama le zarci
+	I twice go-to the market.
+</dl><lx "SA, selma'o catalog">  XE "SA selma=92o"   <cx "erasing phrase, =
selma'o catalog">  XE "erasure: phrase"   <cx "erasing sentence, selma'o ca=
talog">  XE "erasure: sentence"   selma'o SA (<a href=3Dchap19.html>Chapter=
 19</a>)
+<p>
+Erases the previous phrase or sentence.
+<p>
+<dl compact><dt>	<dd>mi klama sa do klama le zarci
+	I go, er, you go-to the market.
+</dl><lx "SE, selma'o catalog">  XE "SE selma=92o"   <cx "conversion, of s=
elbri, selma'o catalog">  XE "conversion: of selbri"   <cx "selbri places, =
re-ordering, selma'o catalog">  XE "selbri place structure: re-ordering"   =
selma'o SE (<a href=3Dchap5.html>Chapter 5</a>)
+<p>
+Converts a selbri, rearranging the order of places by exchanging the x1 pl=
ace with a specified numbered place.
+<p>
+<dl compact><dt>	<dd>le zarci se klama mi
+	The market is-gone-to by me.
+</dl><p>Also used in constructing connective and modal compound cmavo.
+<p>
+<lx "SEI, selma'o catalog">  XE "SEI selma=92o"   <cx "metalinguistic inse=
rtions, marker for, selma'o catalog">  XE "metalinguistic insertions: marke=
r for"   selma'o SEI (<a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Marks the beginning of metalinguistic insertions which comment on the main=
 bridi. Terminated by SEhU.
+<p>
+<dl compact><dt>	<dd>la frank. prami sei gleki [se'u] la djein.
+	Frank loves (he is happy) Jane.
+</dl><lx "SEhU, selma'o catalog">  XE "SEhU selma=92o"   <cx "SEI, termina=
tor for, selma'o catalog">  XE "SEI selma=92o: terminator for"   <cx "SOI, =
terminator for, selma'o catalog">  XE "SOI selma=92o: terminator for"   sel=
ma'o SEhU (<a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Elidable terminator for SEI and SOI. Ends metalinguistic insertions.
+<p>
+<dl compact><dt>	<dd>la frank. prami sei gleki se'u la djein.
+	Frank loves (he is happy) Jane.
+</dl><lx "SI, selma'o catalog">  XE "SI selma=92o"   <cx "erasing, word, s=
elma'o catalog">  XE "erasure: word"   selma'o SI (<a href=3Dchap19.html>Ch=
apter 19</a>)
+<p>
+Erases the previous single word.
+<p>
+<dl compact><dt>	<dd>mi si do klama le zarci
+	I, er, you go to-the market.
+</dl><lx "SOI, selma'o catalog">  XE "SOI selma=92o"   <cx "reciprocity, m=
arking, selma'o catalog">  XE "reciprocity: marking"   selma'o SOI (<a href=
=3Dchap7.html>Chapter 7</a>)
+<p>
+Marks reciprocity between two sumti (like =93vice versa=94 in English).
+<p>
+<dl compact><dt>	<dd>mi prami do soi mi
+	I love you [reciprocally] me.
+	I love you and vice versa.
+</dl><lx "SU, selma'o catalog">  XE "SU selma=92o"   <cx "erasing, discour=
se, selma'o catalog">  XE "erasure: discourse"   selma'o SU (<a href=3Dchap=
19.html>Chapter 19</a>)
+<p>
+Closes and erases the entire previous discourse.
+<p>
+<lx "TAhE, selma'o catalog">  XE "TAhE selma=92o"   <cx "frequency within =
interval, specifying, selma'o catalog">  XE "frequency within interval: spe=
cifying"   selma'o TAhE (<a href=3Dchap10.html>Chapter 10</a>)
+<p>
+A tense modifier specifying frequencies within an interval of time or spac=
e (regularly, habitually, etc.).
+<p>
+<dl compact><dt>	<dd>le verba ta'e klama le ckule
+	The child habitually goes to-the school.
+
+</dl><lx "TEI, selma'o catalog">  XE "TEI selma=92o"   <cx "compound lette=
r marker, selma'o catalog">  XE "compound letter marker"   selma'o TEI (<a =
href=3Dchap17.html>Chapter 17</a>)
+<p>
+Signals the beginning of a compound letter word, which acts grammatically =
like a single letter. Compound letter words end with the non-elidable selma=
'o FOI.
+<p>
+<dl compact><dt>	<dd>tei .ebu .akut. bu foi
+	(=93e=94 =93acute=94 )
+	the letter =93e=94 with an acute accent
+</dl><lx "TEhU, selma'o catalog">  XE "TEhU selma=92o"   <cx "JOhI, termin=
ator for, selma'o catalog">  XE "JOhI selma=92o: terminator for"   <cx "MAh=
O, terminator for, selma'o catalog">  XE "MAhO selma=92o: terminator for"  =
 <cx "MOhE, terminator for, selma'o catalog">  XE "MOhE selma=92o: terminat=
or for"   <cx "NAhU, terminator for, selma'o catalog">  XE "NAhU selma=92o:=
 terminator for"   <cx "NIhE, terminator for, selma'o catalog">  XE "NIhE s=
elma=92o: terminator for"   selma'o TEhU (<a href=3Dchap18.html>Chapter 18<=
/a>)
+<p>
+Elidable terminator for JOhI, MAhO, MOhE, NAhU, or NIhE. Marks the end of =
a mathematical conversion construct.
+<p>
+<dl compact><dt>	<dd>li jo'i paboi reboi te'u su'i jo'i ciboi voboi du
+		li jo'i voboi xaboi
+	The-number array (one, two) plus array (three, four) equals
+		the-number array (four, six).
+	(1,2) + (3,4) =3D (4,6)
+</dl><lx "TO, selma'o catalog">  XE "TO selma=92o"   <cx "parenthesis, dis=
course, selma'o catalog">  XE "parenthesis: discourse"   selma'o TO (<a hre=
f=3Dchap19.html>Chapter 19</a>)
+<p>
+Left discursive parenthesis: allows inserting a digression. Terminated by =
TOI.
+<p>
+<dl compact><dt>	<dd>doi lisas. mi djica le nu to doi frank. ko sisti toi =
do viska le mlatu
+	O Lisa, I desire the event-of ( O Frank, [imperative] stop! ) you see the=
 cat.
+	Lisa, I want you to (Frank! Stop!) see the cat.
+</dl><lx "TOI, selma'o catalog">  XE "TOI selma=92o"   <cx "TO, terminator=
 for, selma'o catalog">  XE "TO selma=92o: terminator for"   selma'o TOI (<=
a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Elidable terminator for TO. The right discursive parenthesis.
+ <dl compact><dt>	<dd>doi lisas. mi djica le nu to doi frank. ko sisti toi=
 do viska le mlatu
+	O Lisa, I desire the event-of ( O Frank, [imperative] stop! ) you see the=
 cat.
+	Lisa, I want you to (Frank! Stop!) see the cat.
+</dl><lx "TUhE, selma'o catalog">  XE "TUhE selma=92o"   <cx "sentence gro=
uping, selma'o catalog">  XE "sentence grouping"   selma'o TUhE (<a href=3D=
chap19.html>Chapter 19</a>)
+<p>
+Groups multiple sentences or paragraphs into a logical unit. Terminated by=
 TUhU.
+<p>
+<dl compact><dt>	<dd>xagmau zo'u tu'e ganai cidja gi cnino  .i ganai vanju=
 gi tolci'o [tu'u]
+	Is-best :  [start] If food, then new.  If wine, then old.
+	As for what is best:  if food, then new [is best]; if wine, then old [is =
best].
+</dl><lx "TUhU, selma'o catalog">  XE "TUhU selma=92o"   <cx "TUhE, termin=
ator for, selma'o catalog">  XE "TUhE selma=92o: terminator for"   selma'o =
TUhU (<a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Elidable terminator for TUhE. Marks the end of a multiple sentence group.
+<p>
+<lx "UI, selma'o catalog">  XE "UI selma=92o"   <cx "emotion, marking, sel=
ma'o catalog">  XE "emotions: marking"   selma'o UI (<a href=3Dchap13.html>=
Chapter 13</a>)
+<p>
+Indicates the speaker's emotional state or source of knowledge, or the pre=
sent stage of discourse.
+<p>
+<dl compact><dt>	<dd>.ui la djan. klama
+	[Happiness!] John is-coming.
+	Hurrah! John is coming!
+</dl><lx "VA, selma'o catalog">  XE "VA selma=92o"   <cx "distance, spatia=
l, selma'o catalog">  XE "distance: spatial"   selma'o VA (<a href=3Dchap10=
.html>Chapter 10</a>)
+<p>
+A tense indicating distance in space (near, far, or neither).
+<p>
+<dl compact><dt>	<dd>le nanmu va batci le gerku
+	The man [medium distance] bites the dog.
+	Over there the man is biting the dog.
+</dl><lx "VAU, selma'o catalog">  XE "VAU selma=92o"   <cx "simple bridi, =
terminator for, selma'o catalog">  XE "simple bridi: terminator for"   <cx =
"GIhA, terminator for, selma'o catalog">  XE "GIhA selma=92o: terminator fo=
r"   selma'o VAU (<a href=3Dchap14.html>Chapter 14</a>)
+<p>
+Elidable terminator for a simple bridi, or for each bridi-tail of a GIhA l=
ogical connection.
+<p>
+<dl compact><dt>	<dd>mi dunda le cukta [vau] gi'e lebna lo rupnu vau do [v=
au]
+	I (give the book) and (take some currency-units) to/from you.
+</dl><lx "VEI, selma'o catalog">  XE "VEI selma=92o"   <cx "mathematical p=
arenthesis, left, selma'o catalog">  XE "mathematical parenthesis: left"   =
selma'o VEI (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Left mathematical parenthesis: groups mathematical operations. Terminated =
by VEhO.
+<p>
+<dl compact><dt>	<dd>li vei ny. su'i pa ve'o pi'i vei ny. su'i pa [ve'o] d=
u
+		li ny. [bo] te'a re su'i re bo pi'i ny. su'i pa
+	The-number (=93n=94 plus one ) times (=93n=94 plus one ) equals
+		the-number n-power-two plus two-times-=93n=94 plus 1.
+	(n + 1)(n + 1) =3D n<sup>2</sup>+ 2n + 1
+</dl><lx "VEhA, selma'o catalog">  XE "VEhA selma=92o"   <cx "interval siz=
e, spatial, selma'o catalog">  XE "interval size: spatial"   selma'o VEhA (=
<a href=3Dchap10.html>Chapter 10</a>)
+<p>
+A tense indicating the size of an interval in space (long, medium, or shor=
t).
+<p>
+<lx "VEhO, selma'o catalog">  XE "VEhO selma=92o"   <cx "VEI, terminator f=
or, selma'o catalog">  XE "VEI selma=92o: terminator for"   selma'o VEhO (<=
a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Elidable terminator for VEI: right mathematical parenthesis.
+<p>
+<dl compact><dt>	<dd><dl compact><dt><dd>li vei ny. su'i pa ve'o pi'i vei =
ny. su'i pa [ve'o] du
+		li ny. [bo] te'a re su'i re bo pi'i ny. su'i pa
+	The-number (=93n=94 plus one ) times (=93n=94 plus one ) equals
+		the-number n-power-two plus two-times-=93n=94 plus 1.
+	(n + 1)(n + 1) =3D n<sup>2</sup>+ 2n + 1
+</dl><lx "VIhA, selma'o catalog">  XE "VIhA selma=92o"   <cx "dimensionali=
ty, spatial, selma'o catalog">  XE "dimensionality: spatial"   selma'o VIhA=
 (<a href=3Dchap10.html>Chapter 10</a>)
+<p>
+A tense indicating dimensionality in space (line, plane, volume, or space-=
time interval).
+ <dl compact><dt>	<dd>le verba ve'a vi'a cadzu le bisli
+	The child [medium space interval] [2-dimensional] walks-on the ice.
+	In a medium-sized area, the child walks on the ice.
+</dl><lx "VUhO, selma'o catalog">  XE "VUhO selma=92o"   <cx "relative cla=
uses, connecting to whole sumti, selma'o catalog">  XE =93relative clause: =
connecting to whole sumti"   selma'o VUhO (<a href=3Dchap8.html>Chapter 8</=
a>)
+<p>
+Attaches relative clauses or phrases to a whole (possibly connected) sumti=
, rather than simply to the leftmost portion of the sumti.
+<p>
+<dl compact><dt>	<dd>la frank. ce la djordj. vu'o noi gidva cu zvati le ku=
mfa
+	Frank [in-set-with] George, which are-guides, are-in the room.
+	Frank and George, who are guides, are in the room.
+</dl><lx "VUhU, selma'o catalog">  XE "VUhU selma=92o"   <cx "operators, m=
athematical, selma'o catalog">  XE "operators: mathematical"   selma'o VUhU=
 (<a href=3Dchap18.html>Chapter 18</a>)
+<p>
+Mathematical operators (e.g. (, (). See MAhO.
+<p>
+<dl compact><dt>	<dd>li mu vu'u re du li ci
+	The-number 5 minus 2 equals the-number 3.
+	5 ( 2 =3D 3
+</dl><lx "XI, selma'o catalog">  XE "XI selma=92o"   <cx "subscript marker=
, selma'o catalog">  XE "subscripts: marker"   selma'o XI (<a href=3Dchap18=
.html>Chapter 18</a>)
+<p>
+The subscript marker: the following number or lerfu string is a subscript =
for whatever precedes it.
+<p>
+<dl compact><dt>	<dd>xy. xi re
+	x sub 2
+	x<sub>2</sub>
+</dl><lx "Y, selma'o catalog">  XE "Y selma=92o"   <cx "hesitation, selma'=
o catalog">  XE "hesitation"   selma'o Y (<a href=3Dchap19.html>Chapter 19<=
/a>)
+<p>
+Hesitation noise: content-free, but holds the floor or continues the conve=
rsation. It is different from silence in that silence may be interpreted as=
 having nothing more to say.
+<p>
+<dl compact><dt>	<dd>doi .y. .y. .djan
+	O, uh, uh, John!
+</dl><lx "ZAhO, selma'o catalog">  XE "ZAhO selma=92o"   <cx "tense aspect=
, selma'o catalog">  XE "tense aspect"   selma'o ZAhO (<a href=3Dchap10.htm=
l>Chapter 10</a>)
+<p>
+A tense modifier specifying the contour of an event (e.g. beginning, endin=
g, continuing).
+<p>
+<dl compact><dt>	<dd>mi pu'o damba
+	I [inchoative] fight.
+	I'm on the verge of fighting.
+</dl><lx "ZEI, selma'o catalog">  XE "ZEI selma=92o"   <cx "morphological =
glue, selma'o catalog">  XE "morphological glue"   <cx "combining words int=
o one, selma'o catalog">  XE "combining words into one"   selma'o ZEI (<a h=
ref=3Dchap4.html>Chapter 4</a>)
+<p>
+A morphological glue word, which joins the two words it stands between int=
o the equivalent of a lujvo.
+<p>
+<dl compact><dt>	<dd>ta xy. zei kantu kacma
+	That is-an-(X - ray) camera
+	That is an X-ray camera.
+</dl><lx "ZEhA, selma'o catalog">  XE "ZEhA selma=92o"   <cx "interval siz=
e, time, selma'o catalog">  XE "interval size: time"   selma'o ZEhA (<a hre=
f=3Dchap10.html>Chapter 10</a>)
+<p>
+A tense indicating the size of an interval in time (long, medium, or short=
).
+<p>
+<dl compact><dt>	<dd>mi puze'a citka
+	I [past] [short interval] eat.
+	I ate for a little while.
+</dl><lx "ZI, selma'o catalog">  XE "ZI selma=92o"   <cx "distance, time, =
selma'o catalog">  XE "distance: time"   selma'o ZI (<a href=3Dchap10.html>=
Chapter 10</a>)
+<p>
+A tense indicating distance in time (a long, medium or short time ago or i=
n the future).
+<p>
+<dl compact><dt>	<dd>mi puzi citka
+	I [past] [short distance] eat.
+	I ate a little while ago.
+
+</dl><lx "ZIhE, selma'o catalog">  XE "ZIhE selma=92o"   <cx "connector, f=
or relative clauses, selma'o catalog">  XE "connector: for relative clauses=
"   selma'o ZIhE (<a href=3Dchap8.html>Chapter 8</a>)
+<p>
+Joins multiple relative phrases or clauses which apply to the same sumti. =
Although generally translated with =93and=94, it is not considered a logica=
l connective.
+<p>
+<dl compact><dt>	<dd>mi ponse pa gerku ku poi blabi zi'e noi mi prami ke'a
+	I own one dog such-that it-is-white and such-that-incidentally I love it.
+	I own a dog that is white and which, incidentally, I love.
+	I own a white dog, which I love.
+</dl><lx "ZO, selma'o catalog">  XE "ZO selma=92o"   <cx "quotation, singl=
e-word, selma'o catalog">  XE "quotation: single-word"   selma'o ZO (<a hre=
f=3Dchap19.html>Chapter 19</a>)
+<p>
+Single-word quotation: quotes the following single Lojban word.
+<p>
+<dl compact><dt>	<dd>zo si cu lojbo valsi
+	The-word =93si=94 is-a-Lojbanic word.
+</dl><lx "ZOI, selma'o catalog">  XE "ZOI selma=92o"   <cx "quotation, del=
imited, selma'o catalog">  XE "quotation: delimited"   <cx "quotation, any =
text, selma'o catalog">  XE "quotation: any text"   selma'o ZOI (<a href=3D=
chap19.html>Chapter 19</a>)
+<p>
+Non-Lojban quotation: quotes any text using a delimiting word (which can b=
e any single Lojban word) placed before and after the text. The delimiting =
word must not appear in the text, and must be separated from the text by pa=
uses.
+<p>
+<dl compact><dt>	<dd>zoi kuot. Socrates is mortal .kuot. cu glico jufra
+	The-text =93 Socrates is mortal =94 is-an-English sentence.
+</dl><lx "ZOhU, selma'o catalog">  XE "ZOhU selma=92o"   <cx "prenex marke=
r, selma'o catalog">  XE "prenex marker"   selma'o ZOhU (<a href=3Dchap16.h=
tml>Chapter 16</a>, <a href=3Dchap19.html>Chapter 19</a>)
+<p>
+Separates a logical prenex from a bridi or group of sentences to which it =
applies. Also separates a topic from a comment in topic/comment sentences.
+<p>
+<dl compact><dt>	<dd>su'o da poi remna ro da poi finpe zo'u da prami de
+	For-at-least-one X which is-a-man, for-all Ys which are-fish : X loves Y
+	There is a man who loves all fish.
+
+<img src=3Dchap21.gif alt=3D[Cartoon] align=3Dcenter width=3D405 height=3D=
405>
+
+<h2>Chapter 21
+<br>
+Formal Grammars</h2>
+<h6>$Revision: 4.0 $</h6>
+</p></center>
+<p>
+<cx "formal grammar">  XE "formal grammar"   <cx "machine grammar">  XE "m=
achine grammar"   The following two listings constitute the formal grammar =
of Lojban. The first version is written in the YACC language, which is used=
 to describe parsers, and has been used to create a parser for Lojban texts=
. This parser is available from the Logical Language Group. The second list=
ing is in Extended Backus-Naur Form (EBNF) and represents the same grammar =
in a more human-readable form. (In case of discrepancies, the YACC version =
is official.) There is a cross-reference listing for each format that shows=
, for each selma'o and rule, which rules refer to it.
+
+YACC Grammar of Lojban</h3>
+<landscape><cx "YACC grammar">  XE "YACC grammar"   <h3>
+<p align=3Dcenter><center>/*Lojban Machine Grammar, Final Baseline
+<br>The Lojban Machine Grammar document is explicitly dedicated to the pub=
lic domain by its author, The Logical Language Group, Inc.<br>
+</center></p>
+grammar.300 */
+<p>
+/* The Lojban machine parsing algorithm is a multi-step process. The YACC =
machine grammar presented here is an amalgam of those steps, concatenated s=
o as to allow YACC to verify the syntactic ambiguity of the grammar. YACC i=
s used to generate a parser for a portion of the grammar, which is LALR1 (t=
he type of grammar that YACC is designed to identify and process successful=
ly), but most of the rest of the grammar must be parsed using some language=
-coded processing.
+<p>
+<h4> Step 1 - Lexing</h4>
+<p>
+From phonemes, stress, and pause, it is possible to resolve Lojban  unambi=
guously into a stream of words. Any machine processing of speech will have =
to have some way to deal with 'non-Lojban' failures of fluent speech, of co=
urse. The resolved words can be expressed as a text file using Lojban's pho=
netic spelling rules.
+<p>
+The following steps assume that there is the possibility of non-Lojban tex=
t within the Lojban text (delimited appropriately). Such non-Lojban text ma=
y not be reducible from speech phonetically. However, step 2 allows the fil=
tering of a phonetically transcribed text stream, to recognize such portion=
s of non-Lojban text where properly delimited, without interference with th=
e parsing algorithm.
+<p>
+<p>
+<h4> Step 2 - Filtering</h4>
+<p>
+From start to end, performing the following filtering and lexing tasks usi=
ng the given order of precedence in case of conflict:
+<p>
+a. If the Lojban word =93zoi=94 (selma'o ZOI) is identified, take the foll=
owing Lojban word (which should be end delimited with a pause for separatio=
n from the following non-Lojban text) as an opening delimiter.
+Treat all text following that delimiter, until that delimiter recurs after=
 a pause, as grammatically a single token (labelled '<a href=3D#y699>anythi=
ng_699</a>' in this grammar). There is no need for processing within this t=
ext except as necessary to find the closing delimiter.
+<p>
+b. If the Lojban word =93zo=94 (selma'o ZO) is identified, treat the follo=
wing Lojban word as a token labelled '<a href=3D#y698>any_word_698</a>', in=
stead of lexing it by its normal grammatical function.
+<p>
+c. If the Lojban word =93lo'u=94 (selma'o LOhU) is identified, search for =
the closing delimiter =93le'u=94 (selma'o LEhU), ignoring any such closing =
delimiters absorbed by the previous two steps. The text between the delimit=
ers should be treated as the single token '<a href=3D#y697>any_words_697</a=
>'.
+<p>
+d. Categorize all remaining words into their Lojban selma'o category, incl=
uding the various delimiters mentioned in the previous steps. In all steps =
after step 2, only the selma'o token type is significant for each word.
+<p>
+e. If the word =93si=94 (selma'o SI) is identified, erase it and the previ=
ous word (or token, if the previous text has been condensed into a single t=
oken by one of the above rules).
+<p>
+f. If the word =93sa=94 (selma'o SA) is identified, erase it and all prece=
ding text as far back as necessary to make what follows attach to what prec=
edes. (This rule is hard to formalize and may receive further definition la=
ter.)
+<p>
+g. If the word 'su' (selma'o SU) is identified, erase it and all preceding=
 text back to and including the first preceding token word which is in one =
of the selma'o: NIhO, LU, TUhE, and TO. However, if speaker identification =
is available, a SU shall only erase to the beginning of a speaker's discour=
se, unless it occurs at the beginning of a speaker's discourse. (Thus, if t=
he speaker has said something, two =93su=94s are required to erase the enti=
re conversation.
+<p>
+<p>
+<h4> Step 3 - Termination</h4>
+<p>
+If the text contains a FAhO, treat that as the end-of-text and ignore ever=
ything that follows it.
+<p>
+<p>
+<h4> Step 4 - Absorption of Grammar-Free Tokens</h4>
+<p>
+In a new pass, perform the following absorptions (absorption means that th=
e token is removed from the grammar for processing in following steps, and =
optionally reinserted, grouped with the absorbing token after parsing is co=
mpleted).
+<p>
+a. Token sequences of the form any - (ZEI - any)=A0<dots>=85</dots>, where=
 there may be any number of ZEIs, are merged into a single token of selma'o=
 BRIVLA.
+<p>
+b. Absorb all selma'o BAhE tokens into the following token. If they occur =
at the end of text, leave them alone (they are errors).
+<p>
+c. Absorb all selma'o BU tokens into the previous token. Relabel the previ=
ous token as selma'o BY.
+<p>
+d. If selma'o NAI occurs immediately following any of tokens UI or CAI, ab=
sorb the NAI into the previous token.
+<p>
+e. Absorb all members of selma'o DAhO, FUhO, FUhE, UI, Y, and CAI into the=
 previous token. All of these null grammar tokens are permitted following a=
ny word of the grammar, without interfering with that word's grammatical fu=
nction, or causing any effect on the grammatical interpretation of any othe=
r token in the text. Indicators at the beginning of text are explicitly han=
dled by the grammar.
+<p>
+<p>
+<h4> Step 5 - Insertion of Lexer Lexemes</h4>
+<p>
+Lojban is not in itself LALR1. There are words whose grammatical function =
is determined by following tokens. As a result, parsing of the YACC grammar=
 must take place in two steps. In the first step, certain strings of tokens=
 with defined grammars are identified, and either
+<p>
+a. are replaced by a single specified 'lexer token' for step 6, or
+<p>
+b. the lexer token is inserted in front of the token string to identify it=
 uniquely.
+<p>
+The YACC grammar included herein is written to make YACC generation of a s=
tep 6 parser easy regardless of whether a. or b. is used. The strings of to=
kens to be labelled with lexer tokens are found in rule terminals labelled =
with numbers between 900 and 1099. These rules are defined with the lexer t=
okens inserted, with the result that it can be verified that the language i=
s LALR1 under option b. after steps 1 through 4 have been performed. Altern=
atively, if option a. is to be used, these rules are commented out, and the=
 rule terminals labelled from 800 to 900 refer to the lexer tokens without =
the strings of defining tokens. Two sets of lexer tokens are defined in the=
 token set so as to be compatible with either option.
+<p>
+In this step, the strings must be labelled with the appropriate lexer toke=
ns. Order of inserting lexer tokens IS significant, since some shorter stri=
ngs that would be marked with a lexer token may be found inside longer stri=
ngs. If the tokens are inserted before or in place of the shorter strings, =
the longer strings cannot be identified.
+<p>
+If option a. is chosen, the following order of insertion works correctly (=
it is not the only possible order): A, C, D, B, U, E, H, I, J, K, M, N, G, =
O, V, W, F, P, R, T, S, Y, L, Q. This ensures that the longest rules will b=
e processed first; a PA+MAI will not be seen as a PA with a dangling MAI at=
 the end, for example.
+<p>
+<p>
+<h4> Step 6 - YACC Parsing</h4>
+<p>
+YACC should now be able to parse the Lojban text in accordance with the ru=
le terminals labelled from 1 to 899 under option 5a, or 1 to 1099 under opt=
ion 5b. Comment out the rules beyond 900 if option 5a is used, and comment =
out the 700-series of lexer-tokens, while restoring the series of lexer tok=
ens numbered from 900 up.  <p>
+*/
+<pre>%token <a name=3Dy501>A_501 </a>	/* eks; basic afterthought logical c=
onnectives */
+%token <a name=3Dy502>BAI_502 </a>	/* modal operators */
+%token <a name=3Dy503>BAhE_503 </a>	/* next word intensifier */
+%token <a name=3Dy504>BE_504 </a>	/* sumti link to attach sumti to a selbr=
i */
+%token <a name=3Dy505>BEI_505 </a>	/* multiple sumti separator between BE,=
 BEI */
+%token <a name=3Dy506>BEhO_506 </a>	/* terminates BE/BEI specified descrip=
tors */
+%token <a name=3Dy507>BIhI_507 </a>	/* interval component of JOI */
+%token <a name=3Dy508>BO_508 </a>	/* joins two units with shortest scope *=
/
+%token <a name=3Dy509>BRIVLA_509 </a>	/* any brivla */
+%token <a name=3Dy511>BU_511 </a>	/* turns any word into a BY lerfu word *=
/
+%token <a name=3Dy513>BY_513 </a>	/* individual lerfu words */
+%token <a name=3Dy514>CAhA_514 </a>	/* specifies actuality/potentiality of=
 tense */
+%token <a name=3Dy515>CAI_515 </a>	/* afterthought intensity marker */
+%token <a name=3Dy516>CEI_516 </a>	/* pro-bridi assignment operator */
+%token <a name=3Dy517>CEhE_517 </a>	/* afterthought term list connective *=
/
+%token <a name=3Dy518>CMENE_518 </a>	/* names; require consonant end, then=
 pause no LA or DOI selma'o embedded, pause before if vowel initial and pre=
ceded by a vowel */
+%token <a name=3Dy519>CO_519 </a>	/* tanru inversion */
+%token <a name=3Dy520>COI_520 </a>	/* vocative marker permitted inside nam=
es; must always be followed by pause or DOI */
+%token <a name=3Dy521>CU_521 </a>	/* separator between head sumti and selb=
ri */
+%token <a name=3Dy522>CUhE_522 </a>	/* tense/modal question */
+%token <a name=3Dy524>DAhO_524 </a>	/* cancel anaphora/cataphora assignmen=
ts */
+%token <a name=3Dy525>DOI_525 </a>	/* vocative marker */
+%token <a name=3Dy526>DOhU_526 </a>	/* terminator for DOI-marked vocatives=
 */
+%token <a name=3Dy527>FA_527 </a>	/* modifier head generic case tag */
+%token <a name=3Dy528>FAhA_528 </a>	/* superdirections in space */
+%token <a name=3Dy529>FAhO_529 </a>	/* normally elided 'done pause' to ind=
icate end of utterance string */
+%token <a name=3Dy530>FEhE_530 </a>	/* space interval mod flag */
+%token <a name=3Dy531>FEhU_531 </a>	/* ends bridi to modal conversion */
+%token <a name=3Dy532>FIhO_532 </a>	/* marks bridi to modal conversion */
+%token <a name=3Dy533>FOI_533 </a>	/* end compound lerfu */
+%token <a name=3Dy535>FUhE_535 </a>	/* open long scope for indicator */
+%token <a name=3Dy536>FUhO_536 </a>	/* close long scope for indicator */
+%token <a name=3Dy537>GA_537 </a>	/* geks; forethought logical connectives=
 */
+%token <a name=3Dy538>GEhU_538 </a>	/* marker ending GOI relative clauses =
*/
+%token <a name=3Dy539>GI_539 </a>	/* forethought medial marker */
+%token <a name=3Dy541>GIhA_541 </a>	/* logical connectives for bridi-tails=
 */
+%token <a name=3Dy542>GOI_542   </a>	/* attaches a sumti modifier to a sum=
ti */
+%token <a name=3Dy543>GOhA_543  </a>	/* pro-bridi */
+%token <a name=3Dy544>GUhA_544  </a>	/* GEK for tanru units, corresponds t=
o JEKs */
+%token <a name=3Dy545>I_545  </a>	/* sentence link */
+%token <a name=3Dy546>JA_546  </a>	/* jeks; logical connectives within tan=
ru */
+%token <a name=3Dy547>JAI_547  </a>	/* modal conversion flag */
+%token <a name=3Dy548>JOI_548  </a>	/* non-logical connectives */
+%token <a name=3Dy550>KEhE_550   </a>	/* right terminator for KE groups */
+%token <a name=3Dy551>KE_551   </a>	/* left long scope marker */
+%token <a name=3Dy552>KEI_552   </a>	/* right terminator, NU abstractions =
*/
+%token <a name=3Dy554>KI_554  </a>	/* multiple utterance scope for tenses =
*/
+%token <a name=3Dy555>KOhA_555  </a>	/* sumti anaphora */
+%token <a name=3Dy556>KU_556  </a>	/* right terminator for descriptions, e=
tc. */
+%token <a name=3Dy557>KUhO_557  </a>	/* right terminator, NOI relative cla=
uses */
+%token <a name=3Dy558>LA_558  </a>	/* name descriptors */
+%token <a name=3Dy559>LAU_559  </a>	/* lerfu prefixes */
+%token <a name=3Dy561>LAhE_561   </a>	/* sumti qualifiers */
+%token <a name=3Dy562>LE_562   </a>	/* sumti descriptors */
+%token <a name=3Dy565>LEhU_565  </a>	/* possibly ungrammatical text right =
quote */
+%token <a name=3Dy566>LI_566  </a>	/* convert number to sumti */
+%token <a name=3Dy567>LIhU_567  </a>	/* grammatical text right quote */
+%token <a name=3Dy568>LOhO_568  </a>	/* elidable terminator for LI */
+%token <a name=3Dy569>LOhU_569  </a>	/* possibly ungrammatical text left q=
uote */
+%token <a name=3Dy571>LU_571   </a>	/* grammatical text left quote */
+%token <a name=3Dy573>LUhU_573  </a>	/* LAhE close delimiter */
+%token <a name=3Dy574>ME_574  </a>	/* converts a sumti into a tanru_unit *=
/
+%token <a name=3Dy575>MEhU_575  </a>	/* terminator for ME */
+%token <a name=3Dy577>MOhI_577  </a>	/* motion tense marker */
+%token <a name=3Dy578>NA_578  </a>	/* bridi negation */
+%token <a name=3Dy581>NAI_581   </a>	/* attached to words to negate them *=
/
+%token <a name=3Dy583>NAhE_583  </a>	/* scalar negation */
+%token <a name=3Dy584>NIhO_584  </a>	/* new paragraph; change of subject *=
/
+%token <a name=3Dy585>NOI_585  </a>	/* attaches a subordinate clause to a =
sumti */
+%token <a name=3Dy586>NU_586  </a>	/* abstraction */
+%token <a name=3Dy587>NUhI_587  </a>	/* marks the start of a termset */
+%token <a name=3Dy588>NUhU_588  </a>	/* marks the middle and end of a term=
set */
+%token <a name=3Dy591>PEhE_591   </a>	/* afterthought termset connective p=
refix */
+%token <a name=3Dy592>PU_592   </a>	/* directions in time */
+%token <a name=3Dy593>RAhO_593  </a>	/* flag for modified interpretation o=
f GOhI */
+%token <a name=3Dy594>ROI_594  </a>	/* converts number to extensional tens=
e */
+%token <a name=3Dy595>SA_595  </a>	/* metalinguistic eraser to the beginni=
ng of the current utterance */
+%token <a name=3Dy596>SE_596  </a>	/* conversions */
+%token <a name=3Dy597>SEI_597  </a>	/* metalinguistic bridi insert marker =
*/
+%token <a name=3Dy598>SEhU_598  </a>	/* metalinguistic bridi end marker */
+%token <a name=3Dy601>SI_601   </a>	/* metalinguistic single word eraser *=
/
+%token <a name=3Dy602>SOI_602   </a>	/* reciprocal sumti marker */
+%token <a name=3Dy603>SU_603  </a>	/* metalinguistic eraser of the entire =
text */
+%token <a name=3Dy604>TAhE_604  </a>	/* tense interval properties */
+%token <a name=3Dy605>TEI_605  </a>	/* start compound lerfu */
+%token <a name=3Dy606>TO_606  </a>	/* left discursive parenthesis */
+%token <a name=3Dy607>TOI_607  </a>	/* right discursive parenthesis */
+%token <a name=3Dy610>TUhE_610   </a>	/* multiple utterance scope mark */
+%token <a name=3Dy611>TUhU_611   </a>	/* multiple utterance end scope mark=
 */
+%token <a name=3Dy612>UI_612   </a>	/* attitudinals, observationals, discu=
rsives */
+%token <a name=3Dy613>VA_613  </a>	/* distance in space-time */
+%token <a name=3Dy614>VAU_614  </a>	/* end simple bridi or bridi-tail */
+%token <a name=3Dy615>VEhA_615  </a>	/* space-time interval size */
+%token <a name=3Dy616>VIhA_616  </a>	/* space-time dimensionality marker *=
/
+%token <a name=3Dy617>VUhO_617  </a>	/* glue between logically connected s=
umti and relative clauses */
+%token <a name=3Dy618>XI_618  </a>	/* subscripting operator */
+%token <a name=3Dy619>Y_619  </a>	/* hesitation */
+%token <a name=3Dy621>ZAhO_621   </a>	/* event properties - inchoative, et=
c. */
+%token <a name=3Dy622>ZEhA_622   </a>	/* time interval size tense */
+%token <a name=3Dy623>ZEI_623  </a>	/* lujvo glue */
+%token <a name=3Dy624>ZI_624  </a>	/* time distance tense */
+%token <a name=3Dy625>ZIhE_625  </a>	/* conjoins relative clauses */
+%token <a name=3Dy626>ZO_626  </a>	/* single word metalinguistic quote mar=
ker */
+%token <a name=3Dy627>ZOI_627  </a>	/* delimited quote marker */
+%token <a name=3Dy628>ZOhU_628  </a>	/* prenex terminator (not elidable) *=
/
+%token <a name=3Dy650>BIhE_650   </a>	/* prefix for high-priority MEX oper=
ator */
+%token <a name=3Dy651>BOI_651   </a>	/* number or lerfu-string terminator =
*/
+%token <a name=3Dy655>FUhA_655  </a>	/* reverse Polish flag */
+%token <a name=3Dy656>GAhO_656  </a>	/* open/closed interval markers for B=
IhI */
+%token <a name=3Dy657>JOhI_657  </a>	/* flags an array operand */
+%token <a name=3Dy658>KUhE_658  </a>	/* MEX forethought delimiter */
+%token <a name=3Dy661>MAI_661   </a>	/* change numbers to utterance ordina=
ls */
+%token <a name=3Dy662>MAhO_662   </a>	/* change MEX expressions to MEX ope=
rators */
+%token <a name=3Dy663>MOI_663  </a>	/* change number to selbri */
+%token <a name=3Dy664>MOhE_664  </a>	/* change sumti to operand, inverse o=
f LI */
+%token <a name=3Dy665>NAhU_665  </a>	/* change a selbri into an operator *=
/
+%token <a name=3Dy666>NIhE_666  </a>	/* change selbri to operand; inverse =
of MOI */
+%token <a name=3Dy667>NUhA_667  </a>	/* change operator to selbri; inverse=
 of MOhE */
+%token <a name=3Dy672>PA_672   </a>	/* numbers and numeric punctuation */
+%token <a name=3Dy673>PEhO_673  </a>	/* forethought (Polish) flag */
+%token <a name=3Dy675>TEhU_675  </a>	/* closing gap for MEX constructs */
+%token <a name=3Dy677>VEI_677  </a>	/* left MEX bracket */
+%token <a name=3Dy678>VEhO_678  </a>	/* right MEX bracket */
+%token <a name=3Dy679>VUhU_679  </a>	/* MEX operator */
+%token any_words_697  </a>	/* a string of lexable Lojban words */
+%token any_word_698  </a>	/* any single lexable Lojban words */
+%token anything_699  </a>	/* a possibly unlexable phoneme string */
+/* The following tokens are the actual lexer tokens. The _900 series token=
s are duplicates that allow limited testing of lexer rules in the context o=
f the total grammar. They are used in the actual parser, where the 900 seri=
es rules are found in the lexer. */
+%token lexer_<a name=3Dy701>A_701   </a>	/* flags a MAI utterance ordinal =
*/
+%token lexer_<a name=3Dy702>B_702   </a>	/* flags an EK unless EK_BO, EK_K=
E */
+%token lexer_<a name=3Dy703>C_703  </a>	/* flags an EK_BO */
+%token lexer_<a name=3Dy704>D_704  </a>	/* flags an EK_KE */
+%token lexer_<a name=3Dy705>E_705  </a>	/* flags a JEK */
+%token lexer_<a name=3Dy706>F_706  </a>	/* flags a JOIK */
+%token lexer_<a name=3Dy707>G_707  </a>	/* flags a GEK */
+%token lexer_<a name=3Dy708>H_708  </a>	/* flags a GUhEK */
+%token lexer_<a name=3Dy709>I_709  </a>	/* flags a NAhE_BO */
+%token lexer_<a name=3Dy710>J_710   </a>	/* flags a NA_KU */
+%token lexer_<a name=3Dy711>K_711   </a>	/* flags an I_BO (option. JOIK/JE=
K lexer tags)*/
+%token lexer_<a name=3Dy712>L_712   </a>	/* flags a PA, unless MAI (then l=
exer A) */
+%token lexer_<a name=3Dy713>M_713  </a>	/* flags a GIhEK_BO */
+%token lexer_<a name=3Dy714>N_714  </a>	/* flags a GIhEK_KE */
+%token lexer_<a name=3Dy715>O_715  </a>	/* flags a modal operator BAI or c=
ompound */
+%token lexer_<a name=3Dy716>P_716  </a>	/* flags a GIK */
+%token lexer_<a name=3Dy717>Q_717  </a>	/* flags a lerfu_string unless MAI=
 (then lexer_A)*/
+%token lexer_<a name=3Dy718>R_718  </a>	/* flags a GIhEK, not BO or KE */
+%token lexer_<a name=3Dy719>S_719  </a>	/* flags simple I */
+%token lexer_<a name=3Dy720>T_720   </a>	/* flags I_JEK */
+%token lexer_<a name=3Dy721>U_721   </a>	/* flags a JEK_BO */
+%token lexer_<a name=3Dy722>V_722   </a>	/* flags a JOIK_BO */
+%token lexer_<a name=3Dy723>W_723  </a>	/* flags a JOIK_KE */	/*
+%token lexer_<a name=3Dy724>X_724  </a>	/* null */
+%token lexer_<a name=3Dy725>Y_725  </a>	/* flags a PA_MOI */
+	/*
+%token lexer_A_905	/* :	lexer_A_701   utt_ordinal_root_906  */	/*
+%token lexer_B_910	/* :	lexer_B_702   EK_root_911   */	/*
+%token lexer_C_915	/* :	lexer_C_703  EK_root_911   BO_508  */	/*
+%token lexer_D_916	/* :	lexer_D_704  EK_root_911   KE_551   */	/*
+%token lexer_E_925	/* :	lexer_E_705  JEK_root_926  */	/*
+%token lexer_F_930	/* :	lexer_F_706  JOIK_root_931   */	/*
+%token lexer_G_935	/* :	lexer_G_707  GA_537  */	/*
+%token lexer_H_940	/* :	lexer_H_708  GUhA_544  */	/*
+%token lexer_I_945	/* :	lexer_I_709  NAhE_583  BO_508  */	/*
+%token lexer_J_950	/* :	lexer_J_710   NA_578  KU_556  */	/*
+%token lexer_K_955	/* :	lexer_K_711   I_432   BO_508  */	/*
+%token lexer_L_960	/* :	lexer_L_712   number_root_961   */	/*
+%token lexer_M_965	/* :	lexer_M_713  GIhEK_root_991   BO_508  */	/*
+%token lexer_N_966	/* :	lexer_N_714  GIhEK_root_991   KE_551   */	/*
+%token lexer_O_970	/* :	lexer_O_715  simple_tense_modal_972   */	/*
+%token lexer_P_980	/* :	lexer_P_716  GIK_root_981   */	/*
+%token lexer_Q_985	/* :	lexer_Q_717  lerfu_string_root_986  */	/*
+%token lexer_R_990	/* :	lexer_R_718  GIhEK_root_991   */	/*
+%token lexer_S_995	/* :	lexer_S_719  I_545  */	/*
+%token lexer_T_1000	/* :	lexer_T_720   I_545  simple_JOIK_JEK_957  */	/*
+%token lexer_U_1005	/* :	lexer_U_721   JEK_root_926  BO_508  */	/*
+%token lexer_V_1010	/* :	lexer_V_722   JOIK_root_931   BO_508  */	/*
+%token lexer_W_1015	/* :	lexer_W_723  JOIK_root_931   KE_551   */	/*
+%token lexer_X_1020	/* null */	/*
+%token lexer_Y_1025	/* :	lexer_Y_725  number_root_961   MOI_663  */
+%start <a href=3D#y0>text_0   </a>
+%%
+<a name=3Dy0>text_0  </a>	:	<a href=3D#y1>text_A_1  </a>
+	|	<a href=3D#y411>indicators_411   </a><a href=3D#y1>text_A_1   </a>
+	|	<a href=3D#y32>free_modifier_32   </a><a href=3D#y1>text_A_1   </a>
+	|	<a href=3D#y404>cmene_404  </a><a href=3D#y1>text_A_1   </a>
+	|	<a href=3D#y411>indicators_411   </a><a href=3D#y32>free_modifier_32   =
</a><a href=3D#y1>text_A_1   </a>
+	|	<a href=3D#y581>NAI_581   </a><a href=3D#y0>text_0   </a>
+	;
+<a name=3Dy1>text_A_1  </a>	:	<a href=3D#y422>JOIK_JEK_422  </a><a href=3D=
#y2>text_B_2  </a>
+		/* incomplete JOIK_JEK without preceding I */
+		/* compare note on <a href=3D#y10>paragraph_10   </a>*/
+	|	<a href=3D#y2>text_B_2   </a>
+	;
+<a name=3Dy2>text_B_2  </a>	:	<a href=3D#y819>I_819  </a><a href=3D#y2>tex=
t_B_2  </a>
+	|	<a href=3D#y820>I_JEK_820   </a><a href=3D#y2>text_B_2   </a>
+	|	<a href=3D#y811>I_BO_811   </a><a href=3D#y2>text_B_2   </a>
+	|	<a href=3D#y410>para_mark_410   </a><a href=3D#y3>text_C_3  </a>
+	|	<a href=3D#y3>text_C_3  </a>
+	;
+<a name=3Dy3>text_C_3  </a>	:	<a href=3D#y4>paragraphs_4  </a>
+/* Only indicators which follow certain selma'o: cmene, <a href=3D#y607>TO=
I_607 </a>, <a href=3D#y571>LU_571 </a>, and the lexer_K and lexer_S I_root=
s and compounds, and at the start of text(_0), will survive the lexer; all =
other valid ones will be absorbed. The only strings for which indicators ge=
nerate a potential ambiguity are those which contain NAI. An indicator cann=
ot be inserted in between a token and its negating NAI, else you can't tell=
 whether it is the indicator or the original token being negated. */
+	|	/* empty */
+/* An empty text is legal; formerly this was handled by the explicit appea=
rance of <a href=3D#y529>FAhO_529 </a>, but this is now absorbed by the pre=
parser. */
+	;
+<a name=3Dy4>paragraphs_4  </a>	:	<a href=3D#y10>paragraph_10  </a>
+	|	<a href=3D#y10>paragraph_10   </a><a href=3D#y410>para_mark_410   </a><=
a href=3D#y4>paragraphs_4  </a>
+	;
+
+<a name=3Dy10>paragraph_10  </a>	:	<a href=3D#y11>statement_11  </a>
+	|	<a href=3D#y20>fragment_20   </a>
+	|	<a href=3D#y10>paragraph_10   </a><a href=3D#y819>I_819  </a><a href=3D=
#y11>statement_11   </a>
+	|	<a href=3D#y10>paragraph_10   </a><a href=3D#y819>I_819  </a><a href=3D=
#y20>fragment_20   </a>
+	|	<a href=3D#y10>paragraph_10   </a><a href=3D#y819>I_819  </a>
+/* this last fixes an erroneous start to a sentence, and permits incomplet=
e JOIK_JEK after I, as well in answer to questions on those connectives */
+	;
+<a name=3Dy11>statement_11  </a>	:	<a href=3D#y12>statement_A_12  </a>
+	|	<a href=3D#y30>prenex_30   </a><a href=3D#y11>statement_11   </a>
+	;
+<a name=3Dy12>statement_A_12  </a>	:	<a href=3D#y13>statement_B_13  </a>
+	|	<a href=3D#y12>statement_A_12   </a><a href=3D#y820>I_JEK_820   </a><a =
href=3D#y13>statement_B_13  </a>
+	|	<a href=3D#y12>statement_A_12   </a><a href=3D#y820>I_JEK_820   </a>
+	;
+<a name=3Dy13>statement_B_13  </a>	:	<a href=3D#y14>statement_C_14  </a>
+	|	<a href=3D#y14>statement_C_14  </a><a href=3D#y811>I_BO_811   </a><a hr=
ef=3D#y13>statement_B_13  </a>
+	|	<a href=3D#y14>statement_C_14  </a><a href=3D#y811>I_BO_811   </a>
+	;
+<a name=3Dy14>statement_C_14  </a>	:	<a href=3D#y40>sentence_40  </a>
+	|	<a href=3D#y447>TUhE_447  </a><a href=3D#y2>text_B_2   </a><a href=3D#y=
454>TUhU_gap_454  </a>
+	|	<a href=3D#y491>tag_491   </a><a href=3D#y447>TUhE_447  </a><a href=3D#=
y2>text_B_2   </a><a href=3D#y454>TUhU_gap_454  </a>
+	;
+<a name=3Dy20>fragment_20  </a>	:	<a href=3D#y802>EK_802  </a>
+	|	<a href=3D#y445>NA_445  </a>
+	|	<a href=3D#y818>GIhEK_818  </a>
+	|	<a href=3D#y300>quantifier_300   </a>
+	|	<a href=3D#y80>terms_80   </a><a href=3D#y456>VAU_gap_456   </a>/* answ=
er to ma */
+/* <a href=3D#y490>mod_head_490 </a>requires both <a href=3D#y450>gap_450 =
</a>and <a href=3D#y456>VAU_gap_456 </a> but needs no extra rule to accompl=
ish this */
+	|	<a href=3D#y121>relative_clauses_121   </a>
+	|	<a href=3D#y161>links_161   </a>
+	|	<a href=3D#y160>linkargs_160   </a>
+	|	<a href=3D#y30>prenex_30   </a>
+	;
+<a name=3Dy30>prenex_30  </a>	:	<a href=3D#y80>terms_80  </a><a href=3D#y4=
92>ZOhU_492  </a>
+	;
+
+<a name=3Dy32>free_modifier_32  </a>	:	<a href=3D#y33>free_modifier_A_33  =
</a>
+	|	<a href=3D#y33>free_modifier_A_33  </a><a href=3D#y32>free_modifier_32 =
  </a>
+	;
+<a name=3Dy33>free_modifier_A_33	</a>:	<a href=3D#y35>vocative_35  </a>
+	|	<a href=3D#y36>parenthetical_36  </a>
+	|	<a href=3D#y34>discursive_bridi_34  </a>
+	|	<a href=3D#y486>subscript_486  </a>
+	|	<a href=3D#y801>utterance_ordinal_801   </a>
+	;
+<a name=3Dy34>discursive_bridi_34	</a>:	<a href=3D#y440>SEI_440  </a><a hr=
ef=3D#y130>selbri_130  </a><a href=3D#y459>SEhU_gap_459  </a>
+	|	<a href=3D#y498>SOI_498  </a><a href=3D#y90>sumti_90   </a><a href=3D#y=
459>SEhU_gap_459  </a>
+	|	<a href=3D#y498>SOI_498  </a><a href=3D#y90>sumti_90   </a><a href=3D#y=
90>sumti_90   </a><a href=3D#y459>SEhU_gap_459  </a>
+	|	<a href=3D#y440>SEI_440   </a><a href=3D#y80>terms_80   </a><a href=3D#=
y451>front_gap_451   </a><a href=3D#y130>selbri_130   </a><a href=3D#y459>S=
EhU_gap_459  </a>
+	|	<a href=3D#y440>SEI_440   </a><a href=3D#y80>terms_80   </a><a href=3D#=
y130>selbri_130   </a><a href=3D#y459>SEhU_gap_459  </a>
+	;
+<a name=3Dy35>vocative_35  </a>	:	<a href=3D#y415>DOI_415  </a><a href=3D#=
y130>selbri_130  </a><a href=3D#y457>DOhU_gap_457  </a>
+	|	<a href=3D#y415>DOI_415  </a><a href=3D#y130>selbri_130   </a><a href=
=3D#y121>relative_clauses_121   </a><a href=3D#y457>DOhU_gap_457  </a>
+	|	<a href=3D#y415>DOI_415  </a><a href=3D#y121>relative_clauses_121   </a=
><a href=3D#y130>selbri_130   </a><a href=3D#y457>DOhU_gap_457  </a>
+	|	<a href=3D#y415>DOI_415  </a><a href=3D#y121>relative_clauses_121   </a=
><a href=3D#y130>selbri_130   </a><a href=3D#y121>relative_clauses_121   </=
a><a href=3D#y457>DOhU_gap_457  </a>
+	|	<a href=3D#y415>DOI_415  </a><a href=3D#y404>cmene_404  </a><a href=3D#=
y457>DOhU_gap_457  </a>
+	|	<a href=3D#y415>DOI_415  </a><a href=3D#y404>cmene_404  </a><a href=3D#=
y121>relative_clauses_121   </a><a href=3D#y457>DOhU_gap_457  </a>
+	|	<a href=3D#y415>DOI_415  </a><a href=3D#y121>relative_clauses_121   </a=
><a href=3D#y404>cmene_404  </a><a href=3D#y457>DOhU_gap_457  </a>
+	|	<a href=3D#y415>DOI_415  </a><a href=3D#y121>relative_clauses_121   </a=
><a href=3D#y404>cmene_404  </a><a href=3D#y121>relative_clauses_121   </a>=
<a href=3D#y457>DOhU_gap_457  </a>
+	|	<a href=3D#y415>DOI_415  </a><a href=3D#y90>sumti_90   </a><a href=3D#y=
457>DOhU_gap_457  </a>
+	|	<a href=3D#y415>DOI_415  </a><a href=3D#y457>DOhU_gap_457  </a>
+	;
+<a name=3Dy36>parenthetical_36  </a>	:	<a href=3D#y606>TO_606  </a><a href=
=3D#y0>text_0  </a><a href=3D#y468>TOI_gap_468  </a>
+	;
+<a name=3Dy40>sentence_40  </a>	:	<a href=3D#y50>bridi_tail_50	</a>/* bare=
 observative or mo answer */
+	|	<a href=3D#y80>terms_80   </a><a href=3D#y451>front_gap_451   </a><a hr=
ef=3D#y50>bridi_tail_50   </a>
+	|	<a href=3D#y80>terms_80   </a><a href=3D#y50>bridi_tail_50   </a>
+	;
+<a name=3Dy41>subsentence_41  </a>	:	<a href=3D#y40>sentence_40  </a>
+	|	<a href=3D#y30>prenex_30   </a><a href=3D#y41>subsentence_41   </a>
+	;
+
+<a name=3Dy50>bridi_tail_50  </a>	:	<a href=3D#y51>bridi_tail_A_51  </a>
+	|	<a href=3D#y51>bridi_tail_A_51   </a><a href=3D#y814>GIhEK_KE_814  </a>=
<a href=3D#y50>bridi_tail_50   </a><a href=3D#y466>KEhE_gap_466  </a><a hre=
f=3D#y71>tail_terms_71   </a>
+	;
+<a name=3Dy51>bridi_tail_A_51  </a>	:	<a href=3D#y52>bridi_tail_B_52  </a>
+	|	<a href=3D#y51>bridi_tail_A_51   </a><a href=3D#y818>GIhEK_818  </a><a =
href=3D#y52>bridi_tail_B_52   </a><a href=3D#y71>tail_terms_71   </a>
+	;
+<a name=3Dy52>bridi_tail_B_52  </a>	:	<a href=3D#y53>bridi_tail_C_53  </a>
+	|	<a href=3D#y53>bridi_tail_C_53  </a><a href=3D#y813>GIhEK_BO_813  </a><=
a href=3D#y52>bridi_tail_B_52   </a><a href=3D#y71>tail_terms_71   </a>
+	;
+<a name=3Dy53>bridi_tail_C_53  </a>	:	<a href=3D#y54>gek_sentence_54  </a>
+	|	<a href=3D#y130>selbri_130   </a><a href=3D#y71>tail_terms_71   </a>
+	;
+<a name=3Dy54>gek_sentence_54  </a>	:	<a href=3D#y807>GEK_807  </a><a href=
=3D#y41>subsentence_41  </a><a href=3D#y816>GIK_816  </a><a href=3D#y41>sub=
sentence_41   </a><a href=3D#y71>tail_terms_71   </a>
+	|	<a href=3D#y491>tag_491   </a><a href=3D#y493>KE_493  </a><a href=3D#y5=
4>gek_sentence_54  </a><a href=3D#y466>KEhE_gap_466  </a>
+	|	<a href=3D#y445>NA_445  </a><a href=3D#y54>gek_sentence_54  </a>
+	;
+<a name=3Dy71>tail_terms_71  </a>	:	<a href=3D#y80>terms_80  </a><a href=
=3D#y456>VAU_gap_456  </a>
+	|	<a href=3D#y456>VAU_gap_456  </a>
+	;
+<a name=3Dy80>terms_80  </a>	:	<a href=3D#y81>terms_A_81  </a>
+	|	<a href=3D#y80>terms_80   </a><a href=3D#y81>terms_A_81   </a>
+	;
+<a name=3Dy81>terms_A_81  </a>	:	<a href=3D#y82>terms_B_82  </a>
+	|	<a href=3D#y81>terms_A_81   </a><a href=3D#y494>PEhE_494  </a><a href=
=3D#y422>JOIK_JEK_422   </a><a href=3D#y82>terms_B_82   </a>
+	;
+<a name=3Dy82>terms_B_82  </a>	:	<a href=3D#y83>term_83  </a>
+	|	<a href=3D#y82>terms_B_82   </a><a href=3D#y495>CEhE_495  </a><a href=
=3D#y83>term_83  </a>
+	;
+<a name=3Dy83>term_83  </a>	:	<a href=3D#y90>sumti_90  </a>
+	|	<a href=3D#y84>modifier_84  </a>
+	|	<a href=3D#y85>term_set_85  </a>
+	|	<a href=3D#y810>NA_KU_810   </a>
+	;
+<a name=3Dy84>modifier_84  </a>	:	<a href=3D#y490>mod_head_490  </a><a hre=
f=3D#y450>gap_450  </a>
+	|	<a href=3D#y490>mod_head_490   </a><a href=3D#y90>sumti_90   </a>
+	;
+<a name=3Dy85>term_set_85  </a>	:	<a href=3D#y496>NUhI_496  </a><a href=3D=
#y80>terms_80  </a><a href=3D#y460>NUhU_gap_460  </a>
+	|	<a href=3D#y496>NUhI_496  </a><a href=3D#y807>GEK_807  </a><a href=3D#y=
80>terms_80   </a><a href=3D#y460>NUhU_gap_460   </a><a href=3D#y816>GIK_81=
6  </a><a href=3D#y80>terms_80   </a><a href=3D#y460>NUhU_gap_460   </a>
+	;
+<a name=3Dy90>sumti_90  </a>	:	<a href=3D#y91>sumti_A_91  </a>
+	|	<a href=3D#y91>sumti_A_91   </a><a href=3D#y497>VUhO_497  </a><a href=
=3D#y121>relative_clauses_121   </a>
+	;
+<a name=3Dy91>sumti_A_91  </a>	:	<a href=3D#y92>sumti_B_92  </a>
+	|	<a href=3D#y92>sumti_B_92   </a><a href=3D#y804>EK_KE_804  </a><a href=
=3D#y90>sumti_90   </a><a href=3D#y466>KEhE_gap_466  </a>
+	|	<a href=3D#y92>sumti_B_92   </a><a href=3D#y823>JOIK_KE_823  </a><a hre=
f=3D#y90>sumti_90   </a><a href=3D#y466>KEhE_gap_466  </a>
+	;
+<a name=3Dy92>sumti_B_92  </a>	:	<a href=3D#y93>sumti_C_93  </a>
+	|	<a href=3D#y92>sumti_B_92   </a><a href=3D#y421>JOIK_EK_421   </a><a hr=
ef=3D#y93>sumti_C_93  </a>
+	;
+<a name=3Dy93>sumti_C_93  </a>	:	<a href=3D#y94>sumti_D_94  </a>
+	|	<a href=3D#y94>sumti_D_94  </a><a href=3D#y803>EK_BO_803  </a><a href=
=3D#y93>sumti_C_93  </a>
+	|	<a href=3D#y94>sumti_D_94  </a><a href=3D#y822>JOIK_BO_822   </a><a hre=
f=3D#y93>sumti_C_93  </a>
+	;
+<a name=3Dy94>sumti_D_94  </a>	:	<a href=3D#y95>sumti_E_95  </a>
+	|	<a href=3D#y807>GEK_807  </a><a href=3D#y90>sumti_90   </a><a href=3D#y=
816>GIK_816  </a><a href=3D#y94>sumti_D_94  </a>
+	;
+<a name=3Dy95>sumti_E_95  </a>	:	<a href=3D#y96>sumti_F_96  </a>
+	|	<a href=3D#y96>sumti_F_96  </a><a href=3D#y121>relative_clauses_121   <=
/a>
+		/* indefinite sumti */
+	|	<a href=3D#y300>quantifier_300   </a><a href=3D#y130>selbri_130   </a><=
a href=3D#y450>gap_450   </a>
+	|	<a href=3D#y300>quantifier_300   </a><a href=3D#y130>selbri_130   </a><=
a href=3D#y450>gap_450   </a><a href=3D#y121>relative_clauses_121   </a>
+	;
+<a name=3Dy96>sumti_F_96  </a>	:	<a href=3D#y97>sumti_G_97  </a>
+		/* outer-quantified sumti */
+	|	<a href=3D#y300>quantifier_300   </a><a href=3D#y97>sumti_G_97  </a>
+	;
+<a name=3Dy97>sumti_G_97  </a>	:	<a href=3D#y483>qualifier_483  </a><a hre=
f=3D#y90>sumti_90  </a><a href=3D#y463>LUhU_gap_463  </a>
+	|	<a href=3D#y483>qualifier_483  </a><a href=3D#y121>relative_clauses_121=
   </a><a href=3D#y90>sumti_90   </a><a href=3D#y463>LUhU_gap_463  </a>
+		/*sumti grouping, set/mass/individual conversion; also sumti scalar nega=
tion */
+	|	<a href=3D#y400>anaphora_400   </a>
+	|	<a href=3D#y499>LA_499  </a><a href=3D#y404>cmene_404  </a>
+	|	<a href=3D#y499>LA_499  </a><a href=3D#y121>relative_clauses_121   </a>=
<a href=3D#y404>cmene_404  </a>
+	|	<a href=3D#y489>LI_489  </a><a href=3D#y310>MEX_310   </a><a href=3D#y4=
72>LOhO_gap_472   </a>
+	|	<a href=3D#y110>description_110   </a>
+	|	<a href=3D#y432>quote_arg_432   </a>
+	;
+
+<a name=3Dy110>description_110  </a>	:	<a href=3D#y499>LA_499  </a><a href=
=3D#y111>sumti_tail_111  </a><a href=3D#y450>gap_450  </a>
+	|	<a href=3D#y488>LE_488  </a><a href=3D#y111>sumti_tail_111   </a><a hre=
f=3D#y450>gap_450   </a>
+	;
+<a name=3Dy111>sumti_tail_111  </a>	:	<a href=3D#y112>sumti_tail_A_112  </=
a>
+		/* inner-quantified sumti relative clause */
+	|	<a href=3D#y121>relative_clauses_121   </a><a href=3D#y112>sumti_tail_A=
_112   </a>
+		/* pseudo-possessive (an abbreviated inner restriction); note that sumti=
 cannot be quantified */
+	|	<a href=3D#y97>sumti_G_97  </a><a href=3D#y112>sumti_tail_A_112   </a>
+		/* pseudo-possessive with outer restriction */
+	|	<a href=3D#y97>sumti_G_97  </a><a href=3D#y121>relative_clauses_121   <=
/a><a href=3D#y112>sumti_tail_A_112   </a>
+	;
+<a name=3Dy112>sumti_tail_A_112  </a>	:	<a href=3D#y130>selbri_130  </a>
+	|	<a href=3D#y130>selbri_130   </a><a href=3D#y121>relative_clauses_121  =
 </a>
+		/* explicit inner quantifier */
+	|	<a href=3D#y300>quantifier_300   </a><a href=3D#y130>selbri_130   </a>
+		/* quantifier both internal to a description, and external to a sumti th=
ereby made specific */
+	|	<a href=3D#y300>quantifier_300   </a><a href=3D#y130>selbri_130   </a><=
a href=3D#y121>relative_clauses_121   </a>
+	|	<a href=3D#y300>quantifier_300   </a><a href=3D#y90>sumti_90   </a>
+	;
+<a name=3Dy121>relative_clauses_121	</a>:	<a href=3D#y122>relative_clause_=
122  </a>
+	|	<a href=3D#y121>relative_clauses_121   </a><a href=3D#y487>ZIhE_487  </=
a><a href=3D#y122>relative_clause_122   </a>
+	;
+<a name=3Dy122>relative_clause_122	</a>:	<a href=3D#y485>GOI_485  </a><a h=
ref=3D#y83>term_83  </a><a href=3D#y464>GEhU_gap_464  </a>
+	|	<a href=3D#y484>NOI_484  </a><a href=3D#y41>subsentence_41   </a><a hre=
f=3D#y469>KUhO_gap_469  </a>
+	;
+<a name=3Dy130>selbri_130  </a>	:	<a href=3D#y491>tag_491  </a><a href=3D#=
y131>selbri_A_131  </a>
+	|	<a href=3D#y131>selbri_A_131   </a>
+	;
+<a name=3Dy131>selbri_A_131  </a>	:	<a href=3D#y132>selbri_B_132  </a>
+	|	<a href=3D#y445>NA_445  </a><a href=3D#y130>selbri_130   </a>
+	;
+<a name=3Dy132>selbri_B_132  </a>	:	<a href=3D#y133>selbri_C_133  </a>
+	|	<a href=3D#y133>selbri_C_133  </a><a href=3D#y443>CO_443  </a><a href=
=3D#y132>selbri_B_132   </a>
+	;
+<a name=3Dy133>selbri_C_133  </a>	:	<a href=3D#y134>selbri_D_134  </a>
+	|	<a href=3D#y133>selbri_C_133  </a><a href=3D#y134>selbri_D_134  </a>
+	;
+<a name=3Dy134>selbri_D_134  </a>	:	<a href=3D#y135>selbri_E_135  </a>
+	|	<a href=3D#y134>selbri_D_134  </a><a href=3D#y422>JOIK_JEK_422   </a><a=
 href=3D#y135>selbri_E_135  </a>
+	|	<a href=3D#y134>selbri_D_134  </a><a href=3D#y823>JOIK_KE_823  </a><a h=
ref=3D#y133>selbri_C_133   </a><a href=3D#y466>KEhE_gap_466  </a>
+	;
+
+<a name=3Dy135>selbri_E_135  </a>	:	<a href=3D#y136>selbri_F_136  </a>
+	|	<a href=3D#y136>selbri_F_136  </a><a href=3D#y821>JEK_BO_821   </a><a h=
ref=3D#y135>selbri_E_135  </a>
+	|	<a href=3D#y136>selbri_F_136  </a><a href=3D#y822>JOIK_BO_822   </a><a =
href=3D#y135>selbri_E_135  </a>
+	;
+<a name=3Dy136>selbri_F_136  </a>	:	<a href=3D#y150>tanru_unit_150  </a>
+	|	<a href=3D#y150>tanru_unit_150   </a><a href=3D#y479>BO_479  </a><a hre=
f=3D#y136>selbri_F_136  </a>
+	|	<a href=3D#y137>GUhEK_selbri_137  </a>
+	|	<a href=3D#y482>NAhE_482   </a><a href=3D#y137>GUhEK_selbri_137  </a>
+	;
+<a name=3Dy137>GUhEK_selbri_137  </a>	:	<a href=3D#y808>GUhEK_808  </a><a =
href=3D#y130>selbri_130  </a><a href=3D#y816>GIK_816  </a><a href=3D#y136>s=
elbri_F_136  </a>
+	;
+<a name=3Dy150>tanru_unit_150  </a>	:	<a href=3D#y151>tanru_unit_A_151  </=
a>
+	|	<a href=3D#y150>tanru_unit_150   </a><a href=3D#y444>CEI_444  </a><a hr=
ef=3D#y151>tanru_unit_A_151   </a>
+	;
+<a name=3Dy151>tanru_unit_A_151  </a>	:	<a href=3D#y152>tanru_unit_B_152  =
</a>
+	|	<a href=3D#y152>tanru_unit_B_152   </a><a href=3D#y160>linkargs_160   <=
/a>
+	;
+<a name=3Dy152>tanru_unit_B_152  </a>	:	<a href=3D#y407>bridi_valsi_407  <=
/a>
+	|	<a href=3D#y493>KE_493  </a><a href=3D#y133>selbri_C_133  </a><a href=
=3D#y466>KEhE_gap_466  </a>
+	|	<a href=3D#y480>SE_480   </a><a href=3D#y152>tanru_unit_B_152   </a>
+	|	<a href=3D#y478>JAI_478  </a><a href=3D#y491>tag_491   </a><a href=3D#y=
152>tanru_unit_B_152   </a>
+	|	<a href=3D#y478>JAI_478  </a><a href=3D#y152>tanru_unit_B_152   </a>
+	|	<a href=3D#y477>ME_477  </a><a href=3D#y90>sumti_90   </a><a href=3D#y4=
65>MEhU_gap_465  </a>
+	|	<a href=3D#y477>ME_477  </a><a href=3D#y90>sumti_90   </a><a href=3D#y4=
65>MEhU_gap_465  </a><a href=3D#y476>MOI_476  </a>
+	|	<a href=3D#y475>NUhA_475  </a><a href=3D#y374>MEX_operator_374  </a>
+	|	<a href=3D#y482>NAhE_482   </a><a href=3D#y152>tanru_unit_B_152   </a>
+	|	<a href=3D#y425>NU_425  </a><a href=3D#y41>subsentence_41   </a><a href=
=3D#y453>KEI_gap_453  </a>
+	;
+<a name=3Dy160>linkargs_160  </a>	:	<a href=3D#y446>BE_446  </a><a href=3D=
#y83>term_83  </a><a href=3D#y467>BEhO_gap_467  </a>
+	|	<a href=3D#y446>BE_446  </a><a href=3D#y83>term_83  </a><a href=3D#y161=
>links_161   </a><a href=3D#y467>BEhO_gap_467  </a>
+	;
+<a name=3Dy161>links_161  </a>	:	<a href=3D#y442>BEI_442  </a><a href=3D#y=
83>term_83  </a>
+	|	<a href=3D#y442>BEI_442   </a><a href=3D#y83>term_83  </a><a href=3D#y1=
61>links_161   </a>
+	;
+
+/* Main entry point for MEX; everything but a number must be in parens. */
+<a name=3Dy300>quantifier_300  </a>	:	<a href=3D#y812>number_812  </a><a h=
ref=3D#y461>BOI_gap_461  </a>
+	|	<a href=3D#y470>left_bracket_470   </a><a href=3D#y310>MEX_310   </a><a=
 href=3D#y471>right_bracket_gap_471   </a>
+	;
+/* Entry point for MEX used after LI; no parens needed, but LI now has an =
elidable terminator. (This allows us to express the difference between =93t=
he expression a + b=94 and =93the expression (a + b)=94 _) */
+/* This rule supports left-grouping infix expressions and reverse Polish e=
xpressions. To handle infix monadic, use a null operand; to handle infix wi=
th more than two operands (whatever that means) use an extra operator or an=
 array operand. */
+<a name=3Dy310>MEX_310  </a>	:	<a href=3D#y311>MEX_A_311  </a>
+	|	<a href=3D#y310>MEX_310   </a><a href=3D#y370>operator_370   </a><a hre=
f=3D#y311>MEX_A_311   </a>
+	|	<a href=3D#y441>FUhA_441   </a><a href=3D#y330>rp_expression_330   </a>
+	;
+/* Support for right-grouping (short scope) infix expressions with BIhE. *=
/
+<a name=3Dy311>MEX_A_311  </a>	:	<a href=3D#y312>MEX_B_312  </a>
+	|	<a href=3D#y312>MEX_B_312   </a><a href=3D#y439>BIhE_439  </a><a href=
=3D#y370>operator_370   </a><a href=3D#y311>MEX_A_311   </a>
+	;
+/* Support for forethought (Polish) expressions. These begin with a foreth=
ought flag, then the operator and then the argument(s). */
+<a name=3Dy312>MEX_B_312  </a>	:	<a href=3D#y381>operand_381  </a>
+	|	<a href=3D#y370>operator_370   </a><a href=3D#y313>MEX_C_313  </a><a hr=
ef=3D#y452>MEX_gap_452   </a>
+	|	<a href=3D#y438>PEhO_438  </a><a href=3D#y370>operator_370   </a><a hre=
f=3D#y313>MEX_C_313  </a><a href=3D#y452>MEX_gap_452   </a>
+	;
+<a name=3Dy313>MEX_C_313  </a>	:	<a href=3D#y312>MEX_B_312  </a>
+	|	<a href=3D#y313>MEX_C_313  </a><a href=3D#y312>MEX_B_312   </a>
+	;
+/* Reverse Polish expressions always have exactly two operands. To handle =
one operand, use a null operand; to handle more than two operands, use a nu=
ll operator. */
+<a name=3Dy330>rp_expression_330	</a>:	<a href=3D#y332>rp_operand_332  </a=
><a href=3D#y332>rp_operand_332  </a><a href=3D#y370>operator_370  </a>
+	;
+<a name=3Dy332>rp_operand_332  </a>	:	<a href=3D#y381>operand_381  </a>
+	|	<a href=3D#y330>rp_expression_330   </a>
+	;
+
+/* Operators may be joined by logical connectives. */
+<a name=3Dy370>operator_370  </a>	:	<a href=3D#y371>operator_A_371  </a>
+	|	<a href=3D#y370>operator_370   </a><a href=3D#y422>JOIK_JEK_422   </a><=
a href=3D#y371>operator_A_371   </a>
+	|	<a href=3D#y370>operator_370   </a><a href=3D#y823>JOIK_KE_823  </a><a =
href=3D#y370>operator_370   </a><a href=3D#y466>KEhE_gap_466  </a>
+	;
+<a name=3Dy371>operator_A_371  </a>	:	<a href=3D#y372>operator_B_372  </a>
+	|	<a href=3D#y808>GUhEK_808  </a><a href=3D#y371>operator_A_371   </a><a =
href=3D#y816>GIK_816  </a><a href=3D#y372>operator_B_372   </a>
+	|	<a href=3D#y372>operator_B_372   </a><a href=3D#y822>JOIK_BO_822   </a>=
<a href=3D#y371>operator_A_371   </a>
+	|	<a href=3D#y372>operator_B_372   </a><a href=3D#y821>JEK_BO_821   </a><=
a href=3D#y371>operator_A_371   </a>
+	;
+<a name=3Dy372>operator_B_372  </a>	:	<a href=3D#y374>MEX_operator_374  </=
a>
+	|	<a href=3D#y493>KE_493  </a><a href=3D#y370>operator_370   </a><a href=
=3D#y466>KEhE_gap_466  </a>
+	;
+<a name=3Dy374>MEX_operator_374  </a>	:	<a href=3D#y679>VUhU_679  </a>
+	|	<a href=3D#y679>VUhU_679  </a><a href=3D#y32>free_modifier_32   </a>
+	|	<a href=3D#y480>SE_480   </a><a href=3D#y374>MEX_operator_374  </a>
+		/* changes argument order */
+	|	<a href=3D#y482>NAhE_482   </a><a href=3D#y374>MEX_operator_374  </a>
+		/* scalar negation */
+	|	<a href=3D#y430>MAhO_430   </a><a href=3D#y310>MEX_310   </a><a href=3D=
#y473>TEhU_gap_473  </a>
+	|	<a href=3D#y429>NAhU_429  </a><a href=3D#y130>selbri_130   </a><a href=
=3D#y473>TEhU_gap_473  </a>
+	;
+<a name=3Dy381>operand_381  </a>	:	<a href=3D#y382>operand_A_382  </a>
+	|	<a href=3D#y382>operand_A_382   </a><a href=3D#y804>EK_KE_804  </a><a h=
ref=3D#y381>operand_381   </a><a href=3D#y466>KEhE_gap_466  </a>
+	|	<a href=3D#y382>operand_A_382   </a><a href=3D#y823>JOIK_KE_823  </a><a=
 href=3D#y381>operand_381   </a><a href=3D#y466>KEhE_gap_466  </a>
+	;
+<a name=3Dy382>operand_A_382  </a>	:	<a href=3D#y383>operand_B_383  </a>
+	|	<a href=3D#y382>operand_A_382   </a><a href=3D#y421>JOIK_EK_421   </a><=
a href=3D#y383>operand_B_383  </a>
+	;
+<a name=3Dy383>operand_B_383  </a>	:	<a href=3D#y385>operand_C_385  </a>
+	|	<a href=3D#y385>operand_C_385  </a><a href=3D#y803>EK_BO_803  </a><a hr=
ef=3D#y383>operand_B_383  </a>
+	|	<a href=3D#y385>operand_C_385  </a><a href=3D#y822>JOIK_BO_822   </a><a=
 href=3D#y383>operand_B_383  </a>
+	;
+<a name=3Dy385>operand_C_385  </a>	:	<a href=3D#y300>quantifier_300  </a>
+	|	<a href=3D#y817>lerfu_string_817  </a><a href=3D#y461>BOI_gap_461   </a=
>
+		/* lerfu string as operand - classic math variable */
+	|	<a href=3D#y428>NIhE_428  </a><a href=3D#y130>selbri_130   </a><a href=
=3D#y473>TEhU_gap_473  </a>
+		/* quantifies a bridi - inverse of -MOI */
+	|	<a href=3D#y427>MOhE_427  </a><a href=3D#y90>sumti_90   </a><a href=3D#=
y473>TEhU_gap_473  </a>
+		/* quantifies a sumti - inverse of LI */
+	|	<a href=3D#y431>JOhI_431   </a><a href=3D#y313>MEX_C_313  </a><a href=
=3D#y473>TEhU_gap_473  </a>
+	|	<a href=3D#y807>GEK_807  </a><a href=3D#y381>operand_381   </a><a href=
=3D#y816>GIK_816  </a><a href=3D#y385>operand_C_385  </a>
+	|	<a href=3D#y483>qualifier_483  </a><a href=3D#y381>operand_381   </a><a=
 href=3D#y463>LUhU_gap_463  </a>
+	;
+
+/* _400 series constructs are mostly specific strings, some of which may a=
lso be used by the lexer; the lexer should not use any reference to termina=
ls numbered less than _400, as they have grammars composed on non-determini=
stic strings of selma'o. Some above _400 also are this way, so care should =
be taken; this is especially true for those that reference <a href=3D#y32>f=
ree_modifier_32</a>. */
+<a name=3Dy400>anaphora_400  </a>	:	<a href=3D#y555>KOhA_555  </a>
+	|	<a href=3D#y555>KOhA_555  </a><a href=3D#y32>free_modifier_32   </a>
+	|	<a href=3D#y817>lerfu_string_817  </a><a href=3D#y461>BOI_gap_461   </a=
>
+	;
+<a name=3Dy404>cmene_404  </a>	:	<a href=3D#y405>cmene_A_405  </a>
+	|	<a href=3D#y405>cmene_A_405  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy405>cmene_A_405  </a>	:	<a href=3D#y518>CMENE_518  </a>/* pause=
 */
+	|	<a href=3D#y405>cmene_A_405  </a><a href=3D#y518>CMENE_518  </a>/* paus=
e*/=09
+/* multiple CMENE are identified morphologically (by the lexer) =96 separa=
ted by consonant & pause */
+	;
+<a name=3Dy407>bridi_valsi_407  </a>	:	<a href=3D#y408>bridi_valsi_A_408  =
</a>
+	|	<a href=3D#y408>bridi_valsi_A_408  </a><a href=3D#y32>free_modifier_32 =
  </a>
+	;
+<a name=3Dy408>bridi_valsi_A_408	</a>:	<a href=3D#y509>BRIVLA_509  </a>
+	|	<a href=3D#y824>PA_MOI_824  </a>
+	|	<a href=3D#y543>GOhA_543  </a>
+	|	<a href=3D#y543>GOhA_543  </a><a href=3D#y593>RAhO_593  </a>
+	;
+<a name=3Dy410>para_mark_410  </a>	:	<a href=3D#y584>NIhO_584  </a>
+	|	<a href=3D#y584>NIhO_584  </a><a href=3D#y32>free_modifier_32   </a>
+	|	<a href=3D#y584>NIhO_584  </a><a href=3D#y410>para_mark_410   </a>
+	;
+<a name=3Dy411>indicators_411  </a>	:	<a href=3D#y412>indicators_A_412  </=
a>
+	|	<a href=3D#y535>FUhE_535  </a><a href=3D#y412>indicators_A_412   </a>
+	;
+<a name=3Dy412>indicators_A_412  </a>	:	<a href=3D#y413>indicator_413  </a=
>
+	|	<a href=3D#y412>indicators_A_412   </a><a href=3D#y413>indicator_413  <=
/a>
+	;
+<a name=3Dy413>indicator_413  </a>	:	<a href=3D#y612>UI_612  </a>
+	|	<a href=3D#y515>CAI_515  </a>
+	|	<a href=3D#y612>UI_612   </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y515>CAI_515  </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y619>Y_619  </a>
+	|	<a href=3D#y524>DAhO_524  </a>
+	|	<a href=3D#y536>FUhO_536  </a>
+	;
+
+<a name=3Dy415>DOI_415  </a>	:	<a href=3D#y525>DOI_525  </a>
+	|	<a href=3D#y416>COI_416  </a>
+	|	<a href=3D#y416>COI_416  </a><a href=3D#y525>DOI_525  </a>
+	;
+<a name=3Dy416>COI_416  </a>	:	<a href=3D#y417>COI_A_417  </a>
+	|	<a href=3D#y416>COI_416  </a><a href=3D#y417>COI_A_417  </a>
+	;
+<a name=3Dy417>COI_A_417  </a>	:	<a href=3D#y520>COI_520  </a>
+	|	<a href=3D#y520>COI_520   </a><a href=3D#y581>NAI_581   </a>
+	;
+<a name=3Dy421>JOIK_EK_421  </a>	:	<a href=3D#y802>EK_802  </a>
+	|	<a href=3D#y806>JOIK_806  </a>
+	|	<a href=3D#y806>JOIK_806  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy422>JOIK_JEK_422  </a>	:	<a href=3D#y806>JOIK_806  </a>
+	|	<a href=3D#y806>JOIK_806  </a><a href=3D#y32>free_modifier_32   </a>
+	|	<a href=3D#y805>JEK_805  </a>
+	|	<a href=3D#y805>JEK_805  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy424>XI_424  </a>	:	<a href=3D#y618>XI_618  </a>
+	|	<a href=3D#y618>XI_618  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy425>NU_425  </a>	:	<a href=3D#y426>NU_A_426  </a>
+	|	<a href=3D#y425>NU_425  </a><a href=3D#y422>JOIK_JEK_422   </a><a href=
=3D#y426>NU_A_426  </a>
+	;
+<a name=3Dy426>NU_A_426  </a>	:	<a href=3D#y586>NU_586  </a>
+	|	<a href=3D#y586>NU_586  </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y586>NU_586  </a><a href=3D#y32>free_modifier_32   </a>
+	|	<a href=3D#y586>NU_586  </a><a href=3D#y581>NAI_581   </a><a href=3D#y3=
2>free_modifier_32   </a>
+	;
+<a name=3Dy427>MOhE_427  </a>	:	<a href=3D#y664>MOhE_664  </a>
+	|	<a href=3D#y664>MOhE_664  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy428>NIhE_428  </a>	:	<a href=3D#y666>NIhE_666  </a>
+	|	<a href=3D#y666>NIhE_666  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy429>NAhU_429  </a>	:	<a href=3D#y665>NAhU_665  </a>
+	|	<a href=3D#y665>NAhU_665  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy430>MAhO_430  </a>	:	<a href=3D#y662>MAhO_662  </a>
+	|	<a href=3D#y662>MAhO_662   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy431>JOhI_431  </a>	:	<a href=3D#y657>JOhI_657  </a>
+	|	<a href=3D#y657>JOhI_657  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+
+<a name=3Dy432>quote_arg_432  </a>	:	<a href=3D#y433>quote_arg_A_433  </a>
+	|	<a href=3D#y433>quote_arg_A_433  </a><a href=3D#y32>free_modifier_32   =
</a>
+	;
+<a name=3Dy433>quote_arg_A_433  </a>	:	<a href=3D#y434>ZOI_quote_434  </a>
+	|	<a href=3D#y435>ZO_quote_435  </a>
+	|	<a href=3D#y436>LOhU_quote_436  </a>
+	|	<a href=3D#y571>LU_571   </a><a href=3D#y0>text_0   </a><a href=3D#y448=
>LIhU_gap_448  </a>
+	;
+/* The quoted material in the following three terminals must be identified=
 by the lexer, but no additional lexer processing is needed. */
+<a name=3Dy434>ZOI_quote_434  </a>	:	<a href=3D#y627>ZOI_627  </a><a href=
=3D#y698>any_word_698  </a>/*pause*/   <a href=3D#y699>anything_699  </a>/*=
pause*/   <a href=3D#y698>any_word_698  </a>
+	;
+/* 'pause' is morphemic, represented by '.' The lexer assembles <a href=3D=
#y699>anything_699  </a>*/
+<a name=3Dy435>ZO_quote_435  </a>	:	<a href=3D#y626>ZO_626  </a><a href=3D=
#y698>any_word_698  </a>
+	;
+/* 'word' may not be a compound; but it can be any valid Lojban selma'o va=
lue, including ZO, ZOI, SI, SA, SU. The preparser will not lex the word per=
 its normal selma'o. */
+<a name=3Dy436>LOhU_quote_436  </a>	:	<a href=3D#y569>LOhU_569  </a><a hre=
f=3D#y697>any_words_697  </a><a href=3D#y565>LEhU_565  </a>
+	;
+/* 'words' may be any Lojban words, with no claim of grammaticality; the p=
reparser will not lex the individual words per their normal selma'o; used t=
o quote ungrammatical Lojban, equivalent to the * or ? writing convention f=
or such text.
+The preparser needs one bit of sophistication for this rule. A quoted stri=
ng should be able to contain other quoted strings - this is only a problem =
for a LOhU quote itself, since the LEhU clossing this quote would otherwise=
 close the outer quotes, which is incorrect. For this purpose, we will chea=
t on the use of ZO in such a quote (since this is ungrammatical text, it is=
 a sign ignored by the parser). Use ZO to mark any nested quotation LOhU. T=
he preparser then will absorb it by the ZO rule, before testing for LOhU. T=
his is obviously not the standard usage for ZO, which would otherwise cause=
 the result to be a sumti. But, since the result will be part of an unparse=
d string anyway, it doesn't matter.
+It may be seen that any of the ZO/ZOI/LOhU trio of quotation markers may c=
ontain the powerful metalinguistic erasers. Since these quotations are not =
parsed internally, these operators are ignored within the quote. To erase a=
 ZO, then, two SI's are needed after giving a quoted word of any type. ZOI =
takes four SI's, with the ENTIRE BODY OF THE QUOTE treated as a single 'wor=
d' since it is one selma'o. Thus one for the quote body, two for the single=
 word delimiters, and one for the ZOI. In LOhU, the entire body is treated =
as a single word, so three SI's can erase it.*/
+
+/*All rule terminator names with 'gap' in them are potentially elidable, w=
here such elision does not cause an ambiguity. This is implemented through =
use of the YACC 'error' token, which effectively recovers from an elision. =
*/
+<a name=3Dy437>FIhO_437  </a>	:	<a href=3D#y532>FIhO_532  </a>
+	|	<a href=3D#y532>FIhO_532   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy438>PEhO_438  </a>	:	<a href=3D#y673>PEhO_673  </a>
+	|	<a href=3D#y673>PEhO_673  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy439>BIhE_439  </a>	:	<a href=3D#y650>BIhE_650  </a>
+	|	<a href=3D#y650>BIhE_650   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy440>SEI_440  </a>	:	<a href=3D#y597>SEI_597  </a>
+	|	<a href=3D#y597>SEI_597  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy441>FUhA_441  </a>	:	<a href=3D#y655>FUhA_655  </a>
+	|	<a href=3D#y655>FUhA_655  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy442>BEI_442  </a>	:	<a href=3D#y505>BEI_505  </a>
+	|	<a href=3D#y505>BEI_505  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy443>CO_443  </a>	:	<a href=3D#y519>CO_519  </a>
+	|	<a href=3D#y519>CO_519  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy444>CEI_444  </a>	:	<a href=3D#y516>CEI_516  </a>
+	|	<a href=3D#y516>CEI_516  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy445>NA_445  </a>	:	<a href=3D#y578>NA_578  </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy446>BE_446  </a>	:	<a href=3D#y504>BE_504  </a>
+	|	<a href=3D#y504>BE_504  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy447>TUhE_447  </a>	:	<a href=3D#y610>TUhE_610  </a>
+	|	<a href=3D#y610>TUhE_610   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy448>LIhU_gap_448  </a>	:	<a href=3D#y567>LIhU_567  </a>
+	|	error
+	;
+<a name=3Dy450>gap_450  </a>	:	<a href=3D#y556>KU_556  </a>
+	|	<a href=3D#y556>KU_556  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy451>front_gap_451  </a>	:	<a href=3D#y521>CU_521  </a>
+	|	<a href=3D#y521>CU_521   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy452>MEX_gap_452  </a>	:	<a href=3D#y658>KUhE_658  </a>
+	|	<a href=3D#y658>KUhE_658  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy453>KEI_gap_453  </a>	:	<a href=3D#y552>KEI_552  </a>
+	|	<a href=3D#y552>KEI_552   </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy454>TUhU_gap_454  </a>	:	<a href=3D#y611>TUhU_611  </a>
+	|	<a href=3D#y611>TUhU_611   </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy456>VAU_gap_456  </a>	:	<a href=3D#y614>VAU_614  </a>
+	|	<a href=3D#y614>VAU_614  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+/* redundant to attach a free modifier on the following */
+<a name=3Dy457>DOhU_gap_457  </a>	:	<a href=3D#y526>DOhU_526  </a>
+	|	error
+	;
+<a name=3Dy458>FEhU_gap_458  </a>	:	<a href=3D#y531>FEhU_531  </a>
+	|	<a href=3D#y531>FEhU_531   </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy459>SEhU_gap_459  </a>	:	<a href=3D#y598>SEhU_598  </a>
+	|	error
+/* a free modifier on a discursive should be somewhere within the discursi=
ve. See <a href=3D#y440>SEI_440   </a>*/
+	;
+<a name=3Dy460>NUhU_gap_460  </a>	:	<a href=3D#y588>NUhU_588  </a>
+	|	<a href=3D#y588>NUhU_588  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy461>BOI_gap_461  </a>	:	<a href=3D#y651>BOI_651  </a>
+	|	<a href=3D#y651>BOI_651   </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy462>sub_gap_462  </a>	:	<a href=3D#y651>BOI_651  </a>
+	|	error
+	;
+<a name=3Dy463>LUhU_gap_463  </a>	:	<a href=3D#y573>LUhU_573  </a>
+	|	<a href=3D#y573>LUhU_573  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+
+<a name=3Dy464>GEhU_gap_464  </a>	:	<a href=3D#y538>GEhU_538  </a>
+	|	<a href=3D#y538>GEhU_538  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy465>MEhU_gap_465  </a>	:	<a href=3D#y575>MEhU_575  </a>
+	|	<a href=3D#y575>MEhU_575  </a> <a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy466>KEhE_gap_466  </a>	:	<a href=3D#y550>KEhE_550  </a>
+	|	<a href=3D#y550>KEhE_550   </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy467>BEhO_gap_467  </a>	:	<a href=3D#y506>BEhO_506  </a>
+	|	<a href=3D#y506>BEhO_506  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy468>TOI_gap_468  </a>	:	<a href=3D#y607>TOI_607  </a>
+	|	error
+	;
+<a name=3Dy469>KUhO_gap_469  </a>	:	<a href=3D#y557>KUhO_557  </a>
+	|	<a href=3D#y557>KUhO_557  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy470>left_bracket_470  </a>	:	<a href=3D#y677>VEI_677  </a>
+	|	<a href=3D#y677>VEI_677  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy471>right_bracket_gap_471	</a>:	<a href=3D#y678>VEhO_678  </a>
+	|	<a href=3D#y678>VEhO_678  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy472>LOhO_gap_472  </a>	:	<a href=3D#y568>LOhO_568  </a>
+	|	<a href=3D#y568>LOhO_568  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy473>TEhU_gap_473  </a>	:	<a href=3D#y675>TEhU_675  </a>
+	|	<a href=3D#y675>TEhU_675  </a><a href=3D#y32>free_modifier_32   </a>
+	|	error
+	;
+<a name=3Dy474>right_br_no_free_474	</a>:	<a href=3D#y678>VEhO_678  </a>
+	|	error
+	;
+<a name=3Dy475>NUhA_475  </a>	:	<a href=3D#y667>NUhA_667  </a>
+	|	<a href=3D#y667>NUhA_667  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+
+<a name=3Dy476>MOI_476  </a>	:	<a href=3D#y663>MOI_663  </a>
+	|	<a href=3D#y663>MOI_663  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy477>ME_477  </a>	:	<a href=3D#y574>ME_574  </a>
+	|	<a href=3D#y574>ME_574  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy478>JAI_478  </a>	:	<a href=3D#y547>JAI_547  </a>
+	|	<a href=3D#y547>JAI_547  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy479>BO_479  </a>	:	<a href=3D#y508>BO_508  </a>
+	|	<a href=3D#y508>BO_508  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy480>SE_480  </a>	:	<a href=3D#y596>SE_596  </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy481>FA_481  </a>	:	<a href=3D#y527>FA_527  </a>
+	|	<a href=3D#y527>FA_527  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy482>NAhE_482  </a>	:	<a href=3D#y583>NAhE_583  </a>
+	|	<a href=3D#y583>NAhE_583  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy483>qualifier_483  </a>	:	<a href=3D#y561>LAhE_561  </a>
+	|	<a href=3D#y561>LAhE_561   </a><a href=3D#y32>free_modifier_32   </a>
+	|	<a href=3D#y809>NAhE_BO_809  </a>
+	;
+<a name=3Dy484>NOI_484  </a>	:	<a href=3D#y585>NOI_585  </a>
+	|	<a href=3D#y585>NOI_585  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy485>GOI_485  </a>	:	<a href=3D#y542>GOI_542  </a>
+	|	<a href=3D#y542>GOI_542   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy486>subscript_486  </a>	:	<a href=3D#y424>XI_424  </a><a href=
=3D#y812>number_812  </a><a href=3D#y462>sub_gap_462  </a>
+	|	<a href=3D#y424>XI_424  </a><a href=3D#y470>left_bracket_470   </a><a h=
ref=3D#y310>MEX_310   </a><a href=3D#y474>right_br_no_free_474  </a>
+	|	<a href=3D#y424>XI_424  </a><a href=3D#y817>lerfu_string_817  </a><a hr=
ef=3D#y462>sub_gap_462   </a>
+	;
+<a name=3Dy487>ZIhE_487  </a>	:	<a href=3D#y625>ZIhE_625  </a>
+	|	<a href=3D#y625>ZIhE_625  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy488>LE_488  </a>	:	<a href=3D#y562>LE_562  </a>
+	|	<a href=3D#y562>LE_562   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy489>LI_489  </a>	:	<a href=3D#y566>LI_566  </a>
+	|	<a href=3D#y566>LI_566  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+
+<a name=3Dy490>mod_head_490  </a>	:	<a href=3D#y491>tag_491  </a>
+	|	<a href=3D#y481>FA_481   </a>
+	;
+<a name=3Dy491>tag_491  </a>	:	<a href=3D#y815>tense_modal_815  </a>
+	|	<a href=3D#y491>tag_491   </a><a href=3D#y422>JOIK_JEK_422   </a><a hre=
f=3D#y815>tense_modal_815  </a>
+	;
+<a name=3Dy492>ZOhU_492  </a>	:	<a href=3D#y628>ZOhU_628  </a>
+	|	<a href=3D#y628>ZOhU_628  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy493>KE_493  </a>	:	<a href=3D#y551>KE_551  </a>
+	|	<a href=3D#y551>KE_551   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy494>PEhE_494  </a>	:	<a href=3D#y591>PEhE_591  </a>
+	|	<a href=3D#y591>PEhE_591   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy495>CEhE_495  </a>	:	<a href=3D#y517>CEhE_517  </a>
+	|	<a href=3D#y517>CEhE_517  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy496>NUhI_496  </a>	:	<a href=3D#y587>NUhI_587  </a>
+	|	<a href=3D#y587>NUhI_587  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy497>VUhO_497  </a>	:	<a href=3D#y617>VUhO_617  </a>
+	|	<a href=3D#y617>VUhO_617  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy498>SOI_498  </a>	:	<a href=3D#y602>SOI_602  </a>
+	|	<a href=3D#y602>SOI_602   </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy499>LA_499  </a>	:	<a href=3D#y558>LA_558  </a>
+	|	<a href=3D#y558>LA_558  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy801>utterance_ordinal_801	</a>:	<a href=3D#y905>lexer_A_905  </=
a>
+	;
+<a name=3Dy802>EK_802  </a>	:	<a href=3D#y910>lexer_B_910  </a>
+	|	<a href=3D#y910>lexer_B_910   </a><a href=3D#y32>free_modifier_32   </a=
>
+	;
+<a name=3Dy803>EK_BO_803  </a>	:	<a href=3D#y915>lexer_C_915  </a>
+	|	<a href=3D#y915>lexer_C_915  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy804>EK_KE_804  </a>	:	<a href=3D#y916>lexer_D_916  </a>
+	|	<a href=3D#y916>lexer_D_916  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy805>JEK_805  </a>	:	<a href=3D#y925>lexer_E_925  </a>
+	;
+
+<a name=3Dy806>JOIK_806  </a>	:	<a href=3D#y930>lexer_F_930  </a>
+	;
+<a name=3Dy807>GEK_807  </a>	:	<a href=3D#y935>lexer_G_935  </a>
+	|	<a href=3D#y935>lexer_G_935  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy808>GUhEK_808  </a>	:	<a href=3D#y940>lexer_H_940  </a>
+	|	<a href=3D#y940>lexer_H_940   </a><a href=3D#y32>free_modifier_32   </a=
>
+	;
+<a name=3Dy809>NAhE_BO_809  </a>	:	<a href=3D#y945>lexer_I_945  </a>
+	|	<a href=3D#y945>lexer_I_945  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy810>NA_KU_810  </a>	:	<a href=3D#y950>lexer_J_950  </a>
+	|	<a href=3D#y950>lexer_J_950   </a><a href=3D#y32>free_modifier_32   </a=
>
+	;
+<a name=3Dy811>I_BO_811  </a>	:	<a href=3D#y955>lexer_K_955  </a>
+	|	<a href=3D#y955>lexer_K_955  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy812>number_812  </a>	:	<a href=3D#y960>lexer_L_960  </a>
+	;
+<a name=3Dy813>GIhEK_BO_813  </a>	:	<a href=3D#y965>lexer_M_965  </a>
+	|	<a href=3D#y965>lexer_M_965  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy814>GIhEK_KE_814  </a>	:	<a href=3D#y966>lexer_N_966  </a>
+	|	<a href=3D#y966>lexer_N_966  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy815>tense_modal_815  </a>	:	<a href=3D#y970>lexer_O_970  </a>
+	|	<a href=3D#y970>lexer_O_970   </a><a href=3D#y32>free_modifier_32   </a=
>
+	|	<a href=3D#y437>FIhO_437  </a><a href=3D#y130>selbri_130   </a><a href=
=3D#y458>FEhU_gap_458  </a>
+	;
+<a name=3Dy816>GIK_816  </a>	:	<a href=3D#y980>lexer_P_980  </a>
+	|	<a href=3D#y980>lexer_P_980   </a><a href=3D#y32>free_modifier_32   </a=
>
+	;
+<a name=3Dy817>lerfu_string_817  </a>	:	<a href=3D#y985>lexer_Q_985  </a>
+	;
+<a name=3Dy818>GIhEK_818  </a>	:	<a href=3D#y990>lexer_R_990  </a>
+	|	<a href=3D#y990>lexer_R_990   </a><a href=3D#y32>free_modifier_32   </a=
>
+	;
+<a name=3Dy819>I_819  </a>	:	<a href=3D#y995>lexer_S_995  </a>
+	|	<a href=3D#y995>lexer_S_995  </a><a href=3D#y32>free_modifier_32   </a>
+	;
+<a name=3Dy820>I_JEK_820  </a>	:	<a href=3D#y1000>lexer_T_1000  </a>
+	|	<a href=3D#y1000>lexer_T_1000   </a><a href=3D#y32>free_modifier_32   <=
/a>
+	;
+
+<a name=3Dy821>JEK_BO_821  </a>	:	<a href=3D#y1005>lexer_U_1005  </a>
+	|	<a href=3D#y1005>lexer_U_1005  </a><a href=3D#y32>free_modifier_32   </=
a>
+	;
+<a name=3Dy822>JOIK_BO_822  </a>	:	<a href=3D#y1010>lexer_V_1010  </a>
+	|	<a href=3D#y1010>lexer_V_1010   </a><a href=3D#y32>free_modifier_32   <=
/a>
+	;
+<a name=3Dy823>JOIK_KE_823  </a>	:	<a href=3D#y1015>lexer_W_1015  </a>
+	|	<a href=3D#y1015>lexer_W_1015  </a><a href=3D#y32>free_modifier_32   </=
a>
+	;
+<a name=3Dy824>PA_MOI_824  </a>	:	<a href=3D#y1025>lexer_Y_1025  </a>
+	;
+/* The following rules are used only in lexer processing. They have been t=
ested for ambiguity at various levels in the YACC grammar, but are in the r=
ecursive descent lexer in the current parser. The lexer inserts the lexer t=
okens before the processed strings, but leaves the original tokens. */
+<a name=3Dy905>lexer_A_905  </a>	:	<a href=3D#y701>lexer_A_701  </a><a hre=
f=3D#y906>utt_ordinal_root_906  </a>
+	;
+<a name=3Dy906>utt_ordinal_root_906	</a>:	<a href=3D#y986>lerfu_string_roo=
t_986  </a><a href=3D#y661>MAI_661  </a>
+	|	<a href=3D#y961>number_root_961   </a><a href=3D#y661>MAI_661   </a>
+	;
+<a name=3Dy910>lexer_B_910  </a>	:	<a href=3D#y702>lexer_B_702  </a><a hre=
f=3D#y911>EK_root_911  </a>
+	;
+<a name=3Dy911>EK_root_911  </a>	:	<a href=3D#y501>A_501  </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y501>A_501   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y501>A_501   </a>
+	|	<a href=3D#y501>A_501   </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y501>A_501   </a><a href=3D#y581=
>NAI_581   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y501>A_501   </a><a href=3D#y581=
>NAI_581   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y596>SE_596  </a><a href=3D#y501=
>A_501   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y596>SE_596  </a><a href=3D#y501=
>A_501   </a><a href=3D#y581>NAI_581   </a>
+	;
+<a name=3Dy915>lexer_C_915  </a>	:	<a href=3D#y703>lexer_C_703  </a><a hre=
f=3D#y911>EK_root_911  </a><a href=3D#y508>BO_508  </a>
+	|	<a href=3D#y703>lexer_C_703  </a><a href=3D#y911>EK_root_911   </a><a h=
ref=3D#y971>simple_tag_971   </a><a href=3D#y508>BO_508  </a>
+	;
+<a name=3Dy916>lexer_D_916  </a>	:	<a href=3D#y704>lexer_D_704  </a><a hre=
f=3D#y911>EK_root_911  </a><a href=3D#y551>KE_551  </a>
+	|	<a href=3D#y704>lexer_D_704  </a><a href=3D#y911>EK_root_911   </a><a h=
ref=3D#y971>simple_tag_971   </a><a href=3D#y551>KE_551   </a>
+	;
+
+<a name=3Dy925>lexer_E_925  </a>	:	<a href=3D#y705>lexer_E_705  </a><a hre=
f=3D#y926>JEK_root_926  </a>
+	;
+<a name=3Dy926>JEK_root_926  </a>	:	<a href=3D#y546>JA_546  </a>
+	|	<a href=3D#y546>JA_546  </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y546>JA_546  </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y546>JA_546  </a><a href=3D#y581=
>NAI_581   </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y546>JA_546  </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y546>JA_546  </a><a href=3D#y581=
>NAI_581   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y596>SE_596  </a><a href=3D#y546=
>JA_546  </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y596>SE_596  </a><a href=3D#y546=
>JA_546  </a><a href=3D#y581>NAI_581   </a>
+	;
+<a name=3Dy930>lexer_F_930  </a>	:	<a href=3D#y706>lexer_F_706  </a><a hre=
f=3D#y931>JOIK_root_931  </a>
+	;
+<a name=3Dy931>JOIK_root_931  </a>	:	<a href=3D#y548>JOI_548  </a>
+	|	<a href=3D#y548>JOI_548  </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y548>JOI_548  </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y548>JOI_548  </a><a href=3D#y58=
1>NAI_581   </a>
+	|	<a href=3D#y932>interval_932   </a>
+	|	<a href=3D#y656>GAhO_656  </a><a href=3D#y932>interval_932   </a><a hre=
f=3D#y656>GAhO_656  </a>
+	;
+<a name=3Dy932>interval_932  </a>	:	<a href=3D#y507>BIhI_507  </a>
+	|	<a href=3D#y507>BIhI_507  </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y507>BIhI_507  </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y507>BIhI_507  </a><a href=3D#y5=
81>NAI_581   </a>
+	;
+<a name=3Dy935>lexer_G_935  </a>	:	<a href=3D#y707>lexer_G_707  </a><a hre=
f=3D#y537>GA_537  </a>
+	|	<a href=3D#y707>lexer_G_707  </a><a href=3D#y596>SE_596  </a><a href=3D=
#y537>GA_537  </a>
+	|	<a href=3D#y707>lexer_G_707  </a><a href=3D#y537>GA_537  </a><a href=3D=
#y581>NAI_581   </a>
+	|	<a href=3D#y707>lexer_G_707  </a><a href=3D#y596>SE_596  </a><a href=3D=
#y537>GA_537  </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y707>lexer_G_707  </a><a href=3D#y971>simple_tag_971   </a><=
a href=3D#y981>GIK_root_981   </a>
+	|	<a href=3D#y707>lexer_G_707  </a><a href=3D#y931>JOIK_root_931   </a><a=
 href=3D#y539>GI_539  </a>
+	;
+<a name=3Dy940>lexer_H_940  </a>	:	<a href=3D#y708>lexer_H_708  </a><a hre=
f=3D#y544>GUhA_544  </a>
+	|	<a href=3D#y708>lexer_H_708  </a><a href=3D#y596>SE_596  </a><a href=3D=
#y544>GUhA_544  </a>
+	|	<a href=3D#y708>lexer_H_708  </a><a href=3D#y544>GUhA_544  </a><a href=
=3D#y581>NAI_581   </a>
+	|	<a href=3D#y708>lexer_H_708  </a><a href=3D#y596>SE_596  </a><a href=3D=
#y544>GUhA_544  </a><a href=3D#y581>NAI_581   </a>
+	;
+<a name=3Dy945>lexer_I_945  </a>	:	<a href=3D#y709>lexer_I_709  </a><a hre=
f=3D#y583>NAhE_583  </a><a href=3D#y508>BO_508  </a>
+	;
+<a name=3Dy950>lexer_J_950  </a>	:	<a href=3D#y710>lexer_J_710  </a><a hre=
f=3D#y578>NA_578  </a><a href=3D#y556>KU_556  </a>
+	;
+
+<a name=3Dy955>lexer_K_955  </a>	:	<a href=3D#y711>lexer_K_711  </a><a hre=
f=3D#y956>I_root_956  </a><a href=3D#y508>BO_508  </a>
+	|	<a href=3D#y711>lexer_K_711   </a><a href=3D#y956>I_root_956  </a><a hr=
ef=3D#y971>simple_tag_971   </a><a href=3D#y508>BO_508  </a>
+	;
+<a name=3Dy956>I_root_956  </a>	:	<a href=3D#y545>I_545  </a>
+	|	<a href=3D#y545>I_545  </a><a href=3D#y957>simple_JOIK_JEK_957  </a>
+	;
+<a name=3Dy957>simple_JOIK_JEK_957	</a>:	<a href=3D#y806>JOIK_806  </a>
+	|	<a href=3D#y805>JEK_805  </a>
+	;
+/* no freemod in this version; cf. <a href=3D#y422>JOIK_JEK_422   </a>*/
+/* this reference to a version of JOIK and JEK which already have the lexe=
r tokens attached prevents shift/reduce errors. The problem is resolved in =
a hard-coded parser implementation which builds lexer_K, before lexer_S, be=
fore lexer_E and lexer_F. */
+<a name=3Dy960>lexer_L_960  </a>	:	<a href=3D#y712>lexer_L_712  </a><a hre=
f=3D#y961>number_root_961  </a>
+	;
+<a name=3Dy961>number_root_961  </a>	:	<a href=3D#y672>PA_672  </a>
+	|	<a href=3D#y961>number_root_961   </a><a href=3D#y672>PA_672   </a>
+	|	<a href=3D#y961>number_root_961   </a><a href=3D#y987>lerfu_word_987  <=
/a>
+	;
+<a name=3Dy965>lexer_M_965  </a>	:	<a href=3D#y713>lexer_M_713  </a><a hre=
f=3D#y991>GIhEK_root_991  </a><a href=3D#y508>BO_508  </a>
+	|	<a href=3D#y713>lexer_M_713  </a><a href=3D#y991>GIhEK_root_991   </a><=
a href=3D#y971>simple_tag_971   </a><a href=3D#y508>BO_508  </a>
+	;
+<a name=3Dy966>lexer_N_966  </a>	:	<a href=3D#y714>lexer_N_714  </a><a hre=
f=3D#y991>GIhEK_root_991  </a><a href=3D#y551>KE_551  </a>
+	|	<a href=3D#y714>lexer_N_714  </a><a href=3D#y991>GIhEK_root_991   </a><=
a href=3D#y971>simple_tag_971   </a><a href=3D#y551>KE_551   </a>
+	;
+<a name=3Dy970>lexer_O_970  </a>	:	<a href=3D#y715>lexer_O_715  </a><a hre=
f=3D#y972>simple_tense_modal_972  </a>
+	;=09
+/* the following rule is a lexer version of non-terminal_815  for compound=
ing PU/modals; it disallows the lexer picking out FIhO clauses, which would=
 require it to have knowledge of the main parser grammar */
+<a name=3Dy971>simple_tag_971  </a>	:	<a href=3D#y972>simple_tense_modal_9=
72  </a>
+	|	<a href=3D#y971>simple_tag_971   </a><a href=3D#y957>simple_JOIK_JEK_95=
7  </a><a href=3D#y972>simple_tense_modal_972   </a>
+	;
+<a name=3Dy972>simple_tense_modal_972	</a>:	<a href=3D#y973>simple_tense_m=
odal_A_973  </a>
+	|	<a href=3D#y583>NAhE_583  </a><a href=3D#y973>simple_tense_modal_A_973 =
 </a>
+	|	<a href=3D#y554>KI_554  </a>
+	|	<a href=3D#y522>CUhE_522   </a>
+	;
+
+<a name=3Dy973>simple_tense_modal_A_973	</a>:	</a><a href=3D#y974>modal_97=
4  </a>
+	|	<a href=3D#y974>modal_974  </a><a href=3D#y554>KI_554  </a>
+	|	<a href=3D#y977>tense_A_977  </a>
+	;
+<a name=3Dy974>modal_974  </a>	:	<a href=3D#y975>modal_A_975  </a>
+	|	<a href=3D#y975>modal_A_975  </a><a href=3D#y581>NAI_581   </a>
+	;
+<a name=3Dy975>modal_A_975  </a>	:	<a href=3D#y502>BAI_502  </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y502>BAI_502   </a>
+	;
+<a name=3Dy977>tense_A_977  </a>	:	<a href=3D#y978>tense_B_978  </a>
+	|	<a href=3D#y978>tense_B_978  </a><a href=3D#y554>KI_554  </a>
+	;
+<a name=3Dy978>tense_B_978  </a>	:	<a href=3D#y979>tense_C_979  </a>
+	|	<a href=3D#y514>CAhA_514  </a>
+	|	<a href=3D#y979>tense_C_979  </a><a href=3D#y514>CAhA_514  </a>
+	;=09
+/* specifies actuality/potentiality of the bridi */
+/* puca'a =3D actually was */
+/* baca'a =3D actually will be */
+/* bapu'i =3D can and will have */
+/* banu'o =3D can, but won't have yet */
+/* canu'ojebapu'i =3D can, hasn't yet, but will */
+<a name=3Dy979>tense_C_979  </a>	:	<a href=3D#y1030>time_1030  </a>
+/* time-only */
+/* space defaults to time-space reference space */
+	|	<a href=3D#y1040>space_1040   </a>
+/* can include time if specified with VIhA; otherwise time defaults to the=
 time-space reference time */
+	|	<a href=3D#y1030>time_1030   </a><a href=3D#y1040>space_1040   </a>
+/* time and space - If <a href=3D#y1040>space_1040   </a>is marked with VI=
hA for space-time the tense may be self-contradictory */
+/* interval prop before space_time is for time distribution */
+	|	<a href=3D#y1040>space_1040   </a><a href=3D#y1030>time_1030   </a>
+	;
+<a name=3Dy980>lexer_P_980  </a>	:	<a href=3D#y716>lexer_P_716  </a><a hre=
f=3D#y981>GIK_root_981  </a>
+	;
+<a name=3Dy981>GIK_root_981  </a>	:	<a href=3D#y539>GI_539  </a>
+	|	<a href=3D#y539>GI_539  </a><a href=3D#y581>NAI_581   </a>
+	;
+
+<a name=3Dy985>lexer_Q_985  </a>	:	<a href=3D#y717>lexer_Q_717  </a><a hre=
f=3D#y986>lerfu_string_root_986  </a>
+	;
+<a name=3Dy986>lerfu_string_root_986	</a>:	<a href=3D#y987>lerfu_word_987 =
 </a>
+	|	<a href=3D#y986>lerfu_string_root_986  </a><a href=3D#y987>lerfu_word_9=
87  </a>
+	|	<a href=3D#y986>lerfu_string_root_986  </a><a href=3D#y672>PA_672   </a=
>
+	;
+<a name=3Dy987>lerfu_word_987  </a>	:	<a href=3D#y513>BY_513  </a>
+	|	<a href=3D#y559>LAU_559  </a><a href=3D#y987>lerfu_word_987  </a>
+	|	<a href=3D#y605>TEI_605  </a><a href=3D#y986>lerfu_string_root_986  </a=
><a href=3D#y533>FOI_533  </a>
+	;
+<a name=3Dy990>lexer_R_990  </a>	:	<a href=3D#y718>lexer_R_718  </a><a hre=
f=3D#y991>GIhEK_root_991  </a>
+	;
+<a name=3Dy991>GIhEK_root_991  </a>	:	<a href=3D#y541>GIhA_541  </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y541>GIhA_541   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y541>GIhA_541   </a>
+	|	<a href=3D#y541>GIhA_541   </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y596>SE_596  </a><a href=3D#y541>GIhA_541   </a><a href=3D#y=
581>NAI_581   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y541>GIhA_541   </a><a href=3D#y=
581>NAI_581   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y596>SE_596  </a><a href=3D#y541=
>GIhA_541   </a>
+	|	<a href=3D#y578>NA_578  </a><a href=3D#y596>SE_596  </a><a href=3D#y541=
>GIhA_541   </a><a href=3D#y581>NAI_581   </a>
+	;
+<a name=3Dy995>lexer_S_995  </a>	:	<a href=3D#y719>lexer_S_719  </a><a hre=
f=3D#y545>I_545  </a>
+	;
+<a name=3Dy1000>lexer_T_1000  </a>	:	<a href=3D#y720>lexer_T_720  </a><a h=
ref=3D#y545>I_545  </a><a href=3D#y957>simple_JOIK_JEK_957  </a>
+	;
+<a name=3Dy1005>lexer_U_1005  </a>	:	<a href=3D#y721>lexer_U_721  </a><a h=
ref=3D#y926>JEK_root_926  </a><a href=3D#y508>BO_508  </a>
+	|	<a href=3D#y721>lexer_U_721   </a><a href=3D#y926>JEK_root_926  </a><a =
href=3D#y971>simple_tag_971   </a><a href=3D#y508>BO_508  </a>
+	;
+<a name=3Dy1010>lexer_V_1010  </a>	:	<a href=3D#y722>lexer_V_722  </a><a h=
ref=3D#y931>JOIK_root_931  </a><a href=3D#y508>BO_508  </a>
+	|	<a href=3D#y722>lexer_V_722   </a><a href=3D#y931>JOIK_root_931   </a><=
a href=3D#y971>simple_tag_971   </a><a href=3D#y508>BO_508  </a>
+	;
+<a name=3Dy1015>lexer_W_1015  </a>	:	<a href=3D#y723>lexer_W_723  </a><a h=
ref=3D#y931>JOIK_root_931  </a><a href=3D#y551>KE_551  </a>
+	|	<a href=3D#y723>lexer_W_723  </a><a href=3D#y931>JOIK_root_931   </a><a=
 href=3D#y971>simple_tag_971   </a><a href=3D#y551>KE_551   </a>
+	;
+<a name=3Dy1025>lexer_Y_1025  </a>	:	<a href=3D#y725>lexer_Y_725  </a><a h=
ref=3D#y961>number_root_961  </a><a href=3D#y663>MOI_663  </a>
+	|	<a href=3D#y725>lexer_Y_725  </a><a href=3D#y986>lerfu_string_root_986 =
 </a><a href=3D#y663>MOI_663  </a>
+	;
+
+<a name=3Dy1030>time_1030  </a>	:	<a href=3D#y624>ZI_624  </a>
+	|	<a href=3D#y624>ZI_624  </a><a href=3D#y1031>time_A_1031   </a>
+	|	<a href=3D#y1031>time_A_1031   </a>
+	;
+<a name=3Dy1031>time_A_1031  </a>	:	<a href=3D#y1032>time_B_1032  </a>
+	|	<a href=3D#y1034>time_interval_1034  </a>
+	|	<a href=3D#y1032>time_B_1032   </a><a href=3D#y1034>time_interval_1034 =
 </a>
+	;
+<a name=3Dy1032>time_B_1032  </a>	:	<a href=3D#y1033>time_offset_1033  </a=
>
+	|	<a href=3D#y1032>time_B_1032   </a><a href=3D#y1033>time_offset_1033  <=
/a>
+	;
+<a name=3Dy1033>time_offset_1033  </a>	:	<a href=3D#y1035>time_direction_1=
035  </a>
+	|	<a href=3D#y1035>time_direction_1035  </a><a href=3D#y624>ZI_624  </a>
+	;
+<a name=3Dy1034>time_interval_1034	</a>:	<a href=3D#y622>ZEhA_622  </a>
+	|	<a href=3D#y622>ZEhA_622   </a><a href=3D#y1035>time_direction_1035  </=
a>
+	|	<a href=3D#y1036>time_int_props_1036  </a>
+	|	<a href=3D#y622>ZEhA_622   </a><a href=3D#y1036>time_int_props_1036  </=
a>
+	|	<a href=3D#y622>ZEhA_622   </a><a href=3D#y1035>time_direction_1035  </=
a><a href=3D#y1036>time_int_props_1036  </a>
+	;
+<a name=3Dy1035>time_direction_1035	</a>:	<a href=3D#y592>PU_592  </a>
+	|	<a href=3D#y592>PU_592   </a><a href=3D#y581>NAI_581   </a>
+	;
+<a name=3Dy1036>time_int_props_1036	</a>:	<a href=3D#y1051>interval_proper=
ty_1051  </a>
+	|	<a href=3D#y1036>time_int_props_1036  </a><a href=3D#y1051>interval_pro=
perty_1051   </a>
+	;
+<a name=3Dy1040>space_1040  </a>	:	<a href=3D#y1042>space_A_1042  </a>
+	|	<a href=3D#y1041>space_motion_1041   </a>
+	|	<a href=3D#y1042>space_A_1042   </a><a href=3D#y1041>space_motion_1041 =
  </a>
+	;
+<a name=3Dy1041>space_motion_1041	</a>:	<a href=3D#y577>MOhI_577  </a><a h=
ref=3D#y1045>space_offset_1045  </a>
+	;
+<a name=3Dy1042>space_A_1042  </a>	:	<a href=3D#y613>VA_613  </a>
+	|	<a href=3D#y613>VA_613  </a><a href=3D#y1043>space_B_1043  </a>
+	|	<a href=3D#y1043>space_B_1043  </a>
+	;
+<a name=3Dy1043>space_B_1043  </a>	:	<a href=3D#y1044>space_C_1044  </a>
+	|	<a href=3D#y1046>space_intval_1046  </a>
+	|	<a href=3D#y1044>space_C_1044  </a><a href=3D#y1046>space_intval_1046  =
</a>
+	;
+<a name=3Dy1044>space_C_1044  </a>	:	<a href=3D#y1045>space_offset_1045  <=
/a>
+	|	<a href=3D#y1044>space_C_1044  </a><a href=3D#y1045>space_offset_1045  =
</a>
+	;
+<a name=3Dy1045>space_offset_1045	</a>:	<a href=3D#y1048>space_direction_1=
048  </a>
+	|	<a href=3D#y1048>space_direction_1048  </a><a href=3D#y613>VA_613  </a>
+	;
+<a name=3Dy1046>space_intval_1046	</a>:	<a href=3D#y1047>space_intval_A_10=
47  </a>
+	|	<a href=3D#y1047>space_intval_A_1047  </a><a href=3D#y1048>space_direct=
ion_1048  </a>
+	|	<a href=3D#y1049>space_int_props_1049  </a>
+	|	<a href=3D#y1047>space_intval_A_1047  </a><a href=3D#y1049>space_int_pr=
ops_1049  </a>
+	|	<a href=3D#y1047>space_intval_A_1047  </a><a href=3D#y1048>space_direct=
ion_1048  </a><a href=3D#y1049>space_int_props_1049  </a>
+	;
+<a name=3Dy1047>space_intval_A_1047	</a>:	<a href=3D#y615>VEhA_615  </a>
+	|	<a href=3D#y616>VIhA_616  </a>
+	|	<a href=3D#y615>VEhA_615  </a><a href=3D#y616>VIhA_616  </a>
+	;
+<a name=3Dy1048>space_direction_1048	</a>:	<a href=3D#y528>FAhA_528  </a>
+	|	<a href=3D#y528>FAhA_528  </a><a href=3D#y581>NAI_581   </a>
+	;
+<a name=3Dy1049>space_int_props_1049	</a>:	<a href=3D#y1050>space_int_prop=
s_A_1050  </a>
+	|	<a href=3D#y1049>space_int_props_1049  </a><a href=3D#y1050>space_int_p=
rops_A_1050   </a>
+	;
+<a name=3Dy1050>space_int_props_A_1050	</a>:	<a href=3D#y530>FEhE_530  </a=
><a href=3D#y1051>interval_property_1051  </a>
+	;
+/* This terminal gives an interval size in space-time (VEhA), and possibly=
 a dimensionality of the interval. The dimensionality may also be used with=
 the interval size left unspecified. When this terminal is used for the spa=
cetime origin, then barring any overriding VIhA, a VIhA here defines the di=
mensionality of the space-time being discussed. */
+<a name=3Dy1051>interval_property_1051	</a>:	<a href=3D#y961>number_root_9=
61  </a><a href=3D#y594>ROI_594  </a>
+	|	<a href=3D#y961>number_root_961   </a><a href=3D#y594>ROI_594  </a><a h=
ref=3D#y581>NAI_581   </a>
+	|	<a href=3D#y604>TAhE_604  </a>
+	|	<a href=3D#y604>TAhE_604  </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y621>ZAhO_621   </a>
+	|	<a href=3D#y621>ZAhO_621   </a><a href=3D#y581>NAI_581   </a>
+	;
+/* extensional/intensional interval parameters */
+/* These may be appended to any defined interval, or may stand in place of=
 either time or space tenses. If no other tense is present, this terminal s=
tands for the time-space interval parameter with an unspecified interval.*/
+/* roroi =3D always and everywhere */
+/* roroiku'avi =3D always here (ku'a =3D intersection) */
+/* puroroi =3D always in the past
+/* paroi =3D once upon a time (somewhere) */
+/* paroiku'avi =3D once upon a time here */
+
+/* The following are =93Lexer-only rules=94, covered by steps 1-4  describ=
ed at the beginning. The grammar of these constructs is nonexistent, except=
 possibly in cases where they interact with each other. Even there, however=
, the effects are semantic rather than grammatical. Where it is believed po=
ssible that conflicts could exist, the grammar of these constructs has been=
 put in the above grammar, even though the lexer/Preparser will actually pr=
event these from being passed thru to the parse routine. (Otherwise we have=
 to put unacceptably fancy code in the PreParser to determine just when the=
se can be passed thru, and when they can't.)
+Constructs in this category include quotes and indicators as defined above=
. (The above grammar handles utterance scope (free_modifier) and clause sco=
pe (gap) applications of the latter, however, and indicators should be allo=
wed to be absorbed into almost any word without changing its grammar.
+<a href=3D#y601>SI_601   </a>, <a href=3D#y595>SA_595  </a>, and <a href=
=3D#y603>SU_603  </a>are metalinguistic erasers.
+<a name=3Dy1100>token_1100  </a>	:	<a href=3D#y698>any_word_698  </a>
+	|	<a href=3D#y503>BAhE_503  </a><a href=3D#y698>any_word_698  </a>
+	|	<a href=3D#y699>anything_699  </a>
+	|	<a href=3D#y698>any_word_698  </a><a href=3D#y511>BU_511   </a>
+	|	<a href=3D#y698>any_word_698  </a><a href=3D#y524>DAhO_524  </a>
+	|	<a href=3D#y698>any_word_698  </a><a href=3D#y536>FUhO_536  </a>
+	|	<a href=3D#y698>any_word_698  </a><a href=3D#y535>FUhE_535  </a>
+	|	<a href=3D#y698>any_word_698  </a><a href=3D#y612>UI_612   </a>
+	|	<a href=3D#y698>any_word_698  </a><a href=3D#y612>UI_612   </a><a href=
=3D#y581>NAI_581   </a>
+	|	<a href=3D#y698>any_word_698  </a><a href=3D#y619>Y_619  </a>
+	|	<a href=3D#y698>any_word_698  </a><a href=3D#y515>CAI_515  </a>
+	|	<a href=3D#y698>any_word_698  </a><a href=3D#y515>CAI_515  </a><a href=
=3D#y581>NAI_581   </a>
+	|	<a href=3D#y612>UI_612   </a><a href=3D#y581>NAI_581   </a>
+	|	<a href=3D#y515>CAI_515  </a><a href=3D#y581>NAI_581   </a>
+	;
+<a name=3Dy1101>null_1101  </a>	:	<a href=3D#y698>any_word_698  </a><a hre=
f=3D#y601>SI_601  </a>
+	|	possibly_unlexable_word (PAUSE)   <a href=3D#y601>SI_601   </a>
+	|	<a href=3D#y20>utterance_20   </a><a href=3D#y595>SA_595  </a>
+	|	possibly unlexable string (PAUSE)   <a href=3D#y595>SA_595  </a>
+erases back to the last individual token I or NIhO or start of text, ignor=
ing the insides of ZOI, ZO, and LOhU/LEhU quotes. Start of text is defined =
for SU below.
+	|	<a href=3D#y3>text_C_3  </a><a href=3D#y603>SU_603  </a>
+	|	possibly unparsable text (PAUSE)	<a href=3D#y603>SU_603  </a>
+erases back to start of text which is the beginning of a speaker's stateme=
nt, a parenthesis (TO/TOI), a LU/LIhU quote, or a TUhE/TUhU utterance strin=
g.
+	;
+*/ %%
+
+YACC Grammar Cross-Reference
+A_501	EK_root_911
+anaphora_400	sumti_G_97
+anything_699	token_1100, ZOI_quote_434
+any_word_698	null_1101, token_1100, ZOI_quote_434, ZO_quote_435
+any_words_697	LOhU_quote_436
+BAhE_503	token_1100
+BAI_502	modal_A_975
+BE_446	linkargs_160
+BE_504	BE_446
+BEhO_506	BEhO_gap_467
+BEhO_gap_467	linkargs_160
+BEI_442	links_161
+BEI_505	BEI_442
+BIhE_439	MEX_A_311
+BIhE_650	BIhE_439
+BIhI_507	interval_932
+BO_479	selbri_F_136
+BO_508	BO_479, lexer_C_915, lexer_I_945, lexer_K_955, lexer_M_965, lexer_U=
_1005, lexer_V_1010
+BOI_651	BOI_gap_461, sub_gap_462
+BOI_gap_461	anaphora_400, operand_C_385, quantifier_300
+bridi_tail_50	bridi_tail_50, sentence_40
+bridi_tail_A_51	bridi_tail_50, bridi_tail_A_51
+bridi_tail_B_52	bridi_tail_A_51, bridi_tail_B_52
+bridi_tail_C_53	bridi_tail_B_52
+bridi_valsi_407	tanru_unit_B_152
+bridi_valsi_A_408	bridi_valsi_407
+BRIVLA_509	bridi_valsi_A_408
+BU_511	token_1100
+BY_513	lerfu_word_987
+CAhA_514	tense_B_978
+CAI_515	indicator_413, token_1100
+CEhE_495	terms_B_82
+CEhE_517	CEhE_495
+CEI_444	tanru_unit_150
+CEI_516	CEI_444
+cmene_404	sumti_G_97, text_0, vocative_35
+CMENE_518	cmene_A_405
+cmene_A_405	cmene_404, cmene_A_405
+CO_443	selbri_B_132
+CO_519	CO_443
+COI_416	COI_416, DOI_415
+COI_520	COI_A_417
+COI_A_417	COI_416
+CU_521	front_gap_451
+CUhE_522	simple_tense_modal_972
+DAhO_524	indicator_413, token_1100
+description_110	sumti_G_97
+discursive_bridi_34	free_modifier_A_33
+DOhU_526	DOhU_gap_457
+DOhU_gap_457	vocative_35
+DOI_415	vocative_35
+DOI_525	DOI_415
+EK_802	fragment_20, JOIK_EK_421
+EK_BO_803	operand_B_383, sumti_C_93
+EK_KE_804	operand_381, sumti_A_91
+EK_root_911	lexer_B_910, lexer_C_915, lexer_D_916
+error	BEhO_gap_467, BOI_gap_461, DOhU_gap_457, FEhU_gap_458, gap_450, GEhU=
_gap_464, KEhE_gap_466, KEI_gap_453, KUhO_gap_469, LIhU_gap_448, LOhO_gap_4=
72, LUhU_gap_463, MEhU_gap_465, MEX_gap_452, NUhU_gap_460, right_bracket_ga=
p_471, right_br_no_free_474, SEhU_gap_459, sub_gap_462, TEhU_gap_473, TOI_g=
ap_468, TUhU_gap_454, VAU_gap_456
+FA_481	mod_head_490
+FA_527	FA_481
+FAhA_528	space_direction_1048
+FEhE_530	space_int_props_A_1050
+FEhU_531	FEhU_gap_458
+FEhU_gap_458	tense_modal_815
+FIhO_437	tense_modal_815
+FIhO_532	FIhO_437
+FOI_533	lerfu_word_987
+fragment_20	paragraph_10
+free_modifier_32	anaphora_400, BE_446, BEhO_gap_467, BEI_442, BIhE_439, BO=
_479, BOI_gap_461, bridi_valsi_407, CEhE_495, CEI_444, cmene_404, CO_443, E=
K_802, EK_BO_803, EK_KE_804, FA_481, FEhU_gap_458, FIhO_437, free_modifier_=
32, front_gap_451, FUhA_441, gap_450, GEhU_gap_464, GEK_807, GIhEK_818, GIh=
EK_BO_813, GIhEK_KE_814, GIK_816, GOI_485, GUhEK_808, I_819, I_BO_811, I_JE=
K_820, JAI_478, JEK_BO_821, JOhI_431, JOIK_BO_822, JOIK_EK_421, JOIK_JEK_42=
2, JOIK_KE_823, KE_493, KEhE_gap_466, KEI_gap_453, KUhO_gap_469, LA_499, LE=
_488, left_bracket_470, LI_489, LOhO_gap_472, LUhU_gap_463, MAhO_430, ME_47=
7, MEhU_gap_465, MEX_gap_452, MEX_operator_374, MOhE_427, MOI_476, NA_445, =
NAhE_482, NAhE_BO_809, NAhU_429, NA_KU_810, NIhE_428, NOI_484, NU_A_426, NU=
hA_475, NUhI_496, NUhU_gap_460, para_mark_410, PEhE_494, PEhO_438, qualifie=
r_483, quote_arg_432, right_bracket_gap_471, SE_480, SEI_440, SOI_498, TEhU=
_gap_473, tense_modal_815, text_0, TUhE_447, TUhU_gap_454, VAU_gap_456, VUh=
O_497, XI_424, ZIhE_487, ZOhU_492
+free_modifier_A_33	free_modifier_32
+front_gap_451	discursive_bridi_34, sentence_40
+FUhA_441	MEX_310
+FUhA_655	FUhA_441
+FUhE_535	indicators_411, token_1100
+FUhO_536	indicator_413, token_1100
+GA_537	lexer_G_935
+GAhO_656	JOIK_root_931
+gap_450	description_110, modifier_84, sumti_E_95
+GEhU_538	GEhU_gap_464
+GEhU_gap_464	relative_clause_122
+GEK_807	gek_sentence_54, operand_C_385, sumti_D_94, term_set_85
+gek_sentence_54	bridi_tail_C_53, gek_sentence_54
+GI_539	GIK_root_981, lexer_G_935
+GIhA_541	GIhEK_root_991
+GIhEK_818	bridi_tail_A_51, fragment_20
+GIhEK_BO_813	bridi_tail_B_52
+GIhEK_KE_814	bridi_tail_50
+GIhEK_root_991	lexer_M_965, lexer_N_966, lexer_R_990
+GIK_816	gek_sentence_54, GUhEK_selbri_137, operand_C_385, operator_A_371, =
sumti_D_94, term_set_85
+GIK_root_981	lexer_G_935, lexer_P_980
+GOhA_543	bridi_valsi_A_408
+GOI_485	relative_clause_122
+GOI_542	GOI_485
+GUhA_544	lexer_H_940
+GUhEK_808	GUhEK_selbri_137, operator_A_371
+GUhEK_selbri_137	selbri_F_136
+I_545	I_root_956, lexer_S_995, lexer_T_1000
+I_819	paragraph_10, text_B_2
+I_BO_811	statement_B_13, text_B_2
+I_JEK_820	statement_A_12, text_B_2
+indicator_413	indicators_A_412
+indicators_411	text_0
+indicators_A_412	indicators_411, indicators_A_412
+interval_932	JOIK_root_931
+interval_property_1051	space_int_props_A_1050, time_int_props_1036
+I_root_956	lexer_K_955
+JA_546	JEK_root_926
+JAI_478	tanru_unit_B_152
+JAI_547	JAI_478
+JEK_805	JOIK_JEK_422, simple_JOIK_JEK_957
+JEK_BO_821	operator_A_371, selbri_E_135
+JEK_root_926	lexer_E_925, lexer_U_1005
+JOhI_431	operand_C_385
+JOhI_657	JOhI_431
+JOI_548	JOIK_root_931
+JOIK_806	JOIK_EK_421, JOIK_JEK_422, simple_JOIK_JEK_957
+JOIK_BO_822	operand_B_383, operator_A_371, selbri_E_135, sumti_C_93
+JOIK_EK_421	operand_A_382, sumti_B_92
+JOIK_JEK_422	NU_425, operator_370, selbri_D_134, tag_491, terms_A_81, text=
_A_1
+JOIK_KE_823	operand_381, operator_370, selbri_D_134, sumti_A_91
+JOIK_root_931	lexer_F_930, lexer_G_935, lexer_V_1010, lexer_W_1015
+KE_493	gek_sentence_54, operator_B_372, tanru_unit_B_152
+KE_551	KE_493, lexer_D_916, lexer_N_966, lexer_W_1015
+KEhE_550	KEhE_gap_466
+KEhE_gap_466	bridi_tail_50, gek_sentence_54, operand_381, operator_370, op=
erator_B_372, selbri_D_134, sumti_A_91, tanru_unit_B_152
+KEI_552	KEI_gap_453
+KEI_gap_453	tanru_unit_B_152
+KI_554	simple_tense_modal_972, simple_tense_modal_A_973, tense_A_977
+KOhA_555	anaphora_400
+KU_556	gap_450, lexer_J_950
+KUhE_658	MEX_gap_452
+KUhO_557	KUhO_gap_469
+KUhO_gap_469	relative_clause_122
+LA_499	description_110, sumti_G_97
+LA_558	LA_499
+LAhE_561	qualifier_483
+LAU_559	lerfu_word_987
+LE_488	description_110
+LE_562	LE_488
+left_bracket_470	quantifier_300, subscript_486
+LEhU_565	LOhU_quote_436
+lerfu_string_817	anaphora_400, operand_C_385, subscript_486
+lerfu_string_root_986	lerfu_string_root_986, lerfu_word_987, lexer_Q_985, =
lexer_Y_1025, utt_ordinal_root_906
+lerfu_word_987	lerfu_string_root_986, lerfu_word_987, number_root_961
+lexer_A_701	lexer_A_905
+lexer_A_905	utterance_ordinal_801
+lexer_B_702	lexer_B_910
+lexer_B_910	EK_802
+lexer_C_703	lexer_C_915
+lexer_C_915	EK_BO_803
+lexer_D_704	lexer_D_916
+lexer_D_916	EK_KE_804
+lexer_E_705	lexer_E_925
+lexer_E_925	JEK_805
+lexer_F_706	lexer_F_930
+lexer_F_930	JOIK_806
+lexer_G_707	lexer_G_935
+lexer_G_935	GEK_807
+lexer_H_708	lexer_H_940
+lexer_H_940	GUhEK_808
+lexer_I_709	lexer_I_945
+lexer_I_945	NAhE_BO_809
+lexer_J_710	lexer_J_950
+lexer_J_950	NA_KU_810
+lexer_K_711	lexer_K_955
+lexer_K_955	I_BO_811
+lexer_L_712	lexer_L_960
+lexer_L_960	number_812
+lexer_M_713	lexer_M_965
+lexer_M_965	GIhEK_BO_813
+lexer_N_714	lexer_N_966
+lexer_N_966	GIhEK_KE_814
+lexer_O_715	lexer_O_970
+lexer_O_970	tense_modal_815
+lexer_P_716	lexer_P_980
+lexer_P_980	GIK_816
+lexer_Q_717	lexer_Q_985
+lexer_Q_985	lerfu_string_817
+lexer_R_718	lexer_R_990
+lexer_R_990	GIhEK_818
+lexer_S_719	lexer_S_995
+lexer_S_995	I_819
+lexer_T_1000	I_JEK_820
+lexer_T_720	lexer_T_1000
+lexer_U_1005	JEK_BO_821
+lexer_U_721	lexer_U_1005
+lexer_V_1010	JOIK_BO_822
+lexer_V_722	lexer_V_1010
+lexer_W_1015	JOIK_KE_823
+lexer_W_723	lexer_W_1015
+lexer_Y_1025	PA_MOI_824
+lexer_Y_725	lexer_Y_1025
+LI_489	sumti_G_97
+LI_566	LI_489
+LIhU_567	LIhU_gap_448
+LIhU_gap_448	quote_arg_A_433
+linkargs_160	fragment_20, tanru_unit_A_151
+links_161	fragment_20, linkargs_160, links_161
+LOhO_568	LOhO_gap_472
+LOhO_gap_472	sumti_G_97
+LOhU_569	LOhU_quote_436
+LOhU_quote_436	quote_arg_A_433
+LU_571	quote_arg_A_433
+LUhU_573	LUhU_gap_463
+LUhU_gap_463	operand_C_385, sumti_G_97
+MAhO_430	MEX_operator_374
+MAhO_662	MAhO_430
+MAI_661	utt_ordinal_root_906
+ME_477	tanru_unit_B_152
+ME_574	ME_477
+MEhU_575	MEhU_gap_465
+MEhU_gap_465	tanru_unit_B_152
+MEX_310	MEX_310, MEX_operator_374, quantifier_300, subscript_486, sumti_G_=
97
+MEX_A_311	MEX_310, MEX_A_311
+MEX_B_312	MEX_A_311, MEX_C_313
+MEX_C_313	MEX_B_312, MEX_C_313, operand_C_385
+MEX_gap_452	MEX_B_312
+MEX_operator_374	MEX_operator_374, operator_B_372, tanru_unit_B_152
+modal_974	simple_tense_modal_A_973
+modal_A_975	modal_974
+mod_head_490	modifier_84
+modifier_84	term_83
+MOhE_427	operand_C_385
+MOhE_664	MOhE_427
+MOhI_577	space_motion_1041
+MOI_476	tanru_unit_B_152
+MOI_663	lexer_Y_1025, MOI_476
+NA_445	fragment_20, gek_sentence_54, selbri_A_131
+NA_578	EK_root_911, GIhEK_root_991, JEK_root_926, lexer_J_950, NA_445
+NAhE_482	MEX_operator_374, selbri_F_136, tanru_unit_B_152
+NAhE_583	lexer_I_945, NAhE_482, simple_tense_modal_972
+NAhE_BO_809	qualifier_483
+NAhU_429	MEX_operator_374
+NAhU_665	NAhU_429
+NAI_581	COI_A_417, EK_root_911, GIhEK_root_991, GIK_root_981, indicator_41=
3, interval_932, interval_property_1051, JEK_root_926, JOIK_root_931, lexer=
_G_935, lexer_H_940, modal_974, NU_A_426, space_direction_1048, text_0, tim=
e_direction_1035, token_1100
+NA_KU_810	term_83
+NIhE_428	operand_C_385
+NIhE_666	NIhE_428
+NIhO_584	para_mark_410
+NOI_484	relative_clause_122
+NOI_585	NOI_484
+NU_425	NU_425, tanru_unit_B_152
+NU_586	NU_A_426
+NU_A_426	NU_425
+NUhA_475	tanru_unit_B_152
+NUhA_667	NUhA_475
+NUhI_496	term_set_85
+NUhI_587	NUhI_496
+NUhU_588	NUhU_gap_460
+NUhU_gap_460	term_set_85
+number_812	quantifier_300, subscript_486
+number_root_961	interval_property_1051, lexer_L_960, lexer_Y_1025, number_=
root_961, utt_ordinal_root_906
+operand_381	MEX_B_312, operand_381, operand_C_385, rp_operand_332
+operand_A_382	operand_381, operand_A_382
+operand_B_383	operand_A_382, operand_B_383
+operand_C_385	operand_B_383, operand_C_385
+operator_370	MEX_310, MEX_A_311, MEX_B_312, operator_370, operator_B_372, =
rp_expression_330
+operator_A_371	operator_370, operator_A_371
+operator_B_372	operator_A_371
+PA_672	lerfu_string_root_986, number_root_961
+PA_MOI_824	bridi_valsi_A_408
+paragraph_10	paragraph_10, paragraphs_4
+paragraphs_4	paragraphs_4, text_C_3
+para_mark_410	paragraphs_4, para_mark_410, text_B_2
+parenthetical_36	free_modifier_A_33
+PEhE_494	terms_A_81
+PEhE_591	PEhE_494
+PEhO_438	MEX_B_312
+PEhO_673	PEhO_438
+prenex_30	fragment_20, statement_11, subsentence_41
+PU_592	time_direction_1035
+qualifier_483	operand_C_385, sumti_G_97
+quantifier_300	fragment_20, operand_C_385, sumti_E_95, sumti_F_96, sumti_t=
ail_A_112
+quote_arg_432	sumti_G_97
+quote_arg_A_433	quote_arg_432
+RAhO_593	bridi_valsi_A_408
+relative_clause_122	relative_clauses_121
+relative_clauses_121	fragment_20, relative_clauses_121, sumti_90, sumti_E_=
95, sumti_G_97, sumti_tail_111, sumti_tail_A_112, vocative_35
+right_bracket_gap_471	quantifier_300
+right_br_no_free_474	subscript_486
+ROI_594	interval_property_1051
+rp_expression_330	MEX_310, rp_operand_332
+rp_operand_332	rp_expression_330
+SA_595	null_1101
+SE_480	MEX_operator_374, tanru_unit_B_152
+SE_596	EK_root_911, GIhEK_root_991, interval_932, JEK_root_926, JOIK_root_=
931, lexer_G_935, lexer_H_940, modal_A_975, SE_480
+SEhU_598	SEhU_gap_459
+SEhU_gap_459	discursive_bridi_34
+SEI_440	discursive_bridi_34
+SEI_597	SEI_440
+selbri_130	bridi_tail_C_53, discursive_bridi_34, GUhEK_selbri_137, MEX_ope=
rator_374, operand_C_385, selbri_A_131, sumti_E_95, sumti_tail_A_112, tense=
_modal_815, vocative_35
+selbri_A_131	selbri_130
+selbri_B_132	selbri_A_131, selbri_B_132
+selbri_C_133	selbri_B_132, selbri_C_133, selbri_D_134, tanru_unit_B_152
+selbri_D_134	selbri_C_133, selbri_D_134
+selbri_E_135	selbri_D_134, selbri_E_135
+selbri_F_136	GUhEK_selbri_137, selbri_E_135, selbri_F_136
+sentence_40	statement_C_14, subsentence_41
+SI_601	null_1101
+simple_JOIK_JEK_957	I_root_956, lexer_T_1000, simple_tag_971
+simple_tag_971	lexer_C_915, lexer_D_916, lexer_G_935, lexer_K_955, lexer_M=
_965, lexer_N_966, lexer_U_1005, lexer_V_1010, lexer_W_1015, simple_tag_971
+simple_tense_modal_972	lexer_O_970, simple_tag_971
+simple_tense_modal_A_973	simple_tense_modal_972
+SOI_498	discursive_bridi_34
+SOI_602	SOI_498
+space_1040	tense_C_979
+space_A_1042	space_1040
+space_B_1043	space_A_1042
+space_C_1044	space_B_1043, space_C_1044
+space_direction_1048	space_intval_1046, space_offset_1045
+space_int_props_1049	space_int_props_1049, space_intval_1046
+space_int_props_A_1050	space_int_props_1049
+space_intval_1046	space_B_1043
+space_intval_A_1047	space_intval_1046
+space_motion_1041	space_1040
+space_offset_1045	space_C_1044, space_motion_1041
+statement_11	paragraph_10, statement_11
+statement_A_12	statement_11, statement_A_12
+statement_B_13	statement_A_12, statement_B_13
+statement_C_14	statement_B_13
+SU_603	null_1101
+sub_gap_462	subscript_486
+subscript_486	free_modifier_A_33
+subsentence_41	gek_sentence_54, relative_clause_122, subsentence_41, tanru=
_unit_B_152
+sumti_90	discursive_bridi_34, modifier_84, operand_C_385, sumti_A_91, sumt=
i_D_94, sumti_G_97, sumti_tail_A_112, tanru_unit_B_152, term_83, vocative_3=
5
+sumti_A_91	sumti_90
+sumti_B_92	sumti_A_91, sumti_B_92
+sumti_C_93	sumti_B_92, sumti_C_93
+sumti_D_94	sumti_C_93, sumti_D_94
+sumti_E_95	sumti_D_94
+sumti_F_96	sumti_E_95
+sumti_G_97	sumti_F_96, sumti_tail_111
+sumti_tail_111	description_110
+sumti_tail_A_112	sumti_tail_111
+tag_491	gek_sentence_54, mod_head_490, selbri_130, statement_C_14, tag_491=
, tanru_unit_B_152
+TAhE_604	interval_property_1051
+tail_terms_71	bridi_tail_50, bridi_tail_A_51, bridi_tail_B_52, bridi_tail_=
C_53, gek_sentence_54
+tanru_unit_150	selbri_F_136, tanru_unit_150
+tanru_unit_A_151	tanru_unit_150
+tanru_unit_B_152	tanru_unit_A_151, tanru_unit_B_152
+TEhU_675	TEhU_gap_473
+TEhU_gap_473	MEX_operator_374, operand_C_385
+TEI_605	lerfu_word_987
+tense_A_977	simple_tense_modal_A_973
+tense_B_978	tense_A_977
+tense_C_979	tense_B_978
+tense_modal_815	tag_491
+term_83	linkargs_160, links_161, relative_clause_122, terms_B_82
+terms_80	discursive_bridi_34, fragment_20, prenex_30, sentence_40, tail_te=
rms_71, terms_80, term_set_85
+terms_A_81	terms_80, terms_A_81
+terms_B_82	terms_A_81, terms_B_82
+term_set_85	term_83
+text_0	parenthetical_36, quote_arg_A_433, text_0
+text_A_1	text_0
+text_B_2	statement_C_14, text_A_1, text_B_2
+text_C_3	null_1101, text_B_2
+time_1030	tense_C_979
+time_A_1031	time_1030
+time_B_1032	time_A_1031, time_B_1032
+time_direction_1035	time_interval_1034, time_offset_1033
+time_interval_1034	time_A_1031
+time_int_props_1036	time_interval_1034, time_int_props_1036
+time_offset_1033	time_B_1032
+TO_606	parenthetical_36
+TOI_607	TOI_gap_468
+TOI_gap_468	parenthetical_36
+TUhE_447	statement_C_14
+TUhE_610	TUhE_447
+TUhU_611	TUhU_gap_454
+TUhU_gap_454	statement_C_14
+UI_612	indicator_413, token_1100
+utterance_20	null_1101
+utterance_ordinal_801	free_modifier_A_33
+utt_ordinal_root_906	lexer_A_905
+VA_613	space_A_1042, space_offset_1045
+VAU_614	VAU_gap_456
+VAU_gap_456	fragment_20, tail_terms_71
+VEhA_615	space_intval_A_1047
+VEhO_678	right_bracket_gap_471, right_br_no_free_474
+VEI_677	left_bracket_470
+VIhA_616	space_intval_A_1047
+vocative_35	free_modifier_A_33
+VUhO_497	sumti_90
+VUhO_617	VUhO_497
+VUhU_679	MEX_operator_374
+XI_424	subscript_486
+XI_618	XI_424
+Y_619	indicator_413, token_1100
+ZAhO_621	interval_property_1051
+ZEhA_622	time_interval_1034
+ZI_624	time_1030, time_offset_1033
+ZIhE_487	relative_clauses_121
+ZIhE_625	ZIhE_487
+ZO_626	ZO_quote_435
+ZOhU_492	prenex_30
+ZOhU_628	ZOhU_492
+ZOI_627	ZOI_quote_434
+ZOI_quote_434	quote_arg_A_433
+ZO_quote_435	quote_arg_A_433
+
+</pre></landscape>
+EBNF Grammar of Lojban
+<cx "EBNF grammar">  XE "EBNF grammar"   <h3></h3>Lojban Machine Grammar, =
EBNF Version, Final Baseline
+<p>
+This EBNF document is explicitly dedicated to the public domain by its aut=
hor, The Logical Language Group, Inc.
+<p>
+Explanation of notation: All rules have the form:=20
+<p align=3Dcenter><center>
+name<sub>number</sub>=3D bnf-expression
+</center></p>which means that the grammatical construct =93name=94 is defi=
ned by =93bnf-expression=94. The number cross-references this grammar with =
the rule numbers in the YACC grammar. The names are the same as those in th=
e YACC grammar, except that subrules are labeled with A, B, C,<dots>=85 </d=
ots>in the YACC grammar and with 1, 2, 3,<dots>=85 </dots>in this grammar. =
In addition, rule 971 is =93simple_tag=94 in the YACC grammar but =93stag=
=94 in this grammar, because of its frequent appearance. <ol>
+<li>Names in lower case are grammatical constructs.=20
+<li>Names in UPPER CASE are selma'o (lexeme) names, and are terminals.=20
+<li>Concatenation is expressed by juxtaposition with no operator symbol.=
=20
+<li>| represents alternation (choice).=20
+<li>[] represents an optional element.=20
+<li>& represents and/or (=93A & B=94 is the same as =93A   | B   | A B=94)=
.=20
+<li><dots>=85 </dots>represents optional repetition of the construct to th=
e left.
+Left-grouping is implied; right-grouping is shown by explicit
+self-referential recursion with no =93<dots>=85 </dots>=94
+<li>() serves to indicate the grouping of the other operators.
+Otherwise, =93<dots>=85 </dots>=94 binds closer than &, which binds closer=
 than |.=20
+<li># is shorthand for =93[free<dots>=85</dots>]=94, a construct which app=
ears in many places.=20
+<li>// encloses an elidable terminator, which may be omitted (without chan=
ge
+of meaning) if no grammatical ambiguity results.
+</ol>
+<dl compact>
+<a name=3Db0></a><a href=3D#y0><dt>text</a><sub>0</sub>=3D	<dd>[NAI<dots>=
=85</dots>]  [CMENE<dots>=85  </dots>#  |  (indicators  &  free<dots>=85</d=
ots>)]  [joik-jek]  text-1
+<a name=3Db2></a><a href=3D#y2><dt>text-1</a><sub>2</sub>=3D	<dd>[(I  [jek=
  |  joik]  [[stag]  BO] #)<dots>=85</dots>  |  NIhO<dots>=85  </dots>#]  [=
paragraphs]
+<a name=3Db4></a><a href=3D#y4><dt>paragraphs</a><sub>4</sub>=3D	<dd>parag=
raph  [NIhO<dots>=85  </dots>#  paragraphs]
+<a name=3Db10></a><a href=3D#y10><dt>paragraph</a><sub>10</sub>=3D	<dd>(st=
atement  |  fragment)  [I #  [statement  |  fragment]]<dots>=85</dots>
+
+<a name=3Db11></a><a href=3D#y11><dt>statement</a><sub>11</sub>=3D	<dd>sta=
tement-1<br>  |  prenex  statement
+<a name=3Db12></a><a href=3D#y12><dt>statement-1</a><sub>12</sub>=3D	<dd>s=
tatement-2  [I  joik-jek  [statement-2]]<dots>=85</dots>
+<a name=3Db13></a><a href=3D#y13><dt>statement-2</a><sub>13</sub>=3D	<dd>s=
tatement-3  [I  [jek  |  joik]  [stag]  BO #  [statement-2]]
+<a name=3Db14></a><a href=3D#y14><dt>statement-3</a><sub>14</sub>=3D	<dd>s=
entence<br>  |  [tag]  TUhE #  text-1  /TUhU#/
+<a name=3Db20></a><a href=3D#y20><dt>fragment</a><sub>20</sub>=3D	<dd>ek #=
<br>  |  gihek #<br>  |  quantifier<br>  |  NA #<br>  |  terms  /VAU#/<br> =
 |  prenex<br>  |  relative-clauses<br>  |  links<br>  |  linkargs
+<a name=3Db30></a><a href=3D#y30><dt>prenex</a><sub>30</sub>=3D	<dd>terms =
 ZOhU #
+<a name=3Db40></a><a href=3D#y40><dt>sentence</a><sub>40</sub>=3D	<dd>[ter=
ms  [CU #]]  bridi-tail
+<a name=3Db41></a><a href=3D#y41><dt>subsentence</a><sub>41</sub>=3D	<dd>s=
entence<br>  |  prenex  subsentence
+<a name=3Db50></a><a href=3D#y50><dt>bridi-tail</a><sub>50</sub>=3D	<dd>br=
idi-tail-1  [gihek  [stag]  KE #  bridi-tail  /KEhE#/  tail-terms]
+<a name=3Db51></a><a href=3D#y51><dt>bridi-tail-1</a><sub>51</sub>=3D	<dd>=
bridi-tail-2  [gihek #  bridi-tail-2  tail-terms]<dots>=85</dots>
+<a name=3Db52></a><a href=3D#y52><dt>bridi-tail-2</a><sub>52</sub>=3D	<dd>=
bridi-tail-3  [gihek  [stag]  BO #  bridi-tail-2  tail-terms]
+<a name=3Db53></a><a href=3D#y53><dt>bridi-tail-3</a><sub>53</sub>=3D	<dd>=
selbri  tail-terms<br>  |  gek-sentence
+<a name=3Db54></a><a href=3D#y54><dt>gek-sentence</a><sub>54</sub>=3D	<dd>=
gek  subsentence  gik  subsentence  tail-terms<br>  |  [tag]  KE #  gek-sen=
tence  /KEhE#/<br>  |  NA #  gek-sentence
+<a name=3Db71></a><a href=3D#y71><dt>tail-terms</a><sub>71</sub>=3D	<dd>[t=
erms]  /VAU#/
+<a name=3Db80></a><a href=3D#y80><dt>terms</a><sub>80</sub>=3D	<dd>terms-1=
<dots>=85</dots>
+<a name=3Db81></a><a href=3D#y81><dt>terms-1</a><sub>81</sub>=3D	<dd>terms=
-2  [PEhE #  joik-jek  terms-2]<dots>=85</dots>
+<a name=3Db82></a><a href=3D#y82><dt>terms-2</a><sub>82</sub>=3D	<dd>term =
 [CEhE #  term]<dots>=85</dots>
+<a name=3Db83></a><a href=3D#y83><dt>term</a><sub>83</sub>=3D	<dd>sumti  |=
 (tag  |  FA #) (sumti<br>  |  /KU#/)<br>  |  termset<br>  |  NA  KU #
+<a name=3Db85></a><a href=3D#y85><dt>termset</a><sub>85</sub>=3D	<dd>NUhI =
#  gek  terms  /NUhU#/  gik  terms  /NUhU#/<br>  |  NUhI #  terms  /NUhU#/
+<a name=3Db90></a><a href=3D#y90><dt>sumti</a><sub>90</sub>=3D	<dd>sumti-1=
  [VUhO #  relative-clauses]
+<a name=3Db91></a><a href=3D#y91><dt>sumti-1</a><sub>91</sub>=3D	<dd>sumti=
-2  [(ek  |  joik)  [stag]  KE #  sumti  /KEhE#/]
+<a name=3Db92></a><a href=3D#y92><dt>sumti-2</a><sub>92</sub>=3D	<dd>sumti=
-3  [joik-ek  sumti-3]<dots>=85</dots>
+<a name=3Db93></a><a href=3D#y93><dt>sumti-3</a><sub>93</sub>=3D	<dd>sumti=
-4  [(ek<br>  |  joik)  [stag]  BO #  sumti-3]
+<a name=3Db94></a><a href=3D#y94><dt>sumti-4</a><sub>94</sub>=3D	<dd>sumti=
-5<br>  |  gek  sumti  gik  sumti-4
+<a name=3Db95></a><a href=3D#y95><dt>sumti-5</a><sub>95</sub>=3D	<dd>[quan=
tifier]  sumti-6  [relative-clauses]<br>  |  quantifier  selbri  /KU#/  [re=
lative-clauses]
+
+<a name=3Db97></a><a href=3D#y97><dt>sumti-6</a><sub>97</sub>=3D	<dd>(LAhE=
 #  |  NAhE  BO #)  [relative-clauses]  sumti  /LUhU#/<br>  |  KOhA #<br>  =
|  lerfu-string  /BOI#/<br>  | =20
+			LA #  [relative-clauses]  CMENE<dots>=85  </dots>#<br>  |
+			 (LA  |  LE) #  sumti-tail  /KU#/<br>  |  LI #  mex  /LOhO#/<br>  |  ZO=
  any-word #<br>  |  LU  text  /LIhU#/<br>  | =20
+			LOhU  any-word<dots>=85  </dots>LEhU #<br>  | =20
+			ZOI  any-word  anything  any-word #
+<a name=3Db111></a><a href=3D#y111><dt>sumti-tail</a><sub>111</sub>=3D	<dd=
>[sumti-6  [relative-clauses]]  sumti-tail-1  <br>  |relative-clauses  sumt=
i-tail-1
+<a name=3Db112></a><a href=3D#y112><dt>sumti-tail-1</a><sub>112</sub>=3D	<=
dd>[quantifier]  selbri  [relative-clauses]<br>  |  quantifier  sumti
+<a name=3Db121></a><a href=3D#y121><dt>relative-clauses</a><sub>121</sub>=
=3D	<dd>relative-clause  [ZIhE #  relative-clause]<dots>=85</dots>
+<a name=3Db122></a><a href=3D#y122><dt>relative-clause</a><sub>122</sub>=
=3D	<dd>GOI #  term  /GEhU#/<br>  |  NOI #  subsentence  /KUhO#/
+<a name=3Db130></a><a href=3D#y130><dt>selbri</a><sub>130</sub>=3D	<dd>[ta=
g]  selbri-1
+<a name=3Db131></a><a href=3D#y131><dt>selbri-1</a><sub>131</sub>=3D	<dd>s=
elbri-2<br>  |  NA #  selbri
+<a name=3Db132></a><a href=3D#y132><dt>selbri-2</a><sub>132</sub>=3D	<dd>s=
elbri-3  [CO #  selbri-2]
+<a name=3Db133></a><a href=3D#y133><dt>selbri-3</a><sub>133</sub>=3D	<dd>s=
elbri-4<dots>=85</dots>
+<a name=3Db134></a><a href=3D#y134><dt>selbri-4</a><sub>134</sub>=3D	<dd>s=
elbri-5  [joik-jek  selbri-5  |  joik  [stag]  KE #  selbri-3  /KEhE#/]<dot=
s>=85</dots>
+<a name=3Db135></a><a href=3D#y135><dt>selbri-5</a><sub>135</sub>=3D	<dd>s=
elbri-6  [(jek  |  joik)  [stag]  BO #  selbri-5]
+<a name=3Db136></a><a href=3D#y136><dt>selbri-6</a><sub>136</sub>=3D	<dd>t=
anru-unit  [BO #  selbri-6]<br>  |  [NAhE #]  guhek  selbri  gik  selbri-6
+<a name=3Db150></a><a href=3D#y150><dt>tanru-unit</a><sub>150</sub>=3D	<dd=
>tanru-unit-1  [CEI #  tanru-unit-1]<dots>=85</dots>
+<a name=3Db151></a><a href=3D#y151><dt>tanru-unit-1</a><sub>151</sub>=3D	<=
dd>tanru-unit-2  [linkargs]
+<a name=3Db152></a><a href=3D#y152><dt>tanru-unit-2</a><sub>152</sub>=3D	<=
dd>BRIVLA #<br>  |  GOhA  [RAhO] #<br>  |  KE #  selbri-3  /KEhE#/<br>  |  =
ME #  sumti  /MEhU#/  [MOI #]  |  (number<br>  |  lerfu-string)  MOI #<br> =
 |  NUhA #  mex-operator<br>  | =20
+			SE #  tanru-unit-2<br>  |  JAI #  [tag]  tanru-unit-2<br>  |  any-word =
 (ZEI  any-word)<dots>=85</dots><br>  |  NAhE #  tanru-unit-2<br>  | |  NU =
 [NAI] #  [joik-jek  NU  [NAI] #]<dots>=85  </dots>subsentence  /KEI#/
+<a name=3Db160></a><a href=3D#y160><dt>linkargs</a><sub>160</sub>=3D	<dd>B=
E #  term  [links]  /BEhO#/
+<a name=3Db161></a><a href=3D#y161><dt>links</a><sub>161</sub>=3D	<dd>BEI =
#  term  [links]
+
+<a name=3Db300></a><a href=3D#y300><dt>quantifier</a><sub>300</sub>=3D	<dd=
>number  /BOI#/<br>  |  VEI #  mex  /VEhO#/
+<a name=3Db310></a><a href=3D#y310><dt>mex</a><sub>310</sub>=3D	<dd>mex-1 =
 [operator  mex-1]<dots>=85</dots><br>  |  FUhA #  rp-expression
+<a name=3Db311></a><a href=3D#y311><dt>mex-1</a><sub>311</sub>=3D	<dd>mex-=
2  [BIhE #  operator  mex-1]
+<a name=3Db312></a><a href=3D#y312><dt>mex-2</a><sub>312</sub>=3D	<dd>oper=
and<br>  |  [PEhO #]  operator  mex-2<dots>=85  </dots>/KUhE#/
+<a name=3Db330></a><a href=3D#y330><dt>rp-expression</a><sub>330</sub>=3D	=
<dd>rp-operand  rp-operand  operator
+<a name=3Db332></a><a href=3D#y332><dt>rp-operand</a><sub>332</sub>=3D	<dd=
>operand<br>  |  rp-expression
+<a name=3Db370></a><a href=3D#y370><dt>operator</a><sub>370</sub>=3D	<dd>o=
perator-1  [joik-jek  operator-1  | =20
+			joik  [stag]  KE #  operator  /KEhE#/]<dots>=85</dots>
+<a name=3Db371></a><a href=3D#y371><dt>operator-1</a><sub>371</sub>=3D	<dd=
>operator-2  <br>  |  guhek  operator-1  gik  operator-2<br>  |  operator-2=
  (jek  |  joik)  [stag]  BO #  operator-1
+<a name=3Db372></a><a href=3D#y372><dt>operator-2</a><sub>372</sub>=3D	<dd=
>mex-operator<br>  |  KE #  operator  /KEhE#/
+<a name=3Db374></a><a href=3D#y374><dt>mex-operator</a><sub>374</sub>=3D	<=
dd>SE #  mex-operator<br>  |  NAhE #  mex-operator<br>  |  MAhO #  mex  /TE=
hU#/<br>  |  NAhU #  selbri  /TEhU#/<br>  |  VUhU #
+<a name=3Db381></a><a href=3D#y381><dt>operand</a><sub>381</sub>=3D	<dd>op=
erand-1  [(ek  |  joik)  [stag]  KE #  operand  /KEhE#/]
+<a name=3Db382></a><a href=3D#y382><dt>operand-1</a><sub>382</sub>=3D	<dd>=
operand-2  [joik-ek  operand-2]<dots>=85</dots>
+<a name=3Db383></a><a href=3D#y383><dt>operand-2</a><sub>383</sub>=3D	<dd>=
operand-3  [(ek  |  joik)  [stag]  BO #  operand-2]
+<a name=3Db385></a><a href=3D#y385><dt>operand-3</a><sub>385</sub>=3D	<dd>=
quantifier<br>  |  lerfu-string  /BOI#/<br>  |  NIhE #  selbri  /TEhU#/<br>=
  |  MOhE #  sumti  /TEhU#/<br>  |  JOhI #  mex-2<dots>=85  </dots>/TEhU#/<=
br>  |  gek  operand  gik  operand-3<br>  |  (LAhE #  |  NAhE  BO #)  opera=
nd  /LUhU#/
+<a name=3Db812></a><a href=3D#y812><dt>number</a><sub>812</sub>=3D	<dd>PA =
 [PA  |  lerfu-word]<dots>=85</dots>
+<a name=3Db817></a><a href=3D#y817><dt>lerfu-string</a><sub>817</sub>=3D	<=
dd>lerfu-word  [PA  |  lerfu-word]<dots>=85</dots>
+<a name=3Db987></a><a href=3D#y987><dt>lerfu-word</a><sub>987</sub>=3D	<dd=
>BY<br>  |  any-word  BU<br>  |  LAU  lerfu-word<br>  | =20
+			TEI  lerfu-string  FOI
+<a name=3Db802></a><a href=3D#y802><dt>ek</a><sub>802</sub>=3D	<dd>[NA]  [=
SE]  A  [NAI]
+<a name=3Db818></a><a href=3D#y818><dt>gihek</a><sub>818</sub>=3D	<dd>[NA]=
  [SE]  GIhA  [NAI]
+<a name=3Db805></a><a href=3D#y805><dt>jek</a><sub>805</sub>=3D	<dd>[NA]  =
[SE]  JA  [NAI]
+<a name=3Db806></a><a href=3D#y806><dt>joik</a><sub>806</sub>=3D	<dd>[SE] =
 JOI  [NAI]<br>  |  interval<br>  |  GAhO  interval  GAhO
+<a name=3Db932></a><a href=3D#y932><dt>interval</a><sub>932</sub>=3D	<dd>[=
SE]  BIhI  [NAI]
+<a name=3Db421></a><a href=3D#y421><dt>joik-ek</a><sub>421</sub>=3D	<dd>jo=
ik #<br>  |  ek #
+<a name=3Db422></a><a href=3D#y422><dt>joik-jek</a><sub>422</sub>=3D	<dd>j=
oik #<br>  |  jek #
+<a name=3Db807></a><a href=3D#y807><dt>gek</a><sub>807</sub>=3D	<dd>[SE]  =
GA  [NAI] #<br>  |  joik  GI #<br>  |  stag  gik
+<a name=3Db808></a><a href=3D#y808><dt>guhek</a><sub>808</sub>=3D	<dd>[SE]=
  GUhA  [NAI] #
+<a name=3Db816></a><a href=3D#y816><dt>gik</a><sub>816</sub>=3D	<dd>GI  [N=
AI] #
+<a name=3Db491></a><a href=3D#y491><dt>tag</a><sub>491</sub>=3D	<dd>tense-=
modal  [joik-jek  tense-modal]<dots>=85</dots>
+<a name=3Db971></a><a href=3D#y971><dt>stag</a><sub>971</sub>=3D	<dd>simpl=
e-tense-modal  [(jek  |  joik)  simple-tense-modal]<dots>=85</dots>
+<a name=3Db815></a><a href=3D#y815><dt>tense-modal</a><sub>815</sub>=3D	<d=
d>simple-tense-modal #<br>  |  FIhO #  selbri  /FEhU#/
+<a name=3Db972></a><a href=3D#y972><dt>simple-tense-modal</a><sub>972</sub=
>=3D	<dd>[NAhE]  [SE]  BAI  [NAI]  [KI]<br>  |  [NAhE]  (time  [space]<br> =
 |  space  [time])  &  CAhA  [KI]<br>  |  KI<br>  |  CUhE
+<a name=3Db1030></a><a href=3D#y1030><dt>time</a><sub>1030</sub>=3D	<dd>ZI=
  &  time-offset<dots>=85  </dots>&  ZEhA  [PU  [NAI]]  &  interval-propert=
y<dots>=85</dots>
+<a name=3Db1033></a><a href=3D#y1033><dt>time-offset</a><sub>1033</sub>=3D=
	<dd>PU  [NAI]  [ZI]
+<a name=3Db1040></a><a href=3D#y1040><dt>space</a><sub>1040</sub>=3D	<dd>V=
A  &  space-offset<dots>=85  </dots>&  space-interval  &  (MOhI  space-offs=
et)
+<a name=3Db1045></a><a href=3D#y1045><dt>space-offset</a><sub>1045</sub>=
=3D	<dd>FAhA  [NAI]  [VA]
+<a name=3Db1046></a><a href=3D#y1046><dt>space-interval</a><sub>1046</sub>=
=3D	<dd>((VEhA  &  VIhA)  [FAhA  [NAI]])  &  space-int-props
+<a name=3Db1049></a><a href=3D#y1049><dt>space-int-props</a><sub>1049</sub=
>=3D	<dd>(FEhE  interval-property)<dots>=85</dots>
+<a name=3Db1051></a><a href=3D#y1051><dt>interval-property</a><sub>1051</s=
ub>=3D	<dd>number  ROI  [NAI]<br>  |  TAhE  [NAI]<br>  |  ZAhO  [NAI]
+<a name=3Db32></a><a href=3D#y32><dt>free</a><sub>32</sub>=3D	<dd>SEI #  [=
terms  [CU #]]  selbri  /SEhU/<br>  | =20
+			SOI #  sumti  [sumti]  /SEhU/<br>  |  vocative  [relative-clauses]  sel=
bri  [relative-clauses]  /DOhU/<br>  |  vocative  [relative-clauses]  CMENE=
<dots>=85  </dots>#  [relative-clauses]  /DOhU/<br>  |  vocative  [sumti]  =
/DOhU/<br>  |  (number  |  lerfu-string)  MAI<br>  |  TO  text  /TOI/<br>  =
| =20
+			XI #  (number  |  lerfu-string)  /BOI/<br>  | =20
+			XI #  VEI #  mex  /VEhO/
+<a name=3Db415></a><a href=3D#y415><dt>vocative</a><sub>415</sub>=3D	<dd>(=
COI  [NAI])<dots>=85</dots>&  DOI
+<a name=3Db411></a><a href=3D#y411><dt>indicators</a><sub>411</sub>=3D	<dd=
>[FUhE]  indicator<dots>=85</dots>
+<a name=3Db413></a><a href=3D#y413><dt>indicator</a><sub>413</sub>=3D	<dd>=
(UI  | CAI) [NAI]<br>   |  Y<br>  |  DAhO<br>  |  FUhO
+</dl>
+<p>
+The  following  rules  are  non-formal:	<dl>
+<a name=3Db1100> =20
+</a><a href=3D#y1100><dt>word</a><sub>1100</sub>=3D	<dd>[BAhE]  any-word  =
[indicators]
+<dt>any-word=3D	<dd>=93any  single  word  (no  compound  cmavo)=94
+<dt>anything=3D	<dd>=93any  text  at  all,  whether  Lojban  or  not=94
+<a name=3Db1101></a><a href=3D#y1101><dt>null</a><sub>1101</sub>=3D	<dd>an=
y-word  SI<br>  |  utterance  SA<br>  |  text  SU
+</dl>FAhO  is  a  universal  terminator  and  signals  the  end  of  parsa=
ble  input.
+
+<h3>EBNF Grammar Cross-Reference</h3>
+<dl>
+<dt>A	<dd><a href=3D#b802>ek</a><sub>802</sub>
+<dt>BAI	<dd><a href=3D#b972>simple-tense-modal</a><sub>972</sub>
+<dt>BAhE	<dd><a href=3D#b1100>word</a><sub>1100</sub>
+<dt>BE	<dd><a href=3D#b160>linkargs</a><sub>160</sub>
+<dt>BEI	<dd><a href=3D#b161>links</a><sub>161</sub>
+<dt>BEhO	<dd><a href=3D#b160>linkargs</a><sub>160</sub>
+<dt>BIhE	<dd><a href=3D#b311>mex-1</a><sub>311</sub>
+<dt>BIhI	<dd><a href=3D#b932>interval</a><sub>932</sub>
+<dt>BO	<dd><a href=3D#b52>bridi-tail-2</a><sub>52</sub>, <a href=3D#b383>o=
perand-2</a><sub>383</sub>, <a href=3D#b385>operand-3</a><sub>385</sub>, <a=
 href=3D#b371>operator-1</a><sub>371</sub>, <a href=3D#b135>selbri-5</a><su=
b>135</sub>, <a href=3D#b136>selbri-6</a><sub>136</sub>, <a href=3D#b13>sta=
tement-2</a><sub>13</sub>, <a href=3D#b93>sumti-3</a><sub>93</sub>, <a href=
=3D#b97>sumti-6</a><sub>97</sub>, <a href=3D#b2>text-1</a><sub>2</sub>
+<dt>BOI	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b385>operand-=
3</a><sub>385</sub>, <a href=3D#b300>quantifier</a><sub>300</sub>, <a href=
=3D#b97>sumti-6</a><sub>97</sub>
+<dt>BRIVLA	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>BU	<dd><a href=3D#b987>lerfu-word</a><sub>987</sub>
+<dt>BY	<dd><a href=3D#b987>lerfu-word</a><sub>987</sub>
+<dt>CAI	<dd><a href=3D#b413>indicator</a><sub>413</sub>
+<dt>CAhA	<dd><a href=3D#b972>simple-tense-modal</a><sub>972</sub>
+<dt>CEI	<dd><a href=3D#b150>tanru-unit</a><sub>150</sub>
+<dt>CEhE	<dd><a href=3D#b82>terms-2</a><sub>82</sub>
+<dt>CMENE	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b97>sumti-6=
</a><sub>97</sub>, <a href=3D#b0>text</a><sub>0</sub>
+<dt>CO	<dd><a href=3D#b132>selbri-2</a><sub>132</sub>
+<dt>COI	<dd><a href=3D#b415>vocative</a><sub>415</sub>
+<dt>CU	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b40>sentence</=
a><sub>40</sub>
+<dt>CUhE	<dd><a href=3D#b972>simple-tense-modal</a><sub>972</sub>
+<dt>DAhO	<dd><a href=3D#b413>indicator</a><sub>413</sub>
+<dt>DOI	<dd><a href=3D#b415>vocative</a><sub>415</sub>
+<dt>DOhU	<dd><a href=3D#b32>free</a><sub>32</sub>
+<dt>FA	<dd><a href=3D#b83>term</a><sub>83</sub>
+<dt>FAhA	<dd><a href=3D#b1046>space-interval</a><sub>1046</sub>, <a href=
=3D#b1045>space-offset</a><sub>1045</sub>
+<dt>FEhE	<dd><a href=3D#b1049>space-int-props</a><sub>1049</sub>
+<dt>FEhU	<dd><a href=3D#b815>tense-modal</a><sub>815</sub>
+<dt>FIhO	<dd><a href=3D#b815>tense-modal</a><sub>815</sub>
+<dt>FOI	<dd><a href=3D#b987>lerfu-word</a><sub>987</sub>
+<dt>FUhA	<dd><a href=3D#b310>mex</a><sub>310</sub>
+<dt>FUhE	<dd><a href=3D#b411>indicators</a><sub>411</sub>
+<dt>FUhO	<dd><a href=3D#b413>indicator</a><sub>413</sub>
+<dt>GA	<dd><a href=3D#b807>gek</a><sub>807</sub>
+<dt>GAhO	<dd><a href=3D#b806>joik</a><sub>806</sub>
+<dt>GEhU	<dd><a href=3D#b122>relative-clause</a><sub>122</sub>
+<dt>GI	<dd><a href=3D#b807>gek</a><sub>807</sub>, <a href=3D#b816>gik</a><=
sub>816</sub>
+<dt>GIhA	<dd><a href=3D#b818>gihek</a><sub>818</sub>
+<dt>GOI	<dd><a href=3D#b122>relative-clause</a><sub>122</sub>
+<dt>GOhA	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>GUhA	<dd><a href=3D#b808>guhek</a><sub>808</sub>
+<dt>I	<dd><a href=3D#b10>paragraph</a><sub>10</sub>, <a href=3D#b12>statem=
ent-1</a><sub>12</sub>, <a href=3D#b13>statement-2</a><sub>13</sub>, <a hre=
f=3D#b2>text-1</a><sub>2</sub>
+<dt>JA	<dd><a href=3D#b805>jek</a><sub>805</sub>
+<dt>JAI	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>JOI	<dd><a href=3D#b806>joik</a><sub>806</sub>
+<dt>JOhI	<dd><a href=3D#b385>operand-3</a><sub>385</sub>
+<dt>KE	<dd><a href=3D#b50>bridi-tail</a><sub>50</sub>, <a href=3D#b54>gek-=
sentence</a><sub>54</sub>, <a href=3D#b381>operand</a><sub>381</sub>, <a hr=
ef=3D#b372>operator-2</a><sub>372</sub>, <a href=3D#b370>operator</a><sub>3=
70</sub>, <a href=3D#b134>selbri-4</a><sub>134</sub>, <a href=3D#b91>sumti-=
1</a><sub>91</sub>, <a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>KEI	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>KEhE	<dd><a href=3D#b50>bridi-tail</a><sub>50</sub>, <a href=3D#b54>ge=
k-sentence</a><sub>54</sub>, <a href=3D#b381>operand</a><sub>381</sub>, <a =
href=3D#b372>operator-2</a><sub>372</sub>, <a href=3D#b370>operator</a><sub=
>370</sub>, <a href=3D#b134>selbri-4</a><sub>134</sub>, <a href=3D#b91>sumt=
i-1</a><sub>91</sub>, <a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>KI	<dd><a href=3D#b972>simple-tense-modal</a><sub>972</sub>
+<dt>KOhA	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>KU	<dd><a href=3D#b95>sumti-5</a><sub>95</sub>, <a href=3D#b97>sumti-6=
</a><sub>97</sub>, <a href=3D#b83>term</a><sub>83</sub>
+<dt>KUhE	<dd><a href=3D#b312>mex-2</a><sub>312</sub>
+<dt>KUhO	<dd><a href=3D#b122>relative-clause</a><sub>122</sub>
+<dt>LA	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>LAU	<dd><a href=3D#b987>lerfu-word</a><sub>987</sub>
+<dt>LAhE	<dd><a href=3D#b385>operand-3</a><sub>385</sub>, <a href=3D#b97>s=
umti-6</a><sub>97</sub>
+<dt>LE	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>LEhU	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>LI	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>LIhU	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>LOhO	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>LOhU	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>LU	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>LUhU	<dd><a href=3D#b385>operand-3</a><sub>385</sub>, <a href=3D#b97>s=
umti-6</a><sub>97</sub>
+<dt>MAI	<dd><a href=3D#b32>free</a><sub>32</sub>
+<dt>MAhO	<dd><a href=3D#b374>mex-operator</a><sub>374</sub>
+<dt>ME	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>MEhU	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>MOI	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>MOhE	<dd><a href=3D#b385>operand-3</a><sub>385</sub>
+<dt>MOhI	<dd><a href=3D#b1040>space</a><sub>1040</sub>
+<dt>NA	<dd><a href=3D#b802>ek</a><sub>802</sub>, <a href=3D#b20>fragment</=
a><sub>20</sub>, <a href=3D#b54>gek-sentence</a><sub>54</sub>, <a href=3D#b=
818>gihek</a><sub>818</sub>, <a href=3D#b805>jek</a><sub>805</sub>, <a href=
=3D#b131>selbri-1</a><sub>131</sub>, <a href=3D#b83>term</a><sub>83</sub>
+<dt>NAI	<dd><a href=3D#b802>ek</a><sub>802</sub>, <a href=3D#b807>gek</a><=
sub>807</sub>, <a href=3D#b818>gihek</a><sub>818</sub>, <a href=3D#b816>gik=
</a><sub>816</sub>, <a href=3D#b808>guhek</a><sub>808</sub>, <a href=3D#b41=
3>indicator</a><sub>413</sub>, <a href=3D#b1051>interval-property</a><sub>1=
051</sub>, <a href=3D#b932>interval</a><sub>932</sub>, <a href=3D#b805>jek<=
/a><sub>805</sub>, <a href=3D#b806>joik</a><sub>806</sub>, <a href=3D#b972>=
simple-tense-modal</a><sub>972</sub>, <a href=3D#b1046>space-interval</a><s=
ub>1046</sub>, <a href=3D#b1045>space-offset</a><sub>1045</sub>, <a href=3D=
#b152>tanru-unit-2</a><sub>152</sub>, <a href=3D#b0>text</a><sub>0</sub>, <=
a href=3D#b1033>time-offset</a><sub>1033</sub>, <a href=3D#b1030>time</a><s=
ub>1030</sub>, <a href=3D#b415>vocative</a><sub>415</sub>
+<dt>NAhE	<dd><a href=3D#b374>mex-operator</a><sub>374</sub>, <a href=3D#b3=
85>operand-3</a><sub>385</sub>, <a href=3D#b136>selbri-6</a><sub>136</sub>,=
 <a href=3D#b972>simple-tense-modal</a><sub>972</sub>, <a href=3D#b97>sumti=
-6</a><sub>97</sub>, <a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>NAhU	<dd><a href=3D#b374>mex-operator</a><sub>374</sub>
+<dt>NIhE	<dd><a href=3D#b385>operand-3</a><sub>385</sub>
+<dt>NIhO	<dd><a href=3D#b4>paragraphs</a><sub>4</sub>, <a href=3D#b2>text-=
1</a><sub>2</sub>
+<dt>NOI	<dd><a href=3D#b122>relative-clause</a><sub>122</sub>
+<dt>NU	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>NUhA	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>NUhI	<dd><a href=3D#b85>termset</a><sub>85</sub>
+<dt>NUhU	<dd><a href=3D#b85>termset</a><sub>85</sub>
+<dt>PA	<dd><a href=3D#b817>lerfu-string</a><sub>817</sub>, <a href=3D#b812=
>number</a><sub>812</sub>
+<dt>PEhE	<dd><a href=3D#b81>terms-1</a><sub>81</sub>
+<dt>PEhO	<dd><a href=3D#b312>mex-2</a><sub>312</sub>
+<dt>PU	<dd><a href=3D#b1033>time-offset</a><sub>1033</sub>, <a href=3D#b10=
30>time</a><sub>1030</sub>
+<dt>RAhO	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>ROI	<dd><a href=3D#b1051>interval-property</a><sub>1051</sub>
+<dt>SA	<dd><a href=3D#b1101>null</a><sub>1101</sub>
+<dt>SE	<dd><a href=3D#b802>ek</a><sub>802</sub>, <a href=3D#b807>gek</a><s=
ub>807</sub>, <a href=3D#b818>gihek</a><sub>818</sub>, <a href=3D#b808>guhe=
k</a><sub>808</sub>, <a href=3D#b932>interval</a><sub>932</sub>, <a href=3D=
#b805>jek</a><sub>805</sub>, <a href=3D#b806>joik</a><sub>806</sub>, <a hre=
f=3D#b374>mex-operator</a><sub>374</sub>, <a href=3D#b972>simple-tense-moda=
l</a><sub>972</sub>, <a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>SEI	<dd><a href=3D#b32>free</a><sub>32</sub>
+<dt>SEhU	<dd><a href=3D#b32>free</a><sub>32</sub>
+<dt>SI	<dd><a href=3D#b1101>null</a><sub>1101</sub>
+<dt>SOI	<dd><a href=3D#b32>free</a><sub>32</sub>
+<dt>SU	<dd><a href=3D#b1101>null</a><sub>1101</sub>
+<dt>TAhE	<dd><a href=3D#b1051>interval-property</a><sub>1051</sub>
+<dt>TEI	<dd><a href=3D#b987>lerfu-word</a><sub>987</sub>
+<dt>TEhU	<dd><a href=3D#b374>mex-operator</a><sub>374</sub>, <a href=3D#b3=
85>operand-3</a><sub>385</sub>
+<dt>TO	<dd><a href=3D#b32>free</a><sub>32</sub>
+<dt>TOI	<dd><a href=3D#b32>free</a><sub>32</sub>
+<dt>TUhE	<dd><a href=3D#b14>statement-3</a><sub>14</sub>
+<dt>TUhU	<dd><a href=3D#b14>statement-3</a><sub>14</sub>
+<dt>UI	<dd><a href=3D#b413>indicator</a><sub>413</sub>
+<dt>VA	<dd><a href=3D#b1045>space-offset</a><sub>1045</sub>, <a href=3D#b1=
040>space</a><sub>1040</sub>
+<dt>VAU	<dd><a href=3D#b20>fragment</a><sub>20</sub>, <a href=3D#b71>tail-=
terms</a><sub>71</sub>
+<dt>VEI	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b300>quantifi=
er</a><sub>300</sub>
+<dt>VEhA	<dd><a href=3D#b1046>space-interval</a><sub>1046</sub>
+<dt>VEhO	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b300>quantif=
ier</a><sub>300</sub>
+<dt>VIhA	<dd><a href=3D#b1046>space-interval</a><sub>1046</sub>
+<dt>VUhO	<dd><a href=3D#b90>sumti</a><sub>90</sub>
+<dt>VUhU	<dd><a href=3D#b374>mex-operator</a><sub>374</sub>
+<dt>XI	<dd><a href=3D#b32>free</a><sub>32</sub>
+<dt>Y	<dd><a href=3D#b413>indicator</a><sub>413</sub>
+<dt>ZAhO	<dd><a href=3D#b1051>interval-property</a><sub>1051</sub>
+<dt>ZEI	<dd><a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>ZEhA	<dd><a href=3D#b1030>time</a><sub>1030</sub>
+<dt>ZI	<dd><a href=3D#b1033>time-offset</a><sub>1033</sub>, <a href=3D#b10=
30>time</a><sub>1030</sub>
+<dt>ZIhE	<dd><a href=3D#b121>relative-clauses</a><sub>121</sub>
+<dt>ZO	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>ZOI	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>ZOhU	<dd><a href=3D#b30>prenex</a><sub>30</sub>
+<dt>any-word	<dd><a href=3D#b987>lerfu-word</a><sub>987</sub>, <a href=3D#=
b1101>null</a><sub>1101</sub>, <a href=3D#b97>sumti-6</a><sub>97</sub>, <a =
href=3D#b152>tanru-unit-2</a><sub>152</sub>, <a href=3D#b1100>word</a><sub>=
1100</sub>
+<dt>anything	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>bridi-tail	<dd><a href=3D#b50>bridi-tail</a><sub>50</sub>, <a href=3D#=
b40>sentence</a><sub>40</sub>
+<dt>bridi-tail-1	<dd><a href=3D#b50>bridi-tail</a><sub>50</sub>
+<dt>bridi-tail-2	<dd><a href=3D#b51>bridi-tail-1</a><sub>51</sub>, <a href=
=3D#b52>bridi-tail-2</a><sub>52</sub>
+<dt>bridi-tail-3	<dd><a href=3D#b52>bridi-tail-2</a><sub>52</sub>
+<dt>ek	<dd><a href=3D#b20>fragment</a><sub>20</sub>, <a href=3D#b421>joik-=
ek</a><sub>421</sub>, <a href=3D#b383>operand-2</a><sub>383</sub>, <a href=
=3D#b381>operand</a><sub>381</sub>, <a href=3D#b91>sumti-1</a><sub>91</sub>=
, <a href=3D#b93>sumti-3</a><sub>93</sub>
+<dt>fragment	<dd><a href=3D#b10>paragraph</a><sub>10</sub>
+<dt>free	<dd><a href=3D#b0>text</a><sub>0</sub>
+<dt>gek	<dd><a href=3D#b54>gek-sentence</a><sub>54</sub>, <a href=3D#b385>=
operand-3</a><sub>385</sub>, <a href=3D#b94>sumti-4</a><sub>94</sub>, <a hr=
ef=3D#b85>termset</a><sub>85</sub>
+<dt>gek-sentence	<dd><a href=3D#b53>bridi-tail-3</a><sub>53</sub>, <a href=
=3D#b54>gek-sentence</a><sub>54</sub>
+<dt>gihek	<dd><a href=3D#b51>bridi-tail-1</a><sub>51</sub>, <a href=3D#b52=
>bridi-tail-2</a><sub>52</sub>, <a href=3D#b50>bridi-tail</a><sub>50</sub>,=
 <a href=3D#b20>fragment</a><sub>20</sub>
+<dt>gik	<dd><a href=3D#b54>gek-sentence</a><sub>54</sub>, <a href=3D#b807>=
gek</a><sub>807</sub>, <a href=3D#b385>operand-3</a><sub>385</sub>, <a href=
=3D#b371>operator-1</a><sub>371</sub>, <a href=3D#b136>selbri-6</a><sub>136=
</sub>, <a href=3D#b94>sumti-4</a><sub>94</sub>, <a href=3D#b85>termset</a>=
<sub>85</sub>
+<dt>guhek	<dd><a href=3D#b371>operator-1</a><sub>371</sub>, <a href=3D#b13=
6>selbri-6</a><sub>136</sub>
+<dt>indicator	<dd><a href=3D#b411>indicators</a><sub>411</sub>
+<dt>indicators	<dd><a href=3D#b0>text</a><sub>0</sub>, <a href=3D#b1100>wo=
rd</a><sub>1100</sub>
+<dt>interval	<dd><a href=3D#b806>joik</a><sub>806</sub>
+<dt>interval-property	<dd><a href=3D#b1049>space-int-props</a><sub>1049</s=
ub>, <a href=3D#b1030>time</a><sub>1030</sub>
+<dt>jek	<dd><a href=3D#b422>joik-jek</a><sub>422</sub>, <a href=3D#b371>op=
erator-1</a><sub>371</sub>, <a href=3D#b135>selbri-5</a><sub>135</sub>, <a =
href=3D#b971>stag</a><sub>971</sub>, <a href=3D#b13>statement-2</a><sub>13<=
/sub>, <a href=3D#b2>text-1</a><sub>2</sub>
+<dt>joik	<dd><a href=3D#b807>gek</a><sub>807</sub>, <a href=3D#b421>joik-e=
k</a><sub>421</sub>, <a href=3D#b422>joik-jek</a><sub>422</sub>, <a href=3D=
#b383>operand-2</a><sub>383</sub>, <a href=3D#b381>operand</a><sub>381</sub=
>, <a href=3D#b371>operator-1</a><sub>371</sub>, <a href=3D#b370>operator</=
a><sub>370</sub>, <a href=3D#b134>selbri-4</a><sub>134</sub>, <a href=3D#b1=
35>selbri-5</a><sub>135</sub>, <a href=3D#b971>stag</a><sub>971</sub>, <a h=
ref=3D#b13>statement-2</a><sub>13</sub>, <a href=3D#b91>sumti-1</a><sub>91<=
/sub>, <a href=3D#b93>sumti-3</a><sub>93</sub>, <a href=3D#b2>text-1</a><su=
b>2</sub>
+<dt>joik-ek	<dd><a href=3D#b382>operand-1</a><sub>382</sub>, <a href=3D#b9=
2>sumti-2</a><sub>92</sub>
+<dt>joik-jek	<dd><a href=3D#b370>operator</a><sub>370</sub>, <a href=3D#b1=
34>selbri-4</a><sub>134</sub>, <a href=3D#b12>statement-1</a><sub>12</sub>,=
 <a href=3D#b491>tag</a><sub>491</sub>, <a href=3D#b152>tanru-unit-2</a><su=
b>152</sub>, <a href=3D#b81>terms-1</a><sub>81</sub>, <a href=3D#b0>text</a=
><sub>0</sub>
+<dt>lerfu-string	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b987=
>lerfu-word</a><sub>987</sub>, <a href=3D#b385>operand-3</a><sub>385</sub>,=
 <a href=3D#b97>sumti-6</a><sub>97</sub>, <a href=3D#b152>tanru-unit-2</a><=
sub>152</sub>
+<dt>lerfu-word	<dd><a href=3D#b817>lerfu-string</a><sub>817</sub>, <a href=
=3D#b987>lerfu-word</a><sub>987</sub>, <a href=3D#b812>number</a><sub>812</=
sub>
+<dt>linkargs	<dd><a href=3D#b20>fragment</a><sub>20</sub>, <a href=3D#b151=
>tanru-unit-1</a><sub>151</sub>
+<dt>links	<dd><a href=3D#b20>fragment</a><sub>20</sub>, <a href=3D#b160>li=
nkargs</a><sub>160</sub>, <a href=3D#b161>links</a><sub>161</sub>
+<dt>mex	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b374>mex-oper=
ator</a><sub>374</sub>, <a href=3D#b300>quantifier</a><sub>300</sub>, <a hr=
ef=3D#b97>sumti-6</a><sub>97</sub>
+<dt>mex-1	<dd><a href=3D#b311>mex-1</a><sub>311</sub>, <a href=3D#b310>mex=
</a><sub>310</sub>
+<dt>mex-2	<dd><a href=3D#b311>mex-1</a><sub>311</sub>, <a href=3D#b312>mex=
-2</a><sub>312</sub>, <a href=3D#b385>operand-3</a><sub>385</sub>
+<dt>mex-operator	<dd><a href=3D#b374>mex-operator</a><sub>374</sub>, <a hr=
ef=3D#b372>operator-2</a><sub>372</sub>, <a href=3D#b152>tanru-unit-2</a><s=
ub>152</sub>
+<dt>number	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b1051>inte=
rval-property</a><sub>1051</sub>, <a href=3D#b300>quantifier</a><sub>300</s=
ub>, <a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>operand	<dd><a href=3D#b312>mex-2</a><sub>312</sub>, <a href=3D#b385>o=
perand-3</a><sub>385</sub>, <a href=3D#b381>operand</a><sub>381</sub>, <a h=
ref=3D#b332>rp-operand</a><sub>332</sub>
+<dt>operand-1	<dd><a href=3D#b381>operand</a><sub>381</sub>
+<dt>operand-2	<dd><a href=3D#b382>operand-1</a><sub>382</sub>, <a href=3D#=
b383>operand-2</a><sub>383</sub>
+<dt>operand-3	<dd><a href=3D#b383>operand-2</a><sub>383</sub>, <a href=3D#=
b385>operand-3</a><sub>385</sub>
+<dt>operator	<dd><a href=3D#b311>mex-1</a><sub>311</sub>, <a href=3D#b312>=
mex-2</a><sub>312</sub>, <a href=3D#b310>mex</a><sub>310</sub>, <a href=3D#=
b372>operator-2</a><sub>372</sub>, <a href=3D#b370>operator</a><sub>370</su=
b>, <a href=3D#b330>rp-expression</a><sub>330</sub>
+<dt>operator-1	<dd><a href=3D#b371>operator-1</a><sub>371</sub>, <a href=
=3D#b370>operator</a><sub>370</sub>
+<dt>operator-2	<dd><a href=3D#b371>operator-1</a><sub>371</sub>
+<dt>paragraph	<dd><a href=3D#b4>paragraphs</a><sub>4</sub>
+<dt>paragraphs	<dd><a href=3D#b4>paragraphs</a><sub>4</sub>, <a href=3D#b2=
>text-1</a><sub>2</sub>
+<dt>prenex	<dd><a href=3D#b20>fragment</a><sub>20</sub>, <a href=3D#b11>st=
atement</a><sub>11</sub>, <a href=3D#b41>subsentence</a><sub>41</sub>
+<dt>quantifier	<dd><a href=3D#b20>fragment</a><sub>20</sub>, <a href=3D#b3=
85>operand-3</a><sub>385</sub>, <a href=3D#b95>sumti-5</a><sub>95</sub>, <a=
 href=3D#b112>sumti-tail-1</a><sub>112</sub>
+<dt>relative-clause	<dd><a href=3D#b121>relative-clauses</a><sub>121</sub>
+<dt>relative-clauses	<dd><a href=3D#b20>fragment</a><sub>20</sub>, <a href=
=3D#b32>free</a><sub>32</sub>, <a href=3D#b95>sumti-5</a><sub>95</sub>, <a =
href=3D#b97>sumti-6</a><sub>97</sub>, <a href=3D#b112>sumti-tail-1</a><sub>=
112</sub>, <a href=3D#b111>sumti-tail</a><sub>111</sub>, <a href=3D#b90>sum=
ti</a><sub>90</sub>
+<dt>rp-expression	<dd><a href=3D#b310>mex</a><sub>310</sub>, <a href=3D#b3=
32>rp-operand</a><sub>332</sub>
+<dt>rp-operand	<dd><a href=3D#b330>rp-expression</a><sub>330</sub>
+<dt>selbri	<dd><a href=3D#b53>bridi-tail-3</a><sub>53</sub>, <a href=3D#b3=
2>free</a><sub>32</sub>, <a href=3D#b374>mex-operator</a><sub>374</sub>, <a=
 href=3D#b385>operand-3</a><sub>385</sub>, <a href=3D#b131>selbri-1</a><sub=
>131</sub>, <a href=3D#b136>selbri-6</a><sub>136</sub>, <a href=3D#b95>sumt=
i-5</a><sub>95</sub>, <a href=3D#b112>sumti-tail-1</a><sub>112</sub>, <a hr=
ef=3D#b815>tense-modal</a><sub>815</sub>
+<dt>selbri-1	<dd><a href=3D#b130>selbri</a><sub>130</sub>
+<dt>selbri-2	<dd><a href=3D#b131>selbri-1</a><sub>131</sub>, <a href=3D#b1=
32>selbri-2</a><sub>132</sub>
+<dt>selbri-3	<dd><a href=3D#b132>selbri-2</a><sub>132</sub>, <a href=3D#b1=
34>selbri-4</a><sub>134</sub>, <a href=3D#b152>tanru-unit-2</a><sub>152</su=
b>
+<dt>selbri-4	<dd><a href=3D#b133>selbri-3</a><sub>133</sub>
+<dt>selbri-5	<dd><a href=3D#b134>selbri-4</a><sub>134</sub>, <a href=3D#b1=
35>selbri-5</a><sub>135</sub>
+<dt>selbri-6	<dd><a href=3D#b135>selbri-5</a><sub>135</sub>, <a href=3D#b1=
36>selbri-6</a><sub>136</sub>
+<dt>sentence	<dd><a href=3D#b14>statement-3</a><sub>14</sub>, <a href=3D#b=
41>subsentence</a><sub>41</sub>
+<dt>simple-tense-modal	<dd><a href=3D#b971>stag</a><sub>971</sub>, <a href=
=3D#b815>tense-modal</a><sub>815</sub>
+<dt>space	<dd><a href=3D#b972>simple-tense-modal</a><sub>972</sub>
+<dt>space-int-props	<dd><a href=3D#b1046>space-interval</a><sub>1046</sub>
+<dt>space-interval	<dd><a href=3D#b1040>space</a><sub>1040</sub>
+<dt>space-offset	<dd><a href=3D#b1040>space</a><sub>1040</sub>
+<dt>stag	<dd><a href=3D#b52>bridi-tail-2</a><sub>52</sub>, <a href=3D#b50>=
bridi-tail</a><sub>50</sub>, <a href=3D#b807>gek</a><sub>807</sub>, <a href=
=3D#b383>operand-2</a><sub>383</sub>, <a href=3D#b381>operand</a><sub>381</=
sub>, <a href=3D#b371>operator-1</a><sub>371</sub>, <a href=3D#b370>operato=
r</a><sub>370</sub>, <a href=3D#b134>selbri-4</a><sub>134</sub>, <a href=3D=
#b135>selbri-5</a><sub>135</sub>, <a href=3D#b13>statement-2</a><sub>13</su=
b>, <a href=3D#b91>sumti-1</a><sub>91</sub>, <a href=3D#b93>sumti-3</a><sub=
>93</sub>, <a href=3D#b2>text-1</a><sub>2</sub>
+<dt>statement	<dd><a href=3D#b10>paragraph</a><sub>10</sub>, <a href=3D#b1=
1>statement</a><sub>11</sub>
+<dt>statement-1	<dd><a href=3D#b11>statement</a><sub>11</sub>
+<dt>statement-2	<dd><a href=3D#b12>statement-1</a><sub>12</sub>, <a href=
=3D#b13>statement-2</a><sub>13</sub>
+<dt>statement-3	<dd><a href=3D#b13>statement-2</a><sub>13</sub>
+<dt>subsentence	<dd><a href=3D#b54>gek-sentence</a><sub>54</sub>, <a href=
=3D#b122>relative-clause</a><sub>122</sub>, <a href=3D#b41>subsentence</a><=
sub>41</sub>, <a href=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>sumti	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b385>operan=
d-3</a><sub>385</sub>, <a href=3D#b91>sumti-1</a><sub>91</sub>, <a href=3D#=
b94>sumti-4</a><sub>94</sub>, <a href=3D#b97>sumti-6</a><sub>97</sub>, <a h=
ref=3D#b112>sumti-tail-1</a><sub>112</sub>, <a href=3D#b152>tanru-unit-2</a=
><sub>152</sub>, <a href=3D#b83>term</a><sub>83</sub>
+<dt>sumti-1	<dd><a href=3D#b90>sumti</a><sub>90</sub>
+<dt>sumti-2	<dd><a href=3D#b91>sumti-1</a><sub>91</sub>
+<dt>sumti-3	<dd><a href=3D#b92>sumti-2</a><sub>92</sub>, <a href=3D#b93>su=
mti-3</a><sub>93</sub>
+<dt>sumti-4	<dd><a href=3D#b93>sumti-3</a><sub>93</sub>, <a href=3D#b94>su=
mti-4</a><sub>94</sub>
+<dt>sumti-5	<dd><a href=3D#b94>sumti-4</a><sub>94</sub>
+<dt>sumti-6	<dd><a href=3D#b95>sumti-5</a><sub>95</sub>, <a href=3D#b111>s=
umti-tail</a><sub>111</sub>
+<dt>sumti-tail	<dd><a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>sumti-tail-1	<dd><a href=3D#b111>sumti-tail</a><sub>111</sub>
+<dt>tag	<dd><a href=3D#b54>gek-sentence</a><sub>54</sub>, <a href=3D#b130>=
selbri</a><sub>130</sub>, <a href=3D#b14>statement-3</a><sub>14</sub>, <a h=
ref=3D#b152>tanru-unit-2</a><sub>152</sub>, <a href=3D#b83>term</a><sub>83<=
/sub>
+<dt>tail-terms	<dd><a href=3D#b51>bridi-tail-1</a><sub>51</sub>, <a href=
=3D#b52>bridi-tail-2</a><sub>52</sub>, <a href=3D#b53>bridi-tail-3</a><sub>=
53</sub>, <a href=3D#b50>bridi-tail</a><sub>50</sub>, <a href=3D#b54>gek-se=
ntence</a><sub>54</sub>
+<dt>tanru-unit	<dd><a href=3D#b136>selbri-6</a><sub>136</sub>
+<dt>tanru-unit-1	<dd><a href=3D#b150>tanru-unit</a><sub>150</sub>
+<dt>tanru-unit-2	<dd><a href=3D#b151>tanru-unit-1</a><sub>151</sub>, <a hr=
ef=3D#b152>tanru-unit-2</a><sub>152</sub>
+<dt>tense-modal	<dd><a href=3D#b491>tag</a><sub>491</sub>
+<dt>term	<dd><a href=3D#b160>linkargs</a><sub>160</sub>, <a href=3D#b161>l=
inks</a><sub>161</sub>, <a href=3D#b122>relative-clause</a><sub>122</sub>, =
<a href=3D#b82>terms-2</a><sub>82</sub>
+<dt>terms	<dd><a href=3D#b20>fragment</a><sub>20</sub>, <a href=3D#b32>fre=
e</a><sub>32</sub>, <a href=3D#b30>prenex</a><sub>30</sub>, <a href=3D#b40>=
sentence</a><sub>40</sub>, <a href=3D#b71>tail-terms</a><sub>71</sub>, <a h=
ref=3D#b85>termset</a><sub>85</sub>
+<dt>terms-1	<dd><a href=3D#b80>terms</a><sub>80</sub>
+<dt>terms-2	<dd><a href=3D#b81>terms-1</a><sub>81</sub>
+<dt>termset	<dd><a href=3D#b83>term</a><sub>83</sub>
+<dt>text	<dd><a href=3D#b32>free</a><sub>32</sub>, <a href=3D#b1101>null</=
a><sub>1101</sub>, <a href=3D#b97>sumti-6</a><sub>97</sub>
+<dt>text-1	<dd><a href=3D#b14>statement-3</a><sub>14</sub>, <a href=3D#b0>=
text</a><sub>0</sub>
+<dt>time	<dd><a href=3D#b972>simple-tense-modal</a><sub>972</sub>
+<dt>time-offset	<dd><a href=3D#b1030>time</a><sub>1030</sub>
+<dt>utterance	<dd><a href=3D#b1101>null</a><sub>1101</sub>
+<dt>vocative	<dd><a href=3D#b32>free</a><sub>32</sub>
+
+Index
+
+  INDEX \e ", " \c "2"  \r =20
+' symbol: and consonant cluster determination in lujvo, 56; definition (se=
e also apostrophe), 31
+&: word for, 416
+(n + 1)(n + 1) =3D n^2 + 2n + 1: example, 437
+(n+1)-th rat: example, 448
+.: word for, 416
+.1010_2 ( 2^{110_2}: example, 451
+: , word for, 416
++1 + -1 =3D 0: example, 436
+-1: example, 433
+1 + 1 =3D 2: example, 435
+10^20: example, 451
+123: example, 432
+12-point: example, 418
+2 + 2: example, 262
+2 rats + 2 rabbits =3D 4 animals: example, 456
+2/7: example, 433
+3 ( 10^8: example, 451
+3 grams: example, 435
+3.1415: example, 433
+4-letter rafsi: definition, 57
+5-letter rafsi: definition, 57
+8 out of ten: example, 448
+a: example, 322
+A gives B to C: example, 421
+A gives BC: example, 421
+a is letteral: example, 422
+A loves B: example, 421
+A selma=92o, 336, 340, 341, 352, 354, 361, 364, 453, 489
+a/an: contrasted with the, 322
+a'a, 302
+abbreviated lujvo and plausibility, 284
+ABC base 16: example, 444
+abduction: example, 316
+Abraham Lincoln: example, 353
+absolute laws, 273
+abstract description, 266
+abstract lujvo, 286; contrasted with abstract bridi, 287
+abstraction bridi: contrasted with component non-abstraction bridi in mean=
ing, 98; effect on claim of bridi, 198
+abstraction contours: compared with contour tenses, 268
+abstraction conversion, 266
+abstraction lujvo: asymmetric, 288
+abstraction(s): achievement, 258; activity, 258; amount, 261; amount contr=
asted with property, 261; concept, 265; connection, 269; creating new types=
, 266; event, 256; experience, 265; forethought connection in, 365; grammat=
ical uses, 255; grouping of connectives in, 365; idea, 265; implicit in sum=
ti, 257; logical connection of, 365; making concrete, 267; mental activity,=
 262; place structure, 255; point-event, 258; predication/sentence, 262; pr=
ocess, 258; property, 259; sentence, contrasted with quotation, 263; simpli=
fication to sumti, 266; simplification to sumti with jai, 267; simplificati=
on to sumti with tu'a, 266; speaking, writing, etc., 263; state, 258; sumti=
 ellipsis in, 256; syntax, 255; table, 269; truth-value, 262; truth-value  =
contrasted with amount, 262; truth-value and fuzzy logic, 262; types, 265; =
vague, 265; with knowing, believing, etc., 262; with wonder, doubt, etc., 2=
64
+abstractors, 502
+accent mark: a diacritical mark, 418; example, 419
+accent marks: proposed lerfu words for, 429
+accented letters: considered as distinct from unaccented, 419
+Acer: example, 63
+achievative event contour, 228
+achievement abstraction: place structure, 259
+achievement abstraction(s): definition, 258; related tense contours, 269
+achievement event: described, 258
+acronym: definition, 423
+acronym names from lerfu words: assigning final consonant, 424
+acronyms: as lerfu strings using =93me=94, 424; using names based on lerfu=
 words, 423
+acronyms names based on lerfu words: omitting bu, 424; using =93z=94 inste=
ad of =93 ' =94in, 424
+activity abstraction: place structure, 259
+activity abstraction(s): definition, 258; related tense contours, 268
+activity abstractor, 258
+activity event: described, 258
+actual events: explicitly expressing, 243
+actual stop: contrasted with natural end, 229
+actuality: expressing in past/future, 244; Lojban contrasted with English =
in implying, 243; marking, 491
+addition: a mathematical operator, 436
+addition operator: contrasted with positive sign, 436
+addition problems: example, 98
+adjective ordering, 89
+adjective-noun combination: with tanru, 84
+adjectives: brivla as Lojban equivalents, 52
+adverbs: brivla as Lojban equivalents, 52
+adverb-verb combination: with tanru, 84
+a'enai, 302
+affirmative answer: quick-tour version, 23
+afraid of horse: example, 179
+after sleep: example, 303
+afterthought bridi connectives: contrasted with forethought bridi connecti=
ves, 338
+afterthought connection: contrasted with forethought for grammatical utter=
ances, 352; definition, 199; of operands, 453; of operators, 453
+afterthought connective(s): as complete grammatical utterance, 352; contra=
sted with forethought connective, 338
+afterthought sentence connection: modal contrasted with tense, 248
+afterthought tense connection: contrasted with forethought in likeness to =
modal connection, 249
+a'i, 302
+ai, 302
+ailment, 282
+Albania: example, 480
+aleph null: example, 434
+Alexander Pavlovitch Kuznetsov: example, 421
+algebra of functions: operator and operand distinction in, 460
+alienable possession: definition, 173
+aliens: communication with, 329
+all persons: example, 398
+allowable diphthongs: in fu'ivla contrasted with in gismu/lujvo, 63
+all-th: example, 447
+alpha: example, 416
+alphabet: Latin used for Lojban, 413; Lojban, 29; words for letters in, ra=
tionale, 413; words for non-Lojban letters, rationale, 413
+alphabetic order, 29
+alternative guidelines, 273
+always and everywhere: example, 231
+ambiguity of tanru, 85
+American dollars: example, 425
+American Indian languages and evidentials, 315
+Amharic writing, 420
+amount abstraction, 261
+amount abstraction and mathematics, 261
+amount abstraction(s): place structure, 262; scale, 262; specifying determ=
ining place with ce'u, 261
+ampersand: example, 416
+ampersand character: word for, 416
+Amsterdam: example, 38
+an: example, 322
+anaphora: definition, 152; pro-bridi go'i-series as, 152; pro-sumti ri-ser=
ies as, 152; pro-sumti vo'a-series as, 158
+anaphoric pro-bridi: stability of, 162
+anaphoric pro-sumti: stability of, 162
+and: as non-logical connective, 353; compared with but, 353; contrasted wi=
th cross-product, 357
+and earlier: example, 364
+and simultaneously: example, 364
+and then: example, 240, 364
+animal doctor: example, 282
+animal patient, 282
+animals: use of fu'ivla for specific, 61
+anomalous ordering of lujvo places, 283
+answers, 469; go'i for yes/no questions, 154; to operator questions, 457; =
to place structure questions, 191; to tense-or-modal questions, 250
+antecedent: for pro-bridi, 151; for pro-bridi as full bridi, 151
+antecedent of pro-bridi: definition, 145
+antecedent of pro-sumti: definition, 145
+anticipated: example, 316
+any: as a restricted universal claim, 399; as a translation problem, 399; =
as a universal claim, later restricted, 400; as an existential claim, 400; =
expressing as existential by variable in subordinate bridi, 401
+any box, 400
+anyone: contrasted with everyone in assumption of existence, 399
+anyone who goes: walks, example, 399
+a'o, 297, 302
+aorist: definition, 223
+apostrophe: and consonant cluster determination in lujvo, 56; as not a con=
sonant for morphological discussions, 49; as preferable over comma in names=
, 33; definition of, 31; example of, 33; purpose of, 31; quick-tour version=
, 12; type of letter in word-formation, 31; use in vowel pairs, 34; variant=
 of, 31
+Appassionata: example, 202
+appropriate standard, 480
+approximate numbers: expressing, 442; expressing some exactness of, 443
+approximately 40: example, 443
+Arabian Nights: as multi-level narrative, 467
+Arabic alphabet: language shift word for, 417
+argument tags: based on tenses (see also sumti tcita), 231
+Armstrong: example, 40
+Arnold: example, 65
+arthropod, 280
+article, 498; number, 435
+articles: cmavo as Lojban equivalents, 50
+Artur Rubenstein: example, 202
+ASCII: application to lerfu words, 425
+aspect: expressing, 228; natural languages compared with respect to, 228
+assignable pro-sumti: explicit cancellation of by rebinding, 162; stabilit=
y of, 162
+assumption: example, 317
+asymmetric lujvo, 287
+asymmetrical tanru, 104; definition, 104
+asymmetrical tanru types: activity + implement-used, 110; cause + effect, =
107; characteristic/detail + object, 106; characteristic-time + event, 110;=
 effect + causative agent, 107; elements-in-set + set, 105; energy-source +=
 powered, 110; general-class + sub-class, 106; inhabitant + habitat, 106; l=
ocus-of-application + object, 110; miscellaneous, 111; object + component/d=
etail, 106; object + place-sold, 110; object + usual-container, 110; object=
-giving-characteristic + other-object, 109; object-measured + standard-obje=
ct, 108; object-of-action + action, 105; object-of-purpose-of-instrument + =
instrument, 107; overriding-property + object-with-implicit-properties, 108=
; possessor + object, 106; product + producer, 109; product + source, 108; =
purpose-of-instrument + instrument, 107; set + element-of-set, 105; similar=
-appearance-object + object, 109; source + product, 107; source-material + =
object, 108; typical-place + object, 109; undesired-object + protection-obj=
ect, 110; whole + part, 109
+at least: contrasted with more than, less than, at most, 443; example, 443
+at most: contrasted with more than, at least, less than, 443; example, 443
+Athens, 258
+attend school: example, 226
+attitude, 297; avoidance of expression, 322; scalar, 305
+attitudes: beginning, 314; ceasing, 314; continuing, 314; empathy contrast=
ed with sympathy, 314; expressing changes in, 314
+attitudinal: example of scale effect, 305; scope, 474; signaling as non-pr=
opositional, 304
+attitudinal answers: plausibility, 313
+attitudinal categories, 306; example of effect, 307; mnemonic for, 307; ra=
tionale, 306
+attitudinal indicator: unspecified, 311
+attitudinal indicators, 297; conventions of interpretation, 311; placement=
 of =93nai=94 in, 311; placement of scale in, 311; quick-tour version, 24
+attitudinal modifiers, 308
+attitudinal phrase, 299
+attitudinal questions, 313; asking about specific attitude, 313; asking in=
tensity, 313
+attitudinal scale: as axis in emotion-space, 306; neutral compared with po=
sitive + negative, 306; seven-position, 305; stand-alone usage, 305; usage,=
 305
+attitudinal scales: rationale for assignment, 304
+attitudinals: a- series, 301; affecting whole grammatical structures, 312;=
 ambivalent emotion words, 300; and logic, 392; at beginning of text, 312; =
attributing emotion to others, 314; benefit in written expression, 312; cat=
egories with nai, 307; categories with scale markers, 307; comparison of me=
aning based on position, 301; complex emotion words, 300; complexity, 310; =
compound, 298; contours, 314; contrasted with bridi, 303, 306; contrasted w=
ith discursives, 317; contrasted with rationalizations of emotion, 306; des=
ign benefit, 306; difficult emotion words, 300; e- series, 301; effect of c=
u'i, 299; effect of nai, 299; emotional contrasted with propositional, 301;=
 emotional/propositional caveat, 302; exceptions, 314; external grammar, 31=
2; grammar of internal compounding, 311; grammar of placement in bridi, 312=
; i- series, 298, 301; internal grammar, complete, 311; logical language an=
d, 302; negative, 304; neutral, 304; non-speaker attitudes, 314; o- series,=
 298; order of, 306; placement for prevailing attitude, 297; placement in s=
entences with =93nai=94, 311; positive, 304; prevailing attitude, 297; prop=
ositional contrasted with emotional, 301; propositional effect on claim, 30=
1; propositional indicators, 301; propositional/emotional caveat, 302; pure=
 emotion, 298; rationale for, 303; referent uncertainty, 312; scale of, 304=
; stand-alone categories, 307; u- series, 298; word-form for primary, 297
+attitudinals and claims, 298
+attitudinals and irony, 299
+attitudinals and truth value, 298
+attitudinals for emotional reaction, 299
+a'ucu'i, 302
+au, 302
+audio-visual isomorphism, 29
+audio-visually isomorphic, 465
+auditoriums, 281
+author of this book, 5
+Avon: example, 196
+ba, 219, 363
+ba'a, 316
+ba'acu'i, 316
+ba'anai, 316
+back-counting pro-sumti, 139
+background noise, 278
+ba'e, 416, 479; interaction with bu, 416
+BAhE selma=92o, 479, 480, 489
+bai, 201
+bai ke, 202
+BAI modal tags: rationale for, 195
+BAI selma=92o, 195, 205, 224, 489; as short forms for fi'o constructs, 195=
; effect of conversion on, 196; form of cmavo in, 197; rationale for select=
ion, 196
+ba'o, 228; as futureward of event, 229; derivation of word, 228; explanati=
on of derivation, 229
+base: assumed, 444; changing permanently, 444; non-constant, 444; specifyi=
ng, 444; vague, 445
+base greater than 16: compound single-digits contrasted with two digits, 4=
45; expressing numbers in, 445; two digits contrasted with compound single-=
digits, 445
+base point: in bases other than 10, 444
+base varying for each digit: separator for, 445
+base-20 arithmetic: remnants of, 461
+basis: example, 317
+ba'u, 318
+be, 93, 197, 262
+BE selma=92o, 93, 489; terminator for, 490
+be'a, 231
+beach: example, 259
+beans: example, 176
+bear wrote story: example, 121
+Bears wrote book: example, 124
+beautiful dog: example, 20
+because: English word, four varieties of, 198
+beckon: example, 291
+be'e, 324
+beefsteak, 285; example, 285
+Beethoven: example, 202
+beetle: example, 280
+beetles, 280
+begin: contrasted with resume, 229
+beginning point: spatial, 231
+BEhO selma=92o, 93, 490
+bei, 93
+BEI selma=92o, 93, 489
+being alive: example, 258
+be'o, 93; effect of ku on elidability of, 94; effect of relative clauses o=
n elidability of, 94; elidability of, 94
+better: example, 293
+between Dresden and Frankfurt: example, 359, 360, 361
+be'u, 310
+beverage: example, 163
+bibliography, 6
+bicycle race: example, 266
+bi'e, 437; effect on following operator, 437
+big boat: example, 55
+big nose: example, 169
+big nose-pores: example, 170
+big person: example, 169
+big red dog: example, 89
+BIhE selma=92o, 437, 490
+BIhI selma=92o, 246, 359, 360, 361, 455, 490; grammar of, 360
+bi'i, 359, 362
+Bill Clinton: example, 275
+binary system: specifying numbers in (see also base), 444
+bi'o, 359
+bi'u, 322
+bi'unai, 322
+black cat: example, 155
+blue: as sad, example, 322
+blue and red: example, 354
+blue house: example, 193, 349
+blue-eyed: example, 284
+blueness: example, 261
+blueness varying: example, 261
+bo, 86, 198, 238, 240, 342, 343, 349, 361, 364, 459, 466; contrasted with =
ke for tensed logical connection, 364; contrasted with tu'e for tensed logi=
cal connection, 364; for right-grouping in tanru, 87; in jeks for operators=
, 361; in joiks for operators, 361; in logical connectives, 342; right-grou=
ping, 343
+bo and forethought connectives, 343
+BO selma=92o, 86, 133, 342, 343, 349, 361, 364, 454, 466, 490
+boat class: example, 73
+boat sailed: example, 233
+boi, 362, 421, 438, 449, 450, 458; effect on elidability of me'u, 449; eli=
ding from lerfu strings, 421; exception before MAI, 458; exception before M=
OI, 449; exception before ROI, 458; in Polish notation, 438; required betwe=
en pro-sumti lerfu string and quantifier, 421
+BOI selma=92o, 421, 490
+bold: example, 418
+bomb destroyed fifty miles: example, 360
+bone bread: example, 38
+books about Lojban, 6
+boring legalities, 8
+borrowing: four stages of, 61
+borrowing from other language: fu'ivla as, 53
+borrowings: fu'ivla form with categorizing rafsi, 61; fu'ivla form without=
 categorizing rafsi, 62; most common form for, 61; Stage 1, 61; Stage 2, 61=
; Stage 3, 61; Stage 3 contrasted with Stage 4 in ease of construction, 62;=
 Stage 4, 62; using foreign-language name, 61; using lojbanized name, 61
+Boston from Atlanta: example, 187
+both dogs: example, 442
+bound variable pro-sumti: stability of, 162
+bovine: example, 285
+bracketed remark, 481
+brackets: use in IPA notation, 29
+breathe: example, 363
+bridi: building from selbri and sumti, 187; compared with predication, 11;=
 concept of, 11; definition, 83, 187; definition, quick-tour version, 26; e=
ffect of alternate form on sumti order, 188; effect of using non-standard f=
orm, 188; exception to sumti place structure in, 188; leaving a sumti place=
 unspecified in with zo'e, 189; leaving end sumti places unspecified in, 18=
9; logical connection with negation, 337; logical connective for, 336; non-=
standard form, 188; omitting the first sumti place, 188; quick-tour version=
, 14; relation to selbri, 83; selbri-first as exceptional, 188; standard fo=
rm of, 188
+bridi connection: use of imperatives in, 353; use of truth questions in, 3=
53
+bridi logical connection: compared with sumti logical connections, 340
+bridi negation: and DeMorgan's Law, 408; and negation boundary, 408; compa=
red with negation between sentences, 404; multiple, 104; na before selbri c=
ompared to naku in prenex, 401; naku in prenex compared to na before selbri=
, 401; relative order with tense, 103; two forms of, 401
+bridi negation and logical connectives, 403
+bridi questions: quick-tour version, 23
+bridi-based comparison: contrasted with comparison with relative phrase, i=
n claims about parts, 204
+bridi-tail: definition, 344, 495
+bridi-tail logical connection: and DeMorgan's Law, 408
+bridi-tail modal connection, 200
+bridi-tails: eliding vau in, 345; forethought tense connection of, 240
+brie: example, 63
+brivla: as one of the 3 basic word classes, 50; consonant pairs in, 53; de=
finition, 52; definition, quick-tour version, 27; from tanru, 55; propertie=
s of, 53; recognition of, 53; relation to bridi, 11; stress on, 40; subtype=
s of, 53; types, 83; types of, quick-tour version, 20
+brivla as selbri, 83
+brivla equivalents, 97
+brivla form: contrasted with cmavo form, 53; contrasted with cmene form, 5=
3
+broda, 151
+broda-series for pro-bridi: compared with ko'a-series for pro-sumti, 151
+broda-series pro-bridi, 151; assigning with cei, 151; use as abstract patt=
ern, 151; use as sample gismu, 151; with no assignment, 151; word-form rati=
onale, 151
+Brooklyn: example, 93
+brothers: example, 355
+Brown: James Cooke, 6; James Cooke, and =93letteral=94, 413
+bu, 414; and compound cmavo, 416; effect of multiple, 416; effect on prece=
ding word, 414; for extension of lerfu word set, 416; grammar of, 416; inte=
ractions, 416, 417; omitting in acronyms names based on lerfu words, 424
+BU selma=92o, 414, 490
+bu'a, 164, 409
+bu'a-series pro-sumti: for bound variables, 161
+bubu, 416
+bu'e, 409
+buffer vowel, 38; and stress, 38; shortening of, 39
+bu'i, 409
+Bulgarian: example, 64
+bu'o, 314
+bu'onai, 314
+but: compared with and, 353; example, 318
+but/and equivalence, 25
+butter is soft: example, 124
+butterfly: social, example, 18
+bu'u, 219; compared with ca, 219
+BY selma=92o, 414, 418, 425, 491; terminator for, 490
+C string: as a symbol for a single consonant, 49
+C/C string: as a symbol for a permissible consonant pair, 50
+C/CC string: as a symbol for a consonant triple, 50
+ca, 219, 232; compared with bu'u, 219; meaning as a sumti tcita, 232; mean=
ing when following interval specification, 221; rational for, 219
+ca'a, 243
+ca'e, 316
+CAhA selma=92o, 243, 491; making sticky, 243; order in tense construct, 24=
3
+CAI selma=92o, 305, 491
+cancellation of indicators, 494
+cancellation of pro-sumti/pro-bridi assignment: with da'o, 162
+cancellation of sumti assignment, 492
+ca'o, 228; derivation of word, 228
+cai, 305
+calculator mathematics: as default in Lojban, 436
+can see: example, 244
+canceling letter shifts, 418
+capital letters: use in Lojban, 415; use of, 29
+capitalization: for unusual stress in names, 66; use in names, 66; use of,=
 66
+captions to pictures, 7
+car goer: example, 279
+cardinal selbri: definition, 446; place structure, 446; place structure ef=
fect from subjective numbers, 448
+cardinality: definition, 125; property of sets, 125
+carried piano: example, 353
+carry sack: example, 205
+carry sack and dog: example, 241
+carry the piano: example, 361
+Cartesian product: with tenses, 246
+Carthage destroyed: example, 317
+case: upper/lower specification, 415
+cat of plastic: example, 160
+Catherine: example, 65
+Cathy: example, 65
+causals: claiming the relation contrasted with claiming cause and/or effec=
t and/or relation, 198; gismu, 197; modal, 197
+cause death: example, 267, 287
+cave: example, 236
+CC string: as a symbol for a permissible initial consonant pair, 50
+CCVVCV fu'ivla: and rafsi fu'ivla proposal, 80
+ce, 354, 355
+ce'a, 418
+cedilla: a diacritical mark, 418; proposed lerfu word for, 429
+ce'e, 208, 347, 399
+CEhE selma=92o, 347, 399, 491
+ce'i, 433
+cei for broda-series assignment: compared with goi for ko'a-series assignm=
ent, 151
+CEI selma=92o, 491
+ce'o, 354, 362
+ce'u, 161, 260, 261; use in specifying sumti place of property in abstract=
ion, 161
+cei, 151, 154, 162; for broda-series pro-bridi assignment, 151
+cessitive event contour, 228
+chapter numbering, 458
+chapter titles: intent of, 4
+character codes: definition, 425
+character encoding schemes: application to lerfu words, 425
+characters: definition, 425; special, 31
+Chelsea Clinton, 275
+chemical elements: use of single-letter shift for, 415
+Chief: example, 66
+child on ice: example, 221
+Chilean desert: example, 80
+Chinese characters: contrasted with alphabets and syllabaries, 420; repres=
enting based on pinyin spelling, 420; representing based on strokes, 420
+choose from: example, 355
+Chrysler: example, 99
+ch-sound in English: representation in Lojban, 31
+ci'i, 434
+CIA: example, 424
+circumflex: a diacritical mark, 418
+ci'u, 204
+claims: contrasted with expression of feelings, 298
+clamshells, 286
+clarity of sounds, 31
+Classical Greek aorist tense: compared with Lojban tense, 223
+clause: subordinate using abstraction, 255
+close-binding, 490
+closed interval, 360; expressed with mi'i, 455
+closings: letter, 325
+cmavo: as one of the 3 basic word classes, 50; compound, 51; contrasted wi=
th rafsi in usage, 61; contrasted with same-form rafsi in meaning, 56; defi=
nition, 50; definition, quick-tour version, 27; diphthongs in, 51; experime=
ntal, 51; for experimental use, 51; interaction list, 485; lack of relation=
 of form to grammatical use, 51; rules for pause after Cy-form, 69; simple,=
 51; stress on, 40, 52; structure of, 51
+cmavo and gismu: major, 53
+cmavo as brivla, 495
+cmavo as selbri: quick-tour version, 20
+cmavo as sumti, 497
+cmavo form: contrasted with brivla form, 53
+cmavo without rafsi: method of including in lujvo, 60
+cmene: algorithm for, 66; alternatives for restricted sequences in, 66; an=
d analyzability of speech stream, 64; as one of the 3 basic word classes, 5=
0; authority for, 65; avoiding impermissible consonant clusters in, 67; con=
sonant clusters permitted in, 66; definition, 64; examples of, 64; final le=
tter in, 66; from Lojban words, 66; method of including in lujvo, 60; prosc=
ribed syllables in, 67; purpose of, 64; rationale for lojbanizing, 64; requ=
irement for pause after, 66; restrictions on form of, 65; rules for, 66; ru=
les for formation, 65; rules for pause before, 68; stress in, 65, 66; unusu=
al stress in, 65
+cmene form: contrasted with brivla form, 53
+co, 95
+CO selma=92o, 491
+cobra: example, 63
+co'e, 158, 164; as selbri place-holder, 158; rationale for word form, 158
+co'e-series pro-bridi, 157
+coffee mixed with tea: example, 359
+coffee or tea: example, 352
+co'i, 230
+coi, 136, 324; quick-tour version, 21
+COI selma=92o, 136, 146, 183, 323, 492; effect on pause before name, 323; =
effect on referent of do, 146; effect on referent of mi, 146; ordering mult=
iple with mi'e, 325; terminator for, 492
+coin heads: example, 447
+Coleoptera, 280
+color standards, 295
+colorimeter, 261
+combining words into one, 507
+comma: definition of, 32; effect on relative clause in English, 171; examp=
le of, 32; main use of, 32; optional, 32; quick-tour version, 12; variant o=
f, 32
+command: contrasted with observative form, 188
+commands: quick-tour version, 22; with ko, 146
+commas in numbers: as numerical punctuation, 433; effect of other notation=
 conventions, 433; with elided digits, 433
+common abstractor, 256
+commutative truth functions, 335
+comparative lujvo: against former state, 293; and seltau presupposition, 2=
93; potential ambiguity in, 292; standardized meanings, 292
+comparatives: use of zmadu in forming, 58
+comparison: claims related to based on form, 204
+comparison with relative phrase: contrasted with bridi-based comparison, i=
n claims about parts, 204
+completitive event contour, 228
+complex logical connection: grouping strategies contrasted, 343
+complex logical connectives: grouping with bo, 343; grouping with parenthe=
sis, 343
+complex movements: expressing, 225
+complex negation: examples, 102
+complex numbers: expressing, 434
+components contrasted with mass: in properties of, 354
+compound base: definition, 445; expressing digits in, 445; separator for, =
445
+compound bridi: definition, 344; logical connection of, 344; more than one=
 sumti in common, 345; multiple with bo, 346; multiple with ke=85ke'e, 346;=
 one sumti in common, 344; separate tail-terms for bridi-tails, 346; separa=
te tail-terms for forethought-connected bridi-tails, 347
+compound bridi with more than one sumti in common: with common sumti first=
, 345; with vau, 345
+compound cmavo: compared with sequence of simple cmavo, 51; definition, 51=
; recognition of, 51
+compound emotions, 306
+compound letter marker, 505
+compound letters: native language, representing as distinct letters, 419
+compound logical connectives: components, 336; naming convention, 336
+compound of gismu: lujvo as, 53
+compound spatial tense: as direction with-or-without distance, 218; beginn=
ing with distance only, 218; effect of different ordering, 218; explanation=
 of, 218; with direction and distance, 218
+compound subscript, 362, 455
+compound temporal tense: beginning with distance only, 220
+compound tense: compared with multiple tenses in sentence, 234; compared w=
ith tense in scope of sticky tense, 234; definition, 218; Lojban contrasted=
 with English in order of specification, 218
+compound tense ordering: Lojban contrasted with English, 218
+compound words, 273
+computer interaction, 484
+concept abstraction, 265
+concept abstraction(s): place structure, 265
+concept abstractor, 265
+concrete terms: use of fu'ivla for, 61
+condescension: example, 308
+confusion: metalinguistic, 321
+confusion about what was said, 321
+conjunctions: cmavo as Lojban equivalents, 50
+connected tenses: negation of compared with negation in connective, 245
+connecting operands: with bo in connective, 361; with ke in connective, 36=
1
+connecting operators: with bo in connective, 361; with ke in connective, 3=
61
+connection: forethought, 494; forethought separator, 495; non-distributed,=
 356; of bridi-tails, 495; of sumti, 489, 496; of tanru units, 495, 496; si=
multaneously modal and logical, 204
+connection of operands: grouping, 454; precedence over operator, 455
+connection of operators: grouping, 454
+connective answers: non-logical, 359
+connective question answers: contrasted with other languages, 353
+connective question cmavo: departure from regularity of, 352
+connective questions: answering, 352; compared with other languages, 353; =
non-logical, 359
+connective(s): as complete grammatical utterance, 352; as ungrammatical ut=
terance, 352
+connectives: for abstractions, 269; table by constructs connected, 366
+connector: for relative clauses, 508
+consonant: definition, 35; effect on syllable count, 35
+consonant clusters: buffering of, 38; contrasted with doubled consonants, =
35; contrasted with single consonants, 35; definition of, 35; more than thr=
ee consonants in, 37
+consonant pairs: in brivla, 53; initial, 36; letter y within, 53; restrict=
ions on, 36
+consonant triples, 37; restrictions on, 37
+consonant-final words: necessity for pause after, 68
+consonants: contrasted with vowels, 33; final, 36; position of, 36; pronun=
ciation of, quick-tour version, 12; restrictions on, 36; syllabic, 34; voic=
ed/unvoiced equivalents, 35; voicing of, 35
+continents: gismu for, 79
+continues: example, 228
+continuitive event contour, 228
+continuous: of tense intervals, 225
+contradictory negation: using naku before selbri, 405
+contradictory negation of modals: explanation of meaning, 207
+contradictory negation of tenses: selma'o allowed with, 242
+contributors to this book, 5
+conversion: accessing tense of bridi with jai, 247; definition, 100, 247; =
effect of multiple on a selbri, 194; effect on BAI, 196; extending scope of=
, 193; modal, 206, 496; of BAI, 195; of operator places, 459; of selbri, 50=
4; scope of, 193; swapping non-first places, 194; swapping with modal place=
, 205; using jai, 101
+conversion and tanru, 100
+conversion into sumti from mekso, 436
+conversion of mekso into sumti, 436
+conversion of operand into operator, 460
+conversion of operator into operand, 460
+conversion of operator into selbri, 457
+conversion of selbri into operand, 456
+conversion of selbri into operator, 456
+conversion of selbri to modal, 493
+conversion of sentence with quantified variables: technique, 407
+conversion of sumti into operand, 456
+conversion of sumti into selbri, 98
+conversion with `ke', 101
+conversion with se: effect of naku negation boundary on, 406
+converted selbri: as different selbri from unconverted, 192; as resetting =
standard order, 193; compared with selbri with FA in meaning, 193; contrast=
ed with other similar selbri, 193; contrasted with selbri with FA in struct=
ure, 193; definition, 192; forming with SE, 192; in descriptions, 193; plac=
e structure of, 192; retention of basic meaning in, 193; to access non-firs=
t place in description, 193
+converting: operand to operator, 500; operator to selbri, 502; quantifier =
to selbri, 500; selbri to operand, 501; selbri to operator, 501; sumti to o=
perand, 500; sumti to tanru unit, 500
+converting sumti to related meaning, 498
+co'o, 324; quick-tour version, 21
+creative understanding, 273
+credits for pictures, 6
+credits for this book, 6
+cross product: with tenses, 246
+cross-dependency, 280
+cross-product: contrasted with and, 357; of sets, 356
+cu, 188, 190, 216; as selbri separator, 188; effect of selbri-first bridi =
on, 190; effect of tense specification, 216; effect on elidability of ku, 1=
22; effect on elidable terminators, 188; necessity of, 188; need for, quick=
-tour version, 19; omission of, quick-tour version, 14; quick-tour version,=
 14; use of, quick-tour version, 14; usefulness of, 188
+CU selma=92o, 188, 492
+cu'e: combining with other tense cmavo, 250
+CUhE selma=92o, 238, 492
+cu'i, 299, 305
+cultural knowledge: example, 316
+cultural words: rafsi fu'ivla proposal for, 80
+culturally dependent lujvo, 322
+cumbersome text, 479
+cu'o, 447
+cup's friend: example, 174
+curious, 263; example, 263; example, 263
+cu'u, 203
+cycles, 230
+Cy-form cmavo: rules for pause after, 69
+Cyrillic alphabet: language shift word for, 417; proposed lerfu words for,=
 427
+da, 162, 392, 473; as a translation for =93something=94, 392; contrasted w=
ith zo'e, 392
+da poi, 394
+da prami da: contrasted with da prami de, 393
+da prami de: contrasted with da prami da, 393
+DA selma=92o, 473
+da'a, 441; default number for, 441
+da'e, 149
+DAhO selma=92o, 162, 466, 492
+da'i, 319
+da'o, 162, 466; for cancellation of pro-sumti/pro-bridi assignment, 162; s=
yntax of, 162
+da-series: after third, 472
+da-series pro-sumti: for bound variables, 161
+da'u, 148
+dai, 314
+De Gaulle: example, 68
+de'a, 229
+decimal point: as numerical punctuation, 433; effect of different notation=
s, 433; in bases other than 10, 444
+deduction: example, 316
+de'e, 149
+default operator precedence: contrasted with mekso goal, 436
+deference: example, 308
+definable pro-sumti, 139; sequences of lerfu words as, 140
+definite numbers: combined with indefinite, 442
+de'i, 204
+dei, 149
+demonstrated potential: expressing, 244
+demonstrative pro-sumti, 140, 147; stability of, 162
+DeMorgan's Law: and bridi-tail logical connection, 408; and distributing a=
 negation, 407; and internal naku negations, 409; and logically connected s=
entences, 408; and moving a logical connective relative to =93naku=94, 407;=
 sample applications, 408
+dereferencing a pointer: with la'e, 134
+derivational morphology: definition, 49
+derogatory terms, 275
+descriptions: and abstractions, 255; as based on first place of following =
selbri, 193; as possessive sumti, 180; based on go'i-series pro-bridi, 155;=
 components of, 120; importance of selbri first place in, 120; non-specific=
, 121; quick-tour version, 19; specific, 121; types of, 120; use of SE in, =
193
+descriptions with lo: teddy bear contrasted with real bear, 122
+descriptor, 498; as part of description, 120
+descriptor for numbers, 499
+descriptors: different implicit outer quantifiers among, 131; implicit qua=
ntifiers for, 129; omission of, 132; purpose of, 120
+de'u, 148
+Devanagari: example, 418
+di'a, 229
+diacritic marks: proposed lerfu words for, 429
+diacritical marks: as lerfu, 418; considered as forming distinct letters, =
419; order of specification within tei=85foi, 419; problem of position, 418=
; problem with multiple on one lerfu, 419; specifying with tei=85foi, 419
+dictionary: superior authority of, 5
+di'e, 149, 358; effect of tu'e/tu'u on, 358
+die after living: example, 232
+digit questions, 449
+digit string: definition of, 458
+digits, 502; cmavo for, 432; list of decimal, 462; list of hexadecimal, 46=
2; names from, 459; rafsi for, 459; rationale for having 16, 444
+digits beyond 9: word pattern, 444
+di'i, 226
+dimension: meaning as sumti tcita, 233
+dimensionality: of walking, 224; order with size in spatial tense interval=
s, 224; spatial, 506
+dimensionality of interval: as subjective, 224
+dimensioned numbers: expressing, 456
+diphthongs: classification of, 34; contrasted with vowel pairs, 34; defini=
tion of, 33; English analogues of, 45; in fu'ivla, 63; IPA for, 33; list of=
, 33; pronunciation of, quick-tour version, 12; specific to cmene, 66; spec=
ific to names, 66
+direct address, 323
+direction: following interval in tense construct, 221; interaction with mo=
vement specification in tenses, 224; order of relative to distance in spati=
al tenses, 217; reference frame for, 224; spatial, 493; specification with =
FAhA, 217; time, 503
+directions: multiple with movement, 225
+disambiguated instance, 276
+disclaimers, 5
+discourse: commentary on, 319; expressing utterance relation to, 317; gest=
ure markers, 319; tone of voice markers, 319
+discrete: of tense intervals, 225
+discursive indicator, 466
+discursives: as metalinguistic claims, 317; contrasted with attitudinals, =
317; definition, 317; discourse commentary, 319; discourse management, 320;=
 embedded, 481; expressing how things are said, 318; knowledge, 319; placem=
ent in sentence, 317; quick-tour version, 25; su'a, 316; word-level, 318
+discursives for consecutive discourse, 317; contrasted, 318
+discursives for managing discourse flow, 320
+discuss in language: example, 357
+distance: order of relative to direction in spatial tenses, 217; spatial, =
506; specification with VA, 217; time, 507
+distributing a negation, 407
+distribution of quantified sumti, 398
+ditto: example, 318
+di'u, 148; contrasted with la'edi'u, 149; contrasted with ta, 148; quick-t=
our version, 21
+di'u-series pro-sumti /r dihuseries, 148
+diversified species, 295
+DNA: example, 424
+do, 146
+do'a, 318
+doctor and then rich: example, 364
+do'e, 197; compared with English of, 197
+dog breathes, 395
+dog house: example, 274
+dog or cat: example, 351
+doghouse: example, 72, 277
+DOhU selma=92o, 137, 323, 492
+do'i, 149; compared with zo'e-series as indefinite pro-sumti, 158
+Don Quixote: example, 68
+Dong: example, 281
+do'o, 146
+do'u, 137, 323
+doi, 136, 138, 323; effect on necessity for pause before name-word, 138; e=
ffect on pause before name, 323; quick-tour version, 21
+DOI selma=92o, 136, 183, 323, 492; quick-tour version, 21; terminator for,=
 492
+double negation: and naku, 407
+double negatives: effect of interactions between quantifiers and negation =
on, 403
+doubled consonants: contrasted with consonant clusters, 35; contrasted wit=
h single consonants, 35
+Doyle: example, 138
+dream: example, 316
+du, 162, 164; as an exception within GOhA selma=92o, 97; compared with me =
in effect, 99; contrasted with dunli, 163, 439; contrasted with mintu, 163;=
 derivation of, 163; grammar of, 435; meaning of, 163; rationale for select=
ion of selma'o for, 162; with complex mekso on both sides, 436
+ducks swim: example, 243
+du'e, 442, 448
+du'i, 204
+dunli: contrasted with du, 163, 439
+Dutch ij: proposed lerfu word for, 429
+du'u, 263, 264
+e, 340; contrasted with fa'u, 356; contrasted with pi'u, 357
+e'a, 303
+Earl: example, 34
+eat bread: example, 441
+eat in airplane: example, 225
+eat themselves: example, 442
+EBNF grammar, 552
+editorial commentary, 480
+editorial insertion, 481; of text already containing sa'a, 321; with sa'a,=
 321
+e'e, 303
+Eek!: example, 301
+ei, 303, 305
+eicai, 305
+eicu'i, 305
+einai, 305
+Einstein: example, 294
+Einsteinian: space-time intervals with 4 dimensions, 224
+eiro'u, 307
+eiru'e, 305
+eisai, 305
+-ek: in name for logical connectives, 336
+ek: definition, 336
+eks: connecting operands, 361; in sumti forethought logical connection, 34=
0
+elementary schools, 281
+Elgin: Suzette Haden and evidentials, 315
+elidability of be'o, 94
+elidability of me'u, 99
+elidable terminators: list, 486
+elided tense: meaning of, 215
+elimination process, 282
+ellipsis: quick-tour version, 14
+elliptical pro-bridi, 157
+elliptical pro-sumti, 157
+elliptical sumti, 157
+elliptical value: contrasted with typical value for sumti, 157
+embarrassment: example, 307
+embedded bridi tenses: effect of main bridi tense on, 235
+embedded discursive, 481
+embedded quotation, 476
+emotion, 297
+emotional categories, 306
+emotional indicators: noticeable effects of, 330
+emotional scale, 305
+emotions: compound, 306; cultural bias of expression, 329; insights, 306; =
marking, 505; marking intensity of, 491; recording using indicators, 329; r=
esearch using indicators, 329; when expressed, 306
+empathy: example, 314
+emphasis, 479; changing by using non-standard form of bridi, 188; marking,=
 489
+emphasis marking, 479
+end of file, 484
+end of text indicator, 493
+endpoints: inclusion in interval, 360
+engineering: example, 14
+English prepositions: contrasted with modal tags in preciseness, 196
+English we: contrasted with Lojban pro-sumti for we, 146
+Englishman in Africa: example, 124
+enough currency: example, 442
+enough-th: example, 447
+e'o, 303; contrasted with pe'u, 324
+equivalents to brivla, 97
+erasure, 482; discourse, 504; multiple word, 483; names, 482; phrase, 503;=
 quotes, 483; sentence, 503; total, 484; word, 482, 504; zo, 482
+error marking: metalinguistic, 321
+e'u, 303, 317; compared with ru'a, 317
+-er: use of zmadu in forming, 58
+=E9t=E9: example, 419
+Eureka!: example, 299
+event abstraction(s): types, 257
+event abstractions, 256
+event contours: achievative, 230; as characteristic portions of events, 22=
8; as sumti tcita, 232; as timeless in perspective, 228; cessative, 229; co=
mpletitive, 229; continuitive, 228; contrasted with tense direction in impl=
ication of extent, 229; definition, 228; division of the event into, 229; i=
mplications on scope of event, 229; inchoative, 228; initiative, 229; inter=
ruption, 229; order with respect to TAhE and ROI, 228; pausative, 229; perf=
ective, 229; points associated with, 229; resumption, 229; resumptive, 229;=
 strings of, 246; superfective, 230; syntax of, 228; temporal contrasted wi=
th spatial, 231
+event contours as sumti tcita: contrasted with direction and distance, 232
+event types: described, 258
+event-relative viewpoint: contrasted with speaker-relative viewpoint, 228
+events: considered as a process, 228; duration, 256; place structure, 257
+everybody loves something: example, 401
+everyone: contrasted with anyone in assumption of existence, 399
+everyone bitten by dog, 396
+everything: expressing with =93ro da=94, 394
+everything breathes: example, 395
+everything loves everything: example, 394
+everything sees me: example, 394
+everything sees something: example, 394
+evidentials: ba'a scale, 316; definition, 315; grammar, 315; in English, 3=
15; indisputable bridi, 315; inspiration for, 315; ja'o contrasted with su'=
a, 316; ka'u contrasted with se'o, 316; placement in bridi, 315; quick-tour=
 version, 25; rhetorical flavor, 315; scales, 315; se'o contrasted with ka'=
u, 316; su'a contrasted with ja'o, 316
+exact number: expressing, 443
+exactly two: example, 443
+example of examples, 5
+examples: structure of, 5
+examples in this book, 4
+except from 10 to 12: example, 360
+existential: mixed claim with universal, 394
+existential claims: definition, 392; restricting, 394
+existential variable: in abstraction contrasted with in main bridi, 400; i=
n main bridi contrasted with in abstraction, 400
+expanding =93no=94 quantifier, 403
+experience abstraction, 265
+experience abstraction(s): place structure, 265
+experience abstractor, 265
+experienced: example, 316
+experimental cmavo: definition, 51; forms for, 51
+explicit magnitude, 250
+exponential notation: with base other than 10, 451; with gei, 450
+exporting negation to prenex: =93naku=94 contrasted with internal bridi ne=
gation, 406; internal bridi negation contrasted with =93naku=94, 406
+expressive power, 290
+external bridi negation: compared to internal bridi negation, 401; definit=
ion, 401
+extrinsic possession: definition, 173
+F.8 base 16: example, 444
+fa, 189
+FA in selbri: compared with converted selbri in meaning, 193; contrasted c=
onverted selbri with in structure, 193
+FA selma=92o, 93, 189, 206, 247, 493; after 5th place, 472; as a reminder =
of place in place structure, 190; avoidance of complex usage of, 191; compa=
red with zo'e for omitting places, 190; effect on place structure, 190; eff=
ect on place structure order, 190; effect on subsequent non-tagged places, =
191; for accessing a selbri place explicitly by relative number, 190; for p=
utting more than one sumti in a single place, 191; syntax of, 189
+FA tags and linked sumti, 93
+fa'a: special note on direction orientation, 253
+face: specifying for letters, 418
+FAhA selma=92o, 217, 242, 493; and direction, 217; contradictory negation =
of, 241; use in specifying space/time mapping direction, 231
+FAhO selma=92o, 484, 493
+false statement: implications of, 337
+fancy E: notation convention, 335
+fancy O: notation convention, 335
+fa'o, 416, 484; contrasted with fe'o, 325; interaction with bu, 416
+fai, 206, 247, 287; as allowing access to original first place in modal co=
nversion, 206; effect on numbering of place structure places, 206
+falsity of mathematical relation: expressing, 440
+fancy A: notation convention, 335
+fancy U: notation convention, 335
+fast talker: example, 17
+fast-talker shoe: example, 17
+father: example, 11
+father mother: example, 55
+fa'u, 354, 356, 470; compared to termsets, 356; contrasted with .e, 356
+fe, 189
+fe=92omi'e, 325
+fe'e, 230, 231, 232; effect of TAhE/ROI with ZAhO on, 231
+feed: example, 288
+feeling, 297
+feelings: expression of contrasted with talking about, 298
+FEhE selma=92o, 230, 493
+FEhU selma=92o, 194, 493
+fe'o, 325; contrasted with fa'o, 325
+fe'u, 194
+fewsome: example, 446
+fi, 189
+fi'a, 191; effect on subsequent untagged sumti, 192
+Fido: example, 396
+fi'e, 203
+field rations: example, 56
+figurative lujvo, 322; place structure, 322
+figurative speech, 322
+FIhO selma=92o, 194, 493; terminator for, 493
+fi'i, 324
+final syllable stress: rules for pause after, 69
+finish: contrasted with stop, 229
+finished: example, 229
+fi'o, 194; and modal conversion, 206; as modal tag, 194; effect on followi=
ng selbri, 194; mixed modal connection with, 205; proscribed for sticky mod=
als, 208; restriction on use, 201; use in adding places to place structure,=
 194
+fi'o constructs: short forms as BAI, 195
+fi'o modal followed by selbri: effect on eliding fe'u, 202
+fi'o modals: negation of by negating selbri, 207; usage in relative phrase=
s, 204
+fi'o tag: relation of modal sumti following to selbri, 195
+fi'o with selbri: meaning of, 194
+first rat: example, 447
+firstly: example, 458
+fish eat: example, 468
+fish on right: example, 222
+fi'u, 433
+five people: example, 178
+five women: example, 178
+flashbacks in story time: example, 237
+fleas, 274
+flexible vocabulary, 53
+floating point numbers: expressing, 451
+flow of discourse: managing with discursives, 320
+fo, 189
+fo'a, 163
+foi, 419
+FOI selma=92o, 419, 494
+folk quantifiers: expressing, 456
+font: example, 418; specifying for letters, 418
+food: use of fu'ivla for specific, 61
+foreman of a jury: example, 146
+forethought bridi connection: as grammatically one sentence, 339
+forethought bridi connectives: contrasted with afterthought bridi connecti=
ves, 338
+forethought bridi-tail connection: special rule for tense, 365
+forethought connection: contrasted with afterthought for grammatical utter=
ances, 352; definition, 199; in abstractions, 365; in tenses, 363; observat=
ives, 347; of operands, 453; of operators, 453
+forethought connections: modal compared with tense in semantics, 249
+forethought connective(s): as ungrammatical utterance, 352; contrasted wit=
h afterthought connective, 338
+forethought connectives: with tense, 364
+forethought connectives and bo, 343
+forethought interval: GAhO position, 361
+forethought logical connectives: within tanru, 92
+forethought logical connectives in tanru: effect on tanru grouping, 92
+forethought mathematical notation (see also Polish), 438
+forethought modal sentence connection, 199; relation to modal of first bri=
di in, 199; relation to modal of second bridi in, 199
+forethought modal sentence connection for causals: order of cause and effe=
ct, 199
+forethought tanru connection, 350
+forethought tense connection: contrasted with afterthought in likeness to =
modal connection, 249
+forethought tense connection of bridi-tails: order of, 240
+forethought tense connection of sentences: order of, 239
+forethought tense connection of sumti: order of, 239
+forethought termsets: logical connection of, 348
+formal grammar, 511
+formal requirement: example, 305
+former market: example, 235
+former state, 293
+formulae: expressing based on pure dimensions, 456
+four =93e=94s: example, 422
+Four score and seven: example, 460
+fourteen =93e=94s: example, 413
+fractions: expressing with numerical punctuation, 433; numerator default, =
433
+fragmentary text, 321
+Frank is a fool: example, 263
+free modifiers: effects on elidability of terminators, 450
+frequency within interval: specifying, 504
+friend's cup: example, 174
+from one to two o'clock: example, 359
+fu, 189
+fu'a, 452
+fu'e, 475
+FUhA selma=92o, 494
+FUhE selma=92o, 475, 494
+FUhO selma=92o, 475, 494
+fu'i, 309
+fu'ivla: algorithm for constructing, 62; as a subtype of brivla, 53; as St=
age 3 borrowings, 61; as Stage 4 borrowings, 62; categorized contrasted wit=
h uncategorized in ease of construction, 62; considerations for choosing ba=
sis word, 64; consonant clusters in, 62; construction of, 62; definition, q=
uick-tour version, 27; diphthongs in, 63; disambiguation of, 64; form for r=
afsi fu'ivla proposal, 80; form of, 62; initial consonant cluster in, 62; m=
ethod of including in lujvo, 60; quick-tour version, 20; rules for formatio=
n of, 62; stress in, 62; uniqueness of meaning in, 61; use of, 61; with inv=
alid diphthongs, 64
+fu'ivla categorizer, 61; for distinguishing fu'ivla form, 64; for distingu=
ishing specialized meanings, 64; selection consideration for, 62
+Fujiko: example, 68
+fully reduced lujvo: definition, 59
+function f of x: example, 423
+function name: lerfu string as, 423
+functional notation: standard, 438
+fu'o, 475
+future event: possible extension into present, 223
+futureward: as a spatial tense, 224
+fuzzy logic and truth-value abstraction, 262
+ga, 408
+GA selma=92o, 336, 338, 339, 340, 341, 352, 361, 453, 494
+gadri: definition, 119
+ga'e, 415
+GAhO position in forethought intervals, 361
+GAhO selma=92o, 246, 360, 361, 494; grammar of, 360
+ga'i, 308
+ga'icai, 309
+ga'inai, 308
+ganai, 339, 340
+ga'o, 360, 362; etymology of, 360
+ge, 408
+ge'a, 450, 452, 453; for infix operations with too many operands, 451
+ge'a gei, 453
+ge'e, 311, 322
+GEhU selma=92o, 175, 495
+ge'i, 352
+gei, 450; as a binary operator, 450; as a ternary operator, 451; rationale=
 for order of places, 451
+gek: definition, 338
+gek bridi connectives: contrasted with ijeks, 338
+geks: connecting operands, 361; in forethought sumti connection, 341; synt=
ax of, 340
+General American, 42
+general sumti: contrasted with operands, 436
+general terms, 295
+German rich man: example, 344
+gerund: using abstraction, 255
+Gettysburg Address: example, 460
+ge'u, 151, 175, 318; effect of following logical connective on elidability=
, 175; elidability of from relative phrases, 175
+gi, 92, 199, 239, 240, 340, 361, 408
+GI selma=92o, 336, 339, 340, 361, 495
+gi'e, 344
+GIhA selma=92o, 336, 344, 346, 352, 364, 495; terminator for, 506
+gihek: definition, 344
+giheks: syntax of, 346
+gi'i, 352
+gik: as name for compound cmavo, 336; definition, 340
+giks: syntax of, 340
+ginai, 340
+girls' school: little, example, 85
+gismu: algorithm for, 75; and cmavo, major, 53; as a subtype of brivla, 53=
; as partitioning semantic space, 53; basic rafsi for, 57; coined, 77; conf=
licts between, 54; creation, and transcription blunders, 76; creation, cons=
iderations for selection after scoring, 75; creation, proscribed gismu pair=
s, 76; creation, scoring rules, 75; cultural, 78; definition, 53; definitio=
n, quick-tour version, 27; ethnic, 79; examples of, 54; exceptions to gismu=
 creation by algorithm, 77; for countries, 79; for languages, 78; for Lojba=
n source languages, 78; geographical, 79; length of, 54; level of uniquenes=
s of rafsi relating to, 57; Lojban-specific, 77; place order, rationale, 29=
5; place structures, 294; place structures, rationale, 294; quick-tour vers=
ion, 20; rationale for, 273; rationale for choice of, 53; religious, 80; ru=
les for, 54; scientific-mathematical, 77; selection of, 53; source of, 54; =
source-language weights for, 76; special, 54; too-similar, 76
+give: example, 11
+give or receive?: example, 191
+giving the horse: example, 260
+global attitudinals, 475
+glottal stop: as pause in Lojban, 31
+glue in lujvo: n-hyphen as, 56; r-hyphen as, 56; y-hyphen as, 56
+go: example, 187
+go to Boston from Atlanta: example, 187
+go to market: example, 215
+go to Paris or Rome: example, 408
+go to the store: example, 4
+go=92i-series pro-bridi: assigning for permanent reference, 154; in narrat=
ive about quotation, 156; in quotation series, 156; in quotations, 156
+go'a, 154
+goal of this book, 3
+go'e, 154
+goer table: example, 85
+goer-house: example, 274
+Goethe, 479
+GOhA selma=92o, 97, 145, 409, 470, 495; as component in tanru, 97; as selb=
ri, 97
+go'i, 97, 154, 318; as affirmative answer to yes/no question, 154; compare=
d with mo in overriding of arguments, 160; contrasted with go'i ra'o, 156; =
contrasted with mi'u, 318
+goi assignment of ko'a-series pro-sumti: use in speech contrasted with wri=
ting, 151
+goi for ko'a-series assignment: compared with cei for broda-series assignm=
ent, 151
+go'i ra'o: contrasted with go'i, 156
+GOI selma=92o, 172, 495; terminator for, 495
+go'i with xu: quick-tour version, 23
+go'i-series pro-bridi, 152; as basis for description, 155; as main-bridi a=
naphora only, 154; as main-bridi anaphora only, exception, 155; as repeatin=
g referent concept, 155; compared with ri-series in word formation, 152; co=
mpared with ri-series pro-sumti in rules of reference, 154; effect of sub-c=
lauses on, 154; effect of sumti of referent bridi on, 154; no'a as exceptio=
n to only main-bridi anaphora, 155; referent of, 154; reinterpreting sumti =
references with ra'o, 156
+go'ixire, 154
+go'o, 155
+go'u, 154
+goi, 150, 154, 162, 421; rationale for non-inclusion in relative clause ch=
apter, 175; use in assigning lerfu as pro-sumti, 152; use in assigning name=
, 152
+good house: example, 92
+grammatical categories: use of upper case for, 5
+grammatical terms: quick-tour version, 26
+grasp water: example, 199
+great soldier: example, 278, 282
+Greek alphabet: language shift word for, 417; proposed lerfu words for, 42=
6
+Greek-Americans own restaurants, 126
+grouping: indicator for, 496; of connection in abstractions, 365; of conne=
ction in tenses, 363
+grouping parentheses, 88
+gu, 339
+GU selma=92o, 339
+gu'e, 92
+GUhA selma=92o, 350, 352, 361, 495
+guhek: definition, 350
+guheks: connecting operators, 361; syntax of, 350
+guheks for tanru connection: rationale, 350
+gu'i, 352
+had earlier: example, 234
+han4zi4: example, 420
+hands in pockets: example, 175
+handwriting: example, 418
+happiness: example, 161
+happy face: example, 416
+has a heart: example, 259
+have never: example, 227
+having: of properties, 259
+healthy: example, 24
+hearsay: example, 316
+heartburn: example, 322
+Hebrew alphabet: language shift word for, 417; proposed lerfu words for, 4=
28
+Helvetica font: example, 418
+hepatitis: example, 60
+hereafter known as: example, 151
+hesitation, 484, 507
+hesitation sound, 484
+hexadecimal system: specifying numbers in (see also base), 444
+hierarchy of priorities for selecting lujvo form, 72
+hiragana: contrasted with kanji, 420; example, 418
+hit cousin: example, 318
+hit nose: example, 318
+hits: example, 11
+Hollywood: example, 127
+Hooray!: example, 299
+hospitality: example, 324
+hours:minutes:seconds: example, 445
+huh?: example, 321
+hundred: expressing as number, 432
+husband and wife: example, 316
+hyphen letter: definition, 59
+hyphens: use of, 59
+hyphens in lujvo: proscribed where not required, 70
+hypothetical world, 301; contrasted with real world, example, 320
+hypothetical world point of view, 320
+i, 198, 238, 465; quick-tour version, 16; regarding forethought bridi conn=
ection, 339
+I selma=92o, 336, 337, 338, 339, 358, 364, 465, 495
+i'anai, 304
+ianai, 297, 303, 319
+IBM: example, 424
+ICAO Phonetic Alphabet: proposed lerfu words for, 429
+ice'o: contrasted with .ibabo, 358
+idea abstraction, 265
+idea abstraction(s): place structure, 265
+identity: expressing with po'u, 174
+identity predicate, 162
+i'e, 304
+i'enai, 304
+ie, 304
+if: English usage contrasted with Lojban logical connective, 337; expressi=
ng real world, 320; meaning in logical connections, 337
+if =85 then: compared with only if, 338; logical connectives contrasted wi=
th other translations, 339
+if coffee: bring tea, example, 353
+if, expressing hypothetical world, 320
+ii, 301
+ijebabo, 363
+ijek: definition, 336
+ijek bridi connectives: contrasted with geks, 338
+ijek logical connective(s): connecting bridi, 336
+ijeks: syntax of, 338
+ijoik: as name for compound cmavo, 336; definition, 358
+imaginary journey: and spatial tense, 217; ending point, 217; origin in te=
nse forethought bridi-tail connection, 240; origin in tense forethought sen=
tence connection, 239; origin in tense forethought sumti connection, 239; o=
rigin of in tense-connected sentences, 239; stages of in compound tenses, 2=
18; starting at a different point, 232; starting point, 217, 232; with inte=
rval direction, 222
+imaginary journey origin: with sticky tenses, 234
+imperatives: and truth, 353; attitude, 308; English contrasted with Lojban=
 in presence of subject of command, 147; quick-tour version, 22; with ko, 1=
46
+implausible, 284
+implicit quantifier: definition, 128; for quotations, 128; on personal pro=
-sumti, 128; on quotations, discussion of, 128
+implicit-abstraction lujvo: definition, 289
+importance of point: scale with ra'u, 320
+in the aftermath: example, 232
+inalienable: distinguishing from alienable, 173
+inalienable possession: definition, 173; expressing with po'e, 173
+inchoative event contour, 228
+incidental association: expressing with ne, 174
+incidental identification: expressing with no'u, 174
+incidental relative clause: as a parenthetical device, 171; definition, 17=
1
+inclusion: property of sets, 125
+indefinite description: as needing explicit outer quantifier, 132; as proh=
ibiting explicit inner quantifier, 132; compared with restricted variable, =
398; definition, 132, 398
+indefinite numbers, 440; combined with definite, 442; effect of pi on, 441
+indefinite portions: subjective, 442
+indefinite pro-bridi, 157; stability of, 162
+indefinite pro-sumti, 140, 157; implicit quantifier for, 140; stability of=
, 162
+indefinite sumti: as implicit quantification, 406; compared to sumti with =
lo, 399; meaning when multiple in sentence, 398; multiple in sentence, 398
+indefinite values: subjective, 442
+index numbering, 500
+indicator scope, 466
+indicator tables: format convention, 298
+indicators, 298; cancellation of, 494; derived from gismu, 298; evolutiona=
ry development of, 329; grammar for compounding, 310; meaning when compound=
ed, 310; placement of, 298; quick-tour version, 24; ramifications, 329; rat=
ionale for selection, 329; scope effect of new paragraph, 466; types of, 29=
8
+indicators derived from gismu: notation convention, 298
+indirect question, 323
+indirect question involving sumti, 265
+indirect questions, 264; =93ma kau=94 contrasted with =93la djan. kau=94, =
264
+indirect questions without =93kau=94, 265
+indisputable bridi, 315
+individual: example, 446
+individual objects: multiple, 123
+individuals: expressing relation with mass formed, 446; expressing relatio=
n with set formed, 446
+individuals into mass: by non-logical connection, 355
+individuals into set: by non-logical connection, 355
+individuals of set: expressing measurement standard for indefinites, 446
+indivisible, 258
+induction: example, 316
+inexact numbers with bounds, 443
+inexact portions with bounds, 444
+infant ducks: example, 244
+inferior: example, 308
+infinity: example, 434
+infix expressions: in operands being used in Polish notation, 439
+infix mathematical notation: shortcomings of, 438
+infix notation mixed with Polish, 455; example, 455
+inflammable: example, 245
+initial consonant pairs: list of, 37
+initiative event contour, 228
+innate capability: expressing explicitly, 243; expressing implicitly, 243
+innate property: extension of from mass to individuals, 243; extension to =
individuals not actually capable, 244
+inner product, 452
+inner quantifier: contrasted with outer quantifier, 129; definition, 129; =
effect of on meaning, 129; explicit, 129; implicit on descriptors, 129; in =
indefinite description, 132
+inner quantifier of sumti: meaning of, 178
+inner sumti: referring to from within relative clause within relative clau=
se, 184
+integral: architectural concept, example, 64; mathematical concept, exampl=
e, 64
+interaction list: cmavo, 485
+interactions between quantifiers and negation: effect, 403
+interjections: quick-tour version, 24
+intermediate abstraction, 267
+intermittently: example, 226
+internal bridi negation: compared to external bridi negation, 401; definit=
ion, 401
+internal naku negations: and DeMorgan's Law, 409
+internal world, 301
+International Phonetic Alphabet (see also IPA), 29
+intersect, 266
+intersection: of sets, 356
+intersection of sets: compared with and, 357
+interval: closed, 360; effect of nai on, 360; expressed as center and dist=
ance, 359; expressed as endpoints, 359; expressing by endpoints with bi'o, =
246; followed by direction in tense construct, 221; forethought, 361; formi=
ng, 490; inclusion of endpoints, 360; open, 360; open/closed specification,=
 494; relation to point specified by direction and distance, 221; relative =
order with direction and distance in tense, 221; specifying relation to poi=
nt specified by direction and distance, 221; spread of actions over, 225
+interval continuousness: meaning as sumti tcita, 233
+interval direction: specifying, 221
+interval properties: meaning as sumti tcita, 233; strings of, 246
+interval size: as context-dependent, 222; meaning as sumti tcita, 233; spa=
tial, 506; time, 507; unspecified, 223; vague, 223
+interval spread: expressing English intermittently, 226; mutually contrast=
ed, 226; negation with nai, 226; with unspecified interval, 226
+intrinsic possession: definition, 173; expressing by using place in some s=
elbri, 173; expressing with po'e, 173
+introduce oneself, 325
+invalid diphthongs: in fu'ivla, 64
+invalid speech: marking as error with na'i, 321
+inversion of quantifiers: definition, 402; in moving negation boundary, 40=
2
+inversion of quantifiers on passing negation boundary: rationale for, 402
+invertebrate, 285
+inverted tanru: effect on sumti after the selbri, 95; effect on sumti befo=
re the selbri, 95
+inverting a tanru, 491
+inverting quantifiers: with movement relative to fixed negation, 406; with=
 movement relative to naku, 405
+ionai, 301
+IPA, 29
+IPA pronunciation: description, 42
+iri'abo, 198
+irony: example, 319; expressing, 319
+irrational number: example, 181
+irrelevant: specifying of sumti place, 157
+isomorphism: audio-visual, 29
+IT: as notation convention in relative clause chapter, 170
+italic: example, 418
+-ity, 259
+iu, 301
+iy diphthong: in cmene, 66
+JA selma=92o, 90, 245, 336, 337, 338, 350, 352, 354, 361, 364, 365, 496
+jabo, 91
+jai, 206, 247, 267, 287; conversion using, 101
+JAI selma=92o, 206, 267, 496
+jai with tense: as equivalent of SE in grammar, 247
+jai without modal: meaning, 206
+James: example, 36
+Jane: example, 65
+ja'o, 316
+Japanese hiragana: example, 418
+Japanese katakana: example, 418
+jargon: use of fu'ivla for, 61
+je, 89, 208
+je'e, 324, 325; contrasted with vi'o, 325
+je'i, 352
+jei, 262; place structure, 262
+jek: definition, 336
+jeks: connecting abstractors, 365; connecting operators, 361; syntax of, 3=
50
+Jesus, 266; example, 266
+ji, 352
+ji'a, 317
+ji'i, 442; effect of placement, 442; with elided number, 443
+Jim: example, 65
+jo'a, 321
+jo'e, 354, 356
+JOhI selma=92o, 451, 496; terminator for, 505
+John and Sam: example, 11
+John is coming: example, 297
+John Jones: example, 65
+John Paul Jones: example, 138
+John says that George goes to market: example, 238
+Johnson: example, 68
+jo'i, 451; precedence of, 451
+joi, 353, 354, 355
+joi grammar: contrasted with eks, 354; contrasted with jeks, 354
+JOI selma=92o, 246, 336, 353, 354, 355, 357, 358, 360, 361, 364, 455, 470,=
 496
+joigik: as name for compound cmavo, 336; definition, 361
+joigiks: connection types, 361; syntax of, 361
+joik: as name for compound cmavo, 336; definition, 354
+joiks: effect of nai on, 358; grouping, 357; syntax of, 360; use of =93se=
=94 in, 355
+jokes, 4
+Jones: John, example, 65
+jo'u, 354, 355; contrasted with ce, 355; contrasted with ce'o, 355; contra=
sted with joi, 355; result of connection with, 355
+j-sound in English: representation in Lojban, 31
+ju'a, 317
+ju'apei, 317
+Judy: example, 294
+ju'i, 324
+ju'o, 319
+Jupiter life: example, 365
+juror 5: example, 181
+ju'u: grammar of, 444
+ka, 259
+ka'a, 196
+ka'e, 243
+kanji: contrasted with alphabets and syllabaries, 420; representing based =
on romaji spelling, 420; representing based on strokes, 420
+ka'o, 434; as special number compared with as numerical punctuation, 434
+katakana: example, 418
+Kate: example, 65
+Katrina: example, 65
+ka'u, 316
+kau, 264, 323; ma kau, contrasted with la djan. kau, 264
+ke, 88, 193, 205, 343, 344, 346, 350, 361, 364; contrasted with bo for ten=
sed logical connection, 364; for conversion of tanru, 101; for expanding sc=
ope of scalar negation, 101
+ke in sumti grouping: where allowed, 344
+KE selma=92o, 88, 343, 344, 350, 361, 364, 454, 496; terminator for, 497
+ke'a, 160, 169; ambiguity when omitted, 161; and abstract descriptions, 16=
1; as referent for relativized sumti, 169; contrasted with ri in relative c=
lauses, 161; effect of omission of, 170; for relativized sumti in relative =
clauses, 160; meaning in relative clause inside relative clause, 184; non-i=
nitial place use in relative clause, 170; stability of, 162; subscripting f=
or nested relative clauses, 161
+ke'a with subscript: use for outer sumti reference, 184
+ke'axipa, 184
+ke'e, 88, 193, 205, 343, 361
+KEhE selma=92o, 88, 343, 346, 497
+ke'i, 360, 362; etymology of, 360
+KEI selma=92o, 98, 255, 497; eliding, 255
+ke'o, 325; compared to ki'a, 325
+kei, 255, 262
+kept on too long: example, 230
+ke'u, 319; contrasted with va'i, 320
+ke'unai, 320
+Khrushchev: example, 68
+ki, 207, 234, 243; with no tense, 235
+KI selma=92o, 234, 497
+ki'a, 321; compared to ke=92o, 325
+ki'e, 324
+killing Jim, 258; example, 258
+ki'o, 433
+kissing Jane, example, 256
+ki'u, 197
+klama, 187; place structure of, 187
+know: example, 263
+know who: contrasted with know that, 264; example, 264
+knowledge discursives, 319; compared with propositional attitudes, 319
+ko, 119, 146; in later selbri place in imperative, 147; in sub-clause of m=
ain bridi, 147; quick-tour version, 22; use for commands, 146; use for impe=
ratives, 146
+ko'a, 150
+ko'a-series: after tenth, 472
+ko'a-series for pro-sumti: compared with broda-series for pro-bridi, 151
+ko'a-series pro-sumti, 150; as assignable, 150; assigning with goi, 150; a=
ssignment with goi as symmetrical, 150; contrasted with lerfu as pro-sumti =
in explicit assignment of, 152; effect on ri-series pro-sumti, 153
+KOhA selma=92o, 145, 260, 469, 497
+Korean: example, 64
+Krishna: example, 68
+ku, 122, 177, 201, 216, 354; as elidable terminator for descriptions, 122;=
 effect of following selbri on elidability of, 122; effect of possessive su=
mti on elidability of, 181; effect on of omitting descriptor, 132; quick-to=
ur version, 19; uses of, 122; with tense, 216
+KU selma=92o, 354, 497; quick-tour version, 19
+ku'a, 354, 356
+ku'e, 438
+KUhE selma=92o, 497
+KUhO selma=92o, 169, 498
+ku'i, 317, 353
+ku'o, 169, 178, 394; effect of relative clause after descriptor on elidabi=
lity, 178; effect of vau on elidability, 181; elidability for relative clau=
ses, 170
+Kzinti: communication with, 329
+la, 119, 121, 129, 137, 138; compared with le in specificity, 121; contras=
ted with lai in implications, 124; contrasted with le in implications, 122;=
 contrasted with lo in implications, 122; contrasted with vocatives, 323; c=
ontrasted with zo, 478; implications of, 121; use with descriptions contras=
ted with use before Lojbanized names, 121
+LA selma=92o, 120, 138, 193, 498; contrasted with LE in use of name-words,=
 138; effect on necessity for pause before name-word, 138; terminator for, =
497
+la'a, 319
+Lady: example, 66
+la'e, 134, 149, 182, 422, 459, 478; as short for =93le selsinxa be=94, 134=
; effect of on meaning, 134
+la'e lu: compared with me'o, 422
+la'edi'u, 149; contrasted with di'u, 149; quick-tour version, 21
+LAhE selma=92o, 133, 149, 182, 266, 478, 498; effect of relative clause pl=
acement with, 182; terminator for, 499
+la'i, 125, 130, 138; as set counterpart of lai, 125
+lai, 123, 130, 137, 138; as mass counterpart of lai, 123; contrasted with =
la in implications, 124
+lambda calculus: operator and operand distinction in, 460
+language shift: based on name + bu, 418; choice of Lojban-lerfu-word count=
erpart, 417; compound, 418; effect on following words, 417; formation of sh=
ift alphabet name, 418; interaction with bu, 417; rationale for, 417; stand=
ardization of, 418
+languages: abbreviations for, 104
+la'o, 61, 416, 479; interaction with bu, 416
+Laplace: example, 65
+large meal: example, 310
+large-base decimal fraction: expressing, 445
+la-series descriptors: compared with le-series in implicit quantification,=
 130
+latent component, 291
+Latin: alphabet of Lojban, 413
+Latin alphabet, 29; language shift word for, 417
+lau, 419; effect on following lerfu word, 419
+LAU selma=92o, 415, 418, 419, 498; grammar of following BY, 426
+le, 119, 129, 177, 354; and specificity, 120; and truth of selbri, 120; co=
mpared with English =94the=94, 120; compared with la in specificity, 121; c=
ontrasted with lo in implications, 122; contrasted with lo in implicit quan=
tification, 131; contrasted with lo in specificity, 121; contrasted with lo=
 in truth requirement, 121; implications of, 120; implicit outer quantifier=
 for, 131; in false-to-fact descriptions, 120; meaning of in the plural, 12=
3
+le contrasted with lo: for relative clause placement considerations, 179
+le nu: definition, 256
+LE selma=92o, 120, 193, 247, 354, 498; contrasted with LA in use of name-w=
ords, 138; terminator for, 497
+learning Lojban: magnitude of task, 53
+Lech Walesa: example, 68
+le'e, 126, 130; relationship to le'i, 126
+left-grouping rule: definition of, 86
+legal jargon: example, 151
+legal system, 262
+legalities: boring, 8
+LEhU selma=92o, 476, 498
+le'i, 125, 130; as set counterpart of lei, 125; relationship to le'e, 126
+lei, 123, 130; contrasted with loi in specificity, 124
+lemon tree: example, 84
+Length ( Width ( Depth =3D Volume: example, 456
+le'o, 309
+Lepidoptera: example, 18
+lerfu: as assignable pro-sumti, 152; contrasted with lerfu word, 413; defi=
nition, 413; reference to, 422; referring to with me'o, 422
+lerfu as pro-sumti: contrasted with ko'a-series in explicit assignment of,=
 152; explicit assignment of antecedent, 152; implicit assignment of antece=
dent, 152
+lerfu juxtaposition interpretation: contrasted with mathematical interpret=
ation, 423
+lerfu pro-sumti: effect on ri-series pro-sumti, 153
+lerfu shift scope: exception for mathematical texts, 423
+lerfu string: as acronym using =93me=94, 424; as function name, 423; as fu=
nction, in mathematics, 438; as mathematical variable, 422; as pro-sumti, 4=
21; as pro-sumti assigned by goi, 421; as pro-sumti, assumption of referenc=
e, 421; as pro-sumti, for multiple sumti separated by boi, 421; as quantifi=
er, 423; as quantifier, avoiding interaction with sumti quantifier, 423; as=
 selbri, 423; as subscript, 423; as utterance ordinal, 423; as variable, in=
 mathematics, 438; definition, 420; in mathematical expressions, 437; inter=
pretation of contrasted with normal mathematical interpretation, 437; inter=
pretation, contrasted with mathematical interpretation, 423; use in mathema=
tics, 422; with numerical selbri, 448
+lerfu word: contrasted with lerfu, 413; for =93 ' =94, 414
+lerfu word cmavo: list of auxiliary, 425
+lerfu word set extension: with bu, 416
+lerfu words: as a basis for acronym names, 423; composed of compound cmavo=
, 414; composed of single cmavo, 414; consonant words contrasted with vowel=
 words, 414; effect of systematic formulation, 414; for consonants, 414; fo=
r vowels, 414; formation rules, 414; forming new for non-Lojban letters usi=
ng bu, 419; list of proposed, notation convention, 426; Lojban coverage req=
uirement, 413; proposed for accent marks, 429; proposed for Cyrillic alphab=
et, 427; proposed for diacritic marks, 429; proposed for Greek alphabet, 42=
6; proposed for Hebrew alphabet, 428; proposed for multiple letters, 429; p=
roposed for noisy environments, 429; proposed for radio communication, 429;=
 table of Lojban, 414; using computer encoding schemes with se'e, 425; vowe=
l words contrasted with consonant words, 414
+lerfu words ending with =93y=94: pause after, rationale, 414
+lerfu words for vowels: pause requirement before, 414
+lerfu words with numeric digits: grammar considerations, 420
+le-series cmavo: as encompassing le-series and la-series descriptors for q=
uantification discussion, 130; definition, 130; rationale for implicit inne=
r quantifier, 130; rule for implicit inner quantifier, 130
+le-series descriptors: compared with la-series in implicit quantification,=
 130
+less: English word, expressing with relative phrases, 203; English word, i=
mportance of relative phrase to, 204
+less than: contrasted with more than, at least, at most, 443; example, 443
+letter: alphabet, 413; contrasted with word for the letter, 413; making a =
word into, 490
+letter encoding schemes: application to lerfu words, 425
+letter shift, 498
+letteral: definition, 413
+letters, 491; non-Lojban, representation of diacritical marks on, 418; non=
-Lojban, representation with consonant-word + bu, 417; non-Lojban, represen=
tation with consonant-word + bu, drawback, 417; non-Lojban, representation =
with language-shift, 417; non-Lojban, representation with names, 416; sound=
 contrasted with symbol for spelling, 417; symbol contrasted with sound for=
 spelling, 417
+le'u, 119, 141, 476; interaction with zoi, 478
+l-hyphen: use of, 62
+li, 119, 141, 435; as converter of mekso into sumti, 436; contrasted with =
me'o, 457; relation to me'o compared with la/zo relation, 457; terminator f=
or, 454
+LI selma=92o, 142, 422, 499; terminator for, 499
+li'a, 318
+ligatured fi: proposed lerfu word for, 429
+LIhU selma=92o, 476, 499
+li'i, 265
+likes more than: example, 203
+lined up, 283
+linguistic behavior, 263
+linguistic drift, 4
+linguistic drift in Lojban: possible source of, 69
+linked arguments, 471
+linked sumti: definition, 93; in tanru, 93
+linked sumti and FA tags, 93
+linked sumti and sumti tcita, 94
+Linnaean, 479
+Linnaean binomials, 479
+Linnaean names: rules for, 67
+li'o, 321
+lion in Africa: example, 126
+lions in Africa: example, 124
+liquefy: example, 289
+list: as a physical object, 355; contrasted with sequence, 355; example, 3=
55
+list of things to do: example, 358
+listen attentively: example, 278
+lists: use of tu'e/tu'u in, 358
+literally, 322
+li'u, 119, 141, 476
+living things: example, 157
+Livingston: example, 317
+LLG, 5
+lo, 121, 129; and truth of selbri, 121; contrasted with le in implications=
, 122; contrasted with le in implicit quantification, 131; contrasted with =
le in specificity, 121; contrasted with le in truth requirement, 121; contr=
asted with loi and lo'i, 125; implications of, 121; implicit outer quantifi=
er for, 131; omission of, 132
+lo contrasted with le: for relative clause placement considerations, 179
+lo'a: contrasted with na'a, 418
+lo'e, 126, 130; relationship to lo'i, 126
+logic: and attitudinals, 392; limits of, 392; resolving ambiguities of =93=
nobody=94, 391
+logic and Lojban: more aspects, 411
+logical connection: effect on elidability of lo'o, 454; grouping strategie=
s for complex cases contrasted, 343; in abstraction(s), inner bridi contras=
ted with outer bridi, 365; in mathematical expressions, 361; in tanru, cont=
rasted with unconnected version, 349; in tanru, expandability of, 349; in t=
anru, grouping with bo, 349; in tanru, grouping with ke, 350; inside an abs=
traction(s), contrasted with outside, 365; interaction with tenses, 363; ne=
gation in connecting more than 2 sentences, 342; of bridi-tail as opposed t=
o tanru, 350; of bridi-tails, forethought, 347; of bridi-tails, restriction=
 on ke, 346; of forethought termsets, 348; of modals, 208; of more than 2 s=
entences, all or none, 342; of more than 2 sentences, forethought, 342; of =
more than 2 sentences, mixed =93and=94 and =93or=94, 342; of more than 2 se=
ntences, things to avoid, 342; of observatives, relation of first places, 3=
45; of selbri, 344; of sumti, grouping with parenthesis, 344; of sumti, res=
triction on ke, 344; of tanru as opposed to bridi-tail, 350; of tanru, cave=
at, 350; termsets, 347; transformation between forms, 340; with bo, precede=
nce, 342
+logical connectives, 333; associative, 341; bridi-tail connection, 345; cm=
avo, format for each selma'o, 336; effect on elidability of ge'u from prece=
ding relative phrase, 175; grouping with bo, 342; in tanru, 89; more than 2=
 sentences, 341; negated first sentence as a potential problem for understa=
nding, 339; observative sentence connection, 345; pairing from left, 342; r=
ationale for multiple sets in grammar, 335; recipes, simplified for logic c=
hapter discussion, 403; relation to truth functions, 334; relative preceden=
ce with me'u, 99; right-grouping with bo, 343; selma'o, enumerated, 336; sy=
ntax rules summary, 366; table by truth function value, 366; tensed, 240
+logical connectives and bridi negation, 403
+logical connectives and negation: caveat for logic chapter discussions, 40=
3
+logical connectives in tanru, 349; ambiguity of, 90; effect on formal logi=
cal manipulations, 90; effect on tanru grouping, 89; usefulness of, 90
+logical connectives within negation: effects of expansion on, 407
+logical language: truth functions, 333
+Logical Language Group: example, 74; relation to Lojban, 3
+logical variables: creating more by subscripting, 410; effect of global su=
bstitution, 393; effect of order in prenex, 394; effect of using multiple d=
ifferent, 393; explicitly placing in outer prenex, 400; for selbri, 409; im=
plicit placement in smallest enclosing bridi prenex, 400; notation conventi=
on, 392; when not in main bridi, 393; with multiple appearances in bridi, 3=
93; with poi, in multiple appearances, 396; with ro, in multiple appearance=
s, 396
+logically connected sentences: and DeMorgan's Law, 408
+logically connected tenses: definition, 363; expansion to sentences, 245; =
with JA, 245
+Loglan, 6
+logograms: words for, 416
+LOhO selma=92o, 499
+LOhU selma=92o, 476, 499; terminator for, 498
+lo'i, 125, 130, 447; as set counterpart of loi, 125; contrasted with lo an=
d loi, 125; relationship to lo'e, 126; with elided quantifiers, 447
+loi, 123, 130; as mass counterpart of lo, 123; contrasted with lei in spec=
ificity, 124; contrasted with lo and lo'i, 125
+Lojban: features of, 3; history of, 3; stability of, 4
+Lojban alphabet, 29
+Lojban letters: IPA for pronouncing, 30; list with IPA pronunciation, 30
+Lojbanistan, 4
+long ago and far away: example, 220
+long rafsi: definition, 57
+long rafsi form: compared with short form in effect on lujvo meaning, 56
+long-sword: example, 283
+lo'o, 454; effect of logical connective on elidability of, 454
+loose association: expressing with pe, 172
+Lord: example, 66
+lo-series cmavo: rationale for implicit inner quantifier, 130; rule for im=
plicit inner quantifier, 130
+lo-series description: caution on exact numbers as inner quantifiers on, 1=
31
+Lottie: example, 138
+lo'u, 141, 416, 476; interaction with bu, 416; interaction with zoi, 478
+love more: example, 260
+lower case letters: use in Lojban, 415
+lower-case: lerfu word for, 415
+lower-case letters: English usage contrasted with Lojban, 415; Lojban usag=
e contrasted with English, 415
+lower-case word: effect on following lerfu words, 415
+lu, 119, 141, 422, 476; contrasted with me'o for representing lerfu, 422
+LU selma=92o, 476, 499; terminator for, 499
+lu'a, 134; effect of on meaning, 134
+lu'e, 134, 264, 459, 478; as short for =93le sinxa be=94, 134; effect of o=
n meaning, 134
+LUhU selma=92o, 267, 499
+lu'i, 134; effect of on meaning, 134
+lujvo: abbreviated, 284; abstract, 286; algorithm for, 70; and consonant p=
airs, 59; and plausibility, 70; and seltau/tertau relationship, 276; and th=
e listener, 70; anomalous, 290; as a subtype of brivla, 53; as suppliers of=
 agent place, 295; asymmetric abstraction, 288; asymmetrical, 278; based on=
 multiple tanru, 70; cmavo incorporation, 274; comparatives, 292; compared =
with tanru, 273; consideration in choosing meaning for, 69; considerations =
for retaining elements of, 70; construction of, 56; definition, quick-tour =
version, 27; design consideration for relationship, 276; dropping elements =
of, 69; dropping NU in implicit abstractions, 288; dropping NU rafsi, 288; =
dropping SE rafsi, 283; examples of making, 72; from cmavo with no rafsi, 6=
0; from tanru, 55; fully reduced, 59; grammar of, 273; guidelines for place=
 structure, 273; implicit-abstraction, 288; interpreting, 276; invention of=
, 57; meaning drift of, 69; meaning of, 56; multiple forms of, 56; NU-dropp=
ing contrasted with SE-dropping, 288; place structure of, 273; place struct=
ure of figurative lujvo, 322; pro-sumti rafsi effect on place structure of,=
 163; quick-tour version, 20; rationale for, 273; recognizing, 59; rules fo=
r formation of, 56; scope abstraction in underlying veljvo, 287; scored exa=
mples of, 72; scoring of, 71; selection of best form of, 71; shorter for mo=
re general concepts, 70; summary of form characteristics, 59; superlatives,=
 292; symmetrical, 278; ultimate guideline for choice of meaning/place-stru=
cture, 69; unambiguity of, 69; unambiguous decomposition of, 56; unreduced,=
 57; unsuitability of for concrete/specific terms and jargon, 61; with =93j=
ai=94, 287; with zei, 60; zi'o rafsi effect on place structure of, 163
+lujvo creation: interaction of KE with NAhE, 286; interaction of KE with S=
E, 286; use of multiple SE in, 286
+lujvo form: consonant cluster requirement in, 59; final letter of, 59; hie=
rarchy of priorities for selection of, 72; number of letters in, 59; requir=
ements for hyphen insertion in, 59; requirements for n-hyphen insertion in,=
 60; requirements for r-hyphen insertion in, 60; requirements for y-hyphen =
insertion in, 59
+lujvo meaning, 274
+lujvo place order, 281; asymmetrical lujvo, 282; based on 3-or-more part v=
eljvo, 282; comparatives, 292; complex relation, 290; elliptical lujvo, 291=
; multi-part with NU, 287; non-overlapping place structures, 290; rationale=
 for standardization, 281; redundant non-first places, 290; superlatives, 2=
94; superlatives as exceptions, 294; symmetrical lujvo, 282
+lujvo place structure: =93ni=94 lujvo, 287; =93nu=94 lujvo, 286; basis of,=
 277; comparative lujvo, 292; cross-dependent places, 280; dependent places=
, 279; dropping cross-dependent places, 280; dropping dependent places, cav=
eat, 281; dropping dependent seltau places, 279; dropping dependent tertau =
places, 280; dropping first place of NU, 288; dropping KE, 285; dropping KE=
hE, 285; dropping redundant places, 276; effect of SE, 278; effect of SE-dr=
opping in tertau, 284; explicated walk-through, 276; guidelines, 273; multi=
-place abstraction lujvo, 287; notation conventions, 276; rationale for sta=
ndardization, 277; selecting tertau, 281; superlatives, 294; when first pla=
ce redundant with non-first, 278; when first places redundant, 278; when fi=
rst places redundant plus others, 278; with =93jai=94 lujvo, 287
+lukewarm food: example, 135
+lu'o, 134; effect of on meaning, 134
+lu'u, 133, 267; as elidable terminator for qualified sumti, 133
+Lyra: example, 138
+ma, 159, 249, 469; as sumti question, 159; for tense questions, 249; quick=
-tour version, 22
+ma'a, 146
+machine grammar, 511
+macron: proposed lerfu word for, 429
+magic square: example, 452
+magnitude: tense, 250
+MAhO selma=92o, 500; terminator for, 505
+ma'i, 224
+MAI selma=92o, 458, 474, 500; exception on use of boi before, 458
+male sexual teacher: example, 74
+man biting dog, 217
+man or woman: example, 333
+manhole: example, 218
+manysome: example, 447
+ma'o, 438, 460; potential ambiguity caveat, 460
+mai, 458, 474; contrasted with mo'o, 458
+man is woman: example, 177
+man-woman: example, 350
+Mao Zedong: example, 68
+maple sugar: example, 63
+maple trees: example, 63
+marathon, 258
+Mars road: example, 193
+Marsha: example, 470
+mass: compared with set as abstract of multiple individuals, 125; contrast=
ed with ordered sequence, 355; contrasted with set in attribution of compon=
ent properties, 125; contrasted with set in distribution of properties, 355=
; expressing measurement standard for indefinites, 446; expressing portions=
 of, 441; expressing relation with individuals forming, 446; expressing rel=
ation with set forming, 446; joining elements into a, 353; rule for implici=
t outer quantifier, 130
+mass contrasted with components: in properties of, 354
+mass name: use of, 124
+mass object: and logical reasoning, 123; as dependent on intention, 124; c=
ontrasted with multiple individual objects, 123; properties of, 123
+mass objects: peculiarities of English translation of, 124
+mathematical equality: expressing, 435
+mathematical expression: abbreviation notation, 431; definition (see also =
=93mekso=94), 431; referring to, 457
+mathematical expressions: connectives in, 361; implicit quantifier for, 14=
2; tensed connection in, 364
+mathematical expressions in tanru, 97
+mathematical inequalities: expressing, 439
+mathematical intervals, 362
+mathematical notation: and omitted operators, 431; and operator precedence=
, 436; forethought (see also Polish), 438; infix, 435; infix shortcomings, =
438; international uniqueness of, 431
+mathematical operators, 436
+mathematical parenthesis: left, 506
+mathematical texts: effect on lerfu shift scope, 423
+mathematical variables: lerfu strings as, 422
+mathematics: use of lerfu strings in, 422
+matne, 124
+matrix: as combination of vectors, 452; definition, 451; use as operand, 4=
52; use of parentheses with, 452; with ge'a for more than 2 rows/columns, 4=
52; with more than 2 dimensions, 452
+matrix column operator, 452
+matrix row operator, 452
+ma'u, 442; with elided number, 442
+mau, 203, 432; avoiding in favor of seme'a, 203
+Mayan mathematics: as a system with base larger than 16, 445
+me, 98, 424, 448; compared with du in effect, 99; effect of MOI on, 448; e=
xplicitly specifying, 325; place structure of, 98; used with names, 99
+ME selma=92o, 98, 448, 500; terminator for, 500
+me/du equivalence, 99
+me'a, 203; avoiding in favor of semau, 203
+measurement scale, 261
+measurements: expressing, 435
+meat slice: example, 285
+medieval weapon, 283
+MEhU selma=92o, 98, 500
+me'i, 443; with elided number, 443
+mei, 446; place structure formed for objective indefinites, 446
+mekso: and literary translation, 460; complex used as quantifier, 454; def=
inition, 431; design goals, 431; list of selma'o for, 461
+mekso chapter: completeness, 431; table notation convention, 431
+mekso goal: coverage, 431; expandable, 431; for common use, 431; for mathe=
matical writing, 431; precision, 431; unambiguous, 431
+mekso goals: and ambiguity, 431; and non-mathematical expression, 431; mat=
hematical notation form, 431
+melting, 295
+membership: property of sets, 125
+mental activity, 263
+mental discomfort: example, 307
+me'o, 142, 422, 457; compared with la'e lu, 422; contrasted with li, 457; =
contrasted with lu=85li'u for representing lerfu, 422; contrasted with quot=
ation for representing lerfu, 422; relation to li compared with la/zo relat=
ion, 457
+metalinguistic comment: with embedded discursive, 481
+metalinguistic commentary, 480
+metalinguistic erasers: within ungrammatical-Lojban quotation, 477
+metalinguistic insertions: marker for, 504
+metalinguistic levels, 481
+metalinguistic levels or reference, 481
+metalinguistic pro-sumti, 140; implicit quantifier for, 140
+metalinguistic words: quick-tour version, 25
+me'u, 98, 448, 449; relative precedence with logical connectives, 99
+mi, 119, 146
+mi'a, 146
+mice: example, 265
+mi'e, 146, 325; contrasted with other members of COI, 325; effect of order=
ing multiple COI, 325
+mi'enai, 326
+mi'i, 359, 455
+minimal list, 273
+mintu: contrasted with du, 163
+mi'o, 146
+mi-series: of pro-sumti, 146
+mi-series pro-sumti: lack of pro-bridi equivalent, 147
+misinterpretation, 286
+Mitsubishi: example, 420
+mi'u, 317; contrasted with go'i, 318
+mixed claim: definition, 394
+mixed modal connection: afterthought, 205; as proscribed in forethought, 2=
05; definition, 204; of bridi-tails, 205; of sentences, 204; of sumti, 205
+mixed with: example, 354
+mo, 160, 470; as selbri question, 160; compared with go'i in overriding of=
 arguments, 160; quick-tour version, 23
+mo'a, 442, 448
+modal bridi-tail connection, 200
+modal causals: implication differences, 197
+modal cmavo: basis in gismu place structure, 210; list of irregular deriva=
tion, 209; position relative to selbri, 104; regular form for derivation, 2=
08; table with English equivalents, 210
+modal cmavo table: format of, 210
+modal connection: simultaneous with logical, 204
+modal connection of selbri: using bridi-tail modal connection, 200
+modal connectives: fi'o prohibited in, 201
+modal conversion: access to original first place with fai, 206; grammar of=
, 206; place structure of, 206; with no modal specified, 206
+modal conversion with fi'o, 206
+modal conversion without modal: as vague, 206
+modal conversions: in descriptions, 206
+modal followed by selbri: compared with tanru modification in meaning, 202=
; contrasted with tanru modification in grammar, 202; effect on eliding cu,=
 201
+modal operand connection, 201
+modal place: definition, 195; on description selbri, 197; rationale for te=
rm name, 195; relation of to selbri, 195
+modal place relation: importance of first place in, 195
+modal sentence connection, 198; condensing, 200; effect on modal, 199; for=
ethought, 199; relation to modal of first sentence in, 199; relation to mod=
al of second sentence in, 199; table of equivalent schemata, 249; with othe=
r than causals, 199
+modal sumti: and FA marking, 195; as first place of modal tag selbri, 195;=
 definition (see also seltcita sumti), 195; effect on place structure, 195;=
 leaving vague, 201; position in bridi, 195; unspecified, 201
+modal sumti connection, 200
+modal tag: and sumti tcita, 94; contrasted with English preposition in pre=
ciseness, 196; definition (see also sumti tcita), 195; fi'o with selbri as,=
 194; for vague relationship, 197; short form as BAI, 195
+modal with no sumti: indicator for, 497
+modal-or-tense question: with cu'e, 250
+modals: compared with tenses in syntax, 248; contradictory negation of, 20=
6; contrasted with tenses in semantics, 248; expanding scope over inner mod=
al connection, 202; expanding scope over logical connection with ke=85ke'e,=
 202; expanding scope over multiple sentences with tu'e=85tu'u, 202; expand=
ing scope over non-logical connection, 202; for causal gismu, 197; importan=
ce of 1st sumti place for sumti tcita use, 248; improving relative phrase p=
reciseness with, 203; making long-scope, 207; making sticky, 207; negation =
of, 206; scalar negation of, 207; termset connection, 200
+modals often attached with relative phrases: list, 204
+modified: of a tanru, 274
+modifier: of a tanru, 274; seltau as, 84
+modifying brivla (see also seltau), 55
+mo'e, 456; terminator for, 456
+MOhE selma=92o, 500; terminator for, 505
+MOhI selma=92o, 224, 501
+mo'i, 224
+Mon Repos: example, 280
+mo'o, 458, 474; contrasted with mai, 458
+moi, 447
+MOI selma=92o, 98, 446, 449, 463, 500; list of cmavo in, 463; use of boi b=
efore, 449
+more: English word, expressing with relative phrases, 203; English word, i=
mportance of relative phrase to, 204
+more than: contrasted with less than, at least, at most, 443; example, 443
+morphological glue, 507
+morphology: conventions for, 49; definition, 49; derivational, 49; simplic=
ity of, 49; symbolic conventions for discussing, 49
+mother father: example, 55
+movement: order in tense constructs, 225; time, 225; with multiple directi=
ons, 225
+movement specification: interaction with direction in tenses, 224
+mu'a, 319
+mu'e, 257, 258; place structure, 259
+mu'i, 197
+multiple compound bridi: restriction on ke, 346
+multiple conversion: avoiding, 194; effect of ordering, 194
+multiple indefinite sumti: effect of re-ordering in sentence, 399; express=
ing with equal scope, 399; meaning, 398
+multiple indefinite sumti scope: in termset, 399
+multiple indicators, 312
+multiple individual objects: contrasted with mass object, 123; meaning of,=
 123
+multiple letters: proposed lerfu words for, 429
+multiple logical connectives: within tanru, 91
+multiple ma: as multiple questions, 160
+multiple mo: as multiple questions, 160
+multiple quantification: effect on selbri placement among sumti, 407
+multiple questions in one bridi: expressing, 160
+multiple relative clauses: attaching with zi'e, 175; connecting different =
kinds with zi'e, 176
+multiple SE: effect of ordering, 194
+multiple speakers, 484
+multiple sumti in one place: avoiding, 191; meaning, 191
+multiple tanru inversion: effect on grouping, 96
+multiple tenses: effect of order in sentence, 235
+multiplication: explicit expression of, 437; implicit expression of, 437
+mu'o, 325
+my: example, 180
+my chair: example, 176
+myth: example, 316
+n people: example, 423
+na, 104, 338, 346, 350, 401, 408; and negation boundary, 408; order in log=
ical connectives with se, 338
+na and tense: multiple, 104
+NA selma=92o, 501
+na writing convention: in eks, 341
+na.a, 341
+na'a, 418; contrasted with lo'a, 418
+na'e, 207; before gu'e, 103; contrasted with na'e ke, 102
+na'ebo, 135
+nago'i: quick-tour version, 24
+NAhE selma=92o, 101, 133, 182, 242, 459, 501; effect of relative clause pl=
acement with, 182
+NAhE+BO: terminator for, 499
+NAhU selma=92o, 501; terminator for, 505
+na'i, 321
+NAI selma=92o, 501
+naicai, 305
+nairu'e, 305
+naisai, 305
+naku, 401; as creating a negation boundary, 405; compared with sumti in gr=
ammar, 405; effect on moving quantifiers, 405; in linked sumti places, 407;=
 multiple in sentence, 407; outside of prenex, 405
+naku negation: rationale for considering an advanced technique, 406
+naku negation boundary: effect on conversion with se, 406
+naku su'oda: as expansion of noda, 403
+naku zo'u, 408; and negation boundary, 408
+name equivalent for =93typical=94: rationale for lack of, 127
+name words: recognition of, 137
+names: algorithm for, 66; alternatives for restricted sequences in, 66; as=
 possessive sumti, 180; assigning with goi, 152; authority for, 65; borrowi=
ng from other languages, 138; examples of, 64; from Lojban words, 66; in vo=
cative phrase, 137; multiple, 138; non-Lojban, 479; pause requirement in le=
rfu words, 416; purpose of, 64; quick-tour version, 13; rationale for lojba=
nizing, 64; requirement for pause after, 66; restrictions on form of, 65; r=
ules for, 66; rules for formation, 65; stress in, 65, 66; stress on, 40; tw=
o kinds of, 137; unusual stress in, 65; uses of, 137; using rafsi, 138; wit=
h LA descriptor, 137; with zo versus la, 478
+names from vowel-final base: commonly used consonant endings, 138
+names in Lojban (see also cmene), 64
+names with la: implicit quantifier for, 139
+name-words: limitations on, 138; pause requirements before, 138; permissib=
le consonant combinations, 138
+naming predicate, 121
+na'o, 226
+natural end: continuing beyond, 230; contrasted with actual stop, 229
+na'u, 456; terminator for, 456; use in asking operator questions, 457
+nai, 206, 226, 241, 299, 338, 340, 346, 350, 358, 360, 361, 408; effect on=
 intervals, 360; effect on joiks, 358; placement in afterthought bridi conn=
ection contrasted with forethought, 339; placement in forethought bridi con=
nection contrasted with afterthought, 339
+name descriptor, 498
+NATO: example, 424
+nau, 238; effect on sticky tenses, 238; syntax, 238
+Navajo: example, 64
+ne, 174, 203; compared with pe, 174
+near the park: example, 232
+nearby in time: example, 220
+necessary detour, 274
+Nederlands: example, 65
+need any box, 400
+negated interval: meaning of, 360
+negating a forethought-connected bridi-tail pair, 347
+negating a forethought-connected sentence pair, 347
+negating a sentence: and truth value, 333
+negation: complex examples, 102; form for emulating natural language negat=
ion, 405; of operand, 459; of operator, 459; of tenses, 241
+negation and logical connectives: caveat for logic chapter discussions, 40=
3
+negation between sentences: compared with bridi negation, 404; meaning of,=
 404
+negation boundary: and zero, 402; effect of moving, 402; forming, 497
+negation cmavo: position relative to selbri, 104
+negation in prenex: effects of position, 401
+negation manipulation: =93na=94 contrasted with =93naku=94 in difficulty o=
f, 407; =93naku=94 contrasted with =93na=94 in difficulty of, 407
+negation of fi'o modals: by negating selbri, 207
+negation of modals, 206; contradictory, 206; scalar, 207
+negation of tenses: meaning of, 241
+negation operator: contrasted with negative sign, 438; contrasted with sub=
traction operator, 438
+negation sumti qualifiers: meanings of, 135
+negations with logical connectives: effects on expansion of sentence, 407
+negative answer: quick-tour version, 24
+negative numbers: expressing, 432
+negative sign: contrasted with negation operator, 438; contrasted with sub=
traction operator, 434
+negator: contradictory, 501; movement from bridi to sumti, 408; scalar, 50=
1; single-word, 501
+nei, 155
+-ness, 259
+new notation, 276
+New York city: example, 174
+New York state: example, 174
+Newport News: example, 138
+news: example, 467
+-ng: Lojban contrasted with English, 41
+n-hyphen: contrasted with r-hyphen in requirements for use, 60; use of, 56=
, 60
+ni, 261, 262
+ni'e, 456; terminator for, 456
+ni'enu'a, 460
+NIhE selma=92o, 501; terminator for, 505
+NIhO selma=92o, 466, 467, 502; quick-tour version, 16
+ni'i, 197
+ni'o, 466; effect on pro-sumti/pro-bridi assignments, 162; quick-tour vers=
ion, 16
+ni'u, 432, 438, 442; contrasted with va'a and vu'u, 438; with elided numbe=
r, 442
+no quantifier: expanding, 403
+no'a, 155; contrasted with other members of go'i-series in possible refere=
nts, 155
+nobody: ambiguous interpretations of, 391; interpretation of, 391; Lojban =
contrasted with English, 391
+noda: expanding to naku su'oda, 403
+no'i, 467; effect on pro-sumti/pro-bridi assignments, 162
+noi, 171, 203
+NOI selma=92o, 169, 502; terminator for, 498
+noisy environments: proposed lerfu words for, 429
+nonagenarian, 293
+nonce word: marking, 489
+non-logical connection: and elidability of terminators, 354; in mathematic=
al expressions, 361; in tanru, distinguishing from connection of sumti, 354=
; of individuals into mass, 355; of individuals into set, 355; of modals, 2=
08; of operands, 455; of operators, 455; of sumti, distinguishing from conn=
ection in tanru, 354; of termsets, 357
+non-logical connectives: effect of nai on, 358; grouping, 357; including t=
ense, 364; intervals, 359; ordered intervals, 359; sentence, 358; syntax ru=
les summary, 366; un-ordered intervals, 359; within tanru, 91
+non-logical forethought termsets: connecting tagged sumti, 358
+non-logically connected tenses, 363
+non-Lojban quotation, 141
+non-Lojban text: rules for pause with, 69
+non-restrictive relative clause: definition (see also incidental relative =
clause), 171
+non-specific descriptions, 121
+non-standard orthographies: caveat, 46; Cyrillic, 46; Tengwar, 46
+non-standard words: marking, 480
+no'o, 441
+normal circumstances, 256
+notation conventions: for Quick Tour chapter, 12
+nothing sits: example, 401
+no'u, 174; compared with po'u, 174; contrasted with po'u, 175
+nouns: brivla as Lojban equivalents, 52
+Nth rat: example, 423, 448
+Nthly: example, 423
+nu, 256, 259, 261; definition, 256; place structure, 257
+NU compared with ZAhO, 268
+NU selma=92o, 98, 255, 256, 257, 262, 263, 268, 365, 502; syntax, 255; ter=
minator for, 497
+nu'a, 97, 457; use in answering operator questions, 457
+nu'e, 324
+NUhA selma=92o, 502
+NUhI selma=92o, 348, 399, 502
+NUhU selma=92o, 348, 399, 502; terminator for, 502
+nu'i, 200, 348, 399
+null operand: for infix operations with too few operands, 450
+null operator: for infix operations with too many operands, 451
+number article: explanation of use, 435
+number questions, 449; answers to, 449
+number sumti: syntax of, 141; with li, 141; with li contrasted with me'o, =
142; with me'o, 142; with me'o contrasted with li, 142
+number words: pattern in, 432
+numbers: as compound cmavo, 432; as grammatically complete utterances, 449=
; as possessive sumti, 180; cmavo as Lojban equivalents, 50; descriptor for=
, 499; English contrasted with Lojban on exactness, 397; expressing simple,=
 432; greater than 9, 432; implicit quantifier for, 142; indefinite, 440; l=
ist of indefinite, 463; list of special, 462; Lojban contrasted with Englis=
h on exactness, 397; meaning when used as quantifiers, 127; on logical vari=
ables, 397; rafsi for, 59; special, 434; talking about contrasted with usin=
g for quantification, 435; using for quantification contrasted with talking=
 about, 435
+numeric digits in lerfu words: grammar considerations, 420
+numerical punctuation, 433; undefined, 434
+numerical selbri: alternative to compensate for restriction on numbers, 44=
8; based on non-numerical sumti, 448; complex, 448; grammar, 448; restricti=
on on numbers used for, 448; special, 446; special, with lerfu strings, 448=
; use of =93me=94 with, 448
+numerical tenses: effect on use of boi, 458
+nu'o, 244
+nu'u, 200, 348, 399
+ny, 437
+NYC: example, 424
+observation: example, 316
+observation evidential: contrasted with observative, 316
+observative: contrasted with observation evidential, 316; definition, 188
+observative form: contrasted with command, 188
+observative with elided CAhA: convention, 245
+observatives: and abstractions, 255; quick-tour version, 15
+ocean shell: example, 286
+octal system: specifying numbers in (see also base), 444
+octogenarian, 293
+o'ecu'i: example, 300
+office or ice-dance: example, 347
+o'i: example, 300
+Old McDonald: example, 32
+old topic, 466
+omitting terminators: perils of, 102
+on right: contrasted with toward right, 224
+on two occasions: example, 246
+on verge: example, 228
+once: example, 226, 458
+one-third of food, 447
+only if: compared with if =85 then, 338
+o'onai: example, 300
+of: in English, compared with do'e, 197
+oi: example, 300
+omission of descriptor: effect on ku, 132
+once and future king: example, 363
+One: the, example, 66
+only: example, 318
+only once: example, 227
+open interval, 360; expressed with mi'i, 455
+opening quotation, 476
+operand: converting from operator, 460; converting into operator, 460; con=
verting selbri into, 456; converting sumti into, 456
+operand connection: afterthought, 453; forethought, 453
+operand modal connection, 201
+operands: connecting, 361; contrasted with general sumti, 436; too few for=
 infix operation, 450; too many for infix operation, 451
+operator: converting from operand, 460; converting into operand, 460; conv=
erting into selbri, 457; converting selbri into, 456; forethought marker, 5=
03
+operator =85ku'e in Polish notation: contrasted with vei =85ve'o, 438
+operator connection: afterthought, 453; forethought, 453
+operator derived from selbri: effect of selbri place structure on, 456
+operator left-right grouping: as Lojban default, 436
+operator precedence: and mathematical notation, 436; effect of pragmatic c=
onvention, 436; generalized explicit specification, 437; in Lojban default,=
 436; plans for future, 458; rationale for default left-grouping, 436; scop=
e modification with bi'e, 437; specifying by parenthesis, 437
+operator precedence in other languages, 436
+operator priority, 490
+operators: analogue of tanru in, 361; connecting, 361; list of simple, 461=
; mathematical, 507
+operators of VUhU: grammar of operands, 436
+opinion: example, 317
+opposite-of-minus: example, 459
+or=94: =93and/or=94 contrasted with =93either =85 or =85 but not both, 334
+order of variables: in moving to prenex, 398
+ordered sequence: by listing members, 355; contrasted with mass, 355; cont=
rasted with set, 355
+ordinal selbri: definition, 447; place structure, 447; place structure eff=
ect from subjective numbers, 448
+ordinal tense, 230
+ordinals: utterance, 474
+orthography: non-standard, 45; relation to pronunciation, 29
+o'u: example, 300
+outer product, 452
+outer quantifier: contrasted with inner quantifier, 129; definition, 129; =
effect of on meaning, 129; for expressing subset, 131; implicit on descript=
ors, 129; in indefinite description, 132; rationale for differences in impl=
icit quantifier on descriptors, 131
+outer quantifier of sumti: meaning of, 178
+outer sumti: prenex for referring to from within relative clause within re=
lative clause, 185; referring to from within relative clause within relativ=
e clause, 184
+over-dot: proposed lerfu word for, 429
+over-ring: proposed lerfu word for, 429
+owe money: example, 346
+p =3D x =3D z: example, 439
+pa, 433
+PA selma=92o, 397, 432, 440, 449, 502; exception on use of boi with MOI, 4=
49; members with rafsi, 460; terminator for, 490
+pa'e, 318, 319
+pa'enai, 319
+paragraph marker, 502
+paragraph separation: spoken text, 467; written text, 466
+paragraphs: effects on scope, 466; separating, 466; separator, 466
+parasitic worms: example, 286
+parentheses: for complex mekso used as quantifier, 454
+parenthesis: discourse, 505; mathematical, 437; textual, 480
+partial quotation, 321
+parts of speech, 50
+passive voice, 16
+past event: possible extension into present, 223
+pastward: as a spatial tense, 224
+paternal grandmother: example, 55
+pau, 322; placement in sentence, 322
+paunai, 322
+pausative event contour, 228
+pause: and cmene, 68; and consonant-final words, 68; and Cy-form cmavo, 69=
; and final-syllable stress, 69; and non-Lojban text, 69; and vowel-initial=
 words, 68; between words, 68; contrasted with stop, 229; contrasted with s=
yllable break, 32; proscribed within words, 68; representation of in Lojban=
, 31; requirement between stressed syllables, 52; symbol for, 416; word for=
, 416
+pause before name: effect of doi, 323; effect of vocatives of COI, 323
+pauses: before vowels, 52; rules for, 68
+pe, 172, 180, 203; as loose association, 172; compared with ne, 174; compa=
red with poi ke'a srana, 172; contrasted with po, 173
+pe'a, 322
+peace symbol, 425
+pe'anai, 322
+pe'e, 347
+PEhE selma=92o, 347, 503
+PEhO selma=92o, 503; terminator for, 497
+pe'i, 317
+pe'ipei, 317
+pe'o, 439
+pei, 313
+percent: as numerical punctuation, 433
+perfective event contour, 228
+perils of omitting terminators, 102
+period: definition of, 31; example of, 32; optional, 32; quick-tour versio=
n, 12; within a word, 32
+permissions notice, 8
+Persian rug: example, 60
+personal pronouns: with ko'a-series for he/she/it/they, 150; with mi-serie=
s for I/you, 146
+personal pronouns for he/she/it/they: English contrasted with Lojban in or=
ganization, 150
+personal pro-sumti, 139; implicit cancellation of by change of speaker/lis=
tener, 162; implicit quantifier for, 128, 139; stability of, 162
+person's arm: example, 173
+Pete: example, 65
+pe'u, 324; contrasted with e'o, 324
+Pheidippides, 258
+phoneme stream, 477
+phonetic alphabet, 29; proposed lerfu words for, 429
+physical distress: example, 307
+pi, 130, 433, 441, 442, 444; effect on indefinite numbers, 441
+pi'a, 452
+piano-moving: example, 123
+pictures: captions to, 7; credits for, 6
+pi'e, 445
+pi'i, 436
+pinyin: as a basis for Chinese characters in Lojban lerfu words, 420
+pi'o, 195
+piro, 130; explanation of meaning, 130
+pisu'o: explanation of meaning, 130
+pi'u, 246, 354, 356; contrasted with .e, 357; use in connecting tenses, 24=
6
+place number: specifying, 493
+place of eating: example, 247
+place structure: adding new places to with modal sumti, 194; definition, 1=
2, 187; effect of FA on, 190; effect of modal conversion on, 206; empty slo=
ts in, 187; explicitly mapping sumti to place with FA, 190; gismu, 294; ins=
tability of, 187; leaving a sumti place unspecified in with zo'e, 189; nota=
tion conventions, 187; omitting places with FA, 190; omitting places with z=
o'e, 189; re-ordering by conversion, 100
+place structure and tanru inversion, 95
+place structure of selbri: determining, 187
+place structure order: effect of FA on, 190
+place structure questions, 191
+plant grows: example, 197, 207
+plants: use of fu'ivla for specific, 61
+plausibility: in abbreviated lujvo, 284
+playgrounds, 281
+pleases, 20
+plural: Lojban contrasted with English in necessity of marking, 120; Lojba=
n equivalent of, 443; meaning of le with, 123
+plural masses: possible use for, 130
+plus negative of: example, 438
+pluta, 193; contrasted with ve klama, 193
+po, 173; as restrictive possession, 173; compared with poi ke'a se steci s=
rana, 173; contrasted with English possession, 173; contrasted with pe, 173=
; contrasted with po'e, 173
+po'e, 173; as intrinsic possession, 173; compared with poi ke'a jinzi ke s=
e steci srana, 173; contrasted with po, 173
+point: event considered as, 230
+point-event abstraction: place structure, 259
+point-event abstraction(s): definition, 258; related tense contours, 269
+point-event abstractor, 258
+pointing cmavo: quick-tour version, 13
+police lineup, 449
+Polish notation: and mekso goals, 431; and use of boi, 438; definition, 43=
8; end-of-operands indicator, 438; explicitly marking as, 439; operands wit=
h infix expressions, 439; operator =85ku'e compared with parenthesization, =
438; separating operands in, 438; vei =85ve'o contrasted with operator =85k=
u'e, 438
+Polish notation mixed with infix, 455; example, 455
+politeness: thank you and you're welcome, 324; you're welcome, 324, 325
+po'o, 317; placement in sentence, 318
+poi, 169, 203, 394, 396; contrasted with voi in veridicality, 177; discuss=
ion of translation, 170; dropping from multiple appearances on logical vari=
ables, 396; syntax of, 169
+pointing: reference by, 147
+portion: on set contrasted with on individual, 131
+portion of whole: expressing, 441
+portion selbri: definition, 447; place structure, 447; place structure eff=
ect from subjective numbers, 448
+positive numbers: explicit expression, 432
+positive sign: contrasted with addition operator, 436
+possessed in relative phrases: compared with possessor, 174
+possession: expressing with po, 173; intrinsic, expressing with po'e, 173;=
 Lojban usage compared with French and German in omission/inclusion, 175; L=
ojban usage contrasted with English in omission/inclusion, 175; quick-tour =
version, 21
+possession not ownership: quick-tour version, 21
+possessive sumti: compared with relative phrase, 180; contrasted with rela=
tive phrases in complexity allowed, 180; definition, 180; effect on elidabi=
lity of ku, 181; relative clauses on, 181; syntax allowed, 180; with relati=
ve clauses on possessive sumti, 181
+possessive sumti and relative clauses: development history, 180
+possessive sumti with relative clauses: effect of placement, 181
+possessor in relative phrases: compared with possessed, 174
+potential: expressing in past/future, 244
+potential events: expressing implicitly, 243
+po'u, 174; as identity, 174; compared with no'u, 174; compared with poi ke=
'a du, 174; contrasted with no'u, 175; relative phrase of contrasted with r=
elativized sumti of, 174
+prayer: example, 281, 290
+precedence: mathematical default, 436
+precise erasures, 483
+predicate answers, 470
+predication: as a relationship, 11; compared with bridi, 11
+predication abstraction, 262
+Preem Palver: example, 98
+pregnant sister: example, 320
+prenex: considerations for dropping, 395; dropping for terseness, 397; eff=
ect of order of variables in, 396; explanation, 392; internal to a bridi, 4=
00; purpose of, 396; removing when numeric quantifiers present, 397; syntax=
 of, 392; use for outer sumti reference, 185
+prenex manipulation: exporting na from left of prenex, 405; importing na f=
rom selbri, 405; moving naku past bound variable, 405; rules, 405
+prenex marker, 508
+prenex scope: for sentences joined by .i, 410; for sentences joined by ije=
ks, 410; in abstractions, 410; in embedded bridi, 410; in relative clauses,=
 410; informal, 410
+prepositions: cmavo as Lojban equivalents, 50
+pretty: English ambiguity of, 87
+pretty little girls' school: forty ways, examples, 112
+previous topic, 467
+pride of lions: example, 456
+primitive roots: gismu as, 53
+principle of consistency: of logical-if statements, 337
+probability .5: example, 447
+probability selbri: definition, 447; place structure, 447; place structure=
 effect from subjective numbers, 448; values, 447
+pro-bridi: as abbreviation for bridi, 151; broda-series, 151; broda-series=
 list, 165; bu'a-series list, 165; compared to pro-sumti as means of abbrev=
iation, 145; definition, 145; go'i-series list, 165; list by series, 165; l=
ist of miscellaneous cmavo used with, 166; miscellaneous list, 165; overrid=
ing sumti of antecedent bridi for, 151; quotation of, 476; scope effect of =
new paragraph, 466
+pro-bridi assignment: explicit cancellation of with da'o, 162; no'i effect=
 on, 162; stability of, 162
+pro-bridi rafsi: as producing context-dependent meanings, 164
+pro-bridi update: flag for, 503
+process abstraction: place structure, 259
+process abstraction(s): definition, 258; related tense contours, 268
+process abstractor, 258
+process event: described, 258
+pronouncement: example, 316
+pronouns: as anaphora, 152; compared to pro-sumti in usage as abbreviation=
s, 145
+pronouns in English: as independent of abbreviations, 145; as noun abbrevi=
ations, 145
+pronunciation: IPA for Lojban, 30; quick-tour version, 12; relation to ort=
hography, 29; standard, 29
+properties: place structure, 261
+property abstraction, 259; use of multiple ce=92u for relationship abstrac=
tion, 260
+property abstraction(s): contrasted with amount abstraction, 261; specifyi=
ng determining place by sumti ellipsis, 259; specifying determining place w=
ith ce'u, 260; specifying sumti place of property with ce'u, 161; sumti ell=
ipsis in, 259
+property abstractor, 259
+property description, 259
+property of loving: example, 260
+proposed law, 283
+proposed lerfu words: as working basis, 426
+propositional: of attitudinals, 301
+propositional attitudes, 262; compared with knowledge discursives, 319
+pro-sumti: and discursive utterances, 481; as possessive sumti, 180; class=
es of, 139; compared to pro-bridi as means of abbreviation, 145; compared t=
o pronouns in usage as abbreviations, 145; contrasted with description, 119=
; da-series list, 165; definition, 145; di'u-series, 148; di'u-series list,=
 164; for listener(s), 146; for listeners and/or speakers and/or others, 14=
6; for relativized sumti in relative clauses, 160; for speaker(s), 146; imp=
licit quantifier for, 139; ko'a-series, 150; ko'a-series list, 164; lerfu a=
s, 152; lerfu string, effect on reference to lerfu itself, 422; lerfu strin=
g, interaction with quantifier and boi, 421; list by series, 164; list of m=
iscellaneous cmavo used with, 166; miscellaneous list, 165; mi-series, 146;=
 mi-series list, 164; quick-tour version, 13; quotation of, 476; rafsi for,=
 163; referring to place of different bridi with go'i-series, 159; referrin=
g to place of same bridi with vo'a-series, 158; ri-series list, 164; scope =
effect of new paragraph, 466; series, 145; ti-series, 147; ti-series list, =
164; typical, 157; unspecified, 157; vo'a-series, 158; vo'a-series list, 16=
5; zo'e-series list, 164
+pro-sumti assignment: explicit cancellation of with da'o, 162; no'i effect=
 on, 162; stability of, 162
+pro-sumti for speaker/listener/others: as masses, 146; relation to joi, 14=
6
+pro-sumti for utterances, 148
+pro-sumti for we: contrasted with English we, 146
+pro-sumti rafsi: anticipated use of for abbreviating inconvenient forms, 1=
63; effect of on place structure of lujvo, 163
+protocol: computer communications using COI, 326; parliamentary using COI,=
 326; using vocatives, 326
+pu, 219, 232; meaning as a sumti tcita, 232; meaning when following interv=
al specification, 222
+pu ge, 365
+PU selma=92o, 219, 227, 242, 362, 503; compared with FAhA, 219; contradict=
ory negation of, 241
+PU tenses: contrasted with ZAhO tenses in viewpoint, 228
+pu'i, 244
+punctuation, 297; in numbers, 433; list of numerical, 462
+punctuation lerfu words: interaction with different alphabet systems, 420;=
 mechanism for creating, 419; rationale for lau, 419
+punctuation marks: cmavo as Lojban equivalents, 50
+pu'o, 228; as pastward of event, 229; derivation of word, 228; explanation=
 of derivation, 229
+pu'u, 257, 258, 268; place structure, 259
+quack: example, 417
+quadratic formula: example, 455
+qualified sumti: contrasted with unqualified sumti, 133
+quality and quantity: example, 365
+quantification: before description sumti compared with before non-descript=
ion sumti, 129
+quantificational pro-sumti, 139; implicit quantification rules, 139
+quantified space, 228
+quantified sumti: different types contrasted for scope for distribution, 3=
99; relative clause scope with, 178
+quantified temporal tense: definition, 226; negating with nai, 227
+quantified temporal tense with direction: Lojban contrasted with English i=
n implications, 227
+quantified temporal tenses: =93once=94 contrasted with =93only once=94, 22=
7; caveat on implication of, 227
+quantified tenses: as sumti tcita, 233
+quantifier, 502; effect of moving naku, 405; explicit on sumti, 127; lerfu=
 string as, 423; on previously quantified variable, 410; on sumti, effect o=
n relative clause, 178; on sumti, expressing inexact amount with, 127; on s=
umti, indicating exact number, 127; with logical variables, 397; with sumti=
, 127
+quantifier scope: in multiple connected sentences, 404
+quark: example, 63
+question pro-sumti, 140; implicit quantifier for, 140
+questions, 469; answering with go'i, 154; connection, 351; digit, 449; fil=
l-in-the-blank, 469; indirect, 264; marking in advance, 322; modal, 492; mu=
ltiple, 470; number, 449, 470; operator, 457; place structure position, 191=
; quick-tour version, 22; rhetorical, 322; selbri, 160, 470; sumti, 159, 46=
9; truth, 469; with "xu, 321
+quick runner: example, 84
+quotation, 475; any text, 508; as possessive sumti, 180; contrasted with m=
e'o for representing lerfu, 422; contrasted with sentence abstraction, 263;=
 delimited, 508; four kinds, 141; grammatical, 499; implicit quantifier for=
, 128, 141; of grammatical Lojban text, 476; of Lojban words, 499; of non-L=
ojban, 477; of parseable Lojban text, 476; of rafsi, 478; of single word, 4=
77; of ungrammatical Lojban text, 476; referent versus symbol, 478; single-=
word, 508; ungrammatical Lojban containing le'u, 477; ungrammatical Lojban =
containing lo'u, 477
+ra, 153; practical referent conventions, 153
+radio communication: proposed lerfu words for, 429
+radix: decimal (see also base), 444
+ra'e, 433
+rafsi: as fu'ivla categorizer, 61; based on pro-sumti, 163; considerations=
 restricting construction of, 58; contrasted with cmavo in usage, 61; contr=
asted with same-form cmavo in meaning, 56; contrasted with words, 61; conve=
ntional meaning for cu'o, 460; conventional meaning for frinu, 460; definit=
ion, 56; definition, quick-tour version, 27; forms of, 57; four-letter, req=
uirement for y-hyphen, 60; lack of, effect on forming lujvo, 60; level of u=
niqueness of relation to gismu, 57; long, 57; multiple for each gismu, 69; =
multiplicity of for single gismu, 57; possible forms for construction of, 5=
8; quotation of, 478; rationale for assignments of, 58; rules for combining=
 to form lujvo, 56; selection considerations in making lujvo, 57; short, 57=
; uniqueness in gismu referent of, 57; use of, 57
+rafsi assignments: non-reassignability of, 58
+rafsi for numbers, 59
+rafsi form: effect of choice on meaning of lujvo, 56
+rafsi fu'ivla proposal, 80
+rafsi space, 58
+RAhO selma=92o, 503
+Ralph: example, 393
+ra'o, 156; for reinterpreting go'i-series pro-bridi sumti references, 156
+rat eats cheese: example, 227, 232
+rat eats cheese in park: example, 247
+rats are brown: example, 125
+rats in park: example, 446
+ra'u, 319; scale of importance, 320
+rau, 442, 448
+real world: contrasted with hypothetical world, example, 320
+real world point of view, 320
+Received Pronunciation, 42
+reciprocal: expression of mathematical, 433
+reciprocal pro-sumti, 158
+reciprocity: expressing with soi, 159; expressing with vo'a-series pro-sum=
ti and soi, 159; marking, 504
+recital rooms, 281
+Red Pony: example, 133, 182
+redundancy: effect on vocative design, 323
+re'e, 307
+re-evaluation of referents: flag for, 503
+reference: ambiguity of ti/ta/tu, 169; and discursive utterances, 481; qui=
ck-tour version, 20; to relativized sumti with ke'a, 169; use of relative c=
lause for, 169
+reference frame: specifying for direction tenses, 224
+reference frame for directions in tenses, 224
+reference grammar, 3
+referent: contrasted with symbol, 478; of operand, 459; referring to with =
la'e, 134
+referent of pro-bridi: definition, 145
+referent of pro-sumti: definition, 145
+reflexive pro-sumti, 139, 158; stability of, 162
+regularly: example, 226
+re'i, 325
+re'imi'e, 325
+relation of first places in logical connection of observatives: rationale,=
 345
+relationship: active/static/attributive compared, 11; as basis of sentence=
, 187; objects of, 187
+relationship abstraction, 260
+relative clause: as part of name, 179; compared with tanru, 172; connectin=
g to relative phrase with zi'e, 176; connecting to whole sumti, 506; contra=
sted with tanru, 172; effect of commas in English, 171; effect of elided ku=
 of relativized sumti, 177; effect of omission of ke'a on, 170; effect of r=
elativized sumti quantifiers on, 178; effect on elidability of be'o, 94; im=
pact of indefinite sumti on placement, 180; impact of la on placement, 179;=
 impact of LAhE on placement, 182; impact of le on placement, 179; impact o=
f lo on placement, 179; impact of NAhE on placement, 182; kinds of, 171; li=
st of cmavo for, 185; on connected sumti, 182; on names, 179; on number, 18=
1; on possessive sumti, 181; on quotation, 182; on vocative phrases, 184; p=
lacement in sentence, 177; placement with vocative phrases, 184; relative c=
lauses within, 184; restricted contrasted with incidental, 171; restricted =
contrasted with incidental in English expression, 171; restrictive (see als=
o restrictive relative clause), 171; syntax with indefinite sumti, 180; use=
 for reference, 169; use in restricting existential claims, 394; use in res=
tricting universal claims, 395; use of ke'a for referral to relativized sum=
ti in, 160
+relative clause after descriptor: effect on elidability of ku'o, 178
+relative clause after relativized sumti ku: meaning, 178
+relative clause after sumti: as common placement in sentence, 177
+relative clause and indefinite sumti: placement considerations, 180
+relative clause and LAhE: placement considerations, 182
+relative clause and le-sumti: placement considerations, 179
+relative clause and lo-sumti: placement considerations, 179
+relative clause and NAhE: placement considerations, 182
+relative clause and names: placement considerations, 179
+relative clause and possessive sumti: development history, 180
+relative clause and quantified sumti: placement considerations, 178
+relative clause before inner quantifier: meaning, 178
+relative clause before relativized sumti ku: meaning, 178
+relative clause marker, 502
+relative clause on complex sumti: Lojban contrasted with English, 183
+relative clause on indefinite sumti: syntax considerations, 180
+relative clause on lo: syntax suggestion, 179
+relative clause placement: considerations for lo-sumti contrasted with le-=
sumti, 179; considerations for simple descriptors contrasted with for quant=
ified sumti, 179; effect on scope, 178; English contrasted with Chinese and=
 Finnish, 178; on sumti with simple descriptor, 178
+relative clause scope: extending to preceding sumti with vu'o, 182; with q=
uantified relativized sumti, 178
+relative clause with possessive sumti: effect of placement, 181
+relative phrase: as an abbreviation of a common relative clause, 172; comp=
ared with possessive sumti, 180; connecting to relative clause with zi'e, 1=
76; contrasted with possessive sumti in complexity allowed, 180; contrasted=
 with relative clause in preciseness, 203; improving preciseness with modal=
s, 203; rationale for, 172; syntax of, 172
+relative phrase marker, 495
+relative phrases with modals: compared to relative clauses in preciseness,=
 203
+relative pro-sumti, 140
+relativity theory: relation to Lojban tense system, 220
+relativized sumti: definition, 169; in relative clauses within relative cl=
auses, 184
+remembered: example, 316
+re-ordering logical variables with se, 396
+repeating decimals: expressing with numerical punctuation, 433; marking st=
art of repeating portion, 433
+replace: example, 289
+representing lerfu: lu contrasted with me'o, 422
+respectively: example, 356; specifying with fa'u, 356; with different rela=
tionships, 358
+restricted claims: definition, 394
+restricted variable: compared with indefinite description, 398
+restrictive relative clause: definition, 171
+restrictive relative clauses: non-veridical using voi, 177; veridical usin=
g poi, 177
+resume: contrasted with begin, 229
+resumptive event contour, 228
+re'u, 230
+revelation: example, 316
+reverse Polish notation: and mekso goals, 431; definition, 452; indicator,=
 494; marker, 452; number of operands, 453; operands of, 453; parentheses i=
n operands of, 453; terminator, 452; use of parentheses in, 452; with too f=
ew operands, 453; with too many operands, 453
+reviewers of this book, 6
+rhetorical question, 322
+r-hyphen: contrasted with n-hyphen in requirements for use, 60; use of, 56=
, 60
+ri, 152; contrasted with ke'a in relative clauses, 161; non-self-reference=
 of, 153; referent of, 152; subscripting for referring further back, 153
+ri'a, 197
+rich and German: example, 356
+ri'e, 309
+righteous indignation: example, 309
+right-grouping in tanru: with bo, 87
+right-grouping rule: definition of, 87
+ri-series pro-sumti, 152; and order of possible referents, 153; assigning =
for permanent reference, 154; compared with ti-series in word formation, 15=
2; effect of ko'a-series pro-sumti on, 153; effect of lerfu pro-sumti on, 1=
53; effect of other ri-series pro-sumti on, 153; effect of ti-series pro-su=
mti on, 153; effect of use on meaning, 153; effect on other ri-series pro-s=
umti, 153; in narrative about quotation, 156; in quotation series, 156; in =
quotations, 156; non-allowable referents of, 153; possible referents of, 15=
3
+ro, 128, 129, 139, 394, 396, 440; as implicit quantifier on personal pro-s=
umti, 128; compared with pa, 440; dropping from multiple appearances on log=
ical variables, 396; effect of order when multiple in sentence, 399
+ro da, 394
+ro prenu, 398
+ro'a, 307
+ro'anai: example, 307
+rock face: example, 231
+ro'e, 307
+roger: example, 325
+ro'i, 307
+ROI selma=92o, 226, 230, 503; effect of ZAhO on fe'e flag, 231; exception =
on use of boi before, 458; scalar negation of, 242
+romaji: as a basis for kanji characters in Lojban lerfu words, 420
+Roman Empire, 258; example, 258
+ro'o, 307
+roi, 226, 458
+room which he built: example, 184
+ro'u, 307
+rounded down: example, 443
+rounded numbers: expressing, 443
+rounded up: example, 443
+rounded/unrounded vowels, 31
+RP (see reverse Polish notation), 452
+ru, 153; practical referent conventions, 153
+ru'a, 317; compared with e'u, 317
+ru'e, 305
+ru'i, 226
+rug: Persian, example, 60
+runner shoe: example, 17
+sa, 312, 416, 477, 483; interaction with bu, 416
+SA selma=92o, 483, 503
+sa'a, 321, 481; editorial insertion of text already containing sa'a, 321; =
interaction with li'o, 321; interaction with sei, 321; interaction with to'=
i, 321
+sa'e, 318
+sa'enai, 319
+sa'i, 452
+said John: example, 481
+salad ingredients: example, 231
+sarcasm: example, 319; expressing, 319
+sa'u, 318
+sai, 305
+Sapir-Whorf effects: and emotional indicators, 329
+scalar attitude, 305
+scalar negation: effect on selbri, 101
+scalar negation of modals: explanation of meaning, 207
+scalar negation of non-logical connective, 358
+scalar negation of tenses: selma'o allowed with, 242
+scale: granular contrasted with continuous, 448
+scale of redness: example, 448
+scale selbri: definition, 447; place structure, 447; place structure effec=
t from subjective numbers, 448
+school building: example, 281
+schooner: example, 83
+scientific names: rules for, 67
+scientific notation: rationale for order of places, 451; with gei, 450
+score: as 20-year span, 460; as alternate base for years, 461
+se, 100, 192, 338, 340, 346, 350, 354, 360, 361, 396, 459, 472; as grammat=
ical in JOI compounds, 355; in logical connective to exchange sentences, 33=
8; order in logical connectives with na, 338; quick-tour version, 16; use w=
ith operators, 459; using to re-order logical variables, 396
+se du'u, 263
+se klama: place structure of, 192
+SE selma=92o, 100, 192, 195, 205, 247, 396, 459, 472, 504; after 5th place=
, 472; effect of multiple on a selbri, 194; effect on place structure numbe=
ring, 192; effect on selbri place structure, 192; extending scope of, 193; =
for converting place structure, 192; quick-tour version, 16; rationale for =
no 1st place conversion, 192; scope of, 193; word formation of cmavo in, 19=
2
+se te, 194
+se writing convention: in eks, 341
+se'a, 310
+seba'i, 204
+sebi'o, 359
+sece'o, 354
+section numbering, 458
+sections of this book, 4
+se'e, 425; and number base convention, 425
+see with eye: example, 202
+see with left eye: example, 194
+sefa'u, 354
+SEhU selma=92o, 159, 482, 504
+se'i, 309
+sei, 321, 481
+SEI selma=92o, 458, 481, 504; terminator for, 504
+selbri: as part of description, 120; brivla as, 83; converting into an ope=
rand, 456; converting into an operator, 456; converting operator into, 457;=
 definition, 83, 187; definition, quick-tour version, 27; lerfu string as, =
423; omitting with co'e, 158; place structure of, 187; place structure of c=
onverted operator, 457; relation to bridi, 83; scalar negation of, 101; wit=
h GOhA, 97
+selbri assignment, 491
+selbri from sumti, 98
+selbri list for quick tour, 13
+selbri logical variables, 409
+selbri place structure: effect on operator formed by, 456; re-ordering, 50=
4
+selbri placement among sumti: effect of multiple quantification on, 407
+selbri questions: quick-tour version, 23
+selbri separator, 492
+selbri to modal converter, 493
+selbri variables: form when not in prenex, 410; prenex form as indefinite =
description, 409; quantified, 410
+selbri-first bridi: effect on sumti places, 188; effect on use of cu, 190;=
 specifying first sumti place in with fa, 190
+self-orientation: example, 308
+selma=92o, 12, 396, 398
+selma'o: cross-reference list of, 489; definition, 50; definition, quick-t=
our version, 27
+seltau: compared with English adjective, 55; compared with English adverb,=
 55; definition, 95, 274; definition of, 84; effect on meaning of tanru, 84=
; filling sumti places in, 93
+seltcita sumti: definition (see also modal sumti), 195
+semantic primitives, 273
+semau, 204
+seme'a, 204
+sentence: basic Lojban, 187
+sentence 10.11: example, 445
+sentence abstraction, 262
+sentence grouping, 505
+sentence separator, 495
+sentences: close grouping, 466; connecting non-logically, 358; connecting =
with tense, 238; forethought tense connection of, 239; joining, 465; separa=
tor for joining, 465; tenseless, quick-tour version, 25
+se'o, 316
+separate questions: quick-tour version, 23
+separately tensed sentences: contrasted with tense connected sentences, 23=
9
+sepi'o, 195
+sepi'u, 354
+sequence: as an abstract list, 355; contrasted with list, 355; contrasted =
with set, 134
+sequence of events: expressing non-time-related sequences, 358
+sequence of tense rules: Lojban contrasted with English, 238
+set: as specified by members, 355; by listing members with ce, 355; compar=
ed with mass as abstract of multiple individuals, 125; contrasted with mass=
 in attribution of component properties, 125; contrasted with mass in distr=
ibution of properties, 355; contrasted with ordered sequence, 355; expressi=
ng measurement standard for indefinites, 446; expressing relation with indi=
viduals forming set, 446; expressing relation with mass formed from set, 44=
6
+set of all rats: example, 447
+set of rats: example, 135
+set operations, 356
+setese, 194
+sets: properties of, 125; rule for implicit outer quantifier, 130; use in =
Lojban place structure, 125
+se'u, 159, 482; as elidable terminator for soi, 159; elidability considera=
tions, 159
+sexual discomfort: example, 307
+sexual teacher: male, example, 74
+shared bridi-tail sumti: avoiding, 200
+sheep breed, 290
+sheep flock: example, 291
+sheepdog, 290; example, 290
+shell worm: example, 285
+shellfish, 285; example, 285
+Sherman tank: example, 60
+shift: single-letter, grammar of, 415
+shift word: canceling effect, 418; for face, 418; for font, 418; for singl=
e letter, 415; scope, 415
+ship sank: example, 314
+shoehorn, 281
+shook stick: example, 145, 162
+short rafsi, 57
+short rafsi form: compared with long form in effect on lujvo meaning, 56
+si, 312, 416, 477, 482; interaction with bu, 416
+SI selma=92o, 482, 504
+si'a, 317
+si'e, 447
+signed numbers: expressing, 432
+signs on numbers: grammar, 433
+Simon says: example, 149
+simple bridi: terminator for, 506
+simple sumti, 119
+simultaneously: example, 364
+sinful: example, 309
+single consonants: contrasted with consonant clusters, 35; contrasted with=
 doubled consonants, 35
+single-letter shift: as toggle, 415
+single-word quotation, 141
+singular me: example, 446
+si'o, 265
+sister pregnant: example, 320
+six-shooter: example, 246
+size: order with dimensionality in spatial tense intervals, 224
+slinku'i test: definition, 62
+slowdown, 258
+smiley face: example, 416; word for, 416
+sneak in: example, 285
+snow falls: example, 233
+snowball's chance: example, 448
+so'a, 440
+social butterfly: example, 18
+Socrates: example, 198
+so'e, 440; meaning of, 441
+so'i, 440
+soi, 159; use in expressing reciprocity, 159; use in expressing reciprocit=
y with vo'a-series pro-sumti, 159
+SOI selma=92o, 159, 504; terminator for, 504
+soi with one following sumti: convention, 159
+some do not go to school: example, 405
+some relationship: example, 409
+somebody: contrasted with somebody else, 393
+somebody loves self: example, 393
+somebody loves somebody: example, 393
+somebody's dog: example, 393
+something: contrasted with someone, 395; expressing using =93su'o=94, 397;=
 unspecified definite with =93zo'e=94, 392
+something is loved by everybody: example, 402
+something sees everything: example, 394
+something sees me: example, 392, 397
+so'o, 440
+so'u, 440
+sounds: clarity of, 31; complex, 31; difficult, 31
+sounds for letters: Lojban contrasted with English, 31
+source languages: use in creating gismu, 75
+south face: example, 231
+sow grain: example, 231
+sowed grain: example, 250
+space: as time-based metaphor, 231; contrasted with time in number of dire=
ctions, 219
+space indicator for interval modifiers, 493
+space interval: compared with time intervals in continuity, 230
+space location: as part of tense system (see also tense, spatial tense), 2=
15
+space movement indicator, 501
+space tenses: quick-tour version, 26
+space/time metaphor: expressing direction mapping for, 231
+spaghetti, 61; example, 63
+Spanish ch: example, 419
+Spanish ll: example, 419
+spatial contours: as sumti tcita, 232; contrasted with temporal event cont=
ours, 231; expressing, 231
+spatial directions: list of, 253
+spatial information: adding to a sentence with tense sumti tcita, 231
+spatial interval: expressing degree of continuity over, 230
+spatial interval modifiers: order in tense, 230
+spatial tense: 4-dimensional interaction with temporal tense, 224; as an i=
maginary journey, 217; as optional in English, 217; as sumti tcita, 232; co=
mpared with temporal tense in elidability, 217; contrasted with temporal in=
 dimensionality, 223; definition, 217; direction, 217; distance, 217; four-=
dimensional, 224; linear, 223; one-dimensional, 223; order of direction and=
 distance specification, 217; order relative to temporal, 219; planar, 223;=
 reference frame, 217; referent of, 217; three-dimensional, 223; two-dimens=
ional, 223
+spatial tense interval: order of size and dimensionality in, 224; order of=
 VEhA and VIhA in, 224
+speaker-listener cooperation, 23
+speaker-relative viewpoint: contrasted with event-relative viewpoint, 228
+speaker's state of knowledge, 319
+specific descriptions, 121
+specific terms: use of fu'ivla for, 61
+specificity: expressing with po, 173
+speech rhythm: for grouping in English, 85
+spelling out words: Lojban contrasted with English, 414
+spiritual discomfort: example, 307
+SQL: example, 424
+square brackets: use of in notation, 5
+standard bridi form: definition, 188
+standard for subjective numbers: specifying, 448
+standard pronunciation, 29
+starting marker, 483
+state abstraction: place structure, 259
+state abstraction(s): definition, 258; related tense contours, 268
+state abstractor, 258
+state event: described, 258
+steady speed, 258
+stereotypical: as not derogatory in Lojban, 126; compared with typical, 12=
6
+stereotypical objects, 126
+Steven Mark Jones: example, 421
+sticky modals: canceling, 208; definition, 207; fi'o proscribed from, 208
+sticky tenses: and CAhA, 243; canceling, 235; definition, 234; effect of n=
au on, 238; effect on future tense meaning, 234; from part of a multiple te=
nse, 235
+stoke cat then rabbit: example, 240
+stop: contrasted with finish, 229; contrasted with pause, 229
+stories: flow of time in, 236
+story tense: Lojban convention contrasted with English convention, 236
+story time: as a convention for inferring tense, 236; definition, 236; rat=
ionale for, 236; tenseless sentences in, 236; with no initial sticky time, =
237
+stress: definition of, 40; effect of buffer vowel on, 38; effect of syllab=
ic consonants on, 34; example, 307; final syllable, rules for pause after, =
69; irregular marked with upper-case, 415; levels of, 40; on cmavo, 52; pri=
mary, 40; quick-tour version, 12; rules for, 40; secondary, 40; showing non=
-standard, 29
+stressed syllable: compared with stressed vowel, 40
+stressed vowel: compared with stressed syllable, 40
+strong quotation, 477
+structure of examples, 5
+structure of this book, 4
+structure words, 50
+su, 312, 416, 477, 484; interaction with bu, 416
+SU selma=92o, 484, 504
+su'a, 316, 319
+su'anai, 316
+subjective amounts: expressing, 442
+subjective numbers: effect on place structure for cardinal selbri, 448; ef=
fect on place structure for ordinal selbri, 448; effect on place structure =
for portion selbri, 448; effect on place structure for probability selbri, =
448; effect on place structure for scale selbri, 448; rationale for effect =
on place structure, 448; specifying standard for, 448
+subjective portions: expressing, 442
+subordinate clause tense: effect of main bridi tense on, 238; Lojban compa=
red with Esperanto, 238; Lojban compared with Russian, 238; Lojban contrast=
ed with English, 238
+subordinate clauses: tense usage rules in English, 237
+subscripted topics, 467
+subscripts, 471; and fuzzy truths, 473; and names, 473; and paragraph sepa=
rators, 474; and pro-sumti, 472; and sumti re-ordering, 472; and tense, 473=
; before main expression, 450; effects on elidability of terminators, 450; =
external grammar of, 449; for sticky tense, 236; internal grammar of, 449; =
lerfu string as, 423; marker, 471, 507; mathematical, 473; multiple as sub-=
subscript, 450; multiple for same base word, 455; on ke'a for nested relati=
ve clauses, 161; on ri, 153; terminator for, 450; to form matrices of more =
than 2 dimensions, 452; use with ke'a for outer sumti reference, 184; use w=
ith logical variables, 410
+subscripts on lerfu words: effect on elidability of boi, 450
+subsets: expressing with outer quantifiers, 131
+sub-subscripts, 450
+subtraction operator: contrasted with negation operator, 438; contrasted w=
ith negative sign, 434
+subtypes of words, 52
+su'e, 443; with elided number, 443
+su'i, 97, 435, 436
+sum of 1: 2, 3, example, 438
+sumti: as having implicit quantifiers, 127; as objects in place structure =
slots, 187; beginning with "ke, 344; between descriptor and description sel=
bri, 180; classified by types of objects referred to, 123; converting into =
an operand, 456; definition, 119, 187; definition, quick-tour version, 27; =
descriptions as, 119; dropping trailing unspecified, 189; explicitly mappin=
g into place structure with FA, 190; for individual objects, 123; for mass =
objects, 123; for set objects, 123; forethought tense connection of, 239; i=
rrelevant to relationship, 157; kinds of, 119; multiple in one place with F=
A, 191; names as, 119; numbers as, 119; omitted first place in selbri-first=
 bridi, 188; order in selbri, 188; order in selbri-first bridi, 188; pro-su=
mti as, 119; quotations as, 119; relation with bridi, 11; re-ordering with =
FA, 190; with explicit quantifiers, 127
+sumti assignment: cancellation of, 492
+sumti connection: afterthought, 340; forethought, 341
+sumti into selbri, 98
+sumti logical connection, 340; compared with bridi logical connections, 34=
0; contrasted with tanru logical connection, 350; rationale for, 340
+sumti modal connection, 200
+sumti place: additional, 489
+sumti placement: variant, quick-tour version, 15
+sumti qualifiers: as short forms for common special cases, 133; elidable t=
erminator for qualified sumti, 133; external syntax of, 133; for negation, =
135; internal syntax of, 133; list of, 133
+sumti questions: quick-tour version, 22
+sumti raising, 266
+sumti reordering: quick-tour version, 16
+sumti tcita: based on event contours, 232; based on spatial contours, 232;=
 based on tense direction, 232; based on tense distance, 232; based on tens=
es, 231; definition (see also modal tag), 195; event contours contrasted wi=
th direction/distance as basis for, 232
+sumti tcita and linked sumti, 94
+sumti tcita and modal tags, 94
+sumti tcita and tense tags, 94
+sumti tcita based on dimension, 233
+sumti tcita based on event contours: relation of main bridi to sumti proce=
ss in, 232
+sumti tcita based on interval continuousness, 233
+sumti tcita based on interval properties, 233
+sumti tcita based on interval size, 233
+sumti tcita based on quantified tenses, 233
+sumti with explicit quantifier: contrasted with sumti without explicit qua=
ntifier, 127
+sumti with lo: compared to indefinite sumti, 399
+sumti with tense: effect of main bridi tense on, 235
+sumti with tenses: quick-tour version, 26
+sumti-based description: definition, 132; inner quantifier on, 132; outer =
quantifier on, 132
+sumti-based descriptions with le: as increasing restricting to in-mind, 13=
3
+sun liquefies: example, 289
+sunburn: example, 259
+su'o, 128, 129, 397, 443; as implicit quantifier for quotations, 128; with=
 elided number, 443
+superfective event contour, 228
+superscripts, 450
+supervising: as a contribution to mass action, 354
+supper: example, 57
+supplementary information, 280
+Susan: example, 481
+su'u, 265
+Sun: the, example, 66
+Svetlana: example, 68
+sword blade, 283
+syllabaries: lerfu word representation, 420
+syllabic consonant: effect on stress, 65
+syllabic consonants, 34; effect on stress, 34; final in word, 34
+syllabic l: considered as a consonant for morphological discussions, 49
+syllabic m: as a consonant for morphological discussions, 49
+syllabic n: as a consonant for morphological discussions, 49
+syllabic pronunciations of consonants: in fu'ivla, 62; in fu'ivla category=
 attachment, 63
+syllabic r: as a consonant for morphological discussions, 49
+syllabication: and names, 39; definition of, 39; examples of, 39; rules fo=
r, 39; variants of, 41
+syllable break: contrasted with pause, 32; representation in Lojban, 32; s=
ymbol for, 416; word for, 416
+symbol: contrasted with referent, 478; for operand, 459; referring to with=
 lu'e, 134
+symmetrical tanru, 111
+symmetrical tanru types: both separately true, 111; one or other true, 112=
; using crucial/typical parts, 112; using more inclusive class, 112
+symmetrical veljvo, 278
+sympathy: example, 299, 314
+ta, 147, 169; contrasted with di'u, 148
+ta'apei, 326
+tables: format of, 5
+ta'e, 226, 324
+tagged sumti termsets: connecting with non-logical forethought connectives=
, 358
+TAhE selma=92o, 225, 504; effect of ZAhO on fe'e flag, 231; scalar negatio=
n of, 242
+tail-terms: definition, 345
+Take care!: example, 22
+Talk!: example, 22
+talker: example, 19
+taller: example, 11
+tan(pi/2) =3D infinity: example, 456
+tank: Sherman, example, 60
+tanru: ambiguity in, 55; ambiguity of, 55, 85; and abstractions, 255; and =
conversion, quick-tour version, 18; and creativity, 55; as ambiguous, 85; a=
symmetrical, 104; combination of, 55; containing mathematical expressions, =
97; default left-grouping of, 86; definition, 83; definition, quick-tour ve=
rsion, 27; expanding, 318; explanation of, 55; explicating, 318; explicitly=
 defining, 318; expression of, 55; meaning of, 85; place structure of, 274;=
 place structure of, quick-tour version, 18; place structures of, 92, 93; p=
ossible meanings of, 274; primary meaning of, 84; purpose, 274; quick-tour =
version, 17; reducing logically connected sumti to, caveat, 350; simple, 83=
; to lujvo, 55; with GOhA, 97
+tanru and conversion, 100
+tanru connection: connotation of non-logical, 354
+tanru connection grouping: guheks unmarked tanru, 350
+tanru conversion: effect on place structure, quick-tour version, 18
+tanru default grouping: quick-tour version, 17
+tanru grouping: complex, 87; effect of tanru inversion on, 96; guheks comp=
ared with jeks, 350; three-part, 85; with bo, 87; with ke, 88; with ke and =
bo, 88
+tanru grouping with JA+BO: effect on tanru grouping, 91
+tanru inversion, 95; definition, 95; effect on tanru grouping, 96; in comp=
lex tanru, 96; multiple, 96; rule for removing, 96; where allowed, 96
+tanru inversion and place structure, 95
+tanru logical connection: contrasted with sumti logical connection, 350
+tanru nested within tanru, 86
+tanru unit: filling in places of, 489
+ta'o, 319
+ta'onao, 320
+ta'u, 318
+ta'unai, 318
+tavla, 14
+te, 100, 192; quick-tour version, 16
+te'a, 437
+technical terms, 5
+TEhU selma=92o, 505
+tei, 419
+TEI selma=92o, 419, 505; terminator for, 494
+telephone conversation: hello, 324
+television, 42
+template, 266
+temporal direction: exception in meaning when following ze'e, 227
+temporal information: adding to a sentence with tense sumti tcita, 231
+temporal tense: as mandatory in English, 215; compared with spatial tense =
in elidability, 217; historical definition, 215; interaction with 4-dimensi=
onal spatial tense, 224; Lojban contrasted with English in necessity, 215; =
order relative to spatial, 219; quantified with direction, 227; real relati=
onship to time in English, 215
+temporal tense elision: compared with spatial tense elision in meaning, 21=
7
+temporal tenses: compared with spatial tenses, 219
+ten: expressing as number, 432
+tense: aorist, 223; as observer-based, 220; as subjective perception, 219;=
 compared with modals in syntax, 248; connected, with negation, 245; connec=
ting sentences in with, 238; contradictory negation contrasted with scalar =
negation of, 242; contradictory negation of with nai, 241; contrasted with =
modals in semantics, 248; effect of different position in sentence, 216; ef=
fect of sticky tense on, 234; emphasizing by position in sentence, 216; exp=
lanation of presentation method, 215; expressing movement in, 224; extensio=
nal, 503; forethought connection in, 363; forethought logical connections, =
246; grouping of connectives in, 363; handling multiple episodes, 236; impo=
rtance of 2nd sumti place for sumti tcita use, 248; in forethought bridi-ta=
il connection, special rule, 365; interval contrasted with point, 221; logi=
cally connected with JA, 245; Lojban contrasted with English in implication=
s of completeness, 223; Lojban contrasted with English in implying actualit=
y, 243; Lojban contrasted with native languages, 215; making sticky, 497; m=
ultiple in sentence, 234; multiple in sentence compared with compound tense=
, 234; negating, 241; non-logical connection of, 246; non-logical connectio=
n of for sub-events, 246; numerical, 458; on embedded bridi, 235; order of =
direction specification in, 217; order of direction, distance and interval =
in, 221; order of distance specification in, 217; order of movement specifi=
cation in, 225; order of spatial interval modifiers in, 230; order of tempo=
ral and spatial in, 219; overriding to speaker's current, 238; point contra=
sted with interval, 221; position in sentence alternative, 216; position of=
 in sentence, 216; possible groupings of, 246; quantified, 226; quick-tour =
version, 25; rationale for relative order of temporal and spatial in, 219; =
relation of interval to point specified by direction and distance, 221; rel=
ation of point specified by direction and distance to interval, 221; relati=
ve order with bridi negation, 103; scalar negation contrasted with contradi=
ctory negation of, 242; scalar negation of with NAhE, 242; scope effect of =
new paragraph, 466; scope of, 234; selbri types applicable to, 215; space-t=
ime dimension for intervals, 224; speaker's current, 238; specifying relati=
on of interval to point specified by direction and distance, 221; static co=
ntrasted with moving, 224; subscripting, 236; sumti tcita form contrasted w=
ith connected sentences, 239; use as sumti tcita, 231; viewpoint of PU cont=
rasted with viewpoint of ZAhO, 228; with both temporal and spatial, 220; wi=
th ku, 216
+tense afterthought connection forms: selma'o allowed, 240
+tense and na: multiple, 104
+tense as sumti tcita: contrasted with tense inside sumti, 233
+tense aspect, 507
+tense cmavo: position relative to selbri, 104
+tense connected sentences: contrasted with separately tensed sentences, 23=
9; forethought mode, 239; importance of =93bo=94 in, 239
+tense connection: equivalent meanings, 240; expansions of, 240
+tense connection of bridi-tails: meaning of, 240
+tense connection of sentences: contrasted with sumti tcita form, 239; orde=
r of, 239
+tense connection of sumti: meaning of, 240
+tense contours: compared with event abstraction contours, 268
+tense conversion: accessing original first place with fai, 247; accessing =
tense of bridi with jai, 247; of temporal tenses, 248; use in sumti descrip=
tions, 247
+tense direction: as sumti tcita, 232; contrasted with event contours in im=
plication of extent, 229; implications on scope of event, 223
+tense direction/distance as sumti tcita: contrasted with event contours, 2=
32
+tense distance: as sumti tcita, 232
+tense forethought connection forms: selma'o allowed, 240
+tense in scope of sticky tense: compared with compound tense, 234
+tense inside sumti: contrasted with tense as sumti tcita, 233
+tense magnitude, 250
+tense on main bridi: effect on embedded bridi tenses, 235; effect on embed=
ded sumti with tenses, 235
+tense questions: by using logical connective question, 250; methods of ask=
ing, 249
+tense questions with ma, 249
+tense selma'o: summary of, 252
+tense sentence connection: table of equivalent schemata, 249
+tense specification: effect on cu, 216; effect on elidability of terminato=
rs, 216
+tense system: and space location, 215
+tense tags and sumti tcita, 94
+tense with elided CAhA: meaning, 244
+tense with no sumti: indicator for, 497
+tense with sumti tcita: asymmetry of, 238
+tensed connectives: in mathematical expressions, 364
+tensed logical connection, 363
+tensed logical connective, 240; forethought, 364; with ke=85ke'e, 241; wit=
h tu'e=85tu'u, 241
+tensed logical connective(s): in ek=85bo, 364; in ek=85ke, 364; in gihek=
=85bo, 364; in gihek=85ke, 364; in ijek=85bo, 364; in ijek=85tu'e, 364; in =
ijoik=85bo, 364; in ijoik=85tu'e, 364; in jek=85bo, 364; in joik=85bo, 364;=
 in joik=85ke, 364
+tensed logically connected bridi-tails, 240; with grouping, 241
+tensed logically connected sentences, 240; with grouping, 241
+tensed logically connected sumti, 240; with grouping, 241
+tensed non-logical connectives, 364; forethought, 364
+tenseless sentences in story time, 236
+tense-or-modal questions: with cu'e, 250
+term: definition, 347
+terminators: eliding ku in non-logical connections, 354
+termset: effect on scope of multiple indefinite sumti, 399; formation, 347
+termset connectives, 503
+termset logical connection: contrasted with bridi connection, 347; contras=
ted with bridi-tail connection, 347; contrasted with sumti connection, 347;=
 unequal length, 348; when used, 347
+termset marker, 502
+termset modal connection, 200
+termsets, 491; compared to fa'u, 356; non-logical connection of, 357
+tertau: definition, 95, 274; definition of, 84; effect on meaning of tanru=
, 84
+te'u, 451, 456, 460
+text: division numbering with -mai, 458; end-marker, 484; structure of, 46=
5; sub-division numbering with -mai, 458
+text quotation: as internally grammatical, 141; syntax of, 141
+thank you: example, 324
+the: contrasted with a/an, 322; example, 322; for talking about numbers th=
emselves, 435
+the destination: example, 193
+the go-er: example, 193
+The men are women: example, 120
+the two of you: example, 132
+there is a Y: expression, notation convention, 401
+thingy: example, 151
+this: adjective expression with ti noi, 148; adjective usage contrasted wi=
th pronoun usage, 148; as utterance reference in English, 148; English, adj=
ective expression with vi, 148; English, pronoun expression with ti, 148; p=
ronoun usage contrasted with adjective usage, 148
+this boat: example, 148
+this book: author of, 5; contributors to, 5; credits for, 6; examples of, =
4; goal of, 3; reviewers of, 6; sections of, 4; structure of, 4
+this/that in English: compared with ti-series pro-sumti, 147
+three bears: example, 133
+three cats white: and two big, example, 410
+three dogs bite two men: example, 398
+Three Kings: example, 99
+three of four people: example, 361
+three or four people: example, 454
+three rats: example, 446
+thus: example, 316
+ti, 119, 147, 169; as pronoun expression for English this, 148
+ti noi: as adjective expression for this, 148
+ti'e, 316
+tilde: a diacritical mark, 418; proposed lerfu word for, 429
+time: as part of tense system (see also tense, temporal tense), 215; as sp=
ace-based metaphor, 231; contrasted with space in number of directions, 219
+time of death: example, 248
+time tenses: quick-tour version, 25
+time travel, 225
+times: explicit expression of, 437; implicit expression of, 437
+ti'o, 458
+ti-series pro-sumti: 3 degrees of distance with, 147; as pointing referent=
s only, 147; compared with English this/that, 147; contrasted with di'u-ser=
ies pro-sumti, 148; conversational convention for, 147; effect on ri-series=
 pro-sumti, 153; lack of pro-bridi equivalent, 148; problems in written tex=
t, 147
+title: specifying with tu'e=85tu'u, 466
+title of book: example, 134
+to, 480
+to movie: house, office, example, 191
+TO selma=92o, 480, 505; terminator for, 505
+to the market from the office, 348
+to'a, 415
+to-do list: example, 358
+to'i, 321, 480
+toi, 480
+TOI selma=92o, 480, 505
+Tolkien: and non-standard Lojban orthography, 46
+tomorrow: example, 282
+tone of voice, 297
+to'o: special note on direction orientation, 253
+too: example, 318
+too long: example, 233; Example, 230
+too many rats: example, 448
+topic/comment: multiple sentence, 468
+topic-comment: description, 467
+topic-comment sentences, 467
+tosmabru test, 71
+to'u, 318
+toward her right: example, 225
+toward my right: example, 224
+toward right: contrasted with on right, 224
+transfinite cardinal: example, 434
+transformations with logical connectives: steps, 408
+traveling salesperson: example, 196
+triumph, 258
+truncation of number: expressing, 443
+truth: in imperative sentences, 353
+truth functions, 333; 16 possible, 333; commutative, 335; creating all 16 =
with Lojban's basic set, 335; fundamental 4 in Lojban, 334; relation to log=
ical connectives, 334; table of logical connectives, 366
+truth questions, 321; answering =93no=94, 351; answering =93yes=94, 351; a=
s yes-or-no questions, 351; contrasted with connection questions, 351; simp=
le, 351
+truth table: explanation, 333
+truth tables: abbreviated format, 334; for 4 fundamental Lojban truth func=
tions, 335; list of 16 in abbreviated form, 334; notation convention, 334
+truth-value abstraction, 262
+truth-value abstraction(s): place structure, 262, 263
+try the door: example, 266
+try to go: example, 95
+ts-sound in Russian: representation in Lojban, 31
+tu, 147, 169; archaic English yon as equivalent of, 147
+tu'a, 134, 266; as being deliberately vague, 134; effect of on meaning, 13=
4; use for forming abstractions, 134
+tu'e, 202, 205, 343, 358, 364, 466; contrasted with bo for tensed logical =
connection, 364; effect on di'e, 358; use in lists, 358
+TUhE selma=92o, 343, 358, 364, 466, 505; terminator for, 505
+TUhU selma=92o, 343, 358, 466, 505
+tu'o, 450, 453; for infix operations with too few operands, 450
+tu'o va'a, 453
+tu'u, 202, 205, 343, 358, 466
+twice today: example, 233
+two brothers: example, 98
+two dogs are white: example, 129
+types and subtypes of words, 52
+typical: compared with stereotypical, 126
+typical Englishman: example, 126
+typical Lojban user: example, 125
+typical objects: and instantiation, 126; determining characteristics of, 1=
26
+typical Smith: example, 127
+typical sumti, 157
+typical value: contrasted with elliptical value for sumti, 157; contrasted=
 with mathematical average, 441
+typographical conventions, 4
+ue, 297
+ugh: example, 359
+ui, 297
+UI selma=92o, 264, 297, 351, 353, 469, 474, 481, 505; extending the scope =
of, 494; quick-tour version, 24
+umlaut: a diacritical mark, 418; proposed lerfu word for, 429
+unabridged dictionary, 480
+unconditional signal, 484
+unconnected tanru: contrasted with logically connected version, 349
+undemonstrated potential: expressing, 244
+under compulsion: example, 201
+under conditions: example, 257
+underlines: example, 12
+unequal termset connection: compared with compound bridi connection with u=
nequal separate bridi-tails, 348
+unfilled places of inverted tanru, 95
+Unicode, 425
+uninterpreted sequence, 477
+union: of sets, 356
+union of sets: compared with or, 357
+units of measurement: expressing, 435
+universal: mixed claim with existential, 394
+universal claims: dangers of using, 396; explanation, 393; restricting, 39=
4, 395
+unqualified sumti: contrasted with qualified sumti, 133
+unreduced fractions: use in granular scales, 448
+unreduced lujvo: definition, 57
+unspecified breed: example, 280
+unspecified direction: temporal contrasted with in spatial, 220
+unspecified emotion, 311
+unspecified level of emotion, 311
+unspecified route: example, 189
+unspecified sumti: non-trailing, 189; using zo'e as place-holder for, 189
+unspecified trailing sumti: dropping, 189
+unstated emotion, 311
+unusual characters: words for, 416
+unusual position, 290
+unvoiced consonants: contrasted with voiced in allowable consonant pairs, =
37
+unvoiced vowel glide: apostrophe as, 31
+upper-case: lerfu word for, 415
+upper-case letters: English usage contrasted with Lojban, 415; Lojban usag=
e contrasted with English, 415
+utterance: expressing relation to discourse, 317
+utterance ordinal: lerfu string as, 423
+utterance pro-sumti: stability of, 162
+utterance pro-sumti (see also di'u-series pro-sumti), 148
+utterances: non-bridi, 471
+u'u, 299; contrasted with uu, 299
+uu, 297, 299; contrasted with u'u, 299
+uuse'inai, 314
+uy diphthong: in cmene, 66
+V: as a symbol for a single vowel, 49
+va, 217
+VA selma=92o, 217, 506; and distance, 217; relation of words to ti, ta, tu=
, 217
+va'a, 438; contrasted with vu'u and ni'u, 438
+va'e, 447
+vague abstraction, 265
+vague abstraction(s): place structure, 266
+vague abstractor, 265
+vague numbers, 128
+vague relationship: modal tag for, 197
+va'i, 318, 320; contrasted with ke'u, 320
+va'inai, 320
+valid speech: marking as error with jo'a, 321
+value abstraction, 262
+variables: logical, 392
+vau, 178, 200, 345, 394; effect on elidability ku'o, 181
+vau for shared bridi-tail sumti: avoiding, 200
+VAU selma=92o, 345, 506
+ve, 100, 192; quick-tour version, 16
+ve klama, 193; contrasted with pluta, 193
+vector: components of, 451; definition, 451; forming, 496; use as operand,=
 452; use of parentheses with, 452
+vector indicator, 451; terminator for, 451
+ve'e, 228
+VEhA selma=92o, 221, 506
+VEhO selma=92o, 506
+vei, 201, 423, 437
+vei =85ve'o: contrasted with operator =85ku'e in Polish notation, 438
+VEI selma=92o, 506; terminator for, 506
+veljvo: definition, 274; symmetrical, 278
+ve'o, 201, 362, 437
+verbs: brivla as Lojban equivalents, 52
+veterinarian: example, 282
+vi, 217; as adjective expression for English this, 148
+vi'a, 224
+vice versa: English, expressing with vo'a-series pro-sumti and soi, 159; e=
xample, 159
+vi'e, 224
+VIhA selma=92o, 223, 506
+vi'o, 325; contrasted with je'e, 325
+virtue: example, 309
+viska, 194
+vo'a, 158
+vo'a-series pro-sumti: use in expressing reciprocity with soi, 159
+vocative, 492
+vocative indicator, 492
+vocative phrase: as a free modifier, 135; effect of position on meaning, 1=
37; elidable terminator for, 137; explicit quantifiers prohibited on, 136; =
forms of, 136; implicit descriptor on, 136; implicit quantifiers on, 136; p=
urpose of, 136; relative clauses on, 184; with complete sumti, 136; with su=
mti without descriptor, 136
+vocative phrase terminator: elidability of, 137
+vocative phrase with name: placement of relative clause on, 184
+vocative phrase with selbri: placement of relative clause on, 184
+vocatives: and definition of =93you=94, 323; contrasted with =93la=94, 323=
; definition, 323; grammar overview, 323; notation convention symbol =93X=
=94, 323; quick-tour version, 21; rationale for redundancy, 323
+vo'e, 158
+voi, 177; contrasted with poi in veridicality, 177
+voiced consonants: contrasted with unvoiced in allowable consonant pairs, =
37
+voiced/unvoiced consonants: restrictions on, 36
+vowel: buffer, 38
+vowel buffer: contrasted with y sound, 38
+vowel pairs: contrasted with diphthongs, 34; definition of, 34; grouping o=
f, 35; involving y, 35; list of, 35; use of apostrophe in, 34
+vowel-initial words: necessity for pause before, 68
+vowels: contrasted with consonants, 33; definition of, 33; length of, 39; =
pronunciation of, quick-tour version, 12
+vu, 217
+vu'e, 309
+VUhO selma=92o, 182, 506
+VUhU operands, 436
+VUhU selma=92o, 435, 436, 444, 507
+vu'i, 134; effect of on meaning, 134; use for creating sequence, 134
+vu'o, 182
+vu'u, 438; contrasted with va'a and ni'u, 438
+VV string: as a symbol for a double vowel, 49
+walk to market: example, 100
+want to be a soldier: example, 255
+wash self: example, 158
+weapon against self: example, 397
+went and bought: example, 363, 365
+what is your name: example, 160
+when: example, 249
+when else: example, 250
+when/where/how: example, 250
+where: example, 249
+whether criminal: example, 262
+who knows what: example, 160
+whole of: example, 441
+whole time interval: expressing, 227
+William: example, 68
+window: example, 336
+wine-dark sea, 274
+word =93abu=94: example, 422
+word =93bu=94: example, 416
+word classes, 50
+word forms: as related to grammatical uses, 49; in Lojban (see also morpho=
logy), 49
+word lists, 273
+word quotation: as morphologically valid, 141; internal grammar of, 141
+words: marking non-standard, 480
+words not in the dictionary, 13
+world's languages, 273
+wrong concept, 281
+x < 5: example, 439
+x{b: d}, example, 362, 455
+x{k}: x sub k, example, 423
+x1: in place structure notation, 187; notation convention, quick-tour vers=
ion, 13
+xe, 100, 192; quick-tour version, 16
+xi, 449, 452, 471
+XI selma=92o, 449, 471, 507
+xo, 449
+X-ray: example, 60
+x-sub-3, 449
+xu, 321, 351, 469; quick-tour version, 23
+y, 484; considered not to be a vowel for morphological discussions, 49; le=
tter; between letters of consonant pair, 53; letter, prohibition from fu'iv=
la, 62; use in avoiding forbidden consonant pairs, 36
+Y selma=92o, 484, 507
+y sound: contrasted with vowel buffer, 38
+YACC grammar, 511
+Yay!: example, 299
+yes/no questions, 321; quick-tour version, 23
+y-hyphen: and consonant cluster determination, 56; and stress determinatio=
n, 56; use of, 56
+yielding the floor, 484
+yon: as archaic English equivalent of tu, 147
+you: defining, 323
+you-cmavo: example, 163
+younger: example, 292
+you're welcome: fi'i contrasted with je'e, 324; je'e contrasted with fi'i,=
 324
+you-talk: example, 163
+z =3D f(x): example, 438
+z instead of ': in acronymic names based on lerfu words, 424
+za'a, 316
+za'e, 69, 416, 480; interaction with bu, 416; use to avoid lujvo misunders=
tandings, 69
+ZAhO compared with NU, 268
+ZAhO selma=92o, 228, 231, 232, 268, 507; contradictory negation of, 241; e=
ffect on fe'e flag for TAhE and ROI, 231
+za'i, 257, 258, 268; place structure, 259
+za'o, 230
+za'u, 443; with elided number, 443
+zai, 418
+ze'e, 227; effect on following PU direction, 227
+ze'eba, 227; meaning of, 227
+ze'eca, 227; meaning of, 227
+ze'epu, 227; meaning of, 227
+ZEhA selma=92o, 221, 507
+ze'o: special note on direction orientation, 253
+zei, 60, 416; interaction with bu, 416
+ZEI selma=92o, 60, 507
+zero: relation to negation boundary, 402
+zero to one: example, 362
+ZI selma=92o, 219, 507; compared with VA, 219
+zi'e, 175; compared with English and, 176; contrasted with logical connect=
ives, 176; use in connecting relative phrase/clause to relative phrase/clau=
se, 176
+ZIhE selma=92o, 175, 508
+zi'o, 157, 163; as creating new selbri, 157
+zi'o rafsi: effect of on place structure of lujvo, 163
+Zipf's Law, 69
+zo, 119, 312, 416, 477; contrasted with la for names, 478; interaction wit=
h bu, 416
+ZO selma=92o, 477, 508
+zo'e, 157, 189, 392; as a translation for =93something=94, 392; as place-h=
older for sumti, 157; as place-holder for unspecified sumti, 189; compared =
with FA for omitting places, 190; contrasted with da, 392; quick-tour versi=
on, 14
+zo'e-series: compared with do'i as indefinite pro-sumti, 158
+zo'e-series pro-sumti, 157
+ZOhU selma=92o, 392, 467, 508
+zo'i: special note on direction orientation, 253
+zo'o, 318, 319
+zo'u, 392, 467
+zoi, 119, 416, 477; interaction with bu, 416; interaction with lo'u/le'u, =
478
+ZOI selma=92o, 477, 479, 508
+zu'a, 217; derivation of word, 217
+zu'i, 157
+zu'o, 257, 258, 268; place structure, 259
+zu'u, 319
+
diff --git a/todocbook/orig/refg915.pdf b/todocbook/orig/refg915.pdf
new file mode 100644
index 0000000..2ada91b
Binary files /dev/null and b/todocbook/orig/refg915.pdf differ

--=20
You received this message because you are subscribed to the Google Groups "=
BPFK" group.
To post to this group, send email to bpfk-list@googlegroups.com.
To unsubscribe from this group, send email to bpfk-list+unsubscribe@googleg=
roups.com.
For more options, visit this group at http://groups.google.com/group/bpfk-l=
ist?hl=3Den.