Date: Sat, 20 Feb 93 10:23:48 -0500
From: lojbab@grebyn.com (Logical Language Group)
Message-Id: <9302201523.AA18316@daily.grebyn.com>
To: nsn@mullian.ee.mu.OZ.AU
Subject: Re: first attempt at a proto-dictionary outline
Cc: cowan@snark.thyrsus.com
X-Mozilla-Status: 0011

! has not been considered for anything.  That idea is a possible one, which
would correspond to ? on question words.  Feel free to propose it to the
list.

Cowan has suggested that we need a beeter way to flag start of sentence v
vusually.  He finds .i is not eye-catchiung enough in flowing text to make
reading of such text easy 9as compared with the capitalization convention norm
in Western languages.  Since sometimes .i is part of a compound, this problem
is heightened.  Part of the problem is that the "." is used so many other
places that are non-punctuational.  One possibility is a double period at
start of sentence.  Another is to start all sentences on new lines (which is 
extremely space inefficient).  The text which particularly caused him to
comment was your ckafybarja text, which in JL is a single paragraph, since you
have no significant conversation nor use of ni'o.  It makes for a looong
paragraph, by English standards (but then you tend to write long paragraphs in
English, too).

>selma'o

the place structure will be being changed to make this work - it is sacred
by now

>IPA attempt

you did a draft, ran it up the flag, and I never heard further mention of it.
If you are satisfied with what you proposed, it is a candidate for inclusion.

>intermingling of lujvo lists

the reason for not doing so is that most of the rest of the lists will not have
place structures, or if so, will not likely have them in the same format.  This
may not be a good enough reason.  On the other hand, if there is a combined
Lojban-order dictionary, then there is good reason to have a separate list
here, so that people can look at teh different kinds of lists sepsarately.
Hence I'm waiting to see.

BTW, I hope you have documented the lujvo that you have weeded out, including
why they were weeded, and the notes as to which piece they were in, if known.
I will probably go with your decisions, if only for time considerations, but
eventually, all need to be run through more than one person, and perhaps the
original poser, as well where applicable.

Note that the list of new lujvo grows with new texts being written, so this is
a never ending job.

>Eaton/TLI lists

I never turn down a volunteer.  This also will need more than one person to
pass.  I did some work a long time ago.  In many cases the Eaton lists include
multiple people kibbitzing on one word, and the net should be that only one or
two of the proposals, if any, survive.  The reviewer should of course feel free
to propose his/her own alternatives, since the Lojban gismu list is so 
different from the TLI list.

Note that these are awful big files (several hundred Kbytes)

>old proposals

These are the hardcopy collections from old issues of The Loglanist, things
we came up with in making decsions on the gismu list for Lojban (many gismu
were eliminated by virtue of someone proposing a good lujvo), the Roget's 
analysis by Athelstan and me, etc.  I'm estimating 100 pages of handwritten 
scribbles, probably with anywhere from 10-30 average per page (I haven't looked
at this mess in a looong time.)

[Back to Eaton]

Ideally, at some time, we will go back and complete the 'Eaton interface' in 
our own version.   It is after all a reasonable standard for coverage of 
the basic lexicaon of Western civilization, even if outdated.  Buit then
we are talking some 7000 concepts, some of which I believe wiol have multiple
lujvo because Eaton did not always recognize that here concepts were not 
unitary (though she did better than a straight word-in-text frequency count).

>names

You are welcome to volunteer.  Mark's list will also be included.  I do not
consider that the arguments Ivan presented are convincing.  If we have to
go to non-Lojban quotes, then we aren't doing Lojban.  People in general 
won;t take to his idealogically pure concept.  They will Lojbanize country
and culture names.  If we don;t do it for them, they will do it themselves,
and the result will be more malglico than what we come up with.  Better to
do what we can (which means taking what you did before, plus anything fuurther
you care to do, adn completing it to the level of accuracyt we can, and then
put in a disclaimer about the limitations).

>Lojban-order dictionary

This is Cowan's preference, rather than a bunch of separate lists, which 
requirte people to know whether a Lojban word is a gismu, a compound, a
cmavo or whatever.  It is based on the perfectly reasonable argument that
most people EXPECT a single dictionary and find multiple separate lists
hard to manage.

The probabale result is that there will be two books, given the page counts
that have turned up.  One will be the combined E-L and L-E lists, and
hence be essential a pure dictionary.  The other would be a reference, and
would have the lists separated out as appropriate.  This will all be based on
what things look like when I first start putting all the pieces together,
rather than merely estimatring them.  Perfect binding a book with more than 
about 500 pages is asking for it to fall apart, and dictionariues get
particularly heavy use (we hope) and need stronger than average binding, hence
a thinner volume.

....
One item not listed that came to mind - you did an Esperanto-Lojban gismu list
once.  I don;t know how satisfied you were with the result (and it might need
to be exapnded into a by-place-structure position list for each place to be
truly useful for an Esperantist) but it would be a start towards breaking
free of pure English-based support for the language.  If we can do this in
our first reference book, it would give us moral high ground - alweays important
in teh IL community %^) (and not a bad precedent either since we seem to catch
queries from many overseas Esperantists)

>etymologies

Most people aren;t interested in these, and to tell the truth, they aren't too
much use as a reference item.  But then, most people aren;t interested in
English etymologies when they use an English dictionary either.  Given the at
least hypothetical claim that the Lojban gismu have a recognition score based
on cognates in the etymology, people who look at the etymologies may gain
some advantage at knowing the etymological roots, especially the non-English
ones that are not otherwise documented even by implication in any public source,
which presumably would be particularly helpful to those who speak one of those
other languages and still may not be able to recognize for themselves the
root/cognate that we chose.

(Note that this also opens us up to the appropriate criticism for when we 
screwed up the words from the other languages, which no doubt is often.  But
I favor academic openness as evidence of integirity.  45 pages is a lot, but
still less than 5% of the whole.  leaving it out hardly solves any problem
and the material needs to be published eventually.  The main weakness with the
current state of the etymologies, is that the non-English source words are
only aailable in their Lojbanized form since other wise requires going back
to the raw paper files to get the word we Lojbanized (which in 4 languages 
is not typically written in the Roman alphabet anyway - my excuse for leaving 
it as it is).

Note however, that there are a large number of people who are especially
attracted to Loglan/Lojban because of this etymological game, even if it is
of questionable acvadmeic worth.  It shows our heart is in the right place,
and this means a lot to some people, especially people who feel Esp-o's
and the other Euroclones' lists to be too culturally biased.

lojbab