Received-SPF: neutral (google.com: 68.230.241.216 is neither permitted nor denied by best guess record for domain of lojbab@lojban.org) client-ip=68.230.241.216;
Message-ID: <50F2D56C.8040405@lojban.org>
Date: Sun, 13 Jan 2013 10:40:28 -0500
From: "Bob LeChevalier, President and Founder - LLG" <lojbab@lojban.org>
Organization: The Logical Language Group, Inc.
User-Agent: Mozilla/5.0 (Windows NT 5.1; rv:13.0) Gecko/20120614 Thunderbird/13.0.1
MIME-Version: 1.0
To: lojban@googlegroups.com
Subject: Re: [lojban] What is the source of gismu *definitions*?
References: <d21ad182-d52e-4a61-8975-1c6f53d75211@googlegroups.com>
In-Reply-To: <d21ad182-d52e-4a61-8975-1c6f53d75211@googlegroups.com>
Reply-To: lojban@googlegroups.com
Precedence: list
Mailing-list: list lojban@googlegroups.com; contact lojban+owners@googlegroups.com
Sender: lojban@googlegroups.com
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: quoted-printable
X-Spam_score: -0.0
X-Spam_score_int: 0
X-Spam_bar: /

la gleki wrote:
> Every Lojbanist understands that gismu denote predicates that are highly
> practical.
> e.g. {pilno} includes a goal as pilno3. Indeed, how can we imagine using
> something without a goal?
>
> My question is who collected those definitions?

Me.

> Was it JCB?

 > How was this gimste formed?

JCB set the place structures for the TLI Loglan words.  His general=20
philosophy of doing so was set forth in his books Loglan 1 and Loglan 2,=20
though he didn't always follow his own principles.

I started with JCB's list, but greatly modified it, both adding and=20
deleting words.  As such, there are half-again as many gismu as there=20
were in TLI Loglan of the time.  In very few cases can I tell you for=20
certain the specific reason I added certain words, though for the=20
culture words I made an attempt to be systematic.  A large chunk was=20
added in 1988 as a result of Athelstan doing a thorough analysis based=20
on Roget's thesaurus, to make sure that we had good coverage of all=20
semantic domains.

During the period from about 1990-1994, I subjected all change proposals=20
to the LogFest attendees, representing the community, for approval.  In=20
the latter two years, a faction emerged favoring the elimination of some=20
gismu and thus keeping the total number constant, in the face of new=20
proposals, if not shrinking.  One last group of new ones was approved,=20
and the list was frozen.  Many years passed before any word was proposed=20
with significant justification, thus suggesting that this decision was=20
correct.  (If no one has really needed a word in 25 odd years of use, it=20
is hard to argue that it is fundamental, even if it might be useful.)

Place structures started with JCB's general pattern.  I attempted to=20
find patterns, and then to make words of similar semantic domain=20
consistent.  Thus all plant and animal gismu were to have a species=20
place.  I eventually got things fairly systematic, though I made some=20
mistakes.  At that point, pretty much no one besides me was looking that=20
closely.

I strongly avoided one-place predicates.

But at one point, I realized I was going too far, ascribing to any=20
possible tool a purpose, and to any object both material and form=20
places.  I backed off from this somewhat.  I thus avoided >5 place=20
predicates.  At about this point, the current concept of BAI started to=20
emerge, and it was realized that a large number of places were=20
superfluous.  I made one last pass, generally reducing many of the=20
excess places I had added.

Is there a changelog of modifications to gismu
> definitions starting from the first edition of loglan?

Not hardly.  I introduced the concept of configuration management in the=20
1988-1994 period, starting to document all changes once a chunk of the=20
language was baselined.  Before it was baselined, documentation was=20
rarely attempted, though there are some cases.  In only a few cases do=20
we even have good copies of the evolving word lists - this was still a=20
primarily paper and pencil project.

> My particular interest here is with the recent discussion of a possible
> new gismu meaning "qua". The corresponding word is of high frequency in
> Mandarin but in European languages it is often confused with words
> meaning {simsa}.
> e.g.
> "as" means both "like" and "qua".
> Russian "=D0=BA=D0=B0=D0=BA" [kak] means both "like" and "qua".

I have no comment on the merits of this, other than to merely observe=20
that many of the world's languages seem to do fine without making a=20
distinction.

The gismu list is baselined.  New gismu are not being considered, and=20
there is no plan to do so in the future, though this could be revisited=20
AFTER the existing language is fully documented.

> Were Mandarin predicates taken into consideration while constructing
> gismu definitions?

Not that I know of.  I did the Mandarin work for Lojban, and I don't=20
know Mandarin.

More importantly, almost no consideration of semantics was involved in=20
gismu-making.  If the basic meaning was generally covered, that was good=20
enough.  It was expected that the meanings and place structures would=20
evolve with usage.  (But by 1997, the community was tired of my and=20
other senior Lojbanists changing the language by fiat.  The community=20
wanted the language to stop changing in that matter.  Completely.  I=20
agreed with them.  We don't change the language by fiat anymore.  The=20
only exception, adopted for byfy use, is that stuff which is so broken=20
as to prevent good documentation of the status quo language, could be=20
changed so as to allow that documentation.  (Since then, sentiment seems=20
to have grown against "usage-based change" which is the other=20
alternative, and one that cannot really be prevented.  People generally=20
are biased against change in language.  They want books that are=20
prescriptive and unchanging, whereas lexicographers strongly consider=20
dictionaries by nature to be descriptive rather than prescriptive.)


I did use some systematic techniques to try to be sure I was picking the=20
correct root, and for a brief time, we had a native Mandarin speaker who=20
looked over what I had done with approval. (A couple of Mandarin=20
speakers since then have also said that the work I did was more than=20
adequate, but they were generally comparing us to Esperanto and other=20
Euroclone languages). I also used 3 different dictionaries in the case=20
of Mandarin in order to be more certain, since Mandarin has such a high=20
weight in Lojban word-making.  Still, there are flaws, and I think my=20
choice for Lojbanization of Mandarin was especially bad, being based=20
solely on the quasi-official Chinese description of the IPA=20
pronunciation of Chinese particles, and the system I used for mapping=20
IPA in other languages.  As a result, Mandarin inputs had too many "a"s=20
representing schwa, and too many fricatives were mapped to s and c,=20
leading to Lojban having a "she sells sea shells" quality that is hard=20
for some speakers, including me, to speak the language quickly and=20
accurately.

But I don't know enough Mandarin grammar to have any clue what subjects=20
and objects any given Mandarin word might require (if any)  I did enough=20
comparative linguistics study to be reasonably confident that my=20
approach was "good enough".

(Arabic is the other language where my word-making rules were systematic=20
but led to a relatively poor result.  And since Arabic has the lowest=20
weight of the 6 source languages, this meant that Arabic influenced=20
relative few words, and its inputs were less useful to Arabic speaking=20
Lojbanists.

JCB may have had some native speaker inputs in the early days, but my=20
general observations on his choices for word-making suggest that they=20
were even more limited and flawed than my efforts.  I know that we had=20
much better dictionaries by 1987 than JCB had in 1955.

lojbab
--=20
Bob LeChevalier    lojbab@lojban.org    www.lojban.org
President and Founder, The Logical Language Group, Inc.


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