Message-Id: <m0pgDVU-0000R4C@xiron.pc.helsinki.fi>
Date:         Tue, 15 Mar 1994 00:16:52 EST
Reply-To:     Nick NICHOLAS <nsn@VIS.MU.OZ.AU>
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         Nick NICHOLAS <nsn@VIS.MU.OZ.AU>
Subject:      Re: What have we learned in developing Lojban?  Anything
              publishable?
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
In-Reply-To:  <199403140631.AA04827@access3.digex.net>; from "Logical Language
              Group" at Mar 14, 94 1:31 am
Content-Length: 3422
Lines: 63

> What have we actually learned from the way people have tried to use
> Lojban and earlier versions of Loglan?

This is a challenging and pertinent question. I presume that the linguists
among us may want to use Lojban in their work; if they do so, they will
certainly need to justify the academic relevance of such a language.

I have no ready answer, but my impression is somewhat somber.

> My best candidate of the moment for something significant learned solely
> through usage of the language and resulting changes to the prescription
> has been the existence of the phenomenon we call 'sumti-raising'.  We
> might also have learned something from the steady increase in use of
> non-logical connectives, but I'm not sure what in particular.  Study and
> usage of many of the more novel features of the language (including
> tenses, attitudinals, and lujvo semantics) is still at too preliminary a
> stage to say that human usage has significantly affected design or
> concept.

Object- and subject- raising have been thoroughly treated in syntax. Our
engineering approach to language, attempting to prescribe presence or
absence of raising according to semantic criteria, is worth an article (so
save up those discussions on place structures!), but I doubt anything
earth-shattering has been achieved; more like an "oh, that's interesting"
data point. Granted, given that there's still on-going discussion, apparently,
on whether causatives have patient roles distinct from the cause event
(a straightforward case of leaving a sumti place "clefted", in our terms),
it's a significant point --- I'm just not sure it hasn't already been argued
by somebody.

The non-equivalence of logical and natural-language connectives
has likewise been long known amongst philosophers of language; McCawley's
book on logic and linguistics has a lot to say on the matter, and indeed
this is a large part of the philosophy of linguistics literature. That we
increasingly use non-logical connectives simply means we are starting to
appreciate this --- but since most of us are informed in logic, we're a
skewed data set.

As far as lujvo semantics are concerned, I am informed that
the semantics of derivation has been done to death; most Lojban compounds
do not in fact involve derivation, and the engineering orientation of
combinatorial predicates is probably interesting, but probably not of
mainstream interest to semanticists.

I suspect one of the more important
contributions Lojban can make (by design!) is in Pidgin linguistics, and
Language Acquisition studies; it is a deviant language for both concerns,
yet even here I'm not sure anything it has to say is earth-shattering. Ditto
for discourse analysis/ text linguistics (which is impinged upon by the
development of Lojban stylistics).

What have we learnt in Lojban? A lot of linguistics, to be sure --- but
most of it the linguists already know. Even more language engineering ---
but that's not linguists' domain per se. There are bound to be some surprises
due in usage, but they'll have to wait for the right data collector.

Since I don't have the overview of Lojban Lojbab has (and since I've
regrettably gotten distant from Lojban, initially by being distracted by
Klingon, now by discovering overwork for the first time in my academic
career), I'm willing (in fact, given my conclusions, eager) to be
contradicted.

Nick, who'll have his files back tomorrow.