Message-Id: <199512192358.BAA21422@xiron.pc.helsinki.fi>
Date:         Tue, 19 Dec 1995 15:43:50 -0800
Reply-To: "John E. Clifford" <pcliffje@CRL.COM>
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: "John E. Clifford" <pcliffje@CRL.COM>
Subject:      lojban changes (were Dialectology and Grammar change and...)
To: Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 5551
Lines: 92

I have just finished working through the sometimes rather edgy
discussions of grammar change proposals, lojban dialects and the
mechanism for changing lojban.  As a survivor of this history, I
feel called upon to throw in my kopecks (I think there may be a
more worthless coin but I can't think of one).

I have learned Lo??an four times ('60,76, 82ish, and 89ish), to
varying degrees of proficiency but always to at least to the point of
doing literary translations with dictionary and grammar (i.e. Latin I
in high school).  I think I can claim to be dedicated to Lojban's
welfare etc. etc. and to be a likely learner and so on and I can
GUARANTEE xorxes and whoever that I am NOT going to learn
the language again until I have a very firm guarantee that the
sucker is not going to change again for at least five years.  That
will give me time enough to get a good run at the Psalter
(including picking up some long-lost Hebrew) and getting the
Lojban-to-logic stuff at least plausibly begun.  So, yes, the needed
Lojban community does hinge on getting a refgram and a
dictionary into people's hands -- until that happens, the most
reliable folks are not going to join in.

I also can say that I have seen several management styles in this
history.  In a sense, the best was the original JCB style (based on
Volapuk and Esperanto).  He put the sucker together, published the
book and said that was the way it was to be.  Unfortunately, JCB
came to listen to the carpers and the reformers and the
perfectionists (and -- embracing all three -- the logicians) and he
opened the language up for discussion and revision and eventually
revolution.  Then he practiced participatory autocracy: he let
everybody talk, ignored everything they said, did what he wanted
to do, and then tried to make it appear that the result was what
everyone had said they wanted.  Happily, (for the most part) JCB
had a clear vision of his language and so what he came up with
worked pretty well, actually solved problems people had raised
("ignored" was maybe too strong a word, he came somehow to
discover for himself what others were trying to point out to him),
and -- down to the Great Morphological Revolution -- kept the
language fairly downward compatible.  Unfortunately (for
governance, but not for the language), it also gave the discussants
the idea that they had some stake in the language.  When they tried
to exercise that assumed stake, they were cut off and informed that
they had no such stake for it was ALL HIS, as indeed it was in
most real senses.

Lojbab has tried for the inversion of this style.  Beyond the basics
that were required by Lojban being a loglan, he has tried to make
the discussants stake a reality.  The discussants have not
cooperated at all, for they have not come to the point or have done
so but rarely.  This was, of course, the pattern in the old days (and
in conlangs generally) but in the old days it did not matter, for JCB
just did what we wanted and so discussion came to an end
effectively (it didn't stop, of course, it just ceased to be a part of
the process -- superfective as ever was).  Lojbab is now faced with
the need to finish the process (well, this phase of it) without the
official steps really having begun: clear proposals succinctly
argued for, with demonstrated need and impossibility of
presentation in the existing format, etc. etc.  Then debate and a
vote or, at least, a consensus.  From pieces of each of these: bits of
a justification, failed attempts to find a mode of expression,
sketches of proposals, he and John must construct the first definite
products.  But how?  Autocratic decision is officially out; the claim
of a consensus immediately draws a slough of new dissidence.
Next time the rules will be clearer, I hope.  For now, only the ex
post facto baselining -- the last thing that got officially agreed to --
can fit in.  But that is clearly flawed in known ways.  Can we live
with it for five years?  Yes and no.  Yes officially and no, we won't
actually pay it that much heed once we get going in the language.
JCB, with characteristic modesty, calls himself the start-giver and
that is just what we need right now.  We'll make do and then agree
among ourselves (maybe peeking at the YACC grammar a bit for
comfort) to do a little differently, as indeed, some have been doing
right along without making any noise about it.  Virtually no Lojban
text on the list is strictly legal, though much of it parses more or
less correctly and is intelligible, but it pushes various edges of the
envelope as the need arises.

So my suggestion is publish and be damned (Wellington).  Solving today's
problems will not make the published text more problem-free and may (as we
have seen with earlier fixes) end up making matters worse a couple of
years down the line, when the data really comes in.  But, with a fixed
target, the data can be generated by a significantly sized community of
serious learners and the first round of reforms -- in five or so years
(hopefully longer, or sooner for overwhelming reasons) -- will be the
better for it.  I suggest that the list for that first reform begins, say,
September 1, 1995.  We can live with guheks (well, I'm not sure I can) and
without fuzzy fussiness for a while, as we have these 40 or 8 years.  And
we can have plenty of time to build our cases for this and that, ready for
January 1, 2001 and the millennium. (Sure, I'd like to see some of the
things I *know* are missing added, but I can wait and get them even
righter in five years.)

pc>|83