GENERAL: concerns of a new Lojbanist about Lojban List discussions

[Logical Language Group <lojbab@GREBYN.COM>]

Richard Schroeder (hedgehog@riscsm.scripps.edu), a relatively new
Lojbanist, and I had the following private exchange.  The issues he
raises are of general interest, especially to the large number of less
experienced Lojbanists that we rarely see posting on this list (but who
occassionally ask me the same questions Richard did.

Among other things, we risk having image problems if the comments
Richard makes are perceived generally.  Ideas on how to deal with the
issues are welcomed.
============================

From: hedgehog@scripps.edu
Subject: Re: Cowan on morphology

>I agree that it would be nice to have more texts aimed at beginners, but
>people don't write such texts.

Too bad that they don't.  I seem to be one of the few people writing to
this list who isn't a professional linguist.  I don't think that I could
even give a linguistic definition of morphology.  I had thought that the
main purpose behind Lojban was the testing of the Sapir-Whorf
(spelling?) hypothesis.  The questions recently raised about the
esthetics and complexity of Lojban morphology (and about xo'e) seem to
miss some of the point about Lojban.  I think that the person who
decides to learn Lojban has decided that he might like to change his
thoughts to match the language.  If it takes a little extra mental
effort to learn the complexity of Lojban then that is just part of the
experiment.  It seems completely backwards for people to complain that
this or that aspect of Lojban doesn't match their preconceptions of what
a language should be.  We are supposed to be questioning some of those
very preconceptions.

I have read through the introductory material on Lojban and now it seems
there is no where to go from here but to memorize word lists (Yuk!).  I
don't want to go to the trouble of learning a gismu list this year to
find out that the list is going to be totally revised next year to match
someone's idea of what the morphology should be.  I don't want to try to
adapt my mode of thinking to the sumti positions of the gismu list only
to find that positions can be cancelled with a xo'e.

It might be much better for Lojban to have a little more effort directed
at encouraging education and propagation and a little less effort
directed at making it the perfect language.  I don't believe that any
language is without its own set of pros and cons.

I have been trying to learn to speak Lojban by communicating with
someone else on the mailing list who is also trying to learn.  I have
felt very discouraged because this is like the blind leading the blind.
When I told my Lojban partner that I was very discouraged and was
thinking about giving up, he told me that our conversations were the
best opportunity to learn Lojban that he has yet had.  I would like to
point out that my partner is also not a professional linguist.

So what's the advice of the Logical Language Group to someone like me?
Should I just give up and let the linguists keep tinkering with the
language in the hopes that it will be completed some day?  Is there some
action that I could take (keeping in mind that I am a rank beginner)
that would help in the production of more basic textbooks for the
learning of Lojban?  I tried to get some less linguistics oriented
discussion started (like how can one use kliru to distinguish between
fluorine and chlorine) but I didn't seem to be able to capture anyone's
interest (which was depressing because you seemed to think that it was a
very good question).

warm wishes,
*****************
Richard Schroeder
hedgehog@riscsm.scripps.edu

The opinions expressed are my own and not necessarily those of TSRI or its
funding agencies.

& r
>From lojbab
To: hedgehog@scripps.edu

Chlorine/Florine WAS a very good question.  I answered it, and probably
several people read it and maybe learned something.  But there is no
controversy, so there was no response.  Now if you felt that my answer
was invalid and continued to challenge, then someone else might throw
their two cents in.  (In general, while less true than it used to be,
people tend to take what Lojbab says to be gospel from ON HIGH and not
to be argued with unless it is grossly wrong.  This is not necessarily
good for a few reeasons, and it means that I try to shut up for
questions asked on the list-at-large for which I think there is liable
to be debate.

As for the debates on gismu morphology, lujvo morphology, le'avla
morphology, and xo'e - none of these should have any effect on anything
you are liable to learn.

The gismu list was baselined some 4 years ago, and there have been two
word changes since, both after public discussion of the reasons why (to
get a decent rafsi for a word that people wanted to use but which was
underused in its then-form), and about 25 additions to the list, none of
which are likely to be considered a split-up in semantic space of
another word (and hence actually changing the meanings).  There are
going to be some minor keyword changes with the dictionary publication
but they are not meaning changes but clarifications; a few earlier
keyword changes were also clarifications, but ones that might have had
some significant semantic import.  But, all in all, less than 3% of the
gismu list has had any change in any of the things that people have
typically emorized (no one has made any significant effort to memorize
place structures, and I don't do so or recommend it myself).

Discussion of changes in the gismu morphology are mainly a honing of our
arguments against the occasional outside critic.  There is not nor will
be any change, while the language is still called Lojban (and I don't
see any splinter groups bothering).

lujvo morphology is similar stable, though it gets debated more, and
criticized more, possibly with a little validity (but a vastly overblown
validity, if it is).  There have been no changes to that morphology
since 1982 except when we adopted the rules for word making formalized
for the version called 'Lojban' (we had some differences in what letters
can go together with and without hyphens from earlier versions).  There
was one special case 'bug' in the original rules that was found when we
codified the Lojban rules, but it is covered easily, and discussed in
the documents describing lujvo making. lujvo- making algorithms are also
not going to change while the language is still Lojban, for all the
debate that appears on the topic.

However, the rafsi assignments that have been on the gismu list and
published separately a few years ago were explicitly NOT a baselined
list, and a revision is about to take hm has been essentially stable since
1982, covering the WHOLE of the language.  However there are aspects,
such as lea'vla (borrowing) making, that have not been well-exercised,
and until the last week were not tested using a computer algorithm.  We
fo one, and any changes to the new set will cause cascading changes
to other rafsi assignments.  Indeed this happened to some extent for
this set of changes, and the volume of rafsi changes implemented with
the baseline is larger than it might otherwise be as a result.  The
stress of making this one change (and the fact that it has taken nearly
a full year to complete and approve it), means that it will not happen
again.  So you can learn the rafsi with confidence, as well.

The Loglan/Lojban morphology algorithm has been essentially stable since
1982, covering the WHOLE of the language.  However there are aspects,
such as lea'vla (borrowing) making, that have not been well-exercised,
and until the last week were not tested using a computer algorithm.  We
found bugs in the algorithm, almost entirely caused by incomplete
specification of the rules.  The 'urgent' issue I talked about a few
days ago is handled in my change proposal by simply defining some
previously undefined terms more carefully, and more restrictively than
they have been presumed to be while undefined.  As a result of the
proposal, probably no exisiting words of the language would change while
there would be a restriction in future words added to the language
through a process that has as yet not even been used by real speakers.
(this process, called type IV borrowing, is sufficiently tricky that we
do not recommend it for ad hoc conversation, but intend it to be done
authoritatively when appropriate, using a computer to check for
validity.  All le'avla made for adhoc use in coversation have been of
types I (quoting the other language), II (turning the foreign word into
a name and then using Lojban grammar to convert it to a predicate), or
III (using some simple Lojbanizing rules and then prefixing a rafsi in
such a way that one never runs afoul of the complications of the type IV
restricitions, which mostly affect the beginnings of words) . The change
being debated, and the problem itself, reflects our conservatism - the
problem arises BECAUSE the type III le'avla have run afoul of those type
IV rules, and the proposed change is the further restrict those type IV
rules so that the type III le'avla that have been made are not invalid.

The Lojban grammar, also baselined, has actually changed more than any
other area of the language under baseline control, but you didn't
mention it as an issue.  It shouldn't be, because even 'major' changes
to the grammar, tend to be unnoticeable to all but the most advanced
users of the language.  Under baseline control, and therfore with lots
of opportunity to debate eabch change, there have been over 50 changes
to the grammar.  Yet only minor aspects of a text written before the
grammar baseline are liable to have changed (and in indeed, I suspect
that more changes have occurred to the cmavo used in the language than
to the grammar that the cmavo implement - but those changes - to grammar
and cmavo, are again mostly additions).

Thus my conclusion is that there should be no fear of memorizing any
aspect of the language design for fear of it changing on you.  One thing
to bear in mind - any thing that those of us who are ahead of you can
manage to agree to change - WE ALSO have to rememorize.  This tends to
keep proposals to an absolute minimum.  I've memorized the most elements
of the language of anyone (as far as I know, I am the only one who made
a significnat effort to learn the rafsi and only my wife and I and John
Cowan have studied the cmavo list systematically).

As for teaching materials - they are in process, and indeed most of the
change proposals that you see are offshoots from that process of
codifying what has not yet been codified, and making sure that the words
that have been written down in the past are actually accurate as they
have come to be understood in actual usage.  The reference book I am
working on, is estimated right now at almost 1200 pages when done if I
include everything, and producing that much paper requires a lot of
checking.  The other books, which will follow, are also partially
revised, and in the case of John Cowan's reference grammar, well-started
on the first version.

I did not notice that you posted this to the net, and I am answering to
you privately.  If you did post it to the net, please post this reply
also (and I encourage you to post if you have not, both query and
answer)

lojbab

From: lojbab
To: hedgehog@scripps.edu

More comment

First of all - none of the rest of us that are spouting off are 'trained
linguists' either.  Cowan and I in particular are totally self-taught,
and I was where you were 6 years ago when I started to split from JCB on
the language - totally ignorant of the jargon of lingusitics.  The
conlang people use a lot of linguistic jargon, though most of them are
also amateurs, hence we tend to so so in response, since we understand -
debates on the linguistic adequacy of the Lojban design are
fundamentally linguistic questions (at least we want them to be), and
hence need to be discussed using the terminology of the field.

The capability to use Lojban for any particular research prupose
requires that we do the best we can to prepare the langauge for that
purpose.  The use of Lojban for sapir-Whorf requires that the language
be as well-defined as possible in the formal, logical propoerties that
are presumed to be its unique feature that will cause S/W effects.
Similarly, the unambiguous grammar that is our main claim to validity in
computer nat lang processing requires us to be intolerant of weakness in
that area.  These areas ARE well-defined for Lojban, but as our most
critical design elements, we like them to bear up under the tightest
scruteny while it is still at least theoretically possible to change -
but the kind of discussion now on the net is not the kind that will lead
to change (except for the le'avla issue which will lead to some minor

making a lujvo that might have a more mushy plce structure definition).

As for another point I caught on rereading - if yo cmavo and
will exist as an option to speakers that alters the logical semantic
properties (hence the first area), and which I will tolerate primarily
because the existence of explicit marking of a deviation from a standard
place structure means that we can more forcefully argue for literal
adherence to the norm when people do NOT mark with "xo'e".  (Note that
nothing about "xo'e" as proposed or implenetd by anyone requires any
change to how you learn or interpret place structures - indeed, it is a
way to do what people want while formally piggybacking on the place
structures as defined, rather than the cleaner must less formal way of
making a lujvo that might have a more mushy plce structure definition).

As for another point I caught on rereading - if you contact Colin Fine
(who I think I identified for you earlier, and in any case he has posted
recently to the list), he is a) in the UK, and b) interested in and
actively supporting Lojban conversation at levels from beginner to
advanced and c) extremely skilled in the language having been expert in
Loglan when I was first learning about it in 1981 (though he later
dropped out for a while due to the inadequacy of the language and the
organization in the early 1980s).  He can support you and others via
normal email, or if you have access, via IRC (internet relay chat) which
is a way of holding real time conversation on the net that somne people
have access to.  (You may not accessing through a BBS, but who knows.)

lojbab
===============
EOT