Comparison: Loglan / Lojban ?

[Logical Language Group <lojbab@ACCESS.DIGEX.NET>]

Nick Summers asked:
>A little while ago some very definite preferences
>were voiced in relation to these two loglangs,
>without any real reasons being given. Would anybody
>care to explicitly state their reasons for preferring
>one over the other? I suppose I am interested in
>opinions related to linguistics or logic rather than
>history/politics.

The split between what is now LLG and The Loglan Institute was essentially
purely political.  However, due to Dr. Brown's copyright claims on every
word of the language, we chose to go back to the most basic level, do our
own linguistics research, and "reinvent the language".

Notwithstanding the comments of others, there has really been very little
drifting apart in one sense.  Almost any text written in current TLI Loglan
can be mechanically translated by word substitution into equivalent Lojban.
You have to be careful about place structures, but there is little in the TLI
language that does not have an exact Lojban equivalent.

Lojban on the other hand IS arguably superior both logically and linguistically,
for a variety of reasons, some of particular interest to conlang designers,
especially those seeking an international audience.  Lojban may effectively
be described as a superset of TLI Loglan - we have added many features that
are linguistically necessary or at least very useful.  In some cases, TLI
has followed the exact same path, a few years later - which is one reason
why "drift" hasn't much happened.

Stylistically, there is much greater distance.  Almost no one actually writes
in TLI Loglan (and there are no known speakers).  The written text that is
published or which appears on the net is riddled with English-origin metaphors
and idiomatic semantics.  TLI writings have an appearance of greater "logicness"
because the writers tend to use the formal logical style more, but since they
do so in a flawed manner, the results is usually more humorous than
communicative.

Lojban has several people who have maintained conversation in the language,
and Nick Nicholas is perceived by others as speaking at fluent speeds,
though he feels he is still one notch short of that level.  Both speech and
writing in the language is "natural" in style, with logical constructs used
only when logical precision is useful.  On the other hand, the basic predicate
nature of the grammar and other features tend to make Lojban usage a tad more
precise than most natural language (stylistically, I mean - people TRY to be
more precise).

Lojban carries through on a couple of Loglan design goals that we believe
TLI has failed to maintain, specifically unambiguous resolution of a speech
stream into specific words, and formal unambiguity in the grammar.  The
arguments on each of these features is fairly technical and TLI continues
to make improvements with each iteration of announced changes to their
language, so I am not absolutely sure how many problems still exist.

TLI Loglan phonology was devised using "taste tests" by native English
speakers.  The phonology therefore abounds in sounds and clusters that
are familiar and easy for English speakers, but TLI abandoned an earlier
more linguistically regular and neutral phonology. The Lojban design returned
to linguistic principles in setting the rules for consonant clusters,
and the result is clusters less likely to be misheard.  Lojban phonology is
also defined in terms of IPA sounds.  As an example of TLI Loglan problems
of this sort, the current edition of Loglan 1 says that "e" before vowels
is pronounced as "eigh" as in "eight", but the supposedly distinct
diphthong "ei" is pronounced as [ay] as in "day".  What the difference is
between these two is most unclear, but the result is that the vowel group
"eio" is indistinguishable from "eo".

The TLI Loglan tense system is simple and probably inadequate for complex
communication needs.  With Lojban, we had the benefit of analysis by
tense logician John Parks-Clifford (pc) formerly the president and chief
linguist/logician for TLI.  The Lojban tense system now includes a
perfective system that has come to be at leats as commonly used as the
simple tenses, as well as a variety of modal modifiers.  The TLI complex
tense system is effectively unanalyzed, and we believe it still to be
grammatically ambiguous, though this cannot be checked since the current
formal grammar is a "trade secret".

Lojban reanalyzed the problem of emotional attitude expression, using among
other things the ideas in Elgin's La'adan, with the resulting language much
richer and hence in a sense less "logical" in that it is more capable of
finely distinguished emotional expression than probably any natural language.

Another area of significant difference is that we worked with an existing
semantic problem involving varying levels of abstraction (the phrase
"object raising" is used in linguistics, and we use "sumti raising" in
the more generalized Lojban case) within the semantics of many predicates.
An example of this is the distinction between "The food is done" vs.
"The preparation of the food is done".  Lojban requires that the former
be marked because "the food" is "raised" from the abstraction "the preparation
of the food".  TLI Loglan has no way to mark this.  The result is semantically
very muddy.

Lojban has a complete system for dealing with numbers and mathematical
expression.  TLI Loglan has an admittedly incomplete "stub" of a design.

These are just a few of the differences.  Lojban allows one to be as
logically and semantically sloppy as one wants, but has added the mechanisms
needed to fix many known problems.

And this is in the final analysis the reason teh languages differ: Lojban
has seen considerable "live" usage, with the results plowed back into the
design before we called the language "done".  Lojban now is frozen for at
least 5 years, and any later changes will probably be made solely by
fluent Lojbanists.  TLI maintains its intention to keep fiddling with the
language design indefinitely, whcih in turn keeps a lot of people from
even trying to learn it.

lojbab
----
lojbab                                                lojbab@access.digex.net
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA                        703-385-0273
Artificial language Loglan/Lojban: ftp.access.digex.net /pub/access/lojbab
    or see Lojban WWW Server: href="http://xiron.pc.helsinki.fi/lojban/";