[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lojban] Comments on the New Policy

At 12:42 AM 11/30/02 +0000, And Rosta wrote:

I support the great majority of Nick's views and the new
policy document. Below are the points where I disagree to
a lesser or greater extent.

A. Experimental cmavo
1. The shape of a cmavo should not determine whether it is
officially documented. If a sufficient number of people agree
that there should be a cmavo with meaning M, and if there are
no available CVV cmavo for M, then N should be assigned to a
CVVV cmavo and documented.
I think we should use the CVV and then the xVV (the reason for reserving xVV seems to be moot given the wide use of CVVV for experimental cmavo these days) before we start assigning words from CVVV. 


2. Many debates about the meaning of a cmavo, C1, come down to
whether the cmavo should have meaning M1 or M2, where both M1
and M2 are legitmate, reasonable, desirable, etc. The easiest
way to settle these debates is to pick one meaning (M1) for
C1 and assign the other meaning (M2) to a spare cmavo, C2
(which of necessity must be a CVVV) cmavo.
It need not necessarily be CVVV. There are a dozen or more free CVVs and almost 30 xVVs available. 


If the current
notion that CVV cmavo are official and CVVV are unofficial
continues, then the debate will not be solved so easily, since
any text using C2 will fail to parse, and many Lojbanists will
dismiss it as nonstandard or experimental Lojban.
Many Lojbanists will indeed dismiss experimental Lojban. Others will use it. Whether experimental cmavo will parse depends on the sophistication of the parser. I suspect it would not be hard to have a parser which queries the user for the selma'o of an experimental word, which means that the only problems would be experimental cmavo that do not fit any standard selma'o. 


The solution is to make both C1 and C2 official and documented.
Believe it or not, I agree with you entirely and proposed the identical solution, which is allowed for in the Board's policy. 


3. The baseline should accept that future study and usage of
Lojban will reveal the need not only for additional fu'ivla but
also for additional cmavo (and perhaps even additional gismu and
rafsi, as witness the problematic absence of a gismu for "intend").
The baseline should assign existing cmavo a clear definite meaning,
but should not be taken as permanently defining what is and isn't
a standard official cmavo.
Then there is no freeze in the language design, and some people have indicated that they will refuse to learn a language that is not frozen. It is true that simple additions are the easiest form of change for a learner to accommodate, but any set of words that is expandible is not frozen (and if the changes are not strictly managed, then the list is not even baselined).
If the language design is so incomplete that we cannot go for even 5 years without adding fundamental structure words to the language, then we have to admit that the language design is just that: incomplete. After all, how many structure words have been added to English in the last *50* years? 


4. If the only objection to the above is that it should be possible
to tell from a cmavo's shape whether it is experimental or official,
then a portion of CVVV cmavo space should be defined as official
(but not necessarily used) and the rest as experimental, just as
is currently done with CVV space.
If that were made necessary by the lack of sufficient CVV space, the byfy has the power to do so - that would be a "fix" to some specific problem - a lack of cmavo space to resolve a definition question. But I don't think there are enough cmavo in question (I hope there aren't) that this would be needed.
In the long-term, after the 5 years, what you say will be necessary. Usage will determine that certain experimental cmavo should be retained because of their usage, and if that is the case, then the usage itself will mandate that we keep the CVVV that exemplified that usage. 


B. Zipfeanism & Lojban's serving its speakers
1. The design of Lojban is such that, quite unnecessarily, it is
difficult to be both concise and logically precise. John has often
said "The price of infinite precision is infinite verbosity", but
this does not apply to logical precision, since it is finite. It
would have been quite possible to design Lojban so that it was more
concise, but concision was never a design goal.
Zipfean concision was indeed a design parameter, since JCB talked of it extensively (and that was my standard for identifying design parameters). 


However, most Lojban
users care a great deal about concision and it is a major factor
influencing their usage. "Saving syllables" is important to most
Lojbanists. At the moment there are very few Lojbanists who care
about Lojban being precise, but I predict that as more Lojbanists
become comfortable with elementary logic, more Lojbanists will yearn
for a Lojban that is both precise and concise.
Perhaps. I yearn for a Lojban that isn't changing at a rate that is noticeable on a month to month basis. 


2. It is natural in language that high frequency words and phrases
get shortened. Low frequency words can be short, but high frequency
words tend not to be long (even if they look long in writing, they
are likely to get shortened in speech). Lojban acknowledges this;
it is the rationale for the rafsi system, and for the notion that
there should be a rough correlation between a lujvo's length and
its frequency. But Lojban has no way to shorten high frequency cmavo
or cmavo sequences.


That is a statement of fact in the design.


4. I therefore predict that as more people are both jboka'e (caring
about usage) and jboskepre (caring about precision), the antizipfeanism
and longwindedness of Lojban will be felt more and more acutely. The
more competent a Lojbanist is, the more acutely the problem will be
felt.
Then no doubt we will have to address it if and when it happens. Sounds like a good topic for the post-baseline all-LOjban discussion, since the competent Lojbanists will be the ones capable of discussing it. 


5. If Lojban is to serve the needs of its speakers (above all, the needs
of those who actually use it and know it thoroughly), it must be willing
to change in a planned, organized, designed way. **Instead of committing
itself to a baseline freeze (i.e. a policy of No Change), the LLG should
have the policy of No Change Without Consensus.**
That is (more or less) the policy for the byfy work, bearing in mind that there is unlikely to be consensus to "fix" what is not broken. After the byfy is complete, then there will be a freeze. If you do not accept the necessity for a 5 year freeze, then you should indeed vote "no" on the policy, because there is no resolution that would satisfy both you and the proponents of a freeze. I can pretty well guarantee that there will be no consensus on a policy that does not include a freeze.
Opposing a long-term freeze entirely is not merely objecting to a detail, but a disagreement with the thrust of the policy as a whole. 

lojbab

--
lojbab                                             lojbab@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA                    703-385-0273
Artificial language Loglan/Lojban:                 http://www.lojban.org