semantics - respond to And

[Logical Language Group <cbmvax!uunet!cuvmb.cc.columbia.edu!lojbab>]

And wrote:
>Lojbab:
>> Furthermore, I'll claim that NO ONE has the competence to define a
>> prescriptive semantics, because there is no unambiguous mode of
>> expression to communicate semantics.  Hence we can only through
>> induction come up with a model of semantics.  I believe that that model,
>> if created, would differ for every human being, since semantics is a
>> function of personal idiolect, not of a language as a whole.
>>>
>> (Translated to English, that means "words mean what I want them to mean"
>> with "I" moving with each new speaker of the phrase.)
>
>These are strong claims that cover natural language, not just artificial
>languages.
>
>(1) I don't know what is meant by "unambiguous mode of expression to
>communicate semantics".
>
>(2) Research in semantics, as in syntax and chemistry, requires a
>certain amount of induction.  When inventing a language, though, one
>makes it up - hopefully with an eye to what goes on in natural language.
>
>(3) Re.  Humptydumtyism:  our success in communicating with each other
>is solid evidence that there is a consensus about the meanings of words.

I answered this somewhat in a message to And copied to the list.  But
some further thoughts that may make my position clearer.

wrt (2), perhaps Lojban is different from other invented languages in
that as a matter of principle, "one" doesn't make it up.  In a sense, at
least two do - a speaker and a listener trying to communicate.  In the
other sense, many are inventing the language because >I< am not the
language inventor.  Loglan went to the public three times, in 1960,
1966-70, and 1975, and the last time, it 'took' to the extent that the
"Loglan Central" that everyone talks about consists of those people who
got interested then and stuck with it, even to resisting JCB's attempts
to take it 'private' again starting in 1982-4.  There is now a community
'inventing' the language.

But, then referring to (3), I disagree.  Among NL speakers, and among
Lojban speakers, there is NEAR consensus about the meanings of words.
But not full consensus.  For anyone who doubts, I can cite a 3 hour
discussion last LogFest as to what a 'book' is, and therefore what
"cukta" should be.  Is a "book" an object with pages (possibly blank)
bound in some manner:  x1 is a book with pages x2 displaying x3 bound in
manner x4, or is a book the contents as well as the medium thereof - a
length of text written by one or more authors and presented to an
audience, in contrast to "vreji", with some intent at long-term
preservation of the expression not implied by "notci".  The place
structure was written to the latter intent, so as to cover the books of
the Bible as well as the Bible, and to enable ancient text on scrolls or
an encyclopedia stored on CD to also be called a book.  The choice is
debatable, and (since I was just looking at cukta, I'm going to add a
place for "preserved in media x5" since the preservation IS significant
to the concept.  The other type of 'book' might be jornypapri
"joined-pages", or selgacrypapri "covered-pages", which could be a
medium to go in the x5 place.  (Other possibilities are welcome.)

But 3 hours of discussion did not achieve these definition as a
consensus.  Indeed, I think there was an agreement to disagree, with the
status quo (rather loosely defined but deemed sufficient at the time for
those of us favoring it) being preserved in the gismu list.  Yet "cukta"
(and "book" for that matter) has been used by those of us present at the
discussion with adequate communication even though there was no
consensus.  (Unless you call conceding to the status quo a consensus.)
But if my change in 1992 is accepted, then the meaning has changed at
least in a formal sense, even though I doubt that my concept of a book
has been changed.  I just realized an aspect when writing this message
that was always significant, but had never been brought out in the cukta
discussions before.

I cannot imagine holding several hours discussion of every Lojban gismu,
much less every Lojban word.  We suffered near-terminal burnout at last
month's weekend meeting which tried just to decide on a small number of
suggestions for changes and clarifications.  Since we ourselves aren't
necessarily going to realize the details of the sense of the word we are
trying to define in the place structure, I think it is far better to let
the community USE the words, hold arguments about those usages as Nick
and Mark do about the word choices in Nick's 'Wallops', and record the
results.  The language will grow much faster than we can record the
discussions, but so what - the point is to have a language - a means of
communication, not a document about a hypothetical language.

I also note that, no matter how well we define the place structures, no
one will memorize all, or even most of them.  People will thus guess, or
otherwise use words other-than as the dictionary intends at some point.
In some cases, a guess will take hold and beecome widespread.  If we are
to assume the standard linguistic dogma that language is what people
use, then when the dictionary contradicts a pattern of usage, it is the
dictionary that is wrong.  Thus a semantics defined in isolation of
usage is meaningless.

lojbab

{P.S. I had more to say responding to And's comments on colors, which we
treated specially, rather after the manner of Bruce's language X, but I'm
going to be trying to cut down on the talking since I'm spending too much
time at it, and I have a book, some very late orders, donation receipts,
a buget, and a JL to get out.  Sorry if I seem a bit quiet for a while.

}