Re: society-language

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

 First the question that came later:

Jim Grossman asks:
>2nd:    However, I can think of a definition of "logical language" that makes
>sense:  namely, a language whose grammar maximally mirrors symbolic logic
>notation.   Inasmuch as the notation is useful, I can't condemn conlangers
>for being interested in creating a speakable form of it, if only for fun.
>
>
>Is Lobjan speakable symbolic logic with lexical items in place of variables?

Among other things Lojban includes the apparatus of predicate logic, and a
predicate grammar is the core of the language structure.  But there is a bit
more than just variables.  We also encode the formal logical connectives,
have the apparatus for strict quantification, have a logic-based attitude to
negation (How do you negate the equaivalent of "Everybody loves somebody").
We added lambda calculus support.

Lojban can serve as a spoken form of the computer language PROLOG, since there
is some capability to encode each of its structures in Lojban, but Lojban has
other features that natural language usage needs, so it is not merely
"speakable symbolic logic".

The original comment was, in part:
Anthony Roberts:
>My point is that "exactness" in language is a very subjective thing,
>especially depending on where you're coming from.  The Japanese do not
>distinguish from the present, future, present perfect, imperfect, or
>present progressive forms; nor do they distinguish between our perfect and
>pluperfect.  But they are no less capable of logic and thinking than anyone
>else.  This begs the question as to whether a perfectly logical language is
>necessary at all, much less possible.  Any language must by definition deal
>in generalizations to a greater or lesser extent or become interminably
>bogged down in itself.

Are generalizations inherently unlogical?

Lojban deals with this by being very specific in the language design as to what
is "logical" and what is "vague".  Lojban metaphors are vague to any degree
necessary, and can even be figurative if specifically marked for same. But
they are still constrained as to structure.  Lojban has  intensional arguments
(i.e. where the arguyment is a description which may not be true of the
referent, but only exists in the speaker's mind.)  There is little that can
be dediced about the truth-value of sucg statements, since they are no more
precise or specific than the speaker intends - and that degree of intension
is ultimately only knowable to the speaker.  But even here the vagueness is
MARKED.

>distinguish singular from plural
I note in passing that thsi distinction is not found in predicate logic.
There is nothing more special about 1 than 2 of something, or vice versa.
The important numbers in logic are "none", "some", "all", and an implicit
"all others" that doesn't even exist explicitly in symbolic logic but is
sometimes required to be inferred.

>Now, while I wouldn't suggest that this extreme degree of generalization is
>what you may want in a conlang, it is important to note that a language can
>only represent thoughts and help translate and organize them (as others
>have noted here); it is not capable of directly mirroring the world or our
>thoughts, no matter how hard we try.  So any attempt at a "logical"
>language immediately becomes bogged down with preconceived western and
>personal philosophical notions of number, aspect, etc.  There is no such
>thing as a "logical" language any more than the fact ...

It becomes uinclear at this point what definition of "logical language" you
are using.  Any sense that a ;ogical language must incorporate only Western
philosophical models of number and aspect seems contrary to logic as well as
to our "logical language" %^)

>I'm not stating the above as gospel truth; as a matter of fact, if anyone
>can give me their own definition of a logical language I won't be pigheaded
>about these assertions.

The definition of Loglan/Lojban from days of yore includes:
1. encompasses the structures of predicate logic
2. is metaphysically parsimonious (it avoids requiring unnecessary distinctions
while allowing all sorts of such distinctions to be optionally made)
3. culturally neutral (in that we avoid patterning things after specific
languages, but rather try to return to some sort of analytical first principles)
This is of course only an ideal, but in Lojban, dominated by English-native
speakers, one of the worst criticisms is to be "malglico" - "damnably English-
like" with the emphasis on the derogatory.

Note that there is nothing in the definition about being precise (vs,. vague)
John Cowan has observed that "the price of infinite precision is infinite
verbosity".  This statement is especially true of Lojban.

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/";