[lojban] Re: some thoughts about lojban use and future

[Michael van der Gulik <lojban-out@lojban.org>]

Hi all.

xah lee wrote:

> The task of having computers understanding natural languages at 
> parsing level is basically solved. For example, today there are lots of 
> online translation softwares from and to many natural languages and 
> they do manage. To go beyond translation to real understanding as to 
> generate a proper response, involves AI that is still pure science 
> fiction.
>  
I think that lojban could be a step in that direction, and not because 
the language is parsable. I personally am learning lojban to improve my 
logical insight. I believe the path to understanding how to create a 
true AI is by understanding how we think.

Also, as a sidenote, a true AI wouldn't need to be programmed to 
understand a language but would slowly learn it like a child would. And 
as a first language for an AI to learn, I believe that lojban is far 
better suited than English.

> I don't think lojban will befit as a language for human communication. 
> Suppose if we have a region to speak lojban from now on, and lojban 
> being the first language of all newborns. I believe in a few 
> generations lojban will quickly loss all its design premises of 
> regularity and mathematical properties. This i believe because humans 
> do not process information as a logical machine, even if all their 
> thoughts are in a logical language.
>  
Umm.. yes and no. If nobody formally learned to read and write, then 
their language would gain and/or lose regularity  and change rapidly 
because the speakers use the language however they wish with no regard 
for the rules of that language. I believe that English as a language is 
evolving much slower than it once did because the speakers of English 
are now receiving proper schooling on the grammer and use of English.

If lojban is taught (teached is a better word..) properly and its 
speakers were keen to use it correctly, then it should become more 
regular rather than less.

> as to the worf(sp?) hypothesis that language effects thought, it is 
> obviously true. I think it is moronic to believe otherwise.
>  
That's a very logical statement! X is obviously true, and for all Y 
where Y is a person that thinks not-X, Y is a moron.

The sapir-worf hypothesis in my mind holds true because in order to use 
the language constructs, the language speakers must learn the underlying 
(underpinning?) theories and techniques of that language. Learning 
affects thought, thus the language learners thought is also affected.

> I think the way for lojban to prosper is for it to remain in the hands 
> of logic community, as it is today. Lojban is a great tool in studying 
> the relation of meaning, logic, and natural languages. It is an aid in 
> the field of philosophy of languages or analytic philosophy. I imagine 
> the utmost achievement for lojban is for example having mathematical 
> treaties in lojban, and for it to be heavily used in AI research, 
> machine to machine or human to machine.
>  
Hooray to that!

Some ideas that I have had:

- Describing mathematical concepts and equations would hypothetically be 
easier in Lojban than English.

- Doing mathematical equations / Logical proofs and theories in Lojban 
rather than mathematic notation could be possible.

- Lojban would make an excellent tool for gathering user requirements 
for software engineering. Hypothetically, one could describe a computer 
system in Lojban, and then perhaps have a computer automatically create 
that system. Like prolog, but even more descriptive.

Mikevdg.