Separating language from biology

[A. Katz]

On Thu, 5 Dec 2002, Dan Everett wrote:

>
> > >Date: Thu, 5 Dec 2002 06:48:53 +0000
> > >From: Dan Everett <Dan.Everett at MAN.AC.UK>
> > >Subject: Re: Conceptually separating language from biology
>
> Right. Fine. But your test ignores semantics. And, moreover, there are
> many people, myself included, who find the evidence strong that language
> and consciousness cannot be separated. But ignore that. Just focus on the
> semantics.
>

I, too, suspect that consciousness has something to do with it, but that's
not useful as long as none of us has a working definition of consciousness
or any tests for its presence or absence.

Insofar as my test ignores semantics, so do all the tests conducted on
human beings. If by semantics you mean what the
speaker "really" meant, or whether he really meant anything at all, our
best clue is language use in context. The AI entity I postulated passed
that test. Do you have a better one to suggest?

The reason reanalysis is a tool for language change is that people don't
always understand utterances the same way. And the reason language is
useful despite these misunderstandings is that language used in context
transmits information despite the differences between and among
individuals. If someone seems to understand us, and we never find out
different, then for all intents and purposes, he really does. That's the
Turing test in action in everyday life.

If a child says: "I want a banana" we all assume that he knows what he's
talking about. If a chimpanzee says the same thing, people ask: "Yes, but
does he have a theory of mind?"

I think we should have a level playing field. Whether it's a human, a
computer program or a chimpanzee, the test for language should be the
same.


       --Aya


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
http://www.well.com/user/amnfn
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++