Conceptually separating language from biology

[Dan Everett]

On Wed, 4 Dec 2002, A. Katz wrote:
>
> Suppose it turned out that one of our Funknet posters was not a
> biological entity at all, but rather an AI program implemented on a
> computer and connected to the net. Let's say none of us were able to
> tell the difference. The poster not only wrote grammatically and
> comprehensibly, but was also able to maintain topic continuity and
> discourse coherence and to interact productively with other posters.
> If this were the case, would you not concede that the AI entity was
> using language? Would it matter that it didn't share our DNA or our
> brain structures?
>
>

Yes, it would still matter. What you have just described is the Turing
Test. Searle effectively showed the inadequacy of this kind of test with
his Chinese room scenario. Ultimately, we have to understand something of
the presence or no of the intentionality of the entity communicating, in
order to tell whether it has semantics or just syntax. Using language
alone is not enough, because, for example, AI programs can use syntax
alone, bypassing intentionality, semantics, and consciousness. And these,
I believe, and Searle convincingly argues, are so far as we know only
found in biological entities.

Dan