Re: Aaron Sloman's "The Irrelevance of Turing Machines to AI" article

From: Wolf Kirchmeir (wwolfkir_at_sympatico.ca)
Date: 08/04/04 

Date: Wed, 04 Aug 2004 10:25:49 -0400

Sergio Navega wrote:

> "Wolf Kirchmeir" <wwolfkir@sympatico.ca> escreveu na mensagem
> news:haPPc.23036$Vm1.458145@news20.bellglobal.com...
>
>>Sergio Navega wrote:
>>[...]
>>
>>>I agree, but there's an important point to be made. The domains where
>>>children usually refine their discrimination abilities are linked to
>>>conceptual categories (as opposed to perceptual categories).
>>>Successive conceptually refined categories (such as "edible stuff",
>>>and then "food", later "liquid food" and finally "milk") are categories
>>>that are refined mostly because of top/down processes. The perception of
>>>stimuli in these cases may be the same for an adult and an infant,
>>>but the former has developed these categories while the latter hasn't.
>>>Perceptual categories (or bottom-up categories) are those which seek
>>>for similarity and clustering based on the raw elements captured by
>>>the senses. After some time, our brain becomes unable to discriminate
>>>all the sort of things that an infant appears to be discriminating
>>>(the example of the japanese children is evidence of such a thing).
>>>
>>>Sergio Navega.
>>
>>
>>Again, granted that "categorisation" takes place. But what, exactly, is
>>it? I submit it's behaviour.
>
>
> And so your investigation of the subject stops. Cognitive scientists
> want to develop abstract (even mathematical) models of categorization.
[...]

a) investigation begins with "categorisation is behavior." Now we have
some idea of what to look for.

b) how does an abstract model of categorisation explain how we do it?
For example, I know only that LISP is a "categorising language," and so
presumably one possible model of categorisation. I played around with
LISP some years ago, and I can tell you that actual people don't
"categorise" that way. If they did, they wouldn't have to learn LISP -
they would speak it. :-) I suspect that any mathematical model of
categorisation would fail in the same way. That's not to say it wouldn't
be useful in design/construction of an intelligent machine. What matters
ultimately is not the structure and processes of the machinery, but the
system's overall function (ie, its behavior.) I agree that mathematical
models are necessary to designing intelligent systems.

IOW, categorisation as a "top-down" process makes sense for building a
model of that process: ie., it shouldd be possible to write a program
instantiating that model. But I don't see much connection with cognitive
psychology (or human behavior, for that meatter.)