Path: utzoo!utgpu!utstat!jarvis.csri.toronto.edu!mailrus!eecae!tank!staff_bob@gsbacd.uchicago.edu
From: staff_bob@gsbacd.uchicago.edu
Newsgroups: comp.ai
Subject: Re: Question on Chinese Room Argument
Message-ID: <2233@tank.uchicago.edu>
Date: 10 Mar 89 18:36:27 GMT
Sender: news@tank.uchicago.edu
Organization: University of Chicago Graduate School of Business
Lines: 97



In article <18073@iuvax.cs.indiana.edu>, dave@cogsci.indiana.edu (David Chalmers) writes...
 
>Answer me these questions.
>  (1)  Do you believe neurons (taken alone) have semantics.
>              [I take it the answer has to be "No."]
>  (2)  Do you believe the brain as a whole has semantics.
>              [I take it the answer is "Yes."]
> 
>Given this, you must accept that semantics can arise out of non-semantic
>objects.  Most of us are a little baffled as to how.  It seems that the 
>only half-way reasonable tack we can take to answer this question is to
>say that what is important for semantics (and the subjective in general)
>is not so much those objects as the complex patterns that they form.
>
> [much deleted] 
> 
>  Dave Chalmers
>  Center for Research on Concepts and Cognition
>  Indiana University

The term 'semantics' comes out of linguistics, and is not a synonym
for the more general term, 'meaning'. One cannot discuss semantics outside
of the context of a language. For a given, well-formed expression in any
language, the semantics of that expression is what it is intended to signify.
Thus, generally, we can say that the semantics of a sentence is its meaning.

I don't know exactly what Mr.Chalmers is trying to say in this passage.
Are we to take a neuron as a well formed sentence in some language that
is nonetheless devoid of meaning? In the context of formal language theory,
neurons do have well defined semantics. We think we know how single neurons
work: for a given set of input values, there is a well defined output.
If I remember correctly, it is possible to characterize these actions with
regular expressions. Therefore, I would have to say that the answer to 
[1] is "Yes".

Mr.Chalmers argument follows from his premise. He assumes that semantics means 
"something more complex" than simple input/output operations, and his
conclusion emerges directly from this (questionable) assumption.

Anyone who has studied compiler writing in general or denotational semantics
in particular is familiar with semantics in the low level sense, but Mr.
Chalmers discludes this usage, either by choice or through ignorance of
the use and meaning of the word 'semantics'. He then goes
on to prove that since he has assumed no low level semantics, but has assumed
high level semantics, that semantics must emerge somewhere in the middle.

To excuse the misuse of the word "semantics", let us substitute some other,
more general term, such as meaning (or perhaps the U-word, understanding). 
Then I would have to say that the answer to [2] is no. It is certainly possible
to claim that meaning does not exist in the brain, it exists in the mind. 
I'm sure that many readers will object to the distinction, and I do not
care to defend it, but my point is this: the Mind/Body problem has never
been solved, and it may well be insoluble. We observe that something called
Mind exists ("I think, therefore I am" is a proof for the existence of Mind),
and we also observe that Minds coexist with bodies. As Mr.Chalmers asserts,
most of us are baffled as to how. The general assumption these days
is that Mind is a by-product of a body, but a recent posting has made the
valid point that we cannot show that a body is not a by-product of Mind.
In any event, I think that one can show many more than half-way reasonable
tacks to take in approaching this problem than Mr.Chalmers has suggested.
I need only list all of the great philosophers from Plato and Aristotle
through DesCartes, Locke, Hume, Kant, Hegel, Schopenhauer, Heidegger and
Sartre to begin to enumerate them.
                           
A lot of people have grappled with this question, none to the general
satisfaction of the rest of humanity. The recent debate in this news
group in re Searle's Chinese Room thought experiment hinges on the
Mind/Body question. Is not the premise of Searles argument just that
understanding only occurs in Mind, and Mind exists neither in the system
(Searle+rules) nor in a computer? IMHO, thia entire debate has revolved around
these two assertions, neither of which can be proven or disproven.

It interests me that so many people of scientific bent show an absolute
distaste for philosophy, yet somehow feel themselves qualified to discourse
at length on the greatest philosophical questions of the ages.

I' reminded of a anecdote from the book "The Dreams of Reason" by Heinz
Pagel, who in his capacity (I believe) as executive director of the
New York Academy of Science once arranged a talk to be given by the
Dali Lama. In the subsequent question and answer session, someone tried
to ask the Dali Lama where intelligent machines fit into his philosophy/
religion/system. The Dali Lama merely responded, "When you have such a 
machine, here, in front of us, then I will answer your question."

What can possibly be gained from this debate over Searle's thought experiment?
Assuming that we could come to some sort of universal agreement about
this (and I really don't think that is possible) will it make a single
iota of difference to the work we're doing? Is this not really a theological
debate, more than anything else? That is to say, aren't the arguments
we have seen to this point really defenses of various faiths in the
possiblilty of machine 'awareness'? Can't we let that debate wait until
we're a little bit closer to something called machine intelligence?


					R.Kohout