Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!rutgers!princeton!mind!harnad
From: harnad@mind.UUCP (Stevan Harnad)
Newsgroups: net.ai
Subject: Re: Searle, Turing, Symbols, Categories
Message-ID: <8@mind.UUCP>
Date: Wed, 22-Oct-86 00:29:21 EDT
Article-I.D.: mind.8
Posted: Wed Oct 22 00:29:21 1986
Date-Received: Wed, 22-Oct-86 22:21:31 EDT
References: <158@mind.UUCP> <150@cwrudg.UUCP> <160@mind.UUCP> <2495@utai.UUCP> <150@uwslh.UUCP>
Organization: Cognitive Science, Princeton University
Lines: 37
Summary: On Trying to Solve the Other-Minds Problem (or any problem) By Definition


lishka@uwslh.UUCP (Chris Lishka) asks:


>	How does one go about testing for something when one does not know
>	what that something is?  My basic problem with all this 
>	[discussion about the Total Turing Test] are the two
>	keywords 'mind' and 'intelligence'. I don't think that what S. Harnad
>	is talking about when referring to 'mind' and 'intelligence' are what
>	I believe is the 'mind' and 'intelligence', and I presume others are
>	having this problem...

You bet others are having this problem. It's called the "other minds"
problem: How can you know whether anyone/anything else but you has a mind?

>	Now, if one is to define 'mind' before testing for it, then
>	everyone will have a pretty good idea of what he was testing for.

What makes people think that the other-minds problem will be solved or
simplified by definitions? Do you need a definition to know whether
YOU have a mind or intelligence? Well then take the (undefined)
phenomenon that you know is true of you to be what you're trying to
ascertain about robots (and other people). What's at issue here is not the
"definition" of what that phenomenon is, but whether the Total Turing
Test is the appropriate criterion for inferring its presence in entities
other than yourself.

[I don't believe, by the way, that empirical science or even
mathematics proceeds "definition-first." First you test for the
presence and boundary conditions of a phenomenon (or, in mathematics,
you test whether a conjecture is true), then you construct and test
a causal explanation (or, in mathematics, you do a formal proof), THEN
you provide a definition, which usually depends heavily on the nature
of the explanatory theory (or proof) you've come up with.]

Stevan Harnad
princeton!mind!harnad