From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!utgpu!jupiter!morgan.ucs.mun.ca!nstn.ns.ca!aunro!ukma!wupost!cs.utexas.edu!uunet!mcsun!uknet!edcastle!aiai!jeff Sun Dec  1 13:05:32 EST 1991
Article 1639 of comp.ai.philosophy:
Xref: newshub.ccs.yorku.ca sci.philosophy.tech:1153 comp.ai.philosophy:1639
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!utgpu!jupiter!morgan.ucs.mun.ca!nstn.ns.ca!aunro!ukma!wupost!cs.utexas.edu!uunet!mcsun!uknet!edcastle!aiai!jeff
>From: jeff@aiai.ed.ac.uk (Jeff Dalton)
Newsgroups: sci.philosophy.tech,comp.ai.philosophy
Subject: Re: Searle (was Re: Daniel Dennett (was Re: Comme
Message-ID: <5707@skye.ed.ac.uk>
Date: 26 Nov 91 19:45:54 GMT
References: <1991Nov14.223348.4076@milton.u.washington.edu> <MATT.91Nov24000158@physics.berkeley.edu> <1991Nov24.195230.5843@husc3.harvard.edu> <JMC.91Nov24194704@SAIL.Stanford.EDU> <5689@skye.ed.ac.uk> <YAMAUCHI.91Nov25203631@indigo.cs.rochester.edu>
Reply-To: jeff@aiai.UUCP (Jeff Dalton)
Organization: AIAI, University of Edinburgh, Scotland
Lines: 70

In article <YAMAUCHI.91Nov25203631@indigo.cs.rochester.edu> yamauchi@cs.rochester.edu (Brian Yamauchi) writes:
>
>(Note that rec.arts.books has been removed from followups.)
>
>In article <5689@skye.ed.ac.uk> jeff@aiai.ed.ac.uk (Jeff Dalton) writes:
>>In article <JMC.91Nov24194704@SAIL.Stanford.EDU> jmc@cs.Stanford.EDU writes:
>>>Let's start out simple.  The reading of a digital thermometer
>>>"stands for" the temperature.
>
>>Because we so interpret it.  The number 32 also stands for a
>>temperature, if we want it to.  It might be argued that there's
>>a direct causal connection between the thermometer and the
>>temperature.  But so what?  If there's a causal connection
>>between the word "temperature" and physical temperature (so
>>to speak) that makes the word stand for the phenomenon, it's
>>surely not the same causal connection as that between the
>>reading of a thermometer and the same phenomenon.
>
>Could you (or anyone else on the anti-AI side) explain what you mean
>by a "causal connection" (or Searle's even-more-mystical term "causal
>powers")?  The "causal connection" between temperature and a
>thermometer is clear, change the temperature and the reading on the
>thermometer also changes.  What does it mean to say that a word "makes
>the word stand for the phenomenon"?

(BTW, I didn't write that a word "makes the word ...", I wrote that 
a causal connection "makes the word ...".  Does that help?)

In any case, nothing very mysterious was involved in my message.  You
employ "causal connections" yourself when you say a robot could detect
temperature directly through thermometers or indirectly through other
forms of sensing.  What would then make some symbol, inside the
computer, stand for the temperature?  Presumably by having it be
connected (however indirectly) to the readings of the relevant
sensory devices.  Since computers work mechanically, so to speak,
the connections would be "causal".

However, having something stand for _the_ temperature isn't very
interesting.  Can a computer have something like our word "temperature" 
that stands for temperature in general?  If so, then presumably it
would again be related to the physical phenomenon.  Whatever this
connection is (we needn't stress "causal"), it isn't the same as
the direct physical connection between a thermometer and the
temperarture.  Indeed, it needs to be shown, and not just assumed,
that the "stands for" that applies to words such as "temperature"
is in any way the same as the "stands for" of direct physical
connections.

Let me add that it's by no means settled, philosophically, how
words manage to refer to the right things when used by humans.
One of the answers advanced against arguments such as Putnam's
"cats and cherries" is that there is a causal connection.  (See
various books by Devitt(?), e.g.)  It's at least plausible.
That "trees" refers to trees might have at least something to
do with the fact that light from trees is received by the
retina, etc, with the consequence that we know something about
what trees look like.

>                     So the robot is "interpreting" the reading of
>its thermometers -- unless you assume, a priori, that machines are
>inherently incapable of "interpreting" anything -- in which case, this
>boils down to an argument over theology...

There's interpreting and interpreting.  You have to show that it's
interpreting in the right sense and not just a pun or equivocation.
Perhaps my pencil is "interpreting" the pressure of my hand holding it
against the paper.  So what?  That I can use the word "interpreting"
for it proves nothing.

-- jd


