From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!think.com!yale.edu!jvnc.net!darwin.sura.net!convex!uunet!psinntp!scylla!daryl Mon Mar  9 18:34:40 EST 1992
Article 4207 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!think.com!yale.edu!jvnc.net!darwin.sura.net!convex!uunet!psinntp!scylla!daryl
>From: daryl@oracorp.com
Subject: Re: Reference (was re: Multiple Personality Disorder and Strong AI)
Message-ID: <1992Mar2.223923.1711@oracorp.com>
Organization: ORA Corporation
Date: Mon, 2 Mar 1992 22:39:23 GMT

cam@castle.ed.ac.uk (Chris Malcolm) writes:

> On the other hand, I think it remains true that what a program does
> is to transform some input data into some output data, and that this
> transformation can only be purely syntactic. This seems to me to pull
> the rug out from under the "English reply". Anyone care to comment?

Whether you call computer processing "syntactic" or not is irrelevant,
of course. The main claim was that, because computers are purely
syntactic, they are not capable of the kind of semantic understanding
that humans possess, but I don't see why any of the supposed
limitations on a computer's ability with semantics do not also apply
to human beings.

Babies are born with no knowledge of English (or at least with no
knowledge of the meanings of particular words, such as "hamburger").
Yet, after two or three years, babies raised in an English-speaking
environment become fairly fluent at English, and we say that they
understand what "hamburger", etc. means. Now, what has happened? How
did the semantics of the word "hamburger" get communicated to the
baby? All that was available to the baby's brain was a sequence of
signals corresponding to (a) people talking about "hamburgers", (b)
the correlation of talk about "hamburgers" with the smell, sight, and
taste of hamburgers, (c) feedback from the baby's incorrect attempts
to use the word "hamburger" correctly. (Note, I am making no
assumptions about what is going on *inside* the baby's brain, I am
only assuming that the only way the baby learns about the world is
through signals to the brain.)

Now, the question arises: is the information that the baby received
sufficient to determine the meaning of the word "hamburger"? Is there
only one meaning of the word "hamburger" that is consistent with that
information?

If the answer to the latter question is "yes", then it would appear
that signal processing is sufficient to determine the semantics of
words, and so there is no a priori reason to believe that a computer
is incapable of it. If the answer is "no" (which I think would be the
position of Putnam et. al.), then we still have no reason to think
that a human can do something that a computer cannot, since we have no
reason to think *humans* can determine the semantics of words.

Daryl McCullough
ORA Corp.
Ithaca, NY


