From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!mips!spool.mu.edu!uunet!mcsun!uknet!edcastle!aiai!jeff Tue Mar 24 09:56:59 EST 1992
Article 4569 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!mips!spool.mu.edu!uunet!mcsun!uknet!edcastle!aiai!jeff
>From: jeff@aiai.ed.ac.uk (Jeff Dalton)
Newsgroups: comp.ai.philosophy
Subject: Re: Chinese room miscellanea
Message-ID: <6434@skye.ed.ac.uk>
Date: 18 Mar 92 18:54:12 GMT
Article-I.D.: skye.6434
References: <1992Mar17.022115.11185@bronze.ucs.indiana.edu> <6417@skye.ed.ac.uk> <1992Mar17.235343.26537@mp.cs.niu.edu>
Sender: news@aiai.ed.ac.uk
Organization: AIAI, University of Edinburgh, Scotland
Lines: 34

In article <1992Mar17.235343.26537@mp.cs.niu.edu> rickert@mp.cs.niu.edu (Neil Rickert) writes:
>In article <6417@skye.ed.ac.uk> jeff@aiai.ed.ac.uk (Jeff Dalton) writes:
>
>>right at the start.  What I object to is when anti-CR folk go beyond
>>"maybe the system understands" to "the system understands".  And I
>
>  Are you just engaging in unnecessary nitpicking here?  Do you really
>expect every sentence to begin with "Under the assumptions of the CR
>room ..."?  You are reading far too much into the absence of a "maybe".
>Many anti-CR people think the whole CR scenario is preposterous,
>and it is only when they temporarily suspend this disbelief that the
>talk about the system.  Why should they add an additional superfluous
>"maybe"?

There isn't a good argument that the system _does_ understand
(arguments that rely on the Turing Test are among the bad ones),
only that Searle has failed to show it doesn't.

Moreover, to refute Searle's argument, it's sufficient to show
he has failed to prove his case; it isn't necessary to show that
his conclusion (the system doesn't understand) is false.

As I said, I regard that -- the possibility of the system
(and hence computers) understanding -- as an open question.

I think people on the AI side do their cause a disservice by
linking it to something they're not in a position to show,
namely that the system understands.

And in general it is not (I'd say) a quibble to distinguish
between showing that an argument is invalid or unsound on the
one hand and proving that its conclusion is false on the other.

-- jd


