From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!usc!rpi!usenet.coe.montana.edu!news.u.washington.edu!milton!jhardy Mon May 25 14:05:18 EDT 1992
Article 5642 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!usc!rpi!usenet.coe.montana.edu!news.u.washington.edu!milton!jhardy
>From: jhardy@milton.u.washington.edu (J. Hardy)
Subject: Re: Turing test and language
Message-ID: <jhardy.705820007@milton>
Keywords: turing test language acquisition new yorker
Sender: news@u.washington.edu (USENET News System)
Organization: University of Washington
References: <1992May12.205205.14441@bony1.bony.com>
Date: Thu, 14 May 1992 05:06:47 GMT
Lines: 54

richieb@bony1.bony.com (Richard Bielak) writes:


>The article discusses at length (after all this is _the_ "New Yorker")
>this case, and others of so called "wild" children, from the
>perspective of linguists.

>The interesting thing is that the girl never really acquired language.
>She learned words, but syntax was beyond her - for example, she never
>learned how to make negatives.

>I thought that if she had been used in a Turing test - playing the
>role of computer - she would have failed miserably. The tester would
>have declared her _not_ intelligent.

>Yet, according to her caretakers, she _was_ intelligent.  She was
>eerily good at non-verbal communications and was excellent in spacial
>reasoning.

>The implication is that the Turing test does not prove or disprove the
>intelligence of the entity tested.

Whoa, hold on a sec there.  All your arguments could show is that the 
Turing test doesn't disprove intelligence.  You haven't said a thing
to suggest that an entity could pass the Turing test and yet be 
non-intelligent.  So, the most you could reasonably claim is that
ability to pass a Turing test isn't a *necessary* condition of 
intelligence.  Frankly I don't know of anyone who has ever claimed that
it was a necessary condition (tho this isn't my area of specialization).
What is generally claimed is that passing a Turing test is a *sufficient*
condition of intelligence.  You've said nothing to suggest that this latter
type of claim is mistaken.

I agree that it would be obviously wrong to maintain that failing a Turing
test is proof of non-intelligence.  Anyone who has spent sometime with an
aphasic can attest to that, and aphasics are far more common than wild
children


Good luck,
and don't drink anything you can see through.


>...richie

>-- 
>* Richie Bielak   (212)-815-3072   | "Your brain is a liquid-cooled parallel  *
>* Internet:       richieb@bony.com | super-computer". He pointed to his nose, *
>* Bang {uupsi,uunet}!bony1!richieb | "This is the fan."                       *
>*    - Strictly my opinions -      |                     - David Chudnovsky - 

jim hardy
jhardy@u.washington.edu



