From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usenet.coe.montana.edu!news.u.washington.edu!milton.u.washington.edu!forbis Wed Feb 26 12:53:54 EST 1992
Article 3953 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usenet.coe.montana.edu!news.u.washington.edu!milton.u.washington.edu!forbis
>From: forbis@milton.u.washington.edu (Gary Forbis)
Subject: Re: Definition of understanding
Message-ID: <1992Feb24.083303.20762@u.washington.edu>
Sender: news@u.washington.edu (USENET News System)
Organization: University of Washington, Seattle
References: <1992Feb22.234830.17713@psych.toronto.edu> <1992Feb23.071810.16573@ccu.umanitoba.ca> <1992Feb24.044654.12505@psych.toronto.edu>
Date: Mon, 24 Feb 1992 08:33:03 GMT

In article <1992Feb24.044654.12505@psych.toronto.edu> christo@psych.toronto.edu (Christopher Green) writes:
>In article <1992Feb23.071810.16573@ccu.umanitoba.ca> zirdum@ccu.umanitoba.ca (Antun Zirdum) writes:
>>>The systems reply says: it's not the man in the room that understands
>>>but the system as a whole: the man, the room, the slips of paper, the
>>>rule books, etc.
>>>
>>>Searle responds: fine. Put the whole system in the man. Have him memorize
>>>the symbols, the rules, etc., and get rid of the room. Have him walk about
>>>like a sort of Chinese deaf-mute who can only communicate via written     
>>>messages. Now you've got the system in the man and can discover whether the
>>>system understands any bettter than did the man-as-part-of-the-system.
>>>You ask him -- the system -- whether it understands Chinese.
>>>He still replies "in his native language" that he doesn't understand a
>>>word of Chinese.
>>>
>>
>>It seems to me that the man cannot respond in any way but to say "Yes I understand
>>Chinese!" He has all the requirements, He can do anything that a native Chinese
>>speaker can. There is no need for him to assume that native Chinese speakers do
>>it in a different way. In short, I have serious doubts that someone can memorize
>>rules for interacting with Chinese speakers and still not understand.
>
>I think you've lost track of the fact that all the rules the man is using
>are purely syntactic. All the Chinese symbols he uses are still purely
>formal and uninterpreted. Thus, what he is missing is their *meaning*.
>And without their meaning, surely he cannot be said to understand them.

It seems to me that one has to first accept the premises that semantics is
not reducable to syntax and machines only manipulating symbols acording
to syntatic rules.  If one truely accept that the machine produced behavior
indistinguishable from human who understand then it is not obvious that
both of these presises can still be consistantly held.

If one accepts the theoretical possibility of the giant look up table then
it is pretty clear that the behavioral content of semantics can indeed be
reduced to syntax.  What is the extra component of semantics which has no
effect on linguistic production?

-------

On a slightly related topic...

In the split brain experiments there were some who could verbally identify
objects when held in one hand and could write what the object was when held
in the other hand.  Would you say that the person didn't understand what the
objects were when verbal communication could not be given?

Likewise, if an individual said in English "I do not understand Chinese" when
asked orally in English if she understood Chinese but wrote "Yes, I understand
Chinese" when viewing the question in Chinese, would you say the person does
not understand Chinese?


 
>Christopher D. Green                christo@psych.toronto.edu
>Psychology Department               cgreen@lake.scar.utoronto.ca
>University of Toronto
>---------------------

--gary forbis@u.washington.edu


