Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!snorkelwacker.mit.edu!bloom-beacon!eru!hagbard!sunic!mcsun!ukc!tcdcs!swift.cs.tcd.ie!ccvax.ucd.ie!tskelly
From: tskelly@ccvax.ucd.ie
Newsgroups: comp.ai.philosophy
Subject: Chinese room and stuff
Message-ID: <12160.274a7a65@ccvax.ucd.ie>
Date: 21 Nov 90 13:00:20 GMT
Organization: University College Dublin
Lines: 67



Good day to y'all,

First order of business is to say that I am a relatively new newser and
have probably missed a lot of the discussion on the subject of Searle's
Chinese room - so, apologies, in advance, to anyone whose toes I step on!!

Now, down to the main business!
From the few postings that I have read on this topic I have to say I'm
very dissapointed with the fact that nearly everybody seems to have
missed the point.
To illustrate, imagine an ENGLISH room! ( I know there are reasons why
Searle chose to use Chinese but I trust the readers of this will understand
them and also understand why I choose to use English!!).
So, in the English room we have an *entity* which takes pieces of paper
with symbols on it and etc. etc. The important point is that the entity
inside the room examines the symbols on the input paper only to apply rules
to them to get a string of symbols for passing back out!
Now, in comes a piece of paper with the symbols
		"WHAT IS 2 + 2?"
After matching up the symbols and the rule book the entity should arrive
at the symbol "4" which it passes back out!
The entity has arrived at this symbol in answer to the input purely by
symbol manipultaion - in other words the rule book will say something to
the effect of " when the symbols '2' and '2' are seperated by a '+' etc.
then the resulting symbol is '4' ". 
However, does an intelligent entity do this?
To be quite candid, I think an intelligent entity does do this sort of
symbol manipulation - ie. very often when asked "what is 2 + 2?" we arrive
at the anser "4" purely by the fact that the symbol "4" has been matched to
that question! BUT! an intelligent entity does not always do it by symbol
manipulation. An intelligent entity can grasp all aspects of the MEANING of
the symbols "2" and "+" - ie. we can visualise what it is to be "2" (2 apples
or 2 computers etc. etc.) and we know what it is to "+" numbers together
(2 apples on one table and 2 apples on another table moved together onto the
same table and recounted give us 4 apples and we know what 4 apples is!)
The entity in the room cannot abstract from the symbol "2" and apply it 
as a concept. All it can do is manipulate it symbolically using rules and in
relation to the symbols surrounding it!
Now, I don't want to labour the point so I'll say no more on that one and
trust that you understand what I mean (which is more than the ENGLISH room
could do!)

Another interesting point that has occured to me is this :
Suppose Searle's Chinese room was created on 1/1/1800 and was sealed
except for the input slot and the output slot. At the time of it's 
creation the rule book was sufficient to manipulate all symbols that existed.
So, everybody was in awe of this "intelligent" room! Now, I come along
and present it with a question concerning "COMPUTERS" or "SPACE SHUTTLES".
Now, to anyone who claims that the room could even begin to manipulate the
symbols "COMPUTER" or "SPACE SHUTTLE" I can only say 'piff'!! And furthermore
it could never manipulate these symbols because no-body can get in to change
the rule book!!!


Anyhow, let's see you pull those arguments apart ( but please address the 
thrust of the arguments and not some silly little points that have very little
to do with what I have said)

Incidently, I am pro AI and I feel that while Searle's Chinese room addresses
a certain aspect of comtemporary views in AI it does not contradict the 
possibility of AI! I would like to see some discussion on other aspects in the
AI line and an end to this flogging of a quite wounded horse!

Stephen Kelly
TSKELLY@CCVAX.UCD.IE