From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!caen!garbo.ucc.umass.edu!dime!orourke Mon Dec 16 11:01:40 EST 1991
Article 2097 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!caen!garbo.ucc.umass.edu!dime!orourke
>From: orourke@unix1.cs.umass.edu (Joseph O'Rourke)
Newsgroups: comp.ai.philosophy
Subject: Re: Searle, again
Message-ID: <40592@dime.cs.umass.edu>
Date: 13 Dec 91 14:47:05 GMT
References: <1991Dec13.030010.15837@oracorp.com>
Sender: news@dime.cs.umass.edu
Reply-To: orourke@sophia.smith.edu (Joseph O'Rourke)
Organization: Smith College, Northampton, MA, US
Lines: 35

Daryl McCullough suggests a technique to convert a stream of Chinese
messages into a stream of chess moves:

>[...]
>      1. Code the Chinese messages in binary.
>      2. Use semantic data compression to remove the redundancy from the
>         resulting bit stream (...).
>These two steps give us a function F from Chinese to bit streams with the
>properties:
>    (a) It has an inverse, F^{-1}.
>    (b) Each bit of the output stream is approximately equally likely to
>        be a 1 as a 0 (...).
>Similarly, we can come up with a function G from sequences of chess moves
>into compressed bit streams. Then the function  H(x) == G^{-1}(F(x)) will
>be a function taking us from a stream x of meaningful messages in Chinese
>to a stream H(x) of meaningful chess moves.
>[...]

This is a fascinating idea, novel to me.  A few comments and questions:
	1. Could this be what Searle had in mind when he suggested
	   that the man in the CR could interpret the symbol manipulations
	   as moves in a chess game?  Is this so well known that it
	   requires no comment?  Or did he have something else in mind,
	   so much simpler that no support is necessary?
	2. Does this only work if the message streams are available
	   in their entirety?  Or could you take the Chinese stream
	   up to time t, work out F & G, and then continue to use
	   the same F & G for Chinese messages beyond time t?
	3. Does H resemble a normal interpretation function: does it
	   map a fixed pattern of Chinese symbols to a fixed pattern
	   of chess moves?  Or would a particlar symbol pattern in the 
	   Chinese stream sometimes map to one set of chess moves, and
	   sometimes to another, etc.? From Searle's description, he 
	   clearly had in mind an interpretation that fixes the meaning 
	   of symbols.


