From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!ccu.umanitoba.ca!zirdum Wed Feb 26 12:54:41 EST 1992
Article 4025 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!ccu.umanitoba.ca!zirdum
>From: zirdum@ccu.umanitoba.ca (Antun Zirdum)
Newsgroups: comp.ai.philosophy
Subject: Re: Definition of understanding
Message-ID: <1992Feb26.004118.4524@ccu.umanitoba.ca>
Date: 26 Feb 92 00:41:18 GMT
References: <1992Feb24.044654.12505@psych.toronto.edu> <1992Feb25.165326.16204@ccu.umanitoba.ca> <1992Feb25.190913.7908@psych.toronto.edu>
Organization: University of Manitoba, Winnipeg, Manitoba, Canada
Lines: 70

In article <1992Feb25.190913.7908@psych.toronto.edu> christo@psych.toronto.edu (Christopher Green) writes:
>In article <1992Feb25.165326.16204@ccu.umanitoba.ca> zirdum@ccu.umanitoba.ca (Antun Zirdum) writes:
>>I think that you missed my point. What is it that you get from me when I write to you
>>on the Net, is it syntax or semantics? Where exactly in your mind is this
>>syntax converted to semantics?
>
>Depends entirely on how you explicate "get from" Why do you assume that
>syntax is "converted" to semantics. That just begs the question. I
>think most semantic theorists would argue that the syntax merely flags
>semantic tokens already in your mind. It is not "converted".

I should not have said 'get from', I meant more along the lines of - you
read my sentences on the net, you have the semantics necessary for the
understanding of these sentences, therefore you understand what I am
talking about! (So far I believe we are in agreement)
Now, when I say the word 'ruka' I presume that this word has no semantic
meaning to you? Now let me add some semantic meaning - 'ruka' means
'hand' in yugoslavian. How did I perform that magic, how is it that
you now have semantics where there were none before?

Seems to me that semantics is just a short way of saying 'having knowledge
of things. The more knowledge you have of a certain thing the more
semantics you have!'
I do not think that semantics is any magic that cannot be duplicated
by a computer running a program.
My question seems to be "How did those 'semantic tokens' get into your mind
in the first place?" Is it through magic, or are you willing to admit
that semantics is gained through syntax. (at least in some cases, I do 
not mean to suggest that your knowledge of color is obtained through
syntax. But then again, a blind man has no knowledge of color, therefore
you should not expect a computer to contain such knowledge if some people
cannot have it!)
>>
>>I am claiming that without interpreting the symbols, the Chinese room has not a prayer
>>in hell chance of actually conversing intelligently. 
>
>Well golly gosh! We agree. Problem is, computers have no way of doing such
>an interpretation (take a look at Fodor's "Tom Swift and his procedural
>grandmother" in _Representations_).

Yes, we agree, but I have a problem with your comment that computers
cannot interprete symbols. If your statement were true there would be
no such thing as a compiler, or better a symbolic math program.
I do not mean to say that a compiler is intelligent, just as I would
not say that a cat is intelligent. (it does what it does very well,
but no more!)
Seems to me that the only difference between the current crop of AI
programs and Human intelligence is the semantic baggage that humans 
carry around (which is gained through a long process of experience!)

Let us take a simple example of computer semantics. Suppose that I
hook up a color detector to a computer (to keep it simple) then
when I point the color detector around it (the AI program) would 
know what color is in front of it! I point it at blue, the AI program
does not know yet what the input it is recieving from the detector
is! Now I type in, you are looking at BLUE. The program now knows
what blue 'means' (it simply means that the input it is recieving
from this register IS 'BLUE'!) To the computer that would be the
meaning of blue! It is no use arguing that it does not understand
the color the same as humans, how do you know that we are 
feeling the same thing when we look at blue?
Now suppose we connect the color detector to a blind mans brain,
when we point it at blue, we tell him that that is the color
blue! He would not have had any similar experience before, therefore
to him that WOULD BE THE MEANING OF BLUE! (agree on this?)

So, to sum up, I am not convinced that computer programs cannot contain
data that is semantics! Not merely syntax!

-- AZ -- zirdum@ccu.umanitoba.ca


