🚀 go-pugleaf

pluto@beowulf.ucsd.edu (Mark E. P. Plutowski) writes:
------
We believe Searle-in-the-box when he claims to not-understand.  

Why?  Don't we have an objective definition of "understanding?"
Why don't we have an objective definition of understanding?
Why do we have to ask Searle whether or not he understands, in order
to ourselves decide whether or not he understands?
------
I reply:

Actually we don't have to believe him.  We can just write to him in
Chinese and see what happens.  In the case where he uses the books and
stuff he will not be able to respond without them.  In the case where
he memorizes the rules, he will not be able to translate the Chinese
into, say, English.

-------------
Mark Plutowski continues:

I am not biased against Searle's basic point: (IMHO)
that the implementation matters.  I just don't see the logic
behind the counter to the systems reply, and would like to 
hear it expressed in objective terms.  

Prove to me the Chinese Room does not understand, without
resorting to asking Searle-inside whether (or not) he does.
------------
I reply:

First of all, the argument does not seem to be, strictly, that
``implementation matters''.  Rather, the argument is that human minds
are doing something in addition to manipulating symbols.

Now as far as the Chinese Room not understanding, if we take the case
where the rules are in books etc., the question is, ``What physical
entity understands, if not Searle?''  I think we would all agree that
books and ink are not things that can understand.  The only actor in
the system is Searle.  If he doesn't understand, what does?  If Searle
memorizes the rules, then he is the ONLY physical entity left that
could be understanding, so I would think you have to ask him, or test
him.

Even if the system prints (in Chinese), ``I understand Chinese'', it
really doesn't matter.  That is because it is easy to produce a set of
rules that will cause a system to print ``I understand Chinese'',
whether it does or not.

If we assume, even so, that there is a ``system'' that understands,
then Penrose, in THE EMPEROR'S NEW MIND, has pointed out that this is
equivalent to postulating a ``mind-stuff'' apart from the physical
entities that exist.  This is known as dualism and most AI folk seem
to want to avoid such a viewpoint.  While the argument as a whole does
not PROVE that the Chinese Room does not understand Chinese, it does
show that no physical entity in the Chinese Room understands Chinese.

-Fred Gilham   gilham@csl.sri.com