X-Message-Number: 4118
Date: Fri, 31 Mar 1995 21:39:35 -0800
From: John K Clark <>
Subject: SCI.CRYONICS Hofstadter

-----BEGIN PGP SIGNED MESSAGE-----

In 4110    (Thomas Donaldson) Wrote:

		>it puts Hofstadter down  several steps in my estimation, for 
		>his lack of  clear thinking
		
Hofstadter has a lack of clear thinking??!! Could you please explain 
what the pronoun "it" in your sentence  refers to. 
		
	       >but it does explain something about how Mr.Clark can consider
	       >that so many different things are symbols that don't look      
	       >like symbols to me at all.
	       
Although you still haven't given us your definition of symbol I
infer that you believe that something is not a symbol unless
somebody is consciously aware of it. The trouble is we are
trying to figure out how consciousness works, so such a restricted
definition of symbol is circular and utterly useless.
	       
	      >when I was working as a math professor in  Australia, I got
	      >very tired of philosophy just from hearing the philosophers 
	      >(ie. people in the Philosophy Dept) discuss things over the 
	      >table at tea time. 
	       
Hofstadter is not a philosopher, he's a math professor, just
like somebody else I know. Hofstadter hasn't engaged in a lot of
vacuous discussion, he has not written a book in over 10 years
(this month he finally came out with a new one), he's been too 
busy writing computer code. 
		
		>Just by introspection, if you think about what you really do
		>when you use symbols, it's clear that you are not using 
		>symbols to work out what you will say or do with symbols. 
		
I like what  Hans Moravec (  ) has said about introspection:
   
    "Introspection is unreliable.  You only have memories about you
     internal dialog.  Why should those be reliable? What
     evolutionary advantage would there be?  What was important to
     your ancestors' survival was how they behaved.  Memories of
     external events, like where your offspring, food or the lion
     was, or who was where on the pecking order, or how to sharpen a
     spear, made a difference, so those were evolutionarily selected
     to be pretty accurate.  Traces of what went through your head
     ten minutes ago could be garbled arbitrarily, as long as you
     could make up an ass-saving story when the alpha male demanded
     to know why you were smirking.  So you're good at making up
     stories about what goes on in your head, but lousy at reporting
     it accurately."
     "You are much better at remembering external than internal
      events, so objective science is a better way to find out what's
      going on inside you than your own subjective feelings."
       

		>Even the simple problem of searching for just the right word
		>can't be done with symbols alone. Witt's [Wittgenstein]        
		>earlier  writings  were all ensnarled in that issue, the 
		>belief that  symbols and language were needed for thought
		>and truth, but he  worked  himself out of it, in the end.
		   
Hofstadter has repeatedly said that Wittgenstein proved that a
rigid, black and white relationship between language and reality
does not exist, so AI programs that use atomic symbolic
primitives that do not change according to context are doomed to
failure. Ludwig Wittgenstein is the reason Hofstadter made
symbols in his experimental programs highly context  dependent,
and that's the reason he criticized Newell and Simon and praised
Rumelhart and McClelland.


				    John K Clark         

-----BEGIN PGP SIGNATURE-----
Version: 2.6.i

iQCzAgUBL3zeyH03wfSpid95AQHsIATw6mQ8+PG2e4Vg7ktsQ7Hk/izBWOWK41Ut
09cHFdOZDC7P4b6+odjzJ1PiG6dXesgY9agPw7eVJ8LxV97kWg0eYEB9w3aNoH0p
r+Y+l4bGfOxegChMnWmrdh/MYxmJfdoRqMNMKDc4tOInBO467uAeBkbEum8C+5ns
L0G3+PkGGuuRd8zlaX15dB6GdDx6F9o3UHgjURgkYWYO52nIF00=
=F7j6
-----END PGP SIGNATURE-----


Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=4118