X-Message-Number: 5459
From:  (Thomas Donaldson)
Subject: once more on Turing tests
Date: Thu, 21 Dec 1995 11:28:27 -0800 (PST)

Hi again!

OK, so you decide to use Turing tests "in the broad sense". Your notion of
Turing test then ceases to have the logical sharpness of the real Turing test.
What it is supposed to mean becomes open to LOTS of interpretation.

As for Hawking, he CAN manipulate his wheelchair and understand how to get
to a window without running into anything. And once there, he can report
the color of something both you and he can see. Just playing with words means
just that: an inability to not only go to the window, but to report some new
piece of information which a human being could. (And finally, of course, we
know that Hawking is a human being because we see his picture and would 
recognize him). I am using Turing tests here not in the broad sense but in
the precise one.

You are quite right that those scientists who try to explain consciousness are
dealing with a subject for which only one person in the universe of each of us
can attest exists. And that has certainly given them problems. Since conscious-
ness also has other senses, however, and those senses can certainly be tested,
they have decided to continue their studies. In one way, such studies suffer
from the same problem as all other science: we can never KNOW ABSOLUTELY that
they are correct. 

Before Turing and before behaviorism, psychologists did studies with themselves
as subjects, trying to analyze just how their perceptions worked, by doing 
tests on themselves. Tulving was one such, but not the only one. How would we
know that they were correct? Because we would see that the same things would
happen when we did the same tests on ourselves. You are right that we can 
never prove the existence of consciousness in anyone else; but if we simply
assume it when conditions seem appropriate, then there are many other 
things we CAN prove on that assumption.

			Best and long long life,

				Thomas


Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=5459