X-Message-Number: 7900
Date: Thu, 20 Mar 1997 08:28:19 -0800 (PST)
From: Joseph Strout <>
Subject: functional equivalence ==> same consciousness

OK, I haven't had time to dig up the Chalmers reference, but since there
seems to be much interest in it, I'll reconstruct it from memory.  This is
the demonstration that two functionally equivalent systems must have the
same conscious experience.

Suppose we have made an artificial visual cortex which is functionally
equivalent to your own.  Regardless of what it's made of, it will give the
same outputs for any set of inputs as your normal, biological cortex.  We
cleverly hook this up to the rest of your brain via a switchbox: with the
switch in one position, your bioligical visual cortex is engaged, and in
the other position, the artificial cortex is used instead.

With your biological cortex engaged, we show you a red apple and ask you
what color you are experiencing.  "Red," you say.  Then we throw the
switch, engaging the artficial cortex.

Now, suppose that this artificial cortex generates different qualia -- for
example, it causes an experience of blueness in place of redness, and vice
versa.  So you are now seeing a blue apple.

But the cortex is functionally equivalent, so it must generate the same
outputs to the rest of your brain (and eventually to the muscles of your
mouth and tongue).  So though you are experiencing blue, when we ask what
you see, you must say "red".  As we flip the switch back and forth, you
must always say that you see only red and have seen nothing but red.
Assuming the artificial cortex does not somehow make you into a liar, this
is absurd -- why would you be forced to say you see red when actually you
see blue?  Our supposition must be false; and this was that the artificial
cortex generates different qualia than the biological cortex.  Therefore,
it must be that functionally equivalent artificial cortex generates the
*same* qualia.  (We argued here using qualia different in kind, but you
could make an analogous argument for no qualia at all.)

The example was simplified by restricting it to just some cortex.  But
surely if having *some* artificial cortex generates the same conscious
experience, then having more of it will not suddenly make your qualia
disappear.  So any artificial brain -- provided it is functionally
equivalent -- will have the same conscious experience as the original.

Note, too, that we are NOT assuming what we want to prove here.
Functional equivalence is a well-defined and relatively weak property,
namely, that the input-output relation is the same.  It says nothing about
what goes on inside; could be ions moving through lipid membranes, or
electrons moving through transistors, or speedy little elves shuffling
around pieces of paper.

Some of these systems, of course, will be impossible to build.  Thomas
believes that no symbol-manipulation system could feel like us, so in
essence, he is asserting that such a machine could not be functionally
equivalent.  Bob goes farther, suggesting that perhaps no machine made out
of nonbiological materials could be functionally equivalent to our brain. 
Could be true (though I have seen no evidence for it).  But at least we
should be able to agree that if some machine IS functionally equivalent to
another, then their conscious experience is the same.

,------------------------------------------------------------------.
|    Joseph J. Strout           Department of Neuroscience, UCSD   |
|               http://www-acs.ucsd.edu/~jstrout/  |
`------------------------------------------------------------------'

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=7900