From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!aunro!ukma!darwin.sura.net!europa.asd.contel.com!uunet!psinntp!scylla!daryl Thu Feb 20 15:21:33 EST 1992
Article 3811 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!aunro!ukma!darwin.sura.net!europa.asd.contel.com!uunet!psinntp!scylla!daryl
>From: daryl@oracorp.com
Newsgroups: comp.ai.philosophy
Subject: Re: Strong AI and panpsychism
Message-ID: <1992Feb17.221311.17990@oracorp.com>
Date: 17 Feb 92 22:13:11 GMT
Organization: ORA Corporation
Lines: 87

David Chalmers writes (in response to Michael Gemar):

>> My concerns are *precisely* with how we use certain concepts.  You
>> want to rule out time-dependent definitions of states, and Putnam
>> would like to include them (at least, as far as I understand his
>> argument).  My problem is I see no way of adjudicating between you
>> two.

> Because we're dealing, essentially, with the pre-theoretical concept
> of "causal organization", and trying to find a good definition of it
> that accords reasonably well with the pre-theoretical notion (as is
> the case with all definitions).  If Putnam's proffered definition has
> the consequence that simple clocks have the same causal organization
> as the brain, then that provides entirely adequate grounds for ruling
> it out as a definition (compare: if someone offered a formal
> definition of "heat" that had the consequence that Antarctica and the
> Sahara were equally hot, then we'd rule it out of court).

You have missed the possibility that "causal organization" might not
be an objective property. For an example of a property that has
flipped in scientists minds from objective, to subjective, and then
back to objective, consider entropy. The earliest definitions of
entropy were in terms of heat, which seemed to be an objectively real
quantity. Later, a definition of entropy was given (by Boltzman?) that
involved, not heat but counting: The entropy S of a system was defined
as the logarithm of W, the number of ways that the system could be
arranged microscopically which produced the same macroscopic state
(determined by the total energy, volume and number of particles). It
turned out that this number was undefined! The reason was that there
was one microscopic state for each choice of positions and momenta for
each particle. This leads to an uncountably infinite number of
microscopic states (and therefore to an infinite entropy).

A way around this problem was to divide the "phase space" into small
chunks, of size delta-x delta-p so that two microscopic states whose
positions differed by less than delta-x, and whose momenta differed by
less than delta-p would be considered to be "the same states" for the
purpose of computing entropy. This led to sensible results for the
entropy, but the results depended on the arbitrary parameters delta-x
and delta-p. Therefore, entropy, which was supposed to be an objective
quantity, depended on the human choice of chunk-size. It seemed, at least
partly, to be a *subjective* quantity; depending on our subjective notion
of when two microscopic states were "essentially the same".

Later, with the rise of quantum mechanics, it turned out that particle
states are quantized; there were only a discrete number of possible
states consistent with a given energy, so the choice of chunk size did
not need to be made.

Anyway, the point is that just because you would *like* the notion of
"causal organization" to be objective, it may be that it simply isn't.

> Perhaps the problem is that you haven't read Putnam's proof?  We do
> indeed possess all the relevant facts about the lump.  All the proof
> requires is that the lump be in different states at different times,
> a requirement that would be satisfied by a structureless ball rolling
> down a hill.  Nothing in the proof involves finding out things about
> lumps that we didn't know before (e.g. that the laws of physics
> guarantee that they are seething with internal structure).

>From another posting, I take it that you don't like Putnam's notion of
implementing a FSA because it allows the implementation of every FSA
by any clock. Certainly there is a notion of implementation that will
prevent a clock from being considered to implement a human brain; for
example, I suggested the notion of "implementing the same process", in
the sense of process algebras. You didn't comment on this suggestion,
but it is what I currently favor. However, it has the consequence that
whether a physical system implements a particular process depends on a
subjective choice of what counts as inputs and outputs (and how to
interpret them).

Your complaint about clocks, that they don't support counterfactuals,
is I think, easily corrected: for example, consider a machine M with a
state determined by a pair: the time, and the list of all inputs ever
made (with the times they were made). If "implementation" simply means
the existence of a mapping from the physical system to the FSA, then
it seems that such a system M would simultaneously implement *every*
FSA. Counterfactuals would be covered, too.

The more people try to avoid panpsychism, the more it seems clear to
me that panpsychism is true. I don't consider that so bad; it would
shift the focus from "Is this system conscious?" to "Under what
interpretations is this system conscious?"

Daryl McCullough
ORA Corp.
Ithaca, NY


