[time 316] Re: Other Minds


Stephen P. King (stephenk1@home.com)
Sat, 15 May 1999 19:29:35 -0400


Dear Jerry,

        I hope my mind is in gear... I just got home from my mind numbing
job... :)

Jerry Malsam wrote:
>
> > > Jerry Malsam wrote:
> > >
> > > What's the difference between "consciousness" and "thought"?
> >
> > Stephen P. King wrote:
> >
> > The difference is reflexion and prediction; thought is the action of
> > modeling expectations given some data. Consciousness is merely the
> > mapping of external data to internal data (in memory). (I think... ;) )
>
> Sure you don't have those backwards? At any rate I don't see the
> distinction.

        I am missing what you see (I'm dyslexic!)... :( The difference between
thought as predictive and consciousness as reflection is shown in the
second order aspect of recursion involved in thought. Thought requires
consciousness but consciousness does not necessitate thought... Thought
is consciousness of consciousness!
 
> >simple questions. I guess the simple questions are the hard ones. How do
> >you know that a "babbling" Chinese man is any more conscious than, oh, say
> >a dog? A dog knows enough to get out of the way of a car (usually), even
> >though it may never have been hit by one. How?
> >
> > We infer that the babbling man is "intensional" in his actions, e.g.
> > that the movement of his mouth are the result of computing an action
> > that is expected to have a reaction. (This computation need not be
> > considered as an object of consiousness!) This inference is grounded on
> > expectations, given past perseptions. The dog (hopefully) moves out of
> > the way of the car because of a reflex, constructed by the blind forces
> > of mutation and selection. (But, can we really say that our actions are
> > not just the same thing with a few more delay loops thrown in!)
 
> Why not infer that the dog is intensional in his actions? The car is a
> recent invention, presumably far to recent to have evolutionarily influenced
> the instinctive response of the dog species. I can understand a reflexive
> response to a lion or bear, but a speeding car? Were there really *that
> many* cattle stampedes throughout history that dogs developed a specific
> instinctual response to them? Ludicrous!

        The car is merely, to the dog, a large moving object, possibly
dangerous. Ok, it is erroneous to assume what is the subjective view of
any observer, rock to man, but we can quantify the amount of information
that can be communicated between arbitrary observers (relative to their
possible configurations?).
 
> Why do you presume his actions are reflexive, and the "babbling" man's are
> intensional?? Because you identify more with the man and not the dog, and
> you know yourself to be "conscious"? Yikes -- scary social implications!
> Careful with that one...

        Again, we measure the Shannon entropy of the man and dog's possible
messages and conclude that the dog's informative behavior is less
expressive (see Peter's definition of expressiveness
http://www.cs.brown.edu/people/pw/papers/bcj1.pdf) or more predictable
given the same third observer. Again Peter's work is relevant here...
 
> And "past perceptions"? Based on what testing? Prove to me that you're any
> more conscious (as opposed to more intelligent) than a dog. Or better yet,
> give me a way to test whether or not the dog is conscious. It's not
> useful to make untestable assertions.

        We look at the range of possible messages that each could communicate
to possible others. This ties into the definition of Shannon entropy. We
have to be behaviorist, since we can only observe the external behaviors
of a Local System, be it an atom of hydrogen of a dog, etc. We infer,
via modeling, the possible internal behavior that could give these
surface features, but we can never assume absolute certainty if we are
limited to a finite period of observation and/or a finite number of
simultaneous examples, e.g. a finite time series and/or an ensemble.
(notice that these look like time and space from the statistical point
of view!)
 
> Have you read the one about the gorilla that was taught sign language? It
> told a trainer the story of the day hunters came and took it from its
> family, and how this experience made it sad. Imagine that -- a conscious
> gorilla!! Same gorilla, before and after it learned sign language. I
> communicate, therefore I am.

        We can communicate therefore we share a world in common within which we
can define ourselves reflexively. If a pair of possible observes can't
communicate, it does not mean that neither exists, per say, they just
would perceive (pair-wise?) disjoint worlds!
 
> Easy question: If you had your life to live over again, starting with the
> *exact* same initial conditions (including your own knowledge -- you didn't
> know anything more than you did the last time), would you do anything
> differently? Assuming a deterministic, if not perfectly measureable,
> universe ... of course not!! A deterministic universe precludes the concept
> of freewill and, by extension, consciousness. We're all utterly products of
> our environment -- no more, no less than the dog. We are both mere
> biological machinery. The argument between intent and reflex is artificial
> and not at all useful.

        The problem with this line of thinking is that it only is correct if
these "*exact* same initial conditions" are even knowable! It can be
shown that any given set of co-moving observers have an (asymptotically
diverging with distance) possible knowledge of each other's behavior. If
we try to set up the situation in your question, we find that there are
more than one set of possible initial conditions that give the same
behavior at any given moment. We model this "moment" in classical terms
as the set of space-like hyperplanes that have the same event in common.
This is a very controversial and difficult idea that needs a lot more
discussion.
 
> The argument for consciousness is the argument for a soul -- the little man
> inside, experiencing the world through your body. You once asked me, "Where
> is your 'soul gland'?" Well?
>
> --Jerry

        There is no 'soul gland', as Vaughan Pratt explains well in his papers
http://boole.stanford.edu/chuguide.html#ratmech, etc. The link between
mind and body is the adaptive mapping between the dynamics of the
'physical' processes of 'brain' and dynamics of the 'information'
processes of mind. There is no clear cut separability of brain and body
and physical phenomena going off to the horizon just as there is no
clear cut separability of sets of information... This also needs to be
looked into...

Onward to the Unknown!

Stephen



This archive was generated by hypermail 2.0b3 on Sun Oct 17 1999 - 22:10:31 JST