From stapp@thsrv.lbl.gov Wed Oct 18 15:45:35 2000
Date: Tue, 17 Oct 2000 17:15:29 -0700 (PDT)
From: stapp@thsrv.lbl.gov

Subject: Re: Where Aaron disagrees with Pat (Was Re: Position Paper...)


Dear Pat,

Many thanks for taking the time to make the detailed criticisms of 
my earlier reply. Your comments sharpen the issues, and allow me 
formulate more incisively, and I hope now convincingly, my 
objections to the "identify theory" position that you are defending.

On Mon, 25 Sep 2000, pat hayes wrote:

[Henry wrote}]
> >The case at hand is the statement "The painful feeling that is
> >"my experiential pain-in-my-Big-Toe" is identically the same thing as a
> >certain reality that is represented in classical physical theory as some
> >physical activity in my brain."

> >In this case a clear meaning can be given to this assertion. Each
> >of my experiences is a component of the total body of experiences
> >that constitutes my knowledge. These experiences are exactly the
> >realities that constitute my immediate knowings. Probably nothing
> >at all can be less ambiguous, as regards what it is, than an enduring
> >pain.
> 

[Pat]
> I disagree with all of the preceding sentences, and think that they 
> embody a basic philosophical error, which I will try to elucidate in 
> more detail below.
> 
> > This pain itself is one reality. The other reality is an
> > activity of a swarm of electrons and protons in my brain.
> 
> This talk of two 'realities' is just word-salad, and very misleading. 
> There is one reality, which can be perceived in a number of different 
> ways and from a number of different perspectives, and described in a 
> number of different ways. But all these differences do not create a 
> multiplicity of realities, any more than being able to walk around a 
> tree creates a world of different tree-realities.

Even if I accept the terminology that, within the framework of classical
physical theory, everything made out of particles and fields is one
single reality, the one physical world, there is still the issue of the
nature and constitution of the different perceptions of that one physical
world.

A language that allows perceptions to be different from the perceived 
physical reality is not a "word-salad", but is rather the result of 
a legitimate use of different words to describe things that might be, or 
conceivably could be, different, in order not to close off by 
linguistic conventions the possibility of even investigating whether 
the two possibily different things are the same thing or not. 


> > .....there is probably no question
> >of sameness that can have a clearer answer than the question of
> >whether an immediate and enduring pain-in-the-toe is exactly the
> >same thing as the reality that is an activity of a swarm of classically
> >conceived electrons, neutrons, and protons in the brain.
> 
> I see no reason why this case is particularly clear. Pains are in 
> fact rather mysterious, and often give us only very partial and 
> incomplete information about where they come from or what their 
> causes are. (I'm just reading 'Phantoms in the Brain', about phantom 
> limbs and related phenomena. Fascinating stuff.)
> 

These disparities between the experiences and what they purport to
be OF, or ABOUT, strongly suggests that the experiences are
different from the physical situation that they seem to be "about".
So what they are, and how they are connected to the classical-physics
conception of physical reality, must be considered to be an initially open
question.

> >When one is considering two different descriptions of the same reality,
> >or two different appearances of the same reality one justifies the
> >claim of sameness by noting the lack of any difference between the
> >two realities that are being described in different ways, or experienced
> >in different ways.
> 
> I'm not sure what you mean by this. Read in one way it is just a 
> tautology; read in a different, more substantial, way, I think it is 
> false. We come to conclusions about identity on many different 
> grounds, many of them quite different from 'noting a lack of 
> difference'. The evening star and the morning star are known to be 
> identical, but not because anyone is confused about the difference 
> between morning and evening.
> 

We have two different names for the two different experiences, but we
have come to understand, via physical investigation, that these two
differently described and differently felt experiences are experiences OF
(or about) two things that within the classical physics ontology  lack any
difference: the two objects of the two perceptions are, within the
classical physics ontology,  the same physical planet.

This is the typical way of justifying the claim of sameness of two 
`objects of perception', which are perceived or experienced in
two different ways. But this argument for `sameness' does not have the
format of a proof that could allow one to equate an experiential
happening, a pain, with some activity in a physical brain, as a brain is
conceived of in classical physical theory. For the defining qualities of 
the experience cannot be said to `lack any difference' from the defining
qualities, within classical physics, of a physical object (or activity):
the defining characteristic of the pain is a horribly unpleasant feeling,
whereas the defining characteristics of any feature of the physical world,
as conceived of in classical physical theory, is in terms of shapes in
spacetime and numbers. These two defining characteristics do not
lack differences. 

> >But in the present case one of the realities is an
> >experience itself and the other is an activity that, according to
> >classical physical theory, is completely describable without asserting
> >that any such thing as an experience exists.
> 
> The preceding sentence commits a basic logical error, which Aaron has 
> called 'nothing-buttery', TWICE. 


The "nothing buttery" assertion is the trivial assertion
that one complete description does not preclude other complete (or
partial) descriptions. This trivial fact is, of course, the basis of a lot
of physics: the first step in the solution of any problem is usually 
to find which one of the infinite set of possible descriptions is best
suited to solving that problem. 

I am not assuming that the `basic' physical description in terms
of the spacetime trajectories of the particles and the values
of the fields, is the ONLY complete description. What I said was 
that the physical brain activity "is completely  describable
without asserting that any such thing as an experience exists." 
I am NOT thereby denying the existence of other (equivalent) 
descriptions, any more then asserting the existence of a complete
description of the form (x,y) denies the existence of a complete
description of the form (x+y, x-y).

> First, it assumes that the reality 
> of experience is completely described in experiential terms (more on 
> this error below) 

I did not assume that!

I said that the *physical reality* is, according to classical
physical  theory, completely describable in a certain way: 

But I did not say an  *experience* is completely describable 
in any way at all.

Experiences can be partially described in many ways, but I would never try
to give a complete description of any experience: it slips away too fast.

> and secondly, it assumes that a classical physical 
> description is thereby a complete description. Both Aaron and I have 
> documented this last error in Henry's thinking to exhaustion, so I 
> won't repeat the old arguments here, but just note the error.
> 

No error here: according to classical physical theory the basic
physical description is ONE complete description of the physical
universe: it is ONE from among an infinitude.

> >There is therefore no logical
> >or definitional basis for asserting that these two realities lack any
> >difference.
> 
> There is no LOGICAL or DEFINITIONAL basis for almost any empirical 
> statement of identity, by definition of 'empirical'. There is no 
> logical or definitional basis for identifying heat with energy of 
> molecular motion, for example. Nevertheless, that is what heat is, a 
> fact which was established over a longish period of experimental 
> investigation and testing of alternative hypotheses. If it had a 
> logical or definitional basis, no empirical investigation would have 
> been necessary.
> 

"Heat", to the extent that it is well defined in classical physics,
and is not a subjective psychological feeling, is defined to
be average kinetic energy in certain idealized conditions of
equilibrium. Whole books and physics courses are built on that 
definition. The definition gets tied via other physical properties
into readings on thermometers, and gas laws etc.. "Heat" is a good
example of a word that, in classical physics, identifies, by definition, 
a certain feature that is entailed by the "basic" description,
but is not an actual element of the basic description.

Empirical studies ruled out other ideas of how 'heat' should be
defined, and have now abundantly confirmed the complex array of
interrelated predictions that follow from this definition of
"Heat".


> >Nor can there be any ontological basis. For the first is a
> >reality that known to exist, first hand, whereas the second is something
> >that is known not to exist.
> >

>An example may be useful.

Experiences  are known to exist: they are the
carriers or bearers of all human knowledge. If experiences did not
exist then human knowledge as we know it would not exist.

But we know for sure that the world as conceived of in classical
physics cannot exist, at least in this part of the world that we do
experiments on.

My point was that accepting a known-to-be-false ontology precludes
the possibility that there could be an ontological identity between it
and something known to exist.

> >
> >Consider a system that consists of four points moving in a Euclidean
> >space. They define 4X3/2=6 pairs of points, and hence six distances.
> >Let us say that a "square" exists whenever four of these distances
> >are equal, and the other two are larger by a factor root 2.
> >(Perhaps one might allow some well specified fractional deviations
> >from these exact conditions.) So as these points move around it would
> >make perfectly good sense to say that at certain times "a square exists".
> >or "squareness" exists.
> >
> >One can then consider the proposition:
> >
> >"Whenever `squareness' exists a `squareness feeling' exists in some
> >person's stream of consciousness."
> >
> >This is a ontological assertion.
> 
> I'm not sure what is meant here by an 'ontological assertion', but 
> this particular assertion is obviously false. I'm sure there are many 
> squares in the world of which nobody is consciously aware. So I fail 
> to see what point is being made here. Is this supposed to be somehow 
> typical of the claim of the identity thesis? If so, it is not even a 
> straw man. Straw hamster would be more like it.
> 

This example is meant to be an excruciatingly simple example of an
assertion that says that when some physical property---of the general
geometric kind that occurs in the classical physical description of the
physical world---occurs (in the brain of a person) then a certain
psychological or experiential property occurs. The point is only to
exhibit in a very simple form an example of the *kind* of assertion 
that the identity theory claims to be true.

> >The property `squareneess' is a property whose existence or
> >nonexistence is fixed purely by the geometric structure of the
> >system in question and the definition of `squareness'. Given the
> >geometric structure (i.e., the locations of the four points)
> >and the definition of `squareness exists"' the answer to the question
> >"Does squareness exist at time t?" is determinate and determined.
> >But the question "Does a `squareness feeling' exist in some person's
> >stream of consciousness at time t?" is not determined just by the
> >geometric structure of the four points and the definition of
> >`A squareness feeling exists in a person's stream of consciousness'.
> >There is an ontological issue that needs to be resolved before one can
> >conclude from the geometric structure, and the definition, whether a
> >squareness feeling exists or not in someone's stream of consciousness.
> 
> Yes, of course. The claim of the identity thesis is that a 
> state of 
> consciousness is identical with SOME physical state of 
> the conscous 
> organism, not that this physical state is somehow similar 
> to the 
> content of the conscious state. The physical description of a 
> conscious state of awareness of a square would be 
> unbelievably 
> complex, probably so long as to be unusable by any 
> physical computer.

By the "feeling of squareness" I mean whatever the 
experience is that occurs when the "squareness property
of the brain" comes into existence: `Similarity' was not
the point.
 
> 
> >This example IS analogous to the case at hand. In the 
> >case at hand
> >the physical structure is asserted to be specified in 
> >accordance with
> >the principles of classical physical theory, and this 
> >means that
> >the physical system is completely specified by geometric 
> >properties
> >alone.
> 
> 1. No, it means nothing of the sort: this term 
> 'geometric properties' 
> is a fantasy of your imagination, Henry, one that you 
> suddenly 
> started using in these emails about a year ago with no 
> warning and no 
> explanation, but which is not found as far as I know 
> anywhere else in 
> the entire world literature, and one that I have repeatedly 
> challenged you to define and offered counterxamples to, 
> but you have 
> never adequately responded.
> 

According to classical physical theory, the physical
state of the universe is specified by giving the
trajectories through spacetime of all of the particles 
in the universe, and the values of all the fields at all
spacetime points. This is ONE way of completely describing
the physical universe for all times. This is a "geometric
description" in the sense that it can be specified by assigning
a finite set of real numbers, and a yes-no value (Is a 
particle here?) to every point in spacetime. Every other
"physical" property is just some feature or aspect of 
this geometric structure: some aspect that is logically 
(i.e., by definitionally) characterized in terms of descriptions of
spacetime arrangements of these elementary values, or of objects 
made up out of these elementary values assigned to spacetime points. 

For example, an object would be an "oak leaf" if most of atoms 
that comprise it have a certain history: they come from physical objects
of a kind ABOUT which English Speaking people have said (or written) 
things like "That (physical object) is an oak tree." These speech acts are
associated with perceptual acts that are OF, or about, physical objects
that have certain geometric shapes and locations and vibratory
qualities, and the speech acts themselves are also qualities of this kind.
The `physical color' of the leaf could be defined to be "green" if certain
vibratory properties hold.  Note that these definitions are all directly
in terms of spacetime behaviours of physical objects. That is the sense
in which they are `geometric'.  A `pain', on the other hand,
is *fundamentally* characterized in terms of a "feeling" that is *itself*
directly defined in terms that are NOT spacetime arrangements of atoms.
The situation as regards physical "heat" (average kinetic energy)
and  "oak leaf" are therefore both completely unlike the
case of a "pain": in the cases of `heat' and `oak leaf' the
DEFINING CHARACTERISTICS are expressible in terms of spacetime
dispositions of various things, over the course of history, but the
DEFINING CHARACTERISTICS of a pain (a horribly unpleasant feeling) is
not geometric in this sense.

> 2. In any case, even if I ignore the fact that this claim 
> about 
> 'geometric' properties is meaningless, the claim that the 
> physical 
> system is "completely" described by any particular 
> collection of 
> properties is obviously false. There are many properties 
> of even 
> mundane physical systems which do not follow from their 
> classical 
> physical description (no matter in what language that 
> description is 
> couched.). The color of an oak leaf, and the fact that it 
> is an oak 
> leaf, do not follow from the physical description 
> (classical or 
> quantum) of the atoms comprising the leaf.  

The "color" of a leaf might mean a quality of the experience
of someone who is observing the leaf. That quality is very 
observer-dependent, and is not a *physical* property of the leaf
itself, within the ontology of classical physical theory.
On the other hand, certain vibratory properties of the
particles and fields that constitute the leaf, and that
that are therefore geometrically localized in the neighborhood
of the leaf, are physical properties of the leaf itself
that have a tendency to cause a certain kind of feeling 
or experience to occur in the stream of consciousness
of normal human observers who are looking at the leaf
in normal daylight, and this feeling goes by the name
of 'experiencing greenness'. The vibratory qualities are 
characterized fundamentally  in geometric terms,  but that 
particular feeling that I call `experiencing greenness' 
is not fundamentally characterized in terms of dispositions 
of particles, fields, and objects in spacetime.

> >The existence or nonexistence of a physical property is
> >determined by the geometric structure and the definition 
> >of that
> >the property in terms of those geometric properties. But 
> >the
> >existence of a corresponding feeling in someone's stream 
> >of consciousness
> >does not follow logically or definitionally just from the 
> >geometric
> >structure and the definition of that feeling.
> 
> You use two different standards in these two sentences. If 
> 'determined by' in your first sentence means the same as 
> the sense of 
> 'follow logically or definitionally' which makes your 
> second sentence 
> true, then the first sentence is manifestly false. 

I am using the same standards in both cases. Whether the object is
circular or square (within certain tolerances) "is determined by..."
and "follows logically and definitionally from" the basic geometric
properties. And whether or not a certain set of physical particles is an
"oak leaf" is both "determined by..." and follows "logically and
definitionally from" the existence or nonexistence 
of a set of spacetime properties that are fixed by the basic
ontological spacetime properties postulated by classical physical theory.

The first case (squareness or  oakleafness) is true, but the second 
(pain) false by this single standard. 

The "squareness of non-squareness" of a physical
object does follow from, and is determined by,
the combination of the geometric data contained the
basic description of the physical universe, and the
defining characteristic of squareness, but the defining
characteristics of a pain (that horribly unpleasant feeling)
may indeed always occur, in actual fact, whenever a certain
physical arrangement occurs. But the occurrence of  the
defining characteristic of a pain does not follow logically 
from the physical data, together with the definition of this
defining characteristic: there is a logical gap.


> If the  second 
> phrase means what the first one must mean in order for the 
> first 
> sentence to be true, then the second sentence is false. 
> You can't 
> have it both ways. Sentences about oak leaves do not 'follow 
> logically or definitionally' from classical physics 
> either: you need 
> a structural account of how oak leaves are made. If we 
> had a 
> structural account of how consciousness was implemented, 
> the two 
> cases might seem exactly analogous.

The truth or falsity of sentences about `oak leaves' do follow from
the basic physics data and the conventional definition of an
"oak leaf" in terms of the historical roots of its component atoms 
as characterized in terms of physical objects identified by
physical human speech acts. The truth of falsity of an assertion
`I feel a pain in my left toe' may be CORRELATED with assertions
of this kind, but the meaning of the pain statement is
not DEFINED by that that kind of statement: it is defined in terms
of the felt character of the experience itself, not in terms of the
apparent spacetime disposition of the (presumed) objects of perception,
and the inferred (via classical physical theory) spacetime properties of
invisible particles. 


> 
> >Whether or not a
> >feeling in someone'a stream of consciousness is an 
> >ontological issues that
> >is not settled just by the geometric structure 
> >and the definition.
> >
> >In short, it does not follow from the definitions that
> >
> >" The P feeling is the same thing as the physical 
> >property P."
> 
> Not from the definitions, no. But then it doesn't follow 
> from any 
> definition of 'oak leaf' that oak leaves are identical to 
> physical 
> collections of atoms. 

I have explained how, within the framework of classical physical
theory, the definition of `an oak leaf' does
allow certain collections of atoms (and the attendant fields)
to be identified as 'an oak leaf'..



> (It's easy to forget that serious 
> scientists 
> still werent all completely convinced of the atomic 
> hypthesis only 
> about a century ago, but I bet even Ernst Mach knew what 
> an oak leaf 
> was. )

Yes, in terms of the space-time dispositions of the
presumed objects of perception. The addition of
the precepts of modern classical physical theory allowed
those ideas to be connected to ideas about atoms.

> (I've already pointed out the error in saying that the 
> P feeling is 
> identical to P)
> 
> >Passing from physical property P exists to a P feeling exists
> >in some person's stream of consciousness requires an ontological
> >assumption that is not part of classical physical theory.
> >Thus within that framework the two properties are logically
> >distinct.
> 
> They are LOGICALLY distinct, in the sense that the claim that they 
> are identical is not logically necessary. But I'm not claiming that it 
> is necessary, only that if entertained as a hypothesis it would 
> elegantly account for a lot of otherwise very puzzling phenomena, and 
> there is no reason to believe it to be false. (Actually the 
> hypothesis connects a P feeling to a physical property Q, where Q is 
> something so complex that it might be practically impossible to 
> state; no matter, the hypothesis is only that it exists.)
> 
> > > Even if classical physics were true it would not be the case that ONLY
> > > things like atoms or sub-atomic particles or physical fields could
> > > produce effects, or control things.
> > > (That's the nothing buttery fallacy.)
> >
> >Hurricanes can certainly cause a lot of damage. And poverty
> >can cause crime. And activities in a person's brain can 
> >cause his body to move in certain ways.
> >
> >But it follows from the assumed validity of classical 
> >physical theory, 
> >and the `fact', within that conceptual framework, 
> >conscious experiences
> >are not the same things as physical activities, 

By, "within that conceptual framework" I mean "without 
inserting any extra ontological hypothesis". Within
this (restictive) conceptual framework  conscious 
experiences are not tied logically into this conceptual 
framework. They are ontologically different, basically, 
from such things  as hurricanes, 
which are existing physical features of the physical 
universe that are merely carved out and identified by 
conventional definitions. Given the (convenhtional)
defining properties of hurricanes, and the physical data, 
one does not need to have an extra postulate in order to 
be able to say that a hurricane is blowing down houses 
in Florida. But it does  require not only physical 
data (e.g. nerve impulses going in various ways)
and the defining characteristics of a "pain" 
(that horrible sort of feeling)  but also an
ontological postulate to deduce from the physical data
and the defining characteristics of a pain the 
conclusion that a conscious experience of pain is occurring:
one needs also a postulate that asserts that this certain set 
of physical conditions (nerve impulses going in various ways)
IS the very same thing as that horribly unpleasant feeling 
that is the defining characteristic of the pain, or always
accompanies it.

> >that a conscious
> >experience cannot "cause" a physical action. For a 
> >sufficient cause of
> >any physical activity resides, according to classical 
> >physical theory,
> >in other physical activities. If conscious experiences 
> >are not

within the restricted classical framework

> >identical to physical activities, and the physical 
> >activities
> >form a causally complete realm, then one can eliminate the
> >conscious experience without affecting any physical 
> >activities. 
> >But a conscious experience cannot properly be called 
> >a "cause"
> >of physical activity if that activity will necessary 
> >occur whether
> >that conscious experience is present or not.
> >
> >Unlike the case of a conscious experience (within 

restrictive

> >classical physics)
> >one cannot eliminate a hurricane without changing some 
> >physical activity. 


> 
> Hey, wait a minute. That isn't a 'fact' within classical 
> physical 
> theory!  Recall that the identity hypothesis is that 
> conscious 
> thoughts (as well as unconscious ones and all other 
> mental phenomena) 
> ARE the same thing as physical activities. That's the point:
> this 
> hypothesis restores the causal efficacy of consciousness by 
> a very 
> simple move, one that you still haven't refuted.


My position has always been that, within the classical physics
framework, conscious experiences are not NECESSARY parts of 
the ontology in the way that hurricanes, leaves, crime, and 
poverty are. An extra ontological postulate is needed to 
bring consciousness into the physical world.

Logically, the physical data and the defining characteristics 
of consciousness can be fixed and, within the classical 
physics framework, conscious experiences could
still be absent. So, within the strict framework of 
classical physics, consciousness is not causal in the 
same way that hurricanes and falling oak leaves are.

So my first point is that within the framework of
classical physical theory consciousness is an ad hoc
superfluity: it is not a natural integral part of the
physical world, but must be brought in by an extra
postulate that is not an integral part of classical
physical theory, as that theory is taught in our
univerities: those nerve impulses could perfectly well 
be doing whatever they are doing without there being 
any associated conscious feelings.

My second point is that there does not seem to be any
point or reason for the feelings to be present: the
physical world would be doing very the same thing whether
those feeling existed or not. So there is no possibility
of a dynamical linkage that would tend, in the course of 
the evolution of the brain, to make to make the feelings
correspond properly to the brain activities. 

My third point is that it is a manifest self contradiction
to postulate, within classical physical theory, that an 
experienced feeling is identical to something 
that is, within classical physical theory, well defined and 
not an experienced feeling: a "feeling" cannot be the
very same thing as a "nonfeeling". That is an absurdity!

I believe that the reason the  debate about "identity
theory" has persisted is that philosophers do not
want to tie themselves down to a particular specific
physical theory: they like to maintain as much generality
as possible. So they have not imposed the formal demand
that the ontological framework be classical physical
theory. Indeed, it is unreasonable to do so in the face
of the certain knowledge that this theory is incorrect.
I claim that within the framework of classical physical
theory the identity theory postulate is self-contradictory.
But the alternative of postulating a causally inert
world of experiences is almost as absurd. The obvious
conclusion to draw is that the framework of classical physical 
theory is inadequate for the treatment of this mind-brain 
issue. But not many philosophers are equipt to seriously
consider the quantum theoretical alternative. So the
debate within the philosophy literature goes on with 
no resolution.

Best regards,


Henry



From stapp@thsrv.lbl.gov Wed Oct 18 15:46:01 2000
Date: Wed, 18 Oct 2000 15:43:23 -0700 (PDT)

Subject: Re: Position Paper (revised) & Reply to Stan

Dear Pat,

This reply adds some details to my reply of yesterday to your
message of Sept 25.

On Tue, 17 Oct 2000, pat hayes wrote:

> Henry, a few comments on your reply to Aaron. Sorry this is delayed, 
> I wrote it a month ago but it apparently never got sent:
> 
> >....
> >in quantum theory those choices are not determined by contact interactions
> >between local bits of matter: some access to the brain as a whole
> >seems to be needed, or at least allowed. This sort of possibility
> >might turn out to be useful in modelling the CAUSE of choice.
> 
> This seems to me to be a thin and delicate distinction, if this is 
> all that the hypothetical explanatory power of quantum over classical 
> physics depends on. 

It is not! Please see "From quantum nonlocality to mind-brain
interaction" starting after nonlocality argument to see
explanatory power of QM.

> Aaron and I might agree that the brain as a whole 
> has causal powers which go beyond those described in terms of the 
> interacting particles of which it is composed. This idea that 
> large-scale structures have properties which arise from their 
> structure, as opposed to arising simply from the causal sum of their 
> constituent parts, is central to Aaron's way of thinking also. But 
> this does not require one to reject classical physics.
> 
> >The problem of our failure to communicate undoubtedly stems from
> >the matter mentioned already above, namely your focus on levels
> >of "explanation", and my focus on the underlying ontological
> >elements, and the structures that are made out of these elements.
> >The high-level structures may enter into interesting relationships
> >with each other, and may be identified and conceptualized by person's
> >seeking to explain features of the world to themselves or others.
> >But within the classical-physics conception of nature I have no trouble
> >distinguishing those high-level structure such as tables, and hurricanes
> >and roses, and computing devices from the ideas about them that human
> >beings use in their explanations.
> 
> I agree, but when it comes to computing devices you have to tread 
> carefully. Computers are rather different from tables, hurricanes and 
> roses in that they contain physical symbols. A physical symbol is a 
> symbol not just because it is described so in a human theory of the 
> machine. It is a symbol because of the way it is causally connected 
> to the rest of the machine and how this in turn is connected to the 
> external environment. These causal connections themselves may (indeed 
> almost always are) mediated in part through other such physically 
> encoded symbols. In a computer, meanings (of physically encoded 
> symbols) can themselves become causally efficacious. The (causal, 
> physical) result of a computation which determines the length of a 
> list, for example, is that a symbol is created which *denotes* the 
> length of the list.

I do not believe that a computer that operates according to classical-
physics principles is *essentially* any different from any other system
that operates according to classical-physics principles. We have designed
it so that it the classical-physics principles ensures that  it will
do certain operations, and OUR description of it brings out these
features. Certainly the notion of symbols is important for a compact
description of how the machine will behave, but the causal structure
rests finally on the behaviours of the tiny parts: the notion
of symbols, though important for a compact description of behaviour
and of structure, is not fundamental. If we were not trying to think ahead
to the problem of human consciousness I do not think that there would  be
any significant challenge to the claim that the high-level description
is just a useful way of compactly describing structural features
and resulting behaviour.


> 
> This is what makes computers special. Every other kind of artifact or 
> machine ever made can be described in such a way that its behavior 
> and its description can be isolated from each other: the machine 
> functions in a certain way, and is described (by describers, ie human 
> beings) in a certain way, and the former is quite independent of the 
> latter. The description may be useful, explanatory, etc., of course, 
> but (as you say) nobody has any trouble distinguishing the actual 
> machine from the ideas about it that people have. But computers are 
> genuinely different, since they really do have theories built into 
> them in ways that make the theories themselves causally efficacious.

They have our theories built into them in ways that allow the physical
mechanism to match our theoretical concepts.  There is nothing hugely
significant about the fact that we can be ingenious in this way, or if
there is some huge significance it is about OUR capacitities to put the
laws of physics to work for us. 
 
> This is not to say that your dichotomy between the machine and a 
> human description or theory of it becomes invalid. One can still make 
> THAT distinction; but one can no longer take the implicit next step, 
> which is that symbolic descriptions have nothing whatever to do with 
> the actual operation of the machine. In the case of computers that 
> second claim is demonstrably false.
> 

The machine functions, by virtue of the laws of physics, so 
as to conform to our ideas. So the functioning does have something to do
with our ideas about certain symbols. There is nothing strange about the
fact that we can exploit the laws of nature, and build machines that
follow certain logical procedures that we have created..



> So computers are genuinely symbolic and also purely physical 
> machines, and they provide an existence proof of the possibility of 
> symbolic machinery existing in a classical universe (well, since they 
> use transistors, maybe not, strictly; but that isnt the kind of role 
> that QT is supposed to play in your argument, right?)

Well, I am primarily interested in deeper quantum features.

>  So, the claim 
> then goes, maybe brains are also this kind of machine. 

But the problem is with *consciousness*.

> We can't prove 
> that, of course; but just entertain it as a hypothesis. It would 
> account for 'thought' as a causally efficacious phenomenon in a 
> purely 'mechanical' universe. 

The "thoughts" that are under discussion are "conscious thoughts".

By "it" I presume you mean the identity theory postutate:
the assertion that "feelings" are identically the same 
thing as things that, according to the basic ontological 
precepts of classical physical theory, are features of 
the motions of billions of particles. 

> It would reject the 'zombie' argument 
> by making zombies impossible. 

Yes,  the identity theory postulate rules out zombies.

> It might account for the structure of 
> our own subjective experience. 

> Overall, it seems to account in 
> general terms for almost everything that needs to be 
> accounted for, 
> and there seems to be no convincing argument against it.
> 

See below.


> >The former ARE space time arrangements
> >of the elementary physical units, and are in principle completely
> >described in physical reality in those "physical" terms, i.e., in terms
> >of spacetime arrangements of the elementary units, whereas the latter
> >ARE "feelings" of various kinds in some person's stream of consciousness,
> >and they are described in psychological terms. Some thinkers have no
> >trouble believing that a "painful feeling" IS a pattern of physical
> >activity, but there is no logical basis for making this assumption.
> 
> There is no LOGICAL reason to make or reject the assumption, but 
> there are persuasive arguments for accepting it, at least as a 
> working hypothesis, and as far as I can see, no persuasive ones for 
> rejecting it. Both you and Chalmers use an argument for rejecting the 
> identity thesis which can be paraphrased in terms of imagination:

I  strictly avoid, like the plague, any argument in terms of 
what can or cannot be "imagined". Chalmers employed such an 
argument, but I never have. 
.......

 
> Now, I don't expect you to just give in and agree with such a 
> challenge, but I do feel entitled to ask: how will you refute such a 
> challenge? What argument (or evidence) can you give which would show 
> that I am wrong? Because unless you have something to say which goes 
> further than the simple 'I can imagine a zombie' claim, we are not 
> having an argument at all, just announcing a difference of opinion. 
> You think feels aren't physical, I think they are, and that's all 
> anyone has to say. Not much to base an entire philosophy of mind on, 
> let alone a dialog between science and religion.
> 

Your "imagination" argument is not against any argument that *I*
have put forth.


> >
> > > > which has now pushed some philosophers to the
> > > > point of trying to convince us that consciousness, as we
> > > > know it, does not exist, or is an illusion, and other
> > > > philosophers to the point of making truth a social construct.
> 
> Who do you have in mind for the first alternative (that consciousness 
> is an illusion)? If it is Dennett, you are grossly mischaracterising 
> his position. If it is someone else, I'd love to know who.
> 

Well, Dennett "Explains Consciousness" without explaining
the experiential qualities that characterize consciousness.

In his debate with Searle (The Mystery of Consciousness: Searle:
1997, NYREV, Inc; p.99) Searle accuses Dennent of denying the 
existence of the subjective data. Dennett, in his reply,
(p 115-119) does not deny this charge, and says that the 
"problem of consciousness is going to remain a mystery until 
we find some such dead obvious intuition (about consciousness) 
and show that, in spite of first appearances, it is false." 
Dennett's candidate for the "fatally false intuition...
[is] the conviction that we know what we are talking 
about when we  talk about *that feeling*... How could anyone 
deny that!? Just watch.... I develop my destructive arguments 
by showing,,, how Searle's proposed "first person" 
alternative leads to self contradiction..."

These words of Dennett seem to me to be saying that he is 
claiming that the conscious feelings that we think we know 
are not what they seem to be, and that is surely a main theme
of "Consciousness Explained". At the end of this book he 
defends leaving out of his explanation of consciousness the 
subjective qualities that are the defining characteristics of
consciousness by saying that an explanation of something 
must explain it in terms of other things. But it then
becomes questionable whether he has explained the qualities
that need to be explained. But by suggesting that "feelings"
are not what they seem to be he can escape from the manifest
absurdity of saying that one's feelings of pain etc.
are identical to things that are not feelings.


> >It seems to me that if one starts with the classical physics ontology,
> >and builds the physical basically out of classical elements then
> >there is the inevitable question of why not just keep the entire
> >physical structure, but say that all these particles fly around
> >in just the way they do, but with no "feels" attached to these
> >motions.
> 
> This is easy to answer. IF the identity hypothesis is true, then the 
> 'feels' aren't 'attached' to the motions (actually more like patterns 
> of motions), they ARE (an aspect of) those motions. So of course one 
> cannot 'remove' them and leave the motions alone.
> 

The difference between "attached to" or "are an aspect of" 
the motions is not the issue. The issue is WHY are the 
feels present at all if they make no difference in these
motions.

> You may reject the identity hypothesis on other grounds, 
> but you cannot use this as an argument against it, since 
> this argument is 
> obviously invalid if the hypothesis is accepted.
> 

The argument here is on the basis of parsimony:
if the "feels" make no difference then parsimony
demands their absense.


I reject the identity hypothesis within the classical
physics framework because within the classical physics
framework `what the brain is' is in principle well defined, 
and its features are defined in terms of the numbers
that characterize the state of the brain. A combination
of numbers might be "accompanied" by a feeling, but it
can no more BE a feeling than a skyscraper can be a baseball.
One can hypothesize that an experienced feeling is "about"
the motions of some particles, or "accompanies" certain 
motions, but what it IS is an experienced feeling, not
an arrangement of motions of billions of particles.

> >The classical laws would say that the motions are all that
> >matter, so there is no logical or causal or ontological
> >constraint or condition that forces the "feels" to be present.
> 
> If the feels are identical to some motions, they are forced to be 
> present by the simple fact of identity. If a table IS a lot of atoms 
> suitably arranged in space, then removing the atoms removes the 
> table, and vice versa.

The table IS a collection of suitably arranged particles.
The name "table" is defined in a way such that the precepts
of classical physical theory---that the physical world
is made out of particles and fields, and that the objects
of perception are collections of these particles and fields, 
moving in accordance with the laws of nature---allow the 
name "table" to identify certain objects of perception that are, 
by virtue of these precepts, understood to be suitably arranged 
and moving sets of particles.

But a "pain" is characterized in terms of a certain 
horribly unpleasant sort of feeling. It is not an object
of perception, but an experience itself. The precepts of 
classical physical theory give ample opportunity for different
experiences to be experiences OF the same object of perception,
but a `feeling' is a *process of experiencing*. 

The precepts  of classical physical theory do not address 
the issue  of the relationship between the 
"process of experiencing" and the particles and fields that 
it postulates the physical world to be made out of. 

Identity Theory is a classical theory of the mind-brain 
connection. It glues the "process of experiencing" to 
the "classically conceived aspect of brain
activity".  It associates experiential realities to 
classically conceived physical realities. That is fine:
the experiential processing certainly needs to be tied
to the particles and fields. But the question is whether
it is rational to say that the relationship is one of 
"identity". Can one rationally assert that a certain
"feelings" is identically the very same thing as an 
aspect of a classically conceived motion? 

In logic we distinguish between *isomorphic connections*  
between certain features of two different (kinds of) things 
and *strict identity*, which can only hold if the two things 
compared are not only things of the same kind, but are
also indistinguishable in every way. But if we use
an overarching conceptual framework that contains
both experienced feelings and the physical realities
that are postulated to exist by the precepts of classical
physical theory, namely realities that are composed of 
fields and tiny particles that are not directly experienced, 
then it appears that the experienced feeling of pain is 
fundamentally and exclusively an experiential kind of thing, 
or process---it is characterized by the fact that it IS
a certain experience---whereas the associated aspect of 
brain activity is a construct made out of component parts that 
are  neither experienced nor experiences.  The two associated
realities are, not merely in how we describe or experience
them, but in their essential natures, not  indistinguishable.

The situation is not analogous to situations in which one 
single basic reality is experienced in two different ways,
so that one has two different experiences OF one 
physical reality. Here a unified whole experience 
itself is one of the two realities being compared, and the 
other is, according to the precepts of classical physical 
theory, made out of the motions of billions of particles.
These two kinds of realities are, by virtue  of their
intrinsic defining characteristics, not indistinguishable.


Best regards,

Henry