From: SMTP%"PSYCHE-D@rfmh.org" 11-OCT-1996 00:05:36.05
To: STAPP
CC:
Subj: Re: Wrapping up QM and C.
Approved-By: PATRICKW@CS.MONASH.EDU.AU
Approved-By: STAPP@THEORM.LBL.GOV
Message-Id: <961010222407.22204f35@theorm.lbl.gov>
Date: Thu, 10 Oct 1996 22:24:07 -0700
Reply-To: PSYCHE Discussion Forum
Sender: PSYCHE Discussion Forum
From: Henry Stapp
Subject: Re: Wrapping up QM and C.
To: Multiple recipients of list PSYCHE-D
We are now wrapping up these discussion of QM and consciousness.
I have spent a lot of time and effort trying to be clear, and to clear
up what seems to me to be some unclear thinking on these matters.
What conclusion can be drawn from all this effort?
Is there nothing but a welter of confusing claims and counter-claims
that can allow a peruser of these exchanges to conclude only that
there are different equally valid points of view?
I think not! Pat has given his view on what has been achieved: very little.
I would like to give mine: I think some important points have been established.
The discussions were obscured by an initial misunderstanding. I made it clear
from the outset that I was making here only the claim that " the principles of
CM do not *entail* the existence of consciousness", not that "consciouness
was *incompatible* with the principles of CM. This weak claim, namely
that "CM does not entail C", I thought to be obviously true, and I
had taken taken it as a secure starting point of the arguments in
my paper "The Evolution of Consciousness" [To appear in the proceedings of
Tucson II, and available from my web site]. It was Sarfatti's mention of
this point that launched this debate. Pat challenged my claim, which I
repeatedly described as point of contention. Then at a relatively late stage
(Sept25) Pat said that: "There seems to be basic flaw in Henry's logic
throughout this discussion". He pointed out the difference between saying
that "CM does not entail C" and saying that "CM is inconsistent with C",
and said that he would happily concur with the first claim: At this point
he said it was trivially true. So, at least in some first approximation,
we had then come into agreement on the issue under discussion:
we both agreed that the principles of CM did not *entail* the existence of
consciousness.
This is, I think, an important concession on Pat's part because it puts in
jeopardy his claim that consciousness is both 1), adequately described
within CM, with no need to go to QM; and 2), simply a complex activity
of classically describable brain, nothing more. For, according to the
principles of CM, brain activity is entailed by the principles of CM.
Hence it would be an apparent problem to admit that these principles
did not entail consciousness, if consciouness is nothing more than
brain activity. Also, zombies would be possible in principle within
the framework of CM, if the principles of CM did not *entail*
consciousness.
On another front, I think I effectively nullified the power of "arguments
by analogy" that have often been raised in this context. Analogies can be
useful to guide our thinking, and to alert us to the ever-present danger
of making mistakes now, by reminding us that mistakes have been made in
the past. But argument by analogy cannot be used to cast out all attempts
to apply reason to a new case. In the end one must look at the details
of the new case.
Going into the details of the present case brings in the need to consider
the role of `boundary conditions' and `bridging principles'. When one
speaks of the principles of CM, and of what can be deduced from them, one
can distinguish three parts of these principles: 1), the `general
mathematical principles', which apply to all worlds that, according to these
principles, are physically possible; 2), the possible `boundary conditions',
expressed within the mathematical framework, which single out individual
cases (or perhaps particular collections of cases) from among the infinite
set of all possibilities allowed by the general mathematical principles;
and 3), the `bridging principles' that associate particular features of the
mathematical description to corresponding experiences or perceptions of
the scientists and technicians that are using the theory.
The development of quantum mechanics forced physicists to pay serious
attention to these `bridging principles'. Following the lead that
Einstein pioneered, in his development of the special theory of relativity,
the quantum theorists focussed on the readings of measuring devices.
(After all, physical theory is a quantitative discipline.) The key move
in the original Copenhagen interpretation was precisely to take the
bridging principles to be associations of the following kind: " `the pointer P
points to the number 8.5', within the mathematical model " corresponds to
"an experience that we could describe by saying that `the pointer P points to
the number 8.5'".
This move achieved two important things. First, it identified certain
(classically describable) features of the QM mathematical model as
what the theory was about: the mathematical theory was construed to be a theory
of predictions about meter readings of this general kind. The mathematical
structure itself had no way of singling one what the theory was about. Second,
the bridging principles then quarenteed the predictions of the mathematical
theory could be interpreted directly as predictions about our experiences.
This formulation links the mathematical theory tightly to our experience,
and resolves in a clear way questions about bridging principles, as they are
used in contemporary pragmatic quantum physics. Given this pragmatic structure,
in which new knowledge is represented by a collapse of the wave function to
a new form compatible with the new knowledge, it becomes an almost automatic
consequence of the attempt to "ontologicalize" this pragmatic structure that
the occurrence of the new thought would be associated with a collapse of the
wave function of the brain to a form compatible with the new thought.
It is, of course, essential to this whole logical structure that the physical
structures that serve as the measuring devices can be represented within
the physical theory that is supposed to represent the physical world. It is
these devices that probe the system under study, and bring the information
up to the level of directly observability.
This brings us up to the present in the discussion with Pat Hayes.
In his latest posting (Oct9) he says, in regard to measuring devices, and
their component parts:
"The concepts of 'pointer'and 'spring' do NOT occur in CM. (If Henry claims
they do, please cite the laws of classical mechanics which mention pointers
and springs.) I can give a description of a particular spring in terms of
CM, but that description already amounts to a 'bridging law' (albeit not
much of a *law*) between CM - which is *entirely* concerned with particles
and fields - and the larger conceptual world which also contains a certain
spring. This is how CM + (bridging assertions - let us drop the 'law')
might indeed entail things about gas guages. That is, CM makes predictions
about other things than particles and fields *when extended by bridging
descriptions which relate these things to particles and fields*. Now,
could there be such bridging descriptions for subjective phenomena? We
don't know, but Henry STILL hasnt given us any logical reason why not. All
he does is exhibit a prejudice by refusing to use subjective vocabulary
when talking about CM but allowing himself to when talking about QM."
I think there is some serious confusion here. The whole analysis here is
going on in someone's thoughts, or perhaps in several people's thoughts.
Part of the thinking is about a mathematical structure, classical mechanics.
This person has also various images from his experiences: images of hurricane,
and billiard balls, and springs and pointers, etc.. I have already emphasized
that the mathematical structures of CM are abstractions from our experiences:
they are abstractions from our experiences of billiard balls and ocean waves
viewed from afar. One has no trouble imagining a particular sort of structure
formulated completely within the framework of classical mechanics that could,
because of its properties within the theory, be called a hurricane, or a
biliard ball, or a pointer, or a spring, or a device that would, if placed
in a box containing a gas, move to a state where the pressure of the gas
would be closely correlated to the position of the pointer. All this would be
set up within the thoughts of some scientist, by thinking about the
mathematical properties supplied by classical mechanics, and other images
of things viewed from afar that could be conceived to be built out of
conclomerates of "particles" which are abstractions of things viewed from
afar.
The point at issue is that if one now introduces into this CM mathematical
model, consisting exclusively of things seen from afar, some structure that
is supposed to be the theoretical image not of a "spring" but rather of
a "human being" then there is *from the principles CM* no entailment of
any consciousness in the CM model of this "human being" seen from afar.
There would be entailment, by the laws of CM, of the "springiness" of
the structure called "the spring", and entailment of the "havoc wreaked",
within the model, by the complex combination of particles and waves called
"the hurricane". And there might even be entailment of "apparently
thoughful behaviour" of the complex combination of particles and fields,
within the model, called "the human being". But there is no principle of
CM that would allow one to deduce that this conglomerate of classical
particles called the "human being", as viewed from afar, had the sort of
experiences that real human beings experience.
I thought that Pat had agreed with me about this.
Pat's argument, given above, said:
"That is, CM makes predictions
about things other than particles and fields *when extended by bridging
descriptions which relate these things to particles and fields*. Now,
could there be such bridging descriptions for subjective phenomena?"
According to the account given above, CM says things about "hurricanes",
and "springs" and "pointers" considered as elements of the CM
theoretical description, without ever bringing in the bridging principles
that relate the elements of the physical theory to the experiences of
real observers. That is not part of standard CM. The whole discussion of
the theoretical model, with its "hurricanes" and "springs", is within the
framework of views from afar: both the "hurricane" and the "particles" are
abstractions from images of things viewed from afar. One can never
deduce from the classical ingredients supplied by CM the "thoughts" of a
structure built from the ingredients of the theory in the way that one can
construct the springiness of a "spring" or the havoc wreaked by the
"hurricane".
One could arbitrarily introduce into CM the same bridging principle that
was needed to make QM work. But the difference in the two cases is that
in the CM case the principle is an ad hoc addition that is not a necessary
part of the theory, and the addition is dynamically inert, because CM is
dynamically complete; but in the QM case the bridging principle is an
essential part of the theory, and is dynamically crucial!
To justify this last asertion one must look at the various interpretations:
the collapse, Everett, and Bohm interpretations.
In the collapse interpretations the crucial thing, in order to get the
right predictions, is essentially that the collapses be to states
that correspond to single experiences, not to states that span several
experiences. So one must either bring in, in a direct way, the demand that
the collapses are to states that correspond to "individual possible
experiences" or achieve the same result in some indirect way.
The Bohr/Heis/vonN/Wigner approach is to accept that "experience" is
ontologically important, and to require directly that collapses be to states
that correspond to possible individual experiences: no other justification
is deemed necessary for this particular collapse rule. Experience *ought to
play* some causally efficacious role: if it did not it would have reason
to exist.
In the Everett interpretation one imagines the wave function to be decomposed
into different orthogonal branches that correspond to "different possible
experiences", and that then in Zurek's words " `consciousness' will perceive
the wave function `branch by branch'". That is, the wave function, which
is one single entity that can be decomposed into branches in an infinitude
of ways, is decomposed according to conscious content, and *probabilities*
are associated with these particular branches as if they were *disjunctive
realities*, even though ontologically the wave contains all of the
branches conjunctively, like the various ripples on the surface of a pond.
This places consciousness in the role of separating into *disjunctive
realities*, to which separate probabilities can be assigned, that which
the wave function gives as *one reality*. This notion of assigning
different probabilities to parts of a reality that all exist together
is really puzzling. I think it is an absurdity. Something "else", other
than the normally evolving wave function, must be introduced in order
to make sense of that, and consciousness is certainly the strongest
contender.
These branches will tend to de-cohere, in the sense that it will be
very difficult to design an experiment that could measure an *interference*
between different branches. But even if such experiments are very difficult to
perform, all of the branches are still present conjunctively, in principle,
unless one modifies the Schroedinger equation. So some extra principle is
needed. This principle could perhaps be stated in some way that did not
explicity mention conscious experience, but it would have to produce exactly
the same result as defining the disjunction in accordance with conscious
content if one is to obtain the orthodox predictions of quantum theory.
In the Bohm model there is the question of why consciousness is associated
with just the single classical universe--which is represented as a single
moving point in the 3N-dimenional space in which the wave function evolves--
rather than with all the branches of the evolving wave function, as in the
Everett model. The only reason to include this moving point in the ontology
is to account for the fact that our experiences do not have the structure
of the evolving wave function of QM: some other structure is needed to
account for the structure of our experience.
Although a lot of work has gone into the effort of trying to keep
consciousness out of quantum theory, and of thus returning to the traditions
of 19th century physics, it is not clear that this has any sound motivation:
consciousness is part of reality, so there is no good reason to try to squeeze
it out of physical theory if has a natural and dynamically efficacious place
there.
Henry P. Stapp
_________________________________________________
Recent papers available at:
http://www-physics.lbl.gov/~stapp/stappfiles.html