From: Henry Stapp <stapp@thsrv.lbl.gov>
To: chrisnunn <chrisnunn@compuserve.com>
cc: Stan Klein <klein@spectacle.berkeley.edu>,
        "jeffery m. schwartz" <jmschwar@ucla.edu>,
        David Hodgson <raeda@ozemail.com.au>
Subject: Re: Where's consciousness?


[Stapp]
Dear Chris,

In a letter to the q-mind moderator you said:
"...
Henry Stapp sent
me a detailed and thoughtful reply.  Among other things, he pointed out
that I had misunderstood his point of view. I would not like my remarks to
be posted unless Henry's reply also appears,.... I hope he'll agree as 
his letter provided a useful clarification which others besides me should 
find helpful."

[Stapp]
Chris, my letter to you was a personal reply to you alone (apart from 
cc's to Jeff and Stan).

You also said in a letter to me:

[Chris]
"I admire your somewhat austere vision, but my fuzzy medical training keeps
getting in the way of understanding it adequately. Am slowly getting there."

[Stapp]
My remarks were terse, and they did indeed describe an austere approach.

I have now gone over my letter to you, and have made some small changes 
designed to make it more accessible to a broader readership, and have 
added some clarifications at the end designed to ease the austerity.
The new version reads as follows:

On Fri, 19 Mar 1999, chrisnunn wrote:
 

> Dear Prof. Stapp,
> 
> You wrote a while ago to Jack that your present thinking is that
> consciousness is confined to people. I know you've stopped contributing to
>  q-mind, but wonder if you'ld be willing to elaborate on this particular
> topic.
> 
> Looking back over what you've been saying, particularly in connection with
> your 'Heisenberg choice' argument, suggests that a principal function  of
> consciousness in your scheme (i.e. registering an increment of knowledge)
> should be fulfilled by any measurement-device/pattern-recognizer  - 
> whether a simple molecule that can, say, respond to  the arrival of a
> photon, or Einstein recognizing the equivalence of gravity and
> acceleration.
> 
> If a principal function of consciousness can be fulfilled by this broad
> range of devices, it's tempting to infer that consciousness itself must
> either be present in the full range or is epiphenomenal and could be
> omitted from your proposals. Of course there's room for all sorts of
> subtleties concerning what sort of 'measurement' might be relevant, what
> constitutes 'recognition' etc.  Is it in relation to these that your idea
> of human consciousness comes in?
> 
> Regards. Chris Nunn
> 

 
[Stapp]
Dear Chris, 

As I have repeatedly stressed, my immediate intention, in the development
of this theory, is to create a theoretical framework suitable for
mind-brain research. Mind-brain research involves two kinds of data
that both conform to Bohr's dictum that science is about relationships
between descriptions that allow us to communicate to our colleagues
what we have done and what we have learned. The first kind of data
are descriptions [expressed in the plain language of our everyday dealings 
with the objects around us, refined by the concepts of classical physical
theory] of what we have done to objects in that world, and what we have 
learned about the common world we share with the human beings around
us with whom we can communicate. These descriptions consists of descriptions 
of how we have constructed and put into place our probing instruments, and 
set the initial conditions, and the descriptions of what we have observed 
about the subsequent behaviour of these devices. These are the contexts in
which quantitative predictions can be made.


In the particular context of mind-brain research these instruments
are probing the physical properties of the body/brain of a human subject.
The second kind of data are descriptions expressed in the language of 
folk psychology, refined by the concepts of academic psychology and
clinical psychiatry. They consist of descriptions, within this language,
made by this human subject, of his "subjective" experiences, i.e., what he is
doing, mentally, and what he is learning about his mental state, and is
able to communicate within this language to the experimenters. Thus we have
the prerequisites for quantitative studies.  

My theory is the assumption that the basic precepts of orthodox
quantum theory [in the form advanced by von Neumann and Wigner, rather
than in the more restrive form known as the Copenhagen interpretation] 
hold in this mind-brain-research context. 
These principles, as I formulate them, are:


1) The *physical* properties of a system are completely represented
   by its quantum state S.
2) This state S represents *certain aspects* of reality. 
3) The aspects of reality represented by the quantum state S behave
   like a body of absolute objective knowledge: it jumps to a new 
   `reduced' form whenever any `knowing' occurs. A `knowing' is an 
   increment in the knowledge represented by S.
4) Each human being is a person. A person consists of a sequence of 
   `knowings' bound together to form a stream of consciousness, plus
   the `physical system' that binds this sequence together.
5) Each physical system is associated with a set of physical variables. 
6) State of a physical system b is defined by the formula:
   S_b = Trace_b S, where Trace_b represents the trace over all degrees
   of freedom EXCEPT those associated with the system b.
7) The Trace operation is the quantum equivalent of the classical-
   physics operation of integration over phase space with the weighting
   factor set to unity: it is the unique operation that represents
   lack of knowledge. [If a bounded operator is represented (in any 
   orthonormal basis) by a matrix, then the trace operation is just the 
   sum of the diagonal elements of the matrix: the sum of the matrix 
   elements Mrc in which r(the row)=c(the column).]
8) Reality itself includes all knowings, and in particular all knowings of 
   the kind that conform to Bohr's dictum about communicability.
9) The occurrence of a knowing K changes the prior state S
   to [P(F) S P(K)], where P(K) is the projection operator that removes from 
   the state S all parts that are incompatible with the increment of
   knowledge K.
10) If the state is S, and a question is posed such that an *affirmative*
   answer would reduce S to PSP, then IF this question is actually 
   posed the state S is first reduced by a von Neumann process I type of
   reduction to PSP + (1-P)S(1-P).
11) If the question mentioned in 10) is actually posed then the
   the propensity for the affirmative answer to appear is
   given by the formula Trace PSP/TraceS.
12) The process that chooses the next question is a real process.
   It is not fixed by S alone: i.e., it is not fixed by those aspect of
   reality that are represented by S. The knowing aspects of
   reality enter into the process of formulating the question that will be 
    posed by the body/brain/mind system that constitutes a human being.
   The mind part is a sequence of discrete experiential events that are called
   "knowings", and the projection operator P(K) associated with a knowing
    K acts on the degrees of freedom of the associated body/brain
13) The theory in this initial stage is restricted to human science:
   it is a theory designed to bring order to phenomena observed
   by human beings, and described by them in accordance with Bohr's
   dictum. That is, in this initial stage the knowing subjects are all
   of the same natural kind, and they can therefore reasonably be assumed 
   to have similar properties, if seriously anomalous individual are
   culled out.


I have already spelled these things before on q-mind: I repeat them here 
because they are relevant to your questions.


Let me then consider your questions. You say:


> 
> Looking back over what you've been saying, particularly in connection with
> your 'Heisenberg choice' argument, suggests that a principal function of
> consciousness in your scheme (i.e. registering an increment of knowledge)
> should be fulfilled by any measurement-device/pattern-recognizer  - 
> whether a simple molecule that can, say, respond to  the arrival of a
> photon, or Einstein recognizing the equivalence of gravity and
> acceleration.
> 


You have it backwards! Consciousness does not register an increment of 
knowledge that a device registers. An increment of knowledge *occurs* in a 
person's stream of consciousness, and this reality, the "knowing" K,
is *represented* in the partial representation S of reality that is called 
the physical world (or the physical system) by the reduction S-->P(K)SP(K).


In particular, I follow the primary insight of the Copenhagen
interpretation, namely that the physical description in terms of the
normally evolving quantum state S does not specify the conditions for
a reduction to occur, and that we should therefore take our knowings
as empirically given realities, and the mathematical formalism as
a representation at least of some (at least approximately) valid
relationships among these empirical realities.

For a "knowing" (hence a reduction) to occur a question must first 
be posed, and the posing of the question involves parts of reality that 
are not represented by S. This is the big problem with quantum theory:
there is no indicator `within what is represented by the physical state 
S' that specifies when and where reduction events occur. According to
orthodox ideas the questions are pose by human beings, who also
observe the answers that nature delivers to the questions they pose.
I adhere to this approach, even though the human being's body/brain
is, in the vN/W formulation, within the physical system represented by 
the physical state S. [In the Copenhagen interpretation the
experimenter/observer, and  his instruments, stand outside the system 
represented by S.]  


Thus I regard reduction as an *effect* of the occurrence of a knowing.
not a *cause* of the knowing. This is in line with the 
main insight of the founders of quantum theory, which was that
knowings are represented by reductions, not vice versa.


In the classical limit, or idealization, the effects of knowings are
squeezed out, and one gets the completely misleading impression
that the mathematical representation provided by physical theory is
essentially isomorphic to the reality itself, rather than being merely a
representation of certain particular aspects of reality that allow us to
represent certain statistical connections between our knowings.


I have focused, above, on the technical structure of the theory.
But some brief remarks on the rationale may also me useful to you.


Consider the role of mathematical representations. In Newtonian mechanics
the mathematical representation was essentially isomorphic to what was 
imagined to be the reality itself: a set of point particles (or almost-point
particles) moving on continuous paths in spacetime. Already in Maxwell theory
there were doubts as to whether the electomagmetic field really existed in the
same real way that the particles of Newtonian physics were imagined to exist.
In quantum theory the connection between the mathematical structure and 
the reality became much more debatable. In the first place, the sudden
jumps did not seem appropriate to "physical reality". Moreover, several
competing ontological pictures were proposed that give either exactly 
the same, or for all practical purposes (FAPP) the same, empirical 
predictions. Hence scientists did not have a clearly-best picture of 
nature herself, as they had had for Newtonian mechanics. Consequently,
most scientist chose to follow the  conservative Copenhagen 
viewpoint, which maintains an essentially agnostic stance on ontological 
issues, and sticks instead to the epistemological ones of what scientists 
can know, and how these knowings are connected to one another. 

 
My approach is to stay as close as possible to this minimalist stance.
I do including the instruments, and all other physical systems, in the 
complete system represented by the quantum state S, in order not to 
have an artefactual splitting of the physical world into parts, and to
allow the theory to cover (in principle) cosmological issues, which
the Copenhagen interpretation cannot.  But I 
apply occam's razor to keep the theory free of all superstructures other 
than what is necessary to carry out the calculations needed to account 
for observed phenomena. This is in line with what I believe to be the 
main illumination provided by quantum theory, namely that remarkable
fact that the classical concept of tiny invisible particles can 
be completely eliminated, at the expense of explicitly introducting our 
experiential knowings directly into the dynamical equations. This
interesting switch eliminates any reliance on that theoretical 
invention of dubious status, and brings our knowings themselves 
directly into the mathematical structure: it automatically produces
a theory of mind-brain interaction suitable for the study of human
beings by human beings.


A final point concerns the connection between vectors and realities. 


A vector (a, b, c) in a three dimensional space R3 is triple of real 
numbers. It can represent a displacement in three-space, or a velocity in 
three-space, or a momemtum in three-space, or the angular momentum of a 
spinning top in three space. Indeed, a vector can "represent" any one 
of thousands of physical properties. But what a vector it IS is a vector, 
pure and simple: it is a strictly mathematical construct. One can
divine from the behaviour of a vector that represents certain features 
of some reality some aspects of that reality, but generally not every aspect
of that reality. Thus one should not expect to find the real quality
"redness" within the state vector that prepresents an evolving brain.
Reality has its intrinsic qualities, some of which are our human knowings.
Contemporary science has access to two aspects of reality: human knowings,
and the properties of reality represented mathematically by our
valid physical theories. Quantum theory tells us how these two aspects fit 
together. It does not suggest the either aspect can be completely reduced 
to the other. I have, in my paper "Attention, Intention, and Will in
Quantum Physics" indicated how, in more detail, these two aspects can
be entwined, with neither reducible to the other.


You said earlier that you thought that my answer to you might be useful 
to readers of q-mind. I no longer see or follow to q-mind. But I have no 
objection if you want to pass this communication to q-mind, as my response 
to your personal request for clarification of my position on these matters.
I would certainly want it to appear, if at all, only as a whole, 
just as I have written it.  


Best regards, 


Henry