Friday, August 8, 2014

A subtle form of dualism





Below we have the austere  and brilliant dialogue between Freeman and Stapp. The former claims to be a physicalist, the latter a dualist.. I contend both are dualists

I refer the reader to the “Chinese room” debate. Searle insisted that his computationalist opponents had a disincarnated notion of running a program that made them ultimately functional dualist. IMO Walter has fallen into the same trap.

In his case, it is more subtle; what he refuses do methodologically is consider how symbols might work in the brain, and in this he follows his mentor Merleau-Ponty. Yet throughout this dialogue Walter is present, a giant symbolic mind with a passionate insistence on all phenomena fitting into third person science. In this he resembles Pat Suppes of Stanford. Henry replies with an insistence on the “ego” that betrays little understanding of cognitive psychology.

IMO, there is a synthesis; insist on the salience of subjective terms that CAN be fit into science. Minkowski’s frames, Von Neumann’s process 1 and others are such. Insist also on an attenuated free will, and the possibility of a Noosphere.  Forget about the nonsense of the “hard problem”; subjectivity will be explored in its heights b spiritual trail-blazers and artists, not scientists. That does not diminish  science; rather it allots it a proper place in society



Again, this is the situation;

1. QM demonstrated limits to “third person” descriptions and re-introduced the observer. It could be argued that Galiliean relativity did something similar, expanded on later b
 Einstein

2. The Qm argument has been over-extended to make every cognitive act seem one of “free will”

3. There is a vocabulary for discussing symbolic products like those of science; that of cog sci with recussion, grammars etc

4. Many “embodied” theorists like Merleau-Ponty have ignored this vocabulary and thus limited themselves to sensorimotor behaviour

5. What the Qm argument shows if we take into account ALL human symbolic behaviour
Is that in the limit there seems to be room for a limited free will

Sean O Nuallain


On Jul 31, 2014, at 3:03 PM, Henry Stapp wrote:
I continue, below, Walter’s clarification of the mapping between my vN-based understanding of the brain process associated with consciousness-collapse, and
Walter’s “Physicalist” understanding of it.
  
Henry

I continue our dialogue. 

Henry describes his "mentalist" (dualist) understanding of the mind-brain relation based on von Neumann's mathematics and contrasts it with his understanding of my "physicalist" (monist) understanding based on my observations of neural activity in humans and animals engaged in intentional behaviors. I believe that we agree that we can focus our descriptions of "matter-mind" on the process by which a stimulus is perceived. For example, a chemical detracted in the nose is categorized as something to eat or as something to run from. Through perception a choice is made on which action to take. 

Henry has stated (below): "In QM the dynamical rule that generates the evolution of the possibilities/potentialities is the Schroedinger equation.  The key issue is how is one (or a subset of) individual attractor(s) picked out for possible actualization.  Walter says “selection by a stimulus.” Copenhagen says “free choice on the part of the experimenter. “ It comes from something beyond the Schroedinger equation." 

Henry goes on to invoke the Born rule: "In QM this choice constitutes a choice of a “projection operator P. Once the observer’s choice is made , then the probability for the answer to the question “Will I experience the “Yes” response to the question represented by P?” is given by the Born rule Prob = Trace P Rho P/Trace Rho). Of course, this way of expressing the choice is designed to convert the description into a logical question embedded in a mathematical formalism."

When stripped of any reference to qualia and consciousness the issues before us, "mentalistic: vs. "physicalistic" appear to me more appropriately "mathematical" vs. "biological", including descriptions on both sides of equations and data in the King's English. 

By the phrase "selection by a stimulus", I refer to the biological phenomenon of "preafference" by which in anticipation of an act of observation the brain sets up in each sensory cortex a memory of previous encounters of what to expect, in the example something to eat, something to fear, and something other. These predictions have the form of an attractor landscape with a basin of attraction for each expected outcome, including an "I don't know but it might be important" attractor. The size of the basin expresses a Bayesian probability of expected occurrence.  If an expected stimulus is inhaled to the nose, it excites a few among equivalent receptors that send their pulses to the bulb. If the pulses put the bulb in the appropriate basin of attraction, they make the selection is by igniting a Hebbian assembly, which generalizes to the category of the equivalent receptors. This is highly significant. It is the first step from matter to mind. The chemical exists in the world and is quickly expunged. The category exists only in the brain of the subject, and its only role is to activate the memory of the stimulus, its category, what it means for the subject, and what action to take by the subject. We know this, because the subject responds appropriately within a fraction of a second, and we can follow the reverberation of the underlying neural activity through its brain. 

Henry's description and mine are complementary, not in opposition. Their relevance to each other stems from the invocation of QM for explanation. The generalization by the Hebbian assembly is prelude to the activation of the appropriate memory, which requires a phase transition of the bulbar neuropil by condensation from a gas-like low energy state of expecting to a liquid-like high energy state of perceiving. As Vitiello has repeatedly emphasized, the quantum change is in the configuration of the microscopic electric dipoles of the neuropil, mostly water, which comprise and embed the mesoscopic neurons and their organelles including microtubules. The quantum change provides the energy and synchrony. The neurons impose the structure of AM patterns that express the remembrance. We propose that the d activity can be modeled in QFT as a Bose-Einstein condensate using the Ginzberg-Landau equation for superconductivity, and the Rankine thermodynamic engine in classical physics. If Henry can adapt the Schroedinger equation and Born's rule to model the data, we will achieve our goal. The relevance of QM to neurodynamics is to provide new concepts from relevant theory in search of new data.

Concerning qualia and awareness, Henry and his authorities following Descartes and von Neumann in believe that they belong to a category of thought differing from matter, consciousness. To paraphrase Laplace, I do not feel the need for that hypothesis. I believe that a global state of oscillation that synchronizes neural firing over a substantial area of cortex is experienced immediately as t it occurs in the action-perception cycle. The knowledge afforded is instantly mobilized in all sensory and motor systems. My evidence is what I experience; I think, therefore I act. Also I infer it from the survival value of having that knowledge, and the emotions I feel in autonomic preparations for action, within the instant that is available for me to decide to act, or not to act. Natural selection certainly favors the agile over the sluggard.

Henry has taken note of the similarity between "physicalist" and "mentalist" QM in respect to the collapse of a probability distribution, and in respect to the finding that the collapse is triggered by an act of observation that is followed by a phase transition. I note two other similarities. First, our evidence shows that neuropil is bistable; it transits at theta rates between the low and high density phases with narrow band oscillation in the latter. In each phase the dynamics has a small signal range of linearity and stationarity, in which cortex stabilizes itself near criticality. Activity in the low phase at rest is Gaussian; non-Gaussian structure emerges from the noise in the high phase at work. These attributes simplify enormously the modeling with K-sets of the multiple co-existing feedback loops among excitatory and inhibitory populations. I understand that QM rests on comparable simplification.

Second, cortical activity has two readily observable modes: the ionic currents of dendritic integration that are observed in the scalar fields of EEG and evoked potentials, and the clouds of pulses that transmit the integrand that are recorded as point processes. Together with intracellular contributions they constitute the vector field of neural activity. Each causes the other in circular causality. The neurophysiologist decides which mode to observe and designs experiments and interpretations according to that choice. It is important for QM physicists to know this duality, because the prevailing neural paradigm is almost exclusively focused on the pulse mode, and that gives a very limited view of neurodynamics. 

I think that these similarities are not coincidental. The brain is equipped with a tool for exploring the world, which is the action-perception cycle. QM is similarly a tool created by human brains for exploring  the infinite world of extreme complexities of interacting particles of unlimited diversity. The similarities should come as no surprise, as witness the MIT graduate student prayer: “Please, God, make the world linear, time-invariant and Gaussian.” We make our observations by batch processing, with clean up and categorization after each shot. The finding of a comparable wave-pulse duality gives us something really interesting and fundamental to think about. I think the duality is a property of our organ of the episteme. Witness the futility of attempts to compute over the real numbers. I think that we can best progress in our discoveries if we accept both the analog and the digital, the continuum and the staccato, the ODE and the random graph, and fit them together with logic and intuition. That is who and what we are: finite beings facing infinite complexity. In other words, in addition to applying QM to brain function, we should apply brain function to understand QM. The EEG data amassed by my students, colleagues and me at unprecedentedly high spatial and spectral resolution establish a new paradigm by which to proceed. 

Walter

From: Walter J Freeman [mailto:dfreeman@berkeley.edu]
Sent: Saturday, July 26, 2014 9:55 PM
To: Henry Stapp
Cc: Stanley Klein; Kathryn Blackmond Laskey; Karla Galdamez; Giuseppe Vitiello; Sue Pockett
Subject: Re: [MoM] Re: a clarification of the two postings: Stapp-Freeman convergence

Dear Henry,
I interweave my reply below.
On Jul 25, 2014, at 4:48 PM, Henry Stapp wrote:

Dear Walter,

In response to your letter below I say:

In vN-QM (and also its Tomona-Schwinger extension to relativistic QFT) a
mesoscopic  physical system (say a subsystem of the physically described brain) is often represented by a “state” rho that can be characterized as a “quantum statistic mixture”.
Typically it is spread out over a large spatial region in some sense.
Then, in association with a conscious experience, this state rho is replaced by a
new, “reduced” state rho’.
Rho is low-energy a state of expectancy in Step 1 of Stage 3 when the brain holds a landscape of attractors representing predicted possible outcomes of an act of observation. Rho` is the selection by a stimulus of one outcome, hence the “reduction” on the phase transition to a high-energy state in Step 2. I see that global vigorous firing as a necessary condition for C, so I think we agree so far.
[HPS: OK! This is already a key point, so let me spell it out in more detail. The vN approach moves the cut “all the way up” so the mind-matter issue becomes a mind-brain issue: the “collapses” are associated with brain actions. The “state” of a pertinent part of the brain representing a set of possible actions is given in QM by a density matrix rho. Each possibility is a “template for a possible action”. But in W’s physicalist version each is an “attractor”. In both theories the “state” is essentially a set of alternative non- interfering possible brain-directed physical actions. In QM the dynamical rule that generates the evolution of the possibilities/potentialities is the Schroedinger equation.  The key issue is how is one (or a subset of) individual attractor(s) picked out for possible actualization.  Walter says “selection by a stimulus.” Copenhagen says “free choice on the part of the experimenter. “ It comes from something beyond the Schroedinger equation. Going into it  a little more detail, in the quantum case this first choice could be a choice of a particular finite interval along a dial in which the center point of a visible “pointer” might lie. The interval has two end points. So this “choice on the part of the observing agent” involves choices of two point from a continuum of points. This choice has zero probability to occur. In QM this choice constitutes a choice of a “projection operator P. Once the observer’s choice is made , then the probability for the answer to the question “Will I experience the “Yes” response to the question represented by P?” is given by the Born rule Prob = Trace P Rho P/Trace Rho). Of course, this way of expressing the choice is designed to convert the description into a logical question embedded in a mathematical formalism. But what is going on psychologically is a remembrance of a certain past experience (or more accurately a chosen blend of several past experiences) that the person wants/intends to re-experience in the new context. This desired experience  should be generated in association with certain motor action, and the person’s choice triggers the brain into constructing the template for action that, if actualized, should produce in the new context the chosen blend of the prior feed-backs. The Born rule is then implemented by the “random” choice of the part of nature, which actualizes the brain state that will in the “Yes” case have the brain-constructed template for action.  This account gives, however, only the general logical form without specifying how it is implemented in the brain, and it brings in a choice in the part of the observer that is tied to a mental realm assumed to exist in that dualistic vN account, whereas Walter is offering a description of how this is done purely physically by the brain alone! The above vN-based account is an elaboration of what I said earlier, below ]
A key feature of the QM dynamics is that the choice of what this new reduced state rho’ will be, is determined by two choices, the first being a personal choice depending on what the person/organism has learned through its life experience, and which is represented by aspects of reality lying outside the system represented by rho, followed by a random choice depending on both that personal choice and a random element of chance.

I would say that the several areas of cortex sustain knowledge of the world and the self in the form of instructions on how to create a remembrance from memory in the phase transition in Step 2. I have no need for the term “represent” and note the ambiguity of your usage, such that I don’t know what you mean. I would say that the construction is making order out of chaos, which for me has more meaning than “a random element of chance” like a throw of dice.
[HPS: The process of “creating order from chaos” ought, in a physicalist interpretation, to be spelled out as a physically described process. You say that the brain initially has many possible attractors, and then, suddenly, there is only one, with  much higher energy density. That is your way of getting the collapse. And since each attractor is associated with its own blend of remembrances the there will be a different feedback associated with each different attractor, only one of which gets actualized. Attractors normally are found in a representation of a complex dynamical situation in which a point representing a subsystem moves in a multi-dimensional space that is not uniform, but has a structure created by other parts of the full physical system, and that cause the point to move in a non-simple way, such that the point sometimes spends a long time near a point called an attractor. So one way to single out one attractor over others, for a while, is by the moving point representing the subsystem to hover about an attractor. But there is also a possibility that the properties of the space in which the point moves can change so that the set of attractors change. So in a proper physicalist description the phrase “creating order from chaos” needs to be spelled out in physical (non-mentalistic) terms.]
 In the ontological version of vN-QM one contemplates the physically described state of the entire universe, and asserts that the personal choice is not determined by it, but rather by the person’s mental aspect, his “ego”. This is a mentalistic QM!  
If the mental and the physical are “aspects” of an inseparable entity, we agree. If the aspects are conceived as separable, then the problem arise of how each “causes” the other.  It appears in what follows below that you separate them. Hence we disagree.
 [HPS: The universally recognized problem is that science must tie the physically described aspects into human experience, which involves properties such as “red”, “sour”, “feeling of intention”, etc. that are described in a different language: the language of psychology.  Abner Shimony, in his contribution to Penrose’s “The large and the Small…” emphasizes that in situations in which we can rationally justify claiming that two seemingly very different description are different descriptions of the same thing there are two different causal routes connecting the single realty into a single realm of experience. One needs to rationally justify an assertion of identify not merely cite analogies, which are notoriously inexact. ]
But there is an alternative possible QM: a physicalistic QM. In this version the personal choice is determined by features of reality that lie outside rho, but inside the full brain of the person, which carries remembrances of past experiences that have not been totally forgotten, but which have been retained in the brain structure created by the Hebbian process of neurons that fire together wire together.   This physicalist QM seems quite reasonable, and is I believe the version of QM that you are using. So the similarity and difference in our approaches are easy to see: I, following vN and the founders, am using mentalistic QM whereas you are using physicalistic QM. That is what I gleaned from your Vision talk, and tried to convey to the audience.
[HPS: Thus there seems to be a close similarity are the practical level, which ought to be useful to neuroscientists trying to better understand how the unified motor-based characteristics of our thoughts are two-way linked to physical processes in our brains]
I am not yet using QM, but with Giuseppe Vitiello am advocating exploration of the descriptions of brain dynamics in terms of QFT, which he distinguishes from QM. We both appeal to experimental data, I to brain waves, you to the Quantum Zeno Effect. Let’s see what Giuseppe has to say about this difference.  
[HPS: The quantum Zeno effect that I use is a mathematical property of quantum theory embedded in the basic mathematic, independently of any empirical data. It was described and tied to Zeno by Misra and Sudarshan.]
Within this way of understanding your work, and in order to allow me to better understand the details, let me ask whether the ‘sparse coding’ aspect is tied to the “cones”, which I am interpreting as a broadcast of information, much like a radio broadcast, from a site at which information is assembled to receivers located throughout the brain. Is it like AM, or like FM? 
The “sparse  coding” is in expectant state with broad spectrum oscillations and abortive phase cones appearing as mini-avalanches characteristic of critical states. It is not a transmission of information in itself but is the spread by anomalous dispersion of the phase of the neuropil, in this instance condensation from a gas-like phase to a liquid-like phase.
[HPS: This change starts at the vertex of a cone and spreads out. Is this vertex a “cortical column”? Does “sparse” refer to the highly non-dense distribution of possible sites of the vertices of such cones? What does “sparse” mean in this context? What physical process picks out this particular site for the origin of the sudden change. I am interested in your answer to this question because it seems to be the question that vN answers by referring to experiential qualities, not merely physical properties. It is the key question of how does a particular physical get selected from the  many. I am not suggesting that no such physical process could exist, but am only asking whether you have a suggestion for what that physical selection process is, and how it fulfills some desiderata. I agree that a switch to a purely physicalistic account would be highly worth pursuing.]
How is the information encoded in the wave?  And what are these inverse waves, which, as I understand you, are supposed to look like information being sucked into preselected sites.    
The wave packet is composed of a vector field expressed in dendritic current (not the scalar EEG field, which is epiphenomenal) and a cloud of pulses like a swarm of insects, each determining the other in circular causality. The completed remembrance from memory is carried in the coexisting amplitude patterns of waves and the local densities of pulses. Hence the “code” is AM of gamma and beta waves in frame rates at the flickering speeds of thoughts.
[HPS: I remain puzzled by the “inverse”  events which if I understand you correctly is an inverse of spreading out from a vertex of a cone in 4D space-time, but is rather a change of state that occurs over a shrinking 4D cone to a vertex located at a final time rather than expanding from some vertex located at an initial time. If such things really exist, empirically, and are as common as the other, then understanding these seemingly anti-causal things must become a top priority for any theory that claims to be rationally coherent.]
In any case, the sudden collapse into new states that your data reveal seems similar to the quantum collapses. But any illumination you can supply on the similarities and differences would be very helpful. I think there is some yet-to-be-identified major missing piece in my understanding of the physical process behind your data.
I think your missing piece is the macroscopic wave packet (I borrowed this term from QM and gave it a neurobiological meaning. It expresses remembrances, not representations.
[HPS: I am very comfortable with the idea that the macroscopic wave packet expresses remembrances. And what we remember include colors and tastes and the odors associated with appearances of skunks, and our intentions. The question is how are you proposing to reduce these  remembrances to purely physical concepts . I am trying to, with your help, become better informed about that.]

Best, Walter

HPS: So we seen to agree on the functional aspects, but you explain (or seek/claim to explain) the functionality in terms of (classical) physically described brain processes, whereas ‘standard QM’, which speaks of “observers” who probe nature in ways that they “freely choose” (i.e.in ways that are not fully fixed by the physically described aspects of reality, and are not “random”, but are based on human-values)

WJF: I agree. The observer as neurobiologist chooses what aspect to observe. The observed scalar fields of electric potential in cortex show the simultaneous coexistence of neural activity in two modes: fields of dendritic potentials from synaptic currents and clouds of axonal action potentials. Each modulates the other in circular causality. Which we observe depends on what spatial location, temporal segmentation, and spatial filter we choose to decompose it. In principle this choice is based on the human-values we attach to the action potential vs. the synaptic potential. The prevailing model is the neural network. The alternative model espoused by my colleagues and me is the population. The activity is what it is. Our interpretation of what and how cortex is doing is determined by our choice of methods of action. In principle this prior choice characterizes the wave-particle duality in QM. My understanding is that Schrödinger used his cat to lampoon the inference that the reality hinges on the observer, not the cat. Am I misinformed?

HPS: Of course, an adequate theory that does not bring in an unnecessary extra realm is superior to one that can rationally do without it. So let me inquire a bit about your theory.
You stress that the abrupt change that we are talking about is not primarily about a transfer of information, but more like a change of state.
So I am likening it to the formation of the ice on the surface of a pond. The water is at a subfreezing temperature, but it is still water, not ice.
Then, say at a single point, but perhaps a sparse set of isolated points the transition to ice begins, and it spreads rapidly over the whole pond.

WJF: I use the analogy of a gas-to-liquid transition, as in formation of a raindrop from a site of nucleation like a grain of dust, from which there is radial expansion. Therefore I expected to see explosion, i.e., negative spatial phase gradients in radians/meters. The ratio of  spatial gradient to temporal frequency in radians/second determined the phase velocity in radians/second. The ratio was a constant equal to the conduction velocity of intracortical axons. What I found was about half of the gradients were implosive, positive. Your analogy of evaporation of the raindrop is apt, but not explicable in terms of neurophysiology.. I rest my case that the finding should be explored by physicists using QFT. Vitiello has shown that this phenomenon is compatible with QM, using the Ginzberg-Landau equation to describe the change of state. Perhaps you can use the Schrödinger equation to describe it.

HPS: Now we normally do not think that the pond feels itself turning to ice, or that a water droplet consciously feels itself changing from gas to liquid. But, of course, no one can know what a pond or a water droplet is feeling, and a brain is far more complex than either a drop or a pond of water. So the empirical evidence is not compelling.

WJF: There is no basis for the introduction here of qualia. The phase cone is not a NCC and does not address consciousness. I am describing the physics of how a condensate forms, not what the NCC of the following wave packet might be, or how to “attach” it to neurons.

HPS: But if “physical” means described in terms of the concepts of Classical (Newtonian-type) physics, then there is a causal gap, because those concepts were designed to not entail conscious feelings; the moving particles of Democritus do not logically or rationally entail our streams of conscious experiences, filled as they are with odors, colors, and mental intentions. One can try to assert that the NCC’s are identical to the feelings. One can assert that the motions of the particles ARE the conscious feelings, but the analogies advanced to defend that opinion are very far from obvious, and have been attacked by very intelligent philosophers. It does not suffice to simply doggedly assert that there is no hard problem: the motions of the Newtonian particles are just too obviously different from our feeling for the two to be allowed be identified without a strong defense. The attempt to eliminate the mental realm must be reconciled with the fact that the empirical evidence upon which science is based lies in the mental realm. So the issue here is not some abstruse academic point but the heart of the problem.

WJF: That is the philosophers’ problem, but it is not the problem I have come to you for. As you say, “That issue has been endlessly debated,“ and with no resolution in sight.  Instead, I ask you to apply your knowledge of QM to my new data, and tell me, are they consistent with QM?

HPS: But the *startling new evidence* is what I understood (perhaps very incorrectly) from your words to be cases in which the change of state starts out at the rim and advances inward to a final point.  Is that correct? And is this change like freezing or a melting? Is that what these inward directed changes are: like the melting of an ice cube, or the evaporation of a water droplet inward from the outside?

WJF: Yes, except that the raindrop hasn’t formed yet, and when it forms it starts from the outside in. For the raindrop, that is nonsense. For cortex it simply reflects the fact that the attractor pre-exists its access and realization, so either way will suffice. What is important is the finite velocity of the cortical phase transition. I think it manifests something like a saddle node bifurcation in which the direction of convergence to the attractor is a matter of chance, which you introduce as a crucial ingredient of choice. This is my intuition as a neurobiologist; you are the mathematician. 

HPS: Of course, even if one does need to bring in our minds, simply because they are the foundation of science, your understanding of the brain dynamics could constitute a giant step foreword the quantum mechanical understanding of the mind-brain connection.

WJF: I agree, because the ‘paradoxes’ of QM reflect more the properties of the human brain than the properties of the universe, so let’s get on with the science and leave this “hard problem” to those who created it. Please attend to my neural versions of the wave-pulse duality. For single neurons it is trivial; for populations it is profound, but it is comprehensible, not a mystery or a paradox.

Equally cordially, Walter
Cordially,
Henry



No comments:

Post a Comment