Traditional models of brain function, which attempt to provide a dynamical account of the specifically conscious aspects of memory invoke global criteria that cannot be admitted to a causal picture in a classical context. But without such global criteria, features of consciousness like the serial character of conscious recall and the apparent lack of neurophysiological modularity in certain kinds of memory cannot be adequately explained. These features find natural explanation in an account of memory that allows classical mechanisms to be supplemented with one operating on quantum theoretical principles. Such models critically depend on the stability of the quantum mechanism and the nature and efficacy of the interface with classical mechanisms. Such an interface is possible in a theory of macroscopic quantum ordered states in which 1) quantum signals are provided with the means to influence meso-scale neural function and 2) the discriminated information inherent in networks of neurons can be usefully translated into a quantum encoding. Stability in a vacuum encoding can be generically ensured and the conditions for its establishment can be met under the stresses of a biological environment. The vacuum parameter coding memory in the quantum system can be determined in terms of physical parameters, forming the basis of a common language for the quantum and classical systems of memory and allowing information flow from one system to the other.

Why Look Beyond Classical Mechanisms? 

Traditional neural network models of brain function have proven to be quite successful in simulating numerous aspects of unconscious cognitive processing so one might ask why we need look any further when it comes to specifically conscious aspects of memory. Indeed if consciousness is understood along emergentist lines, there seem to be any number of proposals to account for its various features. In particular the apparently nonlocal character of memory storage involved in some forms of memory and the predominantly serial manner in which the recall process makes experiential memory available to consciousness have received numerous treatments in a classical framework. Why should these accounts be found wanting?

Traditional models of brain function, which attempt to provide a dynamical account of the specifically conscious aspects of memory invoke global criteria that cannot be admitted to a causal picture in a classical context. But without such global criteria, features of consciousness like the serial character of conscious recall and the apparent lack of neurophysiological modularity in certain kinds of memory cannot be adequately explained. These features find natural explanation in an account of memory that allows classical mechanisms to be supplemented with one operating on quantum theoretical principles. Such models critically depend on the stability of the quantum mechanism and the nature and efficacy of the interface with classical mechanisms. Such an interface is possible in a theory of macroscopic quantum ordered states in which 1) quantum signals are provided with the means to influence meso-scale neural function and 2) the discriminated information inherent in networks of neurons can be usefully translated into a quantum encoding. Stability in a vacuum encoding can be generically ensured and the conditions for its establishment can be met under the stresses of a biological environment. The vacuum parameter coding memory in the quantum system can be determined in terms of physical parameters, forming the basis of a common language for the quantum and classical systems of memory and allowing information flow from one system to the other.

Why Look Beyond Classical Mechanisms?

Traditional neural network models of brain function have proven to be quite successful in simulating numerous aspects of unconscious cognitive processing so one might ask why we need look any further when it comes to specifically conscious aspects of memory. Indeed if consciousness is understood along emergentist lines, there seem to be any number of proposals to account for its various features. In particular the apparently nonlocal character of memory storage involved in some forms of memory and the predominantly serial manner in which the recall process makes experiential memory available to consciousness have received numerous treatments in a classical framework. Why should these accounts be found wanting?

While the character of connectionist explanation is massively parallel and distributed, the character of conscious recall is roughly serial. Consciousness distinguishes certain memories as active, and this distinction passes from one memory to another. Neural networks on the other hand do not, by their very nature, distinguish an active stream of information. The system is disposed to produce particular output given particular input and "memories" are actuated only through this dispositional process. It becomes problematic then to explain why conscious recall of "memories" thus conceived should assume features like serial expression, determined within the system, that cannot simply be read out from a consideration of this output.

The emergentist thesis, that neural systems will give rise, at some threshold level of complexity, to ontological features distinct from behavioral dispositions, suffers from the lack of any criterion by which to decide exactly what will become the property of conscious memory, when the mechanism of consciousness should become operable, or why it should be a necessary consequence of the attendant conditions. Some models have attempted to provide answers to these questions by introducing dynamical criteria. Accounts variously identify the contents of consciousness with "vector activations" "concentric epicenters" or "attractors". Inevitably however such criteria involve global discriminations. Typically the global nature of the criteria can be disguised by implementing a description in terms of phase space, a higher dimensional fictitious space that uses the entire collection of local variables as coordinates. This strategy is often convenient in interpreting the dynamical behavior of systems, but in a classical context such a device cannot be considered causally efficacious. Classical dynamics are determined locally at each point in the system. While global characteristics might handily describe the dynamics from an external point of view, they play no role in determining the intrinsic dynamics. This point will resurface presently in the context of a potentially related problem having to do with the modular character of connectionist explanation.

Some forms of memory have stubbornly resisted explanation by localized storage and recall mechanisms. Much of the success of connectionism in elucidating the workings of the brain has been obtained by modularizing functions and assigning these to localized regions of the brain (see for instance Fodor and Pylyshyn 1988). This methodology was motivated by a recognition of the fact that the discriminatory power of neural networks with respect to input characteristics is read out only in the output since this is the only place where the distributed machinery of the network can be brought to bear. Modularization then allows the intermediate functions of localized areas of the brain to be interpreted as meaningful. While this course has, on most reckonings, successfully accounted for many forms of memory involved in unconscious processing, the kind (or kinds) of memory most directly involved in conscious recall do not appear to be neurophysiologically modular.

The absence of modularization points, in the framework of connectionist thought, to a representation that remains distributed until final output is achieved. Intrinsically-from within the system itself-distributed representation then allows meaningful interpretation only in the final output. Interpretations of distributed processing are, of course, possible but are inevitably made from outside the system. They require the acknowledgement of relations, not explicitly represented in the system, that exist between informational elements located at disparate physical locations and therefore constitute an extrinsic mode of interpretation. Conscious representation however must require an intrinsic mode, since the interpretations of content provided by consciousness are not made from outside the system but rather are provided by the system itself. An explanation of consciousness made in an intrinsic mode would then seem to require the accessibility to and simultaneous availability of the same explicit relationships that facilitate interpretation in the extrinsic mode.

On a classical account, the fulfilment of this requirement is expressly forbidden by the localityconstraint. Locality 1 constrains the speed of propagation of signals to be less than the speed of light and applies indiscriminately to all classical systems. Its relevance on the scale of brain dynamics stems from the immediate corollary that communication over physical distances, however small, must not occur instantaneously. The determination of relationships existing between informational elements that are physically separated however requires such communication if the relations are to be made explicit, simultaneous with the informational elements themselves. One solution to this apparent impasse is to require that the desired relations be expressed locally but such a solution must contend with thermal limits on the possible complexity of a local system, limits far too stringent to allow a realistic description of even very limited conscious processes. Moreover there appears to be no evidence that this is the route pursued in the brain.

What is required then is the means to extend an intrinsic mode of description, in which the elements of the system merely respond in accordance with the principle of locality to the flux of physical conditions in the immediate vicinity, to an extrinsic mode, in which the system itself is capable of providing an interpretation of these activities. Moreover this requirement must apply at every level at which conscious apprehension "binds" information that is not simultaneously available in a local representation. Reinterpreted with this stipulation in mind, the "binding problem" is encountered at a much finer level of psychological structure than is admitted in the traditional reading of the problem. Further, the problem appears not to find an adequate solution in a classical context where the locality constraint is always enforced. Consider for instance the proposal that temporal synchrony in neural firing offers an explanation of binding in visual processing. Since each of the synchronized neurons is firing in response to its own local dynamical situation, oblivious to the fact that other neurons are firing concurrently, temporal synchrony is recognizedonly as a global property, only from outside the system, and therefore must be without causal consequence in the classical context. The encoding becomes local only in the downstream neural dynamics, either through convergence or by the exchange of signals, but at this level there is no longer any reason to anticipate ontological features distinct from connectionist dispositions. Although it originates consequences at the level of neural dynamics, temporal synchrony would seem to be more naturally interpreted as an effect of an underlying mechanism that would explain how these neurons are orchestrated to fire simultaneously and it is presumably this mechanism that encodes for psychological binding. But to be efficacious the mechanism must neither encapsulate information that is available only from an outside perspective-that is, the information must inhere in the system itself-nor reduce to local dynamics with only trivial dispositional consequences. On a classical reading however, these would seem to be the only options.

Quantum Memory in a Subcellular Context

Some of these considerations motivated the suggestion that the success of connectionism in explaining many facets of the working of the brain could not be carried over to all aspects of brain function, and in particular that the character of experiential memory would not find adequate explanation within that framework alone. The necessity of a quantum component to brain function has been suggested, for independent reasons, by several authors . It should be noted that these arguments do not, as some earlier theses did, usethe mechanism of consciousness to patch up long-standing interpretational difficulties in quantum mechanics, but rather employ physical and mathematical concepts to critique the scope of traditional computationalism to explain consciousness and in particular to suggest that no purelyclassical account can be adequate to the requirements of a realistic theory of conscious processes.

Quantum systems need not be subject to the same locality constraints that always apply in the classical realm and, in particular, macroscopic quantum states allow information about global properties of the system to inhere in the system itself. The proposal of a quantum basis for memory might however introduce potential difficulties in accounting for the long-term stability of memory. And the claim that two separate but interacting systems are, taken together, responsible for determining the nature of memory necessitates a bi-directional flow of information between the systems.

The theory situates a quantum component of memory subcellularly in neurons and allows individual memories to be stored in a stable fashion by carrying them into the ground state or vacuum of the system dynamics. This demands that there be a vast multiplicity of different vacuum states to code for different memories. In the quantum mechanics of old this was expressly forbidden by the equivalence theorem, which declared that all vacua were equivalent. Quantum mechanics precluded the possibility of describing transitions from one vacuum phase to another phase exhibiting radically different behavior. The advent of quantum field theory allowed for the first time a realistic treatment of inhomogeneous media exhibiting domains of differing vacuum phase. To enforce the choice of a particular vacuum throughout a coherence domain, long-range correlations must be established. A well-known theorem, the Nambu-Goldstone theorem guarantees that in systems exhibiting spontaneously broken symmetries-systems in which the dynamics are invariant under a transformation that is not respected by the vacuum state-gapless modes are spontaneously excited to enforce the choice of vacuum, thereby establishing long-range correlations. These modes, familiar phenomena in condensed matter and high energy physics, arise spontaneously since no energy barrier need be surmounted.

Polariton Basis

In a standard two-state approximation to the energy level structure of the water molecules in the vicinity of cytoskeletal microtubules, two kinds of vacua arise when the system coupling the electromagnetic field and the water dipole field is diagonalized in a polariton 2 basis. One is the standard vacuum and has the same invariance properties as the system dynamics; the other is a set of degenerate vacua of spontaneous symmetry breaking type, labeled by a continuous parameter.The gapless excitations that arise in the spontaneously broken phase maintain ordered states that manifest long-range correlations.

The spontaneous symmetry breaking vacua are accessible only in a range of physical parameters relating the energy level difference in the two-state description of water to the number density of participating molecules. The polariton model thus determines the values of the vacuum parameter in terms of classically accessible physical quantities. This relation to physical data is essential to allow discriminated information inherent in classical information processing systems to find useful transcription into a quantum encoding and vice versa, essentially providing the basis of a common language for the two systems.

For energy transitions, in the two-state treatment of water molecules, occurring in the far infrared (1012 Hz) one finds a parameter space, that easily accommodates a reasonable range of values for the density of the coherent fraction, even leaving considerable room for large-scale biological impurities.

The dispersion relations for the low-lying excitations in the theory are found by linearizing the system with respect to fluctuations about the spontaneously broken vacuum. These dispersion relations determine the energy spectrum in the polariton basis. The relations describe two different modes. One is a gapless mode with a minimum energy of zero. In the Umezawa model for memory, this corresponds to the mode that maintains the stability of memory. The other mode, an effectively massive mode, has a nonzero minimum energy and corresponds to the mode of memory recall. The ensemble of modes of differing vacuum codings that are concurrently active determine the serial stream of conscious recall in a manner consistent with our earlier considerations. Also, because the minimum energy depends on the value of the vacuum parameter coding for a specific memory, some of these modes will require more energy to be excited; that is, some memories will be "harder" to recall.

The dynamics of the interaction between the electromagnetic field and the water dipole field allow for a phenomenon called superradianceto occur in the coherent phase.The phenomenon is related to spontaneous emission in lasers and allows quantum effects to be propagated up to macroscopic levels in systems involving a large number of participating molecular sites. In neurons, such effects might allow the quantum dynamics to exert subtle control over cellular activities, with microtubules acting in the role of a subcellular optical network.

It would seem that information flow in the other direction is necessitatedby the neural system itself. It is in fact the microtubular network that orchestrates changes to the synaptic "weights" in a connectionist description of neural function. Neurotransmitter vesicles (or their precursors) are conveyed to release along microtubules and other cytoskeletal structures. This suggests that the information needed for the orchestration of these activities at least momentarily inheres in parts of the cytoskeletal network and thereby might be carried into a quantum encoding.

Tuszynski have also pointed to the possibility of classical information processing in the tubulin ensemble composing the microtubule. The system of tubulin exhibits a spin-glass phase, a description often used to describe information processing in neural networks, and the transition to this phase occurs at biological temperatures. Potentially vast increases in processing power make microtubular involvement an attractive evolutionary option that might later have been adapted to make use of the possibilities offered by the quantum dynamics.

Phase Transitions

In order to take advantage of these properties of the spontaneous symmetry breaking vacuum, a phase transition can be effected by means of a polariton condensation into the ordinary vacuum. It has been demonstrated that such a condensation is an indication that one has moved into an inequivalent Fock space built on a new vacuum. The new vacuum is moreover a coherent state of the gapless polariton.The means by which it is currently possible to elicit a phase transition to a coherent state can be classified into one of two categories. The first encompasses the well-known method employed, for instance, in attaining superfluid states: Bose-Einstein condensation. Here the parameter that controls the phase transition is generally temperature 3. Below a critical value, the coherent phase appears.

In the most familiar applications, the critical temperature is often extremely low in comparison to biological temperatures but it is misleading to generalize from these cases. In particular first order calculations depend on particle density and condensate fraction (the fraction of the molecular sites participating in the coherent state) and can easily, without extreme assumptions, assume values greatly in excess of room temperature. Critical temperatures in the theory of magnetism for instance vary considerably, according to the composition of the sample and the nature of the symmetry breaking, ranging from slightly above absolute zero to many times higher than room temperature. Since all temperature dependence is subsumed in the determination of the critical temperature, a discussion of thermal noise would therefore be spurious.

The second category of phase transitions is controlled by an external supply of energy that must be constantly supplied above a threshold value to maintain coherence. This is the kind of phase transition occurring, for instance, in the Frцhlich mechanism (Frцhlich 1968) but is much more familiar as the process governing laser phenomena. Lasers are indicative of the way that coherent states can be maintained by providing a continuous energy supply, a method that has not only been intensively studied from a theoretical point of view but also has been demonstrated experimentally so often as to be commonplace.

In this scheme the external energy provided to the microtubular system might be metabolic energy travelling along the microtubular network. Del Guidice et al. (1985) cite evidence that the water surrounding the cytoskeleton can be brought into an electrically polarized state by very low amplitude, low frequency electric fields when provided with a threshold energy of approximately the same magnitude as is carried by the solitonic waves travelling along the microtubules. Seen in this context, biological systems become hospitable hosts for coherent states: the continuous provision of energy is ubiquitous in the biological world.

Conclusion

It has been suggested here that classical models of memory cannot alone account for the seriality and nonlocality of conscious aspects of memory. These invariably invoke extrinsic elements-descriptions that do not inhere in the system but are applied from the outside-elements that are expressly forbidden in causal accounts given at the classical level. The establishment of a collateral system of memory based on quantum dynamics allows such elements to be assimilated to an intrinsic description without contravening the locality constraint. The thrust of this work has been to provide a framework in which to relate the vacuum parameters that act as codes in a quantum field theoretic treatment of memory to physical parameters. This lays the foundation for a system of communication between classical and quantum levels of encoding, and allows the dual system of memory to encapsulate characteristic features of both modes of operation. Microtubules, perhaps in conjunction with the membrane dynamics of the dendritic network, appear to be most strategically placed to effect the transfer of information between the two systems, and in the polariton basis, the physical variable implicated in mediating this transfer is the number density of dipoles in the coherent phase.

---

Be first to comment this article

Only registered users can write comments.
Please login or register.

Powered by AkoComment Tweaked Special Edition v.1.4.6
AkoComment © Copyright 2004 by Arthur Konze - www.mamboportal.com
All right reserved