| sales@yourdomain.com Tel: 01234 567890 |
|
|
Penrose & Hameroff 2Penrose & Hameroff: 2 1.) Shadows of the Mind - Roger Penrose 2.) Ultimate Computing - Stuart Hameroff 3.) The Large, the Small and the Human Mind - Roger Penrose & Others 4.) Anesthesia & Consciousness 5.) Brainshy - Patricia Churchland 5.) More Neural Than Thou - Hameroff reply to Patricia Churchland SHADOWS OF THE MIND Roger Penrose Oxford University Press: 1994 ISBN 0 9 853978 9 ‘Shadows of the Mind’ is Roger Penrose's second book on consciousness. The early part of the book is largely taken up by his response to the numerous criticism of his first consciousness book, 'The 'Emperor's New Mind' (ENM). However, the late chapters of the book introduce new concepts bring us to the fully developed Orch OR model of quantum consciousness. This involves, firstly, a proposal for a new version of the collapse of the wave function, which links it to fundamental spacetime geometry, and, secondly, suggestions as to how such a process could be physically instantiated in the brain. Responses to Criticisms of 'The Emperor's New Mind' Penrose discusses at length some of the criticisms aimed at his first book. In this, he had argued that Gödel's theorem showed that human mathematicians used some form of understanding that was not based on any algorithm, when they were able to grasp the truth of statements that could not be proved using the axioms of a formal statement. His critics had suggested that while mathematicians could do this, they were in fact using a knowable algorithm present in their brains when they did it. Penrose contests this, arguing that all possible algorithms are defeated by the Gödel problem. In respect to arguments as to whether computers could be programmed to deal with Godel propositions, Penrose accepts that a computer could be instructed as to the non-stopping property of Turing's halting problem. Here, a proposition that goes beyond the original axioms of the system is put into a computation. However, this proposition is not part of the orginal formal system, but instead relies on the computer being fed with human insights so as to break out of the difficulty. So the situation becomes circular, with the apparently non-algorithmic insights required to supplement the functioning of the computer. Penrose further discusses the suggestion of an unknowable algorithm that enables mathematicians to perceive the truth of statements. He argues that there is no escape in any kind of computer or human development from the knowability of algorithms. An unknowable algorithm means an algorithm, whose specification could not be achieved. But any algorithm is in principle knowable, because it depends on the natural numbers, which are knowable, and it is possible to specify natural numbers that are larger than any number needed to specify the algorithmic action of an organism such as a human and a human brain. Penrose says that with a mathematical robot, it would not be practical to encode all the possible insights of mathematicians. The robot would have to learn certain truths by studying the environment, which in its turn is assumed to be based on algorithms. But to be a creative mathematician, the robot will need a concept of unassailable truth, that a concept that some things are obviously true. This involves the mathematical robot having to perceive that a formal system 'H' implies the truth of its Gödel proposition, and at the same time perceiving that the Gödel proposition cannot be proved by the formal system 'H'. It would perceive that the truth of the proposition follows from the soundness of the formal system, but the fact that the proposition cannot be proved by the axioms also derives from the formal system. This would involve a contradiction for the robot, since it would have to believe something outside the formal system that encapsulated its beliefs. The Wave Function and Space-time Geometry In the later chapters, Penrose develops the idea of a form of wave function collapse not involving the conventional concepts of measurement or interaction with the environment. He is discussing the situation where quanta remain isolated from their environment, so that there is nothing in the conventional theory to bring about a wave function collapse. He suggests that as a result of the evolution of the Schrodinger wave, the superpositions of the quanta grow further apart. According to Penrose's interpretation of general relativity, each superposition of the quanta is conceived to have its own spacetime geometry. When the superpositions reach a point where they are separated by the Planck length (10^-35 metres) they become unstable, and the wave function collapses, giving a choice of one or other of the possible space-time geometries for the particle. This form of wave function collapse is proposed to exist in addition to the more convventional forms of collapse in Non-computability, Space-time Geometry & Closed Timelike Lines The significance of this for the study of consciousness is that, in contrast to the conventional idea of wave function collapse, this form of collapse is suggested to be non-random, and instead driven by a non-computable function at the most fundamental level of spacetime. Penrose suggests that there is a non-computational aspect to the structure of spacetime itself. In support of this, he points out that when the physicists Geroch and Hartle studied quantum gravity, they found that there was no algorithm for solving certain problems involving superpositions of four dimensional space-time. Earlier the mathematician A. Markov had shown there was no algorithm for such a problem, and that if such an algorithm did exist, it could solve the Turing halting, for which it had already been shown that there was no algorithm. The possibly non-computable nature of the structure of four-dimensional space-time is deemed to open up the possibility that wave function collapses in the brain could give the mind access to this non-computable feature of fundamental space-time. Penrose also looks at the possibility that so-called closed timelike lines could provide a basis for non-computable processing. This idea has also been developed by David Deutsch. Kurt Gödel (separately from his famous theorem) had been the first to propose that general relativity theoretically allowed closed timelike lines. A closed timelike line could arise around a very massive object, if the continous tilting of a light cone in line with a very strong gravitational field meant that an individual, whose world-line lay within the light cone eventually came back to the point where they started, meaning in this case the point in time, i.e. a point in their past. Penrose thinks that the potential for paradox would prevent this happening in the macroscopic world, but that it would be feasible for it to happen at the quantum level. This area is speculative even by the standards of consciousness theory, but the suggestion appears to be that a quantum computing device circling on a closed timelike line could carry out repeated calculations which it could bring back to a computer lieing in its own past, which might derive some 'insights' from the repeated future calculations. Quantum Activity in the Brain In the later stages of the book, Penrose begins to look at how quantum activity could be instantiated in the brain. The idea that there might be quantum activity in biological tissue was poineered by the phycisist, Fröhlich in the mid 20th century, who argued for quantum coherent activity in biological tissues as a result of metabolic energy. He and others have argued that some form of quantum ordering was required to maintain the orderly sequence of chemical reactions necessary to living organisms. Metabolic energy and dielectric properties are argued to be sufficiently pronounced quantum coherence to arise, and these could come in the macroscopic form of Bose-Einstein condensates. There is some evidence now for the oscillations in tissues predicted by Fröhlich. Penrose considers microtubules as a plausible site for the control of sophisticated activity in organisms. Single cell creatures are seen to navigate and perform other functions without the aid of a brain or a nervous system. The cytoskeleton is believed to be responsible for their capabilities, and microtubules are the most important component of the cytoskeleton. The cytoskeleton can be viewed as analogous to the cell's own nervous system. Penrose discusses the structure of the microtubules. Each microtubule is former out of subunits of the protein tubulin. Each subunit is a dimer of an alpha and beta tubulin. The microtubule is a long tubular structure, and the tubulins are organised into an hexagonal lattice. Each tubulin dimer has two different configurations or conformations. The switching of configurations is thought to correspond to different electric polarisations of the dimer, resulting from the movement of an electron positioned at the junction of the alpha and beta tubulins. Stuart Hameroff, who cooperated with Penrose on the development of the Orch OR theory that emerges in 'Shadows in the Mind', had suggested that signals could be passed along the microtubules as waves of polarisation of the tubulin dimers. Each dimer would be influenced by the polarisation of six of its neighbours, giving rise to effective rules for the conformation of the tubulins, which in turn makes them suitable for the transmission of signals. The microtubules are closely related to the synapses, carrying neurotransmitter molecules to them, and possibly altering their strength as a result. They are also involved in the growth of nerve endings and guiding them towards other cells. They are indirectly connected to dendritic spines, the growth and degeneration of which is an important factor in the plasticity of the brain. Penrose thinks that these structures point to much greater computing power in the brain than had previosuly been realised if as he thinks the tubulin dimers are computational units. However, Penrose is not satisfied with the proposition of increased computing power. The whole argument from the Godel theorem says in effect that no matter how great the computing power, you will get no closer to solving the problem of consciousness. Penrose holds to the argument that the brain must manifest some non-computational property, and that anything that is non-computational will be the result of quantum coherent activity, because this is the only aspect of the universe in which non-computational activity can arise. Because of their influence over the activity of cells and because of their structural suitability for signalling and information processing, microtubules are considered the most likley site for such non-computational activity. Penrose discusses both the plausibility of quantum activity in microtubules, and the type of quantum activity that might be useful for the production of consciousness. This is against the background of the fact that quantum states would normally be expected to decohere in the environment of the brain. Frohlich had argued that metabolic energy and the dieletric properties of biological tissue would be sufficient to maintain quantum coherence at body temperatures. However, the Penrose/Hameroff model places more emphasis on the idea that the microtubules themselves could shield quantum coherence. This is suggested to arise mainly from the ordering of water close to the surface of the microtubules. The structure of the microtubules is also suggested to be favourable for the formation of macroscopic quantum coherence known as Bose-Einstein condensates. Penrose further suggests that the quantum coherent state could extend between microtubules and also between neurons, thus creating a macroscopic quantum structure across a significant area of the brain. It is further suggested that when this macroscopic quantum state undergoes wave function collapse, this will involve the objective reduction (OR), which would connect the microtubules to the non-computational aspect of fundamental spacetime. Numerous articles, mainly published after 'Shadows in the Mind' go into the proposed physical structures in the brain in greater detail, and some of these are reviewed under Penrose & Hameroff sections 3-6. ULTIMATE COMPUTING Stuart Hameroff Elsevier Science Publishers BV 1987 ISBN 0 444 70283 0 The main substance of this book deals with the scope for information processing in biological tissue and especially in the microtubules and other parts of the cytoskeleton. The question of consciousness as distinct from information processing is not given a central role, except when Hameroff discusses the effects of anaesthetics. After the book was written, Hameroff read Penrose's first consciousness book, 'The Emperor's New Mind', met Penrose and suggested to him the possibility that the microtubles might be the seat of consciousness. The Orch OR theory of quantum consciousness arose from this cooperation. Early on in his career, when Hameroff had worked in cancer research, he had realised that cell division was orchestrated by microtubules. When they were first discovered, the cytoskeleton and its most important components, the microtubules, were seen as a support structure for the cell, but later it became apparent that they had wider functions, and were involved in cell growth and splitting and transport of molecules within the cell. They also seemed to be responsible for dynamic organisation within the cell. This implied that the cytoskeleton used some form of computing. Hameroff suggests that sub-units of biological protein, such as the cytoskeleton and organelles could provide the basis for information processing. These sub-units undergo nanoscale conformational oscillations that may be both coherent and coupled to dipole shifts within proteins. It is suggested that this could provide a basis for information processing. Hameroff posits that neurons may not be the basic unit of consciousness or intelligence, but that this could instead lie inside the neurons. He points out that single cell organisms have no nervous system but can perform complicated tasks, which can only be achieved by means of some form of internal processing. He surmises that the same form of processing could exist in cells, including brain cells, within multi-cell organisms such as humans. Hameroff views the functioning of protein as a form of analogue computer. Proteins oscillate between conformational states over periods varying from 10-15 seconds to minutes or longer. In the mid 20th century, the physicist Fröhlich thought that biochemical energy supplied to biomolecular assemblies could result in coherent vibrations. He suggested that a set of proteins on a common voltage gradient such as a membrane or cytoskeleton would oscillate coherently if ATP energy was supplied. This could explain long range cooperative effects, by which proteins and nucleic acids communicate.Protein conformation is a response to a complicated input of temperature, pH, ions, voltage, dipoles etc. Hameroff views each neuron as a computer, and thinks that the cytoskeleton is ideally suited for information processing. Cytoplasm is neither totally regular nor totally random, and is therefore ideal for transmitting information. There is some evidence for cytoskeletal involvement in cognitive processes. Microtubules and tubulins are increased in the brain during periods of learning, memory and experience. Tubulin production is related to learning/training etc in young chickens and rats. It is also suggested that dendritic spines change shape to alter synaptic thresholds and that this is orchestrated by microtubules. The Large, the Small and the Human Mind Roger Penrose with Abner Shimony, Nancy Cartwright & Stephen Hawking Cambridge University Press 1997 ISBN 0 521 56330 5 In the first part of this book Penrose restates the arguments previously discussed in ‘The Emperor’s New Mind’ and ‘Shadows of the Mind.’ This isn’t new ground, but it has the merit of presenting his ideas in a shorter and simpler form. In the latter part of the book, the philosophers Abner Shimony and Nancy Cartwright and the physicist, Stephen Hawking, respond to Penrose’s ideas, and Penrose in turn responds to them. Shimony Shimony’s main objection to Penrose's ideas is that it is unnecessary to drag in the whole question of Gödel and non-computability, in order to refute mainstream ideas about the nature of consciousness. In reply, Penrose agrees that there is a strong argument against the present mainstream theories even without Gödel, but points out that his approach does not merely demolish mainstream materialism, but also tries to provide something to put in its place. Nancy Wright The discussion with Nancy Wright appears to revolve round whether physics or biology should provide the route towards an understanding of consciousness and the ability to provide a programme of experimentation. Penrose’s reply is that the shortcomings that he sees in existing physics are relevant to the state of biology. Stephen Hawking The discussion between Penrose and Hawking reveals a fundamental divergence in their attitude towards the underlying state of the universe. Hawking is closer to the traditional Copenhagen view, in which the quantum level is merely a system for calculations, whereas Penrose views the quantum as an underlying reality. They also disagree as to whether there is a problem with the continuous evolution of the Schrödinger wave. Penrose takes the view that the superposition of progressively larger objects, which is implied by the Copenhagen approach, takes us further and further away from the world as it is actually experienced, and that his theory of objective reduction provides a solution to this problem. In respect of consciousness, Hawking appears to duck the issue, choosing instead to talk about intelligence. Anesthesia & Consciousness Stuart Hameroff Anesthesiology, 2006, 105, pp. 400-12 The article starts by pointing out that the precise mechanism by which anesthesia works is unclear, simply because the precise nature of consciousness is also unknown. Anesthesia is seen as being driven by London forces, the weakest form of van der Waals force, acting in hydrophobic pockets in protein. In this respect anesthetic gases are seen to differ from other drugs, because their action is at the quantum level, while other drugs act at the chemical level. Studies have shown the binding of anesthetic molecules can be enhanced near to hydrophobic sites. In particular, Franks and Lieb in the 1980s demonstrated anesthetic action in hydrophobic pockets, and the preponderance of evidence points to most of the anesthetic action being in such regions. Only 15% of proteins have hydrophobic pockets large enough for anesthetic molecules. This may account for why many brain functions such as autonomic drives and evoked potentials are not closed down by anesthetics. The van der Waal forces involved in the action of anesthetics depends on dipole couplings between atoms or molecules. There are three versions of these forces involving attraction between permanent dipoles, attraction between a permanent dipole and electrons capable of being polarised, and the third type known as the London force, which acts between two normally neutral but polarisable atoms or molecules, with temporary dipoles being created. London forces are sensitive to the distance between electron clouds, and the forces are very weak, but acting collectively they can become strong enough to control the conformation of protein. Proteins perform their functions in the body and brain by changing shape and conformation, involving switching between energy minima. Proteins are linear chains of amino-acids, which fold into three dimensional conformations driven by hydrophobic amino acid-groups. Some of these form hydrophobic pockets in which London forces are able to influence the conformation of the protein. Anesthetics are known to bind not only in the membrane, but also in a number of locations within the neuron, including the tubulin of microtubules. During the 1960s and 1970s the biophysicist, Herbert Fröhlich, proposed that fluctuating dipoles in proteins in the cell membrane or cytoskeleton would synchronously couple, and being pumped by metabolic energy, the proteins would oscillate in a pumped Bose-Einstein condensate. Hameroff mentions some more recent evidence supportive of some form of biological oscillation. Anesthesia produces immobility, amnesia and loss of conscious awareness. Research in recent years has suggested possible sites for all three of these functions, with the spine as a favoured site for immobility, the dorsolateral prefrontal for amnesia, and thalamocortical and intracortical networks for consciousness. A study by John and Prichep showed that loss of consciousness under anesthesia occurred over only 20 ms and involved interruption of the gamma synchrony between the frontal and posterior cortex. Hameroff regards the gamma synchrony as the best established neural correlate of consciousness (NCC). Experiments have shown that the gamma synchrony involves synchronised voltage fluctuations in various regions of the cortex and thalamus. The gamma synchrony is related to dendrite-to-dendrite gap junctions, influences by the dendrite cytoskeleton, rather than axonal synapses. Studies have shown that anesthetics effect dendrites and gamma synchrony more than axons and neurotransmitter release. Hameroff speculates that precise synchrony requires some form of quantum field, since normal brain signalling, even via gap junctions, involves delay. Gamma synchrony is also seen as a candidate to enable the binding process, by which the varied contents of consciousness are conceived as a unity. The gamma synchrony is related to dendrite-to-dendrite gap junctions, influences by the dendrite cytoskeleton, rather than axonal synapses. Studies have shown that anesthetics effect dendrites and gamma synchrony more than axons and neurotransmitter release. Brainshy: Non-Neural Theories of Conscious Experience Patricia Smith Churchland In: Towards a Science of Consciousness II: The 1996 Tucson Discussions and Debates: Eds Stuart Hameroff, Alfred Kaszniak, Alwyn Scott MIT Press 1998 In this paper Churchland seeks to refute the consciousness approaches of Chalmers and Penrose. With reference to Chalmers, who famously characterised consciousness as the ‘hard problem’, Churchland wishes to show that consciousness is no harder than many other outstanding problems in neuroscience, such as motor control, learning or memory. Churchland seems to mock the idea that consciousness may be a different type of problem from these other neuroscience problems. However, with these other problems there is general agreement that however hard these problems may be, they could in principle be solved by a system of algorithms for manipulating energy, protein and other brain materials. What would emerge is a dynamic not in principle different from other aspects of organisms or even inanimate matter. It is less easy to do with consciousness, because what we know about electricity, about protein and about other brain molecules does not allow for them producing a new property not seen elsewhere in the universe. Churchland further attacks the zombie notion, which is essentially the argument that the brain functions of receiving, processing and responding to data could be achieved without the help of consciousness, and without giving rise to consciousness. Consciousness is indeed absent from the standard neuroscience description of the brain, which is causally closed. Churchland tries to evade this by saying that because we can conceive of such a brain does not necessarily mean that it could exist, and therefore we shouldn’t base anything on this argument. It is certainly true that we don’t know enough about consciousness, to say whether or not humans could have evolved without it. But that does not get us away from the fact that brain processes, as described by current neuroscience, do not have a requirement for consciousness, and do not produce it. Subsequent discoveries may show that the brain processes do require consciousness, but that is not the current state of neuroscience. It is somewhat ironic that the mainstream, which does everything it can to belittle consciousness and still more freewill, rushes to its defence when it is suggested that a sophisticated brain might operate without consciousness. Like other mainstream writers, Churchland seeks to fudge the question of qualia. She admits briefly that there are ‘prototypical’ qualia such as pain or blueness, as in the blueness of the sky, but then dives off into discussing grey areas such as thought or experience of limb positions. She asks whether these qualify as qualia. This proves to be rather a sleight of hand, because she now doubles back on the ‘prototypical’ qualia, and claims that they are only a starting point for investigation and not a full characterisation of their class. In this way, she manages to chip away at the qualia problem by introducing categories that might not be qualia, and thus might lead themselves to easier explanation. Even if this approach was successful in the grey areas, it would still leave the ‘prototypical’ qualia of pain and the blueness of blue unexplained, so really Churchland hasn’t progressed at all, although her readers may be left with the impression that she has. Churchland goes on to give us a bit of a lecture on philosophy, and in particular the fallacy of argument from ignorance. Basically she is saying that ignorance about something does not allow one to draw any conclusions about it. One can only draw a conclusion about oneself, to the effect that one is ignorant about the property under discussion. In particular, it is wrong to draw the conclusion that (1) we can never explain the property, (2) that science can never deepen our knowledge of the property, or (3) that the property can never be explained. Only a few modern thinkers such as Colin McGinn support something like the (1) and (2) positions, so the question is really as to whether the third position stands up. In justifying her stance, Churchland targets some straw men, for instance that because we don’t know the cause of a noise in the night, we are not justified in supposing a supernatural or alien origin, rather than gettinging to grips with the possibility of explaining consciousness from existing science. The difference between Churchland’s noise in the night and theories of consciousness is that we are not as ignorant about biology and physics as we are in the case of the noise in the night. We know enough about these to determine the type of things that are possible with them. A system of algorithms instantiated in neurons could in principle drive other neurons to perform brain processes, such as motor control, learning and memory, the precise mechanism of which is as yet unknown, but we know enough about the components of the brain, to know that they do not produce a property not detected in the rest of the universe, and consciousness falls into this category. Penrose/Hameroff Model The last part of Churchland’s paper deals with the Penrose/Hameroff model. Churchland remarks with truth that the details of the Penrose/Hameroff theory are highly technical. This seems too much for her, and she decides to skate round the main issues, but still attempts to refute the theory. Penrose did invoke the Platonic idea of mathematical truth, but in terms of the theory as a whole, this concept could be seen as only an image for what Penrose is proposing. Churchland, however, makes it look like the centrepiece. Her approach to the core of Penrose’s argument about consciousness, that it requires something that is not based on algorithms that can only be found at the quantum level, is garbled. She states that Penrose requires operations at the quantum level, but does not state why. This has the effect of making the whole thing sound improbable, without her having to engage with Penrose’s arguments. Penrose developed a detailed argument for how quantum gravity might be involved, but instead of trying to refute this, Churchland treats us to throw away lines such as ‘quantum gravity were it to exist’ and ‘no adequate theory of quantum gravity exists.’ Of course, scientific knowledge could never progress at all if every hypothesis was treated like this. Meanwhile Churchland offers no reasoned or detailed refutation of Penrose. We are also told that ‘mathematical logicians generally disagree with Penrose’, but their arguments are not presented, so we have no chance to judge. Churchland attempts to disparage the microtubule part of the Penrose/Hameroff theory. She points out correctly that anaesthetic molecules bind to protein receptors in the cell membrane. However, the evidence appears to suggest that these molecules permeate down to other cell proteins including microtubules, so she has hardly made the case against microtubular consciousness on this basis. Strangely she misses the strongest argument against the theory which is the tendency to rapid quantum decoherence in the conditions of the brain. She make think she is referring to this when she mentions the possibility of coherence being swamped by ‘millivolt signalling’. However, the problem is not signalling as such, but the overall activity of the environment. In fact, since this paper was written, microtubules have been shown to be involved in signalling. Subsequent to this the tone of the article sinks to a rather unprofessional level. Any proposal made by Hameroff is ridiculed for being only a ‘might’, a possibility, but how can science develop without ‘might’ proposals. Churchland also seems to think that the microtubule proposal did not explain how it linked to consciousness. This is factually incorrect, with regard to the detailed work of Penrose and Hameroff. More neural than thou (Reply to Patricia Churchland's 'Brainshy') Stuart Hameroff Both in 1996 Tucson discussion and debates In her ‘Brainshy’ paper in 1996, the philosopher, Patricia Churchland, attacked the Penrose/Hameroff model as well as the view points of Chalmers and Searle. Hameroff’s reply in this paper criticises Churchland for ignoring a number of brain features thought relevant to consciousness, including the probabilistic element in the firing of synapses, the role of gap junctions and dendrite-to-dendrite exchanges in brain processing, glial cells and the role of the cytoskeleton. He particularly criticises the lack of mention of the role of the cytoskeleton in regulating the neuron and its synapses. The latter parts of the paper seem to concentrate on redescribing parts of the Penrose/Hameroff model rather than specifically criticising Churchland. This discussion begins with the comment that Churchland is contemptous of Penrose’s Platonism. Hameroff counters by asking, ‘what is fundamental reality’. He remarks that as far back as 1971 Penrose tried to provide a description of the quantum mechanical geometry of space at the Planck scale by proposing quantum spin networks, which are suggested to encode the volumes and areas of physical space, but may also encode non-computational understanding and possibly the qualia. Hameroff also covers the question of anesthetics and consciousness in this article, pointing to evidence that anesthetics act in hydrophobic pockets in protein, which are also seen as a possible site for quantum coherent activity. Reference:- (1) Frank N. and Lieb W. (1997) On the molecular mechanism of general anaesthesia: Tucson discussion and debates (1996) (1982) Molecular mechanism of general anaesthesia: Nature 300: 487-497 (1985) Mapping of general anaesthetic targer sites: Nature 316: 349-81 (2) Halsey M (1989) Molecular mechanism of anaesthesia: General Anaesthetic – Fifth Edition Lamoreaux S, (1997) Demonstration of the Casimir Force: |
|
| quantummind Number 2, Your Street, Smallville, Countyshire, NW3 1SS, United Kingdom - Show map - Tel: 01234 567890 Email: sales@yourdomain.com Copyright © quantummind 2008. Web design by 7Soft.co.uk. |