MAKING SENSE OF CAUSAL INTERACTIONS BETWEEN CONSCIOUSNESS AND BRAIN.
Journal of Consciousness Studies, 9(11), 2002, pp. 69-95.
Max Velmans, Department of Psychology, Goldsmiths, University of London, New Cross, London SE14 6NW, England.
ABSTRACT. My target article (henceforth referred to as TA) presents evidence for causal interactions between consciousness and brain and some standard ways of accounting for this evidence in clinical practice and neuropsychological theory. I also point out some of the problems of understanding such causal interactions that are not addressed by standard explanations. Most of the residual problems have to do with how to cross the “explanatory gap” from consciousness to brain. I then list some of the reasons why the route across this gap suggested by physicalism won't work, in spite of its current popularity in consciousness studies. My own suggested route across the explanatory gap is more subterranean, where consciousness and brain can be seen to be dual aspects of a unifying, psychophysical mind. Some of the steps on this deeper route still have to be filled in by empirical research. But (as far as I can judge) there are no gaps that cannot be filled—just a different way of understanding consciousness, mind, brain and their causal interaction, with some interesting consequences for our understanding of free will. The commentaries on TA examined many aspects of my thesis viewed from both Western and Eastern perspectives. This reply focuses on how dual-aspect monism compares with currently popular alternatives such as “nonreductive physicalism”, clarifies my own approach, and reconsiders how well this addresses the “hard” problems of consciousness. We re-examine how conscious experiences relate to their physical/functional correlates and whether useful analogies can be drawn with other, physical relationships that appear to have dual-aspects. We also examine some fundamental differences between Western and Eastern thought about whether the existence of the physical world or the existence of consciousness can be taken for granted (with consequential differences about which of these is “hard” to understand). I then suggest a form of dual-aspect Reflexive Monism that might provide a path between these ancient intellectual traditions that is consistent with science and with common sense.
I would like to thank the commentators on TA for their many excellent commentaries. To simplify the process of relating these commentaries and my replies to the original text, I will deal with them according to topic in the sequence that these topics are treated in TA. At various points I refer to more detailed treatments of the issues addressed in my recent book Understanding Consciousness (henceforth referred to as UC). A few of the commentaries elaborate on TA and do not require a response. John Kihlstrom for example gives an excellent review of the evidence for the causal effects of consciousness on body/brain, the explanandum of TA, and Todd Feinberg outlines an independently arrived at but similar understanding of consciousness/brain interactions to my own. Of necessity, the bulk of my response is reserved for those who challenge aspects of my analysis, seek clarification, or defend alternative analyses.
Can reductive physicalism be defended?
The causal interactions between consciousness and brain could be easily explained, at least in principle, if consciousness were nothing more than a state of the brain. But, in the Appendix to TA I have given some of my reasons for doubting that science will ever demonstrate human phenomenal consciousness (C) to be nothing more than a state of the brain (B). I accept the widely held view that for any given a given conscious experience there will be associated physical causes and correlates within the brain. However, causation and correlation are not ontological identity. Identity is symmetrical (If B=C, then C=B) and obeys Leibniz's law (if B=C, then all the properties of B must also be the properties of C and vice-versa). Correlation is symmetrical (if B correlates with C, then C correlates with B) but it does not obey Leibniz's Law. Causation is neither symmetrical nor obeys Leibniz's Law. Why does this matter? Suppose that third-person science had established a perfect 1:1 correlation between a given experience C and its physical correlates B. Wouldn't that suffice to establish a reductive identity between them? No, because major differences between the first-person phenomenal properties of C and the third-person physical properties of B would remain. C and its correlate B would of course be intimately related. But an “intimate relationship” need not be a reductive identity. According to dual-aspect theory, B and C are complementary aspects of mind-itself. This would explain both their perfect correlation and their phenomenal differences. So it would provide a better fit to the available facts.
In defence of a “diffident physicalism” Torrance argues that if C and B were identical, that would also explain their correlation and “given the problems inherent in competing theories, asserting an identity relation seems reasonable, in the absence of a better alternative.” Given its better fit to the available facts, I would argue that dual-aspect theory does provide a better alternative and that its so-called problems are not problems at all (see below). I suspect that, on reflection, Torrance would agree that if physicalism can only establish consciousness/brain identity by assuming it, then it is on very weak ground indeed.
Can nonreductive physicalism be defended?
Of the physicalist theories, those of the emergent variety are perhaps the most plausible. It is obvious that higher-order physical properties emerge from the brain's micro-operations. In a sense, conscious experiences also emerge from the brain (in the sense that brain states can be said to cause or correlate with conscious experiences). I nevertheless resist the view that conscious experiences are just higher-order physical states of the brain. Higher order physical states of the brain are likely to correlate with given conscious experiences, but this does not warrant a reduction of the experiences to their correlates for the reasons outlined in the section above. Torrance suggests that it is a little unfair to tar the emergentists with the same brush as the reductionists. Emergentists accept that what emerges may be toto mundo distinct from the processes that have produced them. I agree, and I also agree that there is no problem about treating higher-order emergent properties of the brain as physical if, on commonly accepted criteria, they are physical (e.g. the electrical and magnetic fields detected by EEG and MEG). On the other hand, if, on commonly accepted criteria, conscious experiences would be categorised as mental as opposed to physical, then calling them “physical” becomes a mere relabelling exercise. To given this exercise bite, one would have to show that all the properties that that are normally thought of as “mental” (first-person properties) are in fact physical (third-person properties) otherwise one is simply left with all the problems of why “what it is like to be something” should emerge from or have a causal influence on the physical world. Torrance argues that a weak or wide form of physicalism might nevertheless be coherent and at the very least, tenable (in spite of there being no strong case for it). That may be so. But a physicalism that weak and wide would be empty. All the puzzles of consciousness would simply slip through its net.
In defence, Torrance suggests that there are other reasons “to do with ontological economy, conceptual conservatism, causal closure, and so on, against introducing non-physical properties into the universe.” I accept that ontological economy (simplicity) is desirable, but it has to be balanced against sufficiency, and I would argue that ontological monism combined with epistemological dualism achieves that fine balance. I don't agree that conceptual conservatism should be the order of the day when we are faced with a theoretical orthodoxy about the nature of conscious experience that is so clearly at odds with our actual conscious experience. In any case, the explanatory system in UC does have causal closure as far as I can tell. Crucially, I am not “introducing non-physical properties into the universe.” I am merely suggesting another way to make sense of the phenomenal properties that are observed to be there.
Can nonreductive physicalism be defended against my three threats to the third-person causal status of consciousness?
1. We lack conscious knowledge of the details of the processes that we are supposed to consciously control. According to Van Gulick this does not trouble nonreductive physicalism. As he notes, we often need very little knowledge of a process and of its detailed workings in order to affect, control or initiate it. Use of a computer for example does not require one to know anything about the underlying structure of the operations that execute its commands in machine language. Yet the high level control that we exercise is surely conscious. I agree—and I have used a similar argument to defend my own dual-aspect analysis of causality in TA (p x). However, I think we need to be precise about the sense in which such “conscious control” can be “conscious.” Control can be conscious in the sense that we are conscious of exercising control in this high level (global) way, and for everyday purposes this experience of being in control is veridical (when we think we have voluntary control, we usually do – see discussion of free will in TA). Controlled processing can also be conscious in the sense that it results in a conscious experience. The critical issue, however, is whether first-person conscious phenomenology actually controls or enters into physical processing, which seems to contravene the principle that the physical world is causally closed.
2. The problem of causal closure. If first-person experiences are invariably accompanied by distinct physical correlates, and if the physical world is causally closed, I don’t see how such experiences could exercise causal control—as the relevant control would already be exercised by their physical correlates (see the problem of “overdetermination” raised in the commentary by Chrisley & Sloman). The only escape route for “nonreductive physicalism” is to argue that, one day, science will establish conscious experiences to be nothing more than their physical correlates. As I have argued in the Appendix of TA, however, third-person science is restricted to establishing the neural causes and correlates of experiences. Given that causes and correlates are not identities, this scientific route to establishing a reductive identity is blocked.
3. Consciousness comes too late to affect the processes to which it most closely relates. Scientific claims for the causal efficacy of consciousness are typically based on contrastive analysis. Psychologists commonly contrast preconscious or nonconscious processing of a given type (e.g. preconscious visual processing of input) with conscious processing of the same type (e.g. visual processing where one becomes conscious of input) and then attribute any functional differences between these to the operation of consciousness. In Velmans (1991) I pointed out the fallacy of such attributions. Conscious experiences are a consequence of sophisticated, focal-attentive processing and without focal-attentive processing many forms of complex, novel functioning would not occur. However, the experiences themselves emerge too late to affect that processing. Van Gulick asks “But why should this impugn the commonly accepted causal status of consciousness? Neither folk psychology nor any scientific model of consciousness of which I know supposes that experiences produce the very processing from which they themselves result.” This misses my point. Once it is pointed out that consciousness results from a given process, of course no sensible person would claim it to have a causal role in that process. However, as far as I know, prior to my 1991 BBS target article, no one had pointed this out, claims for the causal efficacy of consciousness based solely on nonconscious/conscious contrasts were legion, and they continue, unabated to this day (see e.g. Baars, Banks & Newman, 2002). I accept, of course, that consciousness might in principle enter into processing that follows its emergence. However, this proposal still has to surmount the problem that the physical world is causally closed (see TA, Note 17 - and discussion above).
Van Gulick asks whether my own view is just a variant of a form of nonreductive physicalism that accepts explanatory pluralism. This would allow higher-level forms of mental organisation to have properties that are best described at that level rather than in the more basic terms of physics, without doubting that such properties are ultimately physical. This applies, for example, to economics. No one is worried by a money/matter problem or seriously advocates money/matter dualism. So why could the same not be true of mind? Van Gulick also points out that my epistemology in TA is presented as first- versus third-person dualism rather than pluralism although he notes that this may well be more a matter of exposition rather than substantive disagreement.
As it happens, these issues interconnect. Rather than contrasting epistemological dualism with epistemological pluralism, in my own work I combine these. That is, I support the view that there are many forms and levels of explanation within both first- and third-person accounts. As Van Gulick notes, pluralism is commonplace in third-person science, with well-defined hierarchies defined by the size and level of organisation of the phenomena, typically ranging from microphysical, macrophysical, chemical, biological, neurophysiological, cognitive/functional, and social levels of organisation. Nonreductive physicalism usually identifies conscious experience with just one level of this hierarchy, typically the cognitive/functional level that, in turn, supervenes on the brain's neurophysiology. In my view this is an oversimplification. Like the forms of material organisation that they accompany, first-person experiences can be described at different levels and may have an ontology at different levels. Some experiences appear to be socially determined, being, in part, social constructions grounded in culture and history (e.g. what it is like to be the Prince of Wales). Others (such as empathy) are quintessentially interpersonal, requiring the presence of at least one other human being (see readings in Thomson, 2001 Between Ourselves). Yet others, such as visual and auditory percepts appear to be largely individual, resulting from the binding of sensory features within an integrated human brain. Under appropriately controlled conditions these features can be further decomposed into the minimally discriminable phenomenal differences studied by psychophysics and so on.
It is important to note that so-called “nonreductive physicalism” is reductive in that it claims conscious experiences to be nothing more than a form of higher order material organisation or cognitive functioning. While dual-aspect theory accepts that every distinct experience has a distinct set of third-person physical and/or functional correlates (at social, personal or subpersonal levels of organisation) it resists the physicalist suggestion that such conscious experiences are reducible to their correlates at any of these levels. Rather, first person accounts of experience and third person accounts of accompanying physical functioning remain complementary and mutually irreducible, whatever the level of organisation.
This brings us to Van Gulick's Point 2. Do I equate the mental perspective with the first-person perspective? No. Like nonreductive physicalism, I treat the first person perspective(s) as a subset of a larger set of mental perspectives, some of which are entirely third person in nature, for example, those aspects of mental functioning described in cognitive psychological accounts of the mind. Unlike nonreductive physicalism, however, I argue that mental processes that have a conscious phenomenology cannot be exhaustively described in third person terms. While it is possible to describe what people do or how their brains function when they have beliefs, desires, etc., in third-person terms, without reference to their first-person perspective it is not possible to describe what they experience. I have given some of my initial reasons for the irreducibility of first- to third-person accounts in the Appendix to TA. As Van Gulick does not take issue with this preliminary analysis I will not offer a deeper defence of it here. Interested readers will find a far more extensive analysis in chapters 3, 4 and 5 of UC (see also my debate with Dan Dennett in Velmans, 2001).
This brings us conveniently to Van Gulick's point 3. As he notes, I share the physicalist commitment to ontological monism, but my dual-aspect view takes the ur reality (the nature of mind in this case) to be neither physical nor mental. Why? Precisely because it has both of these mutually irreducible, first- and third-person aspects. Viewed from the outside, the operations of ur mind appear to be operations of brain. Viewed from a first-person perspective, the operations of ur mind appear to be conscious experiences. Which is it really? If one assumes (as I do) that neither perspective is necessarily illusory or deluded, then the nature of ur mind must support both the views that we have of it. Given this, its nature is better described as “psychophysical” than “physical.”
This also addresses Torrance's suggestion that my position is not really all that different from nonreductive physicalism. He asks, “Doesn’t monism imply unity? So are you not saying that the neuroscientist’s third person facts and the subjective first-person facts are two equally real parts of a single unity? But then, if one side of this unity is physical, mustn’t the other side also be physical (or it’s not a unity)?” He then guesses correctly that, “Perhaps Velmans’ answer to this is that neither the third-person physical facts nor the first-person subjective facts are ultimately real, and that the underlying bedrock of reality is neither the one nor the other. (I guess this is implied by his calling it a ‘dual-aspect’ theory.)”
Viewing the mind-itself as psychophysical rather than physical is more than a simple relabelling exercise. What does this form of dual-aspect monism buy us? As I have argued in UC, “If consciousness and its physical correlates are actually complementary aspects of a psychophysical mind, we can close the “explanatory gap” in a way that unifies consciousness and brain while preserving the ontological status of both. It also provides a simple way of making sense of all four forms of physical (P) and mental (M) causal interaction. Operations of mind viewed from a purely external observer’s perspective (P®P), operations of mind viewed from a purely first-person perspective (M®M), and mixed-perspective accounts involving perspectival switching (P®M; M®P) can be understood as different views (or a mix of views) of a single, psychophysical information process, developing over time. In providing a common psychophysical ground for brain and experience, such a process also provides the “missing link” required to explain psychosomatic effects.” (UC, p 251) (see also TA page x)
Escape from the problem of causal closure. In Van Gulick's words “If the physical factors revealed from the third person perspective give a complete causal explanation of physical events and nothing nonphysical can have a causal impact on the physical, then there does not seem to be any room for other factors viewed from an alternative perspective to act as causes of physical events.” I agree. But we are not interested here in purely physical events. We are interested in the nature of mind, and according to the above, the nature of mind is psychophysical. Unlike “nonreductive physicalism” this analysis of mind and M®P causation is genuinely nonreductive. And it is this that makes it immune to Kim's (1999) point that, if the physical world is causally closed, either the mental reduces to the physical, or it must be epiphenomenal. Unlike “nonreductive physicalism” I do not claim that first-person experiences somehow enter into third-person physical functioning, so I do not need to reduce these experiences to physical events to make good that claim. Within dual-aspect theory there is a more intuitively plausible option. If the mind is genuinely psychophysical, then an entirely third-person physical view of it gives only a partial view of both its nature and its causal operations. Brain states are genuine phenomena (manifestations of ur mind), viewable from a third-person perspective, but conscious experiences are also genuine phenomena (manifestations of ur mind) viewable from a first-person perspective. Descriptions of brain states can be used to give a detailed account of the operations of mind in terms of its physical manifestations. Descriptions of first-person experience can be used to give an account of the operations of mind in terms of its conscious manifestations. For scientific purposes, third-person accounts are more useful. For everyday purposes, first-person accounts are often more useful. Both are required for an account of mind to be complete.
In what sense does this complementary perspectives account advance our understanding of the “hard problems” of consciousness?
In UC and prior work I have argued that some of the problems of consciousness require conceptual advance, some require empirical advance, and some require both. Empirical questions include, “What are the necessary and sufficient conditions for consciousness in the human brain?” and “What are the neural correlates of consciousness?” Questions such as “What is consciousness?” “What is the function of consciousness?” and “How can one make sense of the causal interactions of consciousness and brain?” appear to be largely conceptual. Why? Because the considerable evidence that one can already bring to bear on these questions somehow fails to address them. Each and every one of us has a vast reservoir of conscious experience. Gathering more of it won’t clarify what consciousness is. Extensive contrastive analyses of “conscious” versus “nonconscious” processing have already been carried out in psychological science. These illustrate functional differences between processes that either are, or are not accompanied by phenomenal consciousness. But such contrasts don’t reveal what the conscious phenomenology itself does (see TA). Nor does the massive evidence for mind/body interactions reveal how to make sense of the causal interactions of consciousness and brain. It is the opacity of these questions to further data gathering that makes them “hard.”
The present TA deals only with the last of these “hard problems” (I deal with the other problems and with how they all interconnect in UC). According to Gray, however, these problems are tangential to the real hard problem. Rather, “the Hard Problem can be stripped down to just two questions: how does the brain create qualia; and how does the brain inspect them?” It will be apparent that I do not entirely agree. These two questions are two of many. As it happens, they have third-person aspects that are fully amenable to empirical research. One proper answer to the question “How does the brain produce qualia?” would be to specify the necessary and sufficient physical conditions in the brain for the appearance of conscious qualia. This can be investigated by contrasting physical conditions that are necessary and sufficient for the appearance of qualia with those that are not—a standard method in science. Viewed in third-person terms, “how the brain inspects qualia” can be explained in terms of “how it inspects representations at the focus of attention”, “make them available for report”, and so on.
One might object, of course, that such third-person accounts don't answer the right question. The hard problem is not how one part of the brain might inspect and report on information in another part of the brain. It is how a physical brain could “inspect” a first-person conscious experience! But this is precisely the question that I do address! Ontological monism combined with epistemological dualism makes it clear that one can give a “pure third person account” of brain functioning (in terms of how subsystems in the brain inspect representations at the focus of attention). One can also give a “pure first-person account” of what is going on (in terms of the way that I, a conscious being, can inspect my own conscious experience). An account of “how the brain inspects its conscious experience” can then be seen to be a “mixed-perspective account, involving perspectival switching” (see TA). As Gray notes, “the only satisfying explanation will be one that shows how consciousness is linked to the scientific account that applies to the rest of that world.” In broad terms, that is what this analysis achieves.
Gray objects that, “The standard criterion for whether or not a proposed theory forms part of science is potential falsifiability by empirical observation. I cannot think of any such test of Velmans’ model, nor has he proposed any himself. The same is true, so far as I know, of all other versions of dual-aspect theory, including for example Chalmers' (1996) recent attempt to seek a common basis for the physical and conscious realms in an underlying stuff of ‘information’ (a move Velmans also makes, in his section on ‘the neural correlates of conscious experience’). Thus, Velmans’ proposed solution to the Hard Problem is purely philosophical, which is to say, purely verbal. It purports to tell us what we ‘really’ mean when we say things, respectively, from first-and third-person points of view. We need to move beyond this.” (p x) Rao also makes a similar complaint, when he suggests that I try to “resolve ontological issues by reducing them to epistemological ones.” (p x)
I accept the point that some aspects of my analysis have to do with how to understand “what we ‘really’ mean when we say things, respectively, from first-and third-person points of view”, namely my account of perspectival switching, and mixed-perspective explanations. However the ontological monism, combined with epistemological dualism that underpins this analysis is not just linguistic philosophy. It is a claim about the basic ontology of mind, its manifestations, and about how we can know its nature and its manifestations. For example, the proposal that first- and third-person perspectives of the mind are complementary and mutually irreducible is a claim about mind and how we can know it in the same sense that wave-particle complementarity is a claim about the nature of light, or electromagnetic energy is a claim about the unified nature of electricity and magnetism.
I also accept that my global analysis of how to make sense of the causal relations of consciousness and brain is not a theory about the details of how information encoded in the brain is mapped into conscious phenomenology. However such details can, in principle, be settled by empirical research, which makes them so-called “easy” problems rather than “hard” ones. To make sense of the hard problems we need to think about them in a different way (perhaps in the way that I suggest). But that does not make my entire analysis unfalsifiable, and purely “verbal and philosophical.”  A quick revisiting of my case for dual-aspect theory and my critique of alternative theories will confirm that they are tightly linked to falsifiable, experimental and clinical evidence—to evidence of mind/body interactions (see Kihlstrom), to evidence that the physical world is causally closed, to evidence that phenomenal consciousness comes too late to affect the processes to which it most closely relates (Velmans, 1991), to evidence of preconscious and unconscious processing, and so on. Crucially, unlike the variants of physicalism and functionalism defended by Torrance, Van Gulick, and Chrisley & Sloman, the dual-aspect theory developed in TA also conforms closely to the evidence of first-person experience.
It is instructive to dwell on this last point. Although the dual-aspect analysis in TA, and the Reflexive Monism that underpins it in UC are broad theories about how to make sense of the relation of consciousness to brain rather than theories about the neurophenomenological details, they are first and foremost empirical theories that try to make sense of the combined third- and first-person evidence. If there were convincing third- or first-person evidence that challenges some aspect of these theories, or some clear flaw in how the analyses connect to the data, then I would modify or abandon the theories. Many experimental and theoretical developments could challenge the details of my analysis, for example current attempts to demonstrate that a cognitive unconscious does not exist (Perruchet & Vinter, 2003), or that “qualia” do not exist (see debate with Dennett in Velmans, 2001), or that the neural correlates of consciousness are not representational states that encode identical information (Chrisley & Sloman - see below), or that consciousness is actually a mental field that influences the activities of brain (Libet, in press). I would also abandon my “complementary perspectives” analysis if our everyday insights into the operations of our own minds based on our first-person experiences turned out to be largely wrong.
Compare this with physicalism. Physicalism draws its scientific respectability from its namesake “physics”. However physicalism is a philosophical thesis about the ontological nature of conscious experience, not a field of science. Its claim that first-person phenomenology reduces, without remainder, to states of the brain has no real evidence in its favour (neural causes and correlates are not identities), and massive evidence to the contrary (conscious phenomenology does not resemble brain states). This makes it basically a faith in the all-encompassing nature of third-person science—a commitment to a worldview that is immune to falsifying evidence. If one is looking for an unfalsifiable theory, here it is.
How conscious experiences relate to their physical/functional correlates
At present, we know little about the physical nature of the correlates of conscious experiences. Nevertheless, in UC and TA I suggest that there are three plausible, functional constraints imposed by the phenomenology of consciousness itself. Normal human conscious experiences are representational (phenomenal consciousness is always of something). Given this, it is reasonable to assume that the neural correlates of such experiences are also representational states. For a given physical state to be the correlate of a given experience it is also plausible to assume that it represents the same thing (otherwise it would not be the correlate of that experience). Finally, for a physical state to be the correlate of a given experience, it is reasonable to suppose that it has the same “grain”. That is, for every discriminable attribute of experience there will be a distinct, correlated, physical/functional state. As each experience and its physical correlate represents the same thing it follows that each experience and its physical correlate encodes the same information about that thing. That is, they are representations with the same information structure. I also point out that different representational systems employing different formats can encode identical information without themselves being identical. Neural correlates, for example, might function as representations (encoding identical information to that displayed in their correlated phenomenology) without “mirroring” that conscious phenomenology in any obvious physical sense. While such correlates might be iconic, they could also be propositional, feature sets, prototypes, procedural, localised, distributed, static, dynamic or whatever. The operations on them might also be formal and computational, or more like the patterns of shifting weights and probabilities that determine the activation patterns in neural networks (TA, note 10). 
Given that I do give a supporting case for this in TA (and a far more detailed case in UC pages 236-251) it is hard to understand Chrisley & Sloman's contention that I “move without argument, from the representational nature of experiences to the existence of neural correlates of these experiences which have the same representational content as these experiences.” Nor is it easy to make sense of their claim that I am “either making a rather strong reductionist assumption, or (worse) postulating a dubious causal connection (between the structure of experience and the structure of neural states).” As I have noted, different representational systems can encode identical information without the systems reducing to each other—and the relation between experiences and their physical correlates is, by definition, correlation not causation.
Would the discovery of psychophysical correlations be scientifically useful?
As noted above, even perfect 1:1 correlations between conscious states and physical states would not establish their ontological identity. It is also well accepted in science that correlation does not establish causation. Consequently, even exact neurophenomenological laws that chart the way that given physical correlates map onto given conscious experiences would not be causal laws. If such bridging laws could be found they might nevertheless document invariant, empirical relationships in a precise way—and few, I suspect, would doubt that this would be a major scientific advance. Rakover, however, disagrees. According to him, “correlational laws” are not “natural laws” and cannot fulfil the requirements of measurement that are accepted in science. Consequently he thinks that neurophenomenological laws cannot be used to make sense of the causal relationships between consciousness and brain.
In assessing how Rakover’s commentary relates to my TA it is important to first note that he does not actually address the detailed account that I have given of how to make sense of the causal relationships of consciousness and brain, nor of the way that potential neurophenomenological laws might fit into such an account. However he does offer a critique of the scientific status of neurophenomenological laws as such, and of the use of “information” as a unit of measurement in psychology. As these are important elements of my analysis I will confine my reply to these relevant aspects of his critique. Second, although Rakover gives the misleading impression at the beginning of his commentary that my use of neurophenomenological laws is out of step with psychological practice in that it does not conform to “rules of the game that are accepted by the natural sciences and cognitive psychology” (p x), he goes on to admit at the end of his commentary that my use of such correlational relationships is entirely conventional within psychology and that his real target is psychology! As he concludes on page x, his critique “could also be directed at psychology at large. In comparison with research in the natural sciences, psychological research is limited and does not progress like physics (see Rakover, 1992).”
What are these supposed limitations on psychological research? According to Rakover, neurophenomenological laws do not fulfil the requirement for the “unit equivalency” found in natural laws such as Newton's law of gravity, where the units of measurement found on both sides of the equation S = 1/2GT2 can be shown to be equivalent. Let us suppose, for example, that neuropsychology discovers the exact neural correlates of different subjective aspects of pain phenomenology and manages to express its findings in neurophenomenological laws. In such cases, Rakover asks, “Can it be shown that the combination of the units of measurement on the right-hand side of the pain equation is identical to the combination of the units on the left-hand side? To the best of my knowledge the answer is no.” (p x) I agree. But such an absence of unit equivalency provides yet another argument against reductive physicalism. It has nothing to do with whether or not there are distinct physical/functional correlates of distinct pain experiences, or with whether or not it is possible to chart such relationships precisely in the form of nonreductive neurophenomenological laws that do not require “unit equivalency.”
But would the absence of “unit equivalency” make neurophenomenological laws unscientific? Consider Rakover's doubts about studies of pain phenomenology. Pain is often presented as a paradigm case of a private, subjective, mental event within philosophy of mind. There are many ways to measure the subjective experience of pain, but at the present time no valid “objective” measure of pain experience (in terms of a physiological index) exists. In spite of this, over the period 1960 to 2002, the Medline database lists around 200,000 publications on pain and its alleviation, making it a heavily investigated area of medicine. According to Rakover, such studies are restricted by the absence of “fundamental measurement units” of the kind that obtain, say, for the measurement of length, which sustain properties such as transitivity and additivity. While this is true, it is hardly news to anyone trained in psychological research, where it is taken for granted that whenever numbers are assigned to psychological variables these must be scaled in a way that is appropriate to those variables (reaction time and error rate merit a ratio scale, subjective judgements of magnitude generally merit an ordinal scale, categorical judgements a nominal scale, and so on). Once an appropriate scale is assigned, numbers derived from measurements of behaviour or subjective judgements can be subjected to appropriate statistical analyses, and the results interpreted as supporting hypotheses (or not) in the normal way.
It is true that few relationships between physical and psychological variables have been found to be sufficiently general and orderly to merit the term “law” and even these do not satisfy “unit equivalency.” Perhaps the best example is Stevens’ power law J = kIx where J is the judged intensity of a stimulus (e.g. its brightness or loudness), k is a scaling constant, I is the physical intensity (e.g. specified in lumens or decibels), and x is a constant whose value depends on the modality of the judged stimulus (e.g. for judged loudness, x=0.3). Stevens’ law charts how variations in the physical stimuli are translated into judged changes in the way those stimuli are experienced. Consequently it is “correlational” in precisely the way that Rakover describes. Does this mean that Stevens’ law is unscientific? No. There are countless examples in science where Nature does not fit into the neat conceptual boxes that we have prepared for her, and the psychological and biological sciences have long abandoned the view that the only relationships of scientific interest are fundamental causal laws of the kind found in physics. Functional models in cognitive psychology and compositional accounts of the structure of biological systems are obvious cases in point. In psychophysics, Stevens’ power law may not satisfy unit equivalency, but it nevertheless expresses empirically verifiable relationships between physical dimensions of stimuli and subjective judgements about those stimuli in a precise, systematic way, and it is in that sense unquestionably scientific. Given this, it is reasonable to hope that in some future neuroscience it may be possible to develop neurophenomenological laws with equivalent precision and generality.
Rakover also claims that information cannot be a unit of psychological measurement. But again, few psychologists would agree. It is true that, following George Miller's (1956) seminal paper The magical number 7 ± 2, psychologist have long accepted that human mental processing is often too flexible and varied to be computable in “bits” in the precise Shannon sense. Nonetheless the psychological use of concepts derived from information theory and/or the more general principles of information processing developed within electrical engineering is ubiquitous—to the point that, in cognitive psychology, mental processing is habitually referred to as human information processing. In any case, my own use of the terms “information” and “information structure” relate to a fairly precise use of these terms that is applicable in psychophysics, for example in the study of difference limens (minimal discriminable differences). Such studies document whether or not physically measurable differences in stimuli are translated or not (by sensory/perceptual processes) into consciously perceived differences, that is whether or not information about physical differences is translated into detectable changes in phenomenology. In the same way it is possible to study whether or not physical/functional differences in neural representations of stimuli are translated into detectable changes in phenomenology. Physical/functional changes in neural representational states that are translated may be said to be of the same “grain” as the conscious phenomenology and to mirror its information structure. Rakover doubts that it would be possible to identify such information bearing states, as one cannot remove one's dependence on subjective reports of what is or is not experienced. I agree that one cannot remove subjective reports. However, the combination of subjective reports with triangulating third-person observations of neural states is standard practice in neuropsychology. Investigation of the neural correlates of consciousness is technically difficult, but the field of investigation is already very large (cf Metzinger, 2000). Rather than being questionable science, it is unquestionably normal science.
The ways in which different conscious experiences relate to their physical correlates have to be understood in their own terms. Some properties of these relationships appear to resemble ones that are already well understood in natural science, but, as far as one can tell, no other purely physical system provides an exact homology. Crucially, the relations of experiences to their physical correlates have to be understood in terms of how certain phenomena (the experiences) viewed from a first-person perspective relate to other phenomena (the correlated brain states) viewed from a third-person perspective. By contrast, all the properties of physical systems (conventionally understood) can be viewed from a third-person perspective.
Videotapes and TV screens. Sometimes, however, analogies can help. For example, to understand how experiences and their physical correlates might encode identical information without themselves being identical it is useful to know that such a dissociation between representations and representing systems are commonplace in technology—as in my example of the play “Hamlet” encoded on videotape or displayed on a screen. Given my limited intent, it is hard to understand Chrisley & Sloman's claim that this analogy “misfires.” As they correctly note, “this is not an ontological reduction”. However, according to them, “it is an epistemological one”, and then they go on to claim that, “epistemological dualism is the only thing separating Velmans from the physicalist positions he rejects.” But how can this be?
Videotapes and TV screens encode information in entirely different formats. Even when they encode information about the same thing, they do so in two entirely different ways—which is broadly analogous to knowing about one thing in two different ways. So in what sense is this an “epistemological reduction”? Admittedly, there is one known, the nature of mind, with two (material and phenomenal) aspects, by which it can be known. But, given that I suggest the nature of mind to be “psychophysical,” in what sense is this “physicalist”?
Electricity and magnetism. The same information can be formatted differently, depending on the characteristics of the representing system. If one can specify the different ways that given information is formatted, then it should be possible, in principle, to specify how those different formats map onto each other. In TA and UC I suggest that, in some future neuropsychology it might be possible to specify how the phenomenology of given consciousness experiences map onto to their physical correlates in this way. This might provide a dual-aspect account of the nature of mind in which the relationships between its physical and phenomenal aspects were specified precisely, perhaps with the precision that electrical current in a wire can be related to its surrounding magnetic field.
Chrisley & Sloman confusingly suggests that the duality that I have in mind with the electromagnetism analogy, “is one of aspects, not of ontological character.” What I actually suggest is that the phenomenal and physical aspects of mind specify its (psychophysical) ontological character. Even more confusingly they go on to write, “.. the analogy doesn’t work: electricity and magnetism are not simply two ways of thinking about the same phenomenon, but two different physical phenomena that can be related to each other mathematically.” Given that I never suggest that electricity and magnetism are simply two ways of thinking about electromagnetism (rather than genuine aspects), nor that physical and phenomenal aspects of mind are simply two ways of thinking about mind (rather than genuine aspects) the relevance of this comment to my analysis is hard to understand. They then add to the confusion by going on to write, “In contrast, and crucially, Velmans claims that the difference between first and third person ways of thinking of psychophysical stuff is merely that of differently formatted ways of representing the same information.” I claim nothing of the sort. As noted above, first- and third-person (phenomenal and physical) aspects of mind are not merely different “ways of thinking” about it. Being genuine phenomenal and physical aspects (or manifestations) of mind, they can in principle encode the same information in different phenomenal and physical formats. Chrisley & Sloman go on to note that “the electrical phenomenon is not just an aspect, a way of formatting the same information as that represented by the magnetic way of looking at the situation. There are situations where only the electrical description applies, and other situations where only the magnetic description applies.” I agree—although this again has nothing to do with my analysis of dual-aspect monism or my use of the electricity/magnetism analogy. They go on to conclude that, “Prima facie, this suggests that there are two distinct phenomena involved; to argue that there is actually only one, root phenomenon will require further work from Velmans.” Here I disagree. Electricity and magnetism are indeed distinct phenomena, but the view that they are both manifestations of only one root phenomenon (electromagnetism) is received wisdom in physics. It requires no further work from me.
Wave-particle complementarity. In TA note 13 and UC I note that my dual-aspect analysis of mind also has some interesting resemblances to wave-particle complementarity in quantum mechanics – although, once again, the analogy is far from exact. Quanta either appear to behave as electromagnetic waves or as particles depending on the observation arrangements. And it does not make sense to claim that electromagnetic waves really are particles (or vice versa). A complete understanding of quanta requires both complementary descriptions. First- and third-person observations of mind also depend on very different observational arrangements, so that may help to explain why, from a first-person perspective it takes the form of conscious phenomenology, whereas viewed from the outside it appears to be a brain. Like wave-particle accounts in quantum mechanics, phenomenal and physical accounts of the mind's operations appear to be complementary and mutually irreducible. A complete account of mind requires both.
Note that these distinguishing features of dual-aspect monism contrast sharply with competing analyses of the experience/brain state relationship. Substance dualists maintain that experiences and correlated brain states are entirely different “substances” or “entities”, idealists argue that all physical entities (including brain states) are really forms of mind or consciousness (Rao), and physicalists argue that experiences are nothing more than states of the brain (Torrance, Van Gulick, Chrisley & Sloman). All these positions have well known problems. For example, dualism splits the universe in a way that makes it difficult to get it together again, idealism does not cope well with the apparent, autonomous existence of the material world, and physicalism does not cope well with the phenomenology of conscious experience. I have argued that dual-aspect monism allows one to accommodate first- and third-person evidence in a more natural way that avoids such problems. While the case for this above (and in Velmans, 1991, UC and TA) does not rely in any way on analogies from other branches of science, the parallel with wave-particle complementarity in quantum mechanics is suggestive.
However, according to Chrisley & Sloman, this analogy “is even worse”—although they take issue not with me, but with the founders of quantum mechanics. They write, “More and more physicists and philosophers take the appeal to complementarity as a reductio ad absurdum of particular ontological positions in quantum mechanics. They do not deny the veracity of the data that have led some to conclude that quanta have both wave and particle aspects; but they do deny that the paradox of complementarity is a satisfying way of accounting for that data. There are other, less paradoxical and thus more satisfying metaphysical pictures on offer (e.g. Bohm, 1952; Hiley and Pylkkänen, 2001). To say that your metaphysics of mind is akin to the wave/particle complementarity metaphysics of quanta is just another way of saying that you don’t have a satisfying metaphysics, and choose instead to ‘live with’ the paradoxes.” I think that is being rather unfair to our colleagues in physics. The majority of physicists are more concerned with whether the mathematics of QM accounts for the data, and they think of (exclusive) complementarity as a current, best description of the empirical findings, imposed by the limitations of measurement, rather than “a reductio ad absurdum of particular ontological positions.” Nor is there any emerging consensus about what would be a satisfying metaphysics. As it happens, I share Chrisley & Sloman’s interest in more classical accounts of QM findings (in spite of this being a minority view in physics). According to Bohm and his collaborators, wave-like and particle-like behaviour are manifestations of a unified, grounding reality (Bohm often refers to this as an “implicate order”) just as I have claimed experiences and their physical correlates to be dual-aspects of a unified, psychophysical mind. So adopting a classical metaphysics in QM (in the way that Chrisley & Sloman suggest) would make the analogy with dual-aspect monism even closer!
In sum, let me stress again that analogies have their purposes, but they are not homologies. The analogies that I have used illustrate how phenomenal and physical representational systems might format the same information in different ways, and how phenomenal and physical aspects of mind might be tightly bound to each other without reducing to each other. But I do not claim consciousness to be literally a picture on a TV screen, a magnetic field, or a wave-like QM phenomenon (to claim all three simultaneously would in any case be absurd). The relation of any given conscious experience to its physical correlates has to be understood in its own terms.
A re-examination of what we take for granted.
What has ontological primacy--consciousness or the physical world? In current Western philosophy and science the existence of the physical world is generally taken for granted, while the existence of consciousness is thought to be somewhat mysterious. The physical world is also generally assumed to be the primary reality on which other “emergent” forms of existence such as mind and consciousness depend. Chrisley & Sloman for example, take it for granted that the physical/experiential relationship is asymmetrical. Physical states can exist without accompanying experiences (e.g. in the form of preconscious brain states)—but conscious experiences cannot exist without accompanying physical states. As they note, “The only way to impose symmetry would be to assume (as others have been forced to do, e.g. Chalmers, 1996) that whenever there is a physical phenomenon, there is some experiential phenomenon, however slight or imperceptible or implausible, accompanying it. Panpsychism threatens.”
It is instructive to note however that such opinions about what has ontological primacy and what constitutes a ‘threat’ (to right thinking) are not universal. As Rao points out, very different views about the ontological status and distribution of consciousness and mind dominate in philosophical traditions that have developed in the East. In these traditions, the irreducibility of consciousness to brain states is taken for granted and consciousness, not the physical world, is thought to be primary. In some Indian traditions for example, the physical world is thought to be a projection of consciousness constructed by the mind.
How is it possible that thinkers in the West and the East have come to such very different conclusions? Note that the ontological primacy of either consciousness or the physical world is not obvious from the immediate, empirical “evidence of our senses” for the simple reason that, in everyday life, conscious experience and what we normally think of as the “physical world” co-arise. That is to say, what we normally think of as the physical world just is the 3D phenomenal world that we experience. However, Western and Eastern thinkers have traditionally taken a very different interest in what is experienced. Western “third-person” science has traditionally been interested in experience as a means to an end, namely the nature, control and transformation of the entities and events that such experiences represent (what they are experiences of) and has developed investigative methods and technologies appropriate to these interests. By contrast, Eastern “first-person” philosophy and science has traditionally been interested in the nature, control and transformation of the experiences themselves, and has developed methods appropriate to these aims. I suggest that these different foci of interest and accompanying methodologies partly explain East-West differences of opinion about what has ontological primacy.
It is not altogether surprising that if one’s third-person investigative attention is entirely focused on the material entities and events that one’s experiences represent, one might conclude their fundamental nature to be entirely material. Many external entities and events appear to exist whether they are experienced or not, thereby supporting their ontological primacy and a form of physical realism. In the human brain some processes appear to be accompanied by consciousness while others appear to be preconscious, unconscious or nonconscious, suggesting a physical/experience asymmetry. Viewed from the outside, the material forms of entities and events are evident, but not any accompanying experience, even in other human beings (the problem of “other minds”). Consequently panpsychism looks dubious.
On the other hand, if one’s first-person investigative attention is focused in ever finer ways on conscious experience itself it is not surprising that one might conclude its fundamental nature to be a refined form of consciousness (traditionally a “pure”, contentless consciousness). Conscious experience is in any case “immediately given” and is epistemically primary in the sense that it provides the foundation for the acquisition of all empirical knowledge. Indeed, what we normally think of as the “physical world” just is the 3D phenomenal world that forms part of everyday conscious experience (see above). Conversely, without conscious experience this phenomenal physical world would not exist (a form of idealism), thereby providing grounds for the Eastern view that consciousness has ontological primacy.
Which view is correct? It is not possible to attempt a full analysis in a few lines. However, in UC and TA I develop a dual-aspect, reflexive monism that treads a careful path between taking either a first- or third-person approach to be more privileged or fundamental. Rather, these perspectives are complementary and mutually irreducible. For example, in Velmans (1990a) and UC chapter 7 I suggest that Eastern idealism and Western realism may both be true although they are true about different things. Idealism may be said to apply to the observer-dependent existence of the phenomenal world while realism applies to the observer-independent existence of the entities and events (things themselves) that experienced phenomena represent. Under normal conditions, neither a first- nor a third-person perspective provides a “view from nowhere,” that is a view of the thing-itself as it is in-itself, even if the aspect of the thing-itself under scrutiny is the human mind. Conversely, both investigative routes can lead to deeper knowledge. Third-person science provides a deeper knowledge of the material world, understood in a third-person way. First-person investigations of consciousness provide a deeper knowledge of one’s own mind, understood in a first-person way. My route to this position is an entirely conventional Western one, relying on the normal triangulation of scientific evidence, everyday experience, common sense and theory. Nevertheless, once the implications of this position are fully worked out (in terms of what consciousness is and does, and how it relates to the brain and the physical world) the Reflexive Monism that results takes one a long way from current Western materialism. I conclude for example that
“Human minds, bodies and brains are embedded in a far greater universe. Individual conscious representations are perspectival. That is, the precise manner in which entities, events and processes are translated into experiences depends on the location in space and time of a given observer, and the exact mix of perceptual, cognitive, affective, social, cultural and historical influences which enter into the “construction” of a given experience. In this sense, each conscious construction is private, subjective, and unique. Taken together, the contents of consciousness provide a view of the wider universe, giving it the appearance of a 3D phenomenal world. This results from a reflexive interaction of entities, events and processes with our perceptual and cognitive systems that, in turn, represent those entities, events and processes. However, conscious representations are not the thing-itself. In this vision, there is one universe (the thing-itself), with relatively differentiated parts in the form of conscious beings like ourselves, each with a unique, conscious view of the larger universe of which it is a part. In so far as we are parts of the universe that, in turn, experience the larger universe, we participate in a reflexive process whereby the universe experiences itself.” (UC, p233).
Later, I add, “In this sense, we participate in a process whereby the universe observes itself – and the universe becomes both the subject and object of experience. Consciousness and matter are intertwined in mind. Through the evolution of matter, consciousness is given form. And through consciousness, the material universe is real-ized.” (UC, p280).
It is not possible to summarize the full implications of reflexive monism in a few lines, let alone the case supporting it. However, as Rao notes, my route appears to travel from West to East. His only complaint is that I have not traveled far enough! While I do not have space to deal with how UC relates to various Eastern philosophies in any detail, Rao’s comments provide a welcome opportunity to assess the internal coherence of TA (and UC) from a very different perspective, and it is instructive to address his main points.
From West to East?
Any comparison of Eastern and Western views of “consciousness” and “mind” has to start with a clarification of terms, for the simple reason that in the West and the East the terms “consciousness” and “mind” are habitually used in different ways. As Rao notes, I largely confine my use of the term “consciousness” to phenomenal consciousness—the everyday experience of the external world, the body, and inner experiences (such as thoughts, feelings and so on). Although there are many uses of the term “consciousness” in the West, phenomenal consciousness is arguably closest to its most common usage. Crucially, it is consciousness in the sense of “phenomenal consciousness” that poses “hard” problems of the kind currently discussed in Western philosophy such as “How could conscious experiences affect the activity of brains?” (the subject of TA)
I also largely follow current Western conventions in my use of the term “mind.” As with “consciousness” the term mind has various uses. However, in psychology it is typical to think of the human mind as that which enables us to function in certain ways (to think, to solve problems and so on). Although the details of how consciousness, mind and brain relate are in dispute, there is consensus that “mind” is intimately connected to both brain and consciousness. A major finding of 20th century psychology is that mental processes may or may not “be conscious.” Some processes have associated phenomenal contents, while others are preconscious, unconscious, or nonconscious. Consequently, in Western psychology, “mind” is commonly thought of as encompassing consciousness.
Eastern common usage of the terms “consciousness” and “mind” is somewhat different. However it is not difficult to tease out terminological differences from genuine, theoretical ones. At first glance, the Samkhya-Yoga tradition described by Rao might look very different to Reflexive Monism. In this tradition, consciousness, with purusha  at its center, forms the ground of one's individual being. It is the contentless container within which perspectival, phenomenal consciousness takes form. Mind, unlike consciousness, is physical in that it can be described in material forms and accounted for in physical terms. “The purusha as the center of consciousness is distinct and has unique experiences through its associated mind-body complex. Such observer dependent relativity, in Yoga as well as in Vedanta, is not absolutely given but a transient condition that can be overcome by disciplined practice. The purusha finds itself reflected in the mind illuminating the material forms of the universe. Thus mind becomes as instrument through which the universe reveals itself. Subject-object distinction is not fundamental. It is a contingent manifestation of the mental process by which the universe is revealed.” (Rao, p x)
In my own analysis in UC I am careful to remain within the evidence base currently accepted by Western science, and I tease the modern problems of consciousness away from more traditional concerns with the nature of the “soul” (UC, pp15-16). Consequently, I do not comment on the existence or operations of “purusha”. While I have no doubt that first-person investigative attention can lead to a deeper understanding of mind (see above) I also remain neutral about whether disciplined practice can entirely remove one’s observer dependent relativity, or whether the ensuing conscious state can be entirely contentless (UC, chapter 1, note2). Nevertheless, there are broad similarities between Reflexive Monism and the Eastern view that Rao describes. Like Samkya-Yoga philosophy (and Western materialism) I accept that mind has (third-person) physical aspects that provide an instrument for the formation of phenomenal consciousness. I also accept that both phenomenal consciousness and material aspects of mind are grounded in something deeper, namely a self-revealing universe in which the subject-object distinction is not fundamental (see above). However, in my own analysis the terms I use to refer to what is deeper are different. For example, “consciousness with purusha at its center”, is replaced by the “deeper nature of mind” (or, in Kantian fashion, “mind-itself”).
These different uses of terms partly account for a number of confusions in Rao’s summary of my own theoretical position. Rao notes that both Indian theories and my own make a distinction between consciousness and mind. But he suggests that, “In the Indian view; the distinction is fundamental and primary in the sense that one is not reducible to the other. In Velmans’ view, the distinction is secondary and holds good at the epistemological level and not at the ontological level. Thus consciousness becomes a subcategory or species of the mind.”
In fact, however, I never suggest that “consciousness” interpreted in the broad Eastern sense is an aspect of the material mind interpreted in the narrow Eastern sense (that would indeed be inconsistent with my view that consciousness cannot be reduced to states of the brain). What I actually suggest is that phenomenal consciousness (understood in the conventional Western sense) is an aspect of the deeper nature of mind (mind-itself). The neural correlates of consciousness and other forms of brain functioning provide the complementary, material aspect of mind-itself. Being genuine aspects, both consciousness and brain have an ontology, as well as providing first- and third-person means by which the mind can be known. Consequently Rao is wrong to suggest that the distinction between mind and consciousness in my own work is purely epistemological. And he mistakes my suggestion that mind-itself encompasses consciousness to mean that the material aspect of mind encompasses consciousness. Rather, the deeper, psychophysical nature of mind encompasses both its manifest conscious and material aspects.
If one replaces Rao’s Eastern use of the term “consciousness” with my use of the term “mind-itself” or more broadly “the thing-itself” one immediately clears up a number of other confusions. Rao writes for example, “the distinction between first-person consciousness and third-person consciousness adds little to the clarity of the concept of consciousness. Consciousness is consciousness whether we look at it from a first-person or the third-person perspective. It may manifest different characteristics at different levels of observation, but it underlies all awareness. Consciousness is what makes awareness possible. It is the ground condition for all forms of awareness, like matter which is the ground condition for all the material forms we experience.” (p x)
Viewed in conventional Western terms, Rao’s statement makes no sense, for the reason that phenomenal consciousness cannot be viewed from a third-person perspective (whatever the level of observation). In the West, the terms consciousness, phenomenal consciousness and conscious awareness are often used interchangeably (I do so in my own work). Consequently it makes no sense to suggest that consciousness underlies awareness (it cannot underlie itself). The suggestion that “consciousness is the ground condition for all forms of awareness, like matter which is the ground condition for all the material forms we experience” is also inconsistent with the view that consciousness has ontological primacy over matter (the alternative is ontological dualism). By contrast, “mind-itself” can be viewed from first- and third-person perspectives, does underlie phenomenal consciousness and is the ground condition for both its conscious and material manifestations (thereby avoiding dualism).
Of course, these different terms for what has ontological primacy in the East and in the West (and their corresponding descriptions) also reflect substantive theoretical differences. In Samkya-Yoga philosophy “consciousness with purusha at its center” is the fundamental reality. In Western materialism the physical world is the fundamental reality. In Rao’s opinion I have to choose between these: if I reject the reducibility of conscious experience to brain states, I must accept the primacy of consciousness. Not so. I accept that if one investigates the mind from a third-person Western perspective it will appear to be entirely physical while if one investigates it from a first-person Eastern perspective it will appear to be entirely conscious experience (see above). But, as far as I can judge, neither route to knowledge of the mind is privileged, incorrigible or complete. Rather, first- and third-person routes to knowledge of the mind are complementary and mutually irreducible. Consequently, the “deeper nature of mind” (mind-itself) is better described as psychophysical.
Rao suggests that, in my own analysis, there is an asymmetry between conscious states which do not reduce to states of the brain and nonconscious mental activities which do reduce to brain states. In his view this leads to “the real problem.” He writes, “Velmans acknowledges that consciousness is not reducible to brain states or functions. Yet, he considers consciousness an aspect of the mind. The mind in his view is broader to include nonconscious mental activities as well. Here rests the real problem. Consciousness (subjective experience) is irreducible to neural states or brain functions, whereas the nonsubjective states of the mind are in principle reducible. In the light of such a fundamental difference between them, it is hardly plausible to argue that consciousness is a species or an aspect of the mind. The irreducibility of consciousness to physical states entails that the difference between conscious and nonconscious aspects of the mind is one of kind, primary and fundamental. Reducibility or otherwise of one category into another is an ontological matter and not simply an epistemological issue.” (p x)
Epistemological symmetries and asymmetries between first- and third-person perspectives are important, but I agree with Rao that reducibility is an ontological matter, not an epistemological matter. However, such an ontological asymmetry between conscious states and nonconscious ones would occur only if the nonconscious nature of mind turned out to be entirely physical (as Rao himself assumes). If so, conscious mind would have dual-aspects, but nonconscious mind would only have a physical aspect. As it happens, a similar view to this is held by those “nonreductive physicalists” that adopt property dualism. Whether this leads to a “real problem” depends on whether such asymmetries actually occur in nature or not (if they do, it would be perverse to regard them as a problem). Chrisley & Sloman for example take such asymmetries for granted. As they note, “There is a fundamental asymmetry between the physical and the conscious: Physical laws apply everywhere, both in situations where there is and where there is not consciousness, while the converse does not hold. So there seems to be a primacy of the physical, and one must reply to the idea that it is this physical, causal reality which is always doing all the work.” (p x) Rao adopts the opposite view that a pure consciousness without any material form is the basis of everything, but does not appear to recognize that this produces an inverse asymmetry (in which physical matter becomes secondary to consciousness).
Whether such asymmetries actually occur in Nature is up to Nature—and whether they do or not is largely tangential to the analysis of consciousness that I have given in TA and UC. It is important to note however that, unlike both materialism and idealism, ontological asymmetries are avoidable in dual-aspect theory, which allows for the possibility that mind-itself has a dual-aspect, psychophysical nature irrespective of whether its operations are unconscious, preconscious or conscious.  On this interpretation, the dual-aspect nature of mind is fully manifest only in those aspects of mind that are “conscious”. However, with appropriate investigative techniques, some preconscious and unconscious aspects of the mind can become conscious (in the sense that we can become aware of those aspects or to real-ize their nature). Unconscious and preconscious aspects of mind can also be thought of as psychophysical in the sense that they can have causal effects on both conscious experiences and physical states of the body/brain, for example in the operation of preconscious free will (see TA).
Note that whatever view one adopts about what is primary, one is left with the problem of origins. In the West, we generally accept that the origins and existence of consciousness are somewhat mysterious (when and why did it emerge?). But we habitually ignore the fact that the origins and existence of matter are equally mysterious. Why should there be anything rather than nothing? As the origins of a “psychophysical mind” are also mysterious, the choice between these three positions has to be made on other grounds.
Which view is preferable? Note that there are “hard” problems associated with taking either the material world or conscious experience to be more primary than the other. In the West it is well recognized that taking the material world to be primary leaves one with the problem of consciousness. How could something like an experience emerge from a material world that does not already have it? It is perhaps less well recognized in the East that if one takes the existence of consciousness to be primary one is left with the inverse problem. How could something like an independently existing material world emerge from something like an experience? If the thing-itself and mind-itself are fundamentally psychophysical one avoids such problems. And one can then make sense of mind/body interactions observed in clinical practice and everyday life.
Arbuthnott, K.D. (1995), ‘Inhibitory mechanisms in cognition: Phenomena and models’, Cahiers de Psychologie Cognitive, 14(1), pp. 3-45.
Baars, B. J., Banks, W. P. and Newman, J. B. (2002) (eds.) Essential Sources in the Scientific Study of Consciousness. The MIT press. (in press)
Chalmers, A. F. (1992), What is this Thing Called Science? Open University Press.
John, E. R. (2002), ‘The neuropsychology of consciousness’, Brain Res Brain Res Rev, 39, pp. 1-28.
Kim, J. (1999), Mind in a Physical World, MIT Press.
Libet, B. (2003), ‘Can conscious experience affect brain activity? JCS (in press).
McFadden, J. (2002), ‘Synchronous firing and its influence on the brain’s electromagnetic field: Evidence for an electromagnetic theory of consciousness’, JCS, 9 (4), pp. 23-50.
McFadden, J. (2002), ‘The conscious electromagnetic information (Cemi) field theory: The hard problem made easy?’, JCS, 9 (8), pp. 45-60.
Melzack, R. (1987), ‘The short-form McGill Pain Questionnaire’, Pain, 30, pp. 191-197.
Metzinger, T. (2000) (ed.), Neural Correlates of Consciousness, The MIT Press.
Miller, G. A. (1956), ‘The magical number seven, plus or minus two: some limits of our capacity for processing information’, Psych Review, 63, pp. 81 – 97.
Perruchet & Vinter (2003), ‘The self-organizing consciousness’, Behavioral and Brain Sciences (in press)
Pockett, S. (2002), ‘Difficulties with the electromagnetic field theory of consciousness’, JCS, 9 (4), pp. 51-6.
Searle, J. (1990), ‘Consciousness, explanatory inversion and cognitive science’, Behavioral and Brain Sciences, 13(4), pp. 585-642.
Thomson, E. (2001) (ed.), Between Ourselves: Second-Person Issues in the Study of Consciousness, Imprint Academic.
Varela, F. and Shear, J. (1999) (eds.), The View from Within: First person approaches to the study of consciousness, Imprint Academic.
Velmans, M. (1990), ‘Is the mind conscious, functional, or both?’, Behavioral and Brain Sciences, 13, pp. 629-630.
Velmans, M. (1990a), ‘Consciousness, brain, and the physical world’, Philosophical Psychology, 3, pp. 77-99.
Velmans, M. (1991a), ‘Is human information processing conscious?’, Behavioral and Brain Sciences, 14(4), pp. 651-701.
Velmans, M. (1999), ‘Intersubjective science’, Journal of Consciousness Studies, 6(2/3), pp. 299-306.
Velmans, M. (1995), ‘The relation of consciousness to the material world’, Journal of Consciousness Studies, 2(3), pp. 200-219.
Velmans, M. (2001), ‘Heterophenomenogy versus critical phenomenology: a dialogue with Dan Dennett’, http://cogprints.ecs.soton.ac.uk/archive/00001795/index.html
 As Torrance points out (in his note 2), some philosophers have tried to defend identity theory by arguing that Leibniz's law does not apply to 'referentially opaque contexts'. I might have a twinge in my knee and just not know it to be identical to neural-bodily state S, so I might conclude that they are not identical even though they are. Torrance is cautious about this argument and I share his caution. In some future state of neuroscience we can envision having a given experience C, knowing all about its physical correlates B, and still not being convinced of their identity (given Leibniz's law).
 Note that this block to establishing the ontological identity of conscious states with correlated physical states applies irrespective of the level of organisation of the physical states. That is, the block applies just as much to so-called “nonreductive physicalism” as it does to old-style reductive physicalism. Given this, it is not obvious how Van Gulick's suggestion that “higher level regularities might apply “in virtue” of lower level ones” would actually resolve the problem of causal closure.
 Whether more primitive forms of material organisation are associated with more primitive forms of conscious experience is a separate (controversial) issue that we need not address here. Panpsychists such as Penrose and Hameroff for example suggest that even microphysical events are associated with primitive experiences.
 In Velmans (1990) for example, I defend the conventional cognitive view that many mental states are unconscious and take issue with Searle's (1990) “connection principle” which explicitly links the criterion for being mental to being potentially conscious.
 This also addresses Chrisley & Sloman's point that, “We’d like to think that our conscious states have causal power by virtue of their being the mental states that they are, not by virtue of being identical with some physical state, which itself has, by virtue of falling under physical laws, the true causal power.” Indeed! And that is yet another reason for rejecting any version of physicalism. In the above analysis, conscious experiences are not identical to (correlated) physical states. Nor do they “supervene” on physical states (with the implication that the latter are ontologically more basic). They are first-person manifestations of the operations of our own psychophysical minds. They have causal powers in the sense that any phenomena can have causal powers. Although they only represent the operations of mind-itself (ur mind), from a first-person perspective we can take them to be the operations of mind.
 The need to have both first- and third-person accounts for a complete account of mind makes it clear why such accounts do not face the problem of “overdetermination” (see Chrisley & Sloman).
 Note that falsifiability is one useful criterion of a good scientific theory, but it is not an infallible criterion or the only criterion (other tests include verifiability, explanatory elegance, simplicity, sufficiency, productivity and so on). See Chalmers (1992) for a useful introduction.
 Chrisley & Sloman suggest (in their note 4) that I should say that “the physical aspect must contain at least as much information structure as the experiential aspect” rather than claiming them to have “identical information” as the physical aspect will typically encode more information than the experiential aspect. I do not deny that the brain encodes far more information than that which is manifest in conscious experience, or that this information may support the formation and functioning of the correlates of experience. However, information encoded in the brain that is not encoded in experience is not, in the strict sense that I intend, a “correlate” of that experience.
 As Chrisley & Sloman point out, it important to distinguish the functions that are implemented by a system from the methods it uses to implement those functions. They present this as an issue on which we disagree, suggesting that a strong phenomenal experience/neural correlate “mirroring” is implicit in my analysis. But, as should have been clear from TA note 10, this is actually an issue on which we agree.
 Standard measuring instruments include verbal rating scales, numerical rating scales, visual analogue scales and questionnaires such as the McGill Pain Questionnaire (Melzack, 1987).
 Rakover also complains that phenomenal measurements cannot meet the requirements of objectivity, publicity, and repeatability. I disagree. However this is a large topic on which I have written on extensively, both in this journal (Velmans, 1999) and in UC chapter 8. Given the limitations on space in this reply I ask interested readers to refer to these prior sources.
 If the physical differences can be consciously perceived we can say that information about physically measurable differences has been successfully “transmitted” or transformed into discriminable, phenomenal differences. Note that it is often possible for physical differences in stimuli to be detected in spite of not being consciously perceived (for example if subjects are required to guess). As this is tangential to the point at issue I will not elaborate on it here.
 Chrisley & Sloman go on to claim, “It is this need to distance himself from physicalism which raises the second problem with the analogy: he admits that the videotape and the screen are ontologically distinct, yet he was supposedly defending an ‘ontologically monist’ position! It seems Velmans ends up with the converse of the position for which he was aiming: ontological dualism, but epistemological monism (in the sense that strong assumptions are made about ‘informational mirroring’).” This confused analysis of the intent of my videotape/TV screen analogy needs some unravelling. It is true that conscious experiences and their neural correlates have distinct (phenomenal and physical) characteristics and in that sense may be said to have distinct ontologies. But this does not prevent them being aspects of an underlying, unified mind, thereby making my dual-aspect theory ontologically monist (in the tradition of Spinoza). Nor does the possession of distinct phenomenal and physical characteristics prevent experiences and their correlates from encoding identical information. The videotape/TV screen analogy provides one example of how representational systems can encode identical information without having identical characteristics. It should have been obvious that I did not mean to suggest that brain states are literally a form of videotape and experiences literally a kind of TV screen or that experiences can somehow be decoupled from their physical correlates! Nor does it make sense to interpret the view that one can know (or represent) one thing in two different ways “epistemological monism.”
 As it happens, a psychophysical theory relating information encoded in the brain’s electrochemistry to a pooled, integrated form of the same information encoded in the brain’s electromagnetic field has recently been proposed in this journal by McFadden (2002a,b). According to McFadden this EM field is the physical substrate of phenomenal consciousness (see also Pockett, 2002; John, 2002). While I am not committed to the details of this theory, and do not think it solves the “hard” problem (the EM field would still have to have dual-aspects to bridge the gap from physics to phenomenal experience), it does illustrate the type of theory that I have in mind.
 The fact that one has different (first- and third-person) forms of access to these (phenomenal and physical) aspects of mind does not alter the point that these aspects specify the mind's ontology.
 It is hard to know what Chrisley & Sloman mean by a “magnetic way of looking at the situation.” Unlike them, I do not confound the dual-aspect ontology of mind, or the way information is formatted within its phenomenal and physical aspects, with first- versus third person ways of examining the mind's phenomenal and physical aspects. Likewise, I do not confound the electrical and magnetic manifestations of electromagnetism, or the possibility of encoding information in either electrical or magnetic formats, with the different ways in which we can investigate electricity and magnetism.
 As I point out in TA note 14, I am only concerned here with the broader implications of dual-aspect monism. Consequently, it seems to me useful to suggest that there might be a psychophysical unity underlying the phenomenal and physical aspects of mind, that is broadly analogous to the electromagnetic unity underlying electricity and magnetism. It goes without saying that I am not suggesting that conscious phenomenology is magnetism, or that its physical correlates are electricity. The precise way that given conscious experiences map onto their physical correlates can only be discovered by neuropsychological research and, in this sense, “requires further work.”
 The appearance of the 3D phenomenal physical world is not of course identical to the more abstract world described by physics (quantum mechanics, relativity, string theory and so on). The relation of the phenomenal physical world to the world described by physics is central to a proper understanding of consciousness/material world relationship and I discuss this in depth in UC chapters 6 and 7. As this relationship is somewhat tangential to the issues raised in TA and the commentaries I will not elaborate on it here.
 Rao also raises a number of questions in passing, for example, what is meant by “information structure”, what encodes that information, how does “perspectival switching” work, and in what sense is information viewed from a first- and a third-person perspective complementary. As I have dealt with these issues earlier in this reply I won’t return to them here.
 In the Samkhya-Yoga tradition, Purusha refers to one's true, individual, immaterial essence (also referred to as Atman or soul).
 Rao writes that “Velmans speaks of direct and indirect knowledge, as the Indian theories do. … In Yoga theory, even the so-called first-person experience is indirect, because what the mind presents to consciousness are representations mediated by the perceptual and cognitive systems. Consequently, awareness arising from such mediation is also indirect. In other words, in Velmans, the direct acquaintance is with the representations, whereas in Yoga it is with the things themselves. Such direct knowledge results when the mind detaches itself from the sensory inputs and makes contact directly with the objects, events and processes in the universe. This is what may be labelled as paranormal process distinguished from the normal process in which there is the involvement of the sensory processes.” (p x) However, this doesn’t quite capture the similarities and differences between Reflexive Monism and the Yoga theory that Rao describes. In UC chapter 7 I develop the view that, under normal conditions, we have direct acquaintance with our own experiences but only indirect acquaintance with the things-themselves that such experiences represent. Given that normal experiences are representations, I agree with the Yoga view that they only provide indirect knowledge of things themselves, even when the things themselves that we experience are the operations of our own minds. Nevertheless, contra Kant, I argue that the thing-itself (including mind-itself) is knowable through the representations that we have of it, and the aim of both first- and third-person science is to achieve deeper, more complete knowledge. Knowledge can be gained through direct acquaintance, by experiencing that which one seeks to know, or indirectly, through the use of symbols (description, theory and so on). But it is only through direct experience that things become subjectively real for us (we real-ize what they are like). One can only really know love for example by real-izing what it is like to be in love. This, I think, gets quite close to the Yoga view, with the caution that I remain non committal about the possibility (often suggested in Yoga philosophy) that it is possible for embodied human beings to fully know (real-ize) the thing-itself as it is in itself, that is to have “direct” knowledge in the sense of knowledge that is “perfect and complete”. I also do not comment in UC on the nature and existence of paranormal phenomena.
 In this usage, mind-itself is that aspect of the thing-itself (the ground of being) that forms the basis of the manifest aspects of one’s own mind (i.e. its third-person, material and first-person, phenomenal consciousness aspects).
 Other confusions in Rao’s commentary can be traced to differences in use of terms combined with differences that arise from taking a first-person route to the nature of mind to be more primary than a third-person route. Rao for example claims that equating consciousness with phenomenal consciousness entails confusion “between the contents and the container, between substance and form.” I accept that, viewed from an Eastern first-person perspective, a form of pure contentless consciousness might appear to underlie everyday phenomenal consciousness, and the former therefore is viewed as the container of the latter. However, my own, somewhat different dual-aspect analysis does not confuse contents and container, or substance and form. Rather, the container is mind-itself and the suggested nature of this container is a little different. While I remain open to the view that with appropriate first-person training, the nature of mind itself appears as a form of pure, contentless consciousness, dual-aspect monism would suggest that even a conscious state as basic as this would have correspondingly basic, physical aspects that could, in principle, be discovered by empirical research. As the nature of mind-itself encompasses all its aspects it seems more accurate to describe it as psychophysical.
 My analysis of consciousness (in TA and in UC chapters 1-11) deals largely with phenomenal consciousness in humans and is consequently neutral about whether there is a first-person aspect (latent or manifest) in states other than those that actually have manifestations in (recognisable) conscious experience. UC chapter 12, however, is more speculative and considers the evolution and distribution of consciousness. It compares “discontinuity” theory (that consciousness appeared suddenly at a given point in evolution) with “continuity” theory (that the potential for recognisable consciousness was there from the beginning and evolved in form as matter evolved in form). Although little of my detailed analysis of consciousness depends on it, I argue that the latter is more intellectually elegant, and it fits more naturally into Reflexive Monism. The view that consciousness is a natural accompaniment of material forms also has implications for how one might think about the necessary and sufficient conditions for consciousness in the human brain. Rather than thinking of consciousness as something that is mysteriously added to representations at the focus of attention, it can be thought of a natural aspect of neural information processing as such. Why is it apparently absent in unconscious and preconscious processing? One possibility is that, in the evolution of complex brains with multiple sources of information, massive inhibition of information became a biological necessity to enable focus on information of greatest importance, and with it, inhibition of consciousness. On this view, unconscious and preconscious mental processes have inhibited consciousness. Conversely, information at the focus of attention is subject to release from inhibition (see Arbuthnott, 1995, for a review of the evidence, and the discussion of this and alternative theories in Velmans, 1995). Another possibility is that representations at the focus of attention are activated to a degree that masks any consciousness associated with other representations, rather like an orchestra on stage masks whispers in the audience.
 Many methods have been developed in both the West and the East for gaining conscious access to otherwise nonconscious aspects of mind, ranging from methods to aid recall of unconscious material in cognitive psychology and psychotherapeutic practice, close attention to the phenomenology of otherwise preconscious mental operations (Varela & Shear, 1999), meditative practices in Yoga and so on.
 Due to lack of available space I discuss the notion of “preconscious free will”, introduced in TA, in a later issue of JCS along with commentaries by Libet, Mangan, and Claxton. It is interesting to note, however, that Gray, and Chrisley & Sloman wholeheartedly agree that free will is preconscious as well as conscious, in spite of there being other aspects of my analysis with which they disagree.
 Although, following Rao, I have presented the Eastern view as idealist, it is important to note that there are as many differences in Eastern philosophy about these basic issues as there are in Western thought. The combined material and conscious nature of the thing-itself is well recognised, for example, in major, modern interpretations of Vedanta such as that of Aurobindo.