The Demotion of Consciousness

The Confounding Cognitive Course to Consciousness

As an evangelical evolutionist, I have found much frustration in learning the prevailing views in each branch of philosophy I've encountered -- in that they frequently embrace positions that are simply inconsistent with what we already know about the way humans developed and they way they function. Thankfully, there are also philosophers in each field who strive for scientific consistency, but they tend to be viewed as a radical minority -- probably because there is much about scientific accounts of reality that seems counterintuitive or unpalatable. In the quest to make truth conform to preference, I'm afraid hope springs infernal.

The study of consciousness is no exception. In fact, the view propounded by David Rosenthal suffers in this regard, though I sense in him no willful disregard for fact. It does seem, though, that some facts in plain view get overlooked in Rosenthal's understandable zeal to meet all challenges to his theory.

I suppose one could call Rosenthal's position cognitivist, in that it finds consciousness in thoughts. Creatures that only sense and perceive are perforce unconscious, and it is from thought alone (even if only of a rudimentary sort) that consciousness has emerged. On this view, the cognitive abilities we share (at least in part) with other higher animals is distinct from, and of a different kind than, the faculties of sensation and perception. To their credit, members of this faction typically believe that their theories are consistent with evolutionary theory, and would consider a lack of such consistency to be a liability. So, if they credit what I have to say here, it will be bad news for them indeed.

If I'm assigning black hats to Rosenthal and his ilk, the white hats belong to those whose theories are firmly rooted in evolutionary science -- who see consciousness as the product of more primitive faculties such as memory (e.g., Gerald Edelman) and sensation (e.g., Nicholas Humphrey). In addition to the Good and the Bad, however, there is a third group -- the Ugly, of course -- who indulge in a flagrant dualism (e.g., John Searle) that admits no mechanism for consciousness to evolve; unfortunately, I must limit myself here largely to dealing with one adversarial view -- and Rosenthal is by far the worthier opponent.

Though I'll touch on a number of others, there are three main objections I want to deal with regarding Rosenthal's theories: (1) Contrary to Rosenthal, the higher-order thoughts (HOTs) by virtue of which we are conscious of lower-order thoughts (LOTs) or mental states like sensations aren't thoughts at all in the way he means, but are rather sensory and perceptive in nature -- something he explicitly denies. (2) You can't get here from there. Consciousness necessarily preceded cognition (of the kind important to Rosenthal) in evolution. (3) There are simply better explanations for consciousness, and in the course of examining them, I've synthesized a nascent (but seemingly coherent) theory of consciousness to pit against his. For clarity's sake, I had better take these last two objections in reverse order. But, first, some preamble is required.

Governing Principles

I rely on some important assumptions about evolution. First, at the root of everything is the fundamental algorithm of evolution: random variation and interactive selection. I assume that every new development comes from old "raw material." Mutations hardly ever create new kinds; they duplicate or omit existing structures, or alter them by changing their developmental and/or functional characteristics.

For example, our brain shows clear evidence of having developed in three distinct stages: As vertebrates we began with what is now our brain stem and is often called the 'reptilian' brain; it governs our most primitive survival instincts, appetites, and responses. Then part of that structure was duplicated and the redundant portion developed into the limbic system, also called the 'paleomammalian' brain, which came to specialize in producing emotion and value, mediating learning and memory, and prompting species-typical social behavior. Another fluke 'budding' produced the antecedent to our cerebral cortex (a.k.a. the 'neomammalian' brain), which developed sophisticated sensory and perceptive abilities, came to direct voluntary movements and make plans for them, and eventually acquired conceptual and linguistic capacities.

Secondly, these changes happened incrementally, with each stage having a positive (or at least neutral) influence on reproduction and survival. Parsimony demands that we should look for accounts of consciousness and cognition that admit to intermediate stages -- i.e., that result from typical rather than anomalous evolutionary paths.

In the third place, the brain we have inherited operates, to the best of my knowledge, using a single process that has what we might call functional and metafunctional components. The brain's only function -- from the tiniest neuron to the entire, wonderfully complex system -- is to associate. Sensory neurons, for example, associate specific aspects of external objects with particular rates of their own firing -- by which medium other neurons can associate its output with that of other members in a neuronal group. These groups are linked into sensory maps or other more complex functional structures, and the associations among these units, within and between levels of organization, result in the full repertoire of brain functions -- sensation, perception, memory, and cognition in their many varieties and forms.¹

With the process of association the systems are modified, with connections growing stronger or weaker, some elements changing function, others dying off altogether. This metafunctional modification is so integral to the association process that one might just call the combined mechanism assocification (to coin a term). This should serve as a strong hint that the various familiar modes of cerebral functioning are very closely related (and should also provide insight into how they are related, as we'll see).

Finally, I have learned to be very suspicious of any explanation that appeals to human vanity. I know of no impediment to understanding more insidious and persistent than anthropocentric bias. The cognitive approach to consciousness, because its meaning inheres in that aspect of our minds that we think is most wonderful, is just such an explanation. It could be correct, but if incorrect would still command vehement and tenacious support from adherents whose self-image would be diminished by the admission. As we'll see, much evidence weighs in on the side of a less flattering explanation of consciousness.

HOT: Not So

What is Rosenthal's position on HOTs and consciousness? "Our hypothesis, therefore, is that a mental state is conscious just in case it is accompanied by a higher-order, nondispositional, assertoric thought to the effect that one is in that very state" (1993, p. 12). That is, the HOT is occurrent (not just waiting or 'tending' to happen under the right circumstances), and it asserts something like a belief (to the effect just mentioned). Rosenthal prefers that HOTs be noninferential, but he acknowledges (1992, p. 219, note 18) Gilbert Harman's claim that much of our knowledge comes from non-conscious inference -- which might even contribute to causing the HOTs that make mental states conscious. He also accepts Dennett's objection that sometimes we just don't know if our HOTs are based on inference (note 19), and says that in such cases we might simply be unsure whether the state in question is conscious! I think we might suspect that the state is illusory, but if we can so much as wonder about its metaphysical status, it is conscious.

But an important question has been raised here. What causes HOTs? Just prior to Rosenthal's responses to Harman and Dennett (note 16), he dropped an earlier claim that HOTs are created by the LOTs they are about, saying the claim was "unmotivated." In its place, he offered this: "HOTs are caused by a coincidence of mental factors, many of which are causally independent of the state in question." Isn't this just an appeal to occult qualities? I find no explanation of cause here.

Even if we accept the uncertain pedigree of HOTs, we still need to know their relationship to consciousness. Do they cause it, or merely accompany it? I find no satisfactory answer in Rosenthal's articles. In a footnote (1990a, p. 15, note 11) he explains why HOTs' being noninferential is important for avoiding certain objections: "In particular, one might object that higher-order thoughts do not confer consciousness on the mental states they are about when we come to have those higher-order thoughts by way of exteroceptive observation, or by way of theory." This sounds like a causal claim, but such an important issue shouldn't be relegated to a footnote!

The real force of Rosenthal's argument here goes not to causation but to the mere existence of HOTs. Since I'm committed to something like HOTs myself (differing with Rosenthal over their nature, not their existence), I have no need to take issue with this; in fact, the regular association between such states and the conscious states they are 'about' is good (though not conclusive) evidence of a causal relationship between them -- evidence I can use.

So what are HOTs, and how is it that they provide access to mental states (presumably thereby rendering them conscious)? "There are two broad ways of being transitively conscious of things," says Rosenthal (1993, p. 9). "We are conscious of something when we see it or hear it, or perceive it in some other way, and we are conscious of a thing when we have thoughts about it. Which kind of transitive consciousness is relevant here?" Rosenthal opts for the cognitive variety.

But is the sort of work that HOTs perform cognitive work? The typical HOT is of the form "I am experiencing mental state m." The HOT's job is to assert the existence of some other mental state, and to identify what that state is. If the HOT is not sensory or perceptive, then it requires knowledge of m in order to make either the assertion or the identification. But where might such knowledge come from? What epistemic modes can confer it? We must rule out empirical modes, for they require sensation and perception. That leaves inference, period -- or revelation, if we don't care much about epistemic strictures.

Well, this horse won't get much deader. Let's just note that our HOT's twofold job as expressed above is a fine working definition of sensation and perception.

Rosenthal's objections to what he calls the perceptual model of HOTs are unconvincing. He claims that "perceiving always involves some sensory quality," and that "the qualitative aspect of such higher-order perceiving would itself need to be explained in turn. Moreover, when a sensory state is conscious, we are conscious of the quality of that state, and not the quality of some higher-order perception of the state. And it's unclear how the quality of a higher-order perceptual state could explain our being conscious of the lower-order quality."

I think Rosenthal's problem is this: He has conceived of the HOT as separate from but connected to the LOT it is about, and he imagines that, if it is sensory, it must have a separate sensory quality from that of its related LOT -- or there must be a good reason why the qualities should be identical. Such a reason is readily available, but only if one conceives of the HOT as an extension of the LOT, part of the same active neurological circuit, and partaking of its sensory quality.

The BRR Theory of Consciousness

The view of consciousness I'm developing has key roles for memory, sensation, and perception; cognition turns out to be a combination of these processes. (An unexpected result of this theory of consciousness is that it appears to entail a theory of cognition.) Though it may sound strange at the outset, the general process leading to consciousness is this:

We are awash in sense data from both external and internal sources -- all of which generates mental states (in our sensory cortices, and perhaps in lower levels of our sensory association cortices). I suspect the lowest level of activity that should qualify as a mental state would be that of a neuronal group, whose response to a characteristic stimulus is to oscillate at a particular frequency (Edelman, p. 89). Such oscillations are the vocabulary of the brain.

Only a minute fraction of mental states concerning sense data comes to have phenomenal qualities (qualia, or "what it is like" for an organism to experience those states). This phenomenal subset of our mental states owes its existence to active stimulus provided largely by the reticular activating system (part of the reptilian brain), which somehow identifies which sense data is of importance to us (the theory explains how this might work), and has come to accept a certain amount of direction from higher cortical areas (Dennett, p. 274), conferring the ability to focus attention voluntarily. For convenience, I'll call the reticular activating system and its influences the attentional system.

Perception (of sense data from external sources or internal states) results from an automatic, neural reflex mechanism. When neural oscillations "about" such sense data reach a threshold (though it is not clear to me whether this threshold is one of amplitude, frequency, duration, or some combination) a signal is sent to the attentional system, which responds by returning its own signal. This response kicks the original oscillating system over another threshold (perhaps introducing some new feature such as harmonics of the original frequency), at which point the system develops a phenomenal quality -- a conscious one -- that is experienced in the mind. This exchange of signals should be recognizable as a simple feedback mechanism. The process that leads to conscious sensation, then, is built on the involuntary, automatic activity of the attentional system (hence the name of the theory: Basically Reflex Reverberation) which we learn to control to some degree.

Memory is just "the ability to repeat a performance" (Edelman, p. 102), and while an oscillation is in progress, there is memory of that particular process. Unless the oscillation is driven by external signals, it will eventually die out. If the circuit is stimulated again immediately, the short-term memory will still be there; if not, subsequent events may involve the circuit in different ways, erasing the memory of its erstwhile activity (much like recording new material over old on a cassette tape). If, on the other hand, the oscillation is driven over a prolonged period of time (about 20 minutes), the synaptic connections within the circuit will have become so well established that, even months or years after the fact, stimulation of some part of the original circuit will restore the system's oscillation in something much like its original state. In other words, we have long-term memory. This process of "burning in" short-term memories is governed by the organs of succession: the hippocampus, the basal ganglia, and the cerebellum (Edelman, pp. 105-8).

It is important to note some important relationships and differences among sense data, perception, sensation, memory, and consciousness:

Consciousness is active memory. This is the meaning behind the title of Gerald Edelman's book, The Remembered Present (and reiterated in Bright Air, Brilliant Fire). The physical underpinning of consciousness is reentrant mapping, the bidirectional linking of different neuronal groups, different sensory maps -- even different large-scale subsystems of the brain. It is the existence of neural paths in both directions that allow the structures in question to respond to characteristic stimuli with oscillation of a particular frequency -- the reverberation or feedback loops we have been talking about. But consciousness apparently requires a particular type of memory, as we'll see in a trice.

Sense data need not be conscious. Most of it persists briefly in short-term memory without phenomenal quality and without ever becoming conscious, then dies away.

Sensation is always conscious. For any reverberating mental state involving areas of sensory cortex, it is the activity of those areas that confers phenomenal quality on experience. And the phenomenality -- what the experience is like -- occurs consciously, or it isn't like anything at all. Rule: If it wasn't conscious, it was only sense data (or, in unusual cases, perhaps perception).

Consciousness always involves sensation. This is the audacious (and persuasive, I think) claim that Nicholas Humphrey makes (p. 205). Even states that involve no direct sensory stimulation require activity in some area of a sensory cortex in order to reach consciousness. Dreams, ideas, beliefs, and propositions, when we think of them, typically elicit visual or auditory imagery (such as an internal voice expressing a thought in question).

Sense data in short-term memory can become conscious. On occasion the attentional system suddenly attends to sensory stimuli that had been theretofore ignored. This meets what I think would be the minimum performance conditions to qualify as Ned Block's phenomenal consciousness (consciousness_p), a rather sweeping concept of consciousness that is best described by Block's example of drill noise: Emerging exactly at noon from some state of intense preoccupation, you realise that the sound of a drill outside has been impinging upon you for some time. Block says you were conscious_p of the sound all along, but at noon you also became consciously aware of it -- you had what he calls access consciousness (consciousness_a), the sort of consciousness to which you can actually respond purposefully. Most of us would hardly dignify Block's consciousness_p with the name; something is conscious now if we're aware of it now! But, if we weren't conscious of the sound in the first place, how can we possibly remember it?

The current theory helps to explain: At first, the sound of the drill created mental states in the auditory cortex which reverberated outside of phenomenal awareness, continually refreshed by the ongoing sound stimulus, until noon. Then the attentional system kicked the oscillating system into conscious awareness, first with perception of direct sound, then with memories of how the sound had been associated all along with conscious events. These memories were recalled when portions of the reverberating sensory system excited cortical areas that had been active during that period before noon, thereby sharing in common with the auditory memory a "time stamp" from the organs of succession (and possibly some direct links as well). No, memory need not be of conscious states; in fact, most memory is of short-term sense data that never enters our awareness.

Perception precedes sensation. As already described, perception of sense data is a step along the way to sensation. Humphrey (p. 104) thinks this mechanism evolved as an error detection strategy. A reflex response by the attentional system to incoming sense data verifies that the data is meaningful, and not just noise in the system or a hallucination.

Perception can be unconscious. The favorite example in the literature is blindsight: Subjects who have sustained injury to their occipital lobe experience a blindness in part (or all) of their visual field -- yet they can respond in a number of ways to visual stimuli, thanks to the activity of a separate, primitive visual system whose job is to orient us to stimuli entering our visual field. Humphrey tells of a monkey he studied named Helen (p. 88), whose blindsight allowed her to interact with her environment with what appeared to be normal effectiveness -- until she became upset, at which point she apparently tried so hard to see that she could no longer merely perceive. I'm not sure this is a good example, however, for no sensory cortex is normally involved with this primitive visual system; its function is perceptual, not sensory.

Humphrey (p. 79) cites the fascinating example of skin-vision: A 20 x 20 matrix of 400 tiny vibrators is placed over a 10" square of skin on a blind subject's back. The vibrators are mapped to the visual field of a TV camera so that the matrix represents the TV image "much as a newspaper photograph represents a scene by an array of dots." The subject points the camera by moving his head.

The results surpassed all expectations. With only a few hours' training, blind subjects learned to recognize a range of common objects.... Very quickly they developed the ability to point accurately to objects in space, and to judge their distance and their absolute size (independent of distance).... [Experimenter Paul] Bach-y-Rita quotes an experienced subject engaged in exploring the visual scene with the camera: "That is Betty; she is wearing her hair down today and does not have her glasses on; her mouth is open, and she is moving her right hand from her left side to the back of her head."

What do such subjects experience? If you ask them, they will remember that there was sensation on their backs; but they are normally not aware of it. They perceive objects and acquire information about them, but they apparently don't experience the objects in anything like the way a sighted person does in normal vision.

This suggests a plausible answer to Thomas Nagel's famous question, "What Is It Like to Be a Bat?" Humphrey doesn't miss it, though he waits 40 pages to bring it up: Maybe the bat just knows things about his environment, paying no attention to the auditory jumble of squeaks and echoes in his ears. Maybe being a bat is sort of like being a clairvoyant with dim vision and a case of tinnitus!

But one might wonder if such examples are consistent with a theory that finds consciousness in reverberant areas of sensory cortex. After all, the brain areas processing the tactile information for skin-vision or the auditory information for bats do not typically map things in the detailed three-dimensional way that vision uses -- so how can their activity account for vision-like experience? My best answer is that they probably don't. There is ample evidence that unused cortical areas are frequently coopted to take over brain functions that would normally have been handled by other (now damaged) regions. Since the visual cortex is available (and suited to the job), I think it likely fills the breach. It would then be reverberations there that make the experience conscious. Humphrey disagrees (p. 213): For the skin-vision subject, "when the vibrators tickle the skin of his back, he still represents 'what is happening to me' as being tactile in quality. It is true, as we noted, that he may in fact give all his attention to the perceptual channel, and so mask the tactile sensations altogether...." And if bats "are aware of anything 'happening to me,' it will be of having an auditory experience."

I need to clarify the pedigree of this theory. The original impetus came from Edelman's location of consciousness in memory (p. 207). His evolutionary account builds consciousness from the cellular level up, and is simply stunning. The other significant influence came when Humphrey found consciousness in reverberations of sensory cortical areas. With memory and sensation resulting from the very same oscillating systems, the unification of the two theories was an obvious step. But I disagree with Humphrey's characterization (though not really his description) of the "reaching out" to verify sense data by the attentional system. He personifies it overmuch, I think. "Call it a special sort of 'doing' if you like, call it being the 'author' of recirculating sensory activity" (p. 217). For Humphrey the process is an active, intentional (though unconscious) one performed by the individual. Though I think he'd have to agree with me that the mechanism is actually a reflex one, he would be uncomfortable with it. He already commiserates with the "Is that all?" reaction of critics who find the diminished role of conscious intentionality in his theory unpalatable -- and mine will obviously seem worse.

Other Issues

Explaining How

Functionalist theories aim to explain how consciousness arises, but are criticized for failing to explain why consciousness should result from the processes described. Humphrey frankly admits his to be an identity theory -- claiming that consciousness is a particular sort of physical process -- and the stronger claim meets with stronger objections that the 'why' issue is neglected. Humphrey offers (p. 225) what I think is a good criterion for adequacy of an identity theory (in explaining what consciousness is): "The body side of the equation leaves nothing undesignated that is designated by the conscious side, and vice versa." I hope that the BRR Theory at least offers substantial promise in that regard. Because it draws on the formidable support provided by Edelman and Humphrey, I'm optimistic. In a moment we'll consider the charges that an "explanatory gap" exists between such materialist theories and the phenomena they purport to explain. First, I need to finish with my third objection to Rosenthal: that there is a better account of consciousness than his.

The BRR Theory, like its predecessors, justifies its claims by matching changes in empirically (and publicly) observable neurological processes to the types and nuances of phenomenal experience. Given the necessary limits to the length of my explanation here, I can only hope that I've made a convincing case that the prospects are good for such a reconciliation. How, then, does Rosenthal's HOT Theory measure up? Well, he justifies his claims by matching changes in personal, subjective experience of cognitive states (whose nature is theoretical) to the types of phenomenal experience (not many nuances are explained). We can test the BRR Theory; we can only decide for ourselves whether the HOT Theory "feels right." Personally, I don't think the HOT Theory is even in the same league.

Explaining Why

Absent qualia: Ned Block argues that, if a functionalist theory is correct, it should be possible to duplicate the process mechanically and produce the same qualia found in humans. I think he's right, but this argument loses force in the present case, for several reasons: First, as far as I know, no one has attempted to model the BRR Theory mechanically -- so we have no way of knowing that it would not churn out the most vivid qualia imaginable. Secondly, how could we possibly determine for any such model whether or not such qualia were produced? By their very nature, qualia are inaccessible to all but the individual in whom they are instantiated. Thirdly, Dennett (p. 311) insists -- quite reasonably, I think -- that what he calls a Zimbo (a person who is normal in all respects save that s/he experiences no qualia) would experience a "user illusion": They would believe they experienced conscious phenomena anyway. Finally, Dennett strikes a telling blow against Rosenthal on this issue: "Now what we have just succeeded in imagining, it seems, is an unconscious being that nevertheless has the capability for higher-order thoughts. But according to Rosenthal, when a mental state is accompanied by a conscious or unconscious higher-order thought to the effect that one has it, this ipso facto guarantees that the mental state is a conscious state!"

Why are there qualia at all? Kobes posed a pertinent question in the seminar: "How can constituent representations in a phenomenal world of experience get their content from one another and not from The World?" For lack of space here, I can only outline the evolutionary response: As Hume noted, we have no direct access to external reality. We must rely on sensory information if we are even to form ideas about such things. Biological evolution is largely a story about how organisms developed the means to get and use information about the environment. Part of that story is how the world of experience for organisms is wholly internal (even if "semantic transparency" gives the illusion that we are dealing directly with external objects) and representational. But why should there be anything that "it is like" to exist in that interior world?

For many simple organisms, there is no such experience; they simply have reflex responses to incoming information. Somewhere along the line, however, the system ran up against functional limits: Too many conflicting inputs demanded too many reflex outputs -- the door was closed to further "same direction" development.

I have already described how the paleomammalian and neomammalian brains were 'spun off' from the reptilian brain. They process information in the same Gestalt 'chunks' that the reptilian brain did -- but they take as their inputs the preprocessed, encoded, representational information produced by lower brain structures from raw sense data. Metaphorically, where the reptilian brain produced something like alphabetical letters to represent its environment, the newer structures were forming words, sentences, and paragraphs. Entirely new dimensions of informational organization and complexity became available.

Note that the alternative to developing higher-level Gestalt processes is to develop higher-level logical processes. The difference between the two approaches is expressed by the old saw, "A picture is worth a thousand words." Logical processes would need to describe instead of merely represent things -- a job that is orders of magnitude more difficult. Yet this is the job that Rosenthal requires for the emergence of consciousness. (Yes, we've at last come to my second objection.)

Cognitive processes are distinguished by their internal structure, not by the unified 'face' they present. Thoughts have syntax as well as semantic meaning, whereas perceptions and sensations have the latter only. It is possible, by combining these more primitive processes through the reentrant mapping Edelman explains, to create processes that are truly cognitive in nature and function. But this level of organization had to emerge long after creatures became phenomenally conscious. To insist on cognitive organization as a precursor to consciousness, as Rosenthal does, is to place one's buggy in a pre-equine position....

Explaining Cognition

Cognition deals in propositions and concepts, while the BRR Theory deals in sensations and perceptions. The preceding paragraph outlines how the one group derives from the other, but a bit more detail might be in order. Each level of neuronal organization has a semantic function, and when these are combined in mappings with syntactical structure, we have the stuff of thought. What is also needed is some sort of guidance to order our train of thought -- and this is where the interaction among cognitive forebrain structures, value assigning limbic structures, and the attentional system comes into play.

Discrete concepts emerge from neuronal structures that have many loose links with other areas. Stronger links produce associations among concepts that come to be experienced as related. Relational ideas come from the neuronal systems that, through links to many cognitive areas, have 'recognized' similarities between concepts or other ideas. Such recognition is not miraculous, for the brain is (of course) a selective mechanism: it directs similar input to similar systems. Associations of common conjunction can be very simple and direct (the idea "isoceles," for example, is hardly separable from the idea "triangle") or more remote ("morning newspaper" may be associated with "breakfast" by a common time-stamp from the organs of succession).

Let's look at how a typical sensory mental state can become conscious food for thought: The reverberating neural network (involving parts of the sensory cortex, the organs of succession, and many various tributaries running to brain areas that have been linked by past memory to similar activity) passes the threshold of perception, activating the attentional system, which sends back signals that boost the reverberation further, and which also sends signals that (1) kick into action the appropriate areas in the forebrain (those with memory-mapped links to active sensory areas) and (2) engage a structure in the left forebrain called 'the interpreter', whose job is to ceate coherent 'stories' about what is happening to us (Gazzaniga, p. 230), in making semantic sense of the state in the context of memory-encoded dispositional networks (that store beliefs, concepts, etc.), along with other incoming sensory information (including reports of behaviors being undertaken, many automatically, in response to current conditions). Note that the perception is automatically triggered by reverberation surpassing a threshold; that the "reaching out" by the attentional system is reflex and determinate; and that the interpretation reached -- including content equivalent to "I am experiencing m," is wholly determined by the inputs to the interpreter. The phenomenal aspect of the experience reflects the precise mix of sensory circuits currently resonating above the threshold of sensation.

The BRR Theory also explains dispositional states like beliefs. They 'inhabit' rather complex neural systems that link the brain areas related to all relevant concepts, ideas, supporting beliefs, and values in a single oscillating system. When actively reverberating, they express content something like "I believe that P." When quiescent, they may still be considered to have content, but they don't express it. To get an idea of how complex they are, consider that if they are about objects in any way, then areas of sensory cortex for all modalities by which those object have been represented will be part of the system. If they involve values (and they almost certainly will), they will resonate into limbic regions. They will involve representations in memory for actions motivated by that belief in the past. And the list goes on....

I hope these examples are sufficient to show how others can be explained by the theory.

Parts and Parcels

Kobes proposed in the seminar that when a mental state token m is conscious in (subject or subsystem) S, there is a HOT in S with content "I am in mental state m," such that the state token m is a constituent part of the HOT token. Put another way, m is both mentioned (pointed to) and used (incorporated into the HOT). The "I am in" part of the HOT is functionally implicit, while the "m" part is explicit -- that is, there is no "I" subsystem that is trying to function as the audience in a Cartesian theater of the mind. This is very much the way BRR Theory would have it: The conscious content of experience is something like "m happening here," where "m" and "here" both refer to the activated neurological net that is m, and "happening" refers only to the occurrent nature and temporal width of the experience.

Rosenthal takes a somewhat different stance (1986, p. 469): "[T]he only way for a thought to be about a particular mental state is for it to be about somebody's being in that state. Otherwise, the thought would just be about that type of mental state, and not about the particular token of it." What set of members would constitute the type of state here? All similar states that might be experienced by humans (and/or other sensible creatures), with different tokens belonging to different individuals? This would require conceptualization far beyond the capacity of, say, infants who clearly have such HOTs. I think Rosenthal has the right idea here, but the wrong reason for it.

Conclusion

Obviously, though I have my differences with Rosenthal (and think they deserve to be addressed), my main interest has been in exploring and presenting my BRR Theory.

I do think the BRR Theory has a lot to recommend it, and it has been a real pleasure to see how well it addresses the many tricky clinical cases and 'science fiction' scenarios in the consciousness literature; I only wish there were room to discuss more of them.

Arbib, Michael A., Brains, Machines and Mathematics, New York: McGraw-Hill, 1965.

Armstrong, David M., "What is Consciousness," in The Nature of Mind and Other Essays, Ithaca, NY: 1980.

Carlson, Neil R., Foundations of Physiological Psychology, 2d Ed. Needham Heights, MA: Allyn and Bacon, 1992.

Dennett, Daniel C., Consciousness Explained, Boston: Little, Brown and Company, 1991.

Edelman, Gerald M., Bright Air, Brilliant Fire: On the Matter of the Mind, New York: Basic Books, 1992.

Gazzaniga, Michael S., Mind Matters: How the Mind & Brain Interact to Create Our Conscious Lives, Boston: Houghton Mifflin Company, 1988.

Humphrey, Nicholas, A History of the Mind: Evolution and the Birth of Consciousness, New York: Simon & Schuster, 1992.

Jackson, Frank, "What Mary Didn't Know," in The Nature of Mind, David M. Rosenthal ed. New York: Oxford University Press, 1991.

Levine, Joseph, "On Leaving Out What It's Like," in Consciousness, Martin Davies and Glyn W. Humphreys eds. Cambridge, MA: 1993.

Nelkin, Norton, "The Connection Between Intentionality and Consciousness," in Consciousness, Martin Davies and Glyn W. Humphreys eds. Cambridge, MA: 1993.

Rosenthal, David M., "Two Concepts of Consciousness," in The Nature of Mind, David M. Rosenthal ed. New York: Oxford University Press, 1991.

Rosenthal, David M., "Thinking That One Thinks," in Consciousness, Martin Davies and Glyn W. Humphreys eds. Cambridge, MA: 1993.

Van Gulick, Robert, "Understanding the Phenomenal Mind: Are We All Just Armadillos?" in Consciousness, Martin Davies and Glyn W. Humphreys eds. Cambridge, MA: 1993. I regret that I cannot at this time provide complete citations for these sources::

Block, Ned, "On a Confusion about a Function of Consciousness," 1994.

Dretske, Fred, "Conscious Experience," 1993.

Rosenthal, David M., "The Colors and Shapes of Visual Experience," 1990.

Rosenthal, David M., "Explaining Consciousness," 1993.

Rosenthal, David M., "Why Are Verbally Expressed Thoughts Conscious?", 1990a.

Van Gulick, Robert, "Consciousness, Intrinsic Intentionality, and Self-Understanding Machines," 1988.

Van Gulick, Robert, "A Functionalist Plea for Self-Consciousness," 1988a.

¹Edelman, 1992. The chapter "Neural Darwinism" provides an overview, taking its subject matter and its name from the second volume in the Nobel laureate's trilogy on morphology and the mind (the others being Topobiology and The Remembered Present).