Back to CQ Homepage

Awareness, Mental Phenomena, and Consciousness

A Synthesis of Dennett and Rosenthal

Teed Rockwell

2419A Tenth St

Berkeley, CA 94710

510/ 548-8779 Fax 548-3326


Daniel Dennet has shown, perhaps more than anyone,that philosophers have something to offer to the scientific enterprise. Other philosophers have followed Quine's exhortation to naturalise epistemology, and have carefully studied a variety of scientific specialities. But often these other philosphers of science have implied that such naturalization might require the de-emphasis of philosophy as a separate discipline. Dennett has maintained, even strengthened, his unique identity as a philosopher by demonstrating that even though philosophers do not have a domain of facts they can call their own, they do have two skills which give them fresh ways of approaching everybody else's facts.

1) Philosophers are trained to analyse concepts, to make distinctions which clarify the way ordinary language thinks about fundamental things. Dennett was one of the first philosophers to apply these skills not just to ordinary language but also to the discourses of the various sciences. In "Time and the Observer" (Dennett and Kinsbourne 1992), for example, he revealed that certain scientific experiments were designed to answer a question which was based on an incoherent concept. (the Stalinist-Orwellian distinction) A scientist within the field could have discovered this, of course. But having the ability to design and implement a good laboratory experiment is a different skill from being able to analyze the concepts it presupposes, and it is likely that this pseudo-controversy could have gone on for years if Dennett had not dissolved it.

2) Philosophers are trained to maintain what Wilfrid Sellars called "the eye on the whole". This makes them, not uniquely qualified, but more skillful than most , at coming up with grand syntheses that unify facts from a variety of disciplines in which no one person could possibly acquire expertese. While conceptual analysis can help identify confusions that have inhibited scientific progress in the past, conceptual synthesis points out directions for scientific progress in the future. Such a syntheses is Dennett's Multiple Drafts Model of Consciousness (MDM), of which he says "I am sure there are still plenty of mistakes . . . and I hope they are bold ones, for then they will provoke better answers by others." (Consciousness explained p. xi) In this essay, I will talk about some of these mistakes, in hopes of clarifying what I think is a fundamentally sound theory of mind with many profound implications. One mistake in particular has had far more effect on people's ability to understand Dennett's points than upon the soundess of the theory itself: the assumption that a particular piece of analysis is an essential part of Dennett's grand synthesis.

Many people were first introduced to Daniel Dennett's Multiple Drafts Model of Consciousness through the Behavioral and Brain Sciences article Time and the Observer, which Dennett co-authored with Marcel Kinsbourne, and which became the heart of chapters 5 and 6 of Dennett's book Consciousness Explained. This caused some confusion, because although the MDM is discussed briefly in Time and the Observer , the primary subject matter of that article was a piece of conceptual analysis ( i.e. of the invalidity of the Stalinist-Orwellian distinction) which has no essential connection to the synthetic far reaching theory which is the MDM. Dennett believes that there is such a connection, but this is because he does not acknowledge what I consider to be a significant distinction between three aspects of our concept of consciousness.

Patricia Churchland once described the problem of consciousness as the question "How do you get experience out of meat?". In this paper, I am going to show that this question actually fragments into (at least) three different questions.

(1) How do we get a unified sense of self (out of meat)?

(2)How do we get awareness (out of meat)?

(3) How do we get mental phenomena (out of meat)?

It is my claim that The MDM works best as an explanation for how 1) and 2) are closely related, and that it does so in ways that are revolutionary and profound. The Stalinist-Orwellian distinction, however, is most effectively dismissed by considering question (3), without any reference to questions 1) and 2). Unfortunately, this gets obscured by Dennett and Kinsbourne's decision to use the Stalinist-Orwellian distinction to introduce the MDM. This caused both Dennett and Kinsbourne and their critics to talk past each other on many an occasion, and made it difficult to consider either the Stalinist-Orwellian distinction or the MDM on their own respective merits. For example, when Damasio insists that "a satisfactory model of consciousness should indicate how the dis-integrated fragments operate to produce an integrated self" ( Time and the Observer, p.208) and Velmans asks in the title of his commentary "Is Consciousness Integrated?" (Time and the Observer, p.229), they are really dealing with question (1), which Dennett and Kinsbourne made no real attempt to answer in Time and the Observer, . When separated, the arguments for the MDM ( in other sections of Consciousness Explained), and the arguments against the Stalinist-Orwellian distinction are both accurate and important. But taken collectively as a defense of a single position, they trip each other up unnecessarily.

I do not mean to imply that separating the question of consciousness into these three parts is easy or straightforward. Common sense does not make this separation, and careful analysis and phenomenology may be required to clarify it if it is not intuitively obvious. In most ordinary speech, consciousness, perception and awareness are considered to be more or less synonymous. The sentences "I am conscious of X" "I perceive X" and "I am aware of X" would be assumed to be for all practical purposes interchangeable. But philosopher David Rosenthal formulates a variety of subtle distinctions that give greater precision to these folk psychological concepts. Among other things , he points out that the processes that make us aware of perceptions (which Rosenthal calls Higher Order Thoughts or HOTs ) are distinct from the processes that constitute the perceptions themselves. In his commentary on Time and the Observer, , he points out that "It is a familiar Cartesian doctrine that being conscious is part of what it is for a state or a representation to be mental," and invites us to consider the possibility that the conscious and the mental could be distinct. (see also Rosenthal 1986, 1989, 1990a, b 1993). I will rename his distinction between the mental and the conscious, referring instead to Mental Phenomena and Awareness , because I believe both of these elements to be essential parts of what we call Consciousness, and I think it can be misleading to call only awareness consciousness. By "Mental Phenomena" I mean those entities that are thought of colloquially as inhabitants of our subjective space: perceptions, sensations, emotions etc. I may or may not wish to include things like thoughts and beliefs in this category in other contexts, but for our critique of the Stalinist-Orwellian distinction we will be concerned only with perceptions. My claim (Which I see as a paraphrase of Rosenthal's claim) is that the problem of consciousness cannot conflate the two distinct problems I numbered (2) and (3) above. Almost everyone who has dealt with philosophy of mind has been struck by the mystery of the question of how light rays, retinal firings, and various brain states manage to produce our experiences of light, shape, and color. (question 3). This is a tremendously difficult question, and many people feel that answering it is the sole challenge of the problem of consciousness. But if we reject the Cartesian idea that all mental phenomena are directly and unmistakenly presented to us simply because they are in our mental space, we are also forced to come up with an answer to another very different question. Once our brain does manage to create shapes, colors, sounds, smells, etc. how does it manage to become aware of some of those mental phenomena and ignore all the others? Why are we aware of only some of these things at any given time, and how do we shift from phenomenon to phenomenon?. I believe Rosenthal's HOT theory is the most affective way of dealing with this problem, if it is extended in certain ways that may not be obvious at first glance. If we are willing to assume that HOTS are not necessarily linguistic or symbolic, (a possibility Rosenthal has conceded in conversation), they can be seem as paradigmatic examples of acts of attention: mental acts of particular sort that make awareness possible. Some may not feel the need of such a theory if they accept the widely held assumption that attention is non-declarative and non-assertive, following spontaneously and inevitably from stepping on the stage of the Cartesian Theater. For those who accept that assumption, I would suggest considering some of the following phenomena, which cry out for the sort of explanation that Rosenthal's distinction makes possible.


The Distinction between

Perception and Awareness

It seems undeniable that at any given moment I am perceiving many things of which I am not aware. Because I can focus my concentration on different things in my visual field, as well as shift my awareness from what I see to what I hear, or what I feel or what I smell. It also seems that the various perceptual processes continue to function, even when I am not aware of them. Dennett gives examples that illustrate this.

Are you constantly conscious of the clock ticking? If it suddenly stops, you notice this, and you can say right away what it is that has stopped; the ticks "you weren't conscious of" up to the moment they stopped and "would never have been conscious of" if they hadn't stopped are now clearly in your consciousness. An even more striking case is the phenomenon of being able to count retrospectively in experience memory, the chimes of the clock which you only noticed was striking after four or five chimes. But how could you so clearly remember hearing something you hadn't been conscious of in the first place? (Consciousness explained 137-138)

Dennett refers to these as "mundane and familiar examples, often discussed by philosophers" (p. 137), So it seems that he acknowledges as a heterophenomenological fact that perception seems to go on without our being aware of it, and that awareness can apparently be shifted to and from different acts of perception. It also seems that we can shift our awareness to and from other mental phenomena, such as emotions, thoughts and so on. Sometimes awareness may shift involuntarily as well. Patricia Churchland points out that most of the time we are not aware of what our tongues feel like, but when you read this sentence it is likely that you will become aware of your own tongue without making any effort to do so. (lecture at Stanford University 1993) The spiritual traditions of Hinduism and Buddhism value the ability to focus awareness voluntarily , and have developed exercises to strengthen it. During meditation, it is very difficult to keep one's awareness on a single perceived object (such as one's breathing, or a mantra). What is it that is going on neurologically during meditation, when one's awareness drifts away from one's breathing and then is pulled back again? What part or function of the brain has been developed when one becomes good at this, and is able to focus on a single object for long periods of time?

These questions are given even greater urgency by the fact that neuroscience now tells us that the brain is a massive parallel processor, producing far more mental phenomena than we can be aware of at any moment. Whatever it is that shuts out all of the phenomena in our brains that are temporarily outside of awareness is obviously performing a function that needs to be explained. Consider also the famous experiment that Dennett cites on p.324 of Consciousness Explained, which reveals the blindspot in normal vision. (Figure 1) The only thing that makes the blind spot noticeable is that the experimental procedures gives the subject two separate instructions, which presuppose a distinction between perception and awareness: 1) look directly at the cross, but 2) notice that the dots in your peripheral vision disappear. This distinction is necessary to the experiment because the blind spots are on the fringes of the visual field. Consequently, they can only be noticed if the subject's visual field is centered on one area, and her awareness is to some degree centered somewhere else (i.e. on the visual periphery). If this diagram was shown to a meditator who was given only the instruction "focus your awareness on the cross", that meditator would not separate his awareness from his perception, and probably not even notice the blind spot at all. Surely there is a need to explain why these two functions of looking and noticing appear to be so different from each other, and no theory of consciousness would be complete without attempting one.

Dennett does acknowledge this distinction at least peripherally: He recognizes that we have the ability to probe into and focus on isolated regions of our experience (Consciousness Explained p.138), but gives no non-metaphorical explanation of what a probe is. (we obviously do not probe consciousness with a stick). The fact that Dennett feels the need to use this metaphor shows that he acknowledges that there are mental activities describable as probing which are essential to awareness, and that these activities are somehow distinct from the mental activities that create the contents being probed. This makes necessary a distinction which must be similar in some ways to the one outlined by Rosenthal.

In Time and the Observer , Dennett and Kinsbourne devote themselves to the more traditional question of how mental phenomena (in particular, sensations) are created by the brain. They offer some ingenious answers to parts of that many-sided question, and dissolve some philosophical puzzles that have seriously hampered scientific research. But nowhere do they deal with the question of awareness as distinct from perception. Furthermore, as far as I can tell, the MDM has no essential relevance to the question of how mental phenomena are constituted, although it does have some profound and revolutionary things to say about awareness and self-consciousness. To demonstrate this, I will examine Dennett and Kinsbourne's reinterpretations of the experiments described in Time and the Observer and show that the invalidity of the Stalinist-Orwellian distinction can be accepted without making any reference to the multiple drafts model at all. We can then examine what Dennett has to say about awareness in other parts of Consciousness Explained, and evaluate that part of his theory on its own merits.

Dennett and Kinsbourne on Perception and the Stalinist-Orwellian distinction

Dennett and Kinsbourne introduce the Stalinist-Orwellian Distinction by considering the Color Phi phenomenon discovered by Psychologist P.A. Kohlers. When two stationary dots of different colors appear one after another very fast before our eyes, we have a sensation of a moving spot switching in color from red to green in mid-trajectory. The puzzle is: how does the brain know to put a green spot half-way between the red spot it has already seen and a green spot it hasn't yet seen ? In Time and the Observer , Dennett and Kinsbourne discuss (and disprove) two frequently proposed explanations for this fact, which they label the Stalinist and the Orwellian. These names come from comparing how totalitarian governments lie to how illusions are created by consciousness. A Stalinist illusion is created by altering the experience itself. (the way Stalin would create "show trials" of people he wanted to execute to "prove them guilty". ) An Orwellian illusion is created by changing or eliminating the memory of an experience without altering the experience itself at the time of its occurence. (the way the ministry of truth would rewrite history in Orwell's 1984.) An Orwellian show trial would not have to take place at all, it would only be necessary to rewrite the history books so that they said a trial had taken place, even if it hadn't.

When we talk of events that take place over several days, hours, or even minutes, this distinction is real. Dennett points out that if a mad scientist wanted you to mistakenly think you had seen a woman in a hat at a party last week, he could do this by either creating an hallucination of the woman at the party itself , or by tampering with your memory of the party the next morning. The question of which of these two methods he used is a real one, and is in principle answerable empirically. (If someone found a note in your handwriting at the party saying "Who is that lovely woman in the hat?" it would count as evidence for a Stalinist illusion). But if you are talking about two experiences that follow each other within milliseconds, (such as the experiences of the two dots that make up the color Phi phenomenon) this distinction is meaningless. Nevertheless, Scientist do argue over which of these two explanations is peferable, without realising that "this is a difference that makes no difference".

Where the Stalinist theory postulates. . . an unconscious detection of a red spot, the Orwellian theory postulates a conscious experience of a red spot that is immediately obliterated from memory by its sequel. ( Time and the Observer, p.193).

Dennett and Kinsbourne point out that this distinction is only necessary at the millisecond range if we posit the existence of some kind of Cartesian theater that acts as a finish line or charmed circle of consciousness. If there is a Cartesian theater, there must be a fact of the matter as to whether or not a certain sensation ever occurred within that theater. If there is no Cartesian theater, however, the distinction between Stalinist and Orwellian explanations of the Phi phenomenon "lapses at close quarters"(Time and the Observer 194).

Dennett and Kinsbourne claim that this lapsing is a consequence of one of the fundamental principles of the Multiple Drafts model i.e. that "if one wants to settle on some moment of processing in the brain as the moment of consciousness, this has to be arbitrary." (Time and the Observer, 194). Here we can seen where the conflation of mental phenomena with awareness into the single word "consciousness" causes confusion. This statement is almost true about awareness, if we substitute the word "subjective" for arbitrary, and clarify the differences between those two frequently confused terms. (More on that later.) But the Stalinist-Orwellian distinction can be dissolved without saying anything about awareness itself, because the lapsing at close quarters can be explained by citing two other points about the contents of awareness (i.e. mental phenomena): 1) mental phenomena are emergent network properties of certain neural events that take a certain finite amounts of time to occur. 2) This emergence requires different amounts of time for different types of mental phenomena. Given these facts, we can dissolve the Orwellian-Stalinist distinction without making any reference to the MDM.

When we try to break sensations down to neural events that take less than a certain threshold amount of time, we are no longer dealing with experience, we are dealing with meat. Dennett points this out with another example in his 1992.

. . .There is not a first instant in which you are conscious of a blue disk. The content blue is actually determined by your brain after the content circle and the content that these two contents go together is something that happens fractionally later still. But of course your experience is not one of first realizing that something's happened, then realizing that there's a circle shape, then realizing that it's blue, or that there's blue in the world, then and finally realizing that the blue goes with the circle; You can't distinguish that sequence, but in fact that sequence is occurring in your brain. (p.8)

Because we can't experience these kinds of sequences of brain events, it becomes possible for the brain events that represent our experience of time at the neurological level to occur in a different order from the time we are experiencing. This is because those brain events are not themselves experiences, even though they are constituents of an experience. The Cartesian theater thus creates Stalinist-Orwellian pseudo-distinctions not by giving an unjustified precision to consciousness, but by making a category mistake, in fact the same kind of category mistake that Ryle outlines in the first section of The Concept of Mind. Philosopher Michael Lockwood (1993) explicitly articulates this mistake in a way reminiscent of Ryle's hypothetical visitor to Oxford: He looks at the collection of neurological processes that constitute a sensation and asks which one is the "moment of consciousness" . Dennett's reply is very Rylean; he says that no one of these processes can be that moment, because all of them are. This is also the answer to the question "which building on the Oxford campus is the university?"--"No one of them is, because all of them are." Oxford is made up of parts which are not themselves universities, and mental phenomena are made up of neural events that are not themselves mental phenomena. Claiming that the Stalinist-Orwellian distinction is always valid, even at close quarters, is thus rather like insisting that every part of Oxford university, all the way down to the Dean's cufflinks, is also a university. It is true that experiences can be divided into parts which are themselves experiences, just as Oxford University can be divided up into colleges. But both experiences and universities have to be carved at exactly the right joints to make this happen, and in the case of experiences these joints apparently produce chunks of varying temporal size.

The old empiricist vision of experience as being made up of identical bite size "sense data" was significantly misleading. Within that vision, it seemed inevitable that because one dot-sensation is an experience, two dot-sensations must be two such experiences following each other across a Cartesian stage. But modern neuroscience has taught us that perceptual experiences are not simple atomistic givens, but rather complex neurological constructs. Once we accept this, there is good reason to accept that these constructs could be of varying temporal lengths. The experience of seeing one colored dot obviously takes less time than the experience of seeing the dot in motion, yet the "dot-in-motion" is apparently a whole which cannot be divided into individual "dot-sensations" without producing Stalinist-Orwellian paradoxes. Common sense wouldn't have predicted this, perhaps, but experiments like the color Phi and the Cutaneous rabbit indicate that consciousness is less like a stream and more like a stew, with large chunks of experience of varying size, each of which must be divided in different ways if one is to avoid slipping back into "meathood". The visual experience caused by a single disk reflecting light onto my retina can be separated from the experience caused by another disk two seconds later. But the sequence of neurological events that make up the single disk experience cannot be divided into smaller experiences; it’s parts are neurological, not experiential. The surprising thing we have learned in the laboratory is that when the two disks follow each other quickly enough, they become experientially indivisible in the same way that the single disk is experientially indivisible. It thus becomes possible to have the kind of subexperiental "juggling" that Dennett and Kinsbourne propose to explain the color phi phenomenon (as well as the cutaneous rabbit, and the other examples he gives). Because there is no necessary connection between the order of neurological events and the order of experienced events once we get down to the subexperiental millisecond range, there is no temporal paradox if under unsual laboratory situations these two timelines occasionally diverge.

But this fact does not imply that there is a vague answer to the question "when did I become conscious of some particular event", it only implies that there is a finite answer, and that it is an answer that is different for different kinds of experiences. The sum of the various neural events that constitute the experience of a given blue disk could take exactly 28.37 milliseconds every single time it occurs, and it would still be possible to juggle the contents of that experience within its time frame. The color phi phenomenon could also take exactly the same amount of time whenever it is experienced (say 538.34 milliseconds), and the Stalinist-Orwellian distinction would still be irrelevant to it, once we acknowledge that Color Phi is experientially a whole and does not consist of seeing two discs one after the other very fast.

Note, also, that this solution to the problem takes place without any reference to awareness at all. All of our discussion has been concerned with how mental phenomena are constituted: how neural events that have contents like "blue" and "circle" managed to cohere in such a way as to produce a sensation of a blue circle. But both Dennett and Rosenthal have given us good reasons for believing that the parts of the brain that create sensations like blue circles or ringing chimes are distinct from those parts which perform "probing" functions or generate Rosenthal's higher order thoughts. Dennett's reply to Lockwood is correct; the blue circle in the meta-contrast experiment has no leading edge in perceptual memory. A laboratory subject will be aware of the entire blue disk as a whole--If he focuses awareness on the blue disk. But the issue of awareness as distinct from perception is completely sidestepped, not only by Dennett's explanation, but by most experimenters as well. It is always assumed that the subjects in any experiment are going to be aware of the perceptions under consideration, because they are being paid to do so by the laboratories running the experiment. If one of the subjects did lose awareness of the perception that was being studied, he would have said something like "I'm sorry, I spaced out for a minute, could we do that one again?", and the whole previous run would be dismissed as "noise". So almost all such experiments assume that perception and awareness are operating in harness together, and it becomes very easy to forget that they are capable of operating independently. Dennett does however deal with the question of awareness in other parts of Consciousness Explained, and in the next section I will discuss those.

Probing for the Thimble: Dennett on Awareness

Dennett deals with the distinction between perception and awareness most explicitly in chapter 11 section 3 of Consciousness Explained. Here he gives an example that illustrates the distinction between perception and awareness, and offers an explanation for it. Dennett points out that when children play the game Hunt the Thimble, the whole point is to become aware of what you are already perceiving. The thimble is in plain sight, close enough to bite you if it were a bear, and yet you are not aware of it until that distinctive "aha!" experience occurs. Unlike the experiments discussed in Time and the Observer , the distinction between perception and awareness is crucial here, for it is the very thing that determines whether you have won the game or not.

Dennett's explanation for the hunt the thimble phenomenon requires him to posit a distinction which is both significantly similar and significantly different from the Rosenthalian distinction between awareness and mental phenomena. He acknowledges that there is a difference between 1)being in the foreground of experience 2) being in the background of experience. This distinction bifurcates consciousness itself, and is separate from 3) being merely in the background of the perceptible environment, without being experienced at all. The fact that Dennett acknowledges that there is a difference between 2) and 3), and that he considers both 1) and 2) to be distinct aspects of consciousness, prompts him to make a distinction which is quite similar to Rosenthal's.

Dennett claims that in order for a perception to be in the foreground of my experience, there must be special relationships of intentionality or aboutness which holds between it and some other of my mental states. Before Betsy sees the thimble in this sense, " Even if some representational state in Betsy's brain includes the thimble, no perceptual state of Betsy's is about the thimble yet ". Dennett describes these kinds of intentional relationships as "feed-back guided, error-corrected, gain adjusted purposeful links". Such links require continuing activity to maintain, and make it possible for those perceptions to "serve as a hinge for policy. . . Once I have seen something in this strong sense, 'I can do something about it' or do something because I saw it or as soon as I saw it. ( Consciousness Explained p. 334-335) When Betsy sees the thimble in this sense, not only can she go sit down with the other children, she can decide whether she thinks it is a pretty color, notice whether or not it has a dent on the facing side etc.

This explanation seems to do justice to the phenomenology of awareness described in the first section of this paper, and also seems to fit well with the multiple drafts model as a whole. When a meditator concentrates on a mantra, it is what Dennett calls "an achievement that calls for more than a single momentary informational transaction", which is why it is such a challenge to keep awareness focused on the mantra. It also explains in more concrete terms the "probing" metaphor that Dennett uses to explain shifts in awareness. And it does all of this without requiring the existence of either a single specialized "awareness module" or a central "meaner". One sees the blind spots in the peripheral vision because the visual processes in the brain are relating back to other processes, which include a knowledge about the experimental protocol, which in turn makes possible a higher order thought about the blind spots. Any coalition of brain-states that performs this sort of feedback guided tracking would be able to make higher order thoughts possible. There would be no reason to assume that only one organ in the brain was performing that function, for it could be performed by different parts of the brain at different times.

Or even at the same time, because these tracking process need not all connect back to the same "central meaner". If the "I" which we identify ourselves with is the largest, most active and most coherent set of neural coalitions in our brains, those people whom we speak of as having a strong sense of self or a well-developed self would be ones who tracked the world around them and continuously related it back to an organized set of goals and projects that defined who they were. Such people might also be described as "centered", because they possessed a strong "center of narrative gravity." Their life stories are coherent, like Victorian novels, not random and confused like Beatnik poetry. Most of us are somewhat centered most of the time, which is why the illusion of the central meaner is so persistent. But sometimes during moments of stress or illness, lots of different neural coalitions of approximately equal size are tracking the world in different ways at the same time. At such times our sense of a unified self becomes considerably weakened, and it is more noticeable that the self is not an inescapable Cartesian brute fact, but an achievement that can endure only if it receives constant maintenance.

It might not be immediately obvious that when I sit and observe something "objectively" that I am relating it back to the coalition of "drafts" that define my story. But we must remember that even objectivity is a goal that must be strived towards, and such goals explain why our awareness is focused where it is, and why we see what we see. A meditator concentrating on his breath is in a mental state which is about his quest for enlightenment, and the laboratory subject staring at the color phi phenomenon is in a mental state which is about his desire to be a good laboratory subject. It is this insight which is behind Dennett's assertion that "when a portion of the world comes in this way to compose a skein of narratives, that portion of the world is an observer" (Consciousness Explained 137). To this I would add that an observer comes to compose a skein of narratives, to spin a story out of itself, because it is not just an observer, but an actor as well. Conscious beings are conscious not just because they have a present moment, but because they comprehend that present moment in terms of "projects" which are also projections into a variety of possible futures. What happens when an observer performs an observation, that is, becomes aware of something specific in the vast range of mental phenomena that its brain is always busily synthesizing? It means that he or she (or it) "tracks" a perceived mental phenomenon (such as a perception of a thimble) with some sort of feedback guided link which establishes a relationship between that mental phenomenon and some activity which is part of that observer's story. (such as a game being played at a friend's birthday party.) This tracking need not be performed by a single "consciousness module"; there could be a variety of such links, and there is no reason to assume that they would form a natural kind. Nor need there be a single "central meaner" that the link always goes back to, if the largest neural coalition that arose out of a pandemonium is all there is to the self.

Note, however, that this description of how things are brought to the foreground of experience presupposes that there must be a background of experience as well, and Dennett acknowledges this. He makes a distinction between "the ranks of the merely unconsciously responded to" and being in "the background of conscious experience". (Consciousness Explained p.336) Those items in the background of experiences could be singled out more or less at any moment by an act of intentional tracking or probing, which would make them objects of awareness. They have this kind of immediate accessibility because of past training and experiences. This training can be as simple as being given a look at a thimble before it is hidden, or as complex as being trained to hear the beats in an out of tune piano note. For trained piano tuners, factors like the beating of out of tune strings which before training had only contributed to experience (by giving us a vague sense of out of tuneness or general yuckiness) are now present in the experience ( Consciousness Explained p. 337).

And yet the fact they are present in the experience does not necessarily mean that we are aware of them. If an experienced professional piano tuner went to a party where an out of tune piano was playing, he is not necessary aware of the fact that it is out of tune merely because he has had this training. He could be absorbed in conversation with a friend, and not really listening to the piano. Yet in such a situation, if the piano stopped playing he would probably notice, just as we notice that a clock has suddenly stopped in Dennett's earlier example. For that matter, The piano tuner might be able to tell you, even if you asked him after the piano has stopped, which notes were out of tune. The best explanation for this fact seems to be that he was perceiving the sound of the piano with the full sensitivity of a trained ear, even though he was not aware that he was doing so. This means that we must make a distinction between changes of consciousness caused by development (or attrition) of mental phenomena, and changes produced by probing or shifting awareness.

The essential similarity between Rosenthal and Dennett on this point is obscured by the fact that they both use the word "consciousness" to describe a different half of this distinction. When Rosenthal says that a HOT makes a mental event conscious, what he is saying in Dennett's terms is that an intentional act of tracking or probing has brought something to the foreground of consciousness, and that those things which are in the background of consciousness are only mental, and not conscious. Dennett, however, says that after piano tuners have been trained "they are now conscious of things they were not previously conscious of" (Consciousness Explained p.337)It thus seems that Rosenthal and Dennett really only differ on which aspect of human experience they want to call consciousness. Dennett thinks that the piano tuner who is not intentionally tracking the background music is still conscious of it, even though it is in the background. Rosenthal thinks that the sounds of the piano are only mental phenomena until the piano tuner intentionally tracks them with a HOT, at which point they become conscious. When Dennett considers one's ability to count the sound of chimes in retrospective memory and asks "how could you so clearly remember hearing something you hadn't been conscious of in the first place?" Rosenthal replies that the Chimes were mental before they were probed with a HOT, but did not become conscious until after the HOT.

Both of these descriptions have important strengths, and are not really in conflict with each other. Dennett's description emphasizes that there is not a sharp line between being conscious and being mental, That my so-called "unconscious thoughts" are really subconscious thoughts, because they color my conscious experiences in a variety of subtle ways (unlike the movement of electrons, which will never be present in my experience no matter how much they contribute to it. ). Rosenthal, however reminds us that we do have the ability to bring certain aspects of our experience into focus, and that the difference between being in and out of focus, or shifting from background to foreground, is a phenomenon that needs to be explained. To say that we are conscious of the thimble both before and after it is discovered blurs this distinction unnecessarily. This blurring does lead to some confusion in Dennett's thinking. For example, he uses Rosenthal's term "reportability" to describe being in the background of experiences, but also says "getting something into the forefront of consciousness is getting it into a position where it can be reported on." (Consciousness Explained p.336). What then is the difference between the background and the foreground, given that they both supposedly have reportability? Dennett does not really want to say, because if there is a clear difference he must abandon a substantial part of the current version of MDM.

Dennett's current vision of the multiple drafts model includes "first person operationalism" which "denies the possibility in principle of consciousness of a stimulus in the absence of the subject's belief in that consciousness" This requires that awareness and mental phenomena be conflated to eliminate the possibility of "real seemings". (Consciousness Explained pp. 132-134). If mental Phenomena and awareness are distinct from each other, this would imply that it is possible for us to have mental phenomena without awareness of them. These mental phenomena would thus be real seemings, with an intrinsic nature capable of being misunderstood. But as David Rosenthal points out in his 1993, this first person operationalism is not an essential part of the MDM, and there are many problems that could be solved by abandoning it.

Rosenthal says, contra Dennett, that we can have mental phenomena and not know or believe that we have them. Rosenthal's claim is supported by the example of the piano tuner at the party, who has not probed his consciousness to determine whether the piano is out of tune, but still has the out-of-tune sounds in the background of his consciousness. Can we say that he has no belief that the piano is out of tune, given that no one has asked him to direct an intentional probe towards the piano? He certainly does not have the thought (sometimes called an occurrent or episodic belief) that the piano is out of tune. Does he have a background belief that the piano is out of tune, in the sense that most of us have an unspoken belief that lawyers usually wear shoes? That doesn't seem to be quite right, because there seems to be a sense of discovery that results from probing consciousness to discover the status of the piano. The reason it seems that we already have the belief that lawyers usually wear shoes, is that the answer is obvious as soon as the question is asked; There is no sense of discovery involved in coming to that conclusion. If we accept that it is the act of probing that produces the belief, we must acknowledge that mental phenomena have an intrinsic existence in our experiental domain regardless of what we believe about them.

One way you can have first person operationalism is to say that the act of intentional probing constitutes all that there is to consciousness; that experience has no background at all, only a foreground. Dennett does seem to be advocating this position with the example of the multiple Marilyns, when he says "why should your brain bother importing all of those Marilyns in the first place? Why not let the world store them at no cost, until they're needed?" (Consciousness Explained p.360). This seems to be a specific example of Dennett's radical claim that " there are no fixed facts about the stream of consciousness independent of particular probes" (Consciousness Explained p.138). But this claim requires us to accept the idea that consciousness is identical to what I have called awareness, and jettison Dennett's admission that there is a background to consciousness. What consciousness would be probing would be the world itself, not any realm of mental phenomena. However, if we accept as valid Dennett's discussions of Hunt the Thimble and Piano tuning, we must also accept that there is a background to consciousness, and that what exists in that background is to some degree independent of the particular probes made by acts of intentional tracking.

The fact that awareness and mental phenomena are distinct does not imply that they are unrelated. The reason that the piano tuner has highly developed sensations of musical sounds is that he has directed his awareness towards them in the past, and thus constructed that mental domain which is distinctive to piano tuners. This history of repeated focusing on particular phenomena is what sets the boundary conditions for what can be discovered by any given act of probing, which is why facts of consciousness are not as arbitrary as Dennett frequently implies.

I think that what Dennett means when he says that the moment of consciousness is arbitrary is that it is subjective. Acts of probing and intentional tracking are what create the stream of consciousness out of the vast ocean of mental phenomena, which is why there are no fixed facts about that stream independent of those probes. But both the mental phenomena and the acts that probe into them are governed by the plans and projects of the subject who is the observer/actor. They are thus subjective, but they would not be arbitrary unless the subject in question is leading a life governed only by whims, and not by a center of narrative gravity. The piano tuner can hear those beats because he made it one of his projects to become a piano tuner. This project and others have helped shape a distinctive range of mental phenomena and it is this range which sets the boundaries of what can be discovered by any single act of intentional tracking. This is why first person operationalism is false. Despite Dennett's occasional claims to the contrary, there are true facts about my mental states, constituted by my past history and future projects, and it is possible to get those wrong if one introspects sloppily or self-deceptively.

Once we have abandoned first person operationalism, I think that Dennett's and Rosenthal's theories can be reconciled by saying that what intentional tracking/probing/ HOTs do is bring a mental phenomenon to the center of consciousness. This gives HOTs an important function in consciousness, but does not make them the sine qua non of consciousness. It thus seems more appropriate to use a more specific word, like Awareness, to label the change in consciousness produced by HOTs or Intentional probes, to label the things that HOTs are directed towards Mental Phenomena, and to say that both of these are essential components of Consciousness. Rosenthal's concept of HOTs and Dennett's concept of intentional tracking appear to be more different than they are because they are operating at different theoretical levels: Rosenthal's concept, originating in ordinary language analysis, is very top-down, and Dennett's, being inspired by cognitive neuroscience, is much more bottom-up. Although the concepts of intentional tracking, probing, and HOTs appear to be very different, I believe that careful bridge building could fine tune these ideas to show how they are fundamentally related, and remove any currently existing appearance of dissonance. If we were to indulge in what Dennett calls "slogan-honing" we could describe a Rosenthalian theory of awareness in Dennett's technical terminology:

Awareness is produced by a probing function that brings items that are relevant to one's center of narrative gravity from the background of experience into the foreground of experience.

Such a description of awareness as one component of consciousness would, I think, combine the best elements of Dennett's and Rosenthal's theories.


Block, N.(1992) "Begging the Question against Phenomenal Consciousness" in Behavioral and Brain Sciences 15 number 2 pp. 205-206

Damasio, A.R. (1992) "The Selfless Consciousness" in in Behavioral and Brain Sciences 15 number 2 pp.2O8-9

Dennett, D.C. (1991) Consciousness Explained Boston Little Brown.

____________ (1992) Is Perception the Leading Edge of Memory? Locarno Conference. available as preprint CCS-92-8 from Center for Cognitive Studies Tufts University Medford Masachusetts O2155

Dennett, D.C. and Kinsbourne (1992) "Time and the Observer: The where and when of Consciousness in the Brain" Behavioral and Brain Sciences 15 number 2 pp. 183-247.

Lockwood, M. (1993) "Dennett's Mind" in Inquiry March 1993.

Lloyd, D.(1992) "Towards an Identity Theory of Consciousness" in Behavioral and Brain Sciences 15 number 2 pp. 215-216

MCDermott (1992) "Little Me" in Behavioral and Brain Sciences 15 pp. 217-218

Rosenthal, D. (1986) "Two Concepts of Consciousness" Philosophical Studies 49 pp. 329-359

_____________, (1989) "Thinking that one Thinks" ZIF report no.11 Research group on MInd and Brain, Pedrspectives in Theoretical Psychology and the Philosophy of Mind, Zentrum für Interdisziplinäre Forschung Bielefeld Germany.

_____________,(1990a) "Why are Verbally Expressed Thoughts Conscious?" ZIF Report no.32 Zentrum für Interdisciplinäre Forschung Bielefeld Germany.

_____________, (1990b) "A Theory of Consciousness" ZIF Report no.40, Zentrum für Interdisciplinäre Forschung Bielefeld Germany.

_____________,(1992) " Time and Consciousness" in Behavioral and Brain Sciences 15 pp.220-221

_____________,(1993) "Multiple Drafts and Facts of the Matter" Paper delivered at Meeting of the Society for Philosophy and Psychology Vancouver June 1993

Ryle, G. (1949) The Concept of Mind Hutchison

Van Gulick, R. (1992) "Time for More Alternatives" in Behavioral and Brain Sciences 15 number 2 pp.228-229

Velmans, M. (1992) "Is Consciousness Integrated?" in Behavioral and Brain Sciences 15 number 2 pp.220-221.

(note: all references to any articles in Behavioral and Brain Sciences 15 number 2 are cited as Time and the Observer )