Thinking fish & zombie caterpillars

Michael Tyeinterviewed by Richard Marshall.

Michael Tyeis the jumpin' jack flashman of philosophy of mind, always updating his zap mind with rigorous brooding on the nature of phenomenal consciousness. To do this he has to consider a whole bunch of things - including inverted earths, whether swamp things have eyes, how chinese sounds to the chinese, the beliefs of fish, one eyed zombie caterpillars, camouflaged moths, orgasms, the planet Vulcan and the difference between Keith Richards hallucinating a tomatoe and him hallucinating a unicorn. He writes his books to catch his thoughts as they shoot on by. All in all, he's a funky swell.

3:AM:What made you become a philosopher? Was it really because you just wanted to avoid the lab work of physics? Were there seeds planted way before?

Michael Tye:I went up to Oxford as an undergraduate to study physics. I chose Oxford over Cambridge at the urging of my school physics teacher who was an Oxford man. When I arrived, I found out that, as a physics student, I was expected to spend one day a week in the laboratory. This seemed to me extremely unappealing not only because it would interfere with my social life but also because the practical side of physics was, to my mind, deadly dull. Happily, I discovered that there was a new undergraduate degree — physics and philosophy — that combined theoretical physics with philosophical issues in the foundations of physics as well as pure philosophy. For this degree no practical work was required.

I asked to transfer into physics and philosophy and, in response to my request, I was told to go away and write an essay on truth with particular reference to the Austin/Strawsondebate on the topic. I had never heard of Austin or Strawson at the time, but this I duly did with the help of the college library and, presumably after having not embarrassed myself too badly, I was admitted into the physics and philosophy degree program.

I did not find myself especially interested in the philosophical half of the degree until late in my final undergraduate year. This was due largely to the fact that my college philosophy tutor, who shall remain nameless here, chose not to speak during our tutorials except in the most perfunctory way, preferring instead to stare into the fireplace while puffing on a cigar. (To my knowledge, during his entire career he published just a single essay - on the location of sounds.) I recall his intense dislike of things American. “’Functionalism’,” he would later say to me in an exaggerated, slow and disdainful manner, wrinkling his nose as he did so, “I suppose that is an American word.”

What lured me into philosophy was the sense-datum theory. I did not believe the theory but I found myself very intrigued by it and largely as a result of the writings of Ayer and Austin, I decided to rethink my original plan, which was to do a theoretical physics doctorate. So, that is how I became a philosopher.

3:AM:You are a naturalist philosopher and take science to be very important. So why shouldn’t we just leave it to the scientists to find out about the mind and consciousness? What can philosophers add?

MT:I take philosophers to ask very general questions about the world and ourselves, and also about our relationship to the world and one another through our senses and our social interactions. Philosophers, in my view, should develop their theories in such a way as to respect as much as possible both the latest and best scientific theorising about the world as well as the vast store of wisdom contained in what might be called our “commonsense, everyday theory of things” (dubbed by Sellarsthe “manifest image”).

One way to capture what philosophers of mind, as opposed to scientists interested in the mind, are (or at least should be) trying to do is as follows. Scientists focused on the mind ask: How does this mental faculty work? How does memory work, for example? How are mental images generated? How does shape recognition take place? These questions are ‘how’ questions. They pertain to actual creatures of one sort or another (for example, human beings). They do not pertain to non-actual, possible creatures. And they can be understood, in different instances, to ask for the neurophysiological underpinnings of the appropriate mental faculty or for the computational underpinnings or simply for the more basic psychological components that generate the faculty in the relevant range of creatures.

Philosophers of mind ask: What is such-and such a mental faculty or state? What is it to remember something, for example? What is it to recognise a shape? What is pain? These ‘what’ questions should be understood to ask what is common to all actual and possible creatures that have the relevant mental property or state (remembering something, recognising a shape, etc) in virtue of which they have the property or state. In this way, they are asking about the general nature of the faculty or state.

So, neurophysiology, scientific psychology and artificial intelligence do not directly offer answers to the questions philosophy of mind asks. The philosopher of mind should respect the answers given by scientists to the appropriate ‘how’ questions while insisting that ‘what’ questions remain about which there can be reasonable disagreement - questions whose answers require theorising at a more general level than is found in the individual sciences. Of course with the appropriate theorising, it may be concluded that one of the sciences in particular — for example, neurophysiology — not only provides an account of how the given mental faculties or states are generated in actual creatures (of their realisation in those creatures) but also can supply an account of the general nature of the faculties or states. However, this conclusion is not one that the neurophysiologist, qua neurophysiologist, is in any position to reach. It is not shown to be true directly by neurophysiology and it may reasonably be denied while accepting all that the neurophysiogical account has to say.

3:AM:So what does your theory say the mind and consciousness is?

MT:I am best known for my theory of the nature of phenomenal consciousness. Let me focus on that. I’ll begin by saying a few uncontroversial things about phenomenal consciousness. Of our conscious mental states, some are inherently conscious. That is to say, some of our mental states cannot fail to be conscious. For each such mental state, there is a subjective perspective that goes along with it. This perspective is conferred upon the subject simply by his or her undergoing the mental state. It is captured in everyday language by talk of ‘what it is like’. There is something it is like subjectively to feel pain, to smell vomit, to taste chocolate, to feel elated. Furthermore, what it is like to undergo one inherently conscious mental state can be compared with what it is like to undergo another. For example, what it is like to experience bright red is subjectively more similar to what it is like to experience bright orange than to what it is like to experience dark green.

Mental states that are inherently conscious are standardly said to be phenomenally conscious by philosophers. ‘Phenomenal consciousness’, then, is a feature of mental states. As to which mental states are phenomenally conscious, one not very informative answer is that they are experiences. More helpfully, we can classify the relevant states into at least the following categories: (1) Perceptual experiences, for example, experiences of the sort involved in seeing green, hearing loud trumpets, tasting liquorice, smelling the sea air, running one's fingers over sandpaper.
 (2) Bodily sensations, for example, feeling a twinge of pain, feeling an itch, feeling hungry, having a stomach ache, feeling hot, feeling dizzy. Think here also of experiences such as those present during orgasm or while running flat-out. (3) Felt reactions or passions or emotions, for example, feeling delight, lust, fear, love, feeling grief, jealousy, regret.
(4) Felt moods, for example, feeling happy, depressed, calm, bored, tense, miserable.

As a way of getting at the basic features of my view of phenomenal consciousness, take the case of visual experience. Right now, as you are reading this, you are subject to a visual experience. You are aware of the computer screen before you as well as various other things and a whole host of properties of these items. Suppose I ask you whether you can also be aware of your visual experience. I think that the natural answer to this question is ‘No’. Those who, like me, think that this is the right answer say that experience is transparent to us. When we try to become aware of our visual experiences by introspecting, we seem to ‘see’ right through them so that all we are really aware of is what is outside. Of course, you are aware via introspection that you are having a visual experience. You are aware that you aren’t blind. But you aren’t aware of the experience.

Not only are you not aware of your visual experience, you can’t attend to it. You can switch your attention from the computer screen to the flowerpot on its right, say, but you can’t switch your attention from the screen to your experience. The experience isn’t something on which you can focus your attention at all. All you are really aware of, all you can really attend to, is stuff outside you or at least stuff that is presented to you as being outside you. (The stuff might not really be outside, for it might be that you are undergoing an extended, extremely vivid hallucination.)

What’s the relationship between the stuff outside – the stuff you experience – and your experience? I say that it is stuff represented by the experience. What I mean by this is that it is stuff the experience ‘tells’ you is out there. This has an important consequence. If all you really have access to is stuff outside, then the subjective character of your experience – how it ‘feels’ to you – has got to be understood in terms of the outside stuff, however strange that may initially seem. On this view, what makes two experiences subjectively the same is that they represent the same stuff (roughly) and what makes two experiences subjectively different is that they represent different stuff.

Of course, not everything mental that represents anything is an experience. Think, for example, of the visual system. Vision is highly complex but it begins with representations of what is going on at the retina, representations of light intensity and wavelength. These early representations aren’t conscious at all. What representations have to do to get to be experiences, on my view, is that they have to arise at the right level in the information processing. They have to be available to cognitive processes in the right sort of way. They have to be appropriately ‘poised’. Your visual experience of the computer screen before you and its properties is the evidential basis for your belief that there is a computer screen ahead and a whole host of related beliefs. This evidential basis does not itself involve the application of concepts. Experiences are non-conceptual states given to us by nature, not requiring learning. They are (largely) fixed by our phylogenetic nature. Given how human beings are built, they automatically undergo a certain range of experiences. Given how bats are built, they automatically undergo another.

The idea, then, is (roughly) that experiences are suitably cognitively poised mental states that nonconceptually represent the world. Their phenomenal character – what it is like to undergo them – is given by what they represent (or more carefully, on my current view, by the cluster of properties they represent as being possessed by things out in the world (or in the body, as in the case of bodily sensations)).

This view of experience has come to be called ‘representationalism’. I first advocated a (slightly different) version of it in my book Ten Problems of Consciousnessin 1995.

3:AM:A couple of really fun ideas that help get a grip on some of your key ideas are the idea of an Inverted Earth and the idea of the Swampman. I think many readers will have had the thought that maybe on another planet what I take to be blue they take as yellow and vice versa. And that if a creature was born without a history but with my implanted mental states we’d wonder what their mental states were really. What are the key things these stories are supposed to help reveal? Is it that without a teleological account of mind – i.e. one that tracks the purpose of minds – we don’t get anywhere?

MT:Inverted Earth is an imaginary planet, on which things have complementary colours to the colours of their counterparts on Earth. The sky is yellow, grass is red, ripe tomatoes are green, and so on. The inhabitants of Inverted Earth undergo psychological attitudes and experiences with inverted representational contents relative to those of people on Earth. They think that the sky is yellow, see that grass is red, etc. However, they call the sky 'blue', grass 'green', ripe tomatoes 'red', etc just as we do. Indeed, in all respects consistent with the alterations just described, Inverted Earth is as much like Earth as possible.

In Ned Block'soriginal version of the tale, one night while you are asleep, a team of alien scientists insert colour-inverting lenses in your eyes and take you to Inverted Earth, where you are substituted for your Inverted Earth twin or doppelganger. Upon awakening, you are aware of no difference, since the inverting lenses neutralise the inverted colours. You think that you are still where you were before. What it is like for you when you see the sky or anything else is just what it was like on earth.

But after enough time has passed, after you have become sufficiently embedded in the language and physical environment of Inverted Earth, your experiences and beliefs will come to represent just what the locals’ experiences and beliefs do. You will come to believe that the sky is yellow, for example, just as they do. Similarly, according to Block, you will come to have a visual experience that represents the sky as yellow. For the experiential state you now undergo, as you view the sky, is the one that, in you, now is normally caused by and tracks yellow things. So, the later you will come to be subject to inner states that are representationally inverted relative to the inner states of the earlier you, while the phenomenal aspects of your experiences will remain unchanged. It follows supposedly that my view of phenomenal consciousness is mistaken.

One simple reply that the representationalist can make with respect to this objection is to deny that there really is any change in normal tracking with respect to colour, at least as far as your experiences go. "Normal", after all, has both teleological and nonteleological senses. If what an experience normally tracks is what nature designed it to track, what it has as its biological purpose to track, then shifting environments from Earth to Inverted Earth will make no difference to normal tracking and hence no difference to what your experiences represent. The sensory state that nature designed in your species to track blue in the setting in which your species evolved will continue to do just that even if through time, on Inverted Earth, in that alien environment, it is usually caused in you by looking at yellow things. So, the sky on Inverted Earth will look blue to you even though you call it ‘yellow’ no matter how long you spend there.

On this view, our senses are akin to instruments and our sensory experiences are akin to states of those instruments. Think, for example, of the pointer positions on a speedometer. When the pointer points to ‘60mph’ on the speedometer, it represents that the car is going at 60 mph. This is a consequence of its original design and this is what it will continue to represent even if the speedometer is used for some other purpose. Likewise for our visual experiences.

The suggestion that tracking is telological in character, at least for the case of basic experiences, goes naturally with the plausible view, mentioned earlier, that states like feeling pain or having a visual sensation of red are phylogenetically fixed. On this view, through learning we can change our beliefs, our thoughts, our judgements, but not our basic experiences. Having acquired the concept microscope, say, we can come to see something as a microscope, but we do not need concepts simply to see. Once the receptor cells are matured, it suffices to open the eyes. No learning or training is involved. Basic visual experiences are nonconceptual. Small children see pretty much what the rest of us see. They differ from us in how they see things, in what they see things as. They do not see that the kettle is boiling, the house as being dilapidated, the computer as malfunctioning.

This reply to the Inverted Earth objection, tempting though it is, faces a problem. It entails that accidental replicas of actual sentient creatures lack all experiences. Consider, for example, the case of the swamp creature formed by the chemical reaction that takes place in a swamp after a lightning bolt hits a log there. Swampman, as he is usually known, is an accidental molecule-by-molecule duplicate of some actual human-being, but he has no evolutionary history. On a cladistic conception of species, Swampman is not human. Indeed, lacking any evolutionary history, he belongs to no species at all. His inner states play no teleological role. Nature did not design any of them to do anything. So, if phenomenal character is understood in a teleo-representational way, then Swampman has no experiences. This according to some philosophers, is unacceptable.

The first representationalist about experience to embrace the supposedly unpalateable conclusion was Fred Dretske (in 1995). I now think that Dretskewas right. The question of what to say about the case of Swampman is not restricted to the case of experiences. Ruth Millikanhas said that Swampman does not have eyes, since an eye is an organ whose presence is to be explained in terms of descent from creatures whose fitness was increased in certain specific ways by the operation of that organ. Swampman’s eye-like structures are present merely by chance, and so do not count as eyes. The implications are far-reaching: Swampman doesn’t see anything, assuming seeing constitutively involves using eyes, doesn’t walk, assuming that walking involves use of legs, and legs, like eyes, are individuated historically; and so on.

The surprising character of this view can be diminished by noting that something just like an eye, but which is not an eye, can be as useful as an eye; and likewise for all the other features. Something just like a human, though not human, may be as loveable as a human, and as worthy of a place in our moral scheme. Likewise, something functionally just like an experience may get Swampman around as well even if it is not an experience. So, I am inclined simply to bite the bullet on the Swampman example.

3:AM:It seems that in the disputes between alternative accounts of how to handle alternative earth and Swampman, philosophers draw on intuitions about what would be natural for someone to say and think in these circumstances. But there’s evidence in x phithat suggests that a lot of self-reporting about introspected thoughts is false. So do you worry that cool arguments against Ned Block and Shoemaker et almight be sullied by errors (and theirs too)? Perhaps intuitions are just untrustworthy?

MT:No, I don’t worry. Intuitions are a bit like perceptual ‘seemings’. My belief that I am sitting at a table, drinking coffee is based on how things appear or seem to me. The appearances can, and do, sometimes lead me astray. But they form the basis for my perceptual beliefs about the world. Intuitions play a similar role in connection with philosophical reflection. Certain things seem to me obvious or undeniable and I then try to tailor my philosophical theories to fit or preserve those intuitions. Can they lead me astray?Yes, but so what?

In some cases, intuitions are, I think, influenced by prior philosophical commitments. For example, the intuition in the Swampman case that intrinsic duplicates must feel the same thing, so that my Swamp duplicate must feel just what I do, seems to me to come from a Cartesian, internalist picture of the mind. But then perceptual seemings are sometimes causally influenced by prior beliefs too. For example, how a Chinese speaker sounds to native Chinese speaker is very different from how it sounds to me. Knowledge of the meanings of Chinese words causally influences how the sound stream is segmented and this impacts how it appears or seems to the hearer.

As for xphiand self-reporting of introspected thoughts, my own view is that there is less here than meets the eye. Philosophers have often held a privileged access thesis with respect to conscious, occurrent thoughts to the effect that when introspection is working normally (as it should), thinkers can know what they are (presently) thinking just on the basis of introspection. This is not a thesis about past or remembered thoughts. It is not a thesis about knowledge of the species of a mental state (whether it is a desire or a belief, say). It is not a thesis to the effect that the final authority on what is going on in the mind is the person whose mind it is. To my knowledge, no xphi result shows that there is anything wrong with the privileged access thesis, understood as above.

3:AM:You think that qualia are the heart of the mind body problem. Some, like Dennett in his paper ‘Quining Qualia’ just deny qualia exist. Dennett calls them illusions – but the obvious question is what is it that they are supposed to be illusions of? Others like Richard Brownfind that just an outrageous position. He thinks it’s obvious that qualia exist. So why is this so important to you and what’s your take on the matter? I think you agree with Alex Byrnethat they ‘aint in the head’, but it seems hard for me to understand how they could be anywhere else if they are supposed to be understood as ontological and not epistemic? I guess that makes me kind of some sort of dualist, which isn’t very cool is it?

MT:The term ‘qualia’ is philosophically loaded. In one sense, it is obvious that qualia exist; for it is obvious that people have experiences and feelings and that there is something it is like for them to undergo such states. Often, however, when philosophers talk of qualia they have in mind more than this. According to one view, the subjective character of an experience – what it is like to undergo it – is a matter of the intrinsic, non-representational, introspectible features it possesses. I am opposed to this view. I deny that there are any such features. If I now introspect my experience of red, say, I don’t come across any features of the experience at all. I just come across the color red. And my experience of red is not itself red. Red is what my experience represents. Since I am aware of the subjective character of my experience – how it feels - when I introspect, it follows that the subjective character of an experience is to be understood in terms of the qualities the experience represents rather than the qualities it possesses.

As for Dennett, the most charitable way to read him is as holding that experiences themselves are dispositions to believe various things. So, there are no experiences or qualia, conceived of as non-epistemic, non-belief-laden entities. This view seems obviously mistaken but the alternative is to read Dennett as just denying that there are any such things as experiences and subjective character at all. And that is worse than obviously mistaken – it’s crazy. Hence, Block’s quip that that Dennett’s book, Consciousness Explained, should have been titled Consciousness Ignored.

3:AM:You argue, I think, that if representationalism is right then questions such as whether fish have qualia, or plankton, can be answered. Is that right? So do fish and plankton? And are there creatures that are zombies?

MT:Fish have experiences. Plankton don’t. Honeybees have experiences. Caterpillars don’t. Caterpillars are zombies. Fish have eyes and ears (though not external ears but hidden ear-like structures) and a sense of smell, and the behaviour they engage in on the basis of the information they glean from these senses shows considerable flexibility. Fish learn to recognise markings and patterns, to avoid artificially colored, unpleasant-tasting fish they would normally eat, to solve problems in order to reach feeding places. Fish can find their way through mazes and they defer to other fish that are better at finding their way through when they are in groups. Cumulatively the evidence seems best explained by supposing that fish often make cognitive classifications or assessments, directly in response to the information conveyed to them by their senses, and that these, together with their goals, often determine their behavior.

There may be some reluctance to say that fish have beliefs. Clearly, a fish cannot believe that a hand is dangling in the water. Fish lack the concept hand. And, in general, it seems unlikely that fish share many concepts with us. But the fact that there are striking conceptual differences, that our concepts are much richer and more articulated than theirs, does not show that they lack any concepts at all. Possessing a perceptual concept, in my view, is (roughly) a matter of having a stored memory representation that has been acquired through the use of sense-organs and that is available for retrieval, thereby enabling a range of discriminations to take place. Perceptual beliefs are (roughly) representational states that bring to bear such concepts upon stimuli and that interact in rational ways, however simple, with one another and other representational states the creature generates in response to its needs, thereby determining behavior. In this sense, given the facts adumbrated earlier about fish behavior, it seems to me very plausible to suppose that fish form simple beliefs on the basis of immediate, sensory representations of their environments. These sensory representations, given where they arise in the information processing and the downstream role they play in the production of beliefs, just are experiences.

What is true for fish is true for honeybees. Honeybees have sense organs. For example they have eyes and ‘ears’ and a sense of smell. And they engage in intelligent or apparently intelligent behavior. Take the case of children who are starting to learn the alphabet. They can discriminate different letters, whatever their size or font. Pigeons can be trained to do this also. So too can honeybees up to a point. Some bees can recognise the letter B in different orientations, colors, sizes, and fonts. Apparently, then, they can form novel concepts of items not normally found in their natural habitats.

Here is a striking example of apparently intelligent behaviour in honeybees. The flowers of alfalfa possess anthers that spring forward and deliver a heavy blow in response to pressure. Once honeybees have been hit by the alfalfa anther, they avoid alfalfa like the plague. But if they are taken to a region of many acres of alfalfa and set free there, they are compelled to confront the problem or starve to death. So, they do one of two things: either they learn to identify flowers, the anthers of which have already sprung, and they only alight on the sprung ones or they learn to get at the nectar by chewing through the flower from the side without ever setting off the anther. For the same reasons as fish, then, honeybees have experiences. They are not zombies.

What about caterpillars? Is there anything it is like to be a caterpillar? Different kinds of caterpillars show different sorts of behavior upon hatching. Some, for example, eat the shells of the eggs from which they emerge; others crawl away from their cells immediately. But there is no clear reason to suppose that caterpillars are anything more than stimulus-response devices. They have a very limited range of behaviours available to them each of which is automatically triggered at the appropriate time by the appropriate stimulus. Consider, for example, their sensitivity to light. Caterpillars have two eyes, one on each side of the head. Given equal light on both eyes, they move straight ahead. But given more light on one of the eyes, that side of the body locomotes more slowly. So, when caterpillars move, they tend to move towards the direction of most intense light. This is why caterpillars climb trees all the way to the top. The light there is strongest. Shift the light to the bottom of the tree, and the caterpillar will go down, not up, as it usually does, even if it means starving to death. Remove one of its eyes, and it will travel in a circle without ever changing its route.

Once one is made aware of these facts, there seems no more reason intuitively to attribute phenomenal consciousness to a caterpillar on the basis of how it moves than to an automatic door. The latter responds in a fixed, mechanical way to the presence of pressure on a plate in the floor or ground in front of it, just as the former responds mechanically to the presence of light. There is no learning, no variation in behavior with changed circumstances, no reasoned assessment.

Again, this is the result my theory delivers. Caterpillars do not move purposefully. They do not believe that the light is strongest at the tops of trees. They do not want to get to the strongest light. There is nothing in any of their behavior that seems to require the admission that they have any wants or beliefs. So, caterpillars do not support sensory representations on the basis of which beliefs and desires are formed. None of their representations are appropriately cognitively poised. Caterpillars, then, are not phenomenally conscious any more than are plants. And what goes for caterpillars and plants goes for plankton.

3:AM:Talking of which, does your view preclude a view of Chalmers zombie world – exactly like ours but without consciousness?

MT:Chalmerszombie world is conceptually possible. No amount of a priorireflection alone will show that there can be no such world. Even so, zombie world is metaphysically impossible. That is all that my theory of phenomenal consciousness requires.

Consider the claim that water is H2O. That was a scientific, a posteriori discovery. Water has a certain hidden nature that science has revealed. Is it conceivable that water isn’t H2O? Certainly. It is conceivable, though incredibly unlikely, that scientists have made a mistake about the nature of water. It is even conceivable that the atomic theory of matter is mistaken. But if in fact science has gotten things right and in fact water is H2O, then there is no possible world in which water is not H2O.

I think of my theory about phenomenal consciousness as being somewhat like this. The theory is justified in terms of its overall explanatory power. No claim is made that it is incoherent to deny the theory, that it is inconceivable that the theory is mistaken. Still, if it is true, it is necessarily true – and Chalmer’s zombie world is then metaphysically impossible though it cannot be ruled out a priori.

3:AM:A big debate with lots of legs is the one about whether colours adhere out there in the world or are the subjective products of our minds. How does your theory handle this question? Are roses really red? And how is it analogous to feeling pain? Ned Blockthinks pain is harder to explain on your theory than colour (and that your theory fails) but thinks you are heroic in your efforts. Why is Block wrong?

MT:Yes, roses really are red. That’s a piece of commonsense. It’s also what our experience tells us. We experience colours as covering the surfaces of objects (as with roses) or sometimes filling volumes of liquid (as with wine) or less commonly inhering in films. Unless our colour experiences are radically mistaken, colours must be out there in the world in the sense of being properties of external things. This leaves open, of course, whether colours have a fully objective hidden nature, whether they are simple qualities, or whether they are subject-dependent properties. The last alternative is compatible with colours being properties of objects. It holds that the relevant properties are dispositions the objects have to produce certain subjective responses in perceivers. On this view, red is the disposition an object has to cause a normal perceiver of the object located in standard viewing conditions to have a certain colour experience. And here lies the rub: which colour experience? Obvious answer: the experience of red. But now the proposal is circular. It tells us what red is in part by reference to red.

The circularity can be avoided by saying that the relevant subjective experience is a red* experience, where ‘red*’ denotes an intrinsic, introspectible, subjective property of the experience. Red now is the disposition to produce in normal perceivers in standard conditions red* experiences. The trouble is that there is no such property.

We are driven, then, either to the view that colours have a hidden scientific nature or that they are simple. I opt for the former alternative and I think of colours as reflectances.

Just as colour experiences are perceptual experiences of a certain sort, namely those that represent reflectance patterns, so pain experiences, in my view, are quasi-perceptual experiences that represent certain bodily disturbances. The relevant disturbances involve actual (or potential) tissue damage of one sort or another. The full story here is complicated and I pass over the complications except to note that there is, in my view, an evaluative component to pain experience. Pains feel bad. This is accommodated, on my account, by holding that pain experiences represent the relevant bodily disturbances as being bad for one (as being apt to harm). The flip side of the coin is that of orgasms. They feel good, very good (which is why sex is so popular); and this is because they are experiences that represent different bodily changes as being good for one (as apt to benefit).

Pain experiences are quasi-perceptual in that admit of cases of hallucination and illusion. A pain in a phantom limb is an experience that represents tissue damage in a limb that no longer exists. This is a case of hallucination. A pain in the left arm when the disturbance triggering the pain is in the heart is an experience that erroneously represents the disturbance as being in the arm. This is a case of illusion.

What makes the pain case more complicated than that for colour is that for color we allow that there is an appearance/reality distinction. Something can look red without really being red and something can be red without looking red. This is not true in the case of pain. If something seems/feels painful to you, it is painful. And if it is painful, it seems/feels painful. The explanation is straightforward. Pain is an experience or feeling. Colour is not. This is why there can be no pains in a world without sentient creatures but there can be colors in such a world. If all sentient creatures on earth ceased to exist there would still be red and blue and yellow objects.

3:AM:One of the things that happens in your explanations is that ‘awareness of’ turns into ‘awareness that’. Is this like Jason Stanley’sarguments that ‘knowing how’ is always a case of ‘knowing that’? And like when Fodor disputes Wittgenstein and all meaning holists and connectionists who says you shouldn’t explain meaning in terms of use and how terms function in the whole language web but in terms of propositions? Is inverting the common order of explanation often a good way of beginning to get hold of an idea?

MT:My thesis about awareness is not really like the other theses you cite. Philosophers distinguish between de reawareness – awareness directed on particular things or properties – and de dictoawareness – awareness that so-and-so is the case. I accept this distinction. I am also prepared to accept that in ordinary life we may say such things as the following: when I turn my attention inwards, I am aware of my experience of a tomato, say. But often when we speak this way and use the term ‘of’ in connection with experiences, we really have in mind de dictoawareness, not de reawareness. Consider, for example, my saying of you that I am aware of your penchant for driving fast. This might be understood in a de reway; but a more natural reading is that I am aware that you have a penchant for driving fast.

Earlier, when I said that we are introspectively aware of our experiences, I meant that we do not have de reawareness of them. They are not objects of our awareness. We are certainly aware (de dicto) that we have experiences of cars and trees and people and colours and so on.

3:AM:The economistuses psychology of change blindness in explaining why economists missed the information that would have warned them that the crash was coming. You discuss attention, seeing and change blindness because it helps us think about the character of our visual experiences, helps us think about the role of attention, and about the vehicles of awareness. Can you say how you think we should think about each of these things, and can you say whether this only extends to visual experiences or do we learn things about experience and thought generally?

MT:This is an enormously broad question, so I can only skate the surface here. A natural view to take of consciously seeing a thing is that it is a matter of undergoing a visual experience (using the eyes) that makes one conscious of that thing. But what is it to be conscious of a thing? Evidently, I can be conscious of a thing without being able to identify the thing. Think, for example, of being conscious of a thing located on the other side of thick, distorting, darkened glass. I might be able to see the thing and track its movements without having the faintest idea what it is.

Still even if I can’t identify a thing, if I am conscious of it then it must be marked out or differentiated in the phenomenology of my experience. For suppose that it is not. Then I won’t even be able to mentally point to the thing on the basis of my experience. So, my experience alone does not enable me even so much as to wonder “What is that?” with respect to the thing. Nor does my experience alone enable me directly to form beliefs or make judgments about the thing. The thing thus is hidden from me. I am blind to it. I am not conscious of the relevant thing.

So when am I conscious of a thing? The above remarks suggest the following test: I am conscious of a thing just in case my experience has a phenomenal character directly on the basis of which I can at least ask myself “What’s that?” with respect to the thing (or form some singular belief about it).

Here is another way to motivate the above test. Consider a perfectly camouflaged moth on a tree trunk. The moth is in plain view. Do I see it? What about the blob of white-out on the page of white paper? I have no idea where it is on the paper, as I hold the page before me. Do I see it?

As for the case of simple creatures without the capacity to form propositional attitudes, for example, caterpillars. I deny that they see things around them (in the relevant sense of ‘see’). I do not deny, of course, that such creatures may register or detect things in their environments and thus see them in a weaker sense.

It seems to me that the natural, intuitive view to take is that I am not conscious of either the perfectly camouflaged moth or the blob of white-out. But why not? In the moth case, surely I am not conscious of the moth because my visual experience is not about the moth at all.

It might be replied that this is mere prejudice. My experience is about the moth. But then how does my experience latch onto the moth? After all, the moth is not marked out or differentiated in the phenomenology of my experience. I cannot mentally point to it. So, I can’t even ask myself “What’s that?’ with respect to the moth directly on the basis of my experience. Surely, the right thing to say is that I am not conscious of the moth.

Since I cannot mentally point to a thing unless I attend to that thing, it follows from the above remarks that seeing a thing demands that one be to attend to that thing (without moving the eyes). This, in turn, entails that there will be cases of things in clear view that one does not see, given where one’s eyes are fixated; for there will often be visible things in the field of view on which one cannot mentally focus, given where one’s eyes are fixated. So, one will be blind to these things. Some cases of change blindness are like this.

Philosophers and psychologists who take change blindness to be a purely cognitive phenomenon as opposed to a failure to see any thing in the field of view often implicitly endorse a clear, colour photograph model for the vehicles of conscious awareness. On this view, in the case of two scenes that differ in one respect, if they are seen one after the other, change blindness arises because an inaccurate comparison is made between the two clear snapshots. There is comparison-failure.

A better model of the vehicles of conscious awareness is provided by drawn pictures. Think about the case of an artist drawing a picture of a scene. The artist begins by fixating on some part of the scene. Some things in the scene are left out of her picture altogether. Others are very richly depicted. We might be a bit like such an artist as we see the things around us. Change blindness then could sometimes arise as a result of certain things not being included in the mental sketch that is drawn of the scene before the eyes. In other cases, it could result from a comparison failure as above.

It is worth noting that the drawn picture model of visual experiences allows for the possibility that some things not in the scene are experienced. Artists sometimes embellish the scenes they draw, adding items that are not present. Upon occasion, we might do that too as we see the world around us.

This general approach extends straightforwardly to other perceptual experiences and it makes a connection, already present in my representationalist view of experience, between consciously perceiving a thing and thinking about it; for, as noted above, one cannot consciously perceive a thing unless what can at least ask “What’s that?” with respect to the thing or form some other de recognitive attitude with respect to it. So, one cannot consciously perceive a thing unless one can think about (or form a belief or some other cognitive attitude about) it.

3:AM:So Keith Richards is hallucinating a tomato without there being a particular tomato that he is hallucinating. What is the content of his experience? And would it make a difference if it had been a unicorn?

MT:Consider a case of veridical perception. Suppose that Keith is seeing a ripe tomato. The view of naïve realism is that Keith sees the tomato directly. Keith is in direct contact with the perceived object. There is no tomato-like sense-impression that stands as an intermediary between the tomato and him. Nor is he related to the tomato as I am to a pig when I see its footprint in the mud. He does not experience the tomato by experiencing something else over and above the tomato and its facing surface. Keith sees the facing surface of the tomato directly.

Some philosophers have suggested that to do proper justice to the above thought, we need to suppose that the objects we perceive are components of the contents of our perceptual experiences in veridical cases. This supposition is supported further by the simple observation that if I see an object, it must look some way to me, and if an object looks some way to me, then intuitively it is experienced as being some way. On pain of losing direct contact with the object, that again suggests that the object itself figures in the content of the experience, assuming that experience is representational at all.

In cases of illusion, the perceived object appears other than it is. In such cases, intuitively, the perceptual experience is inaccurate. And it is so precisely because the object is not as it appears to be. The simplest explanation of this, it is natural to hold, is that, where there is a perceived object, a perceptual experience has a content into which the perceived object enters along with its apparent properties.

Once it is acknowledged that in veridical and illusory cases the seen object is a component of the content of the experience and thus that the content itself is singular, a puzzle arises. In standard hallucinatory cases there is no object with which the subject is in perceptual contact and correlatively no singular content. So, again, what is the content of Keith’s experience?

One possible answer is that the content is existential in the hallucinatory case. Keith's visual experience represents that there is something red, round and bulgy before him, and since there is no such thing, his experience is inaccurate. The combination of views that results is unlovely and implausible; for it is forced to postulate a displeasing and radical asymmetry in cases that pre- theoretically seem alike. (My 1995 view that the content of experience is existential in all cases makes the seen object irrelevant to the matching of the experience with the world in the veridical case: all that is needed for the experience to be accurate is that there be some object with so-and-so properties. That now seems to me wrong too. If I see object A, my visual experience – that very experience – would not have been accurate if A had not existed even if some other object that looked just like A had been present.)

Another possible answer is that Keith’s experience has no content at all in the hallucinatory case. Keith is simply sensing a red, round, bulgy sense-datum or he is sensing redly, roundly, and bulgily. Again, the resulting combination of views for the veridical and hallucinatory cases is unlovely and implausible. And there are other well-known difficulties which I shall not rehearse here.

These reflections suggest that in the hallucinatory case, we should say that there is content of the very same sort as in the veridical and illusory cases – content that is just like singular content but with a gap or hole in it where the object is supposed to go. And this is what I said in recent work. However, I have come to think that the very idea of gappy content is incoherent. Let me very briefly sketch the main worry.

Consider the singular content that object O, is red. On the view associated with Bertrand Russell, which I’ll use for illustrative purposes, this content is complex, having object O, and redness as its components. A visual experience (or other mental state) having that content is accurate if and only if O is red. One way to think of the content here - the content that O is red - is as an ordered pair having O as its first member and redness as its second. Another way to think of the content is as a structured, possible state of affairs built out of O and redness.

It was suggested above that where a visual experience is hallucinatory, the content is just like a singular content but with a gap or a hole in it where an object should go. But does this really make sense? On the ordered pair conception of singular content, there must be two items to form the pair. Since a gap or a hole is not an item, or so it seems, there is no first member of the ordered pair and so no ordered pair at all. On the possible state of affairs conception, the relevant complex is structured out of O and redness in the singular case. But in the gappy case, there is no object O. So, how is there a complex entity structured out of its components? There are various things that could be said here on behalf of gappy content but none of them seem to me satisfactory.

Here is another way to think of the content. Consider the thought that Vulcan is a planet. The thought is complex, being composed of the concept Vulcan and the concept planet, combined in a certain way (rather as the sentence “Vulcan is a planet” is composed of the words “Vulcan” and “planet‟ in a certain order). However, the content of the thought need not be complex. The content can be just a set of possible worlds – the set of worlds in which the referent of the concept Vulcan has the property referred to by the concept planet (the property attributed by the thought). Since the concept Vulcan has no referent either in the actual world in any other possible world, the relevant set of possible worlds is empty.

Correspondingly, experiences are complex. They have representational parts. Some parts represent objects seen (if there are any); others represent properties of which the subjects of the experience are conscious. The latter representational parts are arguably more like features (as is the case with real pictures), and so in one sense they are not really parts at all but rather features of parts. But however this is developed further, a complexity in representational structure for experiences need not be reflected in a corresponding complexity in representational content.

Once this point is appreciated, given the difficulties already encountered in trying to understand gappy content, a natural suggestion is that the content of a visual experience is simply a set of possible worlds, namely the set of worlds at which the experience is accurate. On this view, the content of a visual experience is unstructured in the sense that it has no component parts.

This suggestion preserves uniformity in content for all experiences. Experiences, whether they are veridical, illusory, or hallucinatory, have associated with them an appropriate set of possible worlds. An experience, thus, is accurate, if and only if the actual world belongs to the appropriate set of possible worlds. Which is the appropriate set? Answer: the set of worlds at which the objects picked out by representational parts of the experience have the properties the experience aims to attribute to those objects (however this is further cashed out). The objects thus picked out are the objects (if any) that are seen. Where there are no seen objects, as in a hallucination, there are no possible worlds at which the objects picked out by the representational parts of the experience have the experienced properties. So, the set of worlds associated with a hallucinatory experience is the empty set.

It follows that if Keith had been hallucinating a unicorn instead of a tomato, the content of his experience would have been the same, namely the empty set. But, of course, phenomenally – subjectively – the two experiences are very different. On the view being proposed, then, the phenomenal character of an experience is not determined by its representational content. Experiences that are alike in their content can nonetheless differ in their phenomenal character. The thesis of strong content representationalism (that the phenomenal character of an experience is one and the same as its representational content) is, therefore, false. When I wrote Ten Problems of Consciousness, I was a strong content representationalist and I held that the content of visual experience is existential in nature across the board. I no longer hold either view.

3:AM:So there’s an array of formidable philosophers all thinking about what is the nature of consciousness and the mind and qualia and so forth. You’ve all got brains the size of planets. You’ve all read the same papers and know the science. There’s you and Tim Crane and Frank Jackson and Brian McLaughlin and Dan Dennett and Pat Churchlandand Jerry Fodor and a whole bunch more and you are all pretty much equals. But you disagree. Given that none of you are any more likely to be right than anyone else, wouldn’t it be rational for you to modify your positions in the light of knowing that your peers disagree? Or are you all trapped in the cognitive illusion that seems to grant reflexive over-confidence?

MT:Planets vary in their sizes… Seriously, the issue that is being raised here is a very difficult and general one within epistemology. It arises whenever there is disagreement among pretty much equals. I leave the answer to this question to the epistemologists.

3:AM:You have actually changed your mind. In your book Consciousness Revisitedyou argue with your previous self and find new arguments defending materialism. So was there a eureka moment when it suddenly came to you that you were wrong, or was it a gradual slide away from where you used to be, resulting from all the arguments? And as you move forward, are you finding that your thought is still evolving?

MT:The evolution in my thought has been basically as follows. When I wrote Ten Problems of Consciousnessin 1995 and Consciousness, Color and Contentin 2000, I held these three theses: (1) Common Phenomenal Character: Veridical, illusory and hallucinatory experiences can sometimes have the very same phenomenal character. (2) Common Existential Content: Veridical, illusory and hallucinatory experiences all have existential representational content and in some cases have the very same existential content. (3) Strong Content Representationalism: the phenomenal character of a mental state is one and the same as its poised, nonconceptual, existential content (its PANIC, to use my earlier acronym). I gave up (2) around 2006 (for some reasons, and I held for a while in place of (2): (2a) Disjunctivism about Content: Veridical and illusory experiences have singular contents; hallucinatory experiences have gappy contents.

Since I continued to hold (1), I then also gave up (3) about which I had already started to have independent doubts, since it no longer seemed to me to fit well with the transparency of experience. I can no more be aware (de re) of the content of my current visual experience than I can be aware (de re) of any of intrinsic properties, though I can be aware that I am having a visual experience with such-and-such content.

In place of (3), I adopted: (3a) Strong Property Representationalism: the phenomenal character of an experience is one and the same as the complex of properties represented by the experience. A mental state has phenomenal character just in case it is appropriately poised (a functional role condition) and it nonconceptually represents a complex of properties. I was also then a weak content representationalist: (4) Weak Content Representationalism: necessarily, experiences with the same representational content have the same phenomenal character.

Later, in 2009, I came to have doubts about gappy content and so I gave up (2a) and in its place I accepted: (2b) Common Set-Theoretic Content: Veridical, illusory and hallucinatory experiences have as their content a set of possible worlds. Where there is a seen object A experienced as being F, the relevant set is the set of possible worlds at which A is F. Where there is no seen object (the hallucinatory case), so that ‘A’ is empty, the set of worlds is the set of worlds at which A is F, but this set is now the empty set. (2b) also fits well with my view about the content of thought elaborated in my 2012 book with Mark Sainsbury, Seven Puzzles of Thought. Once (2b) is accepted, weak content representationalism goes. So, now, of the theses above, I hold (1), (2b) and (3a).

The transition in my views has been orderly and (I think) well motivated. As to the future, that’s difficult to predict, but I doubt that I will change my views much from the present ones. Like Russell, however, I have no qualms about changing my mind as the arguments dictate.

3:AM:Finally, if atom by atom my body was replaced by silicon functioning as my current carbon and water stuff, would eventually everything go dark for me, or would my consciousness, my mind, survive?

MT:There are two questions that need to be distinguished here. One is whether the phenomenology would change under the above scenario. It is important to appreciate that this question does not presuppose that the original being still exists at the end of the process. What is being asked is simply whether the phenomenology would change, whoever is the subject of that phenomenology. The second question is whether the original being would survive the process of being gradually siliconised.

In the phenomenology case, the question, as posed, is not: is it metaphysically possible that the phenomenology changes – that it suddenly disappears or changes in its character or alternatively fades away? Rather the question, I take it, is to be understood in the same way as the question: if a rock is dropped, would it fall to earth? The relevant possibility is thus nomological. Similarly, for the case of personal survival.

My view is that the phenomenology would remain unchanged. I argue for this view in a forthcoming essay, available for reading on my webpage, entitled 'Homunculi Heads and Silicon Chips: The Importance of History to Phenomenology'.The general idea is that none of my beliefs would change, including my beliefs about phenomenology, and it seems very implausible to suppose that the later beliefs would end up being radically mistaken, as would occur if the phenomenology disappeared or changed dramatically. Here I agree with Chalmers. As for the case of personal identity, it’s tricky. One relevant issue is whether my original brain still exists at the end of the gradual replacement process even though the brain at the end is entirely silicon.

3:AM:Are there any books or films that you’ve found helpful to you outside of philosophy?

MT:Not really. Of books with a vaguely philosophical slant, I enjoyed A Fairly Honorable Defeatby Iris Murdoch; as far as films go, Mementoand The Matrix.

3:AM:And finally, once everyone has finished reading all your great books, are there five other books you could recommend to the androids here at 3:AMso we can delve further into this fantastic realm of thought?

MT:The Conscious Mindby David Chalmers, The Stream of Consciousnessby Barry Dainton, Naturalizing the Mindby Fred Dretske, Perceptionby Frank Jackson and Zombies and Consciousnessby Robert Kirk.

ABOUT THE INTERVIEWER
Richard Marshallis still biding his time.