The Defence of Qualia.

Dr Edmond Wright

In view of the excellent arguments that have been put forth recently in favour of qualia, internal sensory presentations, it would strike an impartial observer - one could imagine a future historian of philosophy - as extremely odd why so many philosophers who are opposed to qualia, that is, sensory experiences internal to the brain, have largely ignored those arguments in their own. There has been a fashionable assumption that any theory of perception which espouses qualia has long since been overcome by a number of 'formidable' objections, in particular, the Homunculus/Infinite Regress Objection, the Solipsism Objection, Austin's Illusion/Delusion Objection, the Ludicrousness-of-Colours-in-the-Brain Objection, the Indirect-Realist-has-to-assume-Direct-Realism Objection, the Impossibility-of-Comparing-Internal-with-External Objection, the Impossibility of Intrinsic Experience, and several more minor varieties of these. It is uncanny how they continue to be repeated, indeed, with a kind of automatism, evidenced by the fact that none of those who repeat them appear to have taken note of the answers to the objections. Indeed, they only appear to refer to those philosophers with whom they agree: it has long been insisted upon in the study of rhetoric that one of the weakest things to do in an argument is to ignore the main points made by one's opponent:

[it is] the wisest plan to state Objections in their full force; at least, wherever there does exist a satisfactory answer to them; otherwise, those who hear them stated more strongly than by the uncandid advocate who had undertaken to repel them, will naturally enough conclude that they are unanswerable. It is but a momentary and ineffective triumph that can be obtained by manœuvres like those of Turnus's charioteer, who furiously chased the feeble stragglers of the army, and evaded the main front of the battle
(Whateley, 1828, 175) Some philosophers have specifically noted how prone anti-qualia people are to attacking imaginary opponents: J. B. Maund complains of their failure to flesh out their arguments by reference to the positions actually being taken (Maund, 1993, 45-6); Jonathan Harrison accuses the anti-qualia cohort of 'learning about Sense-Datum Theory from its critics without reading the work of Sense-Datum theorists themselves' (Harrison, 1993, 20). It is not too much to say that it has become something of a scandal in philosophy that there should be such a blinkered response. One can hardly open an introduction to philosophy today without being taken through the old Objections as if they still retained their force. One wonders how many sceptical undergraduates are being led to question this supposedly received opinion among their mentors.

The distinction between the uninterpreted field and what can be selected from it is perhaps one of the most fruitful to have been made in recent inquiries into the nature of the mind. It can be found in neurophysiological research: John R. Smythies speaks of the essential distinction now being drawn between phenomenal and epistemic perception (Smythies, 1993, 208). In philosophy the field-and-representation distinction was perhaps first adumbrated by C. D. Broad when he spoke of perceptions that were attended to and those which were not: 'A sensum is not something that exists in isolation; it is a differentiated part of a bigger and more enduring whole, viz., of a sense-field, which is itself a mere cross-section of a sense-history ' (1923, 195). If his technical terms look daunting, a better view of their interpretation can be gained if a thought-experiment of J. R. Smythies (1956, 40-42) is taken into consideration: taking up the idea that a television system might be usefully applied as an analogy to the perceptual system, he imagined a situation in which a population of persons were wearing, unbeknownst to themselves, the equivalent of 'virtual reality' hoods, with the difference that the input to those hoods came from two miniature cameras set in front of them. He was able to use this illustration to argue convincingly that those agents would be able to interact in the world without any difficulty, even though they only had an indirect causal access to the continuum existing around them. Apply the situation to what Broad said: the 'sense-field' can be taken to be the screens of the cathode-ray-tubes; the 'sense-history', the sequence of changes upon those screens; a 'sensum' an area actually picked out on that screen. The field/representation distinction can be readily detected, for it is possible to distinguish the actual state of the phosphor cells on the screen from what the observers take themselves to be seeing on it. Furthermore, if one were to make the input on those screen something purely abstract in the painter's sense, a phantasmagoria of computer-generated imagery, the wearer (unaware that he is wearing the 'virtual reality' hood) would be exposed to non-epistemic imagery. This is a demonstration of the empirical possibility of non-epistemic experience that has been commonly denied in current philosophy (Rorty, 1980, 154; Davidson, 1989, 170; Harman, 1990, 39-40; McDowell, 1994, 24-45). One of the very latest books on the philosophy of perception begins with the declaration that ' a sensation of red is first and foremost a sensation of an object that is red' (Clark, 2000, 2).

A significant advance was made by J. B. Maund (1975), who was the first to point out the logical confusions that arise if one tries to use the same descriptive terms of the screen as of what is selected from the screen; it would be like trying to describe the state of the phosphor cells on the TV-screen by means of the terms used to describe what can be seen on it. As ordinary TV viewers we have no immediate way of referring to the states of the screen except by speaking about what things appear on it, even though one can actually get close to the screen and observe the criss-cross matrix, and, if one attends closely enough, one can ignore what the screen is ostensibly representing. Only a neurophysiologist will be able to describe the visual field at that level: for the ordinary observer the field will remain ineffable. The inability to acknowledge this distinction, which the TV Analogy makes plain, has led some to use the ineffability-for-the-observer as an excuse for dismissing the notion of an inner field as occult, when in principle it could be scientifically described.

There has been resistance to the TV Analogy, but this is due to a misunderstanding of what criteria are relevant in the metaphor. As will be argued below, within his theory, TV screens are not really coloured, so that pictorial resemblance is not being claimed; nor could they exist within a brain with eyes in front of them, but that is not being claimed either. The only relevant criterion is their presenting a display which is structurally isomorphic to conditions at the input, that is, in some not necessarily direct ratio, though being different in character, it varies concomitantly with the input. In the case of TV screens, the uncoloured state of their phosphor cells is indirectly isomorphic to the uncoloured input of light-rays into the camera: in the case of human sensing, the coloured state of the neural display is indirectly isomorphic to the uncoloured state of the light-ray input into the eyes.

In the 80's Virgil C. Aldrich was also concerning himself with the logic of representation. He drew attention to the fact that in a picture there are two aspects to consider, what one sees in the picture and the body of the picture itself. In an important article (Aldrich, 1980) he took up the same distinction made by Maund, that between the field in which the picture appears and the field of the picture itself. Aldrich was using the argument against that type of physicalism that rejects phenomena, for he concluded that, since the thrust of physicalism was towards a single field, it could not logically cope with the notion of such a distinction. This correctly points to the problem of how the supposed causal path between 'object' and perceiver is to be described, for Aldrich shows that there is an epistemological break between stimulus and response. This is why he is consistent in arguing that in this 'primary field' (1980, 52) there is no given perceptual awareness. It is rather that the intention of the observer must be taken into consideration when the actual sorting out of what is to constitute a represented object is examined, an assertion that recalls Roy Wood Sellars' dictum that the physical existent is not an object in its own right, but is made so by 'the selective activity of the percipient organism' (1919, pp. 418-19). This is precisely the same as what Aldrich argues (1980, 56). Clearly such an assertion is relevant to the whole issue of how representations can be 'about' anything. This disposes of those who believe that the supporter of qualia must hold to an Object-Causal theory of Perception, when all that need be proposed is a 'whole input/whole-field' causal theory, ignoring at the sensory level the selection-of-entities process, the perceiving itself, which may or may not be being applied.

The question of sentience (or 'qualia', secondary qualities, sensa, the phenomenal) is moving to the forefront of discussion. Nicholas Humphrey's attempt to explain consciousness (1992), for example, has the merit of trying to give the sensory its proper place, whereas Colin McGinn in criticizing him believes that there is little sense in trying to consider how the visual field could be inspected, for he thinks that it would be a matter of examining one's retina from the inside (1992, 18). There has been much consideration, particularly in the philosophy of mind, about what can be said in defence of functionalist and computationalist positions if qualia do exist, but little about the arguments for and against them. Even Thomas Nagel's defence of the subjective (1981) is largely about the difficulties of leaving phenomenal aspects out of the philosophy of mind than an attempt to address their place in the structure of perception.

As regards the common objections themselves, they can all be shown to be attacking the wrong target as far as the present theory is concerned. The keys to overturning them lie in (1) holding to the sensory fields being in a non-epistemic state which is structurally isomorphic to input conditions at the sensory organs, providing bare, non-mental evidence in which natural signs can be detected (see 'sensing as Non-Epistemic' for an explanation of the terms 'non-epistemic' and 'structurally isomorphic') and (2) seeing 'common' perceptions as the projected co-ordination of differing motivated selections from the differing non-epistemic evidence within each person's brain
(see 'Perceiving as Epistemic'). Let us go through the objections one by one.

(I) The Homunculus/Infinite Regress Objection (including (II) the objection that a pictorial resemblance between input and sensory display is impossible)

This is a very old objection. It was first mooted by the 19th-century philosopher Hermann Lotze (Lotze, 1884, 492-3). It claims that any theory which proposes that there is a sensory visual presentation in the brain is doomed to an infinite regress. The argument then goes, if there is a screen in the brain upon which a picture of the outer world is displayed, there would have to be another viewer of the picture, a homunculus with his own set of eyes and therefore another screen in his head, ad. inf. (for a modern statement of this view see Gilbert Ryle 1966/1949, 203). There have been additions to this argument: Alan Millar (in a talk to the Cambridge Moral Sciences Club, 1995) and Zenon Pylyshyn (2002) ask how the inner observer could move his eyes over the picture to inspect various parts of it; J. K. O'Regan and A. Noë ask how there could be 'red neurons' in the brain (O'Regan and Noë, 2002; also for the same kind of question Dennett, 1992, 28; Tye, 1992, 159; Kirk, 1994, 9-10)). This is no advance on the question put over a hundred years ago by F. H. Bradley to Thomas Case, 1888), namely, 'when I smell a smell, I am not aware of the stinking state of my own nervous system' (see Price, 1961, 127). Pylyshyn also asks how there could possibly be a physical surface in the brain to match the physical surfaces outside.

To take vision as our example. The colour registration in the brain 'structurally isomorphic' to the light-ray input at the sensors, that is, it is 'differentially correlated' to it, not necessarily in direct ratio (Sellars, 1932, 86). This implies that sensory phenomena of any kind are utterly unlike what triggers them, so that there is no external 'colour' to match neural colour (i. e. the actually experienced red), no external 'smell' to match neural smell, and so on for all the modalities. There is only a complex causal connection between the whole field and the whole input at the sensory organ.

To get structural isomorphism clear consider this example: there is a structural isomorphism between the sound-track of a movie film, the impulses in the wires leading to the loudspeakers, and the sound which the audience hear, but no one would say that any of these were like each other except in the variation of their intensities. In the same way, the internal colour distribution could be shown empirically to vary concomitantly with the external input, but to be in no way like it. A fortiori, since real external pictures are therefore actually uncoloured, there cannot be pictures in the brain. With that qualification one can readily agree with the objection. But, nevertheless, this theory can still claim without inconsistency that there is a neural-colour registration in the brain. So it is quite true that a pictorial resemblance is impossible but that is no argument against there being a coloured inner visual display. The external world is actually 'unpicturable', as Virgil Aldrich noted over twenty years ago (Aldrich, 1980, 55). Thus this also disposes of (III) the Ludicrousness-of-Colours-in-the-Brain Objection as a case of the fallacy of ignoratio elenchi, that of attacking what has not be proposed.

One thing John Locke pointed out is pertinent here (Essay, II, viii, 6). We still experience a colour when no input is coming into the eyes. If you are suddenly plunged into a completely dark room, you do not experience nothing - you experience Blackness (see Brain, 1951, 10, 15). Now obviously there is no external object for the brain to recognize and yet we are still sensing. Here too is a pure non-epistemic experience, the sort of thing anti-qualians deny. If someone protest that you would know that you were in a dark room, one can easily change the circumstances: you wake after a drugged sleep, not knowing where you are and whether something has been done to your eyes. You would not be able to know anything, except that, though you were still sensing, you were not pereiving. That Blackness is unmistakably an inner experience, because there is no input to the eyes whatsoever. Anyone who is tempted to think that all lightless spaces in all parts of the universe are actually 'black' has some scientific explaining to do! But the problem is no different for those who think that external objects are really coloured.

The approach adopted here also renders useless the complaint that some inner 'eye' would have to scan a field in the manner of the real eye, for the direct sensory experience can be of that very scanning without any supposed movement of a supposed eye (does a TV screen have to move to 'scan' round a cup?). In any case, real eyes have evolved to pick up light rays, which are uncoloured, and there are no light rays in the brain. Therefore, there can be no infinite regress, for no homunculus with eyes of his own is required: the visual sensing in the back of the brain is a direct experience. It is an evidential field from which our motivational module can select gestalts, but this does not require a looking-about over a picture since there is no picture and no eyes to look. Ryle's attempt to maintain that one would have to have another sensation to sense a sensation as an argument (Ryle, 1966/1949, 203) remains as Ayer described it, 'very weak' (1957, 107). Ryle could only think that, in the argument for inner presentations, a literal inner eye was involved that scanned some metaphorical 'mental linen', as he put it.

This counter to the objection also renders explicable the sensing by snakes of infra-red rays or sharks of electrostatic fields. There will be some inner experience to correspond to the input but it will not be like, say, the infra-red rays: perhaps it will be another colour, but, whatever it is, it will be only vary with the input, not match it colour for colour. Infra-red rays like all other rays of the electromagnetic spectrum (including gamma-rays, X-rays, light-rays, and radio-waves) have no colour whatsoever. Nor obviously, since they are heat-rays, can any real heat be in the snake's sense-field in its brain, only something varying concomitantly with the heat. When we see on the television screen films that have been taken through an infra-red camera, we see the warm and hot portions of the view as shades of green; the green is structurally isomorphic to the heat but does not resemble it - does your TV screen get hot?!

Under this theory there is no riddle about the world being uncoloured when no animals were in existence: it never was, nor is, nor can be coloured. We manage reasonably well to act in the world with our internally coloured, structurally isomorphic evidence. There is no riddle about wondering what the difference could be between a ray of 200 nanometres (invisible far-ultraviolet) and one of 420 (visible violet) that has nothing to do with animal eyes, for obviously there is no scientific difference between them except in wavelength and frequency.

The experience of stereoscopic space also must bear no pictorial resemblance to the real space with which it is correlated (stereoscopic space can be turned inside out, see Nakajima and Shimojo, 1981). There is no point therefore in Pylyshyn's claiming that there cannot be a registration because it could not be on 'a physical surface' in the brain, for there is no requirement that the neural registration be on a physical surface similar to that of an external object; structural isomorphism rules out any such similarity. Kant was on the right track (Kant, 1787/1964, p. 69): sensory stereoscopic space is thus not like real space (although in some form it is in real space in the brain. Price (1961/1932, 111) thought it would be 'awkward' if we had to distinguish 'Sensible Space' from 'Physical Space' since he believed that this would force us to conclude that 'sense-data were nowhere in the physical world' - but there is no such entailment. The inner presentation can be structurally isomorphic without matching external space in a direct way and yet still exist in the brain in physical space. Space we are now told by the theoretical physicists may consist of 11 dimensions, so there can be certainly no direct resemblance between real space and our visual 3-D space. Visual 3-D space, in any case can be turned inside out and yet still be isomorphic (Shimojo and Nakajima, 1981, who placed reversing spectacles on a subject so that the right eye's input went into the left eye and vice versa; this turned the world inside out, objects in the Real distance appearing stereoscopically 'close' and vice versa).

(IV) The Objection that Indirect Realists have to assume Direct Realism in order to state their Argument

A typical statement of this objection has been made by Stephen Wilcox and Stuart Katz (Wilcox and Katz, 1984), but there is a clone of it in Hilary Putnam's objection to naturalized epistemology (Putnam, 1982). They claim that it is a self-defeating absurdity for the qualia supporters to espouse Indirect Realism precisely because they are covertly relying on Direct Realism to make the assertion at all. The whole causal chain revealed by the Indirect Realist as proving his case is couched in Direct Realist terms: he is therefore bound to two rival epistemologies, the ostensible one of his argument in which objectivity is ever at one remove, and the concealed one of his practice, which actually presupposes objects. The philosopher or psychologist 'has committed the most common of philosophical follies: he has failed to examine his own presuppositions' (Wilcox and Katz, 153).

The core of this new argument for the place of qualia in human life is their structurally isomorphic nature. They are bare evidence causally connected to the whole-field input at the sensory organs, from which each body, via the processes of motivation, has to learn what to select as significant for action, that is, learn to treat regions of it as 'natural signs'. This includes the entities of the self, other persons, all objects and all recognizable properties. Out of the basic difference at the both sensory and the perceptual levels, an evolutionary advantage for all species, members of the human species have learned how to update (hopefully) each other's selections by means of symbolic communication. The central trick is to bring the differing selections into a rough co-ordination by behaving as if they had achieved a perfect co-ordination. Another way of putting it is to say that, in each spoken sentence, the logical subject of the sentence has to be treated initially as if it referred with absolute precision (in perfect 'synchrony', as the linguists would say). For example, the Speaker might begin with 'The tree in the garden . . .' Once this presupposed perfect superimposition of the differing perspectives has been assumed, then the Speaker can introduce his or her correction of it the logical predicate - completing the sentence with, say, 'has been chopped down.' (The logical subject is not always the grammatical subject, for if the Hearer had asked 'What has been chopped down?', the logical subject would have been a chopping-down of something, and the correcting predicate would have been 'the tree in the garden', a point noted by John Cook Wilson, 1926, 123-6). So the first move in the communication process is for both partners in dialogue to presuppose, to assume that they have together successfully sorted out the Real into the 'reality' of singular true referents. Co-ordination would be impossible without this methodological fiction of absolute coincidence.

It must now be seen why the Objection misses its aim, for the presupposition of singular referents is a initial necessity for communication, but, being a convenient fiction, it is not wholly true, the reason being that each partner in dialogue has his or her own set of criteria of selection, the differences of which cannot all be mutually salient to them. Everyone, in order to speak at all, has to presuppose a 'direct realism' of entities, but none of that implied that there are such perfectly singular entities. Instead what exists are rough co-ordinations on the viscosities and nodules of the ever-changing Real. Objectivity, even of selfhood, is not existence: it is only a socially maintained overlapping of selections from existence. For any 'common' entity, each person has indeed made a selection from the Real, but what is not known is its differences from that of others, and there is also the block that the bare evidence cannot turn into information. Notice the insidious temptation here: each has certainly a portion of the Real that he or she has sorted out, so there is an easy slippage in thought that turns the selection into 'existence'; But what is not known is the degree to which it coincides with others' selections from roughly the same region of the Real. 'I think: therefore I am' is thus fatally mistaken if it claims a true objectivity for the self that speaks it. Nevertheless, to enunciate anything useful, even of the Argument for Qualia, one has to join in the language game of presupposing a perfectly singular entity, but this is a harmless necessity of speaking at all and does not undermine the argument.

As a valuable corollary, we are now able to meet Putnam's objections to naturalized epistemology. First, he said that it begged the question of a Correspondence Theory of Truth, but, if the system is one in which such correspondence has to be mutually projected as a useful make-believe, then his objection loses its force. Second, he said that naturalized epistemology 'presupposed' reason, but his very own words point to the weakness of his argument, since reason is essentially constructed through an active presupposing of universally agreed unities and deductive validities. Lastly, he said naturalized epistemology can find no base for normativity because it was driven back to a circular argument in which the subject strives to be his own guarantee of reliability (Putnam, 1983, 245). But the present theory yields verdicts whose reliability requires mutual co-operation to establish. As it should have become obvious that this theory, New Critical Realism (see for the name, adopted in honour of Roy Wood Sellars, Wright, 1994, 478) is a version of naturalized epistemology, this is a very helpful entailment.

(V) The Solipsism or 'Veil-of-Perception' Objection

The argument claims that to admit that sensory fields were internal to the brain would leave one with no access to external fact. The whole of reality would disappear in a Berkleian dream-world, with a solitary self having no basis for either the reality of what it experiences nor its own reality (John McDowell, 1994, 42).

If sensing is an involuntary, material occurrence, bare evidence containing no knowledge, it can exist without a self being present at all. Indeed, before any learning has taken place in the organism it just exists as evidence that has not been sorted out in any way (innate instinctive reactions are not knowledge). One can also imagine an unfortunate mutation born without any connection between its pleasure/pain system and its sensory fields, that is, there would be no placing into memory of any experience that produced pleasure or pain. No knowledge would be forthcoming, but the sensing would continue as before. So the notion of an internal sensory field does not entail an observing self. A self is one of the entities that has to be learned from the history of the body's experiences. Thus, the accusation that solipsism is implied by such a theory passes it by, for there need be no solus-ipse present while sensing is in progress. Again the objection fails as an example of the logical fallacy of ignoratio elenchi.

A further corollary strengthens the argument. The entity the self is dependent on experience with others for both its inception and its continuing development. There comes a continual adjustment affecting how the sensory fields, especially those which register the states of the body, are to be treated as a 'self'. Whenever such an adjustment takes place at the behest of another person, even of a simple perception of an external object (a perception which, under this theory, is a product of motivation), the self suffering the adjustment has the experience of an external corrector being able to invade his or her own sensory fields and re-order them anew in a way not considered before. This also disposes of (VI) the Impossibility-of-comparing-Internal-with-External Objection (Hoernlé, 1926; Searle, 1983, 59; Kelley, 1986, 108) because it establishes that the only possible comparison is a structurally isomorphic one, for it is this that enables another person to effect the updating of one's own perception. The evidence was there as sensed, but it had not been perceived. The other is able to work from behind the supposed 'veil of perception' to establish and then alter the self observing, proving several things simultaneously:

the existence of the internal field as Real evidence separate from the self; the fact that the other who is acknowledged as altering the nature of the 'self' observing must possess a comparable field with comparable access to what lies 'beyond the veil'; the existence of the external causes at the sensory organs (see for more detail, Wright, 1990, 78). So one can have an internal sensory field without any danger of being trapped behind a 'veil of perception'.

(VII) Austin's Illusion/Delusion Objection

The old Sense-Datum theorists had used what they called the 'Argument from Illusion' to support the claim that sensing was internal. They would point out that there are many occasions when we have the sensory experience of an object and there is no corresponding object in the external world to match it. One might see a mirage of an oasis and there be no oasis at all in actual fact; one sees a galaxy that, because of the time-lag of seven million years, may have not be now where we take it to be, or it may even have gone out of existence. They have pointed out correctly that we see everything as it was in the past, because, no matter how fast light travels, it still takes real time to reach us, so that even things within reach are seen a nano-fraction of a second earlier than what they are at the instant of seeing. One sees a stick that looks bent in water, but we know that it is the appearance is the result of refraction, so the appearance and the reality come away from each other and we have the 'veil of perception'. Then there are after-images, which plainly occur within the visual field, but are not a part of the external world. There are finally the experiences of dreams and mental imagery, and, more peculiarly of hallucination, as of the drunkard who sees pink elephants running over the ceiling.

Austin's complaint is that the Sense-Datum theorists were confusing illusion with delusion. He says 'it positively trades on not distinguishing illusion from delusion' (Austin,
1976/1962, 25). All these illusion cases, he argues, are subtly turned into a delusion case, where, instead of a mix-up over an object, the theory tries to suggest that we are more fundamentally deluded into thinking that there is an immaterial something that interposes itself between ourselves and the world. Illusions, he claims, are misreadings, which on greater familiarity are recognized for what they are, so that you may have thought that the stick was bent, but, once you have had the phenomenon scientifically explained, the illusoriness falls way. To suggest that it does not is to try to turn an illusion into a delusion.

Now one can certainly agree with Austin that the Argument from Illusion, when it confines itself to those cases of misinterpretation like the Bent Stick or that of Time-Lag, does not prove the existence of an inner field. If the sensory fields are brute evidence, then there is no reason why we cannot use such a supposedly 'illusory' state of the field to find out something of use to us. As long as a causal path of some kind, however indirect, even in time, can be traced, something verifiable can be elicited. Someone 'familiarly' acquainted with the refractive indices of various liquids would be able to pass along a row of jars and say 'Alcohol, water, paraffin, amyl acetate . . .' etc., etc. As regards the Time-Lag Argument, nothing prevents an astronomer learning scientific knowledge from observing a galaxy in its state seven million years ago. Even the sun in broad daylight could be described as an illusion since it cannot be precisely in the position where we see it, light from it having taken eight minutes to reach Earth, and the refractive power of the atmosphere shifting it from its 'true' place. The elliptical state of the sensed image of a penny seen at an angle can be taken as good evidence for the viability of the perception of a round penny. So, for these illusion cases this aspect of his argument has to be granted to Austin.

What is really significant about the illusion cases is their pointing to our ability to change our selections from the field. It is obvious that one perceiver can make a different selection out of his or her sensory field from that of another of what had been thought of as the 'same object', for it is not only obvious cases like bent sticks that must be included in the generality here, but actual puzzling ones in which new interpretations are made that hopefully advance our knowledge. Those who try to say that 'all our "seeing-as's" are just versions of "seeing-that"' are like Austin in playing this fact down. In your youth when looking down from a bridge over a stream did you never play at seeing, not the water going under the bridge, but the bridge moving over the water? - so it is not impossible for some early scientist on one special occasion to have seen for the first time, not the sun going down behind some very distant mountains, but the mountains rising over the sun, and consequently come to the novel conclusion that the Earth it was that revolved. Austin could not say of this example that this gestalt-switch was an illusion which 'familiarity' would banish. It looks as though Austin was unconsciously endeavouring to ignore this plain fact of the mutual updating of perceptions. One person's "seeing-as" is never the same as someone else's, though we have to behave as if we are all "seeing-that" in the same way in order to co-ordinate them for communication and action purposes (see Objection III).

But, in particular, Austin is on very insecure ground when he tries to dismiss after-images, mental images, dreams and hallucinations, for this was a positive claim and not a negative one. It is simply a non-argument to try to dismiss dreams as indicative of an inner experience by saying 'dreams are dreams' (Austin, 1976/1962, 27). That is the equivalent of those Jesuits who would not look through Galileo's telescope. Note that logically it is precisely parallel to saying of a TV screen showing a cartoon that the screen cannot be real because it is showing an 'unreal' cartoon. His involved discussion of the meaning of 'real' (64-77) collapses to the same riposte. A cathode-ray-tube's screen can show all kinds of 'unreal, illusory' sights - interference patterns, abstract screen-savers, videos, the output of infra-red cameras, a wax-model of President Bush, computer-created dinosaurs - but not one of these would induce us to say that the screen itself was a 'delusion'. Similarly, the inner sensory field can show dreams, hallucinations, after-images, mental-imagery, memory-images, as well as open-eye vision, and all of them be on a Real inner, non-pictorial display.

One point of glaring weakness in Austin's case is the attempt to downgrade the vividness of dreams and mental imagery. Of dreams, for example, he says,

I may have the experience (dubbed 'delusive' presumably) of dreaming that I am being presented to the Pope. Could it seriously be suggested that having this dream is 'qualitatively indistinguishable' from actually being presented to the Pope? (Austin 1976/1962, 48) It is obvious that Austin was not acquainted with texts in the psychology of dreaming (Hadfield, 1954; Lee and Mayes, 1973). The fact that he himself may not have had any such vivid dreams is dismissable anecdotal evidence (yet it is unlikely that he never had any nightmares; Dennett takes the same stance, for he denies outright that someone could have a hallucination of a ghost that 'talked back'; Dennett, 1992, 7). Dreams and mental imagery can certainly be as vivid as real-life experience and can include tactile, kinesthetic and olfactory sensations. A. R. Luria's 'mnemonist' was able to read off numbers from a mental image of a table of them to order, namely, vertically, diagonally, or horizontally, but what was important for the present argument was the following: he was certainly not just remembering them propositionally, for Luria points out that, where one of them had not been written distinctly on the paper he had had to mentally-image,

he was liable to 'misread' it, take a 3 for an 8, for example, or a 4 for a 9 (Luria, 1969, 23) This is an empirical proof that he was reading from a vivid mental image of the actual writing, iconically stored in his brain apart from the 'propositional' information in the text
(see also for evidence of inspecting vivid aural mental imagery Wright, 1983).

Iconic storage is, of course, non-epistemic storage, an imprint of the whole field regardless of 'what' can be selected from it by the motivated knowledge/perception system. Zenon Pylyshyn denies the possibility of such iconic storage as being beyond the brain's storage capacity (Pylyshyn, 1973, 9). He has no empirical evidence for this unlikely claim; indeed, when Mozart remembered the whole of a piece of choral music that he heard in Rome and was able to write down later in Vienna, it was not a propositional written record of the piece that he was remembering! Many musicians have this ability in varying degrees, and are able to listen with pleasure to their inner 'recordings' of the music - a fortiori, if it is with pleasure, they are appreciating, perhaps for the first time, subtle patternings in the music that they had not picked out before. So it cannot be that they 'tacitly' knew these novel aspects beforehand, as Pylyshyn would be forced to maintain (Pylyshyn, 1981, 24; Wright, 1983)).

With regard to Austin and Pylyshyn, it is worth recalling what Bertrand Russell said of the Old-Behaviourist Watson's denial of mental imagery. Russell, having read that it was found that the older and the more educated you were, the less your ability to call up mental imagery, said, "Professor Watson is a very educated man."

Another empirical fact entirely ignored by the anti-qualia philosophers is the fact that one can dream with one's eyes open. A person exceptionally tired was being driven home by a friend in the pouring rain. As he did not want to fall asleep in his friend's car, he strove manfully to keep his eyes open. This did not defeat his dreaming module, for he found himself dreaming that he was watching an impressive firework display in which a number of amber-coloured rockets went off in succession. He woke up with a start on his friend's suddenly speaking to him, to discover that what he had taken for 'rockets' in his dream was actually the appearance of sodium-vapour lamps seen through the windscreen as the wipers went back and forth. This is an interesting case for it contradicts the claims of those Wittgensteinians like Norman Malcolm who tried to argue away the inner experience of dreams (Malcolm, 1959). Here there was empirical confirmation by another of the Real non-epistemic base of the dream, a succession of brilliant sparking lights, seen as 'sodium-vapour lamps' by the wide-awake driver, and seen as 'amber-coloured rockets' by the dreamer. It certainly seems to point to dreams and open-eye vision being on the same inner display.

Austin could not but think of the world of 'reality' being conveniently divided up between thoroughly singular 'dry-goods', as he liked to put it, and mistaken perceptions, a clear division of truth and falsity. His 'delusion' attack is really the Solipsism Objection is another form, and it has no understanding of the possibility of sensing being non-epistemic, non-mental - just a part of the Real itself, having no necessary relation to the process of knowing, however much it has evolved to present evidence to the motivational knowing module. Those brain-damaged persons known as 'agnosics' (translated from the Greek - 'not-knowers') can see, that is sense, perfectly well but are not able to look, that is, to perceive, to use either their powers of attention or their memory's motivations to select any entity whatever. They are condemned never to interpret the bare evidence. Austin's argument fails utterly to explain their state since he resists the separation of sensing and perceiving.

However, if we think with Piaget as our range of familiar 'objects' being the present state of our co-operative learning venture on the Real (sensory) evidence with which we are presented, then no such clear distinction can be made from the philosophical position. However much we must perform the truth-falsity game in the position of ordinary life in order to co-ordinate these differing selections from the Real that 'objects', that they are singular, have definite boundaries spatial, temporal and qualitative that are
(impossibly) the same for all of us, they never capture the whole of the brute evidence - How could they since the evidence is (valuably) different for each of us at the very moment when we have to behave as if it is not? This was the cogent point made by A. J. Ayer is his answer to Austin (Ayer, 1969, 131), that our perceptions manifestly do not capture all the evidence that is available. For Austin, if you see a pig, that's it - there is nothing more you can learn about it! So there is always more evidence there than we know how to interpret, for there is no end to interpretation. Perceptual selections, in theory, could be moved about infinitely; it is like trying to count all the points in a straight line.

Austin's is an Objectivist argument for it all depends on his assuming that his opponent must be arguing for a pictorial resemblance between external 'objects' and internal 'objects'. There were of course some Sense-Datum theorists who wrote of 'the datum' corresponding to 'the thing' (e. g. Price with his talk of the datum of a red tomato,
1961/1932, 3), though not all of them did (C. D. Broad preferred to think of 'objects as hypotheses'; Broad, 1951/1923, 152). If 'objects' are the co-operative selections from the Real as has been argued in the present theory, an attempt to bring our individual guesses into co-ordination upon the evidence from the undoubted Real, then Austin's complaints do not upset the notion of an internal field. Martin Lean made the same complaints as Austin before him when attacking C. D. Broad, but his argument all depends on the same naïve, even occult belief in the 'commonsense' reality of objects, all without any inquiry into the construction of 'common' sense (Lean, 1953).

What is suspicious about this Direct Realist insistence on the object-as-currently-understood is that it magically sacralizes the received opinion: thus (1) it is authoritarian, privileging the speaker's traditionally favoured objectifications of the world as beyond correction, and (2) it ignores the fact that objectification is actually based on the Idealization of Reciprocity, that 'naive and unreflecting faith' (Rommetveit, 1978, 31) between persons, which makes every identification, including that of oneself, an act of trust in the other. All language is based on this social trust, as was argued in the Times Literary Supplement letter on the opening webpage, so the authoritarian objectivist is trying to turn trustful co-operation into a command-obedience structure with him as the master. But we do not need Hegel to tell us that masters can be corrected by slaves.

Here an argumentum ad hominem is fully appropriate, for those who are fearful of self-correction are at the mercy of a fear-generated narcissism. This ignoring of the trust is exactly what gives their argument the specious force it seems to have, for to say with Austin and Ryle 'talk of deception only makes sense against a background of general non-deception' is merely to repeat the Idealization of Reciprocity - that we have both singled out exactly the same entity - without realising that that agreement is only an ongoing trust which has to acknowledge the risk that both may not have understood it in the same way. Our criteria of selection might be significantly different without either of us knowing. But the Austin-Ryle mantra insidiously suggests that anyone who tries to show that if someone who tries to claim that something in the agreed 'background' of 'non-deception' is actually deceptive is an untrustworthy person. This is where the spurious strength of this bland assertion lies: if you challenge the received opinion, you must be a devious relativist. The truth is that the accuser is the one lacking in trust in his human partner in the language game. He is really no different from the relativist he berates, for both are Humpty Dumptys, their words being what they privately choose them to mean without leaving anything open to mutual correction: for the authoritarian it is the received opinion, for the relativist it is the 'private' meaning. There are no private meanings, for we are not and cannot be wholly aware of the intimate sensory evidence for those meanings in our own heads. To this degree Putnam is right when he says meanings depend upon extension (Putnam, 1988, 49) for the sensory evidence does move concomitantly with the whole-field input, but wrong to believe in given objects; he actually says that one should see science as 'a story of successive changes of belief about the same objects, not as a story of successive "changes of meaning"' (ibid., 12; my emphasis). He should have taken more notice of his own metaphor of 'story', for in a story the transformation at the core of it may even be of the singularity of the 'thing'. This is like David Woodruff Smith's false reliance on a 'determinable x' that remains the same from the old perception to the new (Woodruff Smith, 1982, 52). But this is insufficiently general, for there are many cases of perceptual change in which the original 'singularity' is abandoned and entirely new boundaries are selected (such that there may be two or more 'objects' where one was thought to be, or an overlap with another). Here is an example:

"See that bird on the tree there?" "Yes." "Well, it's really two-and-a-bit leaves." As the ninth-century Indian philosopher Dignaga wrote, 'Even "this" may be a case of mistaken identity' (Matilal, 1986, 332).

This also disposes of (VIII) The objection that Intrinsic Experience is impossible. Gilbert Harman maintains that it is impossible to distinguish sensory features from the object with which they are associated. For example, he says that, when we look at a tree, 'the sensed qualities are always features of the presented tree' (Harman 1990,
39). The reader will probably by now see the form of the mistake here, a carrying of the Idealization of Reciprocity into the authoritarian's superstitious temple: it forgets that objectivity is based on a social trust that accepts risk, ignoring the sensory evidence upon which all such identifications depend, from which others may be making the selection (here of 'the tree') with a hidden and different set of criteria.

Consider another of Helmholtz's observations. Butchers often surround the red meat that they have on sale with green paper. The effect is to enhance the red colour of the meat because of the after-image effect of the green as the eyes move over the red of the meat. Now few, if any, customers are aware of this transformation, and so all would agree with Harman that 'the sensed qualities are always features of the presented meat.' But it was patently not in this case.

Another example. Daniel Dennett believes that once one has taught the cry of the osprey to the point that they have passed 'all the objective testing procedures', then no queer (in the Wittgensteinian sense) private understandings could remain, and public meaning would be guaranteed (Dennett, 1985). However, let the cry of the osprey have characteristic overtones in the 16Hz to 20Hz region. Our pupil, being young, happens to be able to hear these overtones which have been present for him on all the exemplary occasions, though without either his own or his teacher's conscious knowledge (his teacher, being over 40 years old could not hear those overtones at all). On a day of torrential rain when the rest of us only hear the noise of the downpour, the pupil claims to hear the osprey. The majority deny it and are amazed that after his thorough Wittgensteinian training he can come out with such a queer private judgement. The authoritarian objectivist is confused about the word 'public' in the phrase 'public meaning': he thinks it implies the notion of being arrived at by a majority decision, but, on the contrary, it implies the involvement of many persons, one of whom may turn out to know better than the rest !

The same point about the 'object' being an act of trusting co-ordination with the other on the Real evidence, comes out in Austin's treatment of hallucination and after-images. In the case of hallucination, like many another Direct Realist (from Gram, 1983 to Huemer, 2001), he never considers the well-known empirical case of a non-epistemic hallucination that is, an hallucination 'abstract' in the painter's sense, rather like a computer-generated screen-saver, one representing nothing at all. Price
(1961/1932, 28), following Broad (1923, 284) called them 'wild' sense-data. In spite of their being discussed in the debate, Austin and his like completely ignore the possibility, bound as they are to given objecthood, and confine their consideration to instances such as the drunkard's pink elephants; for them, those experiencing an hallucination are never lucid nor can an hallucination be non-objectified. R. J. Hirst says that all those who suffer hallucinations are 'confused' (Hirst, 1959, 44). Moltke Gram is so convinced he never considers the possibility of non-objectified hallucination: 'an imaginary or hallucinatory f is the same kind of f that is encountered in veridical perception' (Gram, 1983, 121). But if they had read their Price (Price, 1964, 18) it is perfectly possible to be fully conscious and critically observant when having a hallucination of something that is not 'an object' at all. can instance the migraine sufferers and others who experience 'fortification patterns', bands of flickering zigzags which slowly expand to cover wide regions of the visual field. The sufferers are not under any 'illusion'; if they are so inclined, they can inspect these non-objectified, that is, non-epistemic, patterns at leisure and attend to different aspects of the show, but what they will then be objectifying will be their own inner sensory experience without any connection to familiar external objects. None of these persons who experience such patterns are suffering from a 'delusion', nor in essence is what they do see propositionally present to them.

It also seems distinctly counter-intuitive to try to maintain that the rest of their visual field is not on an inner display, but the fortification pattern is. The rest of their visual field is being memory-selected into familiar 'reality' - that is all the difference. Just as on a TV screen we can see simultaneously a real person and a cartoon character (as in the Arnold Schwarzenegger film Action Man) without doubting for a moment the reality of the glass screen and the phosphor cells on its back, so too the combination of 'wild' hallucinations and normal perception strongly suggests that one display encompasses all visual experiences.

(XI) The Objections to the Double-Vision Argument Supporters of qualia have often used the fact that one's visual field often shows us double images as a proof that the internal field does not correspond directly to the external world. If we read our Helmholtz, we discover, perhaps to our surprise, that a considerable portion of our visual field is always covered by double images, the reason being that, when we focus on something near, all the background is doubled, and vice versa (Helmholtz, 1968, 177). The Direct Realists are not fazed by this argument, responding in a characteristic fashion that, in the case of double images, we just see one thing twice, that is all (Ryle, 1966/1949, 207; Pitcher, 1971, 41; Gram,
1983, 102-3; Huemer, 2001, 130-1). This is how Ryle puts it, but the same argument is used unchanged by his admirers:

The squinter, aware of his squint, who reports that it looks just as if there were two candles on the table, or that he might be seeing two candles, is describing how the single candle looks by referring to how pairs of candles regularly look to spectators who are not squinting; and if, not being aware of his squint, he says that there are two candles on the table, he is, in this case, misapplying just the same general recipe. (Ryle, 1966/1949, 207) Right through to Huemer, fifty years later - 'There is a single physical object you are seeing' (Huemer, 2001, 130-1) - the argument is as before.

However, to get the matter clear, one has to consider all the empirical facts. When the superimposition of the two eyes comes apart, as in squinting, there is an immediate loss of stereoscopic sensory space for the portion producing the 'doubling'. For the Fingers example, the two fingers are now only 2-dimensional, whereas for the normal superimposition, they were stereoscopically 'solid'. A sensory feature has changed but Ryle and the others make no mention of it. What does this omission imply? That they have completely forgotten - or were never aware - that the two images of 'the finger' were actually different at every point. The difference in the angle of incidence of the two eyes upon the same region of the Real produces a whole-field difference at the level of the continuum of the sensory field. So one cannot describe the 'two' candles as merely the 'same' candle seen in an illusory fashion, when there is a distinct non-epistemic difference in those portions of the visual field as in every other part of it. Notice that we can use Ryle's own unawareness of this as a proof of the non-epistemic state of the field because it was outside his knowledge at the very moment he believed that the 'object' description provided all the characterization necessary. One may observe something of this non-objectified difference by examining those figures constructed by Bela Julesz (1960) in which two squares of apparently random correlated dot patterns produce stereoscopically solid pictures when squinted together: a close inspection at the edge of the place where the 3-D form appears will show a tiny but distinct difference where a section has been shifted over. The same kind of difference is present in 'Magic-Eye' pictures, where the vertical computer-generated bands are different in the fine detail of their composition.

What is particularly relevant here is that, in the 'Magic-Eye' pictures, the 'field-determinateness' of the dots that makes them up is on an entirely different logical level from that of the 3-D forms which appear when we move our eyes away from them while maintaining our fixation. A computerized list of those dot-differences could be made, but it would make no mention of the 'pyramid' or 'leaping dog' that shows up in the 3-D picture. This is equivalent to the difference between the non-epistemic level of the sensing and the object-recognition level. Some anti-qualians have tried to deny the existence of the non-epistemic by saying that its bare presence becomes magically 'ineffable'
(see, for example, Daniel Dennett, 1985, 4): however, just as the point-states of the 'Magic-Eye' pictures could be propositionalized at the level of the field without reference to the 3-D objects, there is no logical bar to some neurophysiologist of the future being able to list the point-states of the inner visual field, in which case it would not be in the least 'ineffable'.

Another empirical fact which demonstrates how the evidence of the non-epistemic can produce surprising epistemic results that do not sit easily with a Direct Realism of objects. Helmholtz reported how the police can check for forgeries of banknotes by putting the note in question into a stereoscope with a genuine note. The fusing of the images of two genuine notes produces no alteration in appearance: the fused picture looks exactly like an ordinary note. The fusing of a forged note with a genuine one produces strange alterations in the stereoscopic appearance of 'the' note: certain letters seem to be standing an inch or two above or below the level of the paper; pictures have solid distortions; the detailed background patterns have strange dips and lumps in them. All these are the result of the mismatches in position of the forger's engravings with those of the genuine note: for example, a 'lump' in the pattern could show that the forged note's pattern was a little too far over to the left in a certain area. What is inexplicable from a Direct Realist stance here is how what appears to be a 'single' note can give non-illusory information about two notes; it is not a case of one thing looking double but of two things looking single! From the New Critical Realist point of view this is merely using the evidence of Real non-epistemic sensing in a perfectly understandable way, a tracing of indirect causal chains. As far as the Direct Realist is concerned there is nothing in the appearance of the forged note to indicate this. He would even have to say of all stereoscopic pairs, that they are only 'illusory' doublings of a 'single' object; I am at a loss to hazard a guess as to what he can say about the forged-note case, the two 'things' looking 'single'.

The non-epistemic difference between the two eyes' versions of the whole field can be made startlingly plain. Cover up the left eye and expose the right eye for a while to dazzling illumination, as of bright sunlight. Then go indoors, remove the blindfold from the left eye, look at a red candle, and squint. You will see a red candle (the left eye's) and a brownish-green candle (the right eye's), the latter being produced by the overstimulation of the cones of the retina. We can still agree with Pitcher that we are seeing 'one thing twice' (Pitcher, 1971, 41) and yet deny that this proves that there is no inner presentation, for we can see a difference that is not accounted for in the language of common 'reality'.

Finally, this objection can be seen to be nothing but a form of Objection III, because in its insistence on the 'common object', it is forgetting Ayer's insistence on the sensory and perceptual differences between person and person, which no amount of performed agreement in action can erase. The non-epistemic sensory experience of the other person, which he or she may be acting on without it being salient, out-in-the-open knowledge to both of you, is present in every 'common' identification, and this is because the evidence cannot turn into information.

References Aldrich, Virgil C. (1980) 'Mirrors, pictures, words, perceptions', Philosophy, 55, 39-56.

Austin, J. L. (1976/1962) Sense and Sensibilia, Ed. G. J. Warnock. Oxford: Oxford University Press.

Broad, C D (1951/1923), Scientific Thought, London: Kegan Paul, Trench and Trubner.

Case, Thomas (1888) Physical Realism, being an Analytical Philosophy from the Physical Objects of Science to the Physical Data of Sens, London: Longmans, Green.

Clark, Austen (2000) A Theory of Sentience, Oxford: Oxford University Press.

Davidson, D. (1989) 'The myth of the subjective', in Relativism: Interpretation and Confrontation, ed. M. Krausz. University of Notre Dame Press, pp. 159-72.

Dennett, Daniel C. (1985) Quining Qualia, Center for Cognitive Studies, Tufts University.

Dennett, Daniel C. (1992) Consciousness Explained, London: Penguin Press.

Gram, Moltke (1983) Direct Realism: A Study of Perception, The Hague: Martinus Nijhoff.

Hadfield, J. A. (1954) Dreams and Nightmares, Harmondsworth: Penguin Books.

Harman, Gilbert (1990) 'The Intrinsic Quality of Experience', in J. Tomberlin (ed.), Philosophical Perspectives, 4: Action Theory and the Philosophy of Mind, Atascadero, California: Ridgeview Pub. Co., pp 31-52.

Harrison, Jonathan (1993) 'Science, souls and sense-data', in E. Wright, New Representationalisms: Essays in the Philosophy of Perception, Aldershot: Avebury, 15-44.

Hoernlé, R. F. A. (1926) Idealism, New York: Doran.

Huemer, Michael (2001) Skepticism and the Veil of Perception, Lanham, Maryland: Rowman and Littlefield.

Humphrey, Nicholas (1992), A History of the Mind, London: Chatto and Windus.

Julesz, Bela (1960) 'Binocular depth perception of computer-gernerated patterns', Journal of the Bell Telephone Company, 39, 1125.

Kant, E. (1787/1964) The Critique of Pure Reason, London: Macmillan.

Kelley, David (1987) The Evidence of the Senses, Baton Rouge, Louisiana: Louisiana University Press.

Kirk, Robert (1994) Raw Feeling: A Philosophical Account of the Essence of Consciousness, Oxford: Oxford University Press.

Lean, Martin (1953) Sense-Perception and Matter: A Critical Analysis of C. D. Broad's Theory of Perception, London: Routledge and Kegan Paul.

Lotze, Hermann (1884) Metaphysics, Oxford: Clarendon Press.

Luria, A. R. (1969) The Mind of A Mnemonist: A Little Book about a Vast Memory, London: Jonathan Cape.

McDowell, John (1994) Mind and World, Cambridge, Massachusetts: Harvard University Press.

Malcolm, Norman (1959) Dreaming, London: Routledge and Kegan Paul.

Matilal, Bimal Krishna (1986) Perception: An Essay on Classical Indian Theories of Knowledge. Oxford: Clarendon Press

Maund, J. B. (1993) 'Representation, Pictures and Resemblance', in Edmond Wright (ed.), New Representationalisms: Essays in the Philosophy of Perception, Aldershot: Avebury, pp. 45-69.

Maund, J. B. (1975), 'The representative theory of perception', Canadian Journal of Philosophy, 5, 41-55.

McGinn, Colin (1992), 'Getting the wiggle into the act', Review of Nicholas Humphrey: A History of the Mind, The London Review of Books, 10th September, 18.

Nagel, Thomas (1981), 'What is it like to be a bat?', in Hofstadter, Douglas R. and Dennett, Daniel C. (eds.), The Mind's Eye: Fantasies and Reflections on Self and Soul, Harmondsworth: Penguin Books, pp. 391-403.

Nakajima, Yoshitaka and Shimojo, Shinsuke (1981) 'Adaptation to the Reversal of Binocular Depth Cues: Effects of Wearing Left-Right Reversing Spectacles on Stereoscopic Depth-Perception', Perception, 10, 391-402.

O'Regan, J. K. and Noë, A. (2001) 'A sensorimotor account of vision and visual consciousness', Behavioral and Brain Sciences, 24:5 (forthcoming).

Pitcher, George (1971) A Theory of Perception, Princeton, New Jersey: Princeton University Press.

Price, H. H. (1961/1932) Perception, London: Methuen.

Putnam, Hilary (1982) Reason, Truth and History, Cambridge: Cambridge University Press.

Putnam, Hilary (1983) Realism and Reason. Cambridge: Cambridge University Press.

Putnam, Hilary (1988) Representation and Reality. Cambridge, Mass.: MIT Press.

Pylyshyn, Zenon W. (1981) 'The Imagery Debate: Analogue media versus tacit knowledge', Psychological Review, 88:1, 16-45.

Pylyshyn, Zenon W. (2002) 'Mental imagery: In search of a theory', Behavioural and Brain Sciences (forthcoming).

Rommetveit, Ragnar (1978) 'On negative rationalism in scholarly studies of verbal communication and dynamic residuals in the construction of human intersubjectivity', in The Social Contexts of Method, (eds.), Michael Brenner, P. Marsh and Marilyn Brenner. London: Croom helm, pp. 16-32.

Rorty, Richard (1980) Philosophy and the Mirror of Nature, Oxford: Basil Blackwell.

Ryle, Gilbert (1966/1949) The Concept of Mind, Harmondsworth: Penguin Books.

Searle, John (1983) Intentionality, Cambridge: Cambridge University Press.

Sellars, Roy Wood (1919), 'The epistemology of evolutionary naturalism', Mind, 27, 407-26.

Sellars, Roy Wood (1932) The Philosophy of Physical Realism, New York: Macmillan.

Smythies, John R. (1956), The Analysis of Sensations, London: Routledge and Kegan Paul.

Smythies, John R. (1993) 'The impact of contemporary neuroscience and introspection psychology on the philosophy of perception', in E. Wright (ed.) New Representationalisms: Essays in the Philosophy of Perception, Aldershot: Avebury, pp. 205-31.

Tye, Michael (1992) The Imagery Debate, Cambridge, Massachusetts: Bradford Books, MIT Press.

Warren, Richard M. and Warren, Roslyn P. (1968) Helmholtz on Perception: Its Physiology and Development, New York: John Wiley.

Whateley, Richard (1828) Elements of Rhetoric, London: John Murray and J. F. Parker.

Wilcox, Stephen and Katz, Stuart (1984) 'Can Indirect Realism be demonstrated in the psychological laboratory', Philosophy of the Social Sciences, 14, 149-57.

Woodruff Smith, David (1982) 'The realism in perception', Noûs, 16: 42-55.

Wright, Edmond (1983) 'Inspecting Images', Philosophy, 58, No. 223, 57-72.

Wright, Edmond (1990) 'New Representationalism', Journal for the Theory of Social Behaviour, 20/2, 65-92.

Wright, E. L. (1994) 'A new critical realism: An examination of Roy Wood Sellars' epistemology', Transactions of the C. S. Pierce Society 30(3):477-514.