Chapter 7: Appearance and Reality

By Mark Johnston
Princeton University

 

It is one thing to repudiate both Projectivism and generalized Response- Dependence as accounts of our relation to the manifest, quite another to understand in detail how experience could reveal or acquaint us with the manifest. Indeed there seems to be a compelling argument that experience cannot reveal the nature of the manifest but only present us with appearances understood as subjective effects of objects whose natures remain unrevealed to us.

 

What is the Good of Sensing?

What if anything makes sensing intrinsically valuable, as opposed to just a useful means for getting around our environment? An important part of what seems to make sensing intrinsically valuable is that as well as providing us with propositional knowledge about just which properties objects have, sensing also acquaints us with the nature of the properties had by those objects. It reveals or purports to reveal what those properties are like. As well as providing a map of our environment which helps us steer our way through the world, sensing also seems to do something that no ordinary map could do, acquaint us with or reveal the natures of the perceptible properties of the things mapped. Explaining as I go what I mean by acquaintance with or revelation of the nature of a property, I shall suggest that acquaintance with perceptible properties is something we non- derivatively want from sensing, and show that sensing seems to provide just such acquaintance. Unfortunately, upon examination, sensing's promise to acquaint us with properties of external things seems to be utterly fraudulent.

Just to fix ideas, let us take some examples. I am acquainted with a certain property of digitally reproduced sound (the sound of CDs), I know or am acquainted with this characteristic property of digitally reproduced sound thanks to having listened to enough of the stuff. Knowing or being acquainted with this property, sometimes called "digital brightness," enables me to recognize digital sound and come out with comparative statements about analog and digital sound, e.g. Don't throw away your LPs, analog is much more pleasing than digital. But what I came to know when I came to know or be acquainted with the brightness of digital sound is not exhausted by my recognitional capacity nor by my capacity to come out with various statements about digital sound. I could have both those capacities without knowing the brightness of digital sound. I could instead be recognizing and comparing on the basis of knowing another characteristic property of digital sound, a certain sucked-out quality in the reproduction of the sound of stringed instruments. By listening, I know what these sound qualities are like. There is a certain way, perhaps impossible to fully characterize linguistically, that hearing represents these sound qualities as being. I not only hear that the Deutsche Grammaphon CD of Herbert Von Karajan conducting the St. Matthew's Passion has these sound qualities. I hear what they are like and so am acquainted with them. Equivalently, the natures of these two sound qualities are revealed to me. Or so it seems.

For another example, take what is often said to be what one knows when one knows what it is like to feel pain. One of the things one knows or is acquainted with is a property or feature of a sensation -- what some call the painfulness of pain -- the feature that makes pain awful. I claim that sense perception in general and vision in particular purports to acquaint us with some properties of external objects in the way that sensation is alleged to acquaint us with features of sensations. The argument we shall consider seems to show that sense perception cannot carry through on this promise.

To locate the idea of acquaintance among more familiar epistemic concerns, consider two familiar philosophical cartoons by which the traditional skeptical problem of the external world is typically presented, the case of the eternal movie buff and the case of the brain in the vat. The eternal movie buff has spent all his life in a dark room watching images on a screen. Never having left his video room he has no idea whether the images correspond to anything outside his room. The brain in the vat is fed a full sensorium by fiendishly clever form of neural stimulation. A computer coordinates the pattern of stimulation so that the brain has a complete and consistent sensory illusion, say as of living an ordinary life in Boise, Idaho.

These bizarre predicaments are typically employed to highlight a skeptical worry about our own predicament. The eternal movie buff cannot be justified in holding any visually generated beliefs about the external world, restricted as he is to mere images which he cannot check against external reality. He can only check experience against experience. But this, so it is said, is also our predicament. We too can only check our experiences against other experiences. It is no more possible for us to attempt to match our experience against external reality as it is in itself, as it is independently of how it is experienced by us. The case of the brain in the vat is supposed to deflate the natural response that we have an epistemic advantage over the eternal movie buff by having a number of potential windows on the world which we can use to triangulate to an external reality as it is in itself. The triangulations of the brain in a vat lead it to false beliefs about a life lived in Boise, Idaho. Now there is no sign in our experience which differentiates our condition from the condition of the brain in the vat. So we possess nothing to rule out the alternative hypothesis that we are brains in a vat. So we are not justified in believing that we are not brains in vats and hence are not justified in believing any of the things we actually do believe about the external world. Or so the argument goes.

Whatever the force of these cartoons in presenting the traditional problem of the justifiability of our beliefs about the external world, and even if their force is undermined by noting that a spontaneous and utterly natural belief is justified or at least does not need justification in the absence of a good case against it, the cartoons also serve to illustrate a deeper epistemic anxiety about our own condition, one overlooked by fixating on the problem of justification.

This other problem of the external world is the problem of acquaintance, the problem of how, given the nature of information transmission, we could be acquainted with the nature of any of the properties of external things represented by our experience. The nature of any signal received is partly a product of the thing sending the signal and partly a product of the signal receiver. It seems that we cannot separate out the contribution to our experience of our own sensibility from the contribution to our experience of the objects sensed. The case of the brain in the vat shows that our experience does not discriminate between many different kinds of external features so long as their effects on our sensibility are isomorphic in certain ways. Therefore, despite the seductive offer that perception makes, we cannot take our perceptual experiences to reveal the natures of external things. For no perceptual experience could at the same time reveal the natures of two things so intrinsically unalike as life in Boise and the inner workings of the vat computer. Conclusion: perceptual experience does not reveal the nature of its causes. In other words, it does not acquaint us with the external features causally responsible for our experience but only with their effects in us. We can of course refer to the external features as the features that are standardly causally responsible for our experiences, thereby making them objects of thought and reference -- but at best we know these features by description -- we know them as the features that are standardly causally responsible for our experiences, whatever those features might be like in themselves. We could put the conclusion this way: relative to the problem of acquaintance, even if we are not brains in vats, things are as bad as they would be if we were brains in vats.1

Both the cartoon of the eternal movie buff and the cartoon of the brain in a vat highlight the problem of acquaintance by inviting us to think of perceptual experience as simply an effect of external causes whose natures are in no way revealed by the experiences they cause. Perceptual experience in no way acquaints the brain or the buff with the nature of the external causes of that experience. In this respect, perceptual experience is unsatisfyingly like morse code reception; both involve interpretable effects at the end of an information- bearing process or signal. But the intrinsic natures of the originators of the signal are not manifest in the signal.

This is a depressing comparison. Perception represents itself as (or is at least spontaneously taken by its possessors as) a mode of access to the perceptible natures of things; a mode of acquaintance with their perceptible properties. When I see the sun setting against the magenta expanse of the sky, I seem to have something about the nature of the sky and the sun revealed to me. I seem not to be merely under their causal influence in a way that leaves completely open what their natures might be like. Just as I take myself as knowing the painfulness of pain as it is in itself, and digital brightness as it is in itself, perception encourages the thought that I know sky blue as it is in itself. Russell expresses just this thought in The Problems of Philosophy: "the particular shade of colour that I am seeing ... may have many things to be said about it ... But such statements, though they make me know truths about the colour, do not make me know the colour itself better than I did before: so far as concerns knowledge of the colour itself, as opposed to knowledge of truths about it, I know the colour perfectly and completely when I see it and no further knowledge of it itself is even theoretically possible."2

The acquaintance with external features which vision seems to provide is something we have reason to value. I think that it is because vision seems to acquaint us with visible properties of external things and thereby acquaint us with the natures of visible things that we take a certain kind of epistemic pleasure in seeing. My pleasure in seeing color is not simply the pleasure of undergoing certain sensory experiences, it is also the pleasure of having access by sight to the natures of the colors and hence access to part of the nature of colored things. The deeper problem of the external world, as it applies to visibilia, is that this characteristic pleasure of seeing seems inevitably to be a pleasure founded in a false belief, a pleasure which philosophical reflection would have me see through.

Once my eyes were covered with bandages for five days. Part of what I longed for in longing to see again was not simply more information by which to negotiate my environment, nor simply more visual sensations. I longed for the cognitive contact with external features which vision seems to provide. It is depressing to conclude that what I longed for -- acquaintance with visible properties -- can never be had, even with the bandages off.

I suspect that such longing for cognitive contact is an important aspect of the longing to know. Consider the metaphors of knowledge: apart from the metaphors of light and sight, they are metaphors of touch and digestion; we grasp things, digest facts, absorb points, assimilate information. (Digestion is of course just touch gone to extremes.) The phenomenology of touch is the phenomenology of getting at the shape and texture of the thing touched and not at all the phenomenology as of having a tactile experience produced at the end of a (very short) causal chain which began with the thing touched. Touch represents itself as providing acquaintance with tactile qualities.

We place an important value on acquaintance, i.e. not just on knowing that external things have certain properties but also on knowing those properties. Indeed it is hard to make sense of having genuine acquaintance with things without being acquainted to some extent with their natures and so with some of their properties. An important part of what seems so bad about the predicament of the brain in the vat is that there is no acquaintance with sensible properties of external objects. Though the subject's experiences are internally coherent they are not revelations of the nature of the computer processes responsible for those experiences.

The consequence that threatens is that our senses don't tell us what bodies are like but only how they appear. Why is that bad?

I cannot give a full account here, but a partial account might begin with a range of emotions -- the aesthetic, sensual and erotic emotions, which are ways of taking pleasure in the revealed sensible natures of things. As such they presuppose that their objects are as they are represented by the senses. The upshot of our indictment of vision and by implication the other senses is that the aesthetic, sensual and erotic emotions should now be replaced by their relational surrogates: instead of taking pleasure in watching my greyhound run I should now take pleasure in being under the causal influence of something whose nature my senses fail to reveal to me but which produces in me certain pleasing appearances as of a greyhound in full stride. Here the pleasure one animal takes in watching another has been replaced by a pleasure taken in certain appearances. A fundamental role of the aesthetic, sensual and erotic emotions -- to draw our attention away from ourselves and our inner lives by pleasing us with the presence of others -- has been philosophically undermined. What is left are just ways of being pleased or displeased with our own reactions to things, whatever those things might be like. The philosophy of perception leads us from what Husserl called the natural attitude towards what we might call the pornographic attitude.

To measure the human cost of this try substituting the relational surrogates in your own life. Try explaining to the mother of a newborn that her sensory perception of the child, which drives her feelings for it, leaves the nature of the child completely unrevealed except for the fact that the child is something or other which causes appearances and feelings in her.

Or better, consider a fast approaching realm in which we are promised an endless variety of ways in which we are turned on by the effects of others upon us. In virtual reality, we are told, the homeliest pair will be able to present to each other as Gregory Peck and Ingrid Bergman. A fun way to spend a half hour, but a hideous way to spend a life. Why hideous? Because inevitably the emotional focus comes to be on whether and in what way the consistent Bergmanesque appearances please. And that is a generalization of the pornographic attitude -- an attitude that is limited and wrong not because it involves sexual pleasure but because it involves taking sexual pleasure in the sensory appearances rather than in the person appearing.

Yet the upshot of our reflections on acquaintance is that we are already inhabitants of a virtual reality. The very idea of the manifest collapses.

Of course, if we had learnt to care for persons not as intelligent animals whose living bodies manifest that intelligence, but as Kantian somethings we know not what, then this would not matter as much. But we haven't and it does.

That is why I believe that if we cannot solve the problem of acquaintance, then the thing to say is that even if we are not brains in vats, we are nonetheless as badly off as brains in vats are.

If that is right, knowledge or justified propositional belief about the external world may remain valuable instrumentally, as input for strategies for getting by, but the core of what made knowledge seem valuable in itself has been rotted out. Knowledge of the external world is left schematic and bloodless, if it includes no acquaintance with properties and hence no acquaintance with the natures of things and hence no real acquaintance with things.

 

The Argument Against Perceptual Acquaintance Revisited

Well, just why is it that perceptual experience cannot reveal the nature of perceptible features of external things?

To run through the specific version of the argument in the case of vision, begin with the obvious causal condition on seeing a feature or property of a thing. If a property of a thing is literally seen then the thing's having that property must be part of the causal explanation of the visual experience which counts as seeing that the thing has that property. If I see the greyness of my dog's coat, as opposed to having a (perhaps accidentally veridical) hallucination of the greyness of the coat, then it will be the case that the visual experience which represents my dog's coat as grey is partly caused by my dog's coat being grey. But if my visual experience is to be a revelation of the nature of the greyness of my dog's coat, then that visual experience must correctly represent how that greyness is. Now the way my visual experience represents greyness as being is fixed by the internal states that I am in at the time of the experience.3 By the causal condition on seeing, those internal states are to be causally explained by my dog's being grey. This means that there is a causal process connecting the state of my dog's coat and my internal states. Barring a pre-established harmony no such causal process will preserve and transmit information so as to secure a nature-revealing match between how some feature of the cause, say the greyness of my dog's coat, is and the way I am caused to represent that feature as being. To see involves having the natures of visible properties revealed by a causal process, but this is just a causal process could not do.

The originally unbelievable conclusion now follows: we cannot see color, because our visual experiences as of the colors of things do not reveal to us what the colors of the external causes of our experience are like.

But if we do not see color, we do not see color difference, and if we do not see color difference, we see neither edges nor colored areas, and if we see neither edges nor colored areas, we do not see surfaces, and if we do not see surfaces, we do not see anything in the material world.4 Our visual experience is then just a "false imaginary glare", simply an arbitrary medium in which the material world is mapped for the purposes of intentional action. The characteristic pleasure of seeing, the pleasure of having the nature of visible properties and visible things revealed to us, is a false pleasure. The promise of vision now appears totally fraudulent.

This indictment of vision, this apparent proof that it cannot be a mode of acquaintance with the visible natures of things depends on an obvious causal explanatory condition on seeing something. If something is literally seen, then it must figure in the explanation for the visual experience, the very having of which counts as seeing the thing in question. If I see the sky darkening, as opposed to having a (perhaps accidentally veridical) hallucination of the sky darkening, then it must be that the visual experience which represents the sky darkening is partly causally explained by the sky's darkening. But if my visual experience is to be a mode of revelation of the nature of the sky's darkening then there must be available aspects of that visual experience which match or show to me the nature of the property of darkening, along with other visible properties of the sky. This is precisely what the causal/explanatory condition on vision seems to preclude. For if we think of the visual experience as partially causally explained by the thing seen and its nature, we must also recognize that the intervening medium, the various features of the ambient and transmitted light, and, most crucially, the particular sensibility of the one seeing are also crucial explanatory factors. Visual experience does not provide us with a way of subtracting from the total quality of the visual experience those qualities due to the viewing conditions and the particular sensibility of the subject, so as to leave as a residuum the qualities which are due just to the things seen. If we think of visual experience as a mental effect of an external thing then we cannot subtract out from the features of the experience the features due solely to the explanatory factor which is the nature of the thing seen.
So how could the nature of the thing seen be revealed or manifest in an experience it only partly explains.

We may extend this argument by relying upon what we know of the empirical details of the processes in virtue of which the thing seen partly explains the seeing of it. Light must travel from the thing seen to the viewer. This causal process is in fact a necessary condition of seeing. Now even if it were possible to isolate the features of an experience exclusively due to thing experienced we still would have no reason to believe that the qualities of the residuum matched and hence revealed the natures of the features of the thing seen. For we have no way of making clear to ourselves how a causal process like reflected light could preserve and transmit information so as to secure a nature-revealing match between intrinsic features of the cause (the nature of the thing seen) and intrinsic features of the effect (the experience).5 To see involves having the visible natures of things revealed by a causal process, but this is just what no causal process can do.

In the same vein, and by way of refurbishing what he takes to be Kant's fundamental epistemological insight, P .F. Strawson writes

The idea that in seeing we only enjoy visual appearances of things is the idea that seeing is not acquaintance with the thing seen and with its visible properties. This is an intolerable conclusion, which no amount of Kantian epistemology really mitigates.

Neither Physicalist or Response-Dependent Accounts of Sensible Qualities Secure Acquaintance. (Omitted)

Well, you know that.

 

The Weakness in the Causal Argument

The causal argument against the possibility of acquaintance turns on an identification of appearings with "appearances" or monadic or internal states of the subject which occur at the end of a causal chain.6 The picture of experience as something at the end of a causal chain involving light, retinal reception and neural processing is simply insinuated into the causal argument, for want of a better alternative. The denial of acquaintance then falls out of the insinuated picture.

But there is a better alternative. Certainly for something to visually appear to me I have to be under its causal influence, it has to produce a change in my visual system and in my brain. That is a necessary condition of my being aware of it. The temptation is then to think of an appearing -- e.g. something's appearing to be some way to me -- as just causal influence plus my enjoying a internal or monadic mental state -- the experience -- whose nature is as much a product of my sensibility as it is of the way the thing seen is. Behind this temptation is a bad argument:7 Causation or more cautiously causal explanation is a necessary condition of experiencing or sensing or appearing, therefore there must then be someway of analyzing experiencing or appearing in terms of this necessary condition and other necessary conditions which together are jointly sufficient for experience. Given this step, another clear candidate for a necessary condition for experiencing is that one is in some mental state. Indeed the very idea of an experience as a mental state can be thought of as experiencing minus causation. But then we can't help but think that the character of the mental state is as much determined by the nature of the subject and the intervening medium as it is by the nature of the thing appearing. Appearings are fatally reduced to appearances plus appropriate causation, and appearances are no more than arbitrary subjective signs of the external things which cause them. The mentalistic conception of experience as a subjective occurrence is thus enshrined. Sensing does not make external things present. It does not acquaint us with the natures of things. It simply registers their effects.

The crucial mistake is the thought that the having of an interior mental event is the other necessary condition to be added to the causal-explanatory condition. How could this be a mistake? Well not every case where there is a necessary condition for the holding of a relation is a case in which the relation is analyzable in terms of that necessary condition plus some others. Consider numerical identity, being the very same thing. The holding of that relation has a clear indisputable necessary condition. A and B are identical only if they have all their properties in common. But it is clearly a confusion to suppose that we can analyze A and B being numerically identical as their having all their properties in common plus having something else in common. That way leads to medieval "haecieties" or thisnesses, which are supposed to lead from a case of mere perfect duplication or similarity to a case of numerical identity. Just as an appearance is what you get when you take away causation from an appearing, a thisness is what you get when you take away from a case of identity mere perfect similarity. But that is a sheer confusion; numerical identity is an unanalyzable relation with a necessary condition for its holding. So also, appearings are unanalyzable relations having as a necessary condition for their holding the causal influence of the object appearing on the subject who is appeared to. Appearances, like haeceities, are the products of a confused reductionist ambition.

What is the relation between an appearing, after all an event, and the sensible qualities of the object that is appearing? The event of some object appearing to a subject, equivalently the event of the subject sensing the object will have some internal complexity. Let an appearing be the coming to hold of a relation between an object, the thing which appears, a subject at a location, who is appeared to, and a property or way in which the thing appears. The idea that sensing or conversely appearing acquaints us with the manifest properties of manifest objects implies that the occupant of the object place in any appearing relation is an ordinary manifest substance, while the occupant of the property place is a manifest property.

So we can think of an appearing as an event which is the coming to hold of relations of the form:

These relations have as the necessary condition of their holding, the holding of a causal-explanatory relation among the constituting matter of the manifest objects, the physical properties which constitute the manifest qualities and the sensory system of the subject. The states of the subject's sensory system are to be causally explained, in the right way, by the matter of the manifest objects and by their physical properties.

This causal-explanatory condition is not simply a necessary condition, it is a constituting condition of the appearing. (But it is not a condition whose holding yields the appearing as an a priori or conceptual matter) The event which is the coming to hold of a relation like

Appearing to (O, Q, S)

is constituted by the event which is the coming to hold of a relation like:

Despite the pleasing parallelism this is not Parallelism in the traditional sense. The manifest substances and their manifest qualities genuinely explain the relevant sensory states of the subject who is seeing them. Manifest substances get to be explainers thanks to their matter, manifest qualities get to be relevant in explanation thanks to the explanatory role of their constituting physical properties.

The crucial pint is this: The causing does not come before the appearing, but rather constitutes it. The appearing is not to be broken down into an appearance on the end of a causal chain. The causal argument does not get a foothold.

In thus allowing for acquaintance with the manifest qualities of manifest substances, everything depends on finding a way to avoid "mentalizing" appearings, i.e., treating them as appearances at the end of causal chains. We must therefore face the major difficulty in the way of the theory of appearing or sensing, namely the argument from hallucination. Many think that here at least there is a decisive argument for a mentalistic conception of appearing or sensing, while others, like J. L. Austin and John McDowell, who dismiss the argument from hallucination seem to me to seriously underestimate its real force.

 

The Argument From Hallucination

It seems obvious that in some sense the very same appearance can be had whether or not there is a non-mental object appearing. Shouldn't we then endorse the Cartesian or mentalistic conception of appearing, according to which an appearing is a private mental event --- an appearance -- caused by an external object?

Here is a case that will serve as our stalking horse. You are undergoing an operation for an aneurism in your occipital lobe. The surgeon wants feedback during the operation as to the effects of the procedure on the functioning of your visual cortex He reduces all significant discomfort with local anaesthetic while he opens your skull. He then darkens the operating theater, takes off your blindfold, and applies electrical stimulation to a well-chosen point on your visual cortex. As a result, you hallucinate or seem to see dimly illuminated spotlights in a ceiling. (You hallucinate lights on in a ceiling. As yet, you are not at all aware of the ceiling of the operating theater.) As it happens, there are spotlights in the ceiling at precisely the places where you hallucinate lights. However, the real lights are turned off, so that the operating theater is too dark to really see anything.

While maintaining the level of electrical stimulation required to make you hallucinate lights on in a ceiling, the surgeon goes on to do something a little perverse. He turns on the spotlights in the ceiling, leaving them dim enough so that you notice no difference. You are now having what some call, after Hyppolite Taine, a "veridical hallucination". You are still having a hallucination because the lights on in the ceiling play no causal role in the generation of your experience. Yet the hallucination is veridical or true to the facts: there are indeed dim lights on in a ceiling in front of you. Moreover, you are not seeing the lights in the ceiling partly because you are hallucinating lights being just where the lights actually are. If you were to move your head a little you might see the dim lights on in the ceiling, in some sense "alongside" the dim lights that you are hallucinating.

In the third stage of the experiment the surgeon stops stimulating your brain. You see the dimly lit spotlights in the ceiling. From your vantage point there on the operating table these dim lights are indistinguishable from the dim lights you were hallucinating. The transition from the first stage of simple hallucination through the second stage of veridical hallucination to the third stage of veridical perception could be experientially seamless. Try as you might, you would not notice any difference, however closely you attend to your visual experience.

At the level of brain states, there will be some causal explanation for this experiential seamlessness. Whether one's brain is stimulated by a survey of the scene before one's eyes or by the direct application of electrical impulses, the effects on one's brain will be very similar in respects relevant to the causation of experience. This explanation in terms of brain states raises another explanatory question, which is our real concern. When we say that either way the effects on the brain are very similar in respects relevant to the causation of experience, we rely upon a picture according to which the differences at the level of brain states make no discernible difference at the level of experience. There will of course be some difference between the brain processes in the two cases. Our idea of some differences as not making a discernible difference at the level of experience begs for a charaterization of what is taking place at the level of experience, such that this or that difference in brain states is not a relevant difference for experience. Accordingly, our question is: How is visual experience constituted so that in a case of hallucination and a case of veridical perception there need be no difference which the subject can discern? In itself, appeal to ever so slightly different brain processes cannot answer that question.

To develop a feel for the question, consider two conflicting answers. The Sense Datum Theorist says that in indiscernible cases of hallucination and veridical perception one is directly aware of mental items, that is, items which would not exist but for the experiences in question. These mental items are indiscernible, or at least not discernible by the subject in the circumstances. The difference between hallucination and veridical perception is that in the second case the mental item is appropriately related to non-mental items --- the lights on in the ceiling -- which the mental item represents. In virtue of that relation holding in the veridical case one is aware of non-mental items by being aware of a mental item. The Sense Datum Theory thus endorses the Cartesian conception of experience. It treats veridical experience as the enjoying of an interior mental state that happens to be appropriately caused. Experience is something that could occur whether or not there are external objects. Even if there are external objects, experience presents a "veil" of mental items between those objects and subjects.8

An alternative answer goes under the name of "Intentionalism." The best version of this view has it that enjoying a visual experience is a sui generis propositional attitude, a relation between subjects and propositional contents concerning various possible scenes. In this way, experience "says" things about one's environment. In simple hallucination the content entertained (what experience "says") is false. In veridical perception the content entertained is true. Veridical perception differs from veridical hallucination in this respect: in veridical hallucination one's coming to entertain the true content is oddly or non- standardly caused. As a result of visually entertaining the same or similar contents the transition from hallucination to perception may go unnoticed.9

Visually entertaining a content to the effect that P is to be distinguished from believing that P and even from having the suppressed disposition to believe that P. When I have a visual experience of a stick half submerged in water the stick appears separated, but I do not believe that the stick is separated. The suggestion of David Armstrong and George Pitcher is that such cases show that experiencing is the acquisition of a possibly suppressed disposition to believe things about the scene before the eyes.10 However, the acquisition of a disposition to believe is only a regular effect of the event of experiencing. It is conceptually possible that the cause occur without the effect, so that the acquisition of a disposition to believe is not a necessary condition of experience. Not only need I not acquire such a disposition, but even when I do acquire it, its presence seems curiously irrelevant. Suppose for example that I am staring at a blank wall, I acquire the possibly supressed but actually unsupressed disposition to believe that the wall is blank. I go on staring at the wall long after I have acquired the disposition. The event of my acquiring the disposition is over, yet I am still experiencing a blank wall. If no more acquiring of dispositions is needed why was it needed in the first place? Nor is the acquisition of a disposition sufficient for experience. One could acquire a suppressed disposition to believe, e.g. that the scene before the eyes is poorly lit, in ways that bypassed experience -- sudden depression, hypnotism, etc.. One would not thereby be experiencing the scene before the eyes as poorly lit. We have no accurate statement of what the right way of acquiring the suppressed disposition to believe is unless we require that it is by way of visual experience.

So seeing or experiencing is not to be reduced to a version of believing. Nonetheless experiencing might be the acquisition of a sui generis propositional attitude, which we have dubbed "visually entertaining a content." If we can make sense of such a propositional attitude then postulating sense data will be adding an idle wheel to the account of how things appear.

To feel the force of this last remark, consider that there is always room for incomplete belief and therefore for erroneous belief about any item of which one is aware. The same goes for any information-registering propositional attitude such as "visually entertaining a content." The forming of belief or any other information-registering propositional attitude about an object of experience is a contingent event. Any such event can go happily or less than happily. So if in seeing the lights on in the ceiling I am aware of a mental array with many dot-like features -- a sense datum as traditionally conceived -- I may immediately entertain an erroneous content concerning how many dimly lit lights there are on in the ceiling. I can get any mental array wrong in just the way I can get any physical array wrong. Suppose there are 20 lights on in the ceiling and thanks to incomplete attention to my 20 dot sense datum I immediately entertain the content that there are 21 lights on in the ceiling. Then surely it will be right to say that it seems to me that there are 21 lights on in the ceiling. The sense datum does not decide how things seem or appear to me, since the content I entertain could be wrong about the details of any sense datum that I am enjoying. It is the contents I visually entertain which determine how things seem to me. Sense data are thus idle wheels if Intentionalism can be made out.
Intentionalism makes up the closest thing to orthodoxy in discussions of perception these days.
11 Unfortunately, I cannot join the orthodox. I do not believe that the relevant notion of the content of experience can be satisfactorily explicated. I shall argue that when it comes to perceptual experience properly so called, representational content and the kind of intentionality that derives from representational content are irrelevant. According to me, perceptual experience is always of a this-worldly, non-mental, non-propositional, non-intentional object. So perceptual experience, as opposed to perceptual belief, is never false, though it is sometimes misleading.

My position depends upon locating the this-worldly, non-mental, non- intentional objects that are the objects of hallucinations, afterimages and the like. These objects will emerge as we examine the other proposals.

 

Harman's Theory of Hallucination

Although I treat the notion of visually entertaining a content to be at the heart of Intentionalism, not all theorists of perception who appeal to intentional items give central play to this notion. For example, Gilbert Harman, following on the work of Elizabeth Anscombe, holds that visual experience relates the subject to intentional objects -- what the experience is about. Let's explore why such a proposal is naturally taken as presupposing the content view.

Harman's answer to our question about experiential seamlessness would proceed as follows. In the case of seeing the lights on in the ceiling the intentional object of your experience is the lights on in the ceiling. In the case of hallucinating lights on in the ceiling the intentional object of your experience is lights on in a ceiling. The intentional objects are "similar" only in the sense that we can confuse one for the other and learn about features of one by attending to the other.

Obviously we need to know what these intentional "objects" are and how they could be similar in the required ways when one of them, the lights in the ceiling, is a collection of specific existing things while the other, lights on in a ceiling, is not.

Anscombe has a straightforward characterization of intentional objects: they are the direct objects of psychological verbs in correct reports of the thought and behavior of subjects.12 It is important to note that Anscombe does not intend this characterization to make intentional objects linguistic items of a particular grammatical kind, e.g., accusatives of a transitive verb. Her explanation of why this is so deserves repeating. Consider whom Jack is said to have hit according to (1).

(1) Jack hit Jill

The object of the verb "hit" in the sense introduced by the question "Whom according to (1) did Jack hit?" is not the word "Jill" but simply Jill. If (1) is false so that (1) is a minimal case of a fiction then the direct object is still not a word but the rather jejune fictional character Jill. Now consider

(2) Ponce de Leon was searching for The Fountain of Youth.
(3) Ponce de Leon was searching for a fountain.
(4) Ponce de Leon was searching for a magical cure for aging
13

Each of (2), (3) and (4) is a correct report of Ponce de Leon's search. Respectively, they relate him to these intentional objects -- The Fountain of Youth, a fountain, a magical cure for aging. For these are the things that (2), (3) and (4) say that Ponce de Leon was searching for. This shows that the phrase "the intentional object of my experience" is strictly speaking ill-formed. Speaking strictly, one should refer to the intentional object of my experience as reported in this or that sentence. Sentences have direct objects and these get associated with experiences when those sentences are correct reports of the experiences in question. Intentional objects are sentence-relative.

Harman's contribution to the argument from hallucination is to focus on the intentional objects of hallucination and seeing. According to Harman, people have been mislead into the Sense Datum Theory because they have supposed that since Macbeth saw or hallucinated a dagger and there was no dagger there, what Macbeth really saw or hallucinated could only have been something mental, the Sense Data Theorist's dagger-like mental item. This would be the same as thinking that because Ponce de Leon was searching for The Fountain of Youth, and there is no Fountain of Youth, what Ponce de Leon was searching for was something mental. That would be a sheer confusion based upon the failure to appreciate how intentional objects work. If (2) is true the intentional object of Ponce de Leon's search was the Fountain of Youth. Clearly (2) can be true even if there is no Fountain of Youth. So also

(5) I was hallucinating lights on in a ceiling

can be true when there are no lights on in any ceiling around me. The intentional object of my hallucination is lights on a ceiling. It is fallacious to infer from this and the fact that there were no lights on that the hallucinated lights on in a ceiling are mental.

How exactly does this talk of sentence-relative intentional objects help explain the possibility of an experientially seamless transition from a case of hallucination to a case of veridical perception? When I see lights on in the ceiling certain sentences are made true by that event, e.g.,

(6) I saw the lights on in the ceiling.

What (6) says I saw is: the lights on in the ceiling. What (5) says I hallucinated is: lights on in a ceiling. These are not the same intentional objects. Harman's example of The Fountain of Youth is misleading in this respect. It is not typical for seentences reporting hallucinations and the corresponding veridical perceptions to share intentional objects. It is not the identity of intentional objects but something about the similar character of the experiences that is at issue. At the linguistic level (5) and (6) exhibit a similarity between a definite description and a corresponding indefinite description. This mildly diverting linguistic similarity has no power to explain how the transition from the experience that (5) describes to the experience that (6) describes could be seamless. The linguistic similarity is clearly a consequence of a non-linguistic similarity between the experiences. What makes for that similarity? Similar sense data? No, at least not for Harman. Similar content visually entertained? Harman would be happier with that, especially if the content were taken to be determined by the similar causal roles of the two experiences. What then is it to visually entertain a content?

 

Visually Entertaining A Content

Our question was: How must we think of visual experience to explain the fact that a transition from hallucinating lights on in a ceiling to seeing the lights on in the ceiling can be undetectable by the subject undergoing this transition? The Intentionalist explains the seamless transition between hallucinating and seeing as follows: Both seeing the lights and hallucinating lights involve similar forms of awareness in that they involve visually entertaining very similar contents. Seeing the lights on in the ceiling differs from hallucinating lights on in the ceiling in that the content entertained while seeing the lights is about items in the scene before the eyes -- the lights -- which are causally responsible in the standard way for the visual experience which is entertaining the content that there are lights on in the ceiling.14

This explanation of seamless transition works only if the Intentionalist can provide for an analysis of seeing in terms of visually entertaining a content under regular causal conditions, i.e., something like:

However, nothing like (7) can be an analysis of seeing. (7) is merely a schema that attaches some meaning to the otherwise unexplicated philosophical jargon "visually entertaining a content" by using the antecedently understood notion of seeing. That this is so is the upshot of considerations of what things can be seen together or taken in by the same visual experience. The constraints on what can be seen together are immediately intelligible as constraints on seeing, but they are not immediately intelligible as constraints on what contents can be entertained. So we cannot make (7) do one of the things distinctive of an anaysis, namely account for why caertain constraints hold on the analysandum.

We cannot, by way of the very same visual experience from the very same viewing position, see b as opaque and see c as wholly behind b. However, we can think, suppose, even truly believe that b is opaque and that c is wholly behind b. So I cannot visually entertain the content that

there is a black mark on the other side of that opaque wall.

though I could perfectly well think or entertain or even believe that content.16 We must therefore read back the principles governing what can be seen together as brute constraints on visually entertaining a content. There are an enormous variety of ways the scene from here could consistently be that I can perfectly well entertain but cannot visually entertain. Think of a small frozen piece of chardonnay floating in the middle of a glass of chardonnay under dim lighting conditions. In dim light seeing the chardonnay in the glass excludes seeing the chardonnay ice floating in the middle of it. In good light, you might see the ice. A host of such examples could be produced to illustrate the principles of opacity, transparency and radiance. I cannot see a green plant through a pink color volume without seeing the greenness of the plant shaded toward the grey range. I cannot see something as a flame and as matt black.

Now the things that I see are supposed to be what the contents that I visually entertain are about. So we must read back all the principles of opacity, transparency and radiance as brute constraints on the contents that can be visually entertained. These constraints are many and various and go far beyond any simple constraint to the effect that one cannot visually entertain inconsistent contents. They will inevitably seem ad hoc constraints on visually entertaining contents unless we are told what is distinctive about visually entertaining a content so that the visual entertaining of such contents is impossible. The Sense Datum Theorist has an answer in terms of how the mental array that is the primary object of awareness is filled in at various "places." Obviously, the Intentionalist who wishes to supersede the Sense Datum Theory cannot have recourse to that explanation.

One developed Intentionalist theory may seem to offer an explanation. John Searle's account of visual experience has the contents "visually entertained" be contents involving the concept of vision. According to Searle, the contents of visual experience are complex self-referential contents. So when I see an opaque barrier before me the content of my visual experience is supposed by Searle to be:

Now while there is nothing impossible about this content:

That there is before me an opaque barrier with a black mark on its back.

There is arguably an inconsistency in the following content:

By building into the content of one's visual experience of some scene the fact that one is having a visual experience of that scene, Searle's Intentionalist account of perception does not leave ad hoc or unexplained the constraints on visually entertaining a content . The explanation is that you cannot visually entertain inconsistent contents.18

Now seeing in particular and perception in general are basic abilities that I share with babes and animals who are not able to conceptualize or represent themselves as seeing things or having visual experiences. Further to the point, I often see things, say in the periphery of my visual field, which I do not notice. I do not identify them as any kind of thing. I am not aware that I have seen them. My only access to the fact that I have seen them might be an inference from the fact that they were there to be seen in an area scanned by my gaze. Not only can I see without noticing what I see, but I can also see without noticing that I am seeing. Like Armstrong's dazed truckdriver I can realize that I must have visually taken in the last stretch of road though mentally I was completely elsewhere.19 Human animals often negotiate their environment by such dumb inattentive gazing. Not only is seeing not necessarily believing, it need not even involve appreciating what is seen20.

It thus seems forced to say that seeing always involves at some level the awareness that you are seeing or the awareness of your visual experience. Yet if entertaining the content

That I am having a visual experience of an opaque barrier and ...

is a way of being aware of the opaque barrier then it should also be a way of being aware of myself and of my visual experience. The opaque barrier is an object of awareness because it is what the intentional content is about. Yet the self- referential content is as much "about" me and my visual experience as it is about the barrier. So one might be tempted to conclude that Searle's Intentionalist account of visual experience demands too much in the way of concomitant awareness of oneself and one's visual experiences. Armstrong, Burge and McDowell do conclude this, only to be told by Searle that this is simply a gross misunderstanding.21 Let me canvass one way in which it might indeed turn out to be a misunderstanding of Searle's position to suppose that it entails concomitant awareness.

 

Non-conceptual Content

There is a position that is gaining currency among Intentionalists that goes under the heading of "non-conceptual content".22 Seeing in particular and sensing in general are basic abilities that I share with babes and animals who are not able to conceptualize or represent themselves as seeing things or having visual experiences. Further to the point, I often see things, say in the periphery of my visual field, which I do not notice. I do not identify them as any kind of thing. I am not aware that I have seen them. My only access to the fact that I have seen them might be an inference from the fact that they were there to be seen in an area scanned by my gaze. Not only can I see without noticing what I see, but I can also see without noticing that I am seeing. Like Armstrong's dazed truckdriver, I can realize that I must have visually taken in the last stretch of road though mentally I was completely elsewhere.23 Human animals often negotiate their environment by such dumb inattentive gazing. Not only is seeing not necessarily believing, it need not even involve appreciating what is seen24.

Consider seeing an opaque surface. I can see an opaque surface while not taking it to be opaque. I may wrongly see it as transparent, as when I am taken in by a trompe l'oeil of a window onto a garden scene. I can also see an opaque surface whether or not I have the concept of opacity. In general, to see an F I need not see it as an F, and so I need not conceptualize, classify or think of it as an F. This is the sense in which visual experience is conceptually undemanding. Yet if we as theorists are to state the conditions under which my visual experience of an F is correct, we must in stating those conditions deploy the concept of an F. We must also deploy the concept of a visual experience, the concept of the subject of the visual experience and the concept of causation. So as theorists we might say such things as:

On the one hand, we are thinking of the visual experience as a state with correctness conditions given by the right-hand side of this biconditional. On the other hand, we do not suppose that John must have or deploy the notions on the right-hand side of the biconditional. John's visual experience has correctness conditions. In that sense it is an intentional state. Yet the content of this intentional state is "non-conceptual," meaning that John's undergoing the visual experience need not involve him in deploying the concepts we need to deploy in describing the correctness conditions of his experience. So babes and brutes and dazed truckdrivers could be having visual experiences with the same content as John's visual experience though they lack or do not deploy the concept of themselves, of visual experiences, or of opacity. Visual experience is thus conceptually undemanding.

So far so good. Yet I think there is something amiss with the notion that visual experience is the entertaining of non-conceptual content, something that threatens any Intentionalism which properly assimilates the fact that visual experience is conceptually undemanding. What makes it the case that a given experience has a particular non-conceptual content rather than another? It is not that the subject is deploying the concepts that make up the content as his mode of thinking about or being aware of the objects of his experience. This would be to falsely model seeing on thinking, the content of which does depend on how the thinker conceives of his subject matter, and so fail to allow for the fact that experiencing is conceptually undemanding. Nor does an experience have a content because it involves a representation in the sense of mental array, sense datum or linguistic description that is more naturally associated with one content or condition of correctness rather than another. This would be to falsely model seeing on the production of written words which determine a sense or content for what is written. The only remaining model of the content of visual experience appears to be this: Visual experience has the representational content it does because of its causal role, i.e., its causal relation to states of the environment and to other mental states of mine, e.g., my beliefs, desires and intentions.

To say how it is that a certain intentional content or correctness condition is properly associated with a given experience even though the subject of the experience need not be aware of the correctness condition, Intentionalism does best to rely upon some version of a causal account of representational content. Now visual experiences have a definite character anyway, i.e., before and whether or not they cause this or that belief, desire or intention. Immediate perceptual beliefs get their content in part because of their connection to visual experience. It is thus natural to look to the states of the environment that standardly cause a given kind of visual experience as the crucial determinant of the content of those visual experiences. (Unactualized causal roles will be considered in due course.)

If we take visual experiences to have content then some visual experiences will be false, i.e., have a content which misrepresents the scene before the eyes. Furthermore, some of these false contents will involve the attribution of properties which nothing has or ever had. Thus I might see the fruit on the table as a very still life, as not changing in any way over a short period. Or I might hallucinate a talking moose. Although things are always changing and no moose talk, the causal theorist may have little trouble elaborating the standard cause story to handle the contents of such experiences. The contents involve the attribution of compound properties, not changing, being a moose and a talker. The properties are logical compounds of properties that are instantiated and can be standardly causally connected with experiences of their instances. The urgent question for any such causal theory of content now becomes: Could a visual experience ever have a character such that the only reasonable content assignable to the visual experience was a content attributing to objects properties that are uninstantiated and uncompounded? If so, then visual experiences do not have representational content by being appropriately caused by states of the environment.

 

Missing Shades

Visual experience could have a character such that the only reasonable content to assign was a content that attributed a property that was uninstantiated and uncompounded. One could hallucinate the missing shade of blue.

There are several stories about the missing shade of blue, so let me make up one of my own.25 Suppose we are mapping the colors in the three dimensional space demarcated by the dimensions of hue, saturation and brightness. The shades of blue that we recognize, catalogue and contrast are bundled together in their own sub-volume of a larger three-dimensional space. We then discover that this sub- volume has a "hole" in it. That is, as we move through the blues along the dimension of brightness the distinctions among the blues in respect of brightness seem suddenly to be cruder or less fine-grained. There are fewer blue samples which are just noticably different. We also find many more graduations in the shades of yellow, green and red along the same dimension of brightness. Try as we might to mix pigments in the familiar ways we fail to generate the intermediate missing shade or shades of blue. We simply cannot find or produce a sample of a blue to match certain yellows, reds and greens of a given degree of brightness and of saturation.

Nonetheless, someone could hallucinate a patch of a missing shade of blue. Fascinated by this experience, he could set upon an even more ambitious project of pigment mixing to realize an instance of the very shade he hallucinated. He might succeed. That he succeeded in realizing the shade he hallucinated might be intersubjectively verifiable. The drug that caused him to hallucinate a missing shade of blue could have this as a regular consequence. Others could take the drug before and after the discovery of the new mix of pigments and thereby verify that the newly realized shade was the shade that the first hallucinator "saw." That is to say that the first hallucination had some definite character that in all likelihood guided the attempt to realize the shade in question.

Let us call the hallucinated shade "Hume blue." It might have been the case that being Hume blue was a property that nothing ever had. The hallucinator might have tired of his project before realizing a sample of Hume blue. Even if that had happened his hallucination would have had a definite character. For its having a definite character could not depend on what subsequently happened.

The Intentionalist's account of the story goes like this. The drug's effect is to get a victim to entertain the content that there is a Hume blue patch before him. That content is a definite content and so can be used to guide and test the success of the project of trying to realize the hallucinated shade.

How is it that the content that there is a Hume blue patch there comes to be associated with the visual experience that is the hallucination of a Hume blue patch? Here causal accounts of content will face an impasse.

Could being Hume blue be a complex property like the property of being a shade with such and such hue, so and so brightness and such and so saturation, a property whose compounds are all instantiated? Let us examine this proposal, which has the form

The proposal has to be taken to be a general account for any shade property, since it is hard to see why one shade should be reducible to a triple of hue, saturation and brightness when others are not reducible in this way. In any case, the proposal would have to be general because the issue about the missing shade of blue is quite general in form. It is adventitious which shades are realized. We might have first inhabited a monochrome world, then hallucinated the chromatic shades, then set about trying to synthesize instances of the chromatic shades, without success.

There is a quite general consideration which suggests that the complexity which the proposal under discussion purports to find in the shades is not a complexity in the shades themselves but a complexity in the similarity and differnece principles which hold among the shades. That is, hue saturation and brightness are dimensions of similarity which hold among the shades because of the ways the individual shades are. The relations of similarity along these dimensions are not definitive of what it is to be this or that shade, as can be shown by certain "rotations" of the shades which nonetheless preserve the relations.

Think of the color solid as made of two cones whose bases touch. The darkest blacks are at one apex, the brightest whites at the other. The fully saturated hues -- red, orange, yellow, green, blue and purple -- are placed around the widest circumference of the cones. As we move toward the center of the solid the hues become less saturated. As we move towards white the hues become brighter, and towards black, darker. The color solid is thus a three-dimensional array in which the shades are placed by relations of similarity in respect of hue, saturation and brightness. If we fix a position in the solid as the position occupied by sky blue then the relations of comparative similarity along the dimensions of hue, saturation and brightness will determine the positions of all the other shades, including any unrealized shades such as Hume blue. In principle, and to any desired degree of specificity, Hume blue could then be picked out in terms of a triple of coordinates which describe its position in the solid. However this specification of Hume blue depends upon a prior anchoring of the hues in a specific position in the three-space that is the color solid.

After all, there could be a missing shade of yellow -- "Hume yellow"--which is just as bright and as saturated as Hume blue. Rotating the hues 180 degrees within the color solid puts Hume yellow at the position that Hume blue was at and vice versa. The rotation preserves all the relations of hue, saturation and brightness among the shades. Short of an anchor to fix one of these positions rather than another, the relations of comparative similarity in hue, saturation and brightness do not distinguish Hume blue from Hume yellow. So there must be at least one anchoring shade M such that the right account of being that shade does not take the form

If something of this form is the right account of any shade then it ought to be the right account of every shade. So I conclude that

is not the right account of being Hume blue. Being this or that shade is not a complex property built up out of hue, saturation and brightness properties. As the rotation argument shows, the shades are qualities whose natures are not exhausted by their relations of similarity and difference to each other26.

If Hume blue is not complex in the way just canvassed it is hard to see what other complex of properties could be a better candidate to make up Hume blue. G.E. Moore suggested the shades are simple properties. If we could prove that then we would have shown that the hallucination of an uninstantiated shade is not visually entertaining the content that something before one has that shade, at least if experiences get to have contents by being appropriately caused by environmental conditions. If that model of bearing a content fails along with the models appropriate to thinking and to writing then we have good reason to look elsewhere for an account of visual experience.

We cannot prove that shades are simple or uncompounded in the relevant sense. The best that can be done is to refute attempts to articulate complexity in the shades.

 

Are Missing Shades Uninstantiated Reflectance Patterns?

Someone familiar with color science might suppose that the best way to handle missing shades is to treat them as light-dispositional properties, complex properties that are themselves uninstantiated but are compounded out of instantiated properties.27 So Hume blue could be a certain reflectance pattern given by percentages of light of various wavelengths reflected back by a surface that would standardly appear Hume blue. The property of having a reflectance pattern is complex and arguably such that the properties that comprise an uninstantiated reflectance pattern are themselves instantiated.

Since no shade has its own unique reflectance pattern, a refinement of this idea is needed. A bundle of distinct reflectance properties is associated with each shade. This is the phenomenon known as metamerism. At most Hume blue could be the disjunction of the reflectance patterns that surfaces would have to instantiate in order to be disposed to look Hume blue. The hallucination of a Hume blue patch would have the content that some patch has this disjunctive reflectance pattern. The hallucination could guide the project of synthesizing a surface which looks Hume blue. The project would succeed when a surface with one of the reflectance properties in the disjunction was synthesized.

Despite its promising beginnings, this account does not survive scrutiny. Color science is not only very informative, it is very surprising. One of its greatest surprises is metamerism; the discovery that at the microphysical level things of the same shade may only have a disjunction of their different surface properties in common. Why is this a surprise? It is a surprise precisely because visual experience presents things of the same shade in a way that immediately encourages the belief that they exhibit some natural ungerrymandered similarity. Now if shades are reflectance patterns and the content of a visual experience of a sky blue wall is that the wall has one of a large disjunction of reflectance patterns then visual experience does not present things that are sky blue as having any natural, ungerrymandered similarity.

If D is the disjunction of reflectance patterns that surfaces have when they standardly appear sky blue then the present proposal takes the content of the visual experience of a sky blue wall to consist of the wall and the property D. So let's denote that content by "<the wall, D>." In taking this content to be "non- conceptual" the Intentionalist is implying that it is not a condition of entertaining this content that one has any way of conceptualizing D. Certainly, no one's visual experience presents D for what it is, a disjunction of reflectances. Consider now the immediate perceptual belief that arises upon the experience of a sky blue wall. What is its content? Obviously, that the wall is sky blue. Now either the belief that the wall is sky blue and the experience of the wall as sky blue attribute different properties of sky blueness to the wall, or they attribute the same property to the wall. If we take the first route we are implausibly characterizing immediate perceptual belief as a continuous series of gross non sequiturs relative to perceptual experience. If we take the second route then another problem immediately arises. Non-conceptual content was invoked to deal with the conceptually undemanding nature of experience in contrast to concept-deploying states like belief. Entertaining the immediate perceptual belief that the wall is sky blue will involve some way of conceptualizing D. This mode of conceptualizing D, whatever it is, cannot plausibly be taken as indicating the disjunctiveness and complexity of D. After all, the disjunctiveness and complexity of the surface causes of experience of the colors of things are surprises that color science has discovered. To avoid a gross non sequitur with respect to the visual experience with the content <the wall, D> the immediately prompted perceptual belief must attribute D to the wall. To leave room for a main way in which color science is surprising the immediate perceptual belief cannot involve a conceptualization of D as disjunctive and complex.

Depending on our favored theory of belief we could either invoke a perceptual analog of a Fregean mode of presentation of D or appeal to "perceptual guises" under which the content <the wall, D> is believed.28 Either way we could claim that the immediate perceptual belief is true if and only if the wall is D, while allowing that the mode of presentation or guise associated with the belief is precisely one which does not register the disjunctiveness and complexity of D. On the Fregean line, the content of the immediate perceptual belief entertained upon seeing a sky blue wall could be represented by the ordered pair <the wall, PMP(D)>, where "PMP(D)" is to denote perceptual belief's mode of presentation of D. On the guise theory, the immediate perceptual belief that the wall is sky blue will involve entertaining the content <the wall, D> under a guise G that does not reveal the disjunctiveness or complexity of D. For our restricted purposes here these two theories really function as notational variants. The points to follow are formulated against the Fregean version, but the reader will see an easy transposition that works against the guise theory.

How is it that a mode of presentation like PMP(D) which gives no hint of the disjunctiveness and complexity of D could nonetheless be a mode of presentation of a property defined as a disjunction of complex properties? It is not that this is in principle impossible. There are models of how this could be so. It is just that it is incredible that these models apply to seeing or hallucinating colored surfaces. PMP(D) might pick out D by being a property characterization like: "the actual surface property, whatever it is like, which is standardly causally responsible for an appearance of sky blue."29 Alternatively PMP(D) might pick out D by being a property characterization like: "The actual surface property, whatever it is like, which is standardly causally responsible for an appearance like this one" Such property characterizations are higher order in that they pick out the property D in terms of its (relational) properties. Such property characterizations are also noncommittal as to the nature of the property to which they refer. Yet not even the closest inspection of immediate perceptual belief turns up such higher order, noncommittal modes of presentation of the shades. Any such higher order or noncommittal property characterization will be too roundabout and hands off to be plausible as the perceptual mode of presentation of a shade that one is concurrently seeing. After all just by seeing shades of sky blue, azure and canary yellow we
can come to know that the property sky blue is more similar to azure in respect of its higher order and intrinsic properties than either is to canary yellow. But the higher-order and non-commital modes of presentation ignore these immediate commitments which visual experience prompts.
30

Furthermore such "modes of presentation" or property characterizations simply state the general form of the physicalist theory of color -- the colors are the actual surface properties standardly responsible for the appearances of the colors. Writing the physicalist theory of color into immediate perceptual belief is not only implausible. If we are to aiming to allow for the surprise which color science prompts, it is self-defeating. In the earliest days of color science the discoveries concerning the physical causes of color led Galileo and Descartes to regard the world as devoid of color. So far from being happy with a gerrymandered physical basis for color, Newton himself supposed that to be this or that shade was to give off light of a characterisitic wavelength. Goethe in The Theory of Color decried the fact that the physical causes of color did not match the phenomenology of visual experience and so ambitiously called for a new science of color on this basis. Were these theorists simply ignorant of the noncommittal nature of the contents of their immediate beliefs about the colors of things? Are we to say that the property characterization PMP(D) has in its turn a higher order mode of presentation that does not reveal the noncommittal nature of PMP(D)? Now we teeter on the edge of absurdity.

The noncommittal and roundabout characterization of a property as the property, whatever it turns out to be, that is standardly causally responsible for the appearance of sky blue is the characterization of sky blue that the physicalist adopts. It is not immediate perceptual belief's characterization of sky blue. If immediate perceptual belief employs a mode of presentation or characterization of sky blue it is more committed than that. It would purport to tell us something about the nature of sky blue rather than simply pick it out as the occupant of some causal role. (Something for example that prompts the thought that sky blue are akin in a way that neither is to canary yellow.) Nor will immediate perceptual belief's characterization of sky blue pick out sky blue as D. Call the more committed and nondisjunctive characterization of sky blue C. Since C tells us something about the nature of sky blue which is not exhausted by D, the wall's being D will not imply that it is C. That is to say that if we treat shades as disjunctions of reflectance properties and represent the non-conceptual content of visual experiences as disjunctions of reflectance properties then we will inevitably represent the move from visual experience to immediate perceptual belief as a gross non sequitur; in our notation, as the step from <the wall, D> to <the wall, C>.31

So far we are at a loss about what the non-conceptual content of visual experience could be and how a causal theorist might find the elements that make up that content in the environment of the subject.

 

Are Missing Shades Dispositions to Produce Content Bearing States?

Some will think that the best way for the causal theorist to associate a content with a hallucination of an uninstantiated shade is to treat the shades as dispositions to appear this or that way. So Hume blue is the disposition to appear Hume blue. It may seem that this property can be treated as a complex property built up out of the relational property of being disposed to produce some appearance in us and the property of being an appearance of Hume blue. The relational property is instantiated whenever we have a disposition to appear this or that way. The property of being an appearance of Hume blue is instantiated by the hallucination of a Hume blue patch. When we put the two properties together have we not solved the problem of associating a content with the hallucination of a Hume blue patch?

No we have not. When the dispositional account of the shades is employed in this way, it implies that when one hallucinates a Hume blue patch the content or truth condition of one's hallucination is that there is a patch that is disposed to appear Hume blue. Notice that the present account of appearing Hume blue treats the phrase "appearing Hume blue" as semantically complex, i.e., as referring to appearing a certain way, namely Hume blue. The only account of Hume blue offered is that Hume blue is the disposition to appear Hume blue, so that the content or truth condition associated with the hallucination is that there is a patch which is disposed to appear to be disposed to appear to be disposed to appear...and so on. The point is not that no one sees or hallucinates anything as having such an ungrounded or regressive feature. No one does, but since "seeing x as F" may involve a referentially opaque predicate position the upshot is unclear. The point is rather that because the content or truth condition associated with the hallucination is ungrounded or regressive in this way, the account under discussion has failed to associate a definite content with the hallucination. This is borne out by the fact that on the account under discussion the content or truth condition of a hallucination of Hume yellow or of cherry red is likewise that there is a patch which is disposed to appear to be disposed to appear to be...and so on.32

The immediate response is that there must be some way to differentially ground these contents. After all, the appearance of the Hume blue patch is ex hyposesi clearly different from the appearance of the Hume yellow patch. How is the dispositionalist to account for this difference without falling back into some version of the Sense Datum Theory? One way of falling back into some version of the Sense Datum Theory is to follow Christopher Peacocke and treat Hume blue as the disposition to produce a Hume blue' patch in some mental array that some subject could enjoy.33 Here "Hume blue'" is to name a property of mental arrays and "Hume blue" is explained in terms of a disposition to produce instances of that property. Clearly this theory is not available for the Intentionalist whose characteristic aim is to eschew visual fields or mental arrays figuring in perception.

There is another way of grounding an account of shades as dispositions to appear. However, as will emerge, this involves a treatment of appearances which renders otiose talk of the content of experience.

 

Unactualized Causal Roles

Jerry Fodor has a famous theory of content which offers an answer to our argument so far. On Fodor's view, the causal roles definitive of content do not have to be actualized. Uninstantiated properties, even uncompounded uninstantiated properties can figure in nomic or non-accidental generalizations that connect them to experiences. According to Fodor, even though Hume blue is uninstantiated, the following could be a nomic truth: Being Hume blue is, under normal conditions, nomically correlated with "Hume blue"-tokenings in the visual system. Because this nomic correlation holds, such tokenings, were they to occur, would have the content that there is a Hume blue thing around.

I do not deny as an abstract possibility that uninstantiated properties can figure in nomic or non-accidental generalizations. It is however a very speculative elaboration of that abstract possibility to claim, as Fodor must, that any uninstantiated shade will under normal conditions be nomically correlated with the corresponding tokenings in the visual system. This is a very strong empirical claim. We can imagine cases in which any normal condition which involves seeing precludes any Fodor-style representation of the shade. We can imagine cases in which as a nomic matter of fact, the uninstantiated shade could not be instantiated. And indeed nature has already imagined such cases for us.
Before we examine one such case, what, for Fodor, is a "Hume blue"- tokening in the visual system? It is not a sense datum, not something of which the subject is directly aware. It does however involve the visual system going into a representational state which guarantees that the subject is having an appearance as of a Hume blue item. That representational state of the visual system may be nothing more than, may supervene upon, a certain pattern of firing of the cones in the retina plus the down-stream neurophysiological consequences of this. Forgetting the rods for the moment, and using the opponent process model of the cones, we have the yellow/blue opponent process subsystem of cones, the red/green opponent process subsystem and the achromatic or black/white opponent process subsystem. How things visually appear with respect to color depends upon just how these systems are activated together and the down-stream consequences of this.

Clearly, not every appearance-determining state that this tripartite system can get into is a state that it can get into under normal conditions of lighting or of viewing or of anything else at all relevant. So for example, there is a state that it can get into by being exposed to bright monochromatic unique green light (500 nanometers in wavelength) in another wise dark room for about twenty minutes. If we then turn the the stimulus off and have the subject look at a small, not-too- bright achromatic surface, he will see a red afterimage. If the after-image is then superimposed on a small red background then something wonderful happens. The subject will then be afterimaging a supersaturated red, a red more saturated or "pure" than any surface red one can see and indeed more saturated than the purest spectral red light (any light in a small band around 650 nanometers).34

On a view like Fodor's, we can say that what makes it the case that the subject is afterimaging a supersaturated red patch, i.e. what accounts for him having that visual experience, is the existence of a "supersaturated red"-token in the visual system. Although we are using words of english to pick out the token, the token itself is nothing more than an articulated state of the visual system, which in fact mainly consists in the extreme fatiguing of the cones sensitive to the middle wavelengths of light.

In Fodor's terms, there will be a "supersaturated red"-tokening in the visual system, but it will not have occured under normal conditions. Don't say that this set of conditions, however odd, is the normal condition for seeing supersaturated red. The subject is not seeing supersaturated red, he is afterimaging it. Nothing external, no light nor surface, is or could be supersaturated red thanks to the basic way the visual system functions. Under normal conditions of the operation of the visual system, you cannot experience a supersaturated red. Under abnormal conditions, you can afterimage an expanse of supersaturated red. Even if we were to call those conditions "normal" in the spirit of a "whatever-it-takes" characterization of normality, this would not help Fodor. For even under those (very abnormal) normal conditons, we do not have an instance of supersaturated red causing "supersaturated red"-tokenings in the visual system. It is monocromatic green light plus a contrast effect which causes "supersaturated red"- tokenings.

Now because we can only have experiences of supersaturated red by way of afterimaging, supersaturated red is a "Hume" red. Moreover, in Fodor's rather weak sense of "nomic," this is nomically so, i.e., it is non-accidentally determined by the very structure of the visual system and by the nature of light. (A fortiori, J.J.C.Smart's account of afterimaging as "having something go on in you which is like what goes on in you when you see such and such." cannot be adequte.) Can it be a "nomic" matter of fact that nothing instantiates supersaturated red and nonetheless also be a nomic matter of fact that under normal conditions of the operation of the visual system being supersaturated red causes "supersaturated red"-tokenings in the visual system? This would save Fodor's theory, but it is hardly plausible. "Under normal conditons of the operation of the visual system, if anything were to be supersaturated red then..." invites us to indulge in a supposition contrary to nomic regularities. We can have no confidence in any way of filling out the consequent. So we can have no confidence in any theory which tells us that a visual experience of afterimaging supersaturated red has the content it does because under normal conditions of the operation of the visual system, if anything were to be supersaturated red then it would cause "supersaturated red" tokenings in the visual system. Fodor's "nomic" connections and unactualized causal roles do not help in associating a content with the hallucinating or afterimaging of expanses of missing shades.
Our attempts to model the content of visual experience have failed. It is not that the subject is deploying the concepts that make up the content of visual experience as his way or mode of thinking about or being aware of the objects of his experience. This would be to falsely model visual experience on thinking, the content of which does depend on how the thinker conceives of his subject matter. It is not that there is a representation in the sense of a mental array or sense datum or linguistic description figuring in the subjects mental life, a representation that is more naturally associated with one content or condition of correctness rather than another. This would be to falsely model visual experience on the production of written words that determine a sense or content for what is written. Finally, it is not because the subject of experience is undergoing an event that is significantly causally connected with what it represents. This would be to falsely model visual experience on signaling or indication as when a beaver waves its tail to indicate or signal danger to its mate. Visual experience is not like any of these things. So what has content got to do with it?

 

The Adverbial View

The temptation to treat visual experience as the entertaining of a content comes from two sources. First, visual experience prompts and in some sense supports immediate perceptual beliefs whose contents characterize scenes. If visual experience is to constitute evidence for the propositional beliefs which immediately arise upon visual experience then it had better be propositional or content bearing. Only a propositional content can entail or be evidence for a propositional content. Compare Ludwig Wittgenstein's remark "How can a proposition follow from a sense-impression?"35 This shibboleth, which is at the heart of the rejection of the idea that something nonpropositional is "given" in experience, becomes much more puzzling once we think through what visual experience involves. If having a visual experience is neither believing nor acquiring the disposition to believe then how can something that is not a belief entail or be evidence for a belief? Either we go with David Armstrong and identify visual experience with immediate belief, thereby mischaracterizing visual experience, or we must rethink the idea that the relation between experience and belief is one of evidential support. Perhaps the function of experience is to make present what immediate perceptual belief is about. This doesn't justify immediate perceptual belief but it makes clear why such belief could have the status of a source of justification for other beliefs.36

A second source of the idea that visual experience is the entertaining of a content is that each visual experience has a definite character. Short of endorsing the Sense Datum Theory, how can that fact be captured without taking visual experience to be contentful? One alternative to both theories is the Adverbial Theory of visual experience. The correct account of visual experience, sometimes called the Theory of Appearing emerges out of the difficulties with this theory.37

On the Adverbial Theory, enjoying a visual experience is sensing in a certain manner, so that when hallucinating dim lights you are being appeared to dimly, radiantly and roundly.38 The adverbs characterize the manner of sensing so as to attach a character to the sensing without requiring that the sensing have an object, material or immaterial. So something of the form

S enjoys an visual experience of an F

gets translated into something of the form

S is appeared to F-ly

which commits us to events of appearing along with manners or ways of appearing but not to contents visually entertained or to sense data-like mental arrays. So there is a manner of being appeared to that is being appeared to Hume bluely. Since this is not entertaining the content that something is Hume blue it is something that can happen to you though you are not deploying the concept Hume blue, though nothing is Hume blue' as Peacocke would have it, and though there is nothing Hume blue to be causally connected with such an experience. Unfortunately, despite avoiding mental arrays, the Adverbial Theory shares the major defect of the Sense Datum Theory.

The Sense Datum Theory has been unattractive for several reasons, depending on its exact formulation. There is the "idle wheel" objection, which is now moot since we have found no satisfactory account of the content of visual experience. Then there is the phenomenological claim that we neither see sense data nor are aware of them when we attend to our visual experience. However, there is another, philosophically more resonant difficulty with the Sense Datum Theory. It treats veridical experience, experience of "external" objects, as experience plus appropriate causation. In this way it recapitulates the Cartesian problematic concerning our contact with the external world. For it models experience as essentially complete whether or not experience has an external object. So it is not inherent to experience that it is a relation to extra-mental reality. Appropriate causation is required as well. I do not think that the real difficulty with this picture of experience is the familiar one concerning the justification of the immediate beliefs based on experiences so conceived. Perhaps most of those beliefs are justified simply because they are reliably caused in us and there is no reason to doubt them. Even if Cartesianism leaves us no worse off with respect to the justification of immediate belief it still wrongly depicts the mind as a interior, private realm in which the mental effects of external things occur.39

Hence the contemporary resistance to the Cartesian picture of experience as essentially complete, whether or not experience has an external object, a resistance which takes the form of Intentionalism about visual experience combined with the so-called "Externalist" claim that visual experience could not have content unless there was an external world. But if enjoying a visual experience is not entertaining a content then the contemporary resistance to the Cartesian problematic is a failure. We must look elsewhere for an account of experience that does not treat experience as essentially complete, whether or not experience has an external object.

As it stands, the Adverbial Theory fails precisely on this score. As well as sentences of the form

S enjoys a visual experience of an F

or equivalently

S is appeared to F-ly

There are sentences of the form

S enjoys a visual experience of x

where x is some object which S sees. How is the Adverbial Theory to handle the seeing of objects? It can reduce the seeing of objects to being appeared to F-ly for an appropriate manner of appearing F. Or it can treat the seeing of objects as primitive. As we shall now show, either strategy recapitulates the Cartesian problematic.

This defect is more obvious given the reductionist strategy. The straightforward way to analyze Sam's enjoying a visual experience of Mount Everest in terms of his being appeared to in some manner is like this:

(8) Sam is enjoying a visual experience of Mount Everest if and only if there is some cluster F of visible features which hold of Mount Everest such that:

(i) Sam is appeared to F-ly,
(ii) Mount Everest is causally responsible in the standard way for Sam's being appeared to F-ly, and
(iii) Mount Everest is the F in the scene before Sam's eyes.


This "conjunctive" analysis displays the fact that enjoying a visual experience of Mount Everest or equivalently, seeing Mount Everest, is a matter of having an experience which one could have if one were simply hallucinating Mount Everest, an experience that in this case is appropriately caused and appropriately satisfied by an external thing.

Suppose instead that we treat enjoying a visual experience of an external object as irreducible, so that the resultant version of the Adverbial Theory has two basic idioms, namely

S is appeared to F-ly

and

S is appeared to F-ly by x.

On this version, the Adverbial Theory admits two general forms of being appeared to, what we might call respectively the monadic form and the relational form. When one hallucinates lights on in the ceiling one is enjoying the monadic form of appearing. When one sees lights on in the ceiling one is enjoying the relational form of appearing. Since they are both forms of being appeared to F-ly (dimly, roundly, in the manner of things in a ceiling, etc.) the transition from one to the other can be experientially seamless. Despite the fact that veridical experience might be on occasion indistinguishable from hallucination, a veridical experience is a case of enjoying the relational form of appearing, a form of appearing which cannot be had without external objects.

This position is unstable. Consider that the cases of seeing dim lights in the ceiling and hallucinating dim lights in the ceiling may have the very same type of immediate cause, a pattern of neurons firing in the visual cortex. In the case of hallucinating the lights, this cause produces an immediate effect. A monadic property comes to hold of the subject: he is appeared to dimly, roundly, in the manner of things in a ceiling, etc. The same immediate cause, down to whatever level of neural detail is desired, can be present in the case of seeing lights in the ceiling. Thus, the same immediate effect should be produced in the case of seeing lights in the ceiling. So here too a monadic property comes to hold of the subject: he is appeared to dimly, roundly, in the manner of things in a ceiling, etc.. By hypothesis, a relational property also comes to hold of the subject: he is appeared to dimly, roundly, in the manner of things in a ceiling, etc. by the lights in the ceiling. Now the position faces a destructive dilemma: either every case of veridical perception is a case in which two qualitatively indistinguishable appearings occur, which is absurd, or the relational appearing that a veridical experience involves is always a matter of enjoying the monadic form of appearing plus appropriate causation and appropriate satisfaction in the manner of (8) above.

The Cartesian problematic is thus reinstated. Veridical experience does involve a form of appearing which can be had whether or not there are external objects appropriately related to the subject.

Evidently it is not plausible to take the "same immediate cause, same immediate effect" principle to imply that in the case of hallucinating lights the neural cause makes it the case that a relational property comes to hold of the subject, namely that he is appeared to redly, etc., by the lights on in the ceiling. That would mean that the hallucinator sees the lights in the ceiling, which he obviously does not. How then are we to restrict the principle so that this absurd consequence does not follow from the principle itself? Here I am guided by some remarks of Harold Langsam. In a similar context, Langsam writes

Langsam's suggestion is that we limit the principle so that it applies only to intrinsic changes in the thing affected. Though one "effect" of the worm's movement in the second case was to make it ten feet from the soda can, this was not an intrinsic change in the worm or the can. My preferred restriction is like Langsam's. The changes that are to be taken as effects in applying the principle are changes in the monadic properties of the things affected. It doesn't follow from the "same immediate cause, same immediate change in monadic properties of the thing affected" principle that every case of hallucination is a case of seeing. The relational property of being appeared to dimly, etc. by the lights in the ceiling cannot be carried across from the case of seeing to the case of hallucination by applying the restricted principle.

Our original question was "How is visual experience constituted so that in a case of hallucination and a case of veridical perception there need be no discernible difference?" Given (i) the restricted principle, (ii) the view that hallucination is a monadic state of the subject and (iii) the view that any veridical experience could have the same immediate cause as a hallucination, the Cartesian or mentalistic conception of experience follows. For together the three premises entail that veridical perception involves a monadic experiential state that could occur whether or not there were external objects appropriately related to the subject. Only by coming to terms with this argument can one "refute the Sense Datum Theory" or more generally the conception of experience which the Sense Datum Theory endorses. So in particular, J. L. Austin did not refute the Sense Datum Theory, let alone extirpate the conception of experience which lies at the heart of it. Austin's coup de gras against A. J. Ayer and H. H. Price was supposed to be that even if cases of hallucination, dreaming, etc. were experiences that did not relate their subjects to external objects, and even if a hallucination could be more or less exactly like a veridical experience, there is simply no argument that the direct object of the veridical experience cannot be Mount Everest or any other external object.41 The restricted principle rearranges the playing field. Given Austin's (reluctant) concessions it follows that every case of veridical perception of external objects involves the monadic or purely mental state present in the corresponding case of hallucination.42

That observation explains why I have not considered the currently fashionable, neo-Austinian dismissal of our question, which goes under the title of "the disjunctive conception of experience." According to the disjunctive conception of experience, one should not look for a common factor -- sense-datum or content or whatever -- in the cases of hallucination and veridical perception.43 All that should be said is that there is a determinable kind of experience -- it seeming to you that you are visually experiencing lights up there on a ceiling -- which divides into two sub-kinds, namely your seeing lights up there on the ceiling and your hallucinating lights in a ceiling. Now if the determinable kind of experience is really its seeming to you that you are experiencing lights in a ceiling, then the disjunctive view is after all just another, higher-order, version of Intentionalism. In both cases, in perceiving the lights and in hallucinating lights, I am entertaining the content that I am seeing lights. In veridical perception the content is true and my entertaining it is appropriately caused. In sheer hallucination the content is false and my entertaining it is not appropriately caused. In veridical hallucination the content is true and my entertaining it is not appropriately caused. Haven't we then found the common factor as between hallucination and veridical perception? It's a higher- order content: that I am seeing lights. If the Disjunctive Theorist tries to deserve his name and reject any such common factor then he leaves us without any account of why the determinable kind of experience -- it seeming to me that I am seeing some lights -- is anything but an arbitrary disjunction of kinds of experience with no common nature.44

This is the position of some Disjunctive Theorists: there is nothing interesting to be said about the relation between veridical perceptions and their corresponding hallucinations. Forget the argument from hallucination. Don't even try to respond to it. There lies confusion and the Cartesian "conjunctive" conception of experience.

I take it that the case of seamless transition is at least a mild embarrassment for this dismissive position. Perhaps the level of embarrassment can be raised by considering a mixed case of hallucination and veridical perception in which you see some lights and at the same time hallucinate others. The hallucination of lights partly determines the character of your visual experience, as does the veridical perception of lights. How is visual experience constituted so that this can be so?

I will not press such objections here, because in any case the dismissive position faces a dilemma concerning hallucination. Is a hallucination a monadic state of the subject? If so, the restricted principle combined with the fact that the immediate cause of any veridical experience could be the same as the immediate cause of a corresponding hallucination implies the Cartesian conception of experience.

What if a hallucination is not a monadic state of the subject, so that the restricted "same cause, same effect" principle does not apply? Then, as we shall now see, the right account of visual experience is neither disjunctive nor conjunctive.

 

A Way Out

We seem stymied. There seems no satisfactory response to the argument from hallucination that does not recapitulate the Cartesian conception of experience. There is a way out. We can recognize that strictly speaking there aren't any hallucinations. What we call hallucinations are really illusions, albeit of a radical sort.

Consider familiar cases of illusion, cases in which ordinary things appear in ways that do not show how they really are. The straight stick in the water appears separated. The coiled rope in the dark looks like a snake. The Moon looks bigger at the horizon. Although the argument from hallucination is sometimes called the argument from illusion, it is clear that one could not validly argue for the Cartesian conception of experience from pairs of veridical experiences and their corresponding illusions. Let's try it. Assume that a case of seeing a snake in the darkened corner of the room could be indistinguishable from a case of seeing a coil of rope as a snake in the darkened corner. Even so, nothing follows about anything in common at the level either of monadic mental states or of relations between the subject and wholly mental items. In the terminology of appearings, all that is happening is that in the first case one is appeared to in a serpentine manner by a snake while in the second one is appeared to in a serpentine manner by a rope. In each case one is enjoying an appearance that is a relation to an external object. So construed these experiences are not such that they could occur without their objects. For all that cases of illusion show, visual experience could be an unanalyzable relation of being appeared to by things.

A hallucination is supposed to be a visual experience in which nothing, or at least no object or state external to the mind, is appearing to the subject. So we considered hallucinating lights on in a ceiling. We supposed the room was initially completely dark, so that nothing could visually appear to the subject. In the same vein, Macbeth hallucinated a bloody dagger and suspected it was " a fatal vision proceeding from a heat oppressed brain." If so, it seems that nothing was appearing daggerishly to Macbeth. Yet I think that both these cases are really cases of illusion, not hallucination.

By way of leading up to this claim, let's examine another kind of case, the phantom forearm example. What is taking place when the amputee feels pain beyond his stump "in his phantom forearm?" One thing that is taking place is that a state of bodily disorder - as it might be, the stump's having nerves left scarred or mangled from the amputation -- is appearing as pain or appearing painfully. This state of bodily disorder that is appearing as pain is also appearing to be where it is not, namely in the space beyond the stump. We do not here have a hallucination but an illusion of location. The illusion mislocates the bodily disorder that appears painfully to the subject.

The pain-in-the-phantom forearm illusion is to be contrasted with the rope- as-snake illusion. In that illusion the way in which the rope appears to the subject is misleading, not with respect to location, but with respect to the properties of the thing at that location. This raises the possibility of a radical illusion that is misleading both in respect of location and in respect of the properties of the object of the illusion. There would be no way of reading off from the illusion any significant information about the object of the illusion. My view is that Macbeth's "seeing" the dagger, the hallucinating of lights in a ceiling, after-images and indeed most dreams are all radical illusions in this sense. They all involve something real appearing in ways that radically mislead as to the nature and location of the thing appearing. Let me lead up to this claim by way of a thought experiment.

 

The Blind Hallucinators

There is nothing incoherent about the idea of a fragmentary form of sensing whose deliverances are not integrated along with other sensory appearings into a working perceptual model of the world. An isolated tribe of human beings could have been genetically determined to be functionally blind from birth except for brief visual flashes every month or so of some scene around them. These visual experiences, because of their brevity and unfamiliarity, would be useless in building up a model of the world. However, the lack of integration of the visual appearances which "flash" now and then does not imply that they are of nothing. They are simply unintegrated visual appearances of what is before the eyes of the one enjoying the flash.

We can extend this thought experiment. Suppose the blind tribe are vivid and frequent hallucinators and have come to rely on their hallucinations as registrations of their internal states. Their environment is provided with variously shaped fruits which contain ingestible chemical precursors for various neurotransmitters. Over time, members of the blind tribe have become adept at detecting by hallucination various conditions of their brains. Some of these conditions they treat by ingesting the appropriate fruit. (Compare a diabetic who senses his blood sugar is low or a nicotene addict who immediately knows when he needs a smoke.) In order to develop the relevant abilities to recognize and discriminate on the basis of their hallucinations what brain state they were in, the members of the blind tribe probably would have to have some other access to those states, say (since we are already in a flight of fancy) by way of a brainscanner with intersubjectively available (tactile?) output. Then initial guesses, taking the hallucination to be of this or that brain state, could be checked and corrected, and the relevant recognitional abilities developed. Thus it could become entirely natural for the blind tribe to think of hallucination as a form of bodily sensation, i.e., a way of sensing their brain states, or equivalently, a way in which their brain states appear to them. Nor need they be confused in this. Detecting their brain states by hallucination could have all the signs of fine sensory discrimination. They would have to learn it. Some could be better or worse than others at doing it. There is room for care and carelessness, success and failure in trying it; and there is the possibility of arriving at correct or incorrect belief about what one is sensing. In these ways sensing slow serotonin re-uptake in one's temporal lobe could be rather like sensing the trace of the oak barrel in a thirty year old Oban.

One day, a member of the blind tribe has a visual flash indistinguishable from a hallucination he had the week before. He reasons that since his flash was of nothing and was indistinguishable from his hallucination then his hallucinations must be thought of as monadic appearances plus appropriate causation by brain states. He has arrived at a "Cartesian" conception of hallucination by wrongly supposing that his flash was of nothing at all. That supposition ignores an alternative possibility, namely that his flashes are fragmentary and unintegrated appearances of the scene before his eyes.

The philosophical treatment of hallucination has ignored just such a possibility. Hallucination, after-imaging and dreaming are for us fragmentary and unintegrated, and hence practically useless, forms of sensing our internal condition. So understood, they provide no obstacle to supposing that all appearances are appearings of some real thing or real state.

But which things or states appear in the case of hallucination?
Contrary to what I once thought there is really no motive for thinking that any well demarcated thing or state appears. In the case of the blind tribe, who deploy when they hallucinate a host of internal recognitional abilities, there is reason to locate this or that specific brain state as the one that appears. But we lack any such developed recognitional abilities. When we hallucinate we are like the members of the blind tribe enjoying a "visual flash". Because they lack any external recognitional abilities there is nothing more specific to be said of them than that they have visual flashes of the scene before their eyes. Compare an infant still enjoying the "booming buzzing confusion" of immediate visual input and as yet without the ability to visually recognize much at all. Still, the infant's visual experiences are of its surroundings. So in the case of our hallucinations the object should be thought of as correspondingly non-specific; say, the hallucinator's own internal condition. With the acquisition of the corresponding recognitional abilities the objects of hallucination can get more specific. For example, I can imagine a sensory physiologist when after-imaging being aware of his retinal states.

So we cannot quite say that Macbeth's "fatal vision" not only proceeds from a heat- oppressed brain but is of a heat oppressed brain. Rather it is a illusory appearance of his own internal condition.

Some find these claims obvious, others are incredulous. Both reactions seem to me mistaken. Incredulity spreads out from the conviction that there was no sense in which Macbeth was visually aware of his internal condition. One line of thought is that if Macbeth were aware of his condition then it would be present to him, it would be a this for subsequent thought, an available object of demonstration for Macbeth and so a potential topic of further thought and talk, as in "Is this my condition which I see before me?" Surely Macbeth's hallucination did not put him in a position to demonstrate his internal condition.

In response, we can distinguish two kinds of demonstration and two corresponding ideas of presence. When I see the rope as a snake, I am still able to demonstrate the rope as a topic for further thought and talk. I can pick it out barely as "this" where my intention is no more than to demonstrate the object of my experience, whatever it turns out to be. I can pick it out as "this snake-looking- thing" or, if I'm not taken in by the illusion, as "this rope." In this last case, I employ a correct sortal or term for the sort of thing in question -- "rope" -- as part of the demonstration, a predicate that associates the correct conditions of identity at a time and over time with the object of my illusion. I can do this because I'm not taken in by the illusion. The illusion itself does not ground any demonstration that employs a correct sortal.

So experience can make objects present to us in different ways. One way it can make an object present is to make it available as an object of bare demonstration, where no correct sortal qualifies the demonstrative. The object is then, at most, a mere "this" for thought. In the best case, the case that most naturally comes to mind when we think of experience as a way in which things are made present, experience provides for a demonstration of objects in terms which employ a correct sortal. This is the case where the rope appears not as a snake but as a rope.

The claim that Macbeth is aware of his internal condition might seem bizarre precisely because it suggests that Macbeth's hallucination enables him to demonstrate his internal condition as a condition of himself. To do that he would have to be aware of his internal condition as his internal condition. Obviously he is not. Equally obviously, this line of thought cannot rule out the claim that Macbeth is aware of his internal condition as, say, a dagger in the air. Macbeth's radical illusion does not put him in a position to pick out the object of his experience in terms of an appropriate sortal. But that is the case for many illusions. In hallucination the object of awareness need not be present in any way which reveals to the subject the sort of thing it is. That is how it is with a "radical illusion."

The classification of hallucinations as radical illusions opens the way for the following position. Experience is always the appearing to a subject of this worldly, non-mental objects. The appearing of an object to a subject is not analyzable in terms of an intermediate relation of the subject to a mental array or a mental content that in its turn "represents" the object. Contrary to the orthodoxy in these matters experience properly so called does not involve representational content. It presents things in a way that is prior to re-presentation. Each appearance has its own character, which may be rendered by specifying the qualitative features of the object appearing. So that what are misleadingly called "qualia" are really qualities of a kind which are had by external objects. Two appearings can involve the same qualities even though they are appearings of very different things. That is what happens when there is an experientially seamless transition from hallucinating lights on in a ceiling to seeing the lights on in the ceiling. When you hallucinate lights on in a ceiling, your experience is of nothing other than your condition.

The private, interior realm, the mind understood as a realm which wholly contains hallucinations, after-images and the like, is the offspring of a confusion of appearings with monadic appearances. "No one can have my experiences." Right. "My experiences are private to me." Right, so long as this is just an awkward way of saying the same thing, i.e. that my experiences are appearings in which I am one of the terms of the relation of appearing. "They exist in a private realm, my mind." Wrong, there is no private realm. My experiences are appearings; events which are the coming to hold of relations between me, certain qualities and the things I experience. Those events are individuated by the terms of the relations; me, the qualities and the objects I experience. This means that no one can have my experiences in just the sense that no one can undergo my innoculation against flu. And it means that my experiences of Prague are as private to Prague as they are to me.

Of course my path though the world is unique. As a result I occupy a unique series of occasions of appearing. But these occasions of appearing, these prospects, these outlooks, these viewing positions, these seats in the concert house or the lecture hall are all publicly available occasions of appearing. Beyond involving this or that subject of appearance there is nothing "subjective" about appearings. Descartes' dreams and hallucinations, from which he spun a conception of the mind as a private, interior realm, were after all fragmentary sensings.

 



Footnotes

 

1. In Reason, Truth and History (Cambridge, 1983), Ch. 2, Hilary Putnam claims that if we were brains in vats then we couldn't mean the standard thing by "WE ARE BRAINS IN VATS" so that we could not formulate to ourselves the traditional problem of the external world. Notice that even if this were so, it would not in any way deal with the deeper epistemic anxiety associated with the problem of acquaintance. Indeed, Putnam's model-theoretic picture of the determination of meaning just capitulates on the problem of acquaintance. I see little interest in a proposed solution to the traditional problem if the proposal implies that the problem of acquaintance cannot be solved.

2. See The Problems of Philosophy (London, 1912), p. 47.

3. The best motivation for this local supervenience claim is given by Ned Block's Inverted Earth example. The example simply mobilizes intuitions. The argument for local supervenience of the higher-order "depictions" of the natures of the colors is that the alternative, namely that the content of these depictive representations of the colors of external objects are fixed by the natures of those colors themselves, makes color misperception much harder than it is.

4. In arguing this I am not assuming that we see edges or areas or surfaces by seeing color difference. I assume only that certain necessary conditions on seeing these things hold. If we could see color and hence color difference and hence shape, then it might be right to count as seeing making out shapes with night-vision goggles which misrepresent all the colors in the scene before the eyes. But if we never see colors, then this does not count as seeing either. All that is going on is the systematic representation in an arbitrary medium of information in the scene before the eyes. This is as good as what we call seeing, but neither is seeing in the sense of visual acquaintance with the scene before the eyes.

5. This claim about causal processes is akin to the same immediate cause, same immediate effect principle appealed to below. Our idea of a causal process is the idea of an extended event, which we could break at any point into an immediate cause and an immediate effect. How the immediate effect is, at least as far as we can characterize that in monadic or intrinsic terms, is determined by how the immediate cause is, at least as far as we can characterize that in monadic or intrinsic terms. This means that how things were with events or states before the immediate cause can only matter insofar as this information is included in how the immediate cause is intrinsically. Our very conception of causal processes seems to rule out the possibility that an experience conceived of as something on one end of a causal process could reveal the nature of something at the other end. An enormous variety of things of varying natures could issue in the same immediate physical causes of experience.

6. Thanks to James Pryor for helping me see this.

7. For this argument see Frank Jackson Perception pp.

8. For a very detailed defense of the Sense Datum Theory see Frank Jackson Perception: A Representative Theory (Cambridge University Press, 1977).

9. Early advocates of Intentionalism include D.M. Armstrong Perception and The Physical World (Routledge Keegan Paul, 1961) and George Pitcher A Theory of Perception (Princeton University Press, 1971). Interest in Intentionalism was recently revived by Gilbert Harman's " The Intrinsic Quality of Experience" in James Tomberlin ed. Philosophical Perspectives 4 (1990).

10. Pitcher, Perception, pp 92-93. Armstrong, A Materialist Theory of Mind (Routledge and Kegan Paul, 1968) pp222-223.

11. See many of the papers in Philosophical Issues,7 (Ridgeview, 1996) a volume on perception edited by Enrique Villanueva.

12. "The Intentionality of Sensation" in R.Butler ed.

13. In a nice unpublished paper on Harman's work, Benjamin Hellie calls the little psychological sentences we have been discussing "tracts" and not "fictions", since they may be true or false. He then considers what exists according to these tracts and finds Harman's intentional objects among them. I do not find it clear that the Fountain of Youth exists according to (2).

14. This explanation of the seemless transition does not suppose that visually entertaining a content involves being aware of the content visually entertained. The common contents are not the objects of awareness. That would be to naively replicate the Sense Datum Theorist's veil of perception -- intermediate objects of awareness between us and worldly items. Instead, entertaining contents under favorable conditions constitutes what it is to see or be visually aware of objects.

15. Does this analysis endorse the Cartesian conception of experience as complete whether or not there are external objects? Not necessarily. A subject's visually entertaining the content that P (where P entails that b exists) may be an event which can occur only if b exists and is seen by the subject. This would be the case if, for example, the content P involves a direct demonstrative reference to b. When the content P is "broadly determined" in this way, i.e., when it is individuated in terms of the object seen, the Cartesian conception need not follow. For we can plausibly maintain that an experience, understood as an event in which a subject comes to entertain a content, is partly individuated in terms of the subject and the content entertained. If we add (as McDowell does, but Burge does not) that the content entertained is essentially "object-involving" i.e. is individuated interms of the object successfully demonstrated, namely b, then the very experience in question could not have occurred in the absense of b. See John McDowell "De Re Senses" Philosophical Quarterly, 34, 1984. Tyler Burge rejects this externalist view of demonstrative content in "Vision and Intentional Content" E.Lapore and R.Van Gulick eds. John Searle and HisCritics (Basil Blackwell, 1991). As I understand Burge's view, whatever Externalism about content there is comes not from demonstative content but from the predicative component of content.

16. Even if someone could see through walls he would not be capable of visually entertaining that content. For if he could see through walls then his visual experience would not represent walls as opaque.

17. Intentionality: An essay in the philosophy of mind (Cambridge, 1983) p 48.

18. Notice that Searle's self-referential contents are not to be identified with what a subject immediately believes on the basis of perception. After all, in the operating room I could be authoritatively told that I am having a hallucination when in fact I am having a visual experience. As a result I could firmly disbelieve that I am having a visual experience of lights in the ceiling caused in me by lights in the ceiling. More extremely, as a result of bizarre but well-honed philosophical convictions I could firmly disbelieve that there are visual experiences or that my visual experiences are "of" anything. None of this matters once we accept the Intentionalist's claim that perception in general and visual perception in particular is a sui generis mode of entertaining a content.

19. Armstrong op. cit.

20. It is Fred Dretske, more than anyone else, who has forcefully argued this point. See Seeing and Knowing (University of Chicago Press, 1969); "Seeing, Believing and Knowing" in Visual Cognition and Action eds., D. Osherson, Stephen Kosslyn and John Hollerbach (MIT Press, 1990).

21. See their papers and Searle's reply in Searle and his Critics, E.Lapore and R. Van Gulick eds. (Basil Blackwell,1991).

22. See Christopher Peacocke "Scenarios, Concepts and Perception" and Tim Crane "The Non-conceptual Content of Experience" in Tim Crane ed. The Contents of Experience (Cambridge, 1992). It was Gareth Evans in Varieties of Reference (Oxford University Press, 1982) who first highlighted this notion.

23. Armstrong op. cit.

24. It is Fred Dretske, more than anyone else, who has forcefully argued this point. See Seeing and Knowing (University of Chicago Press, 1969); "Seeing, Believing and Knowing" in Visual Cognition and Action eds., D. Osherson, Stephen Kosslyn and John Hollerbach (MIT Press, 1990).

25. For David Hume's remarks about the missing shade of blue see A Treatise of Human Nature, ed. L.A.Selby-Bigge and P.H.Niddich (Clarendon Press,1978), 1.1.1.

26. Things are not quite so simple as I have made out in the text. The color solid is not symmetric; e.g., yellow is more similar to white than the other hues. The rotation I envisage would not preserve these assymetries. But, this does not effect the main point. Imagine that we never saw blue or yellow. How could we then distinguish Hume red and Hume green? Not by way of distinctive relations of similarity and difference. (As emerges below we would also have to assume that we have seen no unique green.)

27. For a well-developed account of the colors as light dispositional properties see David Hilbert's Color and Color Perception (Stanford, CLSI, 1987).
28. For an excellent overview of the two theories and their relation to one another, see the introduction and papers in Nathan Salmon and Scott Soames and Propositions and Attitudes (Oxford University Press, 1990)

29. PMP(D) has to be rigidified by the use of a device like "actually" otherwise the content of visual experience, <the wall, D>, will not strictly imply the content of immediate perceptual belief, <the wall, PMP(D)>.

30. There is another strategy, which is rather desperate. We can suppose that the non-commital modes of presentation have the relevant similarity conditions built into them as conditions on any properties turning out to be sky blue or azure or canary yellow. Now visual experience is implausibly being depicted as treating the colors as like properties described exclusively in some legend, properties which exist only if a significant part of the legend is true. Compare the property of being Greek Fire. I discuss such conditionalizing of the modes of presentation of the colors in "A Mind-Body Problem at the Surfaces of Objects" in E.Villaneuva ed. Philosophical Issues, 7, op. cit..

31. In the alternative idiom of guise theory, immediate perceptual belief will habitually be introducing guises for colors, guises which visual experience in no way justifies or anticipates.

32. Compare the argument in Paul Boghossian and David Velleman "Color as a Secondary Quality" Mind 98, 1989, 81-103. Boghossian and Velleman regard this as an argument against any version of the dispositional theory that does not rely upon sense data. I think it has a more restricted scope. A dispositionalist should adopt a third view of visual experiences distinct from Intentionalism and Sense Datum Theory. See my "Visual Experiences" in A.Byrne and David Hilbert eds. Color Perception forthcoming.

33. See Chapter 2 of Sense and Content (Oxford University Press,1983).

34. See Leo M. Hurvich Color Vision (Sinauer Associates Inc., Mass., 1982), pp 187-8. I thank David Hilbert for explaining to me the empirical subtleties surrounding supersaturated red and for the reference to Hurvich.

35. Philosophical Investigations, section 486.

36. There are various ways of developing the idea that sensing makes present what immediate perceptual belief is about. Just to sketch one: suppose that x's sensing y in some manner M; equivalently, y's appearing M-ly to x, is the relation that constitutes sensory awareness. If the subject x then forms the belief that y appears M-ly, he has a belief formed in the presence of the very relation which makes that belief true.

37. See C.J. Ducasse "Moore's Refutation of Idealsim," in Philosophy of G.E. Moore P.A. Schlipp, ed. (Northwestern University Press, 1942) pp225-251; R.M. Chisholm Perceiving (Cornell University Press, 1957) pp 115-225; and Person and Object (Open Court, 1976), pp 46-52; Wilfred Sellars Science and Metaphysics (Routledge and Kegan Paul, 1968) pp9-28.

38. There are a number of difficulties that have been raised about how to handle the semantics of these adverbial modifications. On these difficulties and their solution see Michael Tye "The Adverbial Approach to Visual Experience" Philosophical Review, XCIII, 1984, pp195-225. The theory of appearing differs from the adverbial theory in important ways. For one, the sensible qualities of the thing appearing are elements in the relation of appearing, not manners which qualify the appearing itself. I thank Steven Schiffer for reminding me of this. As he points out if the adverbial theorist is right and there is such a thing as an appearing redly then it is hard to resist the absurd conclusion that the appearing is red. For the following seems like a valid inference form: if x is a F-ing Gly then the F-ing that is x is G. The most important difference between the two theories is discussed int he main body of the text.

39. This metaphysical picture of the mind may in its turn produce a different kind of epistemological worry, a worry which I have elsewhere called "the problem of acquaintance." By treating experience as a mental event at the end of a causal chain, a mere effect of the external cause, experience will then seem arbitrarily related to the nature of the thing experienced. See my "Is The External World Invisible?" in E. Villaneuva Philosophical Issues,7 (Ridgeview, 1995).

40. "The Theory of Appearing Defended" forthcoming in Philosophical Studies. In this very insightful paper Langsam takes himself to be defending, inter alia, the disjunctive view. He also suggests that hallucination could be a monadic state of the subject. See below for an argument that Langsam's principle is actually inconsistent with these views.

41. See Chapter V of Sense and Sensibilia (Oxford University Press, 1960).

42. This is one of several places where I differ from Hilary Putnam, who in his recent Dewey Lectures calls for a non-representionalist account of perception. See pp 469-475 of Lecture 2 "The Importance of Being Austin: The Need for a Second Naivete" Journal of Philosophy, XCI, 1994.

43. For the disjunctive view see John McDowell "Criteria, Defeasibility and Knowledge" op. cit..and Jonathan Dancy "The Argument from Hallucination" Philosophical Quarterly,45, 1995.

44. McDowell does allow "what is given in experience in the two sorts of case (vision and hallucination) to be the same in so far as it is an appearance that things are thus and so." ibid pg.475. But what is an appearance?