INTROSPECTION AND PHENOMENAL CHARACTER
Sydney Shoemaker – Cornell University
There are three different questions that an account of the introspective awareness of phenomenal character should answer. Assuming that awareness involves belief, such an account should include an account of the nature of the belief involved in such awareness. Since the nature of the awareness will presumably depend on what sort of thing it is an awareness of, the account should include an account of what phenomenal character is. And it should include an account of how the phenomenal character of experience issues in beliefs of the appropriate sort.
I have argued elsewhere that introspective awareness should not be construed on the model of sense perception, and in particular that it should not be construed on the model of “object perception.”  It goes with this that introspective awareness is “awareness that” -- in Fred Dretske’s terminology, it is “fact awareness” that is not mediated by “thing awareness.” Thus the belief involved in such awareness is not a belief de a (non-factual) object of introspective awareness. And, correspondingly, phenomenal states should not be conceived as involving the existence of phenomenal objects, e.g., sense-data, suited for being the objects of a quasi-perceptual awareness. I shall not be arguing directly for this claim, but part of my aim is to give an account of the introspection of phenomenal character that is compatible with this claim about introspective beliefs and phenomenal states, and with the corresponding claim about the relation between these (viz, that it is not a quasi-perceptual relation).
One view I hold about the nature of phenomenal character, which is also a view about the relation between phenomenal character and the introspective belief about it, is that phenomenal character is “self-intimating.” This means that it is of the essence of a state’s having a certain phenomenal character that this issues in the subject’s being introspectively aware of that character, or does so if the subject reflects. Part of my aim is to give an account which makes it intelligible that this should be so.
A more substantive view I hold about phenomenal character is that a perceptual state’s having a certain phenomenal character is a matter of its having a certain sort of representational content. This much I hold in common with a number of recent writers, including Gil Harman, Michael Tye, Bill Lycan, and Fred Dretske. But representationalism about phenomenal character often goes with the rejection of “qualia,” and with the rejection of the possibility of spectrum inversion and other sorts of “qualia invesion.” My version of representationalism embraces what other versions reject. It assigns an essential role to qualia, and accepts the possibility of qualia-inversion. A central aim of the present paper is to present a version of this view which is free of the defects I now see in my earlier versions of it.
As I have said, one of my aims in this paper is to give an account of phenomenal states, and our awareness of them, that is compatible with the possibility of spectrum inversion. But what I mean by the possibility of spectrum inversion is something more modest than what is commonly meant. What is often at issue in discussions of spectrum inversion is the possibility of a state of affairs in which the phenomenal character of the visual experiences of two subjects is systematically different when they are in the same circumstances and viewing objects alike in color, and in which the phenomenal character of the subjects’ visual experiences is sometimes the same when the objects they are viewing are different in color, and yet there is no behavioral difference between those subjects – they make the same color discriminations, apply color words in the same way, and so on. Such a case would be a case of behaviorally undetectable spectrum inversion. One reason why the discussion has fastened on the possibility of such cases is that the supposition that they are possible can seem to aggravate the problem of knowledge of other minds. Another reason is that it is the possibility of such cases that seems to imply that qualia, the properties that give experiences their phenomenal character, are not functionally definable. It seems a short step from holding that there can be behaviorally undetectable spectrum inversion to holding that two subjects who are functionally indistinguishable, at the psychological level, can be spectrum inverted relative to each other -- and if this is possible, there can be a psychological difference where there is no corresponding functional difference. It is of course the claim that behaviorally undetectable spectrum inversion is possible that has raised verificationist objections. It is also this claim that has been the target of the empirical objection that our color quality space is asymmetrical in ways that preclude a mapping of determinate shades of color onto other determinate shades of color in a way that preserves the similarity relationships between different shades, preserves boundaries between color categories, and maps unique hues onto unique hues. For some purposes the empirical objection can be finessed by construing the claim about behaviorally undetectable spectrum inversion as saying, not that there could be such inversion in the case of creatures whose color quality space is like our own, but that there could be creatures who perceive colors and whose color quality space is such that for them such inversion is possible. But for my present purposes it is unnecessary for me to claim even this; the claim that behaviorally undetectable inversion is possible plays no role in my argument.. All that I need claim is that it is possible for the visual experiences different creatures have of objects of certain colors to differ in their phenomenal character without any of these creatures thereby misperceiving the colors, and for the visual experiences of two creatures to be alike in phenomenal character when they are viewing things of different colors, again without either of the creatures misperceiving the colors.
If there were creatures whose color qualia were “alien” relative to ours, so that no color experience of theirs was phenomenally like any color experience of ours, and if these creatures were equivalent to us in their ability to perceive colors, then the relation of their color experience to ours would realize the first of the possibilities but not the second. This would be a case of “alien qualia” but not “inverted qualia.”  What is true if there can be either alien color qualia or inverted color qualia is that no phenomenal character is such that an experience’s having that phenomenal character is a necessary condition of its representing a certain color. What additionally is true if there can be inverted color qualia is that no phenomenal character is such that an experience’s having that phenomenal character is sufficient for its representing a certain color.
Anyone who makes the inverted qualia and alien qualia claims about color experience will presumably make the corresponding claims about the perception of other “secondary qualities” – sounds, smells, tastes, and tactile properties like warmth and coldness. Such claims have been made as long as the perception of such properties has been discussed. They are motivated in part by the idea that the phenomenal character of perceptual experiences should be a function not only of the objective states of affairs being perceived but of the nature of the perceptual system of the perceiver. That idea, taken as completely general, would imply that such claims should be made about the perception of “primary qualities” as well. But in the case of secondary qualities the idea derives much of its plausibility from the fact that there seems to be an “explanatory gap” between the phenomenal character of the experiences and what we know independently of this character about the objective properties of the things that cause them. If we ask why something with a certain molecular structure should taste the way bitter things taste to us, or or why something with a certain surface spectral reflectance should look the way blue things look to us, it seems clear that at least part of the answer must have to do with the nature of the perceptual system involved in our perception of these properties. It may indeed seem that an explanatory gap remains even when we bring in the nature of the perceiver. But this remaining gap is an aspect of the mind-body problem, having to do with the relation between states of the brain and the phenomenal character of experiences, and is one that we cannot even formulate without assuming that the phenomenal character of our experiences of things having certain objective properties is fixed by physical states of ours that are caused by things having those properties, and is only contingently related to those properties.
This assumption is supported by what we know about the mechanisms involved in the perception of such properties. We know that what shades of color are perceived as “unique hues” is determined by the rates at which lights of various wavelengths are absorbed by pigments in the cones on the retina, and that because the amount of these pigments varies slightly from one individual to another there are individual differences among normal observers as to what shades are perceived as unique hues – e.g., as to whether something is a unique green or a slightly bluish green.  There is no good reason to say that some of these observers are getting it right and the others are misperceiving. So this can count as a case of behaviorally detectable spectrum inversion. We know that because of the way our perceptual system operates many different combinations of wavelengths are visually indistinguishable (such indistinguishable combinations of wavelengths are called “metamers”), and thus that what we perceive as sameness of color and difference of color is determined in part by the nature of our visual system. If whether things look the same or different with respect to color is determined in part by the nature of our visual system, it would seem that how things look must be determined in part by the nature of our visual system.
If the phenomenal character of perceptual experiences consisted in their representational content, and if this content consisted entirely in the representation of “objective” properties, then experiences that have the same objective representational content would be alike in phenomenal character, whether the perceptual mechanisms of the subjects of these experiences were the same or different. In that case inverted qualia and absent qualia would be impossible. The way I reconcile the claim that inverted qualia and absent qualia are possible with representationalism about phenomenal character is by holding that among the properties represented by perceptual experiences are what I call “phenomenal properties,” and that the phenomenal content of experiences consists in their representation of these. These properties are perceiver relative, in such a way that perceivers can differ in what phenomenal properties they are perceiving the same objects to have while veridically perceiving these objects to have the same objective properties, and in such a way that different perceivers can, when veridically perceiving different objects, perceive the same phenomenal properties even though the objective properties they perceive are different. The idea is that we perceive the colors and other secondary qualities of things by perceiving phenomenal properties that are associated with them, and that individuals can differ with respect to what phenomenal properties are associated in their experience with the same objective secondary properties. Originally I held that phenomenal properties are relational properties that things have only when perceived – so, for example, the “color-like” phenomenal property I perceive when I see something red is a property something has just in case it produces an experience of a certain sort in the creature perceiving it.  More recently I have suggested that phenomenal properties are instead dispositional properties of a certain kind – dispositions things have to produce experiences of certain sorts in one or more sorts of creatures. The role of qualia in the account is to determine the types of experiences in terms of which the phenomenal properties are individuated, and, what goes with this, to bestow on experiences their phenomenal character, i.e., the part of their content that consists in the representation of phenomenal properties.
One needs such an account if one is to combine representationalism about the phenomenal character of experiences with the view that qualia inversion is possible – more generally, with the view that to some extent the phenomenal character of experiences and their objective representational content can vary independently of one another. If phenomenal character is representational content, phenomenal character can vary independently of objective representational content only if there is another sort of representational content that can vary independently of objective representational content.
There is a view about what the phenomenal properties are that makes them aptly called “subjective,” and makes the experiential contents in which they are represented aptly called subjective representational content. This is the projectivist view that they are in fact properties of our experiences themselves, which the experiences falsely represent as instantiated in objects of outer perception – so one’s experiences (correctly) represent the tomato as red by (falsely) representing it as having a certain property that is in fact instantiated in the experience itself and is never instantiated except in experiences. This sort of error theory I am anxious to avoid. On my view, the phenomenal properties, although relational and individuated with respect to kinds of experiences, really do belong to the external things in which we perceive them as being instantiated. So while it seems right to say that the representation of these is not part of the objective representational content of experiences, I am reluctant to speak of it as subjective representational content.
I said that one needs a view of this sort if one is to reconcile representationalism about experience with the possibility of spectrum inversion and the like. So if one already accepts both representationalism and the possibility of spectrum inversion, one has a reason to hold such a view. But it seems offhand that if one already accepts the possibility of spectrum inversion, there is a consideration that gives one a reason to hold this view even if one does not already accept representationalism – and which by giving one a way of reconciling the possibility of spectrum inversion with representationalism, gives one a reason to hold representationalism. It is this I want to look at next.
Earlier I characterized the possibility of qualia inversion as the possibility that the phenomenal character of experience and the objective representational content of experience can, to some extent, vary independently of one another, without this resulting in misperception. So, for example, one might characterize a case of spectrum inversion as one in which one person’s experiences of red are phenomenally like another person’s experiences of green, and vice versa. This should not be put by saying that to one of the people red things look red and green things look green, while to the other red things look green and green things look red. For that of course would imply that one of them systematically misperceives the colors of things. But there is a way of putting this in terms of how things “look” which is not defective in this way – one can say that the way red things look to one of them is the way green things look to the other, and the way green things look to the first is the way red things look to the second. This formulation has the advantage that it does not employ notions – such as those of an experience, of the phenomenal character of experiences, and of the objective representational content of experiences – which might be held to be loaded with questionable philosophical theory. But this way of putting the possibility can seem to lead immediately to the conclusion that the possibility of spectrum inversion requires that our experiences represent phenomenal properties that are distinct from colors. If red and green things look different to the two people, and yet red things look red to both and green things look green to both, and they don’t differ in what other objective properties the things look to them to have, it would seem that there must be a kind of properties other than these colors, and other than any kind of objective properties, such that one of the persons perceives red things to have one property of this kind and green things another, while the other perceives green things to have the first of these properties and red things to have the second. The guiding idea here is that if something looks a certain way to a person, there is, corresponding to that “certain way,” a certain property that thing looks to the person to have. It is this idea I shall be examining and developing in the remainder of this section and in the one that follows.
If this idea is right, it would seem that we can get to phenomenal properties without invoking anything as controversial as the claim that spectrum inversion is possible. For it is absolutely commonplace for things to look different (either to the same person, or to different persons) without there being any misperception and without there being any difference in the objective representational content of the experiences of them. Consider a case in which I look at a table surface that is partly in shadow, or on which there is a highlight. Different parts of the surface will look different to me. There is no misperception here – indeed, given the circumstances, only a malfunctioning of the visual system could result in the parts of the surface not looking different. Supposing that the table surface is in fact uniform in color, and that I am not misperceiving, this is not a case in which the objective representational content of my experience of one part of the surface differs from the objective representational content of my experience of another part of the surface. This will be a case of “color constancy”; the part in shadow, although it looks different from the part not in shadow, does not look to have a different color. Of course, it may look to me as if the one part has, while the other lacks, the objective property of being in shadow. But it would get things backwards if we said that it is because of this that the two parts look different to me. Things can look different to one in the way these parts do when one is in doubt whether this is due to a difference in color or a difference in illumination, and if in the present case one part looks to be in shadow and the other doesn’t, this is partly because they look different in the way they do (and partly because of clues about illumination provided by the context). If their looking different to me in this way consists in there being different properties they appear to have, these won’t be objective properties. They will candidates for being phenomenal properties in my sense of the term.
What properties might these be? One sort of property that is certainly instantiated in such case is the property something has just in case it is currently looking a certain way to someone. I would construe this as the property of causing, in a certain way, an experience of a certain sort. Call such a property an “occurrent appearance property.” This is the sort of property I first took phenomenal properties to be. But a thing’s having such a property is bound to be an exercise of a dispositional property it has – the property it has of being apt to produce experiences of a certain sort in some kinds of observers when those observers are related to it in a certain way. Call such a property a “dispositional appearance property.” This is the sort of property I took phenomenal properties to be in a more recent discussion.
It should be noted that any object will have a vast number of dispositional properties of the sort indicated – even if we put aside the possibility of spectrum inversion, which entails that the object may be apt to appear differently to different observers because of differences in their perceptual systems. For it will be disposed to appear in different ways depending on the illumination conditions, the distance of the observer from it, the way in which the observer is oriented relative to it, and so forth. 
In earlier discussions I have written as if for any sort of observer, each shade of color is associated with a single phenomenal property, in such a way that for an observer of that sort perceiving something to be of that shade of color involves perceiving it to have that phenomenal property. It is clear that if phenomenal properties are any of the “appearance” properties just mentioned, this is a gross oversimplification. Plainly, for any shade of color there will be a large number of different occurrent appearance properties such that, given the right circumstances, perceiving a thing having one of those appearance properties will be sufficient for perceiving it to have that shade. The same is true of the dispositional appearance properties. As noted, an object will have a large number of such properties, and a perceiver will perceive one of these only if she is the right sort of observer and is related to the object in the right sort of way and in the right sort of circumstances. The exercise of these dispositions will always be an instantiation of an occurrent appearance property. And associated with a given shade of color there will be at least as many dispositional properties of this sort as there are occurrent appearance properties associated with it. In fact, there are bound to be more, since sometimes it is one and the same occurrent appearance property that is instantiated in the exercise of the dispositions associated with a number of different dispositional appearance properties.
Among the occurrent appearance properties associated with a color, some will have an especially close relationship to it – relative to a certain sort of observer. It is one or another of these that a thing will have when viewed (by an observer of the relevant sort) under normal or standard viewing conditions. When one speaks of something as looking a particular shade of blue, one is likely to have in mind a way of appearing that belongs to this privileged subset. Call this a “canonical” way of appearing associated with blue. But expressions like “looks blue” seem to be ambiguous. In saying that something looks blue, one may be indicating that if one took one’s experience at face value one would judge that the thing is blue. Following Fred Dretske, call this the “doxastic” sense of “looks blue.” But saying that something looks blue in this sense is in a certain way unspecific about how it looks. It is unspecific about what occurrent appearance property it has, for the statement could be true even if the way the thing looks is not a canonical way for blue things to look – it is enough if the way the thing looks is such that, given the circumstances, the thing’s looking that way indicates that it is blue (and the perceiver is sensitive to this fact). In a different sense of “looks blue,” call it (again following Dretske) the “phenomenal” sense, saying that something looks blue is saying that the way it looks belongs to the privileged subset, i.e., is a canonical way for blue things to look (relative to observers like oneself). And one could say that something looks blue in this sense, looksp blue, even if taking one’s experience at face value would not lead one to judge that it is blue. Under certain circumstances snow looksp blue (or at any rate bluish). This is not a case of misperception, and it is not a case in which one is disposed to think that the snow is blue. If there is a highlight on the table, one might say that part of its surface looksp white – but one is not thereby judging, or even disposed to judge, that it is white, and one also will not think one is misperceiving. On the other hand, it is quite common for white things not to lookp white, i.e., not look the way white things look in normal or standard viewing conditions, even though they do look white in the doxastic sense (do lookd white). Roughly speaking, if F is a color then something looks F in the phenomenal sense if a painter would use F pigment to represent how it looks.
We see here that there is an understanding of “ways of looking” such that a thing’s looking a certain way (namely lookingp that way) is not in and of itself a matter of its appearing to have certain objective properties. Of course, things appearing to have certain objective properties (appearingd certain ways) is grounded in their looking phenomenally (lookingp) certain ways. But the transition from the ways things appear phenomenally to what objective properties they have, and what objective properties they appear to have, rests on certain contingencies. These include facts about illumination conditions and the spatial relation of the perceiver to the object, and facts about how these combine with the objective properties of a thing to determine how it will lookp. The observer needn’t have explicit knowledge of these facts, but she must show an appropriate sensitivity to them. And there is a further contingency: how things with certain objective properties appearp, given illumination conditions and the relation of the observer to the object, depends on what sort of visual system the observer has. This is the point that is dramatized by inverted qualia and alien qualia scenarios.
If the way something looksp to one does not itself amount to one’s experience representing the thing as having certain objective properties, does it amount to representing the thing as having any properties at all? In particular, does it amount to representing the instantiation of any of the properties mentioned earlier as candidates for being phenomenal properties? That there are what I have called occurrent appearance properties and dispositional appearance properties seems to me beyond question. The question is whether any of these can be properties our experiences represent. For unless some of them are, we cannot characterize the phenomenal character of experience in the way I proposed earlier – namely by saying that an experience’s having a certain phenomenal character is a matter of its representing something as having certain phenomenal properties.
If one has a perceptual experience in which one is “appearedp-to” in a certain way, an experience “as of” something appearingp a certain way, then unless one is hallucinating there is something that does appearp to one that way – something that has the occurrent appearance property of so appearing. For such an experience to be veridical, there must be something that has that property, and if there is something having that property, the experience is to that extent veridical. This does not of course exclude its failing to be veridical in other ways. One’s first thought is that it fails to be veridical if the thing fails to be the way it appears to be – e.g., if it looks blue but is not blue. But remember the different senses of “looks blue” distinguished earlier. If something appearsd blue, appears blue in the doxastic sense, then its not being blue would be a failure of veridicality. But if the thing appearsp blue, appears blue in the phenomenal sense, then its not being blue is not sufficient for its failing to be veridical – for it may be that while it is not blue, the circumstances are such that something with its intrinsic properties should lookp blue to an observer like oneself. So the requirement of veridicality could be put as follows: for an experience of a thing to be veridical, the way the thing appearsp must be such that a thing’s appearingp that way to an observer of that sort in those circumstances would not count as misperception. And this amounts to saying that the thing must have a dispositional appearance property of a certain sort. It must be such that in circumstances of this sort it appearsp this way, or can appearp this way, to an observer of this sort with a normally functioning perceptual system who is related to it as this observer is related to it. 
So the veridicality of an experience as of something that appearsp a certain way requires that there be something that has the occurrent appearance property of appearingp that way and also the dispositional appearance property of appearingp that way to observers of a certain kind who view it under certain conditions. This seems a reason for saying that such an experience represents the instantiation of properties of both sorts. It is of course not true in general that if something’s having a certain property is a necessary condition of an experience’s being veridical then that property is represented by the experience. If an experience is of a glass of water, it is a necessary condition of its being veridical that the glass in front of the subject contains hydrogen atoms, but we would not want to say that the experience represents the property of containing hydrogen atoms. But whereas the experience of the glass of water does not, by itself, put the subject in a position to judge that there are hydrogen atoms in front of him, the experience in which something looksp blue does put the subject in a position to judge both that the thing has the occurrent appearance property of lookingp blue and that it has the, or rather a, dispositional property of lookingp blue. And that makes it plausible to say that these are properties of the thing thing that are represented.
There is a difficulty with this that is indicated by my shift from “the” to “a” a couple of sentences back. Presumably the dispositional appearance property a subject perceives on a particular occasion should be one whose associated disposition is exercised on that occasion by the instantiation of an occurrent appearance property. But when something has the occurrent appearance property of appearingp F, there will be any number of different dispositional appearance properties of which this could be the manifestation, and the nature of the experience will not in general indicate of which of these it is in fact the manifestation – for the nature of the experience will not reveal all of the relevant details about the circumstances under which the thing is observed. So the experience will not put the observer in a position to judge that the observed thing has any specific appearance dispositional property – and that seems a reason for saying that it does not represent any such property.
But for any way a thing can appear, there will be the higher order property shared by all things that are disposed, under some circumstances or other, to appear that way to normal observers of one or more sorts situated in one or another way with respect to them. This will be the higher order property of having one or another of the dispositional appearance properties which can manifest themselves in an instantiation of a given occurrent appearance property, and will be a property that a perceived thing will have just in case it has that occurrent appearance property. Call this a higher-order dispositional appearance property. If one has an experience as of something that appearsp F, it is a condition of the experience being veridical that the thing have the higher-order dispositional appearance property of appearingp F. And having such an experience does put one in a position to judge that the object perceived does have that property. There seems a good case for saying that such properties are represented by experiences. And these, and the associated occurrent appearance properties, seem good candidates for being the phenomenal properties my account requires.
But suppose we deny that any of these appearance properties are perceived and represented in our perceptual experiences. It will still be true that it is because things appearp to us in certain ways that we perceive them as having certain properties. And it will still be true that the nature of our experiences, and what we are introspectively aware of in having them, will reflect the ways things appear to us. If anything deserves to be called the phenomenal character of our experiences, it is the part of their introspectable nature that reflects how things look, feel, taste, smell, or sound to us. If this is part of the representational content of our experiences, the appearance properties I have mentioned seem the only candidates for being the properties that are represented. And if this is not part of the representational content of our experience, then representationalism about phenomenal character is false.
But how could this not be part of the representational content of our experience? It could be so only if what we are introspectively aware of here, in being aware of the look, feel, etc. of things, are features of our experiences that are not themselves representational. But how could awareness of intrinsic, non-representational features of experience constitute awareness of the look, feel, etc. of things? One view would be that the features we are introspectively aware of are related to the appearance of things in something like the way the paint we see on a canvas is related to what the painting represents – that our awareness of how things appear is grounded on our awareness of non-representational features of the experience in something like the way our perceptual awareness of what the painting represents is grounded on our perceptual awareness of the lines, shapes and colors on the canvas. This view seems to me false to the phenomenology, and to be avoided for all the reasons that have led to the demise of the sense-datum theory.
But my aim in the present paper is not so much to argue for representationalism about phenomenal content as to argue, first, that if representationalism is to be acceptable it must allow that the properties whose representation bestows phenomenal character are phenomenal properties in my sense, which I now claim to be appearance properties of the sort I have characterized, and, second, that in allowing this a representationalist must allow something that representationalists have usually been unwilling to allow – that the phenomenal character of experiences is independent of their objective representational content (their representation of objective properties) in a way that allows for the possibility of cases of inverted qualia and alien qualia. To complete the case for this I need to say more about how qualia figure in my account. But before I do this I want to say something about how the account applies to the case of pains and other bodily sensations.
One of my claims about introspective awareness is that it is fact awareness unmediated by thing awareness – awareness that which does not involve awareness of any object. One source of resistance to this is the awareness we have in cases of after-imaging and the like. It is certainly very natural to say that my awareness that I see a yellowish-orange after image is grounded on an awareness of an object, namely the after-image. Here I follow the lead of J.J.C. Smart, in his seminal paper “Sensations and Brain Processes”; I affirm the existence of “experiences of after-imaging” while denying the existence of after-images. This is the rejection of the “act-object” conception of sensation. In my view, to hold that “seeing a yellowish-orange after-image” involves the existence in one’s mind of something that is yellowish-orange is to hold a view that can be adhered to consistently only by someone who embraces a sense-datum theory of perception. So I hold that we should not “reify” after-images.
But it is one thing to hold that we should not reify after-images, and another thing to say that we should not reify pains. A more serious obstacle to holding that introspective awareness is fact awareness unmediated by thing awareness is the overwhelming naturalness of saying that pains, itches, tingles, etc. are particular items of which we are introspectively aware, and that it is by being aware of these items – these “objects” – that one is aware of such facts as that one has a pain in one’s foot. It is common to take pains and the like as paradigms of sensations, and sensations as paradigms of experiences – and then it is no wonder that we fall into thinking of experiences as objects that we are aware of in a quasi-perceptual way when we introspect.
This way of thinking is extremely natural, but I think that it is confused. I think that when, as we say, a person has a pain in his foot and feels it, there are two sorts of awareness occurring that need to be distinguished, and which ordinary ways of talking encourage us to conflate. There is perceptual awareness of the condition of some part of one’s body. This involves awareness of the instantiation of a phenomenal property – in the sense discussed above – in that part of one’s body. By perceiving the instantiation of such a property one may also perceive that there is damage of some sort in that part of one’s body – no doubt the biological function of pain and pain awareness intimately involves the perception of bodily damage. Here one can properly be said to be aware of some particular, a region of one’s body, as having one or more properties. This awareness is not simply awareness that; it is, if you like, awareness of an object. But also, this is not introspective awareness; it is perceptual awareness. Now, having this perceptual awareness involves being in a perceptual state, a somatic perceptual state, which represents some part of one’s body as being a certain way. And of this state one can, and normally will, have introspective awareness. This awareness will be like the introspective awareness one has of one’s visual, auditory, tactile, etc. sense-experiences. It is best thought of as awareness that – awareness that one is in a certain somatic perceptual state. The content of this awareness will embed the content of the perceptual state, in the same way that the content of an introspective belief about visual experience will embed the content of the visual experience. The somatic experience is no more in the part of one’s body it represents than the visual experience is in the portion of one’s visual field it represents. And it no more has the phenomenal property it represents as instantiated in a certain part of one’s body than the visual experience has the phenomenal property it represents as instantiated in some object in front of one – at any rate, this is so unless some projectivist view of phenomenal properties is correct. But in the same sense in which the visual experience has a phenomenal character, the somatic experience has a phenomenal character; its phenomenal character is fixed by what phenomenal properties it represents.
Supposing this account is right, how do we map it onto our ordinary talk of pains and itches, and of feeling pains and itches? I think there is no neat way to do so. We speak of pains and itches as located in parts of our bodies. And we speak of them as felt, which seems to make them objects of some kind of perception. But we also speak of them as mental entities. And the instantiation in some part of one’s body of a phenomenal property that can be felt is no more a mental state of affairs than is an apple’s having a phenomenal property that is detectable by sight. The somatic experience of pain is mental, and to that extent is a better candidate for being pain than the phenomenal property instantiation. And it seems to be what we are averse to, which also makes it a better candidate. We are as averse to “hallucinatory” somatic experiences of this sort as we are to veridical ones. But the somatic experiences are not located in the parts of our bodies where we are said to feel pains – if they are located anywhere, they are located in the brain. And the somatic experiences are not felt, just as visual experiences are not seen. So nothing there is seems an ideal candidate for being pain as we ordinarily talk about it. Similarly for itches and tingles. This doesn’t prevent there being quite definite truth conditions for statements about pains, or about itches and tingles. A person counts as being in pain if she has a somatic experience having a certain phenomenal character, i.e., representing the instantiation of a certain phenomenal property at some point in the person’s body. And someone counts as having a pain in a certain location if she has a somatic experience which represents an instantiation of that phenomenal property in that location – even if, as in the case of “referred pains,” the organic cause of the pain is in some other part of the body. Because these truth conditions concern the somatic experiences, these may seem the best candidates for being pains. But if we take them to be the pains, we must allow that it is only in what G.E. Moore called a Pickwickean sense that pains have location, and are felt. And we must be careful not to confuse the introspective awareness of pains so understood, i.e., the introspective awareness of this class of somatic sense experiences, with the perceptual awareness that normally accompanies it, viz the perception of conditions of one’s body.
The confusion we are apt to fall into here has a parallel in the case of other sorts of sensory experience. It is certainly not unheard of for philosophers to confuse our introspective awareness of visual sense experiences with perceptual awareness of phenomenal properties represented in such experience. This happens in sense-datum accounts. What probably makes the confusion harder to resist in the case of bodily sensations is a fact about our interests. Whereas in the case of other sorts of perception our primary interest is not in experiences but in the things they represent, in the case of pains and itches we have at least as direct an interest in the experiences themselves, namely in avoiding or eliminating them. Being in the habit of taking the items of primary interest to be objects of perception, as they are in other cases, we are prone to do this here, thereby blurring the distinction between perceptual experiences and item perceived. So we are apt to accept a sense-datum view about bodily sensation even when we have abandoned it elsewhere. But, as I suggested earlier, taking bodily sensations as our paradigms of experiences can lead one to think that even in the case of vision and hearing our experiences are objects which we can be introspectively aware of as having certain properties – and this is a sort of residue of the sense-datum view. At any rate, that is how I would diagnose my own past thinking about these matters.
I said above that the instantiation in some part of one’s body of a phenomenal property, e.g., in a case of pain, is “no more a mental state of affairs than is an apple’s having a phenomenal property that is detectable by sight.” But does this mean that it is not a mental state of affairs at all? Recall that my earlier discussion settled on two sorts of properties as the best candidates for being phenomenal properties – higher order dispositional appearance properties, and the occurrent appearance properties in which these are manifested. I see no reason not to count both of these as phenomenal properties. The instantiation of a phenomenal property of the first sort is a “mental” state of affairs only in the very thin sense that the instantiation of such properties requires that there be minded creatures capable of perceiving such instantiations. But the instantiation of a phenomenal property of the second sort, an occurrent appearance property, is a mental state of affairs in a much more robust sense – for it requires that the thing having it be actually perceived.
So it is perhaps not out of the question that pains can be held to be both located where we say they are and mental – or at least partly mental. They will be so if they are instantiations of phenomenal properties of the second sort, occurrent appearance properties. Such a view would not save everything we want to say about pain. It would allow for the possibility of pain hallucinations, and would imply, contrary to ordinary ways of speaking, that a “referred pain in the arm” is not really in the arm. But then, neither does the view that pains are somatic experiences of (i.e., representing) such property instantiations save everything we want to say about pain.
But I think it is useful to reflect on the fact that insofar as our perceptual experiences represent instantiations of occurrent appearance properties, they represent states of affairs that are in part mental. For this raises a question about the relation between perceptual awareness and the introspective awareness of perceptual experiences. If the properties represented by perceptual experiences were exclusively objective properties, ones whose instantiation was in no way a state of affairs that involves the existence of minds, then the contents of perceptual awarenesses and introspective awarenesses would be logically independent of each other. But if the properties represented by a perceptual experience include occurrent appearance properties, then part of the content of the perceptual experience seems closely related to the content of an introspective awareness the subject might have. Let it be that I am perceptually aware that something has the occurrent appearance property of lookingp blue, and am introspectively aware that I am “appearedp blue to,” i.e., that I am having an experience as of something that looksp blue. The relation of “That looksp blue to me” to “I am having an experience as of something that looksp blue” seems to be one of conceptual entailment. So the introspective awareness expressed by the second proposition seems implicit in the perceptual awareness expressed by the first.
Now of course one can have a perceptual experience that represents that a thing has a certain property without judging that the thing has that property. This can happen when one’s attention is not on the thing in question, or when the experience of it is just a momentary glimpse. And it can happen when the property is an occurrent appearance property. So it would be wrong to say that whenever one’s experience represents such a property, one is in a position to make the introspective judgment that it does. What seems to be true, however, is that whenever one is in a position to make the perceptual judgment that something has such a property, one is in a position to make the introspective judgment that one’s experience represents the property.
In earlier discussions of phenomenal properties I have insisted that while such properties are in fact relational, we are not normally aware of them as relational. I now want to qualify this. It is true, I think, of the case where one perceives something to have a phenomenal property but does not judge that the thing has it. But the situation is more complicated in the case where one does judge that a thing has a phenomenal property. One thing that remains true is that while the property is relational, only one of the relata, namely the external object, is perceived – so perceiving the property is not a matter of perceiving a relation between two things. And in some cases the judgment one makes is not relational in form; one judges simply “It looks blue.” It may not occur to the person making this judgment that the thing’s looking blue is a matter of its looking blue to her. For one thing, the person very likely is not distinguishing the occurrent appearance property of lookingp blue and the dispositional appearance property of lookingp blue, and of course the thing’s having the latter is not a matter of its standing in a relation to the subject. However, insofar as the subject is ascribing to the thing the occurrent appearance property of lookingp blue, nothing more than reflection on what having this property amounts to should be needed for the person to realize that “looks blue” here means “looks blue to me,” and that the property is relational. And, as suggested above, nothing more than reflection on the content of the latter realization should be needed for the person to realize that she is “appeared blue to,” i.e., has an experience that represents something as lookingp blue.
This gives us one way of cashing out the claim that the phenomenal character of perceptual experiences is “self-intimating.” It is obvious that insofar as perception is a source of knowledge of the environment, it requires the subject to have at least a sensitivity, which may be only subpersonal, to the contents of perceptual experiences -- only so can these give rise to perceptual beliefs with corresponding contents. If it is further the case that perceptual experiences must represent occurrent appearance properties, this sensitivity extends to the representation of these. It would be rash to claim that in general this sensitivity amounts to introspective awareness. But in the case of the representation of occurrent appearance properties, it is arguable that it grounds such awareness. Insofar as experiences representing such properties put one in a position to judge that such properties are instantiated in perceived objects, they put one in a position to judge that they are represented in one’s experience – assuming, of course, that one has the concepts such judgment requires.
As I said earlier, my version of representationalism about phenomenal character differs from other versions in assigning a central role to qualia. I must now try to make clear what that role is.
There is a way of putting my view which on first pass does not involve introducing qualia, and seems to be in line with ordinary ways of thinking. We start by saying that we are perceptually aware of the way things appear, and that this involves being aware of the instantiation of appearance properties, both occurrent and dispositional, where the instantiation of the occurrent ones are manifestations of the [[dispositions that are the]] dispositional ones. The instantiations of the occurrent ones involve the occurrence of states of being-appeared-to on the part of observers, and can be construed as the causing of such states by observed objects; so the dispositional ones can be thought of as dispositions to produce such states under certain circumstances. The states of being-appeared-to are states that represent the appearance properties that cause them or are disposed to cause them.
Here it looks as though the appearance properties are individuated in terms of the states of being appeared to, and that the ways of being appeared to are individuated in terms of the appearance properties. How do we break out of this seeming circle? Could we take the appearance properties as primitive and define the states of being appeared to in terms of them? That obviously will not do. It would not enable us to make sense of the fact that differently constituted observers observe different appearance properties when they perceive the same objects – that what appearance properties a creatures observes when viewing a thing is a function of the nature of the perceiver as well as the nature of the thing. Could we take the states of being appeared to as primitive, and define the appearance properties in terms of them? This won’t do, given that the states of being appeared to are states that have to be specified in terms of their representational contents, namely in terms of what appearance properties they represent.
Could there be a “package deal” definition, say one involving the Ramsey-Lewis technique, which simultaneously defines both the appearance properties and the states of being appeared to? As a first stab we might try to define a particular state/property pair by saying that the state of being appearedp blue to and the property of appearingp blue are the unique state S and the unique property P, such that S represents P and something has P just in case it is producing (or has the disposition to produce in some sort of observers) a state of type S, and in the case of creatures like us something is P only if it looks the way blue things look under optimal conditions. This does not, as it stands, tell us how to apply “is appearedp blue to” to creatures of kinds other than our own. And, what goes with this, it does not provide a way of distinguishing the appearance property in question from the property that realizes it in the case of observers like ourselves – it doesn’t distinguish the dispositional appearance property from the intrinsic surface property that is its categorical base in the case of observers like ourselves, and doesn’t distinguish the occurrent appearance property from the property something has when it is instantiating that intrinsic surface property. It should be noted that it invokes the notion of “looking the way…,” i.e., the notion of sameness of appearance properties. We would need to be able to apply this notion intrasubjectively in order to apply the definition at all. And we would need to be able to apply this notion intersubjectively in order to extend the definition so as to apply to creatures in whose experience the phenomenal property in question is associated with a different color. Supposing we could do this, we could revise the account as follows: The state of being appearedp blue to and the state of appearingp blue are the unique state S and the unique property P such that S represents P, someone is in state S just in case the way something looks to him is the way blue things look to us under optimal conditions, and something is P just in case it is producing state S in some observer (in a way that does not involve misperception), or disposed to produce S in some sort of observer. This is still rough and in need of refinement. But plainly any account of these properties and states needs an account of sameness of state of being appeared to.
Here is where qualia come in. We need a kind of sameness, applicable to states of being appeared to, that is not simply defined as sameness with respect to what property is represented. The only thing that will do is sameness with respect to qualitative character, which I think has to be a functionally defined sort of sameness. Qualia will be the properties of perceptual states in virtue of which they stand in relations of qualitative similarity and difference. In the first instance, the relations of qualitative similarity and difference will be defined in terms of the role they play when they hold intrasubjectively. An important part of this role has to do with what sorts of perceptual beliefs experiences related by these relations are apt to cause, and this is of course intimately related to their role in affecting the discriminatory and recognitional behavior of the creature. Discrimination will require qualitatively different experiences, and recognition will require qualitative similarity between experiences occurring in the same subject at different times. Different experiences of the same objective property may be qualitatively different; but once illumination conditions and the relation of the perceiver to the object are fixed, sameness of objective property represented will go with qualitative sameness. And, of course, sameness of phenomenal property represented will go with qualitative similarity, independently of viewing conditions.
I believe that in order to apply the notion of qualitative similarity and difference intersubjectively one must (assuming physicalism) have recourse to the notion of the physical realization of qualia. Supposing that qualia are multiply realizable, each quale will be associated with a set of properties that can be said to be realizers of it in virtue of the way instantiations of them in the same subject are related – i.e., are such that experiences in which different ones of them are instantiated, but are otherwise the same (at the appropriate functional level of description), are qualitatively alike. If two creatures are alike with respect to what physical properties of their brain states serve as realizers of the qualia involved in their experiences, then an experience of one of them is qualitatively similar to an experience of the other to the extent that the qualia realizers they instantiate are ones whose instantiation would yield qualitatively similar experiences if instantiated in the same subject. While it is the intrasubjective functional role of a property that makes it a quale realizer, its being a realizer means that it realizes the same quale in whatever creature it is instantiated in – so if the same realizer is instantiated in the experiences of two different subjects, or two different qualia realizers are instantiated in them but these are ones that realize the same quale when instantiated in the same subject, the experiences of the two subjects will be to that extent the same.
Qualia can be thought of as the vehicles of the representation of properties of perceived objects. As such, they can be said to represent such properties. In the case of objective properties, it will be only contingently true that a given quale will represent a particular property – e.g., that a particular quale represents red (or red as viewed under certain conditions). Any view which holds that perceptual experiences represent what they do in virtue of causal correlations between types of experiences and features of the environment, or in virtue of an evolutionary history that bestowed on certain types of experiences the function of indicating certain features of the environment, requires a way of typing experiences that makes it a contingent fact that experiences of a given type represent the particular features of the environment they do. It may be thought that all this requires is that the experiences belong to physical types, which physicalist representationalists who reject qualia can readily allow. But I think it is clear that the typing must be functional rather than physical. The types must be so related to each other, and to the rest of the subject’s psychology, as to determine its “quality space,” and so must play an appropriate role in determining its discriminatory and recognitional capacities, and in generating beliefs about its environment. What must be true of different tokens of the same type is that they share a certain causal role; this allows for “multiple realization,” and so for the possibility that tokens of an experiential type might be physically heterogeneous. Spelling out the conditions for type membership requires a functional account of qualitative similarity and difference, and qualia are the properties of sensory states, presumably physically realizable properties, in virtue of which they stand in these relationships. Thus it is that it is qualia that type the experiences, and thus it is that qualia contingently represent the environmental features they do.
But it is necessary, not contingent, that qualia represent the phenomenal properties they do. Phenomenal properties are individuated in terms of the types of experiences they produce, and these experiences are typed by their qualitative character. The qualitative character of an experience, what qualia are instantiated in it, thus fixes its phenomenal character, i.e., what phenomenal properties it represents.
As a representationalist, I hold that our introspective awareness of the phenomenal character of our experience is an awareness of an aspect of their representational content, namely their representation of phenomenal properties. In earlier discussions of this I have maintained that this awareness should not be construed as awareness of qualia, and that qualia are known “only by description.” I now think that this was a mistake.
One source of my former view was the idea that I would be involved in vicious circularity if I both typed experiences by their phenomenal character, i.e., what phenomenal properties they represent, and defined phenomenal properties in terms of what types of experiences they produce – and that typing the experiences in terms of qualia would not avoid the circularity, if the qualitative character of experience and the phenomenal character of experience turned out to be the same. But as I think the discussion in the earlier part of this section shows, what is needed in order to avoid circularity is a “package deal” account which defines phenomenal properties and phenomenal character together in a way that essentially involves a functional account of the sameness of appearance properties and ways of being appeared to – or what comes to the same thing, a functional account of qualitative similarity and difference. And it is compatible with this that qualia, thought of as the properties in virtue of which experiences are qualitatively similar and different, should turn out to be the same as the properties of experiences which are essentially representative of phenomenal properties, and so the properties of experiences we are introspectively aware of when we are aware of that part of the representational content of our experiences which is the representation of phenomenal properties.
No doubt another source of my former view was my taking for granted that qualia are “intrinsic” and non-representational features of experiences – that being, of course, a common way of characterizing them. It is, I think, less than clear what it means to speak of intrinsic features of experiences, given that experiences themselves are states of persons. But it is part of my view that having an experience with a certain qualitative character – one having certain qualia – is an intrinsic feature of a person. If the only properties of things represented by perceptual experiences were objective properties, and if, as I think, we need an externalist account of the representation of objective properties, it would follow that qualia are non-representational – that while they can be “vehicles” of representation, they can be only contingently the vehicles of the representation of particular objective properties. But of course I hold that in addition to representing objective properties, our experiences represent phenomenal properties – occurrent and dispositional appearance properties. About the representation of these my view is internalist rather than externalist. And it is compatible with this that qualia, although internally determined, are essentially representative of such properties.
So I now think that in being introspectively aware of the phenomenal character of one’s experience, being aware of how one is appearedp-to, one can be said to be introspectively aware of the qualitative character of the experience, and of the qualia that make up that character. But I should emphasize again that this awareness should not be thought of as a matter of introspectively singling out an experience and noticing that it has certain features. For there is, I think, no such thing as introspectively singling out an experience and noticing something about it. Experiences are states of persons, not quasi-substances to which their subjects have quasi-perceptual access. What one does have in introspective awareness is awareness that one has an experience with a certain phenomenal character, which is an awareness one has in being aware that one is “appeared to” in a certain way in a certain sense modality. Again, introspective awareness is awareness that.
Part of what I have to say about the belief involved in introspective awareness of phenomenal character is derivative from what I have to say about phenomenal character. Given that phenomenal character is not a matter of mental objects, sense-data or the like, having certain properties, the beliefs this gives rise to are not about such objects. They are beliefs about how things look, feel, smell, etc. – about how one is “appeared to.”
But there is a puzzle here. Given that phenomenal character is an aspect of representational content, beliefs about it should be to the effect that one is in a state having such and such representational content. It is natural to suppose that if the introspective belief is true then its content embeds the content, or part of the content, of the perceptual state; and that in any case it embeds what could be the content, or part of the content, of a perceptual state. The way the perceptual content is embedded should be like the way the content of the belief that it is raining is embedded in the content of the belief that one believes that it is raining. But it has recently been urged, with considerable plausibility, that the content of perceptual states is, at least in part, nonconceptual content. If this is true, it seems plausible that the part that is nonconceptual includes the part that determines the phenomenal character of the state. On the other hand, it is widely held that the content of beliefs is, of necessity, conceptual. And how can a content that is conceptual embed a content that is nonconceptual?
Let the state be one of being visually appeared to in a certain way. The introspective belief will simply be the belief that one is visually appeared to in that way. The content of this belief will certainly be partly conceptual – it will involve the concept of being visually appeared to, or of things looking some way. No one who lacks this concept can have the belief that things look a certain way to her, or that she is visually appeared to in a certain way. What is held to be nonconceptual is the “certain way” one is appeared to, the “certain way” things look – or at least some part or aspect of this certain way. If this is nonconceptual, and is embedded in the content of the perceptual belief, then that content will be in part nonconceptual.
I have nothing much, and certainly nothing new, to offer by way of a characterization of non-conceptual content. What I have in mind is similar to what Christopher Peacocke calls “scenario content” -- a content “individuated by specifying which ways of filling out the space around the perceiver are consistent with the representational content being correct.” On the account I have offered, the “ways of filling out the space” would have to be ways phenomenal properties are distributed in that space. For a person’s experience to have such a content it is not required that the person have the concepts such a specification would require. And even if she does, the content is not in any sense composed of those concepts, in the way the content of the belief that rollerblading is dangerous can be thought of as composed of the concepts of rollerblading and danger.
If one thinks that the content of perceptual experience is non-conceptual, and that the contents of beliefs are conceptual, one might offer an account of introspective beliefs about perceptual experience along the lines of one Richard Heck has recently offered as an interpretation of what Gareth Evans had to say about this. Heck’s proposal is about how we should understand a person’s awareness that things appear a certain way to her. Roughly, it is an awareness, based on an exercise of the conceptualizing capacity that the person uses in making ordinary perceptual judgments about the world, of what one would judge in the absence of certain “extraneous” information – information that bears on whether the experience is veridical. The content of this awareness, and the associated belief, is conceptual. But it “tracks” the nonconceptual content of the experience in question – that to which the conceptualizing ability is applied. As I understand this view, it is never the case that we have a perceptual belief that actually embeds the nonconceptual content of the perceptual experience; the closest we come to this is having a perceptual belief whose content embeds a conceptual content that “tracks” that nonconceptual content.
The awareness of how things “appear” that this view is designed to explain seems to be awareness of how they appear in what I have called, following Dretske, the doxastic sense – the sense in which the way things appear is the way one would take them to be if one took one’s experience at face value. How, along these lines, could one explain awareness of how things appearp, i.e., appear in the phenomenal sense? As noted earlier, something can appearp white without appearingd white – the highlighted part of a table surface may look the way white things look under standard or normal conditions, without its being true that one would judge that it is white if one took one’s experience at face value. In dim light many things lookp black without lookingd black. If one describes something as lookingp F, where F is a color word, one is of course conceptualizing the content of one’s experience in a way. Assuming the possibility of spectrum inversion, an expression like “looksp blue” may stand for different ways of looking in the mouths of observers with different sorts of perceptual systems – so either it expresses a different concept for different sets of perceivers, or it expresses a concept that picks out a particular way of looking only if indexed to a particular sort of perceiver. And it would seem that in order to employ such a concept a perceiver must know, independently of having it, how blue things appearp to her under normal viewing conditions and that this is the same as the way a particular thing appearsp to her. This knowledge involves a representation of a way of appearingp; and if this representation is by way of a concept, it is not a concept standardly expressed by any natural language expressions. John McDowell, who thinks that the content of perceptual experience is always conceptual, holds that the concepts involved in this content are often “demonstrative” concepts. And no doubt one can have a demonstrative concept of a particular way of appearingp – a particular phenomenal property. Such a concept might be expressed on a given occasion by a phrase of the form “The way this looksp to me now.” If, as I think, it is implausible to suppose we have enough such concepts to capture the rich content of our perceptual experience at a particular time, that is a reason for rejecting the view that the representational content of perceptual experience is always conceptual. And if, as I am also inclined to believe, it is implausible to suppose that we have enough such concepts to capture the content of our awareness of the content of our perceptual experience, that is a reason for questioning the view that the embedded content of such awareness is always conceptual – and so for questioning the view that the contents of beliefs is always conceptual.
We certainly have introspective beliefs about how things appear that conform to the Evans/Heck account, and whose contents are entirely conceptual. But I am attracted by the idea that we also have introspective beliefs whose embedded content is nonconceptual, and that these are, or include, beliefs about the phenomenal character of our experience. This idea goes nicely with a view which if true would explain the self-intimation of phenomenal character, and permit some introspective beliefs to be infallible. On this view, there is just one tokening of the content that is both the phenomenal character of the perceptual state and the embedded content of the introspective belief about the phenomenal character of that state. Part of the perceptual state is literally included in the perceptual belief, making it impossible to have the belief without having that part of the perceptual state. While this would make introspective beliefs infallible when their content embeds the part of the content of a perceptual state that is the nonconceptual representation of phenomenal properties, it is compatible with it that all beliefs whose content is entirely conceptual are fallible. The self-intimation of phenomenal content would consist in the fact that we are such that when in a perceptual state having a certain phenomenal character, a certain sort of reflective act results in that state, or part of it, being included in a larger state that plays the functional role of a belief embedding the phenomenal part of the content of that state.
While I am attracted by this view, and think that it is worth developing, I am not at all sure it is correct. The single-tokening view faces the difficulty that an introspective belief about the phenomenal character of an experience can be retained in memory after the experience has ceased to exist; the tokening of the embedded content of the memory belief obviously cannot be identical with the no longer existing tokening of the content of the perceptual experience, and it is unclear how a belief could start by sharing a content tokening with an experience and then come to token that same content in an entirely different way. And any defense of the view that the contents of beliefs can embed nonconceptual contents must take account of the powerful considerations in support of the view that belief content is always conceptual.
What is central to the view advanced in this paper is the claim that the contents of perceptual experiences include representations of properties of external objects (or parts of our bodies) which things have in virtue of appearing to us, or being disposed to appear to us, in certain ways, these being what I call phenomenal properties, and that the phenomenal character of experiences that we are introspectively aware of is this part of their representational content. I think that this view is recommended by the fact that it reconciles representationalism about perceptual content, and the avoidance of phenomenal objects (such as sense-data), with the view that phenomenal character is bestowed by qualia and that inverted qualia and alien qualia are at least conceptually possible. On the present version of this view, unlike the versions in earlier work of mine, our introspective access to the phenomenal character of experiences is equally an access to the qualia that bestow this content, which I hope is a further recommendation of it.
Bermundez, Jose 1998: The Pardox of Self-Consciousness. Cambridge, MA: Bradford-MIT.
Block, N. 1996: “Mental Paint and Mental Latex.” Philosophical Issues, 7, 19-49.
Block, N. 1999: “Sexism, Racism, Ageism, and the Nature of Consciousness”. Philosophical Topics, 26, 1&2, 71-88.
Block, N. and Fodor, J. 1972, “What Psychological States are Not.” The Philosophical Review, 81, 2, 159-181.
Dretske, F, 1993: “Conscious Experience”. Mind, 102:406, 263-283.
Dretske, F. 1995: Naturalizing the Mind. Cambridge, MA: Bradfordd-MIT.
Evans, G. 1982: The Varieties of Reference. Oxford, The Clarendon Press.
Hardin, C.L. 1988: Color for Philosophers: Unweaving the Rainbow. Indianapolis, Indiana: Hackett Publishing Company.
Hardin, C.L. 1997, “Reinverting the Spectrum,” in A. Byrne and D. Hilbert, eds., Readings on Color, Vol. I, The Philosophy of Color. Cambridge, MA: Bradford-MIT.
Harman, G. 1996: “Explaining Objective Color in Terms of Subjective Experience,” Philoosophical Issues, 7, 1-17.
Heck, R. forthcoming: “Non-Conceptual Content and the ‘Space of Reasons”. Forthcoming in The Philosophical Review.
McDowell, J. 1994: Mind and World. Cambridge, MA: Harvard University Press.
Levine, J. 1984: “Materialism and Qualia: The Explanatory Gap.” Pacific Philosophy Quarterly, 64, 354-361.
Peacocke, C. 1992: A Study of Concepts. Cambridge, MA: Bradford-MIT.
Shoemaker, S. 1975: “Functionalism and Qualia.” Philosophical Studies, 27, 291-315.
Shoemaker, S. 1982: “The Inverted Spectrum.” Journal of Philosophy, 79, 7, 357-381.
Shoemaker, S. 1994a: “Phenomenal Character.” Nous, 28. 21-38.
Shoemaker, S. 1994b: “Introspection and ‘Inner Sense’”. Philosophy and Phenomenological Research, LIV, 249-314.
Shoemaker, S. 1996: “Intrasubjective/Intersubjective.” In S. Shoemaker, The First-Person Perspective and Other Essays. Cambridge: Cambridge University Press.
Shoemaker, S. 2000: “Phenomenal Character Revisited.” Philosophy and Phenomenological Research, LX, 2, 465-468.
Smart, J.J.C. 1962: “Sensations and Brain Processes.” In The Philosophy of Mind, V.C. Chappell ed., Englewood Cliffs, N.J.: Prentice Hall. This is a slightly revised version of a paper originally published in The Philosophical Review, 68, 1959, 141-156.
Van Gulick, R. 1993: “Understanding the Phenomenal Mind: Are We All Just Armadillos?”, in Consciousness, ed. by M. Davies and G. Humphrey (Oxfrod: Blackwell), pp. 137-149.
 See Shoemaker 1994b.
 See Dretske 1993.
 This was perhaps an unfortunate choice of terminology, since the term “phenomenal property” is sometimes used to refer to properties of experiences, i.e., as a synonym of “quale,” whereas I am using to to refer to properties of external things. But having made the choice, I will stick with it.
 See Shoemaker 1994a, and Lecture III of Shoemaker 1994b.
 See Shoemaker 2000.
 If we think of the dispositional properties as ones the thing has in virtue of its intrinsic properties, then the specification of the exercise of the disposition will be in conditional terms – it is disposed to appear a certain way to an observer if the observer is of a certain sort and related in such and such ways to it, and if the illumination conditions are such and such. We can also speak of a dispositional property the object has in virtue of its intrinsic properties plus its current situation, minus any facts about observers. Then the disposition is simply a disposition to appear a certain way to an observer of a certain sort related to it in a certain way. Here the dispositional property will be in part relational – having it will involve being in certain lighting conditions, and perhaps being adjacent or proximal to other objects having thus and such intrinsic properties. Dispositional properties of both kinds will be properties the object can have when not perceived; but while a thing can lose a dispositional property of the first kind only by undergoing an intrinsic change, a thing can lose a dispositional property of the second kind without undergoing any intrinsic change, namely by undergoing of change of situation.
 Well, beyond question unless one has a very austere view about what it takes to be a genuine property.
 If a temporary defect in my perceptual system makes something lookp blue, my experience will fail to be veridical, according to the principle stated here. But since the thing does lookp blue to me, won’t it have the occurrent appearance property of lookingp blue? One could define “occurrent appearance property” in such a way that this is true; but I think it is better to make it a requirement of something’s having such a property that the way it appearsp is not due to a defect in the perceiver’s perceptual system.
 The reason for saying “one or more sorts” rather than “some sort” is that I want to allow that the same appearance property can be observed by observers of different sorts. This must be so if spectrum inversion and the like are possible – for individuals who are spectrum inverted relative to each other will have to be of (at least slightly) different sorts.
 Smart 1962.
 It is quite common to hold that pains can be thought as experiences of bodily damage. But it does not seem right to say that the primary content of these experiences is a proposition having to do with bodily damage. What these experiences are, in the first instance, is perceptions representing some part of one’s body as instantiating a phenomenal property, e.g., achiness, which is a sign of bodily damage.
 I am grateful to many people for helping to persuade me that this is so, including Ned Block, Tyler Burge, David Copp, and Tom Nagel.
 This paper is a descendent of a paper I presented at Bowling Green University, the University of Colorado, New York University, and UCLA in the Spring of 2000. I am grateful to the audiences at these places, and in particular to David Copp, George Bealer, Daniel Stoljar, Michael Tooley, Ned Block, Tom Nagel, Tyler Burge, Andrew Hsu, and Michael Thau for comments and criticisms that led me to rethink and extensively revise the paper. Thanks also to Carl Ginet and Benjamin Hellie for comments on the penultimate draft of the present version.