The purpose of the present paper is to explore the relation between the phenomenal aspect of conscious experience and its intentional aspect. I wish to argue that the latter aspect is conceptually dependent on the former. I shall offer an account of the intentionality of conscious experience that will show how it is determined by its phenomenological profile.

In what follows, I shall use the term 'experience' to refer to conscious qualia. So construed, experiences are always of something. There is thus an of-ness relation between experience and something else, for something is always being experienced. If someone says she experiences, and you ask "What do you experience?", to which she answers "Oh, nothing... I just experience", then that person does not understand the English word 'experience'. Experience is, essentially, about something -- it has a content. This simple observation led many people to attempt a theorization of experience squarely within a representational approach to the mind. Furthermore, since we have a strong sense of how to go about naturalizing mental representation (through some story about historico-causal connections between brain and world), to the extent that experiences are also mental representations, they must be naturalizable.

To be accommodated by the representational theory of mind, experiences must have their representational character essentially. For a mental state to be essentially representational, its identity must vary concomitantly with the identity of what it is about (that is, token representational states can be type-identical only if their representational objects are type-identical, and type-different only if their representational objects are type-different). But the problem for this approach to experience is that there seems to be a better candidate for the essence of experience: its phenomenal character, the way it feels -- what it is like -- to have it. It seems that this phenomenal character is what the distinctive experientiality of experience, if you will, is all about. To overturn this consideration, I offer an argument to the effect that, as far as experience is concerned, being in a phenomenal state and being in an intentional state is, appearances to the contrary, one and the same thing.

It is a necessary condition for property difference, I suggest (following David Lewis), that there should be a logically possible state of affairs in which one property is exemplified and the other is not. But there is no pair of possible experiences with type-identical feel and type-different content, or inversely. We have no idea how to make sense of such a possibility. Can we imagine an experience that feels like an experience as of a red apple, but is in fact an experience of a green door? Clearly not. Granted, some unhappy circumstances may conspire to bring it about that an experience of a red apple would be caused by a green door; but that would hardly make the experience in question an experience of a green door, in any pertinent sense of the term. I conclude that the individuations of experiences according to content and according to feel are one and the same, in the sense that for any pair of possible experiences, they are either both intentionally and phenomenally type-identical or both intentionally and phenomenally type-different (they cannot be phenomenally type-identical and intentionally type-different, nor inversely). Therefore, being in a certain phenomenal state and being in a certain intentional state are the same property. The feel of an experience must be precisely a feel of its object or content: the content of an experience is what one feels. Experience thus provides a unique form of felt aboutness.

*

So phenomenology and intentionality, feel and content, distribute convergently, when it comes to experience. It follows that either experiential phenomenology is (appearances to the contrary) a relational property of the experiencer or experiential intentionality is (appearances to the contrary) an intrinsic one. Externalists would obviously go for the former, but I think that the other theoretical avenue -- the suggestion that experiential content must be intrinsic to the experiencer -- is intuitively more compelling. Most internalists about experience seem to share the externalists’ distaste for the notion of internal intentionality, and hence typically resort to an alleged divergence of feel and content. A different sort of internalist is McGinn:

I suggest that this uniqueness McGinn refers to lies in the internality he mentions, and the latter comes down to this, that intentionality is an intrinsic feature of experience. There just isn't the sort of gap between vehicle and content we find in the intentionality of words, tacit beliefs, and the like. The experience encompasses its object, as McGinn puts it -- it really does contain its content. These are all mystifying features of experience -- to the point that some would feel the need to deny their existence, with a somewhat cold-hearted disregard for what is closer to us than anything else -- but I am here merely trying to give a sense of how difficult the problem of experience is. It is unclear how to construe this internality of experiential intentionality, but it seems a necessary condition that it exclude any individuation of experiences that would advert to extra-cranial factors. The puzzle about conscious experience is precisely that the external representational connections to the environment, although there, do not, and cannot, affect the identity conditions of experiences. How could that be?

*

To get a better grip on the problem, let us consider somewhat fanciful cases in which the phenomenal-internal aspect of experience and the representational-external aspect may be alleged to diverge. The cases should not be overly fanciful, however, for then they will fail to threat a mitigated convergence thesis. Thus, if we conjure up cases that involve nomological impossibilities, the problem they will generate could be bypassed simply by urging a convergence of the phenomenal and the representational in nomologically possible worlds, rather than in all metaphysically or logically possible worlds. The cases I shall present will be ones that may perfectly well take place in the actual world. I present two cases: one allegedly involving token experiences with type-identical phenomenology and type-different representational profile; the other allegedly involving, conversely, token experiences with type-identical representational profile and type-different phenomenology.

The first case can be modeled on twin-earth cases: an experience of water and an experience of twin-water are, ex hypothesi, phenomenally type-identical, but they are directed to different things, and are hence representationally type-different. To avoid the possibility that this scenario involves a nomological impossibility, however, let us focus on the actual case of gold and fool's gold. These are different substances, but experiencing them feels the same (which is why fool's gold succeeds in fooling). Gold experiences are therefore phenomenally type-identical with fool's gold experiences. At the same time, they are representationally type-different. This is because samples of gold are type-different from samples of fool's gold, and as I said above, for a mental state to be a representational state, it must have its representational properties essentially, and that means that the identity of the representing state must vary concomitantly with the identity of what is represented.

For the second type of case we can use the inverted spectrum scenario. Here experiences which are directed onto type-identical items in the world nonetheless feel different. Thus, a normal experience of gold feels, in a sense, yellow, but to someone who wears spectrum-inverting lenses it feels purple. Together, these cases allegedly provide a bilateral divergence of intentional object and phenomenal feel. If so, experiences cannot be representational states that are nonetheless internal.

But here our basic intuition that felt aboutness is internal slaps back. Experiences of gold, fool's gold, and inverted gold (that is, gold as seen under the influence of spectrum-inverting lenses) are definitely about something, and what they are about is a piece of the real world. Yet the way they feel fixes the nature of their aboutness, so that gold experiences belong to the same psychological kind as fool's gold experiences and to a different psychological kind from inverted gold experiences. The inevitable conclusion is this: gold and fool's gold must be type-identical objects of experience, and gold and inverted gold must be type-different objects of experience. But since gold and fool's gold are type-different objects simpliciter, and gold and inverted gold are type-identical objects simpliciter, the hold between objects of experience and objects simpliciter must break. That is, a distinction must be drawn between the object as it is in itself and as it is qua experienced. Ultimately, the account of experiential intentionality I am about to canvass turns on just such a distinction. The problem, now, is how to make sense of such a distinction within a naturalist approach to intentionality, one that doesn’t appeal to the suspect notion that experiential objects are ghostlike entities floating about in a sort of private internal space (a Cartesian theater), perhaps projected thereonto by the acts of experience themselves.

*

A sample of gold and a sample of fool’s gold are type-different objects; a sample of gold and a sample of inverted gold are type-identical objects. But why? What makes that the case? A familiar model runs something like this. Every object has many properties. Some it has essentially, most accidentally. The essential properties are properties the object cannot lose without ceasing to be the thing it is. Thus, the essential properties of a chair have to do with its function and its shape; burn the chair down to ashes, so it loses both shape and function, and the chair, as a chair (that is, as the thing it is), goes out of existence. The essential properties of an object thus determine the extension of the object in time, as well as in modal space (they govern, in technical parlance, transtemporal and transworld identity). Crucially, some objects have essential properties independently of us and how we treat them. Gold and water are such: the sort of properties that are essential to them are their underlying natures, which they have independently of us. Therefore, if something does not have the underlying nature gold does, it cannot be gold, even if it looks and feels just like gold; thus, fool’s gold is not the same thing as gold, since it is not atomic number 79. Conversely, if something does have the underlying nature of gold, it is gold, even if it does not look like gold; thus, inverted gold -- which is atomic number 79, but looks purple -- is gold.

Now, what I suggest we do is distinguish between those properties of gold that are essential to it as it is in itself and those that are essential to it as object of experience. As it is in itself, gold is essentially atomic number 79. Qua experienced, however, it is essentially yellow, solid, etc. The gold, qua object of experience, has the same properties the gold as it is in itself has, it's just that different properties in this shared pool count as essential to them. This situation is analogous to the notorious case of the statue and the clay: they share the same (non-modal) properties, but within the set of their shared properties, different proper subsets are essential to the statue and to the piece of clay (e.g., shape is essential to the former but not to the latter).

If we embrace such a model, we can make sense of the type-identity of the objects of gold experiences and fool’s gold experiences and type-difference of the objects of gold experiences and inverted gold experiences. For what makes a sample of gold the object of experience it is are properties it shares with fool’s gold and it does not share with inverted gold. More generally, the essential properties of gold, qua experienced, are precisely those of its properties that give rise to experience that feels the way gold experience feels. That is, only secondary qualities of gold can be essential to it qua object of experience. This can be construed as a general constraint on the essence of all experiential objects as such: an object of experience, qua object of experience, has only its secondary qualities essentially. I now turn to discuss, respectively, the sense in which this suggestion is internalist and the sense in which it is naturalist.

The suggestion is internalist inasmuch as both the nature and the existence of the object of experience is totally dictated by the experience itself. This is because all of its essential properties are properties which are defined according to their experiential effects, what it feels like to experience something that features them. The yellowness of gold, for instance, just is the property of being disposed to give rise to experiences that feel so-and-thus; at least an account of yellowness along these lines has a lot going for it and has been ably defended throughout history. Furthermore, the objects of experience have no existence as such outside the way it feels to experience them. Paint a piece of gold purple and an object of experience has gone out of existence, and a new one has come into existence (just as, upon burning it down, a chair goes out of existence and a pile of ashes comes into existence: the chair ceases to be, qua chair, because it lost its essential properties).

Experiential content is thus fashioned by experience. The externalists individuate experiences relationally, and their objects intrinsically. I suggest to reverse that order: individuate experiences intrinsically (according to their phenomenal properties), and their objects relationally (according to their dispositional properties of giving rise to the phenomenal properties of experiences). Thus the nature of objects of experience turns out to be conceptually dependent on the nature of the way it feels to experience them. We may call the kind of content involved in experience mind-relative content. Note that since experiential content depends definitionally on experiential vehicle, the latter determines the former, and consequently the relation of experiential intentionality supervenes on the intrinsic features of the experience; it is, as they say, locally supervenient.

Although internalist, the account is also naturalist. The way I see it, an account of intentionality is naturalistic iff it supposes both that the intentional object is a natural entity and that the relation it bears to the intentional vehicle is a natural one. I now turn to show that my account meets both conditions. To do that, I shall need a conception of natural objects and a conception of natural relations to work with. Inspired by Kim, I shall suppose that an item in the causal network of space-time is a natural object. And I shall suppose that causality is a natural relation. What I wish to show is that within my account, the objects of experience are items in the causal network of space-time, and that the intentional relation is simply causality.

The object of a gold experience is, on my account, the real piece of gold in front of the experiencer -- which piece of gold is, of course, spatio-temporally locatable. So objects of experience, qua objects of experience, assume determinate spatio-temporal locations, indeed the very same locations assumed by the corresponding objects as they are in themselves. Again, this is analogous to the case of the statue and the clay: although the clay is not a full-blown entity in the way the statue is, it still does not involve any irreal aspect. Furthermore, the set of actual causal relations the experiential object bears to other items in space-time is also the same as that of the object as it is in itself.

As far as the intentional relation holding between the vehicles of experience and the objects of experience is concerned, note that on my account, the relation is the same sort of qualified causal relation postulated by the more standard forms of causal semantics. The set of all the actual causal relations the experiential object bears to other items in space-time is the same as that of the object in itself, and one of these causal relations is a relation to the experiential state itself (as just another spatiotemporal event in the world). Therefore, the causal relation between the object in itself and the experience holds also between the object of experience and the experience. So on the present account, the relation between experiential objects and experiential vehicles is a natural relation, namely causality. Most importantly, this kind of relation is to be contrasted with a sort of internal projection of the objects inside a mysterious experiential space, which is the kind of intentional relation assumed by historical forms of internalism about the intentionality of conscious experience.

*

To conclude, experiential intentionality is a baffling phenomenon to the extent that it is essentially felt aboutness, which means it wraps up together an intrinsic feel and a directedness to the external world. The approach I have tried to develop has been designed to display sensitivity to both aspects of conscious experience and to the complexity of their interlacing. My strategy has been to distinguish the objects of experience from the objects as they are in themselves, and construe the former as mind-dependent natural entities, thereby enabling an internalist semantics for experiences within a naturalistic semantic framework. There is a sense in which the distinction is all too natural, and indeed provides the raison d'etre of higher-order thought and scientific investigation, as intellectual endeavors designed to transcend experiential appearance and reach reality as it is in itself.