THE INTENTIONALITY OF PHENOMENOLOGY AND THE PHENOMENOLOGY OF INTENTIONALITY
Terence Horgan and John Tienson
University of Memphis
[To appear in (Chalmers, ed.) Philosophy of Mind: Classical and Contemporary Readings (OUP, 2002).]
What is the relationship between phenomenology and intentionality? A common picture in recent philosophy of mind has been that the phenomenal aspects and the intentional aspects of mentality are independent of one another. According to this view, the phenomenal character of certain mental states or processes—states for which there is “something it is like” to undergo them—is not intentional. Examples that are typically given of states with inherent phenomenal character are sensations, such as pains, itches, and color sensations. This view also asserts, on the other hand, that the intentionality of certain mental states and processes—their being about something—is not phenomenal. Beliefs and desires are the paradigm cases of intentional mental states. Although they are intentionally directed—i.e., they have aboutness—these mental states are not inherently phenomenal. There is nothing that it is like to be in such a state by virtue of which it is directed toward what it is about.
We will call this picture separatism, because it treats phenomenal aspects of mentality and intentional aspects of mentality as mutually independent, and thus separable. Although there may be complex states that are both phenomenal and intentional, their phenomenal aspects and their intentional aspects are separable. Many philosophers who hold this picture have thought that these two aspects of mentality lead to quite different sorts of problems with respect to the project of “naturalizing the mental.” Proponents of separatism often hold that while the problem of naturalizing phenomenology poses great difficulties, the problem of naturalizing intentionality is much more tractable.
Separatism has been very popular in philosophy of mind in recent decades, and is still widely held. Those who oppose it regard it as a view against which they need to characterize their own positions—a common picture that they must explicitly reject. In this paper we argue that separatism is profoundly wrong. We depart from it quite thoroughly, in ways importantly different from other recent departures. We affirm the following theses, both of which are repudiated by separatism:
The Intentionality of Phenomenology: Mental states of the sort commonly cited as paradigmatically phenomenal (e.g., sensory-experiential states such as color-experiences, itches, and smells) have intentional content that is inseparable from their phenomenal character.
The Phenomenology of Intentionality: Mental states of the sort commonly cited as paradigmatically intentional (e.g., cognitive states such as beliefs, and conative states such as desires), when conscious, have phenomenal character that is inseparable from their intentional content.
In addition to these two theses (henceforth, IP and PI), we advocate another important claim about the interpenetration of phenomenology and intentionality:
Phenomenal Intentionality: There is a kind of intentionality, pervasive in human mental life, that is constitutively determined by phenomenology alone.
We use the expression ‘constitutively determined’ to mean that this kind of intentionality is not merely nomically determined; rather, intentional mental states have such intentional content by virtue of their phenomenology.
So-called “representationalist” theories of phenomenal properties are a currently influential departure from separatism. Although extant versions of representationalism embrace thesis IP, typically they do not embrace thesis PI. Nor do they embrace the thesis of Phenomenal Intentionality, since they hold that intentionality is prior to phenomenology. So our position differs significantly from standard representationalism.
We argue for the three theses set out above (sections 1-3), in part by way of introspective description of actual human experience. If you pay attention to your own experience, we think you will come to appreciate their truth. Our position has important consequences, when combined with the plausible thesis (argued for in section 4) that phenomenology is “narrow,” i.e., it does not depend—except perhaps causally—upon what goes on “outside the head” of the experiencer. One consequence is that there is a kind of narrow intentionality that is pervasive in human mental life—a form of intentional directedness that is built into phenomenology itself, and that is not constitutively dependent on any extrinsic relations between phenomenal character and the experiencer’s actual external environment. A further consequence is that theories that ground all intentionality in connections to the external world, such as causal and teleological theories of intentionality, are deeply mistaken.
1. The Intentionality of Phenomenology
The mental states typically cited as paradigmatically phenomenal have intentional content that is inseparable from their phenomenal character. Let us consider some examples: first, experiences of red as we actually have them. You might see, say, a red pen on a nearby table, and a chair with red arms and back a bit behind the table. There is certainly something that the red that you see is like to you. But the red that you see is seen, first, as a property of objects. These objects are seen as located in space relative to your center of visual awareness. And they are experienced as part of a complete three-dimensional scene—not just a pen with table and chair, but a pen, table, and chair in a room with floor, walls, ceiling, and windows. This spatial character is built into the phenomenology of the experience.
Consider too the experience of seeing an apple on the table, picking up the apple, and taking a bite out of it. There is the look and smell of the apple. Then (as you grasp it) there is the feel of the apple, its smoothness, roundishness, and firmness. Then there is its weight (as you pick it up). Finally there is the feel of the apple in your mouth, followed by the crunching sound, taste, and feel of juiciness as you take a bite. We will not attempt to write the small book one could write describing this simple experience. But we need to note some highlights. First, the look, feel, smell, sound, and taste of the apple are experienced as a unity in space, as all belonging to a single object. The taste is in your mouth; the smoothness and roundishness that you feel—with parts of your mouth as well as your hand—are there, too. Second, it is important to notice that what is experienced tactilely are various spatial properties of the object, not sensations. One has, of course, tactile sensations as well, though one does not normally attend to them. (The tactile sensations are, when noticed, experienced as the sort of things that can only belong to a sentient being.) The properties of smoothness, firmness, etc. are experienced as the sorts of things that can only belong to an “external” object in space. Third, the apple is encountered as moving. The experience is of a temporal object, an object that endures. The same is true when you see another person take a bite of an apple. Experience is not of instants; experience is temporally thick. This is obvious in the case of hearing tunes or sentences, where the temporal pattern is a palpable feature of the experience. The temporal pattern is also a palpable feature of the seen moving apple, though less frequently noted as such. But it is no less true that stationary objects are seen as enduring and as unchanging.
For any experience involving a specific shade of red, one can abstract away from the total experience and focus on the distinctive what-its-like of that shade of red per se—a phenomenal aspect of this total experience that it has in common with innumerable other total experiences that differ in the perceived location of the experienced red or in the shape of the red surface, etc. But even considered in isolation from any total visual-experiential state, the what-it’s-like of experiencing red is already intentional, because it involves red as the intentional object of one’s experience. Again, redness is not experienced as an introspectible property of one’s own experiential state, but rather as a property of visually presented objects. 
Of course, in typical cases of experiencing red, the overall phenomenal character of one’s visual experience is a structurally rich what-it’s-like of experiencing a visually presented scene, a scene that contains a whole array of apparent enduring objects with various properties and relations—including the property redness instantiated on the surfaces of some of these objects. The total visual experience with this overall phenomenal character is richly intentional, since it presents a temporally extended scene comprising various objects that instantiate various properties and relations at various spatial locations relative to one’s center of visual awareness. This total visual experience is also richly phenomenal, because there is an overall what-it’s-like of experiencing the whole scene. (Any visually noticeable alteration in the visually presented scene would be a phenomenal difference in one’s total visual experience.)
Another commonly cited example of phenomenal consciousness is the distinctive phenomenal character of pain. Experiences of pain are thoroughly intentional: pain is experienced as a particular feeling at a certain place in one’s own body. (This is why there can be such a thing as phantom-limb phenomena, in which pain is experienced as located in a limb that has been amputated.)
More generally, the overall phenomenal character of one’s experience includes a structurally rich what-it’s-like of tactilely and kinesthetically experiencing one’s presented body, an apparent body containing a whole array of tactilely and kinesthetically distinguishable apparent parts, many of which are experienced as parts that one can voluntarily move. The total tactile/kinesthetic experience with this overall phenomenal character is richly intentional, by virtue of the complexity of the body as presented. This total experience is also richly phenomenal, since it has the what-it’s-like-ness of tactilely and kinesthetically experiencing one’s whole body. (Any tactilely or kinesthetically distinguishable alteration would be a phenomenal difference in one’s total tactile/kinesthetic experience.)
The full-fledged phenomenal character of sensory experience is an extraordinarily rich synthetic unity that involves complex, richly intentional, total phenomenal characters of visual-mode phenomenology, tactile-mode phenomenology, kinesthetic body-control phenomenology, auditory and olfactory phenomenology, and so forth—each of which can be abstracted (more or less) from the total experience to be the focus of attention. This overall phenomenal character is thoroughly and essentially intentional. It is the what-it’s-like of being an embodied agent in an ambient environment—in short, the what-it’s-like of being in a world.
2. The Phenomenology of Intentionality
We have been describing the intentionality of sensory-perceptual phenomenal consciousness. Let us now focus on the thesis of the phenomenology of intentionality (PI): consciously occurring intentional states have phenomenal character that is inseparable from their intentional content.
Intentional states have a phenomenal character, and this phenomenal character is precisely the what-it-is-like of experiencing a specific propositional-attitude type vis-a-vis a specific intentional content. Change either the attitude-type (believing, desiring, wondering, hoping, etc.) or the particular intentional content, and the phenomenal character thereby changes too. Eliminate the intentional state, and the phenomenal character is thereby eliminated too. This particular phenomenal character could not be present in experience in the absence of that intentional state itself.
Consider, for example, an occurrent thought about something that is not perceptually presented, e.g., a thought that rabbits have tails. Quine notwithstanding, it seems plainly false—and false for phenomenological reasons—that there is indeterminacy as to whether one is having a thought that rabbits have tails or whether one is instead having a thought that (say) collections of undetached rabbit parts have tail-subsets. It is false because there is something that it is like to have the occurrent thought that rabbits have tails, and what it is like is different from what it would be like to have the occurrent thought that collections of undetached rabbit parts have tail-subsets.
The overall phenomenology of these kinds of intentional states involves abstractable aspects which themselves are distinctively phenomenological. For example, if one contrasts wondering whether rabbits have tails with thinking that rabbits have tails, one realizes that there is something common phenomenologically—something that remains the same in consciousness when one passes from, say, believing that rabbits have tails to wondering whether rabbits have tails, or vice versa. It is the distinctive phenomenal character of holding before one’s mind the content rabbits have tails, apart from the particular attitude type—be it, say, wondering, hoping, or believing. This aspect of the overall phenomenology of intentionality is the phenomenology of intentional content.
In addition, there is also a specific what-its-likeness that goes with the attitude type as such. There is a phenomenological difference between wondering whether rabbits have tails on one hand and thinking that rabbits have tails on the other. This aspect is the phenomenology of attitude type. Attentive introspection reveals that both the phenomenology of intentional content and the phenomenology of attitude type are phenomenal aspects of experience, aspects that you cannot miss if you simply pay attention.
One might reply that although there is indeed a phenomenological difference between thinking that rabbits have tails and thinking that collections of undetached rabbit parts have tail-subsets, this difference merely involves the fact that we often think in language. Thus, the phenomenological difference between the two thinking experiences involves not the different contents, but rather the fact that the auditory imagery that goes with thinking that rabbits have tails is different from the imagery that goes with thinking that collections of undetached rabbit parts have tail-subsets.
However, nonperceptual intentionality in normal humans does not always involve language and/or auditory imagery. For instance, conscious, unverbalized beliefs about the locations of nearby unperceived objects are just as ubiquitous in human life as is the explicit or imagistic verbalization of one’s focal train of thought. Think for example, of cooking, cleaning house, or working in a garage or woodshop. In any such activity, you might spontaneously move to retrieve a needed tool that is out of sight. There is something that it is like to think that a certain tool is just there—in that cabinet, say—
but such beliefs are typically not verbalized either vocally or subvocally or by way of verbal imagery. (Your verbal energies might all the while be directed toward ongoing philosophical discussion with a companion, uninterrupted by your selection of an appropriate tool.) You also, of course, frequently have unverbalized thoughts about the locations of objects in distant familiar locations.
In any event, the what-it’s-likeness of intentionality that we are talking about—even when it is specifically tied to certain words in English or some other natural language—does not attach to those words simply as sequences or patterns of sounds, or even as syntactic structures. It attaches to awareness of those words qua contentful; it is the what-it’s-like of hearing or saying those words when they mean just that: that rabbits have tails. So the basic point holds: even if thinking did always involve auditory imagery, the auditory imagery would be intentionally loaded in the experience, not intentionally empty.
This point is illustrated and defended by Galen Strawson (1994). Strawson discusses what he calls “understanding experience”. He contends that understanding and other related kinds of occurrent intentional mental states and processes are very commonly, if not always, laden with distinctive what-it’s-likeness. He points out, for example, the phenomenological difference between hearing speech in a language that one does not understand and hearing speech in a language that one does understand. Imagine two people side by side hearing the same spoken sequence of sounds, with one of them understanding the language and the other one not. At a certain relatively raw sensory level, their auditory experience is phenomenologically the same; the sounds are the same, and in some cases may be experienced in much the same way qua sounds. Yet it is obvious introspectively that there is something phenomenologically very different about what it is like for each of them: one person is having understanding experience with the distinctive phenomenology of understanding the sentence to mean just what it does, and the other is not.
Consider, as a similar example for a single speaker, first hearing “Dogs dogs dog dog dogs,” without realizing that it is an English sentence, and then hearing it as the sentence of English that it is. The phenomenal difference between the experiences is palpable. (If you do not grasp the sentencehood of the “dogs” sentence, recall that ‘dog’ is a verb in English, and compare, “Cats dogs chase catch mice.”)
Consider also hearing an ambiguous sentence. One typically hears it as meaning some one thing in particular, often without realizing that it is ambiguous. There is a phenomenological difference, for example, between hearing “Visiting relatives can be boring,” as a remark about the people who are visiting, vs. hearing it as a remark about visiting certain people oneself. Again, imagine hearing or saying “Time flies” as a cliché about the passage of time, vs. saying or hearing it as a command at the insect races. The actual sound or auditory imagery may be the same, but the total experiences are phenomenally quite different. The sound may have some role that would make it appropriate to call it a vehicle of intentionality, but its meaning what it does, having the intentional content that it does, is an entirely different aspect of the overall phenomenal character of the experience.
In sum: Cognitive intentional states such as consciously occurrent thoughts, and conative intentional states such as consciously occurrent wishes, are phenomenal qua intentional. The overall phenomenal character of such a state comprises both the phenomenology of its specific intentional content and the phenomenology of its specific attitude-type. These are abstractable phenomenal aspects of the state’s unified phenomenal character: the what-it’s-like of undergoing this specific propositional attitude vis-a-vis that specific intentional content.
3. Phenomenal Intentionality
The intuitive considerations in the last two sections can be developed into an argument for the thesis of phenomenal intentionality: there is a pervasive kind of intentionality that is constitutively determined by phenomenology alone. One way to articulate and sharpen this claim is the following. Let two creatures be phenomenal duplicates just in case each creature’s total experience, throughout its existence, is phenomenally exactly similar to the other’s. We can then state the Phenomenal Intentionality thesis this way:
There is a kind of intentional content, pervasive in human mental life, such that any two possible phenomenal duplicates have exactly similar intentional states vis-a-vis such content.
We will call this type of content phenomenal intentional content. The full range of a creature’s phenomenal intentional content is determined and constituted wholly by phenomenology
Consider any creature who is a complete phenomenal duplicate of yourself—its mental life is phenomenally exactly like yours. Assume nothing else about this creature. The thought experiment thus builds in an epistemic “veil of ignorance” about this creature, in order to filter out any factors other than phenomenology itself. So for all you know about this arbitrary phenomenal duplicate of yourself, its sensory-perceptual experiences—including its tactile-kinesthetic experiences and its embodied-agency experiences—might be very largely illusory and hallucinatory concerning the real nature of itself and its surroundings. (It may be helpful to imagine the phenomenal duplicate as a brain in a vat or a disembodied Cartesian mind.) We will argue that you and your phenomenal duplicate share a pervasive kind of intentional content—phenomenal intentional content.
As argued in section 1, sensory-phenomenal states and processes have intentional content that is inseparable from their phenomenal character. These states present an apparent world full of apparent objects that apparently instantiate a wide range of properties and relations, and they present oneself as an apparently embodied agent within that apparent world. Since this kind of intentionality is inseparable from phenomenal character, your phenomenal duplicate will have an apparent world presented to it in exactly the same way.
To make the general point with a representative example, suppose that you have the experience of seeing a picture hanging crooked. Each of your phenomenal duplicates has a phenomenally identical experience. Some of these experiences will be accurate and some will be inaccurate. Whether or not a given duplicate’s picture-hanging-crooked experience is accurate—that is, whether or not things are as the experience presents things as being—will depend upon the duplicate’s actual environment. Thus, the sensory-phenomenal experience, by itself, determines conditions of accuracy: i.e., a class of ways the environment must be in order for the experience to be accurate. In order for such an experience to be accurate, there must be a picture before oneself, and it must be crooked.
That these phenomenally identical experiences all have the same truth conditions is reflected in the fact that each of the experiences is subject in the same way to investigation as to whether it is accurate. For example, you and your phenomenal duplicate each might have the experience of seeming to oneself to be testing one’s perceptual experience for accuracy by making measurements or using a level. You and your phenomenal duplicate each might have the subsequent experience of seeming to oneself to discover that the picture merely appears to be crooked because of irregularities of the wall, or tricks of light. Or, you and your phenomenal duplicate might, in the course of seeming to oneself to be attempting to perform these tests, have the experience of seeming to discover that there actually is no picture—say, by seeming to oneself to discover that one has been looking at a clever holographic image cooked up to make it appear that there is a picture hanging on the wall.
There is also, of course, a sense in which these crooked-picture perceptions of you and your duplicate have different truth conditions. You and your duplicate are looking at different pictures. So the accuracy of your own perception depends on the specific picture you yourself are seeing, whereas the accuracy of your duplicate’s perception depends on the specific, and different, picture that it is seeing. There are thus two ways of thinking of truth conditions: as determined wholly by phenomenology, and as determined in part by items in the experiencer’s environment that satisfy the experiencer’s phenomenology. We return to this distinction in section 5.2.
Your phenomenal duplicate accepts the presentations delivered by perceptual experience—accepts, for example, that there is a picture and a wall—just as you do. These “belief-wise” acceptance states have exactly the same phenomenology, the what-it-is-like of occurrently believing that thus-and-so (where one’s occurrent sensory experience presents things as being thus-and-so). The states also are phenomenologically integrated with those ongoing, richly intentional, sensory-perceptual experiences in exactly the same way as yours. Thus, they are experienced as having the same belief-specific role: the same apparent input conditions, involving apparent deliverances of the apparent body’s apparent senses, and the same apparent effects, involving experiences of apparently acting appropriately with regard to the apparent world as presented. It seems intuitively clear that a belief-wise acceptance state with these phenomenological features is a genuine belief. The phenomenal character of these states, which includes the phenomenology of role, constitutively determines that they are genuine beliefs. And as we argued above, the sensory-presentational content of these states is the same for you and your phenomenal duplicate.
So far we have been discussing perceptual experience and perceptual belief. But since the phenomenal consciousness of your phenomenal duplicate would provide very rich perceptual presentations of a world of numerous apparent objects instantiating numerous apparent properties and relations, such presented items would thereby figure in a wide range of propositional-attitude states whose content goes well beyond the presentations of perceptual experience itself. Here the phenomenology of intentionality—the what-it’s-like of occurrent propositional attitudes as such—enters in full force, quite apart from the presentational content of one’s current sensory-perceptual experience. For your phenomenal duplicate, no less than for you yourself, there would be something that it’s like to wonder whether to cook meatloaf for dinner, something that it’s like to have the thought that there’s beer in the fridge, something that it’s like to hope that one’s spouse isn’t angry that one is coming home late from the Philosophy Department party. These occurrent states in the phenomenal duplicate have all the same “propositional attitude-ish” phenomenology as they do in you. They are experienced as having exactly the same causal role vis-a-vis the phenomenal duplicate’s apparent embodied behavior in its apparent world as you experience them as having. It seems intuitively clear that states with all these features qualify as full-fledged propositional attitudes in your phenomenal duplicate, just as they do in you.
In addition, for each such propositional-attitude state in yourself and in your phenomenal duplicate, the two states have the same phenomenal intentional content, i.e., the same phenomenologically determined truth conditions. Consider, for example, two phenomenologically identical belief-states that you and your phenomenal duplicate would both express by the string of words “The picture behind me is crooked.” In order for such a belief to be true, there must indeed be an object in a certain relation to oneself—behind, no intervening walls, etc.—that satisfies the phenomenologically determined criteria for being a picture, and that picture must be hanging crooked. Considered in this way, your belief and that of your phenomenological duplicate have exactly the same truth conditions. These occurrent states in the phenomenal duplicate, by virtue of having the same phenomenologically determined truth conditions as yours, are thereby subject to the same methods of accuracy assessment: for instance, you and your phenomenological duplicate might each experience turning around to see if the picture is still crooked. If it still appears crooked, you might then experience going through the tests mentioned above. The possibility of such tests is in some sense understood, if not explicitly phenomenologically given, in having the conscious belief that there is a picture hanging crooked behind oneself.
Since your phenomenal duplicate shares with you all the phenomenal intentionality so far described, it thereby possesses significant conceptual resources to speculate and theorize—for instance, about what is very distant spatio-temporally, about what is very small, about the underlying causes of experience and of the apparent ambient environment. It can reason causally, form abstract theoretical concepts, formulate scientific hypotheses and theories, and experience itself as an apparent embodied agent actively engaged (in apparent cooperation with other apparent embodied agents) in the apparent empirical corroboration or disconfirmation of such hypotheses and theories. It can have experiences as of apparently reading about such matters or apparently hearing lectures about them, and thereby can acquire a body of well warranted scientific beliefs about itself and its world. In these respects too, your phenomenal duplicate is like yourself, even though the experiential basis upon which it bootstraps its way up to well warranted, semantically evaluable, scientific beliefs might be highly (or even completely) non-veridical. Thus, for each of your propositional-attitude states about such theoretical entities, your phenomenal duplicate has a propositional-attitude state of the same kind. And for each pair of corresponding states in you and your duplicate respectively, the two states have the same phenomenal intentional content—i.e., the same phenomenally determined truth conditions, linked via the same phenomenally determined “web of belief” to the same kinds of experiential methods of accuracy-assessment.
Virtually everything we have been saying is really just attentive phenomenological description, just saying what the what-it’s-like of experience is like. It is just a matter of introspectively attending to the phenomenal character of one’s own experience. You and your phenomenal duplicate share a pervasive kind of mental intentionality—viz., phenomenal intentionality.
We take it that this thought-experimental argument supports the idea that each specific occurrent intentional state with phenomenal intentional content is constitutively determined by its own distinctive phenomenal character—viz., the what-it’s-like of undergoing that particular attitude-type vis-a-vis that particular phenomenal intentional content. That is, specific phenomenal character determines specific intentional states, provided that the experiencing creature has a sufficiently rich network of actual and possible phenomenal/intentional states. Suppose, for example, that you are now undergoing a psychological state with the distinctive phenomenal what-it’s-like of believing that a picture is hanging crooked on a wall directly behind you. Then you thereby believe that there is a picture hanging crooked on a wall directly behind you; undergoing this phenomenology constitutively determines that you are instantiating that belief-state. Any experiencing creature undergoing this phenomenology would thereby instantiate the belief-state, even if its overall phenomenology is otherwise quite different from your own.
Although each occurrent intentional state with phenomenal intentional content is constitutively determined by its own phenomenal character, at least in the context of a full-fledged cognitive system, it is important to appreciate that this does not mean that phenomenal intentionality somehow guarantees infallible knowledge about what one’s first-order intentional states are. Beliefs about one’s own intentional states are second-order intentional states, and the Phenomenal Intentionality thesis is compatible with the possibility that such beliefs are sometimes mistaken. Indeed, the thesis is compatible with the possibility that some creatures who have phenomenal intentionality—say, certain kinds of nonlinguistic animals—entirely lack any capacity to undergo second-order intentional states at all. What-it’s-likeness is one thing; discursive judgments about it are another. Such judgments are fallible (as are judgments about most anything), even though humans do possess especially reliable capacities to form accurate introspective discursive/classificatory judgments about their own phenomenology.
4. The Narrowness of Phenomenology and of Phenomenal Intentionality
Phenomenology does not depend constitutively on factors outside the brain. Now, it is obvious enough that in normal humans, phenomenology does depend causally on some such factors; but one need only consider how this causal dependence works in order to appreciate the lack of constitutive dependence. First, phenomenology depends causally on factors in the ambient environment that figure as distal causes of one’s ongoing sensory experience. But second, these distal environmental causes generate experiential effects only by generating more immediate links in the causal chains between themselves and experience, viz., physical stimulations in the body’s sensory receptors—in eyes, ears, tongue, surface of the body, and so forth. And third, these states and processes causally generate experiential effects only by generating still more immediate links in the causal chains between themselves and experience—viz., afferent neural impulses, resulting from transduction at the sites of the sensory receptors on the body. Your mental intercourse with the world is mediated by sensory and motor transducers at the periphery of your central nervous system. Your conscious experience would be phenomenally just the same even if the transducer-external causes and effects of your brain’s afferent and efferent neural activity were radically different from what they actually are—for instance, even if you were a Brain in a Vat with no body at all, and hence no bodily sense organs whose physical stimulations get transduced into afferent neural inputs. Among your logically possible phenomenal duplicates, then, are beings whose sensory experience is radically illusory, in the manner of the famous Evil Deceiver scenario in Descartes’ First Meditation—or its contemporary version, the Brain in a Vat.
Thus, phenomenology is narrow, in the sense that it does not depend constitutively on what’s outside the skin, or indeed on what’s outside of the brain. We can now make the central argument:
(1) There is pervasive intentional content that constitutively depends on phenomenology alone.
(2) Phenomenology constitutively depends only on narrow factors.
So, (3) There is pervasive intentional content that constitutively depends only on narrow factors.
That is, the theses of phenomenal intentionality and the narrowness of phenomenology jointly entail that there is kind of narrow intentional content (the kind we have dubbed phenomenal intentional content), pervasive in human life, such that any two creatures who are phenomenal duplicates must also have exactly similar intentional states vis-a-vis this kind of narrow content. Phenomenal intentional content is pervasive and narrow. Any adequate philosophical and scientific conception of mind should accommodate this conclusion.
5. Some Philosophical Morals
We now draw some morals from the preceding discussion, first about strong externalist theories of intentionality, second about how phenomenal intentional content is related to mental reference and to wide content, and third about the extent of the so-called “hard problem” of phenomenal consciousness.
5.1. Strong Externalist Theories of Mental Intentionality are Wrong
We certainly do not deny that there is such a thing as “wide content” in language and in thought. Important lessons have been learned from Kripke, Putnam, Burge, and others about the relevance of the external environment in contributing both to the meaning of certain terms in natural language (e.g., natural-kind terms like ‘water’) and to certain aspects of the content of thought (e.g., aspects of thought that one would express verbally by employing the term ‘water’). But Putnam’s famous slogan that “meaning ain’t in the head” is properly construed as asserting that not all meaning is in the head; it doesn’t begin to follow from this, or from the considerations adduced in support of it, that no meaning is in the head. We will return to the topic of wide content presently.
However, a number of current theories of mental intentionality are strongly externalist: they assert that all intentionality is grounded in causal connections between states of the cognitive system and states of the external world; there can be no intentionality without some suitable kind of actual connection between what is going on in the head and the wider environment. Strong externalist theories of intentionality include (i) causal theories of content that find the necessary connection in the causal antecedents of the state, (ii) covariational theories that find the connection in certain kinds of systematic correlations between occurrences of an internal state and occurrences of an external state of affairs, (iii) teleosemantic theories that look to environmentally situated proper functions that certain internal states possess in virtue of evolutionary design, and (iv) learning-based theories that invoke internal adaptational changes in the creature’s own history.
Given our conclusions in sections 1-4, it follows that strong externalist theories of intentionality are wrong. They are not just slightly wrong, not just wrong in detail. Rather, they are fundamentally mistaken, because they claim to naturalize the entire phenomenon of mental intentionality and yet there is a rich and pervasive kind of narrow intentionality—phenomenal intentionality—that is constitutively independent of external factors. Strong externalist theories therefore cannot successfully naturalize the full phenomenon of mental intentionality, because they utterly fail to aim at one crucial aspect of it. Ideas employed by strong externalists might still have a useful role to play, however, in contributing to philosophical understanding of phenomena like wide content and mental reference, topics to which we now turn.
5.2. Phenomenal Intentionality, Mental Reference, and Wide Content
Suppose Alfred and Bertrand are looking at two different barns, and each of them says, “That’s an old barn.” Do their statements have the same truth conditions? Yes and no. In one way, they have different truth conditions. Alfred’s statement is made true or false by the age of the barn that he is looking at, while Bertrand’s statement is made true or false by the age of the distinct barn that he is looking at. Following recent usage, we will call such truth conditions, which depend on the actual entities referred to in a statement or thought, wide truth conditions. But in another way, Alfred’s and Bertrand’s statements have the same truth conditions. In each case the truth condition is that there must be an actual barn that he is looking at (and not, for example a papier-mache mock up of a barn, or only the facing side of a stage “barn” on a movie set), and that barn must be old. Such truth conditions are narrow truth conditions. They are determined skin-in, so to speak, and are completely determined by phenomenology. In our view, the situation is similar with respect to phenomenologically identical intentional states shared by phenomenal duplicates.
In section 3 we discussed belief-wise acceptance of the deliverances of perceptual experience. Such acceptance is the normal, default attitude. But it can be cancelled. If you have a lump on a finger, then objects that are smooth and flat will feel as though they have a lump where that finger touches them. But you soon learn not to believe that the object is lumpy. There is similar phenomenology of acceptance concerning propositional attitudes. There is a relevant phenomenal difference, for instance, between these two states: (i) believing that Bill Clinton was U.S. President, and (ii) the state you are in when you say (without believing) that Santa Claus brings presents. The salient difference turns on the fact that the phenomenal character of the first state includes the what-it’s-like of accepting the existence of Bill Clinton, whereas the phenomenal character of the second state includes the what-it’s-like of believing that Santa Claus does not exist. Similarly, suppose you hope or fear that an object of a certain description will be found. There is a clear phenomenal difference between the case in which you know full well that there is such an object and the case in which you does not know whether or not there is such an object.
Phenomenal intentional content presents to consciousness an apparent world that goes far beyond what one is conscious of at any one time; presuming so is itself an aspect of the overall phenomenal character of experience. Phenomenal intentionality thereby determines a complex set of presuppositions concerning the existence of, the persistence of, and various features of, the sorts of entities presented in experience: presuppositions about individuals (including flora, fauna, and other creatures like yourself), kinds, properties, relations, processes, and events of that world. For reasons that will become clear, we call these presuppositions grounding presuppositions. They have the phenomenology of acceptance discussed in the previous paragraph. In making a grounding presupposition, one takes it that there really exists an entity of a certain sort; and normally, one also presupposes that the (putative) entity in question has certain specific attributes. If there is an actual entity satisfying that presupposition (or satisfying it near enough), then one’s thoughts that are intentionally directed toward such a putative entity will refer to the actual entity in question; and so the properties of the satisfier will determine whether the beliefs about it are true or false, whether hopes and desires about it are satisfied, and so forth. Thus, wide truth conditions for those beliefs are determined by phenomenal intentionality plus the actual satisfiers of the relevant presuppositions. However, what it takes to be a satisfier of the presuppositions is determined by phenomenal intentionality alone. So, when these presuppositions are included in truth conditions, we get narrow truth conditions that are thereby determined solely by phenomenal intentionality.
Consider, for example, thoughts about individuals. You, your Twin Earth doppelganger, and your Cartesian duplicate all have certain phenomenologically identical thoughts that you each take to be about a person named “Bill Clinton.” Hence these thoughts presuppose the existence of such a person. Your own thoughts are about the actual Bill Clinton. Your Twin Earth doppelganger’s thoughts are about a different person on Twin Earth. You and your Twin Earth doppelganger have thoughts about different individuals, of course, because what a person’s thoughts are about—or refer to—depends not only on phenomenal intentional content, but also on certain relations between the thinker and the thing the thought is about. Your Cartesian duplicate also has thoughts that purport to be about a person named “Bill Clinton,” but since the Cartesian duplicate has not been in the right sort of relations to any such person, the Cartesian duplicate’s thoughts are not about anyone—they lack reference. Referring to something, mentally or linguistically, requires appropriate relations to that thing; but having thoughts that are intentionally directed toward such a thing—thoughts purporting to refer to such a thing—does not.
Straightforwardly, your thoughts about Bill Clinton are made true or false by facts about Bill Clinton, and your Twin Earth doppelganger’s phenomenologically identical thoughts are made true or false by facts about the person who satisfies your duplicate’s corresponding presupposition. There is no person who satisfies your Cartesian duplicate’s corresponding presupposition, so there is nothing that can be a truth maker for its thought that would be expressed by, say, “Bill Clinton is a womanizer.”
The differing truth conditions just mentioned are wide truth conditions. But again, there are two ways of thinking about the truth conditions of the phenomenologically identical thoughts of you and your duplicates. In one way, the truth conditions depend upon what is actually referred to (if anything) in those thoughts; this makes them “wide”. But in another and more fundamental way, the truth conditions are narrow, because what can be referred to in those thoughts is determined by phenomenal intentionality—in particular, by the phenomenally given grounding presuppositions. The thought will have wide content only if something in the thinker’s environment satisfies the phenomenal intentional grounding presuppositions of that thought. That is, wide content is grounded by phenomenologically determined presuppositions, which are an aspect of phenomenal intentionality and hence are narrow.
As a consequence, the strong externalist theories of intentionality discussed in the previous subsection are not wrong just because they leave out a kind of intentionality—viz, phenomenal intentionality. They are wrong because what they leave out is the fundamental kind of intentionality: the narrow, phenomenal kind that is a prerequisite for wide content and wide truth conditions. Because narrow phenomenal content determines wide content, an adequate account of wide content requires a prior account of narrow content.
Because of relevant similarities between singular reference and natural-kind categories, similar observations can be made concerning the narrow and wide truth conditions of thoughts about natural kinds. You, your Twin Earth doppelganger, and your Cartesian duplicate all have phenomenally identical thoughts with the same narrow truth conditions. For all three of you, these thoughts are intentionally directed toward certain small, common furry critters that meow, rub legs, drink milk, etc. For all three of you, these thoughts have the grounding presupposition that there is a natural kind of which these critters are members. But because of differences concerning the satisfiers (if any) of the common grounding presuppositions, these phenomenally identical thoughts have different wide truth conditions. Your own thoughts are made true or false by cats; your Twin Earth doppelganger’s phenomenologically identical thoughts are made true or false by cat-like critters of the kind that she or he and others in her or his community have encountered. Suppose that Putnam’s story in which cats are spy robots controlled from Mars is true concerning Twin Earth: the critters called “cats” on Twin Earth are robots controlled from Twin Mars. Then your belief that cats are animals is true; your Twin Earth doppelganger’s corresponding belief is false. That is, there are wide truth conditions for these thoughts that are partially determined by features in the environment that may be unknown to the thinker. But again: these wide truth conditions, differing as they do for your thoughts and your Twin Earth doppelganger’s phenomenally identical thoughts, are grounded on shared narrow truth conditions.
Your Cartesian duplicate has thoughts that are phenomenally identical to your cat-thoughts, and that have the same narrow truth conditions as yours do. Your Cartesian duplicate’s thoughts, like yours, are intentionally directed toward—and thus presuppose—small furry critters of a certain kind. But there are no such critters that the Cartesian duplicate has encountered, directly or indirectly, so there is no kind to which the Cartesian duplicate’s thoughts refer. This being so, those thoughts do not have wide truth conditions. So the Cartesian duplicate’s thoughts that are phenomenologically identical to your own cat-thoughts lack the kind of wide content that your own thoughts possess and your Twin Earth doppelganger’s thoughts also possess.
5.3. The Whole Hard Problem
We are among those who believe that what David Chalmers (1995, 1996) has called “the hard problem” of phenomenal consciousness is indeed a very hard problem. This is the problem of explaining why it should be that such and such mental state should be like this—that is, why it should have the specific what-it’s-like aspect it does, rather than having some other phenomenal character or having none at all. Presumably what it is like for you to undergo a particular mental state depends nomologically on what is going on in your brain, inside of the transducers. But why what depends on this brain process should be like this, rather than being some other way or being no way at all, seems inexplicable. Standard materialistic treatments of phenomenal consciousness in philosophy and in cognitive science do not close this “explanatory gap” (as it is dubbed by Joseph Levine 1983); rather, they appear to just leave out the intrinsic what-it’s-like aspect of mentality.
In our view, the hard problem is a very pressing and very puzzling conundrum. But its scope is considerably broader than it has often been thought to be. If separatism were correct—i.e., if phenomenology were indeed non-intentional, and intentionality were indeed non-phenomenal—then the hard problem would be limited to the what-it-is-like of non-intentional sensory experience, and would not infect the intentional aspects of mentality. Indeed, discussions of the hard problem often presuppose separatism. But the whole hard problem incorporates phenomenal intentionality. Phenomenal consciousness is intentional through and through.
This adds a dimension to the hard problem that often goes unrecognized. Conscious intentional states are intrinsically, by their very nature, directed toward whatever they are directed toward. Thus, the hard problem includes this: why should a mental state that is grounded in this physical or physical/functional state be by its intrinsic phenomenal nature directed in this precise manner? And this is a very hard problem indeed.
 This paper is thoroughly co-authored; order of authorship is alphabetical.
 For example, Jaegwon Kim (forthcoming) argues, in a way that firmly presupposes the separatist framework, that we can get more than half way to naturalizing the mental, but not all the way. The leading idea is that the intentional aspects of mentality can be naturalized via “functionalization,” but that the phenomenal aspects resist naturalization.
 For instance, Dretske (1995), Tye (1995, 1999), Lycan (1996), and Carruthers (2000).
 Another serious difference is worth mentioning. Representationalists typically regard the problem of naturalizing intentionality as tractable, and they seek to bring phenomenal character within the purview of this putatively tractable naturalization project by construing it as a species of intentionality. We, on the other hand, regard the problem of naturalizing phenomenal character as presently intractable, and we maintain that because of the interpenetration of phenomenology and intentionality, the scope of this intractability includes mental intentionality. Our reasons for holding this view will emerge as this paper progresses.
 For several recent treatments of the relation between phenomenology and intentionality that are similar in spirit to what we say here, see the discussions of Flanagan (1992), Strawson (1994), Siewert (1998), and Loar (forthcoming).
 This formulation is more accurate that the usual, “there is something that it is like to see red” because, we take it, the something-that-it-is-like that is referred to is the something-that-it-is-like of the seen not of the seeing—although, of course, there is something that it is like to see, as opposed, e.g., to hear or imagine.
 See Thomas Reid, Inquiry into the Human Mind on the Principles of Common Sense, Chapter 5.
 For a recent exception, see Lamb (2001).
 See Van Gelder (1999), Varela (1999) for attempts to account for the temporal thickness of experience in cognitive science terms.
 We think that the right view of these matters is at least very close to that expressed by Laird Addis (1989):
The idea of the mental is exhausted in all interesting aspects by (1) states of consciousness [primary mental entities]; (2) sensations, emotions, and perception-related entities [secondary mental entities]; and (3) dispositional mental states [tertiary mental entities—i.e., beliefs, etc., that are not presently occurrent]…. [O]nly states of consciousness are literally intentional entities. On the other hand, what makes the secondary and tertiary mental entities mental is their relation…to states of consciousness. Secondary mental entities cannot exist except as objects of states of consciousness…. (p. 7)
Thus, sensations and sensory qualities exist as, and only as, objects of conscious intentional states. This is essentially Brentano’s view in the famous—but widely unread—chapter of Psychology from an Empirical Standpoint in which he introduces the word ‘intentional’ and distinguishes between mental phenomena and physical phenomena. Brentano’s mental phenomena are Addis’ primary mental entities. Brentano’s physical phenomena are Addis’ secondary mental entities, not physical things.
 We do not know exactly when the phrase, “the intentionality of consciousness” first appeared, but one does well to remember that this was the phrase that characterized issues concerning intentionality early in the twentieth century.
 We are talking about psychological changes discernible to the experiencer—not changes such as the experiencer’s being unknowingly transported to Twin Earth and gradually coming to have ‘water’-thoughts about XYZ that are internally indistinguishable from earlier ‘water’-thoughts that were about H20.
 See Ross (1992) for a nice discussion, congenial to what we are saying here, that focuses not on Quineian indeterminacy but on the “Kripkenstein” thesis (set forth in Kripke 1982) that there is no fact of the matter whether the symbol ‘+’ means plus or quus.
 Part of what makes this aspect of phenomenology essentially a what-it’s-like of holding before one’s mind a specific intentional content is that semantic evaluability is involved: specific truth conditions are attached to it. We expand on this important point in section 3.
 For an excellent discussion that is complementary to what we will say here and that has considerably more detail about phenomenal intentionality, see Siewert (1998), Chapters 7 and 8. For another admirable and pertinent discussion, also complementary to ours, see Silverberg (1995); although phenomenological considerations are less prominent in Silverberg’s discussion, note his emphasis on how things seem, for each member of a group of beings who in effect are phenomenal duplicates respectively inhabiting a variety of different Twin Earthly environments.
 Siewert (1998) emphasizes the idea that a creature’s intentional features are ones for which the creature is assessible for accuracy, and he argues in detail that both perceptual and cognitive experiences are assessable for accuracy by virtue of their phenomenology.
 We are not here presupposing verificationism, or “procedural semantics,” or anything of the sort. The point is that differences in sensory-phenomenal content normally are reflected by differences in confirmation/disconfirmation procedures. Thus, sameness of confirmation/disconfirmation procedures provides strong evidence for sameness of content (even though it certainly does not constitute sameness of content).
 Of course, even if all these first-person tests for accuracy are successfully passed, this does not guarantee that the sensory-phenomenal experience one is testing really is accurate; experiential warrant does not guarantee truth. Also, the reason we talk of seeming to oneself to be performing tests and observing outcomes is that a given phenomenal duplicate of yourself might be one whose experiences—including its accuracy-assessment experiences—are systematically non-veridical. This would be so, for instance, for a phenomenal duplicate who is a brain in a vat.
 There are various further default assumptions involving the intentional objects of perceptual experience that would be psychologically operative in your phenomenal duplicate—normally automatically, as a matter of course—just as they are in you. Examples include default assumptions about experientially presented objects: that these objects have back sides that are not directly presented; that they persist when they are temporarily obscured from view by interposed objects or when one has the experience of looking away from them; that they normally persist when they cease to be presented in experience; and so on.
 Of course one can also think of the truth conditions of these states as involving the actual, and different, pictures referred to phenomenologically by you and your duplicate, if there are such pictures. There are two kinds of truth conditions for propositional attitudes, just as there are for perceptual states. Truth conditions of one kind are determined phenomenologically, as we have been discussing in the text. And truth conditions of the other kind incorporate the respective objects or kinds (if any) in your own and your phenomenal duplicate’s respective ambient environments that are the respective satisfiers of the referring concepts in the respective, phenomenologically identical, thoughts. We explore the relationship between these two kinds of truth conditions in section 5.2.
 The remarks in the preceding footnote apply to these propositional-attitude states too, and to the various components of the relevant web of belief.
 The fact that there is something epistemically special about first-person introspective access to the phenomenal character of experience is, of course, the basis for the kind of reflective inquiry often called “phenomenology.” But being epistemically special does not make such introspective judgments infallible. For insightful discussion of this complex issue, see Chapters 1 and 2 of Siewert (1998).
 Moreover, phenomenology does not depend constitutively on anything beyond phenomenology itself. (Of course, phenomenal character presumably does depend nomically on certain states whose nature is describable in non-phenomenological language—in humans, certain brain states.) In this sense, phenomenal character is intrinsic. We submit that the intrinsicness of phenomenology (as thus understood) is self-evident to reflective introspection. What-it’s-like is what phenomenal consciousness is. And what-it’s-like is what-it’s-like no matter what is going on outside of phenomenology itself.
 At the scientific level, this means that cognitive science should construe the mind as a “control system” for effectively operating a potential body in a potential world, regardless of whether or not the control system is actually embodied or en-worlded in the kind of body and world for which its functional architecture is appropriate. This is a common view among cognitive scientists themselves. Such a scientific enterprise is important and tractable, even though it presupposes intentionality rather than explaining it, and even though it does not address the so-called “hard problem” of phenomenal consciousness (cf. section 5.3 below).
 See Stich and Warfield (1994) for a representative sample of such theories. Although the approach to mental intentionality advocated in Fodor (1987) did acknowledge narrow content, and hence was not a form of strong externalism, it also reflected Fodor’s separatism about the phenomenal and intentional aspects of mentality. Because of this, by our lights his construal of narrow content was insufficiently robust, and was a step down the garden path toward strong externalism. He has since gone further down that path; see Fodor (1990, 1994, 1998).
 Our distinction between narrow and wide truth conditions has some kinship to the approach of so-called two-dimensional modal semantics, which also posits two kinds of truth conditions—one kind narrow and the other kind wide. See Davies and Humberstone (1980), Chalmers (1996 section 2.4) especially pp. 63-5, Jackson (1998 chapters 2 and 3) especially pp. 75-77, and Chalmers (this volume).
 Here and below we talk about thoughts for ease of exposition, but our remarks will apply to occurrent propositional attitudes in general. If one person doubts what another believes, then their propositional attitudes have the same truth conditions—the truth conditions for what is believed by one and doubted by the other. Similarly for other propositional attitudes.
 Your Cartesian duplicate is an exact phenomenal duplicate of you that is in the First Meditation situation, thoroughly deluded. You do not really have an exact phenomenal duplicate on Twin Earth, however, because on Earth people sometimes have the occurrent thought that water is H2O, whereas on Twin Earth they have instead the thought that water is XYZ. But we will use the useful term “Twin Earth doppelganger” for a person who is as much like you as is consistent with this difference.
 It is very common in philosophy of mind to gloss the “intentional directedness” constitutive of intentionality by saying that intentional states have aboutness. (We did so ourselves in the opening paragraph of this paper.) But the word ‘about’ also is often used to express the relation of reference—as we do in the paragraph to which this note is appended, and as we will continue to do below. Both uses can be appropriate in context, but it is important not to conflate them.
 There is a longstanding dispute over whether in such a case we should say that the thought is false or that it merely lacks a truth value, but this dispute does not affect the issues we are concerned with in this paper.
 For a discussion of both similarities and differences between singular reference and natural-kind categories, see Tienson (1986).
 We ourselves have pressed this concern recently; see Graham and Horgan (2000), and Horgan and Tienson (in press).
 Chalmers himself, however, has not presupposed separatism and has left open the question of which aspects of mentality are by their nature phenomenal—and in particular, whether this is true for intentional states like occurrent beliefs and desires. See, for instance, section 3.3 of Chalmers (1996), especially pp. 19-22.
 In “Philosophy as Rigorous Science” Husserl criticizes naturalists for holding that intentional states have natures. Conscious intentional states have essences, he says; they are essentially directed toward what they are directed toward. This is, in effect, the point we are making. When we say that intentional states are directed by their very nature toward what they are directed toward, we do not mean that intentional states have natures in the way in which chemical and physical, and perhaps biological, kinds have natures. Thus, we concur with Husserl’s point, although we do not adopt his terminology.
 We thank William Lycan, Michael Lynch, Brian McLaughlin, Steve Tammelleo, Mark Timmons, and audiences at the University of Arizona and the 2000 Society for Philosophy and Psychology for comments and discussion. Special thanks to David Chalmers and George Graham for their extensive and especially valuable help.