In ‘The Refutation of Idealism’, G.E.Moore observed that, "when we try to introspect the sensation of blue, all we can see is the blue: the other element is as if it were diaphanous" (1922; p.25). Many philosophers, but Gilbert Harman (1990, 1996) in particular, have suggested that this observation forms the basis of an argument against qualia, usually called the argument from diaphanousness or transparency. But even its friends concede that it is none too clear what the argument from diaphanousness—as I will call it—is (Tye 2000; p.45). The purpose of this paper is to formulate the argument, and to assess its merits. My conclusion will be that qualia realists have little to fear from the argument—provided both qualia and diaphanousness are properly understood.
I begin by making a number of preliminary points about the contemporary background of the argument, its different versions, and the proper understanding of its target.
2.1. The Background Debate.
Suppose I am looking at the gray filing cabinet in the corner of my office. This is an experience I am having—the experience of looking at a gray filing cabinet. And we can all agree that there is something it is like for me to have this experience—some phenomenal character of the experience, as it is usually put. Now, what makes it the case that my experience has this character? According to one group of philosophers, qualia realists, to explain phenomenal character one must postulate or assume particular qualities of the experience, qualia. These qualities are taken to be intrinsic to the experience, directly accessible to introspection, and, in some versions of the view, non-functional, non-intentional, and maybe even ineffable, primitive, and non-physical as well. On the other hand, according to another group of philosophers, intentionalists, what explains the phenomenal character of experience is simply the intentionality of experience, or perhaps the intentionality of experience combined with its distinctive functional role.
What I have just said is one way of characterizing the contemporary debate about qualia—at any rate one contemporary debate about qualia—a debate which has attracted a lot of attention in recent philosophy of mind. It is this debate that forms the background of the recent discussion of the argument from diaphanousness. As usually presented, the argument is supposed to tell against qualia realism, and in favor of intentionalism.
2.2. Remarks on Qualia.
While it is simple enough on the surface, the contemporary debate about qualia is in fact rather complex. One sort of complexity derives from the evident obscurity in the notion of qualia. As we have seen, all parties to the discussion agree that experiences have phenomenal character, and that we can sensibly talk about these characters without prejudicing their nature. So one might suggest that qualia simply are these phenomenal characters. But of course, in this neutral sense there is little doubt that there are such properties, and hence there is no hope that the argument from diaphanousness could tell us otherwise. In effect, then, the argument presupposes that we can articulate a more loaded sense of the notion. In this paper, therefore, we will use ‘qualia’ for this loaded notion, and use ‘phenomenal character’ for the less loaded or neutral notion.
The problem is that it is not easy to see how to expound this loaded notion in a way which would command widespread assent. Certainly different traditions tend to use the word ‘qualia’ differently. In Australia, for example, it is common to hear that ‘qualia’ are by definition non-physical. In America, it is common to deny this, and emphasize only that qualia are non-functional or non-intentional properties of experience. In addition, according to some views, qualia are thought of as theoretical features which are postulated to explain an agreed-on fact, the fact of phenomenal character. According to other views, qualia are thought of simply as phenomenal characters of a certain special sort.
In order to impose some order on this rather confusing situation, I will proceed here by stipulation. As I will use the term, qualia are mental or psychological properties of experiences which satisfy at least the following two conditions:
intrinsicness—roughly, the condition that a property satisfies when it is intrinsic to experiences; and
direct awareness—roughly, the condition that a property satisfies when it is such that if one’s experience has it, one in a position to apprehend this directly by introspection.
In accordance with this stipulation, qualia realists are committed at least to the view that experiences have mental properties which satisfy both intrinsicness and direct awareness—qualia precisely are such properties. The argument from diaphanousness is then an argument which tells us that there are no such properties.
One might be concerned that this conception of qualia is too weak. As we just noted, often the notion is thought to include rather more than simply this. And in any case, might someone not hold that there are mental properties of experiences which satisfy these conditions and deny qualia? However, while these concerns are perfectly legitimate, I think we can sidestep them. As I have said, the argument from diaphanousness would tell us if successful that there are no mental properties of experience which satisfy these two conditions. Given our assumption that these two conditions are necessary for some range of properties to qualify as qualia, it trivially follows that the argument if successful would establish that there are no qualia. But it is plain that, from this point of view, it doesn’t matter if there are further features involved in the notion of qualia. Nor will harm be done if the argument tells us more generally that there are no mental properties of other sorts which satisfy these conditions. That would only show that the argument is wider in scope than one might initially imagine.
Alternatively, one might be concerned that the conception of qualia with which I am operating is too strong. For example, Sydney Shoemaker (1996) presents a version of qualia realism that, as I understand it, denies direct awareness. However, while Shoemaker’s version of qualia realism is an important one which deserves extensive discussion, there are three reasons why I want to set it aside here. First, as Shoemaker himself emphasizes, his account of qualia is unorthodox in that by far the dominant conception in the literature is one which accepts direct awareness. Second, Shoemaker is explicitly responding to the argument from diaphanousness. According to him, the way in which a qualia realist should react to the argument is by revising the conception of qualia in such a way that one rejects direct awareness—but this provides further evidence that the conception of qualia that is targeted by the argument from diaphanousness is the one I have set out. Finally, it is worth emphasizing that, however one uses the word ‘qualia’, it remains a substantive issue whether there are properties of experience which are intrinsic and of which we are directly aware. It is this substantive issue which I take to be mainly at issue in the following discussion.
2.3. Two Versions of the Argument.
One sort of complexity in the debate about qualia prompts a certain regimentation of the notion. Another sort prompts us to distinguish two versions of the argument from diaphanousness (Martin 2001). In one version, the argument is a negative argument whose goal is to establish that qualia realism is mistaken. In another version, the argument is a positive argument whose goal is to establish that intentionalism is true.
Now, if the contrast between qualia realism and intentionalism were straightforward this would of course be a distinction without a difference: any argument which would establish the falsity of qualia realism would ipso facto establish the truth of intentionalism. But unfortunately the contrast is anything but straightforward. Qualia realism is in the first instance an ontological thesis to the effect that there are properties of experience which satisfy a certain job description. But intentionalism per se does not seem to take a stand on this ontological issue. All intentionalism says is that the intentionality of experience explains its phenomenal character. So there would appear to be room for versions of intentionalism that are compatible with, or even entail, qualia realism.
To illustrate the problem, consider projectivism about color. According to projectivism, when I look at the gray filing cabinet, my experience represents the cabinet as having a particular quale G, and this very quale is also instantiated by my experience. So, on this view, the quale enters the picture twice: it qualifies the experience, and the experience represents it as qualifying something else, viz., the filing cabinet. (According to projectivism, this quale just is grayness—but we can ignore that aspect of the position.) In addition, on at least some versions of projectivism, what makes it the case that my experience has the content it does is that it has the qualia it does. This version of projectivism is clearly a version of qualia realism. But it is also a version of intentionalism. For it is true on this view that the phenomenal character of my experience is explained by the intentionality of the experience. It is simply that the intentionality of the experience is explained by its qualia!
Perhaps the core idea of intentionalism—that the intentionality of experience explains its phenomenal character—can be developed or extended in such a way that the contrast between it and qualia realism will be made more precise. But whether that can be done or not, from the point of view of an examination of the argument from diaphanousness, it seems best to adopt the assumption that there are at least two versions of the argument. In consequence, I will concentrate here on the first version of the argument—the one that attacks qualia realism—and postpone discussion of the second version until §7.
2.4. A Puzzle about Moore.
The suggestion that the argument from diaphanousness can or should be construed as an argument against qualia, however, raises a further puzzle which must be confronted before we proceed any further. The puzzle starts from the observation that the Moore to whom the insight concerning diaphanousness is credited is the very same Moore who famously defended the sense-datum theory. But—one might say—the sense-datum theory is surely a version of qualia realism! So, barring the unlikely hypothesis that Moore was very confused indeed, it is difficult to see how the diaphanousness of experience is going to result in the denial of qualia.
There are a number of possible ways of responding to this puzzle. One is to question whether it is in fact so clear that sense-data theorists are qualia realists. In view of the difficulty of interpreting the central notions, none might well think that there is, or might be, a legitimate interpretative approach according to which this assumption should or could be denied. Another possibility is to suggest that, despite what one might initially think, a sense-data theorist such as Moore does face a problem of diaphanousness after all. Martin (2001) for example, points out that, according to one way of spelling out the idea behind the diaphanousness argument, experience presents us with a series of mind-independent properties. He goes on to say that a sense-data theorist must deny this aspect of experience and to that extent presents a revisionist conception of experience.
It may be true that suggestions of this sort can be developed. However, rather than investigate further the notions of qualia, sense-data and mind-independence, I want here to finesse the puzzle about Moore rather than confront it directly. For it is possible to distinguish two sorts of qualia realist—a sort typified by Moore and the sense-datum theorist, and a sort typified by contemporary philosophers. For our purposes, it is reasonable to assume that only the second sort is being targeted by the argument.
The distinction between the two sort of qualia realists emerges when we consider what I will call the relational thesis. According to the relational thesis, the phenomenal character of an experience is wholly determined by the objects that one is related to in having the experience. A wide variety of otherwise different approaches to experience might endorse the relational thesis. According to what is sometimes called the naïve approach, for example, in having an experience such as the experience of looking a filing cabinet, one is directly related to the filing cabinet itself. Against the background of this naïve approach, the relational thesis tells us that the phenomenal character of the experience is determined by features of the filing cabinet. According to intentionalism, by contrast, to have an experience is in effect to stand in a relation to some intentional object—say a property or a proposition. Against the background of that approach, the relational thesis tells us that the phenomenal character of the experience is determined by features of the proposition or property that is the intentional object of the experience.
Now, one sort of qualia realist, the sort typified by the sense datum theorist, holds the relational thesis. According to the sense datum theorist, when I look at a gray filing cabinet, what I look at or am acquainted with in the first instance is a mental object which is both gray and filing-cabinet shaped. The properties of this object, according to sense-datum theorists, determine the phenomenal character of my experience. Indeed, on at least one well-established tradition, the qualities of this mental object are the paradigm examples of what qualia are supposed to be. It is clear then that the sense-datum theory endorses the relational thesis. But the sense-datum theory is also a version of qualia realism.
On the other hand, another sort of qualia realist denies the relational thesis. One example of such a position emerges if we contrast the sense-datum theory with its traditional rival, adverbialism. Adverbialism denies the relational thesis since it denies that in having an experience one is related to an object of any sort. According to adverbialism, the experience of looking at a gray object is just the experience of sensing gray-ly, but to sense gray-ly is not to be related to anything. On the other hand, an adverbialist is perfectly entitled to combine this position with qualia realism—the experience of sensing gray-ly might perfectly well have properties which satisfy intrinsicness and direct awareness. So in adverbialism we have a position which might combine qualia realism with the denial of the relational thesis.
Now adverbialism is not a position that many contemporary philosophers are attracted to. But there is a version of qualia realism which agrees with the adverbialism in rejecting the relational thesis. I have in mind the position held at one time by Sydney Shoemaker (1981), and most recently by Ned Block (1990), according to which we sharply distinguish the intentionality of an experience from the qualia it instantiates—the intentional from the phenomenal content of the experience, as it is sometime put. On this sort of view—which we might call the Shoemaker-Block view—experiences generally have intentionality—that is what it means to say that experiences have intentional content. And in addition experiences have qualia, properties which satisfy intrinsicness and direct access and which explain or perhaps partially explain the fact that the experience has the phenomenal character it does. However—and this is the crucial fact for our purposes—these properties are not themselves properties of the objects that one is related to in having the experience. So here we have a second position which denies the relational thesis, but still maintains that there are qualia. Indeed, I think it is fair to say that it is this sort of qualia realist, rather than the sort typified by either Moore or by adverbialism, which is the dominant sort in contemporary philosophy. So by ‘qualia realist’ I will henceforth mean proponents of the Shoemaker-Block view, unless the context suggests otherwise.
How does the distinction I have just drawn— between those qualia realists who endorse the relational thesis and those that don’t—answer the puzzle about G.E.Moore? Well, as we will see presently, the argument from diaphanousness involves precisely the idea that while we are directly aware of objects that we are related to in having experiences we are not directly aware of those experiences themselves, or of the intrinsic features of such experiences. Such a position certainly does threaten the Shoemaker-Block view, since according to that view there are properties of experience of which we are directly aware. But an argument of this sort does not threaten, or at least does not seem to threaten, the sense-datum theorist, since, according to that view, it is not at all obvious that there are properties of experience of which we are directly aware. Instead of convicting Moore of confusion, therefore, I will for the most part limit the following discussion to the opposition between the argument from diaphanousness and the Shoemaker-Block view.
2.5. A Third Version of the Argument
Earlier we distinguished two versions of the argument from diaphanousness, one against qualia realism and other in favor of intentionalism. Having explicitly noted the relational thesis, however, it is natural to go further and suggest a third version of the argument. On this version, the argument is an argument for the relational thesis, or equivalently, an argument against the denial of that thesis.
The first thing to say is that construing the argument this way would certainly widen the scope of our discussion. For one thing, while it is true that there are qualia realists (e.g. Block) who deny the relational thesis, and qualia realists (e.g. Moore) who accept it, it is also true that there are qualia non-realists who deny the relational thesis. A qualia non-realist who rejects the relational thesis, for example, is Paul Churchland (1985). According to Churchland, or at least according to the Churchland of that famous paper, one might envisage a future direction of science in which neuroscientific knowledge becomes such an everyday part of life that one might simply introspect directly the brain states with which various sensory states are to be identified. One might of course question Churchland’s vision of future science, and also the extent to which science so envisioned would impact on our self-conception. However, for our purposes the important point is simply that Churchland is assuming that one might directly introspect intrinsic features of our mental states, and is also not assuming that those states have intentional objects of any sort. So, on at least this dimension, Churchland is to be grouped with Block and not with Harman when it comes to the relational thesis. And yet Churchland is a qualia non-realist.
Similarly, while it is true that there are both intentionalists and non-intentionalists who accept the relational thesis (e.g. Harman and Moore) it is also true that there are intentionalists who deny the relational thesis. An intentionalist who denies the relational thesis, for example, is Lycan (1996). Lycan holds a view which I will call soft intentionalism and which is to be contrasted with the more straightforward development of the position which I will call hard intentionalism. The hard intentionalist holds the relational thesis since, according to that view, in having an experience one bears an attitude to a proposition, and the phenomenal character of the experience is wholly determined by the content of the proposition. Soft intentionalism, by contrast, supposes that the phenomenal character of an experience is determined, not solely by features of the proposition one is related to in having the experience, but in addition by features of the experience itself, including in particular functional features. The more a soft intentionalist emphasizes the functional role of the experience in the determination of phenomenal character, the less likely the position is to endorse the relational thesis. And yet soft intentionalism is recognizable as a version of intentionalism.
So it seems that the relational thesis raises an issue that in a certain sense stands apart from the debate between intentionalism and qualia realism. Nevertheless, I think the suggestion that the argument from diaphanousness can be construed as an argument for the relational thesis is a good one. And this introduces a further complexity into our discussion. How should we deal with this complexity? As in the case of our distinction between the first two versions of the argument, I think an examination of the argument from diaphanousness has no choice but to consider this third version. Once again, however, to keep things manageable, my discussion here will remain focused on the question of whether the argument defeats qualia realism in the sense intended. I will return to this third version of the argument in §7.
3. The Argument
So far we have noted that there are three versions of the argument, and set the version we want mainly to discuss more sharply in focus. In this section I formulate the argument, and discuss which premise in it should properly be called the thesis of diaphanousness.
3.1 What it is
It is widely agreed that the argument from diaphanousness starts from the considerations presented both in the passage from Moore with which we began, and also in this famous passage from Gilbert Harman:
When Eloise sees a tree before her, the colors she experiences are all experienced as features of the tree and its surroundings. None of them are experienced as intrinsic features of her experience. Nor does she experience any features of anything as intrinsic features of her experiences. And that is true of you too…Look at a tree and try to turn your attention to intrinsic features of your visual experience. I predict that you will find that the only features there to turn your attention to will be features of the tree… (1990, p.667).
Both passages are certainly highly suggestive, but what is the argument they suggest exactly?
My proposal is that the argument against qualia implicit in these passages and their supporting texts might be presented as proceeding in three stages. At the first stage one makes a phenomenological point the formulation of which we can simply take over from Harman:
(1) Look at a tree and try to turn your attention to intrinsic features of your visual experience. I predict you will find that the only features there to turn your attention to will be features of the presented tree.
Harman is here both giving instructions (“look at a tree and try to turn your attention…”) and making a prediction about what we will find ourselves doing if and when we obey his instructions. Since both prediction and instruction are supposed to be presented as a neutral and non-committal description of the phenomena, we are so far only at the phenomenological stage of the argument.
At the next stage, one makes a theoretical point about introspection, which is intended to be grounded in the phenomenological point just presented, and perhaps is plausible also on its own terms. There are a number of ways to formulate this point, but I think a reasonable initial formulation is this:
(2) In introspection, one is or becomes aware of the intrinsic features of one’s experience by attending to the objects and properties represented by that experience.
Notice that (2) does not by itself prejudice the issue of whether there are qualia or not in the sense we have intended, that is, whether there are properties of experience which satisfy the job-description of qualia. It simply says that one apprehends the intrinsic features of one’s experience by doing something else.
In the final stage of the argument, one draws the moral from this theoretical point for the debate about qualia. Once again, there might be different ways to state this moral, but a reasonable formulation is this:
(3) In introspection, one is not directly aware of the intrinsic features of one’s experience.
The force of (3) is of course that qualia realism in the sense at issue is mistaken. If qualia realism is true, there must be intrinsic features of experiences of which we are directly aware in introspection. But, by (3), there are no such features.
In summary, the argument from diaphanousness takes us from (1) to (2) to (3), and (3) is something the qualia realist is committed to denying.
3.2 What is 'The Diaphanousness of Experience'?
The formulation of the argument from diaphanousness that I have just offered raises the question about which claim here should be identified with the thesis of the diaphanousness of experience.
On the one hand, one might think that this thesis should be identified with (1) or something like it. After all, the diaphanousness of experience is often thought to be grounded in, or perhaps just to be, a piece of phenomenology. On the other hand, one might think that the thesis should be identified with the ‘by’-claim summarized in (2), where (2) is, as we have noted, a theoretical claim about introspection and about what is involved in being or becoming aware of the intrinsic features of one’s experiences. Perhaps this choice is in some respects terminological, but in any event, so far as I can see the most natural interpretation is to treat (2) as the thesis of the diaphanousness of experience and I will talk for the most part in those terms.
Even if one can afford to be somewhat sanguine about which of (1) or (2) should be treated as the thesis of diaphanousness, it is important to note that one should not be sanguine, or at least should not be so sanguine, about whether (3) should be treated as the thesis of diaphanousness. For suppose that (3) were thought of as the thesis of diaphanousness. Then there would be no interesting question about how there could be an argument from diaphanousness to the denial of qualia realism. For (3) obviously impacts on qualia realism. On the other hand, as will emerge in due course, the crucial question I want to focus on is precisely whether there is a decent argument from the thesis of diaphanousness—that is, (2)—to the denial of qualia realism. On that assumption that (3) were the thesis of diaphanousness, therefore, the central question of our inquiry could not be raised.
Of course one might say that this is all to the good from the point of view of the argument. And perhaps there is some temptation in the literature to identify (3) with the thesis of diaphanousness. But this manoever simply conceals the issue rather than solving it. For the fact is that in most presentations of the argument people begin by making remarks like (1) or (2). The question which will concern us is how it is that these remarks bear on the issue of qualia realism, a question we have formulated in terms of an argument from the thesis of diaphanousness. Dubbing (3) the thesis of diaphanousness makes our question more difficult to express but it does not make it disappear.
3.3. An Alternative Formulation.
The importance of not treating (3) as the thesis of diaphanousness emerges in more detail when we contrast the formulation of the argument that I have offered with an alternative formulation. According to this alternative, the argument should be understood as proceeding, not from (2) to (3), but rather from (2) to (3*):
(3*) In introspection, one is only directly aware of the objects and properties represented by one’s experience.
(3*) is logically distinct from (3). (3*) presupposes, while (3) does not, that there are objects and properties represented by experience and that we are directly aware of them: it simply says that this is all we are aware of. In addition, (3) presupposes, while (3*) does not, that there are intrinsic features of experience: it simply says that we are not aware of them. So it seems that we do here have a genuinely different formulation of the argument: one proceeds from (2) to (3), and the other proceeds from (2) to (3*).
I think it should be agreed that this is an alternative, but it is a formuation I want set aside in what follows. While (3) and (3*) are certainly not logically equivalent, it is reasonable to assume that, in the context of our discussion, they amount to largely the same thing, or at least that one can relatively easily move back and forth between them. On the one hand, it seems reasonable to infer (3) from (3*) given the assumption that projectivism is false: if it is true that one is only directly aware of presented features, and presented features are not also intrinsic features, then one is not directly aware of intrinsic features. On the other hand, it seems reasonable to infer (3*) from (3) given the assumption that one is directly aware of something in having an experience: if one is not directly aware of intrinsic features, and is directly aware of something, then presumably one is directly aware of are presented features. But if that is true, then it seems reasonable to conclude that the distance between (3) and (3*) is not something over which we can make much hay. For this reason I will continue to focus on (3).
Even if (3*) is set aside however, it is important to notice that, as in the case of (3), one must resist the temptatation to identify (3*) with the thesis of diaphanousness. Our focus is on the question of how observations such as Harman’s about Eloise and the tree support (3*). And we have articulated that question by asking how the thesis of diaphanousness has an impact on qualia realism. On the assumption that (3*) were the thesis of diaphanousness, we would be obliged to reformulate our question, even if, as we have already noted, the question would certainly not disappear entirely.
4. Block’s Critique
In recent literature, the philosopher who has done most to criticize the argument from diaphanousness is Ned Block. In this section I set out his criticisms and explain why I think they fail. I will start by pointing out two claims which a proponent of it is not committed to. I will then argue that Block’s criticisms largely rely on mistakenly attributing these claims to a proponent of the argument.
First, a proponent of the argument is not committed to the denial of intrinsicness, the idea that there a psychological properties of experience which are intrinsic to the experience. The argument we just considered certainly involves the suggestion that there are no intrinsic properties which meet direct awareness, but intrinsicness itself is left untouched.
4.1.1. The suggestion that the argument does not involve denying intrinsicness might seem initially strange. Proponents of the argument certainly say things which on the surface suggest the opposite. Concerning Eloise, for example, Harman says that she does not "experience any features of anything as intrinsic features of her experiences" (1990, p. 667, emphasis added). The idea here, I think, is that when we spell out the ways in which Eloise takes the world to be in perception, one might say that she takes the world to contain a tree, with a certain shape and color, which occupies a position at such and such a distance from her and so on. But nowhere does one need to say that Eloise takes the world to contain objects which instantiate intrinsic properties which are also intrinsic properties of her experiences. In short, Eloise takes the world to be a certain way in perception but in spelling out that way one does not need to mention intrinsic features of her experiences.
However, the difficulty here is in moving from a claim of this form—which certainly has a good deal of plausibility to it—to any claim which seems to threaten qualia. One might say that the qualia realist is committed to the idea that qualia exist only if we are directly aware of them in perception as intrinsic features of objects. But put so baldly it is not at all clear why anyone would make such a claim. Similarly, one might say—Harman does say—that qualia realists are guilty of confounding intrinsic properties of objects represented by experience with intrinsic properties of experience. But even if this charge of confusion is true, it would only be an explanation of why people might falsely believe that qualia exist. It does not at all encourage the suggestion that belief in qualia is false to begin with. Finally, one might say that what these claims suggest is that one is directly aware only of intrinsic features of objects and that one is aware of the intrinsic features of one’s experiences by being aware of the intrinsic features of objects represented in experiences. But to say this is simply to embrace a version of the argument already presented.
4.1.2. There is also a second reason why it might seem initially strange that the argument from diaphanousness does not target intrinsicness. Proponents of the argument usually write from a functionalist or intentionalist perspective, and one might think that, from that perspective, it is just obvious that there are no intrinsic psychological features of experience.
The short answer to this is that there is a difference between what a proponent of the argument is committed to qua intentionalist or functionalist and what a proponent of the argument is committed to qua proponent of the argument. My point is only that the argument does not strictly speaking attack intrinsicness, not that people who propound the argument do not attack it as a matter of fact.
The longer answer is that there is in fact no reason at all why an intentionalist or functionalist must deny intrinsicness. The plausibility of the claim that there are psychological properties which are intrinsic to experience depends largely on what one thinks an experience is. On the one hand, one might identify an experience with what Shoemaker (1996) calls its ‘core realizer’. This is the particular or local neural state which, as Shoemaker says, comes and goes as the state comes and goes. Now, on this view, the intrinsicness condition is quite an implausible claim, and, moreover, it is so regardless of whether functionalism or intentionalism is true or not. For, on this conception of experience, to say that some psychological property of an experience is intrinsic to the experience is to say that this property is wholly determined by matters that are internal to its core realizer. But this is highly implausible. At any rate, to assert boldly of any psychological property of an experience that it is determined by matters internal to its core realizer is a wild empirical speculation, something that goes far beyond what anyone could reasonably claim at the moment.
Now, one response to this is to say that psychological properties are not intrinsic to experiences, i.e. to deny the intrinsicness condition. But a better response is to deny that experiences ought to be identified with their core realizers, and to look for an alternative. For there is an alternative available. The alternative is to identify an experience with what Shoemaker calls its ‘total realizer’. This is the total state of the brain or nervous system that one is in when one has a certain experience. On this view, to say that a property is intrinsic to an experience is to say that this feature of the experience is wholly determined by goings-on internal to the total realizer. But this claim is not at all implausible even if one is a functionalist or intentionalist. For this claim amounts to nothing more than the claim that someone who is in that same total or overall brain state will also have the experience—if you like, on this view, the idea that phenomenal character is intrinsic to experience is tantamount to the claim that phenomenal character is intrinsic to the subjects of that experience, rather than simply the experience. This is a claim that has certainly been denied in contemporary philosophy, but it is not one which off-hand is particularly implausible nor one that is in conflict with intentionalism or functionalism.
The moral is that once intrinsicness is to be understood in a way that makes it plausible—that is, when the background conception of experience identifies experiences with total realizers, rather than core realizers— it is not at all clear that functionalists or intentionalists must deny it. But that means—to return to our main line of argument—that there is no reason why a functionalist or intentionalist who endorses the argument from diaphanousness must deny intrinsicness.
4.2. Direct Awareness, not Awareness Tout Court.
Proponents of the argument from diaphanousness are not only not committed to denying intrinsicness, they are also not committed to denying that we are aware of the intrinsic features of our experience. The conclusion of the argument is that one is not directly aware of those features. But, if ‘direct’ is doing any work at all here, we cannot conclude from this that one is not aware of those features. To say that one is not directly aware of something leaves open the possibility that one is aware of it, but to say that one is not aware at all closes off that possibility.
The point that a proponent of the argument is not committed to the denial of awareness of intrinsic features is important when we consider the overall credibility of the argument. After all, if the argument did involve the suggestion that one is not aware in introspection of one’s experience, then it would be fair to say that it should be treated as presenting a paradox rather than as something to which one might look to decide between competing positions in philosophy of mind. For surely it is a datum, something on which everybody can agree, that one can be aware of one’s experience in introspection!
One might respond that this underestimates the radicalness of the proponent of the argument from diaphanousness. A proponent of that argument, one might think, is certainly committed to the claim that one is directly aware of the objects and properties presented in one’s experience. But this should not be interpreted in such a way that it is consistent with the claim that one is also indirectly aware of one’s experiences. Rather it should be interpreted in such a way that it entails or suggests that one is not aware of one’s experiences at all. Moreover, one might suggest, the ‘datum’ that one is aware of one’s experiences derives only from a confusion between the properties presented in one’s experience and the experiences themselves.
However, someone who adopts this radical course is in my view faced with rather serious difficulties. For contrast visual perception with visual imagination. If I were sufficiently adept at visual imagination, I might be able to conjure up in imagination a situation that is identical to the visual experience I am having right now. In other words, it seems quite possible to contrast the situation in which I am visually under the impression that p with a counterpart situation in which I only visually imagine that p. If it were really true that in introspection I am not aware, or could not become aware, of my experiences themselves, it would seem to follow that introspection alone could not allow me to discern these two situations. But surely I can tell the imagination case from the impression case on the basis of introspection alone. In the impression case, I am going to feel inclined to be believe that p. In the imagination case, on the other hand, I am not going to feel inclined to believe that p. The explanation for this, of course, is that perception and imagination have very different functional roles in the mental life of a person: to imagine something is an act of will in the way that to be visually aware of something is not. But, since the respective contents of perception and imagination are in this instance presumed to be the same, it follows that I must be aware, not simply of the intentional objects my experiences, but also of the experiences itself.
We will return to the contrast between imagination and perception at a later stage. For the moment the important point is simply that it is a datum that we are or can become aware of intrinsic aspects of our experience. And so the argument from diaphanousness could not possibly tell us otherwise.
4.3. Block’s Objections.
We are now in a position to demonstrate that Block’s three objections to the argument from diaphanousness misfire.
4.3.1. Block’s first objection is that the argument represents “an error in philosophical method”. He goes on: “Looking at a blue wall is an easy thing to do, but it is not easy (perhaps not possible) to answer on the basis of introspection alone the highly theoretical question of whether in so doing I am aware of intrinsic properties of my experience “(1990, p. 689). As I understand this objection, Block is suggesting that, whatever the argument from diaphanousness is exactly, its premise is based on phenomenology or introspection, and its conclusion is that one is not aware of intrinsic properties. His suggestion is that it is an error to infer from premises to conclusions of this sort.
Block may or may not be right that inferences like this are mistaken, but even so his objection does not touch the argument from diaphanousness. For, at least as I have set it out here, that argument does not conclude with the claim that one is not aware of the intrinsic properties of experience. For consider: we have seen that it is consistent with advancing the argument that one thinks that experiences have intrinsic properties. And we have also just seen that the argument does not attack the idea that we are aware of those properties if they exist and are instantiated. Putting these two points together, we derive the result that the argument does not involve, contrary to Block’s objection, the suggestion that one is not aware of intrinsic properties.
One might respond that the distinction between direct awareness and awareness tout court does not touch the basis of Block’s objection, because that objection can simply be reformulated to accommodate it. On this interpretation, Block’s objection is that we cannot decide on the basis of introspection alone the question of whether we are directly aware of the intrinsic properties of our experiences. However, the response to this version of the objection is that the argument does not ask you to decide this question “on the basis of introspection alone”. The only thing that is to be decided on the basis of introspection alone is the phenomenological point (1). But (1) is supposed to provide evidence for (2), which is not something that one can decide on the basis of introspection alone, but is rather a theoretical claim about introspection. On the other hand, you need (2) to get you to (3), the denial of direct awareness. The argument that Block is attacking effectively cuts out the middle man, that is, it is an argument that moves directly from (1) to (3). But that is evidently not the argument from diaphanousness as we have been discussing it here.
4.3.2 A similar criticism—that the position attacked is not the position defended—is also appropriate when we turn to a second objection mounted by Block against the argument:
Harman relies on the diaphanousness of perception (Moore, 1922), which may be defined as the claim that the effect of concentrating on experience is simply to attend to and be aware of what the experience is of. As a point about attention in one familiar circumstance—e.g., looking at a red tomato, this is certainly right. The more one concentrates on the experience, the more one attends to the redness of the tomato itself. But attention and awareness are distinct, and as a point about awareness, the diaphanousness claim is both straightforwardly wrong and misleading. One can be aware of what one is not attending to. For example, one might be involved in intense conversation while a jackhammer outside causes one to raise one’s voice without ever noticing or attending to the noise until someone comments on it—at which time one realizes that one was aware of it all along. Or consider the familiar experience of noticing that the refrigerator compressor has gone off and that one was aware of it for some time, even though one didn’t attend to it until it stopped (Block 2001 p. 7)
Block is certainly correct that there is a distinction between attention and awareness—indeed, we will come back to this distinction later on. Nevertheless, the objection which I take him to be making in this passage derives again from a mistake about what someone is committed to in advancing the argument from diaphanousness. Block is assuming, I think, that a proponent of the argument is committed to the claim that one is not aware of the intrinsic features of my experience. The point about awareness and attention certainly would defeat that position: from the fact that one does not attend (in the strict sense) to one’s experience, it does not follow that one is not aware of it. But as we have seen a proponent of the argument need not, or should not, adopt the position that one is not aware of the intrinsic features of one’s experiences. What is at issue is direct awareness, not awareness tout court.
4.3.3. Block’s third objection is that the step from (1) to (2) is fallacious. (1) might be true and also conform to (2), but there are plenty of examples of experiences which do not:
[c]lose your eyes in daylight and you may find that it is easy to attend to aspects of your experience. If all experiences that have visual phenomenology were of the sort one gets with one’s eyes closed while awake in daylight, I doubt that the thesis that one cannot attend to or be aware of one’s experience would be so popular. (Block 2001, p. 8).
The crucial fact about an example such as this—elsewhere Block also mentions orgasms and phosphenes—is that here we seem to have an experience which lacks (or might lack) intentionality. In these cases, Block says, experience is not diaphanous, that is, it is not the case that one is or becomes aware of the features of an experience by attending to the object and properties represented by the experience—for in this case there are no such objects and properties! It would thus seem that the step from (1) to (2) is fallacious. Of course, some experience are such that claims like (1) are true of them. Nevertheless, (2) is a massive overgeneralization from these isolated examples.
One thing to say about this objection is that it does not suffer from the sort of problem I identified with the previous two. Unlike the previous two objections, this objection really would attack the argument as I have formulated it. Nevertheless, there are two reasons why I want to set it aside.
First, even if Block is right about closing one’s eyes in daylight, orgasms and so on, Harman still seems to be right about experiences which uncontroversially have intentionality, such as experiences of color. But surely the qualia realist does not want to be maneuvered into the position of saying that color experiences lack qualia. It would be an odd sort of position indeed which postulates qualia but then adds that qualia are only instantiated in cases in which you face the sun with closed eyes, or else are in states of sexual climax!
Second, the question of whether experiences of the kind mentioned by Block lack intentionality is a controversial issue (cf. e.g. Tye 2000). Part of the general issue raised by the debate about qualia with which we began is precisely the question of whether there are sensations which have phenomenal character but lack intentionality—whether there is a distinction between sensation and perception as it is often put. Of course there is a tradition in which it is obvious that there is such a distinction, and arguably ordinary thought recognizes a distinction along these lines. But on the other hand, there have been persistent efforts in philosophy to undermine this distinction. There is first of all the Galilean tradition in which one treats perception as a species of sensation, a tradition partially revived and defended by the adverbialists of the twentieth century. Then there is the sense datum tradition which reduces both sensation and perception to something else – a mental act in which one is acquainted with a sense-datum. Finally there are contemporary intentionalists who try to show that both sensations and perceptions are species of propositional attitude.
In Block’s third objection we are in effect being asked to accept (a version of) the sensation/perception distinction and so reject the argument from diaphanousness. However, whatever is the ultimate truth about that distinction, this is unlikely to be a persuasive response to someone who advances the argument from diaphanousness. For such a person is very likely to rejected the sensation/perception distinction in the first place. So I think we need to leave Block’s third objection aside, and look elsewhere.
5. The Step from (2) to (3).
I have been concerned so far only to set out the argument from diaphanousness, to point out what one is not committed to in advancing it, and to defend it from some criticisms made by Block. I now turn to my own suggestion about what is wrong with the argument. My focus is the step from:
(2) In introspection, one is or becomes aware of the intrinsic features of one’s experience by attending to the objects and properties represented by that experience
(3) In introspection, one is not directly aware of the intrinsic features of one’s experience.
I will argue that this step is fallacious. Even if it is true that I apprehend the features of my experience by attending to the objects presented in that experience, it does not follow that I am not directly aware of those features. In fact, there are two ways of developing this objection. The first proceeds via a closer examination of the direct awareness condition; the second via a closer examination of the notion of attention.
5.1. The First way to Develop the Objection.
In thinking about the direct awareness condition, we need to focus on two questions. The directness question asks: what does it mean to say that one is directly aware of intrinsic features of experience? In other words, the directness question brackets the issue of what it is to be aware of the intrinsic properties of experience and asks instead what it is to be directly aware. The awareness question asks: what does it mean to say that one is directly aware of the intrinsic features of experience? In other words the awareness question brackets the issue of directness, and asks instead what it is to be aware of experience in the first place.
Now, as regards the directness question, I think it is plausible to operate with a rather schematic account of what it is to be directly aware of something, an account which may be extracted from classic discussions by William Alston on epistemic immediacy and Frank Jackson on perceptual immediacy (cf. Alston 1971, Jackson 1977). According to this account, one first defines what it is to be indirectly aware of something, as follows: S is indirectly aware of x just in case one is aware of x by being aware of y, where y is distinct from x. One then defines what it is to be directly aware of something as follows: S is directly aware of x just in case (a) S is aware of x; and (b) S is not (merely) indirectly aware of x. So to say that one is directly aware of experience is to say that one is aware of it but one is not aware of it by being aware of anything else.
As regards the awareness question, the crucial distinction to draw initially is one drawn by Fred Dretske between f-awareness, o-awareness and p-awareness (Dretske, 1999). Dretske introduces the distinction with the example of looking at a moving object. In such a case, one might be aware of the object itself—this is object or o-awareness. Or else one might be aware of the movement of the object—this is property or p-awareness. And finally one might be aware that the object is moving—this is fact or f-awareness. Similarly, in the case of experiences and their intrinsic properties, one might be aware of the experience itself—this is o-awareness. Or else one might be aware of the intrinsic properties of the experience—this is p-awareness. And finally one might be aware that the experience has those intrinsic properties—this is f-awareness. Dretske argues persuasively that these notions are logically independent: one might, for example, be f-aware that one’s experience e has property C, and be neither o-aware of e nor p-aware of C.
In the light of Dretske’s distinction among species of awareness, as well as the schematic account of directness, the bad news is that it now appears that there are no fewer than nine ways to interpret the direct awareness condition:
(4a) S is o-aware of experience e but not by being o-aware of anything else.
(4b) S is o-aware of experience e but not by being p-aware of anything else.
(4c) S is o-aware of experience e but not by being f-aware of anything else.
(4d) S is p-aware of the character C of experience e but not by being o-aware of anything else.
(4e) S is p-aware of the character C of experience e but not by being p-aware of anything else.
(4f) S is p-aware of the character C of experience e but not by being f-aware of anything else.
(4g) S is f-aware that experience e has character C but not by being o-aware of anything else.
(4h) S is f-aware that experience e has character C but not by being p-aware of anything else.
(4i) S is f-aware that experience e has character C but not by being f-aware of anything else.
On the other hand, the good news is that there is a decent case for supposing that it is interpretation (4i) which is the correct one for our purposes. In the first place, as Dretske points out, the usual way of explicating o-awareness or p-awareness is by assimilating them to certain perceptual relations. One is o-aware of the moving object just when one sees it, and one is p-aware of the movement of the object just when one sees its movement. But that suggests that the idea that we are o-aware of experiences or p-aware of properties of experience is simply the idea that a perceptual model of introspection is true, and moreover is true in a rather extreme form. On the other hand, if one wants to avoid the perceptual model, as most contemporary philosophers do, it would seem that the only option is to operate with f-awareness and to adopt (4g-i).  In the second place, of these three interpretations, it is really only (4i) that articulates a reasonable notion of direct awareness. To see this, consider the case in which I am aware of the fact that the filing cabinet is in my office by seeing either it or its shape. This is not the sort of case that is normally classified as a case of indirect awareness. Indeed, in the epistemological literature, cases such as this are the paradigmatic cases in which one has direct awareness of something. And yet both (4g) and (4h) represent cases in which I am or become aware of some fact but not by being o- or p-aware of something else. So it seems natural to set (4g-h) aside and operate only with (4i).
Now, however, it should be perfectly clear that the step from (2) to (3) is illegitimate. On our preferred interpretation, what (3) is telling us is that one is aware of the fact that one’s experience has certain intrinsic features by being aware of some other fact. But (2) provides no grounds at all for that claim. What (2) tells us is that one is or becomes f-aware of the intrinsic features of one’s experience by attending to properties and objects represented in that experience. But to say this is not to say that there is some other fact which is such that when one becomes f-aware of it, one becomes f-aware of those features. On the contrary, it is not to mention any other fact at all.
To illustrate, suppose I come to know or be f-aware of the fact that the filing cabinet is in my office by seeing or attending to the cabinet. As we have just seen, intuitively this is not the sort of case that one would describe as a case of indirect knowledge. On the contrary, this is usually thought of as a paradigm case of direct perceptual knowledge. And indeed, the schematic account of directness delivers this result. For in this case it is not the case that I am aware of one fact—the fact that the filing cabinet is in my office—by being aware of another. Nevertheless, this paradigm of direct knowledge is certainly one in which I know something by seeing or attending to something else—that is I know something about the cabinet by attending to it. So it seems that the idea that I am directly f-aware of something is perfectly compatible with the idea that I am f-aware of it by perceiving or attending to something else. By analogy, therefore, it is hard to see how the mere fact—assuming it to be a fact—that I am or become f-aware of the intrinsic features of my experience by attending to something else entails that here we are in the presence here of anything other than direct f-awareness.
Admittedly, there are points of disanalogy between the two cases. In the cabinet case, I am aware of the fact of the form a is F by being seeing or attending to o—in other words one is aware of a fact by attending to something which is (in an intuitive sense) a constituent of that very fact. I am aware that the cabinet is in the office by being aware of the cabinet. In the experience case, however, I am not aware of a fact of the form a is F by attending to something which is a constituent of that fact. Rather I am aware that my experience has some intrinsic features by being aware of something which is neither my experience nor a feature of that experience.
However, at least given the account of directness I have introduced so far, it is hard to see why this disanalogy matters. For directness in the sense at issue involves a negative claim, and this negative claim is something that both cases share. What is important for directness is the question of whether I come to know a certain fact by coming to know some other fact. If this is not the case, my awareness or knowledge counts as direct. But in neither the experience case nor the filing cabinet case has it been established that I come to know a certain fact by knowing some other fact. Thus it has not been established that (3) is true.
5.2. The Second Way to Develop the Objection.
The suggestion I have just made—that because of considerations of awareness the step from (2) to (3) is illegitimate—faces a number of different responses. Before considering them, however, I want to consider a different reason for supposing that the crucial inference is illegitimate. This reason brings out a further source of complexity in our discussion which I have so far been ignoring.
It is important to distinguish two meanings, or uses, of ‘attention’. In one sense ‘attention’ just means ‘to think about’—let us call this cognitive attention. This is the notion of attention that is at issue when, for example, one says ‘Let us now attend to the second flaw of the argument’ or ‘The program made us attend even more than we had before to the effects of salination on the nation’s rivers’. In another sense, however, ‘attention’ means in the first instance to focus on various items in one’s field of vision, and perhaps also in other sensory fields—let us call this perceptual attention. This is the notion at issue when one says ‘John didn’t notice the pedestrian crossing because he was attending to the spider on the windscreen’, or ‘The space-ship only appears when you focus your attention through the painting, rather than at the painting’.
Now, so far in our discussion, we have been assuming that the notion of attention in play is the perceptual notion. However, it is important also to notice that in the initial phenomenological premise of the diaphanousness argument the word ‘attention’ appears twice. Here is the premise again with emphasis added:
(1) Look at a tree and try to turn your attention to intrinsic features of your visual experience. I predict you will find that the only features there to turn your attention to will be features of the presented tree.
Now, I think it is clear that Harman intends the notion of attention to be univocal here. But that assumption creates a problem for the argument. Suppose that, in its first occurrence, ‘attention’ means perceptual attention. Then Harman’s instructions could only be carried out if one could perceptually attend to the intrinsic features of one’s visual experience. In turn, however, if one is in a position to perceptually attend to the intrinsic features of one’s visual experiences, it would seem that some version of the perceptual model of introspection is true. For to perceptually attend to something is at least to perceive it. This suggests that, unless the perceptual model of introspection is going to be assumed from the start, in its first occurrence, ‘attention’ ought to be interpreted as cognitive attention.
What then of the second occurrence of the word? If we continue with the assumption of univocality, we have no choice but to conclude that the second occurrence is likewise supposed to mean cognitive attention. But this doesn’t seem to be right. The most natural interpretation is that when Harman is making predictions about what will happen when we try to attend he is telling us that we will find ourselves perceptually attending to the tree. So Harman’s prediction is most naturally interpreted as involving perceptual and not cognitive attention.
So far then, we have failed to find a univocal interpretation of the initial premise of the argument. Should we then impose a non-univocal reading on the premise? Of course there is the textual inconvenience that Harman seems to want a univocal interpretation. But perhaps we should ignore this and say that what Harman should have said is summarized in (1a)
(1a) Look at a tree and try to turn your cognitive attention to the intrinsic features of your visual experience. I predict you will find that the only features there to turn your perceptual attention to will be features of the presented tree.
On this interpretation, Harman is instructing us to think about or consider the intrinsic features of our visual experience. And he is predicting that when we try to do this we will find ourselves attending—that is, perceptually attending—to features of the presented tree.
I think this is in fact the most plausible interpretation of the initial premise of the argument. However there is problem lurking here, and this is that once we allow this interpretation to filter through to the rest of the argument, it is plain once again that the inference from (2) to (3) is mistaken. For consider: if (1) is properly interpreted as (1a), likewise (2) is properly interpreted as (2a):
(2a) In introspection, one cognitively attends to (i.e. thinks about) the intrinsic features of one’s experience by (perceptually) attending to the objects and properties represented by that experience
But it is quite clear that (2a) will not result in (3):
(3) In introspection, one is not directly aware of the intrinsic features of one’s experience.
After all, what (2a) says—that is, what (2) says when you interpret it in the way indicated—is that in introspection one thinks about the phenomenal character by perceptually attending to something. But even if this is true, the inference to (3) now looks to be a non sequitur.
The point may again be illustrated by my experience of seeing the filing cabinet. Suppose that I know that I am having this experience, and that I know this directly. In other words, suppose I am directly aware that I am having an experience of seeing the filing cabinet. As we have noted, to be aware of something is not yet to attend to it, and thus to suppose that I am aware of my experience is not yet to say that I attend to it. But suppose now that I begin to think about my experience—to cognitively attend to it. In doing so, I may well focus on the filing cabinet, just as the thesis of diaphanousness predicts. My experience is thus diaphanous to introspection: when I think about my experience of the cabinet, I find myself attending to the cabinet. But none of this provides one with any temptation at all to take back the claim that my knowledge that I am having the experience is in any sense not direct. More generally, (2a) gives us no grounds at all for (3). (2a) tells me something about how I think about what I know. But (3) tells me something about how I know what I know in the first place.
Earlier we saw that Block’s appeal to the distinction between attention and awareness misfired because he was attacking the argument from diaphanousness in the wrong form. What we have just seen however is that there is something importantly right in the suggestion that the argument trades on a conflation between attention and awareness, and thus Block is at least partially vindicated. On our revised interpretation, Moore’s point about diaphanousness—viz., (2)—should be interpreted as (2a). But (2a) is a thesis about attention: it tells us what happens when we try to attend to the intrinsic features of our experience. On the other hand (3) is a thesis about awareness or knowledge: it tells us that we are or become aware of the intrinsic features by being or becoming aware of something else. However, since attention is one thing and awareness is another, the step from (2) to (3) is illegitimate.
6. Replies and Rejoinders
How might a proponent of the argument from diaphanousness reply to the objection that, when one articulates the direct awareness condition in the right way, and when one distinguishes the two notions of attention, it is clear that (2) does not entail (3)?
6.1. A Perceptual Model of Introspection?
The first reply I want to consider rejects the assumption—which we took over from Dretske—that awareness of experience is f-awareness. It is important to notice that this assumption has been critical to the argument. On the assumption that the notion of awareness in play is f-awareness, it becomes rather difficult to establish that my awareness of the intrinsic features of my experience is indirect. For the only way to establish that would be to produce some other fact on the basis of which I become aware that my experience has such and such an intrinsic feature. In effect, what I have been suggesting so far is that considerations of diaphanousness fail to make it plausible that there is some other fact, and thus fail to refute qualia realism.
But what if the notion of awareness in play is not f-awareness but p-awareness? On that assumption, it seems easier to establish that my awareness of intrinsic features of experiences is indirect. For, on that assumption, to show that my awareness of intrinsic properties is indirect one needs to produce, not another fact, but only another property, and then to make it plausible that I am p-aware of the intrinsic features by being aware of this other property. But the considerations of diaphanousness precisely do seem to make this plausible! After all, those considerations make it plausible that I am p-aware of features of my experience by being p-aware of features of objects in my environment. So if it can be made out that I am p-aware of the intrinsic features, it would seem that the argument can be resuscitated.
Now, earlier, our reason for operating with f-awareness rather than p-awareness was that to operate with p-awareness was to commit oneself to the perceptual model of introspection, and that there is ample reason to want to avoid that model. However, in light of what has just been said about p-awareness, one might respond that it is really not clear that the perceptual model is as a bad as its reputation. At any rate, one might respond that I have certainly not shown that it is.
Now is not the time for a full-scale discussion of the perceptual model, but perhaps it will suffice to note that, whether or not the perceptual model is true, a proponent of the argument from diaphanousness is in a particularly weak position to say that it is. For consider again the passage from Moore with which we began. One implication of what Moore is saying here is that my relation to my sensation, whatever that is, is not at all the relation I bear to the blue. However, since I see the blue, we may conclude that I do not literally see the sensation. And, on the assumption that I am aware of my sensation, that is just to say that the perceptual model is false. So it will clearly do no good to appeal to the perceptual model as a potential method of patching up the argument from diaphanousness.
One might reply that the thesis of diaphanousness and the perceptual model of introspection are only in tension if one assumes that one is aware of one's sensation. But couldn't one deny this? As we have noted, proponents of the argument are sometimes interpreted as holding that one is not aware of experiences. And perhaps one might read that into the passage from Moore. After all, the word ‘diaphanous’ at least suggests that experience is not perceived at all. And if introspection just is perception, it follows from this interpretation of the diaphanousness point that I am not introspectively aware of my own experience.
But at this point we are in the position of using a Moorean fact against Moore himself, or at any rate against Moore as we are currently interpreting him. As we in effect noted previously, surely it is a Moorean fact that we are aware of our sensations. Surely we are more sure that we are aware of our sensations than we are of the perceptual model of introspection. To put the point differently, the package of views that Moore on this interpretation is committed to—the perceptual model of introspection plus the diaphanousness of experience—has the result that one is not aware of one's own sensations. But since we evidently are aware of our own sensations, one or both members of this package has to give. Hence a proponent of the argument is in no position to dispense with diaphanousness, it is obvious that what should be dispensed with is the perceptual model.
Alternatively, and more generally, one might suggest that this point misconstrues the role of the perceptual model in the debate about diaphanousness and qualia. Perhaps the point is not that the perceptual model is true, but rather that a qualia realist is committed to the perceptual model, i.e., that the very notion of direct access is somehow bound up with the perceptual approach to introspection. If so the argument from diaphanousness would indeed seem to be on a collision course with qualia realism.
But of course, the problem with this suggestion is that it is completely unobvious that any qualia realist need or should endorse the perceptual model. As we have seen already, the idea of direct awareness can certainly be spelled out in terms of the idea that one is aware of the fact that one’s experience has its phenomenal character not in virtue of being aware of anything else. But that does not involve commitment to the perceptual model. It is true of course that there is a tradition according to which commitment to qualia and commitment to the perceptual model go hand in hand, viz., the sense-data tradition. But as we have noted, the version of qualia realism that we are concerned with is not a version of the sense-datum theory.
6.2. A Non-perceptual Model?
Our discussion of the perceptual model was motivated by the thought that perhaps one could drop the assumption that awareness of experience was f-awareness and operate with p-awareness. And the problem we have run into is that the perceptual model is implausible, or else it is not something which a proponent of the argument can appeal to.
However, one might point out that two things have been run together here: the idea of p-awareness and the idea of the perceptual model. Surely not all p-awareness or o-awareness is perceptual. After all, it seems quite possible that one can be aware of (e.g.) the virtues of Ghandi, without supposing that one sees the virtues of Ghandi. But if that is so, might one not introduce an account of introspection according to which one is p-aware of phenomenal character? In that case, we seem to have the indirectness claim back again. Moreover, we seem to have it back in a form that does not involve us in a discussion of the perceptual model of introspection.
The suggestion that one might be p-aware of something and yet not be perceptually aware of it raises the general topic of non-propositional intentional states. This is an enormous and interesting topic in its own right. Still, I think we can see what is wrong with this method of defending the argument from diaphanousness without discussing that issue in depth. Suppose it is true that there is a notion of introspective awareness which is distinct from perceptual awareness. And suppose also that the awareness at issue is p-awareness rather than f-awareness. Would that help the proponent of the argument respond to our objection? The answer to this question is no. For the fact is that once one draws a distinction between perceptual and introspective awareness, we need to confront a further dimension of difference in the argument from diaphanousness that we have so far been suppressing.
On the schematic account of directness that we adopted from Alston and Jackson, one is indirectly aware of something just in case one is aware of it, but not by being aware of anything else. Now, the suggestion that there are different ways of spelling out the notion of awareness forces us to consider whether, in any putative case of indirect awareness, the awareness in both cases is the same. One dimension of difference that is relevant here concerns whether the notion of awareness that is in play is o-awareness, p-awareness or f-awareness—this is the dimension that we have so far been discussing. However, a quite different dimension comes into focus when one distinguishes between perceptual and introspective awareness. When we focus on that dimension of difference, it seems reasonable to think that the only case in which a case of awareness will count as indirect is when the awareness in both cases is of the same sort. So a case of introspective awareness will count as indirect if it is based on another case of introspective awareness. Similarly, a case of perceptual awareness will count as indirect only if its based on another case of perceptual awareness. On the other hand, in thinking about this sort of issue, we arrive at yet further reason for supposing that (2) does not yield (3).
The issue might be presented compactly if we look back once again at (2):
(2) In introspection, one is or becomes aware of the intrinsic features of one’s experience, by attending to the objects and properties represented by that experience.
Setting aside for the moment the distinctions between attention and awareness, and between f- and p-awareness, one might think that (2) admits of a simple permutation, as follows:
(2b) In introspection, one is or becomes aware of the intrinsic features of one’s experience, by being or becoming aware of the objects and properties represented by that experience.
But (2b) can be interpreted in two quite different ways. On the one hand, we might take the phrase ‘in introspection’ to govern the entire construction. In that case (2b) means something like:
(2ba) In introspection, one is or becomes aware of the intrinsic features of one’s experience, by being or becoming introspectively aware of the objects and properties represented by that experience.
On the other hand, we might take ‘in introspection’ to govern only the sentence that comes before ‘by’. In that case (2b) means something like:
(2bb) In introspection, one is or becomes aware of the intrinsic features of one’s experience, by being or becoming perceptually aware of the objects and properties represented by that experience.
But now the problem is transparent. On the interpretation of direct awareness with which we are working, to derive (3) the proponent of the argument from diaphanousness needs to establish (2ba). For only then will it be true that one introspects the phenomenal character of one’s experience by introspecting something else. One other hand, all (1) seems to support is (2bb), and (2bb) does not yield (3).
Indeed, it is worth pointing out that, on the interpretation we have introduced, the argument is flawed in two quite separate ways. One way is that (2ba) seems not to be supported by (1). But another is that (2ba) looks to be nonsense. Surely it is not merely false, but nonsensical, to say that I can introspect objects in the external world. That would be like saying that one can remember future events. The point is important because proponents of the argument from diaphanousness sometimes talk as if one could ‘introspect’ properties of external objects—something which in my view should raise alarm bells about the cogency of the argument.
6.3. A Different Notion of Directness?
The two replies we have considered so far both respond to our assumption that awareness of phenomenal character is f-awareness. However, the final reply I want to consider criticizes the objections I have made at altogether a different juncture.
My assumption so far has been that direct awareness it to be explicated on the model offered by Alston and Jackson. But—one might point out—there are at least two other accounts of direct awareness available, and these must be disposed of before our objection to the argument from diaphanousness can be assumed to be successful.
The first of these articulates direct awareness in terms of the notion of inference. On this interpretation, the idea behind denying direct awareness might be brought out as follows. I know what my experience is like because I first know what the objects and properties represented in my experience is like, and I infer what my experience is like on the basis of this. On this account, my access to phenomenal character will be direct just in case I know what it is like, but do not know what it is like on the basis of an inference from any other knowledge.
However, there are three responses to this idea. The first is the (ad hominem) point that proponents of the argument from diaphanousness tell us that what is involved here is not an inference (Tye 2000, p.47). The second is that it is in any case implausible to say that I infer facts about my experience from facts about my environment. It is plausible to say, for example, that one might infer the phenomenal character of the experience of others on the basis of their behavior, but given that my own access to the phenomenal character of my mental states is supposed to be very different, inference does not seem to be the appropriate notion here. The third is that, in any case (2) does not entail (3) even on this account of direct awareness. What (2) tells us is that I apprehend in introspection the phenomenal character of my experience by attending in perception to something. But it does not say that I apprehend the phenomenal character by knowing anything.
The second way of interpreting the notion of directness takes up and develops a suggestion of Dretske’s. At one point, Dretske (1999, p. 164) says that one is directly f-aware of some fact F just in case F is constituted by o and P and one is directly o-aware of o and directly p-aware of P. On this understanding, if one is f-aware of the fact that one’s experience e has phenomenal character C, one can only be directly aware of this if one is o-aware of e and p-aware of C. It follows from this that the diaphanousness of experience will easily establish that one is only indirectly aware that one’s experience has any phenomenal character because diaphanousness does suggest that one is not o-aware of e or p-aware of C.
Now, there is no point doubting that, on this account of directness, my awareness of the intrinsic features of my experience is indirect. However, it does not follow from this that we have a version of the argument that is successful. For it should not be at all surprising that one might be able articulate notions of directness according to which my awareness will count as indirect. The point is whether the qualia realist can appeal to a workable notion of directness according to which they are direct. As we have seen, schematic account of awareness that we have introduced does seem to capture such a workable notion. To put the point differently, one can agree that there are senses of directness according to which any particular case of awareness is indirect. But the crucial point is whether the argument from diaphanousness can disprove the idea of direct awareness as that idea is developed by the qualia realist.
The argument from diaphanousness, as I have been considering it, aims to show that there are no psychological properties of experience which satisfy both intrinsicness and direct awareness, and in particular, that there are no psychological properties which satisfy direct awareness. My suggestion has been that the argument is unsound because there is no way to defend the inference from the claim that I apprehend the intrinsic features of my experience by apprehending objects and properties represented in experience to the claim that I only indirectly apprehend those features. As we have seen, there are various ways of understanding what direct awareness and attention might amount to, but on none of these does the ‘by’-claim summarized in (2) support the denial of direct awareness summarized in (3).
I want now to close the paper by returning to the two versions of the argument noted but set aside in §2 and by summarizing the main points.
7.1. The argument construed as an argument for the relational thesis.
In discussing the puzzle about Moore in §2.3, we took note of the relational thesis, the thesis that the phenomenal character of an experience is determined by the properties of the objects that one is related to in having the experience. We saw that the relational thesis crosscuts the debate between intentionalists and qualia realists: hard intentionalists and sense-data theorists endorse it, soft intentionalists and qualia realists such as Block deny it. Nevertheless—as we also noted— it is plausible to suppose that the relational thesis is in the background of the contemporary debate about the argument from diaphanousness. So in this section, I want to ask whether there is a version of the argument that provides any support for the thesis.
This issue is in fact a rather large one, and so I will limit myself here to two points. The first point is that if the thesis of the diaphanousness of experience is going to be a premise in an argument for the relational thesis then it would need to be adjusted slightly. The reason is that the thesis as we have been considering it presupposes that there are intrinsic properties of experience. But the relational thesis would seem to deny this, or at any rate, would seem to deny it as far as phenomenal character is concerned. For suppose that the phenomenal character of an experience is determined by the objects that one is related to in having the experience. Then it is hard to see how the phenomenal character of the experience might count as an intrinsic, or, at any rate, as a non-relational, property of that experience. However, it is not hard to see how one might adjust the thesis to accommodate this complication. Instead of (2) or the variants that we have introduced, the thesis should now be rendered as (2c):
(2c) In introspection, one is or becomes aware of the phenomenal character of one’s experience by attending to the objects that one is related to in having the experience.
This thesis differs from our original thesis in two respects. First, we are now talking about the phenomenal character of experience, rather than specifying that the phenomenal character is intrinsic. Second, the thesis now explicitly includes the picture of experience with which the relational thesis operates, viz., the picture according to which in having an experience one is related to a range of objects. As we noted in §2.3, there are a wide variety of otherwise quite different approaches to experience which conform to the experience thesis, hence there are wide variety of otherwise different approaches to experience which can agree with (2c).
The second, and more straightforward, point is that there is in any case a good reason for supposing that there is no (convincing) argument from (2c) to the relational thesis. For the fact is that, while (2c) might be true—at any rate, nothing I have said militates against it—the relational thesis is false. To illustrate this, reconsider the contrast noted in §4.2 between the situation in which I am under the visual impression that p and the situation in which I visually imagine that p. The relational thesis implies that then the phenomenal character of these two experiences are identical. For according to that thesis, the phenomenal character of an experience is determined by the features of the object that one is related to in having the experience. But in this case there is surely a strong sense in which the phenomenal characters here are not identical. As we noted, in the visual impression case, I am going to feel inclined to be believe that p. In the imagination case, on the other hand, I am not going to feel inclined to believe that p. However, since these inclinations are surely part of the phenomenal character of imagination and perception, it is natural to suppose that what explains the phenomenal character of an experience cannot simply be properties of the objects that one is related to in having the experience—the nature of the experience itself must be included also. But if that is right then the relational thesis is false, and thus there can be no persuasive argument from diaphanousness to the relational thesis.
7.2. The argument construed as an argument for intentionalism.
The final issue I want to consider—and this is the last one I will deal with in the paper—concerns the point noted in §2.2 that, while the argument from diaphanousness is often formulated as a negative argument whose conclusion is that there are no qualia, it is often also formulated as a positive argument whose conclusion is that intentionalism is true. And we also saw that, because the contrast between qualia realism and intentionalism was not straightforward, it was necessary to treat the two versions of the argument separately. In this final section, therefore, I will consider this version of the argument.
We can start with the observation that, given the distinction between hard and soft intentionalism, the only issue that requires discussion is whether the diaphanousness of experience supports soft intentionalism. For the difference between hard and soft intentionalism can be expressed in terms of the relational thesis—hard intentionalism endorses the relational thesis, soft intentionalism denies it. And we have already seen that the relational thesis is false. In consequence, hard intentionalism is false also. Hence there can be no (convincing) argument from the diaphanousness of experience to hard intentionalism. Hence the only issue to be considered is whether the thesis of the diaphanousness of experience provides any support for soft intentionalism. Put differently, the issue is whether the ‘by’-claim summarized in (2) tells us that soft intentionalism is true.
The first thing to say is that this step is certainly not one of entailment. To see this, consider the step from (4) to (5):
(4) One apprehends the feelings of others by attending to their (actual and potential) behavior.
(5) The feelings of others are completely explained by their actual and potential behavior.
It should be clear the step from (4) to (5) is invalid. It is perfectly true that one apprehends the feelings of others by attending to their behavior. But behaviorism—for (5) is nothing else but a formulation of behaviorism—is as false as any philosophical doctrine could be. Since (4) is true, and (5) is false, the inference from (4) to (5) is not valid. But—and here is our objection—the inference from (2) to intentionalism is of the same form as that from (4) to (5).
One might of course agree that the step from (2) to intentionalism is invalid, while at the same time insisting that intentionalism nevertheless provides the best explanation of (2). Here too, however, the analogy with behaviorism is instructive. Suppose the behaviorist says that (5) provides the best explanation of (4), and that is why we should believe it. Our response, I think, would be to grant that behaviorism provides a potential explanation of (4), but to point out that there are others. The other account is roughly that the feelings of others find causal expression in their behavior. Once we have another account of why (4) is true, likewise we may look to another account of why (2) is true. (2) is true—one might say—because if we were in a situation in which our experiences were veridical, the world that we would then find ourselves in would cause us to have certain experiences. In other words, while it might very well be true that (2) represents a striking fact which cries out for explanation, it does not at all follow that intentionalism is the best or only way to explain it.
Of course none of this shows that soft intentionalism is false, nor does it show that soft intentionalism might not draw a certain limited level of support from considerations of diaphanousness. As I have noted previously, the relation of intentionalism to qualia realism, and therefore also the truth of intentionalism, is in fact a highly complex issue. The important point for our purposes, however, is only that the argument from diaphanousness does not provide the short, sharp argument for intentionalism that it is sometimes presented as doing in the literature.
Let me summarize the main points I have tried to make. I began by noting that the argument from diaphanousness comes in three versions: (a) a version which attacks the sort of qualia realism which is prevalent in modern philosophy, typified by the Shoemaker-Block view; (b) a version which supports the relational thesis; and (c) a version which supports intentionalism, the idea that the phenomenal character of an experience is explained by its intentionality. I then argued that there is no version of the argument which is successful.
Alston, W. 1971. ‘Varieties of Privileged Access’, American Philosophical Quarterly, 8 (No.3) 223-41.
Armstrong, D. 1968. A Materialist Theory of the Mind, London: Routledge.
Armstrong, D. 1980. ‘Immediate Perception’. In The Nature of Mind and Other Essays Univ. of Queensland Press, St. Lucia, Queensland, 1980
Block, N. 1990. ‘Inverted Earth’, in Tomberlin, J. (ed.) 1990 Philosophical Perspectives, Vol. 4, Ridgeview, pp.52-79. Reprinted in Block et alia, (eds.) 1997 The Nature of Consciousness: Philosophical Debates, MIT Press.
Block, N. 2000: ‘Mental Paint’. In Hahn, M and Ramberg, B (eds.) Essays in Honor of Tyler Burge, MIT Press. References are to the PDF version.
Broad, C.D. 1925. The Mind and Its Place in Nature. Routledge.
Byrne, A. 2001: ‘Intentionalism Defended’. Philosophical Review, Forthcoming.
Churchland , P. 1985. ‘Reduction, Qualia, and the Direct Introspection of Brain States’, Journal of Philosophy, 82:8-28.
Evans 1982. Varieties of Reference. Oxford: Oxford University Press.
Grice, P. 1962. ‘Some Remarks About the Senses’. In Butler, R.J. (ed.) Analytical Philosophy, Barnes and Noble, New York.
Harman, G. 1990: ‘The Intrinsic Quality of Experience’, Philosophy of Mind and Action Theory: Philosophical Perspectives. 4:31-52.
Harman, G. 1996: ‘Explaining Objective Color in terms of Subjective Reactions’. In Villanueva, E. (ed) 1996 , Atascadero, CA: Ridgeview.
Jackson 1977. Perception. Cambridge University Press.
Jackson 2001. Experience and Representation. MS.
Langton, R and Lewis, D. 1998 ‘Defining “Intrinsic”’, Philosophy and Phenomenological Research, 58:333-345
Lormand 1996. Inner Sense Until Proven Guilty, MS. Available at:
Lycan 1996. Consciousness and Experience. MIT Press.
Martin, M.G.F., 1997, ‘Sense, Reference and Selective Attention: The Shallows of the Mind’, Proceedings of the Aristotelian Society, Vol. 71. Pp. 75-98.
Martin, M.G.F., 1998 ‘Setting Things Before the Mind’. In O’Hear A. (ed.) Current Issues in Philosophy of Mind, Cambridge University Press.
Robinson, W. 1996 Intrinsic Qualities of Experience: Surviving Harman’s Critique, Erkenntnis 47:285-309.
Ryle, G. 1949. The Concept of Mind. Routledge.
Tye, M. 1995. Ten Problems of Consciousness, Cambridge, MA: MIT Press.
Tye, M 2000. Color, Content and Consciousness, Cambridge, MA: MIT Press.
Segal, G 2000. A Slim Book on Narrow Content, Cambridge, MA: MIT Press.
Shoemaker, S. 1981. ‘The Inverted Spectrum’, Journal of Philosophy, 79,7, 357-381.
Shoemaker, S. 1994. 'Phenomenal Character', Nous, 28 (1994), pp.21-38
Shoemaker, S. 1996: The First Person Perspective and Other Essays, Cambridge University Press.
Shoemaker, S. 2001: ‘Introspection and Phenomenal Character’. MS.
Van Gulick, R. 1993, ‘Understanding the Phenomenal Mind: Are We All Just Armadillos’. In Davies, M. and Humphreys, G.M. (eds.) Consciousness: Psychological and Philosophical Essays, Blackwells.
Research School of Social Sciences
Australian National University
Canberra, ACT 0200
Acknowledgments: I presented previous versions of this paper to audiences at the Australian National University, Canberra, Australia, the Language, Mind and World Conference, Tlaxcala, Mexico, the Australian Association for the History, Philosophy and Social Studies of Science Conference, Melbourne, and at MIT. I am very much indebted to all who took part on those occasions, and to the very many others with whom I have discussed this material. Particular thanks to Frank Jackson who was my commentator at the Melbourne conference, and to Philippe Chuard and Dan Lopez for extensive comments.
 Harman does not mention Moore by name, but, as many have noted, it is clear that he has Moore mainly in mind. Presentations and discussion of the argument can be found in Harman (1990, 1996), Tye (1995, 2001), Martin (1998, 2001), Lycan 1996, Jackson 2001. Criticisms of the argument can be found in Block (1990, 2001), Lormand 1996, Robinson 1996, and Shoemaker (1994, 1996). The diaphanousness of experience is well-known in the history of philosophy of mind, receiving attention from Broad 1925, Ryle 1949 and Grice 1962. It has also been widely discussed in the sense datum literature, though I will not take up that issue in detail here; cf. Martin 2001 for discussion. A similar idea about the ‘diaphanousness of belief’ is to be found in Evans 1982, but I will not discuss that issue here. For some recent discussion of the notion as it appears in Evans, see Peacocke 1998. Van Gulick 1993 presents an interesting discussion of diaphanousness, in which he traces the idea, or a similar idea, back to Kant.
 Here is Tye: “I believe that experience is transparent. I also believe that its transparency is a very powerful motivation for the representationist view. I concede, however, that the appeal to transparency has not been well understood” (Tye, 2000; p. 45).
 In speaking of the intentionality of experience, I mean that in having an experience one represents the world as being a certain way. There are a number of important philosophical issues concerning the nature of the intentionality of experience, but I will set these issues aside here. I mention of the functional role of the experience simply to register the fact that many intentionalists (e.g. Lycan 1996) include this factor in the determination of phenomenal character. For some further discussion of this issue, see the discussion of the relational thesis, and the contrast between soft and hard intentionalism, in §2.5 and §7 below.
 Another debate is of course physicalism, but I think the physicalist debate can be profitably separated from the debate about qualia as we will discuss it here and so I will set it aside.
 This oversimplifies matters slightly. Shoemaker is responding both to the argument from diaphanousness and to issues having to do with the inverted spectrum. As we will see below, I think the argument from diaphanousness can be answered even if one adopts the conception of qualia discussed in the text. So, if I am right, at least one motivation for Shoemaker’s view evaporates. It should be noted that Shoemaker’s current view differs from the position presented in 1996. See Shoemaker 2001.
 A puzzle mentioned by a number of different authors including Jackson 2001, Martin 2001 and Tye 2000.
 For some discussion of the Naïve Approach, see Martin 2001.
 The sense-datum theory is sometimes interpreted as a theory according to which the intentionality of the experience explains its phenomenal character (cf. Byrne 2001 for an interpretation like this, but see Jackson 2001 for a dissenting interpretation). So, from this point of view, we have here a second example of a position which combines qualia realism with intentionalism. As in the discussion of projectivism, I assume that intentionalism strictly so-called can be developed in such a way that it is not consistent with these versions of qualia realism. But I will not discuss the issue further here.
 One reason that adverbialism fell out of favor is that it was often presented as an account of the logical syntax of perceptual vocabulary. On that interpretation, the thesis is subject to extremely serious objections; cf. Jackson 1977. But it is tempting to suppose also that another reason that adverbialism fell out of favor is that adverbialists offer no account of the intentionality of experience. Indeed, for adverbialists experiences lack intentionality altogether. For some development of this theme, particularly as it relates to the sense-datum theory, see Jackson 2001.
 The Shoemaker-Block view is also closely allied to views such as that of Peacocke 1983. It is important to note, however, that Peacocke’s position is often construed as version of a sense-datum theory, but the Shoemaker-Block view is not.
 It is important to notice that the difference between kinds of qualia realists that I have just made does not emerge when we consider only the form of words that is used to state the doctrines. Take the statement ‘qualia are properties of experience that satisfy intrinsicness and direct access’. Both Moore and Block might well agree to this form of words. But they would understand it very differently. For Moore, the word ‘experience’ will or might be taken to mean the object that one is related to in having the experience, a sense datum. For Block, on the other hand, and for all qualia-realists who endorse the Shoemaker-Block view, the word ‘experience’ in this form of words will be taken to mean the psychological state of experiencing. Again, since I will operate with the Shoemaker-Block view here, I will follow them in this usage.
 Lycan’s position is sometimes called 'quasi-representationism' or 'quasi-intentionalism'. However, since I want to emphasize the fact that Lycan’s position is a version of intentionalism, I will adopt a different phraseology here. Thanks to Philip Pettit for suggesting the phraseology.
 It is important to stress that the difference between hard and soft intentionalism is mainly a matter of degree rather than kind. Most intentionalists do emphasize the functional role of experience to some extent The crucial issue is to what extent they do so. Cf. Block 2001 including in particular the discussion of whether Harman is a ‘quasi-representationist’ or not.
 One should perhaps emphasize psychological or mental intrinsic features here—but I will largely leave that implicit in what follows.
 My formulation of (2) has been greatly influenced by Martin 2001 and Tye 2000. Tye (p. 51) says for example, that “When we introspect our experiences and feelings, we become aware of what it is like for us to undergo them. But we are not directly aware of those experiences and feelings; nor are we directly aware of any of their qualities. The qualities to which we have direct access are the external ones, the qualities that, if they are qualities of anything, are qualities of external things. By being aware of these qualities, we are aware of phenomenal character.” It is important to note that (2) embodies a certain conception of experience which is friendly to an intentionalist, since it speaks of objects and properties represented by the experience. As we will note in §7, there are more general ways to state the thesis, but I will ignore this aspect of the issue until then.
 One may speak not only of the objects and properties represented by the experience but also of the object and properties presented in the experience. I will sometimes adopt this expression as a stylistic alternative.
 In his commentary on this paper at the conference in Melbourne, Frank Jackson suggested that one might think about the structure of the argument in a slightly different way. According to Jackson, from the premise that one is only directly aware of the objects and properties presented in experience one may conclude that there are no qualia—the reason being that the properties that one are aware of are intentional rather than real, and thus need not be instantiated in experiences. Surely however, qualia are properties which are instantiated in experiences. This interpretation of the argument is interesting and is different from the version that I consider in the text. Nevertheless, I think we can set aside this suggestion here. Given our distinction between (3) and (3*) what Jackson is doing is providing novel grounds to move from (3*) to (3). As noted in the text, however, my focus here is not on the move from (3*) to (3), but rather on the move from (2) to (3), or—something I take to be largely equivalent in the context—the move from (2) to (3*). So far as I can see, Jackson’s point does not affect what I have so say about that transition.
 As Josh Parsons pointed out to me, there is one sort of intrinsic property of an experience which one might say is both a property of the tree and of Eloise’s experience: the property of being such that there is an experience. But it is not clear that one needs to appeal to this property in spelling out how Eloise takes the world to be in visual experience.
 To what extent is the idea that experiences are to be identified with their core-realizers influencing those who deny intrinsicness? This interpretative question is difficult to resolve. However, it is important to note that both Harman 1990 and Block 1990 are operating with a ‘language of thought’ conception according to which an experience is to be identified with a linguistic vehicle of some sort. I take it that this conception of experience is at least close to version which identifies an experience with its core realizer.
 Another alternative would be to deny the identity between experiences and their core-realizers and go for a constitution thesis. Thanks to Martin Davies for pointing out this option to me.
 Two further points concerning intrinsicness should be noted: (1) Philosophers who support internalism in the philosophy of mind, or at least do so in certain specific cases, argue that intentional properties are (or at least are in some cases) intrinsic in the second sense. For a recent general recent defense of internalism see Segal 2000, and for discussion of how the externalism/internalism issue figures in the debate about qualia and intentionalism, see Block 2001 and Jackson 2001. (2) The notion of intrinsicness is in any case extremely puzzling. For a recent discussion see Langton and Lewis (1998) and the references therein.
 There is a further point here that might also be mentioned. Block is contrasting a theoretical question and a question that might answered on the basis of introspection alone. But is not clear that there is any genuine contrast here. Certainly if one lacks a certain concept, one cannot answer on the basis of introspection alone the question of whether the concept applies—but the issue of concept possession does not seem to be relevant to the case Block has in mind. On the other hand, if one possesses the relevant concept, one might well be able to decide the question of whether it applies on the basis of introspection alone. The issue turns on whether the concept is the sort of concept for which introspection provides a proprietary evidential basis. If so, there is no reason to suppose that introspection alone should not furnish sufficient information for one to be able to decide that the concept applies. The problem for Block is that it is not at all clear that the concepts in question are not of this sort.
 For further discussion of the notion of immediacy, see Armstrong 1980.
 The ‘merely’ is required here to rule out the case in which one is both directly and indirectly aware of something.
 The assertion that most modern philosophers are not attracted to the perceptual model might seem strange in light of the fact that that some, like Armstrong 1968 and Lycan 1986, express their allegiance to it. But this strangeness is avoided when we notice that for Armstrong and Lycan the perceptual model at issue is what Shoemaker (1996) calls a ‘broad perceptual model’, that is, on this model one’s awareness that one has an experience of a certain sort is f-awareness. In short, the perceptual model rejected by Dretske is not the theory defended under that name by Armstrong.
 For some excellent discussion of this see Martin 1997. See also Shoemaker 1996 and Martin 1998
 I think this would be a misreading, since Moore is careful to say that the sensation is as if it were diaphanousness. But the interpretative issues are difficult and I will not attempt to address them here.
 One might attempt to defend the package from this criticism by introducing a distinction between direct and indirect introspection on the model of direct and indirect perception. In Berkeley's famous example, I directly hear the clip-clop of the horse’s hooves, but I only indirectly hear the carriage. By analogy, one might say that while I only directly see the blue in introspection I indirectly see the sensation. However, the problem with this suggestion is that the claim that I indirectly see the sensation is only marginally less implausible than the claim that I directly see it. To say that I hear the carriage indirectly is to say that I hear it. It is not to say that it is inaudible. But to say that experience is diaphanousness would seem to imply that I do not see it at all.
 At one point, for example, Tye says that “Via introspection, you are directly aware of a range of qualities that you experience as being qualities of surfaces at varying distances away and orientations…” (2000; p. 47).
 Attentive readers will note that I have switched the order of discussion here.