The Oxford Companion to Consciousness

Q

qualia We normally assume that for certain types of mental states, there is ‘something it is like’ to have them. We suppose, for example, that there is something it is like to see the blue headband of Vermeer’s Girl with a Pearl Earring or to experience the pain of a dull headache. Contemporary discussions of qualia—that is, of the properties in virtue of which such things are true of some mental states (the qualitative states)—have tended to take such examples and such language (‘something it is like’ to have them) as definitive of the subject. Not all contemporary uses of ‘qualia’, however, can be understood in this way.

1. Qualia realism

2. Privileged access

3. The further-fact arguments

4. The problem of other minds

5. Meaning scepticism

6. Alternative approaches to the empirical study of qualia

1. Qualia realism

Some theorists, such as David Rosenthal, use the term in a broad sense that would allow for the existence of unconscious qualia. Others define the term so that only the non-representational properties of mental states count as qualitative properties. Consider a visual experience of two objects viewed at different distances that represents them (correctly) as being the same size. Such a visual experience is an intentional state in that it represents the world as being a certain way and thus has a certain *representational content that can be assessed for accuracy. In this it is normally thought to differ fundamentally from such mental states as pain that are supposed not to represent the world, but simply feel a certain way. But although the visual experience has representational properties in virtue of which the objects are given as being the same size, it will, according to some theorists such as Christopher Peacocke, have non-representational properties as well. These properties (the qualitative content of the experience) include those in virtue of which the closer object seems to occupy a larger area of the observer’s subjective visual field. On the characterization of qualia according to which they are necessarily non-representational, only properties of the latter kind would qualify.

In what follows I shall concentrate on the use of ‘qualitative state’ according to which all that is assumed is that there is something it is like to be in one. The theorists whom I shall call qualia realists take it that a state’s being conscious, its having qualitative content, and there being something it is like to occupy it go hand in hand. What defines qualia realism is, first, the claim that there are such states and, second, that they are not reducible to any other category of mental state. The first claim contrasts with *eliminativism, according to which science will show not what qualia are, but that they do not exist. And the second claim would entail that a view like George Pitcher’s according to which visual perception consists in the acquisition of beliefs would not, by itself, account for its qualitative content. Indeed, this claim would normally be understood to entail the possibility of *zombies—subjects who have all of the same intentional states that we do, but lack any qualitative experience whatsoever. The rationale for this claim is that a reductivist view of qualia that leaves out most of what is intuitively distinctive or essential is not importantly different from eliminativism.

Most qualia realists who are not *dualists also take for granted the truth of *physicalism and scientific realism. Such theorists assume that what exists is what is entailed or presupposed by the truth of the ultimate physical theory of the world. Only those entities exist, it is assumed, that ‘pull their weight’ in such a causal/explanatory account. And it is assumed that qualitative states are such physical entities. It is a natural further assumption of such non-dualist qualia realists that the scientific study of qualitative states will take place either within the framework of a computational cognitive science or a neurophysiological brain science. Those who favour the first alternative endorse functionalist accounts according to which mental states are internal states defined in terms of their causal relations to inputs, outputs, and other internal states. Such accounts liken mental states to states of a computer program understood in abstraction from their physical realizations in specific machines. Those who reject this alternative hold that some entity might have all the same functional states that we do but fail to have our qualitative states (or indeed any qualitative states at all) because those functional states are not realized ‘in the right way’ physically.

A final assumption of qualia realists, which is common but often only implicit, is that there is a close connection between the character of our qualitative experience and the relevant organs of perception. In the case of visual perception, this means that what we are given most directly is governed by the nature of the retinal image. Hence it is something not far removed from the apparent colours, shapes, and relative sizes of external objects. In this, their views bear a marked similarity to those of classical empiricists and 20th century sense-datum theorists.

Qualia realism, together with the physicalist and scientific realist background assumptions, has been one of the most important approaches to qualitative experience for the past several decades and is, arguably, the dominant approach today. However, several fundamental problems have emerged.

2. Privileged access

Central-state materialists such as J. J. C. Smart and David Armstrong identified mental states, including qualitative states, with internal physical states of the brain. We generally take it, however, that we have a privileged relation to our own sensations such as pain. We assume, that is, that, necessarily, if we believe that we are in pain we are, and if we are we believe it. We would understand if a doctor, looking at an X-ray film, said we were mistaken in supposing we had a broken arm. It seems unthinkable, though, that the doctor could be justified in dismissing our claim to be in pain on the basis of a brain scan. But it is difficult to see how this feeling of privilege could be supported if the pain and the belief that one is in pain were independently existing physical states, or even independently existing states that were irreducibly mental. Indeed, even the definitional stipulation, suggested by some functionalists, that something is a pain only if it causes a pain belief and vice versa, seems not to capture the intuition that something is a pain in virtue of its intrinsic nature.

3. The further-fact arguments

The attempt to articulate a principled basis for distinguishing between those physical states that are sufficient for qualitative experience and those that are not has met with serious difficulties not directly related to privileged access. In his well-known *knowledge argument, Frank Jackson claims that if the experiences of a subject, Mary, were limited to those inside a black and white room, she might know all the relevant functional and neurophysiological facts about colour perception and yet (since there is no a priori connection) know nothing about the subjective experience involved. For it seems that when she saw her first red object she would learn a further fact—what it is like to see red. If so, it seems that there are facts over and above the physical and functional facts and that such facts are necessary to account for qualitative experience.

Currently there are two important physicalist responses. The first, due to Laurence Nemirow and David Lewis, holds that on seeing a red object for the first time Mary would gain no new knowledge, but only a new skill or ability—a piece of know-how. There are reasons to doubt, however, whether this really disposes of the problem. What could make it intelligible that from her point of view Mary had discovered or learned something new—something that would justify her saying, ‘So this is what it is like to see red’? Only, it seems, if her new capacity were given as the ability to imagine or visualize red objects, and this ability were understood as itself presupposing an acquaintance with red, would Mary’s response be intelligible.

The second response, put forward by Brian Loar and developed by Ned Block, is that on seeing her first red object Mary would gain a new concept—a *phenomenal concept—of the experience of red. Mary would acquire the ability to entertain new propositions and so, in one sense, new propositional knowledge. But on this account she would gain no access to any new facts. How, though, could we characterize the way in which a brain state must present itself to the subject in order to be picked out most naturally by a phenomenal, as opposed to a neurophysiological, concept? Proponents of phenomenal concepts assume that our ordinary relation to our own qualia, when we pick them out (‘from the inside’) via phenomenal concepts, is direct and demonstrative. Such proponents sometimes put the point by saying that in such cases qualia are their own modes of presentation. If qualia are identical to physical properties, however, this position raises difficulties of its own. For we could have two demonstrative relations to what is by these theorists’ lights the same quale—a demonstrative relation to it as the property of being painful, say, and a distinct demonstrative relation to it as a neurophysiological property of a brain state—and we could fail to believe that they were presentations of the same thing. Thus like the earlier response, this threatens to give rise to another version of the same argument.

Since this problem arises for any identity involving the mental that is a posteriori, it raises the same difficulties for most versions of functionalism that it does for physicalist identity theories. Indeed, the problem arises for any reductivist theory with the exception of those that, like analytic functionalism, postulate a conceptual connection between qualia and the properties to which they are reduced. Such general considerations regarding the modes of presentation of qualitative states are currently being explored in the context of discussions of the property dualism argument. The issues of modes of presentation and phenomenal concepts are also crucial to the discussion of what we might call collectively the further-fact arguments—in addition to the knowledge argument and the property dualism argument, Joseph Levine’s *explanatory gap argument, which turns on the alleged impossibility of explaining the felt quality of pain by reference to neurophysiology, as well as Saul Kripke’s modal argument and David Chalmers’ zombie argument. Thus whether the appeal to phenomenal concepts is a promising response to such objections to psycho-physical identity theories is a matter of current dispute.

4. The problem of other minds

As the further-fact arguments indicate, some of the most serious difficulties for non-dualist qualia realists stem from the assumption that mental-physical identities are a posteriori. This assumption generates another set of problems that are arguably even more difficult. Consider a subject who is like us in all the ways that might be supposed to have an a priori connection with consciousness and qualia, but in other ways differs from us significantly—e.g. a subject who has all of our functional states but who differs from us in how they are realized physically. It seems that there is no way in principle of settling the question whether in such a case there is qualitative experience present. Indeed, it seems that nothing could even count as relevant evidence. Since the subject will have the same inputs and outputs that we do, such a subject will claim to have qualitative states (or at least emit sounds as though such a claim were being made). And certainly this subject will make the same discriminations as a normal subject and pass the same behavioural tests. Moreover, looking inside the head will be no help—we already know by hypothesis that we will find that the behaviour is explained by the same functional states that explain our behaviour but that the explanations at the level of the neurophysiology are different. This has been called the problem of DNA physicalist-functionalism (White) and the harder problem of consciousness (Block).

5. Meaning scepticism

The traditional problem of scepticism regarding our knowledge of other minds leads quickly to an even more extreme form of scepticism—meaning scepticism regarding those mentalistic terms with which we ascribe qualitative states to others. If, as the phenomenal concepts view suggests, the only meaning one associates with the term ‘pain’ is what one gets through acquaintance with the sensation itself, then it seems inconceivable that one could ascribe it to others in the same sense in which one ascribes it to oneself. For if all one has is the ability to recognize pain in the ordinary way when it occurs in one’s own case, then one associates nothing with the experience that could, logically, occur without its occurring in oneself. This problem of meaning scepticism, like the epistemological problem of other minds, is as much a problem for dualist versions of qualia as for their physicalist counterparts. These problems are central to the work of Wittgenstein and have been well articulated more recently by Kripke. Qualia realism, then, threatens to make the scientific study of qualia impossible. Such a conclusion would support an eliminativist approach to qualia.

6. Alternative approaches to the empirical study of qualia

In light of the apparent problems for the most prevalent forms of qualia realism, both dualist and non-dualist, it is significant that there exist a variety of other approaches to the scientific and empirical study of qualia. The work of Albert Michotte in investigating the *subjective impression of a causal connection between events and the precise parameters under which it is created has been extended and generalized by contemporary researchers. Their investigations include the subjective impression of *intentionality and of the exercise of meaningful agency, of the manifestation of specific intentions, and of the pursuit of specific goals and purposes. To say these are impressions is to say, for example, that some events are given in perception as actions in the pursuit of goals. Similarly, current research in the tradition of J. J. Gibson credits subjects with perception in which things are given as having functionally relevant properties and presenting the subject with opportunities. Thus things are given not as mere objects having certain sizes and shapes, but as, for example, potential shelters, hiding places, stairs, or bridges. Such topics overlap with those raised by Oliver Sacks. In his work on the brain stemming from the pioneering studies of A. R. Luria, Sacks describes his subjects not by reference to their neurophysiology, but in the terms in which their ‘worlds’ are given to them. And Sacks’ work suggests that the most basic perception of one’s own body and the surrounding space is not of objective spatial entities. Instead it is of a subjective space and body image that are mutually determining and themselves determining and determined by one’s perception of one’s agential possibilities and opportunities.

What grounds the claim of these alternative approaches to being scientific is an empirical methodology capable of generating significant generalizations across subjects. Though the methodology is non-reductive, there are behavioural criteria for the ascription of intentional states to subjects in the Michotte and Gibson paradigms, and/or the formulation of precise characterizations of the physical parameters determinative of the subjects’ responses. And in the tradition associated with Luria, there are broad generalizations connecting subjective states with states of the brain, as well as generalizations across subjects. Such generalizations contribute to the empirical investigation of the structure of what might be called the subjects’ ‘life-worlds’.

Because the perceptual experiences ascribed in these research traditions are rich, they cannot be characterized in terms of such properties as apparent shapes, colours, or relative sizes. We see a person’s distress directly, and not in virtue of being given the geometric features of the person’s face or body, much less the sorts of apparent shapes and colours that make up some counterpart to the retinal image. The problem of other minds that arose for qualia realism, however, was generated by the lack of a priori or conceptual ties between a subject’s qualia and anything to which we might take ourselves to have direct access, such as behaviour or physical states of the brain. If, however, what is given most directly in conscious perceptual experience has the richness attributed to it in these research paradigms, then, there will be such a priori connections between a subject’s qualia and intentional states—there will be no describing the qualia of a subject’s perceptual experience independently of describing the objects represented, and hence the intentional content. On the side of the subject, then, qualia and intentional states are inextricably tied in a way in which qualia realists—for whom qualia float free of intentional content—evidently cannot allow. Moreover, if our experience is as rich as these traditions would have it, what we are given most directly in our experience of others is not their behaviour, conceived in geometric, spatiotemporal terms, but their actions and their expressions of such things as their hopes and fears. Thus we have in our study of other subjects what the qualia realist cannot supply—a priori or conceptual connections between the qualitative contents of other subjects’ perceptual experiences and what we ourselves are given most directly in our experiences of them. Finally, since the theories are non-reductive, there is no conceptual gap of the kind that generates the further-fact problems.

Such non-reductive accounts also have potential implications for the meaning-sceptical version of the problem of other minds. The problem, as we have seen, is that the radical difference in our mode of access to qualitative states in the first- and third-person cases threatened to undermine the assumption that in both cases our mentalistic terms have the same meaning. This points up the importance of the fact that such accounts as Sacks’ acknowledge, and depend on, deep connections between perception on the one hand and *action, and hence action theory, on the other. And in the context of action theory, some philosophers have suggested that the notion of joint agency is a conceptual possibility and, arguably, a common occurrence. Such a thesis, could it be sustained, would, in the context of a non-reductive treatment of qualia, bridge the apparent logical gap between our access to our own subjectivity and that of anyone else. And the idea of bridging this gap is supported by the fact that our most fundamental access to our own agency is arguably through our access to affordances and opportunities that are perceived as being in the world and as available to others.

Such theories, in effect, intentionalize the sensational aspects of perceptual experience, at the same time that they make qualia in the form of rich perception of such things as affordances essential to such intentional phenomena as agency. And their doing so allows them to draw on current empirical work to answer the further-fact arguments. Such approaches will not, however, automatically answer the points that threatened to make bodily sensations such as pain immune to empirical investigation. But as some philosophers, such as Michael Tye, have argued, pain and other bodily sensations have many of the supposedly definitive features of such paradigmatically intentional states as visual experiences. Most fundamentally, pain is always located, and hence is part of, and partially constitutive of, a *body image. In such a body image we are given to ourselves as having certain bodily characteristics and so in a way that can be assessed for accuracy or inaccuracy. Thus pain seems to satisfy the basic criterion of an intentional state, and it will be accessible to the same sorts of empirical investigations as perceptual states. Because the approaches are non-reductive, however, there can be no such blatant violations of privileged access as someone’s being justified in contradicting one’s claim to be in pain on the basis of a brain scan. Thus reductive and non-reductive scientific treatments of qualia, in the context of some philosophical theses currently under investigation, provide very different conceptual contexts within which such problems as those of privileged access, other minds, and meaning scepticism about the mental may be framed and investigated.