Differences between seeing and believing

It can seem puzzling that there is such a thing as the philosophy of sense perception. Psychology and the neurosciences study the mechanisms by which our senses receive information about the environment. So conceived, perception is a psychological and physiological process, whose underlying nature will be discovered empirically. Since few philosophers these days would presume to interfere with the empirical products of these sciences, the question arises as to the nature of philosophy's role in the study of perception. There is not a philosophy of digestion as there is a philosophy of perception; so what is it, exactly, that the philosophy of perception is supposed to do?

While neither addresses this question explicitly, these two books assume very different attitudes to it. Bill Brewer treats the philosophy of perception as a part of epistemology, the theory of knowledge. His book is concerned with the question of how perceptual experiences can provide reasons for beliefs. He claims that if there are to be any beliefs at all about the empirical world (the world of experience), then perceptions must be reasons for some such beliefs, and their status as reasons must derive from their experiential content. He argues that the content of the relevant kind of experience is demonstrative in character, of the general form "that is thus" (where "that" expresses a singular demonstrative mode of presentation, and "thus" expresses a general demonstrative mode of presentation of some perceptible property). His distinctive and controversial claim is that one only has to "grasp" such a contents in a given context in order to have a reason for belief. The idea is that when the conditions are met for having genuinely grasped (or understood) the content in question, then this suffices to give one a (defeasible) reason to have a belief with that content. There is no more basic account of the content of perceptual experiences available, which leaves their status as reasons open to question. For Brewer, epistemological considerations motivate the philosophy of perception.

John Foster takes a somewhat different approach. He addresses the question of the nature of perception initially as a question in its own right, and not as a part of epistemology. Epistemology later plays a role in deciding between theories that otherwise agree on the nature of perception - this is the approach Brewer rejects. Foster sees the debate as ultimately one between physical realism, that takes the physical world to be fundamentally mind-independent, and idealism, that denies this. Traditionally in the philosophy of perception, realists can be either direct realists, who hold that we directly perceive mind-independent objects in the physical world, and representative realists, who hold that our perceptual access to such objects is mediated by perception of mind-dependent entities. Foster's book is an extended argument for idealism, proceeding by eliminating direct realism, then representative realism. Since, according to Foster, his initial three are the only possible views, only idealism remains. The final section of the book then offers a positive argument for idealism. His idealism does not deny the existence of a physical world, of course, but simply asserts its fundamental mind-dependence.

Each book under review is a significant contribution to the philosophy of perception. Foster's book is certainly unfashionable, not just because the idea that physical reality as ultimately mind-dependent is out of tune with the prevailing contemporary naturalism and realism; it is also because many philosophers of perception would not accept Foster's traditional classification of the possible theories of perception. Many philosophers treat perceptual experiences as states of mind with "intentional content", such as beliefs, desires and intentions; a perception of a vase of flowers on the table represents the world as containing a vase of flowers on the table, in somewhat the way that a belief can represent the world as containing a vase of flowers on the table. (This is not to say that perceptions are beliefs, as some have claimed, but just to illustrate the fact that both perceptions and beliefs have intentional content.) On some views, such "intentional" states of mind can represent that something is the case even when it is not; thus perceptual error can be seen as a case of the kind of misrepresentation that occurs in belief.

Such an intentional theory of perception has something in common with Foster's representative realism, since it allows that a perceptual experience can have its content and phenomenal character independently of the existence of its objects; but it also has something in common with Foster's direct realism, since it allows that when a physical object is perceived, this is not by virtue of perceiving anything else. But the theory cannot be straightforwardly identified with one of Foster's three main kinds, and Foster has trouble with its classification. Rather than let this influence his theoretical taxonomy, he instead argues that the theory is essentially a version of representative realism. But this is not a description of the intentional theory that its proponents would recognise. (Part of the trouble here must derive from Foster's conspicuous neglect of almost all the literature on perception from the 1990s.) This objection to Foster's taxonomy is not an objection to his idealist conclusion. In fact, an idealist conception of the physical world seems perfectly compatible with an intentional theory of perception. Idealism is a view about the fundamental nature of the objects of experience, while the intentional theory is a theory of the states of mind involved in experience. It would be possible to hold that one's experience represents that physical objects are a certain way while also holding, on grounds independent of the philosophy of perception, that these objects are fundamentally mental in nature. My objection to Foster here is not to his idealist conclusions, but to his conception of the space of possible theories of experience.

In many ways, it looks as if the approaches represented by these two books could not be further apart. On the face of it, it seems that Brewer would reject Foster's whole project at the outset, since he regards it as a "datum" that perception presents its objects as mind-independent. One problem concerns what is meant by the objects of perception presenting themselves as "mind-independent". Certainly, when I look at the vase of flowers on the table, it does not seem to me to be dependent on this act of looking, in the sense that it would go out of existence if I closed my eyes. But no serious idealist, Foster and Berkeley included, will think that physical objects such as flowers are dependent for their existence on particular experiences. Idealists hold rather that reality in its fundamental nature is mind-dependent. (It is in this sense that Kant too is an idealist, and since Brewer is not an idealist, it is rather misleading for him to call his kind of approach to perception "neo-Kantian".) It is arguable that the fundamental nature of physical reality is not something that an account of the phenomenology of experience should be expected to yield. And if this is right, then whatever the data for the philosophy of perception are, they should not preclude an idealist conception of its objects.

Foster would presumably agree with this point. In fact, it is in the move from his account of perception in terms of sense qualia, which is common to his idealism and representative realism, to his final idealist conclusion, that he appeals to epistemological considerations. Representative realism, he argues, makes knowledge of the physical world impossible. Given the failure of direct realism, only idealism allows us to have direct knowledge of physical reality. It turns out, then, that both Foster and Brewer, in their very different ways, see epistemological considerations as at the heart of the philosophy of perception. But they disagree fundamentally about how a theory should understand the data of perception, what it is that a theory of perception is a theory of. And this kind of disagreement cannot be resolved or clarified by psychology or neuroscience; only more philosophy will do the job.

Tim Crane is reader in philosophy, University College London.