UC Berkeley

Visual attention plays an important role in epistemology. You come to know about things you see by paying attention to them, and thereby taking them up in thought and belief. Recent work in the cognitive sciences shows that attention also has a quite different role. Attention dictates not only what you take up in thought and belief, but also how things appear to you in visual experience. Attention enhances the neural signal that's processed for attended aspects of a scene, and so highlights those aspects by changing their visual appearance. For example, drawing attention to a colour makes it appear more saturated; drawing attention to a shape makes it appear larger. This raises a challenge for the epistemology of attention. It threatens to show that attention is a systematic source of illusion, rather than a reliable source of knowledge. Indeed, some scientists working in the area draw that conclusion.

To meet this challenge, I argue that visual experience is experience of determinable properties: properties which admit of more specific determinations, as red is determined by crimson, and 90 to 110 feet long is determined by 100 feet long. One determinate property determines many determinable properties. So one determinate shape or colour may take on different appearances, in veridical experiences of its different determinables. And one determinable property has many determinates. So different determinate properties may share an appearance, in veridical experiences of one determinable which they all determine.

I argue that the experimental data about attention and visual appearance are in fact well analysed in this way. For example, attention changes the appearance of a colour, giving it the same appearance as a more saturated but unattended colour; visual appearances remain veridical through this change, because they consist in experiences of different determinable colours, each of which the stimulus really instantiates. More generally, I argue that we need to recognise the role of determinable properties in vision, if we're to understand what experiments tell us about the contents of visual experience.

To show that visual experience is experience of determinable properties, I draw on empirical work about the limits of visual resolution. Visual processes are not sensitive to the finest details of a scene; they're sensitive to determinable properties, not maximally determinate properties. I argue that visual experience is experience of a property only where visual processes are sensitive to that property. I explore the metaphysics of determinable properties, arguing that they have a natural unity which merely disjunctive properties lack. On this basis, I show how visual experience of a determinable property has the phenomenological unity characteristic of experience of a single shape or colour.

With this account of visual indeterminacy in hand, we can also re-assess traditional assumptions about the nature of attention, and about the relationship between attention and visual experience. In traditional Anglophone philosophy, attention is often conceived as a window onto visual experience: attention gives you access to the contents of visual experience, but does not alter them. While this is clearly untenable in light of the empirical findings, a weaker view has attracted recent interest: visual experience is constitutively independent of attention; your seeing as you do does not consist partly in your attending. I make the relevant notion of constitutive dependence precise, and criticise Ned Block's argument for a version of this view, on the grounds that he underestimates the indeterminacy of visual experience.

I then argue that, in some instances, visual experience does depend constitutively on attention - and also vice versa. I defend William James's definition of conscious attention as the `focalization, concentration of consciousness'. So conceived, I propose, attention is not a further mode of consciousness, over and above perceptual experience and thought. Rather, attention consists in a focusing of these modes of consciousness. For example, experiments have shown that attention increases visual resolution: when you attend, you see more determinate properties. Here, I argue, attention and visual experience are mutually constitutive: they consist in one another. More generally, I show how to understand attention in terms of the focusing of conscious cognition and perception. This, I propose, is the form of conscious attention we exploit when we take up what we see in thought and belief, and thereby come to know about it. I conclude by exploring some consequences for intentionalism about visual experience.