Many of the objects that we perceive have an important characteristic: When they move, they change shape. For instance, when you watch a person walk across a room, her body constantly deforms. I suggest that we exercise a type of perceptual constancy in response to changes of this sort, which I call structure constancy. In this paper I offer an account of structure constancy. I introduce the notion of compositional structure, and propose that structure constancy involves perceptually representing an object (...) as retaining its compositional structure over time. I argue that compositional structure is represented in visual phenomenology, and I also assemble empirical evidence in support of the claim that compositional structure is recovered by the visual system. Finally, I draw out consequences of this account. I argue that structure constancy has implications for the predictive capacities of perception, and that the phenomenon places important constraints on viable accounts of both the format and reference frame of visual experience. (shrink)

The orthodox monadic determination thesis holds that we represent colour relations by virtue of representing colours. Against this orthodoxy, I argue that it is possible to represent colour relations without representing any colours. I present a model of iconic perceptual content that allows for such primitive relational colour representation, and provide four empirical arguments in its support. I close by surveying alternative views of the relationship between monadic and relational colour representation.

I motivate and defend a previously underdeveloped functionalist account of the metaphysics of color, a view that I call ‘quality-space functionalism’ about color. Although other theorists have proposed varieties of color functionalism, this view differs from such accounts insofar as it identifies and individuates colors by their relative locations within a particular kind of so-called ‘quality space’ that reflects creatures’ capacities to discriminate visually among stimuli. My arguments for this view of color are abductive: I propose that quality-space functionalism best (...) captures our commonsense conception of color, fits with many experimental findings, coheres with the phenomenology of color experience, and avoids many issues for standard theories of color such as color physicalism and color relationalism. (shrink)

The fact that you see some particular object seems to be due to the causal relation between your visual experience and that object, rather than to your experiences’ phenomenal character. On the one hand, whenever some phenomenal element of your experience stands in the right sort of causal relation to some object, your experience presents that object (your experience’s phenomenology doesn’t need to match that object). On the other hand, you can’t have a perceptual experience that presents some object unless (...) you stand in the right sort of causal relation to that object (no matter how closely your experience’s phenomenology matches some object). According to the continuity thesis, property perception is similar to object perception in these two respects. A standard reason to reject the continuity thesis is the assumption that the environmental properties that a perceptual experience presents are determined by its phenomenal properties (which are not determined by the environmental properties that cause the experience). I maintain that the continuity thesis is false but for a different reason: perceptual experiences present both objects and properties via manners of presentation; but, whereas perceptual manners of presentation for objects are purely relational, perceptual manners of presentation for properties are satisfactional. (shrink)

In the philosophy of perception, representationalism is the view that all phenomenological differences among mental states are representational differences, in other words, differences in content. In this paper I defend an alternative view which I call external sortalism, inspired by traditional adverbialism, and according to which experiences are not essentially representational. The central idea is that the external world serves as a model for sorting, conceptualizing, and reasoning surrogatively about perceptual experience. On external sortalism, contents are construed as a kind (...) of gloss on experiences themselves. We can retain what is attractive about representationalism, namely, that perceptual experiences can be evaluated for accuracy, without problematic commitment to the idea that they bear a substantive, representational relation to external objects and properties and that this relation determines the phenomenal character of experience. (shrink)

Imagine that the person you see through the looking glass is a real person, with her own experiences, living in an environment that is the mirror-reverse of yours. You look at your right-hand glove as you put it on your right hand; she looks at her left-hand glove as she puts it on her left hand. You feel your heart beating on your left side; she feels her heart beating on her right side. You hear a bird chirping out the (...) window to your left; she hears a bird chirping out the window to her right. Is there any difference between your experience—your what-it-is-like phenomenal character—and hers? She is looking at her left-hand glove, the mirror-reverse of you as you look at your right-hand glove. Does she see what you see, a right-hand glove, or the reverse of what you see, a left-hand glove? Since her environment is the mirror-reverse of yours, and she seems to get by in it just fine, you'd probably assume, if you had to guess, that she sees it correctly. She is looking at her left-hand glove, after all. So it stands to reason that she is having an experience as of a left-hand glove, not a right-hand glove. And it stands to reason—it seems obvious—that an experience as of a left-hand glove differs in what it is like from an experience as of a right-hand glove. In this paper, I will argue that this last presupposition, obvious though it may seem, is false. I'll argue that (unless dualism is true) there is no intrinsic difference between what it is like to experience a left-hand glove and what it is like to experience a right-hand glove: if there are any relevant experiential differences, these are either merely comparative (differences in what it is like to compare gloves) or cognitive (differences in what it is like to bring an experience of a glove under a concept like LEFT or RIGHT). So your mirror-reverse twin, who won't differ from you in what she is comparing or thinking, will have the same (type of) experience as you. I'll call the position I am arguing against orientation–experience categoricalism, and the position I am arguing for orientation–experience relationism. Orientation–experience relationism is a surprising thesis, one whose falsity you might take for granted. I will argue for it by cases. Things go wrong for the orientation–experience categoricalist in one sort of way if we imagine that you and your mirror-twin's experiences differ: basically, this leads to absurdity if we imagine that you and she both live at a world with a non-orientable spatial topology (in the sense in which a Möbius strip or Klein bottle are non-orientable). Things go wrong for the orientation–experience categoricalist in a different sort of way if we assume that you and your twin would experience the same, but that someone else might have the experience that we were just imagining that your twin might have. Under these assumptions, we have to accept the existence of partial zombies, who lack the relevant phenomenology but behave as if they have it, at very close, empirically possible worlds. In an appendix, I argue that property dualism is the best alternative for those who cannot bring themselves to accept orientation-experience relationism. (shrink)

Thomas Reid famously claimed that our perceptual experiences reveal what primary qualities are in themselves, while providing us with only an obscure notion of secondary qualities. I maintain that this claim is largely correct and that, consequently, any adequate theory of perception must explain the fact that perceptual experiences provide significantly less insight into the nature of secondary qualities than into the nature of primary qualities. I maintain that neither naïve realism nor the standard Russellian variety of the content view (...) can provide a satisfying explanation; instead, in order to provide a satisfying explanation, we must posit that perceptual experiences represent properties via Fregean modes of presentation. Further, I maintain that we must depart from the standard Fregean variety of the content view in two important respects. First, the relevant modes of presentation must be characterized without appealing to causal relations between perceptual experiences and perceived properties; and second, we must posit that primary and secondary qualities are represented via modes of presentation of different kinds. The resulting view is that primary qualities are represented via perceptual modes of presentation analogous to highly detailed descriptions, while secondary qualities are represented via perceptual modes of presentation analogous to impoverished descriptions. (shrink)

There is variation in how people perceive colors and other secondary qualities. The challenge of perceptual variation is to say whose perceptions are accurate. A natural and influential response is that, whenever there’s variation in two people’s perceptions, at most one of their perceptions is accurate. I will argue that this leads to an unacceptable kind of ignorance.

It is common for an object to present different colour appearances to different perceivers, even when the perceivers and viewing conditions are normal. For example, a Munsell chip might look unique green to you and yellowish green to me in normal viewing conditions. In such cases, there are three possibilities. Ecumenism: both experiences are veridical. Nihilism: both experiences are non-veridical. Inegalitarianism: one experience is veridical and the other is non-veridical. Perhaps the most important objection to inegalitarianism is the ignorance objection, (...) according to which inegalitarianism should be rejected because it is committed to the existence of unknowable colour facts. The goal of this paper is to show that ecumenists are also committed to unknowable colour facts. More specifically, I argue that, with the exception of colour eliminativism, all major philosophical theories of colour are committed to unknowable colour facts. (shrink)

Some truths about color are knowable a priori. For example, it is knowable a priori that redness is not identical to the property of being square. This extremely modest and plausible claim has significant philosophical implications, or so I shall argue. First, I show that this claim entails the falsity of standard forms of color functionalism, the view that our color concepts are functional concepts that pick out their referents by way of functional descriptions that make reference to the subjective (...) responses of perceivers. I then argue that, while some sophisticated forms of color functionalism can accommodate the a priori knowability of a truth like “redness is not identical to squareness,” they can only do so by abandoning color realism, the thesis that colors are instantiated by external material objects. In practice, color functionalists are almost invariably color realists. Thus, given extremely modest assumptions concerning what can be known a priori about color, we should conclude that color functionalism, at least in its typical realist form, is false. (shrink)