Research
Perception
In a series of four papers, I try to sort out some consequences of thinking about perception in information-theoretic terms, inspired in part by Fred Dretske's Knowledge and the Flow of Information (MIT 1981) and in part by John Locke's Essay Concerning Human Understanding.
There is a largely intuitive idea within philosophy and cognitive science that perceptual states are in some sense imagistic while thoughts and other higher, cognitive states are somehow more linguistic in form. If perceptual states are like images, how are they like images? One thought is that perceptual states are mental representations that somehow resemble what they represent, and perhaps they represent what they do because they resemble it. This is a terrible way to think of pictures (see below), one might think, and it is doubly absurd when thinking about perceptual states. If one takes information theory seriously, then content is fixed somehow by information links, which are constituted by the right kinds of causal relations. There is no role for resemblance or anything like it to play in understanding perceptual representation.
In Isomorphism in Information-Carrying Systems (Pacific Philosophical Quarterly 85(4) 2004: 380-95) I argue that this plausible line of thought is misguided. There is an interesting role for a very specific kind of isomorphism between perceptual representations and what they represent to play. Such isomorphisms do not determine the information that a perceptual state carries, but they do determine whether the state makes the information it carries available to the perceiver. That is, perceptual states are imagelike in that they are isomorphic to what they are about in a special way and this is important because it allows such states to make the information that they carry available to perceivers.
Perceptual content, information, and the primary-secondary quality distinction (Philosophical Studies 122(2) 2005: 103-132) uses the result of the previous paper to argue for a kind of Lockean distinction between primary and secondary qualities. It's true that our ideas of colors are not isomorphic to the colors to the extent that our ideas of shapes are isomorphic to shapes. This puts us in a different epistemological situation with respect to these different kinds of qualities. We can know more about shapes on the basis of perceiving them than we can come to know about colors on that basis. This does not treat secondary qualities as metaphysically any different from primary qualities (contra most readings of Locke), but it vindicates his epistemological intuitions concerning colors vs. shapes by appeal to the ways in which perceptual states are isomorphic to what they represent.
What is what it's like? (Synthese 156(2) 2007: 205-229) articulates a kind of perceptual mode of presentation, inspired by the notion of extractability disussed in the previous papers. The perceptual MoP of a property P is, roughly, the set of properties a creature can come to know about perceptually whenever the creature perceptually represents P. I propose identifying these perceptual MoPs with what it is like to perceive a property. Creatures could be alike in perceptually representing, for example, a shade of green, but differ in what it is like to do so because each represents that shade of color under a different mode of presentation. Perceptual MoPs are useful for a number of reasons, among which is that they fit quite well with information-theoretic approaches to perceptual content, while such approaches have been accused of failing to make room for such an idea.
Perceptual content is vertically articulate (American Philosophical Quarterly 44(4) 2007: 357-369) (Penultimate draft) The content of a mental state is vertically articulate when it represents some property, P, as well as some properties that are abstractions from P. For example, the concept scarlet represents a rather specific class of colors, and these colors are all reds, but scarlet does not also represent red. Similarly, the concept square does not represent being quadrilateral, even though all squares are quadrilateral. Perceptual states, by contrast often represent properties across levels of abstraction. This helps to explain how perceptual states perform the function of making information available to perceivers. I show that both a Dretske-style information theoretic approach to perceptual content and a Rosenthal-style appproach to the same both have the consequence that perceptual content is vertically articulate.
These papers, along with some other things I have been working on, might coalesce into a book in the near future. In the meantime, I have been thinking about how to identify the objects of perception from an information-theoretic perspective (Dretske failed in this regard, I think), teleological approaches to semantics, and introspection.
Some of my thoughts about introspection and consciousness will appear in Introspective availability (Philosophy and Phenomenological Research, forthcoming). That paper argues for a new way of handling the uses to which Ned Block puts the Gil Harman's inverted earth example. In short, it claims that while facts about our phenomenal lives are introspectively available, a semantic externalist should expect that we would not notice certain radical changes in our phenomenal lives: precisely those proposed in the inverted earth slow switch cases. This is phenomenologically implausible to many, but it fits well with a semantic externalist's program and it provides a straightforward way to handle inverted earth without abandoning representationalism. The link above is to the penultimate draft.
Artifacts
The best place to get a sense of my approach to artifact representations is a short but fairly polemical piece, Pictorial representation (Philosophy Compass 1(6) 2006: 535-46). (Penultimate Draft) .
My book, On Images (Oxford, 2006) DOI: 10.1093/019929075X.001.0001 ) argues that what it is to be a picture does not fundamentally concern how such representations can be perceived, but how they relate to one another syntactically and semantically. This kind of approach, first championed by Nelson Goodman in his Languages of Art, has not found many supporters in part because of weaknesses with Goodman’s account. That is a shame because a properly crafted structural account of pictures has many advantages over the perceptual accounts that dominate the literature on this topic.
Part I (Chapters 1-5) presents the account and draws out some of its immediate consequences. In particular, it explains the close relationship between pictures, diagrams, graphs, and other kinds of non-linguistic representation. Also, it undermines the claim that pictures are essentially visual by showing how many kinds of non-visual representations, including audio recordings and tactile line drawings, are genuinely pictorial.
Part II (Chapters 6-10) shows that the structural account of depiction can help to explain why pictures seem so perceptually special. That is to say, once we understand the structure of pictorial representational systems, we are in a better position to understand why we perceive them in the ways that we do.
Part III (Chapters 11-12) provides a new account of pictorial realism and shows how accounting for realism relates to an account of depiction in general.
The book is available through Oxford Scholarship Online here or as a real book through the usual channels. I will be expanding this overview of the book in the near future. Recently, Alon Chasid reviewed the book for the British Journal of Aesthetics and Katerina Bantinaki reviewed it for Mind . I will post links to other reviews if and when they become available.
A couple of papers ultimately made their way into the book, too. Image structure (Journal of Aesthetics and Art Criticism 61(4) 2003: 323-40) covers, albeit in less depth, some of the material that became Part I of the book. Pictorial realism as verity (Journal of Aesthetics and Art Criticism 64(3) 2006: 343-54) ultimately became chapter 11 of the book, but this version relies less on other material from the book than the book chapter does.
Artifact expression is forthcoming in New Waves in Aesthetics , edited by Kathleen Stock and Katharine Thomson-Jones (Palgrave Macmillan 2008). Most philosophical accounts of expression sadly do what most philosophical accounts of representation do: they downplay the significance of understanding symbolic structure in favor of trying to understand how consumers of representations make use of them. As On Images suggests, the best way to understand how consumers use representations is first to understand their symbolic structure. The same holds true for expression, which is a form of aboutness in artifacts that relates in interesting ways to representation in artifacts. My proposal is that an artifact expresses some property just in case it seems to have that property insofar as the object indicates that about itself. This treats artifact expression as symbolically akin to the expression of feelings in people. Someone expresses anger when they seem joyous in virtue of seeming to indicate that about themselves (via facial "expressions", movements, etc.) This proposal does a surprisingly good job of accounting for expressive phenomena in artifacts, or so the paper argues. The link above is to the penultimate draft, and I will link to the book when it is published.
More Coming Soon!