Portal de Periódicos da CAPES

Seeing red, the metaphysics of colours without the physics

2005; Routledge; Volume: 83; Issue: 1 Linguagem: Inglês

10.1080/00048400500043936

1471-6828

Michael J. Watkins,

Quantum Mechanics and Applications

Abstract By treating colours as sui generis intrinsic properties of objects we can maintain that (1) colours are causally responsible for colour experiences (and so agree with the physicalist) and (2) colours, along with the similarity and difference relations that colours bear to one another, are presented to us by casual observation (and so agree with the dispositionalist). The major obstacle for such a view is the causal overdetermination of colour experience. Borrowing and expanding on the works of Sydney Shoemaker and Stephen Yablo, the paper offers a solution. Notes 1My comments about error-theories should be taken partly, of course, as tongue-in-cheek. I take the challenges raised by error-theories seriously, but they are not the direct concern of this paper. I should add, however, that the best defence against error-theories is a good offence; the way to respond to error-theories is to offer a viable account of colours, and that is the aim of my paper. For a critical discussion of error-theories, see Watkins [Citation2002]. 2You might mean something less by 'physicalism', of course. You might think that to be a physicalist is just to hold that colours supervene upon, or are realized by, physical properties. If that's the kind of physicalist you want to be, then you have my blessings. I return to this point in the conclusion. 3You might mean something less, or at least something else, by 'dispositionalism'. You might understand 'dispositionalism' in such a way that dispositional properties are intrinsic powers or intrinsic properties that contribute such powers [Shoemaker Citation1984; C. B. Martin Citation1996; Watkins Citation2002]. If you do, then again you have my blessings. 4The view of colours that I defend has its proponents, although they are scarce. John Campbell [Citation1997], for example, responds to the charge that sui generis colours cannot be causally explanatory. But as he admits, his response is only a sketch. He settles for making logical space for colours to play an explanatory role, but he does not provide an account of that role. In what follows, I provide that account. Yablo [Citation1995] presents an account similar to the position developed here. The similarities are not surprising since a central component of my position develops ideas from Yablo [Citation1992]. 5Watkins [Citation2002: Chapters 2 and 3] provides further discussion and criticism of physicalism. 6Byrne and Hilbert [Citation1997b] reply to this objection. 7Watkins [Citation2002] provides a more detailed discussion of causal saliency. 8I thank an anonymous referee for this criticism. 9The discussion above might make it appear that my view is further from Byrne's and Hilbert's than it actually is. So long as colours, at least those colours represented by human colour experience, are thought to be realized by surface spectral reflectances, I have no objection, at least none that I raise here. What I have tried to show is that it is implausible to treat those colours (e.g., red) as surface spectral reflectances or as disjunctive properties having reflectances as disjuncts. Indeed, I might even concede that determinable colours are reflectance-types, as Byrne and Hilbert sometimes state their position. But the relevant reflectance types cannot be thought of as sets of reflectances; instead the relevant reflectance types must be thought of as sui generis, as reflectance types that are not reducible to those, or to sets of those, of interest to colour science. I'll say more by way of comparing my account with Byrne's and Hilbert's in the conclusion. 10I do not intend this claim to apply to everyone who thinks of herself as a dispositionalist (see n. 3). 'Dispositionalism' here is being used to pick out positions according to which colours are relational properties of objects. Even so, you might object to Euthyphro-style arguments. Here's a purported counterexample, for which I thank an anonymous referee: To be popular is to be loved by the multitude. But surely Jones might love something because it is popular, and, let us suppose, everyone loves what Jones loves because, for each of them, they think it is popular. The purported counterexample, it seems to me, is not a counterexample. Perhaps Jones loves it because he believes it to be popular, and perhaps he believes it to be popular because it is. And perhaps everyone loves it because, for each person, they believe that it is popular. But I contend, with Plato, that it makes no sense to suppose that their believing that it is popular was caused by its being popular, and so it makes no sense to suppose that its being loved by the multitude was caused by its popularity. After all, until it came to be loved by the multitude it wasn't popular, and so its popularity could not explain its coming to be loved by the multitude. My point, then, is that redness cannot be the cause of something's looking red (not that it cannot be the cause of anything) if being red is just to be such that it would look red. Intuitively, if redness is causally responsible for something's looking red then redness must be part of the explanation for its being disposed to look red. Watkins [Citation2002: Chapter 4] discusses, in greater detail, the Euthyphro problem as it applies to colours. 11The account of colours that I recommend is similar to, but importantly different from, some current dispositional accounts. See the concluding section of this paper for a comparison and Watkins [Citation2002: Chapter 3]. 12Of course, Kim*'s use of 'Alexander' names Alexander*, Alexander's counterpart in Kim*'s world. But since Alexander and Alexander* share all of the same views, the difference between Alexander and Alexander* is not a difference that makes a difference for this paper. 13I should note that there might be additional problems in applying my specific proposal, and Yablo's general strategy, to mental properties. Mental predicates are sortals whereas colour predicates are not, and sortals, or the properties picked out by sortals, might present special challenges. For a discussion of these challenges, see Eric Marcus [forthcoming]. 14There are interesting philosophical and empirical problems concerning normal observers and normal conditions that I discuss elsewhere [Watkins Citation2002: Chapter 6]. 15Notice that disjunctive physical properties do not satisfy this criterion. Let P be some physical property and let Q be the disjunctive property of having P or R or S. Then it will be true that the instantiation of P necessitates the instantiation of Q. But the causal powers contributed by Q are not a proper subset of those contributed by P. Rather, the causal powers of Q will be the causal powers of P or R or S. If Q is realized by P, for example, then Q will contribute all of the powers that P contributes. 16I thank an anonymous referee for raising this challenge. 17Jonathan Cohen [Citation2003] defends a relational account of colours with which I have greater sympathy, an account according to which colours are thought of as functional properties of objects. My objections to dispositional accounts, I believe, apply against Cohen's account as well. But the debate between us is more difficult to settle and its resolution I must postpone for another occasion. 18This paper is dedicated to the memory of my good friend and colleague Charles Brown. I thank the audiences of Auburn University, Dalhousie University, and the University of Alabama at Birmingham for helpful discussion. I also thank Sydney Shoemaker, Kelly Jolley, Jody Graham, Eric Marcus, Roderick Long, James Shelley, Duncan MacIntosh, Sheldon Wein, and Richmond Campbell for helpful comments. Finally, the paper profited from the comments of three anonymous referees, whom I thank.