In some cases, colors descriminable in simultaneous presentations are indescriminable in serial presentations. Iâ€™ll call such cases â€œKelly Casesâ€ for they play a central role in Sean Kellyâ€™s arguments against the conceptual constitution of perception (hereafter, â€œconceptualismâ€) (Kelly, S. 2001. â€œDemonstrative Concepts and Experienceâ€ The Philosophical Review, 110, 3: 397-420).
Of course, a more accurate description of Kellyâ€™s target is a demonstrative-concepts defense of conceptualism. My intent here is to defend conceptualism without relying on demonstrative concepts.
Kelly cases raise trouble for conceptualism only if accompanied by certain assumptions. One assumption, discussed quite a bit by Kelly, is a re-identifiability requirement on concept possession: in order to have a concept of something, one must be able to identify that something on separate occasions. Another assumption, discussed very little, if at all, by Kelly, is that the perceptual contents in the simultaneous and serial presentations differ only with respect to their time of presentation.
The first assumption doesnâ€™t bother me too much. I question the second assumption.
There are lots and lots of cases in which the context of presentation messes with the discriminability of the colors presented. One of my favorites involves the color contrast cubes depicted below.
Figure 1. This is awesome.
In this image, the â€œblueâ€ tiles on the top of the left cube and the â€œyellowâ€ tiles on the top right are actually neither blue nor yellow but the same shade of gray. See a cool animated demonstration of this over at Dale Purvesâ€™s Lab webpage here.
Itâ€™s open, then, for conceptualism to be protected by treating simultaneous and serial presentations as different contexts that give rise to differences in perceptual content. Of course, the question arises of how to characterize the differences conceptually. It would be consistent with conceptualism to say something like that in the simultaneous half of the Kelly case, the concepts applied are a concept of, say, green plus the concept of difference-in-shade; and that in the serial half of the Kelly case, there is no application of the concept of difference-in-shade.
Note that the defense of conceptualism sketched here is not the demonstrative-concepts defense of McDowell and Brewer that constitutes Kellyâ€™s main target. I envision that the demonstrative-concepts defense would have to say something like that in the simultaneous half of the Kelly case, the concepts applied are the demonstrative concepts that-shade-1, that-shade-2, and the concept of difference; and that in the serial half of the Kelly case, there is no application of the concept of difference. The question arises, however, of what is going on besides a failure of noticing a difference in the serial half. It must be either that (1) the serial case involves neither that-shade-1 nor that-shade-2, (2) only that-shade-1 is applied, or (3) only that-shade-2 is applied.
An objection to this version of the demonstrative-concepts defense that may be raised at this point is that neither (1), (2), nor (3) would constitute the satisfaction of the re-identifiability condition on conceptual content. So, for example, in (2) the shade identified in the simultaneous half of the Kelly case as that-shade-2 is not being re-identified.
Thankfully, Iâ€™m not leaning on demonstrative concepts here, and thus the problem raised is somebody elseâ€™s problem.
[UPDATE (6/19/2007): I really don't like the last three paragraphs of this post.]