KENT JOHNSON and WAYNE WRIGHT COLORS AS PROPERTIES OF THE SPECIAL SCIENCES

We examine the pros and cons of color realism, exposing some desiderata on a theory of color: the theory should render colors as scientifically legitimate and correctly individuated, and it should explain how we have veridical color experiences. We then show that these desiderata can by met by treating colors as properties of the special sciences. According to our view, some of the major disputes in the literature about color anti-realism versus dispositionalism versus reductionism are not well-founded at this stage of scientific inquiry. Our account of color is designed to be of use in the sciences and as such is driven largely by considerations of what the various sciences need in order to proceed appropriately. We argue that a scientific theory of colors need not regard colors as anything more than high-level statistical constructs built out of correlations between color experiences and other phenomena. Do colors exist? Although philosophers have studied this question for centuries, recently it has been the subject of a renewed burst of attention. The number of particular positions taken up in response to this question likely outnumbers the number of philosophers working on it. In general, however, two camps have formed around the issue. On the one hand, the majority of philosophers are color realists. They hold that color is a property of physical objects (e.g., Smart 1975; Peacocke 1984; Averill 1985, Jackson and Pargetter 1987; Johnston 1992; Byrne and Hilbert 1997c, 2003; Tye 2000). On the other hand, many scientists and philosophers are color anti-realists, holding that contrary to our commonsense intuitions, colors are not a part of the world around us; i.e., nothing is actually colored (e.g., Hardin 1988; Boghossian and Velleman 1989, 1991; McGilvray 1994; Cosmides and Tooby 1995; Maund 1995). Unsurprisingly, there is no shortage of disagreements within each group. However, in many ways, the division between realism and anti-realism is as strong as possible, since it amounts to a dispute about whether color properties even exist. We share some sympathies with both camps. In particular, we agree with color realists insofar as we think that colors exist. But we think they exist because they are legitimate scientific constructs of the special sciences. Thus, there is room for us to agree with color antiErkenntnis (2006) 64:139 168 Springer 2006 DOI 10.1007/s10670-005-4310-8