Tuesday, 21 February 2012
Colour Constancy and the Relativities Therein
Derek Brown, Department of Philosophy
12:05 pm, 817R Cathedral of Learning
Abstract: Perceptual relativity refers to the fact that in some sense the ways things appear to us in perception change as a number of perceptual variables are altered. One tradition uses this fact for many purposes, to argue that we do not directly perceive an objective or mind-independent world, that properties such as colours are relative to perceptual circumstance or even mental, and so on. One can claw back from these worries by focusing not on the ways perceived things vary across perceptual circumstances, but instead on the way they don’t, to focus on perceptual constancies. Perceptual constancies are themselves an actively studied area in science, and with respect to my focus, colour, the computational approach to constancy has been particularly influential in philosophy. Thus emerges a familiar dialectic. Some philosophers take constancies to underwrite direct perceptual realism and an objectivist approach to the metaphysics of perceived properties. Others regard constancies as little more than a set of reported judgments we are inclined to make about what we perceive. Both orientations are unnecessarily extreme.
The challenge is to articulate the appropriate amounts (and natures) of what is constant and what is relative in a particular class of constancy perceptions, in my case colour perceptions. My proposed solution is embodied in a Colour Layering Thesis: in colour perception we are regularly given more than one colour along a single line of sight, and in turn often see one colour through a distinct transparent one. Often one such colour is a surface colour, and another an illuminant colour, but filter and volume colours can also be present. To the familiar objectivist ontology of coloured surfaces, films and volumes I thus add coloured illuminants. The idea, resisted by many colour relativists, that we can have varying perspectives on a colour is give a concrete explication: we have a different perspective on a colour when we see it through another colour, or see another colour through it. In addition we must recognize that sometimes we only partially perceive a thing’s colour, for example its lightness but not its hue. Within these constraints we can give a general account of colour constancy and the variety of relativities we find through its study.