Relationalism’s psychosemantic ills

Relationalism about colour (as described in Cohen’s 2004 article on colour properties and his 2009 book The Red and the Real) is motivated by the thought that by identifying colours with relations between objects, subjects and viewing circumstances we can provide the best realist, general, ecumenical accommodation of perceptual variation data. This ‘best’-claim is supported by the charge that one of the view’s ecumenical rivals does uniquely badly in respect of fit with mainstream psychosemantics. This case for relationalism is complicated by Cohen’s later acknowledgement that relationalism has psychosemantic ills of its own. In this paper I argue that Cohen’s proposed remedying of those ills fails.