Secondary Qualities and the A Priori

Abstract

It seems current orthodoxy that one who grasps a concept of a particular secondary quality, e.g. redness, can discover a priori by analysis that the conception is of a secondary rather than a primary quality. I do not mean merely that the concepts of secondary qualities which we actually employ happen to have built into their meanings the a priori condition that the quality in question is a secondary quality. Rather, the orthodoxy is that any conception of a secondary quality must imply that it is a secondary quality, that one can always decide a priori whether a given quality is primary or secondary.' I hope to show that this orthodoxy is mistaken. I do not question that the meanings of our conceptions of colours, for example, have built into them the claim that colours are secondary qualities. However, I shall construct a counterexample to the orthodox view, concepts of COLOURS, such that COLOURS are in fact secondary qualities, yet nothing in the conception of COLOURS implies that they are. I shall then consider the best argument in favour of the orthodoxy known to me and show that it is fallacious. Crispin Wright has identified the core of the distinction between primary and secondary qualities as a distinction between concepts for which our best opinions as to whether they apply to an object track the extension of the concept (concepts of primary qualities) and concepts for which our best opinions determine the extension of the concept (concepts of secondary qualities) (Wright, 1988). Our best opinions as to whether a concept P applies to an object x or not (partially) determine whether or not P applies to x just in case there is a provisoed biconditional2 -hereafter a PBC for brevity-of the form