Qualitative Beliefs, Wide Content, and Wide Behavior

Abstract

It is a common opinion among philosophers of mind that the phenomenal or qualitative character of one's mental states is underdetermined by both actual and potential behavior. The sort of example most frequently used to support this view is the case of spectrum inversion. It is argued, perhaps most cogently by Sydney Shoemaker (1981), that the color experiences of one individual might qualitatively differ from those of another, even in the absence of any possible behavioral differences. Thus, it may be that Jack's color spectrum is inverted with respect to Jill's (e.g., Jack has a sensation of green when and only when Jill has a sensation of red) despite the fact that Jack and Jill exhibit all the same behavior, and have all the same behavioral dispositions. Of course, such an inversion could obtain only under very special conditions. For one thing, to ensure that the qualitative differences between Jack and Jill do not manifest themselves in verbal behavior, their use color words will also need to be inverted. For example, Jack must find it natural to use the word 'red' only in the presence of the color sensation that would prompt Jill to use the word 'green'. Moreover, for any two points A and B along Jack's color spectrum, the distance between A and B will need to be exactly the same as the distance between the corresponding points along Jill's color spectrum. Only then can we ensure that under the same environmental conditions, they will make all the same judgements of relative color similarity and difference-that is, they will always concur in judgements of the form Object A is more similar in color to object B than to object C. Whether all relations of qualitative similarity and difference could ever be preserved in a case of human spectrum inversion is an empirical question regarding the structure of our color experience. However, as Shoemaker notes, this empirical issue does not alter the fact that it is at least metaphysically possible