ABSTRACT
Most philosophers seem prepared to accept the principle that the relation of indiscriminability between phenomenal colors is nontransitive. This chapter argues that if philosophers are prepared to count statistical ensembles of observations as observational data—a quite common practice in science—Nontrans and the arguments which depend upon it must be rejected. Nontrans has been used as a stick with which to beat sense-datum theorists as well as to cast doubt on the possibility of giving a rational reconstruction of the semantics of everyday color terms. In both cases, what is called into question is the coherence of phenomenal color predicates, or, alternatively, the concept of a phenomenal color. An examination of the actual practice of assigning color predicates will show how that practice is based upon the structure of phenomenal color and suggest some of the factors which determine the foci and boundaries of color terms.
