ABSTRACT
This chapter takes the position that metaphor is one of the more important and interesting of the phenomena involved in human communication. According to Reddy, the facts of actual linguistic behavior do not uphold any presumption about the existence of such a priori intuitive categories, either in human language or in physical nature. Like Bickerton, Reddy also rejects any notion of a priori semantic hierarchies for metaphor study in favor of semantic considerations determined purely on the basis of actual language use. A major argument against approaching metaphor from the point of view of selectional deviance is, of course, simply that the existence of these restrictions, along with the form of analysis from which they sprang, have been called into question. The author has seen the word 'castle'; however, take in actual fact something best described as 'the theories of pure syntax' for its referent in the context of the chapter.
