(from Introduction) This book addresses three main topics. First, the general nature of nonliteral language is described. Definitions of many tropes are given, including relatively obscure ones like metonymy and synesis. Second, the nature of metonymy and metaphor is investigated in some depth. Related semantic phenomena, such as literalness, inconsistency, and lexical and case ambiguity, are also examined. Various computational and noncomputational approaches to these phenomena are reviewed. Third, an approach to processing metonymy and metaphor is presented, which has three parts. Part one is a model of language in which metaphor and metonymy and other language phenomena are decomposed into simpler elements. Part two is an artificial intelligence (AI) theory of metaphor and metonymy, called Collative Semantics (CS for short), based on the language model. Part three is an implementation of CS in a working computer program called metaS, which recognizes metaphors, metonymies, and various non-tropes (such as contradiction and redundancy) in short English sentences. (Dan Fass)