This book criticizes the methodology of the recent semantics-pragmatics debate in the theory of language and proposes an alternative. It applies this methodology to argue for a traditional view against a group of contextualists and pragmatists, including Sperber and Wilson, Bach, Carston, Recanati, Neale, and many others. The author disagrees with these theorists who hold that the meaning of the sentence in an utterance never, or hardly ever, yields its literal truth-conditional content, even after disambiguation and reference fixing; it needs to be pragmatically supplemented in context.

The standard methodology of this debate is to consult intuitions. The book argues that theories should be tested against linguistic usage. Theoretical distinctions, however intuitive, need to be scientifically motivated. Also we should not be guided by Grices Modified Occams Razor, Ruhls Monosemantic Bias, or other such strategies for meaning denialism. From this novel perspective, the striking examples of context relativity that motivate contextualists and pragmatists typically exemplify semantic rather than pragmatic properties. In particular, polysemous phenomena should typically be treated as semantic ambiguity. The author argues that conventions have been overlooked, that theres no extensive semantic underdetermination and that the new theoretical framework of truth-conditional pragmatics is a mistake.