Speech acts are actions that are accomplished by uttering a sentence, such as “I promise to be there” or “Please be there.” They convey illocutionary forces such as promising or requesting, which situate the contents of the sentence into a discourse context with reference to the mental states of the speaker and hearer and the goals or actions that the utterance of the sentence accomplishes in the discourse. Their importance in defining the relation between linguistic form and semantic interpretation has long been known, through work by J. L. Austin [1] and John Searle [2]. At the same time, researchers in artificial intelligence interested in modeling comprehension of natural language, such as P. Cohen and R. Perraut, have incorporated an inference system based on illocutionary forces into their model.

This book is a straightforward and clear development of the formalization of speech acts in Searle [2] and Searle and Vanderveken [3]. As a proposal within the philosophy of mind, it assumes that the speech act, and not the proposition or sentence, is the basic unit of understanding of natural language. Semantics is usually based on truth conditions, rather than on contextually varying conditions of reference and mental states.

Vanderveken classes speech act theory as semantics rather than pragmatics, and he makes an interesting and bold attempt to analyze speech acts both in classical truth conditional semantics and in terms of the mental states of both speaker and hearer. The primary interest of this book for computer modeling of natural language understanding is the definition of inferences and entailments among speech acts and the taxonomy of different illocutionary forces and their strengths. These are arranged in partial orderings according to implicational relations among their components. The main proposals are generally intelligible, but familiarity with basic issues in the philosophy of language is helpful (though not essential), as is a basic knowledge of the model theoretic intensional logic used as the metalanguage for semantic translation.

This semantic approach, while not the answer to every puzzle and paradox about speech acts, is on the whole more successful and convincing than some of the pragmatically based accounts. The book as a whole is organized in a clear and perspicuous way, with apt specific illustrations of the abstract terms it discusses.