This work is the culmination of an eighteen-year collaboration between Ken Hale and Samuel Jay Keyser on the study of the syntax of lexical items. It examines the hypothesis that the behavior of lexical items may be explained in terms of a very small number of very simple principles. In particular, a lexical item is assumed to project a syntactic configuration defined over just two relations, complement and specifier, where these configurations are constrained to preclude iteration and to permit only binary branching. The work examines this hypothesis by methodically looking at a variety of constructions in English and other languages.
Hale and Keyser's collaborative work is always interesting. However, I couldn't give this book a 5-star rating for two reasons: 1) there were many places in this book where they provide speculative accounts, which I didn't find very convincing; and 2) their proposal seems to work for English, some native American languages, and probably some other languages too, but I don't see how it would account for causative-inchoative alternation in some other languages like Semitic languages. A good theory of argument structure needs to have that kind of power to account for as many languages as possible.