This book focuses on some of the most important issues in historical syntax. In a series of close examinations of languages from old Egyptian to modern Afrikaans, leading scholars present new work on Afro-Asiatic, Latin and Romance, Germanic, Albanian, Celtic, Indo-Iranian, and Japanese. The book revolves around the linked themes of parametric theory and the dynamics of language change. The former is a key element in the search for explanatory adequacy in historical syntax: if the notion of imperfect learning, for example, explains a large element of grammatical change, it is vital to understand how parameters are set in language acquisition and how they might have been set differently in previous generations. The authors test particular hypotheses against data from different times and places with the aim of understanding the relationship between language variation and the dynamics of change. Is it possible, for example, to reconcile the unidirectionality of change predominantly expressed in the phenomenon of "grammaticalization", with the multidirectionality predicted by generativist approaches? In terms of the richness of the data it examines, the broad range of languages it discusses, and the use it makes of linguistic theory this is an outstanding book, not least in the contribution it makes to the understanding of language change.