The study of syntactic change has been much neglected in the past. Historical linguists have tended to concentrate on phonology, lexis and morphology whilst most theoretical studies of syntax have been deliberately synchronic in intention. In particular, theories of generative grammar have not been responsive to diachronic data and a fortiori have not yielded a convincing account of language change or of the interrelationships between different kinds of change. This study will be of interest to a wide range of linguists. It offers one of the first systematic accounts of a difficult and important topic, with implications for the whole field of linguistics and language study.