Introduction

Rules prescribed by grammar books will have to be distinguished from actual usage in this chapter; although they are connected, there is no easy and obvious relationship between them. Statements by writers of grammar books tend to be the further away from actual usage the more prescriptive the authors are – recent and ongoing change is likely to be classified as ‘mistake'. However, it is not always easy to describe with precision the clash between prescriptive correctness and linguistic reality; DeKeyser (1975) and Rydén & Brorström (1987) have shown for the areas they investigated how substantial the rift could be. Sometimes grammarians also focused their attention on rare items and artificial problems: than who was a combination invented by grammarians, and the existing than whom was too rare to merit much discussion (cf. 5.3.8 below). DeKeyser (1975:235) said he found only two instances in his 19th-century corpus of three million words; he concludes ‘than whom, the target of much criticism in 19th century grammars, is an exceedingly rare construction’.

However, if we want to judge syntactical norms, it is important to realize what the school grammars taught. The continuing impact of a Latin-based concept of what ‘grammar’ was is seen in the short treatment of syntax (reduced to problems of concord and government) in many grammar books; as late as 1841, Latham could state in his English Language that: