Introduction

THIS article is about the contribution of historical linguistics to a reconstruction of the linguistic landscape that the Anglo-Saxons found on their arrival in Britain and to the way in which this linguistic landscape may have influenced the language of the newcomers. The bare bones are well known. British Celtic was spoken, both in the northern and western ‘Highland Zone’ and in the heavily Romanised, eastern ‘Lowland Zone’. Latin was spoken, too, especially in the Lowland Zone, although to what extent remains unclear. It has often been suggested that Latinity was probably very much a question of social status and profession: the Roman military, which came from various parts of the Empire, as well as those frequently dealing with it and with the structures of Roman power must have been fluent speakers of Latin. City-dwellers and estate owners were probably more exposed to Latin than people dwelling in the countryside and working the land. With allowance being made for local and social differentiation, it is plausibly assumed that Latin-Celtic bilingualism must have been widespread.

I propose to refine this picture in two ways. First, by arguing that, in the Lowland Zone, at least, Latin was the predominant language; and second, by presenting a case for distinguishing Lowland British Celtic from Highland British Celtic around 400 AD and for regarding the former as a Latinised variety of British Celtic with stronger connections with varieties of Celtic spoken in northern Gaul than with Highland British Celtic (it is the latter that produced the medieval languages Welsh, Cornish and Breton; the former has no offspring).