place-names provide a rich source of historical information, often for areas and subjects which are otherwise not well documented. The initial stages of their study are the province of the linguist because the original meaning of the majority of names is no longer immediately intelligible. The evolution of the name ‘Eboracum’ through ‘Everog’ and ‘Eoforwic’ to ‘Iorvik’ and thus to modern ‘York’ provides a good example of the complexities involved. Though the explanation of this sequence is a linguistic matter, it clearly has important implications for any historian concerned with settlement continuity and the changing control of this major centre.

England's place-names have been intensively explored by the English Place-Name Society and its publications provide the basic data for most of the country. They show that the majority of English names were formed in the Germanic languages of the Anglo-Saxon and Scandinavian settlers of pre-Norman Britain. It is to the Anglo-Saxons that we owe names ending in -ham and -tun, both indicating a village or farm, whilst the Scandinavian settlers spawned the numerous names in the north and east of the country which contain the elements -by (‘farm’), -thorpe (‘secondary settlement’), and -thwaite (‘clearing’). The relative dating and interpretation of all of these elements is still, however, a controversial issue. Other language groups are less well represented in the onomastic palimpsest; the relative paucity of Celtic names, together with the survival of only a small number of Romano-British names, is, for example, particularly marked and is obviously of relevance to studies of settlement and social change in the 5th and 6th cents. Similarly the advent of a Norman aristocracy added only a few French names, though the representation of their scribal system changed the spelling (and later, by extension, the pronunciation) of many English names. In the post-medieval period, names have continued to be coined; to this group belong most street-names and field-names which are valuable sources on changing patterns of industry, trade, and agriculture.

Scotland's names are linguistically equally complex, though they have been less intensively studied. There is an early stratum of P- Celtic elements (Gallo-Brythonic) which appear to represent the Pictish language: the c.300 names in pit- (‘piece of land’), such as Pitlochry, or the -pevr (‘radiant’) of Strathpeffer provide good examples of a group whose distribution is markedly north-eastern. Old English and Scandinavian forms to the south of the Forth–Clyde line reflect various extensions of Northumbrian and English power after the 7th cent. whilst the distribution of Scandinavian elements like setr (‘dwelling’) and stathir (‘dwelling’; ‘farm’) is the result of Norse activity in the northern and western isles and adjacent coastlands. It is, however, the Gaelic names which give Scottish maps their distinctive appearance. Within these can be seen forms like sliabh (‘mountain’), largely limited to Dalriada and Galloway, which represent a pre-7th-cent. phase. These contrast with the more widespread distribution of other names such as those containing baile (‘hamlet’; ‘farm’) which show the extent of Gaelic usage before the 17th cent. and the language's decline in the Lowlands.

Place-names in Wales are more readily intelligible to native speakers than those elsewhere in Britain for they have shared the general development of the Celtic language. But even here the linguistic make-up is far from straightforward, with interesting forms of Anglicization (e.g. Prestatyn—‘village of the priests’) and traces of Scandinavian naming (e.g. Swansea—‘Sveinn's island’) around the coast.

Richard N. Bailey