The Stories Inside Place Names

AA place name is often described as a photograph of the landscape at the moment it was coined, and the description is attractive enough to have survived a great deal of contrary evidence. Names containing an element for marsh, or for oak wood, or for the crossing of a stream are taken to prove that a marsh, a wood or a ford once lay on the spot, and in many instances they do. The difficulty is that the moment of coining is almost never the moment of first record. A name may be spoken for four centuries before a clerk writes it down, and during that interval the feature it describes may have gone, the community may have been replaced, and the sounds themselves may have drifted far enough to invite a false explanation. What survives on the map is therefore no photograph at all but something copied and recopied, and the historian's task is to work out how many stages deep it lies.

BSofia Verhoeven, who directs the Kelby Institute of Landscape History, has spent much of the past decade testing that suspicion against physical evidence. Her team matched 2,400 settlement and field names from the Wexholt basin against soil surveys and against pollen preserved in the basin's peat, so that the environment implied by a name could be set against the environment recorded beneath it. The agreement was closer than she expected. Of the names built on the element denoting standing water, 71 per cent stand on ground that the soil survey classes as poorly drained, while the pollen cores show woodland at the right dates under 48 of the 89 names built on the element for oak. Verhoeven argues that a correspondence of this order cannot be accidental, and that names may therefore be used, cautiously, to date changes in land use in districts where no written record survives.

CIain Tancred, a landscape historian at Ormsby, accepts the correlation and rejects the conclusion drawn from it. He contends that names are conservative labels which outlive the features they describe, so that a correspondence measured today tells us only that name and landscape once coincided, not when. More awkwardly for Verhoeven, some names appear to record scarcity rather than abundance: in one district Tancred examined, eleven of the sixty names containing the element for wood were first attested after the clearance of that wood, and seem to have been coined precisely because timber had become remarkable. Tancred concedes that his sample is small and drawn from tax rolls which name only those who held land, so that the dwellings of poorer tenants, and any names attached to them, are missing from his figures altogether.

DThe evidence behaves differently again where populations move. Settlers who left the region during the nineteenth century carried names with them, and the pattern of what they carried is instructive: among parties that emigrated together from a single parish, roughly three-quarters of new settlements received a name from home, whereas among families who travelled independently the proportion fell below one in ten, and indigenous names were adopted instead. The lesson is not that transferred names are worthless but that they record the manner of departure rather than anything about the ground the settlers reached. Where the two kinds have not been distinguished, a map of Wexholt names abroad has occasionally been read as a map of Wexholt conditions abroad, a conclusion the names cannot bear. Comparable patterns have been reported wherever a settled population was transplanted in bulk, although the proportions differ considerably from one destination to the next, and no convincing account of that variation has yet been offered by either side of the present argument.

EA further complication is introduced by the very sources that make research possible. Between 1934 and 1958 the national mapping agency regularised the spelling of some 40,000 names, replacing local written forms with versions judged more consistent, and in doing so removed a layer of evidence that had been accumulating for centuries. Spoken forms have proved more durable. A dialect survey conducted between 1978 and 1994 found that in about a third of the villages visited the pronunciation preserved a syllable that the modern spelling had lost, and in a handful of cases it preserved a word appearing in no document at all. The recordings are themselves perishable, since many of the speakers approached in 1978 were already elderly when the interviewers reached their village, and the last of the tapes have yet to be transcribed. Both researchers regard such recordings as the most under-used material in the field, though Tancred notes that they help only where the spoken form has descended independently of the written one.

FWhat, then, may safely be asked of a name? The two disagree less than their exchanges suggest. Verhoeven now accepts that a single name proves very little, and that the method works on populations of names rather than on individual cases; Tancred accepts that where a name, a soil type and a pollen record agree, the coincidence is worth something. The remaining difference concerns what the agreement dates. On Verhoeven's reading it dates the landscape; on Tancred's it dates only the survival of a label. The sensible position is probably that names are strong evidence for what a district was like and weak evidence for who was in it, and that their force in either case depends less on the name itself than on the material that can be set beside it.