Since the early days of dialectological research, many of its practitioners have been at least partly concerned that it should be historical in its approach. By this we mean that dialect geography (the study of regional variations of phonetic and syntactic aspects of speech) should look to the past, and to older living speakers, for analysis.
The philologist Alexander J. Ellis was very keen to seek out older speakers of regional dialects in England for his late 19th Century work On Early English Pronunciation, of which Volume V was the first major dialectological survey to be published in Britain. In it, Ellis mapped out regional variation in British dialects, showing 10 ‘transverse lines’, which would later become called ‘isoglosses’. He remarked that “collecting country words is looked upon as an amusement”, but this was certainly not his own view. Rather, Ellis – and many dialectologists since – viewed the collection of dialects as a matter of urgency. The reasons for this can be linked to the industrial revolution, and since then the resulting continual improvement of infrastructure and communications in Britain.
Ellis was writing at a time when large numbers of British people were becoming much more mobile than previously possible as a result of the burgeoning railways system, and thought – correctly, as we must surely accept now – that this was leading and would continue to lead to the erosion of dialect groups and the boundaries between them. Soon after this, the English Dialect Society, led by Joseph Wright, followed Ellis’ lead somewhat by publishing the English Dialect Dictionary, which attempted to list all currently used non-standard words, as well as those that had been in use during the previous 200 years. This ambitious project might very well be termed ‘linguistic archaeology’, since it seems to concern itself with the past as an end in itself, rather than a means to an understanding of the origins of contemporary language. A larger project, the Survey of English Dialects (SED) was begun by Eugen Dieth and Harold Orton, and published between 1962 and 1978. The fieldwork for this enormous survey, the bulk of which was carried out by Stanley Ellis, deliberately and painstakingly drew its informants from small, rural village communities. It was also stipulated that the informants should be of at least 60 years of age, having been born and bred in their localities, preferably without having spent much time elsewhere during their lives, and having left school very early. These criteria, and the fact that the ratio of male to female informants was 2:1 (because women were thought more likely to alter their speech away from the vernacular towards the standard in the presence of an outsider) were designed to actively avoid a representative cross-section of the communities in the survey, and to create an overview of the history of their dialects. The selection of informants used for Gilliйron’s linguistic survey of France, begun in 1896, was similarly consistent. According to Milroy, “Gilliйron approached his linguistic survey of France by seeking out older male, uneducated speakers who lived in remote rural communities.” Despite Milroy’s’ claim, there seems to be no hard evidence to show that the largely homogeneous selection of informants was deliberately engineered, or that Gilliйron was a necessarily directly responsible for it. Milroy fails to acknowledge the role of his fieldworker, Edmond Edmont, who selected all the informants and conducted all their interviews between 1896 and 1900. However, whatever the reasons for the narrow choice of informants in this survey, it remains pertinent to this discussion because of its findings and consequent considerable influence on later dialect geography.
Both the French survey and the SED show the validity of this type of ‘linguistic archaeology’ in at least one sense: that, if the aim of its practitioners is to identify and record the regional variation of dialects in a language, and to identify the boundaries thereof with a view to preserving a record of the speech of a bygone era, it is certainly worth undertaking.
There is little doubt that, though the data collected in both these surveys, and particularly the SED, is somewhat overwhelming; it has been successful in making such a record. However, say Chambers and Trudgill, “the narrow choice of informants in dialect geography is probably the greatest single source of disaffection for it”. While Dieth and Orton had a clear aim in collecting data for the SED, it is now understandably considered somewhat irrelevant by some, simply because the regional differences and anomalies identified in it are now very largely alien to the current inhabitants of their respective regions. For this and other reasons, dialect geography is not confined to preserving that which is shortly to become obsolete. De Camp’s survey of Jamaican dialects (1971), encountered the problem that attempting to draw clear geographical boundaries between dialects is very difficult when dealing with contemporary variation. In Jamaica it was (and very likely still is) the case that geographical variations in dialect were inextricably linked with social class, and there were two main types of English spoken in Jamaica: ‘Standard English’ and ‘the Dialect’, which are generally spoken by the upper and lower classes respectively. De Camp was adamant that to view these as two discrete dialects spoken by separate groups of people was greatly hindering an understanding of dialect geography in Jamaica, because it was a gross simplification. He insisted that variations in Jamaican English should be viewed as a continuum rather than two groups, and that no individual held a static place in it. He also commented: “Each speaker represents not a single point but a span of this continuum, for he is usually able to adjust his speech upwards or downwards for some distance on it.” Thus De Camp was encountering a problem which had much less impact on the ‘linguistic archaeology’ studies of the SED and the survey of French dialects: that “individual speakers vary within a range which might overlap the range controlled by other speakers” (Milroy).
In considering whether ‘linguistic archaeology’ is a valid description of dialect geography, we should examine dialect geography’s place in the field of linguistics as a whole, and why it exists in the first place. In more recent years, dialect geography has become more closely linked with structural linguistics: “it was a drawback of traditional dialect geography that it tended to treat linguistic forms in isolation rather than as parts of systems or structures” (Chambers & Trudgill). It is no longer thought especially useful to draw isoglosses on the basis of one type of linguistic variation. For example, Orton and Wright’s interpretive map showing the regional variation in words and pronunciations used to describe what most of us now call a ‘newt’ is considered incomplete because it exists in isolation and not in the context of other similar or connected variations, as it should be. An example of structural dialectology is given by Chambers & Trudgill, who show that the locals of Lowestoft and Ipswich in East Anglia differ in their pronunciation of the words ‘road’ and ‘nose’ from the locals of Colchester, a town slightly further south. From this it appears appropriate to draw an isogloss to separate Colchester from the other two towns in this way, but further examination of accents in Lowestoft shows that locals there differentiate between the sounds of ‘road’ and ‘rowed’, as well as ‘nose’ and ‘knows’. Thus the Lowestoft phonological system has two different vowels at a point where the systems of Ipswich and Colchester have only one – and so it is more appropriate to separate Lowestoft from the other two towns. This type of structural variation cannot be shown by examining specific variations in isolation, as practised by Gilliйron and Orton. This clearly does not invalidate the findings of early dialect geographers, and indeed their findings may still be a very useful basis for analysing the structural variation of the dialects of yesteryear, but we can now see that the earlier researchers were not seeing the whole picture.
This placing of dialect geography in context has been cited as a reason for the resurgence of dialect geography in the second half of the 20th Century. Others have noted the rapid technological advances which have allow us to store vast amounts of data with ease, as well as manipulating them: “Computers are… wonderful tools for quickly sorting and matching pieces of information and for performing complex calculations on the results, and these days they are practically unlimited in their ability to store data” (Kretzchmar, Schneider & Johnson). However, I would contend that this reason should be linked with the decline and erosion of local dialects as mentioned earlier, and which Ellis and his followers were so concerned with. Computers, the internet, telephones, television, radio and almost every other medium has contributed to the breaking down of cultural (not just linguistic) boundaries worldwide, and I would contend that increasing worldwide cultural and linguistic ‘sameness’, along with the technological facility for preserving large amounts of data, has increased our desire to preserve the origins of our language – if not to keep them active then at least to keep a record. It has been suggested that the two most universally understood phrases ‘OK’ and ‘coca-cola’ are familiar to three-quarters of the world’s population. There is no research to back up this claim, but in principle I do not doubt the truth of it. This is merely indicative of the erosion of dialect boundaries. Some countries such as Wales have chosen to actively promote the use of their traditional language, since they fear its extinction. Unfortunately I think the ethereal power of the progression of language systems is far too powerful to be combated by these artificial measures, but the human desire to preserve the past is understandable and even commendable. For this reason, while it is not an all-encompassing term, ‘linguistic archaeology’ is an accurate and fair description of at least part of what we otherwise call ‘dialect geography’.