Dialectology Research Paper

1. Beginnings

Its beginning is conventionally dated as 1876, when the German professor Georg Wenker mailed a list of sentences written in standard German to schoolmasters in the north of Germany and asked them to return the sentences to him transcribed into the regional dialect of their school districts. Wenker made further mailings until 1887 by which time he had blanketed the entire nation. He ultimately sent his list to nearly 50 000 schoolmasters, and he received completed forms from 45 000 of them, The schoolmasters, with few guidelines and no precedents, provided Wenker with thousands of regional variants of all kinds: lexical, pronunciation, phonological, morphological, and syntactic. Overburdened by data, Wenker was forced to limit his analyses to a few variants found in a closely circumscribed area of north and central Germany, which he then painstakingly transcribed by hand onto two sets of maps bound as Sprachatlas des Deutschen Reichs and deposited in Marburg and Berlin in 1881 (Wenker 1881).

Thirty years before Wenker, the Finnish polymath Anto Warelius (1821–1904), approached the study of dialect from an entirely different perspective in a scholarly undertaking that remains virtually unknown outside his own country, but which has a prior claim over Wenker as the first truly systematic study. In 1846, Warelius set out on foot from Hamina in southeastern Finland and walked in a northwesterly line, eliciting and transcribing words and expressions from the villagers he encountered across almost 400 km (Rapolo 1969, Chambers 2000).

Warelius remains an obscure figure in the history of dialectology, but this method of eliciting dialect forms in face-to-face interviews with villagers foreshadowed the predominant method used by professional dialectologists for the next 150 years. The influential precedent for the method of direct fieldworker elicitation was set in 1896 by Jules Gillieron, a Swiss professor at the Sorbonne in Paris, in the dialect survey of France. Gillieron trained an obscure grocer, Edmond Edmont, in the craft of phonetic transcription, and dispatched him into the French countryside on his bicycle. Edmont has become a legendary figure in dialectology for his unflagging devotion as he pedaled into 639 villages and interviewed 700 paysans in less than four years. He sent his transcriptions to Gillieron in Paris at intervals, where they were incorporated into a cumulative mapping enterprise, and publication of the maps began in 1901 and ended in 1910 with the thirteenth volume of the Atlas linguistique de France (ALF). Gillieron and Edmont’s atlas (1901–10) stands as an early consummation of dialect geography and became the influential precedent for future projects for almost a century.

2. Impetus

The motivation for these methodical, large-scale studies in the second half of the nineteenth century is conventionally held to be a reaction to Neogrammarian philology, one of the great intellectual currents of the day. The Neogrammarians demonstrated law-like regularities in the historical descent of modern languages from classical ones. One of their tenets was Ausnahmslosigkeit der Lautgesetze ‘sound changes are exceptionless.’ Such a hypothesis, bold and admirably empirical, invites refutation, and a likely source would come from dialect data gathered from highly differentiated, relatively isolated villages.

The view that the first dialectologists rose to the Neogrammarian challenge has romantic resonance, but it is not strongly supported by facts. Neither Wenker nor Gillieron nor any of their successors actually marshaled their data for elucidation, revision or refutation of any overriding linguistic principles, Neogrammarian or otherwise. In practice, the main results of their scholarship were descriptive rather than analytical, and their expressed goals aspired to no more. The Sprachatlas became the preferred medium for dissemination of results, with linguistic tokens carefully arrayed on geographic representations, usually with commentary on secondary variants which had to be omitted because they would overcrowd the map, and occasionally with isoglosses, lines drawn between locations where different variants tended to occur.

Rather than Neogrammarianism, the main impetus for dialectology as a systematic discipline appears to have been Linnaean taxonomy, another forceful current of nineteenth century intellectual history. Dialect data lends itself quite naturally to the taxonomic bent, with numerous gradations subcategorizable as alloforms of -emic categories roughly analogous to genera and their species.

Though taxonomy dictated the form that dialect studies took, the ultimate impetus has a much deeper source in human nature. History is replete with observations about dialect, going back at least three millennia in the Western tradition to the myth of Babel in Genesis 11: 1–9, whereby the Lord punished his ambitious subjects by ‘confusing their speech, so that they will not understand what they say to one another.’ Since Babel there has been a continuous tradition of astute comment on dialect differences which makes a long prelude to the institution of dialectology as a systematic classificatory enterprise. The English scholar Alexander Ellis remarked in 1875 that ‘collecting country words is looked upon as an amusement, not as laying a brick in the temple of science’ (quoted by Chambers and Trudgill 1998, p. 15). Ellis’s spirit thrives, and is evident today in word columns in newspapers, tourist pamphlets on the local ‘twang,’ stand-up mimicry in theaters, news features on slang, and in dozens of other ways. The discipline of dialectology appears to be the scholarly manifestation of this natural human curiosity about language variation.

3. Methods

Until recently, dialectology gave primacy to recovering the oldest extant forms of language. This deliberate bias colored its methodology at every point. Fieldworkers chose their subjects as a judgment sample of NORMs, an acronym for a prototype who was nonmobile, older, rural, and (predominantly) male. For instance, Edmont’s 700 subjects, all of whom were rural, included only about 200 educated people and 60 women. The NORM prototype was purposeful. Lack of mobility guaranteed that their speech would be characteristic of their village. Harold Orton, director of the Survey of English Dialects (SED), England’s nationwide survey on Gillieronian principles for which data was gathered from 1950 to 1961, said, ‘dialectspeakers whose residence in the locality had been interrupted by significant absences were constantly regarded with suspicion’ (Orton 1962, 15–16). Old age guaranteed that their speech would reflect a bygone era. Hans Kurath, director of the Linguistic Atlas of New England (LANE), part of the Linguistic Atlas of the USA and Canada that was begun in 1931 and still continues, declared that ‘since most of the informants … are over 70 and not a few over 80, it would seem that we shall be able to establish the regionalism of the pre-industrial era of New England’ (Kurath 1949, p. 27). Rurality eliminated mobile urban dwellers who were likely to be exposed to other dialects and accents (Chambers 2000). Males were targeted because their speech is less class-conscious than women’s. As Orton put it, speaking about England, ‘In this country, men speak vernacular more frequently, more consistently, and more genuinely than women’ (1962, p. 15).

The basic methodology of traditional (Gillieronian) dialect geography is to send trained fieldworkers to the locations where the subjects live in order to elicit specific items from the subjects guided by a standard questionnaire. The questionnaires are lengthy (Gillieron’s no less than 1500 questions, the SED about 1200, LANE about 700), with questions usually ordered by subject matter (flora and fauna, the barnyard, chores, etc.). Its purpose is to guarantee comparable data from all subjects. The fieldworker is expected to transcribe only the key word or phrase from the subject’s response, no matter how long it might be. Labov summarized the method as ‘a long question from the interviewer and a short answer from the subject’ (Labov 1971, p. 113).

These methods have been subject to criticism from commentators versed in social-science methodology, especially with respect to the reliability of the population sample and the validity of the data (for instance, Pickford 1956). Obviously, NORMs constitute a rarefied minority when the majority population is exactly the converse—mobile, younger, urban, and female (Chambers 1993, p. 134). But criticisms like these largely miss the point about the antiquarian bias that motivates the methods. Raven I. McDavid, a venerated fieldworker on the American project and an articulate apologist, defended the traditional methods with this succinct capitulation: ‘There is a deliberate bias in the choice of communities—in the direction of smaller and often isolated places; there is a deliberate bias in the selection of informants, insistence on local roots and a statistically disproportionate sample of the older and less educated; there is a deliberate bias in the choice of items for the questionnaire, in the direction of the old fashioned rural society, the horsedrawn vehicle, the mule-powered plow and homemade bread. All of these biases are essential to the purpose of the investigation, to push the historical records as far back in time as possible’ (McDavid 1981, p. 71). Even as he articulated this defence, he inadvertently exposed the limits of the tradition at a time when modern technology had eliminated horsedrawn vehicles and social forces were inexorably diminishing the ranks of the uneducated and, indeed, rural society.

4. Structuralist Analyses

With the advent of Saussurean structural linguistics in 1916, language studies took a decisive turn in the direction of synchronic analysis. Taxonomy fell out of favor and intensionality gained adherents, at least in theory. However, dialectology clung to its autonomy, no more moved by structuralism than it had been by Neogrammarianism. Over the years, linguists made several attempts at integrating dialectology into linguistics. For instance, Uriel Weinreich (1954) proposed a level of analysis called the diasystem, which in effect posited a level above the phonemic where ‘diaphones’ could be segmented into the actual phonemes in two (or more) specific dialect systems by means of interpretive rules. Weinreich speculated that speakers must use something comparable to a diasystem for comprehension when they interact with speakers of other dialects.

Later, with the development of generative frameworks, there were similar attempts at integrating dialectology. For example, Keyser reanalyzed traditional atlas data from the Atlantic coast of the USA by demonstrating that several vowel distinctions found in the painstaking tabulations for dozens of speakers could be generated simply by positing two phonological rules and ‘accounting for the formation of new dialect groups through the geographical dissemination of rules’ (Keyser 1963, p. 313). Efforts like these gained few adherents, and they remained peripheral to the linguistic mainstream and had little impact on dialect studies.

More successful were the uses of atlas data in elucidating problems of historical linguistics. For instance, Hall (1943) used data collected in the Sprachund Sachatlas Italiens und der Sudschweiz ([AIS] Jaberg and Jud 1928–40) to demonstrate the persistence of the old political boundary of the Papal States on patterns of dialect differentiation in twentieth-century central Italy. The purposes of historical linguistics were of course consonant with the intentions of traditional dialectology, although the exploitation of the latter by the former offered much greater potential than was actually realized. It did, nevertheless, establish a vital, if minor, tradition.

5. Decline And Rebirth

As post-Saussurean language studies continued to be dominated by synchronic linguistics and entered a period of considerable currency stimulated by the ideas of Noam Chomsky, the failure of dialectology to become integrated caused it to slip ever further into the periphery of language studies. By the middle of the twentieth century, scholarly activities had slowed and become relatively inconspicuous. By 1980, even dialectologists began to realize that reform was necessary. In the first book-length summation of the field, Chambers and Trudgill stated: ‘one might well wonder what the future holds for dialect geography. There is no doubt whatever that the financing of research and publishing for large-scale surveys is a great problem, and there is also no doubt that capable and enthusiastic practitioners of the calibre of those who went before are difficult to find; but it seems most likely that both of these facts follow from another fact, namely, that dialect research has taken a new direction … It is probably safe to say that the future of dialect geography depends upon the ability of its practitioners to embrace and incorporate the concerns and perhaps the methodology of urban dialectology’ (Chambers and Trudgill 1980, p. 23).

Many aspects of dialect geography came under criticism when they were noticed at all, including the utility of the Sprachatlas as a medium for disseminating results. ‘For whom are these expensive volumes designed?’ Macaulay asked, and then he said: ‘Far from being a visual aid the maps are probably a barrier to grasping the pattern of distribution and it seems very unlikely that anyone nowadays would attempt this kind of mapping. Given the sorting power of computers, basic information is much better stored in a form that is suitable to machine processing than in conventional map form’ (Macaulay 1985, p. 180).

With the founders gone, the rearguard reaction to these criticisms was muted. Starting in the mid-1980s, dialectology underwent a renaissance from which it would emerge as a vital and vigorous discipline with greater worldwide currency than ever before and, not coincidentally, greater intellectual depth. The sources were the antidotes identified in the two critiques above: the embracing of powerful new technology and the adaptation of more inclusive research methods.

6. Computerization

From its inception, dialectology had been burdened by an overabundance of data. Dialectology aspired to be, in Saussurean terms, a science of parole, that is, the study of language as a communal effusion, apprehended in actual speech events, imbued by and indeed mediating what Sapir called ‘the apparent lawlessness of social phenomena’ (Sapir 1929, p. 166). But the profusion of data proved daunting, and the methods used in classic dialect geography came into being partly as devices for delimiting the effusion and subjugating its lawlessness, by focusing on a narrow, linguistically conservative population, restricting elicitation to a closed set of items, and transcribing only individual tokens from the flow of speech.

The development of tape-recording as a reliable, affordable technology in the 1950s provided a means for making a permanent record of the flow of speech. However, the immediate impact of tape-recording on dialectology was negligible, coming into use merely for recording traditional elicitation sessions and thus serving mainly to provide archival documentation of the practices of McDavid and other fieldworkers. The most innovative of all regional surveys, the Linguistic Atlas of the Gulf States (Pederson et al. 1986–92 [LAGS]), directed by Lee Pederson, made a significant advance by recording all its interviews so they could be transcribed later under laboratory conditions. Meanwhile, beyond the formal confines of dialectology, tape-recording was embraced as a liberating medium by a few linguists genuinely interested in studying vernacular speech. In this way, tape-recording stimulated the rise of sociolinguistics, an urban dialectology with methods and goals that bore little resemblance to dialect geography.

More significant technology for the revitalization of dialectology came with the development of computers as reliable data-handling devices in the late 1960s, and especially as affordable and accessible universal processors in the 1980s. As Kretzschmar and colleagues put it: ‘ … the development of dialect studies, whether geographical or sociolinguistic, has always been hampered by a superfluity of data … Even smaller surveys have had to settle for selective analysis of their data because the wealth of possibilities for analysis overran the editors’ time and the human capacity for holding in the mind only so much information at once. Computers can help overcome these problems: they are wonderful tools for quickly sorting and matching pieces of information and for performing complex calculations on the results, and … they are practically unlimited in their ability to store data’ (Kretzschmar et al. 1989, p. v). Many dialectologists recognized the computers’ potential for relieving the perpetual problem of overabundance. No less important, many scholars recognized the enormous advantages that would accrue from the computerization of the classic databases of the ALF (Heap 1997), SED (Ramisch 1997, Elmer and Rudin 1997), AIS (Goebl 1993), and the others. Generalizations buried in lists and tables for many decades could be disinterred by judicious computation.

7. Dialectometry

A significant development directly attributable to computational processing is dialectometry, literally, the ‘measure of dialect.’ The word dialectometrie was coined by Pierre Seguy, director of the Atlas Linguistique de Gascogne, one of the regional surveys comprising the Nou el Atlas de la France. Seguy hoped to replace subjective judgments on the occurrence of dialect regions by quantifying the Gascony data in such a way that the regions would be revealed objectively—in effect, they would reveal themselves. In the sixth and final volume of the Gascony atlas (Seguy 1973), Seguy and his associates—mesvele escamarades, he called them—published a set of maps, which linked each of the survey sites to its nearest neighbors by a number representing the percentage by which the two sites differed with respect to the forms elicited there. Seguy conceived the percentages as an index for measuring the ‘linguistic distance’ between the two places. The analyst, by inspecting the distance indices, could draw inferences about dialect regions based on degrees of similarity and difference.

Though Seguy’s methods now look primitive, restricted as they were at the time to univariate comparisons and hand-made cartography, the numerical gradations across the maps rightly impressed many dialectologists as the first approximation of a visual analogue for the dynamics of dialect gradation in space. Since then, dialectometry has advanced to multivariate statistical programs and computergenerated multidimensional displays that simultaneously compare all sites to one another. Dialectometry has become a generic term for all kinds of statistical dialect geography linked, as Goebl says, by a common goal: ‘The preponderant aim of dialectometry consists in discovering, by the numeric combination of many low ranking patterns, higher ranking patterns which have remained hitherto hidden, in order to obtain a systematic insight into the problem of the basilectal management of space by _{HOMO LOQUENS}’ (Goebl 1993, p. 277). Practitioners of dialectometry can be found around the world from Japan to Canada. The numerous research programs remain largely uncoordinated and partly incompatible, but in the next decades the consolidation of methods and goals will undoubtedly crystallize into a powerful branch of dialectology unimaginable in the first century of the discipline.

8. Sociolinguistic Dialectology

The other revitalizing impulse for dialectology is sociolinguistics, which had its effective beginning in the 1960s with the dissemination of William Labov’s fountainhead studies in Martha’s Vineyard and New York’s Lower East Side (collected in Labov 1972). Linguists began studying language variation with freer data-gathering methods, using larger and more representative subject samples, and they sought to understand its social significance with quantitative, acoustic and statistical analyses. The intellectual forebears of sociolinguistics, far removed from Linnaean taxonomy, were post-structural linguistics and the social sciences.

Dialect geographers, not surprisingly, did not immediately embrace the new currents, and it took a generation schooled primarily in sociolinguistics to recognize that the two disciplines were both, in essence, dialectologies. With hindsight, the rise of sociolinguistics can be viewed as a natural response to global social forces that have arisen since dialectology came into being, including urbanization, mobility of all three kinds—geographical, occupational and social —embourgeoisement, mass education and mass literacy. Those forces shrank the traditional domains of dialect geographers by reducing geographically isolated regions and socially isolated subjects to statistical insignificance in all the developed nations and beyond.

Observations of linguistic variation became refocused on social class, network, age, sex, ethnicity, and other social sub-groups as well as region, and the systematicization of those observations required new sets of independent variables and new methods for tracking them. In the social context that came into existence a century after Warelius and Wenker, the remaking of dialectology as sociolinguistic dialectology was inevitable. As Chambers and Trudgill said, ‘A decade or two ago, it might have been possible to think that the common subject matter of dialectology and sociolinguistics counted for next to nothing. Now we know it counts for everything. Dialectology without sociolinguistics at its core is a relic. No serious perspective on dialectology can grant urban research and variation theory less than a central role’ (Chambers and Trudgill 1998, p. 188).

One of the historical consequences of dialectology becoming sociolinguistic is that it has become an integral branch of linguistics. Like other branches of linguistics, its seminal works and formative investigations have been impelled and largely defined by specific issues, that is, by the need to develop or refine or refute particular hypotheses. The defining issues include these (discussed in Chambers 1993), among others: (a) How do the changes dialectologists observe in progress reflect the changes that occur in the history of languages? (b) How, if at all, do the grand concepts of the sociology of language—concepts such as diglossia, language shift, heteronomy, linguistic legislation and planning—affect social interaction, or become encoded emblematically in sociolinguistic variants? (c) How do dialect features diffuse throughout regions, and how does such diffusion affect the linguistic behavior of people in the community? (d) What is the adaptive function of dialect and accent? This last question, in particular, can be posed as a kind of ultimate question for all dialect studies.

There is a long tradition that presupposes dialect and accent to be dysfunctional, starting with the Babelian myth. Dialectology in all its methodological and intellectual guises shows that, on the contrary, in every human society linguistic variation prevails and thrives. It is a profoundly human attribute that is essentially irrepressible. The large, open-ended questions of dialectology can only begin to find their answers in fine-grained, minute research into the social significance of linguistic variation. The answers to those questions will ultimately elucidate the human language faculty. With that as its goal, dialectology constitutes a vital branch of modern linguistics.

Bibliography:

1. Chambers J K 1993 Sociolinguistic dialectology. In: Preston D (ed.) American Dialect Research. John Benjamins, Amsterdam Philadelphia, pp. 133–64
2. Chambers J K 2000 Region and language variation. English World-Wide 21: 169–99
3. Chambers J K, Trudgill P 1980 Dialectology. Cambridge University Press, Cambridge, UK
4. Chambers J K, Trudgill P 1998 Dialectology, 2nd edn. Cambridge University Press, Cambridge, UK
5. Elmer W, Rubin E 1997 The survey of English dialects as an electronic database for research in areal and variationist linguistics. In: Thomas A R (ed.) Issues and Methods in Dialectology. Department of Linguistics, University of Wales, Bangor, Wales, UK, pp. 234–46
6. Gillieron J, Edmont E 1901–10 Atlas Linguistique de la France. Champion, Paris, 13 Vols [ALF]
7. Goebl H 1993 Dialectometry: a short overview of the principles and practice of quantitative classification of linguistic atlas data. In: Kohler R, Rieger B B (eds.) Contributions to Quantitative Linguistics. Kluwer Academic, Amsterdam, pp. 277–315
8. Hall R A Jr 1943 The papal states in Italian linguistic history. Language 19: 125–40
9. Heap D J 1997 La Variation grammaticale en geolinguistique: les pronoms sujet en roman central. Department of Linguistics, University of Toronto: Toronto Working Papers in Linguistics
10. Jaberg K, Jud J 1928–40 Sprachund Sachatlas Italiens und der Sudschweiz. Ringier, Zolfingen [AIS]
11. Keyser S J 1963 Review of Kurath and McDavid, Pronunciation of English in the Atlantic States. Language 39: 303–16
12. Kretzschmar W, Schneider E, Johnson E (eds.)1989 Computer Methods in Dialectology. Journal of English Linguistics 22: special edition
13. Kurath H 1949 Word Geography of the Eastern United States. University of Michigan Press, Ann Arbor, MI
14. Labov W 1971 Some principles of linguistic methodology. Language in Society 1: 97–120
15. Labov W 1972 Sociolinguistic Patterns. University of Pennsylvania Press, Philadelphia
16. Macaulay R K S 1985 Linguistic maps: visual aid or abstract art? In: Kirk J, Sanderson S, Widdowson J D A (eds.) Studies in Linguistic Geography. Croom Helm, London, pp. 172–86
17. McDavid R I 1981 On obtaining spontaneous responses. In: Warkentyne H. J. (ed.) Papers from the 4th International Conference on Methods in Dialectology. University of Victoria, British Columbia, pp. 66–84
18. Orton H 1962 Survey of English Dialects: Introduction. E. J. Arnold; Leeds, UK (SED)
19. Pederson L et al. 1986–92 The Linguistic Atlas of the Gulf States. University of Georgia Press, Athens, GA, Vols 1–7 [LAGS]
20. Pickford G 1956 American linguistic geography: a sociological appraisal. Word 12: 211–33
21. Ramisch H 1997 Dialectological and cartographical features of the computer developed linguistic atlas of England. In: Thomas A R (ed.) Issues and Methods in Dialectology. Department of Linguistics, University of Wales, Bangor, Wales, UK, pp. 224–33
22. Rapola M 1969 Johdatus Suomen Murteisiin. Suomalaisuuden Kirjallisuuden, Seura
23. Sapir E 1929 The status of linguistics as a science. Language 5: 207–14
24. Seguy J 1973 Atlas linguistique de Gascogne. Centre national de la recherche scientifique, Paris, Vol. 6
25. Weinreich U 1954 Is a structural dialectology possible? Word 10: 388–400
26. Wenker G 1881 Sprachatlas des Deutschen Reichs. MS folio. Marburg University