The methods of Dialect Topography in the Golden Horseshoe

Contents

2 Introduction

3 Dialect geography
3.1 The methods of dialectal geography
3.1.1 The questionnaire
3.1.2 The informant

4 Dialect topography
4.1 The informant
4.2 Time-effectiveness
4.3 Survey area: The Golden Horseshoe
4.4 Cartogram of the Golden Horseshoe
4.5 Critical view of the methods
4.5.1 Questionnaire
4.5.2 Distribution of age
4.5.3 Distribution of sex
4.5.4 Distribution of classes
4.5.5 Regional distribution of respondents
4.6 Regionality Index
4.6.1 Critical view on the Regional Index

5 Bibliography

2 Introduction

Dialect topography was invented by J. K. Chambers. He used this method in the Golden Horseshoe (v. 4.3 Survey area: The Golden Horseshoe ) for the first time. “Intended as an alternative to dialect geography” (Chambers 1994:35), dialect topography contains - like dialect geography - methods for surveying dialect variants in a region. In its basics dialect topography accords with dialect geography. “Both provide a macro-level perspective on linguistic variation. Both survey people in a continuous area, making it possible to identify and isolate gross linguistic differences among speakers from region to region. They can also provide the basis for charting linguistic change in subsequent surveys” (Chambers 1994:35-36). Although both methods are very similar, there are two important differences which will be discussed in the following passages.

In order to investigate the influence of non-native speakers on language use in the community, J. K. Chambers invented the Regionality Index (RI). This mechanism allows to identify and compare the languages of different test persons. The relationship between the subjects’ link to the region can be investigated as well as to which extent the subject talks like a local. This method will be discussed in the second part of this study.

3 Dialect geography

In order to understand the results of J. K. Chambers’ studies, it is necessary to have a short look at traditional dialect geography. Not only before the second half of the nineteenth century, the studying of dialect was taking place systematically. Before that time observations of dialect differences were only intuitive and casual. The first to study dialects more closely were the Neogrammarians. Having studied the interrelationship of many modern and classical languages, they started to search for general principles of language change. One of their principle stated: “Ausnahmslosigkeit de[r] Lautgesetze, or ‘sound changes are exceptionless.’[..] The result was the development of Dialect Geography, a methodology or (more accurately) a set of methods for gathering evidence of dialect differences systematically” (Chambers and Trudgill 1980:16-17)

3.1 The methods of dialectal geography

As most other linguistic branches, dialect geography “seeks to provide an empirical basis for conclusions about the linguistic variety that occurs in a certain locale.” (Chambers and Trudgill 1980:24) In order to get the empirical basis for conclusions, data have to be recorded and analyzed.

3.1.1 The questionnaire

The first study of dialect geography refers to Georg Wenker in 1876. With the intention of receiving as much data as possible, he composed a list of forty sentences, written in standard German. He then mailed this list to schoolmasters in the northern parts of Germany with the request of transcribing the sentences into their local dialect and to send them back afterwards. By mailing his questionnaire, Wenker reached nearly 50.000 schoolmasters. About 45.000 persons sent back a completely filled out questionnaire. However, this wealth of data forced to limit Wenker’s analysis. Each sentence of the questionnaire offered several points that could be recorded as regional variants. Therefore, he only looked on certain words in a closely circumscribed area.

Only in the beginning of dialect geography postal questionnaires were of importance. Soon they were replaced by trained fieldworkers, gathering data by interviewing the informants. In 1896, the editor Jules Gilléron worked out a questionnaire with about 1500 items. In the same year, he chose a fieldworker, Edmond Edmont, to interview people and to record their responses in a phonetic notation. For four years, Edmont cycled through the French scenery selecting data by conducting interviews. Periodically, he sent his results to Gilléron, which provided the possibility to start publication almost immediately. In the end he had visited 639 different country sites and had interviewed about 700 different people. The first publishing of Atlas linguistique de la France got underway in 1902. In 1910, the thirteenth and final volume was published.

Today the interviews are conducted by different fieldworkers who all follow the guidelines of one questionnaire. The more fieldworkers are engaged the more time is saved. The results of all interviews, conducted by different fieldworkers under varying circumstances, are comparable by having the same questionnaire for every fieldworker.

3.1.2 The informant

Traditional dialectologists tried to deal with examples of the “most genuine” (Chambers and Trudgill 1980:56) dialect. Therefore, they selected informants who were uneducated, untraveled and old. The influence of non-native speakers was of no importance; the characteristics of a subject were based on a prototype informant known as NORM, an acronym for nonmobile, older, rural males (Chambers and Trudgill 1980:33). However, when social conditions changed, dialectologists tended to handle the informant prototype not that strictly anymore. Hans Kurath, for example, included non-NORM speakers in his Linguistic Atlas of New England. In his survey, he interviewed people of different age and backgrounds. In order to compare the results he categorized the types of informants:

Type I: Little formal education, little reading, and restricted formal contacts.
Type II: Some formal education, usually high school; wider reading, and more social contacts.
Type III: Superior education, usually university; wide reading, and extensive social contacts. […]

Type A: Aged, or regarded as old-fashioned. Type B: Middle-aged, or regarded as more modern (Chambers and Trudgill 1980:34). Although Kurath did not only deal with uneducated old rural people, his informants still had to be nonmobile.

Traditional researchers are looking for the most conservative variety of a dialect because of their philological interests (Chambers and Trudgill 1980:16). However, the most conservative dialect is not necessarily the most ‘typical’ one. In order to get an accurate picture of all the spoken linguistic varieties, the aspects for choosing informants have to be different to the NORMs. The population of today is mobile, younger, urban and no longer dominated by man. It is the absolute opposite of NORMs. A second and important aspect of this issue is that, due to mobility, the number of people fulfilling the NORM-criteria is decreasing.

4 Dialect topography

4.1 The informant

The election of the informants does not follow the NORM scheme. In his study Region and Language Variation (2000), Chambers is interested in speech representatives of society in general. Therefore, his informants are of both sexes, all ages, and all classes. Even other social factors, such as the ethnic background, are of interest (Chambers 1994:36). In his study, he is especially interested in the influence of non-native speakers on the language use in the community. This excludes non-mobility as a criterion for representatives. He uses the traditional assumption that mobility is a great leveler of accent and dialect. Yet, instead of following this, by interviewing NORMs, he actually looks for persons who represent the whole society. Thus, he does not only want to investigate the speech differences that develop by the mix of people from different regions, but he also tries “to discover how the speech features imported by outsiders figure in linguistic variation and change” (Chambers 2000:179) Chambers treats region as an independent variable and invents a mechanism called Regionality Index (4.6 Regionality Index) which might provide “for the first time, an empirical basis for inferring the sociolinguistic effects of mobility” (Chambers 2000:169)

4.2 Time-effectiveness

Chambers’ dialect topography differs not only in the election of informants, but also in its time-effectiveness. His hope is that “future projects, with the Dialect Topography of the Golden Horseshoe as a prototype, could concurrently produce regional dialect atlases from start to finish within three years” (Chambers 1994:35). Studies in dialect geography usually takes some time, especially in former times. Hardly any other study was realized that efficiently as Gilléron’s thirteen volumes of the Atlas linguistique de France (3.1.1 The questionnaire). It only took fourteen years from inception to completion. Compared to other works this was very fast. Harold Orton and Eugen Dieth, for example, needed 28 years until their Survey of English Dialects (SED) was finished. Moreover, there are works that have not even been finished, yet, although they had been started more than fifty years ago. The Linguistic Atlas of the United States and Canada, for example, was funded in 1930. The survey area was divided into several regions which were treated as if they were self-contained projects. Each region had its own directors and fieldworkers. Hans Kurath was the director for the New England states. Yet, he also coordinated each regional survey from his office. The Linguistic Atlas of New England was published between the years 1939 and 1943. Immediately after the interviews in New England were completed, the work on the Atlantic states south of New England was started. However, it was not until 1961 that The Pronunciation of English in the Atlantic States was published. Twelve years later, from 1973 to 1976, The Linguistic Atlas of the Upper Midwest was published in three volumes (Chambers and Trudgill 1980:21-22). Until today, the Linguistic Atlas of the United States and Canada has “advanced to publication in less than one-seventh of the United States” (Chambers 1994:37). However, in almost every region of the United States, interviews have been done and are still being done in some of the central states. Consequently, there are several decades between the interviews of some cases. This leads to the problem that the results of the interviews are not current anymore. In any case, the quicker a linguistic study is completed and published, the more useful it is as a linguistic survey.

In order to precipitate his study, Chambers, surveying a much larger population than traditional dialectologists (4.1 The informant), used several new methods, based on contemporary conditions of our technological age.

First, Chambers took advantage of the fact that today usually everyone is literate. In the last centuries, not everybody was educated and especially the illiteracy of the NORM population was very high. Hence, fieldworkers were sent to the NORMs. However, Chambers used postal questionnaires in order to collect his information. Unlike Georg Wenker (3.1.1 The questionnaire), he did not send the questionnaires to teachers who were intermediaries between him and the NORMs. The postal questionnaires, with 86 questions (Table 1), were sent to the respondents and filled out by them. Chambers mentions “that in this regard dialect topography differs from dialect geography exactly as the U. S. Census differed from 1940, when enumerators interviewed each citizen, and 1970, when most citizen received their forms by mail” (Chambers 1994:38).

illustration not visible in this excerpt

Table 1 Content of Chambers’ questionnaire

[...]