Collecting the data

The principles for the sampling process were to some extent based on the Leeds Survey of English Dialects (SED) criteria for the selection of localities and informants. However, since the SED focused primarily on phonological and lexical data, the fieldworkers in the HD, whose research centred on syntactic study of the dialects, had to adapt different methods of data collection, i.e. tape-recorded interviews of continuous speech. Using a questionnaire as the data collection method would have interrupted the informant's flow of thought and obstructed the original goal of recording long stretches of speech. Ossi Ihalainen (1981: 27) provides another reason for not using questionnaires: "Once the elicitation method becomes too suggestive, the dialect speaker is prone to avoid those features that he knows are stigmatised." The contrary is also possible, where the dialect speaker is aware of the features the interviewer is looking for and thus overuses them.

A male informant weighing
potatoes with the fieldworker

A question that is often raised is whether the fieldworker should be a native of the area. On one hand, the ability to understand and speak the dialect being investigated makes communicating with the informants easier. On the other hand, being too familiar with the dialect features might make the fieldworker act more as an informant than as an interviewer at the cost of objectivity. According to Anna-Liisa Vasko (2005: 40), the informants may actually feel more at ease when the interview is conducted by a non-native fieldworker. Foreigners' language mistakes are more easily accepted, and the informants can thus be more at ease during the talk session, less self-conscious of their own "mistakes".

The sex of the fieldworker was also an issue, since the majority of the fieldworkers in the HD project were female. According to Francis (as cited in Vasko 2005: 40), a female fieldworker would be spared from "improper" lexical items from "the kind of old fashioned rustic who constituted the usual informant in traditional surveys." The interviews conducted by female fieldworkers show very little evidence of any adaptation from the informant's part to the fact that the fieldworker was a woman. Anna-Liisa Vasko reports also that a female fieldworker might be more easily accepted, since older people might have some reservations on inviting strangers, especially men, into their homes (2005: 41).

In order to be accepted, it helped if the fieldworker tried to become an "insider" by, for example, taking part in village happenings, walking around the village and actually living in the village during the fieldwork. Fieldworkers must also have "the basic human qualities of understanding, sympathy and friendliness. It is essential to have patience, to be willing to cooperate, and to have a broad acceptance and understanding of all sorts and conditions of people." (Vasko 2005: 41.)

Selection criteria for localities and informants

Village shop in Landbeach

The data collection begins by first establishing a network of localities to be investigated and then finding suitable informants in these localities. Village post offices were often a great help in finding informants, since they held a list of pensioners, who were potential informants. Shops were also a good source of information, since the shopkeepers usually knew the villagers well. In addition, meeting local people with an interest in village life was of great help, since they could supply the fieldworker with additional information on the potential informants and they could even help out as interviewers themselves. They were also of help in distinguishing unfamiliar dialect forms in the speech of the informants.

A Cambridgeshire informant
presenting farming machinery

The informant selection criteria in the HD were similar to those of the SED. They were based on lifelong residence, social class and sex and age. Naturally, if the informants were lifelong residents of the community, their speech would include most typical features of that area. Short absences such as wartime service or brief stays in other villages in the area were accepted. As in the SED, most male informants were farm labourers or had other, similar jobs. The female informants were usually housewives. Informants were, in most cases, selected from the oldest age group, since a general view in dialectology persists that older speakers are more conservative in their speech (Vasko 2005).

The talk sessions

Local assistants with the fieldworker

It was necessary for the fieldworkers to make the talk sessions as natural and comfortable as possible. They were usually recorded in the informants' homes, where the familiar hustle and bustle in the house made the informant feel more at ease, albeit with a possible decrease in the quality of the recording. The informant was encouraged not to pay attention to the recording equipment, which were impossible to hide especially in the 1970s recordings that were made with reel-to-reel recorders.

Uninterrupted speech data were essential to the syntactic study of the speech. It was deemed important that the interviewer interrupt as little as possible the informant's flow of speech. Minimal interruption lets the informant finish their sentences and thoughts, resulting in a higher possible number of "clear" or non-doubtful items or constructions.

Ossi Ihalainen (right) interviewing

To make the talk sessions more natural, a neighbour, a relative or a close friend might be present (i.e. people who had not been selected as primary informants). If this happened, care had to be taken that they were of similar social status as the primary informant. To increase the naturalness, the informants were encouraged to talk about any topics they pleased at the time of the recording. All this was to ensure that the informants' speech would be as close as possible to their natural everyday speech.

However, the authenticity of the informants' speech was at best only an abstraction. The talk sessions were inevitably somewhat formal, and it must be assumed that the informants have an even more casual speech style that they use when not being recorded. This is the problem described by William Labov as the Observer's Paradox (Vasko 2005: 52). To overcome the paradox, many steps could be taken: creating a familiar interview setting, the presence of a spouse or a friend, interviews with two informants talking to each other, encouraging free conversation without questionnaires, etc.

When the interviews were conducted by local assistants (e.g. Vasko 2005), it was necessary to make sure that the interviewer and the informant were of similar social standing. If a higher-standing local was present, such as a vicar, the informants, being generally of the lower end of the socio-economic scale, would be more conscious of their language and thus the results would not be authentic enough.

The Farmer's Boy

Creating a comfortable and natural atmosphere was paramount in the talk sessions. One informant from Cambridgeshire had become so accustomed to the session and the fieldworker that he didn't mind singing a song at the end of the session.

F.R. from Willingham – The Farmer's Boy (farmersboy.mp3, 3:16, 2.99mb)

For a brief description of the song, click here.