The Problem of Social Class Grouping in Sociolinguistic Research
Bibliographic Data
| ID | 8729950 |
|---|---|
| Authors | Lawrence M Davis (corresponding author) |
| Year | 1985 |
| Volume | 60 |
| Issue | 3 |
| Pages | 214-214 |
| Publication date | 1985-01-01 |
| Peer Reviewed | Yes |
| Open Access | No |
| Type | ARTICLE |
| Venue | American Speech (JOURNAL) |
| Journal identifiers | ISSN: 0003-1283 • E-ISSN: 1527-2133 |
| Publisher | JSTOR (PUBLISHER) |
| DOI | 10.2307/454885 |
| OpenAlex | W2319724167 |
| Language | EN |
| Citations received | 3 |
| References cited | 4 |
SN DAVIS (1982) I POINTED out that some American sociolinguists either have failed to take advantage of statistical methods or have used statistics inappropriately at times. This paper carries that notion a bit further, into the area of how respondents are grouped into social classes.' In some cases, such as the comparison of the linguistic behavior of the sexes, the grouping is a supposedly natural one: males and females. In the case of social class, however, nature is generally not the determining factor, since many sociolinguists use a quantitative scale to rank their subjects. For example, in his study of the speech of black Detroiters, Wolfram (1969) used four criteria to establish his subjects' social class: education, occupation, residence, and income. Wolfram's scale for education is listed here as table 1, and he had similar scales for his other criteria. The result is that Wolfram's subjects received scores ranging from 20 (the highest social class) to 134 (the lowest). His social classes are listed as table 2, and he added the extra precaution of leaving at least seven points between the highest score of a given social and the lowest score of the next higher level (Wolfram 1969, 35-36). Wolfram's decision to use four social classes was based on sociological research done, however, twenty years before. His use of quantitative methods was modeled on those of Labov's classic study of New York City speech (1966). In fact, after Labov, almost all serious studies of social dialects have used quantitative methods, pioneered by Warner, Meeker, and Ellis (1949) to determine social class.2 Table 3 presents some hypothetical data for glottal stop realization for /t/ in words such as water and bottle. The social class rankings of the subjects are also given. Assuming that we have already decided on four classes, we could divide the subjects in table 3 according to what appear to be the more-or-less natural breaks in the socioeconomic class ranks, as in table 4; the classes are ranked here from the highest (class 1) to the lowest (class 4). Then, we would get the results in table 4 for mean, or average, glottal stop realization for the four groups. These results are, in fact, quite neat. We can show the emergence of progressively more nonstandard forms as
Class (philosophy · Genealogy · Linguistics · Sociolinguistics · Sociology · Computer Science · Language, Discourse, Communication Strategies · Linguistic Variation and Morphology · Multilingual Education and Policy · Artificial Intelligence · History · Philosophy
| Unique citing works | 3 |
|---|---|
| Citations per year | 0,07 |
| Citation span | 1986 - 1991 (6) |
| Citation velocity | historical |
| Highly cited | No |
| Citation types | Neutral: 3 |