Stylistic Variation as an Opportunity for Panel Studies of Change: «Solutions»
Given the problems and challenges of situational and metaphorical variability in language use that repeated recordings tend to reveal, what “solutions” for panel studies of language change might we consider? One might be to obviate the challenge by restricting our studies to one interview/recording per subject, or focusing on just the most casual or vernacular style netted, given Labov’s (1966 and elsewhere) contention that that is the most important style for the study of linguistic variation and change. But the one-study solution is an obvious non-starter for those of us committed to the in-depth study of variability and change, and concentrating only on the most casual or vernacular study also leaves much to be desired. For instance, we may not realize that what we recorded in “Danger of Death” or similar contexts is NOT a person’s deepest vernacular until a second recording, sometimes by chance (see Rickford 1987: 153-420, 2014a, 2019: Ch. 4). And it may be important to know whether change occurs across a wide variety of styles, casual AND careful, and in which style it is initiated first and spread most quickly.
Stylistic Variation in Panel Studies of Language Change 87
Ghyselen's Solution: Five Settings
A solution to this issue might be the approach increasingly being taken by some European sociolinguists. For instance, Ghyselen and de Vogelaer (2018, see also Ghyselen 2016) report on a study of Flanders, the northern, Dutch-speaking part of Belgium, in which the linguistic repertoires of ten highly educated women in leper were studied through recordings made in five speech settings (Ghyselen and de Vogelaer 2018:7):
- 1. a dialect test,
- 2. a standard language test,
- 3. a conversation with a friend from the same city,
- 4. a conversation with a friend from a different dialect area, and
- 5. a sociolinguistic interview with an unacquainted interviewer from a different dialect area.
This approach, with multiple recordings of each individual, is likely to yield a more variegated and complex picture of individuals’ longitudinal deployment of their linguistic repertoires. Overall, setting and interviewee- sensitive approaches, as are increasingly being adopted in panel research, provide a much richer basis for a study of variation and change over time than a single sociolinguistic interview.
It still remains to be decided what contextual elements to control and code for in sociolinguistic recordings and analyses of the future. What “solutions” to the challenges posed by stylistic variation would be most helpful for panel studies of variation and change and in computerized corpora from such interviews and recordings? In Rickford (2014a), I suggested that these should “include basic aspects of the setting, scene, participants and perhaps purposes, key and local norms of interaction”, elements that have been shown to fundamentally impact on speaker’s interactional choices.