University of Pennsylvania
Plenary
The cognitive process involved in sociolinguistic variation is the capacity to detect and respond to patterns that are general to the speech community. Minimal awareness is found for new sound changes in progress such as the Northern Cities Shift of North America, with uniform patterns of speech production across the speech community. Maximal sensitivity is found for stable sociolinguistic variables such as English (ING), which exhibit fine-grained social and stylistic stratification in production.
Recent experiments on listeners’ sensitivity to frequency investigated the properties of the sociolinguistic monitor which extracts and evaluates social information from speech production. Subjects heard a series of trials for a broadcasting position, with frequencies of the nonstandard varying from 0 to 100%, and were asked to rate the speaker on a scale of professional competence. For (ING), subjects exhibited a precise logarithmic function in which the effect of each deviation from the expected norm is the proportional increase in the sum total of deviations. This function is independent of the gender, ethnicity and region of speakers and judges. However, younger (adolescent) subjects show less precise responses, conforming to earlier finding that the uniform cognitive response characteristic of the adult community is acquired gradually with age and in proportion to social class status. Sociolinguistic variables which are less stable and less strongly marked, like the vocalization of /r/, exhibit less precise conformity to the logarithmic function.
There is reason to believe that information in the sociolinguistic monitor is stored independently of lexical and morphological information. Lower levels of sensitivity are shown in responses to the experimental reversal of the grammatical constraints characteristic of speech production. This is also evident in non-stochastic variables like the T/V pronouns, where each occurrence is separately evaluated. Although information on this sociolinguistic choice is dependent on morphological and syntactic analysis, the long-term history of the THOU/YOU alternation in English shows no effect of syntax and morphology on the decline of the alternation over several centuries.
Session: William Labov
Saturday, April 5, 2008, 15:45-16:45
room: aula