Linguistics最新文献

Contemporary dialects of Goranî (a Northwestern Iranic language spoken in Kurdistan) exhibit considerable variation in the formation of tense-aspect-mood categories. It has long been recognized that compatibility of the indicative/imperfective prefix mi- or the subjunctive prefix bi- with present-tense verb stems is phonologically conditioned. However, all attempts to identify the specific conditioning environments have failed. In this article, we reexamine the diachronic development of these affixes, showing that subsequent sound changes have obscured the original conditioning environment for the retention of these affixes. Early work on Hewramî (the main variety of Goranî) shows how the pattern of verbs that have lost the prefixes has begun to be extended to the verbs that take these prefixes for synchronically opaque reasons. The least frequent of these verbs were the first to begin the process of extension, showing inconsistent marking in these early works. Moreover, younger generations continue this process of extension using many more unmarked verbs compared to older speakers. The findings of this paper have implications for cyclic developments in the domain of tense and aspect.

Incomplete mastery of dialect grammar engenders 'hyperdialectisms' which may be unconscious errors, but which may also be the result of indexical resourcefulness, viz. the profiling of a regional identity. Fifty younger and older speakers from the Brabantish city of Eindhoven (Netherlands) were first administered an acceptability judgment task containing correct forms and three types of hyperdialectisms featuring gender and number constraints. Following the survey, the same respondents participated in a focus group discussion. Regression analysis on the scaled ratings revealed that all three types of hyperdialectisms were rejected, although it was especially the older respondents (almost all L1 dialect speakers) who were weary of the incorrect forms. Analysis of the focus group data demonstrated that older respondents are consciously aware of the rules of their dialect grammar, and hate it when these rules are violated. Younger respondents showed almost no meta-grammatical awareness, and admitted to 'allowing' the incorrect forms in some contexts because they 'sound Brabantish'. Identity construction, in other words, is at the heart of hyperdialectal usage. Methodologically, this paper makes a plea for the confrontation of quantitative data - which provide the backbone of analysis - with qualitative data that offer access to motives for, and constraints on grammatical variation.

This study contributes to a growing body of research on the social meanings of linguistic variation with particular interest in the cognitive processes governing their emergence. Our research follows in the tradition of Labov et al.'s (2011) work on the sociolinguistic monitor, a cognitive mechanism hypothesized to track quantitative linguistic variation and prompt social evaluations (Labov et al. 2011. Properties of the sociolinguistic monitor. Journal of Sociolinguistics 15(4). 431-463). Previous research shows that L1 English listeners are sensitive to frequency variation, but it is unclear whether this also applies to L2 listeners. This study thus replicates Labov et al.'s (2011) original experiment in a context where English is primarily acquired through L2 instruction. To test the generality of sociolinguistic monitoring, we investigate L2 listeners' sensitivity to quantitative differences in sociolinguistic variation (ing) as well as proficiency-based variation. Since participants were L1 speakers of (Austrian) German, we tested evaluations of varying realizations of /θ/ ([θ]/[s]), /d/ ([d]/[t]), and /w/ ([w]/[v]). Experiments included 135 participants, who rated several versions of newscaster test passages regarding professionalism. Our data shows that both sociolinguistic and proficiency-based variation are monitored and evaluated by L2 listeners, albeit to different extents. This supports the assumption that the focus of the monitoring process is socially meaningful variation that includes L1 sociolinguistic but also L2 proficiency-based features.