1. Combining Different Approaches to Variation
The identification of varieties (dialects)—which requires relating linguistic features to major social categories such as class, age, gender, ethnicity and geographical origin—is one of the main achievements of variationist sociolinguistics. The task is central to the aims of sociolinguistics in two ways. First, it reveals the presence of structured heterogeneity within linguistic variation. Second, it accounts for the social meaning of linguistic variants: through the definition of bounded "social", "regional" or "ethnic" varieties, so-called inter-speaker variation is directly associated with group membership and linguistic features become social identity markers.
Despite its usefulness as an analytic instrument, the notion of linguistic variant nevertheless remains problematic. First, varieties are scientific constructs, and their status in sociolinguistic theory is not clear. In this respect, sociolinguistics faces the same challenges as traditional dialectology, as both approaches establish boundaries within a linguistic reality that is known to be continuously variable (see Peter Auer's postface, this volume). Secondly, the classic variationist method, with its underpinning conception of variation, imposes severe limits on our understanding of the effective social meaning of linguistic variation (Eckert 2000). Variation not only reflects, but also creates, distinctions between social groups. This requires that we investigate the relationship between patterns of variation and how speakers exploit these patterns when they elaborate and perceive effective linguistic practices. In other words, what is needed is a thorough examination of the links between the objective description of variation in terms of linguistic varieties and the speaker's subjective interpretation and experience of this variation.
On the one hand, we need a description of the distinctive linguistic resources speakers draw upon in their social practices, resources which often correspond to [End Page 247] "linguistic dialects" or "linguistic varieties". This in turn raises questions concerning how these concepts are defined , the status we linguists accord them, and the methods we linguists use to describe them.
On the other hand, understanding the extent to which language plays a role in the constitution of sub-entities in a community, and how it reinforces social differentiation, requires an analysis of speakers' representations and attitudes towards both linguistic features and the groups who adopt those features.
This special issue of the Canadian Journal of Linguistics/Revue canadienne de linguistique aims to address these questions. While the contributions in this volume represent diverse theoretical backgrounds, the common thread that unites them is a concern with how linguistic variation is embedded in social and discursive processes. They examine how speakers use linguistic variation to achieve non-linguistic goals, including building social representations (of themselves and others), expressing conversational involvement, creating boundaries between social groups (via stylisation processes), or specifying their relation with interlocutors.
The articles in this volume approach questions of linguistic variety and style from different perspectives, ranging from the more objective (factual, linguistic) to the more subjective (oriented towards speakers' representations). The methods used by the authors, and the kinds of explanation they invoke, are sometimes quite distant from each other. Some of the contributors propose objective criteria to assess linguistic variation, including the measurement of the phonetic distance between regional varieties (Lameli), or the description of the form and functions of a set of regional intonation contours (Selting). Other contributors seek to understand how speakers create subjective boundaries between so-called "ways of speaking", despite the fact that there may be no discernible linguistic evidence for such differentiation (Boudreau and White, as well as Roy).
2. The Circumscription of Regional Varieties
An attempt to circumscribe regional varieties, both from an objective and subjective point of view, is made by Alfred Lameli ("Phonetic measurement and metalinguistic judgment"). The author selects speech fragments produced in authentic speech-events (city council meetings) in two German cities, Neumünster and Mainz, belonging to different dialect areas (Low and Middle German, respectively). These short fragments are subjected to two different analyses: (i) a phonetic measurement of the distance between the regionalized speech and standard German; (ii) an assessment based on native listeners' perceptions of whether the speaker is speaking "standard" or "dialect". The study shows a clear correlation between two variables, namely "phonetic measurement" and "metalinguistic judgment...