Contenu du sommaire : Hétérogénéité et variation : Labov, un bilan, sous la direction de Françoise Gadet

This paper evaluates the theoretical importance of Labov's conception of language : 1) his historical sources around the concepts of variation, variable and variant, including the relationships with dialectology, the Prague School and Martinet ; 2) the way in which his novel concepts of structured heterogeneity, inherent variability and vernacular speech build a specific theoretical construct which can account for variation and change.

This paper deals with the sociolinguistic problem of the spread of changes accross the speech community, with questions like : what are the relationships between internal and social factors within this process ?, how and to what extent, are social forces able to determine the evolution of the structure ? why are women in advance of men for phonetic changes ? how much can social networks modify a language ? how does transmission operate ?

This paper explores Labov's work on phonology and linguistic change during the last ten years. The concepts of inherent variation and heterogeneity are shown to provide an adequate methodological basis for the study of phonetic change. The neo-grammarian controversy is analysed along with some phonological synchronie processes. With an enlargement of the empirical basis, the variationist analysis gives linguists internal and external decision-making criteria with which to evaluate linguistic theories.

The variationist approach to syntax has several inadequacies that arise from both the formalisation and the definition of the variable. In particular, the notion of variation does not capture all of structural heterogeneity. The legitimacy of the concern of integrating heterogeneity into the linguistic system is underlined by a discussion of some recent evolution in syntactic work.

Variation, conceived as multiple realisations of the same linguistic unit, is not a relevant notion beyond phonology or morphology. Thus, in syntax : 1) every syntactic structure is a specific set of relations bearing a specific meaning ; 2) different speakers of the same language make different choices among the possible alternatives ; 3) they do not have the same attitude toward the normative mode of linguistic usage. These points are illustrated with examples taken from spoken French relatives and quantifiers.

Labov has studied Black English within the context of his question, « What makes a linguistic community ? » He establishes two important points : some constructions and some distributions in Black English are related to Creole grammar ; Black and White English are two distinct language communities and they are becoming more and more divergent. The Black linguistic community is evolving its own way.

Why and for whom do linguists build linguistic theories ? One possibility is for the pleasure of theory, another for their relevance to social concerns. Labov showed how variationism could help him working as a linguistic expert in a courtroom. But variationism remains unable to help when the point concerns the production of meaning in interactions. In this paper, alternative sociolinguistic theories are examined for their relevance to social issues.

Labov's sociolinguistic approach to the study of everyday speech implies an implicit sociology that can help sociologists understand social membership. Labov presents a way of looking at social behavior by providing a physical index to social membership using the role of phonic cues in social membership. His two notions — « multi-style speaker » and « non-member » — can be used in the development of a micro-sociology of social interaction.

This paper deals with the historical genesis of the difficult relationships between linguistics and the concept of variation in France from the end of the XIXth century to the triumph of structuralism. The paper shows how scientific discourse took language differences into account, the social basis of its analysis, including the academic actors, their development and careers, and the political representations of these language differences.