Postface: Differentiating and Differentiated Views on Linguistic Representation

This special issue of the Canadian Journal of Linguistics/Revue canadienne de linguistique is devoted to the intriguing notion of représentation linguistique, an idea that developed in French (socio-)linguistics some ten years ago (cf. Jodelet 1993), following earlier work by Anne-Marie Houdebine (1982).

In their preface, Philippe Hambye and Anne Catherine Simon discuss one of the major tenets of this approach, namely the idea that représentation linguistique, in addition to being concerned with how language is represented in the human mind—and thus very much in line with the cognitive turn of linguistics—also implies that as a cognitive entity, language has no self-contained status in the sense of post-structualist, generative grammar. Rather, the way in which we language users look upon language is always socially contextualised. Whatever we know, in the largest sense of the word, about language and varieties of language is inextricably linked to the situations and social groups, the speakers and geographical regions, the linguistic genres and social styles that constitute it. Linguistic representations, understood as everyday knowledge of language, are therefore normatively organised. They have to do with how someone of a certain type (incumbent to a certain social category) is supposed to, or can be expected to speak, given certain typified circumstances, and what it means for him or her to speak in this way.

The notion of linguistic representations has a strong resemblance with what linguistic anthropologists in the Chicago tradition have recently discussed under the rubric of linguistic ideologies (starting with Silverstein 1979); it also links up with the sociolinguistics of prestige in variation studies, the social psychology of language attitudes, and ethno-dialectology. In addition, it raises the fundamental question of what is considered to be part of a language at all and what isn't (a line which will be drawn differently in literate and non-literate societies), where one language stops and the other starts, and, certainly, what is a well-formed sentence of a language and what is not. (The latter, obviously, is an ideological issue just like the preceding ones.) The same questions can be asked with respect to varieties contained within a language (is Alsatian a variety of German? Brasilian a variety of Portuguese? Sicilian a variety of Italian?), and with respect to variants of a linguistic form (is nonante a word of French, or only quatre-vingt dix? is Weltanschauung a word of English?). The point is not to find linguistically well-informed answers to these questions of representation; rather, the goal is gain an understanding of how everyday "languagers" go about asking and answering such questions.

An issue that is central to sociolinguistics is the convergence or divergence between the social-cognitive linguistic constructs people carry around in their mind, and their observable linguistic–communicative practices. Often, representations are categorical while practices blurred (and therefore in need of being investigated statistically). This difference is a necessary one because the application of sociolinguistic knowledge to verbal activities is always context-sensitive and context-creating: it has a surplus of meaning which is generated in the situation and does not derive from prior knowledge. "Following the rules", as ethnomethodologists have insisted for decades, is an active, indexical business. In language change, representations are regularly out of phase with observable behaviour; Labov's notion of hypercorrection nicely makes this point already at the level of terminology (Labov 1972).

The papers in this special issue draw upon a variety of theoretical frameworks. The French tradition of représentation linguistique competes with more traditional attitudinal research (Lameli) and with analyses of sociolinguistic style in the tradition of Penelope Eckert (Moore). However, taken together, they offer a good panorama of possible approaches to, and issues that can be raised within, research on linguistic representations. Margret Selting's paper, for instance, deals with the very simple yet fundamental issue of what belongs to a variety, in this case, a regiolect of German. In the domain of prosody, it is particularly difficult to draw a sharp line between one variety and another related one. In addition, lay knowledge about intonational differences between varieties...