First of all we want to thank the general editors of the Nordic Journal of Linguistics for this opportunity to present to the readers of one of the most important linguistics journals in Northern Europe some specimens of sociolinguistic research. When we were approached by the general editors, we wholeheartedly agreed that this would be a good idea, although – or rather, precisely for this reason – sociolinguists in general have not been keen on publishing in NJL in the past. We have a modest hope that the present issue will change this.
NJL is the brainchild of the Nordic Association of Linguists and it is rather obvious that, from the start, this association included more grammarians and theoretical linguists than sociolinguists. Be that as it may, this editorial introduction could be the place to broach some of the issues separating the two traditional readerships from each other in order to scrutinize the reasons for the division and ask whether they are still valid.
The split between structuralist linguistics in all its shades and colours, on the one hand, and, on the other hand, sociolinguistics – for this purpose including all branches of non-autonomous linguistics, i.e. discourse and conversation analysis, ethnomethodologically inspired linguistic analyses and quantitative and qualitative sociolinguistics (proper) – goes back to discussions about the autonomy of linguistics. The autonomy of linguistics is a stance which comes in two guises. One is an ontological claim that language is an organ which is biologically and neurologically as separate as these things may come and that therefore the phenomenon of language should be approached and studied as a natural kind. The other, a rather more common stance, is that it is strategically wise to try to unravel the secrets of the phenomenon of language keeping the research questions as free of connections to other realms of knowledge as possible. Arguments against the first position, often attributed to Noam Chomsky, may be found primarily in the writings of Michael Tomasello and the long tradition of socio-cultural developmental psychology which he belongs to (Lev S. Vygotsky, Michael Cole, and others).
The arguments against the second position, that linguistics ought to be kept, for strategic reasons, as connection-free as possible, again come in two guises. The first is that it is impossible to study such central features of language as variation and change if we do not adopt an integrated view of language as immersed in, indeed central to, societal factors. If linguistics is seen as a social science or even a science holding one of the important keys to understanding the inter-relations between humans, nature and society, giving up the autonomy of linguistics seems to be the inescapable conclusion. This is not to say that further arguments of what this means in terms of which questions may be asked and answered by a non-autonomous linguistics are settled in advance.
The second version of resistance to strategic autonomy is more ecumenical in nature. This position gives up any vision of linguistics as one discipline and accepts a peaceful co-existence of all paradigms concerned with language as in principle valid answers to in principle fundamentally different questions. We suspect that any head of department would feel inclined to accept this stance as the key to avoiding any reversion to the linguistic wars of the past. While we appreciate this motivation, we would like to point out that there is a sort of complacency and evasion of debates across paradigms inherent in the ecumenical approach that may resemble intellectual laziness. Thus, researchers actively identifying themselves as sociolinguists should never feel securely insulated from the criticism of those who address each other as theoretical linguists. If that were the case our disciplines would grow completely apart to the detriment of both strands.
What, then, are the issues that could constitute a common meeting ground for both strands? First of all, a feature obvious to any empirical linguistics: variation. In early European structuralism much effort was spent on getting through the variable outskirts in order to reach the invariable centre of language structure. This strategy has led to a number of important and permanent insights but it has also drawn attention away from the fact that variation is a universal feature of language use and as such merits explanation precisely because it is a central part of the mystery: How do language users end up with grammars of the structural properties of language through a subconscious analysis of actual use when this use always manifests widespread and significant variation? A number of theoreticians of so-called usage-base grammar have speculated that the early structuralist focus on structure as systematically discarding variation should be replaced by grammars which integrate variation in their basic structure. We are not in a position to take a stance here but we do know that the central issue of variation in recent years has led to a number of contacts across the great divide between theoretical linguistics and sociolinguistics, supported, among other things, by the initiative Nordic Language Variation Network (http://nlvn.uit.no/), funded by Nordforsk. In the future we look to NLVN's recent offspring, the N'CLAV initiative (Nordic Collaboration on Language Variation Studies), chaired by Maia Andreasson of the University of Gothenburg, for pursuing this dialogue.
The second issue which must lead to a dialogue between paradigms is the ancient, indeed traditional, problem of language change. For sociolinguists, variation has always been seen as the key to change. A structuralism which removes variation in order to reach more or less historically stable structures will need to answer the question of how and why such structures change at all. This is the central argument in the early and seminal paper on the empirical foundations of a theory of language change by Weinreich, Labov & Herzog (Reference Weinreich, Labov, Herzog, Lehmann and Malkiel1968), which is still an intellectual pleasure to read. The pleasure is enhanced by the knowledge that this paper has served as a guideline for a lot of sociolinguistic research studying the various problems outlined: The Constraints problem, the Transition problem, the Embedding problem, the Evaluation problem and the Actuation problem (op. cit.: 183ff.).
The two issues of variation and change cannot be ignored by any empirical study of language and this is good news for a dialogue across paradigmatic borders which may enhance all participants' reflexive knowledge of what Ferdinand de Saussure was after: what the linguist does (when he practices his trade as a linguist, that is).
The four papers assembled in this issue of NJL have several things in common beyond coming from sociolinguists. First of all they are written by a new generation of scholars who have just finished their Ph.D. theses (Randi Solheim) or did so some years ago (Jenny Nilsson), or are about to defend them (Janus Spindler Møller, Toril Opsahl). It is a good sign for the future of sociolinguistics that we are also able to introduce Signe Schøning who has not yet finished her Danish MA (but has already graduated from Lancaster University as an MA).
Secondly, the authors share the conviction that their analyses must be based on looking at either interaction between informants and interviewers or interaction among informants themselves. All of them prefer to base their insights on qualitative analyses of interaction so that Toril Opsahl's appeal to the idea of (some version of) interactional sociolinguistics may be seen as emblematic for the whole issue.
Thirdly, the progression in the issue highlights an important tradition of working with dialect levelling and methodological themes. We wish to discuss this here since it cannot be taken for granted, with respect to readers outside the Nordic countries (indeed for any reader unacquainted with more than one of them), that dialect levelling is an issue worth spending that much effort on.
Sociolinguistics in all the Nordic countries was born as a more or less distant cousin to dialectology. In traditional dialectology the Neogrammarian tradition was kept alive, if not actually kicking, by a constant referral to the posited forms of the reconstructed early stages of the language studied as the undisputed reference point for geographical patterns of isoglosses. Common to all dialectological research, whether traditional or modern, is that it is empirical and thus has to grapple with fundamental problems of ‘getting from use to structure’: what we get from the informants is their linguistic reactions to our input, i.e. language use. And what we want to disclose is the structure of the dialect or the character of this particular mixture of features. In this respect the dialectologist is placed at the heart of the linguistic endeavor as such: he or (more often) she tries to disentangle the important differences from the less important ones. Instruments used for this purpose are ‘loans’, ‘contact phenomena’ and very importantly ‘corruption’ of the dialect, that is, levelling, whether we know which standard is levelled to or not.
Obviously, the levelling process was a theme right from the start for any ‘one foot in the grave’ dialectology since the selection of informants crucially depended on finding the right speaker to represent the language. Again this is no different in principle from any field worker who tries to get at (the structure of) any spoken language.
Randi Solheim, in her paper on the ‘new industrial city’ of Høyanger in an old dialect speaking community in Western Norway, outlines the stages of a process of dialect contact involving both incoming local speakers and speakers migrating from the east of Norway who have a completely different dialect background. Solheim convincingly describes how the various stages involve differentiation and focusing so that a new norm is created. Moreover, she shows that this norm has recently become subject to the dialect levelling process common to large parts of Western Norway. One importance of this paper lies in its contribution to the new debate on the role of identity, following Peter Trudgill's stimulating and controversial statements of what Lars Brink and Jørn Lund in their early theories of language change call the Napoleonic principle: numbers of speakers are the sole necessary and sufficient ingredient in an explanation of what the outcome of a dialect contact situation is; identity formation does not count (Trudgill Reference Trudgill2008). Trudgill's example is colonial new-dialect formation and he states his point forcefully: ‘I see no role for identity factors in colonial new-dialect formation’ (Trudgill Reference Trudgill2008:243). Solheim, in contrast, does and argues her point in the final section.
Jenny Nilsson directly tackles the question of what linguistic reasons we might have to posit a change of dialect levelling in the west of Sweden. Nilsson distinguishes two questions: Has the dialect changed? Has the originally dialect speaking speech community changed? The first question may be answered by looking at the status of the 24 linguistic features that Nilsson has separated as useful for an indication of the distinctive character of the variety. How many of them are still in use and how are they used? The second question must be answered by looking at the community of speakers: How many of the speakers have productive access to these 24 dialect features and to what extent do they actually use the standard variants instead? How many are only able to decode the features as dialect features, but do not themselves productively use them? The paper is thus concerned with an important methodological caveat that we only compare likes when we want to establish change.
In her paper on multiethnolect in Oslo, Toril Opsahl may be presenting a new but central problem for sociolinguists in Europe and elsewhere. When we look at the favourite work place of sociolinguistics, i.e. the metropolis, one obvious distinguishing feature is its multilingualism as a speech community. Whether we are in Toronto, Canada, where numerous diverse immigrant communities have kept their ‘heritage languages,’ or in Oslo, Norway, the fact is that there are many languages other than Norwegian around. In Høyanger the various dialects of Norwegian met and the result was a koineization process. By contrast, in Oslo, as in Stockholm and Copenhagen, immigrants' first languages meet the local majority Nordic language. The outcome of this meeting may be termed a multiethnolect if and only if informants with various linguistic backgrounds including having the majority language as their first language adopt this variety in particular circumstances. Opsahl studies primarily one feature of this multiethnolect in Oslo, namely the use of the discourse marker wolla. Through an analysis of a number of instances from two corpora, the Upus corpus of youngsters and the comparable youngsters from the NoTa corpus at the University of Oslo, we learn about the various functions of this one word and some of its equivalents.
Finally, Signe Schøning & Janus Spindler Møller take us on a trip through a number of self-recordings. For many years the original ‘observer's paradox’ of William Labov, i.e. that we want to know what the informants say when we are not there to record it, was central to any methodological debate within sociolinguistics. Schøning & Spindler Møller show that when they furnish informants with a recorder, the informants' first task is to explain to the co-participants what this is about. Often informants will play down their own role as being just on a mission for someone else – i.e. the researchers – but sometimes the recorder is seen as an enviable object of prestige. At other times it is seen as a severe obstacle to talking freely, thus actually corroborating that there are limits to what informants want recorded. The gist of the paper is that self-recordings are valid in their own right, and while that point is well taken we still wait for the grand comparison between data obtained in one and the other way, that is through traditional sociolinguistic (or for that matter ethnographically-inspired) interviews and self-recordings. If there is no difference, this is good news for traditional sociolinguistic methods but if there is – which is more likely – we may need to develop traditional data collection methods by including self-recordings as indeed a number of young sociolinguists in the Nordic countries have already done.
We hope that this issue will whet the appetite of the NJL audience for more.
