The Analysis of Medium-Sized Language Communities

F. Xavier Vila and Vanessa Bretxa

Looking at Languages in Between

It may come as a surprise to many that sociolinguistics, understood in a broad sense as the discipline that studies the relation between language and society, has so far been unable to agree on a basic typology of linguistic communities. More than half a century ago, Ferguson's Diglossia (1959) established the outlines of what was expected to be the first step towards a more general classification (Ferguson, 1991). However, in spite of the paper's huge success, and also the calls made by other authors in this direction (e.g. Einar Haugen's (1971: 25) advocacy of a 'typology of ecological classification', as well as Gumperz, 1962, or Kloss, 1966), the truth is that the efforts to establish a clear, systematic and comprehensive sociolinguistic classification of languages and linguistic communities around the world have been relatively unsuccessful. In other words, sociolinguistics has still not produced a typology that classifies language communities and/or their linguistic ecologies according to a widely accepted set of features. Also, to put it in in Peter Mühlhäusler's terms, 'To understand why so many individual languages are disappearing requires an understanding of the ecological conditions that sustain complex language ecologies' (Martí et al., 2005: 45).

One of the reasons for this failure may lie in the complex relation of sociolinguistics with the language construct, and with some of the main concepts associated with it. The main focus of research of sociolinguistics as a discipline is linguistic diversity (Coulmas, 2005), and most introductions to the field make clear from the very beginning that the notion of language itself is polysemic, ambiguous, difficult to define and even 'a fallacy' (Simpson, 2001: 31). It is indeed commonplace that every introductorycourse to sociolinguistics reminds the novice that language borders are often impossible to delineate in purely linguistic terms; that languages show very disparate degrees of internal structural difference; that mutual intelligibility is not a safe indicator of the 'language' versus 'dialect' divide; and that, in actual terms, the distinction between a language and a dialect is a contingent sociohistorical compromise rather than an immutable structural fact, to the extent that some scholars propose rejecting the notion of language altogether (Blommaert, 2010).

In such a context, it is not surprising that many of the existing typologies of linguistic situations have not adopted languages or language communities as their analytical frame, but rather polities, and especially sovereign states (cf. Bastardas & Boix, 1994; Laitin, 2000; Spolsky, 2004). These typologies attempt to classify polities according to the number of languages spoken in each country and the official status and function of each, and thus have a strong legal, politological approach to the analysis of sociolinguistic situations. Indeed, one does not have to subscribe whole-heartedly to the oft-quoted saying that 'a language is a dialect with an army and a navy' to consider that (sovereign) states constitute one of the main factors to be taken into account when analysing any given sociolinguistic ecosystem. In fact, states play such a central, decisive role in language policy in contemporary times that any classification that ignores their existence and impact is doomed to failure. Besides this, empirical quantitative analyses are often impossible across state borders, for the basic statistical data crucially needed for sociolinguistic analysis are usually provided by public administrations, and therefore depend strongly on existing political and administrative borders. Consider Europe, for instance. Comparative analysis of the sociolinguistic reality of the European languages has become much easier since the Union – a sui generis political entity, but a polity at the end of the day – decided to take on the task of obtaining comparable data in all of its Member States by means of the Eurobarometers. Before then, sociolinguistic comparisons across countries had to deal with the arduous task of putting side by side the results from disparate data-collecting methods based on vastly different premises (cf. Extra & Gorter, 2008). It is no coincidence that the root of the term 'statistics' is 'state'.

Important as they undoubtedly are, politologically oriented classifications of languages and language groups and situations do not in themselves exhaust the possibilities of classifying languages and linguistic situations. On the one hand, they do not necessarily capture some crucial aspects of a particular language community such as the degree of intergenerational transmission, language use in socioeconomic spheres, cultural production and consumption, and the ideological positioning of its speakers vis-à-vis other languages. On the other hand, almost by definition, research structured on the basis of political borders finds it difficult to deal with phenomena that go beyond them, and languages do go beyond borders at least in three different senses. First, the borders of the almost 200 independent states in the world do not coincide with those of the 5000 to 6000 languages (still) spoken. Second, even if nation states have striven to make their citizens linguistically homogeneous, people move around and take their linguistic repertoires with them. Finally, people communicate more and more across borders. A glance at Fischer's (2011) map of the world language communities of Twitter should suffice to convince the reluctant that next to the politologically oriented classifications, we need more refined sociolinguistically oriented comparative analyses, that is, analyses that pay attention not only to sovereign states, but also to people(s) and communities.

In fact, there are a number of classifications that are more community oriented, and therefore link themselves, in a more or less ambiguous way, with the historical meaning of language community mentioned above. Most of these classifications (although not all – think of the concept of ethnolinguistic vitality; Ehala, 2010; Giles et al., 1977; Harwood et al., 1994) tend...