‘Mediaeval Europe, in contrast with today’s world, was not a place of official vernacular languages. By and large written communication was in Latin. The grammar, which was taught as a basic subject in the curriculum of the trivium, was Latin grammar (Murphy, 1974). The vernacular tongues, even when used in written form, were not considered grammatical, nor was the spelling of their vocabularies standardised. In this context, there were no right and wrong ways to write the vernacular; and in most cases it simply was not written. The pressures to standardise spelling, to establish correct grammars and to teach an approved form of the native tongue were to come much later. Michel Foucault (1972) has compared the emergence of grammar as an academic discipline in the eighteenth century with the development of medicine and economics at the same time. In each case, the academic study was developing in the context of the emerging modern state, which was imposing uniformity and order on its citizenry and which, according to Foucault, was “a disciplinary society” (1986, p. 206).
In the Middle Ages, according to Douglas Johnson, “it was undoubtedly difficult for the ordinary person in one part of France to be understood in another part of France” (1993, p. 41). Indeed, the situation persisted well into the nineteenth century in France (Braudel, 1988). One can imagine mediaeval peasants’ relation to their patterns of speaking. They would share ways of talking with fellow members of their village. They would recognise these patterns – and perhaps distinctive words – especially when encountering fellow villagers away from home. The documents of Montaillou recount one villager, Arnaud Sicre, a shoemaker working in San Mateo, overhearing a woman entering the workshop and speaking “the tongue of Montaillou” (Ladurie, 1978, p. 286). He put down his tools to ask whether she did indeed come from Montaillou. The ‘tongue’ may have been distinctive, yet it was also comprehensible to those living in neighbouring regions, which would also have their own recognisable ways of speaking. Some words would be unfamiliar to outsiders, whilst others would not be. As one travelled further from one’s home village, the ratio of unfamiliar phrases to familiar ones would rise, with problems of communication increasing. If one travelled to a particularly inaccessible village, one might find few common phrases. In the case of fourteenth century Montaillou, Ladurie writes that there was a continuum of communication between Occitania and Catalonia.
In travelling between villages and along the continuum of communication, there would be no point at which the peasant would imagine that they had passed through a linguistic boundary, separating one distinct tongue from another. Moments of intelligibility might get fewer, dribbling away entirely in distant horizons. The travelling peasant, however, would not stop to ask ‘do these people speak the same language as myself?’, as if there was an actual point at which the ratio between the familiar and the unfamiliar became critical and the speech pattern changed from one grammatical essence to another. This essentialism, by contrast, is insinuated into the core of modern common sense about language. We would want to know whether the speech of Montaillou should be categorised as a dialect of Occitan and whether the inhabitants of San Mateo really spoke a variant of Catalan. We assume the reality of underlying different deep grammars. If the modern political map, unlike its mediaeval equivalent, contains precise boundaries, so too does the modernly imagined map of speech. The assumptions of this imagined mapping are easily projected on to other cultures and other times. Clifford (1992) recounts how anthropologists typically assume that each village, or each tribe, which they study, has its own unique language.
The modern imagining of different languages is not a fantasy, but it reflects that the world of nations is also a world of formally constituted languages. The disciplinary society of the nation-state needs the discipline of a common grammar. The mediaeval peasant had no official forms to complete, inquiring whether the respondent speaks Spanish or English. No acts of parliament decreed which language was to be used in compulsory public education or in state broadcasting; nor would the mediaeval subject have dreamt of ever going to war over such matters. The questions about language, which today seem so ‘natural’ and so vital, did not arise. To put the matter crudely: the mediaeval peasant spoke, but the modern person cannot merely speak; we have to speak something – a language.
Languages and Boundaries
A world of different languages requires the constitution of categorical distinctions. A problem confronts anyone who attempts to make distinctions between one language and another. Not all the speakers of a language speak in the same way. Thus, some differences of speaking have to be classified as being instances of different languages and some will be classified as differences within the same language. The notion of ‘dialect’ becomes crucial to maintain the idea of separate languages: it seems to account for the fact that not all speakers of a language speak the same way. The word ‘dialect’ did not gain its linguistic meaning until the early modern period (Haugen, 1966a). Previously, the linguistic problems, which the Word addresses and seems to solve, did not arise. The inhabitants of fourteenth-century Montaillou did not worry whether their tongue was a ‘dialect’ of a wider language, or whether it was a separate language: the shoemaker was interested in knowing whether he shared a birthplace with the woman, not whether they spoke ‘the same language’.
The idea of a dialect had little use before nation-states started establishing official ways of speaking and writing. Differences between languages and dialect, then, became hotly contested political issues, as well as concerns for the discipline of linguistics. If it seems obvious to us that there are different languages, it is by no means obvious how the distinctions between languages are to be made. Suppose one stipulated that speakers of the same language understand each other; and that speakers of different languages do not. This would imply that all the variants (or dialects) of a single language are mutually intelligible, and that different languages are mutually unintelligible. Linguists have emphasised that there is no simple criterion for determining mutual intelligibility. How much comprehension should count as intelligibility? Where on the continuum of comprehensibility is the boundary between understanding and non-understanding to be drawn?
Even if such a criterion could be applied, it would lead to very different distinctions from those which are conventionally accepted and which seem so solid to speakers and non-speakers alike (Comrie, 1990; Ruhlen, 1987). There are instances of ‘different’ languages, such as Danish, Norwegian and Swedish, which are mutually intelligible. As Eriksen (1993) points out, the spoken language of Norwegian cities like Bergen and Oslo is closer to standard Danish than it is to some of the rural dialects of Norwegian. As well as the problem of different languages which are mutually comprehensible, there is the problem of languages which encompass mutually incomprehensible dialects. Thus, speakers of both Gheg and Tosk dialects imagine themselves to speak the common language Albanian, although the dialects are mutually incomprehensible (Ruhlen, 1987).
More is at stake in drawing the boundary of a language than linguistics. The battle for hegemony, which accompanies the creation of states, is reflected in the power to define language, or in what Thompson has called the power “to make meaning stick” (1984. p. 132). This power resides not merely in the imposition of certain words or phrases, but also in the claim of languages to be languages. The middle class of the metropolitan areas typically will make their meanings stick as the official language, relegating other patterns within the national boundaries to ‘dialects’, a term which almost invariably carries a pejorative meaning. As Haugen (1966a) suggested, a ‘dialect’ is frequently a language which did not succeed politically: for example. Piedmontese was relegated to the status of dialect after Tuscan succeeded in becoming the language of Italy.
Nationalists, in attempting to create a separate nation, often will create a language as a distinct language, although they might claim to be creating the nation on the basis of the language. as if the latter were an ancient, ‘natural’ fact. When Herder was praising the German language as the soul of the German nation, he was arguing to bring both – the language and the nation – into existence, whilst treating both as if age-old. The speech of the territory that was to become Germany comprised several mutually unintelligible ways of talking, none of which had succeeded yet in establishing its status as the ‘correct’ form of an overall German language. At that time. Prussians spoke Low German and “learnt High German as a second language” (Hawkins, 1990, p. 105). In the following century, with the rise of Prussia, ‘standard’ German was to emerge as the north German pronunciation of southern High German.
Again and again, the boundaries between languages, and the classification of dialects, have followed the politics of state-making. Where national boundaries are established, then, the differences in speech patterns either side of the boundary are more likely to be seen as belonging to distinctly different languages by the speakers themselves, their national centres and the world in general. When the Dutch went their way politically, their form of lower Franconian was to become a separate language, in contrast to other forms which have become known as dialects of German (Schmidt, 1993). Galician. spoken in Spain, and Portuguese, spoken across the border, are now generally thought to be distinct languages. In linguistic terms, French and Italian merge into each other, but the speech patterns on the French side of the border are likely to be seen as dialects of French, and those on the Italian side as dialects of Italian (Ruhlen, 1987). Similarly, Friul in Northern Italy is similar to Romanesch in Switzerland, but, again, national boundaries reinforce a sense of linguistic separateness (White, 1991). The creation of Norwegian is instructive. The decolonisation from Denmark was marked by a struggle for language. First, the state of Norway was to declare its own language, creating a spelling to match so-called Norwegian patterns of talking, rather than Danish ones. Then, there was the internal battle between two rival patterns of speech, the Riksmal and Landsmal, both having claims to be considered as the proper Norwegian (Haugen, 1966b). In all these cases, professional linguists have tended to fall into line with accepted practices, accepting Norwegian and Danish as different languages, High and Low German as variants of the same language etcetera (Comrie, 1990). As Ruhlen (1987) admits, because there are no purely linguistic criteria for classifying languages, linguists follow common beliefs about identifying similar and different languages.
The common practices of naming languages tend to emerge through struggles for hegemony. And what is made into a common practice can, under certain circumstances, be unmade or become a locus of struggle. For example, Italian law makes a distinction between koine (dialects) of Italian and full-blown minority languages. Friulan and Sardinian activists campaigned for years to have their speech recognised in law as official languages. Successive central governments, fearful of separatism and the cost of grants for minority languages, resisted their demands. In the debate whether Friulan and Sard were dialects or languages, both sides have their expert linguists, contesting the other side’s characterisation of what constitutes a language and what is merely a dialect (Petrosino, 1992). More dramatically, the Turkish government officially denies that its Kurdish citizens are Kurds and that there is a Kurdish language: the Kurds really are “mountain Turks”, who have forgotten their native, Turkish tongue (Entessar, 1989).
One might suppose that nationalist movements, seeking to form separate states, will seek to convert dialect into language. The power of writing down a way of speaking should not be underestimated: it provides material evidence for the claim that a separate language exists. In order to highlight differences from the ‘official’ governing language, separate spellings might be adopted and these might highlight the area’s distinctive way of talking. These spellings, written on public notices and used in mythic poetry, will proclaim the uniqueness of the speech and its status as a language. Sometimes different orthographies can divide mutually intelligible ways of speaking, as in the case of Serbian and Croat, and also Urdu and Hindu. The status of the writing, however, can be contested or officially branded as dialect. In 1994, for the first time since the 1872 Scottish Education Act banned the use of lowland Scottish (or Lalland) in schools, Glasgow University accepted a dissertation written in Lalland: topic of the ‘dissertation’ was ‘Scots spellin’. Significantly, the university senate only agreed to accept the thesis on the understanding that its writing be classified as a dialect of English, not as a separate language (Guardian, 8 July 1994).’
‘Conflicts over language are commonplace in the contemporary world. They are comprehensible to ‘our’ common sense: reports about French and Flemish speakers in Belgium, or Urdu and Hindu speakers in India, do not occasion puzzlement. Such conflicts are not just struggles about language, but importantly they are conducted through language (as well as through violence). In this respect, the universal, or international, aspects of nationalism are crucial. Without common notions, which can be translated across particular languages and dialects, the conflicts would not be pursued in their nationalist forms. Foremost amongst such notions are the ideas of ‘language’ and ‘dialect’ themselves. These terms must be reproduced in every language which is used by its speakers to claim that they possess a separate language, and that, in consequence, they are a separate nation, Whose internal differences of speech are merely differences of ‘dialect’.
Notions of language and dialect are not the exclusive property of ‘extremists’, who pursue narrow national dreams. They are part of ‘our’ common sense. This has methodological and political implications. Nations may be ‘imagined communities’, but the pattern of the imaginings cannot be explained in terms of differences of language, for languages themselves have to be imagined as distinct entities. If nationalism is to be studied as a widespread ideology, and if nationalist assumptions are to be found in common-sense notions about what a language is, then nationalism should not be projected on to others, as if ‘we’ are free from all its effects. In addition, the assumptions, beliefs and shared representations, which depict the world of nations as our natural world, are historical creations: they are not the ‘natural’ common sense of all humans.
At other times people did not hold the notions of language and dialect, let alone those of territory and sovereignty, which are so commonplace today and which seem so materially real to ‘us’. So strongly are such notions embedded in contemporary common sense that it is easy to forget that they are invented permanencies. The mediaeval cobblers in the workshops of Montaillou or San Mateo might, with the distance of 700 years, now appear to us narrow, superstition-bound figures. But they would have found our ideas on language and nation strangely mystical; they would be puzzled why this mysticism could be a matter of life and death.’
Banal Nationalism, Michael Billig, 1985.