Calabrese, Relevant southern features here are NC > NN (monno vs mondo < MUNDUM ‘world’, piommo vs piombo < PLUMBUM ‘lead’); characteristic patterns of both tonic and atonic vowel development; use of postposed possessives (figliomo vs mio figlio ‘my son’); extensive use of the preterit; etc. A number of features mark off Tuscan from its neighbours: absence of metaphony (umlaut); -VriV-> -ViV-(IANUARIUM > gennaio, cf. Gennaro, patron saint of Naples); fricativisation of intervocalic voiceless stops – the socalled gorgia toscana ‘Tuscan throat’ – which yields pronunciations such as [la harta] la carta ‘the paper’, [kauo] capo ‘head’, [lo hiro] lo tiro ‘I pull it’; etc. Such divisions reflect both geographical and administrative boundaries. The La Spezia-Rimini line corresponds very closely both to the Apennine mountains and to the southern limit of the Archbishopric of Milan. The line between central and southern dialects approximates to the boundary between the Lombard Kingdom of Italy and the Norman Kingdom of Sicily, and to a point where the Apennines broaden out to form a kind of mountain barrier between the two parts of the peninsula. The earliest texts are similarly regional in nature. The first in which undisputed vernacular material occurs is the Placito Capuano of 960, a Latin document reporting the legal proceedings relating to the ownership of a piece of land, in the middle of which an oath sworn by the witnesses is recorded verbatim: sao ko kelle terre, per kelle fini que ki contene, trenta anni le possette parte Sancti Benedicti ‘I know that those lands, within those boundaries which are here stated, thirty years the party of Saint Benedict owned them.’ The textual evidence gradually increases, and by the thirteenth century it is clear that there are wellrooted literary traditions in a number of centres up and down the land. These are touched on briefly by the Florentine Dante (1265-1321) in a celebrated section of this treatise De Vulgari Eloquentia, but it is the poetic supremacy of his Divine Comedy, rapidly followed in the same city by the achievements of Petrarch (1304-74) and Boccaccio (1313-75), which ensured that literary, and thus linguistic, pre-eminence should go to Tuscan. There ensued a centuries-long debate about the language of literature – la questionedella lingua ‘the language question’, with Tuscan being kept in the forefront as a result of the theoretical writings of the influential Venetian (!) Pietro Bembo (1470-1547), especially his Prose della volgar lingua (1525). His ideas were adopted by the members of the Accademia della Crusca, founded in Florence in 1582-3, which produced its first dictionary in 1612 and which still survives as a centre for research into the Italian language. Meanwhile, although the affairs of day-to-day existence were largely conducted in dialect, the sociopolitical dimension of the question increased in importance in the eighteenth and nineteenth centuries, assuming a particular urgency after unification in 1861. The new government appointed the author Alessandro Manzoni (1785-1873) – himself born in Milan but yet another enthusiastic non-native advocate of Florentine usage – to head a commission, which in due course recommended Florentine as the linguistic standard to be adopted in the new national school system. This suggestion was not without its critics, notably the great Italian comparative philologist, Graziadio Ascoli (1829-1907), and a number of the specific recommendations were hopelessly impractical, but in any case the core of literary usage was so thoroughly Tuscan that the language taught in schools was bound to be similar. Education was, of course, crucial since the history of standardisation is essentially the history of increased literacy. On the most conservative estimate only 2.5 per cent of the population would have been literate in any meaningful sense of the word in 1861, although a more recent and moreper had 91.5 per cent by 1961, the centenary of unification and the thousandth anniversary of the first text. Even so, there is no guarantee that those who can use Italian do so as their normal daily means of communication, and it was only in 1982 that opinion polls recorded a figure of more than 50 per cent of those interviewed claiming that their first language was the standard rather than a dialect. Yet the opposition language/dialect greatly oversimplifies matters. For most speakers it is a question of ranging themselves at some point of a continuum from standard Italian through regional Italian and regional dialect to the local dialect, as circumstances and other participants seem to warrant. Note too that the term dialect means something rather different when used of the more or less homogeneous means of spoken communication in an isolated rural community and when used to refer to something such as Milanese or Venetian, both of which have fully fledged literary and administrative traditions of their own, and hence a good deal of internal social stratification. Another significant factor in promoting a national language was conscription, firstbecause it brought together people from different regions, and second because the army is statutorily required to provide education equivalent to three years of primary school to anyone who enters the service illiterate. Indeed, it is out of the analysis of letters written by soldiers in the First World War that some scholars have been led to recognise italiano popolare ‘popular Italian’ as a kind of national substandard, a language which is neither the literary norm nor yet a dialect tied to a particular town or region. Among the features which characterise it are: the extension of gli ‘to him’ to replace le ‘to her’ and loro ‘to them’, and, relatedly, of suo ‘his/her’ to include ‘their’; a reduction in the use of the subjunctive in complement clauses, where it is replaced by the indicative, and in conditional apodoses, where the imperfect subjunctive is replaced by the conditional, and the pluperfect subjunctive is replaced by either the conditional perfect or the imperfect indicative (thus standard se fosse venuto, mi avrebbe aiutato (‘if he had come he would have helped me’) becomes either se sarebbe venuto, mi avrebbe aiutato or se veniva, mi aiutava, the latter having an imperfect indicative in the protasis too; the use of che ‘that’ as a general marker of subordination; plural instead of singular verbs after nouns like la gente ‘people’. Some of these uses – e.g. gli for loro, the reduction in the use of the subjunctive and the use of the imperfect in irrealis conditionals – have also begun to penetrate upwards into educated colloquial usage, and it is likely that the media, another powerful force for linguistic unification, will spread other emergent patterns in due course. Industrialisation, too, has had its effect in redrawing the linguistic boundaries, both social and geographical. In addition to the standard language, the dialects and the claimed existence of ita-liano popolare, there are no less than eleven other languages spoken within the peninsula and having, according to one recent but probably rather high estimate, a total of nearly 2.75 million speakers. Of these, more than two million represent speakers of other Romance languages: Catalan, French, Friulian, Ladin, Occitan and Sardinian. The remaining languages are: Albanian, German, Greek, Serbo-Croat and Slovene. Amidst this heterogeneity, the Italian national and regional constitutions recognise the rights of four linguistic minorities: French speakers in the autonomous region of the Valle d’Aosta (approx. 75,000), German speakers in the province of Bolzano (approx. 225,000), Slovene speakers in the provinces of Trieste and Gorizia (approx. 100,000), Ladin speakers in the province of Bolzano (approx. 30,000). Yet French (and Occitan – approx. 200,000) and German speakers outside the stated areas are not protected in the same way. Norof closely related the two in turn being sub-branches of the Rhaeto-Romance group. The recognised linguistic minorities are, not surprisingly, in areas where the borders of the Italian state(s) have oscillated historically. In contrast, the southern part of the peninsula is peppered with individual villages which preserve linguistically the traces of that region’s turbulent past. It is here that we find Italy’s 100,000 Albanian, 20,000 Greek and 3,500 SerboCroat speakers, as well as a number of communities whose northern dialects reflect the presence of mediaeval settlers and mercenaries. Sardinia too contains a few Ligurian-speaking villages and 20,000 Catalan speakersin the port of Alghero as evidence of former colonisation. More importantly, the island has almost 1,000,000 speakers of Sardinian, a separate Romance language which has suffered undue neglect ever since Dante said of the inhabitants that they imitated Latin tanquam simie homines ‘as monkeys do men’. What he was referring to was the way in which Sardinian, both in structure and vocabulary, reveals itself to be the most conservative of the Romance vernaculars. Thus, we find a vowel system with no mergers apart from the loss of Latin phonemic vowel length; an absence of palatalisation of k and g; preservation of final s (with important morphological consequences); a definite article su, sa, etc. which derives from Latin IPSE rather than ILLE. Old Sardinian also maintained direct reflexes of the Latin pluperfect indicative and imperfect subjunctive. and the language is one of the few not to retain a future periphrasis from Latin infinitive + HABEO, using instead a reflex of Latin DEBERE ‘to have to’, e.g. des essere ‘you will be’. On the lexical side we have petere ‘to ask’, imbennere ‘to find’ (cf. Lat. INVENIRE), domo/domu ‘house’, albu ‘white’, etc. (contrast It. chiedere, trovare, casa, bianco). The presence of Italian outside the boundaries of the modern Italian state is due totwo rather different types of circumstance. First, it may be spoken in areas either geographically continuous with or at some time part of Italy, as in the independent Republic of San Marino (population 30,000), enclosed within the region of Emilia-Romagna, and in Canton Ticino (population approx. 325,000), the entirely italophone part of Switzerland. Both have local dialects, Romagnolo in San Marino and Lombard in Ticino, as well as the standard language of education and administration. Elsewhere, the historical continuity is reflected at the level of dialect, but with the superimposition of a different standard language. Thus, in Corsica (population approx. 280,000) the dialects are either Tuscan (following partial colonisation from Pisa in the eleventh century) or Sardinian in type, but the official language has since 1769 been French. The same situation obtains for those Italian dialects spoken in the areas of Istria and Dalmatia now part of Slovenia and Croatia. The second circumstance arises when Italian, or more often Italian dialects, has beencarried overseas, mainly to the New World. In the USA about one million Italian speakers constitute the second largest linguistic minority (after Hispano-Americans). They are concentrated for the most part either in New York, where they are mainly of southern origin and where a kind of southern Italian dialectal koine has emerged, and in the San Francisco Bay area, where northern and central Italians predominate, and where the peninsular standard has had more influence. Italian language media include a number of newspapers, radio stations and television programmes. The current signs of a reawakening of interest in their linguistic heritage amongst Italo-Americans are paralleled in Canada and Australia, each with about half a million Italian speakers according to official figures. There were also in excess of three million émigrés to South America, mostly to Argentina, and this has led, on the River Plate, to the developmentItalian in the Australia had its origins in the language of an underprivileged and often uneducated immigrant class, in Africa – specifically Ethiopia and Somalia and until recently Libya – Italian survives as a typical relic of a colonial situation. Ethiopia also has the only documented instance of an Italian-based pidgin, used not only between Europeans and local inhabitants but also between speakers of mutually unintelligible indigenous languages. The position of Italian in Malta is similarly due to penetration at a higher rather than a lower social level. Research is only now beginning into the linguistic consequences of the postwar migration of, again mainly southern, Italian labour as ‘Gastarbeiter’ in Switzerland and Germany. Finally, two curiosities are the discovery by a group of Italian ethnomusicologists in 1973 in the village of S˘tivor in northern Bosnia of a community of 470 speakers of a dialect from the northern Italian province of Trento, and the case of a group of émigrés from two coastal villages near Bari in Puglia, who settled in Kerch in the Crimea in the 1860s and whose dialectophone descendants died out only in the late twentieth century.