The Romance Languages

Martin Harris, Nigel Vincent

• Publisher: Routledge
• Year: 2003
• ISBN:
  9781134712281
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Languages
• Index: Languages & Linguistics

Available again, this book discusses nine Romance languages in context of their common Latin origins and then in individual studies. The final chapter is devoted to Romance-based Creole languages; a genuine innovation in a work of this kind.

1 The Romance Languages

Martin Harris

The Romance languages, whose history, structure and present-day distribution are the subject of the chapters which follow, share a common source: their development in each case may be traced back to Latin. Latin for its part developed from a form of Italic spoken originally in a number of small communities in Latium (Lazio) in central Italy, probably settled by Proto-Latin speakers around 1000 BC. The Italic branch of Indo-European appears to have been brought to the peninsula towards the end of the second millennium BC, and included Oscan (spoken over much of southern Italy at least until the time of the Pompeii disaster, as graffiti clearly testify), Umbrian (spoken in the north Tiber valley) and a number of other more or less well known varieties in addition to the Latin group of dialects. The label ‘Latin’ may be said to refer initially to this group of related dialects (including, for instance, Faliscan, spoken around what is now Cività Castellana, some fifty miles from Rome on the north bank of the Tiber), but it soon came firstly to designate the speech of Rome — attested since the sixth century BC — and then to be used as an increasingly broad cover term for a range of related varieties differing along temporal, geographical and social dimensions (see below). Latin was, as we have seen, bordered to the south and east by cognate tongues, while to the north its principal neighbour was the non-Indo-European Etruscan. Farther north still, by the fourth century BC — the time at which Rome was establishing her dominance in central Italy — the Po plain had been settled by speakers of varieties of Celtic (p. 3), a separate Indo-European family, but one which bears a number of striking structural parallels to Italic. In the extreme south, on the other hand, Greek was a recurrent source of external but still Indo-European influence.

As Roman military, political and economic influence spread during the period of expansion of the Roman Empire, firstly within Italy and then beyond, the Latin language also flourished, coming to be spoken in much of western and central Europe, and western north Africa, only Greek (spoken in the eastern Mediterranean and the Middle East) being a serious linguistic rival within the imperial boundaries. In particular, in addition to those areas of Europe which are currently Romance-speaking, much of southern Britain, the rest of what is now Belgium, Holland, much of Germany, Austria, Yugoslavia and Albania, and a fairly narrow coastal strip in what is today Morocco, Algeria, Tunisia and Libya were also Latin-speaking to a greater or lesser extent. The subsequent retreat of Romance from part of the territory it once occupied — which has of course in recent centuries been vastly more than offset by its expansion overseas — came about mainly in the period after the end of the Empire in the west during the fifth century AD (earlier north of the Danube; cf. p. 23), largely to the profit at that time of Germanic and Slavic languages within Europe and, from the later seventh century onwards, of Arabic in north Africa.

Despite the military and political collapse of the Western Empire and the subsequent loss of territory by Romance, Germanic and Slavic actually made less headway within Europe than might have been expected. In Iberia, for instance, the incoming Visigoths were already Latin-speaking before they arrived, and retained many aspects of the civilisation they found there, with themselves now in a dominant role; the continued use of Romance is therefore hardly surprising. In northern Gaul, to take a second example, a Catholic Frankish kingdom under Clovis emerged by the end of the fifth century, in which Latin was established from the outset as the language of both religion and administration and where a Romance vernacular — with a significant Frankish overlay — rapidly began to develop. The persistence — or reintroduction — of Romance in the area of present-day Rumania, on the other hand, is more difficult to account for, and is discussed in some detail below (pp. 2–3). The survival or otherwise of Romance when political mastery passed into other hands can be ascribed partly to the extent and profundity of earlier Latinisation and partly to the density and pattern of settlement of the newcomers; the use by the Christian church as its ‘official’ language of Latin/Romance (what some scholars call ‘Late Latin’ is the same as what others call ‘Early Romance’) is certainly also a relevant factor. It is with the subsequent fate of Romance in those areas where it did persist that this chapter is primarily concerned.

Of course, as was indicated earlier, even when Roman power was at its height there was not one single homogeneous form of Latin used by all speakers throughout the Empire: social and regional variation, particularly in the spoken language, would have been apparent at all times. There would, for example, have been considerable differences between the speech of Cicero and that of his slaves, or between the Latin spoken by a Roman provincial governor and that of his subjects, and the question of how such varieties should be distinguished and denominated is discussed below (pp. 26–7). During the period between the collapse of the Empire in the west and the emergence of the first Romance vernacular texts in various parts of Europe, one must envisage a situation in which this ever-present variation within Latin was accentuated as the language developed in ever more divergent ways in different localities. There are three main reasons for this. The first is simply the general tendency towards linguistic fragmentation inherent in the language acquisition process, counterbalanced at all times by the need to communicate with others within a shared speech community. Given the loss of a single uniform education system, and given the increasing separation of various groups of Romance speakers from one another, particularly after the rise of the Moslems in the eighth century shook the cohesiveness of the Western Romance world, this shared speech community must have grown progressively smaller for most speakers; thus the pressures offsetting fragmentation weakened and dialectalisation proceeded apace.

Secondly, there were already during the Empire incipient divergences between the Latin of various provinces, partly at least because of the different language or languages which were spoken (and often continued to be spoken for centuries) in various regions before Latin became the predominant language. Thus there are, for example, considerably more words of Celtic origin in contemporary French and north Italian dialects than in Spanish, standard Italian or (even more so) Rumanian, reflecting the Celts’ domination before Rome's expansion both of Italy north of the river Po (Gallia Cisalpina: ‘Gaul this side of the Alps’) and of most of present-day France (Gallia Transalpina: ‘Gaul beyond the Alps’). One representative example may be found in derivatives of a Latinised Celtic word RUSCA ‘ bark’ (cf. Welsh rhysg ‘rind’), surviving with various meanings ranging from ‘peel’ and ‘skin’ through ‘bark’ to ‘cork’ and ‘(cork) bee-hive’ in Gallo-Italian dialects, throughout Gaul, and in Catalan. Within Iberia there seem to have been several languages spoken in various parts before the arrival of the Romans, including (in addition to Celtic) both Basque, a non-Indo-European language still spoken in the western Pyrenees on either side of the Spanish–French frontier, and also another language or language family, Iberian, of unknown provenance and genetic relationship. Very often, the precise source of a word peculiar to all or part of Iberia is unclear; for this reason, those lexical items found in Ibero-Romance that are clearly of long standing and which are apparently neither of Latin nor Celtic origin — such as Sp. and Port, cama ‘bed’ (cf. pp 118,165)—seem best labelled simply ‘pre-Romance’. In much of central and southern Italy, most of the ‘substrate’ languages were themselves, as we have seen, of the Indo-European Italic group closely related to Latin itself, although there are limited traces of the influence of Etruscan. From the eighth century BC, there was significant settlement by speakers of Greek in southern Italy and Sicily, with some borrowing of lexical items into early Latin (p. 75), but while one or two Greek-speaking communities survive to this day (p. 19), the effects of this on local Romance dialects appear to have been minimal. As for possible pre-Roman influences on Rumanian, these are lost in the mists of time, partly because the present-day location of Rumanian is very probably not identical with that where Latin was first learnt (p. 23) and partly because we know virtually nothing of any pre-Roman languages in this entire area (cf. pp. 412–13).

The third reason for the increased linguistic divergence following the break-up of the Empire lies in the languages of the conquerors, whether immediate or subsequent. Thus one expects, for instance, to find most words of Germanic origin in French, particularly in those dialects — such as Walloon — nearest to the eventual Romance–Germanic frontier, with fewer in those other Romance-speaking areas where Germanic settlement was less dense. In Spanish (and to a lesser extent Catalan and Portuguese), one finds a substantial Arabic element (p. 119), reflecting the occupation of significant parts of the peninsula by Arabic speakers for nearly eight centuries, while in the case of Rumanian the constant contacts with Slavic and other non-Romance languages have led to a substantial non-Romance lexical element in the language even in everyday vocabulary (pp. 413–14).

The previous paragraphs have discussed the problem of linguistic divergence as though it were exclusively a lexical phenomenon: this is of course far from the case. Much has been written about the extent to which phonological, morphological and syntactic differences can be attributed to substrate or to adstrate factors, but in very few, if any, cases is general agreement reached. The pronominal use of on ‘one’ (< HOMO ‘man’) in French (p. 221) is a structure once regarded as certainly of Frankish origin; but while the parallel with modern German man ‘one’ (cf. Mann ‘man’) is indisputable, the direction of any influence between Germanic and Romance — and indeed whether such influence need be postulated at all — remains a contentious issue. Very often too, one finds that exactly the change or pattern under discussion is to be found also in some other Romance variety, or indeed in a totally different language family, in a situation in which the postulated external influence is wholly lacking. Such an instance is the passage of prevocalic initial [f] in Castilian (and Gascon) via [φ] and [h] to θ (‘the loss of initial [f]’), a development once confidently attributed to the influence of Basque or a Basque-related substrate, but parallelled in part at least in a number of southern Italian dialects, where a comparable cause cannot of course be adduced. In short, most non-lexical divergences, in Romance and elsewhere, seem best attributed to internal linguistic evolution, though of course the ‘selection’ of one change rather than another may be unconsciously favoured by structures found in other languages still actually in use in a multilingual community.

All of these reasons, then, led to the emergence of a number of linguistically distinct areas within the Romance-speaking world. The process of fragmentation, however, went much further. As we have seen, a language as it evolves is subjected always to two conflicting pressures simultaneously: the pressure towards convergence or homogeneity, which facilitates communication within a perceived speech community, and the pressure towards divergence or heterogeneity (p. 3), caused by the very nature of the language acquisition process, which ensures that no one generation or individual learns their native language in exactly the form in which it has been internalised by their elders. Enhanced social mobility, a high level of education and greater frequency and range of travel and communication strongly favour the former pressure, as the recent retreat of many nonstandard dialects clearly indicates; social and geographical immobility on the other hand, with very limited possibilities for education and travel, favour dialectalisation, with each community developing a form of language peculiar to itself, as part of a strong local identity. This process of course never goes so far as to prevent communication with those in the near vicinity; local dialects range along a spectrum, even in districts perceived as being on either side of a major dialect division. (Consider, for instance, the gradation from French to Tuscan via a whole set of French, Franco-Provençal and Gallo-Italian dialects spoken in adjacent parts of France, Switzerland and Italy.) Nevertheless the particular social context of the period between the end of the Roman Empire and the beginning of the Middle Ages did bring about marked linguistic divergence, the dialectal consequences of which remain, albeit now often rather marginally, in all of the present-day Romance-speaking areas within Europe.

From the early part of the Middle Ages, however, at least in the western part of the Romance-speaking world, the first signs of a new phase of linguistic evolution could be discerned, namely the gradual emergence in a particular area of one dialect more favoured for various reasons than any other; from these favoured varieties, at different speeds in different territories, a series of national languages has developed. The precise timing and result of this development, which affected written forms of the language markedly sooner than the everyday spoken idiom, depended on a whole variety of historical factors, in particular the establishment or otherwise of a nation-state in a given region and the policy, explicit or otherwise, of the linguistically dominant group towards those whose native form of speech was other than theirs; these factors are considered in detail below. At this stage we will simply contrast by way of example the fates of Portuguese and Occitan, the former now a major world language and the latter having little official status even in those areas of rural southern France where it is still in use. Portuguese, originating from the Galician dialect spoken in the north-west of the Iberian peninsula, came to be the language of an area which since the mid-twelfth century has been — apart from a brief period from 1580 to 1640 — politically independent of Spain, and has flourished accordingly, whereas Occitan, despite the high standing of medieval ‘Provencal’ (p. 16) as the literary language of the troubadours and the fact that Occitan dialects were once spoken over more than a third of France, could not compete with the strong desire which developed in the highly centralist post-revolutionary France for there to be a single national language. We return to this theme at several points in what follows.

We shall now look in turn at each of the branches of the Romance family of languages.

The Romance of Iberia

Within the Iberian peninsula, the major early division, apparent (with the usual caveat about dialect gradation) as early as the ninth century, was between Catalan on the one hand — which had and has close affinities with Occitan north of the Pyrenees (p. 16) and whose speakers were within the Frankish domains for several centuries — and the other dialects of Spain and Portugal, collectively referred to as Hispano-Romance. This latter group includes both the dialects of the Christian north (limited in the tenth century roughly to the northwestern third of the peninsula) and those of the Arab-dominated south, collectively known as Mozarabic. In linguistic terms we may observe that the eight centuries from the first Arab invasion in 711 near Gibraltar (< Arabic gebel al-Tariq ‘mount of Tariq’) to their final expulsion from Granada in 1492 can be characterised as a period involving firstly the gradual divergence of the dialects of the Arabised south from those of the north and then, slowly at first but later with greater speed, the recapture of Mozarabic- or Arabic-speaking territory by speakers of ‘Christian’ northern dialects. Simultaneously with these developments, we find at first the familiar process of linguistic fragmentation between the Christian kingdoms and then the gradual emergence of two of the resultant dialects, Castilian and Portuguese, to become in due course the national languages of Spain and Portugal.

More specifically, we may observe that as the Reconquest got underway, there was a range of Hispano-Romance dialects, traditionally grouped, largely because of the political divisions of the time, into four, these being, from west to east, Galician, Leonese, Castilian, and (Navarro-)Aragonese, with Catalan still further to the east. Speakers of each of these dialects gradually reoccupied territory to the south, but the central Castilian-dominated swathe gradually grew broader, to the point of cutting off the southward expansion of Catalan, Aragonese and Leonese at points close to Alicante in the east and Badajoz in the west, with a substantial strip further west christianised by speakers of Galician-Portuguese, who reached and recaptured the Algarve by the mid-thirteenth century, at which point modern Portugal may be said to have taken shape. At first, it was largely Mozarabic that these incoming dialects replaced (albeit possibly with some residual influence from Mozarabic on the dialects of Andalusia), but in much of the southern third of the country, from the latter part of the twelfth century onwards, it was often non-Romance languages, in particular Arabic and Berber, which gave way to Castilian, coming in from the north and now very much in a dominant position. Furthermore, along those eastern and western flanks of Spanish territory initially reconquered by speakers of Aragonese and Leonese respectively, Castilian gained ground fairly rapidly, a process helped no doubt by the fact that the differences between the dialects at that time were significantly less than those found now between standard (Castilian) Spanish and those forms of Aragonese and Leonese which continue to be spoken today.

There are three questions arising from this greatly truncated account which need a brief response. Firstly, why did the Christian dialects of the north have such an easy task in defeating both the Mozarabic forms of Spanish spoken in the reconquered areas and also the non-Romance languages of the occupiers? Secondly, why, within the central group of dialects, did Castilian fairly early emerge as dominant? And thirdly, why and to what extent have Galician and Catalan escaped Castilian hegemony?

The first of these questions is relatively easily answered: Mozarabic could not compete in prestige with the speech of the newcomers, and given the ‘religious crusade’ nature of the Reconquest, this was clearly even more true of Arabic or Berber. Further, in Arab Iberia, Mozarabic had the status of a spoken patois, the languages of culture and administration having been Arabic and, to a significant extent, Hebrew. All in all, the victory of the northern forms can readily be explained, and the principal long-term effect of Mozarabic on Spanish and Portuguese was as a medium whereby a considerable number of lexical items of ultimately Arabic origin passed in due course into the two national languages of the peninsula.

The second question reflects simply the central role played by Castilian in the Reconquest within what is now Spain. After the recapture of Toledo in 1085, and in particular after the reunion of León with Castile in 1230, this pre-eminence increased, to the point where, when Mozarabic was abandoned, as we have already seen, in favour of the language of the incomers, it was in fact, except of course where the new ruling elite spoke Portuguese or ...

Table of contents

• Front Cover
• Half Title
• Title Page
• Copyright
• Contents
• Abbreviations
• Preface
• 1. The Romance Languages
• 2. Latin
• 3. Spanish
• 4. Portuguese
• 5. Catalan
• 6. French
• 7. Occitan
• 8. Italian
• 9. Sardinian
• 10. Rhaeto-Romance
• 11. Rumanian
• 12. Romance Creoles
• Maps
• Index