Langues et civilisations
à tradition orale

          Le CNRS           Accueil SHS          Autres sites CNRS          
cnrs paris3 inalco paris3 paris3
Intranet Lacito English

  Accueil > Colloques > Colloque "Migrations" - Workshop Migrations

Colloque "Migrations" - Workshop Migrations

5-7 septembre 2007, Porquerolles (Var, France)
coord. F. Jacquesson

– Titles & Abstracts – (14 September 07)


Peter Bakker - Reconstruction of migration from South Asia to Europe: linguistic and genetic data on the Roma toup
The Indic affiliation of the Romani (Gypsy) language was established already in the 18th century. As there is no necessary link between linguistic and biological/genetic features, an Indian origin of the Romani speakers could only be assumed on the exclusive evidence of the Romani language - which is not sufficient proof. Additional, solid historical, cultural or genetic evidence is needed.
Contemporary linguistic data on the Romani varieties in Europe, combined with knowledge of documented changes in the languages of Asia with which Pre-Romani was in contact during the migration of its speakers, make a detailed reconstruction of a migration from India to southeast Europe possible, including tentative dating of movements through Asia.
Only in the past five years a number of empirical studies have appeared by geneticists on the Roma. Both overall genetic features, and rare hereditary pathologies have been studied - as well as studies trying to improve the health situation of this underprivileged group. Most of these studies took linguistic data into account, both in general terms and/or on the level of dialectal groupings. The resulting data shed light on the Indian origin of the Roma (now beyond doubt), and also to some extent the migration routes between India and Europe, and the dialectal diversification and gene flow in Europe. Most genetic data - to the extent that they can be interpreted - appear to supplement and corroborate linguistic data. Dating the outmigration from India, however, is not easy, with the linguistic and genetic data.
In my paper I will summarize the main studies by geneticists, and discuss some of the linguistic implications on the basis of what is known from Romani linguistics.

Pierre Darlu - Inference of migration pattern from space-time surname distribution toup
The analysis of the distribution of surnames has been widely used to describe structure and movement of populations at various geographical and temporal scales, by taking advantage i) of the fact that familial transmission of surname is attested, at least since the XIIIth century and depending on the places, and ii) of their large geographic diversity allowing to possibly recognise their origins. Moreover, historical surname data can be easily obtained from parish records, birth registers, notarial deeds, land registries or others sources, providing large amount of data on successive generations. However, the handling of such data requires specific processing and statistic modelling which involves methods often developed from population genetics.
Such approaches will be exemplified by surname data collected from different places and times. The selected examples will be intended to raise various questions connecting migration process and historical, sociological, or political changes: mobility of people and their settlement during the Middle Age (through the Irminon’s polyptique, 826); stability or local exchange between rural populations in Savoy from the XVIIIth to the XXth century; rural exodus and immigration from foreign country (Flanders, Italy) to labour market area, in France at the turn of the XIXth century

Francesco d'Errico - Archaeological evidence for migrations and cultural interaction in the
Middle and Upper Palaeolithic
Human migrations during the Middle and Upper Palaeolithic have been mostly identified on the basis of anthropological, genetic, and palaeoclimatic evidence. Genetic data advocates the hypothesis that modern human populations originating in Sub-Saharan Africa or Eastern Africa colonised during the Middle Stone Age the entire African continent. Anthropological data have been used to propose that Modern humans moved to the Near East ca 100,000 years ago and were replaced in this area at the beginning of the last ice age by Neanderthals coming from Europe. Depopulation and subsequent recolinisation of large areas of Eurasia and Africa are predicted as a consequence of the Toba super-eruption, which took place at ca 73,000 BP. A migration from Africa of modern human populations 50,000 years ago and their subsequent colonisation of the Arabic Peninsula, the Indian continent, the Far East and Australia are inferred by genetic evidence and has become a common place in the scientific literature. A recurrent recolonisation of Northern Europe by Neanderthals from Mediterranean refugia at the end of each glacial period has been also put forward. Paleogenetic and anthropological data are used to suggest that modern humans colonised Europe 40,000 years ago with little or no genetic input from local Neanderthals. Modern genetic variability would support an Eastern origin of Gravettian populations or an expansion of the Magdalenian from a Pyrenean refugium after the last Glacial Maximum.
We will first present a review of these alleged cases of migrations and show that the supportive archaeological evidence is either scant or can, in a number of cases, be rather used against the scenario proposed by other disciplines. We will then present results of three attempts to model colonisation, population interaction and eco-cultural niche expansion in the Middle and Upper Palaeolithic: the colonisation of Iberia by the Aurignacian, the interaction between late Neanderthals and modern humans at an European scale based on personal ornamentation, and the identification of the northern limit of the human range during the Last Glacial Maximum by applying the Genetic Algorithm for Rule-Set Prediction to archaeological and palaeoclimatic data.

Pascale Dollfus - Peopling and migrations : some remarks based on research in Ladakh (India's western Himalaya) toup
Ladakh lies embedded in the mountains of the Karakoram in the north-west, the Himalayas in the south-west, and the Trans-Himalayas at its core. Located at the western end of the Tibetan plateau, the region is inhabited mainly by a population of Tibetan culture: in the eastern part (i.e. Leh District), a great majority of the inhabitants are followers of Tibetan Buddhism, while in the western part the inhabitants are Muslims. Data on early peopling remain inadequate. No large-scale digs have been carried out and there has been very little systematic research to verify dates, locations and the stratification of settlements. Just how, when, and where the people inhabiting Ladakh today originated remains a mystery. There are many varied and contradictory theories, and we still have a long way to go before settling these issues. These peopling hypotheses, however, are largely worded in terms of the theoretical assumptions of past centuries. Some are based on naive Western evolutionism while others reflect the desire of writers to link the origin of Himalayan peoples to the people mentioned in classical Greek or Indian sources.
In this paper, I attempt to illustrate the possible pitfalls of taking what is commonplace at face value, and to do this I have chosen 3 accepted ideas related to peopling and migrations to illustrate my sentiments. As an anthropologist, in order to create a narrative of the past, I look beyond the textual material for clues in the practices engaged in and in oral traditions.
1 - Isolated behind high ranges, mountain peoples are more archaic and mix less than those inhabiting plains and valley floors.
2 - Farming on barren land such as in Ladakh depends on large-scale waterworks calling not only for great hydraulic expertise, but a great workforce (or to use a Marxist vocabulary, requiring a high level of production forces), then government control for ensuring a proper distribution of the water, that leads to an absolutist managerial state (Karl Wittfogel's theory on "hydraulic civilizations"
3 - Nomadism is but primitive survival, a stage in evolution following hunting-gathering and leading thence to agriculture and sedentarisation.

Michael Dunn & Ger Reesink - Linguistic traces of the colonization of New Guinea toup
It is well known that the linguistic classification ‘Papuan’ does not presuppose genealogical unity. Various attempts have been made to classify the non-Austronesian languages of the greater New Guinea area (including various islands in Eastern Indonesia in the west and the Solomon Islands in the east), for example, Greenberg (1971); Wurm ed. (1975), but these have met with serious criticism. The recent comparison of pronoun forms by Ross (2005) has yielded a tentative classification of a major Trans New Guinea family and some 22 smaller groups and isolates. Our goal is to determine the historical facts which explain similarities among languages where reason exists to suspect shared history (whether contact phenomena or shared ancestry in the deep past), but where the comparative method cannot be applied. Following the pilot study (Dunn et al. 2005) that investigated possible historical scenarios for both Oceanic and Papuan languages of Island Melanesia on the basis of abstract structural features, we coded stratified samples of 20 putative TNG and 16 non-TNG languages for 200 binary features. Using a number of different computational and statistical techniques we calculated network relations for these languages, and compare these results with the conservative classification obtained by Ross, and with migration proposals of Wurm (1975).

Jean-Marie Hombert [& Gérard Philippson] - Bantu Migrations: Population movements and contacts toup
Around 500 Bantu languages are spoken to-day in sub-saharan Africa. It is widely accepted that the populations speaking these languages started to migrate from an area close to the present day boarder between Nigeria and Cameroon some 4000 years ago. These migrations allowed for the spread of Bantu languages as well as Bantu culture and technology.
The first part of our presentation will illustrate to what extend linguistic reconstructions can help us to speculate on the chronology of several crucial developments of Bantu societies. We will focus in particular on agriculture and on iron technology.
During their eastward and southward movements, these Bantu speakers came into contact with other populations. The descendants of these populations have either been assimilated to Bantu speaking groups or have remained hunter-gatherers. Two of these groups, the Pygmies and the San (bushman) are well known and present an interesting puzzle for historical linguists:
     - the contact between Bantu and Pygmies resulted in the apparent disappearance of a putative language used by pygmy hunter-gatherers before contact with Bantu agriculturalists
     - the contact between Bantu and San speakers did not lead to the death of the languages spoken by the hunter-gatherers ; they even influenced Bantu languages (for instance by incorporating clicks in some southern Bantu languages)
In the second part on this presentation, we will show that these two superficially opposite situations are in fact quite parallel and that the main difference is a question of chronology: Bantu have been in contacts with pygmies for more than 3000 years but only over 1000 years with San speakers. San languages are quickly disappearing and it is likely that none of them will still be around at the end of this century. We will also show that “pygmy languages” have influenced Bantu languages spoken in the equatorial forest but in order to understand this point we have to accept that there was never a stage with a single pygmy “proto-language”.
Finally, we will compare our linguistic claims with the results of recent genetic studies on Bantu, Pygmy and San populations to evaluate to what extent the dispersal of Bantu languages and cultures correspond to actual movement of populations vs. spread of languages and techniques.

Alf Hornborg - The Arawakan expansion in pre-columbian Amazonia: rethinking "Migration" in terms of ethnogenenis toup
The Arawakan languages of South America at the time of European contact represented the most widely dispersed linguistic family on the continent, ranging from Cuba to Bolivia. The expansion of Arawakan languages has generally been attributed to riverine migrations of prehistoric populations through Amazonia, and was explained by Donald Lathrap in 1970 as a consequence of these populations’ adoption of manioc horticulture, which would have provided a demographic advantage in territorial competition with non-horticulturalists. However, this basically biogeographical model of the Arawak expansion in terms of simple demographic displacement does not consider what social and cultural theory might have to suggest on the matter. An anthropologically informed account would need to consider socio-cultural factors such as language shifts, multilingualism, intermarriage, politics, prestige, and the strategic construction of cultural identity (ethnogenesis), particularly along the major rivers that have been posited as corridors of migration, but that were more obviously trade routes conspicuously often dominated by Arawak-speaking traders. The Arawakan “migrations” definitely involved some movement of people, but probably in a smaller scale and different way than previously postulated. The widely dispersed Arawakan dialects encountered by Europeans may be testimony not so much to prehistoric population movements as to the integration of a regional trade network spanning most of the Amazon Basin and linking it to the Andes and to the Caribbean.

François Jacquesson - Questions about migrations and peopling. Attila and Theodoric toup
In order to criticize usefully the simplistic idea we often have of « migrations », I describe two apparently opposite cases drawn from the classic treasure of migrations in stories and histories. Attila and the Huns were not a brutal and sudden wave devastating Europe. Actually, Huns were know before (and even after!) Attila's campaigns as faithful mercenaries. Several Roman officers of high rank knew Attila, sometimes personnally, or had friends ammong the Huns. Moreover, during the final clash in Champagne in 451, most troops in Attila’s army were not Hunnic, but Germanic; the same was true in the Roman army. Therefore, it is very difficult to see the Huns, or even Attila’s campaigns, as « a people » or a « nation » on the move. Theodoric and his Goths came to reign in Italy at the Greek emperor’s request. Theodoric knew him well, could speak Greek with him, because he has passed years in Byzantium. The whole effort of Theodoric the Goth, when in power, was to make the two groups, his and the Roman, equal in law and in fact. Most of his work was to assert and support what remained of Antiquity. A few years after his death, the Byzantine emperor took over in Italy, and imposed the new fashions, which we can consider as the first marker of Middle Ages. Attila was left out of the inner circle of Romano-Greek power, Theodoric stepped inside. The people they had with them knew two different fates. But together, Attila & Theodoric show that human migrations are not to be explained out only by climate or genes: they are social phenomena, deeply set in history.
L’idée de migration implique bien souvent, dans l’imagination, celle d’une armée en marche : il suffirait d’y ajouter des enfants et des femmes. Cette fantasmagorie politique est née de l’exploitation, dans l’Europe des nationalités à partir du 18e siècle (et même avant), des « grandes invasions », couplées au mythe biblique de la Dispersion sur la Terre. Examinons deux cas très différents de migrations dans ce contexte, celui d’Attila et celui de Théodoric. Le cas des Huns, examiné d’un peu près, laisse la conviction non pas d’un raz-de-marée déferlant sur l’Europe, mais d’une manoeuvre politique traditionnelle des bureaux de l’empereur à Byzance, jouant des relations de clientèle entre les grandes familles de l’empire et les princes des régions extérieures, souvent déjà installés en dedans des frontières. Les Huns n’ont été longtemps qu’une population où recruter des mercenaires. En 451, lorsque les troupes d’Attila furent « arrêtées » en Champagne, elles en recrutaient à leur tour, et les troupes étaient essentiellement germaniques, aussi bien du côté des « Romains ». Le cas de Theodoric et de ses Goths est encore plus saisissant. Il avait passé plusieurs années à Byzance, et parlait grec. L’empereur lui suggère d’aller remplacer Odoacre (autre « germain ») qui devenait gênant sur le sol italien. Theodoric se prend au jeu, essaie d’intégrer les Goths et les Romains en une société nouvelle, équilibrée. Il prolonge en un effort désespéré la culture antique. Quelques années après sa mort, l’empereur byzantin fait reconquérir l’Italie et abolit l’œuvre de Theodoric. Attila est resté hors du cercle du pouvoir impérial, Theodoric est rentré dedans. L’un et l’autre montrent deux cas très différents de contact entre sociétés. Mais l’un et l’autre montrent ensemble qu’une migration n’est pas un fait massif ni naturel. C’est un drame humain complexe, susceptible d’avenirs très divers.

Alain Kihm - Creole languages and migration : the exception that doesn’t prove the rule toup
Creolization, a rare linguistic phenomenon, appears to be quite separate from the expected linguistic consequences of migration. Whereas migrations, common events in the world’s history, result in people switching to the main language of their destination, or imposing their own language, or (more rarely) becoming stable bilinguals, creolization means the creation of new languages (except for the vocabulary), and it is related to a unique historical event: the European conquest of the world between the 16th and 19th centuries. It occured with people moving not in real space, but mentally, from one socio-cultural world to another. When people did move in real space, it was not migration, but deportation, violently imposed to slaves, under duress when indentured labour was the institution. Creolization therefore depends on a number of necessary conditions, which together build up a scenario. The scenario will be detailed and illustrated with two case studies.

Djamel Kouloughli - L'expansion de l'arabe hors de son territoire d'origine toup
L'arabe, dernière des langues sémitiques à émerger sur la scène historique (au VIIème siècle, portée par l'expansion de l'Islam), a pourtant pratiquement submergé le territoire des autres langues sémitiques d'Asie, ainsi que celui du copte en Egypte et du berbère en Afrique du Nord. Ce processus ne s'est cependant déroulé ni de façon instantanée ni de façon homogène, et n'a réussi (à des degrés variables d'ailleurs) que dans des conditions spécifiques où le facteur migratoire semble partout avoir été décisif. C'est à l'examen des conditions générales de ce processus que sera consacrée cette communication.

Andrew K. Pawley - Island colonizers: patterns of movement, interaction and diversification in the Austronesian settlement of Remote Oceania toup
This paper will reflect on the patterns of movement, interaction, and linguistic and social diversification associated with the first human colonizers of the islands of Remote Oceania (consisting of those parts of Melanesia that lie east of the main Solomons group, plus Polynesia and Micronesia). The colonization of Remote Oceania, beginning just over 3000 years ago, is one large chapter in the complex story of the dispersal of the Austronesian-speaking peoples. The latter dispersal began in Taiwan about 5000 years ago, and in time led to the fragmentation of Proto Austronesian into a family of more than 1000 languages, spread two-thirds of the way around the world, and to as many distinct societies. I will first ask what we know of the culture and society of the first colonists of Remote Oceania, at what enabled them to successfully colonize so many island groups so rapidly, and at what halted their movement eastwards for 1000 years after they settled western Polynesia. Then I will review a longstanding puzzle concerning linguistic and societal diversification in Remote Oceania. Why is there such a marked contrast in the extent of diversity – as measured by the number and geographic extent of indigenous languages – between two regions that were first settled at about the same time by peoples of similar culture and language? One the one hand, the major island groups of Western Polynesia have just one language each and the large Fijian archipelago has just two (though these two both exhibit extensive dialectal variation). On the other hand, most of the island groups of Eastern Melanesia have a much greater degree of diversity. For example, Vanuatu has more than 100 indigenous languages and New Caledonia/Loyalties have about 25, all restricted to a small geographic area. It is clear, from the genealogical relationships of the languages, that diversification in each of these regions has very largely developed in situ and is not (with a few exceptions) due to the arrival of new languages from outside. I will consider a range of factors that may have played a part in creating these regional differences, such as the size and physical geography of island groups, population growth and distribution, and changes in political structure and social norms, in canoe technology, and in patterns of interaction between dispersed populations following initial settlement of a region.

Philippe Ramirez - Tales of origin, cultural encounters and political claims in Northeast India toup
Is the history of migrations, as written by scholars, drawn up using only linguistic and archaeological material, while disregarding folk histories? We’ll consider the case of North-East India, described as one of the two gateways supposedly used by the main migratory flows which led to the present peopling of India. The Brahmaputra Valley acts both as the interface and the superimposition of Gangetic "Hindu" India and the hill tribes of South-East Asia. Thus the history of migrations is confronted here with a particularly critical constraint: migrations are a central reference for ethnic groups in their growing assertion. Any work on migrations thus necessarily leads to enquiring about the nature of the groups involved and about how they position themselves with regard to their current identity. Ethnic politics give rise to myths which feed or influence in one way or another “scientific” history. In return, the scenarii drawn up in the academic world are prestigious models on which ethnic elites can draw. The illusion is maintained on both sides, however, that writing these histories constitutes a relatively free exercise whereby each individual draws on effected linguistic and genetic knowledge, on historiography and indigenous myths to offer a likely scenario. Feedback between ethnic models and academic models represents one of the multiple interactions at stake in drawing up migratory history and whose importance I would like to emphasise here.
L'histoire des migrations dans sa version scientifique, s'élabore-t-elle à partir des seuls matériaux linguistiques et archéologiques, isolément des traditions populaires ? On considèrera l'exemple du Nord-est indien, décrit comme une des deux portes qu'auraient empruntées les grands flux migratoires à l'origine du peuplement actuel de l'Inde. Cette région est plus que toute autre celle de l'interface entre l'Inde gangétique « hindoue » et les tribus montagnardes d'Asie du Sud-Est. L'histoire migratoire doit ici faire face à une contrainte particulièrement critique : les migrations constituent une référence centrale des entreprises de légitimation de la part de groupes ethniques en pleine affirmation. Tout travail sur les migrations renvoie donc obligatoirement à une interrogation sur la nature des groupes considérés et sur leurs positionnements identitaires actuels. La politique ethnique produit des mythes qui alimentent ou influencent d'une manière ou d'une autre l'histoire à vocation scientifique. En retour, les scenarii établis dans le monde universitaire constituent des modèles de choix dans lesquels vont puiser les élites ethniques. De part et d'autre pourtant se maintient l'illusion que l'écriture de ces histoires constitue un exercice relativement libre par lequel chacun puiserait dans des savoirs linguistique et génétique établis, dans l'historiographie et les mythes indigènes, pour faire émerger un scénario vraisemblable. Les feedbacks entre modèles ethniques et modèles universitaires représentent l'une des multiples interactions en jeu dans la fabrication de l'histoire migratoire et dont on voudrait souligner l'importance ici.

Lolke Van der Veen - The Fang coming all the way down from Egyptland? Or how the walls of a myth may come tumbling down toup
The Fang expansion is a fairly well-documented demic expansion that took place within western Central Africa over the last millennium. Some 1000 years ago, a group of farming villagers left the Yaunde region moving southwards. Part of them, the Proto-Fang, also invaded the Gabon area. Their wave-like (and multidirectional) expansion intensified during the 17th century A.D. and came to an end only in the first half of the last century.
The language spoken nowadays by the Fang is found in Cameroon and most of northern Gabon. Fang comprises northern and southern dialects, as well as some intermediate varieties (Medjo Mvé, 1997). In Gabon, it has a strong impact on the surrounding languages and is currently replacing several of them (e.g. the Shiwa language (an A83 dialect) spoken by the Makina).
“Fang, their language and their culture came all the way down from Egypt (or Sudan).” This theory was elaborated by Rev. Trilles (1912, 1931) in the early 20th century on the basis of Fang’s oral tradition (Mvet), and several linguistic, cultural and physical traits attributed to this group by external observers. It has become popular believe, especially among Black African scholars, where this believe often also takes a strong ideological dimension accusing (white) Egyptologists of falsifying ancient History. It suffers, however, from important methodological and theoretical weaknesses. Guthrie (1948), Hombert & al. (1989) and Medjo Mvé (1997) have shown that the Fang language presents all the traits of a regular Bantu language. There is absolutely no evidence of a non-Bantu substratum. Most cultural anthropologists also reject the alleged parallels and consider the “Out-of-Egypt” theory to be untenable. Recent analysis of the mtDNA and Y-chromosome variation in Gabonese and Cameroonian populations does not support a non-Bantu origin for Fang either. Fang, as a whole, are not significantly different from the other farming populations of the area. However, the study of the Y-chromosome diversity reveals the presence of a rather unexpected haplogroup, which raises new intriguing questions, not only for Fang but also for several other Bantu-speaking groups of villagers and therefore also for part of the western Bantu expansion.

Vanhove, Martine - Migrations in the Maltese Islands: a linguistic melting pot toup
Malta, a strategic island situated in the center of the Mediterranean Sea, has been submitted to various foreign rules, and constant migrations throughout its history. If at the linguistic level, nothing has survived from the pre-Arab period, the Maltese language, althought a direct heritage from Maghrebi Arabic, has been heavily influenced by Romance languages (Sicilian, Italian), and more recently English.
This presentation will highlight the most important geographical and historical facts that explain for the present state of the language and will show how its different layers are connected with the different migration and conquest waves.

Imprimer Contacter le webmestre Plan du site Crédits Accueil