Conceptions of Ethnicity in Early Medieval Studies
Walter Pohl   (p. 13-24 in: Debating the Middle Ages: Issues and Readings, Ed. Lester K. Little and Barbara H. Rosenwein, Blackwell Publishers, 1998)

Recently, the problem of ethnicity has been one of the most widely discussed topics in early medieval studies. From the historian's perspective, the discussion on ethnicity owes its decisive impulse to Reinhard Wenskus (1961; his approach was elaborated in the monographs by Herwig Wolfram in 1979 and Walter Pohl in 1988). Traditional research has taken the meaning of the terms "people" or "tribe" for granted. In this view, a "people" is a racially and culturally highly homogeneous group sharing a common descendance and destiny, speaking the same language and living within one state. Peoples (and not individuals or social groups) were often seen as factors of continuity in a changing world, as the real subjects of history almost immutable in its course, indeed more a natural than a historical phenomenon. Their fate was described using biological metaphors: birth, growth, flowering, and decay. This historical conception was rooted in the national movements of the nineteenth and twentieth centuries, and it had its share in encouraging all kinds of chauvinist ideologies. The idea that anything apart from one people living in one state was an anomaly (and should be corrected by all means) was, tacitly or explicitly, supported by many historians. Even today, after centuries of modern nationhood, the identity of people and state is the exception and not the rule, as the examples of Switzerland, Austria, the Germans, the Jews, the Arabs, the United States, or the Soviet Union show. Today's nationalist movements in many eastern European countries have rediscovered the nineteenth-century ideal of the homogeneous nation-state; it is sad to see that after so many tragedies it has brought about, some more seem to follow, and often in the name of history.

This situation explains the crucial importance of early medieval studies for the conceptions and preconceptions of ethnicity. Nations that for some reasons felt that they fell short of the "one people, one state" doctrine looked to those sombre times for a justification of their claims. The existence of Romans, Germans or Slavs in the fifth or seventh centuries became important arguments in an endless series of national struggles, culminating in the bizarre revival of the fair and reckless Germanic hero that lured an entire people into the Nazi Holocaust.

That the peoples in the Migration Period had little to do with those heroic (or sometimes brutish) cliches is now generally accepted among historians. But still, ethnic terms carry their load of emotions and preconceptions and tend to evoke misleading ideas. Even if we try to substitute "people" or "tribe" with the contemporary terms ethnos or gens for scientific use, we do not escape this methodological problem. It is remarkable enough that we still seem to rely upon biological metaphors for instance, when we speak of "ethnogenesis."


Even the Marxist ethnologist Bromley (1974: 69) has coined the term "ethnosocial organism (ESO)" to describe the interdependence of ethnicity with the social and political sphere. This type of imagery can help to express the complexity of ethnic structures; but it should not be misunderstood as placing them in the realm of nature, far beyond the reach of history.

It has become clear enough now that ethnic units are the result of history. It may be discussed if ethnicity, in a very general sense, has been a basic organizing principle from times of old, a position that the English sociologist Anthony D. Smith (1986: 6ff) has labelled "primordialist," vs the "modernist" view that sees the nation as a relatively new phenomenon. But single peoples (or ethne, to use the technical term I do not dwell on the problem if ethnos and "people" - Volk - narod cover the same range of phenomena) can have a beginning and an end; their composition changes; and their development is not the result of inherent "national" characteristics, but is influenced by a variety of political, economic, and cultural factors. Of course, ethnic changes are mostly a question of the longue durée ["long haul", a phrase often used by the French historian Fernand Braudel]; they scarcely even become obvious to contemporaries. In this respect, the Avars mark an exception it was noted how quickly they disappeared without leaving a trace (in a well-known passage of Nestor's chronicle, but also in a lesser-known letter of Nikolaos Mystikos cf. Pohl 1988: 323). But this "relatively persistent character of ethnic features" (Bromley 1974: 61) should not obscure its historical dimension.

Secondly, early medieval peoples were far less homogeneous than often thought. They themselves shared the fundamental belief to be of common origin; and modern historians, for a long time, found no reason to think otherwise. They could cite Isidore of Seville's seventh-century definition: "Gens est multitude ab uno principle orta" ("a people is a multitude stemming from one origin"); it has often been ignored that Isidore continues: "sive ab alia natione secundum propriam collectionem distincta" ("or distinguished from another people by its proper ties" - Isidore, Etymologies, IX, 2, i). Natio, in those days, was a near-equivalent to the term gens, whereas populus carried a connotation of a political body or a Christian community (cf. Loek 1990). It is hard to render the meaning of propria collectio; but I think it is as good as any modern definition trying to pin down the elusive characteristics of ethnicity. It was Reinhard Wenskus in his comparative study of German ethnogeneses who worked out some of the mechanisms of collectio, of collecting and holding together a gens, an early medieval people; and he made it clear that the idea of common origin was a myth. This myth, however, was an essential part of a tradition that shaped the particularity of the gens, its beliefs and institutions. A relatively small group guarded and handed on this tradition and set it up as a standard for much larger units; Wenskus calls this group Traditionskern (the term "kernel of tradition" had already been used in a similar sense by H. M. Chadwick in 1912, as my collegue Andreas Schwarcz in Vienna observed). This conception fits very well with recent sociological theories that see the ties within a given group based on a common interpretation of symbols (Girder 1982; Smith 1986 calls this a mythomoteur). These broader units were often of various origin; the Lombards in Italy, for instance, incorporated Gepids, Suevians and Alamans, Bulgarians, Saxons, Goths, Romans, and others. This "polyethnic" composition was generally observed; Alcuin, for instance, congratulated Charlemagne on having subdued "gentes populique Hunorum" the peoples of the Huns (i.e. the Avars); and his contemporary, Paul the Deacon, could list the components of the gens Langobardorum (Pohl 1988: 215; Paulus Diaconus, Historia Langobardorum 2, 26).

The third important discovery in the dynamic and often contradictory character of ethnic affiliations. The Russian ethnologist Shirokogoroff (1935) had already written in the 1930s


that we can describe ethnos as a process rather than a unit (cf. Daim 1982: 65). Ethnic boundaries are not static, and even less so in a period of migrations. It is possible to change one's ethnicity (otherwise the Indians would still be the only Americans we know). Even more frequently, in the Early Middle Ages, people lived under circumstances of ethnic ambiguity. We may cite Edica and his son Odoacer as striking examples; at different stages of their career they were taken to be Huns, Sciri, Turcilingi (or Thuringians), Rugians, Eruls, and even Goths, while Odoacer made his career as a Roman officer and eventually became king of Italy. We do not have to sort out the "errors" from this list to arrive at an "authentic" ethnic background; presumably, the two princes used a polyethnic background for widespread and flexible claims to the political loyalty of their partners and followers. However, they belonged to a social class in which ethnicity mattered. We have scarcely any proof that the lower strata of society felt part of any large-scale ethnic group (Geary 1983 remarks that our sources only put members of the upper classes including the armies into ethnic categories); it is more likely that their identity was rooted in smaller local groups, like clans or villages. To advance socially thus meant to grow into a dominating group with high prestige, to copy its lifestyle. Processes of assimilation produced a variety of transitional stages (Mühlmann 1985; 26f; Pohl 1988: 219). Also, there were individuals who were Avars or Lombards in a fuller sense than others who claimed to be so; and one could easily be Lombard and Gepid, or Avar and Slav, at the same time. Usually one of these names denoted the higher, constitutional unit, the other one a subgroup that had stuck to the reminders of another ethnic tradition. Which of these affiliations prevailed often depended on the situation: this is why Geary (1983) has called ethnic identity a "situational construct."

Consequently, we cannot expect to classify peoples in the same way that Linnaeus classified his plants. In the language of philosophy, ethnic terms are not classificational but operational terms; ethnic groups cannot be deliminated from each other clearly, and their reality has to be constantly reproduced by human activity (Oeser 1985; Girtler 1982; Pohl 1988: 14f). Therefore, we do not have to look for ethnicity as an inborn characteristic, but as an "ethnic practice" that reproduces the ties that hold a group together. In the political sphere, this means political actions and strategies that we can partly reconstruct from literary sources; on a cultural level, it denotes a rich variety of objects and habits that serve as expressions of ethnic identity. Later on, I shall return to the methodological implications for historical and archaeological research; now I want to give a few examples for the role and the contradictions of ethnicity in early medieval central and eastern Europe.

Avars, Bulgarians, Slavs, Romans: Types of Ethnicity

Maps of Europe in the early Middle Ages usually show clear ethnic boundaries; only in some cases do they allow for areas of overlapping populations. Coloured arrows denote routes of migration; although these sometimes become quite intricate (as in the case of the Goths) they nonetheless establish an equation between the bearers of identical names in different periods and areas. This picture may help the general orientation, but at the same time it obscures the variety of ethnic groups, the different forms of cohesion and their interdependence.

A map of Europe east of the Elbe, Enns and Adria rivers in the seventh century would show Avars, Bulgarians, Slavs, and Byzantines. In a certain sense we could call these four the dominant peoples of the period. A closer look, however, makes it clear that none of them was a "people" in the modern sense of the word, that ethnicity meant something different for


each of them; and that they all were tied to each other in a complicated pattern of interdependence that shaped the very form of their ethnic existence.

The world they all lived in had developed in the course of Antiquity; indeed, it was the creation of the Imperium Romanum. Perhaps it seems paradoxical to see the "mass of Slavs," as the Byzantines disdainfully called them, as an offspring of the Roman world; even more of a paradox is Patrick Geary's brilliant statement: "The Germanic world was perhaps the greatest and most enduring creation of Roman political and military genius" (1988: vi). But the very bipolarity between civilization and the barbarians, between the Imperium and the gentes, was a result of the Roman system. It attracted generations of barbarian warriors and profoundly changed barbarian societies; Roman gold, goods and symbols became inevitable factors of prestige among barbarians, social inequalities and tensions grew, and many a rural population was militarized by the enticing new possibilities to fight for or against the Romans who had accumulated such incredible wealth. Late Antiquity saw a veritable explosion of this "pull to the centre"; the dramatic confrontation between the Empire and the barbarians spread all over its provinces. But the first victim was not the Empire itself but its Germanic periphery that virtually collapsed in the course of the sixth century. German gentes either followed the lure of the Empire and managed to gain control of a part of its tax system paying the price of a certain assimilation (or, for many, subjection by a half-romanized warrior class). Or they lost their independence (with the exception of Scandinavia, and, partly, the continental Saxons and Frisians).

By the end of the sixth century, the old bipolarity had been reestablished on a "lower" level. New peoples had taken the role of the Germans in the vast areas outside and increasingly inside the ancient Danube frontier. On the one hand, Slavic populations proved rather impenetrable to Roman influence. On the other hand, there were those who took the place of the "outer" Germanic (and Hunnic) warrior aristocracies, living outside of the provincial framework of the Empire but inside its system of balancing power. The emperor had to keep the balance (Theophylaktos even polemically compares him to a referee in sports), financing a number of various armies and bands of more or less barbarian warriors and furnishing them with Roman titles and honours, playing them off against each other and thereby securing a kind of deadlock among them. Technically, Goths, Vandals, Franks, even Huns were not enemies of the Roman Empire they were its (initially external) members and federates, and their attacks were more of an upheaval than an invasion by foreigners (Wolfram 1990). None of their kings could have kept his position without the revenues from the Roman tax system, whether he took them by force or by treaty, whether he was charged with guarding part of it (again, usually by treaty), or was paid off directly.

The heirs of the Germans in this system were the Avars and the Bulgars. Their ethnogeneses in Europe would not have been possible without the Empire. It is characteristic that the first thing the Avars did when they came near the Caucasus on their flight from Central Asia was to send an embassy to the aging emperor Justinian. That took place sometime in winter 558/59, and they struck the usual deal: the Avars were to fight for the Empire against unruly gentes and in turn would receive annual payments and other benefits. Indeed, for 20 years to come the Avars, under their Khagan Baian, fought Utigurs and Antes, Gepids and Slavs, whereas their policy towards the Empire relied more on negotiation than on war. Toward 580, Baian had established his supremacy over practically all groups of barbarian warriors along the Balkan frontier. This gave him a monopoly only Attila had enjoyed before him for a few years. Despite growing difficulties with unruly groups of Slavs, his sons managed to preserve his position until the failure of the siege of Constantinople in 626. They used it to increase the pressure on the emperor and the


payments in gold and kind reached the record sum of 200,000 solidi shortly before 626. These treasures allowed the Khagans to win the loyalty of their growing army; nowhere could a barbarian warrior hope to gain more prestige than by following the Khagan with his legendary victories and his dazzling wealth (see Pohl 1988).

The ethnogenesis of the Avars followed the pattern laid out by this strategy. Their "kernel" had crossed a considerable part of Central Asia to escape from the Turks, who had just destroyed the two leading steppe empires, the Juan-juan and the Hephtalites. It is pointless to ask who exactly the forefathers of the European Avars were (see the detailed attempts of Haussig 1953 and Czeglèdy 1983 and the methodological objections by Pohl 1990 and 1988: 27ff). We only know that they carried an ancient, very prestigious name (our first hints to it date back to the times of Herodotus); and we may assume that they were a very mixed group of warriors who wanted to escape domination by the Turks. The charismatic tradition that they made their own proved a very powerful unifying factor, indeed a self-fulfilling prophecy. This tradition was indissolubly linked to the Khaganate. It maintained an absolute monopoly on the name "Avar" right until the end; it is remarkable that our sources do not call anyone else an Avar. Secessionists breaking out of the Khagan's dominion were known as Bulgarians; even in the case of the renegade Kuver around 680, whose following was initially composed of Roman provincials, Avars, and Bulgarians. After the fall of the Avar Empire around 800 the name disappeared within one generation. This did not mean that the Avars had all disappeared; a contemporary source around 870 calls the population of Pannonia "the people that have remained from Huns (Avars) and Slavs in these parts" (Conversio Bagoariorum et Carantanorum, c. 6; Pohl 1988: 325). It simply proved impossible to keep up an Avar identity after Avar institutions and the high claims of their tradition had failed.

The Bulgarians had roughly the same political and cultural background as the Avars: the world of the steppe with its specific forms of organization. And yet Bulgarian ethnogeneses took a very different path. From the end of the fifth century, Bulgarian groups (and other groups following a similar model of organization, usually carrying names ending "-gur" like Utigurs or Onogurs) appear in the sources. They are characterized by a marked discontinuity; Khan Kuvrat managed to unify the majority of Bulgarian groups under his rule north of the Black Sea only for a brief period in the middle of the seventh century. While Asparuch and his successors built up the Bulgarian state in the Balkans, Bulgarians stayed behind under Khazar domination, others lived in the Avar Khaganate, and many had joined the Byzantines. Kuver and Mavros failed in their attempt to erect a Khaganate of the Sermesianoi Bulgarians around Thessalonika, and the Alzeco-Bulgarians enjoyed an autonomous existence in the Lombard duchy of Benevento; later on, another Bulgarian ethnogenesis took place in the Volga region. Usually, these groups can be distinguished by their leaders. Whereas the Avar Khagans are usually only presented in the sources by their title (we have only a single name, Baian), we know all the important Bulgarian leaders by name; and they bear very different titles. Bulgarians often did not mind living under non-Bulgarian rule; but even in this case, they usually managed to preserve their ethnic identity for a long time.

In the world of the steppe warriors, Avars and Bulgarians represent complementary principles of organization. It is pointless to postulate any homogeneous ethnic substrate as a basis for these differences, for both gentes were the heirs of the polyethnic barbarian environment of the sixth century that had split up into so many bands and petty kingdoms. Of course, organizing a large group of warriors and their following always meant setting off an ethnogenesis; only ethnic bonds, supported by traditional myths and rites, could be strong enough to hold such a group together, to give it a structure that could resist failure. But even


so, it was by no means invulnerable. Misfortune led to erosion; after 626 many "Avars" turned to the new Slavic regional powers at the periphery of the Khaganate, or they became Bulgarians. Only in exceptional cases can these phenomena of migration and change of ethnic identity be traced in our sources as in the famous Kuver story in the Miracula Sancti Demetri (c. 2, 5; cf. Pohl 1988: 278ff).

If Avars and Bulgarians were, in a certain sense, the heirs of the Germanic military aristocracies between the Alps and the Black Sea, the Slavs took up another line. Traditional Germanic societies had virtually split up into an active, aggressive part that stimulated and fulfilled the Roman demand for military manpower on the one hand; and those staying behind as deprived peasants, unable to cling to the ancient rural traditions that gave their existence a name and a meaning. This gap was filled by the Slavs. They reestablished the old bipolarity between a more or less self-sufficient "barbarian" periphery and the Roman world. It has often been argued that their apparent "primitiveness" made them a victim to any foreign intervention. In reality, it was exactly the refusal to build up stable concentrations of power and to tolerate the establishment of a durable military kingship that, in the long run, secured the success of the Slavs. Avars and Bulgarians conformed to the rules of the game established by the Romans. They built up a concentration of military power that was paid, in the last resort, from Roman tax revenues. Therefore they paradoxically depended on the functioning of the Byzantine state. The Slavs managed to keep up their agriculture (and a rather efficient kind of agriculture, by the standards of the time), even in times when they took their part in plundering Roman provinces. The booty they won apparently did not create a new military class with the greed for more and a contempt for peasant's work, as it did with the Germans. Thus the Slavic model proved an attractive alternative for the lower classes on both sides of the old Danube frontier. What's more, it proved practically indestructible, in spite of all its defeats. It is characteristic that the final breakdown of the Danube limes, which had been kept against Goths, Huns, Gepids, Bulgarians, and Avars, took place after a series of Byzantine campaigns against Slavs. Emperor Maurikios wanted to continue the purge, but the army had learned its lesson well enough: the exasperated soldiers mutinied, marched on Constantinople, and overthrew the emperor. The war that could not be won was given up (Pohl 1988: 121ff, 159ff). Of course this does not mean that the early Slavs lived in a purely egalitarian society. In some regions (for instance, north of the Danube in the second half of the sixth century), tendencies towards the formation of a military kingdom appear; but these kings could never compare with the power of an Avar or Bulgarian leader, and they were systematically attacked by Byzantines and Avars alike.

The Slav way of life again represents a model complementary to that of the Avars and Bulgarians. Slav traditions, language, and culture shaped, or at least influenced, innumerable local and regional communities: a surprising similarity that developed without any central institution to promote it. For the theory of ethnicity this constitutes an important example: should we speak of one Slav ethnos (and we have reason to believe that a certain conscience of Slavic identity existed)? Apparently ethnicity operated on at least two levels: the "common Slavic" identity, and the identity of single Slavic groups, tribes, or peoples of different sizes that gradually developed, very often taking their name from the territory they lived in. These regional ethnogeneses inspired by Slavic tradition incorporated considerable remnants of Roman or Germanic population ready enough to give up ethnic identities that had lost their cohesion.

Even so, "Romans" survived as an important ethnic factor in eastern Europe. It is perhaps the most contradictory form of ethnic identity we know from the Early Middle Ages. In the first place, "Roman" was the Empire, an empire that had established itself as a singular form


of ruling "the world," of supremacy among the kings and princes. For more than a millennium to come, one could not be an emperor without being a Roman emperor. This political principle, that, to its defenders, belonged more to the realm of the history of salvation than to that of human history, had some astonishing effects: a Roman empire of the Franks and Germans, a "Third Rome" in Moscow, Seljuk princes named after the Romans whose neighbours they were; and, of course, the Roman Empire of the Greeks (which we call Byzantine by a modern name). For indeed, the Byzantines called themselves Romaioi, Romans, although they mostly spoke Greek, which they soon came to call the "Roman" language, and their capital was the "New Rome," Constantinople (Koder 1990). To them, being Roman gave them a special position among the gentes, the ethne: Roman identity meant belonging to an empire both earthly and heavenly, an example of civility and salvation. Even in the fifth century contemporaries believed it had forever brought ethnic conflicts to an end. This Romanitas already posed severe problems of delimitation to contemporaries; for us, it is almost impossible to define, as great numbers of barbarians lived inside the Empire, became Christians, and could even become highest Roman officials. In a certain sense Theodoric the Ostrogoth, who grew up as hostage in Constantinople and went through a splendid career as a magister militum, consul, and patrician, was a Roman; in another respect, he remained a barbarian. Large areas of ambiguity surrounded the notion of Romanitas.

On the other hand, even outside the Empire the Roman-Christian tradition in a variety of forms proved an important factor of ethnic aggregation. Again, the Kuver story can serve as a proof of how Roman provincials in a barbarian environment (in this case, as prisoners and slaves of the Avars) preserved their Roman identity. Roman enclaves in the Alps reached different stages of a local "Roman" ethnogenesis, as did the Aquitainians (Rouche 1990). The most striking example, of course, are the Romanians; centuries after the fall of the Balkan provinces, a pastoral Latin-Roman tradition served as the point of departure for a Valachian-Roman ethnogenesis. This kind of virtuality ethnicity as hidden potential that comes to the fore under certain historical circumstances is indicative of our new understanding of ethnic processes. In this light, the passionate discussion for or against Roman-Romanian continuity has been misled by a conception of ethnicity that is far too inflexible.

Some Methodological Conclusions

Many of the problems of early medieval ethnicity are, of course, unsolved. Still, it is possible to propose a few methodological principles.

1. "Ethnicity was not an objective phenomenon . . . but it was likewise not entirely arbitrary" (Geary 1983). Therefore, we cannot expect to find direct and objective proofs of ethnic identity. The subjective ethnic consciousness of an individual or a group of people was the decisive factor; but usually we do not know what group they felt they belonged to. Yet the specific form of ethnic consciousness gave objects an ethnic dimension. Archaeological objects or data can allow hypotheses about the "ethnic practice" that once gave them their context and meaning.

2. Many of the objects found by archaeologists may be direct or indirect expressions or symbols of ethnic identity especially if they were left behind in ceremonial form (e.g. in burials) but there is no object or group of objects that is ethnically unequivocal. As far as the many attempts to list distinctive and ethnically typical features have shown, almost anything can take or lose ethnic significance. Only an ensemble of objects and habits can serve as a basis for ethnic interpretation.


3.  The written sources have preserved a great number of ethnic names that become concrete when political events or cultural expressions are recorded in connection with one of them. Many of these names are, of course, topical (like "Scythians" for the Huns). Others are names given and used by foreigners (as Venidi for the Slavs). It is possible that one and the same group is recorded under different names; but also that two groups are justly or not subsumed under the same name. The complexity of the relation between peoples and ethnic names should not be reduced too easily. Likewise, we should be very cautious in identifying gentes bearing the same name in different contexts. As long as there is no apparent link between the two groups of the same or similar name, no direct connections should be postulated far less an identity, direct origin, or common ethnic substance.

4. Archaeological cultures and ethnic groups often coincide, but they cannot always be expected to be fully identifiable. Even more so, political boundaries, ethnic territories, linguistic groups, and areas of a certain material culture should not be mixed up, for they do not necessarily have the same extension.

5.   Different types of ethnicity should be kept in mind. They pose different problems of interpretation. As has been shown, we may assume that Avar and Slav ethnicity was based on radically different socioeconomic models; but ethnogenetic processes sometimes tend to obscure such distinctions. Thus, we cannot be sure if some Avar-style warriors would not have called themselves Slavs, or vice versa; if we want to be precise, we could speak of an individual following this or that model. However, we can reach relatively high statistical probability in such cases.

In the case of the Avars, a great number of grave finds may rather easily be attributed to the leading strata of warriors following the Avar Khagan. Quite characteristically, the discussion in this case has paid much attention to the ethnic composition of this group. To contemporaries, this question apparently was much less significant they were usually content to call Khagan's warriors Avars, even if they were of Cutrigur or Slav origin. To a certain extent, we may adopt a similar view. In spite of many attempts, no clear ethnic or socioethnic divisions within the Avar Empire have yet been proved; we may assume a very dynamic situation, showing all stages of acculturation.

In the case of the Slavs, it is much harder to trace groups mentioned in the sources in archaeological finds. Whereas the Avar warriors sought to display their splendor even after death, Slavs were often more modest; in other cases, they copied Avar or Byzantine apparel. Cremation also contributed to obscuring what they left behind. Thus, only areas of compact and stable Slavic culture can easily be identified, like north of the Carpathians; in other regions, for instance the Eastern Alps, sixth- to eighth-century Slavic culture remains quite unknown; and Slav components in mixed populations still are very hard to sort out. In those cases where clearly Slavic elements appear in conjunction with Avar features, as in the plains north of the Danube and east of the Morava, this should not be interpreted politically (Pohl 1988: 290f); again, this should be seen as a sign of dynamic processes of acculturation.

Bulgarians have long been a favourite for the attribution of warrior-type finds throughout eastern Europe. Not in all cases is this methodologically acceptable. We cannot always be sure what kind of material culture corresponded to the Bulgarian tradition and type of organization. Especially within the Avar Empire, this is not yet sufficiently clear. Hungarian archaeologists have tried to prove that middle and/or late Avar culture had been introduced


by a Bulgarian invasion; historical arguments for this hypothesis have turned out to be rather untenable (Pohl 1987, 1988; 282ff; to support the hypothesis, among many others, Bona 1988), and the cultural changes discovered so far cannot prove per se the appearance of a new people.

The most difficult problem is posed by finds of late Roman or Byzantine style. It is almost impossible to prove that the person who wore or used these objects really was and felt Roman. Byzantine culture and its various provincial and barbarized derivatives penetrated all early medieval cultures; barbarian leaders especially used all kinds of Byzantine objects as symbols of prestige and luxury goods. But again, archaeology and literary history both mirror one of the fundamental conditions of early medieval ethnicity: the gentes could only define themselves vs the overwhelming reality of a polyethnic, late Roman state and civilization. The degree of dependence on or refusal of Roman models shaped early medieval Europe; it laid the basis for success or failure of a great number of ethnogeneses. It was not any ethnic quality (force, number, talent) that decided their fate; it was their ability to adapt to an environment going through rapid changes, and to give this adaptation a credible meaning rooted in tradition and ritual. Slavs or Bulgars succeeded because their form of organization proved as stable and as flexible as necessary; the Avars failed in the end because their model could not respond to new conditions. The medieval peoples and their cultural characteristics were the result, not the condition, of this complex historical process.


Bona, I. 1988: Die Geschichte der Awaren im Lichte der archäologischen Quellen. Settimane di Studio del Centra Italiano 35: 437ff.

Bromley, J. 1974: Soviet Ethnology and Anthropology Today. The Hague and Paris.

Chadwick, H. M. 1912: The Heroic Age. Cambridge.

Czeglèdy, K. 1983: From East to West. Archivum Eurasiae Medii Aevi 3: 25-126.

Daim, F. 1982: Gedanken zum Ethnosbegriff. Mitteilungen der Anthropologischen Gesellschaft in Wien 112: 58-71.

Geary, P. 1983: Ethnic identity as a situational construct in the Early Middle Ages. Mitteilungen der Anthropologischen Gesellschaft in Wien 113: 15-26.
1988: Before France and Germany. The Creation and Transformation of the Merovingian World. New York and Oxford.

Girtler, R. 1982: "Ethnos," "Volk" und andere soziale Gruppen. Mitteilungen der Anthropologischen Gesellschaft in Wien 112: 42-57.

Haussig, H. W. 1953: Theophylakts Exkurs über die skythischen Völker. Byzantion 23: 275-436.

Koder, J. 1990: Byzanz, die Grechen und die Romaiosyne - eine "Ethnogenese" der "Römer"? In Typen der Ethnogenese, ed. H. Wolfram and W. Pohl, vol. I, pp. 103-11. Vienna.

Loek, F. 1990: Ethnische und politische Terminologie bei Jordanes und Einhard. In Typen der Ethnogenese, ed. H. Wolfram and W. Pohl, vol. I, pp. 147-52. Vienna.

Mühlmann, W. 1985: Ethnogonie  und Ethnogenese. Abhandlungen  der Rheinisch-Westfälischen Akademie der Wissenschaften 72: 9-38.

Oeser, H. 1985: Methodologische Bemerkungen zur interdisciplinaren Problematik der Ethno- und Glottogenese. In Entstehung von Sprachen und Voxikern, ed. P. Sture Ureland, pp.1-6. Tübingen.

Pohl, W. 1987: Das awarische Khaganat und die anderen Gentes im Karpatenbecken (6.-8. Jh.). Südosteuropajahrbuch 17: 4152.
  1988: Die Awaren. Munich.
  1990: Verlaufsformen der Ethnogenese Awaren und Bulgaren. In Typen der Ethnogenese, ed. H. Wolfram and W. Pohl, vol. I, pp. 113-24. Vienna.


Rouche, M. 1990: Peut-on parier d'une ethnogenese des Aquitans? In Typen der Ethnogenese, ed. H. Wolfram and W. Pohl, vol. I, pp. 45-51. Vienna.

Shirokogoroff, S. M. 1935: Psychomental Complex of the Tungus. London.

Smith, A. D. 1986: The Ethnic Origins of Nations. Oxford.

Wenskus, R. 1961: Stammesbildung und Verfassung. Cologne and Graz.

Wolfram, H. 1979: Die Goten. Munich.
1990 Das Reich und die Germanen. Berlin.