Florin Curta

   One of the most persistent stereotypes about the early medieval history of Eastern Europe holds that the Slavs, at the time of their migration, were a "polyarchic tribal society with no elevated notion of sovereignty." No Theodoric arose among the Slavs to gather their scattered communities into a state and attempt a symbiosis with the Greco-Roman civilization of Byzantium.1 My intention in this paper is to take a fresh look at the early Slavic society in the light of written evidence. I will then focus on the applicability of the modern concept of chiefdom to Slavic society and compare Slavic leaders with "classical" examples of big-men and great-men, on the basis of a theory of symbolic power. By emphasizing the mechanism of the accumulation of power in the hands of the Slavic "kings," I will consider the archaeological evidence and how feasting may have promoted the growth of social inequality. I shall raise the question of whether current historiographical models are appropriate for understanding early medieval societies.

   The notion of the politically "primitive" Slavs of the early Middle Ages derives from Procopius' frequently-cited description of the Sclavenes and the Antes in the mid-500s:

   Some have argued that "democracy" is derisively applied here to what, in Procopius' eyes, might have been the opposite of Byzantine monarchy.2 Others called Procopius an unqualified witness, who could not distinguish between acephalous societies and "primitive democracies."3 Some others, particularly among Soviet historians, believed Procopius to have described what is now known

   One of the key arguments for interpreting early Slavic society as a military democracy is the chief's retinue of warriors.7 According to a sixth-century source, the Avars attacked in 578 a Sclavene leader named Daurentius and "the chiefs of his people" (hosoi en telei tou ethnous) (Menander the Guardsman, fr. 21). Some argue that this particular passage points to a tribal aristocracy, whose authority was presumably based on wealth differentials.8 That Daurentius was a warrior leader is beyond any doubt, but no evidence exists of the council of the elders, one of the institutions both Morgan and Engels viewed as a necessary condition for the existence of a military democracy. Nor can Menander the Guardsman's evidence be used to postulate the existence of a political hierarchy, in which the power of the military leader was checked by that of the "chiefs of his people." When Procopius refers to "the people" or to public affairs, there is no indication of chiefs (Wars 14.21-22). Where chiefs appear, there is no indication of their separation from the agrarian substrate.

   In Engels's terms, military democracy was a form of social organization typically associated with the dissolution of communal ownership and with the emergence of private ownership and exploitation, based on tribute and clientship.9 Chiefs set themselves apart from the agrarian substrate and ruled through a retinue of warriors. The warrior chief or king controlled and exploited the farming communities through tribute and taxation.10 Marxist theorists, however, tend to use a restricted definition of economic interest, one that does not acknowledge that the notion of "economic calculations" should be extended to all the goods, material and symbolic, without distinction, that present themselves as rare and worthy of being sought after in a particular social formation.11 The current literature depicts chiefdoms as institutions which depended on the interlocking of three major components of power: control over economy, military force, and ideology.12 It is precisely economic control that is absent from any description of early Slavic society. There are, however, clear cases of accumulation of "symbolic capital."13 This is particularly true in the episode of a Sclavene chief narrated by Michael the Syrian (X 21). During their raid into Greece, in the early 580s, the Sclavenes carried off on carts the holy vessels and the great ciboria from devastated churches. In Corinth, however, one of their leaders took the great ciborium and using it as a tent, made it his dwelling. The Sclavene chief clearly seems to have grasped the symbolic potential of the otherwise useless stone ciborium, shaped as it was like a canopy over a throne.14 This further suggests that, at least in this case, simple accumulation of "material capital" cannot account by itself for the

   An alternative to the "military democracy" model is the "segmentary society."15 When historians speak of the "segmentary society" of the Slavs, they usually refer to the late sixth- or early seventh-century military treatise known as Strategikon, whose author claimed that Sclavenes were unable to fight a battle standing in close order or present themselves on open and level ground (XI 4.9 and 19).16 This lack of strategy, he argued, was a direct consequence of their rudimentary political organization:

   "Lack of government," it has been argued,18 refers to a segmentary lineage system. The underlying idea of such a system is that the functions of maintaining cohesion, social control, and some degree of "law and order," which normally depend on specialized agencies with sanctions at their disposal, can be performed with tolerable efficiency, simply by the "balancing" and "opposition" of constituent groups. Evans-Pritchard described this as an "ordered anarchy."19 What is usually referred to as "segmentary society," however, is one that is in some sense structured in terms of descent, in terms of lineage.20 Lineages are social entities, which emerge only in a social situation characterized by competition between groups.21 Marshall Sahlins has argued that a segmentary lineage system was a predatory organization characteristic of societies in migration, for, as a social means of intrusion and competition in an already-occupied ecological niche, it developed specifically within a tribal society struggling with other tribes.22 Just as the African Tiv of more recent times,23 Sclavenes reacted violently against any attempts to impose upon them rulers from the outside. The author of the Strategikon knew that Sclavenes and Antes were "both independent, absolutely refusing to be enslaved or governed, least of all in their own land" (XI 4.1). He reports that while Sclavenes may in fact unite to attack or repel an enemy at one time, at another they will fragment into feuding factions, quarreling over land or personal injuries. Once the defensive objectives which had impelled confederation were accomplished, the confederation dissolves again into its several segments, and the leaders who had emerged now return to social oblivion or retain only local influence.

   Can we apply the model of a segmentary lineage system to the Sclavene case? In other words, was the early Slavic society structured in terms of descent? Inspired by Pierre Clastres's model of the "Society against the State," some historians would infer a segmentary system from the presumed absence of social mechanisms contributing to the consolidation of royal authority.24 Lineage theory and segmentation are not, however, the same thing. Although we do not know what mechanisms were responsible for the descent structure of the early Slavic society, we may be in a better position to identify elements of social segmentation.

   It has been observed that a segmentary structure of society involves a segmentary structure of space, the minimal unit of which represents the "primary

   Archaeological evidence supports this inference of small-sized segments. Sixth- to seventh-century settlements excavated north of the Danube river, in present-day Romania, Moldova, and south-west Ukraine, include only a small number of features per occupation phase, ranging from ten to fifteen.27 The most common type of residential structure in all these settlements is the sunken-floored building of almost square plan with an oven or hearth placed in one of the corners. Since the average area of a sunken-floored hut is 20 square meters, it could hardly have sheltered more than 5 persons, the size of a nuclear family. However, since lineage is an "emic" category,28 there is no way settlement features per se could tell us anything about social relations between the families who lived in them.

   It also remains unclear to what extent Byzantine sources provide reliable empirical facts about actual behavior. It is logical to believe that the author of the Strategikon had a better (perhaps first-hand) knowledge about Sclavenes than did Procopius. But when he claims that Sclavenes were always at odds with each other, he is repeating a well worn topos, used by many before him, including Procopius.

   The model of the "segmentary society" also ignores historical process. It is very unlikely that the Sclavene society remained "frozen" in its "primitive," segmentary stage after contact with the Empire. Despite claims to the contrary,29 Byzantine sources make it clear that Sclavenes had their own "kings." In fact, by its ideological definition of political action as the result of segments in balanced opposition to one another rather than as the affairs of particular individuals, the theory of segmentary lineage structures would imply the emergence of men entrusted with considerable authority and wielding great political power.30 As long as political leadership remains personal and does not become institutionalized into an office, an ideological assertion of egalitarianism can co-exist with considerable political inequality.31 Any attempt to convert symbolic into material capital may have resulted, in the case of the militantly egalitarian Sclavene "democracy," in periodic purges of would-be "tyrants," such as those clearly attested by Pseudo-Caesarius. Pseudo-Caesarius, a Monophysite monk writing in the 560s, is the first author to refer to Sclavene chiefs, who, according to him, were often killed at feast or on travel by their kinsmen.32 Pseudo-Caesarius used this example to show that the Sclavenes were living by their own law without the rule of anyone (anegemoneutoi), a remark which dovetails with the evidence of other sources.33 That the purge of would-be tyrants took place during feasts further

   Ever since Elman Service defined them as "redistributional societies with a permanent central agency of coordination," chiefdoms have been viewed as the prevailing form of social organization in early medieval Europe, existing first beyond the Roman frontiers and then persisting into the migration period.34 It is reasonable to ask, therefore, whether all Sclavene "kings" of the sixth and seventh century were truly chiefs. The terminology employed by Greek sources is very complex and difficult to interpret. Later sources constantly applied the term archon which designated a ruler with full, regionally organized authority.35 Hegemon is a term Menander the Guardsman employed frequently in reference to Saracen, Utigur, Alan, Frankish, Turkic, and Slavic leaders. This suggests that those whom he also called hosoi en telei tou ethnous were not subordinates or in any way inferior to Daurentius, the Sclavene chief attacked by Avars in 578 (see above), but enjoyed a similar status and membership in what seems to have been a tribal confederation. Other sources, such as Theophylact Simocatta, the author of Book II of the Miracles of St Demetrius and the author of the Strategikon employ the word rex borrowed from the late Roman administrative jargon to refer to independent barbarian leaders.36 Such leaders had significant power over their fellow tribesmen, a feature easily recognizable in the case of Musocius, the Sclavene rex mentioned by Theophylact Simocatta (VI 9.1). It is interesting to note that Menander the Guardsman and the author of Book II of the Miracles of St Demetrius referred to hegemones and archontes in plural, whereas rex was a title bestowed upon individuals, who were often known by name (Musocius, Perbundos). This suggests that there were many Sclavene leaders at any one time, but not all of them wielded the same kind of power. Pseudo-Caesarius' leaders who were killed at feasts or on travels, arguably by their fellow tribesmen, were not on a level with "king" Musocius, explicitly said to have had "subjects."

   This variety of leadership types may be best described, in anthropological terms, as the coexistence of three different sorts of power. Anthropologists distinguish chiefs, whose powers are largely ascribed and coincide with the privileged control of wealth, from big-men, whose powers are largely achieved and derived from the manipulation of wealth, and great-men, whose powers may be largely ascribed or achieved, but are not based upon the control of wealth.37 The distinction between chiefs and big-men goes back to Marshall Sahlins, who depicted the typical Melanesian leader as a "big-man," because he achieved his position in a context of egalitarian ideology and competition, and his Polynesian counterpart as chief, because he succeeded to a hereditary position in a context of social hierarchy.38

   Big men are leaders who organize feasts and festivals, daring warriors and commanders in warfare, orators, or men of authority who arbitrate disputes within the community. Some dominate by their physical strength, particularly in contexts where leading warriors are politically important, some by force of character.39 The concept of big-man has been applied outside of Melanesia when achievement rather than ascribed leadership status is under discussion.

   More recently, Maurice Godelier has proposed that the big-men system be-

   Ardagastus, as described by Theophylact Simocatta, fits the model of a great man well. He had a remarkable physical size and strength, which helped him escape from being captured by Romans in 593 (VI 7.3). He had a "territory" of his own, which Priscus' troops devastated in that same year (VI 7.5). It is interesting to note that the inhabitants of this chora are never referred to as his subjects, only as "Sclavene hordes" or his "followers" (VI 7.1 and 9.6; cf. VI 9.1). Ardagastus may have been a warrior leader, organizing raids into the Roman provinces. Warriors from afar may have come to his "territory" and joined him in his plundering expeditions.41 No mention is made of a village and, if we are to believe Theophylact Simocatta, Ardagastus was on the point of launching a new raid across the Danube, when Priscus' attack took him by surprise. That Ardagastus was a real threat for the Romans, is indicated by the fact that Priscus directed his first operations beyond the Danube against his chora. His power was undoubtedly achieved, with his remarkable physical strength at the basis of this political prominence. He had already begun to build a name for himself, when Priscus' expedition north of the Danube frontier put an end to his career. Though he may have survived the Roman aggression, Ardagastus fell back into social oblivion, for nothing further is reported about him in the otherwise well documented events of the following decade.

   Can we bestow the title of great men upon other Sclavene leaders? The case of Peiragastus is much more difficult to decide than that of Ardagastus. Peiragastus is briefly mentioned by Theophylact in relation to Peter's campaign north of the Danube, in 594. Theophylact calls him "the tribal leader of that barbarian horde" (VII 4.13) and then "their brigadier (touton taxiarchos)"(VII 5.4). He may have been a simple commander in warfare, not a true chief. Mention is made of forces under his command, but unlike Ardagastus, Peiragastus had no "territory." Immediately after his death in battle, Sclavenes "turned to flight" and Romans were concerned with pursuing them, not with ravaging neighboring villages which may have existed in the area. Peiragastus and his "horde" had come from afar in what might have been an expedition against Peter's troops. It is very unlikely, therefore, that he was anything other than a warrior leader.

   The association between Pseudo-Caesarius' leaders and feasting suggests they were big-men. The piling up of debts and the prospect of future gains for all supporters are held to be the critical aspects for understanding how competitive feasting fosters the emergence of wealth-accumulators.42 Big-men also play a key role in "making" groups. Their oratorical interventions during meetings, together with private persuasion, transform actions that would otherwise be construed as merely personal into collective ones as well. This seems to have been

   More than twenty years earlier, another rex, Musocius, had "subjects" whom he could send to reconnoitre or to give assistance to refugees from neighboring territories. Musocius' chiefdom was more limited in territory than Samo's. Theophylact's account of Priscus' campaign of 593 shows clearly that, in order to shatter his chiefdom, Roman troops simply needed to capture Musocius and to devastate his village.

   Big-man leadership, with its emphasis on competitive feasting is also visible in the archaeological evidence, particularly in settlement patterns. Ever since Gordon R. Willey introduced the concept, settlement pattern analysis has been viewed as a useful interpretative tool in detecting in archaeological cultures the reflections of various institutions of social interaction and control.45 Decisions as to how to organize the use of space within residences and settlements may indeed have been influenced by the group's socioeconomic organization. This influence, however, would be mediated by the kind of activities performed on a site at a given time. This shows up in excavation as an activity area, defined as an archaeologically consistent, spatially clustered, association of artifacts in a precisely dated archaeological horizon.46

   At Seliste (Moldova), two groups of buildings with sunken foundations were located on either side of a central place.47 The eastern group produced all of the needles, most of the amphora sherds, and all of the clay pans found on site. By contrast, all arrow heads, awls, and dress accessories (beads and bow fibula) were found in the western group. Furthermore, three of the five buildings in the east had no heating facility, which suggests they were used not as dwellings, but as workshops. The almost exclusive association of clay pans and amphora sherds with the eastern sector also indicates that certain activities were performed there involving the consumption of special foods.48

   Though on a comparatively smaller scale, the site at Bucharest-Soldat Ghivan Street shows an organization very similar to that of Seliste.49 Settlement features, all of them with clay ovens, were placed around a large area devoid of any structures. A large building on the northern side produced all of the tools and weapons found on site, while a neighboring structure supplied the only fragments of clay pans. No such artifacts turned up on the southern side. The bow fibula and the potsherd with an incised cross found in building 12, and the handmade lamp from neighboring building 15, stand in sharp contrast with the artifact distribu-

   The site at Davideni (Romania)52 was divided into two groups of sunken-hut buildings, presumably separated by a creek (Figures 1-2). The larger group to the north includes the biggest structures found on site, but also one of the smallest buildings, which was located in the middle of a central, open area. Though too small to accommodate a family, this structure had two heating facilities, a stone oven and an open hearth, and produced no tools and no dress accessories, only sherds of clay pans. Most other buildings surrounding the central area were equipped with two heating facilities and produced large numbers of clay pans. Most tools found on site, as well as a bow fibula and a double-layered comb were found in this area. There is only one structure with two heating facilities in the smaller group of buildings, to the south. This sector, however, produced three open-air ovens. Judging from the intrasite distribution of artifacts, the central area on the northern side of the settlement may have been a locus of industrial activities, such as smelting and, possibly, production of dress accessories. It was also an area of special activities involving consumption of special foods, since clay pans were more frequently associated with features equipped with two or three ovens which were located in this region. At Davideni, the distribution of clay

   This analysis of the intrasite distribution of artifacts on sixth- and seventh-century Romanian sites thus reveals a systematic organization and use of space. The most important characteristics of this organization are the presence of a central, open area, and the polarization of the artifact distribution. The central area may have been an arena for ceremonies involving the processing and consumption of special, cereal-based foods. As the center of intervillage social, political, or economic events, this area may have acquired a special public character as the symbol for the community as a whole. As suggested by the exclusive presence of such markers of social status as bow fibulae, the central area was, however, not only a locus of communal activity, but also an arena of social competition, a "beyond-the-households context" for the display of symbols of leadership.53

   The end of the sixth century was a period of increasing competition among Sclavene leaders. The author of the Strategikon knew that there were many Sclavene "kings," "always at odds with each other" (XI 4.30), a useful political detail for any Roman general who may have found himself in a position to make war against any of them. What were the stakes of this competition, we can only guess. The archaeological evidence, however, suggests that shortly before and after 600 A.D. symbols of personal identity came into higher demand. At this time bow fibulae found in Romania (Figure 4), Crimea, and Mazuria display the greatest number of links in their ornamental patterns. Long-distance connec-