Empires and Barbarians

by Peter Heather

  Central Europe

  Slavic expansion into central Europe between the Elbe and the Vistula was equally thorough. The key proof text is a short and unpretentious document of staggering historical importance: the so-called Anonymous Bavarian Geographer. Dating from the 820s, it was written by an anonymous geographer working somewhere in Bavaria. It surveys and names the Frankish Empire’s neighbours between the Rivers Elbe and Oder, and even attempts to give some indication of their relative power, each unit being given a rating in terms of the numbers of ‘cities’ (civitates) that it comprised (Map 18). What these cities may have looked like is a point we will return to in Chapter 1. The central point is that all of these units have Slavic names. We know from other sources that some Slavic-speakers had even penetrated west of the Elbe at certain points before the rise of the Carolingians, but these immigrants were never numerous enough in these regions to challenge the dominance of Germanic-speaking Saxons and Thuringians. The Anonymous gives out pretty much at the River Oder, and knowledge of areas still further to the east, between the Oder and the Vistula, was perhaps not yet common in Carolingian Europe of the early ninth century.26

  A fuller picture of the extent of Slavic domination in central Europe had to wait until the Ottonian era of the tenth century, when the third of the Frankish imperial dynasties stretched its tendrils of dominion eastwards from the Elbe. In 962, a nascent Polish state suddenly appears in the historical record, providing unimpeachable evidence of Slavic domination of its territories between the Oder and the Vistula as well, with Arab sources confirming the point. From the mid-tenth century at the latest, then, all of north-central Europe between the Elbe and the Vistula was now the domain of Slavic-speakers. Indeed, the fact that historical documentation for lands east of the Oder is available only for the tenth century surely shouldn’t be taken to mean that Poland was Slavicized later than Bohemia or Moravia. What we’re looking at are the dates when interaction between these lands and western European imperial power gathered momentum, not the moment when they first came to be occupied by Slavs. The overall revolution in north-central Europe effected by Slavic expansion is plain to see. In the first half of the millennium, all this territory between the Elbe and the Vistula had been dominated by Germanic-speakers.27

  But if Carolingian and Arab sources between them document the total Slavicization of central Europe by c.900 AD, they provide little insight into the chronology or nature of the historical processes responsible for it. With the end of the western Empire in 476, historical light on north-central Europe – fitful at best, in the Roman period – is pretty much extinguished for the next three hundred years. All that the written sources preserve are a few vignettes that shed a little light on spreading Slavic domination through the uplands of central Europe: the extension of the Carpathians westwards to meet the Alps. The first refers to events of the year 512, when, as we saw in Chapter 5, the unfortunate Heruli began their long march to Scandinavia. According to Procopius, they first passed ‘through the land of the Slavs’. Most likely the Heruli left the Middle Danube by the valley of the River Morava, the main natural route north out of the Great Hungarian Plain. If so, Slavs were already then established in what is now Slovakia. This conclusion is supported by a second incident, of 543. In that year, a Lombard prince named Hildegesius attacked east Roman forces with six thousand warriors, most of whom were again Slavs. Since the Lombards were still living at that point in the Middle Danube, before the arrival of the Avars, it seems probable that he recruited his Slavs from the edges of that region – the Morava valley again, or somewhere nearby. Our third marker dates to the end of the sixth century, when Bavarian militias had to fight off Slavic attacks in both 593 and 595. So within half a century of the Hildegesius incident, Slavic groups are documented another 250 kilometres to the west, on the fringes of Bavaria.

  A similar vision of the western extent of Slavic expansion by the early seventh century is provided by one of the most famous episodes in early Slavic history, the adventures of Samo, our Frankish merchant turned Slavic prince. In the course of his colourful life, which involved – amongst other feats – siring twenty-two sons and fifteen daughters by his twelve Slavic wives, it emerges that by 630 the Slavic Sorbs were established on the borderlands of Thuringia.28 This would suggest that they were entrenched somewhere in the southern Elbe region. They had by this date, according to the Frankish chronicler Fredegar, a ‘long-standing relationship’ with their Thuringian neighbours, which would date the Sorbs’ occupation of these lands to c.600 at the latest. From these few references it is possible to get some sense of a westward Slavic penetration through central Europe, working roughly along the line of the northern hinterlands of the Carpathian Mountains and the Alps in the course of the sixth century (Map 18). But this is all the sources give us, and there is nothing here about the northern lowlands or the shores of the Baltic.

  The archaeological evidence, such as it is, broadly confirms the picture. As we have seen, Korchak-type material assemblages probably first emerged in the outer arc of the Carpathians in the later fifth century, but then spread over a much wider area. To the west, they diffused right around the outskirts of the Carpathians and on through the central European uplands as far west as Bohemia and adjacent areas of the southern Elbe region. An additional cluster of Korchak remains has also been excavated further to the north-west, in Mecklenburg and Lusatia (Map 18). The archaeological pattern here is rather different from that of most of the Balkans. Instead of a few isolated finds of Korchak ceramics or the odd burial, the central European uplands have thrown up entire Korchak cultural complexes. Not just stray Korchak items, but an entire Korchak way of life – including agricultural production methods and patterns of social connection – came to be reproduced in these areas.

  When they were first identified, the Korchak materials of Bohemia and Moravia were dated to the mid-fifth century. But it is now clear that Korchak remains in Bohemia date to no earlier than the second half of the sixth. Brzezno is the oldest Korchak site identified so far, and its remains date from no earlier than c.550. This is entirely in line with the new dating evidence from Moravia, a little further east, where, again, Korchak materials have now been shown to have appeared no sooner than c.550 at the absolute earliest. Dendrochronology has also provided precise dates for the Korchak sites in the Elbe–Saale region, west of Bohemia. Here, too, they have extended the received chronology. The Elbe–Saale remains used to be allocated to the late fifth century or the early sixth; their earliest materials have now been dated to no earlier than the 660s.29

  The geographical spread of Korchak materials across south-central Europe thus amplifies the picture of Slavic expansion suggested by the stray historical references. The new chronologies have also put paid to older theories that an initial Slavic penetration into the Elbe region in the later fifth or sixth centuries was followed by a second wave of migration in the seventh. This hypothesis had in mind a potential parallel with the Serbs and Croats and the Balkans. It was based, however, on the appearance of brand-new types of pottery in the Elbe region, which were finished on a slow wheel rather than entirely hand-formed. The geographical spread of the subtypes of this pottery broadly coincides with the main tribal confederations known from the Carolingian and Ottonian eras (Map 18): the Wilzi (Feldberg pottery), the Lausitzi (Tornow pottery) and the Sorbs (Leipzig pottery). It used therefore to be thought that the appearance of the new pottery types marked the arrival in the region of these tribal groups. Dendrochronology has shown, however, that the sites containing these wheel-turned pottery types date not from the late sixth and the seventh century, but from the later eighth and ninth. By this date, Carolingian narrative coverage of the region is more than full enough to rule out the possibility of any further large-scale migration. The new pottery types therefore represent the spread of new ceramic technologies among Slavs already indigenous to the Elbe region. The later dating also makes much better sense of the fact that some of the pottery resembles
eighth-century Carolingian ceramics, by which they were clearly influenced.30

  From all these materials, therefore, a clear enough picture emerges of a ribbon of Slavic settlement extending westwards from the northern hinterland of the Carpathian Mountains as far as the northern reaches of Slovakia in or around 500 AD. About fifty years later, a Korchak-type material culture penetrated south into the river valleys around the Middle Danube, and pushed on westwards to Bohemia. Another fifty years further on, and Slavic groups were both threatening the fringes of Bavaria and establishing themselves in the Elbe–Saale region.

  So far so good; but we have not yet got to the heart of the Slavic takeover of central Europe. As we have seen, ninth- and tenth-century sources demonstrate that in this era Slavic-speakers dominated the entire North European Plain between the Elbe and the Vistula as far north as the Baltic. But this is a much bigger area than that encompassed by our ‘thin’ ribbon of Korchak sites along the central European uplands and part of the way up the Elbe, and the historical evidence only comes on stream after Slavs were well established here. So what do the archaeological materials reveal of the process of Slavicization in north-central Europe?

  A first stage seems to be reflected in the so-called Mogilany group of sites from the Cracow region of south-eastern Poland. They are probably best viewed as a local variant of the Korchak-type sites found in nearby areas of the Carpathians, which they strongly resemble. Mogilany sites produce a range of similar handmade ceramics, and are marked by the familiar sunken floored huts with stone-built ovens. As yet, and this is the only reason they have been given a different name, no cemeteries have been found alongside Mogilany hamlets. No dendrochronological dates are available for this group, so its dating has at the moment to rely on an older method. This was based on the fact that, in most of central Europe, the largely undatable Korchak-type remains succeed the materially richer and hence chronologically more helpful materials generated by its previously dominant Germanic speakers, before the phenomenon of culture collapse set in. The end date for these Germanic-type cultures in any given area, therefore, can provide a useful earliest possible date for the advent of Slavic settlement there, so long as two conditions apply. First, the immigrant Slavs must not have coexisted with the Germanic-speaking groups responsible for the richer material culture being used to provide the date. But, second, there has to have been no lengthy interval between the disappearance of Germanic materials in the area and the arrival of Slavs.

  Both conditions are potentially problematic, but the approach does work reasonably well where it can be tested against dendrochronological information in the south. There is no evidence, for instance, that the previously dominant Germanic cultures of northern and eastern Slovakia, and north-eastern Moravia, continued in existence past the year 500. In southern Slovakia and Moravia, together with Lower Austria (Austria north of the Danube), on the other hand, enough later Germanic materials have been found to suggest that they continued in use there until c.550. Bohemia also continued to generate Germanic-type cultures until a similar date.31 These chronologies are broadly consistent with the new scientific dates for the earliest Slavic settlements in these regions, suggesting that it is still worth applying the method to regions where more scientific dates are currently lacking.

  In the Roman period, Cracow, home of the Mogilany group, fell within the southern expanses of the old Przeworsk system. Its collapse coincided, as we saw in Chapter 5, with the rise of Hunnic power sometime in the first half of the fifth century. An imported fibula brooch found at the Mogilany site of Radziejow Kujawski can be dated to the later fifth or the very early sixth century, and the start of a second and distinct cultural phase within the group is marked by the appearance of metalwork datable to c.600 AD excavated at Mogilany pit number 45, providing an earliest possible date of the sixth century for what went before. It seems likely enough, then, that Korchak-type Slavic speakers spread into the power vacuum created by Przeworsk culture collapse in south-eastern Poland in the late fifth or the earlier sixth century, soon after they first became visible in the Carpathian region.32

  The early medieval period in most of what is now Poland is not marked, however, by the widespread dissemination of such Korchaktype remains. Areas north of Lublin, eventually extending as far west as the Elbe, saw the development of another regionally distinct archaeological system: the so-called Sukow-Dziedzice culture (Map 18). Up to now, a clear boundary has been drawn between this second set of sites and those of the Mogilany type. Although some of the smaller pot-types of each group are identical, the larger pottery is entirely different in shape, and a much wider repertoire of forms was used here than by potters working more directly in the Korchak tradition. Some of the pots even look like handmade imitations of those previously in vogue in the same lands during the period of the Germanic-dominated Przeworsk culture. The sunken log cabins (Grubenhäuser) which are such a distinctive feature of Korchak areas have also not generally been found in Sukow-Dziedzice lands. Apart from an isolated group on the fertile loess-type soils of Mazovia, Kuiavia and Celmno, the major house-type identified so far is an above-ground wooden cabin. This different building tradition has provided one of the planks of the argument that Poland became Slavic-speaking via an entirely different trajectory of historical development from that which was working itself out in Moravia, Bohemia and the southern Elbe region. To some scholars, the Sukow-Dziedzice and Korchak cultures look so different that the former must have been generated by an entirely separate Slavic-speaking population. According to different views, this population was either indigenous to Poland – having been long submerged under a Germanic elite – or moved into Poland after 500, and not from the Carpathians but from a second ‘Slavic homeland’ outside of Korchak-dominated Carpathian areas – perhaps Byelorussia. According to either of these views, the Slavicization of central Europe so evident in Carolingian sources was the product of two simultaneous but independent waves of Slavic expansion: Korchak-type populations from the Carpathians, and Sukow-Dziedzice from Byelorussia or from within Poland itself.33

  The tail end of old nationalist agendas seems to be lurking behind this determination that Poland should have had its own unique trajectory towards Slavdom. In particular, the idea that house-types can provide a secure means of so absolutely distinguishing two population groups has been undermined by some recent excavations. These unearthed sunken huts in three areas where they had been unknown: at Wyszogrod, Szarlig and Zmijewo. At Wyszogrod, moreover, contemporaneous sunken huts and surface cabins were found on the same site. These discoveries make it likely that continuing investigations will uncover Korchak-type sunken huts more generally within Sukow-Dziedzice territories, eroding the apparently clear line that used to be drawn.34 That said, because of the uncertainties of the linguistic evidence, it is perfectly likely that Slavic-speakers were more widely dispersed north of the Carpathians and east of the Vistula in the later fifth century, with the Korchak Podolians being no more than one subgroup among them. It is also entirely likely, that this broader Slavic-speaking population, if it existed, would have later become involved in the broader Slavicization of areas such as Poland. The much wider range of pot forms in use in Sukow-Dziedzice areas is very striking, and strongly suggests that, unlike the Mogilany group, Sukow-Dziedzice has to be seen as a more specific phenomenon than merely another local Korchak Polish variant. This could be because its Slavs had different origins, but, as we shall see in a moment, it may have more to do with conditions the immigrants encountered when they arrived in Poland.

  How quickly the new Sukow-Dziedzice cultural form spread across the area between the Vistula and the Elbe is difficult to say, since the internal chronology of the system has not yet been established. Germanic culture collapse in more northerly Przeworsk and Wielbark areas had occurred by 500 or shortly afterwards, which is consistent with a stray reference in a work by the east Roman historian, Theophylact Simocatta, which might just about indicate that some Slavs had reached
the Baltic Sea by the 590s (or, frankly, might mean nothing at all). On the other hand, scientific dates for Sukow-Dziedzice sites in Lusatia, in the former DDR, indicate that these belong to a rather later period. Scandinavian metalwork found with Sukow-Dziedzice remains at Rostow Karkow provides a terminus post quem for that site of just after 700 AD. Absolute dendrochronological dates from the actual site of Sukow-Dziedzice itself and a number of well holes in the same region have likewise provided eighth-century dates. These dates apply only to the westernmost Sukow-Dziedzice territories, and, since there is good reason to suppose that the Slavic spread worked generally from east to west, do not necessarily contradict a sixth-century date for some of the Polish materials. For the moment, however, that is the best that can be done, although more scientific investiations will certainly provide more information in due course.35

  In broad outline, therefore, the spread of Slavic-speaking domination across the whole of north-central Europe, documented in the Carolingian era, seems to have converged on the Elbe from two different directions, if not necessarily from two points of departure. One line of advance is marked by the ribbon of Prague-Korchak sites running through the uplands of central Europe into Bohemia – and even onwards, in places, west of the Elbe. This trajectory of advance extended over about a century, between c.500 and 600. A second line of advance is marked by the spread of the Sukow-Dziedzice culture across the North European Plain, which was equally successful. It spread eventually, if more slowly than used to be thought, as far west as the River Elbe, which it had reached apparently by c.700. Much here remains obscure, but a broad outline of the initial Slavicization of Europe west of the Vistula can be sketched in from a mixture of literary and archaeological sources.