Empires and Barbarians

by Peter Heather

  Where linguistic evidence can give us very little help, however, is with chronology. We know the Slavic language family emerged relatively recently, but what does that mean? Some experts argue that the split with Baltic-speakers began only in the middle of the first millennium AD, at the precise moment when Slavic-speakers begin to appear in our sources. Others would place it much earlier – by maybe even a thousand or more years. This difference of opinion matters when it comes to trying to understand the Slavicization of Europe which unfolded after c.500 AD. If we should be envisaging very few Slavic-speakers at that date because the linguistic split was just beginning, so that Europe’s Slavic-speakers may have amounted to no more than the Sclavenes and Antae of Korchak and Penkovka fame, then the broad Slavic domination of Europe achieved by c.900 AD has to be accounted for from an extremely restricted demographic base. If, on the contrary, the Slavic linguistic family had emerged much earlier, the Sclavenes and Antae might only be two particular subgroups from within a far larger Slavic-speaking population. At this point, there is no way to be certain, but most of the experts would place the emergence of the Slavic language family much further back in time than the mid-first millennium AD, and it does make much better sense of the broader evidence for Slavic expansion to suppose that Slavic-speakers were not just restricted to Moldavia and Wallachia at that point.14 Nonetheless, it is worth keeping in mind both possibilities when trying to comprehend the explosion of Slavic dominance along its three main trajectories: south into the Balkans, west and north to the Elbe and Baltic, and east and north to the Volga and the fringes of the Arctic tundra.

  The Balkans

  For Slavic expansion into the Balkans, there is a relatively full selection of broadly contemporary east Roman and Byzantine historical sources. Until recent archaeological materials came online, they provided much the earliest body of information of any quality about early Slavic history. As a result, and this always happens when too many clever people have been studying a limited amount of information for too long, the subject area came to resemble a famous chess match, each intellectual gambit with its well-rehearsed counter. We have no need, fortunately, to become entangled in these set-pieces, since the broader outlines of Slavic expansion into the Balkans are clear enough.

  As we have just seen, Slavic raiding into the Balkans increased in scope and ambition towards the middle of the sixth century. In 547/8, a large raiding party spread south-west from the Danube through Illyricum as far south as the major Adriatic port of Epidamnos (Dyrrhachium). Procopius reports that these raiders captured many strongholds, a phenomenon not previously witnessed. The success encouraged further attacks. The next year, three thousand Slavs crossed the Danube and advanced on the River Hebrus. There they defeated some local Roman forces and captured the fairly major settlement of Topirus, by luring the city’s garrison into an ambush. Some thirteen thousand male inhabitants are said to have been killed in the subsequent sack, with many women and children taken prisoner. The year 550 then saw an unprecedentedly large force move south past Naissus, with the highly ambitious aim of capturing Thessalonica, the heavily fortified regional capital of the western Balkans. Eventually the raiders turned aside, moving through the mountains into Dalmatia and scattering in front of the major Roman army that was on its way north through the Balkans to complete the conquest of Ostrogothic Italy. When the army had passed, the raiders doubled back to the western Balkans, defeating a second, improvised Roman force at Hadrianople. Following this victory, the raiders spread to within a day’s march of the imperial capital of Constantinople itself.15

  There is no good evidence, though, that any of these Slavs were actually settling on a permanent or semi-permanent basis inside the imperial frontier at this point. The Antae were granted the old Roman fortress of Turris by treaty in 540, but this was north of the Danube and the whole point of the arrangement was to block further raiding on the part of the Sclavenes. Some Slavic place names, perhaps, figure in lists Procopius supplies of Balkan forts repaired or built by the Emperor Justinian (527–65), but, if so, the fact they are attached to forts might suggest that they were the outcome of authorized settlements of Slavic recruits into the Roman army rather than any proper migration as such. In any case, Slavs were not operating in sufficient force in these years to attempt a formal conquest of any part of the Balkans, or to capture major centres such as Thessalonica.16 The overall situation was radically transformed from about 570, however, by the rise of the Avar Empire.

  The Avars figure so strongly in what follows that they require some introduction. They were the next major wave of originally nomadic horse warriors, after the Huns, to sweep off the Great Eurasian Steppe and build an empire in central Europe. Thankfully we know rather more about them than about the Huns. The Avars spoke a Turkic language and had previously starred as the dominant force behind a major nomadic confederation on the fringes of China. In the earlier sixth century they had lost this position to a rival force, the so-called Western Turks, and arrived on the outskirts of Europe as political refugees, announcing themselves with an embassy that appeared at Justinian’s court in 558. The Emperor saw them as a new pawn in the great diplomatic game of divide and rule by which he sought to prevent really serious trouble in his north-eastern approaches. This, however, proved hopelessly over-optimistic. Not content with the role assigned them, the Avars quickly created an imperial power block of real menace. Attaching Bulgar nomads to their train, by 570 they had relocated to the Great Hungarian Plain, the old stomping ground of Attila, where they added Gepids to a growing list of conquered subject peoples. Their arrival also prompted the Lombards to leave for safer Italian domains on the other side of the Alps.17 If all this wasn’t enough, the arrival of the Avars also marks a watershed in Slavic history.

  Like many of their Middle and Lower Danubian neighbours, the Slavs of the Carpathian region found themselves targets of aggressive Avar ambition. The Antae seem to have suffered particularly at their hands in a punishing campaign of 604, which destroyed their political independence. On one level, the rise of the Avars meant that some Slavic groups now sought to move south of the Danube permanently, to escape their domination. In this area, massive Avar attacks on the east Roman Empire, particularly widespread in the 570s and 580s and again in the 610s, also provided such Slavic groups with much greater opportunity to pursue these ambitions free from Roman counterattack. At the same time, the Constantinopolitan authorities were having to defend their eastern territories in Syria, Palestine and Egypt against Persian and then Arab assault. The latter were a much richer source of tax revenues than the war-torn Balkans and always received – naturally enough – a higher priority.

  The new era announced itself in the 580s. The Emperor Maurice (582–602) was embroiled in a major war with the Persian Empire in the Near East, which sucked most mobile Roman forces away from the Balkans and allowed the Avars to launch a series of severe and wide-ranging attacks in Thrace. At the same time, Sclavenes mounted successive highly destructive campaigns in Thrace and Illyricum, the first really threatening assault upon Thessalonica, regional capital of Illyricum, occurring in 586. In the same year, ‘the fifth year of the Emperor Maurice’, one of the famous texts, the Chronicle of Monemvasia, even reports that Slavs took over all the Peloponnese except for an eastern coastal strip that remained in east Roman hands. According to the Chronicle, this caused a mass evacuation of ‘all the Greeks’ from the captured zones: the citizens of Patras went to Rhegia in Calabria (southern Italy), those of Argos to the island of Urok, the Corinthians to Aegina, the Spartans to both Sicily and to Monemvasia itself, a rocky, defensible peninsula in the southern Peloponnese.

  The terminal Slavicization of the Peloponnese, however, did not happen so early. The Chronicle of Monemvasia is a late text and, although preserving some authentic information, it kaleidoscopes the process of Slavic settlement. In the 590s, with the Persian War successfully won, Maurice was able to counterattack in the Balkans. Diplomatically, he paid the Ant
ae to attack the raiding Sclavenes, while his armies inflicted major defeats on the main Avar host in 593–5 and again from 599. In 602, his forces were even operating north of the Lower Danube, mounting a series of pre-emptive strikes which destroyed some whole Slavic groups. Letters of Pope Gregory I from the same period demonstrate that Church structures were restored in Illyricum generally, and in the Peloponnese in particular. While they certainly occurred, therefore, initial Slavic settlements of the 580s were swallowed up by Maurice’s counterattacks.18

  But this wasn’t the end of the story. From 604, repeating the pattern of the 580s, Maurice’s successors Phocas and Heraclius found themselves embroiled in a war with Persia, which by the early 610s was going diabolically badly, with control lost of pretty much the whole of Egypt, Palestine and Syria. Every military resource available had to be turned eastwards, opening the way to further Avar and Slav attacks on an unprecedented scale. In 614, disaster struck. Thessalonica avoided capture by a whisker. Salona, on the other hand, the largest Roman centre on the Dalmatian coast, fell into Avar and Slav hands, along with many of the Empire’s key cities in the northern Balkans, such as Naissus and Serdica. The action then spread as far south as the Peloponnese, when – amongst other things – Slavic raiders took to coastal waters in vast flotillas of dugout canoes. Constantinople itself eventually came under threat in a week-long Avar siege in 626. Alongside this military assault, Slavic settlement was gathering momentum.19

  Heraclius eventually won his war with Persia, but was immediately faced with the rise of militant Arab Islam. In contrast to the 590s, there was no opportunity this time to repair any of the damage done to the fabric of Roman life in the Balkans. Consequently, the disasters of 614 marked the definitive collapse of the Danube frontier of the old east Roman Empire, and paved the way for Slavic settlement across most of the Balkans: all the way from the Dobrudja in the north-east to the Peloponnese in the south-west. It is impossible to reconstruct a detailed narrative of this settlement process, but a series of vignettes, provided by various sources, leave us in no doubt as to its scale. In Macedonia in the northern Balkans, the Miracles of St Demetrius shows that large-scale Slavic settlement in the region of the Strymon River around Thessalonica was well established by the mid-seventh century. From one of its episodes it emerges that several Slavic groups were settled in the vicinity of the city by about 670, a point confirmed by later events. In the late 680s, the Byzantine Emperor Justinian II was able, if temporarily, to take the offensive in Macedonia, subduing the Slavic tribes of the region and restoring central imperial control. As part of the process he transferred reportedly as many as thirty thousand Slavs to Asia Minor. The reports also find some archaeological reflection. Seventh-century Macedonia and adjacent areas to the north did not see the spread of fully formed Korchak-type cultural systems across their landscapes, but many isolated discoveries of Korchak materials have been made in cemeteries and find-spots across Serbia and Croatia – Bakar Muntjac, Osijek, Stinjevac, Vinkovci.20

  Further east, in Thrace, Slavic settlement is equally well attested. When the first Bulgar state was established north of the Haemus Mountains in c.680, seven Slavic tribes already inhabited the region. They were resettled in an arc in the uplands around what became the Bulgar heartlands on the Danubian plain. Here the pattern of archaeological remains is different from that in Macedonia. Isolated Slavic ceramics, mixed with indigenous materials, have been discovered in sixth-century levels in cemeteries and rural zones around some of the fortresses of the frontier region, particularly Durostorum and Bononia. But excavations in northern Bulgaria have also uncovered sites such as Popina, where Korchak-type materials appear with no admixture of foreign imports. This and related sites used to be dated to the sixth century, but have now been shown to be later, postdating the definitive collapse of the Danube frontier in 614, which clearly marked the beginning of full-scale Slavic settlement in this part of the Balkans too. In archaeological as well as historical terms, the situation was then transformed by the arrival of the originally nomadic and Turkic-speaking Bulgars, but these further developments sat on the back of an earlier, large-scale Slavic settlement.21

  Literary and archaeological evidence also attests to a substantial Slavic presence further south, right in the heart of what is now Greece and the Peloponnese. The Miracles of St Demetrius mention in passing more Slavs, the Belegezitae, established near Thessaly and Demetrias. Later texts specifically mention other Slavs in the Peloponnese, not least the Milingas and Ezeritae in the vicinity of Patras, who in the early ninth century revolted against the tribute payments imposed upon them by a (slightly) resurgent Byzantine state. The archaeological echoes of this Slavic presence more closely resembled those of Macedonia in the north-west Balkans than those in Thrace in the north-east. Just a few, relatively isolated, finds of Korchak materials have been made, with no sign that the immigrant Slavs imported with them a complete material-cultural system. And some of the materials that used to be attributed to them probably had other origins anyway. A cemetery at Olympia, for instance, turned up twelve armed cremation burials of individuals interred in Korchak-type funerary urns. These are in all probability east Roman soldiers, if perhaps of Slavic origin, rather than independent immigrant Slavs. More convincing Slavic ceramics have been found at Argos, Messina and Demetrius, and Greece as a whole, like the rest of the Balkans, has thrown up a selection of the ‘fingered’ style of fibula brooch which was often, but not exclusively, sported by Slavs in the early medieval period. There are other possible explanations for this relative lack of Slavic materials. Above all, the first classical archaeologists, who were completely uninterested in medieval remains, ravaged most of the major Greek sites in the late nineteenth and early twentieth centuries and simply threw anything post-classical away. Nonetheless, it does appear that, again, the advance of Slavic groups into Greece proper did not generate entire Korchak-type material-cultural systems.22

  By the mid-seventh century, Slavic settlement was already affecting more or less the entire Balkans, but this is perhaps not yet the full story. According to one source, the north-west Balkans saw a further distinct wave of Slavic settlement. The De Administrando Imperio of Constantine Porphyryogenitus records that a first wave of undifferentiated Slavs originally settled in the lands now largely divided between Croatia, Serbia, Montenegro and Macedonia as Avar subjects, at the time when Avar rule was establishing itself in central Europe (from c.560 onwards). They were followed somewhat later, but still in the time of Heraclius (610–41), by two, more-organized, Slavic groupings – the Serbs and Croats – who arrived from the north to expel most of the Avars from that region (causing the others to submit) and establish their own rule instead, over Serbia and Dalmatia respectively. In the case of the Croats, the De Administrando preserves two versions of the story, one obviously Byzantine, the other Croat. These vary – as you might predict – on whether the Croats were invited to the Balkans or acted on their own initiative, and whether or not they promised to acknowledge Byzantine overlordship as a condition of settlement.

  The stories are famous, but it is difficult to know what to make of them. Serb and Croat nationalists have long cherished them as the origin stories of their ‘peoples’, arriving as fully formed units in the Balkans landscape. The problems they pose, however, are obvious. By virtue of being unique, they lack corroboration. They also occur in a comparatively late source, the De Administrando being a mid-tenth-century text, and their telling has a distinctly legendary tone: the Croats are led south by a family of five brothers. Not surprisingly, they have often been rejected outright. On the other hand, tenth-century Arab sources confirm the existence of other Serbs and Croats north of the Carpathians at that point, and there is nothing inherently impossible in the general action outlined. If it is accepted that they possess a kernel of truth, the stories suggest that some more organized Slavic groups asserted their independence from Avar rule by moving south into the Balkans and establishing some kind of a relation
ship with the Byzantine state before the death of Heraclius. Indeed, the northern Serbs (or Sorbs) themselves threw off Avar domination – if perhaps temporarily – in alliance with an ex-Frankish merchant, Samo, in about 630; that is, precisely in the reign of Heraclius. This was, in fact, a moment of general crisis for the Avar Empire following its huge defeat at the siege of Constantinople in 626, and the consequent loss of prestige for its ruling Khagan. Substantial numbers of its Bulgar subjects also escaped Avar domination by fleeing into Italy at this point, so that the idea that other Slavic groups were doing the same, either with or without a Byzantine invitation, is perfectly plausible.23

  But if this much is plausible, the seventh-century Serbs and Croats were not whole peoples responsible for the complete repopulation of these parts of the Balkans. As we have seen, the better-documented instances of first-millennium migration have never thrown up a case of total demographic replacement: some indigenous population elements always survive. And there is in fact a possible extra twist to this story. The group names ‘Serb’ and ‘Croat’, together with some of the personal names reported of their leaders, might derive from Iranian rather than from the Slavic language group. It has been suggested, therefore, that both groups may have originally been dominated by cores of Iranian nomads.24 This is not inherently impossible. It could have come about, for instance, by Slavic groups established in the northern Black Sea region becoming part of a military confederation dominated by Iranian nomads. There is not the slightest shred of narrative evidence to support such a view, but this is how nomads like the Huns tended to operate on the fringes of Europe. That the Serbs and Croats asserted their independence at Avar expense in the reign of Heraclius, perhaps around 630 when their Empire was in crisis, and that the Byzantines used them as part of a broader strategy for limiting Avar power in the Balkans, all seems likely enough. Whether we should envisage them as already entirely Slavic at this point, or as a structured confederation with distinct groups of Iranian-speaking nomads at their cores, is entirely unclear. It is also unclear whether their arrival represented a further major wave of Slavic immigration into the north-western Balkans, or whether they functioned essentially as an organizing element for Slavic groups already present there but formerly subject to Avar domination. If the latter, this would make them not unlike the Bulgars of the eastern Balkans.25