Empires and Barbarians

by Peter Heather

  This conclusion is important in itself, but there is a bigger point here too. Given the higher level of documentation provided by Ammianus Marcellinus, the events of 376 provide an important test case, illustrating what might also have been possible in other less well-documented instances involving Germanic groups of the late Roman period. It cannot just be assumed that all migratory phenomena of this era took the same form, and some certainly did not. But if we think of 376 as round one of the traditional Völkerwanderung, then it is reasonable to think that it saw a large-scale movement being undertaken not by a single ‘people’, but by a coherent mass of population. And the picture is drawn for us by a well-informed contemporary who was evidently not slave to an ideological blindness about barbarians on the move. It also makes good sense given both the broader history of the Gothic world as itself the product of a migration into the Black Sea region, and the spread of political power and military capacity in contemporary Germanic society. To the predatory migration flow building up from small-scale activities into much larger forces, which was characteristic of the third century, we can add a second form of predatory (or, in the Goths’ case partly predatory) migration: the massed, mixed group. This is an important interim conclusion to keep in mind when considering the second stage of Roman frontier collapse.

  MOVEMENT OF THE PEOPLES

  About thirty years after the knock-on effects of Hunnic invasion destroyed Roman frontier security in eastern Europe, its frontiers in central Europe were plunged into similar turmoil. And unlike 376, when there was only one major frontier crossing, this second crisis had several distinct components. First, in 405/6, the Germanic King Radagaisus led a large and again, seemingly, mostly Gothic force into Italy. The sources are fragmentary, but these intruders came from west rather than east of the Carpathians, since they crossed into Italy via its eastern Alpine routes without passing through the Balkans. Also unlike the Tervingi and Greuthungi, Radagaisus did not stop to ask permission. His was a totally uninvited intrusion.30

  Second, at more or less the same time, a large and disparate grouping of barbarians left broadly the same region as Radagaisus’ force, but moved west along the line of the Upper Danube rather than following the latter south across the Alps. This group consisted for the most part of Vandals, Alans and Sueves, although there were numerous smaller population fragments attached to it as well. The Vandals (in two separate groups – the Hasdings and the Silings) had already appeared west of the Carpathians opposite the Roman province of Raetia (part of modern Switzerland) in 401/2. The Iranian-speaking Alans, originally steppe nomads, had occupied lands east of the River Don as recently as c.370. The identity of the Sueves, however, is more problematic. This term appears in Roman sources of the early imperial period, but not between about c.150 and 400. It most likely designates some of the Marcomanni and Quadi, who had formed part of the old Suevic confederation and who had been settled in the Middle Danubian region, again west of the Carpathians, since the early Roman period. More Sueves certainly occupied this same region in the fifth century, and, as Constantius II discovered in 358, the various kings of these peoples were in the habit of forming temporary political alliances amongst themselves. Drawing on these highly disparate sources of manpower, this combined unit eventually forced its way across the Upper Rhine frontier on to Roman territory. The traditionally accepted date is 31 December 406.31

  Third, the same era also saw two rather less dramatic incursions. In 407/8, shortly after the Rhine crossing, a force of Huns and Sciri led by a Hunnic leader called Uldin invaded east Roman territory in the Lower Danubian frontier zone. Formerly a Roman ally, Uldin had been established north of the river in this region since c.400. Then fourth, by 413, the Burgundians had moved a significant, if shorter, distance west to the River Rhine. In the third and fourth centuries they had built a power base in the Main region, east of the Alamanni. Somewhere between 405/6 and 413, they leap-frogged their old neighbours and established themselves both on and beyond the Roman frontier line in the area of modern Worms and Speyer. This represented a displacement of about one hundred and fifty kilometres from their fourth-century abodes (Map 8).32

  The surviving information about this second bout of frontier collapse is much less illuminating than that for the first (c.376–80) because we lack a surviving historical source of the calibre of Ammi-anus Marcellinus. Had it survived in full, the History of Olympiodorus of Thebes, a diplomat in the employ of Constantinople, would probably have told us much of what we want to know, but unfortunately we have only his account of events from c.408 to the sack of Rome in August 410 (though this bit is more or less complete).33 It gives great insight into some of the consequences of frontier collapse, but not into its origins. Hence it is no accident that historical debate has focused largely on the initial events on the frontier. Recent discussion, though, has allowed some common ground to be established between all parties, and brought into sharper focus the points of disagreement.

  Traditionally, all of these invasions were seen as part of the Völkerwanderung, the ‘movement of peoples’. The Vandals, Alans and Sueves were each whole ‘peoples’, large groups of men, women and children. How large, exactly, was always a bit mysterious, but certainly several tens of thousands of individuals. The Hasding Vandals are reported to have lost 20,000 warriors in a hard fight against some Franks even before they got across the Rhine. And given that the ratio of warriors to total population was generally reckoned at something like 1:5, this implied a total force for just the Hasding Vandals of well over 100,000 (since they clearly weren’t wiped out by the Franks). Two sources also give figures of seventy and eighty thousand respectively for the number of warriors that could be fielded by the Vandal/Alan coalition and the Burgundians, while Radagaisus is given a total following in the hundreds of thousands.34

  No one now believes that the size of forces implied by these figures can be correct. The Burgundians proved in practice never more than a second-rate power, whereas an army of eighty thousand would have made them overwhelmingly strong, and another source anyway gives the same figure as the size of their total population.35 But there is substantial consensus that the military forces deployed by these invading groups had to be significant, with several of them individually fielding warrior groups in the ten-thousand-plus range, just like the two main concentrations of Goths in 376. The scale of the destruction they wrought within the Roman system makes no sense otherwise, and the more specific figures confirm it.

  On the Roman side, the cumulative effect of fighting all these invaders shows up in an army listing (the distributio numerorum) of c.420. As A. H. M. Jones has demonstrated, this document shows that something like eighty regiments – close to 50 per cent – of the west Roman field army were ground into the dust between 395 and 410. Some of this damage surely occurred in fighting civil wars, on which more in a moment, but much of it was inflicted in the heavy fighting with the different invaders that followed after 405/6. More specifically, Stilicho, commanding general and effective ruler of the western Empire, had to put together a force of thirty regiments (numeri), plausibly fifteen thousand-plus men, just to attack Radagaisus. One of the few fragments from the earlier part of Olympiodorus’ History also records that, on defeating Radagaisus, Stilicho drafted twelve thousand of the better warriors in the Gothic leader’s following into the Roman army, confirming that this intrusion mustered well over ten thousand warriors, or quite plausibly twice that number and more.36

  For the coalition that crossed the Rhine, the one figure worth worrying about is provided by Victor of Vita, who records that when the Vandals and Alans among them crossed to North Africa they were mustered into seventy groups of notionally one thousand people (not warriors) each, making a total population size of seventy thousand – except that Victor also notes that this was a ruse designed by its leader, the Hasding Vandal King Geiseric, to make outsiders think the group larger than it was. Victor was a North African bishop writing a few decades after Geiseric captured Cart
hage in 439, but he was working primarily for a North African audience that had had to live with the Vandals and Alans. There is a reasonable case for thinking, therefore, both that he knew what he was talking about and that he had to remain on this point within the bounds of contemporary plausibility. A total Vandal/Alan population of something over fifty thousand – allowing for the exaggeration – would imply again well over ten thousand warriors, and the move to North Africa had been preceded by heavy losses in Spain. When it crossed the Rhine in 406, then, the group is likely to have been considerably larger, not least because the Sueves then formed part of it.37 The possibility for argument is endless, but the narrative of the groups’ activities, and the indications we have both of Roman counterforces and of group size, are all pretty consistent with one another. At least two of the units caught up in the central European frontier collapse could field anything up to twenty thousand warriors, perhaps a few more, and this does seem now to be widely accepted.38

  Although large, it is evident that the nature of the forces on the move was not so simple as the traditional characterization of them as ‘peoples’ would suggest. The Vandal Alans and Sueves were a brand-new alliance, not a people, and the same is true of the Sueves as a group, while the Vandals originally came in two distinct sub-units: Silings and Hasdings. And Silings, Hasdings, Alans and Sueves each originally came with their own separate kings. Radagaisus’ force may have been, similarly, a new alliance, although he seems to have been its only king, while the Huns and Sciri led into the Empire by Uldin were also a new political unit of the post-376 era.39

  Women and children are mentioned just explicitly and often enough, and in a wide enough range of sources, to suggest their presence. The wives and children of some of the followers of Radagaisus, who eventually found themselves drafted into the Roman army, were, we are told, quartered as hostages in a number of Italian cities. For the Vandals, Alans and Sueves we have no evidence contemporary with their initial moves across the Rhine, but a group of Alans operating in Gaul by the early 410s had its women and children in tow. And when the Vandals and Alans moved on to North Africa in 429, they were certainly then moving in a mixed body. The women (and hence their children) could have been acquired since 406, and some probably were; but this seems an unlikely and unnecessarily complicated way to account for them all, especially since we have explicit evidence elsewhere, not least in relation to the events of 376 where the phenomenon is now generally accepted, that Germanic and Alanic groups did on occasion move with families. This makes it likely enough that women and children were already present in 406. The fact that different sources can squabble over whether eighty thousand represents a total figure for the Burgundians or a count of just the warriors implies the same thing about this group. Even if they were not ancient ‘peoples’, the evidence very strongly indicates that we must still figure on them as mixed groups of tens of thousands.40

  Two more points have also won general acceptance. First, despite the varied trajectories of their intrusions into the Roman world – Radagaisus into Italy, the Vandals, Alans, Sueves and Burgundians up to and across the Upper Rhine, and Uldin into the northern Balkans – it is right to regard the participants as a clustered group. For although they went in different directions, all were to be found, just before they attacked, on or around the fringes of the Middle Danubian plain of modern Hungary, west of the Carpathians.

  Second, it was shortly after these departures that Huns in large numbers first moved into the same Middle Danubian region. It used to be thought that the Hunnensturm had swept west of the Carpathians as early as 376. But this was based on a misreading of the Roman poet Claudian who reports Hunnic attacks only through the Caucasus and not over the Danube in 395 (contrary to what has sometimes been thought), and on a miscasting of the Hunnic leader Uldin caught up in the events of 405–8. He was clearly a relatively minor figure, not a conqueror in the class of Attila the Hun. Between them, these observations indicate that the main body of Huns remained north and east of the Black Sea up to c.400 AD, and yet by 411/12 at the latest, and quite possibly 410, many had established themselves west of the Carpathians.41 Together these points of agreement nicely define the historical problem posed for us by the collapse of Rome’s central European frontiers in the first decade of the fifth century. Everyone accepts the large scale of the intrusive military forces involved in the action, most agree that there were women and children along too, that the crisis had its epicentre on the Great Hungarian Plain and that the Huns moved on to the plain shortly afterwards. But if this much is generally agreed, the underlying causes of the invasions remain hotly disputed.

  In 1995, having identified the Middle Danubian origins of most of the barbarian groups caught up in the crisis of 405–8 and established that Huns are first found there in large numbers soon afterwards, I argued that the collapse of Rome’s central European frontiers was best understood as as a rerun of 376, as it were, this time played out west rather than east of the Carpathians. Similarities in the nature of the migration units and the precise chronology of the Huns’ advance into Europe suggested to me that the crisis of 405–8 was caused by a number of Rome’s other barbarian neighbours having decided that they would prefer to take their chances in the Roman Empire rather than face the uncertainties of dealing with the Huns, echoing the choice made by the Gothic Tervingi and Greuthungi in 376. In other words, the crisis had fundamentally non-Roman origins and was caused by developments in Barbaricum.42

  Two recent studies have taken an alternative approach, locating the key causes of the crisis inside the Roman world, in a combination of evolving Roman policies towards outsiders and the politically dislocating effects of the division of the Empire into eastern and western halves. In his Barbarian Tides, Walter Goffart considers it possible that Constantinople may have encouraged Radagaisus’ invasion of Italy so as to distract Stilicho from his immediate ambition to take back from the eastern Empire control of parts of the Balkans (Roman east Illyricum) which had traditionally belonged to the west but were currently being ruled by the east. More generally, however, he argues that changes to barbarian perceptions of Roman policy and to the actual power of the Roman state, rather than the Huns, were the prime cause of the crisis. On the one hand, the continued authorized survival on Roman soil of the Goths who crossed the Danube in 376 as semi-autonomous political communities decisively increased the range of ambitions at play in Barbaricum. It raised the prospect for other frontier groups that they might enter economically more developed imperial territory without having to give up their group identity and cohesion. They were encouraged in this idea, the argument continues, because, at the same time, the west was – or was perceived to be – growing weaker. Both the actual and perceived weakness stemmed from the fact that, after the death of the Emperor Theodosius I in 395, a real separation grew up between the two halves of the Empire, ruled by different advisers in the names of Theodosius’ two minor sons, Arcadius in the east and Honorius in the west (ruled by Stilicho). This offered outside groups the prospect of being able to exploit imperial disunity to increase their chances of prosperity and survival on Roman soil.43

  A related line of argument has been put forward by Guy Halsall, who contends that two usurping western emperors of the late fourth century, Magnus Maximus (383–7) and Eugenius (392–4), stripped the north-western Rhine frontier of Roman troops so as to deploy them for their – ultimately failed – civil wars with the eastern Emperor Theodosius. Western troop losses in these conflicts were heavy, especially at the battle of the Frigidus in 394, and after 395 when he was in effective control of the west, the generalissimo Stilicho did little to restore the situation north of the Roman Alps because he was much more interested in pursuing his quarrels with rivals in Constantinople for control of the entire Empire. By the early fifth century, therefore, defence on the Rhine was largely dependent upon the goodwill of local barbarian client kings; and this was only one aspect of a more general withdrawal of Roman state control which
also manifested itself in the closing of the Trier mint after the fall of Eugenius in 394, and the transfer of the capital of the Gallic prefecture from Trier south to Arles. For Halsall, this withdrawal had a further effect of particular relevance to the crisis of 405–8. Coin flows to some sites in the Roman north-west were disrupted from the time of Eugenius onwards, and Halsall suggests that this extended into a decline or even interruption in the normal diplomatic payments that had been flowing across the frontier to the Empire’s semi-subdued clients for centuries. With their own political power structures thus threatened, these leaders instead moved their followers directly into Roman territory from 405 onwards, to seize the wealth that they needed to keep themselves in power. For both Goffart and Halsall, developments within the Empire thus prompted the Middle Danubian barbarians to move on to Roman soil, and the Huns then moved into the power vacuum they left behind.44

  Some of the factors identified in these arguments certainly had a major influence on how the crisis played itself out. There is a distinct strand of evidence that the advantageous terms granted to the Tervingi and Greuthungi in 382 were responsible for changing perceptions of what kind of deal it might be possible to negotiate from the Roman state. In the late 390s, the revolt in Asia Minor of some allied Gothic troops under a leader called Tribigild seems to have drawn initially upon resentments of other barbarians in Roman employ that they had not been granted such good terms. Synesius of Cyrene was already claiming in 399, likewise, that the treaty of 382 (specifically as modified in further negotiations between Alaric and Eutropius in 397) had led at least one other group of outsiders to ask for admission into the Empire on similar terms.45 Divisions between the eastern and western halves of the Empire hindered any coordinated Roman response. From autumn 405, Stilicho, effective ruler of the west, was, as we have seen, in dispute with Constantinople over the control of Illyricum, even threatening war over the issue. In these circumstances, there was no prospect of any eastern assistance for the west as its central European frontier began to collapse – not, at least, until after Stilicho fell from power in the summer of 408. Some military and financial assistance then followed, but by this stage the barbarians were well established on west Roman soil.46