Empires and Barbarians

by Peter Heather

  The Amal-led Ostrogothic takeover of Italy was as much of a political shock as the Vandal–Alan takeover of North Africa, but perhaps less of a social one, since on the whole Theoderic pursued a more conciliatory policy towards the Roman landowners of his new kingdom, although this should not just be taken for granted. At one point during the conquest, he threatened Romans who did not support him with the loss of their lands. After the conquest, though, he devoted much of his reign to establishing good relations with them, and many preserved their elite status under his rule. Like Geiseric, Theoderic – it seems – endowed his important followers with substantial landed estates, together with rights to annual donatives, but this process of enrichment seems to have been accomplished without the kind of expropriations seen in Proconsularis.

  Gothic settlement was clustered in three particular areas of the Italian peninsula (Picenum and Samnium on the Adriatic coast and between Ravenna and Rome, in Liguria on the north-west of the north Italian plain, and in the Veneto in the east), and what appears mainly to have been transferred was the ownership of landed estates, quite likely from public as much as from private sources. The tenant farmers who had customarily done the work may well have been left in place. As far as we can tell, these transfers caused no massive socioeconomic dislocation. The size of Theoderic’s Italian kingdom perhaps made it easier to find the requisite portfolio of estates, in fact, than would have been the case in a smaller one. Nonetheless, the immigrant Gothic military elites did form a new, politically dominant force. Some Roman elites participated in court politics, but, to judge by our narrative sources, the migrant Gothic contingent remained dominant in such crucial political areas as royal succession and war-making. The Ostrogothic seizure of Italy seems to have had only a limited socioeconomic impact, then, even if its political effects were much larger.25

  In the Visigothic and Burgundian kingdoms the position was similar, but with one important variation. Laws issued in the Burgundian kingdom indicate that affected Roman landowners had to hand over two-thirds of their estates, whereas the corresponding figure in the Visigothic kingdom was one-third. Again there is much about this evidence that requires careful discussion – too much to engage in here – but one especially relevant factor is that, at its fullest extent from the mid-470s, the Visigothic kingdom ran from the banks of the Loire to the Straits of Gibraltar. It was many times larger, in other words, than the Burgundian kingdom centred on the Rhône valley. If you stop to think about it, it is hardly surprising that more expropriation was required to satisfy the incoming outsiders in a smaller kingdom, where there was a more restricted range of landed assets available. That variation aside, however, the evidence suggests that broadly comparable patterns prevailed everywhere outside of Britain and north-eastern Gaul. In all the successor kingdoms created by partial elite replacement, the immigrants generated a dramatic enough political shock as the given territory passed under new management, but the indigenous population would have experienced these immigrations as ‘mass’ only in distinct and limited areas, where substantial amounts of land passed into migrant hands: Proconsularis, for instance, and the three clusters of settlement in the Ostrogothic kingdom. For the Visigothic and Burgundian kingdoms, the best we can do is guess as to the areas of densest settlement on the basis of place names and/or archaeological evidence.26

  From the perspective of the immigrants themselves, moreover, sanitary terms like elite transfer, partial or otherwise, entirely fail to capture the nature of the action. Even where they were ultimately successful in establishing themselves in conditions of prosperity, there is no mistaking the scale of dislocation and trauma that preceded settlement. Between 406 and 439 the Vandal–Alan coalition moved in two stages, first from the Great Hungarian Plain to southern Spain (2,500 kilometres), then a further 1,800 kilometres south and west, before their eventual capture of Carthage, the latter involving crossing the Straits of Gibraltar. The Alans amongst them, of course, had made an earlier, westerly, trek of 2,000 kilometres sometime between c.370 and 406, from east of the River Don to the Great Hungarian Plain. These distances are huge and would have been punishing for the groups’ weaker members – the old, the young and the sick – especially since they tended to be covered in great leaps: the Don to Hungary, Hungary to Spain, and – if perhaps less of an ordeal – Spain to North Africa. Each of these leaps was a traumatic event that would have caused substantial loss just in itself. In aggregate, such journeys may even have distorted age profiles within the groups, killing off old and young in disproportionate numbers. Among modern political migrants forced to move long distances, exhaustion combines with the propensity for disease generated by overcrowding to make the experience a deadly one. Close to 10 per cent, or a staggering hundred thouand, of the refugees from Rwanda in 1994 succumbed en route to cholera and dysentery. Most of our first-millennium migrants were making more calculated moves and were probably better prepared, but modern comparisons warn us not to underestimate the degree of loss inherent in movement alone.

  These losses were greatly increased by conflict with the Roman state. The initial Rhine crossing, for instance, may have caused more losses to the indigenous populations of Gaul and Spain than to the migrants. But, as we have seen, in the late 410s Roman counterattack destroyed one of the two Vandal groups as an autonomous unit (the Silings), and so savaged the Alans – up to that point the most powerful of the invaders – that they submitted to Hasding leadership. The subsequent Vandal–Alan conquest of North Africa, to judge by the admittedly sparse sources available, probably did not involve losses on the same scale, but in its early stages, at least, was hard fought.27 Sheer distance and heavy military conflict combined to make the process traumatic in the extreme for the migrants, even if they were ultimately able to insert themselves as a new elite in the richest provinces of Roman North Africa.

  Similarly with the Ostrogoths, the Visigoths and the Burgundians. Take particularly the Ostrogoths, who ended up, of course, as conquerors of Italy. This staggering achievement laid open to them the region’s wealth, which they proceeded to enjoy. It is not so obvious, at first sight, that their migration involved anything much in the way of trauma, but between 473 and 489 they (or at least the original Amal-led Pannonian group) first had to trek almost a thousand kilometres south, from Hungary to Thessaly. By 476, they had notched up a further 500 kilometres, moving north-east to the banks of the Danube. In 478 and 479 they then moved another 1,200 kilometres, first south to Constantinople, then west again (past Thessalonica) to Dyrrhachium on the Adriatic coast. As part of an agreement with the east Roman state in 482/3, they appear to have been resettled six hundred kilometres to the north-east, on the banks of the Danube again. The mid-480s then saw a contingent of Gothic soldiers campaign for the Emperor Zeno against the usurper Illus in Asia Minor (another 1,100 kilometres each way), before the whole group first besieged Constantinople (400 kilometres from the Danube) and then made its 1,500-kilometre trek to Italy. If Jordanes is to be believed, all of this was prefaced in the mid-450s with the small matter of a 700-kilometre journey from east to west of the Carpathians.

  There was also plenty of conflict. The group suffered no defeats on the scale that savaged the Vandal–Alan coalition in Spain, but the Amal-led Goths variously fought the east Romans, the Thracian Goths (initially) and the usurper Illus – in 473/4, 476/8, 478–82 (including, in 479, the loss of a baggage train), 484, 486/7 and 489–91. All in all, the trauma was still severe.

  The experiences of the Visigoths were similar, in terms of distances covered, to those of the Ostrogoths (from the eastern Carpathians to south-western Gaul), and they also suffered substantial losses. The moves undertaken by the Burgundians were much more modest: from southern Germany to the Rhine, and then south into the Lake Geneva region and beyond. But this relative absence of geographical drama was more than compensated for by much larger losses, particularly at the hands of the Huns in the 430s. All the migrants suffered huge dislocation to the pattern of their li
ves, and whatever the view from the Roman side of the fence, their experiences can only be characterized as mass migration.28

  Migration and Development

  The interconnections between migration and prevailing patterns of development in the period of Roman collapse were multifaceted and profound. Straightforwardly, the greater wealth of Roman provincial economies exercised an enormous pull on both of our main sets of shorter-distance migrants, the Franks and the Anglo-Saxons. To judge by the fortifications already required to keep them out in the fourth century, this wealth disparity had been calling from across the Rhine and the North Sea for some time. And once the power of the central Roman state collapsed, what was in many ways its natural consequence simply worked itself out. Militarized outsiders, previously excluded from Roman territory by force of arms, were able to take possession of the more developed economic zones within their vicinity, and the rest, as they say, is history.

  Disparities of development also partly explain the choices of the longer-distance migrants. Their motives were often highly complex, involving a large negative component, since they were regularly moving into territories of which they had little direct knowledge and where the Roman state was still alive and kicking, so that they were facing a much greater military threat. This being so, it took an extraordinary kind of stimulus – that provided by the dramatic events surrounding the rise and fall of the Hunnic Empire – to get them moving. In both of the main initial pulses of Hun-generated migration, 375–80 and 405–8, the motivation was political and negative: the desire to avoid coming under Hunnic domination. Later on, as the Hunnic Empire unwound, motivations remained equally political, if now more positive. The Amal-led Goths in the 470s in particular, but also Lombard groups slightly later on, moved on to or towards Roman territory to secure a larger share of the greater wealth available there. But even for the initial migrants of 375–80 and 405–8, the pull of Roman wealth had some role to play. Hunnic aggression may have prompted their initial decision to leave their established homelands, but at that moment it was open to them to seek new homes either inside the Roman Empire or somewhere outside of it. The fact that most chose the former again reflects the pulling power of greater Roman wealth. One predominant feature of the period can reasonably be understood as barbarian outsiders from beyond the limits of imperial, developed Europe using immigration as a device by which to access a share – for good and ill – of its greater economic wealth.29

  Equally important, prevailing levels of development explain what at first appears to be the very peculiar nature of the migration units undertaking at least the longer-distance moves. Because contemporary non-Roman European societies could field specialist warrior retinues only in their hundreds, any migration that involved taking on the Roman state would necessarily include women and children as well. As we have noted for earlier eras, only by recruiting from a broader range of militarized manpower, some of which naturally came with families, could military forces of sufficient size be generated for these ambitious enterprises. The families’ participation was itself facilitated by another feature of economic development in barbarian Europe. Prevailing agricultural regimes, at least among Germanic groups east of the River Oder – those falling within the areas of Przeworsk and Cernjachov cultures – meant that central and south-eastern European populations retained an inherent tendency to shift settlement site. Shifting settlement every generation or so in search of more fertile fields was obviously a completely different kind of movement from the major migratory episodes of the fourth and fifth centuries. Comparative migration studies suggest, however, that this would still have made them more open to being persuaded to participate in the large, carefully organized expeditions that were required to move into Roman territory with any hope of success.

  There is also, I think, one further level on which an interaction between migration and development profoundly shaped the action. What in the end made it possible for some of the first, longer-distance migrants of 375–80 and 405–8 to succeed on Roman soil was their capacity to renegotiate their group identities quickly enough to create new groupings that were too large for the Roman state easily to destroy. The Tervingi and Greuthungi who appeared on the Danube in the summer of 376 were each substantial in their own right, and they clearly got very lucky at the battle of Hadrianople two years later. No other confrontation of the era produced anything like such a one-sided result, and it bought the Goths breathing space in the form of the semi-autonomy granted them in the treaty of 382. At that point, the Roman state had by no means renounced hopes of renegotiating these terms at the point of the sword sometime in the future, so as to bring them much more into line with its traditional treatment of immigrants. One key element in frustrating these hopes was the unification brokered in Alaric’s reign between both of the original Gothic groups of 376 and the survivors of Radagaisus’ later attack on Italy – the process that created the Visigoths. In a subsequent treaty negotiated with this enlarged force in 418, the Roman state was forced to acknowledge that it could not seriously hope to destroy it in its present configuration – as a unified and autonomous Gothic grouping on the Roman side of the frontier line. Similar processes of unification, likewise, made it possible for the Vandal–Alan coalition and the Amal-led Ostrogoths to survive and flourish. In all of these cases, had the unifications not happened, the Roman state could have mopped up piecemeal a more motley collection of smaller groups, as was the fate of many smaller groupings even in this era. The capacity of immigrants to reconstitute themselves into entities capable of fielding 10–20,000 warriors, or just a touch more, was critical to survival.30

  This means that the longer-term transformation of Germanic society examined in Chapter 2 is entirely relevant to the migration processes of the fourth and fifth centuries. If the Huns had arrived in the first century instead of the fourth, and pushed the kinds of Germanic group that then existed across the Roman frontier, the overall result would have been very different. Because of the smaller size of the primary political units that existed in the first-century Germanic world, too many of them would have had to become involved in too complicated a process of political realignment to make the creation of entities capable of mustering twenty thousand or more warriors at all likely. The kingdom-forming groups of the fifth century were created out of no more than half a dozen, sometimes only three or four, constituent units. Precisely because these base units were by this date so much larger, the twenty thousand-plus warriors that long-term survival in the face of Roman power broadly required could be rounded up much more easily. To get that many Germanic warriors facing in the same direction in the first century, you would have had to unite maybe a dozen of the smaller and highly competitive units that then existed, and the political problems would have been correspondingly huge.

  Even so, the processes of political unification which unfolded among the immigrants in the late fourth and fifth centuries were anything but plain sailing. Each of the new kingdom-forming supergroups was generated out of something like four or five constituent units, and each had its own pre-existing leadership. This meant that there were always at least four or five potential leaders for these new confederations, all but one of whom had to be eliminated in some way. The threat of Roman violence played a huge role in generating broad acceptance of the need to create the new confederations, and some of the potential contenders were willing, it seems, to resign their claims on that basis. Among the Ostrogoths, Gesimund (or Gensemund) the son of Hunimund was remembered as an independent Gothic warband leader who decided to support the Amal dynasty from whom Theoderic came, rather than press his own claims. But other contenders were much less amenable and had to be suppressed. Other members of the Hunimund line pressed their own claims with vigour, in addition to which the Amal line had to defeat that of another Gothic king, Vinitharius, before the group entered east Roman air space in 472/3, and the rival Triarius line at the head of the Thracian Goths after that. Clovis’ unification of the Franks, likewise, involv
ed eliminating at least seven rival kings. It is not possible to explore parallel leadership struggles in the same kind of detail for some of the other new confederations, but they certainly occurred.31

  Furthermore these struggles for power among the newly united supergroups were themselves interacting with patterns of development, because they too were dependent upon the greater resources that were available in the Roman context. The smaller size of barbarian political units beyond the Roman frontier was not accidental, but structural, reflecting the amount of redistributable wealth available for kings to build up loyalty among their followers; and even much of this, as we have seen, usually came from Roman sources. But the greater wealth available on Roman soil, whether extracted directly or indirectly, changed the parameters of political possibility among the barbarians, enabling their leaders to conciliate and control much larger groups of followers. Even the only seeming counter-example, when you examine it, isn’t one. One large supra-regional power was constructed in non-Roman Europe in late antiquity: the fifth-century Empire of Attila the Hun. But even this could not support itself from the resources of the non-Roman economy. Rather, it was dependent on broad flows of Roman wealth in the form of plunder and tribute, and collapsed when these were cut off. Non-Roman levels of economic development by themselves were insufficient to sustain political structures of the scale necessary for carving out a successor state on Roman soil.32