These suspicions have been confirmed by the comparative evidence about nomadic lifestyles gathered more recently by anthropologists. There are of course almost as many differences between different nomad groups as there are nomad groups in the first place. According to the types of grazing and animals available, practices and organization vary enormously. But there are nonetheless some important features in common, and one of the key ones is that nomads do not usually either move at random, or that far – long-distance treks being punishing for both humans and animals. Eurasian groups observed at first hand in the twentieth century, for instance, tended to move a limited distance twice a year between designated blocks of summer and winter grazing. In the case of the Khazaks, before Stalin sedentarized them, this distance was about seventy-five kilometres. Stock-raising subgroups then slowly cycled their herds around within the pasture blocks, keeping their distance from one another so that the grass had time to grow after each subgroup’s visit. Other parts of the population, in the meantime, occupied fixed camps and some even grew crops. The purpose of the longer-distance moves in this regime is to connect two blocks of grazing land, neither of which could provide year-round support. Summer pasture, typically, might be up in the hills where it was too cold for grass in winter; winter pasture in reasonably adjacent lowlands where heat and the lack of rainfall limited grazing in the summer months. Essentially, nomadism builds two landscapes into a complete grazing portfolio. In this set-up, movement fulfils a designated function and could never just be random. A nomadic existence is potentially fragile anyway, highly dependent upon rainfall in what are by definition marginal landscapes; but setting off into the wild blue yonder without knowledge of a potential destination’s carrying capacity or, equally important, established rights to graze there, would have been to invite economic disaster.7
What this means, of course, is that the intrusions of Huns into the Alanic-dominated world north-east of the Black Sea, and then subsequently into the heart of Europe, cannot be viewed – as J. B. Bury did, for instance, in a famous set of lectures given in the 1920s – as a natural extension of their nomad economy. The Huns did not just meander around the Great Eurasian Steppe until they happened to come across its western edge north of the Black Sea and take a liking to it. The decisions to switch their centres of operation westwards – in two distinct stages separated by about a generation – must have been taken for specific reasons, and carefully calculated. The potential gains of these moves had always to be balanced against the dangers of failing to find, or – more likely – establish, rights over sufficient grazing for their flocks at the new destinations.8
As to what reason or reasons led the Huns to move westwards, no easy answers are available. Roman sources are of little use. Ammianus’ view that attacking other barbarians was just something that came naturally to Hunnic megabarbarians does not get us very far. The available evidence does suggest three factors, however, two possible and the other more certain, that made it generally likely that Hunnic groups would want to move west. One of the possible factors is climate change. Around the year 400 AD, western Europe was basking in a climatic optimum, with long hot summers and plenty of sunshine. But what was good for western Europeans was less good for the world beyond the Don, where the same climatic optimum meant that there was less summer rainfall to make the grass grow. Given these conditions, it would be only natural to expect greater competition for grazing among steppe nomads, and the modern world provides us with a nasty parallel for what can happen. At the heart of the Darfur conflict are Sudanese nomad populations driven out of their old homelands as global warming turns pasture into desert. The trouble with applying this argument to the fourth century, however, is that, for the moment at least, it is impossible to know how severe or, indeed, limited the effects of fourth-century climate change actually were. There are no precise data. And in their absence, the chances are that any effects were fairly marginal. But as we shall see in subsequent chapters, a sequence of nomadic groups exploded out of this same steppe in the mid-to-late first millennium, and more were to follow, which strongly suggests that Eurasian nomadism was not facing any fundamental ecological challenge. And in any case, like the Tervingi and Greuthungi when faced with the Hunnic menace, Huns under ecological pressure could have moved in any of several directions, and adducing climate change would still leave us having to explain why they moved westwards.
The other possible factor is political revolution. At least two of the nomadic groups that followed the Huns out of the steppe into Europe in the later first millennium did so, in part, because they were under political and military pressure from other nomadic groups to their east. The sixth-century Avars were on the run from the Empire of the Western Turks, while the ninth-century Magyars moved from north of the Black Sea to the Great Hungarian Plain because of the attacks of Petchenegs. In the absence of specific information about the western steppe in the fourth century, it would be foolish to rule out the possibility that the Huns too were facing this kind of pressure.9
But even if we allow the Huns a negative element to their motivation deriving from a combination of potential climatic and political factors, there is no doubting that this coexisted, as has proved to be the case in so many flows of migration, with some very positive reasons for moving west. Roman sources describing the Huns’ initial impact on the outer fringes of the Empire offer no substantial explanation of what was going on, but later materials are highly suggestive. From c.390 and particularly the 420s onwards, we find Huns engaged in a variety of activities in relation to the Roman world. Sometimes they raided it. A huge raiding party targeting both the east Roman and Persian Empires passed through the Caucasus in 395, before the Hunnic main body had moved on to central Europe, and there are indications of other smaller raids in this era besides. Sometimes Huns served the Empire as mercenaries. As early as the 380s, the activities of a body of Huns and Alans led to diplomatic confrontation between the western Emperor Valentinian II and the usurper Maximus. In the 400s, likewise, Uldin provided military support for Stilicho, before his ill-advised incursion into east Roman Dacia. With the arrival of Huns in large numbers in central Europe from c.410 onwards, however, mercenary service reached its apogee. They were possibly already providing major military support to the de facto ruler of the western Empire, Flavius Constantius, in the 410s, but it was in the time of Aetius, from the 420s, that they became a crucial bulwark of the western Empire. Not only did Aetius use their support to keep himself in power against Roman rivals, but they were also deployed to keep in check the aggressive ambitions of the other barbarian groups now well established on western imperial territory: most notably in major campaigns against the Visigoths and the Burgundians in the 430s. Then, finally, as Hunnic power grew in the time of Attila, the Huns turned from raiding and mercenary service to large-scale invasion. Two massive attacks on the east Roman Balkans, in 442 and 447, were followed by invasions of Gaul and Italy in 451 and 452.10
What all of these activities had in common was that they were different methods of tapping into the greater wealth available within the more developed economy of the Mediterranean-based Roman world. Raiding, obviously enough, was all about movable shiny stuff and other forms of negotiable booty, and this too was the point of mercenary service. For all his Hunnic connections – and Aetius had spent three years among them as a hostage – they did not fight for him without receiving generous payment. And even Attila’s invasions were undertaken with cash in mind. We have very detailed accounts of the diplomatic contacts that preceded and followed these attacks, and Attila’s central concern was always the size of the diplomatic subsidy he could secure. Extra territory and other types of gain were of only marginal interest.11 If it is legitimate to import this vision of the Huns’ basic attitude towards the Roman Mediterranean back to the 370s, and there is no obvious reason why not, then the Huns’ decisions to move westwards in two stages make complete sense. Increased proximity to the political centres of the Roman world in no
rthern Italy and Constantinople meant greater opportunities for extracting a share of Roman wealth. In other words, the Huns were acting like the Goths and the other largely Germanic-speaking predators of the third century AD: their migrations were a response to fundamental inequalities of wealth. Like the Goths, they were moving from the less developed outer periphery of the Empire, and perhaps from beyond even that, into richer inner zones where there was a wide variety of wealth-generating opportunities available to groups able, like themselves, to deploy military force of sufficient potency.
It is also possible to say something useful about the developing nature of the Hunnic migration flow. No source gives us figures for the size of Hunnic migration units, but all the contemporary evidence indicates that the initial expansion into the northern Pontus was carried forward essentially by warbands: small groups of all-male warriors. Vithimer, the king of the Greuthungi whose death sparked the move of the Goths to the Danube in 376, fought many skirmishes – multas clades – against the Alans whom the Huns had displaced into his realm. This strongly implies that, while hugely destabilizing in aggregate, no individual engagement at this point was that large. Ammianus also records that Vithimer was able to hire some Huns to help him fight off the Alans. This has sometimes been discounted as a copying error, but there is no good reason to believe so. The report fits into a context where multiple small-scale warbands were operating on a more or less individual basis. The fact that Vithimer’s predecessor Ermenaric was able to resist the Huns ‘for a long time’ (diu) also suggests a sequence of smaller engagements rather than a set-piece confrontation. In similar vein, we find Huns operating in a variety of places and employing a variety of strategies for self-advancement as Rome’s eastern European frontiers collapsed.
Aside from the Huns who fought for Vithimer, others are recorded raiding the lands of the Tervingi (twice), signing up as mercenaries with some Alans to fight with the Tervingi and Greuthungi against Rome south of the Danube in 377, and raiding the Empire off their own bat from north of the Danube with Carpo-Dacians in tow in the early 380s.12 There is every reason to suppose that these were all independent groups of Huns, not the same one popping up in different places, and none of the recorded action requires military forces of any great size. One of the specific things Ammianus says about the Huns of this era in his digression, in fact, is that they were not governed by kings but by ‘improvised leaders’. This is a slightly slippery phrase whose meaning has been much debated, but again it fits well with a picture of small independent Hunnic units. It is also striking that this era threw up no Hunnic leaders who were individually significant enough to be mentioned by name.13 This recalls the first small-scale phases of Slavic and Viking raiding in, respectively, the sixth and ninth centuries. In both of these cases, it was only as raiding groups increased in scale that individual leaders came to be named.
But if the Hunnic expansion behind the crisis of 376–80 was being powered by warbands, the collapse of Rome’s central European frontier a generation later saw migration on a much larger scale. A hint that the size of Hunnic groups operating on the fringes of the Empire was growing is already there in the sources before this second crisis. Around the year 400, contemporary Roman sources finally mention a Hunnic leader by name: Uldin. He was powerful enough to provide useful military assistance to the Empire on occasion, with a following composed of Huns and Sciri. But although given to the occasional boast that his power stretched from where the sun rose to where it set, events put him firmly in perspective. His attempt to seize east Roman territory was defused without military action when his leading followers abandoned him, and at that point he disappears from our sources to where the sun of history doesn’t shine. This is not the career profile of a genuine predecessor of Attila. To my mind, Uldin’s sudden and otherwise inexplicable switch from ally to invader strongly suggests that his power base was not strong enough for him to hold his own in the face of the new Hunnic groups who became dominant there from c.410 onwards, almost certainly because these newcomers were turning up in larger and more organized bodies.14
The evidence for this is straightforward. When the east Roman diplomat and historian Olympiodorus visited the newly arrived Huns in the Middle Danube region in 411/12, he found them ruled by multiple kings ranged in order of precedence. At the time of the visit, the Huns had been in central Europe for only a handful of years, with no time for such a complex political order to emerge from a mass of independent warbands, and, in fact, a similar system is documented among another group of fifth-century steppe nomads, the Akatziri. It is overwhelmingly likely, therefore, that the second stage of Hunnic migration westward was actually led by the kings that Olympiodorus encountered. Indeed, given the numbers of Germani and others that the Huns displaced from the Middle Danube in the process – many tens of thousands, as we have seen – it is doubtful that a series of independent warbands could have mustered enough force to take over this new landscape. The kings’ presence makes it apparent that the move from north-east of the Black Sea to the Great Hungarian Plain had been accomplished by much larger and more organized social units than the warband activity that underlay the earlier crisis of 376–80.15
Overall, therefore, the evidence suggests that Hunnic migration into Europe took a form we have encountered before, in the third century, and will encounter again in the ninth. The initial impulse came from warbands on the make, without their having had, at this early date, any necessary intention to migrate. But when the warband activity proved highly profitable, larger and more organized groups became involved, probably aiming to maximize the amount of wealth that could be extracted by actually seizing total control of the landscape. In this case, the Huns’ later actions suggest that the attraction was not the land of the Middle Danube in terms of its agricultural potential (the attraction of England, eventually, for ninth-century Danes or eleventh-century Normans), but the fact that it was conveniently placed for maximizing profits via closer ties of various kinds with the Roman world. As a result, small-scale raiding north of the Black Sea elided into a population flow of steadily increasing momentum, until large-scale group migration emerged as the logical mechanism for maximizing profits by seizing control of the Great Hungarian Plain.
The exact size of the Hunnic groups involved in these two main phases of migration is unknowable. The kind of Gothic political unit whose stability was undermined by the aggregate action of Hunnic warbands and displaced Alans in the first phase of c.370 AD could field perhaps ten thousand warriors in total. But it is hard to extrapolate from this to the size of any attacking Hunnic force, and for two reasons. First, the Hunnic assault was indirect. Political stability north of the Black Sea was undermined over a long period by multiple raids and small-scale attacks, not head-on confrontation, and in the end it was Hun-generated upheavals among the Alans, rather than the Huns themselves, that led the Gothic Greuthungi to take their momentous decision to move in 375/6. So we don’t have to be thinking of anywhere near enough Huns to defeat ten thousand Goths in a set-piece battle. Second, like the nineteenth-century Boers, the Huns enjoyed a telling advantage in military hardware. One of their characteristic weapons was the composite reflex bow, long known on the steppe. Now, however, they employed a longer bow – up to 150 centimetres rather than the usual 100 – than had previously been seen on the western steppe. This gave them longer-range hitting power whose effects are visible in the rhetoric of Roman sources. These report Huns able to devastate the ranks of their Gothic opponents while themselves staying safely out of range. The Huns’ other characteristic weapon was a long cavalry sabre, which could do an excellent job of mopping up at closer quarters once the opposition ranks had broken.16 But exactly how big an advantage the bow gave them is uncertain. Flintlock rifles allowed the Voortrekkers to operate highly effectively against odds of about 10 to 1. Commandos numbering in their hundreds could rout Zulu and Matabele forces in their thousands at almost no cost to themselves. With this much advantage, an entire Gothic cl
ient state could have been defeated by groupings of no more than about a thousand Huns. But even the Huns’ longer bow was probably not as big an advantage as a rifle.
There is no direct evidence, either, for the size of the larger forces that Hunnic kings led on to the Great Hungarian Plain. To judge by Mongol analogies, each Hunnic warrior required many ponies to remain fully mobile. This perhaps provides an indirect indication of the total possible size of Hunnic forces, since it has been calculated that the Great Hungarian Plain could provide grazing for no more than about a hundred and fifty thousand horses. Extrapolating backwards, this number of horse could serve somewhere between fifteen and thirty thousand Hunnic warriors, which is perfectly plausible, but obviously no more than a guess. Lacking better information, I would suspect that the expansionary raiding of c.370, which did not take on the full might of the Gothic client states directly, of course, was undertaken by war parties of a few hundred, and the large-scale group move into central Europe of the early fifth century by a force somewhere in the region, again, of ten to twenty thousand warriors. But this too is only a best guess, and others could legitimately produce very different estimates.
Empires and Barbarians Page 29