Internal limitations must be given their due weight, then, but anyone who argues that they played a primary role in the Empire’s collapse and that the barbarians were no more than an irritant hurrying the process along, has to explain exactly how the imperial edifice could have crumbled in the absence of massive military assault from outside. And this, it seems to me, is very hard to do. It is not that the late Empire had a perfect political system. It encompassed many centrifugal tendencies, even before the advent of the barbarians, and some outlying areas were a lot less integrated into its structures than were the Mediterranean heartlands. Britain, in particular, showed a marked tendency to throw up dissident political movements, and to judge by the amount of banditry recorded there, north-western Gaul (Armorica) may have been similar. What happened with these revolts is instructive. First, they flared up only when there was instability at the centre; and all the Empire had to do was to send out a modest expeditionary force – in the case of Britain – to bring the province back into the fold. In 368, Count Theodosius, father of the first emperor by that name, managed the task with only four regiments.21 For the Empire to have fallen apart on its own, therefore, a critical number of localities would have had to rebel simultaneously, each carrying with it enough of the Roman army to make it impossible for the centre to reconquer the rebels piecemeal.
Such a sequence of events, analogous to those which broke up the western half of the Carolingian world in the ninth century, is impossible to imagine in the fourth, precisely because the Roman Empire differed in some fundamental ways from the Carolingian. In the Carolingian Empire, the army consisted of local landowners leading armed contingents of their own retainers, whereas the Roman Empire operated with a professional army. When localities broke away from the Carolingian Empire, they already had their own ready-made armies. Roman landowners, by contrast, were civilian, and had to struggle to put together enough of a force in their locality to defend themselves from predation from the centre. Not only Britain, therefore, but northern Gaul, Spain and North Africa would have had to break away simultaneously to make internal collapse conceivable, and there is no sign of centrifugal pressure within the late Empire on anything like this scale. To my mind, rather than talking of internal Roman ‘weaknesses’ predestining the late imperial system to collapse, it makes more sense to talk of ‘limitations’ – military, economic and political – which made it impossible for the west to deal with the particular crisis it faced in the fifth century. These internal limitations were a necessary factor in, but not by themselves sufficient cause for, imperial collapse. Without the barbarians, there is not the slightest evidence that the western Empire would have ceased to exist in the fifth century.
Exogenous Shock
IN BRINGING this study of the destruction of the Roman west to a close, there is one final line of thought that I would like to explore. The exogenous shock, referred to earlier, had two components: the Huns who generated it, and the largely Germanic groups who caught its momentum and whose invasions fatally holed the west Roman ship of state. As far as we can tell, there is no deep-seated reason why the Huns should have moved into the lands north of the Black Sea at the precise moment they did. In the ancient and medieval periods, the Great Eurasian Steppe threw up from time to time militarily significant pulses of population. Sometimes these pressed eastwards towards China, sometimes westwards into Europe. The dynamics of this movement are still insufficiently understood for us to have a clear idea of any general underlying reasons that might explain why these pulses occurred when they did, or whether each had its own entirely individual explanation. In the case of the Huns, we can do no more than outline a few possibilities. These range from the environmental (the steppes becoming drier, and so less able to support livestock); to sociopolitical change; to military contingency (their having a more powerful bow). But as things stand, we have no more idea why the Huns moved west in the later fourth century than why the Sarmatians did the same around the birth of Christ.22
The Huns themselves, though, were only part of the problem. The more immediate and damaging component of the Hunnic crisis was the largely Germanic groups who forced their way across the imperial frontier in the two major waves of 376–80 and 405–8. If we can’t get any further with the Huns, the interaction between steppe nomad and Germanic agriculturalist merits more attention, because its effects were, from a broader historical perspective, unique. In the first century AD, Sarmatian nomads similarly assaulted some Germanic-dominated agricultural societies at the eastern end of the Carpathians, and some of these Sarmatians eventually moved, like the Huns would, on to the Great Hungarian Plain. Despite these similarities, however, the arrival of the Sarmatians generated no knock-on effects remotely resembling the exodus of Goths, Vandals, Alans and others on to Roman soil some four hundred years later.23 Why was this?
The likely explanation for this difference lies in the transformation of the Germanic world that had occurred between the first and the fourth centuries. As we saw in Chapter 2, first-century Germania was divided into many small, competitive political units, whose overall poverty was such that the Romans didn’t think them worth conquering. At this time Germania could put together raiding parties and larger defensive alliances that might well successfully ambush a Roman army wandering in a forest, as Arminius did with Varus’ legions in AD 9. But it possessed no political structure capable of standing up to Roman might and diplomatic manipulation in extended open conflict. By the time the Huns arrived, much had changed. An economic revolution, above all in agricultural production but also in certain manufactured goods, had generated both a much larger population and new wealth. Social stratification had increased, with a dominant free class, hereditary princes and armed retinues. This social change manifested itself at the top in the form of more robust political structures. By the fourth century, subsections of the Alamanni and Goths, amongst others, functioned as client states on the fringes of the Roman world. For the most part complaisant, they could nevertheless take action, when they thought it necessary, to limit the demands that the Empire made upon them.
As Germanic groups moved on to Roman territory to escape Hunnic aggression, this long-standing process of sociopolitical amalgamation acquired new momentum. One of the most important – and much discussed in this book – but least thought about phenomena of the fifth-century narrative is that all of the major successor states to the west Roman Empire were created around the military power of new barbarian supergroups, generated on the march. The Visigoths who settled in Aquitaine in the 410s were not an ancient subdivision of the Gothic world, but a new creation. Before the arrival of the Huns on the fringes of Europe, Visigoths – and don’t let any old-style maps of the invasions convince you otherwise – did not exist. They were created by the unification of the Tervingi and Greuthungi, who had arrived at the Danube in 376, with the survivors of Radagaisus’ force who attacked Italy in 405/6. Alaric’s ambition brought the survivors of all three groups together, and created a new and much larger grouping than any previously seen in the Gothic world.24 The Vandals who conquered Carthage in 439, likewise, were a new political entity. In this case, the new unit was generated out of just one pulse of migration, the invaders who crossed the Rhine at the end of 406. These originally comprised a loose alliance of two separate groups of Vandals – Hasdings and Silings – an unknown number of Alanic groups (the largest force), and Suevi, who were probably the product of a renewed alliance among some of the Germani of the Middle Danube. Under Romano-Gothic military assault in the mid-410s, a new entity emerged; the Siling Vandals, and various Alans attached themselves to the Hasding Vandal ruling line.
At a later date, the emergence of a Frankish Gallic kingdom was made possible only by a similar realignment among the Franks. The Franks have not figured much in our narrative of Roman imperial collapse, essentially because they are an effect rather than a cause of the process. They start to feature as a significant force on Roman soil only from the mid-460s, by which time R
oman power was already on the wane in northern Gaul. That their unification was intimately linked to Roman collapse isn’t demonstrable, but it’s highly likely. In the fourth century, Roman policy towards the Franks’ southern neighbours in the Rhine frontier region, the Alamanni, was in part directed towards preventing the build-up of militarily dangerous political confederations. If the same was true of the Franks, it will have become significantly easier for political amalgamation to have occurred among the Franks as Roman power declined in the region. And certainly, the force of Frankish warriors that Clovis used after about AD 480 to bring about a united Gallic kingdom from the Garonne to the Channel, was created by the unification of at least six separate warbands. To that inherited from his father Childeric, Clovis added those of Sigibert (and his son Chloderic), Chararic, Ragnachar and Ricchar (brothers, but seemingly with their own followings) and Rignomer.25 In the same way, the Ostrogoths, who deposed Odovacar in 492 to create the last of the successor states, were also a recent creation. Theoderic the Amal, first Ostrogothic king of Italy, completed the process begun by his uncle Valamer. In the 450s Valamer united some Gothic warbands, much as Clovis did among the Franks, to create one of the successor kingdoms to the Hunnic Empire in the Middle Danubian region. At this stage, the group numbered upwards of 10,000 or so. In the 480s, Theoderic united this force with another of more or less the same size: the Thracian Goths, previously settled in the eastern Balkans. It was this united force that then conquered Italy.26
It is worth taking a closer look at the process of reorganization into larger and more cohesive units from which the successor kingdoms sprang. In all cases, unification took place amidst a cacophony of dynastic rivalry. On the one hand, the process was fired by warband leaders readily killing each other off. Clovis, in particular, seems to have enjoyed the merry crack of axe on skull, and individual feuding was certainly rampant. On the other hand, although killing each other had always been popular with Germanic warband leaders, this had never before produced such major realignments in their society. Just as important as the leaders’ individual ambitions, therefore, were the attitudes of the warriors witnessing the spectacle. Gregory of Tours’ account of Clovis’ unification of the Franks emphasizes that, with pretty much every assassination, the dead leader’s followers declared themselves ready to ally with Clovis. And they did, of course, have a real choice. This applies equally to all the other unifications. The Visigoths were created not only by Alaric’s ambition, but also by the willingness of most of the Tervingi and the Greuthungi, plus the defeated followers of Radagaisus, to attach themselves to his standard. The Vandal coalition, as we saw, came into existence when the Siling Vandals and the Alans decided to throw in their lot with the Hasdings and the Ostrogoths on the back of positive responses to the individual successes, over two generations, of Valamer and Theoderic. In some of these cases, we know of a few individuals who decided not to join the new alliances. Rather than focusing just on the leadership struggles, then, we need to think about the choices being made by the Germanic freemen, whose decisions turned the usual leadership rivalries into a process of political unification.27
We know from the available information that the Roman Empire played a critical role in this process on two levels. First, as the pre-eminent military power of the age, over the centuries it had developed tried and trusted methods for undermining the independence even of immigrants it welcomed. Faced with such power combined with the Empire’s self-image as a society superior to all others, many of the immigrants newly arrived within the Empire became immediately aware of very good reasons, whatever their past divisions, to join forces. The Tervingi and Greuthungi were already cooperating as early as the summer of 376, when Valens tried to divide and rule them by allowing only the Tervingi into the Empire. Those in Radagaisus’ following who were sold into slavery immediately after his defeat, or saw their wives and children massacred in Italian cities after the assassination of Stilicho, were quick to grasp the logic of attaching themselves to Alaric’s following. And it was after major defeats that the Siling Vandals and Alans joined the Hasding Vandals, precisely to resist more effectively the campaigns being mounted against them by Constantius. The creation of the Ostrogoths, likewise, was marked by a thrilling moment in the summer of 478 when the emperor Zeno tried to make Theoderic the Amal fight the Thracian Goths. The emperor pretended that he would lend him a substantial force to help him defeat his rivals, while actually wanting the two Gothic forces to do each other serious damage, before sending in an imperial army to mop up. In the event, despite the two groups’ leaders being at loggerheads, the rank and file refused to fight, well aware of the path to mutual destruction that Zeno had mapped out for them.28
Second, the Roman Empire operated a powerful redistributive tax machinery. This fact was exploited by Goths and others who made the Empire – more and less willingly – recognize them as allies, or picked off pieces of it in the form of revenue-generating city territories, to secure a level of income that was not available outside the Empire. For all its economic advances, the Germanic world of the fourth century remained relatively unproductive compared with the Empire. As we saw in Chapter 7, gold only appears in any abundance in Germanic burials from the time of Attila, who had exacted it in unprecedented amounts from the Roman state. For the adventurous, the Roman Empire, while being a threat to their existence, also presented an unprecedented opportunity to prosper. When it came to exacting riches by force, alien groups who could mobilize large armed forces again stood a better chance of achieving their aim. The precise admixture of fear and anticipated profit varied, but one way or another, a heady cocktail of the two fired all these migrants towards unification. There is a very real sense in which, once the Huns had pushed large numbers of them across the frontier, the Roman state became its own worst enemy. Its military power and financial sophistication both hastened the process whereby streams of incomers became coherent forces capable of carving out kingdoms from its own body politic.
This argument also, I think, can be taken one step further. If the Huns had arrived in the first century AD instead of the fourth, and had pushed Germanic groups of the kind that then existed across the Roman frontier, the result would have been very different. Because of the smaller size of their political units in the first century, too many of them would have had to be involved in too complicated a process of realignment to make the creation of large alliances at all likely. The three or four, maybe half a dozen, units that made up each fifth-century supergroup provided enough manpower to create a military force of 20–30,000 warriors – probably the minimum for long-term survival. To get that many Germanic warriors pointing in the same direction in the first century, you would have had to unite perhaps up to a dozen rival units, and the political problem involved would have been huge. This, I would argue, is why the Sarmatian movements in the first century created so much less of an impact than the Huns’ did 300 years later.
The transformations separating fourth- from first-century Germanic society are thus a crucial factor in the story of western collapse. But what caused them? Why and how did this society change so radically?
Of the internal dynamics operating within Germanic society in these centuries, the sources – all Roman, of course – give no more than a hint. Tacitus in the first century and Ammianus Marcellinus in the fourth both mention violent struggles taking place between different groups of Germani, with no Roman involvement, and there is no reason to think this exceptional. Be that as it may, to my mind the key is the relations between Germania and the Roman Empire, on many levels, some of which we have touched on. With no judgement implied on their relative merits – let’s not forget that the Roman Empire had central heating, but saw nothing amiss in feeding human beings to wild animals for the pleasure of the multitude – the Germanic world can be said to have been a relatively simple society located at the edge of a more complex one. The close geographical proximity of such disparate entities was bound to promote precisely the kind o
f changes that we have observed in the Germanic world.
The most obvious relationship, and one that has attracted plenty of attention from archaeologists, was economic, and the evidence for substantial economic exchange between Germanic societies and the Roman Empire is impressive. High-quality items of Roman manufacture became early on in the period a feature of rich burials in the far Germanic world beyond the frontier zone. Within the frontier zone, about two hundred kilometres wide, more ordinary Roman products were part and parcel of everyday life. In return, the written evidence suggests, the Roman Empire consumed large quantities of raw materials from across the frontier. At one point in the fourth century the emperor Julian used punitive diplomatic treaties to extract from various Alamannic groups wood, foodstuffs and manpower (both slaves and recruits for his army), on other occasions, such goods and services were paid for. Roman frontier garrisons had for centuries served as centres of demand for nearby German economies. The perishable exports produced by the Germanic world are not archaeologically visible, but they certainly generated enough wealth to matter. A major slave trade, for instance, operated out of Germania. As early as the first century AD, Rome’s neighbours on the Rhine were using Roman silver coins as means of exchange, and even when 300 years later relations between the Empire and the Tervingi were more distant, trading stations remained open. We know, too, that it was common for individuals from beyond the frontier to sign up with the Roman army and then return home with their retirement bonuses.29
The Fall of the Roman Empire: A New History Page 55