Empires and Barbarians
Page 25
But there is no evidence, in fact, that Constantinople encouraged Radagaisus’ attack on Italy, and divisions between east and west Rome help explain only the subsequent course of the crisis, specifically why no eastern assistance was forthcoming until 409, not why the barbarians crossed the frontier in the first place. Nor do the changing perceptions of the barbarians provide sufficient explanation. The Vandals, Alans and Sueves still crossed the Rhine on 31 December 406, despite the disasters that had befallen Radagaisus’ force the previous summer. It took a while, but Stilicho had eventually put together a Roman army large enough to confront Radagaisus, and the result was a total Roman victory. As we saw, Radagaisus himself was captured and executed, large numbers (reportedly twelve thousand) of the higher-status warriors were recruited as auxiliaries into the Roman army, and so many of their lesser and less fortunate peers were sold into slavery that the bottom fell out of the slave market.47 Quite clearly, then, no deal analogous to that offered the Goths in 382 was on the table in the Roman west in the first decade of the fifth century. The fact that the Vandals, Alans and Sueves decided nonetheless to cross the Rhine suggests that some other factor was also at play in their thinking.
Whatever else it was, I’m pretty confident that Halsall’s proposed Roman withdrawal from the north-west does not provide the answer. For one thing, the evidence that there really was such an evacuation is not compelling, being largely an argument from silence. Many commentators date the transfer of the Gallic prefecture to Arles after 405, seeing it not as cause but as consequence of the Rhine invasion.48 Furthermore, there were enough Roman troops left in the north-west for yet another western usurper, Constantine III, to launch a putsch which took him from Britain in early 406 to the Alps and the brink of total rule of the west in 409. It was also the wrong barbarians who invaded, if interrupted diplomatic subsidies really had anything much to do with it (and we don’t actually know that the subsidies were interrupted: this too is an argument from silence). Roman diplomatic payments, as we know, went above all to the major barbarian groupings right on the frontier: namely, working our way round the frontiers of the western Empire – Franks, Alamanni, Marcomanni, Quadi and Sarmatians. The invasions of 405–8 did not for the most part draw on these frontier barbarians. The Sueves of the Rhine coalition probably fell into this category – if they really were Marcomanni and Quadi by another name – but all the others were either from the east, far beyond the western Empire’s diplomatic network (Radagaisus’ Goths and the Alans of the Rhine coalition), or from the regions behind the main frontier clientele (Burgundians and both groups of Vandals). Interrupted subsidy payments should have affected Franks and Alamanni most of all, but these groups conspicuously stayed put.49
This argument could be taken further, but there is yet another decisive problem in supposing a withdrawal of Roman power from the north-west to have triggered the frontier collapse of 405–8. The first of the invasions, the attack of Radagaisus (405/6), didn’t actually affect the north-west. It powered its way across the Alps into northern Italy, where it is not possible to argue there had been any reduction of central imperial power. In fact, any troop withdrawals from the northwest would only have strengthened imperial military capacity in Italy. If a reduction in Roman power in the north-west was the prime cause of the invasions of 405–8, why did the first invasion go in a different direction?
More revealing, in my view, is a closer look at the identity of the barbarians caught up in the crisis. The available sources are not good enough to allow us to reconstruct a detailed situation map for the fourth-century Middle Danube, but we can sketch in the basic outlines: Marcomanni and Quadi north and west of the Danube bend, Sarmatians from different groups (Limigantes and Argaragantes) either side of the River Tisza. Further north were to be found Vandals and other Germanic groups, but they did not impinge directly on the frontier action in the fourth century.50 When this distribution is compared with the invaders who emerged from the region after 405, it becomes clear that the Middle Danube had already seen a huge political-cum-demographic convulsion before the outpourings across Rome’s central European frontiers.
Vandals first appeared on Stilicho’s radar a few years before 405–8, in the winter of 401/2, when their presence nearby posed something of a threat to the peace of Raetia, more or less Roman Switzerland. This neighbourhood had emphatically not been their home in the mid-fourth century, when they were to be found the best part of six hundred kilometres further north-west, in the northern Tisza region and Slovakia, right out on the fringes of the Middle Danubian plain and old Roman Dacia.51 Their initial relocation to the fringes of Raetia, while nothing compared with subsequent marches to Spain and North Africa, was nonetheless a substantial move in itself.
That Radagaisus’ coalition, which certainly included some Goths, should have invaded Italy from west of the Carpathians reinforces the point. One or several of the many Gothic groups known from the fourth century were presumably drawn upon to make up the Gothic contingent in Radagaisus’ following. But no Goths inhabited land west of the Carpathians at that time. Likewise, the Alans: historical sources are entirely unambiguous that when they crossed the Rhine, they were the largest single component of the mixed invasion force. In other words, many Alans had come to occupy territory west of the Carpathians by about 405. But again, no Alans inhabited this region in the fourth century. Up to c.370, their westernmost stamping grounds were located around fifteen hundred kilometres further east, on the far side of the River Don.52 Different Alanic subgroups (their political structure seems to have encompassed many largely autonomous units) had begun to move west on the tails of the retreating Tervingi and Greuthungi from the mid-370s. One group of Alans, in alliance with some Huns, joined the Goths in the Roman Balkans in the autumn of 377 and even fought at Hadrianople. More Alans were encountered by the Emperor Gratian in the north-west Balkans in the summer of 378, who incorporated the same or yet more Alans into the western field army in 380.53 Things then quietened down, at least in our sources, but Alans on the move to the west were a major part of the first frontier crisis in the years after 376, and some continuation of this phenomenon is necessary to explain why there were so many Alans west of the Carpathians by 406. The observation is only reinforced by the fact that Uldin’s mixed power base, which also crossed into Dacia from somewhere on the fringes of the Middle Danube, consisted of Huns and Sciri.54 Neither of these groups shows up in the fourth century, even on the eastern fringes of the Middle Danube. The Burgundians and the Sueves, if the latter were indeed Marcomanni and Quadi, were hugely in the minority, therefore, in becoming involved in the crisis of 405–8 as long-established inhabitants of the Middle Danube and its environs.
Such a degree of population displacement was entirely abnormal in the hinterland of Rome’s frontiers. Group movements in the frontier region were usually controlled by the Romans extremely tightly. As we saw in Chapter 3, when members of just one Sarmatian subgroup, the Limigantes, returned in 359 to the sector of the Middle Danube frontier from which they had been expelled the previous year, Constantius II reacted decisively because of the propensity for disturbances beyond the frontier to spill over on to Roman territory.55 The arrival of so many newcomers in the Middle Danubian region immediately before the crisis of 405–8 completely dwarfs the amount of disruption faced by Constantius fifty years previously. Two substantial groups of Vandals, very large numbers of Alans, at least the Gothic element of Radagaisus’ coalition, and the Huns and Sciri of Uldin were all newcomers to the Middle Danube. So the frontier penetrations faced by the western Empire in 405–8 were the product of an equally large, if not actually bigger, crisis beyond the frontier itself. Something profound must have been going on there to cause all these groups to relocate themselves west of the Carpathians, even before they made their better-documented moves on to Roman soil.
So what was it? None of the factors relating to developments internal to the Roman Empire satisfactorily account for this major concentratio
n of armed groups and their dependants in the Middle Danube region before 405–8, though they certainly help explain what happened next – why the west received no eastern help before 409/10, and why attacking through Gaul proved a better option than invading Italy. In 1995, I argued that it was the second stage of Hunnic movement into Europe that had prompted this gathering of the clans west of the Carpathians, and to my mind this still provides much the likeliest explanation. Not only does the chronological correlation between their advance to the heart of Europe and the departure of our invaders from the Middle Danube plain suggest it, but, as we will explore in the next chapter, a close look at the migratory patterns of the Huns themselves provides two strong planks of further support. First, the Huns had pressing reasons of their own to want to move into central Europe, making it highly unlikely that they were merely exploiting a power vacuum that had already been created there by the departure of the Vandals and others. Second, the Huns’ treatment of neighbouring populations who got in their way made it reasonable for those neighbours to want to escape. It is thus entirely comprehensible that a second westward shift in the centre of Hunnic operations from the Black Sea to the Middle Danube, which clearly did occur in the early fifth century, should have had the effect we observe in the run-up to 405–8: causing potential new subjects to move out of its way. Not only is this the simplest explanation for the build-up of immigrants west of the Carpathians, it is also the most cogent and compelling. The proposed alternatives utterly fail to explain what the bulk of the invaders of 405–8 were doing west of the Carpathians in the first place.
Given such a strong likelihood that this crisis was a rerun of that of 376, only this time west rather than east of the Carpathians, we should not be surprised that the sources suggest some similar observations about the detailed operation of the migration processes involved in the later case. Many of this second wave of migrants, like the fourth-century Goths before them, had an established history of relocation. The one exception, it would seem, were the Sueves (assuming, again, that this term does designate various subgroups of the Marcomanni and Quadi), who had not moved anywhere before participating in the Rhine crossing. The Alans, on the other hand, were originally nomads – but this needs a bit more comment. Nomads, contrary to the received images of random movement over vast distances, typically make relatively restricted and cyclical moves between well-established blocks of summer and winter pasture. This is an entirely different phenomenon from the geographical dislocation witnessed in the late fourth and early fifth centuries, when families and flocks were moved hundreds of kilometres from long-established haunts. As with the late second- and third-century Goths, an inherent capacity for movement, engendered by the less rigid attachment of their agricultural economy to any particular territory than we are generally used to in the modern world, will also have been a factor in making this relocation possible. And in any case, by the time the various Alanic subgroups involved in the Rhine crossing had reached the Middle Danube, the jumping-off point for the events of 406, they had recently made one long trek from east of the River Don, so that a properly migratory – rather than merely nomadic – habit had already gathered momentum amongst them.56
The same is also true of Radagaisus’ Goths. They will have shared some of the past experiences of the Tervingi, being another of those concentrations of Germanic-dominated military power generated by the third-century migrations to the Black Sea. Since Goths are not found west of the Carpathians in the fourth century, the Gothic followers of Radagaisus must have made at least one move in the recent past from the Pontus to the Middle Danube, on the eve of what was to prove their ill-fated journey to Italy. The Vandals had not moved as far as the Goths in the third century, but did extend their control, from the time of the Marcomannic War onwards, south from northern and central Poland to parts of former Roman Dacia in upland Transylvania. They, again, must also have made an initial move west from this region to the fringes of the Alps, where their presence was noted in 402. In large measure, therefore, round two of the Völkerwanderung encompassed population groups with firmly entrenched migration habits, who were more likely to respond to major threats and opportunities by moving again.
The range of motivations in play among these later migrants, likewise, was probably similar to those of the Goths of 376. What we cannot reconstruct, since the date of the Huns’ entry en masse into the Middle Danubian region is uncertain, is how immediate a threat they faced. Whether they needed to leave their old abodes in more of a hurry than had the Tervingi in 376 is unclear, but this doesn’t change the fact that their motives for moving were substantially political and negative. They too were looking for new and safer homes. The influx of substantial numbers of Goths, Alans and Vandals on to the Middle Danubian plain would have been enough in itself, of course, quite apart from any Hunnic pressure, to generate political problems within the region itself. If the return of one relatively small Sarmatian subgroup to the frontier area was enough to destabilize the situation in 359, a mass influx of outsiders can only have caused political chaos.
But, as was also the case in 376, this does not deny that the immigrants also had their eye on the potential economic and other gains that might come their way from a well-organized relocation on to Roman soil. If increased pressure from the Huns made it imperative to move somewhere, then, as in 376, perceptions of its likely advantages turned the migrants’ thoughts towards the Empire. Two further observations are also worth making. First, finding a new home outside the Empire would not have been easy. The smaller concentration of Tervingi who retreated from the Danube in 376 rather than pursue their Roman visa applications further, for instance, relocated themselves in upland Transylvania or its western fringes. But to secure this new territory, they had to expel some Sarmatians already in residence. These latter, in turn, spilled on to Roman soil.57 Similarly, while en route for the Rhine in 406, the Vandals, as we have seen, had a bruising encounter with some Franks, in which they are said to have lost the unbelievable figure of twenty thousand dead – which we can reasonably take as representing a genuine memory of a hard fight. Germania was not full of fertile land ready for the taking, and given that you would have to fight for a new home wherever you went, at least Roman territory had the attraction of greater economic development. And, like the Tervingi in 376, most of the second wave of migrants had enough knowledge of the Empire to be well aware of these potential advantages. An active field of information, in other words, may have turned the discussions of our later migrants towards an imperial option, just as, even in 376, hopes of economic predation were operating alongside the Goths’ genuine fear of the Huns. Second, there is every reason, as we have seen, to suppose that the survival of the Goths of 376 as a semi-autonomous unit on Roman soil provided a further incentive for the displaced groups of the early fifth century to try the Roman option.
None of these immigrants should have been in any doubt, though, that their ambitions for a place in the Roman sun would meet with heavy resistance. If doubt there was, the fate of Radagaisus’ force must have quickly defused it. Given that their migration was an attempt to force the Roman Empire into making concessions in their direction, then each group needed to field a powerful military force. This meant, again as in 376, that freemen (or their Alanic equivalents)58 had to be recruited. For the same reasons as in 376 (and in the later stages of the third-century Germanic and the ninth-century Viking expansions), the only possible migration unit was the large grouping of ten thousand-plus warriors, many accompanied by their families. The immigrants’ clear perception of the dangers of their enterprise is also visible in some of the alliances they put together for the purpose. The Sciri footsoldiers sold as slaves and distributed as coloni (farmers) in the aftermath of Uldin’s defeat probably had no choice, and the sources on Radagaisus’ following are not good enough to make comment worthwhile.59 But the massive alliance of Vandals, Alans and Sueves was an entirely new combination of groups that had not even been near-neighbours in the
fourth century. At this point it was clearly still a loose alliance, but even this much cooperation must have taken a great deal of brokering. And not everyone seems to have been persuaded that it was the right move. It has been suggested that enough Siling Vandals stayed put to give their name to modern Silesia, and, more convincingly, that Sueves in large numbers still inhabited the Middle Danube region long after the migrant Suevi of 406 had reach northwest Spain.
So determined and so thoroughgoing was the Empire’s resistance to these new migrants that some of them altered their initial strategies. Uldin’s force was picked apart by diplomacy, when the east Roman negotiators managed to win over some of his key supporters without a fight. These were offered attractive positions in the Roman military, one presumes, while many of the less fortunate Sciri were consigned to servitude on Roman landed estates. The fate of Radagaisus’ force was similar. Again, some of his higher-status supporters abandoned ship, doing a deal whereby they were drafted into the Roman army. This time, however, the scale was different. The twelve thousand ‘of the best’ of Radagaisus’ warriors who were drafted into Stilicho’s army may have had it in mind from the beginning that being part of a larger migrating group might be a useful means of eventually cutting their own deals with the Roman authorities. But, just as likely, side-swapping was a stratagem employed only when the brute reality of overarching Roman military power became clear, as Stilicho and his field army approached.60