One of the motives was simply to plunder Roman communities en route. In 395, Alaric was a new Gothic leader and had to secure his power base. Putting his followers in the way of funds answered this need, and we have no reason to suppose that our sources are lying when they describe the Goths’ slow trot south into and around Greece as an extended booty raid.76 But that was only part of its purpose. Alaric also needed to force the Roman state into accepting revisions to the treaty in the Goths’ favour. Mostly we hear little of the substance of these negotiations, but where the sources are more detailed, as they are for Alaric’s second sojourn in Italy between 408 and 410, it emerges that the key issues were full recognition of his leadership, possibly symbolized by granting him some kind of Roman office, the degree of economic support that the Goths would receive from the Roman state, and the finding of a suitable settlement area. Underlying this was a concern to extract a truly unconditional acceptance of the Goths’ basic right to exist as a semi-independent entity on Roman soil. In 382, the Roman authorities clearly had at least one pair of fingers crossed behind their backs. When the imperial spokesman Themistius rose to justify the treaty in front of the Senate in January 383, he closed his speech by looking forward to the time when all signs of separate Gothic identity would disappear.77
All of these Gothic prerequisites for a lasting peace agreement had to be dragged unwillingly from a Roman state that, for centuries, had enjoyed sufficient military hegemony never to have to accept long-term coexistence with a barbarian power on its own soil. Winning concessions was never easy, therefore, as the better-documented episodes of diplomatic exchange again show. Repeatedly between 408 and 410 Alaric appeared on the verge of a settlement, only to see it torpedoed by imperial intransigence. He showed enormous patience, famously reducing his demands to an absolute minimum before allowing his forces to sack Rome when even these lesser demands were rejected. This time, migration had the purpose both of inflicting damage on imperial assets so as to pressure the Empire into an agreement, and of moving the Goths to the location that offered the best chance for longer-term diplomatic success. The Grecian holiday that Alaric took in 395–7 was an attempt to force the eastern Empire to negotiate, and eventually he succeeded. In 397, the ruling regime in Constantinople, headed by the eunuch Eutropius, cut him a suitable deal. But making these concessions was extremely unpopular in some elite imperial circles, and one of the issues that contributed to Eutropius’ own downfall in 399. A sequence of regimes followed that had in common the determination not to negotiate with Alaric, whose concessions were withdrawn.78 This closing-off of the east sparked Alaric’s next migratory venture: the Goths’ first intrusion into Italy, in 401/2. This used further migration as a means of pressuring the western half of the Empire into doing a deal. But Stilicho was able to fend off Alaric’s advances militarily, and the Goths, caught in limbo, returned to the Balkans with neither half of the Empire willing to negotiate.
The situation was changed only by the intrusion of outside factors. The impending collapse of his central European frontier left the western generalissimo Stilicho desperately in need of military manpower. Having already confronted a Vandal threat to Raetia in the winter of 401/2, he was aware that a highly explosive situation was building up in the Middle Danube, as Goths, Vandals, Alans and other refugees from the Huns moved west of the Carpathians. This made him turn towards Alaric’s Goths as possible allies.79 When Stilicho was eventually deposed in the summer of 408, essentially because of his failure to deal with the mixture of invasion and usurpation that from 405 had ripped the western Empire apart, Alaric had already negotiated an understanding with him, and now pushed his followers back on the road to Italy, ostensibly to collect what he was owed. More fundamentally, the current disarray in the west made it much more likely that he would be able to extract a suitable deal there than in the east.
The Goths stayed in Italy for the next three years, and got close to agreement at certain points. In the end, however, imperial intransigence starved them out, and now under Alaric’s brother-in-law and successor Athaulf they headed off to Gaul, again in search of the right combination of circumstances to force a lasting settlement. There, finally, between 416 and 418, the bare bones of a new agreement emerged. The Goths were given a prosperous area for farming and settlement in the Garonne valley of Aquitaine, much richer than any part of the Balkans but more distant from the imperial centres of power in northern Italy, and their leader received full Roman recognition. But they were given none of the gold payments or appointments to high office within the political structure of the Roman state that had featured in Alaric’s most ambitious proposals between 408 and 410. Physically and politically they had been banished to the fringes of the Roman world. The Goths agreed to fight on occasion, as before, for the Roman state, and were employed in Spain against the Rhine invaders.80
Strange as such behaviour might appear from a modern perspective, the punctuated migrations of Alaric’s followers after 395 had their own logic. There is nothing in any of this – and certainly not the final form of the 418 agreement – that requires us to see the core of his support as having been drawn other than from the 382 Goths. They had been attempting to force the Roman state, or one half of it, into a lasting agreement, and their relocations were designed to manoeuvre them into the kind of political and geographical context from which a suitable settlement could be negotiated. What we are witnessing again, in fact, is the inescapable influence exercised by Roman state structures on the Goths’ migratory process. Throughout this long period of movement, lasting for nearly twenty years, they were twisting and turning in an attempt to gain sufficient leverage to force the Empire to change centuries-old policies. In the end, it took the crisis of 405–8, and above all the Rhine invaders’ annexations in Spain, to make the west Roman authorities receptive to the Goths’ advances.
In the emergence of this agreement, one development in particular played a role of special importance. Within the Goths’ extended odyssey, stretching from the Balkans to Aquitaine, there were some lengthy periods of relative stability: during 397–401 and again during 402–7 in the Balkans, during 408–11 in Italy, and during 412–15 in southern Gaul. In total, then, maybe only about five and a half of the twenty-odd years from original revolt to settlement in the Garonne were spent in long-distance relocation. Nevertheless, this was an extraordinarily testing and stressful period, and, as you might expect, Alaric’s force did not just proceed unchanged from point of departure to final destination. With the benefit of hindsight, we know that this extended odyssey ended satisfactorily enough. But facing hard marches and food shortages – especially in Italy in 410/11 and again in Gaul in 414/15 – and the constant threat of Roman counterattack (not least in confrontations with its field armies in 395, 397, and twice in 402) the Goths did not know that the end result was all going to be worth such a huge effort.
Whereas older narratives took the existence of Alaric’s original force for granted, more recent accounts have emphasized, rightly, that its membership changed substantially between 395 and 418. The idea that it rose and fell according to his followers’ estimates of Alaric’s likelihood of success, indeed, has now become something of a commonplace. In reality, the evidence for a steadily increasing membership is much better than that for its supposed decline. A trawl through the sources throws up a handful of individuals of high status who switched their allegiance to Rome, probably accompanied by their personal military retinues, on being defeated in the ongoing struggles for political pre-eminence that periodically preoccupied the Goths. Sarus and Fravittas, whom we have already met, fall into this category, as, seemingly, does a certain Modares. These men belong to a very specific category, however, providing no evidence that the substantial membership of the group ebbed and flowed. Otherwise, the only reference we have to Alaric losing support is from a Roman spin doctor working for Stilicho, desperate to find some way of salvaging his employer’s reputation when the latter had failed to defeat the Goths in
battle in 402. His airy claim that Alaric’s followers were abandoning him in droves can carry little weight.81
Even so, the evidence for renegotiated identity is incontrovertible. To start with, the immigrants of 376 had come across the Danube in two separate groups: Tervingi and Greuthungi. This distinction disappeared, in my view by 395, in another by 408. But the date is a matter of detail. North of the Danube, the Greuthungi and Tervingi had been entirely separate political entities. Within a generation of crossing the Danube, the distinction disappeared.82 Two had become one, and further additions of manpower followed. Outside Rome in 409 Alaric received two major reinforcements. After the overthrow of Stilicho, a major body of barbarian soldiery from the Roman army of Italy, closely allied to the generalissimo, threw in their lot with Alaric when their families, quartered in various Italian cities, were massacred in a pogrom. It is overwhelmingly likely that these were in large part the men who had, just four years before, followed Radagaisus to Italy before swapping sides in the diplomatic coup that led to their former leader’s downfall and death. Alaric’s Goths were also joined outside Rome by a very large number of slaves. I suspect many of these had the same origin, given that so many of Radagaisus’ less fortunate followers had been sold into slavery in 406. But no doubt there were others, from a variety of origins, besides.83 We are a very long way here from the old billiard-ball view of Gothic migration explored in Chapter 1.
In the course of Alaric’s career, then, a new and much larger political unit was created on the hoof in the years of renewed movement after 395. Why this happened is straightforward, I think, even if the negotiations behind the process are nowhere reported for us. The former military allies of Stilicho joined Alaric simply because of Roman hostility. They had been prepared to contemplate a long-term future as Roman allies, having abandoned Radagaisus with this in mind. Stilicho had offered them an attractive deal, perhaps something like the terms Eutropius had granted Alaric in 397. But on Stilicho’s fall, inherent Roman hostility towards ‘barbarians’, manifested in the attacks on their families, led to a change of mind. The necessary preconditions for the unification of the Tervingi and Greuthungi, likewise, were created by their joint campaigns against the Roman state from 376. Again, the process was not a smooth one. After their joint victory at Hadrianople, the groups split up again in the winter of 379/80, not least because feeding the united force was proving problematic, but probably also because each had its own leadership that was not about to give way to the other – which any definitive unification would necessarily have entailed.84 But as the narrative makes clear, this new and much bigger military-political entity was primarily created to fend off Roman power, and without imperial pressure would surely never have emerged. Not only did the cracks in Roman political structures direct the precise moves made by the Goths between 395 and 418, but the pressure of Roman military power had the effect of pushing a number of originally separate immigrant groups together just to survive. There are many complementary examples of those who failed to learn the lesson, and suffered as a consequence. Isolated Gothic raiding parties were destroyed in the course of the Hadrianople campaign, while whole breakaway subgroups were subdued and resettled on more normal Roman terms.85 The only way to prosper on Roman territory was to hang together in sufficient numbers and with sufficient political cohesion to prevent the Roman state from hanging you individually.
Alaric’s Goths provide us, therefore, with an excellent example of contingent group identity in action. Most of the constituent elements of the force seem to have been Gothic, but a shared Gothic cultural identity, if this really existed in the fourth century – and it may have – was not a prerequisite for group membership. We know at least of some Huns whose membership of the new group appears to have been permanent, and the origins of the slaves who joined Alaric outside Rome is a thoroughly moot point.86 But none even of the Gothic contingents had formed part of the same political unit before their entry on to Roman soil. It was Roman military pressure that had brought the Tervingi and Greuthungi together at Hadrianople, and that made Radagaisus’ more fortunate followers conclude that their initial choice of a Roman option was a mistaken one so that their interests would be better served by attaching themselves to Alaric’s command. On the other side of the imperial frontier, Roman aggression was not so fierce nor so sustained as to cause such a large group to form, but on Roman soil all these Goths had to unite so as to survive as an independent entity. This is, in fact, a classic pattern. Outside pressure often provides the necessary catalyst for active group identities to form.
We have no information on the negotiations between the groups that preceded their unification, but given their previously separate political histories these can’t always have been easy, as the number of high-status Goths forced out of the group and into Roman service confirms. But this, of course, is why outside pressure was required to make it happen at all, and doesn’t mean that the resulting group identity, forged in the fires of war, was fundamentally weak. If it had been, the Roman state would have been able to prise it apart (as it did with the forces of Uldin and Radagaisus); but even the subsequent diplomatic setback, then famine, and the extinction of its initial leadership line were not enough to cause the new group’s unity to collapse. And in this crucial point we find a second reason why the Roman state was willing in the end to do a deal. The Gothic force assembled in Gaul in the 410s was much larger, and, thanks to continued conflict with the Roman state, more cohesive than any Gothic political unit previously seen.87 The Romans were forced to accept by 418 that a deal had to be done, therefore, not least because the force Alaric had created was now too large to be destroyed.
For all the problems of the evidence, then, the action that unfolded from the outbreak of Alaric’s revolt down to the settlement in Aquitaine in 418 is best understood as the immigrants of 376 taking to the road again in search of a better future and picking up en route reinforcements from some of the migrants of 405–8. As we have seen before when discussing Germanic society in this era, army versus people is a false dichotomy. In a world whose economic and political structures could support only restricted numbers of specialist warriors, recruiting for any enterprise that required large armed forces automatically brought freemen and their families into the picture. To have any chance of success, Alaric had to convince large numbers of Goths that it was in their interests to up sticks and move again. But, as we have also seen, the immigrants’ aims were destined to be fulfilled only when they were able to recruit from a still wider body of support. The new political identity thereby created may have drawn in part on preexisting cultural similarities among the various Gothic groups who joined the new enterprise, but cultural similarity was by no means crucial. The Vandal–Alan alliance shows that entities with a strong political identity could be built out of constituent groups with utterly different backgrounds. Much more important than cultural similarity was the hostile presence of the Roman state.
The analysis offered above satisfactorily explains, I think, all the oddities of the action, which the alternative proposition simply cannot. The sophisticated political agendas on display and, above all, Alaric’s need for a settlement area do not sit well with the mercenary-band model. Adopting it would also raise the question of where Alaric might have found such a massive reservoir of specialist warriors.
Many of these points also apply to those other great practitioners of repeat migration: the Rhine invaders of 406. You’ll be relieved to know that there’s no need to rehearse the army-versus-people argument again in relation to their history after their initial settlement in Spain in 412. This would be trying the reader’s patience, and the sources are anyway less informative. Whatever view you form of Alaric’s Goths, therefore, will tend to spill over into your understanding of the Vandals, Alans and Sueves. Suffice to say that the one detailed, broadly contemporary source with any claim to authority that we have does picture the Vandals and Alans moving on to North Africa with wives and families in tow.8
8 And, for similar reasons to those explored in the case of the Goths, there really are no good grounds for doubting it.
In other more precise respects, however, the migration processes of both Goths and Rhine invaders correspond more closely with what comparative migration studies might lead us to expect. Logistics, naturally enough, played a key role in shaping the individual moves. Alaric’s Goths hit the road with a huge wagon train. This meant that they were confined to land routes and the Roman road network, which, particularly in the Balkans, greatly restricted the choices of route and helped dictate, for instance, the Goths’ circulatory itinerary between 395 and 397. An inability to secure sea transport also nipped in the bud Alaric’s plans to ship his force to North Africa after the sack of Rome in the autumn of 410, and eventually made it possible for the Romans to blockade his people in southern Gaul and cut them off from food supplies. The Vandals and Alans also moved with wagons while on land, but fared better than Alaric’s Goths in their eventual bid to cross to North Africa. Part of the reason for this lay in the fact that they had had longer to prepare. Alaric considered moving to North Africa only when his sieges of Rome failed to bear fruit in the form of a diplomatic settlement. But he dropped the plan within just a few months, in the late summer and autumn of 410. The Vandals and Alans, by contrast, had been mounting wide-ranging campaigns right across the Iberian peninsula for over a decade before taking ship to North Africa. This gave them plenty of time to organize the necessary shipping, and, again unlike Alaric in 410, in 429 they were not facing the imminent possibility of an imperial counterattack. This meant that they could afford to move themselves across the straits of Gibraltar piecemeal, and hence needed fewer ships, since there was no danger of those left behind being attacked while waiting to be transported.
Empires and Barbarians Page 27