THE FALL OF ATTILA’S EMPIRE is an extraordinary story in its own right. Up to about AD 350, the Huns had figured not at all in European history. During 350–410, the only Huns most Romans had encountered were a few raiding parties. Ten years later, Huns in significant numbers had established themselves west of the Carpathian Mountains on the Great Hungarian Plain, but they still functioned mostly as useful allies to the Roman state. In 441, when Attila and Bleda launched their first attack across the Roman frontier, the ally revealed his new colours. In forty years, the Huns had risen from nowhere to European superpower. By anyone’s standards, this was spectacular. But the collapse of Attila’s Empire was more spectacular still. By 469, just sixteen years after his death, the last of the Huns were seeking asylum inside the eastern Roman Empire. Their extinction would cause deep reverberations in the Roman west.

Empire to Extinction

RECONSTRUCTING the collapse of Hunnic dominion in central Europe is a tricky proposition. Our old friend Priscus told the story in some detail, but since there was little diplomacy involved in the fall, his account hardly made it into Constantine VII’s Excerpts concerning Embassies (see p. 306). For the most part we have to rely on one of the most intriguing historical works to survive from late antiquity: the Gothic History, or Getica, of Jordanes, whose voice we have already heard in earlier chapters. About ten pages of text (half of it notes) in the standard edition provide the only coherent existing account of the fall of Attila’s Empire.1

Jordanes was a man of Gothic descent living in Constantinople around the year 550, so he was writing nearly a century after the events we’re interested in. At this point he was a monk, but had previously served as a secretary to a Roman commander on the Danube, so was not without relevant experience. He tells us in the preface to the Getica that his history of the Goths is largely an abridgement of a lost history written by an Italo-Roman called Cassiodorus. Cassiodorus was adviser to Theoderic the Amal, Ostrogothic king of Italy, in the 520s. Jordanes says that he had access to Cassiodorus’ history for just three days when compiling his own and that, as he puts it, ‘the words I recall not, but the sense and deeds related I think I retain entire’. Some have sensed something a bit fishy in this, arguing either that Jordanes had much greater access to his model than he pretends, or that he had very little to do with him and was trying to use Cassiodorus’ name for his own purposes. These hypotheses founder, however, on their proponents’ failure to come up with a convincing reason for Jordanes to have lied.2 I am confident that he is broadly telling the truth in claiming to have followed Cassiodorus’ outline closely. The Getica corresponds well enough with the few things we know from elsewhere about Cassiodorus’ history.3

But even if Jordanes’ preface is not disguising some massive deception, this doesn’t make the Getica a reliable source. Cassiodorus wrote his history of the Goths for the court of the Ostrogothic king, Theoderic the Amal, and this has a significant bearing on the narrative of Hunnic collapse that has come down to us in the Getica. Above all, and as you might expect, it is a thoroughly Gotho-centric account. Only the story of the Goths removing themselves from Hunnic overlordship is told in any detail in its pages, and even the Huns appear only incidentally. More specifically, Cassiodorus had to tell his Gothic history as his particular Gothic king wanted it told. As a result, it contains two historical distortions.

First, it claims that all the Goths who didn’t flee from the Huns in AD 376 by crossing over into the Roman Empire immediately fell under Hunnic control. This is nonsense. We actually know of seven groups of Goths, other than the Greuthungi and Tervingi who sought asylum from the emperor Valens in 376 (and there is no reason to suppose that even this list is exhaustive):

1.   The Amal-led Goths, who were under Hunnic control by the time of Attila and were presently ruled by Theoderic.

2.   The Goths of Radagaisus who invaded Italy in 405/6 and eventually became part of Alaric’s new Visigothic group (see Chapter 5).

3. The Goths of Pannonia, detached by Roman military action from Hunnic hegemony in the 420s, and resettled by the Romans in Thrace; quite possibly the ancestors of group 6 below.

4. The Goths of a king called Bigelis, who unsuccessfully invaded the east Roman Empire sometime between 466 and 471.

5. The Goths operating in the train of Dengizich, son of Attila, when he invaded east Roman territory in the late 460s.

6. A large group of Goths already settled in Thrace as Roman allies in about 470.

7. Two other, smaller, Gothic groups established in enclaves around the Black Sea: the Tetraxitae of the Cimmerian Bosporus and the Goths of Dory in the south-western Crimea.4

In concentrating solely upon group 1, therefore, the Getica’s historical vision substantially simplifies Gothic history.

Second – and closely related to the first point – the Getica overstates the historical importance of the Amal dynasty from which Theoderic, Cassiodorus’ employer, was descended. By dividing the Goths into those who were conquered by the Huns in 376 and those who fled, the Getica can maintain that the Amal family had long ruled every Goth who did not enter Roman territory during the reign of Valens. The Amals were later responsible for the creation of the Ostrogoths, as mentioned earlier, but this happened between about 460 and 490. Nothing suggests that the Amal dynasty had been anything like as prominent before it acquired this new power-base. Parvenu dynasts often pretend that they are not parvenus at all, and Theoderic was a case in point. Cassiodorus’ letters consistently refer to Theoderic’s family as a ‘purple dynasty’; this perspective permeated Cassiodorus’ history – hence its presence in the Getica. Furthermore, there is no reason to suppose that our list of seven groups is exhaustive: there were many Gothic ‘royal’ families competing at the heads of their individual warbands.5 In reality the fall of the Hunnic Empire was rather more messy than Jordanes makes out.

As the Getica tells it, the origins of Hunnic collapse lay in a dispute over succession between Attila’s sons soon after their father’s sudden death. At least three of the sons figure in different sources as important leaders in their own right – Dengizich, Ellac and Hernac – but we have no idea of how many there were in total, or of whether all, or only some, were potential candidates for their father’s position. The quarrel soon degenerated into civil war, which resulted in one Germanic subject group, the Gepids under their king Arderic, throwing off Hunnic domination. This presumably meant that the Gepids refused to pay any more tributes or to answer demands for military service. The rebellion was not taken lying down, the Getica tells us, and the outcome was a battle on an unidentified river in Pannonia called the Nedao:6

There an encounter took place between the various nations Attila had held under his sway. Kingdoms with their peoples were divided, and out of one body were made many members not responding to a common impulse. Being deprived of their head, they madly strove against each other . . . And so the bravest nations tore themselves to pieces . . . One might see the Goths fighting with pikes, the Gepids raging with the sword, the Rugi breaking off the spears in their own wounds, the Sueves [Suevi] fighting on foot, the Huns with bows, the Alans drawing up a battle-line of heavy-armed and the Herules of light-armed warriors. Finally, after many bitter conflicts, victory fell unexpectedly to the Gepids.

This is good breastplate-ripping stuff, but not very informative even if the outline story is plausible enough. Clearly, dynastic strife was the norm within the royal family of the Huns, once power became more centralized in the fifth century. We saw in Chapter 7 that royal refugees from previous succession struggles had ended up inside the Roman Empire in the 440s, for instance, and some were returned for execution. Jordanes is also unlikely to have given the Gepids a starring role unless it was impossible not to, especially since there was no love lost between Goths and Gepids by the sixth century.7

What’s not at all clear, though, is who was on whose side in the battle, and whether there was just one big battle or a series of smaller ones. Jordanes is also a bit vague on the outcome of all this violence. He baldly reports that ‘by his revolt [Arderic] freed not only his own tribe, but all the others who were equally oppressed’. But how precisely this liberation happened is open to question. When, in the battle (or battles), Attila’s son Ellac was killed, Jordanes reports, the others immediately abandoned their homes in the Middle Danube and made for lands east of the Carpathians and north of the Black Sea, handing out freedom to all the Huns’ subjects, no matter whose side they had fought on.8 By about the year 460, the position of the major powers in and around the Middle Danubian Plain, in so far as we can reconstruct it, was more or less as follows (map 15). The Amal-led Goths occupied an arc of territory south of the River Danube in former Roman Pannonia, stretching from Lake Balaton towards the city of Sirmium. The Gepids controlled the north-eastern stretch, including much of the old Roman province of Dacia abandoned in the third century. Between the two were the Suevi north of the Danube bend, plus the Sciri, Herules, Rugi and Sarmatians/Alans. According to a literal reading of Jordanes, thanks to the revolt of the Gepids all of these groups rapidly converted from Hunnic subjects into independent kingdoms. There are enough hints in fragments preserved elsewhere, however, and in odd details of Jordanes’ account, to make it clear that, again, this is much too simple a picture.

The idea that the Huns suddenly disappeared from the Carpathian region in 453/4, for instance, is deeply misleading. In the later 450s and early 460s, they twice intervened west of the Carpathians against the Amal-led Goths in Pannonia, as Jordanes himself tells us,9 and in the later 460s Attila’s remaining sons were still able to launch attacks into the Roman Empire across the Danube. If, as Jordanes reports, the Huns did leave the Middle Danube after the battle of the Nedao, they didn’t go far. And while Nedao may have freed the Gepids, it clearly didn’t free everyone. When the Huns, under Attila’s son Dengizich, attacked the east Roman Empire for the last time in 467/8, there were still substantial numbers of Goths in his following, Priscus reports.10 Jordanes also tells us that Dengizich had mobilized several groups – Ultzinzures, Angisciri, Bittugures and Bardores – for his second attack on the Amal-led Goths.11 This doesn’t mean that Nedao was not a significant turning-point, but it does demonstrate that Hunnic power over the other population groups of the Carpathian region wasn’t suddenly extinguished.


The path to freedom of the Amal-led Goths, and most of the Huns’ subjects, was not quite what Jordanes implies, either. No sudden moment of liberation freed everyone at the same time. As we have seen, there were at least three separate groups of Goths under Hunnic dominion at Attila’s death, and there had earlier been a fourth (group 3/6, p. 353), detached from Hunnic control by east Roman action and resettled in Thrace in the 420s. Group 1 had escaped by the later 450s, group 4 by the mid-460s, while group 5 never escaped at all, participating in the Huns’ final attack on the Empire in 467/8. We have no equivalent information for the Huns’ other subject peoples, but behind each individual group name – Suevi, Rugi, Herules, Gepids, Alans and so on – there may likewise have been several independent political units who threw off Hunnic dominion at various points between 453 and 468.

Nor should we assume that each of the separate units that emerged from the wreck of the Hunnic Empire already had its own smoothly functioning leadership at the time of Attila’s death. The Getica reports that this was true of the Amal-led Goths, claiming that Valamer the Amal, Theoderic’s uncle, had been a trusted right-hand man of Attila and that the Amal dynasty’s pre-eminence over Group 1 was beyond challenge. There are good reasons for doubting both claims. Jordanes himself reports that for forty years under Hunnic hegemony, before the appearance of Valamer, this supposedly unchallengeable dynasty hadn’t actually ruled any Goths at all. He also tells some interesting stories about a supposedly Hunnic ruler by the name of Balamber, who defeated several Gothic rulers, in particular Vinitharius and Hunimund. Many chronological inconsistencies fizzle out once it is recognized that the accounts of Balamber’s exploits probably describe how Valamer first consolidated his hold over the Amal-Goths. Balamber doesn’t appear in any other sources; and in Greek, Valamer is written ‘Balamer’. The stories tell of him defeating two rival Gothic ruling lines in the persons of Vinitharius and Hunimund, together with the latter’s son Thorismund. Gesimund, the brother of Thorismund, accepted Valamer’s overlordship rather than continuing the contest, while Thorismund’s son Beremund fled west into the Roman Empire.

Instead of an Amal dynasty with a unique, long-established prestige at the time of Attila’s death, then, we need to envisage several competing petty Gothic warlords, each with their own warbands. It was Valamer, it seems, who first united them, in some instances by direct military action (as in the killing of Hunimund); in others, as with Gesimund’s surrender, by conciliation; and in yet others, by a mixture of the two – Valamer killed Vinitharius, then married his granddaughter.12 My best guess is that all this political restructuring happened after the death of Attila. The process generated a much larger Gothic force, better able to resist Hunnic domination, and it is hard to think that Attila in his pomp would have tolerated it.13

Quite clearly, then, not all of the Huns’ subjects came in neat units, with established leaderships ready and waiting to recapture their independence as soon as the great man died. The Gepids perhaps did, and this might explain why they were able to regain their independence so quickly. But other groups that we see asserting their autonomy after Attila’s death had been generated only recently: on the hoof, as it were, around the leadership of new men. The emergence of the kingdom of the Sciri, for instance, was far from straightforward. In the 460s, they were ruled by the same Edeco whom we met in the last chapter as one of Attila’s trusted inner circle, the man the east Romans had tried to bribe into assassinating the then Hunnic leader. Edeco was supported by two sons, Odovacar and Onoulph. As the Hunnic Empire collapsed, Edeco clearly managed to reinvent himself, turning from trusted Hunnic henchman into the king of the Sciri. Interestingly, he probably wasn’t a Sciri by birth. His sons are described as having a Scirian mother, but he himself is labelled as either a Hun or a Thuringian. The latter – being more specific – is perhaps more likely to be correct. What qualified Edeco for leadership of the Sciri was not his origin, then, but a marriage alliance probably with the daughter of a Scirian bigwig, combined with his pre-eminence at Attila’s court. For the other groups we have no information; but I suspect that plenty of this kind of political reordering went on in the mid- to late 450s before the successor kingdoms to the Hunnic Empire could emerge into the light of history.14

Putting all these fragments together suggests a rather different account of the collapse of the Hunnic Empire from that given by Jordanes. If the reassertion of independence on the part of at least some of the subject peoples had to be preceded by major political readjustments, this tells us that the Hunnic Empire eased towards extinction as the Huns gradually lost control of those peoples.

The emergence of the new independent groups then set in motion the final stage in the process of Hunnic extinction. The Huns had gathered most of them together on the Great Hungarian Plain, this unprecedented concentration of armed groups creating there a hugely powerful war machine.15 In the Roman period, the area had been divided between just Sarmatians, Suevi and Vandals – Roman policy took great care to prevent overcrowding in the immediate frontier area, for fear that it would lead to violence. The removal of Hunnic domination created just the situation that these old Roman policies were designed to prevent: a concentration of competitive armed groups in a relatively small area. So battles for independence naturally evolved into a fight for regional hegemony in the 460s, as the new kingdoms took each other on in a struggle for mastery on the Danube.

Again, the only coherent narrative is to be found in the Getica, which of course presents it as a triumph for the Amal-led Goths.16 As Jordanes tells it, these quickly came to blows with the Suevi, over whom they won a great victory. The Suevi then stirred up the other regional powers against the Goths, particularly the Sciri, who managed to kill Valamer in a first bout of fighting. The Goths, however, took a ferocious revenge, destroying the Sciri as an independent power. This led most of the rest – the Suevi, the remaining Sciri, Rugi, Gepids, Sarmatians ‘and others’ – to unite against the Goths. The result was a second great battle, on a second unidentified river in Pannonia, the Bolia, where, as Jordanes tells us:

The party of the Goths was found to be so much the stronger that the plain was drenched in the blood of their fallen foes and looked like a crimson sea. Weapons and corpses, piled up like hills, covered the plain for more than ten miles. When the Goths saw this, they rejoiced with joy unspeakable, because by this great slaughter of their foes they had avenged the blood of Valamer their king.17

Other sources provide just about enough information to confirm Jordanes’ version. A fragment of Priscus’ history records that, before the showdown, the Sciri and Amal-led Goths both sent embassies to Constantinople to try to procure east Roman assistance.18The destruction of the Sciri also figures in other sources. But whether and to what extent the Amal-led Goths were always victorious, we don’t really know.

The violence and instability only began to ease off a little in the region as some of the competing groups were eliminated. The Scirian kingdom lost its independence in the late 460s, and in 473 the Amalled Goths left the area to try their luck in the east Roman Empire. None of this came soon enough, however, to save the sons of Attila. As the events of the 450s and 460s unfolded, their position was fatally undermined. Each assertion of independence meant that another subject people had stopped paying their annual tributes. This was bad enough, but then the new kingdoms started to take the initiative, looking to maximize their positions at the expense both of each other and of the Huns. The transformation from victors to victims is well illustrated in the two wars that the sons of Attila fought, according to Jordanes, against the Amal-led Goths. In the first they attacked them as ‘fugitive slaves’, with the aim of reasserting their own hegemony and tribute rights. In the second, they were seeking to prevent some of the smaller groups settled in Pannonia from falling under Gothic dominion.19 All the other major groups we hear about were doing much the same, so that the Huns’ power-base was steadily eroded.

By the mid-460s the two surviving sons, Dengizich and Hernac, were desperate. The loss of subject peoples, combined with the increasing empowerment of groups like the Amal-led Goths, left their position north of the Danube untenable. The only option open to them was to seek an accommodation with the Roman Empire. But Dengizich got it wrong – perhaps he demanded too much. In 469 he was defeated by the Roman general Anagastes, and his head publicly displayed at Constantinople. Hernac and his followers, perhaps less greedy, were eventually resettled beside the Danube in northern Dobrudja (modern Romania), and some other Hunnic remnants settled in and around the fortresses of Oescus, Utus and Almus. Independent Hunnic power north of the Danube had ended. The demise of Attila’s realm had been swift and total.

Riding the Tiger

DESPITE ITS MANY limitations, then, the Getica’s account allows us to reconstruct some of the key stages in the process of Hunnic collapse. Over the years, many explanations have been offered for this extraordinary phenomenon. Historians of earlier eras tended to argue that it was testament to the extraordinary personal capacities of Attila: the Empire could only exist with him at the helm. Edward Thompson, by contrast, rooted the Huns’ demise in the divisive social effects of all the wealth they acquired from the Roman Empire.20 There is something in both of these theories. Attila the Hun, as we have seen, was an extraordinary operator, and no doubt the gold extracted from Rome was not distributed entirely evenly among his people. But a full understanding of the Hunnic Empire must turn on its relations with its largely Germanic subjects. As already suggested, it was the ability to suck in so many of these militarized groups that underlay the sudden explosion of Hunnic power in the 420s–40s. After Attila’s death, likewise, it was his successors’ increasing inability to maintain control over those same groups that spelled their own decline.

The key starting-point is that the Hunnic Empire was not generally enrolled voluntarily. All the evidence we have suggests that non-Hunnic groups became caught up in it through a combination of conquest and intimidation. In the time of Attila, the Akatziri were the latest to fall into the Empire’s orbit. We took in the first half of the story in Chapter 7, when the east Roman ambassador gave the best gifts to the wrong king. Priscus tells us what happened next:

Kouridachus, the senior [king of the Akatziri] in office . . . called in Attila against his fellow kings. Attila without delay sent a large force, destroyed some, and forced the rest to submit. He then summoned Kouridachus to share in the prizes of victory. But he, suspecting a plot, declared that it was hard for a man to come into the sight of a god . . . In this way Kouridachus remained amongst his own folk and saved his realm, while all the rest of the Akatzirian people submitted to Attila.21

Attila then sent his eldest son to rule over the conquered. The passage reveals that while Attila was capable of deft political manoeuvring when the occasion demanded, the basic tool of Hunnic imperial expansion was military conquest. It was, of course, to avoid Hunnic domination that the Tervingi and Greuthungi had come to the Danube in the summer of 376 in the first place. And it was after a savage mauling at the hands of the Huns in the 430s that the Burgundians also ended up in the Roman Empire. All this is consistent with the fact that there was, as we have seen, one way, and one way only, of quitting Attila’s Empire: warfare.22

We don’t have all the information we might like on relations between the Hunnic conquerors and their various subjects. Pride of place has tended to be given to a story told by Priscus, often seen as illustrating the ethnic and social mobility that was possible in the Hunnic Empire. While hanging around Attila’s camp, Priscus ran into a well dressed Hun who greeted him in Greek. Upon inquiry, the ‘Hun’ turned out to be an ex-Roman prisoner, a former merchant captured at the fall of Viminacium in 441. In the share-out that followed he had been assigned to Onegesius and had fought in subsequent campaigns, against both the Romans and the Akatziri. He did well, won lots of booty, which he passed on to Onegesius, and was consequently freed. He’d then taken a Hunnic wife and was now a trusted companion of his former master, accustomed to dining with him. Thus a slave who did well in battle could win his freedom and be accepted in fairly exalted Hunnic circles. Not so commonly quoted is another story exposing the other side of master-slave relations under the Huns. Also during his stay at Attila’s court, Priscus saw the gibbeting of two slaves who had taken the opportunity offered by the turmoil of battle to kill their master. And in fact, most of the Huns’ subjects were exploited in a variety of ways and kept firmly in their place.23

A revealing fragment of Priscus’ history records an incident in 467/8 during Dengizich’s last attack on the east Roman Empire, when a mixed force of Goths and Huns was picked apart by Romans; they reminded the Gothic contingent of exactly how the Huns generally behaved towards them: ‘These men have no concern for agriculture, but, like wolves, attack and steal the Goths’ food supplies, with the result that the latter remain in the position of slaves and themselves suffer food shortages.’24 Taking the subject peoples’ supplies was, of course, only part of the story. They were also used, as we have seen, to fight the Huns’ wars. Few civilian prisoners are likely to have been very good at fighting, and casualty numbers during Hunnic campaigns were probably enormous. Priscus’ merchant-turned-Hun certainly prospered, but his was no doubt an unusual story.

Clearly, then, the Hunnic Empire was an inherently unstable political entity, riven with tensions between rulers and ruled. Tensions of a different kind also existed between the subject peoples themselves, who had a long history of mutual aggression even before the Huns appeared. This particular instability tends to receive little coverage from historians because most of our source material comes from a Roman, Priscus, and dates to the time when Attila’s power was unchallengeable. Cast the net wider, though, and the evidence rapidly gathers itself. The greatest strength of the Hunnic Empire – the ability to increase its power by quickly consuming subject peoples – was also its greatest weakness. The Romans, for instance, were happy to exploit, whenever they could, the fact that these subject peoples were not there of their own free will. In the 420s, the east Roman counteraction against the rising Hunnic power in Pannonia was to remove from their control a large number of Goths whom they then settled in Thrace.25 And an early fragment of Priscus tells us:26 ‘When Rua was king of the Huns, the Amilzuri, Itimari, Tounsoures, Boisci and other tribes who were living near to the Danube were fleeing to fight on the side of the Romans.’ This dates to the late 430s, after Rua had achieved considerable success, indicating that even success wasn’t enough to guarantee the quiescence of subject groups. The start of a new reign was a moment of particular stress. The first campaign of Rua’s successors, Attila and Bleda, when they came to power in 440, was not against the Romans: ‘When [at the start of their reign] they had made peace with the Romans, Attila, Bleda and their forces marched through Scythia subduing the tribes there and also made war on the Sorogsi.’ Reasserting your overlordship over subject groups, once you had established your supremacy, was probably the first priority for any new ruler of the Hunnic Empire.

The conflicts that arose after Attila’s death were not exceptional, then, but inherent in the relationship between the Huns and their subjects. When they could, Hunnic leaders tried to ensure that the Romans wouldn’t stir up trouble for them in this quarter. In their first treaty with the east Romans, when the latter wanted peace on the Danube so as to be able to pursue their ambitions in North Africa, Attila and Bleda were able to ensure ‘that the Romans should make no alliance with a barbarian people against the Huns when the latter were preparing for war against them’. Unlike the Roman Empire, which spent centuries dissipating the tensions of conquest turning their subjects – or, at least, the landowners among them – into full Romans, the Huns lacked the necessary stability and the bureaucratic capacity to run their subjects directly.27 Instead of revolutionizing the sociopolitical structures of the conquered peoples or imposing their own, they had to rely on an indigenous leadership to continue the daily management of the subject groups. As a result, the Huns could exert only a moderate degree of dominion and interference, and even that varied from one subject people to another. The Gepids, as we have seen, had their own overall leader at the time of Attila’s death, and so were quickly able to assert their independence. Other groups, like the Amal-led Goths, first had to produce a leader of their own before they could challenge Hunnic hegemony. Some, like the Goths in thrall to Dengizich when he invaded east Roman territory in the 460s, never managed to do so. But even these, still dominated by Dengizich in 468, had their own subchieftains.

If the sources were more numerous and more informative, I suspect that the narrative would show the Hunnic Empire peeling apart like an onion after 453, with different subject layers asserting independence at different times, in inverse relation to the degree of domination the Huns had previously exercised over their lives. The two key variables were, first, the extent to which the subjects’ political structure had been left intact; and second – I strongly suspect but cannot prove – their distance from the heartland of the Empire where Attila had his camps. Some groups, settled close to the Huns’ own territories, were kept on a very tight rein, with any propensity to unified leadership suppressed. Groups living further away preserved more of their own political structures and were less readily controlled. In the time of Attila, the Franks and the Akatziri defined the geographical limits of his marginal influence, while groups in between such as the Thuringians, Goths, Gepids, Suevi, Sciri, Herules, Sarmatians and Alans faced differing degrees of closer control.28

Archaeological evidence from Attila’s Empire offers us a further perspective on relations between its subjects and rulers. As we saw in Chapter 7, this mainly takes the form of Germanic or seemingly Germanic cemeteries; a striking feature of the excavated material is the contrast between the large number of unfurnished burials and a smaller number of rich ones. These rich burials are not just quite rich: they are staggeringly so. They contain a huge array of gold fittings and ornamentation, the stars of the collections being the cloisonné gold and garnet jewellery in which the stones are mounted in their own gold cases to give an effect not unlike mosaic. This kind of work would later become the mark of elites everywhere in the late and post-Roman periods. For instance, the style of the cloisonné jewellery found in the Sutton Hoo ship burial of the early seventh century in East Anglia originally gained its hold on elite imaginations in Hunnic Europe.29 One burial at Apahida (modern Transylvania) produced over sixty gold items, including a solid gold eagle that fitted on to its owner’s saddle. Every other piece of this individual’s horse equipment was likewise made of gold, and he himself was decked from head to foot in golden jewellery. There are other similarly wealthy burials, as well as others containing smaller numbers of gold items.30

The presence of so much gold in Germanic central and eastern Europe is highly significant. Up to the birth of Christ, social differentiation in the Germanic world manifested itself funerarily, if at all, only by the presence in certain graves of larger than usual numbers of handmade pots, or of slightly more decorative bronze and iron safety-pins. By the third and fourth centuries AD, some families were burying their dead with silver safety-pins, lots of beads, and perhaps some wheel-turned pottery; but gold was not being used to distinguish even elite burials at this point – the best they could manage was a little silver.31 The Hunnic Empire changed this, and virtually overnight. The gold-rich burials of the ‘Danubian style’ mark a sudden explosion of gold grave goods into this part of Europe. There is no doubt where the gold came from: what we’re looking at in the grave goods of fifth-century Hungary is the physical evidence of the transfer of wealth northwards from the Roman world that we read about in Priscus and the other written sources. The Huns, as we saw in the last chapter, were after gold and other moveable wealth from the Empire – whether in the form of mercenary payments, booty or, especially, annual tributes. Clearly, large amounts of gold were recycled into the jewellery and appliqués found in their graves. The fact that many of these were the rich burials of Germans indicates that the Huns did not just hang on to the gold themselves, but distributed quantities of it to the leaders of their Germanic subjects as well. These leaders, consequently, became very rich indeed.

The reasoning behind this strategy was that, if Germanic leaders could be given a stake in the successes of the Hunnic Empire, then dissent would be minimized and things would run relatively smoothly. Gifts of gold to the subject princes would help lubricate the politics of Empire and fend off thoughts of revolt. Since there are quite a few burials containing gold items, these princes must have passed on some of the gold to favoured supporters.32 The gold thus reflects the politics of Attila’s court. (It’s nice to think that the prince buried at Apahida may have been one that Priscus encountered.) Equally important, the role of such gold distributions in countering the endemic internal instability, combined with what we know of the source of that gold, underlines the role of predatory warfare in keeping afloat the leaky bark that was the Hunnic ship of state.

First and foremost, success in warfare built up the reputation of the current leader as a figure of overwhelming power. Witness the case of Attila and the sword of Mars. But there is every reason to suppose that military success had been just as important for his predecessors. A reputation for power brought with it the capacity to intimidate subject peoples, and it was also military success, of course, that provided the gold and other booty that kept their leaders in line – although the speed with which subject groups opted out of the Empire after Attila’s death suggests that the payments did not compensate for the burden of exploitation. In contrast to the Roman Empire, which, as we have seen, attempted to keep population levels low in frontier areas so as to minimize the potential for trouble, the Hunnic Empire sucked in subject peoples in huge numbers.33 The concentration of such a great body of manpower generated a magnificent war machine, which had to be used – it contained far too many inner tensions to be allowed to lie idle. The number of Hunnic subject groups outnumbered the Huns proper, probably in a ratio of several to one. It was essential to keep the subject peoples occupied, or restless elements would be looking for outlets for their energy and the Empire’s rickety structure might begin to crumble.


WE HAVE ARRIVED at a very different perspective on Attila the Hun. As is often the case, the factor that made him so powerful was at the same time his greatest liability. The military force that brushed aside the armies of the east Roman Empire in the 440s was itself highly unstable. The victories with which it provided him cemented Attila’s control in the short term, but it was riven with internal tension: further victories were essential, to maintain his dominance. Should his reputation start to crack, then his subjects would desert into the welcoming arms of the Romans. Attila was the greatest barbarian conqueror in European history, but he was riding a tiger of unparalleled ferocity. Should his grip falter, he would be mauled to death.

To my mind, this in turn explains his otherwise mysterious turn to the west at the end of the 440s. Between 441 and 447, Attila’s armies had ransacked the Balkans except for some small areas protected by two major obstacles: the Peloponnese because of its geographical isolation, and the city of Constantinople because of its stunning land defences. The eastern Empire was on its knees: the annual tribute it was having to pay out was the largest ever expended by a factor of ten. The Huns had squeezed out of Constantinople just about everything they were likely to get; at the very least, further campaigning against it was bound to run into the law of diminishing returns. But there on the Hungarian Plain Attila sat, still surrounded by a huge military machine that could not be left idle. With nothing to attack in the Balkans, another target had to be found. Attila turned to the west, in other words, because he’d exhausted the decent targets available in the east.

This suggests a final judgement on the Hunnic Empire. Politically dependent upon military victory and the flow of gold, it was bound to make war to the point of its own defeat, then be pushed by that defeat into internal crisis. The setbacks in Gaul and Italy in 451 and 452 must anyway have begun to puncture Attila’s aura of invincibility. They certainly caused some diminution in the flow of gold, and some of the outlying subject peoples may already have been getting restive. Quite likely, Attila’s death and the civil war between his sons provided just the opportunity they were looking for. Overall, there can be no more vivid testament to the unresolved tensions between dominant Hunnic rulers and exploited non-Hunnic subjects than the astonishing demise of Attila’s Empire. The strange death of Hunnic Europe, however, was also integral to the collapse of the western Empire.

A New Balance of Power

INSTEAD OF ONE HUGE power centred on the Great Hungarian Plain, its tentacles reaching out towards the Rhine in one direction, the Black Sea in another, the Roman Empire both east and west now found itself facing a pack of successor states. Much of the time fighting amongst themselves, they also pressed periodically upon the Roman frontier. As the Empire became ever more deeply involved in the fallout from the Hunnic collapse, the nature of Roman foreign policy on the Danube frontier began to change. In confronting their new situation, the Roman authorities had two priorities. They needed to prevent the squabbling north of the Danube from spilling over into their own territory in the form of invasions or incursions, while safeguarding that what emerged from the chaos should not be another monolithic empire.

The loss of the full text of Priscus’ history prevents us from telling a continuous story from the Roman perspective, but the essence is easy enough to distil. The surviving sources refer to overflows of various kinds on to Roman territory, the result of the ferocious struggle for Lebensraum on the other side of the Danube. Into the western Empire large numbers of refugees now flooded, individuals and groups who had decided that life south of the river looked preferable to the continuing struggle north of it. The most famous of these refugees was Odovacar, son of Edeco and prince of the Sciri. After the Amalled Goths destroyed the Scirian kingdom, he moved into Roman territory with a band of followers, heading first for Gaul and then for Italy, where he signed up with the Roman army. His lead was followed by many others of less distinguished origins. By the early 470s, the Roman army of Italy was dominated by central European refugees: Sciri are specifically mentioned, along with Herules, Alans and Torcilingi, who had all been recruited into its ranks.34 The surviving sources give us no numbers and no precise dates for the population moves that had brought them to Italy. This perhaps suggests that we should think in terms of a steady flow of immigration and recruitment, rather than a single large-scale influx, although factors such as the destruction of Scirian independence presumably accelerated the process.

If some groups, displaced in dribs and drabs, were merely fleeing the carnage north of the Danube, others were seeking to create their own enclaves on Roman soil – perceiving this, it would seem, as an easier option than continuing to compete on the Hungarian Plain. By the mid-460s, a number of groups were finding the competition too hot to handle, and three separate incursions on to east Roman territory took place in quick succession. In 466 or just afterwards, the Gothic king Bigelis (of the fourth group mentioned on page 353) led his followers south of the Danube, where he was defeated, Jordanes tells us.35 At more or less the same time a band of Huns led by a certain Hormidac raided Dacia, penetrating as far as the city of Serdica. There they were defeated by the east Roman general Anthemius.36 It was at this point too that Attila’s son Dengizich made his play for a piece of east Roman territory; as we have seen, he too failed to prosper. The arrival of these armed bands more or less coincides with the wars between the Amal-led Goths and their rivals on the Middle Danubian Plain, and, like the smaller flow of refugees into the western Empire, was perhaps caused by this new upsurge of violence.37

At the same time the new kingdoms were also, to an extent, carrying on from where the Huns had left off. Thanks to one of the two surviving fragments from Priscus’ history that deal with the aftermath of the fall of Attila’s Empire, we know that Valamer and his Goths invaded the east Roman Empire to extract an annual subsidy from it. By the early 460s, Priscus records, this amounted to 300 pounds of gold38 – a much smaller amount than was extracted by Attila at the height of his power (2,100 pounds) and less than half that paid to the Hun at the start of his reign. But it was not an insignificant sum, and if Valamer were to succeed in expanding his power-base further, there was always the chance that he would up his demands, just as the Huns had done. Since the authorities in Constantinople were probably having to pay annual subsidies to some of the other successor kingdoms as well, they had to tread very carefully. The new kingdoms had the potential to amalgamate into something just as nasty as Attila’s Empire. Some insight into Roman attitudes towards this potential problem is provided by the other relevant fragment to survive from Priscus’ history.39 During the interval between the first and second bouts of fighting between the Goths and the Sciri, both sides sent embassies to Constantinople asking for assistance. No one wanted to aid the Goths, but opinion was divided as to the best course to take. One counsel was that the Romans should keep out of the conflict entirely. Eventually, it was decided to give limited support to the Sciri. Jordanes ignores this dimension of the post-Attilan conflicts, but it’s clear that all sides were not only manoeuvring with and against one another, but trying to secure Roman support as well. The fact that no one in Constantinople wanted to back them attests to the increasing power of the Amal-led Goths, who were the closest thing to a new superpower.

The Romans greeted the death of Attila as the dawn of a new era. On the night of the great Hun’s death, the eastern emperor Marcian is said to have had a happy dream in which he saw Attila’s bow broken in two.40 However, the disappearance of a rival superpower proved not to be the end of all troubles, but a development that spawned a whole series of new problems. The prospect of a further clash of empires had vanished only to be replaced by many complicated regional conflicts with serious implications for both halves of the Roman world. And I strongly suspect that those we hear about in our motley collection of sources represent no more than the tip of the iceberg. Furthermore, the many and varied problems of refugees and invaders were as nothing compared with the broader consequences of the crash of Attila’s Empire. Above all, it destroyed the balance of forces on which, by the mid-fifth century, the western Roman Empire had come to depend.

The Fall of Aetius

AS WE SAW in Chapter 6, the emperor Valentinian III, son of Flavius Constantius and Galla Placidia, came to the throne in 425 at the age of six. He had been put there by the armies of the eastern Empire, and had never really held the reins of power. An eight-year domination by his mother, who eventually failed in her balancing act between the commanders of the several western army groups, had given way to that of Aetius. This man’s extraordinary military acumen during the 430s would both keep the western Empire afloat and cement his own hold on power. At fourteen a Roman youth was notionally an adult and could make legally binding decisions about property, but at this age in 433 Valentinian was nowhere near ready to compete for power with a tough and experienced general, especially when the Empire faced so many military problems. And by the time he might have been able to exercise authority, five or six years later, Aetius’ position was fully consolidated. By 440 it was the general, not the emperor, who was making the key decisions about policies and appointments – the very state of affairs that Placidia had laboured to avoid.

Thus, trapped within patterns of power over which he had no control, the notional emperor of the Roman west found himself a mere figurehead. The drudgery of such an existence is easy to underestimate. Never venturing out of Italy, Valentinian spent his time shuttling between Rome and Ravenna, his routine alternating between a private life replete with the trappings of almost limitless wealth, and state occasions. An emperor’s job, as we have seen, was to embody the core ideologies of the Roman state. He was expected to encapsulate the superhuman, indeed God-ordained, nature of the Roman world order, displaying in his ceremonial self the divine support that had called the Roman Empire into being. As the star turn in the many ceremonies, processions, Christian masses and audiences, he could never let his halo slip. And what he had to officiate at, day in and day out, was supremely tedious in its repetitiveness. The Empire being the epitome of the one-party state in action, public disagreement was not tolerated. Unity was all. Ceremonies were relentlessly orchestrated to bring this point home. It was under Valentinian, it will be remembered, that the Theodosian Code was introduced to the Senate (see p. 124). Valentinian was spared this particular performance, but it was typical of what he had daily to endure. The acclamations that probably prefaced every major imperial ceremony involved 245 shouts of approval from the assembled senators. A brief experiment I have just run with my eleven-year-old son reveals that you can shout about eighteen such acclamations in a minute, so that the ceremony for the Code would have taken at least forty minutes – and that’s not allowing for fatigue setting in and slowing things down towards the end.

Valentinian’s predecessors had experienced the same daily grind, but they at least had the satisfaction of making policy decisions and appointments behind closed doors once the spectaculars were over. We have already witnessed the frustration that such a lifestyle engendered in Valentinian’s sister Honoria: an affair with her estate manager, an unwanted pregnancy and a dangerous liaison with Attila the Hun (see Chapter 7). Nor was it easy for Valentinian to change things. Life is difficult for royal minors who reach adulthood only to find that they still remain marginal to the exercise of power. They may throw caution to the winds, like the seventeen-year-old Edward III, who at midnight on 19 October 1330 broke into Nottingham castle to remove his mother Queen Isabella, arrest her lover Mortimer, and seize the reins of power. But most royal minors are not so daring, and in the 440s Aetius was the young emperor’s only bulwark against the Huns.

If there was nothing that Valentinian could do about his frustrations in the 430s and 440s, the collapse of the Hunnic Empire brought a wind of change blowing through western court circles. By 450 or so, two bones of contention had arisen between Aetius and his emperor. On 28 July that year the eastern emperor Theodosius II had died after a fall from his horse. Valentinian was of the Theodosian dynasty, married to one of Theodosius’ daughters, Eudoxia, and it was Theodosius’ forces who had put him on the western throne in a determined restatement of the unity of that dynasty (see Chapter 6). Theodosius had been its last male representative in the east, his only son Arcadius having predeceased him. Hearing of his cousin’s death, Valentinian had the idea, so we are told, of going to Constantinople to assert his claim to rule the entire Roman world as sole emperor. Aetius set himself against the plan. It was certainly ill conceived. Valentinian had no contacts in Constantinople, and eastern political circles were not about to welcome him. Matters there were ordered by Theodosius’ sister Pulcheria, who had been a strong voice throughout her brother’s reign. Eventually she married a staff officer by the name of Marcian. On 25 August it was Marcian who became the new emperor of the east. Valentinian had missed his chance, such as it was, and Aetius’ opposition to his plan continued to rankle.

The second disagreement between the two concerned marriage alliances. Valentinian’s union with Eudoxia, produced only two daughters: Eudocia (born in 438 or 439) and Placidia (born between 439 and 443). By the early 450s, after fifteen years of marriage, it was unlikely that the imperial couple would have any more children. This meant that the succession to the western Empire was up for grabs, and the likeliest route to securing it would be marriage to one or other of Valentinian’s daughters. As we saw in Chapter 6, Eudocia had been betrothed to Huneric, son of Geiseric king of the Vandals, as part of the peace deal of the 440s, and he was not a serious contender for the throne. It was thus Placidia who became the key to the future of the Roman west, and Aetius worked hard in the early 450s to persuade Valentinian to betroth her to his son Gaudentius. Such a marriage would have cemented Aetius in power, making it extremely likely that Gaudentius would succeed Valentinian. Given the lack of a male Theodosian heir, marriage into the dynasty would have been sufficient to confer legitimacy, especially as the same procedure had just been followed in Constantinople. Whether, in pushing for the marriage, Aetius was responding to a perception that the eastern succession issue had already weakened his hold over Valentinian, is unclear. But the proposal certainly increased the emperor’s already festering resentment at the extent to which he was being marginalized within his own Empire.41

Moreover, with the death of Attila and the collapse of his Empire, Aetius now seemed much less critical to Valentinian’s survival, and it was the emperor, not Aetius, after all, who embodied imperial continuity. For the first time since reaching adulthood, Valentinian could dare to contemplate life without his generalissimo. Aetius perhaps sensed the danger, which might be another reason why he risked adding the marriage issue to Valentinian’s list of grievances. For all its emphasis on consensus, sharks always lurked in the deeper waters of Roman imperial politics; now, individuals in the emperor’s entourage caught the first faint scent of blood. Of the plot that eventually brought Aetius down we are pretty well informed, thanks again to the labours of Constantine VII Porphyrogenitus. An account survives in another of his works: Excerpts concerning Plots. The fall of Aetius is preserved in a fragment from the history of a certain John of Antioch, but he was a late compiler and probably drew primarily on the history of Priscus. So it is again the Priscus–Constantine axis that tells us what we want to know.

There were two main conspirators. The first was a Roman senator of high birth named Petronius Maximus. He had begun his career before Aetius came to power, but was clearly considered an Aetian loyalist. Between 439 and 441 he held the important post of Praetorian Prefect of Italy, and was named consul for a second time in 443 – both appointments taking place during Aetius’ pre-eminence.42 The second was drawn from the A-list of likely suspects in any Roman palace plot: the eunuch head of the emperor’s household Heraclius, the primicerius sacri cubiculi (Chief of the Sacred Bedroom). Armed with two issues with which to work on Valentinian, and aided by the fact that the Hunnic threat had receded, the plotters did their worst.43

As Aetius was explaining the finances and calculating the tax revenues, with a shout Valentinian suddenly leaped up from his throne and cried out that he would no longer be abused by such treacheries . . . While Aetius was stunned by this unexpected rage and was attempting to calm his irrational outburst, Valentinian drew his sword from his scabbard and, together with Heraclius, who was carrying a knife ready under his cloak . . . fell upon him.

Attacked simultaneously by emperor and eunuch, on 21 or 22 September 454 Aetius lay dead in the palace. His fall was followed by the usual round of bloodletting. Chief among the victims was Aetius’ current Praetorian Prefect of Italy, a senator by the name of Boethius, grandfather of the famous philosopher.

Valentinian had waited until his thirties, but he had finally broken free. Unfortunately for him, he was not nearly as successful as the young Edward would be some 900 years later at rallying support afterwards. For one thing, the conspirators soon fell out among themselves:

After the murder of Aetius, Maximus paid court to Valentinian hoping that he would be made consul, and when he failed to achieve this, he wished to become Patrician. But Heraclius . . . acting from the same ambition and not wishing a counter-balance to his own power, thwarted Maximus’ efforts by persuading Valentinian that, now he had freed himself from the oppression of Aetius, he should not transfer his power to others.

Old habits die hard, and even after Aetius’ death Valentinian was not really in charge. The challenge was on to run him, especially as he had no male offspring, which meant that, in the longer term, the imperial succession remained an open race. Once it became clear that he was getting nowhere by persuasion, Maximus turned again to deadlier methods, this time suborning two guards officers, Optila and Thraustila, who had been close to Aetius. Priscus relates that, on 16 March 455:

Valentinian decided to go riding [in Rome] on the Campus Martius . . . When he dismounted from his horse and was walking off to practise archery, Optila and his followers . . . attacked him. Optila struck Valentinian across the side of the head and, when he turned to see who had struck him, felled him with a second blow to the face. Thraustila cut down Heraclius, and both of them took the emperor’s diadem and horse and rode off to Maximus.

So perished Valentinian, less than six months after the murder of Aetius. This is the kind of political anarchy that always followed regime change in the Empire. After years of of autocratic rule, albeit in this case more a regency, there was no ready-made regime-in-waiting. As usual, a coalition had been hastily constructed by individuals who had no intention of sharing power with one another afterwards. But if the pattern of Aetius’ fall was nothing out of the ordinary, and the fact that it failed to generate an immediate successor hardly surprising, other features were highly particular. Fascinating in this respect is the obituary for Aetius, originally appearing in Priscus’ history immediately after the murder:

Through his alliance with the barbarians, he had protected Placidia, Valentinian’s mother, and her son while he was a child. When Boniface crossed from North Africa with a large army, he out-generalled him . . . Felix, who was his fellow general, he killed by cunning when he learned that he was preparing to destroy him at Placidia’s suggestion. He crushed the [Visigoths] who were encroaching on Roman territory, and he brought to heel the [Bagaudae] . . . In short, he wielded enormous power, so that not only kings but neighbouring peoples came at his order.

As obituaries go, it’s pretty succinct, and it captures the mix of plotting at court and campaigning in the field that was the reality of Aetius’ political life. What is especially interesting is the mention in its opening words of Aetius’ dependence on an alliance with ‘barbarians’. Not just any barbarians, but one group in particular: the Huns. As the passage suggests, Aetius’ career was founded upon his Hunnic alliance. It was the Huns who sustained him when he seemed about to lose civil wars – first in 425 as the usurpation of John unravelled, and again in 433 when Boniface defeated him at their first confrontation. And as we saw in Chapter 6, Hunnic troops played a central role in his restoration of order in Gaul in the 430s, particularly in his defeats of the Burgundians and Visigoths. Aetius’ death was far more than one man’s tragedy. It also marked the end of an era. The death of Attila and disappearance of the Hunnic Empire not only made it possible for Valentinian to contemplate life without Aetius, it also undermined the delicate balance of powers by which Aetius had kept the western Empire in business. Aetius without the Huns had been surplus to requirements. His successors needed to find a new mechanism for sustaining the west.

Brave New World

THE KEY TO understanding the new political order brought on by the extinction of Hunnic power is provided by virtually the first act of the short-lived regime of Petronius Maximus.

Having murdered Valentinian III on 16 March 455, he was proclaimed emperor the following day. His hands had barely grasped the imperial sceptre when he sent an ambassador to solicit the support of the powerful Visigoths, who had been settled in south-western France since 418. The man he chose was one of his newly appointed military commanders, perhaps commanding general in Gaul (magister militum per Gallias), Eparchius Avitus. Avitus was a Gallic aristocrat of impeccable fortune and education. Descended from high office-holders, he was related to a network of important families, and his estates centred on Clermont-Ferrand in the Auvergne. He had served with distinction under Aetius in the campaigns against the Norici and Burgundians in the 430s, then followed this up with a spell as supreme civilian administrator in Gaul – Praetorian Prefect – between 439 and 441. At that point he left office, possibly through natural rotation or because he fell out with Aetius, to return to prominence about a decade later. He then played a major role in negotiating the Visigothic assistance that helped Aetius repel Attila’s assault on Gaul in 451.44 In every way, therefore, Avitus was an excellent choice. Close to Aetius, but not too close, he had a good track record and connections with both the Gallic aristocracy and the Goths.

From Avitus himself, no writings have survived. As more than partial compensation, however, we have a collection of poetry and letters from his son-in-law, a certain Gaius Sollius Modestus Apollinaris Sidonius (who has already been cited in this book). The name is generally shortened for sanity’s sake to Sidonius. As his marriage alliance with the family of Avitus might suggest, Sidonius derived from Gallic landowning stock of similar standing – its main estates were situated around Lyon in the Rhône valley. His father had been Praetorian Prefect of Gaul himself about a decade after Avitus, holding the post in 448/9.45 In the past, Sidonius’ writings tended to get a rather bad press. At a time when any decent-thinking chap valued the standards of the classical Latin (first-century BC or AD) he was brought up on, the complexities and allusiveness of Sidonius’ work could only aggravate, if not shock. Compared with the clarity and matter-of-factness of, say, Caesar, his love of showing off seemed the height of decadence. Writing at the end of the Victorian era, Sir Samuel Dill passed this judgement:

[Sidonius] is essentially a literary man, of the stamp which this age of decadence [the fifth century] most admired. He is a stylist, not a thinker or inquirer. There is little doubt that he valued his own compositions not for their substance, but for those characteristics of style which we now think most worthless or even repulsive in them, the childish conceits, the meaningless antitheses, the torture applied to language so as to give an air of interest and distinction to the trivial commonplace of a colourless and monotonous existence . . .46

Even in translation, Sidonius can drive you crazy with his inability to call a spade a spade, and there’s no doubt he spent a lot of time trying to say things in as complicated a way as possible. One of his later letters contains a nicely illuminating comment, delivered at a moment when he thought that the literary audience he had been educated to address had gone for ever: ‘I am putting together the rest of my letters in more everyday language; it is not worth embellishing phrases which may never be published.’47But it is not fair to judge fifth-century style by first-century standards, and more recent commentators on late Roman Latin (not to mention late Roman Greek) have been less quick to condemn the stylistic complexities that were the height of artistic chic in the fourth and fifth centuries.48 An age that can see chain-sawed cows in preservative as art is by definition unlikely to judge other artistic endeavours by rigid universal standards.

In any case, the issue of whether Sidonius wrote ‘good’ Latin or not is beside the point, since there is no doubting the historical importance of his oeuvre. The earliest of his extant writings date from the mid-450s, the latest to about 480, but the bulk fall into a twenty-year period after 455. He knew pretty much everyone who was anyone in southern, especially south-eastern, Gaul, and the great and the good figure prominently in his letters, which, unlike those of Symmachus, don’t hesitate to discuss matters of political substance when appropriate. His poems, or some of them, are equally important. Sidonius was significant enough to be involved in politics, and for emperors to court him for his support, but he was not important enough to have to face execution when their regime collapsed. Recognized as one of the leading stylists of his age, he served a succession of emperors who drew on his talents as a writer of panegyrics – keynote speeches – in their praise. We have met such texts before, and while they certainly don’t tell the truth as you or I might recognize it, they have the huge virtue of giving us access to the world as particular regimes wished it to be portrayed. Sidonius, like Themistius and Merobaudes before him, was a propagandist.

From Sidonius’ account it emerges without a shadow of doubt that Petronius Maximus sent Avitus to the Visigoths to solicit their support for his regime. Sidonius, of course, dressed this bald fact up a little. As he portrays it, the Visigoths, after hearing of the murder of Valentinian III, were preparing to launch a hostile takeover bid for the entire Roman west, when news of the approach of Avitus filled them with sudden panic:49

One of the Goths, who had reforged his pruning-hook and was shaping a sword with blows on the anvil and sharpening it with a stone, a man already prepared to rouse himself to fury at the sound of the trumpet and looking at any moment with manifold slaughter to bury the ground under unburied foes, cried out, as soon as the name of the approaching Avitus was clearly proclaimed: ‘War is no more! Give me the plough again!’

You can see why those brought up on the tenets of classical Latin might find Sidonius’ verbiage annoying, but the rhetoric is anything but pointless. It gives us a clear picture of his father-in-law as the one man able to dissuade the Visigoths from launching war. The same imaginary Goth goes on to declaim that, far from being mere onlookers, his people will now lend their military assistance to the new regime – and precisely because it is sponsored by Avitus: ‘Nay, if I have gained a right knowledge of you [Avitus] in action before this, your auxiliary trooper will I be; thus at least I shall have permission to fight.’ What strikes you here is the exaggerated presentation of Avitus’ importance. Earlier in the poem, likewise, when talking of Aetius’ successes of the 430s, Sidonius excels himself: ‘He [Aetius], glorious in arms as he was, did no deed without you [Avitus], although you did many without him.’ Avitus no doubt performed useful service to him, but Aetius managed perfectly well without him in the 440s, when the latter slipped out of office. There can be no disputing that Aetius was the dominant partner.

But irritation at Sidonius’ hyperbole must not distract us from the historical significance of Petronius Maximus’ first move as emperor. Both Flavius Constantius and Aetius had strained every political sinew to prevent the Visigoths from increasing their influence within western imperial politics. Alaric and his brother-in-law Athaulf had both had visions, if fleeting, of the Goths as protectors of the western Empire. Alaric had offered Honorius a deal whereby he would become senior general at court, and his Goths be settled not far from Ravenna. Athaulf married Honorius’ sister and named his son Theodosius. But Constantius and Aetius, those guardians of the western Empire, had resisted such pretensions; they had been willing to employ the Goths as junior allies against the Vandals, Alans and Suevi, but that was as far as it went. Aetius had preferred to pay and deploy Huns to keep the Goths within this very real political boundary rather than grant them a broader role in the business of Empire. Avitus’ embassy, which, as Sidonius makes clear, sought from the Visigoths not just peaceful acquiescence but a military alliance, reversed at a stroke a policy that had kept the Empire afloat for forty years.

The immediate aftermath only reinforces the point. While Avitus was still with the Visigoths, the Vandals under the leadership of Geiseric launched a naval expedition from North Africa which brought their forces to the outskirts of Rome. In part, its aim was fun and profit, but it also had more substantial motives. As part of the diplomatic horse-trading that had followed the frustration of Aetius’ attempts to reconquer North Africa, Huneric, eldest son of the Vandal king Geiseric, had been betrothed to Eudocia, daughter of Valentinian III. On seizing power, however, in an attempt to add extra credibility to his usurping regime, Petronius Maximus married Eudocia to his own son Palladius. The Vandal attack on Rome was also made, then, in outrage at being cheated, as Geiseric saw it, of this chance to play the great game of imperial politics. Hearing of the Vandals’ arrival, Maximus

panicked, mounted a horse and fled. The imperial bodyguard and those free persons around him whom he particularly trusted deserted him, and those who saw him leaving abused him and reviled him for his cowardice. As he was about to leave the city, someone threw a rock, hitting him on the temple and killing him. The crowd fell upon his body, tore it to pieces and with shouts of triumph paraded the limbs about on a pole.50

So ended the reign of Petronius Maximus, on 31 May 455; he had been emperor for no more than two and a half months.

When the imperial capital was sacked for the second time, the damage sustained was more serious than in 410. Geiseric’s Vandals looted and ransacked, taking much treasure and many prisoners back with them to Carthage, including the widow of Valentinian III, her two daughters, and Gaudentius, the surviving son of Aetius.51 Upon hearing this news, Avitus immediately made his own bid for the throne, declaring himself emperor while still at the Visigothic court in Bordeaux. It was later, on 9 July that year, that his claim was ratified by a group of Gallic aristocrats at Arles, the regional capital. From Arles, not long afterwards, Avitus moved on triumphantly to Rome and began negotiations for recognition with Constantinople. The senior Roman army commanders in Italy – Majorian and Ricimer – were ready to accept him because they were afraid of the Visigothic military power at his disposal.52

A new order was thus born. Instead of western imperial regimes looking to keep the Visigoths and other immigrants at arm’s length, the newcomers had established themselves as part of the western Empire’s body politic. For the first time, a Visigothic king had played a key role in deciding the imperial succession.

The full significance of this revolution needs to be underlined. Without the Huns to keep the Goths and other immigrants into the Roman west in check, there was no choice but to embrace them. The western Empire’s military reservoirs were no longer full enough for it to continue to exclude them from central politics. The ambition first shown by Alaric and Athaulf, and later by Geiseric in his desire to marry his son to an imperial princess, had come to fruition. Contemporaries were fully aware of the political turn-around represented by Avitus’ elevation. Since time immemorial, the traditional education had portrayed barbarians – including Visigoths – as the ‘other’, the irrational, the uneducated; the destructive force constantly threatening the Roman Empire. In a sense, with the Visigoths now having served for a generation as minor Roman allies in south-western France, the ground had been well prepared. Nonetheless, Avitus’ regime was only too well aware that its Visigothic alliance was bound to be controversial. This is nowhere better demonstrated than in the writings of Sidonius, in particular in a letter penned by him from the court of the Visigothic king Theoderic II in the early months of Avitus’ reign. Sidonius’ letters are in no sense private documents. He wrote them in the expectation that their contents would be circulated. They were, in short, an excellent mechanism for disseminating a point of view among fellow Gallic landowners.53

Written to Avitus’ son Agricola as a description of life at the Visigothic court, it opens with a portrait of Theoderic: ‘In his build the will of God and Nature’s plan have joined together to endow himwith a supreme perfection; and his character is such that even the jealousy which hedges a sovereign has no power to rob it of its glories.’ We then hear about the king’s day. Having started with a prayer or two, he spends the morning receiving embassies and settling cases; then, in the afternoon, perhaps a little hunting, at which, as in all else, he excels. In the evening comes the main meal:

When one joins him at dinner . . . there is no unpolished conglomeration of discoloured old silver set by panting attendants on sagging tables; the weightiest thing on these occasions is the conversation. The viands attract by their skilful cookery, not by their costliness. Replenishment of goblets comes at such long intervals that there is more reason for the thirsty to complain than for the intoxicated to refrain. To sum up: you can find there Greek elegance, Gallic plenty, Italian briskness; the dignity of state, the attentiveness of a private home, the ordered discipline of royalty.54

The letter closes with a little joke at the king’s expense. After dinner Theoderic liked to play a game of dice, and would show the proper spirit by protesting if he perceived that his rival was letting him win. On the other hand, should you want a favour done, Sidonius notes, the thing to do was to let the king win, but without his noticing what you were up to. This bit of patronizing aside, Sidonius’ message could not be clearer. Theoderic II was not your run-of-the-mill barbarian, driven by his senses, addicted to alcohol and the next adrenalin rush. He was, in fact, a ‘Roman’ in the proper sense, one who had learned reason and self-discipline, who ran his court, his life – indeed, himself – in the time-hallowed Roman manner. He was a man one could do business with. I have no idea what life was really like at the Visigothic court, but to justify Avitus’ association with Theoderic, Theoderic had to be presented as possessing all the virtues, and Sidonius duly obliged. The revolution was gathering pace. Barbarians were being presented as Romans to justify the inescapable reality that, since they could no longer be excluded, they now had to be included in the construction of working political regimes in the west.

At first sight, this inclusion of the alien would not seem to be a mortal blow to the integrity of the Empire. Theoderic was Roman enough to be willing to play along; he saw the need to portray him as a good Roman in order to satisfy landowning opinion. There were, however, a couple of very big catches which made a Romano-Visigothic military alliance not quite the asset you might initially suppose. First, political support always came at a price. Theoderic was entirely happy to support Avitus’ bid for power, but, not unreasonably, he expected something in return. In this instance, his desired reward was a free hand in Spain where, as we have seen, the Suevi had been running riot since Aetius’ attention had been turned towards the Danube in the early 440s. Theoderic’s request was granted, and he promptly sent a Visigothic army to Spain under the auspices of Avitus’ regime, notionally to curb Suevic depredations. Hitherto, of course, when the Visigoths had been deployed in Spain, it was always in conjunction with Roman forces. This time, Theoderic was left to operate essentially on his own initiative, and we have a first-hand – Spanish – description of what happened. The Visigothic army defeated the Suevi, we are told, capturing and executing their king. They also took every opportunity, both during the assault and in the cleaning-up operations that followed, to gather as much booty as they could, sacking and pillaging, amongst others, the towns of Braga, Asturica and Palentia. Not only did the Goths destroy the kingdom of the Suevi, they also helped themselves uninhibitedly to the wealth of Spain.55 Just like Attila, Theoderic had warriors to satisfy. His willingness to support Avitus was based on calculations of profits, and a lucrative Spanish spree was just the thing.

Second, the inclusion of barbarians into the political game of regime-building in the Roman west meant that there were now many more groups manoeuvring for position around the imperial court. Before 450, any functioning western regime had to incorporate and broadly satisfy three army groups – two main ones in Italy and Gaul, and a lesser one in Illyricum – plus the landed aristocracies of Italy and Gaul, who occupied the key posts in the imperial bureaucracy. The desires of Constantinople also had to be accommodated. As in the case of Valentinian III, should western forces be divided between different candidates, eastern emperors disposed of enough clout and brute force to impose their own candidate. Though too far away to rule the west directly, Constantinople could exercise a virtual veto over the choices of the other interested parties. Incorporating this many interests could make arriving at a stable outcome a long-drawn-out business.

AFTER THE COLLAPSE of the Hunnic Empire, the Burgundians and Vandals were the next to start jockeying for position and clamouring for rewards. The Burgundians had been settled by Aetius around Lake Geneva in the mid-430s. Twenty years later, they took advantage of the new balance of power in the west to acquire a number of other Roman cities and the revenues they brought with them from their territories in the Rhône valley: Besanc¸on, le Valais, Grenoble, Autun, Chalon-sur-Saône and Lyon.56 The Vandal–Alan coalition’s sack of Rome in 455, as we have seen, betrayed a desire to participate in imperial politics. On the death of Valentinian, Victor of Vita tells us,57 Geiseric too, expanding his powerbase, seized control of Tripolitania, Numidia and Mauretania, together with Sicily, Corsica and the Balearics. Allowing just some of the barbarian powers to participate in the Empire massively complicated western politics; and the greater the number, the harder it was to find sufficient rewards to generate long-term coalition.

A strong sense of the underlying tensions that made the regime of Avitus essentially unstable emerges from the second of Sidonius’ poems to survive from this period. On 1 January 456, when the emperor assumed the consulship in Rome, his ever loyal son-in-law was called upon to make a speech on his behalf. It began, not surprisingly, by establishing the emperor’s overwhelming suitability for office. In doing so, Sidonius took the opportunity to make some pointed comparisons. In particular, he dismissed Valentinian III as a ‘mad eunuch’ (semivir amens), and contrasted his style of leadership with the military and political skills that Avitus brought to the job. Turning to the key issue of Avitus’ relationship with the king of the Visigoths, Sidonius handled this potentially explosive subject with subtlety, but his intent was clear enough. First, he argued with vigour that Avitus had never been one to cosy up to the Visigothic court. He had been there as a young man, as everyone knew, in the 420s, when ‘[the Visigothic king] desired exceedingly to have you [Avitus] as one of his own, but you scorned to act the friend rather than the Roman.’58 Sidonius then focused on one small incident in the 430s when Avitus took a terrible revenge against a marauding Visigoth who had wounded one of his servants:

When first they approached, breast to breast and face to face, the one [Avitus] shook with anger, the other [the Goth] with fear . . . But when the first bout, the second, the third have been fought, see! The upraised spear comes and pierces the man of blood; his breast was transfixed and his corselet twice split, giving way even where it covered the back; and as the blood came throbbing through the two gaps the separate wounds took away the life that each of them might claim.

Translated into English (or even into Latin), Sidonius is saying that Avitus found the Visigothic bastard who’d hurt his man, and ran him so far through with his spear that it came out the other side. Translated into politico-speak, the message is that Avitus was no Visigoth-loving traitor but a true Roman who had given the barbarians as hard a kicking as even the fiercest hawk could desire.

All of this was addressed to the suspicions of Sidonius’ audience of Italo-Roman senators and generals, as was, above all, his account of the new emperor’s elevation. On hearing of the deaths of Aetius and Valentinian, the Visigoths had begun to plan their own wars of conquest.59 Then into the Visigothic camp strode Avitus, and everything changed. By his presence alone he spread panic among them, and such was their fear of him that the Visigoths’ immediate impulse was to try to please him by engaging in a military alliance. But whether Avitus should declare himself emperor was his decision alone. As for the Visigothic king, Sidonius has him say:

We do not force [the purple] on you, but we do beg you; with you as leader I am a friend of Rome, with you as emperor I am her soldier. You are not stealing the sovereignty from any man; no Augustus holds the Latian hills, a palace without a master is yours . . . My part is only to serve you; but if Gaul should force you, as she has the right to do, the world would love your command for fear it would otherwise perish.

We see from this special pleading, and the allusion to the power vacuum in Italy, exactly where the audience’s political sensitivities lay. To the Italians, the audience Sidonius was now addressing, Avitus might appear no more than a creature of the Visigoths after the pattern of Priscus Attalus under Alaric and Athaulf. The speech responded by insisting that Avitus was his own man. You only had to look at his long history of smacking the Visigoths around. He had also taken the purple, if unwillingly, because he was the only man who could command their obedience. In these straitened times, the barbarians’ military power was necessary to the safety of the Empire, but Avitus remained a true Roman.

It was a good try. And so much for the claim that Sidonius lacked ideas. But the Italian audience, particularly the army men amongst them, were having none of it. The sources insist, as we have seen, that the Roman army of Italy only ever tolerated Avitus because he had the military backing of the Visigoths. When, in 456, the Visigoths became too deeply embroiled in Spain to intervene any further in Italy, the two main Roman commanders, Majorian and Ricimer, withdrew their allegiance. On 17 October that year they gave battle to the few forces Avitus could scrape together – presumably remnants of the Roman field army of Gaul – outside the city of Placentia in northern Italy. Avitus was beaten, forced to become the city’s bishop, and died shortly afterwards in mysterious circumstances.60

We see here, then, in a nutshell the problem now facing the west. Avitus had the support of the Visigoths, the support of at least some Gallic senators, and of some of the Roman army of Gaul. But faced with the hostility of the Italian senators, and especially of the commanders of the Italian field army, the coalition didn’t stand a chance. By the early 460s, the extent of the crisis in the west generated by the collapse of Attila’s Empire was clear. There were too many interested parties and not enough rewards to go round. Constantinople, however, had decided on one last throw of the dice.

If you find an error or have any questions, please email us at Thank you!