CHAPTER 6
Barbarians and Romans
Goths, Vandals and Huns
During the half century following the death of Theodosius I, the empire had to contend with many barbarian groups, but the most significant were the Goths, the Vandals and the Huns. Each of these groups owed its ability to create serious difficulties for the empire, at least in part, to having forceful leaders in the persons of Alaric, Geiseric and Attila respectively. The analytical narrative which follows therefore focuses on imperial dealings with each of these men in turn.1
Gothic federates had formed an important component of Theodosius’ expeditionary force to the west in 394 and had played a major role in the victory at the River Frigidus (p. 27 above). It was perhaps lack of recognition of that role which prompted some of these Goths, on their return to the Balkans in 395, to rebel against imperial authority, under the leadership of one of their officers, Alaric, who seems to have resented not being given a formal Roman command. No doubt supplemented by other Goths from the region, Alaric’s forces plundered widely through the Balkans, and since the eastern field armies from the Balkan region had not yet returned from Italy, the authorities in Constantinople could do little to stop them. Stilicho did intervene with the substantial forces at his disposal, but since Rufinus feared that this was a pretext for Stilicho to march on Constantinople, Arcadius was persuaded to order Stilicho to relinquish those of his troops from the eastern field armies and return to Italy – which Stilicho duly did (at the cost, however, of Rufinus’ life (p. 90 above). The following year, Alaric moved south into Greece, unopposed because eastern forces were preoccupied with a major Hunnic invasion across the Caucasus into Anatolia. In 397 Stilicho again intervened, but before he was able to deal a decisive blow against Alaric, he suddenly withdrew his forces to Italy – perhaps because Eutropius, fearful of Stilicho’s intentions, instigated Gildo’s rebellion in north Africa (the precise chronological relationship of these different events is, however, not entirely certain) (pp. 83–4 above). At the same time, Eutropius seems to have appeased Alaric by granting him some sort of command in the eastern army, usually thought to be that of senior general in the Balkans, as well as probably providing pay and rations for his forces. This would certainly help to account for the relatively quiescent state of Alaric’s Goths over the next three to four years.2
That changed in late 401 when Alaric and his forces moved to the north-west Balkans and crossed the Julian Alps into northern Italy (facilitated, no doubt, by the recent destruction of the forts in this area during Theodosius’ campaign against Eugenius and Arbogast).3 Suddenly, Constantinople’s Gothic problem had become Stilicho’s. The most plausible explanation for Alaric’s decision is that his agreement with Eutropius in 397 lapsed in 399 when the latter fell from power, and Eutropius’ immediate successors were unwilling to continue arrangements; when the turmoil at Constantinople created by Gainas in 400 (above, p. 92) failed to produce a regime more amenable to Alaric’s wishes, he may have calculated that he might achieve success more easily by turning his attention westwards.4 Initially, that proved not to be the case, since Stilicho achieved sufficient success against Alaric in a battle at Pollentia in north-west Italy on Easter Day 402 to convince the Goths to withdraw back to the north-west Balkans, where they remained for the next few years. Perhaps the most important consequence of these events concerned the location of the imperial court. In the period before Pollentia, the Goths had threatened Milan, and although Honorius was not there at the time, it was enough to persuade him that Ravenna would make a better residence. While the defensive advantages of Ravenna were not as great as often assumed – among other things, it had no natural supplies of fresh water – it did offer greater security than Milan, but above all its coastal location on the Adriatic facilitated communications with the east.5
Alaric’s return to the north-west Balkans brought only temporary respite to Italy, for in 405 another substantial body of Goths and other barbarians, this time from outside the empire, crossed the middle Danube and advanced into northern Italy, where they plundered the countryside and besieged cities and towns. The decision of their leader Radagaisus to split his forces into three groups helped Stilicho to contain the threat, eventually cornering Radagaisus’ group near Florence and starving them into submission. However, the imperial government was clearly struggling to find enough military manpower to cope with these different invasions, since Stilicho had to recall units from Britain and Gaul, and a law issued at Ravenna in early 406 took the very unusual step of offering slaves their freedom and money if they enlisted (Claud. de bello Getico 414–29, Cod. Theod. 7.13.16).
Pressures on Stilicho increased further when another body of barbarians, comprising Vandals, Sueves and Alans, crossed the Rhine into Gaul at the end of 406 (p. 115 below),6 and a rebellion in Britain spread to Gaul under the leadership of the usurper Constantine in 407 (p. 84 above). All these developments prevented Stilicho from implementing a plan to take troops into the western Balkans (perhaps with a view to detaching from Constantinople a region which could help to relieve the west’s shortage of recruits) – but not before he had persuaded Alaric to prepare the way by advancing there with his forces. When Stilicho then found himself unable to follow with Roman troops, Alaric marched back into Italy and demanded financial compensation for his efforts to the tune of 4,000 lbs of gold. While not a vast sum by the standard of senatorial incomes (cf. Olympiodorus, fr. 41,2), Stilicho’s insistence in early 408 that the senate should find the money to pay Alaric met with fierce resistance as a matter of principle, and although Stilicho’s request was eventually conceded, this episode played a major part in generating the opposition which eventually led to his downfall later that year (p. 84 above).
Those responsible for Stilicho’s elimination did not retain their influence over Honorius for long, but since they were resolutely opposed to negotiating with Alaric, their short tenure of power was to have disproportionate consequences. When his attempt to negotiate some sort of modus vivendi with the new regime was rejected, Alaric tried to place pressure on Honorius by blockading Rome during the winter of 408/9. The resulting starvation induced the senate to send envoys to Ravenna pleading with the emperor to grant Alaric’s demands – money, food and a senior command in the army. After provisionally agreeing to Alaric’s terms in return for lifting the blockade, Honorius then reneged on a part of the terms (specifically the senior army command), prompting Alaric not only to besiege Rome again in late 409, but also to proclaim a leading senator, Priscus Attalus, as a rival emperor, from whom Alaric then received the appointment as senior general which Honorius could not bring himself to bestow.7
A now desperate Honorius initially offered to share power with Attalus, but when this proposal was rejected, he made ready to escape from Ravenna by sea to the east, until his resolve was restored by the unexpected arrival by ship of 4,000 troops from Constantinople. Eventually in the summer of 410 he offered further negotiations with Alaric, whereupon Alaric deposed Attalus and advanced towards Ravenna. While waiting nearby for negotiations to commence, however, Alaric and his retinue were attacked by an independent body of Goths led by a certain Sarus who had long been connected with the court of Honorius. Whether Sarus was acting on imperial orders or on his own initiative is unclear, but Alaric assumed the former, and immediately withdrew southwards, where he besieged Rome for a third time. This time, however, his aim was not to force Honorius back to the negotiating table, but rather to vent his frustration. In a little over a fortnight, the city was captured and then subjected to three days of plundering, before Alaric withdrew southwards, perhaps hoping to cross over to north Africa.8
Far from representing some great success for the Goths, therefore, Alaric’s sack of Rome in 410 was, as commentators have noted, ‘an irrelevance, forced on him by the failure of his other policies’, and ‘an admission of defeat’.9 From an imperial perspective, too, the fact that Rome had long ceased to be the emperor’s residence and the focal point of court politics meant that, at a practical level, the significance of the event was also much reduced. Nonetheless, there was no escaping its symbolic importance, as the first time that a foreign army had occupied the city since the Gallic sack of 387/6 BC – when Rome was still only one state among many in the Italian peninsula and was still more than a century away from acquiring the beginnings of a territorial empire. This helps to explain the famously apocalyptic reaction of one contemporary to news of the event, referring to it as the time ‘when the brightest light of the world was extinguished, the Roman empire was decapitated, and the whole world perished in one city’ (Jer. Comm. in Ezech. 1, praef.). At the same time, there was a ground swell of pagan opinion blaming the empire’s adoption of Christianity for the fate of the city – a reaction which provoked Augustine to write his hugely influential apologetic work De Civitate Dei (The City of God).10 As for the city itself, it certainly experienced destruction of buildings, as well as loss of population – besides those killed or taken prisoner by the Goths, many wealthy inhabitants fled overseas to north Africa and the east, while numerous slaves seized the opportunity to escape. However, there is a limit to how much damage can be done to a city of Rome’s size in three days, and the city did recover, with a succession of public structures in the city centre restored over the course of the next decade or so.11
As for the Goths, their attempt to cross to Sicily en route to Africa was thwarted by a storm which wrecked the fleet they had assembled, and while returning northwards, Alaric himself fell ill and died, whereupon his brother-in-law Athaulf was chosen to succeed him as leader. Under him the Goths eventually advanced into southern Gaul, where they fought against both Roman and Vandal forces. In an effort to strengthen his position vis-à-vis the Romans, he married Galla Placidia (the daughter of Theodosius I who had been captured at Rome in 410) in 414; intriguingly, at the ceremony not only the bride, but also the groom, was attired in Roman dress (Olympiodorus, fr. 24). It is reported that, under Galla’s influence, Athaulf declared it his aim to ‘restore and extend the Roman empire by the might of the Goths … and to be remembered by posterity as the author of Rome’s renewal’ (Oros. 7.43.6–7). However, the opportunity to see what that might have meant in practice was foiled the following year when Athaulf was murdered by a resentful retainer. As a result of the energetic activities of the Roman general Constantius (pp. 84–5 above), who blockaded the Goths in northern Spain, Athaulf’s successor Vallia was forced by the threat of starvation to come to terms in 416, agreeing to hand over Galla Placidia and to fight the Vandals and other barbarians in Spain in return for supplies of food. After two years of effective campaigning, he negotiated Roman agreement to the Goths’ settling permanently in Aquitania, although he himself died before this agreement was implemented.12 The arrangement suited the Roman authorities since Aquitania was geographically peripheral to the centres of political power in the west. Unlike Vallia, or indeed Athaulf, his successor Theoderic, a son-in-law of Alaric, was to enjoy an extended period as ruler of the Goths, from 418 until 451 – a longevity which enabled him to develop an increasingly independent Gothic state in south-western Gaul (p. 178 below).13
The Gothic sack of Rome may be the most enduring image associated with this period in the popular imagination, but it was another barbarian group who were to play the most significant role in determining the fate of the western empire – the Vandals. That role was not to become apparent for some decades following their crossing of the Rhine at the end of 406, but even during those intervening years before they invaded north Africa their presence in Gaul contributed to the fall of Stilicho, the rise of Constantius and the trajectory of the Goths.14 A decade or so later, in 418, their future may have looked doubtful as they reflected on the recent destruction of one sub-grouping of Vandals by Vallia’s Goths while the other, surviving sub-grouping found itself confined to the north-west of the Iberian peninsula. However, the withdrawal of the Goths to Aquitania, followed by the death of Constantius in 421 and the extended period of instability in Italy after Honorius’ death in 423 (p. 85 above), provided these surviving Vandals with the opportunity to move southwards into the wealthier regions of southern Spain and consolidate their position there. Nonetheless, they are unlikely to have been viewed by the imperial authorities as a serious threat at this stage.15
The accession of a new Vandal ruler in 428 in the person of Geiseric – ‘among the German leaders of his time … unquestionably the ablest’16 – prompted a bold initiative which opened up an alarming new scenario in the west: in 429 (probably), Geiseric led the Vandals across the Straits of Gibraltar into north Africa, attracted no doubt by the economic resources of the region and by the limited Roman military presence.17 Advancing eastwards, they soon encroached on the more heavily populated and wealthy region of Numidia. In 431 a Roman army comprising forces from Italy and the east arrived under the command of Aspar, but these forces proved unable to defeat the Vandals and eventually came to terms with them in 435. The Vandals were permitted to retain control of part of Numidia and the regions further west, while also paying an annual tribute to the imperial government which, crucially, retained control of Carthage. However, Geiseric soon took advantage of Aetius’ preoccupation with affairs in Gaul to resume his eastward advance until, in October 439, the Vandals captured Carthage. They quickly mobilised the shipping resources of the city to attack Sicily and provoke panic about a possible seaborne expedition against Italy and even Constantinople.
Even without this unexpected naval threat, the Vandal occupation of the north African provinces as far east as Carthage was a blow of incalculable importance to the western empire, for the provinces in question included the wealthiest in the western Mediterranean, based on their production of grain, olive oil and wine. Their occupation by the Vandals meant the loss of significant revenues for the government, with all that implied for its ability to maintain the army.18 It is not surprising therefore that the government in Constantinople quickly began coordinating another rescue package, in the form of a major expeditionary force combining eastern forces from the lower Danube with western forces. In 441 this force assembled in Sicily, to be met by a request for negotiations from Geiseric, who then dragged them out over the ensuing months until the empire’s opportunity was lost. Fortuitously for the Vandals, a Hunnic invasion on the lower Danube forced Constantinople to withdraw its forces from Sicily, and Ravenna then had to agree (in 442) to a far from satisfactory settlement with Geiseric which recognised Vandal possession of Carthage and adjacent regions in return for Vandal recognition of imperial control of the western regions of north Africa – effectively, the mirror image of the settlement of 435. The importance of this episode cannot be overstated, for had the expedition proceeded and managed to defeat the Vandals, the western empire’s financial well-being would have been placed on a much sounder footing, with all that that implies for its ability to sustain its military efforts in the years to come. The actual outcome, on the other hand, left the west with a serious handicap which risked more and more becoming fatal the longer it was not reversed.
It was the Huns who were, to a significant degree, responsible for that outcome and who also presented the eastern empire with increasingly serious problems of its own during the first half of the fifth century. Unlike the Goths and Vandals – both Germanic-speaking groups who, despite their movement through parts of the empire in the early fifth century, were from socio-economic backgrounds involving settled villages and arable farming – the Huns were nomadic pastoralists from Central Asia. Their skill in horse-manship and archery gave them military potential, but the herding of animals which underpinned their way of life meant a social structure involving dispersed, impermanent settlement which normally militated against the development of centralised political authority. However, their advent in the lower Danube basin in the late fourth century allowed them to establish hegemony over various settled barbarian groups living in the region – for even after Valens’ admission of so many Goths to the empire in 376, there remained substantial numbers north of the Danube who soon fell under Hunnic rule. The resulting control of significant human and material resources facilitated the emergence of Hunnic leaders able to present significant challenges to Constantinople.19
The first warning of this worrying development came in 408 when the Hunnic leader Uldin, who had been responsible for the death of Gainas some years earlier (p. 92 above), invaded Thrace, albeit without being able to sustain the threat. This episode may have been enough to prompt the decision to construct impressive new land walls for Constantinople, completed in 413 (Fig. 9).20 At about the same time, the imperial court was sufficiently concerned to use an embassy to the Huns as a cover for the assassination of a Hunnic leader named Donatus.21 It was in 422, however, that the direct consequences of the Huns’ proximity to the empire were first felt in a serious way. Relations with Persia were disturbed only twice during the fifth century, with short-lived conflicts occurring in 421–2 and 440,22 and the Hunnic leader Rua took advantage of the first of these to invade Thrace and force Constantinople to purchase his withdrawal by promising an annual payment of 350 lbs of gold over the next fifteen years. Rua employed a similar tactic in 434 when significant numbers of imperial troops were absent from the Balkans helping to defend Carthage against the Vandals, although on this occasion the fortuitous death of Rua, apparently through a lightning strike, afforded the empire some respite.
Figure 9 A section of the Theodosian Walls, Constantinople. © Nevit Dilmen/Wikimedia Commons
Rua’s death, however, signalled the emergence into leadership of his nephews Bleda and Attila, the latter of whom was to become a formidable enemy of the empire over the next two decades. While perhaps not as able as Geiseric, and certainly not as long-lived – Attila died in 453, whereas Geiseric continued until 477 – Attila is better known primarily because the pre-eminent historian of this period, Priscus of Panium, wrote a detailed account of an embassy to Attila in 449 in which he participated (Priscus, fr. 11, 2). Bleda and Attila’s first opportunity to aggrandise themselves at Constantinople’s expense arose in the winter of 439/40 when, in anticipation of the forthcoming expedition against the Vandals, Roman envoys agreed to double the annual payment to 700 lbs of gold in return for assurances of peace on the lower Danube. However, once those forces assigned to the Vandal expedition had been withdrawn, the Huns found excuses for reneging on their recent agreement and began raiding Balkan communities. The recall of the imperial forces from Sicily put an end to this and an uneasy truce ensued until 447, when Attila (now sole ruler, following his murder of Bleda in 445) once more went on the offensive, this time with devastating effect. Imperial forces were unable to resist, and Constantinople had no choice but to grant Attila’s demands, including a trebling of annual payments to 2,100 lbs of gold. Despite the burden this placed on imperial finances, resources were found to invest in the construction of the so-called ‘Long Walls’ in Thrace, with a view to enhancing the protection of Constantinople’s hinterland, on which the city relied for its water supply via aqueduct. An attempt to solve the Hunnic problem by the expedient of assassinating Attila in 449 went badly awry, but the fallout was not as bad as it might have been, since Attila was beginning to shift his focus westwards.23
This shift, whose rationale remains unclear,24 resulted in the Huns invading Gaul in 451 and advancing towards Orleans until confronted by Aetius, leading a coalition of Roman and barbarian forces, among whom Goths were most prominent. In the ensuing battle of the Catalaunian Plains, Aetius forced Attila to retreat, but not before suffering serious losses, including that of the Gothic ruler Theoderic. After regrouping over the winter, Attila invaded Italy the following year, capturing many of the major cities in the north, until a combination of food shortage, disease and harrying by Aetius induced him to withdraw. Any plans he may have had to return were forestalled by his sudden death in 453, apparently from a brain haemorrhage. While helpfully removing the source of some of the empire’s most taxing problems in recent years, this fortuitous development was to generate another whole set of difficulties for the east arising from the break-up of Attila’s empire.
Vandals, Goths and Franks
Of the three major barbarian leaders of the first half of the fifth century, only the Vandal leader Geiseric now survived, and while he lived (until 477), the Vandals continued to create significant difficulties for the empire. However, as already intimated, the death of Attila also created new instabilities in the Balkans, out of which another, different group of Goths, led by Theoderic the Amal, emerged as a powerful force in imperial affairs. Finally, in the closing decades of the fifth century, a new barbarian group known as the Franks began to impinge forcefully on Gaul.
The settlement of 442 by which Valentinian III acceded to the Vandal occupation of Carthage and the more prosperous parts of north Africa also seems to have included provision for the betrothal of his elder daughter Eudocia to Geiseric’s eldest son Huneric – a move which has interesting implications for Geiseric’s attitude to the empire, and for the political capital to be gained from marriage into the imperial family. In the event, nothing further happened in this respect until Valentinian’s death in 455. Upon receipt of this news, Geiseric led a large seaborne force to Italy and proceeded to sack the city of Rome in a much more comprehensive manner than Alaric’s Goths had done in 410. Ironically, the booty he carried back to Carthage included treasures from the Roman sack of Jerusalem in 70, but also Valentinian’s wife and daughters, and at some point after their arrival in Carthage, Eudocia was finally married to Huneric. This union produced a son, Hilderic, who was to play a crucial role in the eventual imperial conquest of north Africa by the Emperor Justinian in the sixth century (p. 259 below).
Valentinian’s death does appear to have been the trigger for a more aggressive policy towards the empire generally on Geiseric’s part, with Vandal forces occupying Sicily, Sardinia, Corsica and the Balearics. This in turn prompted the western emperor Majorian (457–61) to prepare another attempt to drive the Vandals from north Africa. However, Geiseric launched a pre-emptive strike which destroyed the fleet Majorian had prepared in ports along the southeast coast of Spain, and in the aftermath of this debacle Ricimer had Majorian arrested and executed (p. 94 above). For the next few years, Geiseric led his fleet in annual raids along the coastline of Sicily and Italy (Priscus, fr. 39, 1), but his extension of this raiding into the eastern Mediterranean, together with his persecution of orthodox Christians in north Africa (pp. 187–8 below), eventually provoked a response from Constantinople (Procop. Wars 3.5.22, 3.6.1). In 467 the Emperor Leo persuaded Ricimer to support the elevation of the eastern aristocrat Anthemius to the vacant imperial throne in the west in preparation for the largest expedition yet to regain north Africa from the Vandals. An armada from Constantinople under the command of Leo’s brother-in-law Basiliscus combined with western forces to advance on Carthage.
Once again, however, Geiseric proved adept at using diplomatic delay to forestall the attack, and then deployed fireships to wreak havoc among the densely packed imperial fleet as it lay moored off the coast of north Africa. The cost of this disaster was put by contemporaries at somewhere between 7.5 and 9 million solidi – a truly enormous sum ‘which probably exceeded a whole year’s revenue’ for the eastern empire.25 This failure also had other significant repercussions. As well as effectively sealing the fate of the western empire, it left deep psychological scars in the memories of the elite in Constantinople for many years to come, ensuring that the Vandals escaped further external interference for more than half a century – a development further aided by the death of Geiseric in 477. While his successors continued to persecute orthodox Christians, in other respects they pursued less provocative policies towards Constantinople, giving their attention, as they had to, to the growing problems presented by local Moorish polities (p. 260 below).
In addition to leaving his successor Zeno with a near-empty treasury, Leo also bequeathed him a highly volatile situation in the Balkans. Following Attila’s death in 453, the various barbarian groups who had been subject to his rule in the lower Danube basin seized the opportunity to rebel against Hunnic rule, successfully defeating Attila’s sons in a battle at the River Nedao in 454. Of these newly free groups, the most important were Goths. During the late 450s and 460s, one group of these Goths settled in the former province of Pannonia in the north-western Balkans, while a second group settled in Thrace with a formal status as federate troops. Aspar made a point of cultivating ties of patronage with this second group, with the result that his assassination by Leo in 471 (p. 99) provoked a revolt on their part. Leo was only able to restore order in the region by agreeing in 473 to a number of concessions, including an annual payment of 2,000 lbs of gold and an imperial generalship for their leader Theoderic Strabo (‘the squinter’) – a mark of status much valued by barbarian leaders in this period because of the way it enhanced their authority among their own retainers. In the meantime, however, the other group of Goths had taken advantage of Leo’s preoccupation with this revolt to advance from Pannonia to Macedonia, where the threat they posed persuaded Leo to grant them land on which to settle and farm.26
As the chief beneficiary of Aspar’s demise, Zeno could hardly expect Theoderic Strabo to view him favourably, so it is no surprise to find Strabo and his Goths supporting Basiliscus’ usurpation in 475 (p. 100 above). During his enforced exile in Isauria, however, Zeno had begun to develop links with the other group of Goths as a counter-weight, and upon his regaining power in 476, he transferred the privileges previously enjoyed by Strabo and his Goths to this other group and their leader Theoderic the Amal.27 This Theoderic was to be a figure of increasing importance in the decades to come, and it is logical to link his success, at least in part, to the formative experiences of his youth. He was the son of Theodemer, one of the leading Goths during the years immediately after Attila’s death, and in the early 460s, at the age of seven or eight, Theoderic had been sent to Constantinople as a hostage, to guarantee an agreement between Leo and his father. The status of hostage in antiquity did not usually entail the degree of close constraint and deprivation associated with modern usage of the term, and Theoderic spent the next ten years in the imperial capital living a lifestyle appropriate to his elite status and receiving a formal Roman education in Greek (and perhaps Latin) language and literature – a process Theoderic must have viewed favourably, since a later panegyrist believed he would win his approval with the observation that ‘Greece educated you in the lap of civilization (civilitas)’ (Ennodius, Pan. 3). Theoderic will therefore have been well known at the imperial court and have developed a good understanding of imperial politics and culture, as indeed he himself acknowledged in a later letter to Zeno’s successor: ‘With divine help I learned in your republic the art of governing Romans with equity’ (Cassiod. Var. 1.1).28
Zeno did not, of course, transfer Strabo’s privileges to Theoderic the Amal without a quid pro quo – namely, that the Amal and his Goths act against Strabo and his Goths. However, when the Amal came increasingly to suspect that Zeno was simply playing the two groups of Goths off against one another, he broke off ties with the imperial court, forcing Zeno to reach an accommodation with Strabo (478). That accommodation, however, proved very short-lived, when Strabo backed the attempted usurpation of Marcian the following year (p. 100 above). After its failure, Strabo tried again to capture Constantinople in 481 on his own account, but was repulsed and, fortuitously for Zeno, died soon after in a freak accident. Although this might have been the cue for Zeno to focus his attention on the elimination of the other Theoderic, his developing difficulties with Illus in the early 480s (p. 100 above) prompted him instead to renew his association with Theoderic the Amal, albeit at the cost of granting the latter not only an imperial generalship, but even the consulship for the year 484. While the latter was certainly ‘an unprecedented honour for a barbarian leader’,29 its novelty was no doubt tempered by knowledge of the Romanisation which Theoderic had undergone in his youth.
From his new position of strength, Theoderic proceeded to eliminate Strabo’s son and successor Recitach, thereby paving the way for the amalgamation of Strabo’s Goths with his own, but although he did assist Zeno in the early stages of his response to Illus’ rebellion, Theoderic himself turned against Zeno in 486, perhaps fearing more double-dealing from the emperor. When Theoderic moved to attack Constantinople itself in 487, Zeno averted the immediate danger by buying him off with gold, but it was not until the following year that a definitive solution emerged when Theoderic decided to accept Zeno’s proposal that he go west and seek greener pastures in Italy by overthrowing Odoacer, whose formal status had remained anomalous since 476 (p. 97 above) and who had been in communication with Illus during his revolt. Of course, as far as Zeno was concerned, it did not really matter which of the two won the ensuing contest: either way, Theoderic and his Goths would be removed from the Balkans and, Zeno hoped, the complexities of affairs in the eastern empire would be simplified immeasurably.30
In the event, it was Theoderic who prevailed. After hard-fought victories in the field against Odoacer near Verona in 489 and Milan in 490, Theoderic besieged him in Ravenna. Eventually, in early 493, shortage of food forced Odoacer to negotiate his surrender, with assurances of his safety and perhaps even shared rule. Within less than a fortnight, however, Theoderic had murdered his rival during a banquet, apparently justifying his treachery with a claim of vengeance on behalf of relatives killed by Odoacer in years gone by.31
It remained to be seen how far Theoderic would follow Odoacer’s precedent with regard to maintaining Roman traditions and respecting the rights of the senatorial aristocracy (Ch. 9 below).
The post-imperial west within which Theoderic had to establish himself involved various barbarian groups who have previously featured – Vandals in north Africa, Goths and Burgundians in Gaul, and Sueves in Spain. There had, however, been an important new addition in recent decades – the Franks. Another Germanic-speaking group, they had been a presence on the lower Rhine from the later third century and received attention during the campaigning of the emperors Constantine and Julian in fourth-century Gaul. It may have been in the context of diplomatic dealings between the empire and the Franks during the fourth century that the Romans planted the idea that the Franks shared a common origin with them from the Trojans, an idea which found elaboration in later sources from the seventh century.32 In the second half of the fourth century, Franks (like other barbarians) held high rank in the Roman army, notably Silvanus, Bauto, Richomer and Arbogast. There is also evidence for Franks serving as rank-and-file soldiers, most famously the unnamed individual commemorated on an epitaph as ‘a Frankish citizen [and] a Roman soldier in arms’.33
While Goths, Vandals and Burgundians were active in Gaul in significant ways at various points during the first half of the fifth century, Franks barely register in the (admittedly patchy) sources. During the second half of the century, however, they become increasingly prominent, notably in the context of the last outpost of Roman authority in northern Gaul associated with the general Aegidius and his son Syagrius. Although Aegidius was unable to prevent Franks from seizing control of Cologne and Trier in the mid-450s, he had, by the early 460s, gained their support as allies against the expansionist Goths in Aquitaine, with the Frankish leader Childeric helping Aegidius to defeat the Goths in a battle at Orleans in 463. Following Aegidius’ death in 465, Syagrius increasingly exercised independent power from his base at Soissons – he had little choice, given the turmoil in Italy (pp. 94–6 above) – so that one source later referred to him as ‘king of the Romans’ (Gregory of Tours, Hist. 2.27). In 482 Childeric (about whose activities nothing else is known) died and was buried at Tournai, where what appears to have been his grave was discovered in the seventeenth century, complete with rich grave goods including a seal ring bearing his name and the title rex (‘king’) (most of these finds were, unfortunately, subsequently stolen).34 His son and successor, Clovis, soon began to pursue a more aggressive, expansionist policy in northern Gaul, resulting in the defeat and death of Syagrius in 487, and setting the stage for the conflict between Clovis’ Franks and the Goths in southern Gaul which was to be so important in defining the future of France (pp. 178–80 below). From the perspective of the empire, however, the death of Syagrius was significant as extinguishing the last vestige of Roman authority in the west.35
Over the course of the fifth century, then, imperial power in the west contracted as various barbarian groups established themselves in different regions. The Gothic groups led by Alaric and Theoderic played major roles in these developments, but it was the Vandal occupation of north Africa, with its economic resources, which proved to be most significant. While the wealthy east was also an attractive target for barbarian groups, geography and other fortuitous circumstances combined to make the eastern empire less vulnerable. Egypt, the wealthiest region in the empire (pp. 233–4 below), was never exposed to serious external threat during the fifth century, while the empire’s most powerful neighbour, Persia, was preoccupied with problems along its own northern frontier for much of this period, leaving the other eastern provinces in peace.36 This allowed Constantinople to focus on maintaining the lower Danube frontier, while also trying to provide aid to the west. As a result, it was inhabitants of the western half of the empire who were more likely to find themselves having to deal at first hand with the consequences of these significant and often rapid changes, as explored in the final section of this chapter.
Romans and barbarians in a changing world
For many ordinary inhabitants of the Roman empire in the fifth century, their encounter with barbarians obviously occurred in the context of warfare, whether it be barbarian raiders plundering farm-steads or more organised assaults on urban centres (cf. Fig. 10). Although Roman commentators on such events were often unable to resist the rhetorical opportunities which they offered for hyperbole and apocalyptic drama, there can be no doubting the significant loss or destruction of property and the incidence of violence, rape and death at the hands of foreigners in vulnerable parts of the empire during this period.37 One particular dimension of this which involved ongoing interaction with barbarians was enslavement, the frequency of which is reflected in the increasingly common involvement of bishops in organising the ransoming of captives.38 Interestingly, however, not all prisoners of war were keen to escape barbarian control. The historian Priscus, who participated in an embassy to the Huns in 449, famously reports encountering in Attila’s camp north of the Danube a former inhabitant of the empire who was apparently very happy to remain where he was. A merchant who had been captured by the Huns when they sacked the Danubian town of Viminacium in the early 440s, he had become the slave of one of the leading Huns, and after proving his valour in battle and gaining much booty, he was given his freedom, married a barbarian woman, enjoyed a comfortable life and had no desire to return to the empire with its heavy taxes and corrupt system of justice (Priscus, fr. 11.2 [lines 407–53]).39 While it is hard to believe that this was a common experience or attitude, it provides an interesting caveat against making blanket assumptions.40
For those within the empire who retained their lives and freedom, there remained a variety of less violent contexts in which interaction with barbarian incomers occurred. As already noted (p. 115 above), some barbarian groups were induced to provide military resources for the empire in the first half of the fifth century, and one consequence of this was that landowners sometimes found themselves having to provide temporary billets for barbarian soldiers. Paulinus of Pella, a grandson of the eminent Gallic aristocrat Ausonius, referred to Goths billeted on landowners in the Bordeaux district who protected their hosts and their property against destruction by other, independent Goths in 414, while his own estate, which lacked a Gothic billet, was pillaged (Eucharisticon 281–90). In the middle of the century, another Gallic aristocrat, Sidonius Apollinaris, ruefully remarked on the experience of having to host a group of Burgundian soldiers in his house:
[I find myself] situated among a lank-haired band of soldiers, having to endure conversation in Germanic, maintaining a serious expression while repeatedly complimenting the singing of a Burgundian glutton, his hair smeared with rancid butter … You don’t have to put up with the stink of garlic and onion emanating from ten breakfasts first thing in the morning … or with the pre-dawn invasion of so many oversized giants that even the kitchen of Alcinous would struggle to cope! (Carm. 12)
Fortunately for Sidonius, billeting was only a temporary arrangement, but there are also references in the sources which suggest that barbarian groups in the fifth-century west received land on a more permanent basis. This is a subject with important implications for the character of Roman–barbarian relations in this period which continues to generate lively academic debate – a reflection, at least in part, of the difficulties of the extant sources. The seminal work in this debate is Walter Goffart’s 1980 study Barbarians and Romans, AD 418–584: The Techniques of Accommodation.41 An older strand of scholarship had proposed a link between, on the one hand, late Roman laws on the billeting of soldiers, which referred to troops being given a third of a house (on a temporary basis), and, on the other, provisions in early barbarian law codes which mentioned the division of land into thirds, and on this basis it was suggested that barbarian groups in the fifth-century west were granted thirds of estates in regions where they settled as a sort of extension of billeting principles.
Figure 10 Wooden carving of a besieged city being relieved (45 × 22 cm) (Egypt, early fifth century). Roman infantry approach from the left, driving away mounted barbarians who have been besieging the city, while the defending garrison lines the top of the wall. The buildings of the city rise up behind them, with three large figures on the left, who are perhaps protective saints. © bpk/Skulpturensammlung und Museum für Byzantinische Kunst, SMB/Jürgen Liepe
Although not without its critics, this view gained broad acceptance until the publication of Goffart’s study, which was particularly significant for two reasons.42 First, he showed that the posited link between Roman billeting practice and barbarian settlement was untenable: essentially, billeting involved temporary use, whereas fifth-century arrangements were more permanent. Secondly, and most significantly, he argued that the references to ‘thirds’ in the context of barbarian settlement related not to ownership of land, but to receipt of the taxes due from that land. If right, this made a significant difference as to who bore the brunt of the costs of barbarian settlement, with further implications for the nature of Roman–barbarian relations. If barbarian groups were given land, then it was landowners who lost out, with resulting resentment towards the incomers and the government; if, on the other hand, the barbarians received a portion of the taxes on land, then it was the imperial treasury which suffered. Imperial finance officials would not be happy, but they were unlikely to protest.
One of the attractions of Goffart’s thesis is that it helps to explain the surprising lack of evidence for complaints by landowners which one might otherwise have expected if the settlement of barbarians had involved expropriation of land on a large scale.43 At the same time, there has remained for many scholars an important obstacle to accepting Goffart’s thesis – namely, that a number of sources seem to refer unequivocally to fifth-century barbarians receiving land.44
Goffart has more recently sought to address this objection by arguing that the terms used for ‘land’ in the sources could have an array of connotations, ranging from ownership of land, to rental of land, to the fiscal liability of land, and that the sharp contrast drawn between ‘land’ and ‘taxes’ in many discussions of his thesis sets up a false opposition.45 However, while recognition of the ‘multivalency’ of the term ‘land’ allows a more nuanced approach to the issue, it does not necessarily follow that every reference to ‘land’ in this context must refer to its fiscal liability, as Goffart would apparently have it.46 Goffart also argues for a consistent application of the same principles in all the different regions of the west where barbarians were settled in the fifth century – in his own formulation, ‘one size fitted all’.47 But a good case has recently been made for the possibility that arrangements varied from context to context, depending upon specific circumstances. By way of illustration, the earliest instance, the establishment of Goths in south-west Gaul in 418, took place at a time when imperial authority still carried weight, and the Roman general Constantius who oversaw the arrangements is unlikely to have instigated large-scale expropriation of landowners; in these circumstances, reassigning of tax revenues from land is a plausible scenario. The Vandals, on the other hand, acquired control of north Africa by conquest, in which situation the seizure of estates was much more likely to have occurred, perhaps supplemented by other arrangements. However, whereas they were not concerned to conciliate estate-owners (many of whom were absentee landlords anyway), it was much more important for Theoderic in Italy to win over the senatorial aristocracy (cf. p. 182 below), and so an arrangement involving tax revenues could well have carried greater appeal there.48 This more flexible approach offers a better prospect of doing justice to the changing circumstances of the period and the difficulties of the evidence, while also taking some account of Goffart’s insights.
A rather different issue which is also relevant to the subject of Roman–barbarian relations in this period is that of religion. Most of the major barbarian groups which settled within the empire during the fifth century had converted to Christianity, but they usually adopted the Arian version at odds with the Nicene Christianity which prevailed within the empire from the reign of Theodosius I onwards.49 It is ironic that a step which was presumably intended to aid the integration of barbarian incomers should have ended up being a potential source of division. The Goths are the only group about which any detail is known concerning the circumstances of their conversion, and since it occurred around the time of their entry to the empire in 376, it is unsurprising that it was the Arian version favoured by the Emperor Valens which they embraced (p. 33 n.29 above). On the other hand, it is unclear why other groups such as the Vandals who entered the empire rather later, when Arianism was no longer in receipt of imperial favour, also adopted Arian Christianity, unless perhaps they were simply following the Gothic example.50 This divergence in creeds certainly had the potential to create barriers between barbarian immigrants and provincial inhabitants, especially if the barbarian leadership adopted an aggressive stance towards Nicene Christians, as was the case with the Vandals in north Africa. Even if much of the initial Vandal aggression was actually more to do with acquiring the church’s wealth than with an excess of religious zeal,51 it will still have militated against integration. On the other hand, there is much less evidence for open religious hostility from and towards the Goths in Gaul, which may have allowed greater scope for the slow but steady impact of another, integrative factor.
That factor was intermarriage, a high-profile example of which has already been noted – that of the Gothic leader Athaulf to the Roman princess Galla Placidia in 414 (p. 115 above). This instance was of course primarily a political marriage, of which there are further examples during the fifth century,52 but there is also evidence for the phenomenon occurring at lower levels of society where political considerations will have been irrelevant. At the same time, the subject of Roman–barbarian marriages has been seen as problematic because of an imperial law from the early 370s which seemed to forbid any such unions: ‘For none of the provinciales, of any rank or status, may there be a marriage with a barbarian wife, nor may any female provincialis marry any of the gentiles’ (Cod. Theod. 3.14.1). Although this is sometimes interpreted as reflecting official concerns about racial mixing or about the risk that such marriages might encourage disloyalty to the empire, the most plausible exegesis has focused on provinciales as a term referring to a particular category of citizen liable to serve on a provincial council and gentiles as a term referring to a specific category of barbarian settler liable to military service – in other words, legal statuses both of which entailed obligations to the state in one form or other. The concern of the law would then be to prevent mixed unions which might confuse the status of any offspring and potentially help them to evade their obligations, whether civic or military – which was undoubtedly a more general anxiety on the part of the late Roman authorities. If the law had this much more restricted focus, it strengthens the likelihood of increasingly widespread intermarriage between Romans and barbarians during the fifth century, with all that that implies for the growing integration of barbarians residing in imperial territory.53
1. For more detailed overviews of the often complex developments treated here, see P. Heather, The Fall of the Roman Empire, London: Macmillan, 2005; G. Halsall, Barbarian Migrations and the Roman West, 376–568, Cambridge: Cambridge University Press, 2007.
2. For discussion, see J. Matthews, Western Aristocracies and Imperial Court, AD 364– 425, Oxford: Clarendon Press, 1975, 270–3; P. Heather, Goths and Romans 332–489, Oxford: Clarendon Press, 1991, 199–208.
3. N. Christie, From Constantine to Charlemagne: An Archaeology of Italy, AD 300– 800, Aldershot: Ashgate, 2006, 325–6.
4. Heather, Goths and Romans, 206–8.
5. For differing emphases on the factors involved, see A. Gillett, ‘Rome, Ravenna and the last western emperors’, PBSR 69 (2001), 131–67, 159–65; Christie, Constantine to Charlemagne, 332–3; D. M. Deliyannis, Ravenna in Late Antiquity, Cambridge, Cambridge University Press, 2010, 46–8.
6. The proposal to advance the crossing to the end of 405 (M. Kulikowski, ‘Barbarians in Gaul, usurpers in Britain’, Britannia 31 [2000], 325–45) has been vigorously challenged by A. R. Birley, The Roman Government of Britain, Oxford: Clarendon Press, 2005, 455–60, and N. McLynn, ‘Poetic creativity and political crisis in early fifth-century Gaul’, JLA 2 (2009), 60–74, at 61 n.3.
7. It was around this time that Honorius is reported to have ‘sent letters to the cities of Britain, urging them to fend for themselves’ (Zos. 6.10.2), thereby signalling the formal end of Roman rule there. For further discussion, see D. Mattingly, An Imperial Possession: Britain in the Roman Empire, 54 BC–AD 409, London: Allen Lane, 2006, 529–39.
8. For a superb, more detailed account of the complex twists and turns of the period 408–10, see Matthews, Western Aristocracies, 284–306.
9. Matthews, Western Aristocracies, 301; M. Kulikowski, Rome’s Gothic Wars from the Third Century to Alaric, Cambridge: Cambridge University Press, 2007, 177.
10. For discussion of its genesis, see G. O’Daly, Augustine’s City of God: A Reader’s Guide, Oxford: Clarendon Press, 1999, 27–33.
11. Details in B. Lançon, Rome in Late Antiquity (tr. A. Nevill), Edinburgh: Edinburgh University Press, 2000, 39.
12. For the major debate which this settlement (and that of other barbarian groups) has generated, see below, pp. 128–32.
13. For further detail on the events of this paragraph, see Matthews, Western Aristocracies, 307–28, Heather, Goths and Romans, 219–24.
14. P. Rousseau, ‘Inheriting the fifth century: who bequeathed what?’ in P. Allen and E. M. Jeffreys (eds), The Sixth Century: End or Beginning?, Brisbane: Byzantina Australiensia 10, 1996, 1–19, at 4–5.
15. For many years, the most detailed study of the Vandals has been C. Courtois, Les Vandales et l’Afrique, Paris: Arts et métiers graphiques, 1955, but see now A. Merrills and R. Miles, The Vandals, Oxford: Wiley-Blackwell, 2010, who emphasise the Vandals’ lacklustre military record prior to 422 (50–1).
16. J. B. Bury, History of the Later Roman Empire from the Death of Theodosius I to the Death of Justinian, London: Macmillan, 1923, 1.246.
17. Merrills and Miles, Vandals, 52.
18. For the fiscal significance of the loss of north Africa, see C. Wickham, Framing the Early Middle Ages: Europe and the Mediterranean, 400–800, Oxford: Oxford University Press, 2005, 87–8.
19. For good introductions to Hunnic society, see J. Matthews, The Roman Empire of Ammianus, London: Duckworth, 1989, 332–42, and C. Kelly, Attila the Hun: Barbarian Terror and the Fall of the Roman Empire, London: Bodley Head, 2008, 7–54, 238–48.
20. Lying 1.5 km to the west of the Constantinian wall (with substantial portions still extant), they increased the city’s area by 5 km2, reflecting its population growth, but their defensive priority is evident in their height (11 m), thickness (4.8 m), and provision of towers (96, c. 70 m apart) and advance wall and trench (details in W. Müller-Wiener, Bildlexikon zur Topographie Istanbuls, Tübingen: Wasmuth, 1977, 286–7).
21. The episode is briefly described in Olympiodorus, fr. 19, with some scholars arguing that Donatus was a Roman refugee, rather than a Hunnic leader; for discussion of this episode and Roman double-dealing more generally, see A. D. Lee, ‘Abduction and assassination: the clandestine face of Roman diplomacy in late antiquity’, International History Review 31 (2009), 1–23.
22. Details (with sources) in G. Greatrex and S. Lieu, The Roman Eastern Frontier and the Persian Wars, AD 363–630, London: Routledge, 2002, 36–45.
23. For important revisions to the traditional chronology of the 430s and 440s, see C. Zuckerman, ‘L’Empire d’Orient et les Huns: notes sur Priscus’, T&M 12 (1994), 160–82, at 160–8; for a good overview of the basis of Attila’s power, see M. Whitby, ‘The Balkans and Greece 420–602’, CAH2 14.701–30, at 704–12.
24. Attila was, however, clearly well informed about western affairs via a steady stream of envoys, merchants and others: Rousseau, ‘Inheriting the fifth century’, 14.
25. M. F. Hendy, Studies in the Byzantine Monetary Economy, c. 300–1450, Cambridge: Cambridge University Press, 1985, 223.
26. Full details and discussion in Heather, Goths and Romans, 240–71.
27. Heather, Goths and Romans, 272–8.
28. For a good discussion of this case as part of a wider phenomenon, see J. Shepard, ‘Manners maketh Romans? Young barbarians at the emperor’s court’ in E. Jeffreys (ed.), Byzantine Style, Religion and Civilization, Cambridge: Cambridge University Press, 2006, 135–58, who notes an important limitation to Theoderic’s education, which evidently did not succeed in converting him from Arian to Nicene Christianity (although it may have been responsible for his later moderation in religious policy). See also J. Moorhead, Theoderic in Italy, Oxford: Clarendon Press, 1992, 13–14.
29. Heather, Goths and Romans, 300.
30. For further details on the complicated events of this and the previous paragraph, see Heather, Goths and Romans, 278–308, with further valuable comment on the background to Theoderic’s decision to invade Italy in Moorhead, Theoderic, 17–19.
31. Moorhead, Theoderic, 19–31.
32. I. Wood, The Merovingian Kingdoms, 450–751, Harlow: Longman, 1994, 33–5. Cf. Sid. Apoll. Carm. 7.501–2, where, in a speech delivered at Rome in 456, Sidonius claimed that the Romans and Goths shared a common ancestor in the person of Mars.
33. CIL 3.3576 = ILS 2814, from Aquincum on the middle Danube. For Frankish units in the fourth-century Roman army, see James, Franks, 39.
34. For Childeric’s tomb, see S. Lebecq, ‘The two faces of King Childeric: history, archaeology, historiography’ in W. Pohl and M. Diesenberger (eds), Integration und Herrschaft: Ethnische Identitäten und soziale Organisation im Frühmittelalter, Vienna: Österreichischen Akademie der Wissenschaften, 2002, 119–32.
35. For the early history of the Franks, see E. James, The Franks, Oxford: Blackwell, 1988, 34–84, and Wood, Merovingian Kingdoms, 35–41; for a significantly different chronology, see G. Halsall, ‘Childeric’s grave, Clovis’ succession and the origins of the Merovingian kingdom’ in R. Mathisen and D. Shanzer (eds), Society and Culture in Late Roman Gaul, Aldershot: Ashgate, 2001, 116–33.
36. Apart from the short-lived Roman–Persian conflicts in northern Mesopotamia in 421–2 and 440.
37. B. Ward-Perkins, The Fall of Rome and the End of Civilization, Oxford: Oxford University Press, 2005, 13–31; A. D. Lee, War in Late Antiquity: A Social History, Oxford: Blackwell, 2007, 133–46 (with 167–73 for provincials sometimes experiencing violence at the hands of Roman troops).
38. W. Klingshirn, ‘Charity and power: Caesarius of Arles and the ransoming of captives in sub-Roman Gaul’, JRS 75 (1985), 183–203; C. Rapp, Holy Bishops in Late Antiquity, Berkeley: University of California Press, 2005, 228–32.
39. While Priscus’ subsequent discussion with the man about the relative merits of life under imperial and barbarian rule has echoes of a formal rhetorical exercise, this need not cast doubt on Priscus’ report of the encounter itself.
40. For other evidence, see E. A. Thompson, ‘Barbarian invaders and Roman collaborators’, Florilegium 2 (1980), 71–88.
41. Goffart has more recently responded to critics and restated and refined his views in Barbarian Tides: The Migration Age and the Later Roman Empire, Philadelphia: University of Pennsylvania Press, 2006, 119–86, and in ‘The techniques of barbarian settlement in the fifth century: a personal, streamlined account with ten additional comments’, JLA 3 (2010), 65–98 (the latter particularly prompted by the next item).
42. For a clear overview of earlier scholarship and a careful assessment of Goffart’s 1980 study, see Halsall, Barbarian Migrations, 422–47, supplemented by his ‘The technique of barbarian settlement in the fifth century: a reply to Walter Goffart’, JLA 3 (2010), 99–112.
43. Goffart, Tides, 183–6, summarises the half dozen instances of complaints by property-owners against barbarians, arguing that their detail does not support an assumption of widespread dispossession.
44. Halsall, Barbarian Migrations, 430–1.
45. Goffart, Tides, 126–7.
46. Halsall, ‘Reply’, 107 (‘multivalency’ is his term).
47. Goffart, ‘Streamlined account’, 93.
48. See Halsall, Barbarian Migrations, 436–47, for further detail, with Halsall, ‘Reply’, for his entertaining response to Goffart’s critique; also Moorhead, Theoderic, 32–5, for the specific situation in Italy.
49. For the major exception of the Franks (whose conversion to Nicene Christianity did not occur until the early sixth century), see p. 189 below.
50. For the Vandals, see P. Heather, ‘Christianity and the Vandals in the reign of Geiseric’ in J. Drinkwater and B. Salway (eds), Wolf Liebeschuetz Reflected, London: Institute of Classical Studies, 2007, 137–46, at 143; for an overview of the inconclusive argument about the Ostrogoths (i.e., the Goths whom Theoderic the Amal led to Italy), see Moorhead, Theoderic, 89–90.
51. Merrills and Miles, Vandals, 181.
52. Moorhead, Theoderic, 84–5.
53. R. Mathisen, ‘Provinciales, gentiles, and marriages between Romans and barbarians in the late Roman empire’, JRS 99 (2009), 140–55.