Figure 2.1 The artistic patronage of the late Roman aristocracy: a marriage casket adapted for a Christian couple, Proiecta and Secundus, from the Esquiline treasure, Rome, late fourth century. Trustees of the British Museum
2
The questions dealt with in this chapter have been transformed by a deluge of publications in recent years. Attention has shifted from the problem of the late Roman army and its use of federate troops to the barbarians themselves, the impact on them of their contacts with Rome, the development of concepts of ethnicity as they came under Roman influence, Rome’s usually unsuccessful, or at least short term, attempts to deal with the issue, and the gradual emergence of discrete groups and eventually of the early medieval kingdoms. Modern studies of migration and ethnic identity are also influential, and have been used as a corrective to the old stereotypes of barbarian invasion, based on prejudiced Roman sources, and sharp distinctions between civilized Romans and the barbarian ‘other’.1 These newer approaches also imply a critical reading of the very influential Getica, the mid-sixth century Latin work written in Constantinople by Jordanes, which employs the familiar technique of genealogy to claim for the Goths a mythic beginning in Scandinavia before their encounter with the Roman empire.2 Equally, and in contrast with much earlier scholarship, it has now been suggested that the overall ethnic term ‘Germanic’ is better avoided. On the newer and sceptical model, the barbarian groups did not come with their ethnicity ready made but developed it under the influence of contacts with Rome.3 A similar process is argued to have taken place in relation to the Arab federates in the east in the sixth century (Chapter 9). Rome had indeed long used and integrated barbarians, even while attempting to assert Roman identity over foreigners. Numbers, cultural transfer and identity formation are all key issues now, as are debates about the interpretation of material evidence. Another development has been a new understanding of frontiers and their functions, which has moved away from the old notion of a clear-cut fortified barrier.4 The concept of hordes of barbarians pouring in and overwhelming the empire has been well and truly discarded. In the light of these developments, Roman policy towards the barbarians has also had to be rethought. But as we have seen, these changed approaches have also gone along with a re-assertion of what is often called ‘the fall of the Roman empire in the west’. From the perspective of this book this needs to be placed in the context of greater continuity in the east in late antiquity, and these issues will recur in the following chapters.
Figure 2.1 The artistic patronage of the late Roman aristocracy: a marriage casket adapted for a Christian couple, Proiecta and Secundus, from the Esquiline treasure, Rome, late fourth century. Trustees of the British Museum
The fifth century saw one of the most famous non-events in history – the so-called ‘fall of the Roman empire in the west’, which according to traditional views took place in 476, when the young Romulus Augustulus, the last Roman emperor in the west, was deposed and replaced by Odoacer, a Germanic military leader, who followed Ricimer (457–72) and Orestes (475–6) as power-broker. Odoacer differed from his barbarian predecessors in that he did not attempt to rule through a puppet emperor; he sent an embassy of Roman senators to the Emperor Zeno (474–91) in Constantinople asking to be given the prestigious title of patricius. Zeno had only recently secured his own throne from the serious threat mounted by Basiliscus, and the emperor’s reply was equivocal; the deposed Julius Nepos, who was also now seeking his aid, had been placed on the western throne with eastern support in 473. However, Odoacer satisfied himself with the title rex, and henceforth the only emperor was the eastern emperor in Constantinople.5 Zeno had other problems to contend with, including dealing with two powerful Gothic leaders, Theodoric the Amal and Theodoric Strabo (Chapter 1), but eventually used the former to put down Odoacer (493); Theodoric promptly succeeded him, founding the Ostrogothic kingdom and ruling Italy until 526. The date 476 has traditionally provided a convenient point at which to place the formal end of the Roman empire, and Procopius of Caesarea begins his history of Justinian’s Gothic war (535–54) by recounting the history of Italy from that point. Gradually, though not immediately, the eastern empire came to terms with the fact that it was left alone as the upholder of Roman tradition, and invented its own myths of translatio imperii to justify its new role; this included the claim that the palladium of Rome had been buried under the great statue of Constantine in Constantinople.6 But the year 476 has no significance in the context of the economic and social changes that were taking place in the period; it is doubtful whether even the population of Italy at first noticed much difference. The changes which were taking place were long term and multiple, part of a gradual process which ate away at Roman territory in the west through settlement and force of arms, and which made it increasingly difficult for the Roman government in the west to field an adequate army or prevent the erosion of its tax-raising powers;7 the loss of the rich provinces of North Africa was very serious, and in Gaul and elsewhere the local elites often had no choice but to make their own accommodation with the settlers. Sidonius Apollinaris, bishop of Clermont-Ferrand in the 470s, and himself from the landowning class, exemplifies the dilemmas which now faced the old Roman elite, caught between trying to maintain a remembered lifestyle while coming to terms with their barbarian neighbours.8 The Life of Severinus of Noricum (Austria) by Eugippius, set in the 470s, depicts a kind of no man’s land where previous structures had broken down.9 Eastern emperors tried on several occasions during the mid-fifth century to intervene in western affairs, but this became less and less feasible as time went on, so that the eventual invasion of Italy under Justinian seems extraordinary, if not even quixotic (Chapter 5). In political terms, the fall of the feeble Romulus Augustulus was entirely predictable. But identities were complex, especially after the death of Attila in 451.10 Odoacer was the son of Edeco, a Hun or Thuringian, leader of the Sciri and one of Attila’s close allies. But he was only one in a long line of generals who had held the real power in the western empire since the late fourth century. When one of the first and most powerful of these, Stilicho, the Vandal magister militum of Theodosius I and regent for his son Honorius, fell in 408, suspected of treason (above, Chapter 1), he was succeeded by Romans in the high positions of magister utriusque militiae and patricius; but real power still lay with barbarian generals, in particular Aetius (c. 433–54). After the murder of Valentinian III in 455, his successor Avitus, a Gallic senator, was defeated by the Sueve Ricimer and an uneasy period followed before Majorian was officially proclaimed emperor in 457, only to be killed by the same Ricimer four years later. Again Ricimer was kingmaker, but his undistinguished choice, Severus, who had not been ratified by the eastern emperor Leo, died in 465, again leaving the west without an official ruler. When Leo imposed Anthemius, his own choice, the rivalry between Anthemius and Ricimer became first a scandal and then the occasion for open hostilities, in the course of which Anthemius was killed (472). Ricimer’s final choice for emperor was Olybrius, the Roman husband of Valentinian III’s daughter Placidia (see Chapter 1); but both Olybrius and Ricimer died before the year was out. The nominee of the Burgundian Gundobad was deposed by Julius Nepos with the encouragement of the Emperor Leo, only to be deposed in his turn in favour of the ill-fated Romulus Augustulus. It is a dreary and confused story, in which the principal players vary between barbarian or Roman commanders and members of the civilian aristocracy, with the eastern emperor invoked at times for the sake of respectability and at times attempting to interpose his own choice. Only occasionally did these power struggles at the top have a direct impact on government; Majorian (457–61), for instance, issued reforming legislation, but soon fell at the hands of Ricimer. There was no western Leo or Zeno. No western emperor had succeeded in establishing strong government after the death of Theodosius I, and while the eastern government in the later fifth century under Marcian and Anastasius had become progressively more civilian in style, the exact opposite happened in the west. Nor could the western government be said to represent strong military rule; on the contrary, both the territories occupied by the western empire and the Roman army had itself by now suffered fragmentation on a major scale. These processes are closely interconnected, with roots reaching back to the fourth century, but they will be treated separately here for the sake of clarity.
Romans and Barbarians: the Late Fourth Century Onwards
We have inherited a dramatic view of the Roman empire in the west as being submerged by successive waves of northern barbarian invaders. In fact, interaction with peoples from beyond the Rhine and the Danube had been a fact of life since the Marcomannic wars of Marcus Aurelius in the late second century, and indeed earlier. Until the mid-fourth century, however, it had generally been possible to contain them by a judicious deployment of force and diplomacy. These were settled peoples with social hierarchies. The arrival on the scene in 376 of the Huns, a nomadic people perhaps originating in the steppe-lands of modern Khazakhstan,11 was a decisive moment on any view; as we have seen, Ammianus believed that it forced the Tervingi and Greuthingi to cross the Danube into Roman territory and led to their settlement in the Balkans by the Emperor Valens.12 The Greek and Roman sources depict the event in lurid colours, but the Goths were neither a terrified rabble nor part of a great wave of invaders sweeping over the Roman empire. Complex social and economic factors lay behind their appearance in later Roman history, and when they came, they came as an organized military force. Only two years later came the battle of Adrianople (378), a blow that Rome never forgot. The Roman defeat was the signal for other barbarian leaders to cross into Roman territory. Alaric and his Visigoths entered Italy in 401, were defeated by Stilicho in 402, but returned in 408 to sack Rome two years later (Chapter 1). In 405 a certain Radagaisus collected a large barbarian army from across the Rhine and Danube and invaded Italy; on his defeat by Stilicho, 12,000 of them were enrolled in the Roman army.13 But from then on groups of Alans, Vandals and Sueves were on the move across Germany and Gaul and into Spain and almost at the same time the usurper Constantine moved from Britain into Gaul.14 The numbers involved are hard to assess on the basis of the patchy sources. Heather estimates the size of Radagaisus’ force alone at 20,000 fighting men, which implies 100,000 including the women, children and others who travelled with them.15 The story is complex, and the course of events confused by rivalries between different groupings, not to mention the problems presented by the sources. By the late 420s, however, as we saw in Chapter 1, the Vandals under Gaiseric crossed the Straits of Gibraltar into North Africa, reached Augustine’s see at Hippo by 430 and took Carthage in 439. They were able to sack Rome in 455 and take Sicily in 468. The situation in the northern provinces was less clear-cut, and unlike the Vandal occupation in North Africa, did not cut off the critical food supply to the city of Rome; nevertheless Roman government and defence were crucially eroded. In the difficult conditions of the first decade of the fifth century, Zosimus tells us that the defence of Britain was formally abandoned by Honorius: ‘Honorius sent letters to the cities in Britain, urging them to fend for themselves.’16 Some of the troops in Britain, who had apparently supported usurpers before 406, remained in the province, but there was no longer a central authority, and Saxon raids now exacerbated the already confused situation. The rapid disappearance of Roman towns in Britain after several centuries of Roman rule is only one of the many puzzling features of the period.17 In mainland Europe, the fifth century saw a lengthy jostling for position as different groups competed against each other and with Rome for land and influence. The west suffered more than once from the greater ability of the east to avert the danger by financial and diplomatic means, most conspicuously in the case of Alaric and the Visigoths, who were allowed by the eastern government to build up their strength in the Balkans, only to use it against Italy, demand large amounts of gold and silver and eventually sack Rome.18 The sack itself, while perhaps not as destructive as it might have been, came as an enormous psychological blow to Christians and pagans alike, and caused many leading members of the Roman aristocracy to flee. But the chance event of Alaric’s own death shortly afterwards, like that of Attila in a similar situation later, saved Rome from the possibility of long-term occupation. The Visigoths moved north under Athaulf and eventually ended up in Aquitaine, after a series of confusing episodes during which Galla Placidia married successively Athaulf and Constantius, magister militum and co-emperor in 421, and gave birth to the future Valentinian III (425–55).19 The aftermath depended on the changing configurations of barbarian groupings, and their respective success in dealing with the imperial government (and vice versa). Various means were used. Further settlements were made c. 440 by the magister militum Aetius, of Alans in Gaul and Burgundians north of Geneva. Meanwhile, a new threat was posed by the Hun king Attila, who, having already extracted large subsidies, crossed the Danube in the early 440s, defeated the Roman armies sent against him on two occasions and succeeded in obtaining even higher annual payments of gold. He eventually turned towards the west, accepted the advances of Valentinian III’s sister Honoria and demanded half the empire. The battle between the forces of Attila and Aetius on the Catalaunian Fields which followed in 451 was a temporary check, but did not prevent the Huns from invading Italy.20 Again, the western empire had a lucky escape, for Attila’s death (Chapter 1) brought the break-up of the Hun empire and removed the danger.
Figure 2.2 Composite buckle, Ostrogothic style, inlaid with cloisonné garnets, green glass and shell in gold cellwork, late fifth–early sixth century. Found in a female grave near the gate of the church of St Severin, Cologne. Trustees of the British Museum
From now on, as the western government became progressively weaker, and it became less and less possible to sustain any coherent policy in relation to barbarian settlement. Even in the vacuum left by the death of Theodosius I in 395, Rome still occupied the centre in the shifting game of barbarian movements; by the end of the century no Roman emperor was left in the west, and we can see the first stage in the development of the early medieval kingdoms.21 The first to be established was, as we have seen, that of the Vandals in North Africa. However, it was untypical in that it was overthrown by the imperial armies under Belisarius in AD 534 and replaced by well-established Byzantine rule lasting at least in part until the late seventh century. North Africa, in fact, represents the success story of Justinian’s policy of reconquest; the irony was, however, that in contrast with the long history of Roman Africa before the arrival of the Vandals, the restored imperial province was governed by easterners from Constantinople whose language of administration was Greek (Chapter 5). The longest lasting of the Germanic kingdoms was that of the Franks, established by their king Clovis (481–511) after their defeat of the Visigoths at Vouillé in 507 and lasting until 751. Although it was the Franks who gave their name to modern France. Clovis’ descendants are usually known as the Merovingians.22 They found a vivid chronicler in the late sixth-century bishop Gregory of Tours, whose History of the Franks is our main source, remarkable not least for its unrestrained cataloguing of the bloodthirsty doings of the Frankish royal family.23
Map 2.1 The Mediterranean world, early sixth century
Gregory provides a colourful account of the conversion and baptism of Clovis: the king’s wife Clotild was already a Christian and tried unsuccessfully to convert her husband, but his reaction when her first son died after being baptized was one of anger:
If he had been dedicated in the name of my gods, he would have lived without question; but now that he has been baptized in the name of your God he has not been able to live a single day!
(HF II.29)
The king was finally converted after successfully praying to the Christian God for victory on the field of battle against the Alamanni, and was then baptized by bishop Remigius of Rheims, who, we are assured by Gregory, had raised a man from the dead.24 The scene of the king’s baptism was spectacular:
The public squares were draped with coloured cloths, the churches were adorned with white hangings, the baptistry was prepared, sticks of incense gave off clouds of perfume, sweet-smelling candles gleamed bright and the holy place of baptism was filled with divine fragrance. God filled the hearts of all present with such grace that they imagined themselves to have been transported to some perfumed paradise. King Clovis asked that he might be baptized first by the Bishop. Like some new Constantine he stepped forward to the baptismal pool, ready to wash away the scars of his old leprosy and to be cleansed in flowing water from the sordid stains which he had borne so long.
(HF II.31)
In this mass spectacle, more than 3,000 of his army were said to have been baptized at the same time.
In Italy, the Ostrogothic kingdom founded by Theodoric lasted until 554 when its last king, Teias, was finally defeated by Justinian’s general Narses after nearly twenty years of warfare (Chapter 5). But the arrival in Italy of the Lombards in 568 meant that Byzantine control in Italy was not to last for long, except in a limited (though still important) form from the late sixth until the mid-eighth centuries under an exarch based at Ravenna.25 After 568 the situation in Italy was confused and fragmented, and it was in this period that the popes, especially Gregory the Great (590–604), acquired much of their enormous secular influence and economic power.
Ostrogothic Italy retained many continuities with the Roman past,26 among them the survival of many of the immensely rich and aristocratic Roman families who continued to hold office under the new regime. It is a remarkable fact that the Roman senate survived during the fifth century, through all the political changes, and its members continued to be appointed to traditional offices and to hold the western consulship even under the Ostrogoths; the consulship was in fact ended by Justinian himself in 541.27 Many of these Roman families were extremely wealthy, and Procopius, who describes the Gothic wars in detail from the eastern point of view, particularly identifies with this class, most of whom lost their land and position, and many of whom were reduced to a pitiable state by the Justinianic war, unless they were able to flee to the east where they often also possessed estates.28 Like many others of this class, Cassiodorus, whose highly rhetorical and bureaucratic Latin letters (Variae), many written as Theodoric’s quaestor, are another of our main sources for the period, was one of the prominent Italians who left for Constantinople.29 Before that, he had written a Gothic History, used by Jordanes in his Getica,30 and after the wars ended and he had returned to Italy, the Institutes, a set of precepts on Christian learning, and other theological works. Cassiodorus founded the monastery of Vivarium on his family estate near Squillace, which was to become one of the most important medieval centres for the copying and preservation of classical texts. A traumatic event had taken place in relations between the Ostrogoths and the Roman upper class in Italy in AD 523–4, when Theodoric had unexpectedly turned on and eventually executed two of its most prominent members, Symmachus and Boethius, author of the Latin classic, the Consolation of Philosophy. The case was sensational – Symmachus held one of the most prestigious names among the late Roman aristocracy, while Boethius’ two sons had both been given the consulship and he had been consul himself in 510 and was Theodoric’s magister officiorum. Boethius’ Consolation was written in prison as he mused on his fate; he imagines himself visited by the Lady Philosophy, and engages in extended discussion of fate, free will and the fickleness of fortune, and includes a number of long poems, which are of themselves of great interest.31 But the deaths of Symmachus and Boethius were exceptional; Theodoric seems to have shared the general respect for Roman tradition, and the Ostrogothic regime was not in general oppressive.32
Figure 2.3 Coin of Theodoric the Ostrogoth (d. 526). Trustees of the British Museum
The defeat of the Visigoths by Clovis at Vouillé in AD 507 put an effective end to their kingdom in Gaul, which had had its capital at Toulouse since 418, and to the descent of the Balt dynasty which had ruled since Alaric I at the end of the fourth century.33 In the troubled period which followed, Theodoric, whose daughter had married the son of the Visigothic Alaric II, intervened, and Visigothic rule passed temporarily into Ostrogothic hands. More important in the longer term, however, was the movement of the Visigoths into Spain, which had already happened before the end of the fifth century; there, especially from the time of the Ostrogothic Theudis (431–48), they were to establish a kingdom which, despite some Byzantine success in the context of Justinian’s reconquest, lasted until the arrival of the Arabs in the early eighth century.34
Barbarian Settlement, the Roman State and the Early Medieval Kingdoms
With the establishment of the barbarian kingdoms we pass into the traditional realm of early medieval history. But the continuities are such that it can also be argued that the period up to the later sixth century was still part of a surviving Mediterranean world of late antiquity. Despite the obvious changes in settlement patterns in the west, the available archaeological evidence seems to show that long-distance exchange and travel still went on, even if in reduced form.35 The western kingdoms retained many Roman institutions, and even saw their relation with the emperor in Constantinople in terms of patronage. Their kings received Roman titles, and the former Roman upper classes survived in substantial numbers and adapted themselves in various ways to the new regimes. One who adapted, as we have seen, was Sidonius Apollinaris, of whom Gregory of Tours writes:
He was a very saintly man, and as I have said, a member of one of the foremost senatorial families. Without saying anything to his wife he would remove silver vessels from his home and give them away to the poor. When she found out, she would grumble at him; then he would buy the silver vessels back from the poor folk and bring them home again.
(HF II.22)
Both Gregory of Tours, the historian of the Franks, and Gregory’s contemporary and friend Venantius Fortunatus, himself a Merovingian bishop and the author of Latin poems on political and contemporary subjects, came from this class, as did Pope Gregory the Great.36 Germanic law existed in uneasy juxtaposition with Roman; the Ostrogothic kingdom had one law for the Goths and another for the Roman population, while successive Visigothic law codes, beginning with the Code of Euric (c. 476) and the Romanizing Lex Romana Visigothorum of Alaric II (506), followed by an extensive programme of law-making in the Visigothic kingdom of the sixth and seventh centuries, gradually brought about a unification of the German and the Roman traditions.37 The eastern government pursued a pragmatic policy, knowing that it was in no position to impose a western emperor, but not admitting (or, no doubt, believing) that the current regimes were permanent. When the time came, it was ready to use one against another. The fact that the Goths in Italy, like the Vandals and, at this period, the Visigoths, were Arian was, perhaps paradoxically, a help to imperial diplomacy, for it made it possible to represent Justinian’s invasion of Italy in 535 in religious terms. Seeking aid from the catholic Franks, the emperor wrote:
The Goths have seized Italy, which is our possession, by force, and have not only refused to return it, but have committed wrongs against us which are past endurance. For this reason we have been forced to go to war against them, a war in which both our common hatred of the Goths and our orthodox faith dictates that you should join us, so as to dislodge the Arian heresy.
(Proc., Wars IV.5.8–9)
The imperial rhetoric was backed by gold, and by the promise of more if the Franks agreed; not surprisingly, perhaps, they were not to prove very loyal allies.
In studying the process of barbarian settlement in the territory of the western empire, we must distinguish between formal grants made by successive emperors and governments and the longer process of informal settlement patterns. In practice, a continuous process of settlement reaching back at least to the fourth century had long ago undermined Roman control of the west and, through the use of non-Romans as troops, had eroded any sense in which there could still be a single Roman army. Control of the land, and therefore of tax revenues, was also seriously affected (see below).38 Contemporary literary sources written from the Roman point of view are imbued with anti-barbarian stereotyping, and give only a very imperfect and one-sided picture of the process and extent of settlement, and consequently historians have turned to the evidence from archaeological finds, especially those from graves, as markers of different barbarian ‘cultures’. However, this approach has also been challenged as too simplistic and as methodologically unsound; reading off ethnicity from grave goods can be as deceptive as taking the literary sources at face value.39 The reasons for settlement might vary greatly, from invasion and imperial grants of land to resettlement through service in the Roman army, and it is often difficult to identify the reasons in individual cases. In the same way, it is often impossible to connect known historical events such as invasions, or even in some cases longer-term settlement mentioned in literary evidence, with available archaeological remains. The newcomers often tended to take over the customs of the existing provincial population, making traces of barbarian settlement even harder to detect. There are obvious resonances, even if also differences, with issues of migration and settlement in today’s world. In the present case, despite the many difficulties and controversies surrounding the archaeological evidence, a steady process of small-scale cultural and demographic change had been taking place in the western provinces long before the formation of the barbarian kingdoms as we know them. The scale of this process, with the concomitant economic factors, was such that by the mid-fifth century the former Roman villas in the western provinces had in many cases been abandoned or gone into decline, and the role of the former Roman landowning class been transformed.40 In the western provinces the Roman government was not so much faced with discrete incursions as with a slow but steady erosion of Roman culture from within. The process was not of course understood in these terms by contemporary writers, who paint a lurid picture of Romans versus ‘barbarians’; for this reason contemporary interpretations of highly charged events such as the battle of Adrianople and the barbarian settlements which followed it are particularly liable to mislead.41 The moral and political explanations given in the literary sources are not adequate to explain what was happening on a broader scale, and indeed, most of the long-term changes lay outside government control. Yet it was these changes, rather than any political events, which would in the long run detach these areas from effective imperial rule, and fatally so once that control passed from the hands of a weak western emperor to those of a government in far-away Constantinople.
The impact of this process on the late Roman economy was profound (see Chapter 4).42 But wealth also played a direct role in the empire’s dealings with barbarians in the fifth century in the form of the subsidies paid by the Roman government to various groups, either as reward for quiescence or as inducements to go elsewhere; again modern parallels are striking. Although the eastern government was better placed to make use of this device than the western (Chapter 1), and was still making large payments to some groups in the late sixth century (Chapter 8), the policy proved useful at different times to both.43 When the new emperor Justin II cut off subsidies to the Avars in 566 this was highly provocative, but Justinian’s use of subsidies was scathingly criticized by the conservative Procopius:
On all his country’s potential enemies he [sc. Justinian] lost no opportunity of lavishing vast sums of money – on those to East, West, North and South, as far as the inhabitants of Britain and the nations in every part of the known world.
(Secret History 19)
The convenient practice of using barbarian troops as federates for the Roman army, a prominent feature of this period, was also expensive, and their maintenance could involve money as well as supplies.44 But the amounts of gold that might be involved as subsidies or payments to barbarian leaders were large: in AD 408, for example, Alaric demanded 4,000 lb of gold for his recent operations on behalf of the imperial government in Epirus. The example of Alaric and his Goths also shows how easily clever barbarian leaders could play off east and west. The Visigoths are said to have invaded Italy in 401 because the eastern government had cut off their regular subsidies.45 Why this should have happened is not clear, but Thrace was also threatened at the time by Goths under Gainas and other barbarians who are described as Huns; at any rate, Alaric saw more advantage in moving against Italy, where he was alternately fought and bought off by Stilicho. The latter’s dangerous policy of attempting to buy the service of Alaric and his troops ended when he himself fell in AD 408; but when this happened and Alaric’s demands for payment in return for retreating to Pannonia were rejected, and he besieged Rome (408–9), he fixed the price of movements of food into the city at 5,000 lb of gold and 30,000 lb of silver.46 The cat-and-mouse game continued, and we find Alaric’s successor Athaulf alternately plundering Italy and fighting on the Roman side in Gaul.
When Athaulf became king, he returned again to Rome, and whatever had escaped the first sack his Goths stripped bare like locusts, not merely despoiling Italy of its private wealth but also of its public resources.
(Jordanes, Get. 31)
His marriage to Galla Placidia was another kind of barbarian manipulation:
Then Athaulf set out for Gaul, leaving Honorius Augustus stripped of his wealth, to be sure, yet pleased at heart because he was now a sort of kinsman of his.
(ibid.)
In 418 what remained of the Gothic army of Alaric was settled on Roman land in Aquitaine: ‘they received land in Aquitaine from Toulouse to the ocean’.47 The twenty or more years of plundering, negotiating, bargaining and fighting before the Gothic settlement vividly demonstrate the ambiguities, the cost and the dangers with which the Romans were faced in their attempts to deal with the barbarians. Nor is it clear on what terms the land was granted, or later settlements were made; indeed, they must surely have varied from one case to another.48 The traditional view is that the barbarians, beginning with the Visigoths, were to be entitled to a share of the land on which they were settled, in the surprisingly high proportion of two-thirds to one-third. Examples would be the settlements of Alans and Burgundians in 440 and 443 (Chron. Min. I.660) and Ostrogothic Italy, where, however, the share may have been one-third rather than two-thirds; the rent paid on the share thus received was itself known as ‘thirds’ (tertiae).49 But there are many uncertainties, arising not least from discrepancies among the sources and a lack of hard evidence. A generation ago Walter Goffart proposed a quite different reading of the evidence from the later law codes, according to which it was not the land itself, but the tax revenues from the estates which were divided between barbarians and Romans,50 and this has given rise (and still does) to intense discussion. Controversy surrounds the meanings of the Latin terms hospitalitas and sors, and the evidence is very incomplete; in practice, arrangements probably changed with changing conditions, and while land does seem to have been at issue in the settlement with the Visigoths in 418, it may not have been in the case of the Ostrogoths and does not accord with evidence from Cassiodorus.51 By contrast, there is no evidence from northern Gaul, for example, to tell us about the arrangements which were made there. The reality was surely more varied than has usually been allowed in modern debate, and the fifth century in any case only marked the beginning of a much longer process.52 But while the settlement of barbarians may not have represented an ‘existential struggle’,53 it did spell the end of the Roman empire in the west. The question may indeed be asked why these groups, once settled, did not integrate fully and simply become absorbed. But they had by then begun to develop their own identity, and the answer may be that it was simply too late.
Barbarians and the Late Roman Army
It has often seemed as though it was the Roman army that was spectacularly unable to defend the western provinces; one historian has called his chapter on the fifth century ‘The disappearance of an army’.54 What had happened to the Roman army, and why it does it seem to have been so unsuccessful? Older assumptions were that the recruitment of barbarians into the army was one of the factors that led to poorer performance. This is a factor mentioned in the contemporary sources, together with complaints about weakened frontiers. The latter is usually blamed on a particular emperor – thus the pagan historian Zosimus lays most of the blame on the Christian Constantine. Soldiers are regularly depicted in the sources from the fourth century onwards as debauched, ‘soft’ and undisciplined. The late Roman practice of billeting soldiers in towns often lies behind such criticisms, and indeed in the early empire, citizens of the more peaceful provinces had rarely seen soldiers at first hand, much less experienced their rough behaviour.55 The anonymous author of the treatise De Rebus Bellicis (late 360s) already complains about the high cost of the army and the weakening of frontier defence (De Rebus Bellicis 5), and the soldiers settled on the frontiers known as limitanei are frequently blamed for alleged poor performance.56
The fact that these complaints come in so stereotyped a form indicates that their form has much to do with the prejudices of the contemporary sources. But the army of the late fourth and fifth centuries was certainly different from that of the early empire.57 Many of the changes, such as the stationing of troops in or near cities rather than in large masses on the frontiers, stemmed from the fact that under the reforms of Diocletian and Constantine the late Roman army was paid in supplies as well as in cash: there was a simple need for troops to be near the sources of collection of the taxes in kind which were now among their chief sources of pay. Provisioning the army and paying the soldiers in kind involved enormous logistical efforts, and from the fourth century there was a gradual reintroduction of pay in cash, especially gold, and particularly in the west.58 According to the sources, Diocletian had strengthened frontier defence installations throughout the empire, but studies of the fortress of Lejun in Jordan and elsewhere show that the size of late Roman frontier fortresses and of the legions which manned them was far smaller than in the early empire. The late Roman army was the product of gradual evolution rather than of sudden change, and this evolution arose from a combination of different reasons, though it is true that the effects were felt acutely in the late fourth- and fifth-century west where fragmentation and progressive settlement, with a consequently lower tax revenue, were serious factors. The loss of the North African provinces to the Vandals also had a major impact on the resources of the east. However, by the sixth century, interruptions in army pay were also a constant complaint in the eastern sources, and the government was finding it increasingly hard to keep up numbers; it was able to field only small forces even for its prestige endeavours in Italy. Roman and Persian military dealings in the sixth century were also hampered by the fact that troop numbers in the eastern frontier areas seem to have been reduced (see Chapter 5).59 By this stage also, barbarian bands known as bucellarii had come very near to being the personal retainers of individual generals, and reliance on mounted archers increased, part of a trend towards cavalry which had been taking place gradually over a long period, partly in response to the threat posed by Sasanian heavy-armed cavalry; however, Belisarius in Africa and Narses in Italy in the sixth century both still had a majority of infantry under their command.60
By the fifth century the Roman army included a high proportion of barbarian troops (Elton estimates that one in four soldiers in field armies were non-Roman). The recruitment of barbarians was certainly not new, but from the late fourth century units of barbarian troops had constituted a crucial element in late Roman military organization, even though they are for the most part not listed in the Notitia Dignitatum. This in itself tells us that the Notitia, which gives a set of ‘paper’ figures, is an unreliable guide to the nature of the Roman army as it was in practice. Barbarians could appear in several different guises – as units, in relatively small groups or as individuals enlisted by commanders for individual campaigns. In any of these cases, they had to be paid, whether through the annona, the official distribution in kind to the troops via the tax system, or directly in money and supplies. In the past such barbarian troops had frequently been drawn from outside the empire, but with the process of barbarian settlement a fundamental change took place and they came more and more from within the confines of the empire itself. There is little direct evidence that such troops were any less effective than Romans.61 Some contemporaries naturally thought otherwise, and there was much contemporary concern during the aftermath of the battle of Adrianople; the military treatise of Vegetius, probably of this period, also reflects this conservative view. But if rank-and-file barbarian soldiers were not usually a problem, it was otherwise with powerful individuals, and we have already seen the power gained by individual barbarian generals who rose to hold the highest office of magister militum: Stilicho, Ricimer and Odoacer are the most conspicuous examples, and they posed a rather different threat.62 This too began in the fourth century, and barbarian officers are frequently mentioned in the military narrative of Ammianus covering the years 353 to 378. In the ranks, barbarians held a variety of statuses, including those of laeti and gentiles, both referring to groups of settlers with an obligation to military service, foederati, individually recruited barbarians or units enrolled through treaties,63 and dediticii, prisoners of war from beyond the frontiers. In practice there were probably barbarian troops in all the many different units of the army. Rather than appealing to a general drop in manpower (which is hard to establish), the explanation for this change is probably simply that it was easy. Barbarians were available in large, if not massive, numbers, and utilizing them in the army was a convenient way of deciding what to do with them and, it was hoped, also of neutralizing any capacity they might have for disruption; moreover the process did not interfere with the interests of the landowners who were emerging as more and more powerful in this period.
The Late Roman Army
Barbarian invasion is one of the classic explanations put forward for the fall of the empire. It further implies the ineffectiveness of the late Roman army to contain the situation.64 One issue is that of size: how large an army was at the disposal of the late Roman state? While calculations based on the Notitia Dignitatum (which lists the eastern army establishment c. 394 and the western one of c. 420) are difficult to make, they seem to suggest a size well over 400,000 or even more, depending on one’s interpretation.65 The mid-sixth-century writer John the Lydian gives a figure of over 435,000 (De mens. I.27), and later in the sixth century Agathias gives a total of 645,000 (Hist.V.13). The latter must be much too high even as a paper calculation, and it is simply incredible that the empire could have sustained so vastly increased an army. As we have seen, the Notitia also fails to take into account the very large proportion of barbarian federate troops who actually did much of the fighting, and Agathias admits that by his own day the actual overall size had been reduced to 150,000: ‘whereas there should have been a total effective fighting force of six hundred and forty-five thousand men, the number had dropped during this period to barely one hundred and fifty thousand’ (Hist. V.13). From the fifth century at least, the western government was simply no longer in a position to control the empire by military means. We must therefore conclude that the high figures tell us little or nothing about actual troop deployment; it is more important to understand the fragmentation of the army into several field armies and border commands and the limits on its effective deployment than to rely on overall size.66 Roman urbanism and the Roman presence in the Balkans had suffered a severe decline by the end of the sixth century, Slav invasions brought more insecurity, and towns in the east often preferred even in the Justinianic period to make their own terms with Persian armies; the same pattern was repeated after the failure of Roman troops against the Arabs at the River Yarmuk in 636 (Chapter 9). Similarly, the changes in, or, as Roman writers saw it, the progressive weakening of the frontier system, should also be seen in the contexts of the long-term transformation of local settlement-patterns and of economic and social change. For contemporaries the concept of the frontier was an emotive issue; a simple equation was made between failing to keep up the frontier defences and ‘letting in the barbarians’. Diocletian was remembered for having strengthened the frontiers by the building and repair of forts, Constantine for having ‘weakened’ them by supposedly withdrawing troops into a mobile field army:
Constantine destroyed this security [i.e., Diocletian’s alleged strong frontier defence] by removing most of the troops from the frontiers and stationing them in cities which did not need assistance, thus both stripping of protection those being molested by the barbarians and subjecting the cities left alone by them to the outrages of the soldiers, so that henceforth most have become deserted.
(Zos., New Hist. II.34)
The actual situation was much more complex. Although the literary sources are unsatisfactory and the archaeological evidence hard to assess overall, the latter shows clearly enough the steady development of installations such as watch-towers and fortified stores-bases whose functions included ensuring the supply-system to such forward troops as remained, as well as watching and if possible controlling barbarians within Roman territory. It was now impossible to maintain a defensive line which could really keep barbarians outside the empire, and a variety of local expedients recognized contemporary realities.67 The expedients chosen differed very much from one part of the empire to another, depending on the terrain and the nature of the threat; in northern Gaul a series of coastal forts had gradually come into being over a very long period; in North Africa the so-called fossatum Africae to the south was no help against Vandals arriving from across the Straits of Gibraltar; in the east, where there had never been a fortified line as such, the desert zones on the one hand and the powerful military organization and aggressive policies of the Sasanians on the other, presented a totally different situation. The many defensive installations in the eastern frontier region in the later empire may in any case have been designed not only for defence against invaders from outside the empire but also for the maintenance of internal security.
The apparently successful defence system of the early empire had worked largely because in most areas there had been no serious threat; once, however, nearly all the erstwhile frontier was under pressure there was no serious chance of maintaining it in the same way, and recourse was made of necessity to whatever best fitted local conditions. The change is best seen in the case of the northern provinces, where the old concentrations of force on the Rhine and Danube can now be seen to have been replaced by a fragmented and complex mixture of ad hoc and often unsuccessful defences. In the confused conditions of the fifth century it must often have been difficult to know exactly not only who was defending and who was attacking but also what was being threatened. Political factors compounded the local ones. In the fifth century, when, as we have seen, real power was often held by Germanic military leaders, the official abandonment of Britain by Honorius had been preceded by the suppression of Constantine III, who had been proclaimed by the soldiers in Britain; it was followed by the proclamation of another counter-emperor, Jovinus, at Mainz, whose support seems to have lain among Burgundians, Alans and Franks. In the confused conditions which followed, the elevation in 421 of Constantius, who had defeated Constantine III and married Galla Placidia, marked merely another passing event in a situation in which it must for much of the time have been difficult simply to know who was who.
When for some periods in the west, at least, it is hard to see the Roman army as anything other than a variety of different units without unitary structure or control, it is hardly surprising that the organization, supply and command of the diverse units which made up the late Roman army in the empire as a whole should have proved so difficult. Even if we take a less robust view of the actual numbers of troops, the sheer maintenance of the army can be seen to have posed a variety of problems in the fifth century, of which cost was only one. Once barbarian settlement was allowed and encouraged, the old frontiers no longer even pretended to keep out barbarians in any meaningful sense, while the growing presence of barbarians within the empire, combined with the activities of leaders such as Alaric and Gainas, meant that the army itself was hardly any longer an army of ‘Romans’. Difficulties of recruitment in the face of the mounting power of landowners and their unwillingness to release labour, supply problems and the weakening of government structures, especially in the west, all contributed to make the late Roman army (if one can still describe it in such unitary terms) difficult, and in the west impossible, to maintain and control. Most of the literary sources harked back to supposedly better days which had gone for ever, but the sources also had a rhetoric of their own. When Synesius in Cyrenaica, who had lived with the bitter realities of provincial life for himself, says with tired resignation ‘Pentapolis is dead’, that is one thing;68 but when conservative historians such as Zosimus, or Procopius, who also tended to be the most vocal, fail to understand the depth of the structural change that had taken place, and prefer to lay the blame on moral factors or individuals, we need to be fully aware how far such judgements have been conditioned by the education and cultural background of the writers.
The Erosion of the West
Late antiquity was a time of profound change, and if the impact of the barbarians was at times a matter of military conflict, it was also characterized by a gradual movement and seepage of new peoples into the former imperial territories. Even if we give up the old-established model of enormous numbers of invaders swamping the existing population, the effects of these changes were fundamental. No state in history can survive unchanged for ever; all are dependent on external factors as well as internal ones and so it was with Rome.69 Since the reasons for this continuous migration of northern peoples remain obscure, one might be tempted to conclude that the voluminous historical literature on ‘decline and fall’ has in fact failed in its attempt to explain the end of the Roman empire in the west. But simple explanations are always inadequate for complex historical change. The negative attitudes of the Romans themselves towards barbarians, and their own tendency to see the problem in very black and white terms contributed largely to the problems and made serious integration and acculturation of barbarians more difficult. At the same time the process of barbarian settlement in the western provinces, whether ad hoc or officially encouraged, and the recruitment of barbarian troops in the Roman army, brought profound changes to social, economic and military structures which were in many cases already precarious, the nature of which was not readily understood by contemporaries and which they had few means of controlling. It was not a matter simply of invasion or conflict, but of development and dynamic relations. But we must also remember that the east in the fifth century, even while undergoing similar processes to those in the west, and facing similar dangers, supported a strengthened civilian government and increasingly prosperous economy, and kept its administrative and military structures sufficiently in place to be able to launch offensive wars in the west on a large scale under Justinian; this fact alone should be enough to make us remember the critical importance of context and local differences in explaining historical change.