Version classiqueVersion mobile
OpenEdition Books

Governare e riformare l’impero al momento della sua divisione : Oriente, Occidente, Illirico

Umberto Roberto
Laura Mecella

Parte seconda - La difesa dell’unità mediterranea

East & West in the Fifth Century

Peter J. Heather

Texte intégral

  • 1 Division of Empire : Demougeot 1951, with the crucial extension of the argument in Goffart 1981. F (...)

1Two propositions concerning East Roman attitudes towards the breakdown of centralised imperial rule in the western Mediterranean have become quite firmly entrenched in some recent scholarly literature. First, it is now regularly argued that the imperial authorities in Constantinople were content more or less to stand by and watch as the western Empire collapsed. There are many components to this line of thought, but it was postulated more than a generation ago now that the period after the death of the Emperor Theodosius I saw the definitive partition of the old unified Roman Empire into two more or less independent halves. Then, building in some new ways on this starting point, Walter Goffart subsequently argued that, because of its propensity to throw up dangerous usurpers who would also seek power in Constantinople, the eastern political establishment came to regard its ostensible western counterpart much more as a threat than a natural ally. There was consequently a recalibration of priorities in the east, where outside ‘barbarian’ powers came to be seen as much less of a problem than a powerful western Empire. As a result, the eastern authorities were perfectly content to allow intrusive bodies of non-Roman military manpower to annex different parts of the western Empire’s tax base, since this undermined its capacity to be a source of dangerous usurpation. The aim was probably not the complete destruction of the Roman west, which could not be foreseen, but the new attitudes in Constantinople constituted a major contributory factor to western collapse, because they made the Roman east unwilling to contribute serious resources to the task of propping up the west1.

  • 2 Croke 1983; cf. its recent acceptance by Wickham 2009, p. 79.

2The second proposition fits in quite neatly with the first, although it in fact evolved separately. This argues that there is no sign that the eastern imperial authorities saw anything of great significance as having occurred in late summer 476, when Romulus Augustulus was deposed, in what has traditionally been viewed as the final act in the destruction of the Roman west. At the time, the deposition was viewed in Constantinople as no more than the latest in a series of swift regime changes, and it was only two generations later, in the 520s, that Odovacar’s coup d’etat came to be identified as the final moment of western imperial dissolution, and then for reasons which were entirely to do with affairs of that later period2. This is obviously a much more precise kind of argument than the broad revisionist sweep offered by the first proposition, but the purpose of this paper is to reconsider both of these lines of argument. In my view, there is good reason to argue both that the deposition of Romulus Augustulus marked a major break in established East Roman policy lines towards the political situation in the western Mediterranean, and that those policy lines – in general terms – had encompassed a degree of practical assistance for the ailing Roman west which was substantial, even if it had never amounted to the proverbial blank cheque.

East & West

3The place to start is by laying out the full range of known Constantinopolitan interventions in fifth-century west Roman affairs. What follows is essentially a list, with some items requiring rather more comment than others.

402 (& 407/8)

  • 3 See Bayless 1976 on the misunderstanding. For the general circumstances of Alaric’s first invasion (...)

4Since this used to be taken as fact, it is important to underline that there is no evidence at all that Alaric was positively sent west into Italy from the East Roman Balkans by the imperial authorities in Constantinople. One line from the poet Claudian was understood to demonstrate this, but this interpretation misconstrued the relevant passage. More generally, the evidence suggests that Alaric made the move entirely off his own bat, when it became clear, following the fall of the eunuch Eutropius in 399 and the years afterwards, that the Constantinopolitan political establishment, whomever was in charge, was not going to be willing to replicate the highly advantageous relationship the Goth had enjoyed with it from 397, when for two years he held the post of Magister Militum per Illyricum3.


  • 4 Soz., H.E. IX, 8, 6; cf. Zos. VI, 8, 2. Here and in the remaining elements of the list which follo (...)

5New fight was put into the Emperor Honorius by the arrival of an East Roman fleet in the port of Ravenna. He was about to flee to the east in the face of the combined pressure being placed upon him by the usurper Constantine III and Alaric’s second Italian expedition, but the fleet brought 4000 soldiers which encouraged him to hold on at a moment of great crisis. As subsequent events showed, there in fact remained enough military resources available to his regime within Italy, properly handled, to re-establish his hegemony over much of the western Empire. Once Alaric’s Goths, now led by Athaulf, had been starved out of Italy in 411/12, Flavius Constantius was able to mobilise these forces first to defeat the usurpers and then to force the Goths into submission. Britain fell out of the western imperial system at this point, and parts of Spain remained under the control of Vandals, Alans, and Sueves, but at least two further political generations of life were put back into the western Empire by the imperial comeback of the 410s.4.


  • 5 Refs. as PLRE II, Ardaburius, p. 137f. and ibid., John, p. 594f.

6The regime of Theodosius II mounted a major expedition under the command of Ardaburius and Aspar to put Valentinian III on the western throne following the deaths in quick succession of Fl. Constantius (Constantius III) and Honorius, and the usurpation of the primicerius John5.

427 (or 421)

  • 6 Theoph., A.M. 5931; cf. Procop., Vand. I, 2, 39-40 with the discussion of Croke 1977, p. 360. The (...)

7An east Roman intervention expelled Hunnic forces from Pannonia, traditionally part of the western Empire. The event gets only one line in the Chronicle of Marcellinus Comes, so that its size and significance is hard to judge, but this may well have been the occasion when a large force of Goths was detached from Hunnic hegemony and relocated to Thrace, to provide the manpower for the Thracian Gothic foederati who would be a major feature of the east Roman military establishment down to the 470s. If so, the brevity of the surviving record must not be allowed to hide the substantial nature of the action6.


  • 7 Refs. as PLRE II, Aspar, p. 164ff. A recent discussion of these events is provided by Merrills and (...)

8East Roman troops under the command of Aspar operated in Roman North Africa against those surviving elements of the Vandal-Alan coalition led by Geiseric which had crossed over the Straits of Gibraltar from Spain in 429. No detailed narrative of the action has come down to us, but Aspar’s intervention was part of a moderately effective containing operation which led Geiseric to accept an initial treaty in 435/6. This restricted him only to some of the less desirable provinces of North Africa : Mauretania and Numidia, rather than Proconsularis and Byzacena7.


  • 8 The so-called Law of Citations of 426 – Cod. Theod. 9, 43, 1 – is an excellent example of a wester (...)

9A major legal reform project begun in 429 culminated in the publication of the Theodosian Code in 437/8. The commission was based in Constantinople, but western legal sources were consistently drawn upon, including some highly significant recent western rulings, so that the project, while eastern-led, must be understood as a joint one. The final solemnisation of the marriage alliance between Valentinian III and Theodosius’ daughter Eudoxia in 437 was tied into the ceremonies orchestrated around the publication of the finished Code further to emphasise a commitment to imperial unity at this point8.


  • 9 Kaegi 1968, p. 28f. provides an excellent account with full refs.

10Large numbers of eastern troops, taken mostly from the Danube front, were moved to Sicily as part of a joint east-west expedition whose object was the reconquest of Carthage and the richest North African provinces which Geiseric had seized in 439 when he broke the earlier treaty of 435/6. The expedition never sailed because Hunnic attacks in the Balkans forced the return of its eastern troops, but there’s no doubting the size of the eastern commitment, not least because it was the absence of the Danubian forces which allowed the Huns sufficient leeway to launch their assault9.


  • 10 Nov. TheodII 2, 2-3 of October 1 447. Its validity for the west was confirmed by Nov. Val. III 2 (...)

11This year saw the first exchange of new imperial laws between east and west (Novellae) after the completion the original Theodosian Code project. We have explicit evidence only of an eastern compilation of laws being sent to the west, and nothing vice versa. But the Novel of Theodosius II which introduced and ratified the new collection explicitly requested that Valentinian III should send a similar compilation of his news laws to Constantinople in return. Theodosius’ intention was to initiate a regular exchange of new legal rulings between the two halves of the Empire10.


  • 11 Hyd., Chron. 154 with the argument of Maenchen-Helfen 1973, p. 137f. vs. Thompson 1948, p. 148. Th (...)

12Eastern forces attacked the central European heartlands of the Hunnic Empire while Attila and the majority of his striking forces were engaged in their own expedition into northern Italy. E.A. Thompson denied that the Marcian referred to in our brief Chronicle source for this event could be the eastern emperor, arguing that Constantinople was too frightened of the Huns to risk Attila’s wrath. He was, on the contrary, an unknown western commander. As others have argued subsequently, however, this is a decidedly forced interpretation, and most scholars have been happy to accept that Constantinople launched a diversionary counterattack of some kind while Attila was busy in Italy, as well as providing Aetius with some direct military assistance within Italy itself11.


  • 12 For a more detailed discussion & full refs., see Heather 2005, p. 390-407.

13The Eastern Emperor Leo negotiated with the western general and king-maker, the Patrician Ricimer, to put Anthemius, an eastern general and son-in-law of the former eastern Emperor Marcian, onto the western throne. Constantinople then mounted a huge seaborne expedition under the command of Leo’s brother-in-law Basiliscus to attempt to reconquer the richest provinces of North Africa from the Vandal-Alan coalition of Geiseric12.

14What broad conclusions emerge from this inventory of east Roman engagement with the western Empire in the fifth century?

  • 13 Brooks 1893 remains an excellent account of these court manoeuvres.

15Quite clearly, the imperial authorities in Constantinople were never willing to write their western counterparts a blank cheque. In the end, the eastern political establishment stood aside as what had remained a substantial rump of the western Empire fell apart in the years between Basiliscus’ defeat (468) and Odovacar’s coup d’état (476), at a point when the east still had significant military and financial resources to hand. If Zeno was not in the most comfortable of positions in the summer of 476, having just regained the throne from Basiliscus after an eighteen month hiatus, neither was the eastern Empire on the point of collapse13. There were distinct limits, therefore, to Constantinople’s commitment to the cause of Empire in the west. Nonetheless, even on a quick read through, the list does amount to a record of substantial engagement in support of the fifth-century west, with serious amounts of assistance being provided on a number of separate occasions. Some further reflections strengthen this initial impression.

  • 14 An excellent introduction to Constantine’s 50 volume project, of which only 4 partly or wholly sur (...)

16To start with, I doubt very much that the list is comprehensive. It has to be reconstructed from disparate sources of a wide variety of kinds, largely taking the form of very brief entries in western Latin chronicles, and longer narrative histories in Greek from the east. But the former, by their very nature, omit many matters, and the second, while originally, I suspect, amounting to a pretty full account of east-west relations in the fifth century – via the sequential combination of the histories of Olympiodorus, Priscus, Malchus and probably Candidus (lying behind the material extant in the text of John of Antioch) – survive only in extracts made in the tenth century under the supervision of Constantine VII Porphyryogenitus. And because the surviving extracts come from a small number of volumes of Constantine’s project, with very specific titles which closely dictated the matter to be excerpted (De Legationibus and De Insidiis in particular), they tend to deal only with very specific types of action. This has to be an argument from silence, but I think it highly likely that the record of other moments and possibly even whole types of assistance – for instance in the form of cash advances – has not survived, especially in the early 410s and again after Valentinian III was put on the western throne in the later 420s, when relations were demonstrably close between east and west14.

  • 15 I would not myself be so dismissive. Establishing a regime in power with whom the east could work (...)
  • 16 Procop., Vand. I, 6, 1; Cedr. I, 613 Bekker; cf. Lyd, Mag. III, 43. Kaegi 1968, p. 44ff. provides (...)

17More positively, on some occasions, the amount of eastern assistance committed to the west was extremely large. We have few details, but the expedition of 424/5 to enthrone the young Valentinian was clearly substantial. This could be dismissed as mainly an exercise dynastically-inspired civil war, and hence of little if any benefit to the western Empire as a whole, so that the two expeditions against the Huns in 441/2 and 468 arguably provide better examples15. Famously, that of 441/2 never managed to leave the island of Sicily, but the eastern commitment was large enough to destablise its Danubian frontier, being exploited by Attila and Bleda to launch the first major attack of their joint reign. Obviously the record here is mixed, since the authorities in Constantinople ordered the troops to return at this point – a further confirmation that the eastern cheque was never a blank one – but there is no doubting the number of troops and hence the financial expense of the effort, however fruitless it turned out to be. With the Hunnic menace now safely out of the way thanks to a series of large-scale subject revolts, the way was clear by 468, finally, for an unfettered attempt to reconquer Carthage. The recorded figures suggest an extraordinary level of commitment. Procopius states that Leo employed 100,000 troops, and spent altogether 130,000 pounds of gold, while Cedrenus reckons the number of ships at 1,113. Some of these figures are suspiciously round, but that’s no excuse for doubting the basic size of Constantinople’s commitment to the task, strongly suggestive of an attempt actually to destroy the Vandal-Alan alliance of Geiseric (an end that the western Empire itself had once come close to achieving in Spain in the 410s). We know with hindsight that the expedition failed, but Belisarius’ success in the early 530s, with a considerably smaller force, shows that it might have succeeded and the 468 expedition was certainly large enough to do so. The effort left the east Roman treasury still registering ‘empty’ five years later on the death of its organiser, the Emperor Leo16.

  • 17 Croke - Crow 1983 and Whitby 1986, with refs., discuss the evident East Roman need to refurbish Me (...)

18The scale of the east’s commitment to western defence has to be reckoned larger still in my view, when proper account is taken of the overall strategic context. When thinking about how much assistance to give to the west, the different regimes in Constantinople needed to balance western need against the requirement that their own territories should not be left unguarded. And, in fact, the fifth century posed the east a whole series of pressing strategic problems. Above all, the Persian front could not be left unguarded. Persia was not only the East’s traditional enemy, but a successful Persian attack on the central Mesopotamian front would quickly inflict damage one of the eastern Empire’s key revenue-generating regions: the Fertile Crescent. In the fifth century, famously, relations with Persia were relatively peaceful, compared to the fourth and sixth centuries, but there were moments of hostility nonetheless, and relations with Persia were never so good that particularly the Mesopotamian frontier could be left unguarded. In the late fourth century army listings of the Notitia Dignitatum, something like one third of the entire military establishment of the total Empire – east and west – was stationed on the eastern front. I imagine – although there is no hard evidence – this was allowed to decline somewhat during the fifth century (certainly defensive fortifications required a major refurbishment at the start of the sixth), but troop numbers could never have been allowed to fall very far. Although Persia faced its own periodic nomad threat particularly from the Hephthalite Huns, if Constantinople’s eastern front had become too weak, then there is every reason to suppose that the Persians would have taken advantage of it to score some quick and profitable victories17.

  • 18 The evidence is collected and analysed in Heather 2001.
  • 19 On the rise of Hunnic power, see e.g. Thompson 1948; Maenchen-Helfen 1973; Heather 2009, esp. ch.  (...)

19If that were not enough, the fifth-century also saw an unprecedented level of threat emerge on the eastern Empire’s northern, Danubian front. Before the late fourth century, Constantinople faced the strategic problem in that quarter of managing a series of small-to-medium-sized, largely Germanic-dominated client states north of the river. Broadly-speaking, one major campaign per generation (every twenty to twenty-five years) was required to keep these entities operating within lines acceptable to the Roman authorities, backed with carefully structured lines of economic and cultural diplomacy18. The rise of Hunnic power in south-eastern and then central Europe, by contrast, united an unprecedented quantity of the warrior groups at the heart of these client states under Hunnic dominion. By the time of Attila and Bleda in the 440s, the Hunnic Empire included several previously independent groups of Goths, Rugi, Suevi, Gepids, Sciri, and Alans amongst many others. Its overall structure was always fragile, since conquest was the primary means by which unification had been achieved, and many of the new Empire’s subjects were only too happy to take every opportunity to break away from Hunnic domination. Nonetheless, particularly in the 430s under Rua and in the 440s under Attila, an entity had come into being which could directly challenge Roman hegemony on the Danube. 447 represented the nadir, Attila inflicting sequential defeats on both the eastern Empire’s Thracian and Praesental field armies, but, from the mid 420s onwards, any help for the west had to be balanced against the opportunity that major troop transfers in that direction – even temporary ones – might offer to the Huns19.

  • 20 The evidence is collected and discussed in Heather 2009, 238ff.
  • 21 A detailed discussion of this period can be found in Heather 1991, pt. 3.

20Nor, indeed, did the heightened level of threat on the Danube disappear with Attila’s sudden death in 453, although it certainly changed nature. At that point, the enforced unity created by Hunnic dominion was progressively overturned, and, as this process unfolded, probably in the five years after Attila’s death, the threat of any really large armies bursting onto east Roman territory from north of the Danube subsided. But, again, the process of Hunnic collapse was itself a violent one with serious consequences for Constantinople on two levels. First, the struggles threw various uninvited refugee groups onto east Roman territory in the late 450s and early 460s, who then needed to be dealt with in some way20. Second, as new winners began to emerge from the process, such as the Amal-led Goths in Pannonia, their leaderships showed predatory intent towards East Roman wealth, and began to manoeuvre against it in order first to extract and then increase annual gold subsidies. At the end of one invasion of east Roman Illyricum, in the early 460s, for instance, the Amal-led Goths increased their annual subsidy to 300 pounds of gold per annum, and, ten years, later exploited political disarray in Constantinople to throw themselves onto east Roman territory for a more extended, fifteen year period, to increase the flow of Roman wealth in their direction. Not until Theoderic led them out of the Balkans again in autumn/winter 488/9 did the level of threat subside, by which time central imperial authority in the west had ceased to exist21. Throughout pretty much the whole of the fifth century, therefore, any aid sent to the ailing west was being provided despite the fact that Constantinople faced threats on two other fronts: an implicit Persian menace to the eastern Empire’s jugular, and an extremely unstable situation on the Danube.

  • 22 There has been a tendency in some recent literature to suppose that all the subject peoples chose (...)

21Sometimes, at least, this assistance was also appropriate and coherent. A reliable garrison for Ravenna in 409/10, for instance, was exactly what Honorius needed in order to find some political breathing space from the immediate pressures placed upon him by the manoeuvres of Constantine III and Alaric. Re-endowed with one point of security, the emperor was then in a position to find a military eminence grise of sufficient talent to mobilise the considerable forces still available to the regime within Italy first to restore its control there, and then, of the vast majority of the Roman west. This was achieved by Fl. Constantius in Honorius’ name from 411, but, without the east’s assistance, Honorius would probably have fled to Constantinople and the western Empire could easily have fragmented at this point. Attacking the Hunnic homelands left vulnerable by the despatch of a major long-distance expedition in 452 was also a classic mechanism employed on several other occasions by eastern imperial regimes to disrupt an attacking force, by playing on the natural concerns of warriors for their families. In this case, however, there was also a further dimension. Many of the Huns’ subjects had been incorporated only unwillingly into Attila’s Empire, and, as our sources demonstrate, took every opportunity to attempt to throw off Hunnic domination22. Attacking Attila from behind in 451 potentially generated another context for a major loss of Hunnic subjects (such as happened in 427 when a Gothic force was detached from Hunnic rule Pannonia and resettled in Thrace).

  • 23 In both Italy after the sojourn of Alaric’s Goths in 408-11 and Numidia after the Vandal occupatio (...)

22But the clearest evidence of what can in my view reasonably be considered strategic thinking in Constantinople is in my view again provided by the great Vandal expeditions of 441/2 and 468. If the situation was already serious beforehand, the Vandal seizure of Carthage and the richest provinces of North Africa in 439 plunged the western Empire into deep crisis. The consequent loss of tax revenues directly and immediately threatened the size of western military establishment that could be maintained, which in turn meant that still greater opportunities for expansion (with further losses of tax revenue and hence troop numbers) would fall open to the Visigoths and other centrifugal forces already established on western soil, most of them with recent origins beyond the imperial frontier, and which even the pre-439 western military establishment had struggled to contain. The immediate context of Vandal success in 439 was the three year war Aetius had had to wage against the Visigoths from 436 to prevent them adding substantial new territories to those granted them in the Garonne valley in 418/19. Against this broader backdrop, attempting to land a knock-out blow on the Vandals made perfect strategic sense. At one and the same time, it would both have removed one of the major centrifugal forces established on western soil, and added an extremely valuable flow of potential revenues back into the western Empire’s coffers. Proconsularis, where the Vandals and Alans had settled and redistributed a sizeable quantity of landed estates, would not have produced much tax revenue for some time, but landholding patterns in Byzacena and Numidia had been left intact, so that a substantial flow of revenues might be expected from these provinces more or less immediately23. No other available option had the same potential to arrest the erosion of west Roman central imperial power.

  • 24 This is the only point on which I substantially disagree with Kaegi 1968, p. 29-31.

23Nor, perhaps more surprisingly, is there much sign of any great difference in the willingness of the various eastern regimes to provide assistance to the west, when, in principle, there might very well have been. Theodosius II, of course, was expressing a strong dynastic commitment when putting his young cousin Valentinian III on the western throne in 424/5, but, during his reign, help was not confined merely to keeping a Theodosian on the throne. The expeditions under Aspar to North Africa to the 430s and again to Sicily in 441/2, preparatory to an attempt to recapture Carthage, both showed a wider interest in the general health of the west Roman polity. It has been suggested that Constantinopole’s interest in the west then tailed off under Theodosius II’s non-dynastic successor, the Emperor Marcian, since his reign (450-457), saw no large-scale expedition to the west. On reflection, however, I think this is probably mistaken. Marcian was willing to launch at least a spoiling attack in 452 to help force Attila’s withdrawal from northern Italy. Disease also played its part in the Huns’ retreat, as famously did Pope Leo I, but this was no small act to have undertaken. And, after 452, political instability in the west set in so fast with the deaths in quick succession of Attila (453), Aetius (454) and Valentinian III himself (455) that there was simply no opportunity to negotiate and mount a major expedition to the west before Marcian’s own death in 457. Given the pace of ancient communications, where everything moved at an arthritically slow 40 kms per day, you have to calculate that a lead-in time of two to three years was required to undertake the political negotiations, and then to move troops and assemble the necessary logistic supplies for a major military intervention in the west. It was, for instance, only two years later, in 441 that the counter-expedition even began to gather in Sicily following Geiseric’s seizure of Carthage in 439. From this perspective, it is simply the case that no western regime between the deaths of Attila and Marcian lasted long enough for the East to be able to think of doing anything major to strengthen the strategic position of the west. Our sources are also far too lacunose to make it at all a safe conclusion that smaller-scale assistance was not again being provided in the meantime24.

24Under the Emperor Leo, the approach of Constantinople remained broadly similar to what we can observe under Theodosius II and Marcian. Leo was certainly not prepared to support just any ruler of the west. Libius Severus was never recognised at all, and Majorian only, it seems, after a considerable delay. Leo was also much happier, naturally enough, to provide weighty support for his own candidate: the Emperor Anthemius. Nonetheless, there was no dynastic link between these two men, yet still Leo was willing to put together an armada to attempt to reconquer Carthage in 468. As I have argued elsewhere, I do think this expedition was part of the price that Anthemius had negotiated on leaving Constantinople for the west, and that Leo had political interests in wanting him gone in the mid 460s, since he was a potential candidate for the throne in what was at that point a highly unstable situation, Leo having no heir. But if dynastic calculations always played a part in the decision-making process, so, evidently, did a broader commitment to the west which transcended changes of regime in Constantinople, and was not dependent on close familial ties between rulers of east and west.

25The argument must not, of course, be overstated. The help provided by Constantinople was always highly calculated, with a distinct element of self-interest, and never exceeding the limits of strategic safety, which, as in 441/2, always put the security of the East first. But to expect anything else, I think, would be unrealistic. And within these parameters, it does seem to me that the Eastern Empire’s record of assistance to the west in the fifth century is actually a perfectly respectable one. Help was provided not just when the west was run by a fellow member of the same ruling dynasty, but much more generally, and, in fact, quite consistently across the sequence of regimes of under the overlordship of three different eastern Emperors : Theodosius II, Marcian, and Leo I. It was also on occasion large-scale, generating very considerable costs for the east, and, in strategic terms, well-directed. Reconquering Carthage and its surrounding provinces from the Vandals was the single military move which promised to achieve the maximum benefit for the western Empire.

  • 25 An excellent introduction in English is provided by a combination of Nelson 1992; Goldberg 2006; C (...)

26In particular, Constantinople’s record of assistance to the fifth-century west stands up extremely well, I would argue, in comparison to kinds of manoeuvring observable within the confines of the ninth-century Carolingian Empire as Charlemagne’s creation came to be divided first between the three surviving sons of Louis the Pious in 840, and then between their various descendants subsequently. Each of the many territorial divisions which followed (as did the run up to the first partition while Louis himself was still alive in the 830s) saw major bouts of direct, full-scale battle between the different protagonists. This was supplemented by another kind of conflict when the rulers of different parts of Charlemagne’s former Empire stirred up problems for their fellow Carolingians by providing financial and military support for younger nephews and then cousins, as the latter started to demand a share of their fathers’ domains. A third type of conflict was also occasionally evident in the form of the ruler of one part of the Empire negotiating directly with groups of outsiders (usually Vikings) to cause trouble for a fellow Carolingian in charge of another portion25. Little of this kind of direct military conflict between dynastically-related (or even unrelated) Roman rulers is evident in the fifth century after the crystallisation of the division between east and west in 395, and the two large-scale intrusions onto Roman territory of outsiders in the 370s and again in the first decade of the fifth century.

  • 26 On Stilicho’s manoeuvres, see Heather 1991, ch. 6. 424/5: refs. as above note 5.
  • 27 The only ‘successful’ division of the Empire in the fourth century – measured as one which generat (...)

27The entire fifth century, between 400 and 476, saw just one threatened and one actual bout of internecine East-West Roman conflict with overtones of barbarian involvement. In 405, Stilicho was gearing up to try to reattach East Illyricum to the western Empire with Alaric’s assistance (I think because he wanted a broader military alliance with Alaric whose followers were occupying the region at the time) when first Radagaisus’ invasion of Italy, and then the Rhine crossing derailed his plans. In 424/5, likewise, the usurping western regime of the primicerius John recruited Hunnic assistance for its pending struggle for survival against the campaign of Ardaburius and Aspar, but the campaign was concluded before the Huns could arrive. Those two instances apart, the period saw no major bouts of east-west conflict. There is also no evidence whatsoever that Constantinopolitan rulers directly encouraged any of the ‘barbarian’ invasions of the western Empire’s territories, or directly negotiated with former outsiders now established on western territory with the intention of undermining the power of any west Roman central imperial authority26. The Carolingian pattern of much greater conflict between the separate sub-rulers of a once-united Empire is much more applicable to the fourth century, before the formal division of the Roman Empire in 395, when civil war was a regular feature of the imperial system, even sometimes when different parts of the Empire were ruled by members of the same dynasty, and different contenders would on occasion stir up ‘barbarian’ neighbours to cause trouble for their rivals (the barbarians of this period, of course, being established beyond rather than within the defended frontier)27. Viewed from this perspective, it would seem that far from stirring up trouble between east and west, the formal division of the Empire generally worked to limit internal conflict between the two halves of the Empire, refocusing political rivalries from full-scale civil war between those at the head of each part, to much more limited – and much less damaging – struggles for power within each of the two imperial courts, while relations between the two were generally good.

  • 28 Only the Duchy of Normandy among post-Carolingian entities was constructed around the military pow (...)

28Bringing in the Carolingians may seem like an odd twist of argument, but I think it is perhaps to the point that Walter Goffart started life as a Carolingian historian. His famous article on east and west Rome in the fourth and fifth century never makes an explicit comparison, but it seems worth at least suggesting that his view of east-west relations was influenced by the behaviour of the rulers of Rome’s Carolingian successor in the ninth century. There, once the Empire had been divided, the rulers of its various portions did largely (but not completely) act as though they were the rulers of separate and for the most part rival realms. And in the Carolingian case, it was these continuous internal rivalries which fundamentally undermined the existence of Charlemagne’s Empire, while the role of outside attackers – whether Vikings, Magyars, or Saracens – was decidedly marginal to the process of imperial dissolution. The overall effect of Goffart’s argument, therefore, is to recast the fall of the Roman west broadly along the well-established lines of a Carolingian model of imperial collapse. One consequent problem, however, is that the nature of relations between Constantinople and the west in the fifth century, as explored here, does not fit at all well with the ninth-century model. Where ninth century Carolingian rulers were, on the whole, content to stir up any amount of trouble for their relatives, the Roman east gave direct, large-scale, and entirely apposite assistance to the west. Nor, indeed, are some of the broader dimensions of the ninth-century model so easily applicable either. Post-Carolingian Europe for the most part ended up in the hands of wide variety of indigenous elites, who had come to the fore in an extended internal political process of struggle and civil war, but the vast majority the successor states to the western Roman Empire were founded around the military power of intrusive outsiders (Goths, Burgundians, Vandals, Franks etc.)28. This is, of course, the subject of another paper, but it does on the face of it suggest that outsiders had a larger role to play in the fifth century than they did in the ninth, where internal rivalries of various kinds clearly did hold centre stage.

The Year 476

  • 29 Croke 1983 engaging with Ensslin 1948; Wes 1967; Momigliano 1955.

29Turning now to the second of the propositions, it is entirely correct to state that no contemporary Roman author – eastern or western - before Marcellinus Comes wrote his Chronicle in 517/18 explicitly identifies the deposition of Romulus Augustulus as the end-point of the western Empire. That is not exactly the same thing as saying that no one before Marcellinus had seen this event as one of great or even epochal significance, and, in fact, the kinds of broader conclusion that have been drawn from this initial and entirely persuasive observation about Marcellinus’ Chronicle – that the events of 476 passed by contemporaries generally unnoticed – mark a distinct step beyond the main argument being put forward in the original paper from which they derive, and were not central to its purposes. In that study, Brian Croke’s main purpose was to put a final nail in the coffin of an historical argument, mounted mainly by Ensslin and Wes (but partly supported too by Momigliano), that the lost Roman history of Q. Aurelius Symmachus, the consul of 485, was the original source behind Marcellinus’ view of the significance of the events of 476. According to this hypothesis in its fullest form, Symmachus wrote his history in the time of Theoderic the Ostrogoth, and its emphasis on a lost Roman past was one strand in a broader effort on the part of its author, his famous son-in-law the philosopher Boethius, and other supportive Roman senators, to encourage the eastern half of the Roman Empire to reconquer Italy and return it to a properly Roman political affiliation. In this view, it was these treasonable connections which eventually led to the downfall and executions of both Symmachus and Boethius late in Theoderic’s reign. Croke’s study used Marcellinus’ view of the year 476 as the starting point for a highly successful demolition of what is in fact a chain of increasingly implausible hypothesis required to make all this hang together, and to argue – entirely convincingly – that it was Marcellinus’ own view, and not something taken from Symmachus, that the western Empire had ended with the deposition of Romulus Augustulus29.

  • 30 Croke 1983, p. 114f.

30So far, so good: but establishing this fundamentally historiographical point is, as Croke himself explicitly notes, not remotely the same thing as arguing either that Marcellinus’ Chronicle was the first moment that any Roman had ever viewed the deposition of Romulus Augustulus as the final moment of western imperial demolition, or that contemporaries at the time had not regarded the events of 476 as momentous30. Both of these propositions – not so much argued for by Croke himself, but reflecting the broader reception of his paper in subsequent scholarship – are extremely doubtful. In my view, it is relatively straightforward to show both that the ending of central western imperial authority was well-recognised in the 470s, and that it profoundly affected Constantinopolitan policies towards the western Mediterranean.

  • 31 Refs. as above note 14. 

31To start with, it is worth recalling just how much of an argument from silence it is to suppose that no one in Constantinople prior to Marcellinus had thought of 476 as marking the end of the western Empire. Before Procopius’ history of the reign of Justinian, our long run of contemporary classicising Greek accounts of fifth- and early sixth-century events survive, as we have already seen, only in fragmentary form, and in relation to highly specific subject areas, particularly ‘embassies’ and ‘plots’31. It is hardly surprising, therefore, that no sustained treatment survives in the historical record of the broader Constantinopolitan response to the deposition of Romulus Augustulus. More positively, when you look at the available evidence as a whole, there are considerable indications, even if they are sometimes indirect, that the coup of 476 and the handful of years leading up to it were recognised as a highly significant moment of political transformation both inside the eastern imperial court at Constantinople, and more broadly in the Roman west itself.

  • 32 Malch., fr. 14.

32The single most important piece of evidence survives in Malchus of Philadelphia’s account of a pair of embassies which arrived in Constantinople in the autumn of 476, shortly after Odovacar’s successful coup. One was from Odovacar himself, the other ostensibly from the Roman Senate acting in the name of Romulus Augustulus, although Odovacar had carefully coordinated the effort, so that the senators were primed to say exactly what their new overlord wanted. In these embassies, both the senators and Odovacar himself explicitly stated that the western Empire had come to an end, and, indeed, part of their mission was to hand over western imperial regalia – which it was treasonable for anyone but a legitimate emperor to wear – to the eastern Emperor Zeno32. Odovacar’s purpose here, I take it, was to try to head off any possibility that Zeno might actively support a restoration of the deposed Julius Nepos, currently sulking in Dalmatia – Nepos had plenty of eastern contacts and also sent an embassy at the same time to Constantinople asking for assistance (as Malchus also tells us in the same fragment) – but that doesn’t mean that both Odovacar and the Senators of Rome were not also extremely clear in their own minds that the whole concept of a western Roman Empire was now an anachronism. Indeed, by the autumn of 476 this had become a straightforwardly sensible conclusion to draw, since the western half of the former Roman superstate was now divided between a series of successors, many of them headed by former outsiders to the Roman system such as Goths, Vandals and Alans, Sueves, Burgundians, Franks and Anglo-Saxons. Such was the extent of political fragmentation by 476 (although boundaries remained in flux) that none of the successors looked remotely imperial in character and extent, certainly not Odovacar’s realm which encompassed only Italy and its immediate surrounds. Monolithic Empire had given way to a series of (emerging) regional powers, and contemporaries were not stupid enough to miss the point. The western Empire was over, and both Odovacar and the greater landowners of Italy were perfectly well aware of the significance of this transformation.

  • 33 Malch., fr. 14 with Croke 1983, p. 115 on the coinage.

33If it was entirely obvious to all the interested parties in the former Roman west that events had drawn a line under the continued existence of Empire, Zeno’s ostensible response to Odovacar’s approach was more guarded. Instead of accepting the generalissimo’s declaration that the western Empire was over, the emperor responded that the west still had an emperor in the person of Julius Nepos who had been driven out of Italy the year before, and that Odovacar should both receive him back into Italy and look to him for the title he desired. At the same time, however, Zeno said that he would have granted Odovacar the rank of Patrician had Nepos not done so – clearly Nepos had already made some kind of a grant, perhaps as part of an exploratory move to see if the new ruler of Italy would receive back – so the emperor did not reject the generalissimo’s approach outright. And up to a point, at least, Odovacar was willing to obey orders. He made no moves to receive Nepos back into Italy, but the coinage of Italy continued to acknowledge him as Augustus up to his death by assassination in 48033. At first sight, Zeno’s guarded response might suggest that, in 476, Constantinople was still holding on to the continued importance of the concept of a western Empire, but a bit more thought suggests that such a conclusion is mistaken. Substantial changes in eastern behaviour at the same time signal that, whatever Zeno’s reasons for fending off Odovacar’s approach in the way he did, a continued belief in the viability of a west Roman imperial state was not one of them.

  • 34 Euric: Jord., Get. 237, with the detailed account of Wolfram 1988, p. 181ff. On the full consequen (...)

34Most important, the years after 468 saw no further major eastern interventions in west Mediterranean affairs. The entirely coherent strategy of conquering Vandal Africa to return the richest available revenue source to the rump western imperial taxbase, the only policy which offered any hope of putting significant new life back into the western Empire, was definitively abandoned after the defeat of Basiliscus. The sources report that the treasury in Constantinople had still not recovered from the costs of that expedition at the time of the Emperor Leo’s death in 474, but lack of cash was not the only reason that no further Vandal expeditions were launched or even planned. Above all, the expedition’s defeat quickly generated a major adjustment in prevailing balances of political power in the remaining territories of the western Empire. As soon as news came through of Basiliscus’ defeat, Euric ‘suddenly’ (so Jordanes tells us) realised that he no longer needed to take account of potential western imperial power, and began a series of major conquests. In half a decade from 468, he extended the boundaries of the original realm he had inherited, confined just to the Garonne valley and the city of Narbonne (added in 462), northwards to Tours on the Loire, eastwards to the old provincial capital at Arles, and southwards right across the Iberian peninsula to the straits of Gibraltar. This created a new Visigothic kingdom which was significantly more expansive than the territories still controlled from Rome and Ravenna. In the same era, others, too, were busy annexing everything they could lay their hands on. Burgundian expansion southwards down the Loire gathered a certain momentum until the Visigoths halted it, and Frankish groups were busy further north. The prospect of a western imperial comeback via a successful conquest of Vandal Africa had just about kept these centrifugal forces in check through the late 450s and earlier 460s, and Anthemius had still ruled a substantial imperial rump as late as 467/8. But Basiliscus’ defeat drew a line under any possibility of imperial renewal, and these territories were quickly divided up between various predatory rivals in its immediate aftermath34.

  • 35 473/4: Malch., fr. 5. On the final peace treaty, see now Merrills - Miles 2010, p. 122-124.

35In a very real sense, therefore, Basiliscus’ defeat unleashed the final dissolution of the western Empire, and this was clearly recognised in Constantinople, where it generated a fundamental change in policy towards the west. At the same time as Zeno was giving his guarded answer to the ambassadors of Odovacar and the Roman Senate, the Emperor was also in the middle of finalising a definitive peace treaty with the Vandals. Negotiations had begun as early as 473/4 when Zeno’s ambassador got as far as sorting out a truce and some exchanges of prisoners. Held up, perhaps, by internal political problems (Basiliscus’ coup exiled Zeno from Constantinople for eighteen months or so from January 475), negotiations for a full and final peace treaty did not come to fruition until late 476. There had, however, been no further hostile eastern moves towards Vandal Africa after 468, and the delay must not be allowed to hide the colossal significance of Constantinople’s change of tack. Re-conquering Vandal Africa was the one move which might have put some serious life back into the western Empire. The decision to make peace with the Vandals in the aftermath of Basiliscus’ defeat is itself a clear sign that Constantinople too had given up hope that there was any serious possibility of a western imperial survival, whatever Zeno might have said to Odovacar on the subject of Julius Nepos. It was chronological coincidence that peace with the Vandals finally came in exactly the same year as the deposition of Romulus Augustulus, but even if a final peace had been negotiated a year or two earlier, this would not change the point. By the early to mid-470s, it was crushingly obvious not only in the west (explicitly, at the very least, to Euric, Odovacar, and the Roman Senate – and implicitly to the rulers of the Franks and Burgundians) that the western Empire was past its sell-by date, but also to the eastern imperial authorities in Constantinople, whose acquiescence in the face of the new, brute strategic realities was firmly signalled by their acceptance of the necessity of negotiating an overarching peace deal with the Vandals35.

  • 36 On the carefully-crafted Romanness of Theoderic, see, amongst many possibilities, Heather 1993.

36The point is confirmed by some other profound changes in behaviour. More symbolic perhaps, but important nonetheless, the period after Romulus’ deposition saw no further exchanges of new legal rulings between east and any power of the new west. Famously, the regimes of both Odovacar and particularly Theoderic the Ostrogoth tried to stay within a Roman ideological framework. They maintained the tradition of appointing consuls, most of whom were recognised in Constantinople, and passed any new laws in the form of edicts, which more junior Roman magistrates than emperors had long had the right to do. They neither claimed the right to issue fully-fledged novellae, however, nor were any new imperial laws made in the Constantinople transmitted to their domains. Even if both regimes maintained some kind a façade of formal Romanness (which can be observed, likewise, in the fact that exchanges of public letters also continued at intervals between eastern emperors and the Roman Senate), the formal legal fragmentation makes it clear that after 476, as far as Constantinople was concerned, there no longer existed a Roman imperial authority structure in the west36.

  • 37 He was hailed as semper augustus by the Senator Caecina Mavortius Basilius Decius (ILS 827), and t (...)
  • 38 Jones 1962 is a valiant attempt to sort out the details of Theoderic’s precise constitutional posi (...)

37Even when a new power of imperial stature appeared in the west, this attitude did not change. From 511 onwards, a point that is sometimes missed, Theoderic was ruling Italy and Sicily, Spain, southern Gaul, Dalmatia, and parts of Noricum and Pannonia directly, and exercising hegemony over the Vandal and Burgundian kingdoms. He had, in short, put back together a very considerable chunk of the former western Empire, ruling far more of it than had Anthemius, the last western Emperor to receive direct eastern assistance. Theoderic was also well aware of the extent of his achievements and clearly wanted to be recognised as western emperor, adopting a style of rule, both in terms of ceremonial and propagandistic self-projection, which made the point entirely clear. In several rounds of negotiation, however, Constantinople refused to recognise Theoderic’s imperial pretentions and, when it could, sought to rein them in, making his successors agree, for instance, that the emperor’s name should be read out first in proclamations and that imperial statues, rather than those of the Gothic royal house, should be placed on the right in the position of honour. At least one of Theoderic’s senatorial subjects was ready to hail his ruler with the uniquely imperial title of semper Augustus, but Constantinople not only rejected the Goth’s pretentions, but consistently worked to undermine his power as circumstances offered opportunity, initially concocting an alliance with his Frankish rival Clovis in the first decade of the sixth century, and then taking advantage of a succession crisis in the 520s, caused by the early death of Theoderic’s chosen heir, to stir up religious dissentions within the kingdom and seduce the Vandal and Burgundian kingdoms from their Gothic allegiance37. No matter how powerful and how apparently Roman a political structure emerged in the western Mediterranean after 476, therefore, Constantinople was no longer willing to recognise it as properly imperial, and, on the contrary, was now entirely ready to exploit the west – for the first time – as a dumping ground for its own problems. Much scholarly ink has been expended debating the precise terms on which Theoderic originally moved into Italy, but the central point is that Zeno’s prime motivation was to remove him and his Gothic military following from the east Roman Balkans. After a decade and a half of intermittent conflict and periodic assassination attempts, relations between emperor and Goth had reached a complete impasse, to which overthrowing Odovacar offered a mutually convenient solution. The perspective which has sometimes been anachronistically applied to Alaric’s arrival in Italy really does apply to that of Theoderic. By the 480s, Italy was no longer a peer centre of Empire, but a convenient disposal site for unwanted, eastern barbarians38.

38Two broad conclusions follow from this reconsideration of East-West Roman relations in the fifth century. First, it was not until 518 that the year 476 was specifically identified in any of our surviving texts as the end point of Empire in the west, and Brian Croke is entirely convincing that this was indeed the view of Marcellinus Comes himself, rather than a reflection of some bizarre, historiographic plot hatched within Ostrogothic Italy. Nonetheless, there is every reason to suppose that leading contemporary Romans of both east and west fully understood that events of epochal significance were unfolding around them in the mid-470s. For west Romans, the evidence is entirely explicit. Those located in Italy itself were participants in diplomatic contacts which went to Constantinople straightforwardly to proclaim the end of Empire in 476, while the Roman elites of central and southern Gaul and Spain just lived through the direct experience of passing from the control of a rump imperial state to that of different successor state kings of recent non-Roman origins in the few years since the defeat of Basiliscus’ armada in 468. That the East Roman authorities too were conscious of the end of Empire is less explicit, but no less certain. For one thing, they actually received the Roman senatorial embassy which told them so, and their acknowledgement of the reality of this fact is evident above all in the end of any further attempts to reconquer Vandal Africa.

39My other main conclusion is that, contrary to a perception which has taken hold in some recent scholarship, the eastern half of the Roman Empire did in fact provide significant assistance to the west in the years after 400, and that there is no evidence at all that its leadership (across multiple imperial reigns and individual regimes) at all preferred to work with the barbarian dynasts who now controlled parts of the Roman west, or to use them to weaken the power of the western Empire, as a guard against potential imperial rivals. There is, as we have seen, some evidence that emperors and would-be emperors of the fourth century were willing to use externally situated barbarians in this way, but none at all of the same kind of behaviour once some of these outsiders began to settle on western territory in the fifth century. Indeed, the whole tone of relations between the two halves of the Empire altered substantially in the fifth century, compared to the fourth. Where western-based emperors and usurpers regularly confronted and sometimes even defeated eastern-based emperors in the fourth century, and tension between notionally joint rulers was the norm rather than an exception, relations in the fifth century were normally good, with only occasional bouts of tension and just the one head to head conflict initiated by Constantinople in the mid-420s.

40Thinking more broadly about this extraordinary transformation, I think one can make the argument that the strikingly good relations of the fifth century probably were facilitated by the fact that the western Empire’s tax revenues and hence military power had been damaged by its losses of territory to intrusive outsiders. In its crucial first two decades, permanent and temporary losses of tax base in Britain, northern Gaul, and Spain meant that no western emperor could stand any longer as a credible rival to a Constantinopolitan peer, and hence the Roman west no longer posed the same kind of potential threat to eastern politicians. But that – it must be emphasised – is not remotely the same thing as arguing that Constantinople somehow manufactured barbarian annexations of western territories in order to achieve this outcome. On the contrary, we have no evidence of direct contacts between Constantinople and the leading barbarians of the west, and plenty of evidence, as we have seen, of eastern help for west Roman rulers struggling to limit the damaging effects of these outside settlements. To my mind, the conclusion is inescapable that, despite past rivalries and conflicts, the political establishment of the Roman east saw its western counterpart as a very special partner, to whom it was closely tied by bonds of shared history and cultural commonality. Constantinopolitan-based imperial regimes were thus willing to expend major resources in trying to prop up the Roman west, for as long as there was any chance that a properly Roman imperial power could still be maintained there.


Bayless 1976 = W. N. Bayless, The Visigothic Invasion of Italy in 401, in Classical Journal, 72, 1976, p. 65-67.

Brooks 1893 = E. R. Brooks, The Emperor Zenon and the Isaurians, in English Historical Review, 8, 1893, p. 209-238.

Costambeys 2011 = M. Costambeys et al., The Carolingian World, Cambridge, 2011.

Croke 1977 = B. Croke, Evidence for the Hun Invasion of Thrace in A.D. 422, in Greek Roman & Byzantine Studies, 18, 1977, p. 347-367. 

Croke 1983 = B. Croke, A.D. 476: The Manufacture of a Turning Point, in Chiron, 13, 1983, p. 81-119.

Croke - Crow 1983 = B. Croke, J. Crow, Procopius and Dara, in Journal of Roman Studies 73, 1983, 143-159.

Demougeot 1951 = E. Demougeot, De l’unité à la division de l’Empire romain 395-410: Essai sur le gouvernement impérial, Paris, 1951.

Dignas - Winter = B. Dignas, E. Winter, Rome and Persia in late antiquity: neighbours and rivals, Cambridge, 2007.

Dodgeon - Lieu 1991 = M. H. Dodgeon, S. N. C. Lieu, The Roman Eastern frontier and the Persian Wars (AD 226-363) : a documentary history, London, 1991.

Ensslin 1948 = W. Ensslin, Des Symmachus Historia Romana als Quelle fur Jordanes, Munich, 1948 (Sitzungsberichte der Bayerischen Akademie der Wissenschaften, phil.-hist. Kl.).

Goffart 1981 = W. Goffart, Rome, Constantinople, and the Barbarians in Late Antiquity, in American Historical Review, 76, 1981, p. 275-306.

Goldberg 2006 = E. J. Goldberg, Struggle for Empire: Kingship and Conflict under Louis the German 817-876, Ithaca, 2006.

Greatrex - Lieu 2002 = G. Greatrex, S. Lieu, The Roman eastern frontier and the Persian Wars : a narrative sourcebook, London, 2002.

Heather 1991 = P. J. Heather, Goths and Romans 332-489, Oxford, 1991.

Heather 1993 = P. J. Heather, The Historical Culture of Ostrogothic Italy, in Teoderico il grande e i Goti d'Italia. Atti del XIII Congresso internazionale di studi sull'Alto Medioevo, Spoleto, 1993, p. 317-353.

Heather 1996 = P. J. Heather, The Goths, Oxford, 1996.

Heather 2001 = P. J. Heather, The Late Roman Art of Client Management & The Grand Strategy Debate, in W. Pohl, I. N. Wood (ed.), The Transformation of Frontiers from Late Antiquity to the Carolingians. Proceedings of the Second Plenary Conference, European Science Foundation Transformation of the Roman World Project (Brill), Leiden, 2001, p. 15-68.

Heather 2009 = P. J. Heather, Empires and Barbarians: Migration, Development, and the Birth of Europe, London, 2009.

Honoré 1998 = A. M. Honoré, Law in the Crisis of Empire, 379-455 AD: The Theodosian Dynasty and its Quaestors, Oxford, 1998.

Humfress 2007 = C. Humfress, Orthodoxy and the Courts in Late Antiquity, Oxford, 2007.

Innes 2007 = M. Innes, Introduction to Early Medieval Western Europe, 300-900: The Sword, The Plough and the Book, London, 2007.

Jones 1962 = A. H. M. Jones, The Constitutional Position of Odoacer and Theoderic, in Journal of Roman Studies, 52, 1962, p. 126-130.

Kaegi 1968 = W. E. Kaegi, Byzantium and the Decline of Rome, Princeton, 1968.

Lemerle 1971 = P. Lemerle, Le premier humanisme Byzantine, Paris, 1971.

Maenchen-Helfen 1973 = O. J. Maenchen-Helfen, The World of the Huns, Berkeley, 1973.

Matthews 1975 = J. F. Matthews, Western Aristocracies and Imperial Court A.D. 364-425, Oxford, 1975

Matthews 2000 = J.F. Matthews, Laying down the law: a study of the Theodosian code, New Haven, 2000.

Merrills - Miles 2010 = A. Merrills, D. Miles, The Vandals, Oxford, 2010.

Mitchell 2007 = S. Mitchell, A History of the Later Roman Empire AD 284-641, Oxford 2007.

Momigliano 1955 = A. Momigliano, Cassiodorus and the Italian Culture of His Time, in Proceedings of the British Academy, 41, 1955, p. 215-248.

Nelson 1992 = J. Nelson, Charles the Bald, London, 1992.

PLRE II = J. R. Martindale, The Prosopography of the Later Roman Empire, II, A.D. 395-527, Cambridge, 1980.

Prostko-Prostynski 1994 = J. Prostko-Prostynski, Utraeque Res Publicae: The Emperor Anastasius’ Gothic Policy (491-518), Poznan, 1994.

Rollason 2012 = D. Rollason, Early Medieval Europe 300-1050, Harlow 2012.

Sirks 2007 = A. J. B. Sirks, The Theodosian code: a study, Friedrichsdorf, 2007.

Thompson 1948 = E. A. Thompson, A History of Attila and the Huns, Oxford, 1948.

Wes 1967 = M. A. Wes, Das Ende des Kaisertums im Westen des romischen Reichs, The Hague, 1967.

Whitby 1986 = L. M. Whitby, Procopius and the Development of Roman Defences in Upper Mesopotamia, in H. Kennedy, P. Freeman (ed.), The Defence of the Roman and Byzantine Near East, Oxford, 1986, p. 717-735.

Wickham 2009 = C. J. Wickham, The inheritance of Rome: a history of Europe from 400 to 1000, London, 2009.

Wolfram 1988 = H. Wolfram, History of the Goths, trans. T. J. Dunlap, Berkeley, 1988.


1 Division of Empire : Demougeot 1951, with the crucial extension of the argument in Goffart 1981. For varying degrees of positive reception of Goffart’s ideas in some recent Anglophone textbooks, but they are all positive to some extent, see e.g. Mitchell 2007, p. 102 ff.; Innes 2007, ch. 2; Rollason 2012, p. 27-29.

2 Croke 1983; cf. its recent acceptance by Wickham 2009, p. 79.

3 See Bayless 1976 on the misunderstanding. For the general circumstances of Alaric’s first invasion of Italy, see Heather 1991, p. 206-213.

4 Soz., H.E. IX, 8, 6; cf. Zos. VI, 8, 2. Here and in the remaining elements of the list which follow, I generally follow the seminal discussion of Kaegi 1968, ch. 1, whose influence on what follows should be taken as read unless otherwise signalled although, for reasons of economy, I will not provide a precise reference at each footnote. In this instance, Kaegi’s discussion – Kaegi 1968, p. 16f. – already provided a necessary corrective to the dismissive comments of Demougeot 1951, p. 493, and for fuller discussion of the ‘comeback’ which Honorius’ survival of this moment of crisis made possible, see Matthews 1975, ch. 13-14; Heather 2005, p. 233ff.

5 Refs. as PLRE II, Ardaburius, p. 137f. and ibid., John, p. 594f.

6 Theoph., A.M. 5931; cf. Procop., Vand. I, 2, 39-40 with the discussion of Croke 1977, p. 360. The year could have been 421 rather than 427 because Theophanes says that the Gothic resettlement happened in the 19th year of Theodosius II, whose reign, at different points, Theophanes dates both from 402 (when Theodosius became Augustus) and 408 (when he became sole Augustus).

7 Refs. as PLRE II, Aspar, p. 164ff. A recent discussion of these events is provided by Merrills and Miles 2010, p. 52ff.

8 The so-called Law of Citations of 426 – Cod. Theod. 9, 43, 1 – is an excellent example of a western law whose potential impact in the east from 438 was enormous (even if its original significance was more limited, as argued by Matthews 2000, p. 221; cf. Humfress 2007, p. 67). For partly contrasting accounts of the legal reform process behind the Theodosian Code, see Honoré 1998; Matthews 2000; Sirks 2007.

9 Kaegi 1968, p. 28f. provides an excellent account with full refs.

10 Nov. TheodII 2, 2-3 of October 1 447. Its validity for the west was confirmed by Nov. Val. III 26 of June 3 448.

11 Hyd., Chron. 154 with the argument of Maenchen-Helfen 1973, p. 137f. vs. Thompson 1948, p. 148. The scale and precise direction of the East Roman manoeuvre is, however, unknown.

12 For a more detailed discussion & full refs., see Heather 2005, p. 390-407.

13 Brooks 1893 remains an excellent account of these court manoeuvres.

14 An excellent introduction to Constantine’s 50 volume project, of which only 4 partly or wholly survive, is Lemerle 1971, p. 280-288. On the likely origin of John of Antioch’s (fragmentary) material on the reign of Zeno in the history of Candidus, see Heather 1991, p. 237-238.

15 I would not myself be so dismissive. Establishing a regime in power with whom the east could work was a necessary precondition for further cooperation, in the same way that Fl. Constantius dealt with usurpers first in order to be able to unite the military establishment of the west against the Goths of Alaric/Athaulf and the Rhine invaders of 406 (refs. as above note 5). In this case, the eastern intervention against the Huns in Pannonia, if that occurred in 427 (note 6) should be viewed as a further outcome of putting Valentinian III on the western throne, since it may not have happened without that precursor.

16 Procop., Vand. I, 6, 1; Cedr. I, 613 Bekker; cf. Lyd, Mag. III, 43. Kaegi 1968, p. 44ff. provides more or less a maximalist commentary, but even more guarded accounts recognise that the expedition must have been ‘massive’: Merrills - Miles 2010, p. 121-122.

17 Croke - Crow 1983 and Whitby 1986, with refs., discuss the evident East Roman need to refurbish Mesopotamian defences in c. 500 AD. A good overall introduction to Roman/Persian relations in late antiquity is generated by a combination of Dodgeon - Lieu 1991; Greatrex - Lieu 2002; Dignas - Winter 2007.

18 The evidence is collected and analysed in Heather 2001.

19 On the rise of Hunnic power, see e.g. Thompson 1948; Maenchen-Helfen 1973; Heather 2009, esp. ch. 4-5.

20 The evidence is collected and discussed in Heather 2009, 238ff.

21 A detailed discussion of this period can be found in Heather 1991, pt. 3.

22 There has been a tendency in some recent literature to suppose that all the subject peoples chose to become fully integrated and fully enfranchised ‘Huns’ under Attila, and then chose to cease being so afterwards. In my view this massively understates the evident instability and internal contradictions of the Hunnic Empire; for a full discussion, see Heather 2009, p. 227ff. with refs.

23 In both Italy after the sojourn of Alaric’s Goths in 408-11 and Numidia after the Vandal occupation of 435-9, the Empire had to concede massive tax reductions (of up to 8/9ths of the normal revenues) to areas brought back under the imperial wing after ‘barbarian’ occupations: Cod. Theod. 11, 28, 7, 12; Nov. Val. III 13. A restored Proconsularis would presumably have seen similar levels of revenue disruption, but the Vandals had maintained Roman taxation structures elsewhere.

24 This is the only point on which I substantially disagree with Kaegi 1968, p. 29-31.

25 An excellent introduction in English is provided by a combination of Nelson 1992; Goldberg 2006; Costambeys et al. 2011. But there are of course equally good alternatives in all the major European languages.

26 On Stilicho’s manoeuvres, see Heather 1991, ch. 6. 424/5: refs. as above note 5.

27 The only ‘successful’ division of the Empire in the fourth century – measured as one which generated no known conflict – was that between the brothers Valentinian and Valens, which only lasted for about ten years from 464. Otherwise, all the non-dynastic Tetrarchic and dynastic Constantinian divisions generated consistent political tension between notionally joint rulers, and regular civil war, while Theodosius I found it impossible, subsequently, to share power. Famously Julian accused his cousin Constantius II of stirring up Alamanni against him; the evidence is discussed in Heather 2001.

28 Only the Duchy of Normandy among post-Carolingian entities was constructed around the military power of a new elite which had recent origins in groups of intrusive outsiders.

29 Croke 1983 engaging with Ensslin 1948; Wes 1967; Momigliano 1955.

30 Croke 1983, p. 114f.

31 Refs. as above note 14. 

32 Malch., fr. 14.

33 Malch., fr. 14 with Croke 1983, p. 115 on the coinage.

34 Euric: Jord., Get. 237, with the detailed account of Wolfram 1988, p. 181ff. On the full consequences of Basiliscus’ defeat right across the Roman west, see Heather 2005, p. 407ff.

35 473/4: Malch., fr. 5. On the final peace treaty, see now Merrills - Miles 2010, p. 122-124.

36 On the carefully-crafted Romanness of Theoderic, see, amongst many possibilities, Heather 1993.

37 He was hailed as semper augustus by the Senator Caecina Mavortius Basilius Decius (ILS 827), and the imperial pretentions of his ceremonial protocols emerge from the negotiations of the 530s recorded at Procop., Goth. I, 6, 2-5. On the extent of Theoderic’s imperial pretentions and their fundamental reality after 511, see Heather 1996, ch. 8. An excellent account what can be reconstructed of the Goth’s various rounds of diplomacy with Constantinople is Prosto-Prostynski 1994.

38 Jones 1962 is a valiant attempt to sort out the details of Theoderic’s precise constitutional position, but this was a moving target (see previous note) not least because the overriding necessity had been to get him out of the Roman Balkans: Heather 1991, ch. 9.


King’s College London -

© Publications de l’École française de Rome, 2015

Conditions d’utilisation :