The Roman Empire did not end overnight, in fact its decline, like its rise, was a very prolonged affair. One can understand Gibbon’s decision to continue the story all the way down to the Turkish occupation of Constantinople in 1453, though it also seems a great mistake, since the Roman-Byzantine state can scarcely be called an empire at any time between the late seventh and the early eleventh century. What is not rational is the decision to write as if the Roman Empire ended as early as the sixth or even the fourth century ad.
There are also those who think that the word `decline’ should be considered taboo in this context, but no thinking historian of political power will hesitate over that issue. The power of the Roman state contracted dramatically, and, though it is a commonplace that all empires end, this particular contraction merits an explanation – or rather, as we shall see, a set of explanations.
The phenomenon we are considering took place in two major phases, the first between the 370s and the 430s – at the outer limit 455 – the second between the 560s and the years 636-41. In both cases, it is possible, and necessary, to consider antecedent causes, but these were the crucial decades. So our questions are the following: why in the first place did the Roman state lose so much of its power in the face of Germanic invasions over two generations between the 370s and the 430s, with the result that the Roman Empire in the west became a kind of ghost empire with very little power within its own former territories? This debate started as soon as the event took place, indeed before it was over, and started again during the Enlightenment. And why did the revived Roman Empire of Justinian, which included at its height Italy, almost all of North Africa and even some Spanish territory, as well as most of the old empire in the east, stumble after his death and finally collapse, as an empire, in the following century?
One central problem is military defeat and withdrawal; the other is the nature of power relations within Rome’s surviving domain. And how are these two sets of relationships connected? Should we suppose that the changed nature of power inside the Roman world had a major effect on the Romans’ failure, in either of the periods identified above? And conversely, how did these two periods of external failure change the power relations of the Romans with each other?
From the 370s to the 430s then. Few things are more difficult in late-antique history than to know why, in the western half of the empire, the Roman military and the Roman government failed: the sources and the archaeology simply never tell us enough, and modern historians have invented some unlikely tales (about the size of the Roman army, for example, and about the supposed absence of conflict). We can say very little that is useful about the German invaders: they obviously knew how to make war very effectively, and we may imagine that they had learned some practical lessons from the Romans – but how much any of this was a recent or decisive development is scarcely knowable. And of course we need to understand how the eastern emperors, their ministers, and their armies managed to fare decisively better. Did internal vulnerabilities of some kind make a difference, or were the external forces in the west simply stronger than Rome’s enemies in the east?
It is no use arguing that the German invasions were not a hostile takeover or that there was no crisis. The central government in the west gradually lost control of the provinces. When a leading historian argued that there cannot have been much of a crisis because the fall of Romulus Augustulus in 476 had little contemporary impact, it was scarcely probative, since by 476 the actual western crisis was already over. More recent scholarship, best exemplified by Bryan Ward-Perkins, has re-established the indeed rather glaring fact that the power of the Roman Empire in the west had by 476 largely disappeared. The Vandal conquest of the Maghreb provinces, culminating in the capture of Carthage in 439, made a major recovery impossible, for financial as well as military reasons. What is most intriguing here is that scholars lost sight of these facts, from the 1960s onwards, during an exceptionally prolonged period of bourgeois western comfort when the impact of the state’s military performance on the individual was relatively limited. In the late fourth and early fifth centuries, by contrast, it was impossible, almost anywhere in the Roman Empire, to be unaffected by war.
The problems that most contributed to military weakness in the 250s and 260s – unstable government at the centre and inadequate funds to maintain larger military forces – had gone into remission, but after 337 they to some extent returned. At least thirteen men claimed the title of Augustus in the forty years after Constantine’s death. In the 370s it could be said by an informed observer that `the vast expenditure on the armies . has thrown the entire system of tax payments into difficulties’ (Anonymous, On Warfare [De rebus bellicis] 5.1, trans. E. A. Thompson, with details; see further Ammianus Marcellinus 20.11.5 on the same problem in 360). But for all the money that was spent on them, these armies failed to defend the empire’s western half. How exactly should we estimate the performance of Rome’s military forces over this sixty-year period (370s to 430s)? For several decades now, some contemporary writers have taken to denying – astonishingly – that anything went badly wrong, which seems to have been a curious and distorted echo of the `explosion’ of late antiquity that has occurred in academia over the last half-century. One such scholar writes in an authoritative publication that `the performance of the army during the decades after 378 was by no means an unmitigated disaster’, which invites three main responses. First, you must not write rhetorical history – if the Roman army was not `an unmitigated disaster’ in this period, let us at least admit the obvious fact that it was a disaster. Second, `the performance of the army’ is too narrow a concept to allow us to assess Rome’s military effectiveness: we need to include among other factors the political leadership, and those who did not fight but should have fought. And third, no one doubts that there were good soldiers fighting for the Roman state – not enough of them, however, and they were too often fighting against their own side.
The pollyannaish view is indefensible, in fact absurd. Vegetius, probably writing under Theodosius (379-95), knew better (Epitome of Military Affairs 1.7, `so many defeats’). The government of the western empire (based not in Rome now but in Milan, then from 402 onwards in Ravenna) steadily lost control of large parts of its territory from the 380s onwards, mainly to Germans. After the disastrous Roman defeat at Hadrianople in 378 and subsequent events, large sections of the Balkan provinces had to be conceded to Goths whose periodic willingness to fight in alliance with Rome did not make them Roman subjects; the Pannonian provinces (even Savia, quite close to Italy), Lower Moesia, Thrace, and Macedonia were all deeply affected. Theodosius nonetheless accepted them, under their new young leader Alaric, as allies in his war against the usurper Eugenius, which culminated in the Battle of the Frigidus (in western Slovenia) in 394. Being disappointed of a Roman command, Alaric proceeded to do extensive harm in Thrace, Greece, and Epirus. In about 398, in spite of this, the court of Constantinople appointed him to military command of the diocese of Illyricum (Claudian, Gothic War 535-9, Against Eutropius 2. 214-8, etc.), from which, in 401, he invaded northern Italy. Until that happened, Constantinople had probably seemed to be in more danger from barbarians than Italy was; the Asia Minor provinces were often in almost as much trouble as the ones on the Danube, and there were Goths living in Phrygia by the 390s. But though Alaric was not immediately victorious south of the Alps, Rome’s highly competent commander Stilicho was unable to dislodge him.
Alaric’s wars in and close to Italy in the years 401 to 410 were by no means always successful, indeed the battle he lost to Stilicho at Pollentia in north-western Italy in 402 may have involved severe Gothic losses (the matter is debated). But the whole peninsula as far south as Rome itself suffered badly during these campaigns. There is really no sign that Alaric was aiming at bringing down the empire of Rome (though Claudian, Gothic War 530, may imply such an interpretation); what he claimed in 409 was resources and territory within the empire (Zosimus, New History 5.48; cf. Orosius, History against the Pagans 7.38.2). But these Italian campaigns of his, combined with the invasion of other very numerous Goths under Radagaisus in 406, left a trail of destruction, most intense in the north. That is plain enough both from the texts and from the archaeological records of the Italian towns and countryside.
None of the western European provinces was secure after 406: Vandals, Alans, and Suebi effectively took control of large stretches of the German and Gallic provinces in 407-8, before moving to Spain. Alaric famously seized Rome on August 24, 410, and then proceeded south to Calabria. After his death later in that year, his successor Athaulf eventually in 412 took the Goths in question to Gaul. Meanwhile rebellious Bagaudae once more caused trouble in north-western Gaul. The consequences of all this loss of power were severe, and can only be minimized by scholars who are unable to visualize the effects of large-scale invasions. And as the government lost control of tax-paying areas such as Pannonia, so it could afford fewer well-equipped soldiers and thus was less able to defend the old frontiers, which in turn gradually led to still further losses – a vicious circle in short.
In 418 or 419 Rome ceded to the Visigoths an area in the Garonne Valley of south-western France, which the latter steadily expanded over the next two generations until it included much of France and Iberia. In 429 a force of Vandals crossed the straits into the African provinces, and promptly headed eastwards towards the region’s greatest wealth, reaching Hippo in 430 and taking Carthage in 439. They used the ships they captured there to raid Sicily and, in 455, Rome itself, with very severe effects,. In 442 Valentinian III had recognized their North African claims. Italy, with Sicily and Sardinia, gradually came into German hands in mid-century. The western emperors who nominally ruled between 455 and 476 – Petronius Maximus, Avitus, Majorian, Libius Severus, Anthemius, Olybrius, Glycerius, and Romulus `Augustulus’ – ruled at most over parts of Italy and Gaul, had very few troops of their own, and had no prospect of improving their position. Romulus was not even worth murdering when he was deposed. The men of power on the Roman side, Aetius (from 435 to 454) and Ricimer (457 to 472), the latter half-Suebian and half-Goth, did not attempt to become emperors; nor did they succeed, in spite of their evident ability, in arresting Rome’s military decline.
What the Germanic invasions meant on the ground has to a certain extent been disputed, but at least it is clear that they meant vastly diminished Roman revenues. And there is really ample evidence that in very many districts death and destruction, including the destruction of fixed capital, were widespread. Refugees were evidently numerous, especially in Illyricum. The effects of the invasions were all the worse because the invaders often fought among themselves (Goths against Vandals in Spain in 413-15, and many other such conflicts in the Iberian peninsula described in the Chronicle of the Spanish Hydatius) – thus many places suffered from more than just a one-time conquest. Gaul, Raetia, Noricum, Italy, Spain, Mauretania, and Numidia were all deeply affected. Archaeology trumps any desperate attempt to discount the textual evidence. The constant trend was for the central government and its agents to leave the local communities to their own devices, as far as defence was concerned: thus Romanus, commander of Rome’s forces in north Africa, refused to help Lepcis and the other coastal cities when they were attacked in the 360s by the Austuriani from inland (Ammianus Marcellinus 28.6; not a simple story) (and later see for instance Valentinian, Novels 9 [ad 440]).