Ancient History & Civilisation

Chapter 2

The Roman World – The Western Empire in the Fifth Century

Overview. Decline and fall? How much has it been exaggerated, and was it inevitable?

The basic argument is that the fall of the Western Empire was not inevitable, despite its comparative structural weaknesses that made it more vulnerable than the Eastern Empire. The latter survived in various, increasingly Graecicised forms until 1453 under a continuous line of Emperors: so why did the West collapse? Was it the long Western Empire frontier with the Germanic tribes to the north, open to penetration as soon as the Rhine and upper Danube were crossed? The East only had a shorter frontier on the lower Danube, and marauding tribes could be stopped at the Bosphorus and Hellespont. The East’s Gothic invasions from 376 saw the East’s army severely damaged at the Battle of Adrianople, yet the enemy were contained in the Balkans. The massive raids by Attila and his Hunnic-led empire in the 440s were similarly confined to the Balkans, as were the rampant Ostro[East]goths in the 470s and 480s. The mid-sixth century raids by the Kutrigur and Utrigur Bulgars reached the walls of Constantinople at a time when Justinian had secured control of Italy and part of Spain; this Balkanborn Emperor could not preserve his homeland from widespread ravaging despite an exhaustive programme of fortification testified to by Procopius.1 New arrivals in the Hungarian basin, the Avars, engaged in semi-permanent warfare with the Empire over the Danube valley and then Thrace from 568. We know little of how this endemic insecurity damaged agriculture and reduced the availability of peasant soldiers for the army, but it must have been a serious problem, and by the 580s the ravaged Balkans were being settled permanently by the Slavs. None of these attacks by a locally powerful foe ranged right across the East to its permanent disruption; and Asia Minor remained secure apart from one bout of Hunnic raiding south from the Caucasus, which even reached Syria, around 400.

But in the West three major Germanic tribes’ crossing of the Rhine in 406 led to permanent barbarian settlement in Gaul and Spain and in due course North Africa. This was not the case in the East, despite the mass movement of the Tervingi and Geuthungi Gothic peoples into the empire in 376–8 that was similarly militarily successful. The initial Gothic autonomous ‘federate’ tribal state in the Balkans, conceded to them by the East in 382 as they were too powerful to be evicted, was not a permanent solution, as seen on both sides. A Roman revival was hoped for by the Eastern orator Themistius in his up-beat propagandist account of the treaty to their Senate, where the Goths were portrayed as defeated and as turning into peaceful farmers; and a desire within the Gothic leadership for further pressure on the Empire was shown by the next Gothic leader, Alaric, in his aggressive behaviour in 395.

But had the Goths’ joint leadership of the 370s or a friendly Gothic ruler like Fritigern been in place, would this attack have occurred at all? After the mid-390s Alaric shifted his activities West, and Gainas, an over-powerful Gothic general who did secure supreme military command in Constantinople, was soon killed. The same Eastern ability at containment applies to the next two Gothic tribal states, both ruled by a Theodoric, in the Balkans in the 470s and 480s, as Emperor Zeno induced their unifier Theodoric the Amal to invade Italy in 490. But in the West, the Goths followed their wanderings across Italy, southern Gaul, and Spain by securing a ‘federate’ state in Aquitaine in 418, and other parts of Gaul and Spain fell away too. Britain was abandoned to its own devices in 410, and the Vandals (part of the 406 Rhine coalition) moved on from Spain into North Africa in 428 and secured its capital, Carthage, in 439.

Losses in the West were thus permanent, and each one weakened the state’s revenues, and ability to field an army, further. In turn, this encouraged further attacks. It was true, however, that the decline of Western power was not a smooth downward curve but came in sharp bursts. Each was precipitated by a specific political crisis. Due to personal charisma and military power, the Western supreme commander Aetius (in power 433–54) was able to call on the semi-independent Germans of Gaul to aid him against the invading Attila in 451, and central and northern Gaul were ruled by a mixture of Roman and German authorities until his murder in 454 saw the Goths turning on the fatally weakened Empire and extending their domains. Arguably, Aetius’ influence over his allies in Gaul was personal, not institutional, and he had a valuable past knowledge of Germans and Huns alike that aided his success. He had been an exile in the Hunnic state in the early 430s and used them as allies to regain power in the Empire. His vigorous campaigns against Germans and peasant brigand rebels (‘bacaudae’) gave him the respect of individual leaders, in an era when personal ties were crucial to such warlords.

Independence or autonomy for German polities in the Western Empire thus did not mean an end to Roman power or influence, with successive generations of Gothic leaders hankering after adopting Roman lifestyles or gaining Roman political influence. The Romanised social behaviour of Gothic king Theodoric II (reign 453–66) was praised by his Roman clientemperor Avitus’ son-in-law Sidonius Apollinaris, a contemporary Gallic poet-aristocrat who scrupulously aped classical literary culture. Indeed, the piratical Vandals in North Africa after Gaiseric’s time (post-477) adopted a sybaritic Roman lifestyle, to which their military decline after the 460s was to be attributed. In central and southern Gaul the Goths seem to have lived separately from the Romans, shunning the towns, and to have preserved their own culture and traditions, as described by Sidonius Apollinaris. The Franks mainly settled in less urbanised northern Gaul, and the only semi- Romanised sybarite ruler with cultural pretensions (as seen by Gregory of Tours) was Chilperic of Soissons, who died in 584. Landed estates in at least two ceded Roman areas, Gothic Aquitaine in 418 and Italy in 476, were formally divided between the two peoples.

In 408 the rebellious Alaric the Goth initially sought his late foe Stilicho’s supreme Roman commandership-in-chief from the supine Western Empire, and set up Attalus as his own puppet Emperor; the sack of Rome only followed the failure of his plans. He was seeking blackmail money to pay off his armies, not the destruction of the Empire, and made huge but manageable demands for gold and silver from the weak government of Emperor Honorius, then sought to replace it. His brother-in-law Athaulf spoke of wanting to fuse Roman and Gothic peoples into one state according to a story which reached the historian Orosius, and married Honorius’ kidnapped half-sister Galla Placidia.2 His murder in a private feud ended this attempt to set up a German-Roman state based at Narbonne, and his successors were driven west into Aquitaine in 418.

After the disasters of 454–5, the Goths of Toulouse used their military supremacy in Gaul to impose their own nominee, Avitus, as the new Emperor in leaderless Rome, thus seeking to influence the state rather than revolt against it. Avitus was a former supreme civilian official in Gaul so he knew the Gothic leadership, and had been sent to seek their alliance by the new Emperor Petronius Maximus after the latter had his predecessor Valentinian III murdered. This plan was forestalled by the Vandal attack on Rome. With Petronius in flight and killed, and Rome sacked by their rivals, the Goths then installed Avitus in his place. But the point is that now (455) it was the weakening Empire seeking Gothic military help, which it had firmly resisted when it was Alaric attempting to force his military assistance on the Empire in 408–10. In 416–18 Constantius III had been insistent on containing the Goths in far-away Aquitaine and recovering Princess Galla Placidia for himself, but after 454–5 the Goths, and then German generals in the Western army, were the senior partner in any Romano-German alliance.

Attila also sought to blackmail the Eastern Empire into sending him huge subsidies and gifts rather than conquering it in the 440s, though he did annex the middle and lower Danube valley from it too. He was aided prodigiously by sheer luck. His attacks and advantageous treaty with the East in 441–2 followed the departure of part of their army to fight Gaiseric in North Africa, and in 447 the walls of Constantinople and other cities were damaged by a massive earthquake. His open aggression towards the West in 451 followed an appeal from the disgruntled Princess Honoria for his hand in marriage which the government disowned, an excuse for a politically logical attack, but still useful to him. The flattering servility and massive bribes offered by the latest Eastern embassy to him, led by the supreme civilian official, ‘Master of Offices’ Nomus, had bought the East a temporary reprieve. Allegedly he had grudges against a Western banker for keeping plate promised to him or maybe was also bribed by Gaiseric the Vandal. He had already considered attacking Persia instead, according to Eastern envoy Priscus, but the geography was prohibitive as he would have to cross the Caucasus.3 His choice of Gaul, not Italy where Honoria could be found, shows practicality; it was easier to cross the Rhine than the Alps. The nature of this steppe-based state was clearly based on warfare by restless nomads, unlike the relatively settled German lands bordering on the Danube and Rhine frontiers, where farming not pastoral herding predominated and the Germans had long been semi-integrated into the Roman world as mercenary-supplying vassals.

A leader like Attila needed constant success and loot to keep his followers contented, and the Empire was the richest source of both. Indeed, as of the Romano-Hun negotiations of 411 there had been several Hunnic kings; the sole rule of Attila was a novelty. This meant that Attila’s power depended partly on his success in imposing unity as a war-leader, and partly in his role as the sole conduit of loot (or Roman bribes) to his warriors. War was more useful to him than peace and the Empire had far more gold than his German neighbours, though if they extorted huge Roman subsidies he could channel these as sole negotiator with the Empire. Buying him off permanently was an unlikely result of Roman appeasement diplomacy, given the way he shamelessly raised his demands year by year. Possibly the East paid, rather than fighting, in 442 and 447 due to temporary strategic weakness, not out of fear or military incompetence. Its army was occupied elsewhere on the first occasion, and the earthquake had struck on the second. Had Attila been satisfied with the results of blackmail on East and West alike he would still have needed targets to conquer, and we have seen that he considered Persia.

Botched plans by Eastern chief minister Chrysaphius to assassinate him in 449 and the apparent appeal to him by Honoria exacerbated tensions, but any wiser Western submission would have left Attila with a problem of keeping his warriors occupied. He would probably have sought other excuses for aggression and the East could hardly afford to pay him any more; his demands had already risen ten-fold in a decade. But his court included Romans as well as Germans and Huns, with his secretary being the Roman Count Orestes who was later to become father of the West’s last Emperor. It is too simplistic to present a notion of an irrevocable ‘Romans vs. Germans and Huns’ estrangement leading to the latter all pursuing a settled policy of seizing Roman territory. Rather, the more aggressive Germanic and Hunnic leaders made use of the opportunities that presented themselves in the decades after the first Danubian crossing in 376. The nature of newly established dynastic sole rulers, first Alaric, then Attila, in peoples used to no or multiple kingship encouraged the successful warlords to wage war and secure success and loot which benefited them personally.

It should be remarked here that the allegedly irrevocable, hostile Gothic crossing of the Danube by the Tervingi and Gaethungi in 376 was a refugee problem, a response to the loss of their steppe lands to the Huns, not anti- Roman aggression. The contemporary historian Ammianus claimed that Emperor Valens was pleased with their arrival as providing thousands of useful Gothic military recruits, at a time of rising tension with Persia (he was at Antioch in Syria preparing for war). He had previously negotiated successfully with these peoples as dependant allies at the end of a three-year war in 369, albeit probably forced to moderate his terms by the need to relocate east to a Persian war over Armenia. The Romans had been using their Danube neighbours for this purpose, and admitting thousands of agriculturalists to boost their denuded farming communities, for centuries. Constantine secured large numbers of recruits from the Goths in 331, and his son Constantius II did the same with the Sarmatians in 358–9. In recent years, one leading Gothic king (Athanaric of the Tervingi) had tried to limit, not extend, Gothic dependency on and supplies of troops to the Empire in the 369 treaty; the Hunnic attack forced a re-think as the Goths now needed sanctuary. The mass immigration in 376 was not a new phenomenon, either; the Empire had admitted thousands of Carpi from the Danube in 300. The main difference with the 376 phenomenon was that on the latter occasion the Goths obstinately stayed under the direct control of their own war-leaders; the Romans usually hastened to split bodies of armed immigrants up into manageable numbers under Roman command. Presumably this normal practice was Valens’ intention for 376–7 too, but was hampered by circumstances such as the sheer number of the Goths and probably the lack of Roman troops to supervise them at a time of war with Persia.

As of 376–7 the Goths were interested in land and food, not attack; the situation only turned ugly after they were moved on South to local Roman commander Lupicinus’ base at Marcianopolis and the Gaethungi crossed the Danube unilaterally to join the Tervingi. Lupicinus and other officials seem to have been operating a ‘black market’ in food-supplies and their extortion bred resentment. Valens should have sent reliable officials to avoid this in such a delicate situation. Lupicinus then panicked and tried to murder the Gothic leaders at a banquet, a logical move to decapitate the threat and hopefully force the leaderless Goths to obey Roman orders. Instead the targets escaped and war resulted, with Valens hundreds of miles away and unable to react quickly. The attempted strike at the enemy leadership was to be repeated, equally unsuccessfully, by chief minister Chrysaphius attempting to murder Attila in 449.

When Valens did arrive and march into Thrace in July 378, he seems to have expected to meet only around 10,000 Goths who he outnumbered, but faced at least twice or thrice that; possibly he had not heard that the Geuthungi had now linked up with his initial foes, the Tervingi. The size of the Gothic cavalry charge onto his army as it attacked the Gothic camp near Adrianople on 9 August then precipitated disaster.4Was his defeat therefore due to over-confidence or faulty scouting? It is arguable that what distinguished the disaster of 376–8 from successful Roman management of mass-immigration in 331 and 358–9 was that on the first two occasions the Emperor had been on the Danube with an army to supervise the process; in 376–8 Valens was in Syria and left it to under-resourced and corrupt military officials. The resulting damage to the Empire was permanent, but it was not an unavoidable invasion of the Empire by hostile barbarians.

The overall amount of Germanic looting and pillaging has also probably been played up by rumour and apocalyptic exaggeration by Christian writers, to whom the catastrophic collapse of the Christian Empire was a sign of God’s disfavour and portended the Last Days foretold in Revelation. In 395–6 the Goths ranged at will across the major sites of ancient Greece, sacking Eleusis, Sparta, and Olympia and blackmailing Athens into paying ransom, a major psychological blow to the Empire.5

In 402 Alaric attacked the Western capital at Milan by surprise, forcing the court to take refuge permanently in the inaccessible marshes of Ravenna, hardly the situation of a militarily confident government. 6 Thereafter Alaric returned to an uneasy role as a ‘federate’ ally based on the Illyrian border of East and West, playing them off against each other. An independent leader, Radagaisus, invaded Italy on his own in 405 and was defeated. Although our account of the attack (by Zosimus) is garbled it seems that he had nothing to do with Alaric’s Goths but crossed the upper Danube from Bohemia. The West was thus starting to attract copycat opportunistic invasions, and on 31 December 406 a multi-ethnic German coalition crossed the Rhine. Led by the Vandals and also including the Alans and Suevi, they rampaged at will across Gaul and produced apocalyptic comments about the end of civilization from local writers (e.g. Prosper); the lack of Roman Imperial military re-action led to the commander in Britain, Constantine (III), taking action unilaterally and claiming the throne. A revolt against his authority by his general Gerontius then enabled the Germans to move on into Spain, which was divided between them without any need to consult the Empire.

In 408 the murder of Stilicho left the West open to another invasion of Italy and threats to pillage Rome. Alaric shamelessly raised the stakes of protection money for leaving, and eventually lost patience. The Goths’ sack of Rome in August 410 was a relatively disciplined and organised affair, with the Christian, albeit heretic Arian, Goths treating the churches and the Papacy with some respect. Indeed it was a result of Alaric’s blackmail of the government in Ravenna failing to extort the pay-off he expected, not a longterm plan.7 If the Western military high command had not been decimated by the anti-Stilicho purge in 408 he would have been unlikely to reach Rome at all. He had after all simply been attempting to secure power within the Roman ‘system’ as commander-in-chief to his own new puppet-emperor, Attalus. But the psychological effect was immense, with St. Jerome in distant Bethlehem summing it up as symbolising the destruction of the world.

In reply to the pagan reaction that it was the gods’ revenge on the Empire for abandoning them, St. Augustine of Hippo wrote ‘De Civitate Dei’ arguing that the real ‘City of God’ was the new, spiritual Christian world not an earthly city. This was not a new reaction to the difficulty of fitting in the spiritual world of Christianity to a state that had initially persecuted it, and abandonment of the ungodly secular society was a desirable course for the virtuous Christian long before 410. But the sack of Rome gave Augustine an opportunity to establish a theological basis for the separation of the aims of Christianity and of the state, and to place the former as infinitely preferable. This fed into the claims of the Papacy to religious authority and prestige in place of the Emperor as lord of Rome, although Constantine had already given the Popes supreme jurisdiction over their ecclesiastical subordinates in the Western part of the Empire, effectively as ‘Patriarchs of the West’.

The Vandals’ sack in 455 was more brutal and secured a far greater haul of loot, but also opportunistic, and unlike Alaric, Gaiseric was not likely to be bought off before his forces attacked the city.8 Like Attila in 451, he used the excuse of wanting the implementation of a promise (this time in a formal treaty) of an Imperial heiress, Valentinian III’s daughter Eudocia, destined for his son Hunneric but unlikely to be delivered willingly to a barbarian. In political terms, it was extremely implausible that Gaiseric would have secured the Imperial succession for Hunneric. Even if the son-less Emperor had been forced to marry his elder daughter to Hunneric to avoid war, or after the murder of Valentinian his successor Petronius Maximus had done so, the succession would not have passed to Hunneric. The main political aim of Gaiseric in 455 was probably to forestall Petronius’ planned alliance with the Goths (via Avitus’ embassy), which could lead to a Romano-Gothic attack on the Vandals in North Africa. Had the alliance been implemented and Gaiseric not reacted, the Vandals would probably have faced the same dangerous level of attack from north, east, and west as they had in 441–2 with the Eastern Empire able to join in with greater German participation than earlier thanks to Attila’s death.

The written evidence suggests that what came to be known to much later centuries as the eponymous ‘vandalism’ by the Vandals in Rome and elsewhere, systematic and deliberate destruction, was an occasional rather than a commonplace occurrence. At most, Gaiseric collected all the valuable moveables he could and stripped the roofs from temples in Rome to carry off the precious metals. Most damage to the fabric of the Empire’s cities and towns was done gradually, not by concentrated barbarian assault. Across the West, buildings collapsed over decades for lack of maintenance rather than being pulled down by German attackers, and it is now suggested that the evidence of fires in excavated villas (e.g. in Britain) is not necessarily due to arson by passing Germans. Nor did hordes of Goths storm the walls of Rome in 410; the gates were opened for them by runaway slaves. In 455 Petronius Maximus fled the city and Pope Leo surrendered sooner than face a massacre.

There was widespread insecurity and anarchy, at least in some areas where governmental authority had collapsed, e.g. the mid-fifth century middle Danube written about by the local St. Severinus9. The decline in building standards of what little new works were undertaken, and the use of wood not stone, in the fifth and sixth centuries West suggests an inability to find adequate craftsmen or materials10. If this is not physical ‘decline’ into an atomised society, what is? But it should be remembered that in less affected areas such as mid- and southern Gaul, the local Romanised aristocracy were still in existence as a cultured, Latin-speaking elite and running the Church throughout the sixth century. The world of the 590s historian Bishop Gregory of Tours was post-Roman politically, but not culturally, and the Church remained a strong bond with the city of Rome. Even in seventh century Anglo-Saxon England the international links of the Catholic Church, restored to the Germanic kingdoms there from the time of St. Augustine’s mission in 597, could allow for the imposition of Theodore, a Greek, as Archbishop of Canterbury, who came from distant Tarsus in Cilicia in 669.11

The fall of the Western Empire was not the end of the international world of a Mediterranean-centred Church. Indeed, the concept of ‘Roma Aeterna’ as the centre of the civilised world now applied to spiritual rather than political leadership, and was played up by Pope Gregory the Great, who was from an old Senatorial family but with a monastery established in his ancestral mansion. The collapse of the central institution of the Senate did not occur in 476, as it was still functioning and given practical autonomy in Rome by the Romanophile Gothic king Theodoric from 493. It only went into eclipse after the disruption of the wars between Eastern Empire and Goths over Italy in 537–54, when Rome was captured several times and Gothic leader Totila once evicted its declining population.12

The thesis of a weaker Western army open to greater recruitment from unreliable German troops and Germanic supreme commanders has also been suggested as damaging to the West; the West had a Germanic supreme infantry and cavalry commander (‘magister utiusque militiae’) and effective regent, Stilicho, in 395–408 and eventually fell victim to more German generals after 455. But the East’s army also relied on extensive Germanic recruitment, as in 331 (Goths) and 359 (Sarmatians). The East’s senior German officers included one man who briefly held supreme military power in the capital (Gainas in 399–400) and one who served as military commander and chief minister (Aspar, 450–467). Both were murdered and their partisans massacred, as was Stilicho; but after Stilicho’s fall the powerless Western court was at the Germans’ mercy in 408–10. The East, however, fought off its Germanic challengers after their similar coups in 400 and 467. After Gainas and Aspar were killed their surviving troops were left at large in Thrace but could only plunder the countryside. Did the West face a more concentrated and resource sapping Germanic challenge than the East? Did its geography make attack easier and its containment more difficult?

Problems of state structure and control. Religion, governance, and failings in the structure of the bureaucracy

Much has been written about the top-heavy bureaucracy and court, and the vast armies imposed on the Empire by Diocletian at the end of the third century13. A new governmental system was set up after he took power in 284 (East) and 285 (West). He ruled for twenty years in a rare period of stability, though it is unclear how much was entirely his work as opposed to further innovation in the early fourth century. This was a time when Roman taxpayers’ ability to fund this had been weakened by decades of insecurity and disastrous losses of manpower to the plague in the early 250s. Greater centralization was a logical reply to the multiplicity of revolts by provincial military commanders in the 250s and 260s. The Empire also faced a much more organised and determined military threat in the East than before, from the time that the aggressive Sassanid regime and its armies replaced the decentralised Parthian government in Mesopotamia and Persia in the mid-220s.

A small household Imperial entourage in the early Roman Empire had become a large, Persianised staff with a strict hierarchy and protocol, reflecting the religious overtones of a semi-divine Emperor, living in a Sacred Palace surrounded by quasi-religious ritual, who came to be associated with the cults of ‘saviour’ gods. Initially the old Romano-Greek pantheon was preferred, with Diocletian identifying himself and his colleague Maximian with Jupiter and the monster-destroying hero Hercules. An alternative version linked the Emperor to the cult of the sun god, ‘Sol Invictus’, to whom Aurelian, restorer of unity and conqueror of autonomous regimes in Gaul and Syria in the 270s, built a massive temple in Rome14. This identification of the Emperor with the state’s protective gods later transferred to Christianity under Constantine the Great, with the Emperor as the ‘thirteenth Apostle’ and his court as the reflection of Heaven.15 As a result, the Church was co-opted into the State bureaucracy, with a hierarchy of bishops under district metropolitans and supreme Patriarchs reflecting the civil bureaucracy. The government also set the correct form of worship and the only legal form of Christian doctrine, as shown by Constantine at the Council of Nicaea in 325. The doctrine established there is still the basis of Catholic and Orthodox theology. Constantine enjoyed debating and pronouncing on theology, as approved by his Church panegyricist Bishop Eusebius. This was a new move for a Roman ruler, as the traditional Roman cults were rituals lacking in doctrine.

As a result, the mid and later fourth century and the fifth century saw Emperors changing the state Church’s doctrine to fit in with their personal beliefs, and enforcing their subjects’ adherence. Constantius II backed Arianism in the 340s and Valens did likewise in the East in the 360s. In the 440s Theodosius II backed Monophysitism. Success for one particular doctrine depended on its promoters having the Emperor’s ear; the Arian Bishop Valens was a major influence on Constantius II and in the 420s Theodosius II had to be dissuaded from imposing the controversial Christology of his new Patriarch Nestorius. Defeated at the Church Synod of Ephesus thanks to the forcefulness of the ultra-Catholic Egyptian bishops, Nestorius was banished to Syria and his sect flourished only outside the Empire, especially in Sassanid Persia. (It later developed an offshoot in China.) In the later 440s, Theodosius became a convert to the mainly Syrian and Egyptian Monophysite doctrine and attempted to impose it on his capital. In the manner of Henry VIII (or later Stalin and Chairman Mao), if the ruler changed his doctrinal beliefs everyone was required to follow suit.

On all such occasions, the death of the Emperor responsible saw an orthodox Emperor restoring Catholicism, his bishops denouncing the defeated doctrine as heresy and banning it. The triumph of Christianity in attracting State backing under Constantine opened up opportunities for rival interpreters of doctrine to win over Imperial support and lobbying by clerics at Court, and those who lost out could be portrayed as traitors to their Emperor. Such treason was a matter of dates and luck. The long career of Catholic champion St. Athanasius, Patriarch of Alexandria 328–73, saw him alternately in favour under Catholic rulers like Constantine and exiled under the Arian Constantius II and the pagan Julian. Arianism (called after the Alexandrian theologian Arius) only triumphed with those German tribal polities that had adopted it outside Roman borders in the fourth century, when Bishop Ulfilas had been able to proselytize North of the Danube. Constantius II backed Ulfilas’ mission, and re-settled him and some of his Gothic converts when they were driven out by the Tervingi leadership in 348. These tribes (principally Goths and Vandals) kept to this doctrine after they moved into the Empire, and this caused difficulty when they took over Catholic provinces.16 Some persecuted the locals (most notably the Vandals in Africa), though this could be linked to fears of the Catholics’ political intrigues with the Eastern Empire rather than heresyhunting intolerance.

It is a plausible theory that the Arian beliefs of the Goths ruling southwestern Gaul (and later Spain) helped to alienate them from the local Catholics and to keep them at arms’ length from the latter’s clergy and townbased bishoprics. Greater unity with the local population followed their conversion in the 580s. In contrast, historians have seen the choice of Catholic baptism by Frankish conqueror Clovis as inaugurating centuries of co-operation between Frankish rulers and their Church. The reconquest of Arian-ruled Vandal Africa by Justinian in 533 could be played up as a righteous liberation from heresy. But did the different faiths of Germanic conquerors and local ex-Roman subjects really exacerbate their mutual alienation to a dangerous extent? Or is this just a modern assumption based on the swift collapse of the Arian-led German states in Italy and Africa in the face of Eastern invasion?17 Would the conversion of the Goths and Vandals to Catholicism by a different set of missionaries in the fourth century have produced stronger post-Roman German kingdoms? The Papacy in Rome had more autonomy under the Arian Theodoric, living in distant Ravenna, than when it was supervised after 537 by Justinian and Theodora.

The failure of the religion of peace and love to tolerate its own dissidents, about which St. Athanasius was unapologetic, was duly ridiculed by the apostate Emperor Julian.18 ‘Paganism’ (a term from ‘pagus’ meaning ‘countryside’, i.e. the rural remnants of ancient Romano-Greek religious cults), was initially the persecutor of ‘unpatriotic’ Christians under Decius, Valerian, and Diocletian but from the 380s was exposed to legal bans on its public practice and seizure of temple property by Theodosius. This led to alienation between the aristocratic practicers of the ancient cults in Rome and the government in Milan and Ravenna. The Roman nobility, however, had been disconnected from court life and service since the Emperors moved out of the city in the mid-third century anyway. If there was any disastrous loss of competent manpower to governments from the end of allowing senators to serve as provincial governors, this was a mid-third century law by Valerian’s son Gallienus not a result of pagan alienation from a Christian Empire. Constantine re-employed the aristocrats. The reason for Gallienus’ measure seems to have been state fear of well-connected, rich aristocrats using their provincial armies to stage revolts. The Late Empire created a table of ranks with honorifics dependant on service, like Peter the Great in Russia.

Arguably the failure of the State to tolerate any dissident religious sects alienated many potential supporters and led to unnecessary diversion of time and resources to persecution, as well as to the flight into hiding or exile of the defeated. But was this only temporary and played up by the polemical partisans of the victims, such as the senior Catholic clergy under Constantius and Valens? Did most people quietly conform, as they had done during the Diocletianic persecutions? It has also been argued that the emergence of monasticism led to many religious people dropping out of state service (or paying taxes) to live a holy life in the wilderness and save their souls not the state. St. Martin of Tours, a soldier turned monk turned bishop, was one famous example; and the persecutions of Christians in the early 300s saw thousands of dissidents fleeing to live as monks in the Egyptian deserts.19

How much this factor mattered in the less monasticised West is uncertain. The peasant flight from the land due to oppressive taxes, demands for supplies to the army, and legal semi-serfdom as tied tenants (‘coloni’) was more of a long-term problem. The high profile wealthy who rejected their secular lifestyle to set up monasteries, live as hermits, or go on pilgrimage to the Holy Places, e.g. Paulinus of Nola, ex-Empress Laeta, and the heiress Melania, were played up by approving supporters like St. Jerome, but were numerically few. There were not many clerics, exempt from tax, and not contributing to civic society according to critics from Gibbon onwards, compared to the state bureaucracy. Did they really make a difference to the state?

In political terms, the new order did pose one major problem. The Emperor was now a much more remote and semi-divine figure than the first citizen of the early Empire or even the rough-and-ready military Emperors of the mid-third century, when a brutal survival of the fittest and frequent revolts meant that only the most competent (or luckiest) rulers survived long. The men who saved the Empire in the 260s, 270s, and 280s, such as Claudius II, Aurelian, Probus, Diocletian, Maximian, and Constantius I, were usually experienced soldiers of humble birth. But once order was restored, the Emperors were more often immobile in their capitals than fighting in the field, with limited knowledge of the world beyond their court, though this did not apply to the autocratic Constantine the Great, Valentinian I, or Theodosius. They had their own failings: Valentinian seems to have been over-suspicious and persecuted alleged plotters, and Constantine and Theodosius gave full support to brutal and corrupt ministers such as Ablabius and Rufinus. This was not new, as shown by the activities of Sejanus under Tiberius, Tigellinus under Nero, and Cleander and Perennis under Commodus, and in all cases the misrule of individual favourites at Court did not affect the lives of ordinary citizens. But now the State was under greater threat and needed competent officials to advise a Palace-bound ruler. The danger of all the ritual acclamations at public ceremonies (even the Games) and obsequious language from courtiers to the sacred Emperor was that the latter would confuse the image of harmony at Court with the reality of life in his Empire.

It is noticeable that the records we have of the most senior Palace-based ministers, and great provincial bureaucrats headed by the Praetorian Praefects, suggest a quick ‘turnover’ rather than a system of long-serving men who could grow in experience and competence. Late Roman bureaucratic office had to be bought, and applications for favours from these high officials had to be paid for. Official salaries were small – hence the predominance of officials from the landed ‘gentry’ class, like the mid-fourth centuryAntiochene orator Libanius, who could ‘pay their way’. A case-study of the East’s senior bureaucracy in the career (and complaints) of the sixth century John Lydus suggests that raising money was a major concern of officials, to pay off the debts from one’s previous promotion and to afford the next one.

Did all this lead to what we would nowadays categorise as a lack of concern for competence or (in senior ranks at the Palace) policy-making? It may not have been vital to the Empire’s survival, as were military mistakes, but it arguably led to such venal incompetents in high office as Rufinus. Rufinus did considerable damage to the Eastern leadership in the early 390s. His most eminent victims, ex-finance minister and current Praetorian Praefect Tatianus (exiled) and his son Proculus (executed), were the sort of capable and honest ministers who the Empire needed in the crises after 395. Zosimus’ attack on Rufinus’ intrigues, cruelty, and corruption, which Theodosius did nothing to halt, is hardly just political partisanship, and is an indictment of Theodosius’ abilities in civilian rule20. Would East and West have co-operated better against Alaric at the crucial moment of military crisis in 395 but for Rufinus, or would Stilicho have undermined any independent-minded Eastern chief minister?

There is also a highly significant trail of evidence pointing to the unwise political advice offered to impressionable Emperors by their (eunuch) chief chamberlains, from Eutropius under Arcadius to Chrysaphius under Theododius II and Heraclius in the West in 454, who advised Valentinian III to murder Aetius. Such household intrigue was not new, but it was more dangerous at a time of severe crisis, particularly if it concerned foreign policy (murdering Attila) or the army (murdering Aetius). Court favourites destroyed both Stilicho and Aetius.

After 395 no Eastern Emperor commanded his troops in person and the only Western Emperor to do so in 395–457 was the short-lived Constantius III. This probably gave him a reduced insight into the problems facing his government, let alone the lives of its ordinary citizens, as bad news could be filtered by his courtiers and he had a slower reaction to crisis. The sons of Theodosius the Great, Arcadius and Honorius, and the latter’s nephew Valentinian III seem to have been particularly politically inert and vulnerable to court factions. But this was a personal, not institutional failing; all three were weak characters. The charge that a remote and suspicious Emperor living in the closed world of the court was listening to bad advice, following the dictates of flatterers, punishing loyal and competent ministers at the behest of intriguers, and neglecting honesty for servility was made strongly against Constantius II in the 350s, and played up by his overthrower, Julian.21 It was not a new phenomenon, but the nature of court life made this threat to capable leadership from the centre greater than before, though it did not weaken the Eastern Empire, perhaps more used to autocratic rule by pre- Roman dynasts, as it did the Western.

Division of the Empire – militarily necessary, or extra problems?

Diocletian divided the Empire into East and West, with two senior ‘Augusti’ and two junior colleagues, ‘Caesars’ who would in due course succeed them. This was meant to avoid struggles over the succession by pre-nominating competent adult heirs but predictably collapsed into chaos. He and his partner Maximian abdicated in 305 and their deputies Galerius and Constantius I took over as the senior Emperors (‘Augusti’); however the new deputy Emperors never secured recognition and Maximian’s and Constantius’ sons joined in the struggle. The succession reverted to the ruler’s immediate family, brothers or sons, if available, with each ruler having his own court and mobile ‘comitatus’ field army. Constantine the Great, who had first challenged Diocletian’s system in 306 by claiming the succession to his father Constantius I, defeated Maximian’s son Maxentius (usurping ruler of Italy and Africa) in 312; Maximian, who had staged a come-back and then deserted his son for Constantine to betray him too, was also eliminated. Constantine reunited the Empire in 324 but left it in 337 to his three sons by Maximian’s daughter Fausta: Constantine II, Constantius II, and Constans. He intended to leave two sub-states to his nephews Dalmatius and Hannibalianus, but they were murdered.

The eventual reunion of the Empire under Constantius II (353–61) and his nephew Julian (361–3) was followed by the extinction of the dynasty in 363.This was avoidable: Julian was only thirty-two, had had two marriages without children, and was killed by an arrow in a skirmish on the River Tigris while invading Persia. Arguably, if the militarily competent Julian, who had defeated the massive Alemannic attack on Gaul in 355 despite minimal military experience and so was a natural general, had been able to complete his Persian campaign in 363, withdrawing in good order, the Empire would not have been weakened by the surrender of vital fortresses West of the upper Tigris which his successor Jovian agreed to.

Evidently panicking with the Persian cavalry pursuing his demoralised army, the Emperor Jovian abandoned this frontier region (the modern south-east Turkey) in return for peace, and so opened the way for extra Persian pressure on the Roman frontier, and Armenia, from a much more secure position than formerly. Major fortresses that had protected Syria and Armenia, such as Nisibis, were lost and the Empire forced to divert troops and Imperial policy-making to planning a military recovery. Was it thus that Valens was fatally distracted from the Danube in the 370s and the Empire put at greater risk of a disaster like Adrianople? Would Persia’s threat have been more containable but for the long-term disaster of Jovian’s treaty, or would an over-confident Emperor have neglected the Danube for the chance of glory in Mesopotamia anyway?

Jovian accidentally suffocated from paint fumes. His elected successor Valentinian, a capable Danubian general, took the West and handed the East to his brother Valens, against advice to empower an experienced minister or general. Valentinian died of an apparent fit of rage while shouting at some insolent German envoys during a Danube campaign in 375, aged fifty-four. If he had been alive in 378 he was experienced enough to have done better than his son in dealing with the Goths in Thrace, if not to destroy them.22 Would Valens have waited for his more experienced brother to arrive to join him before he tackled the Goths in autumn 378, unlike he did for teenage Gratian?

Valentinian I was succeeded in the West by his sons Gratian (sixteen) and Valentinian II (four), and after the violent deaths of both (and their usurping successors) his empire fell to Valens’ successor in the East, Theodosius the Great. The latter had been appointed by Gratian to clear up the Gothic war in 379, but Valentinian I, who executed Theodosius’ father, would not have given him this chance. Theodosius died within months (January 395) and left the two halves of the Empire to his two sons; this time the division was permanent. But the division could have been as ephemeral as the previous multiple divisions of 337 and 375, or Theodosius never gained the West at all had Gratian or Valentinian II survived and left their thrones to sons. For that matter, the triple division of 337 could have lasted and three Empires, not two, emerged with their capitals at Trier (to face the Germanic threat on the Rhine), Milan (to guard the Danube), and Constantinople (the East). The notion of a careful, planned permanent division of the Empire in 395 owes much to hindsight.

The State and army: too heavy a burden? Or irrelevant to the question of survival?

The multiple courts of a number of Emperors, ruling as colleagues, and their ministers were followed down the administrative hierarchy by a large number of provinces, many more than in the early Empire, plus over-governors, ‘vicars’, ruling groups of provinces. All these men needed officials, and the armies of over 300,000 men now included both the field armies and local frontier garrison ‘limitanei’. The precise size of the army is unclear, as it can only be estimated from the early fifth century administrative summary in the ‘Notitia Dignitatum’, which may have been out-of-date or reflecting an idealised picture. But it was clearly much larger than the twenty-nine legions (120–150,000 men?) of the early Empire. Now each province had its own military commander (‘dux’, leader or general, hence our ‘duke’); the civilian governors (‘praeses’) had been stripped of troops to discourage revolt. The army apparently took up two-thirds of the Imperial budget.23

Service in many professions such as the army and farming was meant to be hereditary to ensure continuity of manpower in an Empire denuded of manpower by war and plague. The costs of the Imperial courts (there was usually more than one Emperor after 286), officialdom, and armies made the tax base shoulder more costs than it had in the second century, coinciding with difficulties in payment. The always fragile monetary economy was weak, as seen in the hyper-inflation of the later third century and the drastic regulatory measures taken to counter-act it. It is noticeable that there was both a reluctance to serve in urban civic offices, the middle-class social and economic basis of city life who could not escape the burden of tax, and a flight from the land by agricultural workers, defying the legal requirement for hereditary service in both cases. The dramatic decline in the commissioning of inscriptions commemorating public works by the provincial ‘curial’ middle class has been taken as a definitive sign of their impoverishment, flight from their responsibilities, and probably a lack of concern for the early Empire’s pattern of civic improvement. There was certainly greater fourth-century spending on private than public life.

Indeed, the declining ability of the over-taxed urban middle classes to fund urban building projects and civic public life seems to have added to the shrinkage of towns and lack of new public works compared to the early Empire; the most impressive new building was carried out by the wellfunded and staffed Church. All this testified to a much different state of society from the relatively under-governed and prosperous first and second centuries AD. The new government interference in private life, prescribing belief and hunting down dissidents, first Christians, then heretic Christian sects and pagans, added to the sense of permanent crisis. But case-studies of places such as fifth century Syria and fourth to fifth century North Africa have shown that these areas were certainly still prosperous, with flourishing agriculture and farms, and even in raid-hit Britain there were grander villas than ever before in the fourth century. Recent estimates of the late Roman population and prosperity have painted a far less gloomy picture than was once assumed. On this basis the populace and the tax-hit ‘curial’ classes were better able to sustain the burden of taxation. Possibly earlier reckonings relied too heavily on Gaul and the Rhine valley, badly hit by the third century invasions.

Germans and Romans

The Empire was restored as a viable structure by Diocletian and his successors, but was still vulnerable to insecure successions and civil wars. Then from the 370s a series of massive incursions by ‘barbarian’ Germanic tribes seeking new homes, commencing with the Goths, added to the constant pressure of frontier wars against smaller, usually disorganised Germanic attacks from the north and a ‘cold war’ that erupted into sporadic conflict with Rome’s only major rival state, Sassanid Persia, to the east. The defeat and death of Eastern Emperor Valens at the hands of the Goths at Adrianople in 378 broke the myth of Roman military supremacy and encouraged further tribal attacks, besides leaving a permanent autonomous Gothic presence within the Roman borders that the Empire had to accept.

When the next chance for Germanic attack arose on the death of the strong Eastern Emperor Theodosius in January 395, revolt by the Goths followed; and incursions from north of the Rhine and Danube resumed. The Empire was permanently divided into the East, with its capital at Constantinople, and the West, with its official capital at Rome but the Emperor’s court now at Milan or Ravenna, and help from one half of the Empire to the other was not guaranteed. In 406 a coalition of tribes crossed the Rhine into Gaul; they had done so before, most notably in the 270s and 350s, but this time they could not be driven back.24

But was this permanent irruption of Germans into Roman lands the inevitable precursor of the fall of the Western Empire? The apocalyptic notion of a ‘Fall of Rome’ has been debated, as Roman civilization, its social and economic structures, its Church, and use of its Latin language survived across much of the post-Roman West; the main break in social and economic continuity occurred in the seventh century with the Arab invasions and collapse of Mediterranean trade. But the end of formal Imperial authority and the administrative structure of the Late Empire is undeniable, commencing with the outer provinces like Britain and the Rhineland.

The notion of a massive ‘volkwanderung’ of hordes of hundreds (or at least scores) of thousands of Germans across the Empire from 376 was played up by romantic nationalists in the nineteenth century, but is now largely discredited.25 For one thing, there had been substantial Germanic settlement within the Empire and recruitment to its army already in the fourth century, an era characterised by shortages of manpower in the Empire after the plagues and wars of the third century. As we have seen, there had been assimilation of bodies of thousands of Germanic warriors (and/or farmers) from across the Danube into the East after 331 and 358/9. There were Germans at Court and in high military command by the 360s, as shown by their names. Valentinian II as well as Honorius had a German commander-in-chief (Argobast).

Nor were the German peoples who invaded the Empire necessarily neatly divided into permanent and recognisable tribes, all strictly delineated by descent and recognising unchallenged war-leaders. What is attested by archaeology is that certain widespread Germanic agricultural settlements North of the Danube disappeared after the fourth century, e.g. the ‘Cernjachov’ culture in Wallachia and the Moldova-Ukraine region (largely peopled by Goths) and the ‘Przeworsk’ culture of Bohemia, Slovakia and southern Poland (the Vandal area). The abandonment of these settlements is obviously linked to large-scale immigration into the Empire, whether in the specific invasions testified to by the literary sources or more gradually. The culprits for the mass-movement were presumably the nomadic, horse based Huns.

Names that were later adopted by separate tribal kingdoms are used to define particular military groups who fought the Empire in the later fourth century and the fifth century, e.g. the Goths (East and West), Franks, and Vandals; but this is not to say that their armies consisted solely of men from that tribe. It is more likely that there was a fluid situation of assorted warriors (plus their families) from mixed Germanic backgrounds supporting the most promising leader available, with a successful leader attracting a growing coalition of warriors who might become a permanent, coherent grouping if they secured a geographical ‘state’ to settle in. Identity was as much ‘cultural’ as ethnic – by choice as well as descent successful warbands became the allied kingdoms that fought the Empire and were then sometimes recognised by treaty as occupying particular Roman territory. As described by the Roman historian Tacitus in the first century AD, Germanic kingship was divided between ephemeral military commands, created for particular wars by the election of the best or most forceful war-leader, and permanent judges and religious leaders. All came from hereditary noble or royal lineages; the former only emerged when needed.26

There is no continuous record of any dynasty ruling a stable kingdom of one particular Germanic people for centuries before the invasions of the Empire from 378, even when the name of a people is continuous. The ancient lineage of the Goths’ leaders was probably exaggerated by their sixth century propagandist Jordanes, back-dating its length to add to their descendants’ prestige. But Rome preferred to deal with individual kings, who it could control with gifts or military pressure, than with anarchic bands of warriors, and so built up its allies beyond the Rhine and Danube as rulers from the first century. All the evidence suggests that these kingships, e.g. of Maroboduus in Bohemia in the early first century, were personal and temporary. Arguably, indeed, it was the need for permanent military leadership as the Germanic peoples invaded and settled in the Empire that led to permanent kingships; Alaric’s relatives ruled the ‘Visi[West]goths’ from the 400s. But there were undeniably Germanic hereditary elites by the fourth century, with Athanaric the Goth (fl. 376) being the son of a former noble hostage in the Empire and 350s Alemmanic leader Serapio called after the god Serapis by his royal hostage father. The Alemannic-led invaders of the Rhine frontier in 355–6 had seven kings, and the Tervingi and Geuthungi Goths of 376–8 had multiple leadership too. Athanaric and others were linked to previous kings (i.e. war-leaders) or judges, but not to any long-lasting, permanent patrilineal rulership.

On an archaeological level beyond the Roman frontier, there is evidence of elite, maybe royal, residences in fourth century tribal territory lacking in earlier centuries, such as Urach, in Alemmanic territory. New wealth, and goods from the Empire, was clearly concentrated in a few hands, and status could be hereditary. Arguably, this posed an extra danger to the Empire that had been absent in the first or second centuries, as such leaders could look to war with Rome to boost their reputation and leadership. An agricultural revolution of greater arable yields has also been traced, as seen in the 1960s at the excavated fourth century settlements of Feddersen Wierde (Germany) and Wijster (Netherlands)27. Did this produce greater survival rates for the populace and a hunger for lands, leading to more pressure on the Empire?

All this is suggestive of a probability of extra problems for the Roman frontier and new aggression by an emerging royal elite of Germanic warleaders. But their tribes’ leadership in the fourth century was multiple, even when attacking the Empire (355–6 on the Rhine and 376–8 on the Danube). Individual kingship only emerges in the fifth century, among peoples traversing or settling in the Empire, and did not emerge among the fifth century Ripuarian Franks who were not in contact with Rome. Would these dynasties have emerged but for a permanent military confrontation with or settlement within the Empire?

Less successful warbands broke up and less successful tribes were amalgamated into their rivals’ kingdoms; the fifth century is littered with the names of leaders without their own tribal following (e.g. Odovocar) and tribes who disappeared from history (e.g. the Heruls). In Gaul, the Merovingian dynasty originally only ruled a small section of the Frankish peoples, based around Tournai in Belgium in the mid-fifth century. Many other Frankish groups (e.g. the Ripuarians on the Rhine) and their kings fell victim to Clovis and his sons after 481.28

Indeed, it should be recognised that the empire of Attila the Hun, the supreme charismatic non-Roman leader and loot-plunderer whose state collapsed after his death, was not a coherent grouping of Huns alone. The latter were described by contemporaries as Mongolian in feature, attacked the Goths on the Ukrianian steppes in the 370s from the East, and were assumed to be unholy savages (the ‘scourge of God’) by fifth century writers from their appearance and ferocity. Ammianus in 376 already recognised them as a distinct, aggressive, non-Germanic people relying on cavalry archers not the usual foot-soldiers.29 Attila himself seems to have turned on violence as a matter of policy, and received the East Roman embassy of Priscus in 449 coolly enough despite a recent Roman assassination-plot. But the state created by Attila from the Rhine to the steppes in the 430s and 440s consisted of Germanic tribes he had subdued as much as his own Mongolian steppe horsemen, and many of them revolted and resumed their tribal identity in 454. Other nomadic states generally referred to as Hunnic existed on the steppes and raided the Balkans through the later fifth century and into the sixth century, but ethnically are now known as proto-Bulgar. It is possible that their newly sedentary settlement on the Hungarian plain after 400 speeded up the Huns’ centralization into a state, changing their manner of leadership.

476 – a convenient but misleading date?

The end of the Western Empire is conventionally dated at 476 when the final Emperor exercising authority in Italy, Romulus ‘Augustulus’, was deposed; but in fact there had been a vacancy in a resident Emperor in the region in the 460s after the death of the shadowy Libius Severus in 465; commander-inchief Ricimer ruled alone, recognising the Eastern Emperor as his sovereign. The interregnum ended with the arrival of Anthemius with an army from the East in 467. After the deposition of Romulus in 476 the Germanic successor- state of Odovacer recognised the titular authority of the Eastern Emperor Zeno, a return to the situation of 465–7, not an innovation. It turned out to be permanent, with the regalia sent to Zeno in Constantinople, but he could have named a new Emperor as Leo I had done in 467.

Indeed, the real authority of a Roman Emperor, sitting in Constantinople, was restored to Italy, North Africa, and even part of Spain in the 530s-550s by Justinian’s generals and did not end in Rome itself until the eighth century. Ravenna, seat of Eastern Roman power in Italy, did not fall to the Germans until 751. The last Eastern Emperor to visit Rome was Constans II in 663, and he was resident for some years in Sicily not Constantinople. The Eastern Emperors regarded themselves as fully Roman despite residing in a Greek city on the Bosphorus and from the early seventh century using Greek as the official language of State, the modern term ‘Byzantine’ for the Eastern Empire is a seventeenth century invention.30 In parts of southern Italy, the Eastern government exercised authority until the Norman conquest in 1071.

The notion of a Western Empire falling and the creation of a medieval, post-Roman world in 476 would have seemed inaccurate to its contemporaries. As far as Justinian, ruling from the Cadiz area across the Mediterranean world to the Euphrates, and his heirs were concerned the Roman Empire was still functioning. It was indeed as a Roman Emperor of the ‘Romaoi’ that Constantine XI fell defending the ramparts of Constantinople against the Ottoman Turks on 29 May 1453; that date was the end of the Roman Empire. The term ‘Rumelia’ for the ex-Roman Balkans continued to be used by the Ottoman state, and the restored state of Greece provocatively numbered its first King Constantine (reigned 1913–17, 1920–2) as ‘XII’ not ‘I’.

The timing of the end of Rome was thus seen as different in the East than in the West. And other questions need to be asked about the undoubted end of a separate Imperial-led governmental structure in the West in the 470s. How much did the contemporary chroniclers exaggerate the extent of disruption to ordinary life in the Western provinces caused by the invasions? Would the post-378 Germanic invasions, whose numerical scope has been revised downwards in recent decades, have occurred but for a lapse in strong, unified Roman leadership? And were they bound to lead to the collapse of Imperial authority across the entire Western empire?

For example, the shocking sack of Rome by Alaric’s Goths in August 410 was traumatic but not politically decisive; it was followed by a revival in Imperial power in the central lands of the West under strong leadership by Constantius III in the later 410s. He tackled the Germans within the frontier as well as assorted Roman pretenders systematically, and restored a measure of peace by 418, only to die within months of gaining the co-emperorship. Britain, the Rhine frontier, parts of Gaul and Spain, and from the later 420s North Africa were already or subsequently lost, but after a bout of civil war unity was restored by new Imperial commander-in-chief Aetius from 433–54. He preserved Rome’s military predominance over its allies in Gaul and Spain, secured Eastern help to tackle Gaiseric, and fought off Attila the Hun (with Gothic help) in 451. Usefully, he had once lived among the Huns as an exile and understood ‘barbarians’ better than many haughty Roman aristocrats. The failure of Attila’s invasion of Italy in 452, probably due to plague as he did not suffer military defeat, and his death in 453, which caused his empire (a tribal confederation owing personal allegiance to one war-leader) to collapse seemed to leave the Western Empire intact within its shrunken post-420s borders.

Apart from Gaiseric the Vandal and his large Mediterranean fleet the Western Empire faced no serious military challenge, and even after Gaiseric had sacked Rome in 455 his realm came close to defeat by a massive Eastern Roman naval attack in 468. As far as military collapse was concerned, was the irreversible destruction of the Empire only dateable to 455 or 468? Even after that, Odovacar could have named a new puppet-emperor in 476 and kept the ‘kingdom of Italy’ in being under a titular Emperor until he was overthrown by Theodoric in 491–3.

A vicious circle of gradual collapse?

The final collapse of Imperial power outside Italy, and the reduction of the tax-starved government to being the victim of its German-dominated armies, only followed the murders of Aetius in 454 and his killer, weak Emperor Valentinian III, in 455. Even after the sack of Rome by Gaiseric’s Vandals later in 455 one competent military leader, Emperor Majorian, could revive Roman power to a limited extent in its central lands in 457–62. The East brought a new Emperor (Anthemius) and major military aid in 467–8, and its incompetent commander’s defeat in North Africa was arguably the final blow to the Empire’s security as a viable government controlling Italy and southern Gaul.31 Even after that, the over-powerful German military commanders at the Western court (Ricimer and then Gundobad) controlled Italy and kept up a series of puppet-emperors, a regime that could have survived for decades.

The decline of the central government’s control of its provinces was thus sporadic, and could be (temporarily) reversed by determined action by Constantius, Aetius, and Majorian; but each new crisis led to further aggression by its enemies against a state whose ability to fight back was weakening. It led to a vicious circle through the years 395–476; each loss of extra provinces and revenues weakened the Empire further, thus emboldening its ambitious neighbours to attack again. And on each occasion it was less likely to fight back successfully. It is no coincidence that the death of Theodosius in 395 was followed by Alaric revolting in the Balkans and playing off East against West. The murder of generalissimo Stilicho in 408 led to Alaric attacking Rome, the deaths of Constantius and Emperor Honorius, both in 421, led to civil war, and the Vandal attack on Africa, and the murders of 454–5 led to Gaiseric’s attack on Rome.

Potential alternatives and the crucial moments of Roman collapse

The proposal considered is that a crucial preservation of strong leadership in the later fourth and early-mid fifth centuries in particular should have seen off the challenges of the Germanic peoples to the unity of the state. Outlying provinces and their resources might have fallen away under the pressures from outside from around 395, but the core Empire would have remained intact if there had been domestic stability and an unbroken run of powerful military leaders from Theodosius the Great to Stilicho to Constantius III to Aetius to the latter’s heirs (e.g. Majorian). The collapse of the Western Empire was brought about by a series of internal crises and invasions from the death of Theodosius, with the first establishment of an autonomous German ‘federate’ kingdom within Roman borders (382) by the Goths in Thrace being followed by its successful attacks on both East and West, crucially at odds with each other at the time, from 395 to 410.

It is of paramount importance to remember that in the crisis of 395–7, when Alaric roamed at will across Greece and the south-western Balkans, the Western commander-in-chief Stilicho (regent for Theodosius’ younger son Honorius), a Vandal by birth, so looked down on by aristocratic Romans, brought aid to the East but was not trusted by their government. He was suspected of wanting to rule the East as well, Honorius’ elder brother Arcadius being mentally feeble and a puppet of his ministers. His panegyricist Claudian referred to him as entrusted with the regency of both Empires by his late master Theodosius I, a claim evidently not accepted by Rufinus who Stilicho’s German allies in the Eastern army then killed in late 395 after returning from service with Stilicho. He was also suspected of not destroying Alaric in Greece while he had the chance in order to use him as a military weapon against the East. Alaric’s army was able to survive a halfhearted Roman blockade in Epirus and secure another grant of a ‘federate’ state on the Eastern-Western borders. From this vantage point Alaric was able to invade the West in 402, and despite defeating him at Pollentia near Milan Stilicho failed to destroy him again. But had Stilicho or a competent Eastern general cornered and destroyed Alaric’s men, most obviously during the time they were trapped in Epirus, the Gothic survivors would have been reintegrated under direct Roman military command or driven into flight to the Danube. The Gothic troops of Gainas in the East met this fate in 400, after their leader and his closest officers were killed in a coup in Constantinople. Italy would then have been safe from the attacks that Alaric was to launch from Illyria in 402 and 408.

Alaric’s survival as a major military threat to East and West through the period 395–408 was fortuitous and partly due to internal Roman politics, namely Stilicho’s ambitions to rule the East; and when Arcadius died in 408 Stilicho apparently preferred to plan taking over the Eastern regency for Theodosius II to using the Western army against the invading Germans in Gaul. This gave the latter a breathing-space to spread further afield, whereas in the 270s and 350s the Empire had acted swiftly to contain them. In 407 Stilicho was planning some sort of move on Eastern-held Illyricum and Epirus with Alaric, not sending troops north to Gaul. Indeed, court resentment at Stilicho’s behaviour was stirred up by the courtier Olympius and caused the killing of him and his military partisans, who would have included quite a few competent officers in a violent purge in August 408.

This then left the West with a demoralised and faction-ridden army to face Alaric’s next attack, while the lack of state help to Gaul had led to the commander of Britain taking charge of the defence and proclaiming himself Emperor (Constantine III). The resulting civil war paralysed the West further, Alaric’s success in blackmailing the Western government into paying him off in 408–9 and then sacking Rome and ravaging Italy in 410 thus owed much to internal Roman politics not inevitable German military superiority. The West recovered after 410 under a competent new general, the later Constantius III, but the Goths under Alaric’s heirs had to be bought off with another ‘federate’ state (in Aquitaine) and the Vandals and other tribes remained at large in Spain. After another Western Roman civil war in 423–5 the Vandals were able to move on into Africa and revive the old threat of a naval power based at Carthage to Italy. Had there been no civil war then due to Constantius III not dying in 421, he would have been at liberty to attack the Vandals in the rear in Spain or even also reinforce Africa when it was attacked. The actual course of events saw the West’s two post-425 generals, Aetius and Boniface, at loggerheads in these years to the detriment of any counter-attack.

The major Germanic invasion-crisis on the Rhine of 406–7 was not the first such attack, and though its scale is unclear it involved several peoples and probably tens of thousands of fighting men. There were enough of them to spread out across Gaul and to move into northern Spain as mercenaries hired by Constantine III’s ex-general Gerontius and divide the peninsula up. Large-scale German invasions had been halted by the Emperor Probus in 276–7 and by Constantius II’s nephew, the ‘Caesar’ Julian, in 355–6. But this time Roman weakness enabled the invaders to remain at large within the Empire thereafter. The gradual loss of outlying provinces to the invaders meant loss of their tax revenues and a spiral of diminishing Roman resources, out-of-control Germanic kingdoms facing a weakening Imperial army, though strong leadership under Constantius III and Aetius delayed military-political collapse until 455, and a situation where each barbarian success emboldened more ambitious war-leaders to defy the Empire.

As of circa 420 the Empire had been stabilised by Constantius III, though analysis of the ‘Notitia Dignitatum’ indicates that 97 of the West’s 181 regiments had been raised since 395 and only 84 ‘originals’ from the C4th survived – an indication of disruption to half the army. The Rhine was worst hit. This stabilisation was then hit by the deaths of Constantius and Honorius, and the struggle between Aetius and Boniface to be Western commander-in-chief in the late 420s – leading to the Vandals invading Africa. The latter could not be expelled, and delivered the devastating blows of seizure or Rome’s grain-supplies and in 455 the second sack of the city. After Gaiseric’s attack, the new Emperor Avitus had to be provided by the Goths and his successor Majorian was chosen by the German general Ricimer – both signs of the West’s catastrophic decline in 454–5 (when it finally lost control of central and northern Gaul). Anarchy undermined trade and security, the State lost its ability to impose order outside Italy, and by the 460s the government was at the mercy of its German allies and riven by coups. Much of this could have been averted with strong and continuous Roman leadership, which was lacking at court for most of the period after 395.

This was not inevitable. Theodosius the Great, for example, died aged 47 and could easily have survived for another decade or two as sole ruler of the Empire. Given his military successes against the Goths (limited) in 379–81 and Western Roman rivals in 388 and 394, he was unlikely to have faced or been defeated by a challenge under Alaric (his loyal subordinate as of 394). Alaric or his followers may have revolted against Theodosius – suspected of exposing them to heavy losses in battle deliberately in 394 – in any case. But victory was unlikely. A competent military leader of the East in 395–7 could have defeated Alaric, which Stilicho failed to do; but the Empire had a feeble young ruler who never left his capital and a circle of vicious, feuding ministers like Rufinus (renownedly corrupt) and the eunuch chamberlain Eutropius. Stilicho’s murder and the Western military collapse in 408 owed much to the distraction of the Eastern crisis; though he might not have tackled the Germans in Gaul himself if he was free (to avoid being undermined at Court in his absence) he could have sent a capable subordinate.

The Western civil war of 407–12 was thus avoidable, and if Constantius III had not died in 421 he would have succeeded his brother-in-law Honorius in 423 and the West avoided a civil war then. Constantius was probably under fifty when he died; he could easily have ruled the West as ‘stand-in’ for his son Valentinian III into the 440s. A long-lived Theodosius, Stilicho, or Constantius would still have had to face the threat from the Huns on the steppes, which developed into a new ‘empire’ of anti-Roman tribes raiding the Empire in the 430s independent of Roman politics. But the West would not have lost control of the manpower or taxes of Britain, parts of Gaul and Spain, and later North Africa, and would have been in a much stronger position by the 450s.

Ultimately the first breach in the Empire’s defences and major blow to its prestige was the East’s military humiliation by the Goths in 378, in which the incompetence of its Emperor Valens played a major part. He owed his position to the fortuitous election of his brother Valentinian, who then chose him as his colleague, to rule the Empire in 364, following the sudden deaths of Emperors Julian and Jovian, and indeed Valentinian had been warned against choosing him.32 Moreover, the influx of Goths across the lower Danube, fleeing the Huns on the steppes, led to corrupt Roman officials mishandling them, extorting payments for inadequate supplies, and driving initially peaceful if armed incomers into revolt. As seen above, the venal black marketeer Roman commander Lupicinus then failed to decapitate the Gothic leadership and turned Gothic suspicion into rebellion. Valens was preoccupied with Persia, and unlike the situations in 331 and 359 was unable to handle the influx of militarily coherent, armed immigrants personally or to send enough troops to outmatch them.

Once rebellion broke out he marched into Thrace to tackle the Goths without waiting for reinforcements from his nephew Gratian in the West, though the latter was held up by trouble on the Rhine so Valens may have been anxious to strike before winter fell. The resultant battle was on 9 August. Valens was asked not to attack by the cautious Western general Richomer, commander of the advance-party sent to aid him by Emperor Gratian, but listened to the more confident Eastern general Sebastian instead.33 But Valens could have been deposed earlier. If a well-supported Constantinople revolt by Julian’s cousin Procopius had succeeded in toppling him in 365–6 he would not even have been Emperor at this point, though Gratian was unlikely to have assisted a usurper who had deposed his uncle. A different ruler in the East who was more interested in the Danube than in Persia in the 370s, better handling of the refugee Goths in 376–8, or a victory instead of defeat at Adrianople would have prevented the Goths establishing an autonomous state in the Balkans, which the East had to concede in 382. It would thus have prevented Alaric’s career of unchecked defiance within both Empires.

If you find an error please notify us in the comments. Thank you!