Ancient History & Civilisation


AD 284–305

Reign of Diocletian

AD 303–11

The Great Persecution

AD 306–37

Reign of Constantine

AD 313

Constantine’s Edict of Toleration

AD 325

Council of Nicaea

AD 361–3

Reign of Julian

AD 376

Valens allows parties of Goths to cross the Danube, beginning the events leading to the defeat of the eastern Roman army at Adrianople

AD 395

On the death of Theodosius the empire is ruled by Honorius in the west and Arcadius in the east, both of them minors

AD 409–75

Progressive conquest of Iberian peninsula by Visigoths

AD 410

Rome sacked by the Goths

AD 429

Vandals invade Africa, capturing Carthage in ad 439

AD 435–8

Compilation of Notitia Dignitatum

AD 442–52

The Huns, led by Attila, ravage the Balkans, Gaul, and Italy

AD 455

Rome sacked by the Vandals

AD 476

Last western emperor deposed by Odoacer the Ostrogoth



When Polybius of Megalopolis decided to make a record of the most significant events of his own day, he thought it was appropriate to begin by demonstrating on the basis of the facts, that the Romans did not win a great empire in the six hundred years following the foundation of the city, even though they were regularly at war with their neighbours for all this period. On the contrary, that occurred only after they had captured a part of Italy and then lost it once again after the invasion of Hannibal, the defeat at Cannae, and only after they had actually seen from their walls the enemy threatening. Only then did they begin to be so favoured by fortune, that in less than fifty-three years, they took control not only of all Italy but all Africa as well. The Iberians of the west submitted to them. Setting out on an even greater project they crossed the Adriatic, conquered the Greeks, and dissolved the empire of Macedon, capturing alive their king and carrying him off to Rome as a prisoner. Nobody could attribute this success to human might alone. The explanation must lie in the immutable plan of the Fates, the influence of the planets, or the will of God that favours all human enterprises so long as they are just. For these things establish a pattern of causation that leads future events to come out in just such a way, that shows just how right they are who believe that human affairs are subject to some kind of Divine Providence. So that when their energy is aroused, they flourish; but when they become displeasing to the gods, their affairs decline to a state like that which now exists. The truth of this proposition will be demonstrated by the events I will now relate.

(Zosimus, New History 1.1.1–2)

Emperors and Christians

The Emperor Diocletian ruled from AD 284 to 305. Constantine I from 306 to 337. They were not the only emperors who reigned during this half-century, and civil wars characterized the reigns of both. But the length of their reigns is a sure sign of increased imperial stability. Another sign is that the frontiers, although never peaceful, held. The military crisis that had nearly resulted in imperial meltdown seemed to have been averted. It was once common to write about the reigns of these two emperors in terms of transformation, reform, and recovery. But that is too simple. The institutions of the Roman Empire were indeed transformed: but many of their ‘reforms’ were unsuccessful, and the recovery was partial and, in the west, short-lived.

One transformation in particular affects all histories of this period. During the second and early third centuries AD the religious diversity of the empire had gradually resolved into a world of competing religions. How that happened is the subject of the next chapter, but its consequences have to be explored here. During the nadir of the military crisis, the 250s, the emperors Decius and Valerian had each tried to use general hostility against Christians to create a wider sense of imperial unity. Diocletian’s response was more extreme. His Great Persecution was a systematic attempt to eliminate Christianity, and it traumatized great swathes of the empire between AD 303 and 311. Constantine’s tactic was the opposite, to first tolerate the new religion, then protect, sponsor, patronize, and eventually seek to regulate and unify it through an ecumenical council held in 325 at Nicaea. History still remembers Diocletian as the Persecutor, Constantine as the Convert. Greek and Roman historians took radically different views of these events, depending on whether they embraced the new religion (as did the bishop Eusebius of Caesarea who invented a church history and wrote a panegyrical life of Constantine) or whether they deplored the abandonment of the ancestral religion, as did Zosimus whose verdict stands at the head of this chapter.

The divided reaction of historians mirrored the divided response of the empire’s elite. Before the end of the third century AD there had been many kinds of historical writing in Greek and in Latin—local and global histories, total histories of Rome, contemporary histories, and histories that were more like a series of imperial biographies placed end to end. Some historians and biographers stressed the mythological and the marvellous, others were closer to satire and scandal-sheets. But all reflected a set of common idealsabout the role of the emperor. Those ideals were a blend of Greek ideologies of kingship, and Roman notions of good citizenship. Good emperors were just, were successful in battle, deferred to tradition, respected the rights (especially the property rights) of the elite, were modest, merciful, and did not raise new taxes. Their sex lives were dull and unimpeachable. Bad emperors had all the opposite vices: think Caligula, Nero, Domitian, or Commodus.1 Now one new criterion trumped all the others. How did he stand with the Church? Was he a persecutor or a protector, and then later was he orthodox or a heretic?

No historian was neutral. Christians celebrated Constantine as a saint and the second founder of the empire, and they savagely condemned those emperors, like Diocletian and Galerius, whom they remembered most of all as persecutors. Perhaps this is understandable for the generation that lived through the Great Persecution and Constantine’s patronage of the Church. Lactantius was an African rhetor, summoned by Diocletian to teach at his eastern capital of Nicomedia and then sacked in the persecution of 303. But he moved west, and survived to tutor the eldest son of Constantine. Towards the end of his life, Lactantius composed the grisly On the Deaths of the Persecutors which recounted the gruesome punishments God reserved for Galerius, Diocletian, and the others. Eusebius and Lactantius offered a new vision of imperial history as part of God’s unfolding plan. Conversely, those writers who were not Christians deplored the decreased support for civic cults, the casual licence given to acts of violence against their temples, and what they saw as the ruinous consequences of abandoning the gods.2

But if we set aside—for just a moment—the matter of their contrasting attitudes to the emergence of competing religions, neither Diocletian nor Constantine was entirely unlike the emperors who had preceded them in the last years of the third century AD.

Both, to begin with, were soldier emperors. Like many of the emperors and would-be emperors who rose and fell during the third century, Diocletian (Diocles by birth) originated in the Balkans. Nothing certain is known about him until his entry into history as the commander of the Emperor Numerianus’ bodyguard in 283. Numerianus’ father Carus was a praetorian prefect who had rebelled against Probus in 282. Carus was killed in 283—we do not know how or by whom—and Numerianus ruled for only one year before being murdered in his turn by his own praetorian prefect, Aper, but it was Diocles whom the army hailed as Augustus. So far, so conventional. Equally conventional was Diocletian’s first campaign, against Carus’ other son (and Numerianus’ co-emperor) Carinus whom he defeated and killed in 285. For most of the next decade he fought, first on the eastern frontier, then on the upper Danube, then back in the east, while his ally Maximian, who came from a similar lowly background also in the Balkans, served first as his Caesar and then as his fellow Augustus, mostly on the western frontier. The collaboration became formalized and more complex in 293 when the two adopted two younger generals, Galerius and Constantius, as their Caesars. The four emperors (the tetrarchs) successfully worked together until 305 when, on Diocletian’s initiative, the two Augusti stepped down, the two Caesars replaced them, and appointed two new Caesars. For most of Diocletian’s more than twenty-year-long reign he and his fellow emperors moved back and forth between bases along the northern and eastern frontiers, and for most of the time they were at war with the enemies of Rome.

The success of the tetrarchs depended in part on the achievements of earlier soldier emperors like Gallienus and Aurelian. Roman armies were now better adapted to war against the northern barbarians, the cities of the east were now fortified bases, and, unlike the emperors of the mid-third century, Diocletian and Maximian were able to fight most of their wars on the frontier or on foreign territory. The great innovation was solidarity within the imperial college. The succession of coups and failed coups which ultimately brought Diocles to power was largely suppressed, although it took a while to control Britain. Diocletian invested heavily in ceremonial and titles, but perhaps it was his military success that ensured he faced fewer challenges. Once secure it was possible for him to make other changes, building more defences, and increasing the size of the army, while in order to support this he modified its command structure, and the way provinces were governed and taxed. These changes were not the implementation of a grand plan, but the cumulation of pragmatic expedients. Many built on the more successful experiments of earlier emperors, all of them were focused on the needs of the army.

The joint abdication of 305 marked the end of consensus. Even before Diocletian’s death, probably in 312, the carefully plotted succession plan began to unravel. Among the changes were the death of Constantius in 306, only a year after his elevation to the rank of Augustus, and the succession of his son Constantine. Constantine the Great himself died in 337, but he was sole emperor for only the last of his three decades in power. Before then, relations between the emperors shifted back and forth for a decade, a decade in which Constantine himself campaigned against the Franks. By the end of 312 a new pattern had began to emerge. Galerius was dead (devoured by worms if you believe Lactantius, but not before he had formally ended the Great Persecution with an Edict of Toleration); Constantine had won a decisive victory over his rival Maxentius at the battle of the Milvian Bridge which gave him control of Rome; and he had formed an alliance with Licinius, who was able to eliminate his rival Maximin Dia the next year. The alliance with Licinius was a stormy one, but it was not until 325 that Constantine was able to defeat and execute him and rule alone. By then Constantine had celebrated twenty years in power, and begun the creation of a great new capital on the site of Byzantium, to be called Constantinople. The final decade of his life was divided between wars on the Danube. and trying to create a new college of emperors from his three surviving sons (he had had a fourth, Crispus, executed in 326) and nephews. Like Diocletian he had spent much of his reign engaged in foreign wars.

It is difficult to say whether Diocletian spent more energy trying to suppress Christianity than Constantine expended on trying to reconcile its factions. Early after his public sponsorship of Christianity he was drawn into the bitter Donatist schism in Africa, and one reason for the Council of Nicaea was an attempt to develop a single view on the nature of Christ, a response to what became known as the Arian heresy. Christian writers, Eusebius above all, focused their attention on Constantine’s relations with the Church, his personal journey, his building projects, and the Council. Yet like Diocletian before him, he was also concerned with changing the military and civilian command structure, with raising taxes, and with changes to the coinage. Diocletian and Constantine were both extraordinarily successful members of a new species of emperor, one that had emerged during the third century. It is convenient to call them soldier emperors, but they were also exceptional managers who seem to have thought of the empire first, rather than the city of Rome, let alone the Senate and people. Neither spent much time in the interior of the empire or Rome itself. The old aristocratic orders of senators and equestrians were marginal to their attention. As for religion, perhaps they were motivated by feelings of resentment or passionate conversion. Who can say? But their policies—persecution, toleration, and promotion alike—were all about imperial unity. Constantine in particular had plenty of opportunities to play the zealot, especially against heretics, but he resisted them all. Bishops felt themselves influential at his court—and so some were—but it is difficult to identify any area of policy where Constantine’s commitment to Christ did not also serve his vision for the empire.

The only emperor after Constantine who was not a Christian was his nephew Julian, born just six years before Constantine’s death in 337. Julian’s childhood was lived against the background of civil wars conducted among Constantine’s heirs. Within a year of Constantine’s death two of his nephews had been murdered and the empire was divided between Constans, Constantine II, and Constantius II. Constans defeated and killed Constantine II in 340 and for ten years divided the empire with Constantius II. Constans himself was killed by a usurper in 350. By 351 Constantius II was sole Augustus, a position he held until his death in 361. For much of this time Julian was kept out of public life. But when his brother Gallus was made junior emperor, with the title of Caesar, in 351 his own return to public life must have seemed inevitable. Gallus was executed for treachery in 354. The next year Julian was made Caesar in turn and given a command in Gaul. We can only speculate on the effects on Julian of this history of familial murders and intrigue. But we do know that in his twenties, partly as a result of his own reading and partly under the influence of the philosopher Maximus of Ephesus, he rejected the (Arian) Christianity in which he had been brought up and secretly embraced—not too strong a term in his case—a very idiosyncratic and highly intellectualized version of what he regarded as the ancestral religion, a broad polytheism in which the gods of the Romans, the Greeks, and the Jews all had their place. He called this Hellenism. It is difficult for us to avoid the name the Christians gave it, paganism. But the cults of the ancestral gods never formed the kind of connected organized entity we usually mean by religion except in the imagination of Christian writers. It is an irony of Julian’s vision that the nature of the paganism he tried to restore and institutionalize, both its cosmological coherence and the charitable institutions he wished to encourage, is one of the clearest testimonies of his Christian upbringing.

Given the history of Constantine’s family no one can have been surprised that in 361 Julian rebelled against Constantius II. Only the latter’s death prevented another civil war. But when it transpired that Julian was not only not a Christian, but was a passionate advocate of the ancestral religion, the empire went into shock. Along with wars against Persia and hostilities with his brothers, Constantius II had also been embroiled in the great religious controversy of the age, inspired by the teaching of Arius, that Jesus, the Son, was completely subordinated to the Father. Constantine had tried to impose a compromise at the great council of bishops assembled at Nicaea in 325, but the controversy simmered on. Now, suddenly, all that was swept away. Julian’s court honoured Neoplatonist philosophers, not bishops. During his brief reign he wrote feverishly about his ideas, tried to ban Christians from teaching, restored funding to civic cults, planned to rebuild Jerusalem, and attempted to reorganize the old cults as a kind of counter-church. The opposition he faced from all sides showed the deep penetration of Christian ideology and the empire, especially among the ruling classes. Would Julian have made more progress if he had not died in 363 of a wound sustained in a new Persian war? It is impossible to say. As it was, the memory of ‘the Apostate’ was reviled, and his successors threw themselves enthusiastically back into their struggles with the bishops over orthodoxy.

Did that settle it? The reign of Julian seemed to have showed that although the old gods still had some devotees, Constantine’s transformation of the empire’s institutions had gone too far to be reversed. But proponents of the ancestral religion had one last moment in which to denounce Christ and his followers. For in the century that followed Julian’s death, a slow-moving disaster engulfed the Roman Empire. Diocletian, Constantine, Constantius II, Julian, and his successor Jovian all fought wars against the Persians and these continued into the fifth century. These wars consumed resources and lives without leading to any radical shifts of power between the ‘brother emperors’. Over time the two empires would come to seem more and more alike.3 Conflict between them would continue off and on until the Persian Empire was destroyed by the Arabs in the seventh century, while the Romans narrowly escaped the same fate. But the avalanche fell from the north, not the east. The Roman recovery at the end of the third century had seen Gallienus, Claudius II, Aurelian, Probus, and finally Diocletian campaigning against various northern peoples. The invasions of the empire had been stopped, but at the cost of the surrender of Trajan’s Dacian provinces in what is now Romania. The empire was now bordered by peoples transformed by generations of contact with Rome, contact that included trade and military service as well as war. There were even missionaries operating north of the frontier. The Goths were partly converted to (Arian) Christianity in the middle of the fourth century. But at the end of the century these peoples themselves found themselves under pressure from the north and east.

The new arrivals were the Huns. All reports present them as completely unlike Romans or Goths, a highly mobile nomadic people, moving westward very rapidly, in the same way that Cimmerians and Scythians had in the early Iron Age and Tartars and Mongols would in the Middle Ages and the early modern period. They clearly terrified the settled agricultural communities they conquered. They first appear in the 370s, somewhere north of the Black Sea, and pressed westwards. It is impossible to assess quite how novel or serious was the threat posed by the Huns. The empire remained weaker than it had been before the third-century military crisis. The Goths experienced pull factors—their desire for wealth as well as security within the empire—as well as the push factors generated by the Hunnic invasion. Either way, groups of Goths sought permission to cross into the empire and in 376 the emperor Valens allowed them in. Could he have stopped them? Again it is impossible to know. But when two years later he faced them in battle at Adrianople he lost his life and most of the eastern army. A chain reaction set in, more groups crossed the Rhine and Danube frontiers, and emperors were increasingly reduced to making concessions to some groups of northerners as the only defence against others. By the 450s the Huns were campaigning under their king Attila in the Balkans, in northern France, and even in Italy. On Attila’s death in 453 his fragile empire collapsed under a mixture of internal rivalry and rebellion from its subjects. In this too it resembles other nomad empires throughout history.4 But the political landscape they left behind was changed forever.

By the end of the fifth century, half the Roman Empire was occupied by barbarian kingdoms. The military power and fiscal resources of the eastern empire were crippled beyond repair. Back in the darkest days of the Mithridatic War, the Romans had briefly lost control of all territory east of the Adriatic. Now, 500 years later, all the territory west of the Adriatic had been lost. The city of Rome itself had been sacked twice by barbarian hordes. The empire that Rome had created had ballooned in the late Republic to encompass the whole Mediterranean. Now the balloon had deflated, leaving Rome on the outside. The historian Zosimus—writing around AD 500 in Constantinople, the city founded on the Bosporus by Constantine as New Rome and now the sole capital of what was left of the empire—set out to chronicle Rome’s decline and fall, presenting it as a matching narrative to Polybius’ account of Rome’s rise to empire. For Zosimus, the decline in Rome’s fortunes was the direct consequence of Constantine’s disastrous desertion of the traditional gods of Rome.

A New Empire?

Historians today frequently contrast the empire of the fourth and fifth centuries with that of the first three centuries AD. We write of ‘late’ or ‘later’ empire, or in French haut (high) as opposed to early or bas (low)-empire. Some of these labels indicate an unarticulated assumption that the empire of the first three centuries was primary, and later versions of Roman power secondary. Terming late antiquity ‘post-classical’ is an explicit statement to this effect, as is writing of Rome’s inheritance or of the transformation of the classical world. Historical change is, of course, the only real constant. No age stays the same. But some ages are more conscious of change than others, and modern phraseology reflects some ancient preoccupations.

The later Roman Empire was, in some respects at least, self-consciously past its prime. A group of historians writing in Greek of whom the most important are Eunapius, Olympiodorus, and Zosimus are for this reason often termed ‘classicizing’.5 Taking Greek literature produced in the fifth and fourth centuries BC as stylistic models was not in itself new. The second-century AD orators of what we, following their third-century biographer Philostratus, term the Second Sophistic were given that label for precisely this reason. They strove to speak an ‘Attic’ Greek that was quite different from the spoken language even of the elite who had studied these works at school.6 Latin loanwords were shunned and so were Greek neologisms. Plutarch’s Parallel Lives, composed in the late first century AD, paired great Greeks with great Romans. The Greeks were mostly figures from the fifth and fourth centuries BC, and none post-dated the Roman conquest: the Romans were all figures from the Republic. Two classical eras were set in parallel while the early empire was deemed post-classical for both Greece and Rome. Arrian of Nicomedia governed Cappadocia in the reign of Hadrian, but wrote in the style of Xenophon. Cassius Dio, another Bithynian Greek senator this time of the Severan period, composed aRoman History in which the thought and language of Thucydides is never very far below the surface.7 But the classicism of late antiquity was different. Zosimus’ engagement is more elegiac, as if the classical past is gone forever. Moreover, Zosimus, Eunapius, and the others were self-consciously not writing a Christianized version of history, a version that was in some ways becoming the official narrative. Late Latin literature too was preoccupied with its relation to a Latin canon, one based around the works of Cicero and Virgil.8 That was already true of the Gallic panegyrists in the late third century.

By the second half of the fourth century AD Ausonius and his contemporary Ammianus Marcellinus, the greatest Latin historian of late antiquity, offer perspectives quite similar to their Greek counterparts. Again, it was not new to focus on the classics: early imperial Latin was preoccupied with them and a series of explicitly post-Virgilian epics were composed by Roman senators in the course of the first century AD. But by late antiquity, the age of Augustus and Trajan seemed far away. Ammianus, a pagan member of the military elite, chronicled the noble but futile exercise of traditional virtues and especially the reign of Julian, whom he greatly if not uncritically admired. Much of his narrative seems very familiar—intrigues at court, battles against barbarians in the north, the great rumbling hostility with Persia—but the world he moves through is already partly ruined, its cities sacked, its Senate decayed, its civilization haunted by literary ghosts of a happier time.

That pessimism, largely expressed by representatives of the propertied classes, needs to be tempered against the realities of the restoration of stability by warrior emperors, pagan and Christian alike. Civic elites and the Senate were not generally winners in the reorganization of the empire. Senators had lost their role in government, and had little access to the emperors ruling from Trier or Sirmium or even Milan and Constantinople. The growing bureaucracy around the court alienated many, especially in the west, and taxation was heavier.9 Many of the distinctive features of the empire of the fourth and fifth centuries could be seen as the culmination of long-term trends, either established during the generation-long military emergency of the mid-third century, or even earlier. The development of the imperial court and its transformation into a mobile institution can be traced back to Hadrian and Marcus Aurelius. Once military affairs drew emperors to the frontiers, the role of the Senate inevitably became marginalized. Senators who wished to take a role in public life became generals and courtiers. Imperial edicts replaced senatorial decrees as a source of law and ambassadors now came to the court, not to Rome, for practical reasons. Precedents can be traced back to the reign of Augustus.

Yet there was something new, and it was recognized as such. The empire that emerged from the crisis was ruled by a group, or college, of emperors. Each was based in his own court, and each had responsibility for a share of the army and of the provinces. Each was, as a result, able to act more effectively in relation to his subjects: how they did it is documented in a vast amount of documentation.10 The number of provinces had grown slightly over the third century. Diocletian increased their number and grouped them together into larger units, assigning oversight of the administration to one or other of the praetorian prefects. Eventually there were four great prefectures each managing a quarter of the empire and a wing of the bureaucracy that gathered taxes and organized government directly, with less and less involvement from either senators or local elites.11 That move towards a more centralized form of government built on developments in the tax system that I have already described. It also helped the emperors extract more revenue to supply slightly larger armies. Probably unintentionally, it also made some regions effectively self-sufficient: the taxes needed for a region’s defence were typically raised locally. That development would make fragmentation easier in the future.

The changes introduced by Diocletian and modified by Constantine were unpopular with many members of traditional elites. But the growing bureaucracy provided many opportunities for educated provincials and even for some barbarian leaders to join in a new ruling class. Once recruited they were rapidly socialized and created their own traditional ways of doing things, ways they would staunchly defend against further changes in the sixth century.12 Roman bureaucracy was probably not fantastically efficient. This was not a modern state, patronage remained important and corruption endemic.13 But its internal dynamics had changed.

The multiplication of emperors and so of courts responded to the realization that a single emperor could not be everywhere he was needed, and that in his absence usurpers would appear. Like so much else it evolved from contingent modifications of earlier expedients. Diocletian provided the model when he adopted Maximian, first as Caesar, and then promoted him to Augustus. The use of the title Caesar to mean junior emperor and/or Augustus-in-waiting went back to the reign of Vespasian. It had been used by other emperors, along with adoption, to secure the succession. Diocletian’s idea for extending this stability was that each Augustus would adopt a new Caesar, creating an imperial college of four, the tetrarchy. Each Augustus would eventually retire, to be replaced by his Caesar who would in turn adopt his own helper and successor. Yet sharing imperial rule was not in itself new. Marcus had ruled as co-Augustus with Lucius Verus in the middle of the second century, and Severus briefly ruled with both his sons whom he seems to have expected to succeed him as co-emperors. Diocletian’s only real innovation, the idea of two parallel dynasties renewed by periodic adoptions, was also the only component of the new package that was not successful. After the civil wars of the early fourth century, Constantine ruled for a while as sole emperor. On his death all three of his sons succeeded, to begin their own struggles for power. After this the number of emperors varied according to the chance of civil war and political fortune. Constantius II was sole Augustus from 350 until 361, as were each of his two short-lived successors, Julian and Jovian. But on the latter’s death in 364 the brothers Valens and Valentinian ruled together, and two more members of the family ruled as boy emperors before the dynasty was extinguished in the chaos that followed the battle of Adrianople in 378. Theodosius I ruled as sole emperor, but by his death in 395 both his sons had been made his colleagues. At no point did collegiate rule represent a division of the empire into two, three, or four separate states, and co-emperors were mostly related. Generally the system worked, if the aim was to reduce the incidence of civil wars and usurpations. The western frontiers seemed stronger for being ruled from Trier or Milan, and there was no repeat of the events of the 260s when troops, resources, and attention had been drawn off to the east. Ironically, the fall of the west was triggered by the defeat of the eastern army. Yet one effect of the multiplication of emperors was a potential lack of coherence in policy. That was exposed most dramatically in the years after Adrianople when each court seemed most interested in expelling the Gothic immigrant groups from their own sphere of influence.


Fig. 19. Porphyry statue of the Four Tetrarchs at the Basilica di San Marco, St Mark’s Square, Venice, Italy

The fact that there were earlier precedents for many of the innovations of the fourth century did not mean the package was unchanged. How far Diocletian and Constantine felt themselves to be innovators is another question, but maybe not a very important one. Perhaps the real significance of the military crisis of the third century was that it intensified the process of experimentation: emperor after emperor sought new solutions, liberated by necessity from considerations of tact and custom. Decius and mass persecution, Gallienus’ military reforms, Aurelian and the cult of Sol Invictus, Diocletian and an edict on maximum prices were all initiatives taken in desperation, but they were not crazy ideas. Senatorial accounts of some of the soldier emperors are very vicious, marking the social distance that had emerged between court and Senate. Macrinus was presented as more or less a barbarian. Yet some were very effective.14 A few experiments were disasters: debasing the coinage to fund larger armies when tax revenues were depleted by usurpation was a short-term fix that triggered inflation and weakened the monetary system fatally. Diocletian had to devise a new currency system that was then modified by Constantine in order to repair things. That worked, but his edict on maximum prices was unenforceable. Yet other experiments, such as the development of a more mobile field army, were more successful. The net effect was to create an effective empire for the fourth century, and the basis of a smaller one that would survive for many centuries more.

The final component of the new empire was presentational. Ceremonial became more elaborate.15 Ruler and courtiers dressed in extraordinary costumes, as we can see from the mosaics at Ravenna and Istanbul. Contemporary accounts notice the sudden increase in court rituals. The Emperor Julian wrote a satire entitled the Caesars in which each of his predecessors is caricatured as they arrive for a Saturnalian banquet held on Olympus by Romulus. When Diocletian turns up he makes a grand entrance in fabulous costume, surrounded by a chorus of the other tetrarchs.16 The image of the emperor as fellow citizen, who wore simple togas woven by his wife and daughters, was banished forever.17 Extravagant triumphs were celebrated by (among others) Aurelian, Diocletian, Constantine, Constantius II, and Theodosius I.18 Anniversary celebrations were held for the tenth- and twentieth-year anniversaries of an emperor’s reign. Thirty-year anniversaries, tricennalia, were celebrated in Rome by both Constantine and Theoderic the Ostrogoth. The arrival of an emperor in a city generated a major festival. Gigantic statues were built and vast intimidating palaces. Approaching the imperial presence inspired real awe. Religious ritual was harnessed to these ends, as emperors tried to set themselves back in the centre of the cosmos. Severus had organized Saecular Games in AD 204 but Philip organized another series in 248 to celebrate the thousand-year anniversary of the city. Decius attempted to organize a mass sacrifice, a supplicatio, to be carried out by all citizens of the empire. This may well have been the point at which the number of Christians in the empire first became apparent, as the first general persecution followed in AD 250 directed against those who had failed to sacrifice. Sacrifice certificates, proving participation, have been found from Egypt. The idea of organizing empire-wide rituals was a new one, even if it was in some ways a logical consequence of Caracalla’s expanding the citizen-body to incorporate most inhabitants of the empire. Aurelian found time in 274 after his victory over Palmyra to found a temple of the Undefeated Sun in Rome, with a new college of pontiffs dedicated to its worship. The sungod had been closely associated with the person of the emperor on coinage since the early third century. The conquest of Palmyra provided enough booty, and perhaps some imported statuary, to create the magnificent temple. Diocletian associated each Augustus and his Caesar with a divine patron, Jupiter in his case, Hercules in the case of Maximian, aiming to create Jovian and Herculean dynasties. Perhaps it was with similar ideas of creating religious unity that he also initiated the Great Persecution which earned him such condemnation from the Christian polemicists Lactantius and Eusebius.

Military success generated its own legitimacy. Once individual emperors survived longer than the average, then they seem automatically to have become less challengeable and so were able to achieve more. The tide was already turning in the 250s. Gallienus ruled 253–68, and despite the disasters of his reign, including the capture and execution of his father and co-emperor Valerian in 260, he was remembered for a series of successful campaigns against the Alamanni, and the first major steps towards the creation of a mobile field army. Aurelian, who was implicated in Gallienus’ murder, ruled 270–5 and triumphed over the Gallic emperor Tetricus and Queen Zenobia of Palmyra, before being murdered himself. Diocletian and Constantine, just by surviving longer, restored more coherence to the imperial system and prestige to the position of emperor.

An Evolving Empire

All imperial systems evolve over time. It is no surprise that the Roman Empire of the fourth century AD was not identical to that of the first century. To many of its inhabitants it probably did seem a success story. Civil wars recurred but must often have seemed like rather violent reshuffles within the imperial college, as when the sons of Constantine jockeyed for control or when the most successful of them, Constantius II, faced a coup from the nephew he had raised to the rank of Caesar. Those conflicts were generally short-lived and seem not to have seriously disrupted the management of the empire at other levels.

The empire had evolved through a combination of long-term changes, many unplanned, like the spread of Roman habits of thought and lifestyle, and some, like slow social mobility barely noticed by contemporaries; incremental modification (such as minor redeployments of troops or changes to the fiscal system); and some dramatic responses to crisis. It was now rather different in structure from what had been created by Augustus at the moment when serious expansion stopped and the transition from conquest state to tributary empire was first achieved. The Augustan empire had had concentric structure. Rome was at the centre, then Italy inhabited by a few million Roman citizens and exempt from much taxation, and beyond that the broad zone of taxpaying provinces administered by the propertied elites of cities new and old, enclosed by an emerging frontier zone where citizen armies stared outwards at barbarians and inwards at potential rebels. By the fourth century the difference between Italy and the inner provinces was gone, and virtually all the inhabitants of the empire were citizens. It has aptly been compared to a vast nation-state.

The cultural distance between the provinces had been reduced massively at the level of the elite who shared the essential components of a common lifestyle wherever they lived. Educational systems divided the provinces where the educated spoke Latin from those where Greek was dominant: crudely put the line bisected modern Libya and the Adriatic and then turned north-east to separate the Latin-speaking Danube provinces from the Greek world. Yet at the highest level the division was not absolute and anyone who sought a career in the bureaucracy or a place in the Senate could speak Latin. The common aristocratic lifestyle that had emerged in the second century persisted, more now in rural residences than in cities perhaps, but still attested by amazing mosaics, collections of statuary, and literatures that celebrated the symposium and knowledge of the classics. Spending on civic monuments had declined along with cities.19 Most of this is observable through archaeology and hardly figures in literary texts, yet the differences must have been evident to travellers like the soldier Ammianus who served on the Rhine and the Persian frontier and had visited Rome and Antioch, and to many others.

Below the level of the elite, material culture suggests greater regional differences. Only a part of these were the legacy of the huge diversity of the societies that had been incorporated into the empire by force half a millennium before. At the end of the last and beginning of the first centuries, there had been a phenomenal provincial demand for Roman-style products, from temples to tableware, and bronze statues to wine and olive oil. Italian and then other Mediterranean producers grew briefly rich supplying these goods to the new provinces, especially those outside the Mediterranean world. It is possible to trace the subsequent spread of technologies including brick-making and the use of waterproof concrete, viticulture and arboriculture, stock-raising and fish-pickling, and the production of fine pottery. These were put at once to local uses, and as local production was established so regional styles began to diverge. If the lifestyles of the elite were converging across the empire, the opposite was true for many others, although the situation was complicated by imitation of the local wealthy by the upwardly mobile, and by empire-wide travel by merchants, pilgrims, and soldiers. The pattern of unified high culture and diverse local cultures is one reproduced in early empires around the globe. Members of the elite were in any case more mobile than their subordinates, and changes in the governmental structure of the empire only enhanced this.

Rome in the early empire has been characterized as ‘government without bureaucracy’: the imperial aristocracies of the senatorial and equestrian order provided a tiny number of governors and generals, overseeing a world of cities run by local elites. What held the system together was a shared sense of aristocratic culture. The new bureaucracy of the fourth century was much larger and recruited from those who were educated but generally not aristocratic. It had a complex internal hierarchy of ranks, was divided into departments, and the whole system was focused on the praetorian prefects and the courts in capitals that were located in the frontier zone. The aristocratic propertied classes that ruled cities in the peaceful interior maintained their social eminence but began to come under financial pressure and were now potentially alienated from the government of the empire. This was true even of Rome, which emperors visited rarely and never made their base. Rome remained a cultural centre, and the wealth of the senatorial aristocracy in the fourth century was fabulous, but its relationship with the new centre of power was precarious. During the late fourth century emperors in Milan communicated with the Senate via the prefect of the city, and the Senate sent ambassadors and petitions to the court just as great provincial cities had done in the early empire.

Some precious documents present us with an image of imperial style in this period. One is the Notitia Dignitatum of which a gloriously illustrated copy made in the sixteenth century survives. The original was composed at the end of the fourth century and parts revised in a complicated manner that is still obscure, through the early fifth century. Each page is devoted to a different office in the hierarchy—praetorian prefects, regional vicars, provincial governors of all kinds—and then the parallel military hierarchy including armies, forts, and weapons factories. The information contained is exactly what a centralized empire would need, but the fact it was illustrated so expensively and yet would be out of date almost at once suggests it also offered a kind of panorama of power. Each entry is also preoccupied with those details that matter most to insiders: precise titles, the orders of precedence and seniority, the number of postal warrants assigned to each official. And the fact it includes western and eastern officials and units in a period in which the empire was effectively divided into two shows it depicts an order of dignities that was idealized and ideological as much as practical.

A second key document is the Theodosian Code, a compendious record of all imperial edicts issued since the start of the reign of Constantine, compiled between 435 and 438 and then distributed in both halves of the empire on the orders of the second emperor of this name.20 The code illustrates both the strengths and limitations of imperial government. The capacity to plan and execute such a project in such a short time shows how the imperial bureaucracy could be put to work and produce results. Yet the fact that to compile this collection it was necessary to write to provincial governors all over the empire asking for copies of any edicts they had filed away shows an astonishing lack of record keeping. The desire to make a complete collection, to order it logically by sixteen themed books, and to remove contradictions and inconsistencies perfectly expresses both the aspiration to rational, universal rule and the gap between that ideal and reality. That it was carried out as late as the 430s shows that despite the disasters of the generation following Adrianople, the empire and imperial society was definitely one thing and not two in the minds of the emperor and his staff. Starting the record with the reign of Constantine is also a rare testimony to an ancient concept approaching our modern one of ‘a later Roman empire’. Finally, Theodosius’ great project was identified immediately with imperial monarchy. The clearest sign is not the elaborate ceremony with which the emperor presented a copy to an ambassador of the western Senate attending him in Constantinople; nor the adulatory acclamations by senators when they received the work, a record of which forms a preface to the work; but the fact that in the centuries to come whenever barbarian warlords across the former western empire tried to convert themselves into hereditary monarchs with courts and ceremonial of their own, one of the first things they did was to issue their own codes of law.


Fig. 20. An image from the late antique Notitia Dignitatum

Civilization without Empire? The Collapse of the West

Following the death of Valens at the battle of Adrianople in 378, the dependence of the emperors on their former enemies was greatly increased. For a generation, bribery, diplomacy, and threats of force were employed by both western and eastern emperors in a series of attempts to contain and deflect the Goths. But the weakness of the empire was now obvious. The Goths eventually headed into Italy, and Rome was sacked in 410.21 The emperors retrenched to protect the interior provinces. Even before the Gothic sack, Britain had been abandoned.

Most of the other groups now entering the empire originated in that broad band of peoples who knew Rome well from long acquaintance.22 Some crossed the Rhine —Vandals, Sueves, and others—heading through Gaul to Spain. The Goths moved on to Aquitaine where they were settled as ‘guests’ in 418. From there they expanded their power into Spain, driving other groups ahead of them. The Vandals crossed into Africa in 429 and ten years later captured Carthage, the second greatest city of the Roman west. Meanwhile Franks, Burgundians, and Huns joined a confusing struggle for northern Gaul. As it happens a mass of literature written in late antique Gaul has survived, and through it we can trace the stages by which faith in the emperors was lost, the provinces slipped from Roman control, and local accommodations were made between the landowning classes and their new rulers. Following Aurelian’s suppression of the separatist Gallic emperors who had ruled the region from AD 260 to 275, the tetrarchs had paid more attention to this part of the empire. Trier, on the Moselle, became an imperial capital, and was endowed by Constantine with a great palace and basilica, imperial baths (which were never finished), and other monuments, many of them surviving to this day.23 A series of panegyrical speeches from this period illustrate the efforts of local aristocrats to draw imperial favour to their cities. From the court at Trier we have the poems of Ausonius, a teacher of grammar and rhetoric from Bordeaux who served both as a governor and as tutor to the boy emperor Gratian in the 370s before rising to the consulship. His poetry describes his relatives and colleagues in Bordeaux, the landscape of the Moselle Valley, but most of all the urbane life of the educated in the last generation of the western empire. During the early fifth century that world changed by stages.24 It became less and less easy to tell barbarian warlords adopting Roman titles from Roman generals behaving like local dynasts as they struggled to protect their own regions.25Everywhere communities sought local protectors. Many aristocrats entered the Church. Some embraced ascetic disciplines, while others continued to exercise social authority in their cities as bishops. The letters of Sidonius Apollinaris offer a finely nuanced picture of the move from Ausonius’ world of educated aristocrats playing sophisticated literary games to a world of churchmen interceding for their people with warrior kings.26

By the middle of the fifth century, the Roman Empire in the west was limited to Italy and parts of southern France. A Vandal fleet from Africa sacked Rome again in 455.27 The last western emperor was deposed by his barbarian ‘guests’ in 476, and his place taken by a Gothic king, one of the many barbarian leaders on whom western emperors had come to depend. Eastern emperors were powerless to intervene, and were compelled to use diplomacy in the west to free up resources for defence in the north and east. No single moment of crisis was recognized as the end of what we call the western empire. But it was obvious enough to Zosimus.

West of the Adriatic and north of the Balkans a new world of barbarian kingdoms had replaced the Roman provinces. Their societies were quite unlike those their Iron Age ancestors had lived in in central Europe.28 During their time on the frontiers, new social structures had emerged. Most ‘barbarian’ rulers were Christian, and their idea of kingship was in many ways modelled on their image of the Roman emperor. Goths, Vandals, and Burgundians relied at first on tax systems descended from those established by Diocletian.29 They ruled from Roman cities where they repaired monuments and they created courts at which they patronized Roman scholars and churchmen. Throughout the fifth and sixth centuries their administrations largely depended on an elite who in education and cultural outlook were as Roman as their ancestors had been. Some of these scholars put their scholarship to work reconstructing the ancient traditions of their new masters, combining tribal traditions with Greek mythology to do so.30 The warbands gradually mutated into armies, the chieftains into landholders. Successive kings issued law codes, just as the Emperor Theodosius had, if on a rather smaller scale.31 Some of those law codes enshrined the principle of multi-ethnic states, each people using their own laws. Like Roman emperors the kings squabbled with bishops and they were drawn into disputes over heresy. Roman civilization continued in some ways very much as before, until the arrival of Franks and Lombards from the north in the sixth century and Arabs in the seventh. But the empire was gone.

Further Reading

Tim Barnes’s New Empire of Diocletian and Constantine (Cambridge, Mass., 1982) is the basis for understanding the transformation of Roman government at the end of the third century, and is much more than a companion piece to his Constantine and Eusebius(Cambridge, Mass., 1981). The institutions of the empire are described in detail in A. H. M. Jones’s Later Roman Empire (Oxford, 1964): how they worked in practice is the subject of Christopher Kelly’s Ruling the Later Roman Empire (Cambridge, Mass., 2004). John Matthews’s Roman Empire of Ammianus (London, 1989) offers a vivid picture of the empire before the disasters of Adrianople, one that encompasses politics and society. Fergus Millar’s A Greek Roman Empire (Berkeley, 2006) offers a new view of the early fifth century. A particularly useful set of essays is included in Simon Swain and Mark Edwards’s Approaching Late Antiquity (Oxford, 2004).

Alongside these studies focused on politics and institutions, late antiquity has emerged as a vast field of cultural history. Peter Brown’s World of Late Antiquity (London, 1971) was in some ways the manifesto for this approach. His own voluminous writings, and those of his students and associates, have explored with subtlety the rich material offered by Christian writings. His Augustine of Hippo (rev. edn. London, 2000) shows just how much can be mined from this seam. Perhaps the best conspectus of late antiquity is offered by Late Antiquity: A Guide to the Post-Classical World, edited by Brown, Glen Bowersock, and Oleg Grabar (Cambridge, Mass., 1999).

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