Ancient History & Civilisation


15 BC–AD 9

Main period of Augustan wars of conquest in Europe, ending with the loss of three legions in the Teutoberger Forest

AD 43

Invasion of Britain begins under Claudius

AD 66–70

The Jewish War

AD 69

Year of the Four Emperors. First civil war since Actium

AD 82–3

Domitian campaigns on the Rhine

AD 85–9, 101–2, 105–6

Domitianic and Trajanic wars against the Dacians

AD 106

Arabia annexed

AD 114–17

Trajan’s Parthian war. On his death Hadrian withdrew from the new province of Mesopotamia

AD 166–80

Marcus Aurelius’ Marcomannic Wars on the Danube frontier

AD 193–7

Civil wars leading to establishment of Severan dynasty

AD 226

Sassanians overthrow the Parthians in Persia

AD 241–72

Reign of Shapur I as Emperor of Persia

AD 249–52

Reign of Decius

AD 250s

Increasing raids across Rhine by Alamanni and other groups

AD 260

Capture and execution of Valerian by the Persians

AD 260–8

Reign of Gallienus. Gallic emperors and rulers of Palmyra allowed to become in effect autonomous. Among other disasters the Franks sack Tarraco (264), the Herculi sack Athens (267), and the Goths sack the sanctuary of Artemis at Ephesus (268)

AD 268

Claudius II defeats Goths at Naissos

AD 270–5

Reign of Aurelian. Campaigns against Vandals, Alamanni, and Iuthungi, builds a new wall around the centre of Rome, suppresses Palmyrene revolt, defeated Gallic separatist emperor Tetricus and triumphed over both in 274. Organized evacuation of Dacian provinces

AD 284

Accession of Diocletian



From Caesar Augustus’ day to our own time there have been nearly two hundred years. Over that time the Roman people might seem to have grown old and impotent, owing to the laziness of the Caesars, except that under the rule of Trajan it has stirred its limbs and, against everyone’s expectation of the old age of empire, it has revived, almost as if it were young once more.

(Florus, Epitome of Roman History 1 Preface 8)

The Laziness of the Caesars

The rise of the emperors coincided closely with the effective end of Roman expansion. Contemporaries noticed, and they complained. Tacitus, writing in the early second century, lamented the length of time the conquest of Germany was taking.1 Florus, writing around the same time, divided Roman history into four ages, each corresponding to one stage of a man’s life. Rome’s childhood had been under the kings, and its adolescence was the Republican period up until the outbreak of the first Punic war. From then until the reign of Augustus, Rome pacified the entire world. ‘This was at once the youth of empire and the robust maturity of Rome.’ But in the succeeding period Rome had grown old. The biological analogy no longer seems such a good way to describe imperial history, but the reign of Augustus still seems to mark a rupture.

That break is naturally a simplification. The emperors continued to celebrate their foreign victories on an unparalleled scale. They continued to fight wars after the death of Augustus, and even conquered a little more territory. But—leaving to one side a general trend to extend the system of provinces over regions previously governed by allied kings—there were only a few areas of genuine expansion. Britain was invaded in Claudius’ reign and conquered (very slowly, and cautiously) over the rest of the first century AD. In south-west Germany, the frontier was advanced under Domitian from the Rhine across the Black Forest to the Neckar Valley. Trajan conquered part of what is now Romania in the early second century. He also invaded Mesopotamia, modern-day Syria, and Iraq. But the scale of these expansions seems feeble compared with the audacious conquests of the 60s and 50s BC or even to those of the heyday of Augustan expansion between 15 BC and AD 9 when his sons fought their great wars across Europe. Besides, all the new territories except Britain were rapidly lost again. Mesopotamia was handed back to the Parthians by Trajan’s successor Hadrian. The conquests of Domitian and Trajan in Europe were lost in the military crisis of the third century AD. Future emperors campaigned beyond the Danube and the Euphrates, now and again, but the imperial frontiers hardly moved.


Fig 16. A detail of Trajan’s Column showing triumph of the emperor after the first campaign against the Dacians

Many of the core institutions of the Augustan empire built on the experiments of Caesar, Pompey, and even earlier generations. Alongside developments in taxation we can set the practice of extending the citizenship; a growing dependence on local elites to run their own communities; and the practice of using city-states as a basis for government, even in areas where they had never existed before. One way to describe the transformation of the empire at the turn of the millennium is to say that over a long period the Roman state experienced a tension between two divergent tendencies. The first we could label the pursuit for glory, the second the desire for security. A move in one direction made other moves that cohered with it more feasible, and so more likely. The disasters of the middle second and early first century BC—the Cimbric invasion, the Social and Mithridatic Wars, and the rise of the populares—had pushed the state a long way in the direction of security. Hegemony had been converted to empire, and new institutions developed to rule it. By contrast the accelerating competition between leaders from Sulla to Augustus pushed the empire towards the more risky pursuit of glory. After Actium security always trumped glory.

Perhaps this choice was never made consciously. Roman ideology certainly did not always reflect the new logic of empire. The idea that world conquest was a realistic and laudable aim was repeatedly restated, most influentially in the classroom, where it was encoded in the classics of Latin literature as defined in the last century BC. Imperial monuments and imperial ceremonial also reproduced the incidental music of a conquest state, long after it no longer suited the plot of Roman history.2 But detailed investigation of the military system of the empire—investigations made possible by a wealth of epigraphic data—reveals a set of mutually supporting institutions that were well geared to preserving peace, and that consequently made further conquests difficult and costly. It has even been argued that the emperors made so much noise about conquest to compensate for their reduced willingness to attempt it.3 Trajan was in some senses a throwback, a proof that an individual emperor need not be trapped by the role. Yet his reign, with its futile and short-lived conquests, is also a demonstration that expansion no longer suited the Roman Empire very well.

Augustus had offered an earlier object lesson when he decided on the conquest of Europe. Some historians have tried to excuse this on the grounds of geographical ignorance: maybe he did think the world was smaller than it was, or had no decent maps? None of this is believable. The diameter of the globe had been estimated two centuries earlier. Geographical works of the day included descriptions of India and mentioned the distant silk people, the Chinese. Africa had been circumnavigated half a millennium before Augustus. Much more likely, Augustus’ European front was a tactical error driven by short-term political difficulties. A number of Augustus’ challengers in the 20s BC had been commanders of major provinces. Did he dare let someone else achieve victories in the Balkans or south Egypt? Besides, it was still unclear what Augustus’ own role in the state would be once he had restored civil peace. Triumphs for his heirs might help their succession plans too. External warfare was evidently still popular in Rome, although it is noticeable that from this point on fewer and fewer Italians fought in the legions. So Augustus fell back on the pursuit of glory when the logic of his situation demanded more efforts to create security. The poetry written in his court advertised future campaigns, promising conquests in Britain and Parthia, India and Scythia, while monumental art offered images of a world already conquered.4

But where to start? Persia was a dangerous enemy which had defeated Crassus and humiliated Antony, and did not seem to want further conflict. The tribes of northern Europe must have seemed an easy option: Caesar had conquered all Gaul in only eight years with as many legions. Could Britain and Germany offer anything other than easy victories? At first the victories did seem easy, and the armies of Drusus and Tiberius rolled around the Alps, campaigned up to the Danube, and across the Rhine as far as the Elbe. But rapid progress was deceptive. Conquered Pannonia rebelled in AD 6, and it took three years to restore order.

Almost immediately a Roman army of three legions was massacred in Germany in AD 9. The Varian disaster—named after the general made to carry the can—was a trauma that echoed through the history and literature of the last years of Augustus’ reign. The astrologer Manilius used it as an example of the terrible catastrophes foretold by comets. The actual defeat was a running battle lasting several days: the site was recently located at Kalkriese near Osnabrück and has been painstakingly reconstructed. It cost the Roman army nearly 10 per cent of its manpower. All territory east of the Rhine was abandoned. A half-built Roman city has recently been found at Waldgirmes, offering eerie witness to the sudden change of direction. A great new province had been planned, and construction had begun on a network of civil communities just like that created in Gaul after Caesar’s conquest and in Pontus by Pompey. Roman accounts, reflecting the official line, blame Varus for behaving as if he was in a conquered province, spending his time giving justice, imposing taxation, and dispersing his troops among native communities. But he was certainly following orders from the emperor. After disaster struck Rome pulled back to the Rhine, Waldgirmes was abandoned, and with it all the territory up to the Elbe. Conquest was never formally renounced, but it was not resumed in Augustus’ reign. Another imperial prince, Germanicus, visited the site of the disaster during the next reign but his campaigns were also—if less dramatically—unsuccessful. The north began to be quietly written off as ungovernable, poor, undesirable. Strabo, writing under the Emperor Tiberius, reported that Britain too would not pay the cost of its occupation.5 Appian, a second-century historian, went so far as to claim alien nations had begged to be admitted to the empire, but the emperors had turned them down.6 These were all lies.

For the emperors had much to gain from security, and there were easier routes to glory. Did Augustus ever realize that the real motor for the wars of Pompey and Caesar had been competition? And now there were no rivals. Emperors did not plan for peace, but they learned from their mistakes. The risks of failure never made up for the potential rewards offered by military expeditions. Much more could be achieved by diplomacy. And emperors soon found out how to massage the news from the frontier, playing up minor successes, suppressing news of reverses, claiming any victories for themselves, and blaming generals on the ground like Varus when things went badly wrong. The laziness of the Caesars was a very pragmatic response to the absence of rivals.

A World without History?

The choice of security over glory makes for unexciting history. Or so the senators of the imperial age affected to think. Court intrigues and imperial assassinations actually make for rather good drama. The histories written by the senators Tacitus and Dio, and the scandalous biographies of the courtier Suetonius and his late imperial successors, have inspired their share of racy novels and movies. But it is true that it is difficult to find a clear narrative in the political history of the first two centuries AD. Institutionally, culturally, and economically, there were slow changes.7 The citizen body expanded and with it Roman law stretched over more and more of the empire’s subjects: that process of assimilation continued long after the Edict of Caracalla.8 The greatest cities grew and acquired their complements of monuments. The rich grew richer, building great rural residences and endowing festivals, temples, theatres, and bathhouses in their native cities.9 Trade flourished across the urban network, and between the great ecological divides into which the empire was divided. All these changes were real, but few were visible to contemporaries, and they were not the subject of history as the ancients understood it.

There were gradual changes in the style of imperial rule too. The monarchical nature of the emperors’ rule became more overt. As the emperors ruled from their itinerant courts they seem to have gradually developed a preference for direct over indirect control, and a greater reliance on state officials rather than aristocratic former magistrates. Augustus had travelled widely, but few of his first-century AD successors spent long away from Rome. The expeditions of Trajan and the restless travels of Hadrian were in some senses optional, and Antoninus Pius who reigned from 138 to 161 spent most of his time in Rome. The situation began to change with Marcus Aurelius, who succeeded him until his death in 180. Marcus and his co-emperor Lucius Verus (161–9), and then his son and successor Commodus, who ruled until his assassination in 192, were all compelled to spend long periods of their reigns on the frontiers. So did all the Severan emperors who ruled Rome between AD 193 and 235. These necessary displacements were accompanied by the emergence of a new and more openly monarchical style. Away from senatorial sensibilities emperors could rule like the kings they had always been.

By the early fourth century there were up to four different courts at any one time, strung out along the northern and eastern frontiers of the empire. Wherever the courts rested for a few years—Trier or Antioch, Sirmium or Ravenna—magnificent palaces were created with great bathhouses, hippodromes, and reception areas. The empire came to be divided into four great prefectures, each headed by a praetorian prefect whose administration extended down into a growing number of provinces. Senators no longer played much part in the administration of the empire, and the Senate itself (or senates after Constantine created a second one in his new capital on the Bosporus) became peripheral to the political system. Ambassadors sought out emperors in the camps; law making had to be by edict rather than senatorial decree; consultation—even for form’s sake—was no longer practical. Emperors were away from Rome for years, and then decades. Constantius II, who with his brothers succeeded Constantine in 337, did not visit Rome untilAD357. Before that new order came about, however, the empire had to face a much more serious military emergency, one that in some senses lasted for two entire generations.

The Early Imperial Security System

Historians sometimes write of Rome as having a Grand Strategy,10 but there is no real sign that any such thing was ever planned or implemented. The deployment of the legions was accidental, the evolving product of incremental changes made in response to immediate needs. When expansion ended, Augustus’ campaigning armies had stopped in their tracks. Troops would be moved for imperial expeditions like Claudius’ invasion of Britain, or if they were needed to crush a rebellion or participate in a civil war. Over time the legions gravitated to the points of stress. Spain was far from the frontiers and relatively peaceful, and as a result the size of its garrison was progressively reduced. On the Danube and in the east troop numbers increased slightly, but always limited by the salary bill. The emperors knew where the troops were, how many they were, and what they were owed. But there is no sign they made use of this information to plan for anything but the short term.11

It is not always clear exactly where the armies were deployed in the early years of Augustus’ reign, except that they were kept out of Italy, where only the Praetorian Guard—made loyal through extra pay—were allowed in the capital. But a decade into Tiberius’ reign we happen to have a snapshot, from a passage in Tacitus’ Annales, of the locations of what were now twenty-five legions.12 They were concentrated overwhelmingly on the northern and eastern frontiers: facing across the Rhine and Danube, that is, and stationed along the long frontier with Parthia and its vassals. Only token forces remained in Spain, Africa, and Egypt. Many other provinces were effectively unarmed. Legions of heavy armed infantry formed the core of the early imperial army. They resembled the Republican armies in terms of their equipment and battlefield tactics. Auxiliary units of cavalry, light infantry, missile troops supported them, alongside teams of engineers and other specialists. There were naval bases in the Mediterranean, and in time fleets were established on the major rivers. That pattern stayed more or less unchanged until the middle of the third century, although the number of legions increased to thirty-three and some were moved to newly conquered regions like southern Britain and Dacia. The total strength was always very small. At its height the early imperial army numbered less than 200,000 legionaries and perhaps as many support troops, to protect and control an empire of between 50 and 100 million inhabitants.

Keeping the army small was a financial necessity given the huge share of the imperial budget it consumed. This meant it had to be highly efficient. The campaigning armies of the late Republic and the reign of Augustus had marched and camped in strength, but a different disposition was needed for their new roles. The logic of the imperial frontiers was all based on establishing a communications advantage over its opponents. Little by little the frontiers became a dense network of bases—some very small—signal stations, barriers, control points, and, most important of all, roads. Hadrian’s Wall provides an excellent example of the kind of system that developed in the second century, but the ditches and ramparts that so impress us today were among the latest and least vital components of a frontier system. A precious collection of letters from Vindolanda reveals the collection and processing of information on what was happening beyond the frontier; the management of provisioning; the constant redeployments of soldiers along the frontier; and communications back into the province.13 Signal stations ran down the coasts watching for raids by sea. Groups of scouts operated far north of the wall, and local leaders were cultivated with gifts that Romans called subsidies. A slightly different system of signal stations and forts passed news on barbarian movements up and down the upper German Limes, and in the pre-desert of North Africa military installations were different again. Physically the frontiers developed in a piecemeal way, adapting to local circumstances not a central model, but the guiding logic was the same.

The vast majority of military units were based on the frontiers. Yet soldiers were ubiquitous in the empire. Detachments provided protection for provincial governors and procurators, for messengers and tax collectors, for movements of grain and cash, for the managers of imperial mines and quarries with their workforces of slaves and criminals, and for the authorities of the larger, more unsettled cities. Centurions in particular acquired a whole range of functions we do not normally associate with the military. They acted as district officers in the northern provinces, can be found on detachment organizing provisioning and as the most recognizable representatives of Roman government depicted in the New Testament. Particularly trusted soldiers served as beneficiarii consulares, effectively aides de camp to governors. The early empire had no civil service, and there were some jobs that it was impolitic to entrust to imperial slaves and freedmen.


Fig 17. Hadrian’s Wall

Imperial propaganda proclaimed that the soldiers protected the provinces. That picture is only half true. The legions were also the emperors’ ultimate weapon against both provincial rebellions and aristocratic usurpers.14 The mass demobilization of Roman soldiers at the end of the Triumviral period had been driven by financial and political priorities. Armies were dangerous, unpaid ones doubly so. Now they would be composed of career soldiers who served for twenty or more years, loyal to the emperors, not their commanders.15 On retirement the emperors provided each man with a substantial bonus, so long as he had proved himself loyal. Commanders, drawn from the senatorial and equestrian orders, came and went. It was the centurions, risen from the ranks, who provided continuity of command and military expertise. Ceremonial was deployed to bind the soldiers to the imperial family. Army units celebrated imperial birthdays, and worshipped the emperors along with the signa, their standards. A third-century calendar from the eastern frontier post of Dura lists endless holidays marking the anniversaries of imperial family members, some long dead. Emperors and their sons took care to visit the legions, and even commanded them when it was safe to do so.

Mostly the system worked. Civil wars only broke out in 68 and 196 when entire dynasties had been extinguished. A few rebellious generals discovered to their cost the depth of loyalty the legions felt to the imperial family of the day.16 That loyalty meant that the legions could also be deployed against provincial rebels. During AD 69 the legions of the Rhineland were deployed first against Vindex, a senator from southern Gaul who had rebelled against Nero, and then against the Batavians of the lower Rhine who had tried to use the Roman civil war that followed his death as an opportunity to secede. The legions were in principle recruited from Roman citizens, but Italians rarely joined up after Augustus’ reign. The main sources of recruits were first the Roman cities of the inner provinces and later the camps themselves. One sign of the success of the emperors’ investment in maintaining the loyalty of the troops (and perhaps too of the socialization effects of long service) is that legions almost never made common cause with neighbouring provincial populations. Troops recruited in Gaul and Germany were content to be deployed against British, Gallic, and German rebels, and the army of Numidia to march against African usurpers.17

Mostly, however, the soldiers inhabited a world of their own. The larger camps of northern Britain, the Rhineland, the Danube provinces, and Africa eventually came to resemble cities, equipped with monumental walls and gates, stone-built amphitheatres, bathhouses, and shrines. Formally soldiers could not marry and camps were organized in barrack blocks arranged in precise parallel lines, but the artefacts and clothes found in them shows there were women and children in these communities too, as well as in the informal villages called canabae that grew up alongside them. In Syria and Egypt soldiers mostly lived in cities in any case. Documents from Dura show them marrying and buying land and in general assuming the sort of roles in local communities that their relatively good pay and excellent connections could secure them.18

The Roman Empire had no Grand Strategy, but it nevertheless developed a frontier system quite like those of many other tributary empires. The most common point of comparison is China’s Inner Asian Frontier, the limit between the areas controlled directly by Chinese officials and a great periphery within which subject nations shaded out into barbarian allies and enemies.19 China, like Rome, enjoyed some technological advantages over its neighbours, although in some periods there was something of an arms race as China sought to prevent technology transfers to the barbarians. China, like Rome, was able to provision and supply its troops from an intensely farmed and taxed hinterland. The Chinese too deployed a mixture of linear barriers and garrisons of regular soldiers with irregular allies, and they too sought an information advantage over their opponents. The information system extended deep into the provinces, to the imperial court or courts. It was not Grand Strategy that preserved either empire, but the tactical advantages given by information superiority.


Map 5. The third-century crisis

Crisis on the Frontiers

The Roman system had, naturally, its weaknesses too. One consequence of depending on an infantry army based at the edge of empire was that it was slow to respond to disasters in the interior. Roman troops were generally successful against provincial rebellions during the first two centuries AD because most rebellions occurred relatively near the frontiers, and the rebels were usually stationary and had no fortifications. Order was generally restored within months. The Jewish war lasted as long as it did because the Jews possessed fortresses. Another problem was that more mobile enemies, once they broke through the frontier, could outrun Roman armies. When the frontiers did collapse, as they did in the third century, and raiding parties penetrated as far as Athens and Ephesus and Tarragona, they found rich cities with no defenders and often no functioning defensive walls. Civic populations learned the lesson. During the late third and fourth centuries, city after city built defensive circuits, sometimes dismantling earlier monuments to create safe zones in the middle of once extensive cities. One third-century emperor, Aurelian, actually built walls around the centre of the city of Rome. One of his predecessors, Gallienus, had also begun developing a more mobile army, one based on cavalry, which could act as a rapid response force. The use of troops of this kind, alongside the legions, became more and more evident in the changed conditions of war on the northern frontier. Constantine too was credited with an enlargement of their role. Meanwhile, at the eastern end of empire Romans and Persians were both developing armies based on heavy cavalry. Where urban fortifications were also improving as fast as techniques of siege warfare, the military landscape came to look more and more medieval, a world of knights and castles amidst a landscape of peasant villagers.

The gradual social and economic transformations of the first three centuries AD were not confined within the political limits of the empire. The political economy of the empire may be thought of as a vast redistributive system that drew resources from all over the interior and spent them at the frontiers, mostly as army pay. The court was the other main recipient and this too was increasingly located at the edge rather than the geographical centre of the empire. Perhaps no provincial societies were transformed as utterly as those on the frontiers. The effects can be traced in the spread of new cults, of epigraphy and technology, and in the apparent prosperity of areas that had once been marginal. Nor were these effects confined to Roman subjects. The eastern frontier bisected a chain of caravan cities with ancient shared traditions of language, cult, and commerce. Greek, Aramaic, and its sister languages were spoken in a great arch that stretched from the Mediterranean to the Persian Gulf. By the late third century there were populations of Jews, Christians, and Manichaeans on both sides of the Romano-Persian frontier.

Rome’s northern frontier had bisected other peoples with shared prehistoric cultures. The existence of a frontier zone promoted connections. Population densities were low in this region, relative to the Mediterranean world, and there were no cities beyond the Roman provinces. But the agricultural potential was high. Commerce, including slaving, crossed the frontier, and there is archaeological evidence for Roman manufactures in a broad band 50–100 kilometres (around 30–60 miles) wide stretching back from the frontier.20 There were technology transfers too. The political units and ethnic groups of this zone seem to have been quite unstable. Roman writers sometimes blamed this on differences of temperament and characterized barbarism as a deficiency in the stability of settled, urban societies. Perhaps the larger hegemonies were intrinsically temporary. Yet these populations were not nomadic, and it is possible their political fragmentation was actively managed by Rome, which (like China) gave subsidies to their friends, took hostages, sheltered exiled princes, and generally tried to extend their control well beyond the limit of the provinces. Periodic raids and expeditions by both sides were simply one part of a complex relationship. From early in the first century AD there is also evidence for the recruitment of ‘barbarians’ to serve in the Roman army, and some rose to high ranks. Romans often found themselves facing enemy armies commanded by former Roman soldiers, like Arminius who had led the massacre of Varus’ legions in AD9, in fact, andmany spoke Latin. By the second century, many societies bordering the Roman frontier were locked into a series of interdependent relationships with Roman power. Certainly the leaders of the groups we hear of in the third century, such as the Alamanni on the upper Rhine, and the Goths on the lower Danube, knew the Roman Empire well.21

That world began to change in the late second century AD for reasons that are fiercely disputed. One school of thought sees the transformation of barbarian societies as the main cause of the collapse of the frontiers. The long relationship with Rome created better organized and equipped enemies who knew very well the riches the empire had to offer. Eventually Rome lost the arms race and the frontiers folded. Others see the change originating in Rome’s increasing military commitment on the eastern front. As troops were withdrawn from the west to serve first against Persia, and then in successive civil wars, the delicate balance on the western frontiers collapsed: Alamanni, Franks, and others walked into provinces that were effectively undefended. Yet others see the origin of the crisis in obscure movements on the distant Steppe, where truly nomadic peoples, especially those that Romans later came to know as the Huns, pressed hard on the settled barbarians of temperate Europe pushing some populations, like the Goths, south and west onto the Roman frontiers. Large population movements certainly occurred within Europe in some periods, and had intruded into Mediterranean world on several occasions, including the Gallic sacks of Rome and Delphi in the fourth and third centuries BCand the Cimbric wars and the Helvetian migration at the end of the Republic. Various combinations of all these factors might, naturally, be imagined. The problem is simply that we know very little of movements so far beyond the Roman frontier.

One traditional narrative of the crisis begins in the late second century AD with Marcus Aurelius’ wars against the Marcomanni and Sarmatians, wars that kept him occupied for years on the northern frontier. The new provinces he allegedly contemplated creating would have been to the west of the three Dacian provinces founded by Trajan earlier in the century. But unlike Trajan’s wars, these did not result from an imperial initiative. The Marcomannic Wars began with a German invasion of Italy in 166 and continued, with only short periods of remission, until 175. A new war drew Marcus back in 177 and he was still campaigning on his death in 180. Commodus abandoned the war rather than finishing it. The frontier evidently held, even while Roman armies were distracted by civil wars in the 190s. Renewed activity on the northern frontier began on the Danube in the 230s with raids on Black Sea vassals of Rome and then on the Roman province of Moesia. Gothic warbands raided Dacia and the Danube provinces in the 240s. Decius, who ruled between 249 and 252, briefly repelled them, but was killed in a counterattack. Goths continued to raid through the 250s. Who were these groups, not yet united into a single force or nation? One possibility is that the Goths originated among populations who had lived on the borders of Trajan’s new Dacian provinces, peoples who had recently undergone processes of social transformation of the kind experienced by other German-speaking groups on the Rhine much earlier.22 Whether wars on the lower Danube led to Roman neglect of points further west, or whether now-invisible pressures moved east to west over the century, the security crisis spread. On the middle Danube, the Sarmatians raided Noricum, Rhaetia, and Pannonia in the later 250s. At the same time there were raids by the Alamanni across the Rhine into Gaul and down into Spain. During the 260s Tarraco was sacked by the Franks, Athens by the Heruli, and Ephesus by the Goths. The Roman recovery began finally in 268 when Claudius II defeated the Goths at Naissos: thereafter it was surprisingly fast. Aurelian expelled the Iuthungi from Italy in the early 270s and Probus repelled the last major invasion of Gaul across the Rhine in 276.

War on Two Fronts

The task of restoring normal relations with the northern peoples was conducted in deadly counterpart with a deterioration of relations on the eastern front. It is common to blame this on the appearance of a new Persian dynasty, the Sassanians, in 226 AD and the aggression of the Emperor Shapur (241–72), who fought several wars against Rome, defeated the Emperor Philip in 240, seized the city of Antioch in 256, and captured and executed the Emperor Valerian in 260. But the Romans bore some responsibility for all this. Again the story can be traced back to the 160s. After Trajan’s conquest and Hadrian’s withdrawal from his new province of Mesopotamia there had been peace with the Parthians until the joint reign of Marcus and Lucius, when Roman armies once again invaded Persia, without much provocation.23 Severus did the same a couple of decades later. Roman aggression did a good deal to destabilize the Parthian dynasty, creating an opportunity for the Sassanian takeover. It is difficult to tell now whether Persia exploited Rome’s difficulties in the north or vice versa or whether the security system of the early Roman Empire was simply incapable of dealing with threats on so many fronts.

What is clear is that the inability of the emperors to defend the great cities and unarmed provinces of the interior of the empire led to a crisis in their legitimacy. One index of failure was that in the period 235–84 more than twenty emperors reigned. The exact number depends on how many rebels are considered as short-lived rulers. A second index of failure was geographical fragmentation. Local rulers, client kings, and army commanders took over responsibility for protecting their immediate localities. When Aurelian came to power in 270, most of Gaul, Spain, and Germany had been ruled from the Rhineland for over a decade, and the monarchs of the caravan city of Palmyra in Syria controlled much of the Near East, even including Alexandria. Usurpations had been attempted in Africa, on the Danube, in Egypt, and in Asia Minor. Successful and unsuccessful usurpers alike were drawn from the military classes, their links to their armies personal and contingent on their continued success. Civil war and failure at the frontiers fed off each other. Only military success could restore legitimacy and reverse the fragmentation of authority. Valerian’s son Gallienus (253–68) achieved some external successes. Aurelian (270–5), who had expelled the Iuthungi from Italy, went on regain control of Egypt (272), to suppress the secessions led by Palmyra (273) and the emperors of Trier (274). His successors inflicted more defeats on the Germans. Carus finally carried the Persian war into Mesopotamia and captured the Persian capital Ctesiphon. He died on campaign in 283 and his successor Numerian withdrew, but within a year he had been replaced by Diocletian, who ruled until his abdication in 305. During his long reign he too fought on the Danube and against Persia, and had to assert his power in Egypt and against rebels in the west. He left behind a completely reorganized Roman Empire. The period from the end of the Severan dynasty in 235 to the accession of Diocletian in 284 is sometimes known as the Anarchy. Any ‘crisis’ that lasts for half a century would inflict a huge cost on institutions. Diocletian’s empire did indeed need a new coinage, a new taxation system, and a new administration as well as a new military system. Under Constantine it acquired a new capital and a new religion too. But the late Roman Empire was not created in a revolution. Well before Diocletian’s reign a new ideal of the emperor had emerged, crowding out the productions of senatorial historians and Greek panegyrists. This emperor was a soldier rather than a fellow citizen, and he was surrounded by spectacular ceremonial and ferocious justice.

For us, looking back with hindsight, the most amazing aspect of this story is not that the crisis occurred, but that the empire survived it at all. The energy of the soldier emperors was clearly one factor, but there were other sources of strength not appreciated at the time. Consider, for example, the commitment of the empire’s elite to its continued existence. The ‘Gallic emperors’ who controlled Gaul, and at times the Spanish and British provinces, between 260 and 274 are a case in point. The main figures, Postumus, Victorinus, and Tetricus I, were all soldiers and all apparently descended from rich local families. Their support was drawn from both local nobles and the army of the Rhine. Their ‘empire’ originated in a revolt against Gallienus, but its main efforts were directed at survival and the preservation of vested interests. Following the successes of first Claudius II and then Aurelian, provinces, cities, and then even the last of the emperors rejoined the central empire. Throughout the secession the political propaganda, known mainly through coinages, was utterly Roman. At the other end of the empire the fierce resistance put up by Greek cities drew on even older allegiances. Publius Herennius Dexippus, a historian who organized resistance at Athens, presented his efforts as just the latest episode in a long history of Athenian resistance to the barbarian. The survival of these allegiances is impressive testimony to the durability of the identities created in the early empire. The empire survived because, when it seemed about to come apart, the ruling classes and many of its subjects chose to participate in its rescue.

Further Reading

The evolution of the Roman military machine is surveyed in a number of essays in the second volume of the Cambridge History of Greek and Roman Warfare (Cambridge, 2007) edited by Philip Sabin, Hans van Wees, and Michael Whitby. Adrian Goldsworthy’sRoman Army at War (Oxford, 1996) is a lively account of how it worked in practice. The evolution of a stable frontier, and its social and economic consequences, is the subject of C. R. Whittaker’s Frontiers of the Roman Empire (Baltimore, 1994). Debate still largely responds to Edward Luttwak’s controversial but stimulating Grand Strategy of the Roman Empire (Baltimore, 1976). A good selection of these responses is included in John Rich and Graham Shipley’s collection War and Society in the Roman World(London, 1993). Benjamin Isaac’s Limits of Empire (Oxford, 1990) discusses the role of Roman armies in controlling the provincial populations they claimed to protect.

The complex history of the third-century crisis is covered in the usual reference works, but there is a particular good account in David Potter’s The Roman Empire at Bay (London, 2004). A good sense of what historians are arguing about at the moment is given by the collection Crises and the Roman Empire (Leiden, 2007) edited by Olivier Hekster, Gerda de Kleijn, and Danielle Slootjes.

If you find an error or have any questions, please email us at Thank you!