Chapter 1
The Goths had a momentous impact on Roman history, appearing as if out of nowhere in the early decades of the third century. When we first meet them, it is in the company of other barbarians who, together, made devastating incursions into the eastern provinces of the Roman empire. The mid third century, particularly from the 240s till the early 300s, was an era of constant civil war between Roman armies, civil war that in turn encouraged barbarian invasions. Contact with the Roman empire, and particularly with the Roman army, had helped to militarize barbarian society, and opportunistic raids all along the imperial frontiers exploited Roman divisions and distraction in the civil wars. When the Goths first appear, it is in this world of civil war and invasion. Unfortunately for the modern historian, it is not always easy to distinguish third-century Goths from other barbarians. The problem stems from the way ancient writers talked about barbarians in general and the Goths in particular.
‘Scythians’ and Goths
To the Greek authors who wrote about them, the Goths were ‘Scythians’ and that is the name used almost without exception to describe them. The name ‘Scythian’ is very ancient, drawn from the histories of Herodotus, which were written in the fifth century B.C. and dealt with the Greek world at the time of the Persian Wars. For Herodotus, the Scythians were outlandish barbarians living north of the Black Sea in what are now Moldova and Ukraine. They lived on their horses, they ate their meat raw, they dressed in funny ways, and they were quintessentially alien not just to the world of the Greeks, but even to other barbarians nearer to the Greek world. Greek historical writing, like much of Greek literary culture, was intensely conservative of old forms, and canonized certain authors as perfect models to which later writers had to conform. Herodotus was one such canonical author and his history was regularly used as a template by later Greek historians. In practice, this meant that authors writing 500 or 1,000 years after Herodotus talked about the world of their own day in exactly the same language, and with exactly the same vocabulary, as he had used all those centuries before.
For Greek writers of the third, fourth and fifth centuries A.D., barbarians who came from the regions in which Herodotus had placed the Scythians were themselves Scythians in a very real sense. It was not just that classicizing language gave a new group of people an old name; the Greeks and Romans of the civilized imperial world really did believe in an eternal barbarian type that stayed essentially the same no matter what particular name happened to be current for a given tribe at any particular time. And so the Goths, when they first appear in our written sources, are Scythians – they lived where the Scythians had once lived, they were the barbarian mirror image of the civilized Greek world as the Scythians had been, and so they were themselves Scythians. Classicizing Greek histories often provide the most complete surviving accounts of third- and fourth-century events, and the timelessness of their vocabulary can interpose a real barrier between the events they describe and our understanding of them.[2]However, the testimony of our classicizing texts sometimes overlaps with that of less conservative writings that employ a more current vocabulary. Because of such overlaps, we can sometimes tell when actions ascribed to Scythians in some sources were undertaken by people whom contemporaries called Goths.
Map 2. The Roman Empire at the time of Septimius Severus.
The Earliest Gothic Incursions
Because of this complicated problem of names in the sources, we cannot say with any certainty when the Goths began to impinge upon the life of the Roman empire, let alone precisely why they did so. The first securely attested Gothic raid into the empire took place in 238, when Goths attacked Histria on the Black Sea coast and sacked it; an offer of imperial subsidy encouraged their withdrawal.[3] In 249, two kings called Argaith and Guntheric (or possibly a single king called Argunt) sacked Marcianople, a strategically important city and road junction very near the Black Sea.[4] In 250, a Gothic king called Cniva crossed the Danube at the city of Oescus and sacked several Balkan cities, Philippopolis – modern Plovdiv in Bulgaria – the most significant. Philippopolis lies to the south of the Haemus range, the chain of mountains which runs roughly east-west and separates the Aegean coast and the open plains of Thrace from the Danube valley (all cities are shown on map 4 in chapter four). The fact that Cniva and his army could spend the winter ensconced in the Roman province south of the mountains gives us some sense of his strength, which is confirmed by the events of 251. In that year, Cniva routed the army of the emperor Decius at Abrittus.[5] Decius had persecuted Christians, and Lactantius, a Christian apologist of the early fourth century, recounts with great relish how Decius ‘was at once surrounded by barbarians and destroyed with a large part of his army. He could not even be honoured with burial, but – despoiled and abandoned as befitted an enemy of God – he lay there, food for beasts and carrion-birds’.[6]
The Black Sea Raids
Gothic raids in Thrace continued in the 250s, and seaborne raids, launched from the northern Black Sea against coastal Asia Minor, began for the first time. What role Goths played in these latter attacks is unclear, as is their precise chronology. The first seaborne incursions, which took place at an uncertain date between 253 and 256, are attributed to Boranoi.[7] This previously unknown Greek word may not refer to an ethnic or political group at all, but may instead mean simply ‘people from the north’. Goths did certainly take part in a third year’s seaborne raids, the most destructive yet. Whereas the Boranoi had damaged sites like Pityus and Trapezus that were easily accessible from the sea, the attacks of the third year reached deep into the provinces of Pontus and Bithynia, affecting famous centres of Greek culture like Prusa and Apamea, and major administrative sites like Nicomedia.[8] A letter by Gregory Thaumaturgus – the ‘Wonderworker’ – casts unexpected light on these attacks. Gregory was bishop of Neocaesarea, a large city in the province of Pontus, and his letter sets out to answer the questions church leaders must confront in the face of war’s calamities: can the good Christian still pray with a woman who has been kidnapped and raped by barbarians? Should those who use the invasions as cover to loot their neighbours’ property be excommunicated? What about those who simply appropriate the belongings of those who have disappeared? Those who seize prisoners who have escaped their barbarian captors and put them to work? Or, worse still, those who ‘have been enrolled amongst the barbarians, forgetting that they were men of Pontus and Christians’, those, in other words, who have ‘become Goths and Boradoi to others’ because ‘the Boradoi and Goths have committed acts of war upon them’.[9]
Ten years later, these assaults were repeated. Cities around the coast of the Black Sea were assaulted, not just those on the coast of Asia Minor, but Balkan sites like Tomi and Marcianople. With skillful seamanship, a barbarian fleet was able to pass from the Black Sea into the Aegean, carrying out lightning raids on islands as far south as Cyprus and Rhodes. Landings on the Aegean coasts of mainland Greece led to fighting around Thessalonica and in Attica, where Athens was besieged but defended successfully by the historian Dexippus, who would later write an account of these Gothic wars called the Scythica.[10] Though only fragments of this work survive, Dexippus was a major source for the fifth- or early sixth-century New History of Zosimus, which survives in full and is now our best evidence for the third-century Gothic wars. As Zosimus shows us, several imperial generals and emperors – Gallienus, his general Aureolus, the emperors Claudius and Aurelian – launched counterattacks which eventually brought this phase of Gothic violence to an end. Gothic defeat in 268 ended the northern Greek raids, while Claudius won a smashing and much celebrated victory at Naissus, modern Niš, in 270.[11]
Aurelian and a Problematic Source
In 271, after another Gothic raid across the Danube had ended in the sack of several Balkan cities, the emperor Aurelian (r. 270–275) launched an assault across the river that probably had considerable success. Aurelian was an extremely capable soldier, and one who spent his five-year reign in continuous motion from one end of the empire to the other, rarely out of the saddle, and rarely pausing between campaigns. A Gothic war is entirely in keeping with the evidence for Aurelian’s movements, and a late fourth-century collection of imperial biographies which we call the Historia Augusta records that Aurelian defeated and captured a Gothic king named Cannobaudes.[12] Here, however, we run into the sort of problem with the sources that we will encounter more than once in the pages that follow. The Historia Augusta is the only Latin source we have for large chunks of third-century history, and even where it refers to events known from Greek historians, it often preserves details that they do not. If it could be trusted, its circumstantial and anecdotal content would be invaluable. Unfortunately, the whole work is heavily fictionalized, its anonymous author sometimes using older – and now lost – texts as a jumping off point for invention, sometimes making things up out of thin air. The biographies of late third-century emperors are the least reliable part of the work, and some of them contain no factual data at all. For that reason, even though he appears in many modern histories of the Goths, we cannot be entirely sure that this Gothic Cannobaudes was a real historical figure.
In this case, however, we are able to confirm at least part of the Historia Augusta’s testimony from another type of evidence altogether, because inscriptions make clear that Aurelian did definitely campaign against Goths. From a very early stage in Roman history, whenever a Roman general won a victory over a neighbouring people, he would add the name of that people to his own name, as a victory title. When the Roman Republic gave way to the one-man rule of the empire, the honour of such victory titles was reserved for the emperor, and whether he won a victory personally, or whether a general won it in his name, it was the emperor alone who took the victory title. In this way, a Persian campaign would allow the emperor to add the title Persicus, a campaign against the Carpi would make the emperor Carpicus, and so on. Since these victory titles became part of the emperor’s name, they were included in the many different types of inscriptions, official and unofficial, that referred to the emperor. This provides a wealth of information for the modern historian, because victory titles often attest campaigns that are not mentioned by any other source. Thus in the chapters that follow we will sometimes be able to refer to a particular emperor’s Gothic campaign only because an inscription happens to preserve the victory title Gothicus – as in the present case, Aurelian’s use of the name shows that he did in fact fight against the Goths and felt able to portray that campaign as a success. We can also infer that success from the fact that his Gothic victory was still remembered a hundred years later, and from the rather limited evidence for Gothic raids in the decades immediately following his reign: although we hear of more seaborne raids in the mid-270s that penetrated beyond Pontus deep into Cappadocia andCilicia, after that Goths disappear from the record until the 290s, by which time major changes had taken place in the empire itself.[13]
Explaining the Third-Century Invasions
As the past few pages have demonstrated, the earliest evidence for Gothic invasions of the empire is not well enough attested to allow for much analysis, but that does not mean we should underestimate its impact. The letter of Gregory Thaumaturgus gives us a rare glimpse into just how traumatic the repeated Gothic raids into Asia Minor and other Greek provinces could be. But it does not answer basic questions of causation: what drove these Gothic raids, what made them a repeated phenomenon? The Graeco-Roman sources are content to explain barbarian attacks on the empire with an appeal to the fundamentals of nature itself: to attack civilization is just what barbarians do. That sort of essentialist explanation can hardly be enough for us. Rather, we need to seek explanations in the historical context. Now it happens that the third century was a period of massive change in the Roman empire, which saw the culmination of social and political developments that had been set in motion by the expansion of the Roman empire in the course of the first and second centuries A.D. Against this background, the first appearance of the Goths and the Gothic raids of the third century become comprehensible. Roman expansion had transformed the shape of Europe and the Mediterranean basin. It affected not just the many people who became Romans for the first time, but also the political constitution of the empire and even the many different peoples who lived along the imperial frontiers. One by-product of these changes was a cycle of internal political violence in the third-century empire that produced and then exacerbated the instability of the imperial frontiers.
The Roman empire had been a monarchy since the end of the first century B.C., when Augustus (r. 27 B.C.–A.D. 14), the grand-nephew and adoptive heir of Julius Caesar, put an end to a full generation of civil war that had ripped the Roman Republic apart. Augustus brought peace to the empire, but it came at the expense of the free competition amongst the Roman elite that had created a Roman empire to begin with. In its place, Augustus founded an imperial dynasty that lasted until A.D. 68. By that year, when the regime of the detested emperor Nero collapsed and he himself committed suicide, three generations had passed since the end of the Republic. The imperial constitution was fully entrenched – what mattered most was the relationship of the emperor to the powerful clans of the Roman elite, particularly the senatorial families of Rome itself, who now competed amongst themselves for the emperor’s favour and the offices and honours it bestowed. Until 68, emperors had been made at Rome, and loyalty to the dynasty of Augustus had been an essential element in their creation. The civil wars of A.D. 68/69 changed that forever: their eventual victor was Vespasian, a middle-aged commander born of a prosperous but undistinguished Italian family and raised to the imperial title in the eastern provinces of the empire, just as some of his immediate rivals had seized the purple in Spain or Germany. This revealed what Tacitus called the arcanum imperii, the ‘secret of empire’ – that an emperor could be made outside Rome.[14] Italy remained the centre of the empire, but it was no longer the sun around which provincial planets revolved. These provinces increasingly had a life of their own and political influence that could, in time, impose itself on the Italian centre.
To be sure, the provinces might be very different from one another, and they might stand in different relationships to the imperial capital in Rome. Some provinces, like Spain, southern Gaul, or the part of North Africa that is now Tunisia, had been part of Rome’s empire for a century or more. Others, like Britain, much of the Balkans, or what is now Morocco were only a generation away from their conquest by Roman armies. Well into the late third century, these different provinces continued to be governed according to many differing ad hoc arrangements that had been imposed on them when they were first incorporated into the empire. But all the imperial provinces were more and more integrated into a pattern of Roman life and ways of living, much less conquered territories administered for the benefit of Roman citizens in Italy. Indeed, the extension of Roman citizenship to provincial elites was an essential element in binding the provinces to Rome. As provincial elites became Roman citizens, they could aspire to equestrian or senatorial rank, and with it participation in the governance of the larger empire. Already by A.D. 97, a descendant of Italian immigrants to Spain named Trajan had become emperor. Trajan’s successor Hadrian was likewise of Spanish descent, while his own successor and adopted son came from Gallia Narbonensis, the oldest Roman possession in Gaul.
Roman Citizenship and Roman Identity
These provincial emperors are the most impressive evidence for the spread of Roman identity to the provinces, but the continuous assimilation of the provincial elites into the Roman citizenship was ultimately more important in creating the sense of a single empire out of a territorial expanse that stretched from the edge of the Arabian desert to Wales, from Scotland to the Sahara. These imperial elites could communicate with one another, linguistically and conceptually, through a relatively homogeneous artistic and rhetorical culture. This culture was founded on an educational system devoted almost exclusively to the art of public speaking, the rhetorical skills that were necessary for public, political life. Mainly Greek in the old Greek East, frequently Graeco-Roman in the Latin-speaking provinces of the West, this elite culture nurtured an aesthetic taste devoted, in Greek, to the fashions of the Classical and early Hellenistic period and, in Latin, to those of the very late Republic and early empire. It thereby provided a set of cultural referents and social expectations shared by Roman citizens and Graeco-Roman elites from one end of the empire to the other, and allowed them to participate in the common public life of the empire at large, even if they came from wildly divergent regions.
The use of Roman law, which came with the acquisition of Roman citizenship, provided a framework of universal jurisdiction that, for the elites who used it, also overcame regional differences. Because of the growing elite participation in the Roman world and its governance, those lower down the social scale began in time to feel some measure of the same integration, helped along by the hierarchies of patronage that permeated the whole Roman world. The cult of the Roman emperors, and of the personified goddess Roma, was another effective means of spreading the idea of Rome and participation in a Roman empire to the provinces. Greg Woolf has examined in detail how incorporation into an ordered network of provincial government – with the assimilation of local elites into Roman citizenship – could transform an indigenous society.[15] In northern and central Gaul, less than two generations after the organization of the local tribal territories into a Roman province, both old Celtic noble families and the larger Gallic population had learned to express traditional relationships of patronage and clientship, power and display, in Roman terms, eating off Roman tableware, living in Roman houses, and dressing as Romans should. The same process is observable in the Balkans, at a slightly later date but at the same relative remove from the generation of the conquest. In the Greek world, ambivalent about its relationship to a Latin culture that was younger than – and partially derivative of – Hellenic culture, assimilation was more complicated, but even if Latin culture had little visible presence, the sense of belonging to a Roman empire was very strong in the ancient cities of the East.
This convergence on a Roman identity within the empire culminated in a measure taken by the emperor Caracalla in A.D. 212. Caracalla was himself the heir of an emperor from Africa – Septimius Severus, a man who could attest indigenous Punic ancestry in the very recent past. Much given to giganticism and delusions of grandeur, Caracalla undertook all sorts of massive building projects, and it is in this light that we should understand his decision to extend Roman citizenship to every free inhabitant of the empire in 212. The effects of this law, which we call the Antonine Constitution from Caracalla’s official name of Antoninus, were varied. It both acknowledged the convergence of local elites on a Roman identity and encouraged its continuation, but it also created the dynamic of political violence which dominated the middle and later third century. Once all inhabitants of the empire were Romans, any of them could actively imagine seizing the imperial throne if they happened to be in an opportune position to do so. This was a radical step away from the earlier empire in which only those of senatorial status could contemplate the throne. The Graeco-Roman reverence for rank and social status was extraordinary, and there was a world of difference between accepting the son of a provincial senator as emperor and accepting a man whose father had not even been a Roman citizen. And yet by the middle of the third century, such recently enfranchised Romans not only seized the throne, but their doing so quickly ceased to occasion surprise and horror among the older senatorial nobility.
Warfare and the Rhetoric of Imperial Victory
If the expansion of citizenship and the broadening definition of what it meant to be Roman permitted such men to imagine themselves as emperor, it was increasing military pressures that made their doing so practicable. Much earlier, in the era of Augustus when Roman government was for the first time in the hands of one man, the security of monarchical rule was by no means guaranteed. The authority of the emperor – or princeps, ‘first citizen’, as Augustus preferred to be called – rested on a number of constitutional fictions related to the old public magistracies of the Republic. More pragmatically, however, the authority of Augustus and his successors rested on a monopoly of armed force: that is to say, it rested on control of the army. Empire could not exist without army, and it is hardly an exaggeration to say that the whole apparatus of imperial government developed and grew ever more complex in order to redistribute provincial tax revenues from the interior of the empire to the military establishments on the frontiers. These armies were the ultimate sanction of imperial power, and they needed not only to be paid but also to be kept active: soldiers were far less inclined to mutiny or unrest when they were well supplied and occupied in the business they were trained for, rather than in more peaceable pursuits. This made periodic warfare consistently desirable.
The regular experience of warfare, in turn, fed into the pre-existent rhetoric of imperial victory and invincibility which provided part of the justification for imperial rule: the emperor ruled – and had the right to rule – because he was invincible and always victorious in defending Rome from its enemies. Thus even after imperial expansion stopped early in the second century, the need for Roman armies to win victories over barbarians was ongoing. The result was a constant stream of border wars, which allowed emperors to take victory titles and be seen to fulfill their most important task – defending the Roman empire from barbarians and from the eastern empire of Parthia, the only state to which Roman emperors might reluctantly concede a degree of equality. As we shall see in a moment, the militarization of the northern frontier had for many years had a profound effect on the barbarian societies beyond the Rhine and Danube, but at the start of the third century, a more acute transformation took place on the eastern frontier, again as a result of Roman military intervention.
From Parthians to Persians on the Eastern Frontier
Caracalla is the pivotal figure here as well. In 216 he invaded the Parthian empire, the creation of the central Asian dynasty that had displaced the Hellenistic Seleucids as the rulers of Iran and Mesopotamia during the last centuries B.C. Since the defeat of the Republican general Crassus at Carrhae in 53 B.C., Parthia had possessed an iconic quality as the mortal enemy of Rome that was not matched by the actual strength or competence of the Parthian monarchy. A Parthian war might be a significant ideological goal for a Roman emperor – it avenged Crassus, imitated Augustus, and followed in the heroic footsteps of Alexander the Great – but victories in Parthia could actually be quite easy to win. The Parthian empire was fractious, and its kings faced almost continuous revolts in their far-flung eastern provinces. Thus when Caracalla determined to luxuriate in the easy triumph of a Parthian victory, he unwittingly destroyed the Parthian monarchy. It was replaced by a much more dangerous foe, a new Persian dynasty known as the Sassanians. A Persian nobleman, Ardashir (r. c. 224–241), revolted against the overlordship of the crippled Parthian dynasty and by the middle of the 220s had defeated the last Parthian king.
Under Ardashir’s son Shapur Ⅰ (r. 240–272), the Sassanian monarchy not only imposed itself upon the old Parthian nobility and the subject peoples of the Parthian empire, but also undertook repeated assaults on the Roman empire – Greek and Roman authors attributed to him the ambition of restoring the ancient Persian empire of the Achaemenid dynasty, which had been conquered by Alexander the Great 600 years earlier. Caracalla was murdered in 217 while still on his Parthian campaign, but the new Sassanian Persia became the chief focus of his imperial successors. Not only was there the continued lure of a prestigious Persian victory, there were sound strategic reasons for the imperial focus on the East: Persian raids on the eastern provinces – unlike barbarian attacks on other frontiers – threatened the permanent annexation and removal of the provinces from imperial control. Yet the relentless draw of Persia might distract imperial attention from problems on other fronts, and failure against Persia could be fatal to an emperor’s hold on his throne – the last Severan emperor, Alexander Severus, was murdered after failures on the Persian front, and innumerable third-century emperors faced usurpations in distant provinces as soon as they had turned their attention to the East. The rise of the Sassanians was therefore one of the catalysts for the third century’s cycle of violence. When, as had not been the case a hundred years earlier, a claim on the imperial throne could be contemplated by any powerful Roman and not just the great senatorial generals who had dominated the politics of the second century, then even a minor local crisis – a mutiny, say, or a Persian or barbarian raid – might prompt the local population or the local troops to proclaim a handy leader as emperor to meet the crisis. Having accepted the imperial purple, the new emperor had no choice but to defeat and replace whoever was presently claiming the title. Civil war was inevitable in those circumstances, and the pressures of civil war left pockets of weakness on the frontiers which neighbours could exploit. In consequence, for almost fifty years, a vicious cycle of invasion, usurpation and civil war became entrenched, as even the briefest survey of the mid third century will suggest.
Usurpation, Civil War and Barbarian Invasions
When Alexander Severus was killed in 235, rival candidates sprang up in the Balkans, in North Africa and in Italy, the latter promoted by a Roman senate insistent on its prerogatives. Civil war ensued for much of the next decade, and that in turn inspired the major barbarian invasions at which we have already looked, among them the attack by the Gothic king Cniva that ended in the death of Decius at Abrittus in 251. Decius’ successors might win victories over such raiders, but the iron link between invasion and usurpation was impossible to break. This is clearly demonstrated in the reign of Valerian (r. 253–260), who was active mainly in the East, and that of his son and co-emperor Gallienus (r. 253–268) who reigned in the West. Our sources present their reigns as an almost featureless catalogue of disastrous invasions which modern scholars have a very hard time putting in precise chronological order.[16] We need not go into the details here, and instead simply note the way foreign and civil wars fed off each other: when Valerian fought a disastrous Persian campaign that ended in his own capture by the Persian king, many of the eastern provinces fell under the control of a provincial dynasty from Palmyra largely independent of the Italian government of Gallienus. Similarly, every time Gallienus dealt with a threat to the frontiers – raids across the Rhine into Gaul, across the Danube into the Balkans, or Black Sea piracy into Asia Minor and Greece – he was simultaneously confronted by the rebellion of a usurper somewhere else in the empire. Thus Gallienus had to follow up a campaign against Marcomanni on the middle Danube by suppressing the usurper Ingenuus, while the successful defence of Raetia against the Iuthungi by the general Postumus allowed him to seize the imperial purple and inaugurate a separate imperial succession which lasted in Gaul for over a decade.[17] Even when Gallienus attempted to implement military reforms to help him counter this cycle of violence, the reforms themselves could work against him: he created a strong mobile cavalry that allowed him to move swiftly between trouble spots, but soon his general Aureolus, who commanded this new force, seized the purple for himself and Gallienus was murdered in 268, in the course of the campaign to supress him. As we have now come to expect, his death inspired immediate assaults on the frontiers, by ‘Scythians’ in the Balkans and across the Upper Danube into the Alpine provinces as well.
Again, a full list of invaders and usurpers is an arid exercise and one unnecessary here. The successors of Gallienus – Claudius, Aurelian, Probus, and their many short-lived challengers – faced the same succession of problems as their predecessor had done. Claudius successfully defeated an invading army of Scythians twice, at Naissus and in the Haemus mountains, and won for himself the victory title Gothicus which assures us that those Scythians were Goths.[18] We have already seen that Aurelian won a Gothic campaign, but his energies and attentions were constantly distracted by other invasions, some reaching as far as Italy, and by the civil wars in which he suppressed the independent imperial successions in Gaul and the East. Aurelian fell to assassins, and so too did his immediate successor Tacitus, the latter struck down while in hot pursuit of Scythian – perhaps Gothic – raiders deep in the heart of Asia Minor.[19] Though Probus managed to hold the throne for a full six years, he too was killed in a mutiny that broke out in the face of yet another Balkan invasion, and his praetorian prefect Carus was proclaimed emperor by the legions.[20]
The Accession of Diocletian
We get our first indication that authors of the fourth century had come to understand the connection between internal Roman dissension and barbarian invasion with reference to the death of Probus. As the historian Aurelius Victor put it, writing around 360: ‘all the barbarians seized the opportunity to invade when they learned of the death of Probus’.[21] In response, the new emperor Carus left his elder son Carinus in charge of the western provinces and led an army against the Quadi and Sarmatians on the middle Danube before launching the invasion of Persia during which he met his end – supposedly struck by lightning, perhaps the victim of assassination.[22] The accession of Diocletian at Nicomedia in 284 prompted the inevitable war against Carinus. The latter had restored the Rhine frontier in 283, but by marching east to face Diocletian he allowed new barbarian raids on the Gallic coast. Carinus was defeated and killed at the battle of the Margus in 285, and in that same year, the victorious Diocletian campaigned against the Sarmatians on the Danube. He also appointed a colleague in the imperial office, a fellow soldier named Maximian, who campaigned on the Rhine.[23]
This was a significant step and one with major repercussions for the longevity of Diocletian’s regime. By appointing a co-emperor with whom he was on good terms and who would regard him as his benefactor, Diocletian hoped to give himself the breathing space needed to secure his hold on the throne and prevent rival usurpers appearing in parts of the empire where he could not be himself. The plan worked to a degree, although it took time. Only the appointment in 293 of two caesars, or junior emperors, allowed Diocletian and Maximian to suppress several provincial revolts and secure the frontiers. The evidence of these efforts is visible all along the imperial frontiers, for instance in the so-called Saxon shore forts along the Channel and North Sea coasts of what are now England, France, Belgium and the Netherlands. More important for the history of Roman relations with the Goths is the Diocletianic programme of fortification along the Danube. This consisted both of brand new constructions, as at Iatrus, and also of enlarged and refurbished early imperial fortifications, as at Augustae and Oescus.
Diocletian and the Goths
Such improvements were not simply measures of passive self-defence – they were also bases from which imperial campaigns could be supplied and supported. Already in the 280s, Diocletian and Maximian showed a renewed imperial willingness to campaign beyond the frontiers, and Maximian’s general Constantius – his caesar after 293 – won several spectacular victories against Franks on the lower Rhine. Meanwhile, Diocletian campaigned on the Danube against Tervingi and Taifali, winning victories in 289 and again in 291. That campaign is significant for us because it marks the first appearance of the Tervingian name in Greek or Latin writing. Our source is a panegyric – a speech in praise of the emperor Maximian, delivered in Gaul in 291 – and it refers to the Tervingi as pars Gothorum, which is to say, a section of the Goths.[24] As we shall see in the following chapters, the Tervingi were throughout the fourth century the most important subdivision of the Goths. They were the Gothic group with which the Roman empire had the most regular dealings, and for that reason they are the one about which we know the most. It was the Tervingi with whom the emperor Constantine would conclude a lasting peace in the 330s; descendants of these same Tervingi made up the majority of the Goths who crossed the Danube into the Roman empire in 376, eventually taking part in the Balkan settlements from which Alaric himself would emerge.
For all these reasons, therefore, this first hint of the Tervingi’s existence will automatically seem significant to the modern historian of the Goths. We cannot, unfortunately, tell just how important these third-century Tervingi were at the time, particularly as they are mentioned in the same breath as the Taifali, a group of barbarians who often appear together with the Goths in later sources, but always in an inferior position. What is more, this couple of lines in the panegyric of 291 is the last we hear of the Tervingi or any other Goths for more than a decade. By that point, the internal politics of the empire had changed dramatically yet again. As we shall see, the joint reign of Diocletian and Maximian broke the vicious political cycle of the preceding half century. In the process of doing so, they reinvented the governmental system of the Roman empire, strengthening the central government and laying the foundations of a political system that lasted for several hundred years. Just as important, by finally establishing a secure hold on the imperial office, Diocletian and his colleagues were also able to secure more stable relations with barbarian groups along the frontiers. We will return to the government of Diocletian and to the imperial frontiers in chapter four, paying particular attention to the lower Danube. There, by the 320s, the Goths were unquestionably the dominant political force immediately beyond the frontiers, a position they had achieved partly because the emperors wanted them to. In the meantime, however, we must turn to an important interpretative question which is raised by our discussion of third-century invasion and civil war.
If, as we have suggested, the middle of the third century can be defined by this constant cycle of internal and external violence, we are still left to ask why it was that barbarian groups along the northern frontiers could exploit imperial weakness, and particularly imperial rivalry, so successfully and widely. After all, this ability was something quite new, unknown in the early empire, when imperial generals could rampage at will through the land beyond the imperial frontiers. Then, the central European lands beyond the Rhine and Danube were a patchwork of very small political units that could be brought together for coordinated action only for very short periods of time, if at all. That is the situation depicted in the classic account of Tacitus’ Germania, written in A.D. 98, and corroborated by the political history of the period. The later second and the third centuries stand in very sharp contrast to this early imperial picture. Now, beginning in the 160s and 170s, barbarian groups along the northern frontiers challenged the empire in ways that had always eluded them previously, and did so on a scale never before seen. If we are to understand this exponential growth in the ability of Rome’s northern neighbours to pose a threat to the empire, we need to look at the social and political history of barbarian Europe. There, during the first and second centuries, society was transformed in ways that paralleled changes inside the imperial provinces.