Ancient History & Civilisation


The Roman Empire from Diocletian to Alaric



Emperors come to power – Diocletian 284 and Jovian 363


The military monarchy


The age of Diocletian and Constantine


The emergence of Constantine


Constantine's successors


Julian from Gaul to Persia


Valentinian, Valens, and the northern frontier


The Goths and the Battle of Adrianople


The reign of Theodosius I


The western dominance of Stilicho, the rise of Alaric the Goth, and the fall of Rome


The consolidation of the eastern empire


In the spring of 283 the Roman emperor Marcus Aurelius Carus died before the walls of the Persian city of Ctesiphon, struck by a bolt of lightning. His son, M. Aurelius Numerianus, who had accompanied him on campaign, was hailed to succeed his father by the army, but soon perished himself. The sources say that he was taken ill with an eye infection and confined to a litter, in which he was ambushed and killed by the praetorian prefect Aper. News of the assassination was kept from the troops, and the victim's body was hidden by Aper until the moment when he could seize power himself, but the crime was betrayed by the stench of the putrefying corpse. The story reached its dénouement on November 20, 284, at Nicomedia in Bithynia, when a council of army officers chose one of their number, Valerius Diocles (later Diocletianus), as the new Augustus. Diocletian addressed the troops for the first time at his inauguration, called upon the sun god to avenge the death of his predecessor, and struck a dagger into Aper, who was standing next to him. He swore an oath that he himself had played no part in the death of Numerian or willed his own accession to power. The coup was not yet complete, for Carus' elder son Carinus still retained control of the western part of the empire. Diocletian led his forces into Illyricum and confronted Carinus at the battle of Margus, near Viminacium, a short way east of modern Belgrade. Carinus succumbed, not apparently to Diocletian's forces but on account of the hatred he had incurred in his own ranks. We are told that he had denounced and condemned innocent men and seduced the wives of his officers. At the end he was slain by one of his tribunes, whose wife he had debauched.

This version of the accession of Diocletian derives from a group of fourth-century Latin accounts, which were largely drawing on a single source.1 It depicts the key elements in imperial accessions during the third and fourth centuries: the selection of the candidate by an inner circle of military officers, the ruler's address to his troops, calling on the support of the gods for the new regime, and their acclamation of the candidate.2 These and other procedures can be identified in many other imperial proclamations of the period. However, the accounts of Diocletian's accession raise problems that recur throughout the study of the later Roman Empire, which should be highlighted from the outset. Firstly, and transparently, the narrative is almost wholly apologetic for the new regime. It suggests that Diocletian was a reluctant nominee to imperial power and had no part in the death of his predecessors; on the contrary he had actually avenged Numerianus' presumed assassin, Aper. He might even have suffered defeat by Carinus' forces, had these not turned against their leader in justified outrage at his excesses. This was the authorized version of events, and it is full of implausibilities. How had the murder of Numerian passed unnoticed? Why had Aper failed to seize the initiative that he had created by his crime? Why in any case had he allowed Diocletian to be acclaimed Augustus, if this is what he had plotted for himself? If Diocletian was genuinely a reluctant ruler, not responsible for overthrowing his predecessors, why should he have led an army against Carinus in Illyricum, instead of offering loyal support to him? Why then admit almost to losing the decisive battle and attribute his opponent's fall to his own moral degeneracy? A precise answer to all these questions is beyond our reach. Imperial coups d'état were not the subject of investigative journalism at the time, and the intrigues and plots which brought down or created emperors spawned self-interested rumor and accusations, not dispassionate enquiries. However, a reasonable reading of the evidence would acquit Aper of conspiracy and suggest that he had been the senior military officer who remained loyal to house of Carus, but had been killed by the new man whose cause had been promoted by a rival group of army officers. The logic of usurpation then determined Diocletian's attack on Carinus, while the story that he had been killed by one of his own officers diverted attention from the crime of the usurper to those alleged against the incumbent emperor.3

Behind the false certainties offered by the official sources, there are important features of the narrative that no propagandistic distortion could obscure. Firstly, the drama of Diocletian's accession was acted out across the main axis of power that joined the northern and eastern frontiers of the Roman Empire (Map 3.1). The route that led Carus into battle against the Sassanian Empire and ended before the walls of Ctesiphon had been trodden before him by the emperors Trajan in 114, Lucius Verus in 165, and Septimus Severus in 198. It was to be followed again by Galerius in 298, and most famously by Julian in his ill-starred Persian campaign of 363. Nicomedia, at the hub of communication between the Balkans and Asia Minor, had grown steadily in importance since Trajan's day and was soon to become Diocletian's own imperial capital.4 Finally the decisive encounter between the rival rulers occurred in the middle Danubian region, midway along the main military highway between the western and eastern provinces. This was the critical zone for imperial civil wars, in military terms a much more significant frontier than the outer boundaries of the empire. The conflict of Diocletian and Carus at Margus presaged the battles between Licinius and Constantine at Cibalae in 316, the army of Constantius and Magnentius at Mursa in 351, and Eugenius and the forces of Theodosius at the river Frigidus in 394.


Map 3.1    The empire of the tetrarchs c.300

Army commands and military power gave usurpers the capacity to seize political control of the Roman Empire. The emperors themselves were serving military officers, and they achieved their positions with the help of demonstrable support from the army; both from the officers who organized the dynamics of accession, and from the regular troops, whose raucous shouts and acclamations formally bestowed the title of Augustus on new rulers. Events between 283 and 285 were defined by the use of extreme political violence: covert and public assassinations; the manipulation of judicial processes, and civil war. In these struggles for power the winners took all, and losers were not permitted to survive to mount a future challenge.

The accession of Diocletian is often held to have ended the so-called period of military anarchy, which had lasted since the fall of the Severan dynasty in 235, and brought an era of stability to the Roman world. However, much remained unchanged by the new order. We need only to look ahead to the years 363–4, when the events of eighty years before were virtually replayed.5 The emperor Julian led a large military expedition to Ctesiphon. After initial successes against Sassanian forces he too died on Persian soil, struck by an unknown hand during a skirmish as his troops withdrew northwards beyond the river Tigris. No Caesar was on hand to succeed him, and the officers of the imperial army, which combined forces from the western and eastern empire, wrangled over the succession. The first name to be proposed was a compromise figure, the aged praetorian prefect Salutius Secundus, but he disqualified himself (or perhaps was seen as unacceptable by the officer circle) on the grounds of his advanced years and lack of military qualifications.6 A younger candidate then emerged, perhaps as the candidate of junior officers rather than the high command. Flavius Iovianus was the senior officer of the regiment of domestici, the palace guard, who had escorted the body of Constantius from Tarsus to Constantinople in 361 and may well, therefore, have been related to the imperial family.7

Jovian's reign proved shorter even than that of Numerian. In the crowded days after Julian's death the Roman army withdrew from its exposed position beyond the Tigris. A route was chosen through desert country across Upper Mesopotamia to Nisibis, and the march brought the army to near starvation. The price of its rescue was a peace treaty which ceded much of Mesopotamia, including the city of Nisibis, to the Sassanians, and relinquished Rome's title to a group of small provinces beyond the Tigris, which had been acquired in 298. Jovian himself drew back to Antioch in Syria and then made a winter journey across Anatolia towards Constantinople. He ceremonially assumed the consulate on January 1, 364, at Ancyra, and among other legislative acts, rescinded Julian's famous law which forbade Christians to be professional teachers (CTh. 13.3.6). In February he continued his journey and reached the small town and staging post of Dadastana in eastern Bithynia on February 17. The following day he was found dead in his bed, allegedly suffocated by fumes from a brazier which had been placed in his newly decorated bedroom.

Once again the main imperial highway (later to be known to Christians as the Pilgrims' Road to Jerusalem) witnessed a bizarre imperial death, which was followed, like that of Numerian eighty years earlier, by the choice of a new emperor (Ammianus 26.1–5). The entourage of officers again considered candidates, and opted for Valentinian, who had recently been appointed by Jovian to command a legion which was stationed in Ancyra. The new man was rushed to join the army at Nicaea, where he mounted the platform to be acclaimed Augustus by the troops. Within a month he co-opted his brother Valens to be his colleague, and in May both set out westwards for Naissus on the boundary of Thrace and Lower Moesia, where they parted ways. Valens returned to the East, while Valentinian made his way to the western provinces, where he spent the remaining eleven years of his reign securing the troubled frontier from Germanic attacks and incursions.

The events of 363–4 are recorded in detail by the greatest Latin historian of the fourth century, Ammianus Marcellinus, who had accompanied Julian's Persian expedition. In contrast to the situation that followed the fall of Numerian and Carinus, alternative versions of the deaths of Julian and Jovian circulated freely. Neither Jovian nor Valentinian had been directly implicated in his predecessor's death, and they had no need to produce a sanitized version of their succession to power. However, the main features of the events of 363/4 uncannily matched the precedent of 283/4. The theater of political activity extended across the same long and narrow stage, the military roads that extended from eastern Mesopotamia to the central Balkans. In each case an emperor died on an eastern campaign, their successors succumbed on the highway across northwest Asia Minor, and the emperors after them were elevated to power in the leading cities of Bithynia, Nicomedia, and Nicaea. Once again a cabal of officers moved to present their chosen candidate to be hailed by the campaigning army. Diocletian, Jovian, and Valentinian all came from similar backgrounds. They were middle-ranking military men, aged in their thirties or early forties, who had been born in the provinces of Illyricum: Diocletian from Salona in Dalmatia, Jovian from Singidunum in Moesia, and Valentinian from Cibalae in Pannonia. None of this may be regarded as coincidental. These narratives reflect both the realities of power and the political and military imperatives which shaped life at the highest levels of the Roman state of the later third and fourth centuries.

What are the lessons of this prelude? Reconstructing a historical narrative for the fourth century is a necessary task, but an almost impossible one. Apart from the twenty-five-year period covered by Ammianus Marcellinus, which included the reign of Jovian, the narrative sources are brief, and usually tendentious or confused. Gibbon made the point in his famous adieu to Ammianus when his own account reached the period after the battle of Adrianople in 378: “It is not without the most sincere regret that I must now take leave of an accurate and faithful guide, who has composed the history of his own times without indulging the prejudices and passions which usually affect the mind of a contemporary.”8

Careful sifting of the evidence usually makes it possible to reconstruct events in chronological order, and to identify time, place, and actors in historical transactions. But questions of causation or motivation are always obscure, subjective, and distorted. The way that most of our sources presented great events in imperial history was usually determined by the victors, who created a narrative and expounded an interpretation that justified their own role in events. This is particularly true at moments of regime change. The higher the political stakes, the greater the likelihood of spin and distortion.

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