12
DIOCLETIAN, CONSTANTINE AND THE CREATION OF THE LATER ROMAN EMPIRE
The quickest way to make sense of the tetrarchy is to look at its coins. As the third century progressed, and claimant followed claimant with increasing speed, the portraiture of the coinage declined in both complexity and representational specificity. At mid-century, and especially on the bronze sestertii, the portraits of emperors like Decius show the sort of individuality that we can trace all the way back to the Hellenistic tradition of naturalistic coin portraiture, but already before that, the silver-washed antoniniani of Caracalla and Severus Alexander look more schematic and stylised; by the time of Claudius, Aurelian and Probus, sestertii had disappeared and portraits on the antoniniani no longer look as if they are meant to depict real people, but are instead stylised statements of rulership. The stylisation changes slightly from reign to reign, but with their beards, helmets and armour, they remain little more than abstract images of imperial power and authority. Much the same can be said of the early portraits of Diocletian. But then, in 293, he and Maximian not only took on two new junior emperors, they also reformed the mints and introduced an entirely new type of coin, whose ancient name is unknown, but which we call the follis. This medium-sized bronze type – at around 10 grams, double the weight of an antoninianus and half that of an old sestertius – was minted across the empire, in a whole series of new mints and in a variety of provinces. On these folles, the four emperors – the augusti Diocletian and Maximian and the caesars Constantius and Galerius – are indistinguishable. In some types they appear in military garb, in others in civilian, but their portraits are all the same: short full beard and moustache in the typical military fashion; heavyset, even jowly, faces; eyes fixed on the middle distance with no attempt at portraiture; and necks more like those of bulls than human beings. As an image of the unity and ubiquity of imperial purpose, tetrarchic coinage would be hard to better.
But these new coins were just one small part of a much larger Diocletianic programme of reform implemented by the tetrarchy. These reforms were an ongoing response to the political traumas the tetrarchs had experienced working their way up through the fractious military hierarchy of the post-Gallienan decades, a response that Diocletian was imaginative enough to modify when something seemed not to be working. The first element in the new tetrarchic programme, with which we ended chapter 11, was the multiplication of the imperial office and the creation of what we can call a college of emperors. Flavius Constantius, who had served as Maximian’s praetorian prefect and had been married into his family since 289, was made caesar on 1 March 293, probably at Mediolanum. He immediately launched an assault on Carausius’s Gallic strongholds and routed him, even taking the naval stronghold of Bononia (modern Boulogne), the main staging post for sailing to Britain. On the same day in March 293, probably at Sirmium in the Balkans, Diocletian proclaimed C. Galerius Maximianus as caesar in a parallel move, marrying him to his own daughter Valeria in June. Constantius would be the senior caesar, as Diocletian was the senior augustus, and the fiction of divine kinship already established for Diocletian and Maximian was extended further, with Constantius joining the ‘Jovian’ line of Diocletian and Galerius the ‘Herculean’ line of Maximian. Both caesars added a new nomen, Valerius, to their titulature, confirming their relationship to Diocletian. This division of government, with its publicly legible ideology, would make it easier to put an emperor on the spot wherever one might be needed. The eastern pair would begin to move against Persia, the western would continue to stabilise the frontiers and suppress usurpers, of whom Carausius had not been the last. Indeed, the defeat of Carausius in battle led to his murder by a man named Allectus, of whom very little is known save that he assumed Carausius’s imperial title and began to mint coins in his own name.
Crushing usurpations as they sprang up would never, in itself, be enough to ensure stability, and Diocletian realised that. His new tetrarchic dispensation was meant to make such attempts much more difficult to stage in future than they had been in the past. A key move in that direction was redrawing provincial boundaries and thereby multiplying the number of imperial administrators, with a view towards both improving oversight and weakening the capacity of any one governor to cause trouble or win for himself a power base with which to challenge the ruling emperor. As we saw throughout chapter 11, the military side of imperial government had become more and more prominent, and likewise more specialised, as the third century progressed. The Julio-Claudian and Antonine high imperial model, in which all members of the governing class were at least theoretically able to serve both as civil administrators and military commanders, had given way to a system in which military commands went to career soldiers, many of them risen through the ranks, and more often than not promoted on the basis of ability rather than birth. That was the background of all the emperors after Gallienus. At least three of the tetrarchs – Diocletian, Maximian and Constantius – had come up through precisely that career path, and though little certain is known of Galerius save his parentage (his parents were peasants from near Romulianum on the Danube and his name at birth was Maximinus), he had undoubtedly served in the army.
So predominant was this career trajectory that fourth-century writers could imagine a moment in the third century when someone had made a conscious decision to exclude senators from high office (they attributed the choice to Gallienus, an irony given that he more than any other emperor of the period had portrayed himself in a traditionally senatorial vein). In fact, of course, the process was a natural one that grew out of the need to deal with the various short-term crises of the third century. Diocletian, understanding the value of a professional army, formalised some of the trends that had been emerging over the previous decades. He began by separating military from civilian commands altogether, thus marking off a divide between career paths through the officer corps and career paths in civil administration. The immediate value of this was not merely to confirm the professional skills of the relevant commanders, but also to ensure that those who commanded troops in battle would not at the same time be responsible for paying them their salaries. By separating these two functions, and relying on rivalry between soldiers and bureaucrats, Diocletian reckoned that any would-be usurper would have a hard time paying an army; and since an unpaid army was not a loyal one, the perennial third-century threat of usurpation would be severely limited.
Important as that move was, there was more. Diocletian radically revised the imperial administrative map, chopping the old Severan provinces up into more than a hundred much smaller provinces (as an example, Gallia Lugdunensis and Gallia Belgica were both split in half, doubling the number of provinces in that region of north central Gaul). These were administered by civilian governors of differing rank, the prestige accruing to each province developing into a very definite hierarchy over time. The idea was to provide greater oversight of regional affairs and stronger direct connections to the imperial government than had been possible in the old Augustan – or even the Severan – provincial system, with its enormous provinces and imperial government relying almost entirely upon local municipalities and local big men to mediate its interests. Where necessary, these provinces were assigned separate military commanders of middling rank, known as duces, further separating military from civilian governance. Sometime between 297 and 305, these new provinces were grouped into larger units for the purposes of fiscal administration and taxation by both the palatine treasuries and the staff of the praetorian prefects. These groups of provinces were known as dioceses and seem to have been supervised on an ad hoc basis, by the prefect himself or by a deputy, until, by 314 at the latest, each was put under its own official, who was known as a vicarius.
The prefects, too, multiplied under the Diocletianic system, as was necessary given the multiplication of the emperors. But they also now lost most of their military functions, becoming instead the most senior administrators in a civilian hierarchy, a process that became final in the second decade of the fourth century. A prefect was assigned to each augustus, and perhaps to each caesar, and travelled with his imperial court. Because these courts were so mobile, a prefect’s sphere of activity was coterminous with that of the augustus (or caesar) he served and was not territorially fixed. By keeping the highest ranking administrators with them, but then fixing groups of provinces into dioceses that could be supervised by lesser officials on the prefects’ behalf, the tetrarchs maintained the flexibility and adaptability of their collegiate emperorship but also provided for hierarchical and territorial supervision.
Change in other government bureaus accelerated as well. We have already mentioned the importance of jurists and lawyers in the period after Caracalla’s citizenship edict, as Roman law was extended to people who had no practical tradition of using it. Even in the darkest days of rapid imperial turnover, the quotidian labour of instructing citizens on how they could and could not use the new laws to which they had become subject went on. Over time, the imperial bureaus into which such questions filtered, and by which they were answered, had changed. By the time of Carus, if not before, the old Antonine officials ab epistulis and a libellis had come to be known as magistri epistularum and magistri libellorum, and it was the occupants of these offices who had the most interest in regularising the application of Roman civil law across the empire.
Early in the tetrarchic decade, several of these magistri undertook major efforts to codify the laws that had been issued by the emperors of the previous century or more. The so-called codices Gregorianus and Hermogenianus, named after magistri libellorum of Maximian and Diocletian respectively, collected and organised imperial rescripts in coherent volumes. Their goal was to provide a more comprehensible statement of the standing law of the land than could be gleaned from the random consultation of whatever documents happened to be available in a given time or place. The Gregorian code collected rescripts of pre-tetrarchic emperors that its compiler believed to still have the force of law in 291, the date of its issue. A few years later, the Hermogenian code collected the rescripts of Diocletian, some of which had clearly been drafted by Hermogenianus himself. The dissemination of these codes was quite wide at the time, though only the Gregorian code survives – in seventeen small parchment fragments of the fifth century, recovered from the binding of a medieval book. Instead, for the most part, the rescripts collected in both codes survive in the great sixth-century code of the emperor Justinian that eventually supplanted them and rendered their further preservation otiose.
Change in the legal bureaus of the imperial government was matched by changes in the financial administration. The old separation of aerarium and fiscus had long since ceased to work in any regular fashion, and the capture of Decius’ treasury at Abrittus must have accelerated their dysfunction. Instead, two new bureaus, the res privata and the sacrae largitiones, were created, though whether they developed to their fullest form before the end of the tetrarchy is open to dispute. The sacrae largitiones, whose staff travelled with the imperial courts, administered mining operations throughout the empire and paid soldiers their salaries and any extraordinary donative payments; they levied taxes on precious metals and oversaw the imperial arms and clothing factories. The res privata, into which the old aerarium and fiscus were merged, oversaw imperial estates throughout the empire, managed their leases and collected their rents; they also monitored properties that for one reason or another were forfeited to the emperor.
As we saw at the start of the chapter, the money these officials worked with had undergone thoroughgoing reform alongside the rest of the government’s functions, and this new coinage was struck at a large number of mints. In the high empire the mint at Rome provided the official coinage for the entire empire; supplementary economic needs were met by the local civic coinages that continued to be minted in the eastern provinces right into the last decades of the third century. Thus, imperial coinage minted anywhere but the city of Rome was almost always a sign of usurpation or political crisis. Both those phenomena, of course, multiplied during the third century and many legimitate emperors, not just local usurpers, struck official imperial coinage at regional capitals like Antioch, Siscia or Viminacium as circumstances required. The multiplication of imperial mints thus became somewhat normalised and the distribution of tetrarchic mints is striking: Londinium, Trier, Lugdunum, Arelate, Ticinum, Rome, Carthage, Aquileia, Siscia, Heracleia, Thessalonica, Cyzicus, Nicomedia, Antioch and Alexandria, with Sirmium joining the list in the immediately post-tetrarchic period. Not every one of these mints struck coins in every currency, and not every tetrarchic mint survived the tetrarchic period, but many that were established under Diocletian continued to operate into the fifth century and beyond. They were meant to allow for tighter control of the money supply and the taxation that it was meant to bring in.
It is therefore no surprise that in 296 Diocletian conceived, and by 297 had created, a new census system that was meant to be uniform across the empire. It was to run on a 5-year cycle, with registered persons being assessed for taxation as uniform units of people and land (caput and iugum, poll tax and property tax respectively). Three tax cycles of five years made up an ‘indiction’ of fifteen, and this 15-year tax cycle in turn became a calendrical system that long outlived Roman government in large parts of the Latin west. Taxation had always been one of the most visible ways in which imperial government interacted with its subjects, and it now became less random and far more predictable. It also became more uniform, and the new fiscal regime had very distinct long-term effects in the east and the west that we will consider in chapter 15.
In many ways, the Diocletianic reforms were the culmination of the process of equestrianisation that we have seen evolve in earlier chapters. They involved an aspiration to greater control and oversight, a sense that system was not only desirable but also necessary, and most importantly that system was achievable. We shall see this theme repeated in later tetrarchic history, in fiscal contexts attempting to harmonise taxation and fix prices, and in attempts to impose uniformity on religious practice, if not belief. As with so many of Diocletian’s reforms, we can scour earlier reigns for precedents and find them, particularly in the third century, but those were expedients meant to deal with the exigencies of specific moments. What makes Diocletian’s experiment so surprising, and gives it its enduring interest, is the way it aimed to create a total system, one that could account for most possibilities and allow a response to every contingency to originate close to its source, even if that response had then to make its way up to one of the prefects or an emperor for decision. This aspiration to totality provided real flexibility, but it also brought in an increasingly totalising rhetoric of the state and its power, one that was articulated as a series of good/bad binaries. Heated rhetoric and exclusionary language is intensely characteristic of the fourth-century empire, and it is quite clear that this language often shaped the actions of those who used it. As a cultural propensity it was exaggerated still further after the emperor Constantine’s conversion to Christianity, but its origins lie here, in the far-ranging reforms of Diocletian.
It would take years for the full implications of Diocletian’s many changes to make themselves felt. The immediate effect was not social and cultural but political: it revolutionised tetrarchic activity, allowing for multiple actions on multiple fronts. Thus, having won his victory over Carausius immediately upon his promotion, Constantius began to do what Maximian had earlier failed to do – build a fleet that could spearhead an invasion of Britain and suppress Carausius’s murderer and successor Allectus. That this project took the better part of two years shows how difficult seaborne warfare could be in antiquity, and is also a reminder that Roman control of Britain was always a greater challenge than anticipated.
By 296, Constantius had readied two fleets, one of which he commanded personally, the other sailed under the command of his praetorian prefect Asclepiodotus, who seems to have done for Constantius what Constantius had once done for Maximian. Asclepiodotus’s fleet landed in Hampshire, Constantius’s probably in Essex or Kent, and, while the emperor made for the provincial capital at Londinium, Asclepiodotus brought Allectus to battle and routed his forces, leaving the usurper dead on the battlefield – becoming in the process the last praetorian prefect on record to hold a military command. Constantius and Asclepiodotus were then able to impose the tetrarchic model of provincial division on Britain, creating four provinces where in the Severan system there had been only two. Although we know the names of the four British provinces (Britannia I, Britannia II, Flavia Caesariensis and Maxima Caesariensis), we know neither their boundaries nor their respective locations. Despite that ignorance, we can see the way in which the tetrarchic model responded to circumstance, imposing new ways of doing things as the four rulers extended control and stable governance to new areas. In a nicely symbolic manner, the sole extant inscription of Carausius is a milestone from near Lugovalium (modern Carlisle), which in 297 was turned upside down and re-engraved in the name of Constantius Caesar.
Maximian, the least militarily successful of the tetrarchs, spent most of the years after 293 in Italy and North Africa, and the administrative pairing of those dioceses would become increasingly characteristic of the fourth- and early fifth-century empire. Africa had been relatively neglected by the third-century emperors, but it had also been relatively untouched by military crises. Prior to Caracalla’s enfranchisement of the whole imperial population, Africa had been more than usually a patchwork of communities of differing status, the colonial, municipal and peregrine rubbing up against one another at much closer quarters than was the case in the other urbanised parts of the empire, such as Spain or southern Gaul. During the third century, formerly less privileged or unprivileged communities in Africa raced to bring themselves into line with such long-standing citizen colonies as Carthage and Hippo Regius. By the time of the tetrarchy, urban and enfranchised Africa Proconsularis was increasingly matched by similar landscapes in Mauretania, Numidia and in the fertile enclave of Tripolitania, where the desert came closer to the Mediterranean than it did elsewhere in North Africa. In the west, the Rif remained a virtually insuperable barrier to communication. Tingitania – the tip of modern Morocco – remained much closer to Spain across the straits of Gibraltar than it did to the rest of North Africa, and Diocletian made it a part of the Spanish, not the African, diocese. But the new Diocletianic provinces of the African diocese – from west to east, Mauretania Caesariensis, Mauretania Sitifensis, Numidia, Proconsularis, Byzacena and Tripolitania – would be one of the intellectual and economic powerhouses of the coming century.
As often happened along Rome’s less populous frontiers, however, this thriving Roman civilisation created imitative and competitive impulses along its edges. Just as we have seen increasing sophistication and organisation among the Alamanni, Franci and Goths along the northern frontiers, so along the African pre-desert and desert frontiers, tribes of Mauri that had not been incorporated into the imperial governmental system, but who had all sorts of informal relations with the urban world nearer the coast, become more active in the later third and the fourth centuries. At the same time that Constantius was reconquering Britain for the tetrarchic government, Maximian was to be found in Africa, fighting against unspecified Mauri, perhaps the Quinquegentanae (a tribe that is also attested in an unreliable source as having supported a usurper named Julianus who most likely did not exist). A better source is a fragmentary poem, preserved in a single papyrus, that seems to have been an epic account of Maximian’s Moorish campaigns. However, like most of Rome’s frontier wars, what the poets make into epic was just one of many endlessly repeated police actions. Except in rare circumstances (almost always the imperial army’s distraction in civil war), the tribal polities of the Rhine, Danube and African pre-desert could never muster organised forces on a scale to threaten a Roman campaign army in the field.
The eastern frontier was altogether different. There, the Sasanian kings, when not distracted by internal challengers or attacks on their northeastern satrapies, could frequently best Roman armies in the field, using a combination of the Iranian nobility’s heavily armoured cavalry, massed archers drawn from subject tribes, mounted archers from the steppe and peasant levies from the agricultural regions. The reigns of Probus and Carus, and the early reign of Diocletian, had seen the Persian state deeply divided against itself. As we saw in chapter 11, Shapur I’s eldest son Ohrmazd I had ruled for only one year, and his successor Varahran I for only three. The latter’s son, Varahran II (r. 276–93), was just a child when he acceded to the throne and his reign was never secure. Openly challenged by his cousin Ohrmazd, based in the eastern satrapies, he was also permanently threatened by the resentment of his uncle Narseh. The latter was the last surviving son of old Shapur and a one-time king of Armenia: he rankled at the sight of an untested boy on his father’s throne. For a time, however, enough of the great feudal lords supported the young Varahran to render the question moot, and the power of the Mazdean priesthood in the state is the dominant theme of the 280s. Kardir, Varahran’s mobad mobadan (‘priest of priests’) – modelled on the title shahanshah (‘king of kings’) – was the dominant figure in his regime.
We met Kardir in chapter 11, but we need to consider him more closely here, because the manner in which he persecuted non-Zoroastrian believers in the Persian empire is directly relevant to the course of events in Rome. Temple priests had a tradition of wielding great power in the Near East from very early in Mesopotamian history. The Iranian world had long since adopted these Mesopotamian traditions, vesting many functions of imperial administration in the priesthood since the time of the Achaemenids. Ardashir and Shapur, though the latter was himself relatively tolerant, had carefully nurtured the Mazdean priesthood, which accumulated ever greater wealth and control of royal decision-making as Shapur aged. Kardir, remarkably, was the only non-royal figure to inscribe his image and an account of his deeds on the monuments at Naqsh-e Rustam, where he boasts of having protected Zoroastrians in the lands of the Romans. With Shapur dead, he used his personal dominance at Varahran’s court to impose a policy of persecution on Persia’s Christians and other non-Zoroastrians, most particularly the Manichaeans.
One of the most interesting religious cocktails of later antiquity, Manichaeism was the creation of a single man’s bizarre inspiration. Mani was born in 216 at Mardinu in Mesopotamia, though both of his parents were Persians and probably devotees of the gnostic Mandaean religion. He experienced a revelation at the age of twelve, when a heavenly ‘twin’ appeared, urging him to leave the community into which he was born, and he experienced this same vision again, in 240, when he was twenty-four. From that moment, he devoted himself to spreading the newly revealed truth, continuing to experience revelations through the year, when he began to preach, just at the end of Ardashir’s reign. His first destination was the Kushanshahr, where he encountered the Buddhist art and texts of Gandhara, before returning to Fars and converting Shapur’s own brother Mirshah to his new faith. In the fourth year of Shapur’s reign, Mani had an audience with the shahanshah himself, and thereafter preached widely in the empire’s western provinces. Though the Mazdean priesthood was infuriated at it, Shapur authorised this proselytising and made Mani a member of his court; he would go on to accompany the king in his wars in Rome’s eastern provinces. Mani sent one of his chief followers to the Mesopotamian cities to convert Christians, another into eastern Iran, whence Manichaeism would spread widely in Central Asia. When Shapur died in 273 and his successor Ohrmazd shortly thereafter, the Mazdean priesthood saw its chance for revenge and pounced. While he was in Mesopotamia, Mani was summoned to the court of Varahran and there interrogated by the king himself, who accused him of all sorts of anti-social behaviours. Bound in chains early in 276, he died of hunger and exhaustion after a month of imprisonment.
His teachings, however, would not die out. Their appeal, like that of many other salvific faiths, was the access they gave believers to the possibility of being saved by the secret wisdom that belief imparted. Written mainly in Syriac (some in Parthian), his teachings try to explain the dualist world of ‘two principles’, good and evil, within which man must live. This resembles the Mazdean belief in a perpetual struggle of Ohrmazd and Ahriman, but incorporates elements of pantheism and a very elaborate mythology of the fall of man that resembles Christianity – something that made Mani’s teachings particularly attractive to Christians whose worldview was already prepared for many features of Mani’s cosmology. Ultimately, through the practice of asceticism and the coming of revelation, the fallen human can be reunited with his higher self and rise up into the realm of light. Mani granted that Buddha, Zarathustra and Jesus were all his predecessors, but that he had surpassed them because he wrote his own revelation.
His followers were divided between the ‘elect’ and the ‘hearers’. The elect led much stricter lives than did the hearers: they had to keep pure in thought and word, maintain a strict vegetarianism and do no harm to animals, and abstain completely from sex. This staunch asceticism and self-denial is what gave Manichaean divines their appeal, much like Christian holy men. Hearers, to maintain their position in the faith, were required to fast every Sunday, and undergo a full month of fasting once a year. It was the combination of ostentatious other-worldliness with the promise of salvation that set Mani’s revelation apart from Zoroastrianism, while the secret, gnostic elements in his system, and the hierarchical promise of reaching a place among the elect, was a powerful attraction to many Christians. We will see that Roman emperors saw Mani’s revelation as being just as threatening as did the Mazdean priesthood, but the hostility of neither could suppress the spread and elaboration of Manichaean theology from one end of Eurasia to another, in time reaching as far as China, where many otherwise lost Manichaean texts are now preserved in Chinese translation.
Yet though it helped entrench the role of the Mazdean priesthood in Sasanian political life, Mani’s death did not make Varahran’s hold on his throne materially more stable. In fact, the disarray within the Persian ruling elites only subsided somewhat in the early 290s, when the rival regime of Varahran’s cousin Ohrmazd was finally suppressed. Varahran now styled himself shah of Seistan along with his other titles, indicating the submission of the eastern part of the empire. He died, naturally so far as we know, in 293. If Varahran’s rival Ohrmazd had in fact survived his defeat, it was now that he ruled, however briefly, in his own right, but only in the central part of the empire. Meanwhile, the claims of Varahran’s very young son were pressed by a faction of the Persian nobility, led by a certain Vahunam. The boy’s accession as Varahran III, however, angered as much of the nobility as it pleased. It may be that they foresaw the continued dominance of Kardir and his priests, or it may be that the essential function of the shahanshah – leadership in the martial virtues which defined his ability to rule – could simply not be exercised by so young a boy. In consequence, they exercised their prerogative of choosing among Sasanian claimants and looked around for an alternative.
In another of the remarkable inscriptions to survive from ancient Iran – this one from Paikuli – the monarch who overthrew the young Varahran III explains how he came to power. That monarch was none other than Narseh, the former king of Armenia and the last surviving son of Shapur the Great, by now a man in late middle age. Narseh’s inscription is not the triumphal account that we found in Shapur’s Naqsh-e Rustam text. It is instead an explanation for his actions in seizing the throne, or rather an extended apologia assuring his audience that his act of usurpation was in fact entirely legitimate. The location of the inscription speaks almost as loudly as do its words, for Paikuli is in Asuristan, the Parthian and Persian province of the lower Euphrates and the heart of urbanised Mesopotamia, where the vast majority of the shahanshah’s subjects lived. (Although etymologically derived from ‘Assyria’, Asuristan was the ancient Babylonia far to the south of the old Assyrian heartland, which was known in late antiquity as Adiabene.) Although written in Middle Persian, rather than Parthian or Greek, the inscription was clearly meant to justify Narseh’s reign to the widest possible audience.
The text is quite fragmentary, and has been reconstructed in various ways that pre-occupy and divide specialists, but the gist of the story is clear. Narseh records how a delegation of the nobility came to him in Armenia and begged him to take the throne that was his due. Those he lists are an interesting array, for they clearly represent the fringes of the Sasanian realm, rather than its centre: we find among them Gilan on the south-western coast of the Caspian and all of the easternmost regions that had until recently supported Ohrmazd against Varahran II. The great Parthian hereditary houses go unmentioned, as do the core regions of Fars, Khuzistan and Asuristan. Even behind the studied ambiguity, we can presume that the child Varahran III was the candidate of a powerful court faction, while Narseh offered an appealing alternative for men who had been excluded from power under Varahran II.
It seems that Narseh marched out of Armenia into Iran and then led an army into Asuristan. There, at Paikuli, where he later set up his inscription, many of the great lords ‘both Persian and Parthian’ came over to his side, including members of the Varez, Karen and Suren clans, as did the mobad mobadan Kardir. Varahran III’s support melted away, Narseh marched on Ctesiphon, where Vahunam was defeated in battle and Varahran was himself killed. Narseh, aged though he was, had proved vigorous. He seems to have received the submission of all the provinces and satrapies of the empire – the inscription mentions regions as distant as the Kushanshahr, Turan and Khwarezm on the Amu Darya, as well as the Lakhmids of Arabia – but in truth the internal history of the Sasanian empire is a blank for most of Narseh’s reign. Not so his frontier with Rome, however, for two years after his victory at Ctesiphon and secure now on his throne, he launched an attack on the empire.
Narseh began by invading Armenia in 295 or early 296, driving out the Arsacid Tiridates whom Diocletian had installed half a decade before. He then marched south into Roman Mesopotamia. There, late in 296 or early in 297, he was met by a tetrarchic army under the direct command of the caesar Galerius, who had been in the region since 295. Details are scant, but it is clear that Galerius was completely routed and Narseh took control of the lands between the upper Tigris and Euphrates that had been in Roman hands for decades. Galerius’s defeat was a bad one, a point Diocletian hastened to drive home: meeting his defeated caesar upon his return to Antioch, he staged a procession in front of his own and Galerius’s campaign armies, riding in the imperial coach while Galerius, clad in his full imperial regalia, trod the course before him on foot. It is a testimony to Diocletian’s authority over both his soldiers and his caesars that a man as proud and dangerous as Galerius would later prove permitted himself to suffer such treatment and that his soldiers would tolerate it. But Diocletian attached no blame to the army for its defeat – the emperors were universal, universally powerful, and if the young Herculius had failed the elder Jovius, then in good Roman fashion he would receive the punishment his divine father meted out.
Galerius now conceived a well-concealed hatred for Diocletian, and would in time repay him for his humiliation. But in the short term, the message of unity was clear and the defeated caesar redoubled his efforts, collecting another army on the Danube in 297 and preparing to launch a new campaign against Narseh in the following year. His fellow tetrarchs were similarly busy. As we mentioned above, part of Diocletian’s reform plans included the introduction of a new system of taxation with uniform poll and land taxes. To collect them first required a new census, which was highly disruptive of the old ways of doing things in many parts of the empire. One such place was Egypt, where the new census procedures led directly to usurpation. One L. Domitius Domitianus claimed the purple between August and December 297, minting both local Alexandrian tetradrachms and folles on the Diocletianic imperial model. Diocletian led his army against the usurper and, although Domitianus himself was dead by the end of 297, one of his correctores, Aurelius Achilleus, carried on the revolt in Alexandria. He must have done so in Domitianus’s name, for he minted no coins of his own, but he held out against a siege in Alexandria until late spring 298.
While Diocletian campaigned in Egypt in 298, Galerius resumed the Persian war. He invaded Armenia and then carried on down the course of the Tigris through Media and Adiabene into Asuristan. There was heavy fighting throughout the campaign, and at least two pitched battles against Narseh in which the Romans were victorious. In the second, Galerius actually captured the shahanshah’s camp, taking his harem and some of his family, along with a great deal of treasure. The king’s wife was escorted back to Daphne, in the Antiochene suburbs, and treated with the dignity due to a queen. Narseh retreated to Ctesiphon to rally support and regroup, and it is a measure of his surprising authority that even so major a defeat did not lead to a coup against him.
Rather than continue to fight, Narseh opened negotiations with the Romans, wintering in Mesopotamia in 298–9. In the spring of 299, he was met in Adiabene by Diocletian himself. Peace talks were conducted on behalf of the Romans by Sicorius Probus, who occupied a new palatine office created by Diocletian, that of magister memoriae. This magister supervised a bureau (or scrinium) of scribes who handled the responses to any requests and petitions in which the emperor took a specific personal interest, while other magistri supervised the more routine business of government. On the Persian side, the negotiators were Narseh’s close adviser Apharban, and Hargbed, one of the great Iranian nobles. They argued that the Roman and Persian empires were the two lamps of humankind: ‘as with two eyes, each should be adorned by the brightness of the other and not forever be seeking the other’s destruction’.
As a statement of political theory that might have been accepted as true, but the treaty that Sicorius negotiated was heavily one-sided. Narseh received back his wife and family, but few other concessions. Rome got back its long-standing prerogative of appointing the Armenian king, this time yet another Tiridates. The treaty also required Narseh to surrender a great deal of territory that had been contested between Persia and Rome since the Severan age – not just the Mesopotamian steppe between the Tigris and the Euphrates, but also the mountain regions of Sophanene, Arzanene, Corduene, Zabdicene and Ingilene that commanded access into Armenia. Nisibis was to be the sole point of commercial contact for traders between the two empires, and its prosperity in the following half century is well known. Not all of these provinces would remain in Roman hands for long, but the Mesopotamian steppe became the focus of Roman military defence and was transformed into a chain of fortress cities, from Edessa and Carrhae just beyond the Euphrates, via Resaina on the Khabur, Amida and Nisibis, to Singara. For much of the fourth century, confrontations between Romans and Persians were concentrated on this broad stretch of strategic land, Rome relying upon the great cities to stop Persian armies and bog them down in debilitating sieges. Those confrontations lay in the future: Galerius’s victory led to a forty-year peace between the two great powers of the world.
Further to the south, in the Syrian desert, Diocletian created an elaborate fortification system known as the Strata Diocletiana, along a military road running from Resafa on the Euphrates down past Palmyra (the latter a shadow of its former self) and into the northern Arabian desert. The earliest Arabic inscription yet known is from the city of Bostra in north-western Arabia, along the Strata Diocletiana, and is from about this time. Entirely by chance, it reveals another consequence of the victory of Galerius over Narseh: the Lakhmid king of Hira, Imru’ulqais, changed sides. Having long served Shapur and his successors, he now joined the Romans and moved his seat of power westwards, to Bostra. Hira itself remained a Persian client under other Lakhmids, but the habitual allegiance of some Arabs to Rome and some to Persia would become one of the main political dynamics of later antiquity.
At the end of the third century, and in the aftermath of Galerius’s momentous victory, it must have seemed as if the tetrarchic experiment had fulfilled most, if not all, of its explicit goals: although Diocletian’s ambitions had not gone unchallenged, his grand project to harmonise and make uniform the administration of empire looked not just feasible, but very nearly accomplished. And yet it is clear that Diocletian did not think he had gone far enough. On the contrary, still more centralising regulations, for the economy and for the state religion, issued from the court right at the turn of the century.
The extent of Diocletian’s ambitious reshaping of Roman government becomes clearer when one looks at four measures he took between 301 and 303, each written in the bombastic style that the professional jurists at his court were now coming to perfect. In September 301, he retariffed the coinage, as is demonstrated by a fragmentary inscription from Aphrodisias in Caria, an important and well-excavated city in Asia Minor. The edict that enacts this measure states that the imperial coinage would, from 1 September of the year, be worth double its face value; consequently, debts to the imperial treasury contracted prior to that date could be repaid at the old prices, while those contracted since would have to be paid at the new tariff. So far, so good, and presumably a measure meant to combat the inflation that the economy seems to have been suffering at this time. The second part of Diocletian’s plan, however, was both more original and more problematic. In late November or early December 301, he issued a far-reaching measure that we call the Edict on Maximum Prices, which happens to be the best attested Latin document from the Greek east. The law it preserves was an attempt – an almost demented attempt, as it proved – to revolutionise the imperial economy. With the hyperbole that is the standard currency of late Roman edicts, the Edict on Maximum Prices sets out to regulate the price of an extraordinary range of goods of all sorts, and at different and precise stages of manufacture. It does so because the greed of nefarious individuals had damaged the very people whose job it was to protect them, the servants of the state and the soldiers. That is, we have here yet another tetrarchic attempt at regularising and regulating the affairs of the empire in a single, systematic way across the length and breadth of an amazingly diverse polity. Some such attempts succeeded, others did not. The impact of this one was catastrophic.
Local conditions had long made some goods cheaper in one part of the empire than in another. This was inevitable, given how much the costs of transport, particularly transport by land, added to the end value of goods. That meant that many things were intrinsically more valuable in one place than another, with the exception of highly perishable items such as fresh foodstuffs, the range of which was limited by the threat of spoilage in those pre-refrigeration days. Opportunities for profitable arbitrage were thus ubiquitous. Demanding that there be a universal maximum price for every item in the empire, as Diocletian’s edict did, was a practical impossibility, not least because where we can calculate actual prices and compare them to the prices set out in the edict – which is to say, mainly in Egypt – the prices in the edict are much lower. We can of course explain this in terms of the government trying to save itself money on the many things it had constantly to buy for its own use, though that is perhaps to give too much credit for rational economic thinking to a regime that so prized uniformity for its own sake. And, whatever the motive, the attempt ended in failure. The price edict survives in numerous epigraphic copies, some complete, some fragmentary, but we have only one literary source – the Christian Lactantius, whose loathing of Diocletian was total – that mentions it at all. For what his hostile testimony is worth, Lactantius claims that the edict caused violence over ‘small and cheap things’ and led to a continued rise in prices. We may well believe that it did worsen inflation and spark widespread rioting but, despite Lactantius, it is hard to know how widely any attempt was made to enforce the edict. In theory, tetrarchic laws were issued in the name of all four emperors and valid across the whole empire; in practice they depended upon the willingness of the different members of the imperial college to endorse one another’s measures, and even more on provincial governors’ willingness to enforce them. That means that, despite its breadth of dissemination, the Edict on Maximum Prices may well have been a dead letter the moment it was issued, and there is some ambiguous evidence that the edict was repealed without fanfare within the year. But the damage it had done, and the chaos it had introduced, was lasting. The imperial economy, and indeed the value of the imperial currency itself, took several decades to stabilise.
The Edict on Maximum Prices is revealing of the tetrarchy’s aspirations to total control, but another such measure was still more significant. Diocletian, famously, launched the last empire-wide persecution of Christians, known fittingly enough as the Great Persecution. Ever since Gallienus offered toleration to Christians after his father’s defeat in Persia, adherence to Christian beliefs had undergone a very dramatic expansion. Numbers are impossible to come by, though they are bandied about freely by scholars, and the evidence is fragile enough to make controversy endless and irresolvable. Probably, Christians represented a majority population in parts of the Greek-speaking empire and formed substantial minorities in a few urbanised regions of the Latin west. By the early 300s, there is evidence of communities self-segregating in the east, with some villages entirely Christian, others entirely pagan, in the manner of modern American suburbanites self-segregating by race and party political taste.
Within Christian communities, a clear hierarchy of clerical grades was beginning to emerge, and the bishops who led each Christian community corresponded widely with one another across the empire. Shortly before the year 300, the first council of bishops whose rulings (‘canons’) are preserved met at Elvira in southern Spain, passing judgement on all sorts of issues that arose when Christians needed to get along in a world of many other religions. Among the empire’s educated population, Christianity was just one option among many, and was certainly not the religion of the majority even though monotheism of one sort or another was now the elite norm. That said, the evidence does not support the once commonly held belief that Christianity was predominantly confined to the lower echelons of society. Indeed, Christianity was widely enough practised that both the army and the civil service included numbers of open Christians, and it is reported that Diocletian’s wife and daughter were among their sympathisers. It was precisely that visibility that incited persecution.
Beginning with an incident at court in 299, Diocletian enacted increasingly harsh penalties against Christians in government, spurred on by his caesar Galerius. These ended in an edict imposing universal sacrifice to the gods of the state and specifically outlawing Christian worship. The actual causes of the anti-Christian hysteria – for that is clearly what it was – at the courts of Antioch and Nicomedia are hard to disentangle, given the weight of later Christian apologetic and the wholesale destruction of anti-Christian books from the reign of Constantine onwards. Nevertheless, we know that Diocletian’s court, like that of many earlier emperors, was a place of learning in which Christian and non-Christian intellectuals mingled and taught, and in which the patronage of such men was a signal part of the emperor’s duties. Elsewhere in the empire, at Antioch, Alexandria, Caesarea and Rome, imperial officials took part in the soirées of celebrity thinkers, enjoying their debates as entertainment suitable to men of their social status. Philosophical and religious debate was at one level cultured entertainment, but at another it was deadly serious, particularly where the philosophical and the religious overlapped. When sincere beliefs are also a cause for rivalry among intellectuals, the poison native to academic debate could seep into more perilous contexts, accidentally but suddenly turning political. Ever since the Julio-Claudian and Flavian periods, when Stoic philosophy and opposition to autocracy were equated, philosophers and theologians could, occasionally, influence government or draw down its wrath.
The tetrarchic persecutions may in part have been triggered by just this sort of politically charged intellectual debate. The basic story of how the persecution began is simple. During a ritual haruspicy in the praetorium at Antioch in 299, Christian officials had crossed themselves and the haruspex proved unable to take the auspices from the entrails of the sacrifical animal. Galerius himself was present at this event, after which an oracle of Apollo at Daphne issued the statement that the ‘righteous’, or Christians, were forcing him to compose false oracles. Galerius was enraged, and he was already an enemy of Christians. Diocletian’s immediate response was to order the flogging of any courtier who refused to sacrifice to the gods. Early in 300, the order was extended to take in the whole of the eastern army. Other anti-Christian oracles – Zeus of Dodona and several theurgic oracles of Hecate – also began to circulate, playing upon the fears of an emperor who believed, like everyone, that the safety of the state was in the hands of the divine. Galerius continued to expend considerable energy on convincing Diocletian to mount a full-scale persecution, but it was not until 303, when the oracle of Apollo at Didyma issued another anti-Christian pronouncement, that Diocletian was convinced. That February, he ordered Christians to ‘return to the institutions of their ancestors’. By placing braziers and altars in all courts of law, Diocletian hoped to ensure that all inhabitants of the empire, Christians included, participated in sacrifices to the gods by burning incense on their altars. Those who did not would be punished.
That raw chronology does not get to the intellectual background of events. It is clear that during the 290s, Christian apologists were deeply exercised in defending themselves against philosophical attacks on their beliefs, the tenor of which is recorded in some sources. Eusebius of Caesarea, in particular, quotes hostile pagan writers who asked, given Christian atheism and apostasy from the gods of the state and their ancestors, ‘to what kind of punishments would they not be justly subjected’? The author of the lines quoted by Eusebius is certainly the Neoplatonist Porphyry, a man who positioned himself among Greek intellectuals as the true heir to the philosopher Plotinus, universally acknowledged as the most brilliant Greek thinker of the third century. Plotinus was a Neoplatonist, and late Platonic philosophy, unlike most post-Classical philosophical schools, had always had a dogmatic side that could conceive of there being just one single path to truth. Porphyry’s attacks, both on rival philosophers claiming similar pedigree and on Christians, were widely read and seem to have circulated among the more cultured officials of the eastern civil service, some of whom became strong supporters of persecution. We know that, before the outbreak of the Great Persecution of 303, the imperial official Sossianus Hierocles (his actual post at the time is unclear) and a philosopher who may have been Porphyry himself argued the case against the Christians before Diocletian.
Galerius, for his part, had every reason to stoke Diocletian’s fears and worries, for Constantius, his fellow caesar on the other side of the empire, was if not himself a Christian then at least quite sympathetic to their presence in his court and family. Constantius had a daughter named Anastasia, which is diagnostically a Christian name, and if we could be certain that the name was given to her at birth, we would have much greater certainty about Constantius’s own predilections. Regardless, Constantius implemented only the mildest clauses of the persecuting edict in his territory. Attacking Christianity might usefully undermine Constantius, which was very much in Galerius’s interest, because while he and Diocletian had no sons, both Maximian and Constantius did. These, respectively named Maxentius and Constantinus (though we will call him Constantine henceforth, as is conventional), were being groomed for the succession, so that if one of the augusti were to die, Constantius would, as the senior caesar, succeed him and then promote one of the two heirs apparent – presumably his own son Constantine – to the vacant caesarship, leaving Galerius in that lower rank himself. But if Christians were ejected from public roles then both Constantius and his son would be out of contention, opening the way for Galerius. Fears for the safety of the state, philosophical disagreements, the petty rivalry of intellectuals and the deadly serious politics of imperial succession thus prepared a toxic miasma of hatred, misunderstanding and wilful misrepresentation.
The edict that resulted – promulgated at Nicomedia at the end of February 303, in much of the east in March, and by May in Africa and elsewhere in the west – was nothing if not damning. It ordered Christian churches to be pulled down and Christian scriptures burnt, condemned Christian imperial freedmen to slavery, and stripped Christian civil servants of their rank, thus opening them up to consequences that would be just as dire. In legal matters, Christians lost any recourse in actions taken against them, nor could they initiate prosecutions against their assailants. In sum, the edict was designed to make being a Christian intolerable, rendering Christians defenceless against charges and inviting neighbour to prey upon neighbour, in the way that delatory regimes have always done. Denunciation became profitable, and that was one point of the decree. Indeed, with a sort of malign glee, the initial edict was promulgated on the feast of the Terminalia, a festival celebrating the boundaries that kept the Roman state safe, that was traditionally marked by the sacrifice of a black lamb – a macabre joke at the expense of a religion whose central figure was often represented in the form of a shepherd or indeed a lamb.
As always, the success of this edict depended upon its enforcement by imperial officials, which varied widely across the empire. The caesar Constantius did the bare minimum possible to keep up the appearance of participating in his senior colleague’s initiative, tearing down churches, but burning no scriptures and enforcing no other bans. In Maximian’s territory, only in North Africa did the efforts of local officials ensure serious attention to enforcement. In the east, by contrast, enforcement was vigorous and violent, all the more so after Christians tore down the edict in Nicomedia as soon as it was posted and then set fire to the imperial residence. That arson prompted a second edict, ordering the arrest of the church’s leaders, but this was only promulgated in the east – it was entirely ignored in the west.
There were victims everywhere, of course, and the few genuine martyr acts that survive illustrate both enthusiastic and utterly indifferent prosecution by imperial administrators. Some municipal officials made sure that Christians in their midst would be forced to declare themselves; others did their best to let Christians get away with the form rather than the substance of compliance, for instance by accepting for the pyre any book handed them by a known Christian without verifying its content. Where officials cared to enforce the law strictly, there were many deaths. Palestine and Syria, where Galerius was nearby and men loyal to him in charge, saw something of a bloodbath, enough so that parts of Syria and Commagene may have erupted in open rebellion. North Africa, where the governor C. Annius Anullinus harboured anti-Christian sentiments as virulent as those of Galerius, suffered badly and the behaviour of North African Christians in the face of persecution would fuel a schism in their church that dragged on for more than a century. In Italy, by contrast, the deadlier side of the edict could be evaded, while in Britain, Gaul and much of Spain, the loss of corporate property certainly exceeded that of life and limb.
Although undoubtedly the most numerous, Christians were not the only group targeted for persecution under the tetrarchy: on 1 March 302, Diocletian issued an edict against the Manichees, a measure that was not merely about conformity, but about xenophobia as well. This law is preserved intact, with all its rhetorical preamble, whereas most laws of this period survive only in edited forms in later codes. A work known since the sixteenth century as the Collatio legum Mosaicarum et Romanarum (‘Comparison of Roman and Jewish Law’) but properly entitled the Lex dei quam praecepit dominus ad Moysen (‘The Law of God Which the Lord Taught to Moses’) is an early fourth-century Jewish text that juxtaposes passages from the Torah with Roman law from second- and third-century juristic texts and imperial edicts from the Gregorian code. One of its sections, on astrologers, magicians and Manichees, preserves the Diocletianic edict condemning Manichaeism.
Mani’s faith was a product of the cultural diversity of the eastern Persian world, and of the support that a prince like Shapur had been willing to give it. Narseh, when he became shahanshah, ruled in a similarly tolerant vein when it came to religious matters, and allowed followers of Mani to practise their religion freely after the years of persecution by Kardir and the Mazdean priesthood. This meant that, in Roman minds, Manichaeism seemed not just foreign, but specifically Persian – an enemy religion in a literal sense – and Mani’s followers could be seen as fifth columnists of the Persian kings and their ‘depraved’ religion a threat to the Roman state. In Diocletian’s edict, that connection is made explicit: Manichaeism is referred to as a new religion, ‘recently sprung up from the Persian people, our enemy’, bringing the ‘accursed laws and customs of the Persians’ to innocent Romans. The governor Julianus, to whom the law is addressed, should burn Manichaean leaders alive along with their holy books, their followers should suffer capital punishment and the confiscation of their estates, while Manichees in imperial service are condemned to the Phaenensian or Proconessan mines, the equivalent of a death sentence. So far as we know, this edict was never revoked, and occasional persecution of Manichees continued throughout the fourth century, under both pagan and Christian emperors.
Although many Romans continued to flirt with the religion, whose dualism and salvific promise was as mystical and attractive as Christianity, imperial law never ceased to view Manichees, like practitioners of black magic, as an inherent threat to the state.
It is tempting for the modern historian to downplay the religious side of the tetrarchic persecutions and see in it a cynical political manoeuvre on the part of Galerius, or a passing storm exaggerated in retrospect by Christian writers. But that would be wrong. Galerius really did hate Christianity, while Diocletian, more than many emperors, had a heavy ideological investment in the traditional religion of the state. Yet both in the end reluctantly ceased persecution, having realised that Christianity really was too widespread, and Christians too widely represented among the powerful, to eliminate. It is not that Galerius and Diocletian were on the wrong side of history, though some confessional historians would to this day argue that point. Rather, Christianity was simply too large a phenomenon to submit to their totalising mania. Late in 303, amnesty was declared throughout the empire, no doubt to the displeasure of the more committed persecutors. But there was a festival to celebrate: blood would not be shed during the vicennalia of the augusti.
Late in 303, with the persecution of Christianity in abeyance across the empire, Diocletian and his court travelled to the territory of his fellow augustus Maximian. The two then progressed together to the city of Rome and on 20 November 303 celebrated their vicennalia, the twentieth anniversary of their joint reign: in a rewriting of past history typical of the later empire, Maximian and Diocletian shared the same regnal reckoning, despite Maximian’s actually having become augustus a full year after Diocletian. It was meant to be a happy occasion, and no doubt for many there it was: the tetrarchs had imposed a stability on the empire such as had not been seen for two generations. What went on in Rome between the two augusti is unclear, but decisions were certainly made for the future. Although there have been many modern discussions about a tetrarchic ‘theory of government’, all rely on inference from later events that may have owed more to chance than design. What is clear is that for some time, perhaps since 293, there was an understanding among the tetrarchs that at some point the augusti would retire, to be succeeded as such by the caesars, and that two new caesars would then be appointed.
We cannot know if a retirement date had long been anticipated, or whether in 302 or 303 Diocletian decided that the time had come and that Maximian now needed to be told as much. But that there had been retirement plans in the works for some time is made clear by the massive palaces that the tetrarchs were building. The best preserved and understood of these is Diocletian’s at Split in Dalmatia, which must have been designed as a retirement residence, given its distance from the main centres of imperial government. Galerius’s palace at Romuliana, now Gamzigrad in Serbia, was on a similar scale. A third tetrarchic palace has been excavated in Spain under the high-speed rail station in modern Córdoba (Corduba), which destroyed most of the remains, although its speculative attribution to Maximian is dubious. It had thus long been intended that a new college of emperors would continue to run the empire from Antioch, Nicomedia, Sirmium, Mediolanum and Trier, while their former colleagues lived out an honourable but discreet retirement far enough away to cause no trouble. Diocletian’s plan for systematic retirements would guard against the unexpected and disruptive death of an augustus, but it implies no formal theory of government or, as is sometimes claimed, a hostility to dynastic succession.
Diocletian was too smart a man to believe that competent male heirs, if such existed, could be passed over; they would surely try to seize what they were not given. The plan was thus that Maximian’s adult son Maxentius and Constantius’s adult son Constantine would become the new caesars once Diocletian and Maximian stepped down as augusti and Constantius and Galerius became the senior pair, so both Constantine and Maxentius spent much of the 290s at Diocletian’s court, being groomed for power. But when the imperial retirements were announced rather more than a year after the vicennalia, on 1 May 305, the succession plan had changed – wholly to the benefit of Galerius.
Events are obscure, but during the celebrations of the vicennalia at Rome, Diocletian fell badly ill. He left Rome on 20 December 303, his second and last visit to the eternal city. Marching back to the Balkans via Ravenna, he met Galerius, who had not gone to Rome for the ceremonies. In the spring, Diocletian was evidently well enough to take joint command of a punitive war against the Carpi along the lower Danube. Maximian likewise made a journey to the Balkans a few months later, after Diocletian had departed for Asia, and met Galerius at Sirmium. The two men argued violently, which may suggest that Maximian had learned of the plan to alter the succession and disliked it. Constantius, throughout all this, had been studiously silent, which bespeaks either deep political wisdom or later whitewashing by his victorious son. Regardless, the western caesar would be present beside his divine father Maximian on the fateful day in May 305. The latter had returned to Rome after his angry encounter with Galerius, and there celebrated the secular games in 304. Diocletian was in Nicomedia by August of the same year but, those facts apart, little is known of this period – save that a fourth persecuting edict was issued in the east, ordering all subjects to sacrifice, in a return to the policy that had been abandoned for the sake of the vicennalia.
The renewed persecution was probably the work of Galerius, who gained an ascendancy over Diocletian after the latter’s illness. He stood beside Diocletian on 1 May 305 outside Nicomedia, where the army had been gathered to hear a very special announcement. On a dais, in front of a column honouring Jupiter, Diocletian announced his retirement with tears in his eyes, pleading age and exhaustion. He would leave the direction of affairs to others, and appoint new caesars. Constantius’s son Constantine was there, too, and the gathered crowd must have expected his proclamation to be the next step. Instead, Diocletian proclaimed as caesars Severus and Maximinus, relatives and cronies of Galerius. Maximinus, likewise present at Nicomedia, was brought forward and took off his civilian cloak before Diocletian garbed him in his own purple mantle, effectively resigning the imperial office and handing it to another. In Mediolanum, on the same day, a parallel scene was playing out, in which Maximian and Constantius garbed Severus in the purple and Maximian stepped down into retirement. Nothing like this had happened before in Roman history, a voluntary resignation and peaceful handover of power between rulers. It is impossible to exaggerate the strangeness of it all.