`There have never been such earthquakes and plagues, or tyrants and emperors with such unexpected careers, which were rarely if ever recorded before. Some of these men ruled for quite a long time, others held only transient power; some hardly reached the title and fleeting honour before they were deposed. In a period of sixty years the Roman empire was shared by more rulers than the years warranted.' - Herodian, middle of the third century.'
In the half-century between the murder of Alexander Severus and Diocletian's victory over Carinus, over sixty men claimed imperial power. A precise figure is impossible, since in some cases it is unclear if a rebel leader actually aimed at the throne, or even existed at all. In 2004 a coin was found in Oxfordshire bearing the image and name of the extremely short-lived usurper Domitianus. Briefly mentioned in the literary sources, he was said to have rebelled against Aurelian in 270 or 271. Only a single coin had previously been found with his name, and this had widely been dismissed as a forgery. Now it is clear that he made a bid for imperial power, probably somewhere in the western provinces, and lasted long enough to have coins minted in his name.'
Domitianus was one of many would-be emperors who only survived for a matter of weeks. Gallienus lasted longer than anyone else, if his seven years ruling jointly with his father are added to his eight years of sole power. Otherwise, Postumus' nine years represents the longest reign of an emperor without a colleague, albeit over only part of the empire. Both Gallienus and Postumus were murdered by conspiracies involving their own senior officers and staff. This was by far the commonest fate of enduring emperors and short-lived usurpers alike.
The years from 235-285 are often characterised as a time of anarchy and the whole era is dubbed the `Third Century Crisis'. Certainly, the rapid turnover of emperors was in marked contrast to the first and second centuries - between 31 Bc and AD i8o there were only sixteen emperors (seventeen, if Lucius Verus is included as Marcus Aurelius' colleague). Yet this in turn differs from the fifty-five years following the death of Marcus Aurelius. There were ten emperors (eleven, if Geta is included) from 180 to 235 - and the number could actually be increased if the handful of short-lived usurpers who rose against the later Severans is included. Things were clearly far worse in the middle of the third century, but it was mainly a question of degree. Similarly, although Diocletian ruled with colleagues for two decades, several usurpers emerged during these years, and for a while Britain and parts of Gaul broke away in a repeat of the earlier `Gallic Empire'. Civil war remained a common event, even if it was never again quite so frequent.
The `Third Century Crisis' used to be painted in the bleakest terms. It was a time when the Romans suffered defeat after defeat at the hands of new, more powerful foreign enemies. Persian armies took Antioch, fleets of Gothic pirates plundered Greece and Asia Minor, and other barbarians burst across the frontiers into Gaul, Italy and Spain. For years on end much of the east was controlled by the monarchs of Palmyra, and the western provinces broke away under the rule of a succession of their own emperors. There were also outbreaks of plague over wide areas, which may well have rivalled the Antonine epidemic in their virulence and the number of victims. At the same time the economy collapsed, as successive emperors massively devalued the coinage to pay for their wars. Society also changed, some of the poorer citizens in rural areas being reduced to little more than serfs. All of this was accompanied by a crisis of belief, as everywhere people abandoned the old faiths for new religions and wild superstitions.
However, academic fashions change and today few would accept such a grim picture, for the evidence can be understood in other ways. Some would argue that crisis is simply the wrong word, since it is clear that there was never any question about the survival of the empire in some shape or form. Yet no one doubts that this was a period of profound change. Diocletian's empire looked and functioned very differently to the one familiar to Marcus Aurelius. Before moving on to look at the system he created, it is worth pausing to consider some of the wider changes that seem to have occurred.3
Render Unto Caesar: Money and the Economy
Emperors were expected to spend lavishly. Almost the first act of any new ruler was to promise a cash gift (or donative) to the army. This was as true for Marcus Aurelius as it was for any usurper seeking power in later years. Even in normal circumstances maintaining the army was the greatest single expense in the imperial budget. Frequent warfare - most especially campaigns that were fought inside the empire and so produced little plunder - greatly added to its overall cost. On the whole, the troops remained loyal to any established imperial family, as long as their rule seemed reasonably effective and they did not suffer too many serious defeats. It was much harder for new emperors to gain this level of loyalty, which only added to their urgent need to treat the soldiers generously. Severus raised army pay for the first time in over a century, and Caracalla increased this again just a few years later. Macrinus struggled to cope with the cost of this and his clumsy handling of the situation prompted his rapid fall.
The emperor was the greatest landowner in the world, the imperial estates including vast areas of farmland, as well as such things as mines, throughout the provinces. Whatever its origins, this quickly ceased to be a personal possession and passed from emperor to emperor regardless of whether there was any family connection. Property confiscated from executed opponents was often added to this, but it was obviously dangerous for any ruler to kill too many rich and influential men. New taxes could be introduced - Dio believed that Caracalla's grant of citizenship was essentially a money-raising venture - or old ones reinterpreted in the government's favour. This, too, was dangerous. The revolt of the Gordians against Maximinus was provoked by resentment at the exactions of the official collecting taxes in Africa. On the whole, the level of taxation throughout the provinces remained remarkably static.4
Most of the methods of raising extra revenue quickly were dangerous for any emperor, and especially one who had recently seized power. Yet the need was always there to spend, and it was not just a question of the army. Maintaining even the small-scale administration of the empire cost money and the number of imperial bureaucrats grew steadily during this period. There were also many other things an emperor was expected to pay for. At Rome itself there were the distributions of free and subsidised food, public entertainments, as well as the maintenance of public buildings and the construction of new ones. Emperors were expected to be generous to communities and individuals. If they were perceived as mean and grasping, then that was an invitation for more lavish challengers to emerge. To make matters worse, a good number of third-century emperors could not even call upon the entire imperial resources, since parts of the empire were often under the control of rivals. One of the reasons for Gallienus' survival for so long was that he kept control of the prosperous North African provinces with their substantial imperial estates. Similarly, it was not until after his death that Zenobia took over Egypt, whose grain continued to supply a good part of Rome's needs.
Financial problems appear time and again in the record of the third century. Emperors controlled the minting of gold and silver coinage and there was a constant temptation to make their resources go further by debasing the currency. The change is most striking in the silver coinage. In Trajan's day, a silver denarius contained just over 9o per cent silver (the denarius was the penny of the Authorised Bible, hence the abbreviation before decimalisation of pence as d). Under Marcus Aurelius the percentage of silver as opposed to base metals dropped below 75 per cent at a time when the empire was ravaged by warfare and disease. Septimius Severus increased army pay and let the silver content of his coins drop to 50 per cent. Caracalla introduced a new silver coin, known as the antoninianus and probably worth 2 denarii, although its weight was only equivalent to one and half of these coins. This was abandoned under Elagabalus, but reappeared in the joint reign of Balbinus, Pupienus and Gordian III. By this time the silver content had fallen close to 40 per cent. Around the middle of the century the decline rapidly accelerated and was as little as 3.5-4 per cent silver by the time Aurelian came to power.5
The impact of this massive debasement is very hard to judge since, as is usual with the Roman economy, we lack any useful statistics. It is clear that prices by the end of the third century were massively higher than they had been at the end of the second - in some cases by several hundred per cent. There had never been inflation on such a scale in Rome's earlier history. However, the bulk of our evidence for day-to-day rises in costs comes from Egypt, which uniquely had its own system of coinage until Aurelian's reign, when it was brought into line with the rest of the empire.
Prices rose, perhaps suddenly. Farmers, both large- and small-scale, may have been insulated against the consequences of this, since they still had produce to sell or trade. Those who like to see barter as a major part of the rural economy tend to claim that large areas of the empire were unaffected by changes to the currency. Similarly, people involved in manufacture had products to sell, assuming that they, like the farmers, still found a market. Change probably occurred slowly, as the increased number of coins would have only started to circulate gradually, and perhaps people had time to adapt. Yet it was not the only factor. Wars were costly, perhaps especially civil wars, which involved the movements of armies through normally peaceful areas at a time when military discipline tended to slacken. Many suffered; a few probably profited. The population may have been declining - certainly, there must have been substantial local drops in numbers as a result of outbreaks of plague, even if over the long term there was little overall change.'
Ultimately, we simply do not know. It is worth bearing in mind that if we had the same amount of evidence for the twentieth century as we have for the third, then we would not have any real idea of the scale of the Great Depression or the impact of two world wars. For instance, Japan's and Germany's growing prosperity would doubtless be seen as inexorable and unbroken in the course of the century. Any talk in literary fragments of the devastation caused by war would doubtless be dismissed as wild exaggeration. It is clear that there were plenty of very wealthy people in the Roman world at the opening of the fourth century. This need not necessarily mean very much - after all, some people remained rich and prosperous throughout the Great Depression. There are no figures to tell us whether the number of wealthy individuals was smaller in the fourth century compared to the second. There were the very poor in both periods, and those at every stage in between, but again we know nothing of numbers or proportions in the overall population.
There are some indications that there was less disposable income around. There is a dramatic drop in the number of surviving inscriptions from the middle of the third century. Since these frequently commemorated civic benefactions by city magistrates and other leading local figures, this suggests that such things were also becoming far less common. While it is possible that it was simply the inscribed memorials that stopped, the archaeological record does confirm the impression that grand new public buildings were becoming rare. It is worth noting that most cities already had their bath houses, theatres, amphitheatres, basilicas and temples, and so perhaps no longer needed new monuments. Yet the ideology of the local aristocracies in the earlier periods would have seen this as a challenge to build still grander structures, even if this meant risking bankruptcy, like the cities Pliny described from his time as governor in Bithynia early in the second century. The one thing every sizeable city had acquired by 300 was a substantial circuit wall. That so many of these were made from material taken from demolished buildings does not suggest widespread prosperity.?
There are also signs that long-distance trade declined. Maritime archaeologists have discovered far more Roman shipwrecks dating from the first century BC through to the second century AD than for any subsequent period. We need to treat this conclusion with a degree of caution - most finds have come from the western Mediterranean because more teams have operated there. Yet it is clear that trade in this area peaked during the early Principate. Perhaps a different pattern will emerge over time as more work is done along other coastlines. Yet excavations on land also suggest that from the third century onwards there was a substantial decrease in the number of objects circulating that had been manufactured outside the province. In part, this was a sign of strong regional development. In Gaul, Britain, Spain and elsewhere there were now locals with the skill to manufacture fineware pottery and glassware, or to lay a mosaic floor. More exotic luxuries such as silks and spices continued to be transported over huge distances - once again, there is no way of knowing whether or not the quantities of such things had changed. However, the trade with India and beyond seems to have declined sharply in the third century, only to revive in the fourth.8
In some areas cities shrank in size and population, while villages and farms were abandoned. The fortunes of individual communities had always fluctuated and smaller ones had vanished in the past for various reasons, but in areas such as north-western Gaul such failure and decline undoubtedly became far more common in the second half of the third century. Other areas most directly exposed to frequent warfare also suffered. Dura Europos was abandoned and Palmyra's prosperity dwindled along with its independence. Some regions may already have been in decline before the disturbances of the third century. Italian agriculture was at its peak of productivity in the early Principate, until lucrative markets were lost as the provinces developed. Gauls who made their own wine no longer wanted the produce of Italian vineyards in such great quantities. When trade was booming even the less fertile land tended to be cultivated for profit, but by the second century such fields were no longer worth the effort of farming and we start to hear of deserted land. Spain flourished later than Italy, in part because it came to supply olive oil to Rome and also because of the widespread popularity of the pungent fish sauce, or garum, that it produced. In time, its farmers faced competition from other regions, most notably North Africa.
The African provinces were undoubtedly one of the greatest success stories of the third century, although even here there was considerable variation from area to area. On the whole North Africa was rarely disturbed by large-scale warfare - apart from the rebellion of the Gordians, the region was almost untouched by civil war and only parts of it were exposed to barbarian raids - and its economy flourished. Irrigation schemes kept large areas fertile to a degree rarely seen before or since, producing vast surpluses for sale, especially for the markets in Italy and Rome. Its cities show every sign of prosperity, evidenced by the continued construction of grand public buildings. Palestine and other parts of the Syrian provinces also seem to have prospered and show signs of a numerous population, especially in rural areas. Britain was another region that seems to have done well, and it is probably no coincidence that it similarly was generally peaceful. In contrast to Africa, a number of its cities show signs of decline, but this may well have more to do with the local aristocracy remaining at heart more rural than urban. Many villa sites in Britain were at their grandest in the early fourth century.
Fortunes varied from region to region and there were also marked differences within the same provinces. As more archaeological evidence is acquired it will almost certainly suggest an even more complex picture. There is no doubt that some areas suffered badly in these years and went into decline, while others fared much better. A few individuals and whole communities even in the worst affected regions did very well indeed. This is almost always true, even in apparently the bleakest of conditions. It is also probable that there were many tragedies, as those unable to cope with the changes in the economic climate - or subject to any one of many short-term misfortunes - failed and lost everything. Such risks were always there in any commercial enterprise, but became more common in these disturbed times. Those for whom the margins of success and failure were narrow were inevitably most likely to succumb, as indeed they are in any age. In Africa there were plenty of wealthy local men willing to take on civic office and to fund festivals and building work. In other regions there may have been far fewer such men willing or able to keep the cities going.
Roman currency had become hugely debased - the denarius was abandoned and replaced with new silver coins, while bronze coinage vanished altogether for a while - but in many respects it had long been a token coinage. All coins still carried the head of an emperor and throughout the empire were accorded value depending on their denomination and not the actual content of precious metal. (Outside Rome's borders things were different, and there is some evidence for a preference for older, purer coins.) At times there seems to have been concern over whether or not coins bearing the image of defeated emperors would be honoured, but the authorities were generally keen to stress that all properly minted coins were acceptable.'°
In the year 300 the economy of the Roman Empire was certainly more sophisticated and robust than anything that would be seen in the same regions for well over a thousand years. Without statistics we cannot say how it compared to the conditions of the second century, but it is unlikely to have been stronger and most probably had declined, probably by a large margin. The evidence of traces of pollution in the polar icecaps from the Roman period do suggest a massive boom in the first and second centuries AD, which fell sharply in the third. In most cases, for instance, in the cases of lead and copper, the same levels would not be reached again until the nineteenth century. A number of mine workings in Spain and Britain were abandoned in the late second or early third centuries, even though they still contained extensive deposits. Combined with the loss of the mines in Dacia, it does seem that the empire was producing and refining significantly smaller quantities of minerals. Underlying the problems of the coinage was a real shortage of silver and bronze."
Some of the third-century decline proved temporary and a few regions enjoyed greater prosperity than at any other time, but elsewhere the picture was far less rosy. In the end, the essential fact of the Roman Empire was its sheer size and wealth. Variations between different regions and over time were inevitable. Overall the economy probably did decline, and quite possibly the population had fallen. In spite of this the empire was still massively wealthier and more populous than any other state or people in the known world."
Senators and Knights
Senators remained extremely wealthy - some of them fabulously so - and continued to have influence, but their political role changed profoundly during the third century. Gallienus was said to have banned senators from holding military commands. This was probably an exaggeration, and it is unlikely that there was a specific decree, although he may well have contributed to a well-established trend. Young men from senatorial families stopped serving as the senior tribune of a legion and senators were also no longer given command of a legion as legate. They continued to govern provinces, and in some cases controlled garrisons of troops, but over time ceased to hold command over major armies. The tradition of the chief agents of the state mixing civil and military posts dated back at least to the beginning of the Republic eight centuries before. Now it ended."
Instead, the trend was towards professionals or specialists. Men followed either a career in the army or one in administration and law. All of the senior posts in the army - and indeed the majority of those in the administration - passed to equestrians. To modern eyes professionals ought to be markedly more competent than amateurs. Therefore, on average, equestrians, who spent most of their career with the army, ought to have been better commanders than senators, who interspersed a few years of military experience in a much less specialised career. If so, then the rise of equestrian officers to exclusive control of the senior ranks of the army can be seen as necessary for military efficiency. It is often tied to the emergence of new, allegedly greater threats from Sassanid Persia and the Germanic tribes uniting into confederations under the pressure from new arrivals like the Goths. The Senate's amateur commanders were simply not up to dealing with such tough opponents and so emperors pragmatically turned to experienced equestrian officers instead.14
Like many theories, this emphasises the role of external pressures in forcing profound change on the later Roman Empire. Yet we have already seen that the scale of the threat posed by the Persians and Germans in the third century has been greatly exaggerated. Their successes had far more to do with Roman weakness from internal disorder than the inherent failings of the army and frontier systems. It is also very hard to see any noticeable difference in the level of competence displayed by senatorial and equestrian commanders. The evidence is lacking to say much about standards of generalship in the third century, but the 'professional' army commanders in the better recorded fourth century do not seem any more capable than the `amateur' senators of the Principate. Both led their troops in much the same way."
It is all too easy to impose modern ideas of specialism on the Roman world. Equestrian officers were professional in the sense that their career consisted only of military posts - they did not receive any specialist training. Some rose to higher ranks after service as more junior officers, most notably as centurions, but many were directly commissioned to command a force of 500 or more men (most usually an auxiliary infantry cohort). They had experience of long service, over time in progressively more senior posts, and it was clearly assumed that they would learn as they went along. The same assumption had also been made about senatorial officers, save that they had less time to acquire the skills to command. There is not a shred of evidence that competence was the main criterion for selecting equestrian officers for promotion. In fact, it would be most surprising if patronage did not play the key role in the process. There was also the question of loyalty to the emperor.
Under the Principate, equestrians commanded the praetorian guard and also governed Egypt and led the legions stationed there. No emperor wished the only military force in Rome, or control of the province that supplied the bulk of its grain, to fall to a senator who might easily become a rival. Equestrians could be trusted because they lacked the social status and political connections to make a bid for the throne - at least without first becoming a senator. Septimius Severus rose to power after the murder of two emperors and four years of civil war during which he disposed of three rivals. Unsurprisingly, he greatly increased the troops at his immediate disposal by augmenting the guard units and stationing II Parthica near Rome. This and the other two legions he raised were all commanded by equestrians, as was his new province of Mesopotamia. At the same time many existing provinces were split in two. Senatorial legates were left in charge, but none commanded more than two legions. Senators at the head of powerful military forces were potential threats, most of all to emperors who had themselves only recently gained power through armed force. Equestrians were seen as less dangerous, although over time this ceased to be true.16
Caracalla disliked the company of senators and preferred to be surrounded by army officers. Hence, after his murder the equestrian Macrinus was able to convince those on the spot to accept him as emperor. As a body the Senate quickly acknowledged his claim. He had an army, and they did not, while the memory of Severus' purges of their class were recent. After this, it was the exception rather than the rule for a new emperor to come from the Senate. Several were infants, but the overwhelming majority were equestrian army officers. As senators ceased to command armies it became even less likely that any of them would be elevated to imperial power.
Only a handful of third-century emperors died a natural death - Septimius Severus and Claudius II are the only two where this was certainly the case. The majority died at the hands of their own subordinates. Assassination and open rebellion by a usurper with an army were frequent occurrences. The men who seized power in these years were obviously ambitious or, in the case of the teenage emperors, had ambitious people behind them. They were also inevitably nervous and insecure. Not one lasted long enough to feel fully safe, still less to found a stable dynasty.
Frightened people will tend to trust only those they know well. Family were preferred and generally proved reliable - one of the main exceptions being the fragile relationship between Elagabalus and Alexander. Otherwise, men chose former colleagues and subordinates. Therefore, equestrians who had spent their careers with the army tended to choose other army officers to command their armies, govern provinces and fill senior posts in the administration. They did not know many senators intimately, and in any case were wary of those who could boast of senior magistracies, wide connections and great wealth. There was no sudden exclusion of senators from government, it was just that over time all senior appointments, and especially military commands, came to be filled by men from outside the order. Had any emperor lasted several decades in power, then it is possible that he would have felt confident enough to begin trusting more senators with such important posts. This never happened, although our sources claim that some rulers, for instance, Alexander and Tacitus, showed great public faith in the Senate .17
The great irony was that in the long run the exclusion of senators from high commands made emperors more, rather than less, vulnerable to challenges. In the first and second centuries only senators were deemed suitable for the throne if there was no acceptable candidate from the imperial family. Furthermore, only the more distinguished members of the order were likely to succeed - at most a few dozen men out of a council of boo. Emperors succeeded by establishing a working relationship with the Senate, tolerable to both sides. Septimius Severus eventually managed to do this, even if his methods were at times brutal.
It was much easier to control the Senate than the larger and more diverse equestrian order. Senators returned to Rome throughout their career, mixed socially and inter-married. An equestrian officer was most unlikely to know men of a similar rank serving with the army in distant provinces. It was now far easier to become emperor, since a man only needed to gain the support of the troops on the spot. Persuading the rest of the empire that his rule was better than any alternative was much harder. The Senate had remained a force for political cohesion long after it lost genuine political independence. Yet its value in this respect was dependent on emperors treating the body with respect and employing senators in the vast majority of senior posts. This was the way Augustus had smoothed the feelings of Rome's elite, helping them to accept the passing of the Republic. For emperors it was a voluntary convention, but it had generally worked well for over two hundred years.
The replacement of senators by equestrians at the head of Rome's armies had nothing to do with military requirements. Still less was it caused by concerted demands from an increasingly prosperous and numerous class for a greater role in public life, for the equestrian order lacked any sense of corporate identity. It had always been possible in the past for the most successful individuals to be enrolled in the Senate. The rise of equestrians to dominate high office in fact led to increasingly stark divisions within the order between those of higher and lower status. The dominance of equestrian officers was a result of the desire of a succession of newly made and normally equestrian emperors to surround themselves with men they felt could be trusted. That this was largely an illusion is shown by the number murdered by such `trustworthy' subordinates.
As the Senate lost its central role in public life, so Rome itself also diminished in importance, although both were still powerful symbols of the empire's greatness. The Senate continued to meet, debating and passing motions - usually for very formal praise of the emperor. Senators still had influence, and also filled administrative posts in Italy and governed lesser provinces. Rome's population continued to be pampered with free and subsidised food and drink, enjoyed spectacular entertainments in huge venues like the Colosseum and Circus Maximus, and took their pleasures in massive public bath houses. People rarely saw an emperor, and even when one was resident in the city, he was most unlikely to enter the Senate House. Rome remained the largest city in the empire, but it and its Senate had drifted to the margins of political life. Power rested with the emperor and his court, and these rarely came anywhere near Rome.
Surviving the Crisis
Emperors spent little time in Italy, and even less in Rome itself. They frequently went to war, in some cases campaigning every year. Rule of a single emperor became unusual. In some cases this was involuntary, for instance, when rival emperors emerged who were too strong to overcome at the time. More often it was a matter of choice. Emperors with sons tended to elevate these to joint rule almost as soon as they seized power. If they were still infants this marked out a clear successor and so gave the promise of long-term stability. Adult sons could be sent to deal with problems in another region while the father was busy elsewhere.
For many scholars this development shows that it was no longer possible for a single emperor to rule the entire empire. Most also claim that from at least the time of Marcus Aurelius, emperors were expected to preside over major wars in person. They were also expected to win. If there was more than one war at the same time, then there needed to be more than one emperor. Therefore, the trend was towards the division of imperial power, and ultimately this would lead to the split between eastern and western empire at the end of the fourth century. Such analysis not only relies on hindsight, but assumes that this major change was a reasonable, perhaps even inevitable, response to new problems. Once again, this is bound up with the belief that in the third century the empire faced far greater threats from the outside than ever before."
Change was inexorable and stimulated by external pressure. Similar reasoning has often been used to claim that the fall of the Republic and creation of the Principate by Augustus was equally inevitable. If he had not done this, then someone else much like him would have created an essentially similar monarchy. This sort of analysis robs the individuals whose decisions and deeds shaped the process of any independence of action, perhaps even of responsibility. More importantly, a closer examination of the course of events paints a very different picture. Many of the underlying assumptions prove deeply questionable and any sense of inevitability vanishes.
In the second century Trajan spent several years of his reign on campaign, and Hadrian carried out long tours of the provinces. Both also spent a good deal of time in Italy, and Antoninus Pius never left there. The joint rule of Marcus Aurelius and Lucius Verus had nothing to do with any perceived need for two emperors. When Verus died he was not replaced for seven years until Commodus was considered old enough. It had long been normal to mark out a successor by sharing power with him. Commodus ruled alone, as did Septimius Severus until he wished to make clear that his sons would succeed him. Caracalla and Geta inherited jointly because an adult son could not be safely passed over.
Marcus Aurelius took his role as emperor extremely seriously and spent much of the second half of his reign on campaign. This was out of his sense of duty rather than any obligation. Commodus saw things differently and never again went on campaign after he returned to Rome from the Danube. Severus waged war often. At first his opponents were Roman and then subsequently he led major operations in the east and in Britain - the regions that had supported his rivals. There may have been good reasons for these wars, but Severus was also concerned to create a bond with the troops in each area and to win glory - Augustus had been similarly aggressive in the first years of the Principate. Caracalla craved an image as a tough soldier and preferred life on campaign to public life back in Rome. Winning great victories always had valuable propaganda value - hence the taking of triumphal names like Ger manicus, Samarticus or Parthicus, and the prominence of symbols of victory on coins."
Military victories were especially attractive to insecure emperors. Success in war showed their competence and, with its implications of divine favour, suggested that their rule was legitimate. Well-established emperors could take credit for wars won by their governors just as Antoninus Pius had done. The less confident needed to preside over the success in person. They were also reluctant to let anyone else gain glory that might rival or surpass their own. The division of the provinces into smaller units anyway ensured that governors had smaller armies at their disposal and these were not always sufficient to deal with a major problem. The third century saw the growth of extraordinary commands - for instance, Philip's brother Priscus over much of the eastern frontier, and the similar rank later granted to Odaenathus. This was necessary because provincial governors now operated on too small a scale to cope in some circumstances. Yet it was inherently dangerous. Time and again a local success by a governor or other commander prompted him to declare himself emperor."
This was the root of the need to share imperial power, far more than an increased threat from outside. Sharing rule with a colleague meant that there was someone else who could - hopefully - be trusted to take command of substantial forces in another part of the empire. Trustworthiness, however, did not ensure competence. Nor did it prevent the colleague from being murdered and a challenger taking over command of his troops. No emperor was ever truly secure in the third century. Long ago, Emperor Tiberius had described ruling the empire as equivalent to `holding a wolf by the ears'. He had the advantage of succeeding a father by adoption who had ruled for forty-five years, and himself reigned for more than two decades. No ruler lasted that long between Severus and Diocletian."
The Roman army often receives much blame for its role in the disruptions of the third century. Clearly, none of this could have happened without the willingness of Roman soldiers to fight and kill each other. Yet the army did show a clear preference for established dynasties, just as it had done throughout the Principate. Enough of the troops were willing to back Elagabalus, in spite of his extremely dubious claim, because he was seen as a Severan. Similarly, it took major failures on the part of Severus Alexander before a coup succeeded. None of the other emperors could lay much claim to deep loyalty, especially since so many of them had in the first place been involved in killing their predecessor. The army wanted stability, when it could be sure of being paid and the more ambitious could win promotion. Failing that, soldiers usually wanted to be on the winning side, hence the number of emperors killed when their army was approached by that of a stronger rival.
Soldiers of all ranks certainly took advantage of the chaos of civil war, which inevitably relaxed discipline. There were plenty of opportunities for looting and extortion. It was also possible to rise very quickly to high rank. Elagabalus' leaders promised to promote to the vacant rank any of Macrinus' men who killed a senior officer refusing to defect to them. The best prospects of all were open to officers closest to the emperor. If they joined a conspiracy to arrange for his murder then they could expect rewards from his successor. There were opportunities, but there were also great risks. Backing the wrong side could prove fatal, and certainly risked damaging a career, so that men had to judge very carefully whether to stay loyal or switch allegiance to a challenger. The majority of emperors died at the hands of men who had been close to them. Trust was a precarious thing for the ruler and his subordinates. Fear of execution prompted several assassinations from Commodus and Caracalla onwards.22
It was a vicious circle, as each new assassination or rebellion by a usurper - no matter how quickly it collapsed - made a renewal of civil war more likely. A humiliating defeat in a foreign war was likely to provoke an emperor's rapid fall. Others were murdered because of alleged pursuit of their officers' wives. Some died simply because ambitious followers believed that they or an associate could take imperial power for themselves. Emperors were always more worried by Roman rivals than foreign enemies. Even the Persians stood no real chance of seizing large areas of Roman territory. German warriors might raid Italy, Spain or Asia Minor, but they could not stay there.
In the third century the Roman Empire wasted much of its strength fighting itself. Its military system suffered severe dislocation and became less able to deal with foreign threats. Augustus had created a system that veiled the power of the emperor. It was not a clear, constitutional position and so there was no formal arrangement for succession. This was a weakness, but the general stability of the first two centuries make it hard to see this as responsible for the problems of the third century. Chance played more of a role than historians might care to acknowledge. It happened that Marcus Aurelius was survived by a young son who was simply not up to the job of being emperor. Pertinax need not have misplayed his hand and been killed, and Septimius Severus might have lived longer and perhaps taught his sons how to work together and to rule well. Had Commodus, Caracalla, Elagabalus or Alexander been older when they came to power, then they might have proved more capable.
Decisions by successive emperors made the situation worse. The gradual exclusion of senators from military commands and their replacement by equestrians was intended to make emperors safer. In the long run it did the opposite, and emperors came more and more from the ranks of senior equestrian army officers. Similarly, reducing the size, and therefore garrisons, of provinces weakened the capacity of the governors to deal with major operations. Emperors had to go themselves or give sufficient extraordinary power to allow a subordinate the resources to deal with the situation, hoping that it would not also give him the ability to rebel. Not until Diocletian was a measure of imperial control reasserted. Far fewer emperors were assassinated after 285. Civil wars became somewhat less common, although they tended to be much bigger and more costly when they did break out. Things were not as bad as they had been in the half century before he came to power, but we should not exaggerate the change. The stability of the first and second centuries would never be repeated.