Ancient History & Civilisation


The Roman Empire had a long life: from 30 BC to AD 476 in its ‘imperial’ phase, a matter of five centuries; several more centuries before as a Republic; and 1,000 years longer as the ‘East Roman’ or ‘Byzantine’ Empire centred at Constantinople. It is no surprise that conditions changed during that time, or that the methods of selecting emperors should alter. Change in the Roman Empire was always slow, though it clearly did take place. On the question of the methods of the succession of the emperors, change tended to be irregular and abrupt. It is perhaps more surprising that the basic elements involved in that selection and succession should still be the same at the end of that long period as they had been at the beginning. These were the previous emperor, the Senate and the army, and at times the Guard and the Roman population, though the Guard was essentially negative and the Romans had little influence for much of the time. Nevertheless, within that time, the importance of each of the elements did change and the interplay between them continued throughout the five centuries between Augustus and Ricimer. Each in turn seemed to control the process, but the others had a considerable influence as well.

There was a basic tension, or conflict, built into the process of finding an imperial successor, for the three essential players in the process had different priorities and claims to authority. The army required a commander, preferably one they already knew before he became emperor, and preferably one who was reasonably generous in his handing out of pay. The Senate, on the other hand, required a politician, a man who was familiar with social and political conditions in the city of Rome and one chosen by the senators themselves. This was a reflection of the Senate’s original powers, the fact that the emperor was in effect a magistrate and that magistrates in the Republic had been elected. The Senate claimed an indefinite right of election, which shifted rather into one of investiture and acceptance, but remained a senatorial requirement which, throughout the Empire, was one of the marks of ‘legitimacy’.

The influence or intention of the previous emperor was probably always the strongest element in the selection of his successor. This brought immediate conflict with the Senate, and less so with the army, for the automatic choice by an emperor of the man to succeed him would always be his son, if he had one. This hereditary impulse was the basic social building-block of Roman (and every other) society. So whereas the Senate claimed a residual right of confirmation (or refusal, presumably), the normal assumption would be that the emperor’s successor would be his son or his nearest male relative. In that case, the Senate would have no role in the choice.

Normally the Senate would accept such hereditary succession, though it could at times resurrect its claim to confirmation with awkward results, but the main problem was that from the time of Augustus to the mid-third century (almost three centuries), only three emperors – Vespasian, Marcus Aurelius and Septimius Severus – had sons who could inherit. Whatever the reasons for this – a genetic condition, the unhealthiness of Rome or homosexuality – it is an extraordinary sequence of procreative failure.

Augustus’ attempts to establish a dynasty were thus repeated by almost every emperor who came after him. His own failure to beget a son forced him into elaborate expedients for devising a succession system, and these reappeared throughout the history of the Empire in one guise or another. His one child was a daughter, so the pool of possible successors widened to include his marital relations. He resorted repeatedly to the adoption of ever more distant relations as his son or sons to pretend to an hereditary system. His innovation of suggested emperor-regents such as Agrippa and Tiberius was designed to safeguard the actual succession to those within his own family, and it was an expedient that was resorted to repeatedly by his more distant successors. It culminated in the near institutionalization of the office of patrician-cum-magister militum, or army commander, in the fifth century. Ricimer was a direct political descendant of Agrippa: both were rulers by virtue of their military prestige and the support of and their support for the actual emperor, and both married into the imperial family; right at the end Orestes was intending to rule for Romulus as regent for the under-age emperor.

The essential drawback of heredity as a succession system was fairly soon apparent. Gaius/Caligula arrived at the throne with virtually no training for rule, and his peculiar genius was to see that his actual powers were limitless; his fate then showed what happened to a man attempting to exercise that unlimited power. He was followed by an emperor who was widely regarded as being physically unfit, and then by a wilful child. Hereditary succession could produce monsters just as easily as competent men, but Gaius’ realization of the sheer power at the emperor’s disposal meant that the potential for monsters was significantly increased and the only way of dealing with such a phenomenon was by murder.

All these issues were in fact apparent from the start. Julius Caesar’s murder came from his ambition and his sidelining of the traditional system of politics in Rome. Augustus finally, largely by luck, chose his stepson as his successor after neglecting him or favouring others for decades; Tiberius nominated his nearest male relative, who turned into the tyrant Gaius.

It was necessary for Augustus to keep the Senate informed of and involved in his plans, for the Senate still had much of the prestige and experience of its Republican predecessor. Yet Augustus treated it brutally, executing and exiling senators, and repeatedly purging it of men with whom he disagreed, thereby demonstrating his imperial power over it. Even so, it retained a central role in the determination of the succession all through the century after his death. In the civil warfare of 68–69 it came through virtually unscathed in authority, if not in personnel, but it over-reached itself in 96–97. By taking on a central role in elevating Nerva, the Senate effectively claimed the ultimate right to decide the imperial succession. The senators were relying on the prestige of their House and on the effectiveness of their emperor, but politics in the Roman Empire was founded less on influence and more on physical force. It was soon revealed that the Senate’s authority was in fact very limited and the Senate’s emperor could be forced by threat of death to do the generals’ will. Then the near-brutal insistence of the army commanders in conclave on the Danube that Nerva must adopt Trajan as his successor showed exactly where the real power in the Empire lay, and effectively reduced both the Guard – which had developed exaggerated ideas of its influence – and the Senate to subordination. For the next century and a half, until 238, the Senate’s role was as a ratifying body only: the decision on the succession was always made elsewhere, by the reigning emperor until 192, then by the army and the reigning emperor, either together or in competition.

The dethronement of the Senate was, however, not solely to the advantage of the army but mainly to that of the sitting emperors. From Trajan to Commodus the successor was always a man designated by the emperor before his death. This had also been the case before Nerva, of course, with the Flavians, but they had gone through the old motions, originating with Augustus, of ensuring that the designated successor accumulated many consulships, thereby involving the Senate in the process, but Trajan and Hadrian did not bother to designate their successors till the very end of their lives (or even, in Trajan’s case, after his own death). In both cases, the choice was a senator, but neither man had been singled out in the traditional way beforehand; both Trajan and Hadrian were consulars, but with only one consulship each to their names when they became emperor. There was a considerable number of other men who had more consulships at the time of their accession; that is, the prestige of such offices was much less from Trajan’s accession onwards. In theory, of course, this allowed the emperors to select the most suitable candidate; in fact, it turned out to be very messy in both cases: Trajan mounted a coup against Nerva, and Hadrian’s search for a successor was murderous. Then when an emperor had an heir of his own body and so had an obvious successor in place from birth, it turned out to be Commodus.

Trajan, Hadrian and Marcus all paid more attention to the army than to the Senate in considering the imperial succession, the result of the generals’ coup that brought Trajan to power. Septimius and Caracalla took up that emphasis and expanded it, even further reducing the influence of the Senate. This did not remove the Senate’s pretensions, however, but attempted revivals of senatorial authority in 193 and under Alexander Severus were only temporary, yet did succeed in 238 and for several decades after that. This was not a revival at the expense of the army so much as one at the expense of the imperial office. The army, like the old Praetorian Guard, whose apparent power was similarly crushed by Septimius with little effort, was always better at knocking down emperors than finding new ones. The trouble with senatorial selection was that the Senate’s criteria for suitability – in particular age and senatorial experience – did not impress the army, and were no longer those that were the best for the Empire in the third century. The rapid removal of emperors between 235 and 284 gave plenty of scope for senatorial influence, but when an emperor was required in the midst of a military campaign the choice had to be left to the army in the field, and this prioritized military command skills above all else, though such men were less effective as politicians. So, for that half-century, the balance of influence between army, Senate and ruling emperor oscillated; in the end it came down on the army’s side, first under Carus and his family, and then Diocletian, none of whom paid any heed to the Senate.

The Tetrarchy was, in effect, a military dictatorship. Diocletian had total disdain for the Senate, but carefully maintained his grip on the army and his appointed imperial colleagues. His experimental reforms, of which his organization of the imperial succession was one of the strangest, was just the sort of centralizing and controlling measures to be expected of a general indulging in politics. His succession scheme was apparently designed to promote competent generals from within the army and then cement them into a rigid system, enhanced by intermarriage. Yet it failed at its first test, knocked over by Constantine and Maxentius in favour, once again, of the hereditary principle. The aim of Augustus three centuries earlier had been to establish a dynasty and it had repeatedly emerged with later emperors as well; only with Constantine could it be said that a clear case of a dynastic succession had been finally and successfully established, yet Constantine’s dynasty only lasted for three generations.

The Senate’s interest in becoming involved in the succession never died, despite the triumph of the army and of the hereditary principle. With these being the determining factors, the scope for senatorial influence faded in the fourth century, but in the end in the West in the fifth century, it partly revived. Between 409 and 475 senatorial candidates for the throne repeatedly emerged, as pretenders, as brief rulers or as actual emperors: Priscus Attalus, Petronius Maximus, Libius Severus and perhaps Avitus and Olybrius. Not since the Gordians and Valerian had so many senators been able to reach the top.

This, however, was also one of the problems that overwhelmed the Western part of the Empire at that time. The last two decades of the Empire in the West, from the time of Valentinian’s murder of Aetius in 454, saw another period of oscillating influences as each interested group attempted to install its own man as emperor: the Eastern government, the army commander (magister militum or patrician) and the Senate. These were the same elements that had been Augustus’ concern five centuries earlier. Then they had been the ruling emperor, the army and the Guard, and the Senate; in the fifth century they emerged as the emperor in Constantinople, the army’s warlord and the senators. The balance struck by Augustus between these groups and interests had rocked back and forth, with each having its time of predominant influence ever since. Yet the inherent conflict between them was never resolved and was still being played out in the 470s.

It proved in the end to be an unsustainable tension, even though it had by then operated for five centuries. The final victory went to the army, in the person of Odoacer in 476. It is curious that it was the demand for land for the soldiers that was one of the main problems provoking Odoacer’s break with the Empire, for this had been one of the main demands of the several generals whose power had brought the Republic to its knees, and solving it for the time being had been one of Augustus’ main achievements. In both cases it was the reluctance of the Senate to comply with the army’s demands that provoked the generals into action.

However, the army in the fifth century was no longer the army of the Republic, or even of two centuries before, one composed of citizens or aspirant citizens. Now it was a barbarian war band, and one liable to be all too easily defeated by a competing war band. Therefore the establishment of full control by the army commander meant that what the three groups were fighting for – the Empire itself – was brought down in ruins around them. This was not the inevitable result. In the first two decades of the century dynastic and senatorial emperors had survived rather longer than the army’s candidates. In the East, where the senatorial and army influences were so much weaker, the dynastic principle, as modified by coups d’état at irregular intervals – in effect the normal system since the death of Gaius – remained the norm for the next 1,000 years.

There was no real reason why this could not have been the case in the West as well. Even as late as 470 the Western emperors’ authority extended throughout Italy and its borderlands and into much of Gaul and Spain, and the barbarian settlers within the frontiers were federates – part of the imperial system – and in several cases more than willing to work to sustain the Empire. Recovery, as the third-century crisis had shown and as Justinian demonstrated later, remained perfectly possible. It was the constant disputes over the powers and the person of the emperor in the West that were debilitating, far more so than any fighting with the barbarians. The effective secession of much of the Empire outside Italy was due to the absence of an emperor to whom the Romans in those territories could direct their allegiance, and this made it all too easy for the barbarian kings – notably Geiseric and Euric – to seize and expand their lands and to repudiate imperial authority. Allegiance could not be given to a barbarian magister militum, especially when he was liable to dethrone or murder his nominal imperial master and then to leave the throne empty for long periods. It was the disabling tension between emperor and warlord in Italy that paralysed the Western Empire. Odoacer’s decision to eliminate the emperor was perfectly rational and permitted a revival of power in Italy, but it was too late to revive the whole of the West; a powerful emperor might have eliminated the warlords, as Valentinian III attempted to do, with the same result. On the other hand, the system of a warlord and an ineffective emperor that had existed since Honorius’ time had elements of stability in it and could well have lasted indefinitely. Yet the disagreement between Odoacer, Orestes and the Senate persuaded Odoacer that a roi-fainéant (‘do-nothing king’) kind of emperor was a useless appendage. The emperorship in the West was abolished, and along with it the Empire itself, in this casual manner, as a result of a personal-political disagreement.

The continual disputes over who should be emperor were all effort that could have been more profitably directed elsewhere. In the 60s the success of Corbulo in war in the East was felt by the profoundly unmilitary Emperor Nero to be a threat, and his order for the general to commit suicide was one of the factors that drained the last support for him; in AD 96 senators and others conspired to murder the Emperor Domitian just as he was about to set off to the frontier to lead a major trans-frontier offensive aimed at the conquest of at least Bohemia, one that could well have succeeded. This would have been a change to the northern frontier of much greater strategic significance than Trajan’s conquest of Dacia a decade later (and that had to be abandoned later); it was prevented by the senatorial plot. In 238 Maximinus’ trans-frontier intentions were sabotaged by the Senate’s recognition of Gordian I and II and then its elevation of Balbinus and Papienus. Emperors were repeatedly struck down in the midst of attempts to eliminate the problem of the Eastern frontier, sometimes by disease, occasionally by wounds, often by a murderous plot among their own officers: Germanicus, Trajan, Lucius Verus, Caracalla, Gordian III and Julian, to name but a few. This may be a sign of imperial over-extension, but the agent of striking down was all too often the Senate. It is difficult to accept that senators had the welfare of the Empire as a whole at heart.

Incidentally it is noticeable that the beneficiaries of the murders of emperors were rarely those who carried out the killing. The typical case is the first: the killing of Gaius; he was murdered by a soldier, but then his fellows were at a loss as to what to do until Claudius was found in the Palace. Claudius’ killing, if it was murder, did benefit Nero and his mother, though Agrippina, the presumed murderess, was herself soon murdered by her son. Galba’s murder only benefited its author Otho for a few months. Vitellius’ killing scarcely benefited Vespasian, who had won the civil war already. Domitian’s murder did benefit Nerva, who was promoted as the next emperor, but he proved a fairly weak ruler and held power for less than a year before the army coup that made Trajan his heir also sidelined Nerva himself. Commodus’ murder briefly brought Pertinax and Didius Iulianus to the throne and they were probably involved in the plot to kill him, but the eventual beneficiary was Septimius, who had not been involved in the killing.

So it went on: the killers of Commodus, Elagabalus and Maximinus did not benefit. The killers of Geta, Caracalla, Macrinus and Alexander Severus did, from anything from a year to six years, but this is scarcely a mark of great success. In the third century, murders of emperors became routine, but they were often followed by conclaves of generals to choose a new ruler and the eventual chosen one was normally not involved in the original plot. The beneficiary of Geta’s death was his co-ruler Caracalla, and it was Constantius II who benefited from the murders of his brothers but he was not party to the deeds. In the fifth century it could be said that Petronius Maximus benefited by his murder of Valentinian III, but only for a couple of months; the beneficiaries from Petronius Maximus’ own killing (by the Romans) were the Vandals (who sacked Rome). Killers of the emperors from Majorian onwards tended to be non-emperors like Ricimer.

This was therefore a most inefficient way of changing rulers. If the murder was a spur-of-the-moment deed like that of Gaius or Commodus it only produced confusion; if it was the result of a plot, the new ruler’s regime was irremediably stained from the start and, of course, it only revealed that it was possible to gain power by a plot and murder. To those who feared or were disgusted by a ruling emperor, however, his killing was the only way of changing things. The absence of any less lethal method of replacing a ruling emperor made murder the only option, unless it was a rising in rebellion and instituting a civil war, in which many more people would die than by the assassination of a single man.

The common factor in all the murders and crises was the failure of the Roman governing system to devise an intelligible and workable system of imperial succession. It is, after all, not something that is too difficult to arrange. Hellenistic kingdoms achieved it from an even less convincing constitutional basis than that of Augustus; the barbarian kings camped in the territory of the Empire had done so; mediaeval Christian states succeeded in doing so; modern democracies have managed it; even empires have produced workable systems, though they have often found it much more difficult than more compact states, since by definition a great empire lacks the legitimacy of a traditional kingdom or city and the rewards of seizing power are always greater. It is a mark of the basic incapacity of the Roman political system: at first the monarchy failed, then its successor Republic (eventually doing so repeatedly over a period of a century and more), and then its method of imperial succession also failed. It did not much matter which of the possible schemes of succession could be settled on: dynastic succession, primogeniture, election for life or for a period of time, or military dictatorship. It was the failure to settle on any of these that was one of the major causes of the fall of the Empire.

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