The murder of Gaius was a particularly bloody case of regime change, but the transmission of imperial power in Rome was often murderous. Despite the impressive survival rate of the emperors (fourteen rulers in almost 200 years is one testament to stability), the moment of succession was fraught with violence and surrounded by allegations of treachery. Vespasian in 79 CE was the only emperor in the first two dynasties to die without any rumours of foul play surfacing. Gaius, Nero and Domitian met obviously violent ends. There were rumours of murder surrounding the deaths of all the others. The names, dates and details change, but the story remains the same. Some said that Livia poisoned Augustus to ease Tiberius on to the throne; Tiberius was widely believed to have been poisoned or smothered to make way for Gaius; Agrippina is supposed to have dispatched her husband Claudius with some poisoned mushrooms in her successful bid to make her son Nero emperor; and some said that Domitian had a hand in the early death of Titus – contrary to a hopeful story in the Talmud which claims that after Titus destroyed the Temple in Jerusalem, a gnat flew into his nostril and gradually ate away his brain.
Many of these stories must be fiction. It takes a lot to believe that the elderly Livia would have painstakingly smeared poison on figs still growing on a tree, then tricked her husband into eating them. But true or not, together they underline the uncertainty and danger in the transmission of power. The message was that succession almost never happened without a struggle or a victim. This was a pattern projected back to the myths of the early kings too: they enjoyed long reigns, but only two of the seven died natural deaths. Why was it so difficult? And what solutions did the Romans find?
The first Augustus intended to make one-man rule permanent and to keep it in the family. But the series of deaths among those marked out as his heirs and the lack of any surviving sons from his marriage to Livia dogged his plans. Succession throughout the first dynasty continued to be fraught, as different claims from different sides of the Julio-Claudian family tree clashed. But the problems were bigger than that, and they would not have disappeared even if the imperial couple had produced half a dozen healthy boys.
Augustus was trying to invent from scratch a system of dynastic succession, against the background of a fluid set of Roman rules about the inheritance of status and property. Crucially there was no presumption in Roman law that the firstborn son would be the sole or principal heir. The standard modern system of primogeniture is a fail-safe mechanism for removing any doubt about who should succeed, although – by making the order of birth the only criterion – it risks some decidedly unsuitable incumbents on the throne. In Rome, the eldest male child of the emperor would have had a certain advantage in trying to follow his father, but no more than that. A successful claim to power also rested on behind-the-scenes manoeuvres, on the support of key interest groups, on being groomed for the part and on the careful manipulation of opinion. It also depended on being in the right place at the right time. The only reliable way to guarantee a peaceful transition was to have the new emperor on the spot to take over the old Augustus’ signet ring as he breathed his last, with no awkward gap. That is what the rumour-mongers realised: most of the allegations of poisoning under the Julio-Claudians present the murder not as part of a plot to spring some new candidate into power but as an attempt to get the timing right and to ensure a seamless takeover for the man already marked out as the likely successor.
These uncertainties about how to establish a legitimate claim to rule also help to explain the peculiarly murderous image of the Roman imperial court, where danger seems to have lurked on every fig and such an atmosphere of suspicion prevailed that Domitian is said to have had the palace walls lined with reflecting stone so that he could see who was coming up behind. Without any agreed system for the transmission of power, every relative counted as a potential rival of the emperor or of his likely heir – and it followed that those in the penumbra of the imperial family found themselves in a very perilous position indeed. Many of the stories may well be more fantasy than fact; the Roman elite was not by nature particularly cruel and ruthless, even if that is the image they have in film and fiction. What was ruthless was the fundamental logic of imperial succession. Tacitus captures that, with characteristic cynicism, in describing the events of the beginning of Nero’s reign in 54 CE. ‘The first death under the new emperor,’ he starts, implying that there were many more to follow, was that of Marcus Junius Silanus Torquatus, the governor of Asia. He was a man of no ambition whatsoever, so shamelessly apathetic, Tacitus explains, that Gaius had aptly nicknamed him the Golden Sheep. But his death was inevitable, and the reason obvious: ‘He was a great-grandson of Augustus.’
There were alternative routes to power. One was exactly what the first Augustus had tried to preclude: elevation by the army. In 41 CE the Praetorian Guard in Rome had played the leading part in putting Claudius on the throne. In 68 CE, to quote Tacitus again, ‘the secret of imperial rule was revealed, that an emperor could be made somewhere other than Rome’. ‘Somewhere other than Rome’ is a euphemism for ‘by the legions in the provinces’, as each of the four rival claimants to replace Nero was backed by army units from different provinces. Within eighteen months, Vespasian was raised to power in the East, with no connection by birth to the Julio-Claudian dynasty. It is clear, however, that he and his supporters felt that military force alone was not enough to secure his position. Despite the down-to-earth image he later projected, at the beginning of his rule widespread reports of the miracles he had worked underpinned his claims to the throne. In Egypt, just before his proclamation as emperor, he is supposed to have restored sight to a blind man by spitting on his eyes and to have cured another man’s withered hand by standing on it. Whatever carefully manipulated display lay behind these reports (and whatever the uncanny similarity with a far better known miracle worker of the first century CE), eyewitnesses are said to have vouched for the miraculous cures years later, long after Vespasian’s death.
The praetorians continued to influence imperial succession; certainly, no one would have been able to hold on to the throne if the troops in the city actively opposed him. But in the period up to 192 CE they never again engineered quite such an open coup as they had in 41 CE, nor in that period did the legions in the provinces ever again create an emperor. That is partly because from the end of the first century CE – after a brief interlude of relatively unproblematic succession in which Vespasian had been followed by his two natural sons – an alternative route to the throne was devised, which appeared to get round some of the earlier difficulties: adoption.
Adoption in Rome had never been principally a means for a childless couple to create a family. If anyone just wanted a baby, they could easily find one on a rubbish heap. Adoption among the elite had always been a means to ensure the transmission of status and property and the continuance of the family name in the absence of surviving sons. Those adopted were more likely to be distinguished adolescents or young adults than babies, whose high risk of death made them an unwise investment. That is how Scipio Aemilianus, for example, the friend of Polybius and conqueror of Carthage in 146 BCE, the natural son of another famous Roman commander, Aemilius Paullus, ended up in the Scipio family.
It was not at all surprising that Augustus and his successors in the Julio-Claudian dynasty used adoption, as other elite families sometimes did, to mark out their favoured heir among the wider group of relatives. Hence Augustus adopted his grandsons and, when they died, did the same thing with Livia’s natural son, Tiberius; Claudius likewise adopted his wife’s son, Nero. But from the end of the first century CE there was a new pattern. When Domitian was assassinated in 96 CE, the senate offered the throne to the elderly and childless Nerva – a safe pair of hands presumably. Between Nerva and Marcus Aurelius heirs to the throne were selected and adopted without obvious concern for family relationships. Some had no link to the existing emperor by blood or marriage at all, or only a remote one, and they came from further afield. Trajan, the first such adoptee, was originally from Spain; the families of others came from either there or Gaul. They were the descendants of early Roman settlers abroad, who had probably married into the local communities, rather than from the indigenous population. But, in a way that dramatically fulfilled the Roman project of incorporation, they made the point that the emperor could come from the provinces of the empire.
This new system, which operated for most of the second century CE, was sometimes presented as a major shift in the ideology of political power, almost a meritocratic revolution. Gaius Plinius Caecilius Secundus (now called ‘Pliny the Younger’, to distinguish him from his uncle ‘the Elder’) justified the procedure in precisely those terms, in a speech delivered to the emperor Trajan: ‘When you are about to hand control of the senate and people of Rome, the armies, the provinces, the allies to one man alone, would you look to the belly of a wife to produce him or search for an heir to supreme power only within the walls of your own home? … If he is to rule over all, he must be chosen from all.’ Tacitus, also writing during the rule of Trajan, echoes those sentiments in a speech he put into the mouth of Servius Sulpicius Galba, one of the claimants who briefly held power after the death of Nero. Just a few days before his death, elderly and without an heir, Galba looked for someone outside his family to adopt as a successor. Tacitus’ words ostensibly justify that decision in 69 CE; but they really belong to the world of imperial adoption in his day: ‘Under Tiberius and Gaius and Claudius,’ he makes Galba say, ‘we Romans became the inheritance of just one family … Now that the Julio-Claudian dynasty is over, adoption will select only the best. For to be descended and born from emperors is pure chance, and is rated no more highly.’
These are fine words, and they suggest a new style of reflection on the nature of the emperor’s power and qualities. In practice too, the adoptive system occasionally worked smoothly. On the death of Nerva in 98 CE, Trajan’s succession was so guaranteed that the new emperor did not even return to Rome from Germany for more than a year. But it was not the perfect solution that some of the glowing ancient accounts make it seem. To read between the lines, it is clear that the praetorians had pressured Nerva into adopting Trajan (Pliny’s speech lets out rather awkwardly that Trajan had been ‘forced’ on the old man), and the legions massed with Trajan on the Rhine might well have been a factor too. And when Trajan died, almost twenty years later, whatever really happened, the reported machinations are very much on the Julio-Claudian model: there were rumours of poisoning, the adoption of Hadrian was announced only at the very last minute, and some suspected Plotina, Trajan’s wife, of manipulating the succession in Hadrian’s favour and concealing the death until all arrangements were in place.
Besides, despite the splendid meritocratic rhetoric, adoption was still treated as a second-best means of succession. When Hadrian wrote a little poem in honour of Trajan, he preferred to call him the descendant of Aeneas rather than the son of Nerva – a fantasy of genealogy that perhaps also hints at Trajan’s overseas origin. Pliny ended his fulsome speech in praise of Trajan with hopes that the emperor would in due course have sons and that his successor would indeed come from ‘the belly of a wife’. And when Marcus Aurelius was the first emperor for more than seventy years to produce a son and heir who survived childhood, that son succeeded him without there being any pretence of searching for the best man for the job. The outcome was disastrous. Commodus’ assassination in 192 CE was followed by the intervention of the praetorians and of rival legions from outside Rome and by another round of civil war, which marked the beginning of the end of the Augustan template of imperial rule.
Roman emperors and their advisors never solved the problem of succession. They were defeated in part by biology, in part by lingering uncertainties and disagreements about how inheritance should best operate. Succession always came down to some combination of luck, improvisation, plotting, violence and secret deals. The moment when Roman power was handed on was always the moment when it was most vulnerable.