Regime Change, Resistance, and Reconstruction
IMPERIALISM ANCIENT AND MODERN
THE NATURE OF ROMAN IMPERIALISM
This chapter addresses problems arising from the presentation of the Roman Empire in much modern literature as a largely benign power. Recent world events remind us of the potential messiness of imperial adventures designed to bring about regime change.1 The events of the conquest period in Britain will be reassessed, with a particular focus on the dismantling of the client kingdom that lay at the heart of the Roman decision to invade.
A simple definition of empire is “rule over very large territory and many peoples without their consent” (see the introduction to the present volume). The Roman Empire was just such a system of territorial domination, social power, and economic exploitation, and in these respects it shares common characteristics with other empires. Power is the key concept for understanding what empires share in common and in moving beyond observations about the obvious differences between ancient land empires and modern capitalistic and maritime empires.2 While recognizing that there are profound differences in the scale and operation of modern imperialism, my starting point is that all empires share a common basis in the enforced domination of lands, seas, and peoples. This nonconsensual nature of imperial rule is of the highest importance, because the existence of an empire predetermines the emergence of resistance to its power.3
CLIENT RULERS IN THE ROMAN WORLD
Of the entire area which is subject to the Romans, some is ruled by kings, some they rule under the designation ‘provincial’ territory, appointing governors and tax collectors to the inhabitants. There are also free cities, some of which attached themselves to the Romans as friends from the outset, while to others the Romans themselves granted freedom as a mark of honour. Some dynasts, tribal chieftains and religious rulers are also subject to the Romans; these people regulate their lives along traditional lines.4
Alongside the territorial domination of areas formally annexed, the Roman Empire made considerable use of hegemonic methods of control.5 The recognition of individual client rulers and the formal assignation of territories to their control was a standard element of imperial rule in the Late Republic and the Early Principate.6 The technical title of these rulers was rex sociusque et amicus populi Romani, often abbreviated as “friendly kings,” a status that enshrined a mass of ambiguities of power.7 Although these kingdoms differed in their scale and their specific relations with (and obligations to) Rome, there was a sense in which they were seen as somewhat equivalent to provinces with military forces.8 Witness also the notion that the principalities were “limbs and parts of the empire.”9 The British Empire in India was later to accomplish a similar semantic trick in integrating the numerous minor principalities within the loose embrace of the Raj.10
Regulated by treaties, such friendly kings were an economical way to rule territory and to extract tribute. However, they were not designed (or destined) to last, and the story of imperial expansion in the last century BC and first century AD is in large measure that of a sequence of client kingdoms being annexed by the Roman state when they had served their purpose or outlived their utility to Rome—notably, in the early Augustan and Flavian periods.11 The moment of annexation often coincided with the death of a friendly king and the decision of the Romans not to recognize a local successor. A number of friendly kings bequeathed their possessions to Rome in the event of dying without an heir, or sought Roman guardianship for their heirs.12 This was often linked to Roman protection against rivals or neighboring states, but it was also a tacit recognition of the realities of their treaties with Rome. From the Augustan age onward the position of heirs to a client kingdom was increasingly subject to explicit approval by the emperor.13 Many client rulers in waiting spent periods within the Roman Empire for education, and those who were ousted by internal coup or as a result of being forcibly retired by Rome could generally count on a pension and lands within the provincial territory.14 The relationships between Rome and friendly kings were thus essentially individual ones similar to the patron/client links between individual members of Rome’s elite classes.15
From this perspective the client kingdom existed primarily as a personal gift between the Roman state (or princeps) and an individual; it did not necessarily involve Roman recognition of the autonomous or semiautonomous existence of a particular people. The pragmatic nature of these relationships is well summed up by Edward Luttwak: “The client states needed constant management: unsatisfactory rulers had to be replaced … and successors had to be found for rulers who died.”16
However, the “economy of force” achieved by Rome in governing territory by proxy was often revealed to be a false economy in the medium term. Routine deployment of troops may have been avoided, but often at the cost of large-scale crisis management later. A recent analysis of the background to the Jewish revolt under Nero has highlighted the fact that the lengthy series of changing arrangements for the rule of Judaea under the Julio-Claudian emperors—varying from client kingdom to direct rule and back again—prevented consistency of “authority of power.” The arbitrary comings and goings of Roman nominated kings and imperial officials exacerbated fault lines in Jewish society, undermining traditional hierarchies and elites. Rome’s “long hesitation” about how best to govern the region thus contributed to the horrific savagery of revolt when it came.17
Recognition of individuals as client kings was not necessarily restricted to members of the local ruling group of families.18 A prince whose father’s kingdom had been annexed might be accorded a neighboring kingdom to rule—as in the case of Juba II, son of the last king of independent Numidia, who became king of Mauretania under Augustus in 25 BC.19 Forging marriage alliances with clients was also good politics for the empire, as the union of Juba II and Cleopatra, daughter of the last queen of Egypt, indicates.20
FIGURE 3.1 Gold coin of Augustus, featuring a (Germanic) barbarian chief offering up his child to the emperor. (RIC 1, 201a: Werner Forman / Art Resource.)
Figure 3.1 illustrates the theme of the offering of native princes as “hostages” for education in the Roman army and at the imperial court.21 This gold coin of Augustus encapsulates both the importance and the effect of power in an imperial society. A (German) barbarian leader offers up his own child as a hostage to the emperor. This presents an idealized image of the friendly king and the sacrifices expected of him in his relations with the imperial power.22 Here we see that power is not simply about dominance—though that is, of course, part of the significance of the scene. We are also witness to the quality of power to change behavior in both predictable and in less tangible ways. Roman power in this situation did not simply control; it acted as a creative force in the societies in contact with Rome, creating dynamic new possibilities for native princes and rulers. Yet, as we shall see, in encouraging expectations and ambitions among clients, there were also dangers to imperial stability.
The basic ground rules of client kingship as far as we can reconstruct them were:
1. Clients were formally bound to the empire by treaties and provided hostages and other sureties of good behavior.
2. All things being equal, Rome favored heirs of client kingdoms who were known quantities (often those who had spent time at Rome as hostages) and who had already espoused Roman values and culture to some degree.
3. The appointment of a client king required his formal acceptance at Rome; a fait accompli of a local strongman seizing power would result in conflict unless he could obtain the explicit authority of a Roman client.
4. Some client kings were able to secure the succession of their heirs, and this appears to have been most straightforward to achieve in periods when territorial expansion was limited (as late in the reign of Augustus or under Tiberius).
5. Client kingdoms were expected to make a financial and military contribution to the empire, as well as covering their own costs of governance.
6. Failure to abide by the terms of their treaties, unauthorized military actions against third parties or other resistance to Rome could result in the renunciation of a client and Roman annexation or replacement with another more pliable client ruler.
Above all, we need to recognize that friendly kings could require a high level of input from the empire. Despite its sometime tendency toward taking ad hoc decisions and to being reactive to events, the Roman Empire required a more proactive basis to the relationship with its clients.
REGIME CHANGE AND THE ROMAN EMPIRE
Does an invasion need to be hostile to the inhabitants of a country being invaded? Modern experience of United Nations actions, for example, should tell us that such actions can have an altruistic aim.23
The reasons for regime change were no doubt varied, but we need to be particularly skeptical of a model that sees the Roman Empire as analogous to the United Nations. There has been something of a revival of views in recent years about the altruistic or disinterested aims of the Roman Empire, as exemplified by Martin Henig’s comment above, and a prime goal of this paper is to argue that this is largely the result of uncritical reasoning.
Unsurprisingly, the historical winner (that is, Rome) tended to present a negative image of any client rulers that it deposed. There is no doubt that Roman history features many major ‘rebels’ who had previously enjoyed a close relationship with Rome (Jugurtha of Numidia, Arminius the German, Boudica the Briton).24 It is easy to get carried along by the imperial rhetoric concerning the character flaws of these individuals, and so to conclude that their resistance was due to deep-seated anti-Roman sentiment. Yet many of the greatest rebels against Rome had a previous history of close relations and their resistance thus needs explaining in the context of changing relations with the imperial power. In several cases one can see that military conflict resulted from Rome’s decision to annex client states or to effect regime change rather than being due to initial disloyalty by the client.
Just as the British and American justifications of their pursuit of regime change in Iraq have been shown to be based on shallow and “sexed up” claims,25 so our ancient sources on similar Roman interventions are also presumptively suspect. There is, of course, no doubt that Saddam Hussein was a monster, but it was as a client of the West that he became so. For example, the main period of Iraqi use of weapons of mass destruction, in the war with Iran and against the Kurds, occurred at the time of Saddam’s closest rapprochement with the United States. The moral paradox here simply emphasizes the political contingency of empires.
In the same way, we need to look at Rome’s relations with clients and native rulers on the periphery of its empire with a more critical eye. Threats to the security of empire could on occasion be ascribed to anti-Roman sentiment, but in a number of important cases we should consider the crises to have arisen as a result of sudden breakdowns in what had formerly been close rapprochment with client rulers. Personalities aside, it is possible that complications arising out of the processes of regime change and power transfer represent a systemic weakness of empires.
THE KINGDOMS OF CUNOBELIN AND VERICA AND THE ROMAN INVASION OF BRITAIN
The political history of Britain between the campaigns of Caesar in 55–54 BC and the Claudian invasion of AD 43 is shadowy in the extreme.26 Many accounts of the period start with disclaimers about the extent to which it is possible to construct conventional narrative.27 Yet out of the diaphanous factual fabric much has been manufactured. From the coin series issued by late Iron Age peoples, numismatists have endeavored to trace dynastic links between named individuals, constructing plausible but unprovable genealogies for some of the major peoples.28 That has not stopped historians and archaeologists from attempting to flesh out the biographies of these ghostly individuals.29 In reality the factual basis of many conventional accounts is illusory and rests heavily on some basic but untested assumptions.30 Roman and Iron Age experts alike have a habit of adopting without question the idiom of the classical sources (and modern colonial regimes) in talking of “tribes,” “civilization,” “barbarians,” and so on.31 Most modern accounts are punctuated by such stereotypes—for instance, a repeated insistence on “Catuvellaunian aggression” against neighbors and Romans.32
FIGURE 3.2 Map of Britain on the eve of the Claudian invasion of AD 43, showing approximate territories of the eastern and southern kingdoms.
Iron Age specialists tend to take a more archaeological approach but are themselves somewhat wedded to the small scraps of information provided by the ancient literary sources and coin series.33 Much recent work on late Iron Age Britain has focused on the delineation of significant differences between southeast England and the rest of the British archipelago.34 It is now widely recognized that, at the time of the Claudian invasion of AD 43, the southeastern region was dominated by two major kingdoms and that these had (or had previously had) some sort of formal client relationship with Rome (see fig. 3.2).35 There is similarity here with the situation observed in other colonial contexts, wherein an expanding empire has a considerable impact on peripheral societies, leading to the emergence of higher forms of statehood.36 The “eastern kingdom,” with its main proto-urban centers (oppida) north of the Thames at Colchester (Camulodunon) and St. Albans (Verlamion), is often associated with the Trinovantes and Catuvellauni people, but is perhaps better conceived as the kingdom of the House of Cunobelin. The southern kingdom, with known centers at Silchester (Calleva) and Chichester, is generally associated with Atrebates, but again may more accurately be considered the House of Verica. These sophisticated states were built on the personal power of their ruling dynasties and the close ties the latter had with the Roman Empire.37
The ostensible reason for the Claudian invasion according to Cassius Dio was the flight across the Channel of Verica (or Berikos) as a result of an uprising in Britain;38 he apparently persuaded Claudius to send an expedition. As noted already, Verica appears to have been king of a substantial territory south of the Thames around AD 10–42. There is an interesting symmetry with the account of Caligula’s planned campaign of AD 40, when another refugee British prince was received by the emperor. Adminius was a prince of the ruling house of the eastern kingdom of Cunobelin to the north of the Thames.39 The consensus view of these events has tended to be that the expansionism and profound anti-Roman tendencies of the eastern kingdom had got out of control, exacerbated by their annexation of neighboring territory (some commentators anachronistically and hyperbolically write of a growing “Catuvellaunian empire”).40 The flight of Verica and Adminius are seen as representing the expulsion of recognized Roman clients or of pro-Roman elements within the court of Cunobelin. Henig gives a particularly lopsided view of events, taking the Roman sources on trust and amplifying their innuendos:
When, at the end of Cunobelin’s reign, his philo-Roman son Adminius, was driven out of Britain, to surrender himself to Gaius Caligula … it might appear that the situation was very different from that of 30 years before when Tincomarus was expelled from the southern kingdom. What had changed was that Cunobelin’s other sons, Caratacus and Togodumnus, were bent on continuing an aggressive policy and building up a unified realm in the south-east. … Claudius could only nod thoughtfully, realising that the situation in Britain was a potential threat to the security of the neighbouring parts of Gaul.41
The same tendency to embellish the threadbare evidence is demonstrated also by John Wacher, who writes,“In [Cunobelin’s] declining years … power was increasingly transferred to his sons. One of these Adminius, held pro-Roman views … two others … Togodumnus and Caratacus revived the indifference towards Rome. By the time Gaius had succeeded Tiberius … this pair had sufficient authority over their father to cause Adminius’ expulsion.”42
Although blamed by most modern commentators on the expansionism of the eastern kingdom, the flight of Verica to Rome in about AD 42 was actually attributed by Dio to “internal unrest” within his southern kingdom.43 Modern textbooks generally present this as the final straw for Rome in its deteriorating relationship with a rogue state. This puts all the responsibility for the Roman invasion of Britain on the native princes, with Henig—among others—claiming that the restoration of the House of Verica in the southern kingdom was a prime objective of the invasion in AD 43. As we shall see, there is no doubting that the initial and prime target of the invasion was the annexation of the eastern kingdom. The fundamental assumption in most modern accounts concerns the presumed anti-Roman outlook of Cunobelin and his sons Caratacus and Togodumnus.44 But how did the breakdown in relations come about, and was Rome blameless in this situation? What if the depiction of the House of Cunobelin as fundamentally and unwaveringly anti-Roman was a deliberate falsehood?
Despite the lack of more explicit testimony relating to the events of the late AD 30s and early 40s, I think some suggestions can be made for an alternative narrative to that most widely disseminated today. I suggest that the transformation from friendly king to “rogue state” was late, sudden, and precipitated by unilateral Roman actions.
My understanding of the Roman conquest is built on a series of observations and suppositions that I shall acknowledge at the outset, though I think each can be justified:
1. The status of the eastern and southern kingdoms was as formal client kingdoms of Rome.
2. The increasing age of Cunobelin and Verica by the late AD 30s (both had been in power since about AD 10) was likely to have focused attention on the issue of political succession in both kingdoms.
3. The death of Tiberius in AD 37 marked an end of a nonexpansionist era in Roman foreign relations and put Roman territorial annexation on the agenda.
4. Roman actions provoked the crisis in relations; Gaius (Caligula) announced Rome’s new intentions by assembling ships and troops for a British campaign at Boulogne, and building a lighthouse there.
5. Verica and Adminius may both have been pro-Roman voices in their respective kingdoms, but the fact that neither was restored as a client king postconquest suggests that they were useful pretexts rather than part of Rome’s plans.
In the rest of this chapter I shall consider these key terms one by one.
CLIENT KINGDOMS
In the period between 30 BC and AD 43 a wide range of innovations appeared in the Iron Age coinage produced in southeastern Britain, heralding the emergence of two major kingdoms.45 The use of Latin legends and the inclusion of imagery drawn from the visual language of imperial power are particularly interesting developments. We know that Caesar demanded hostages (obsides) and tribute payments from the British rulers as well as guarantees from them regarding their future conduct and acknowledgment of Rome’s suzerainty.46 It has sometimes been considered unlikely that such arrangements endured for long in the prolonged period of civil wars and internal strife that beset the Roman Empire.47 The coinage evidence as interpreted by John Creighton suggests a different picture—that parts of Britain were effectively considered as integrated client states within the empire. Archaeological evidence is also starting to accumulate to suggest a much higher engagement with Roman material culture and continental innovations in social and religious behavior.48
The absence of these British obsides from their homelands could be prolonged. They established high level contacts in Roman society, they were very often given a period of service as officers in the Roman army, and they were firsthand witnesses of imperial politics at the center of power. The major innovators in the Iron Age coinages in Britain were thus most likely princes who had spent a period of time in Rome—or, at any rate, inside the empire. The seeming coincidence of approximate accession dates for both Cunobelin and Verica around AD 10 might hint at a late Augustan revisiting of the British situation and installation of two men with a particularly good exposure to and understanding of the new visual vocabulary of power at Rome.49 However, one can also point to the earlier installation of Tincomarus and Tasciovanus around 30–25 BC as equally critical moments when new iconographic programs appeared in the British coinage, heralding the initiation of two long-lived reigns. The Roman sources highlight the importance of these relationships in the reign of Augustus.50
There is significant overlap between some of the imagery employed by British client rulers such as Tincomarus of the southern kingdom and that on the coins of Juba II and his son Ptolemy of Mauretania. The similarities among coinages of a number of Augustan clients are so close as to suggest at the very least a keen awareness of what other client rulers were doing.51
SUCCESSION
The dynamics of this contact situation were inherently unstable, as a political settlement based on individual power and agreements with Rome required renegotiation and renewal every time leadership changed, whether as a result of the incumbent’s death or his overthrow. The transfer or prolongation of power in the next generation of a friendly kingdom could make headline news at Rome, especially if the emperor’s preferred candidates were not immediately accepted—as may have happened in Britain around 27–26 BC.52 Client kingship was something that had to be agreed by Rome, so anyone seeking to establish their own rule, no matter what their support in Britain, would have to seek recognition from Rome.53 In succession matters, Rome may often have had a favored candidate, though it is clear that the diplomatic and military situation did not always resolve itself to Rome’s liking, and we know of at least two British rulers who fled to Rome under Augustus in addition to those on the eve of the Claudian invasion.54 The point is that such upheavals are most likely to have occurred at moments of succession to the two major client kingdoms that Rome recognized in Britain rather than being a continuous problem.
In a similar way, a famous series of inscriptions from the forum area at Volubilis in Mauretania refer to “altars of peace” being erected at colloquia between the Roman governor and the chiefs of one of the leading peoples of the Atlas Mountains, the Baquates.55 Although at one time interpreted as evidence for recurrent warfare between Rome and hostile forces on its Moroccan frontier, more recent appraisals have tended to conclude that these represented periodic reaffirmations of a client relationship with successive chieftains.56
What was the relationship between the two British client kingdoms? Where is the evidence to suggest that one was attacking the other in the time leading up to the Claudian invasion? Overlapping with the latter stages of Verica’s coinage is the series produced for another ruler called Epaticcus, centered on the northern part of Verica’s zone of influence.57 The iconography of the coins follows that of the eastern kingdom, though the denominations match the normal pattern of the southern kingdom, with gold and silver but no bronze issues. Epaticcus thus signaled a dynastic link to the ruling group of the eastern kingdom. This has often been presented as the prime evidence that the eastern kingdom was expanding in defiance of Roman wishes at the expense of its southern rival, provoking Verica’s fall.58 There are alternative scenarios to consider. If the eastern kingdom had indeed become predominant over the southern, there is a possibility that Rome had given explicit encouragement and support for this, perhaps favoring a situation where it dealt with a preeminent kingdom rather than maintain the potentially tricky balancing act of supporting two rival powers equally. It is also entirely feasible (and perhaps more likely in terms of Roman diplomatic management of its clients) that Epaticcus appeared as ruler at Calleva as a result of an arranged dynastic marriage brokered by Rome; there are certainly many other examples of Rome arranging such matches between neighboring client states. It is surely significant that the few coin issues ascribed to Caratacus, who may have succeeded Epaticcus in this territory just ahead of the Roman invasion of AD 43, continued Roman-style themes, perhaps still in the hope that Rome would recognize his legitimacy as a client ruler.59 It is thus by no means certain that the fall of the House of Verica was the equivalent of Saddam’s invasion of Kuwait in provoking an imperial military response.60
The reign of Cunobelin marks the apogee of the late Iron Age kingdoms in Britain.61 His coinage is estimated to have exceeded more than a million struck pieces, and dwarfs all the other pre-Roman kingdoms. He appears to have ruled over a wide region with Roman support and approbation. The Roman historian Suetonius described Cunobelin as rex Britannorum (king of the Britons), which implies that he ruled over a large group of British peoples with imperial blessing.62 His client status is also indicated by the use of the rex legend on his coins.63
Cunobelin’s coinage was the most classical of all the kingdoms. The proliferation and range of the imagery employed overall is astonishing. By identifying with the iconographic program of Augustus and the reinvention of Roman culture that took place in the Augustan age, the British dynasts both proclaimed their allegiance to Augustus and constructed their own identity as powerful rulers.64 The coins reflect the close integration of these kingdoms within the Roman world; the complexity of the imagery would have made the message of the coins wholly accessible to a relatively small group in society or perhaps had been consciously designed with an external audience in mind. Indeed, some of the Cunobelin and Verica issues feature what are clearly imperial portraits of Augustus, Tiberius, or Gaius and the voluntary inclusion of such images must indicate an expression of allegiance.65
Succession issues are also indicated by the use of F or FIL, for filii (“son of”). These coins were a way of showing a sense of belonging to the Roman world and of seeking legitimation of power and authority by emulation of programmatic Roman imagery and wordplay. In presenting real or imagined dynastic links to other rulers, the kings reinforced their own authority. It is conventional to take literally the family links implied by the coin legends, but while it is not implausible that Cunobelin, for instance, was the son of Tasciovanus, we should be cautious. After all, if the Roman Empire was their model, the relationship of Octavian to Caesar (or of a succession of presumptive heirs to Augustus) was that of adoptive son, not blood relation. Perhaps it is safer to assume that the claims of filiation on the coins were a legitimation device indicating “of the family of” or “descendant of,” following good Roman precedents.66
Given that Rome often wished to play kingmaker in succession issues, it was by no means certain that a client kingdom would have passed in a simple father-to-son pattern. The expression of such dynastic links may also have articulated the hope that the wishes of the present incumbent would be given due weight by Rome.
Client rulers were controlled and manipulated in a variety of ways, but interference in the succession process was probably the most significant and contentious area.67 It was clearly in Rome’s interests when a client king died to have him replaced by one of the educated hostages from the pool held at Rome. But hostages were sometimes at a disadvantage in terms of not always being on the spot when the ruler died, and were also potentially out of touch with the bulk of the people within the kingdom. To counterbalance these disadvantages, Rome made certain requirements of client rulers about their succession: that arrangements were openly discussed with Rome and the nominated successor approved by Rome, new client kings were expected to travel to Rome to be formally recognized, and so on. It was an exceptional case, as with Herod, for the emperor to grant a client king the privilege of appointing his own successor.68
The system did not always work smoothly, of course, and Rome could, if necessary, threaten annexation against client states keen to appoint a king other than a candidate endorsed by Rome. There might be a lengthy interregnum while the issue of succession was sorted out to the satisfaction of all.
ANNEXATION
The conquest of AD 43 saw the eradication of the eastern kingdom and the start of wider campaigning in Britain, coupled with the recognition of several new client kings. There was no return for Verica, the southern kingdom being handed over (though perhaps only after an interval of several years) to a certain Togidubnus, while new clients were recognized as rulers of the Iceni and Brigantes.
The identity of Togidubnus (or Cogidubnus, as some books still present him on the basis of the less likely version of his name in the classical sources) is a conundrum. It has generally been accepted that he was a British prince who happened to have a rather similar name to Togodumnus, the son of Cunobelin. The conventional view, on Dio’s apparent testimony, is that Togodumnus was killed in battle early in the AD 43 campaign. However, as I have hinted at in Imperial Possession, Dio may have misinterpreted his source in reporting the death of Togodumnus, an idea now convincingly developed by John Hind to suggest that far from being killed in battle, Togodumnus was effectively “turned” by Rome and persuaded to collaborate.69 This clears the way for correlating Togodumnus with Togidubnus, the client king postconquest. The possibility is very attractive, and is consistent with what we know of Roman use of clients elsewhere, while also highlighting the contrasting fates of the two premier princes/kings who led the opposition to the Roman invasion in AD 43. While Togodumnus was potentially persuaded to collaborate, his brother Caratacus fought on resolutely, eventually to be paraded at the triumph of Claudius in Rome.70
The kingdom assigned to Togodumnus/Togidubnus in the later AD 40s was eventually annexed peacefully (probably quite early in the reign of Vespasian), but some of the worst military problems in the first century AD attended the termination of other British client kingdoms, that of Prasutagus of the Iceni and Cartimandua of the Brigantes. The Boudican revolt erupted when the Romans overrode the wishes of the Iceni and annexed the territory of the client state after the death of king Prasutagas, ignoring his will that had attempted to leave only half his territory to Nero and the rest to his daughters.71 The Brigantian problem is presented in the sources as the fall out from a sordid palace scandal, when Queen Cartimandua replaced her husband Venutius in her bed with his armor bearer, Vellocatus.72 Venutius is sometimes presented as fundamentally anti-Roman, but he had in fact been described as the British leader after Caratacus who “excelled in military skill. … For a long time he was loyal and enjoyed the protection of Roman arms.”73 The marital discord seemingly opened up fault lines within Brigantian society, perhaps focusing on different readings of Rome’s long-term intentions in northern Britain.74
So whose fault was the invasion of AD 43? Was it a catastrophic miscalculation by the ruling family of the eastern kingdom, or was it the result of political duplicity on the part of Rome, turning on a long-term ally? We shall never have the evidence to answer all our questions about the underlying story of Cunobelin’s sons: Caratacus and Togodumnus who resisted the invasion (though with Togodumnus possibly being persuaded to switch allegiance at an early stage), Amminius (or Adminius) who was forced to flee to Gaius Caligula after a quarrel with his father, and unnamed other brothers captured by Ostorius Scapula in AD 52.75 However, it is surely significant that following his final defeat and capture by Ostorius Scapula after almost ten years of warfare, Caratacus was spared from execution after being paraded in Claudius’s triumph.76
Tacitus implied it was the oratorical skills of Caratacus that gained his reprieve (and likely pensioned retirement within the empire), but if he had been a long-term friend of Rome prior to AD 43 there may have been an element of bad conscience behind Claudius’s decision.
My reconstruction of events thus differs from the consensus view in putting greater responsibility on Roman imperial politics rather than on intransigence and unilateral action on the part of the House of Cunobelin. The territory under the influence of the eastern dynasty was exceptionally large—as already noted, Suetonius referred to Cunobelin as “king of the Britons,”77 implying control of several distinct peoples in southern Britain. I have argued herein that there is actually no evidence to suggest that the expansion had been attained in defiance of Rome’s wishes. Cunobelin had a long and prosperous reign as a friend of Rome. But he chose a bad moment to die and to create a succession issue, at a time when two new emperors came to the throne in Rome in rapid succession, each needing some quick military glory to establish their popular reputation. Client kingdoms, no matter whether loyal or not, were an attractive target in such circumstances, as if handled well, the glory and booty could be had fairly cheaply and quickly become victories, provided that some at least of the native elite order could be persuaded that their interests would be best served by compliance. Again, the decision to reward Togodumnus/Togidubnus with a new client kingdom south if the Thames, if the correlation of the names is correct, is much more plausible if Cunobelin and his sons were not directly responsible for provoking a crisis that necessitated Roman intervention.
It is thus an interesting paradox that annexation, when it came, should focus on the British client kingdoms that had been most closely aligned with Rome in the period between the campaigns of Caesar and Claudius. The accession of Caligula in AD 37 created the need for military glory to bolster his prestige. The first result of this political pragmatism was the execution of the king of Mauretania, Ptolemy. Mauretania erupted in revolt and Claudius was still mopping up the resistance in the early years of his reign.78 Under the rule of the antiexpansionist Tiberius, the succession of Cunobelin’s sons may well have been agreed in principle, but trouble and suspicion would have flared up quickly if, following the accession of Gaius, Roman missives suddenly become evasive on the issue of who was the approved heir. Cunobelin was evidently dead by the time Plautius landed with his invasion force in AD 43, with Caratacus and Togodumnus now cast in the role of rebels. Given Rome’s invasion planning since at least AD 39, it is questionable whether Rome left them much of an alternative. Annexation here looks to have been a unilaterally “done deal.”
ACTIONS
The evident deterioration in relations between the House of Cunobelin and Rome in the late AD 30s and early 40s was thus arguably a direct consequence of the realization in Britain that Rome intended to annex the kingdom on Cunobelin’s death. Our sources imply that it was the British kings who became difficult and broke agreements, but it is at least as likely that it was the Romans who tangibly instituted a change in policy and whose actions preparing for the invasions sparked a crisis before Cunobelin was even dead. Although the detail of the abortive campaign under Gaius Caligula is fuzzy, there is little doubt that the construction of a lighthouse and improvements to the harbor at Boulogne, the assembly of an invasion fleet and the massing of troops on the Channel coast was an unambiguous threat that signaled a fundamental change in relations between the superpower and its British satellites.79 The flight of another son of Cunobelin, Adminius, to Rome was perhaps a consequence of Roman attempts to convince some in the ruling elite to collaborate. It is likely that the same policy may have also affected the southern kingdom of Verica, or have had repercussions for him if he was perceived as accepting a Roman takeover as inevitable, since he was forcibly ousted just prior to the conquest of AD 43.
Adminius and Verica were used as pretexts for invasion in AD 40 and 43, respectively, but significantly neither appears to have been returned to power as a Roman client once the invasion was launched. Although in the event only the eastern kingdom was directly annexed in AD 43, with the southern kingdom being returned to the care of a new client ruler, one senses that realpolitik lay behind Rome’s actions. Recent attempts to suggest that the Roman conquest was invited and encouraged by a large percentage of the Atrebates and other peoples threatened by the eastern kingdom’s expansion seem to be far too accepting of a Roman perspective of these events.80 If the new client king over the southern kingdom was in fact Togodumnus, son of Cunobelin and prince of the eastern kingdom, the events of the AD 40s appear in an altogether different light. Faced with a strong British force in the early stages of the invasion, the persuasion of one of the sons of Cunobelin to revert to a position of friendship to Rome—with the cynical promise one imagines of a client kingdom to follow—was a classic example of the pragmatic “divide and rule” mentality of imperial expansion.
The sequel to the invasion of AD 43 was far more than the annexation of a single client kingdom, in that a larger project for the conquest of the British archipelago was rapidly launched. The economic opportunities offered by newly annexed territories, in terms of control of natural resources and construction and supply contracts, can be readily paralleled in other imperial contexts. I remain dubious about the extent to which imperial grievances with the British client kings were anything more than a useful excuse for the habitual Roman pattern of exploitation and domination.
The origins of this chapter lie in an unpublished paper presented at the 2005 Roman Archaeology Conference held in Birmingham, England. I am grateful to Richard Hingley for the invitation to participate in his session on imperialism, and for his comments on the paper that have enabled me to develop and expand on it for publication here.
1 This chapter was drafted in the direct aftermath of the U.S.-led “regime change” in Iraq.
2 Power has been a key concept in the attempts of international relations specialists to make sense of the Roman world; see Fitzpatrick 2005; and James 2006, 141–49.
3 Said 1986, 151; Sarup 1993.
4 Strabo, Geography 17.3.24.
5 Luttwak 1976, 13–40, remains an interesting and strategic slant on hegemonic rule.
6 Braund 1984 is still the standard study; see also Braund 1988.
7 Braund 1984, 23–37.
8 Tacitus, Annals, 4.5; Luttwak 1976, 30.
9 Suetonius, Augustus, 48, membra partesque imperii; Braund 1996, 88.
10 See Ferguson 2004, 172–73, for a map of British India and the native princely states, and 208–12 on their exploitation by British imperial administration.
11 Braund 1984, 165–80.
12 Ibid., 129–64.
13 Braund 1996, 85–89; Creighton 2000, 169–70.
14 Allen 2006; Braund 1984, 9–21, 165–74; Creighton 2000, 89–94.
15 Wallace-Hadrill 1989a.
16 Luttwak 1976, 39.
17 Curran 2005.
18 Jacobson 2001.
19 Plutarch, Caesar, 55.2.
20 Dio, Roman History, 51.15.6; Jacobson 2001, 25.
21 I must thank Bruce Hitchner for suggesting this wonderful image for all the publicity for the lecture series. It is also illustrated and discussed in Roymans 2004, 243.
22 Braund 1984.
23 Henig 2002, 38.
24 See Wells 2003, 105–24; Hingley and Unwin 2005, 41–61.
25 See Runciman 2004 on the Hutton and Butler reports.
26 See Mattingly 2006c, 47–84, for my overview of the period.
27 Braund 1996, 67–69.
28 Van Arsdell 1989 is a good example of this tendency to push genealogy and chronology to the limits. For changing views on the significance of so-called Celtic coinage, see Aarts 2005; Cunliffe 1981; Haselgrove 1987; Haselgrove and Wigg-Wolf 2005.
29 Peddie 1987, 15–22, is typical of many.
30 Todd 1999, 17–54, and Wacher 1998, 13–18, make the best of the material, but the approach seems flawed to me.
31 See the discussion in Mattingly 2006c, 26–38.
32 Wacher 1998, 14. The axiom is dubious on two counts: first, the “aggression” may be simply Rome’s post facto justification for invading rather than a reality, and second, the conceptualization of a “Catuvellaunian tribe” is at odds with the impression that power in preconquest southern Britain was vested more in individuals with whom Rome had a special relationship than in peoples per se. It is reasonable to infer that there were people called
Catuvellauni who were under the rule of Cunobelin, but they were one of several regional populations associated with his kingdom and it is by no means certain even that they were a dominant group within the kingdom, as is often assumed.
33 See Cunliffe 2005, 149–77, for a recent synthesis (and cf. Cunliffe 1991); Pryor 2003, 433–44, weakens his brilliant revisionist approach by taking too much of the source evidence on trust. The papers in Haselgrove and Moore 2007 offer the most authoritative synthesis of current thinking on late Iron Age society.
34 Champion and Collis 1996; Gwilt and Haselgrove 1997; Haselgrove 1999, 130–33; 2004; Mattingly 2006c, 68–84.
35 Creighton 2000 is the best account.
36 Whitehead 1992.
37 Creighton 2006, 19–31.
38 Dio, Roman History, 60.19.1–2.
39 Suetonius, Caligula, 44.2.
40 Braund 1996, 101; Wacher 1998, 15.
41 Henig 2002, 33–34.
42 Wacher 1998, 15 (all highly speculative and accepting at face value the hints of motive/justification in Roman accounts of the invasion).
43 Dio, Roman History, 60.19.1.
44 The pair are described as “hot-headed” in Wacher 1998, 15, but on what grounds?
45 Here again Creighton 2000 is a fundamental study; see also Braund 1996, 76–90, for the historical background.
46 Cicero, Letters to Atticus, 4.18.5; Strabo, Geography, 4.5.3. On the role of obsides in Roman “empire management,” see Allen 2006.
47 Frere 1974, 39–45, and Salway 1981, 48, for instance, suggest that there was some hiatus before Augustus renewed relations around 16 BC.
48 Creighton 2006, 35–45; Manley and Rudkin 2005.
49 Creighton 2000, 95–125.
50 Strabo, Geography, 4.5.3.
51 Creighton 2000, 118–24.
52 Dio, Roman History, 53.22, 53.25.
53 Braund 1984, 24–27; 1996, 84–86.
54 Augustus, Res Gestae, 6.32; Suetonius, Caligula, 44.2.
55 IAM 2.348–350, 356–61, 384, 402.
56 For the traditional interpretation, see Frezouls 1957 and 1980. Cf. Mattingly 1992; Shaw 1987.
57 Creighton 2000, 104–5, 111–12.
58 See, among others, Salway 1981, 70.
59 Creighton 2000, 219; 2006, 30.
60 We might note, however, that Saddam mistakenly thought he had been given the nod by the U.S. ambassador ahead of sending in his troops.
61 Cunliffe 1981, 37–39, 82–84; Creighton 2000, 109–11.
62 Suetonius, Caligula, 44.2.
63 This runs contra the views of Salway 1981, 56.
64 Creighton 2000, esp. 80–125.
65 Creighton 2000, 176–88.
66 This has been argued already in Creighton 2000, 170–73.
67 Braund 1984, 24–26.
68 Josephus, Jewish Antiquities, 15.343, 16.92; Braund 1984, 26–27.
69 Dio, Roman History, 60.21.1; see Mattingly 2006c, 100, and Hind 2007, 96–100, for detailed discussion and proposed reappraisal of Dio’s text and Togodumnus’s presumed death.
70 Hind 2007, 99–100.
71 Hingley and Unwin 2005, 43–47; Mattingly 2006c, 106–8.
72 Tacitus, Histories, 3.45.
73 Tacitus, Annals, 12.40.
74 Creighton 2006, 33–34.
75 Tacitus, Annals, 12.35–36.
76 Ibid., 12.36–37.
77 Suetonius, Caligula, 44.2.
78 Ibid., 46; Dio, Roman History, 59.25.
79 Dio, Roman History, 59.25.1–3; Suetonius, Caligula, 44.2, 46.1; Todd 2004a, 44–46.
80 Henig 2002.
