I INTRODUCTION
In the Res Gestae, Augustus takes pride in his achievements at the expense of kings of foreign gentes: leading nine kings and their children in triumph (4.3); being able to impose kings on the people of Armenia Major (27.2); receiving embassies, refugees and hostages from kings, and sending kings to gentes upon their request (31–3). Importantly, as presented in the Res Gestae, all eastern kings with whom Roman relations continue after the victory at Actium are from far-away lands, east of the Euphrates. By contrast, Augustus claims that he ‘regained’ the lands west of the Euphrates, the Greek-speaking provinces east of the Adriatic, from the kings who held them (i.e. Mark Antony and his coalition).Footnote 1
This dichotomy is both misleading and instructive. Misleading, since Augustus pardoned most of the Anatolian and Levantine monarchs from Antony's coalition and allowed them to stay in power. Moreover, even in the case of monarchs who were executed in the aftermath of Actium, Augustus reinstated their family members during his visit to the east in 20 b.c.e.Footnote 2 Instructive, since, as will be demonstrated below, although Augustus retained the same dynastic families that had ruled in the Roman east in the late Republic, their relations with the Roman centre were now fundamentally changed.
In this paper I will propose a new model to understand the relations between the imperial centre and these eastern monarchs, consisting primarily of the dynasties of Thrace, Pontus, Cappadocia, Commagene, Emesa and Judaea. I will argue that Augustus transformed this specific group of ‘friendly kings’ into a new kind of Roman administrators, which I term ‘provincial monarchs’. These monarchs formed a cohesive elite whose behaviour will be shown to be meaningfully distinct from both their late republican predecessors and contemporary kings east of the Euphrates. Continued reliance on Braund's fundamental (but far broader) model of the ‘friendly king’, combined with a recent trend towards individual case-studies and a focus on relations between kings and their subjects, have thus far obscured the specificity of this group, and their role within Augustus’ broader agenda.Footnote 3
The proposed model challenges the common view of the Julio-Claudian east as a confusing amalgamation of ad hoc arrangements, administrative fluidity and lingering power structures from the Hellenistic and late republican period.Footnote 4 Rather, I will argue that the flexibility and diversity in administrative forms, especially the division of territories between rule by imperial legates and procurators on the one hand and provincial monarchs on the other, should be considered primarily as a novelty of the Principate, a hitherto overlooked aspect of the ‘Augustan Revolution’.
I will suggest that both ‘direct’ and ‘indirect’ rule in the Julio-Claudian east were equally rooted in Augustus’ constitutional and administrative experiments in the 20s b.c.e. Specifically, I will demonstrate how, starting in these years, Augustus took several complementary and unprecedented steps to change Rome's relations fundamentally with eastern provincial monarchs, allowing Augustus and his heirs subsequently to employ them as a distinct and novel type of administrator.Footnote 5 Provincial monarchs were raised to be personally loyal to the domus Augusta and expected to form a check on the potentially subversive power of the senatorial imperial legates, while the latter in turn formed a check on the monarchs.Footnote 6 While certain developments find parallels in the west, the fundamentally different provincial geopolitics, frontier challenges and cultural heritage of the western regions led Augustus to find other solutions there.Footnote 7
The model will then be used to offer a fresh explanation for the changes in Roman imperialism in the east under the Flavians. Rather than looking outward to ‘frontier policy’ or ‘grand strategy’,Footnote 8 the proposed model suggests that personal relations and concerns with legitimation in Rome may have been primary factors in the changes to provincial administration. The political expedience of provincial monarchy, as well as the socialising mechanisms inherent within it, led the later Julio-Claudians to maintain this system by proactively appointing new monarchs. The aid of provincial monarchs in the rise of the Flavians, although demonstrating the continued importance and vitality of the system, also proved its undoing. The new emperors did not have the same social and political incentives to appoint new monarchs as their predecessors. Instead, they may have come to realise that close relations with provincial monarchs hindered their attempts to legitimate their hold on power, leading to a gradual phasing-out of the system.
II THE CREATION OF PROVINCIAL MONARCHY
As noted by Fergus Millar, Strabo, a contemporary of Augustus, reports that kings and other dynasts were included in the imperial provinces granted to Augustus in 27 b.c.e.Footnote 9 Yet it would be a mistake to infer that all allied kings of Rome were likewise considered part of the imperial provinces.Footnote 10 Describing the contemporary state of the empire in Book 6, Strabo in fact contrasts the land within the Phasis and Euphrates rivers, which he claims was under the power of Romans and dynasts appointed by the Romans, to the lands beyond, including the peoples of Armenia (Major), Iberia and Albania, whom the Romans needed to constantly keep in check:
ἅπαντα τὰ ἐντὸς Φάσιδος καὶ Εὐφράτου πλὴν Ἀράβων τινῶν ὑπὸ Ῥωμαίοις ἐστὶ καὶ τοῖς ὑπ' ἐκείνων ἀποδειχθεῖσι δυνάσταις. Ἀρμένιοι δὲ καὶ οἱ ὑπερκείμενοι τῆς Κολχίδος Ἀλβανοί τε καὶ Ἴβηρες παρουσίας δέονται μόνον τῶν ἡγησομένων, καλῶς δὲ κρατοῦνται· νεωτερίζουσι δὲ διὰ τὰς τῶν Ῥωμαίων ἀπασχολίας.
All the peoples within the Phasis and Euphrates, except certain of the Arabians, are under the power of the Romans and of the dynasts appointed by them. The Armenians and those lying above Kolchis (the Albanians and Iberians) need only the presence of leaders, and act well when ruled; but they revolt when the Romans are otherwise occupied.Footnote 11
Tacitus likewise appears to have been aware of the distinction. Describing the state of the empire in 23 c.e., he contrasts all of Syria ‘up to the River Euphrates,’ which was controlled by Roman legions (and auxiliaries), with Iberia, Albania and their neighbouring kingdoms, who were passively ‘protected against external empires’ by the Romans:
cetera Africae per duas legiones parique numero Aegyptus, dehinc initio ab Syriae usque ad flumen Euphraten, quantum ingenti terrarum sinu ambitur, quattuor legionibus coercita, accolis Hibero Albanoque et aliis regibus, qui magnitudine nostra proteguntur aduersum externa imperia.
The rest of Africa was controlled by two legions, and Egypt by a similar number, and the whole area thereafter, which is encompassed within the mighty arc of land from the start of Syria right up to the River Euphrates, by only four legions, its neighbours being the Iberian, Albanian, and other kings who are protected by our greatness against external empires.Footnote 12
By contrast, in the following sentence Tacitus presents the monarchs in Thrace as actively guarding a part of the empire, just as the Roman legions do in Pannonia and Moesia: ‘et Thraeciam Rhoemetalces ac liberi Cotyis, ripamque Danuuii legionum duae in Pannonia, duae in Moesia attinebant’ (‘Rhoemetalces and the children of Cotys were holding Thrace, while two legions in Pannonia and two in Moesia [were holding] the bank of the Danube’).Footnote 13 We should note that the Thracian monarchs share with the Danube legions the active verb of the sentence: each of these were holding a different part of the empire.Footnote 14 The typical Tacitean pithiness here contrasts with his choice to add the relative clause to the end of the previous sentence. Tacitus clearly wishes to articulate the relationship of the trans-Euphratean kings to Rome as different from the full integration alluded to in his treatment of the Thracian dynasty in parallel to the legions, as can likewise be inferred from the fact that he completely glosses over the contemporary presence of monarchs (with their armies) in Asia Minor and Syria.Footnote 15
It would appear that Strabo and Tacitus were aware that under the early Principate monarchs in the eastern provinces became more meaningfully integrated into the empire, in ways that their neighbours east of the Euphrates were not. We can in fact trace multiple systematic steps taken by Augustus to bring about this shift, which turned provincial monarchs into a new kind of administrative class, distinct both from their late republican predecessors and from kings and dynasts east of the Euphrates. First, as demonstrated by Raggi, Augustus appears systematically to have granted citizenship to Eastern monarchs who hitherto lacked it, including the rulers of Emesa, Thrace, Commagene and the Tarcondimotids of Cilicia. This unprecedented step (there are no known cases of ‘friendly kings’ receiving citizenship before Augustus), which Augustus probably took during his visit to the east in 20 b.c.e., stands in remarkable contrast to Augustus’ reportedly sparing attitude towards citizenship grants in general.Footnote 16 Thanks to these grants, these eastern monarchs were now not merely Roman, but Iulii.Footnote 17
Second, we should note that the dynasties Augustus pardoned and reinstated in the decade after Actium were initially an eclectic group. Some had ruled in parts of the east for generations (e.g. in Commagene) but many others were ‘new men’ who came to power thanks to Roman support in the previous generation (e.g. Polemo in Pontus, Archelaus in Cappadocia, Herod in Judaea).Footnote 18 Suetonius famously notes that Augustus encouraged monarchs to marry each other.Footnote 19 The prosopographic studies of Richard Sullivan show the particular applicability of this statement to the eastern provincial monarchs: repeated marriages between dynasties created an ‘eastern dynastic network’ of a handful of specific families, namely the ruling dynasties of Thrace, Pontus, Cappadocia, Commagene, Emesa, and Judaea — largely the same families whom Augustus appears to have systematically made Roman citizens.Footnote 20 The children of this ‘network’ would then often be raised in Augustus’ own household at Rome, in another break with late republican practice.Footnote 21
The experience of childhood in the imperial household appears to have been of particular importance to prospective eastern provincial monarchs, and this practice continued beyond Augustus’ reign. With the later Julio-Claudian emperors, we can observe how the rise of a new emperor to power was accompanied by the appointment of childhood friends, descended from the Augustan network of eastern families, as provincial monarchs. Gaius offers the clearest case: of the six or seven provincial monarchs he is known to have appointed shortly after taking power, at least four, and probably five or more, knew him from childhood.Footnote 22 When Claudius took power, he awarded new territories only to monarchs who grew up in the imperial court, the majority of them being his own childhood friends.Footnote 23 Nero is known to have assigned territories to only three provincial monarchs soon after taking power; at least one of them (Agrippa II) and possibly a second (Aristoboulos) grew up in the imperial household.Footnote 24 By systematically granting Roman citizenship to these monarchs, encouraging the formation of a network of marriage alliances between them, raising their children in the imperial household and allowing only their family members to serve as future monarchs, Augustus transformed a diverse group of eastern potentates into a new, socially cohesive and distinct class of Roman administrators.
As administrators, provincial monarchs’ route to power in the provinces paralleled that of their contemporary imperial legates, rather than their late republican predecessors. The latter typically first succeeded to power in their kingdom (or usurped it), and only then petitioned the senate for retroactive recognition.Footnote 25 By contrast, provincial monarchs were appointed by the emperor. Augustus and his successors exercised much discretion regarding both whom to appoint and what territory to assign them. A dynastic ‘claim’ to a territory, i.e. having an ancestor who ruled that specific realm, might have formed a consideration in the emperor's decision, and appears as a justification in epigraphical and historiographical discourse.Footnote 26 Yet such ‘claims’ were neither sufficient nor necessary. No principle of inheritance (e.g. primogeniture) appears to have been generally applied, and capable adult sons could be passed over in favour of other relatives.Footnote 27 Moreover, provincial monarchs, although members of the ‘eastern dynastic network,’ were routinely assigned territories to which they had no personal ancestral claim whatsoever. Josephus explicitly reports this with regards to some of the territories Claudius gave Agrippa, though he does not claim Claudius’ choice was unusual or merits special explanation:
προσθήκην τε αὐτῷ ποιεῖται πα̑σαν τὴν ὑπὸ ̔Ηρώδου βασιλευθεῖσαν, ὃς ἦν πάππος αὐτοῦ, Ἰουδαίαν καὶ Σαμάρειαν. καὶ ταῦτα μὲν ὡς ὀφειλόμενα τῇ οἰκειότητι τοῦ γένους ἀπεδίδου· Ἄβιλαν δὲ τὴν Λυσανίου καὶ ὁπόσα ἐν τῷ Λιβάνῳ ὄρει ἐκ τῶν αὐτοῦ προσετίθει.
He also added to Agrippa's dominions all the other lands that had been ruled by King Herod, his grandfather, namely, Judaea and Samaria. He restored these lands to him as a debt due to his belonging to the family of Herod. But he also added Abila, which had been ruled by Lysanias, and all the land in the mountainous region of Lebanon as a gift out of his own territory.Footnote 28
Many other cases are known or can be surmised: so Augustus gave Armenia Minor and a part of coastal Cilicia to Archelaus I in 20 b.c.e.; Gaius gave Antiochus IV parts of coastal Cilicia in addition to his ancestral Commagene; Claudius gave Chalcis to Agrippa's brother Herod; and Nero gave Armenia Minor to Aristoboulos.Footnote 29 The title assigned to the monarch, either full-fledged kingship or a lesser title such as τετράρχης or δυνάστης, was likewise neither inherited nor determined by territory, but a symbolic honour awarded by the emperor,Footnote 30 who reserved the power to reward with full kingship monarchs initially appointed with lesser titles.Footnote 31 Moreover, though Strabo's assertion that some territories were considered particularly suitable for monarchic rule may be correct,Footnote 32 it is nevertheless the case that emperors could and did appoint provincial monarchs to rule over territories previously governed by imperial legates or procurators, just as they appointed the latter to replace deceased or deposed monarchs.Footnote 33 Provincial subjects were well aware of the emperor's flexibility in this regard, and could petition him to be ruled by either form of administration.Footnote 34
As with imperial legates, the emperor could terminate a provincial monarch's appointment at any time. This could be a punishment, when the monarch was suspected of either domestic mismanagement (e.g. Archelaus of Judaea) or conspiracy with Rome's enemies (the Parthians most of all — e.g. Antipas and Antiochus IV of Commagene).Footnote 35 However, removal could also be decided upon with apparently no blame attached to the monarch, as can be surmised in cases where the removed monarch was immediately given a new realm to rule.Footnote 36
Provincial Monarchs vs External Kings
Observing some of the traits of provincial monarchs outlined above, some scholars assert that all ‘friendly kings’ functioned as Roman administrators under the early Principate.Footnote 37 However, such a generalisation goes too far. As seen above, imperial authors, and Augustus himself in the Res Gestae, distinguish between the eastern imperial provinces and the ‘external kings’ beyond them. This conceptual distinction is underpinned by significant historical differences in the position of the monarchs on either side of the Euphrates. Provincial monarchs were appointed by the emperor and ruled at his pleasure. Such monarchs were therefore very keen to advertise their loyalty to Rome and the imperial family by adopting epithets such as φιλορώμαιος and φιλόκαισαρ, fostering the imperial cult, naming cities after members of the imperial family, presenting the emperor's portrait on their coinage, and so forth.Footnote 38 External kings, by contrast, came to power, and tended to maintain it, independently of Roman emperors’ authority. Accordingly, they did not advertise loyalty to Rome in similar fashion. They were not usually Roman citizens, nor raised in Rome, and they had at most weak ties to the dynastic network of provincial monarchs. These were not ‘limbs and parts of the Empire’Footnote 39 who could simply be given orders by the emperor, but had to be negotiated with, persuaded, threatened or coerced.
Full treatment of these external kings cannot be attempted here. I will, however, provide an example by way of illustration: King Pharasmanes of Iberia.Footnote 40 Pharasmanes succeeded his father in 35 c.e. There is no indication that this happened with Roman consent, let alone Roman initiative. Rather, it is reported that Tiberius merely exploited the opportunity to convince Pharasmanes to invade Armenia Major and instate his brother Mithridates as king in that realm (appeasing the latter for the fact that he did not inherit Iberia).Footnote 41 Alternatively, Josephus reports that Tiberius resorted to bribes to induce the involvement of Iberia and Albania in this venture,Footnote 42 a claim which may find some corroboration in archaeological finds that might indicate Roman gift-giving diplomacy in this region in the first and second centuries c.e.Footnote 43 Pharasmanes evidently changed his mind some years later, and invaded Armenia in order to depose his brother and install his son instead. Messengers from the legate of Syria demanding that he retreat were apparently ignored.Footnote 44 It was only after the Parthians invaded Armenia, and expelled Pharasmanes’ son, that he allied himself with Roman interests once more, in a move that Tacitus plainly considers opportunistic.Footnote 45
This pattern of behaviour clearly contrasts with that of provincial monarchs, as demonstrated by the conduct of King Cotys in this affair. Cotys, appointed to rule neighbouring Armenia Minor by his childhood friend Gaius,Footnote 46 was invited to intervene in the conflict by leading figures from Armenia Major. Tacitus reports, however, that Cotys was ‘restrained’ by a letter from the emperor.Footnote 47 Importantly, it would seem that this effective order was issued, and obeyed, before Cotys had taken any significant action. It is not difficult to imagine that it was Cotys himself who reported the Armenian emissaries’ request to the emperor, and asked for instructions.
Borderline Cases: The Bosporan Kingdom and the Nabataeans
Despite the conceptual distinction stressed here, there may have been ambiguous cases, or, at any rate, cases where the emperor could hope to integrate an external kingdom fully into the empire by ‘provincialising’ its monarchy.Footnote 48 The Bosporan Kingdom may perhaps offer an example for such a belatedly provincialised monarchy. The position of this kingdom at the time of Augustus’ visit to the east in 20 b.c.e. is ambiguous. King Asander, who originally came to power against Roman wishes, was perhaps later recognised,Footnote 49 and used the epithet φιλορώμαιος on his inscriptions.Footnote 50 His coins, however, make no overt reference to Rome, nor do those of his queen Dynamis, which are contemporary with Augustus’ visit.Footnote 51 It is unknown whether Augustus granted Roman citizenship to either.
In the following decade, Agrippa, visiting the east, responded decisively to the spontaneous succession/usurpation of a new Bosporan king: by threatening a Roman invasion, he imposed Polemo, pre-existing provincial monarch of neighbouring Pontus, as king, and married him to the long-reigning queen, Dynamis, thus joining Bosporus to the provincial dynastic network.Footnote 52 In consequence of this intervention the coinage of Bosporus started featuring imperial portraiture for the first time, and it may also be commemorated by the naming of cities after members of the domus Augusta (Agrippia and Caesarea).Footnote 53 The dates of both Dynamis’ and Polemo's deaths are uncertain,Footnote 54 but the fact that a new king, Aspourgos, came to power within a year of Tiberius’ succession may suggest that Bosporan kingship was now awarded under imperial initiative.Footnote 55 It is also notable that Aspourgos, too, married into the eastern dynastic network, further reinforcing the notion that the Bosporan kingdom was successfully ‘normalised’ as a provincial monarchy.Footnote 56
Only one aspect might give us pause: the indication that Tiberius granted Aspourgos Roman citizenship.Footnote 57 This may suggest that Augustus had not granted citizenship to Aspourgos’ father, King Asander, confirming the suspicion that Bosporus was an outlier before Agrippa's intervention. Alternatively, it may be that Aspourgos was not Asander's son, but a usurper whom Tiberius recognised in return for full integration into the provincial monarchic system.Footnote 58 In any event, the fact that succession in Bosporus in 37/8 c.e. again closely followed imperial succession in Rome is suggestive of the kingdom's subsequent stability as a provincial monarchy.Footnote 59
The Nabataean kingdom, by contrast, may offer a case where Augustus’ hopes for a conversion of an external kingdom into a provincial monarchy fell short. This kingdom, too, held an ambiguous position early in the Augustan period. As noted recently, its royal discourse differed in most respects from that of its provincial-monarchic neighbours. As Andreas Kropp observes with regard to Nabataean coins: ‘There is no allusion to Rome in either text or image, and not so much as a hint of even the existence of the Roman empire’.Footnote 60 There is likewise no indication that any Nabataeans were granted citizenship by Augustus, or were raised in Rome.Footnote 61 While this may be a matter of gaps in our evidence, the use of Nabataean Aramaic rather than Greek legends on Nabataean coins offers a striking contrast with the provincial monarchs’ carefully maintained Hellenistic persona of kingship (Section III below).Footnote 62 On the other hand, Obodas II (c. 30–9 b.c.e.) did preserve the diadem as his dominant symbol of kingship, potentially indicating some remaining affinity to a regional Hellenism.Footnote 63
Obodas’ energetic minister (ἐπίτροπος), Syllaeus, was apparently eager to foster close ties with Roman elites, offering himself as guide for Gallus’ Arabian campaign and later visiting Augustus in Rome.Footnote 64 Syllaeus was also reportedly keen on joining the budding provincial dynastic network.Footnote 65 It is possible that Augustus was planning to have this ‘new man’ succeed Obodas and thus transform Nabataea into a provincial monarchy. This may explain his reported outrage at the fact that Aretas, a member of the native dynasty, succeeded Obodas without seeking imperial permission.Footnote 66 Augustus may have briefly considered retaliating by either recognising Syllaeus as a counter-king or by granting Nabataea to the neighbouring provincial monarch, Herod (as with the imposition of Polemo on Bosporus); if so, he thought better of it.Footnote 67
Within a few years of Aretas’ succession Nabataea's position as an external kingdom was firmly set. Aretas abandoned the diadem in favour of the laurel wreath as a symbol of power, in stark contrast to the imagery used by provincial monarchs and in apparent defiance of Rome, where the laurel wreath was the exclusive purview of the imperator. Far from being φιλορώμαιος or φιλόκαισαρ, he introduced the epithet RḤM ‘MH (‘lover of his people’).Footnote 68 Though he did marry a daughter to a neighbouring provincial monarch, he himself took no such queen.Footnote 69 Late in the reign of Tiberius, Aretas even invaded the territories of the recently deceased tetrarch Philip — now presumably under direct Roman rule — and may have penetrated deeper still into the province of Syria, possibly occupying Damascus, and apparently retreating only after the governor Vitellius threatened war on Tiberius’ orders.Footnote 70 Our extant Greco-Roman sources are quite informative on the provincial-monarchic arrangements of Gaius and Claudius,Footnote 71 but do not mention Aretas’ death in 40 c.e. and his succession by his son Malichus. The Nabataean realm was by that time well established as an external kingdom, whose domestic administration was not an imperial concern.Footnote 72
III PROVINCIAL MONARCHY AS AN AUGUSTAN ARCANUM IMPERII
It is remarkable that Augustus should lay the foundations for the system of provincial monarchy outlined above in the 20s b.c.e., precisely when he was busy experimenting with constitutional and administrative reforms for the empire as a whole. In fact, as we have seen, Strabo indicates that monarchs formed a part of the ‘first settlement’ of 27 b.c.e. and were subjected to the power of the emperor as part of the imperial provinces. Thus modern accounts that describe the imperial provinces in the east as consisting exclusively of imperial legates and procurators, either ignoring the presence of monarchs or treating them as a historical remnant, do not give us a full picture.Footnote 73 We should rather stress that the portion of the east assigned to Augustus’ management in 27 b.c.e. was a composite of two parallel power structures: ‘direct rule’ through imperial appointment of legates and prefects, ruling under Roman titles, and ‘indirect rule’ by imperial appointment of provincial monarchs. These monarchs were members of a newly created eastern elite, Roman citizens typically raised in Rome, but ruling under distinctly uncivic and un-Roman titles — either as kings, or under ad hoc Greek titles which never received a Latin translation.
Importantly, this parallel structure did not correspond to any permanent territorial division, and lands shifted back and forth between the two types of administrators, as noted above. Braund observes that monarchs in the east tended to rule territories which were bordered by, or formed enclaves within, provinces under direct rule,Footnote 74 and Barrett notes that the same monarch could be assigned to administer several distinct territories simultaneously, geographically separated by vast tracts of land under direct Roman administration.Footnote 75 Thus if a geographic pattern is to be observed at all, it is the lack of any consistent territorial separation between the two systems — or rather the consistent geographic mix of both systems, the dynamic patterning of provincial monarchies within and between areas controlled by ostensibly Roman officials.
Famously, by defining most of the frontier provinces as ‘imperial’, the settlement of 27 b.c.e. allowed Augustus to control the bulk of the legions and personally appoint their commanders. In the east, the number of legions was relatively small, and they were complemented by significant auxiliary forces provided by the provincial monarchs, likewise personally appointed by the emperor.Footnote 76 However, Augustus evidently did not consider personal ties a sufficient precaution against conspiracy. Senatorial legates in particular could aspire to seize power like the generals of the late Republic, and therefore additional precautions were needed to make sure only Augustus and his family members held supreme power. As is well known, after 19 b.c.e. only members of Augustus’ own family celebrated triumphs.Footnote 77 Similarly well known is the decision that strategically important Egypt would be administered by an equestrian, and that senators and high-ranking equestrians would need special permission even to visit Egypt, a policy which Tacitus aptly describes as one of Augustus’ dominationis arcana (‘secrets of despotism’).Footnote 78
I propose that the system of provincial monarchy outlined above should be understood as another such dominationis arcanum. The geographic spread of provincial monarchs throughout the east, and the relatively high proportion of the eastern forces under their command, meant that any senatorial legate's attempted coup could only hope to succeed with the cooperation of many of these monarchs (as Vespasian's later success demonstrates).Footnote 79 Provincial monarchy was therefore fashioned and maintained in a form designed to render such a scenario unlikely while Augustus and his family remained in power, and to minimise its potential damage should it occur.
It is in this light that we should understand the measures to guarantee provincial monarchs’ personal connections and loyalty to the Julio-Claudians, and also the fact they were encouraged to marry each other, but never intermarried with senatorial and equestrian elites.Footnote 80 More importantly, this interpretation offers a good explanation for the consistency in maintaining the monarchs only as monarchs, keeping the forms of their power distinctly un-Roman and uncivic. The titles (and accompanying honours) of the monarchs, both the notorious rex/basileus and other, technical-sounding, Greek titles (not coincidentally, never translated into Latin), gave the monarchs’ power a foreign and uncivic appearance, a perfect foil for Augustus’ own persona as civilis princeps. Footnote 81 Their dynastic marriages and dynastic pedigree, though in fact largely an Augustan creation, nevertheless strengthened this flavour of ‘oriental’ and Hellenistic kingship. An alliance of a Roman senatorial commander with such eastern kings would be politically toxic. In this period of deep thought and experimentation with the running of the empire, Augustus surely remembered how he himself had demonised Antony's alliance with ‘the east’ and its monarchs only a few years earlier; he must have realised the advantages of maintaining the outer appearance of this monarchic east, while changing its essence to his advantage.
In other words, by creating and maintaining provincial monarchy, Augustus produced an effective ‘divide and rule’ dynamic as a systematic feature of the Roman east. The alienating monarchic persona, and the relative weakness of personal ties between ‘eastern’ monarchs and senatorial legates, created a latent tension which allowed the two groups to act as mutual checks on each other. Thus when the imperial legate of Syria, Marsus, came upon a meeting of provincial monarchs (c. 43 c.e.), he did not join in, but rather hastened to disperse the meeting, and sent the monarchs back to their respective realms, at a safe distance from each other and from himself.Footnote 82
It is evident that the toxicity of the eastern monarchic persona in Roman politics persisted over the long term. Dio reports that Kings Agrippa and Antiochus were accused of being Gaius’ τυραννοδιδάσκαλοι (‘tyrant-teachers’).Footnote 83 The accusation is absurd if taken literally — Agrippa and Antiochus enjoyed no political power before Gaius himself took power and appointed them as kings; their de facto or de jure powers as kings were in no way more ‘tyrannical’ than those of imperial legates, and far less than Gaius’ own powers. It was rather their personae as Hellenistic kings, their outward uncivic appearance and their royal pedigree, that were considered a bad influence.
Importantly, this foreign monarchic persona was not natural or inevitable; it was artificially and deliberately maintained and advertised. As we have seen, this was achieved by the careful maintenance of the specific pedigrees of the monarchs and the clear distinctions of their titles. Moreover, certain ceremonial gestures may also have been meticulously fostered to establish this persona. A passage in Dio, describing the aftermath of Claudius’ rise to power, offers a useful illustration:
τῷ γὰρ Ἀγρίππᾳ τῷ Παλαιστίνῳ συμπράξαντί οἱ τὴν ἡγεμονίαν (ἔτυχε γὰρ ἐν τῇ Ῥώμῃ ὤν) τήν τε ἀρχὴν προσεπηύξησε καὶ τιμὰς ὑπατικὰς ἔνειμε. τῷ τε ἀδελφῷ αὐτοῦ Ἡρώδῃ τό τε στρατηγικὸν ἀξίωμα καὶ δυναστείαν τινὰ ἔδωκε, καὶ ἔς τε τὸ συνέδριον ἐσελθεῖν σφισι καὶ χάριν οἱ ἑλληνιστὶ γνῶναι ἐπέτρεψεν.
He [Claudius] enlarged the domain of Agrippa of Palestine, who, happening to be in Rome, had helped him to become emperor, and bestowed on him the ornamenta consularia; and to his brother Herod he gave the ornamenta praetoria and a principality. And he permitted them to enter the senate and to express their thanks to him in Greek.Footnote 84
Agrippa and (presumably) Herod grew up in Rome, in Claudius’ household.Footnote 85 The two were in Rome at the time of Gaius’ assassination, and were in a position to aid Claudius in taking power. If Josephus is to be believed, Agrippa was well known and well regarded enough to serve as an intermediary between Claudius and the Senate.Footnote 86 When rewarding the two for their service, Claudius chose to bestow upon them some of the highest civic honours a Roman citizen could receive, the ornamenta consularia and the ornamenta praetoria.Footnote 87 And yet Dio thinks it worthy of mention that they were given permission to enter the senate and thank Claudius in Greek. I find it hard to believe that the two, having grown up in Rome, did not have sufficient command of Latin to be able to declaim adequately what must have been a scripted and ceremonial text. Rather, their use of Greek in this instance served a performative function (as did the grant of a dispensation to use it). Agrippa and Herod, now in their public role as eastern monarchs, had to become ostensibly foreign and estranged from Roman politics and its ‘republican’ institutions. The emperor gave them power in the uncivic form of kingship, and their address to the senate in this capacity mandated the performative use of Greek.
This pseudo-Hellenistic persona of eastern provincial monarchs, their function as a cohesive elite of intermarried families, and the wide discretion which the emperors held as to the title, territory and timing of their monarchic appointment, were all characteristics honed by Augustus to serve his goal of monopolising power. This particular combination of features took into account and exploited the specific geopolitical and strategic constraints and opportunities offered by the Roman east. A brief examination of the west will demonstrate why a full parallel cannot be found there.
IV PROVINCIAL MONARCHS IN THE WEST?
As argued above, thanks to the system of provincial monarchy, the new eastern monarchic elite, personally loyal to the domus Augusta, both complemented and formed a check on senatorial legates. Augustus could hence afford to leave a smaller number of legions on the Parthian frontier and have greater confidence in his ability to maintain overall control over the eastern provinces, even while leaving them without the physical presence of members of the imperial family for many years at a time.Footnote 88 The situation west of the Adriatic was quite different. For one thing, there was no equivalent abundance of allied kingdoms, rich with resources and politically potent cultural legacy, to be similarly exploited. Nor were the frontier challenges focused on negotiation with a single main adversary. These radically different conditions necessitated a different approach, including the deployment of a larger number of legions under the direct oversight of Augustus or immediate family members. Even as Augustus was finishing his tour of the east, Agrippa was busy suppressing revolts in Gaul and Spain. The campaign to finish the subjection of the Alpine tribes and thus to secure the strategic passes was renewed in 17 or 16 b.c.e., and Tiberius and Drusus were quickly given a leading role. Following Lollius’ defeat, Augustus personally spent three years in Gaul, and members of his family would go on to command the Rhine legions for most of the rest of his reign. Campaigns on the Danube were likewise personally supervised by members of the domus Augusta. Footnote 89 Augustus’ consistency in appointing men of his family to high commands was, of course, important in building and maintaining their prestige in Rome,Footnote 90 but likewise helped ensure the loyalty of the vast armies on the European frontiers. Under these conditions, there was less of a need to create a system like that in the east.
Thus even in the rare cases where forms of indirect rule were available and could be usefully exploited, they took a very different form. In the aftermath of the Alpine campaigns, Augustus allowed Cottius to remain as ruler of his ancestral realm in the western Alps. His dynasty, which continued to rule in the region down to the time of Nero, could be considered to fit Braund's broad model of ‘friendly kings’.Footnote 91 However, the form of rule, and especially its discursive manifestation, is clearly distinct from the eastern provincial monarchs. The incorporation of Cottius I into the empire involved not merely receiving Roman citizenship, but exchanging the title of king for that of prefect. Even in his self-presentation to his own subjects, on the arch he built in Susa, Cottius presents himself as the son of a king, but a prefect himself.Footnote 92 The iconography of the monument (with togate figures, lictors, suovetaurilia sacrifice and so forth) is likewise fully ‘Romanised’ and avoids any overt symbols of foreign monarchic power.Footnote 93 Furthermore, the precise title under which Cottius ruled (praefectus ceiuitatium) was also used by Roman commanders assigned to rule over other Alpine tribes.Footnote 94 Thus, while the Cottian Alps differed from neighbouring areas in being controlled, de facto, by a hereditary native dynasty, under Augustus the outward forms of their rule were transformed so as to be part of a homogenised system, which downplayed, rather than emphasised, the differences between direct and indirect rule.Footnote 95 In the absence of a broad dynastic network, with local rulers who lacked the toxic cultural prestige of eastern kings, and with members of the domus Augusta safely in command of the legions, there was neither the possibility nor the need to foster a ‘divide and rule’ dynamic on the model of the east.
The kingdom of Mauretania, in North Africa, provides a more compelling comparison with eastern provincial monarchies, and Octavian-Augustus’ treatment of Juba II in many ways foreshadows his conduct towards eastern provincial monarchs. Led in triumph as a child following the death of his pro-Pompeian father, Juba II was raised in Rome. Augustus probably granted him Roman citizenship before appointing him to rule Mauretania in 25 b.c.e., half a decade or so before his new settlement of the east.Footnote 96 As would increasingly be the case in the east in later years, the appointment was made despite Juba's lack of strong hereditary claims to his new realm.Footnote 97 Moreover, as often noted, Juba and his son and successor Ptolemy displayed many behaviours typical of eastern monarchs, such as euergetism towards old Greek cities, occasional use of Greek on their coinage (though Latin predominated), and the (re-)foundation of Caesarea-Iol, the only western equivalent to the naming of cities after members of the imperial family (common in the eastern monarchies).Footnote 98
These similarities must, however, be weighed against significant differences. As we have seen, one key aspect of Augustus’ reinvention of provincial monarchy was the fostering of marriage ties, creating an ‘eastern dynastic network’. The dynasty of Mauretania stands apart from this network, with none of its known members of either sex forming lasting marriage ties with eastern provincial monarchs.Footnote 99 An explanation for this idiosyncrasy may be suggested by the marriage that Augustus did arrange for Juba, with Cleopatra Selene, daughter of Antony and Cleopatra. Cleopatra Selene's unique position as the last of the Ptolemies, and as the only monarch with family ties to elite Romans (including Augustus’ own family!) must have weighed heavily with Augustus.Footnote 100 Marrying her to Juba, and sending the couple to rule a realm at the western edge of the empire to which neither had strong ancestral claim, allowed an honourable position for Cleopatra (to whom Augustus was now a patron and benefactor) while keeping her away from both senatorial elites in Rome and her traditional power bases in the east. Viewed in this way, the Mauretanian ‘exception’ to Augustus’ dynastic policy can be neatly explained: Cleopatra's own genealogy undermined the cultural and political dichotomy between monarchs and senatorial elites which Augustus was keen to maintain in the east. The unique position of Cleopatra's son King Ptolemy as kin to the emperors may likewise explain Mauretania's continued dynastic isolation in the next generation.
In addition to the unique genealogy of its monarchs, there are indications that the legal-administrative position of Mauretania may likewise have been sui generis. As noted above, one of the key features of eastern provincial monarchy under the Principate was that dynastic succession ceased to be taken for granted, but the emperor exercised wide discretion in replacing a deposed or deceased monarch.Footnote 101 By contrast, Ptolemy seems to have become co-ruler in his father's lifetime.Footnote 102 Furthermore, we lack any sign of imperial discretion regarding Ptolemy's elevation as sole king on Juba's death. In fact, we are told that the senate recognised him well after he became sole ruler — rewarding him for his support in war — in a ceremony whose archaism is noted by Tacitus.Footnote 103 And yet this continuity with republican forms of conduct is echoed in the neighbouring province of Africa, which was maintained as a ‘public’ province, and was the only province of its kind where the proconsul continued to have a legion at his command.Footnote 104 In this remote corner of the empire, with its relatively small armies, Augustus could tolerate the survival of republican forms of administration, both ‘direct’ and ‘indirect’, which he took pains to phase out in the strategically important east.Footnote 105
With his victory at Actium, Octavian-Augustus succeeded in unifying the Roman empire under his rule. In the quest to stabilise this novel autocratic regime, it is reasonable to expect certain empire-wide commonalities. In his approach to indirect rule, some innovations were applied throughout the empire, such as a willingness to grant Roman citizenship to native dynasts.Footnote 106 Nevertheless, it is equally clear that Augustus was highly pragmatic and flexible in his approach, and took regional needs and opportunities into account; his careful fostering of the budding dynastic network in the east (complete with Hellenistic royal persona), combined with idiosyncratic and unprecedented modes of appointment and territorial flexibility, was specifically tailor-made to favour the monopolisation of power by the Julio-Claudians.Footnote 107 It is for this reason that provincial monarchy thrived in the east to the end of the Julio-Claudian period, but disintegrated under the Flavians, as will be explored below.
V TITUS, BERENICE AND THE END OF PROVINCIAL MONARCHY
It has long been observed that the nature of Roman imperialism in the east seems to have changed under the Flavians. Most noticeably, we hear of no more ‘allied kings’ west of the Euphrates by the early second century c.e.Footnote 108 Some scholars, believing that indirect rule in the east was in long-term decline since the beginning of the Principate, consider this to be merely the end of a long process. However, such notions tend to be based on the assumption that ‘client kingship’ is a relic of republican times, destined to fade with the consolidation of the empire.Footnote 109 The model offered above attempts to challenge this view. Moreover, even allowing the premise of long-term decline, one must at the very least concede that a qualitative shift occurred under the Flavians. The Jewish war, the civil wars following Nero's death and the Roman invasion of Commagene all reveal an east in which provincial monarchs still played a fully functional and significant part.Footnote 110 If Josephus’ figures for the Jewish War can be believed, monarchs provided about a third of the total forces in that conflict.Footnote 111 Other monarchs, not participating in the suppression of the revolt, might have had forces substantial enough to allow Vespasian to leave minimal legions on the Parthian frontier.Footnote 112 The fact that no provincial monarchs seem to have been left in power west of the Euphrates by the time of Domitian's death therefore cries out for explanation.Footnote 113
Many scholars hypothesise a Flavian policy of systematic shift to direct rule throughout the east. This shift, the argument usually goes, was part of a general strategic reorganisation of the Parthian frontier in the aftermath of Nero's concession of Armenia to the Arsacids. Scholars typically note the shift to direct rule in Commagene (with parts of Cilicia) and in Armenia Minor early in Vespasian's reign. Some go on to cite other roughly contemporary changes, such as the introduction of legions to Cappadocia and administrative reshuffles in Asia Minor. On this basis scholars have inferred a thoroughgoing strategic reform under Vespasian, in which all the remaining provincial monarchs (including those for whom the circumstances and timing of their end is unclear) were similarly deposed in favour of direct rule.Footnote 114
A closer examination of the evidence, however, raises doubts. Our sole testimony for the circumstances of the end of the monarchy in Commagene (Joseph., BJ 7.219–43) clearly states that Vespasian was not proactive and strategic, but reactive and focused: after the Syrian legate Caesennius Paetus alleged that Antiochus IV was conspiring with the Parthians, Vespasian allowed Paetus to invade Commagene (with the aid of two kings!) and depose Antiochus. Some modern historians dismiss Paetus’ allegation as a ‘pretext’ used by Vespasian to justify what was in truth a strategic reorganisation of the frontier.Footnote 115 However, as Roman emperor, Vespasian could end a provincial monarch's tenure on no more serious pretext than was required for an imperial legate,Footnote 116 and one should not therefore dismiss the positive evidence pointing to concrete ad hoc reasons for the change in Commagene.
The only other monarchy whose shift to direct rule can be confidently dated early in Vespasian's reign is Armenia Minor. Our evidence is purely numismatic: the latest coins of King Aristoboulos can be dated to 70/1 c.e., and subsequent coins use a new provincial era in Nicopolis beginning in 71 or 72 c.e.Footnote 117 Modern scholars are quick to assume that this monarch, too, was deposed for the same strategic reasons as Antiochus of Commagene. However, while Vespasian may have deposed Aristoboulos, and this could have been a response to ‘strategic’ considerations, neither assumption is at all certain. For one thing, Aristoboulos may have died (there is no certain record of him after this time). Alternatively, he may have been removed due to local or personal considerations.Footnote 118
All this is not to deny that the Flavians were concerned about the Parthians (Josephus’ account of the invasion of Commagene in fact suggests as much).Footnote 119 The unprecedented introduction of two legions to Cappadocia makes sense in this context, as does perhaps the rapid succession of Syrian legates in the early 70s c.e. and the infrastructural projects they oversaw in Syria.Footnote 120 Furthermore, it is reasonable to interpret the provincial boundary-changes in Asia Minor during this decade as indications for imperial rethinking and concern with administrative arrangements.Footnote 121 It is by no means self-evident, however, that such measures must have entailed a systematic shift to direct rule. Previous emperors seem to have added territories to monarchs in preparations for Parthian campaigns.Footnote 122 Nor should provincial redistricting necessarily entail a shift to direct rule: Vespasian himself appointed a new monarch in part of Cilicia Tracheia after uniting the two Cilicias into a single province.Footnote 123 The fact that Agrippa II appears to have remained in power down to the late 80s or early 90s c.e. (probably until his death of old age) also challenges the notion of such a systematic shift.Footnote 124
In the final reckoning, we know of two monarchies that ended early in Vespasian's reign for reasons that cannot be determined with confidence (Commagene and Armenia Minor); of one that was subsequently created (Alexander in Cilicia); and of one that certainly lasted far beyond Vespasian's reign (Agrippa II). The fate of at least three other monarchs in power at the beginning of his reign is unknown;Footnote 125 there are some indications that the total number of eastern monarchies in the early 70s c.e. might have been still greater, casting further doubt on our ability confidently to detect any ‘provincialising’ trend soon after Vespasian took power.Footnote 126
Concern over the Parthians, therefore, though probably present, cannot be taken for granted as either a necessary or sufficient explanation for the end of provincial monarchy west of the Euphrates by the death of Domitian. The model of provincial monarchy presented above suggests a novel approach to the problem, which may yield a different explanation. In what follows, I interpret some aspects of the Flavians’ rise to power and the early years of their reign in this light, and suggest some possible conclusions regarding the end of provincial monarchy. This reconstruction also draws connections between aspects of early Flavian history rarely considered in tandem by modern scholars, offering avenues for further study.
The End of Provincial Monarchy: A Tentative Reconstruction
I argued above that a systematic ‘divide and rule’ dynamic was integral to the dual nature of the eastern imperial provinces as organised by Augustus and maintained by his successors. This was based on two pillars: the fostering of strong personal loyalty of monarchs to the Julio-Claudians, while at the same time rendering them politically toxic to senators through their carefully maintained Hellenistic royal persona. However, with Nero's death and the end of the Julio-Claudians, one of the two pillars of this system was lost.
Elites in Rome may have been slow to appreciate the fundamental change, as demonstrated by the different reactions of Titus and King Agrippa II to Galba's death. Titus, who grew up in Claudius’ household and was the son of one of the major commanders in the east, was clearly considered capax imperii and was sent by Vespasian to Rome to pledge his loyalty to Galba (perhaps with a hope of adoption).Footnote 127 However, after learning of Galba's death and the ensuing civil war, Titus felt he must avoid Rome, and returned to Palestine instead.Footnote 128 By contrast, Titus’ travel companion, Agrippa II, who had also spent much of his youth in Claudius’ household and was also a commander of an eastern army (albeit significantly smaller than Vespasian's), evidently did not face a similar problem, and could continue on to Rome without being expected to declare his loyalties clearly one way or another.Footnote 129 Even at this late stage, the Hellenistic persona of the provincial monarchs endured: to the senatorial elites, Agrippa was a ‘foreign’ king and thus outside the senatorial ‘game of thrones’, while Titus was a senatorial son of an imperial legate.
Such a view proved dangerously outdated, however, in the wake of fully fledged civil war. The monarchs had significant forces to contribute, and probably also substantial financial means.Footnote 130 With no Julio-Claudian in play, no claimant could take their loyalties for granted. The existing cooperation between several eastern monarchs and Vespasian in suppressing the ongoing Jewish revolt must have offered a convenient basis for an alliance.Footnote 131 Titus’ affair with Berenice, which seems already to have been ongoing at this point, was surely considered convenient to all sides.Footnote 132
The Eastern monarchs, an asset in civil war,Footnote 133 became a liability once the civil war was won and the Flavians turned to securing legitimacy and stability. The central place of the victory in the Jewish War in Flavian propaganda may well have been attempted as a corrective to the embarrassing fact that the Flavians took power with the aid of ‘foreign’ eastern monarchs.Footnote 134 Naturally, we hear of no monarch present at the celebrations in Rome.Footnote 135 To the Romans a king was welcome in a triumph only as vanquished, not as victor.Footnote 136 It is in this context that we should understand the fact that Titus left Berenice behind when he returned to Rome in 71 c.e.Footnote 137 The presence of the foreign queen would have been detrimental to the Flavians’ quasi-Augustan campaign of Roman restoration and rebirth.Footnote 138 The harsh treatment of Antiochus IV of Commagene in the following year probably did not hurt the Flavians’ effort at re-branding.Footnote 139 It should therefore come as no surprise if the Flavians did not appoint new monarchs in this period, even if some vacancies were created as elderly monarchs died.Footnote 140
In 75 c.e. the Flavians were apparently more confident in their position and it seems an attempt was made to reverse course. Agrippa II and Berenice were welcomed in Rome. Berenice now lived openly as Titus’ mistress, and Agrippa, her brother, was honoured with ornamenta praetoria.Footnote 141 It is probably also at this time that the deposed Antiochus IV and his children were welcomed in Rome and treated ‘with every honour’.Footnote 142 This was the likely occasion for the marriage of Iotape (PIR2 I 48), daughter of Antiochus IV, and the Herodian-Cappadocian Alexander (PIR2 A 500). Vespasian evidently felt confident enough to make his first monarchic appointment, sending this Alexander to rule a part of Cilicia.Footnote 143
This attempt to return to Julio-Claudian patterns of eastern policy proved difficult, however, under the changed circumstances.Footnote 144 Berenice's presence in Rome as Titus’ mistress was highly unpopular, resulting in a public outcry.Footnote 145 Rumours spread of an eventual marriage and the affair was taken as a sign that Titus would be a ‘bad’ emperor.Footnote 146 Realising that he could ill afford to be portrayed as the consort of a second Cleopatra, Titus dismissed Berenice upon taking power, if not sooner.Footnote 147 By doing so Titus re-cast himself in the role of an Aeneas-Augustus rather than an Antony, as part of a general reform of his conduct and image.Footnote 148
Thus, on one hand, the Flavians’ rise to power with the backing of a senatorial-monarchic coalition proved that the system of provincial monarchy could no longer be relied upon as a bulwark against revolution from the east. On the other hand, the Flavians’ subsequent experience in power, especially the affair of Titus and Berenice, showed that the monarchic persona remained highly politically toxic, and perhaps especially so for the Flavians: unlike Augustus, they rose to power while fighting with, not against, eastern monarchs. It was Titus, not his rivals, who took an eastern queen for a lover.
In this context, it is no surprise that in the years following the dismissal of Berenice, under the rule of Titus and Domitian, we do not hear of any additional provincial monarchs being appointed. It was much easier and safer to let the system gradually disintegrate as the older monarchs died out. One therefore does not need to assume a poorly attested proactive policy of provincialisation. It is far more likely that provincial monarchy gradually ended by natural means, leaving little direct evidence in our sources. By the time of the death of Rhescuporis of Bosporus c. 93 c.e., provincial monarchy must have all but died out.Footnote 149 The exceptional decision to allow Rhescuporis’ son Sauromates to succeed him, thereby guaranteeing the survival of the Bosporan kingdom, may have taken the Bosporus’ remote location into consideration. It is, however, an open question whether the same decision would have been made had Rhescuporis died a decade earlier, when the memory of Berenice was fresher and the system of provincial monarchy still very much a geopolitical reality.
VI CONCLUSIONS
As the forms of Roman provincial administration and Roman government changed in the decade following the battle of Actium, so did the role of monarchs in the provincial east. I have offered a new model highlighting some key features of the relations of these provincial monarchs to the emperor and their place in the administration of the east in the early Principate. I argued that there was a fundamental difference, probably recognised by contemporaries, between the position of these monarchs and that of allied kings in earlier periods, or in areas outside the provinces, particularly those east of the Euphrates (Section II). I have further argued that Augustus played a key role in constituting this new system, and have interpreted its creation as a hitherto overlooked aspect of the Augustan revolution. Capitalising on the politically toxic image of eastern monarchy in Roman discourse, Augustus fostered a ‘divide and rule’ mechanism intended to minimise the threat to his personal hold on power and to stabilise the eastern part of the empire (Section III). This Augustan response to the specific conditions and cultural legacy of the east differed from his treatment of other regions of the empire (Section IV).
This model and interpretation consequently offer fresh explanations to a key development in Roman imperialism in the east: the contrast between the prominence of eastern monarchs in the Julio-Claudian period, and their marginalisation and disappearance under the Flavians (Section V). Rather than considering the problem in terms of a ‘Grand Strategy’ which looks outward to Roman relations with external powers, I suggested shifting our gaze inwards, to the problems of imperial stability and domestic competition over power. We can see that the Julio-Claudians benefitted from the stabilising mechanism and internal checks inherent in using two parallel systems of administration in the east. Furthermore, the later Julio-Claudian emperors were socialised to maintain this system, thanks to their common childhood friendships with monarchs-to-be. The Flavians, by contrast, did not undergo the same socialisation. Moreover, in the years following their rise to power they gradually came to appreciate the toll of the system on their own public image, due to the retrospectively embarrassing role that ‘foreign’ monarchs had played in their violent rise to power and Titus’ affair with Queen Berenice. This growing comprehension did not trigger a proactive policy of provincialisation, but rather manifested itself passively in the marginalisation and neglect of the system. Unlike republican ‘client kingship,’ provincial monarchy under the Principate required repeated proactive maintenance through imperial appointments of new monarchs. Growing Flavian reluctance to perform this duty can sufficiently explain the end of the system within a generation.
