7  Roman religion and Roman Empire

What was the impact of Roman religion on the provincial communities of the Roman empire? We have already discussed the spread of so-called ‘oriental religions’ outside the city of Rome itself. But what of the ‘official’ cults? How far did the inhabitants of the empire acquire Roman religious identities? Was the impact of Rome strikingly different in different parts of the empire? Was it different at different periods? Or among different classes and groups of people? As we shall see, the historical development of imperial religion produced some remarkably idiosyncratic effects (the emperor Augustus, for example, depicted in traditional Egyptian style as a pharaoh offering cult to Egyptian gods), as well as some curiously ironic enigmas (as when the Roman governor of Egypt circulated the emperor Claudius’ message to the Alexandrians that they should not worship him as a god – with a covering edict calling him precisely that, ‘our god Caesar’).1 In what follows, we shall explore such representations as part of the operation of imperial power across the Roman world.

    The point of this chapter is to show, first, that Roman imperialism did make a difference to the religions of its imperial territory; and, second, to explore how we might trace the impact of Roman religion outside Rome, principally in the period after the reign of Augustus. Of course military conquest and the imposition of foreign control (whether in the form of taxation, puppet government or military occupation) inevitably impacts on cultural life – both in the imperial centre and in the provincial territories. No one can be culturally unaffected by imperialism. But its impact comes in a wide range of forms, and is experienced very differently by the parties involved – whether conquering or conquered, peasant or aristocrat, the native resistance or the local collaborators. Imperialism is, besides, constantly re-interpreted in culture and religion, as we can see very simply in the different images of the emperor himself that are found throughout the Roman provinces – not just the relatively standardized portraits on the coins that flood the Roman world, but the (to us) almost unrecognizable images from Nile sanctuaries with the emperor in the distinctive guise of Egyptian pharaoh or Ptolemaic king. Religion and culture are regularly put to work on imperialism’s behalf, incorporating the conquering power into local traditions. But at the same time religion and culture may always work against imperialist power, in reasserting the distinctiveness of native traditions against the forces (whether military or cultural) of occupation. It is a plausible suggestion that ‘native’ rebellions in the Roman empire tended to fight under the banner of local deities.2

    Within these different perspectives, we shall delineate some of the most characteristic features of Rome’s religious impact on the empire. Rome did not generally seek to eradicate ‘native religious traditions’ nor systematically to impose her own religious traditions on her conquered territories. (Roman religion’s identity as a ‘religion of place’ – strongly focussed on the city of Rome – would anyway make unlikely any wholesale direct export.) On the other hand there was borrowing and interchange at various levels between Roman cults and religious practices through the empire; Roman gods, for example, or at least (and this may not be the same thing at all) gods with Roman names, were widespread across the imperial territories, throughout our period. But such borrowing was not the same everywhere, from Scotland to the Sahara. We shall disentangle some of the factors which affected the impact of Roman religion on the world outside Rome, and how that impact was experienced. These factors include formal political rights and privileges (communities of Roman citizens outside Rome being much closer to the religion of Rome itself than non-citizens), as well as wealth and class (local élites in the provinces showing greater interest in ostensibly Roman deities than their poorer compatriots). But we shall also consider how different religious and cultural traditions in the conquered territories affected patterns of ‘Romanization’: the Roman authorities treated the Jews, for example, differently from the Druids; while the western part of the empire, with only a limited history of urban culture on the ‘classical’ model, imported Roman institutions, (whether willingly or not) more directly than did the eastern part; there, by contrast, Greek civic life and cults often worked towards the accommodation of Rome; there, the Roman conquerors found religious traditions that they recognized as already like their own, or even as the ancestors of their own.³

    All these factors (and others, as we shall see) intersect – and sometimes conflict – to produce the complex pattern of Rome’s religious influence on its empire. In section 1, however, we concentrate on the legal status of the different provincial communities and their constitutional relationship with Rome. In one sense influence literally radiated from Rome: it was strongest in Italy itself and in the camps of the Roman army, wherever it was stationed; elsewhere it was felt in proportion to the official Roman status of the town or group concerned, and their formal links to Rome itself. If some Roman pressures were exercised everywhere, the particular legal status of the community made a fundamental difference to its religious life. Let us explain first how these statuses differed.

    Throughout the first two centuries A.D. (until, that is, the emperor Caracalla completed the restructuring of such distinctions by granting citizenship to most of the free population of the empire) there were three principal types of provincial community under the empire: coloniae, municipia, and towns without any specifically Roman status at all. Roman coloniae were, with the army, the main context in which the Roman religious system was replicated abroad.4 Coloniae were communities of Roman citizens settled outside Italy. In the middle Republic they were mostly landless citizens from Rome itself, and in the first centuries B.C. and A.D. mostly ex-legionaries who received land in return for their military service; these foundations ceased altogether after the early second century A.D. They were designed to be clones of Rome in all respects: Latin was the official language, even when they were established in the Greek world; some coloniae made a point of boasting ‘seven hills’, just like Rome. So too, in their religious institutions, these ‘mini-Romes’ abroad explicitly mirrored the institutions of the capital.

    In the Latin west (especially in North Africa, Spain and Southern France) there was also a second category of towns with Roman status, known as municipia.⁵ These towns had been granted the so-called ‘Latin right’ by the Romans, which meant that individual members of the community gained some of the rights of Roman citizens and their ex-magistrates automatically became full citizens. It seems that when they received this status the new municipia also received a new constitution directly from Rome. After Vespasian granted the Latin right to towns in Spain, these municipal constitutions were standardized (under Vespasian’s son, Domitian); fragments of seven copies of the standard municipal regulation survive from Spain which clearly show the direct influence of Roman practice on institutions outside Rome.

    Communities without Roman status fell into two main types: towns in the East, whose principal language was Greek, and whose own ancient religious traditions were deeply embedded in the fabric of urban life; and towns without municipal status in the West, often of much more recent foundation, in areas that were themselves more recently conquered, or without a long history of loyalty to Rome. Both these types of community (though much more commonly and directly those in the west) borrowed elements from the Roman system, though less directly than coloniae or municipia; and their religious institutions might be subject to Roman regulation. We shall discuss the religious impact of Rome on communities without Roman status in section 2.

    Of course, as we have already implied, juridical status – even if a useful starting point – was not everything. The adoption and adaptation of Roman religious custom by local communities depended on much more than constitutional position (and on more, for that matter, than any of the other factors that we have so far mentioned): individual interests within the province, local perceptions of cultural and religious identity, calculations of advantage, no doubt, in relation either to the Roman government or the ‘native’ élite, or both. Besides, the religious practice or beliefs of individuals might always (as at Rome itself) go against the grain of the regulations laid down for the community as a whole. Just as there must have been some individuals in municipia or coloniae who lived in a resolutely non-Roman religious world, so too in towns without any formal Roman status, there must have been some whose religious experience was in many respects Roman.

    The religious history of other empires may also help us at the outset to understand the pattern of religious influence in the Roman empire. The religious impact of the centre in the periphery of empires varies greatly, often depending on how integrated the empire is (a theme we touched on in chapter 5). In some cases the central power makes stringent religious demands on its dependent territories. This is particularly clear, for example, in the highly integrated Inca empire (where the nobility directly administrated their provincial territories and there were strong reciprocal obligations between rulers and subjects). Significantly, the Inca transported the images of the major gods of the vanquished to their capital and, in return, the subjects were compelled to accept new, Inca, idols and to maintain places of worship in the same manner as in the capital. A hostile contemporary account claims that the eleventh Inca king killed all priests of the subject peoples and destroyed even their less important shrines – because the priests had refused to give him information. At all events the Incas attempted to create a tightly centralized system, of administration and religion.⁶ Likewise, in the late nineteenth and twentieth centuries, the massive overseas expansion of Japan was accompanied by the export of state Shintoism: shrines were built, for example, in Formosa, Korea and Manchuria, not just for the Japanese residents overseas, but as part of a new-found ‘world civilizing mission’.7 This is a model not far different from the British empire, which also attempted to impose its own religion (Christianity) on its colonies and to eradicate ‘unacceptable’ or ‘uncivilized’ native religious practices, though unsystematically, with varying energy and many varieties of self-deception and chauvinism.⁸

    The Roman empire, on the other hand, operated according to a quite different pattern. In general, it was relatively diffuse and unintegrated – neither systematically imposing its own cults on the conquered, nor systematically removing the cults of their subjects to the capital.⁹ But, as we have already implied, one aspect of integration was particularly important within the vast geographical and political extent of the Roman empire: that is, Roman citizenship. The bearers of Roman citizenship were, it seems, expected to recognize Roman gods; an expectation which overlaps neatly with the juridical status of the different communities we have outlined (from coloniae as full citizen communities to towns without Roman status, which might have included no Roman citizens at all). Despite increasing religious choices in the imperial period, the identity of religion and state was maintained: those who counted as ‘Roman’ in civic terms counted as ‘Roman’ in religious terms too.

    In exploring these issues, we shall consider the process by which local and Roman gods apparently merged with each other and were often referred to, and presumably worshipped, under a composite title. In Roman Britain, for example, as in many other provinces in the West, we find a wide variety of these hybrids, ‘Mars Alator’, ‘Sul Minerva’ and so on.¹⁰ In most cases, however, we have only the record of a mixed divine name; we can only guess what that name meant, which deity (Roman or native) was uppermost in the minds of the worshippers, or whether the two had merged into a new composite whole (a process often now referred to as ‘syncretism’); we do not know, in other words, how far the process was an aspect of Roman take-over (and ultimately obliteration) of native deities, how far a mutually respectful union of two divine powers, or how far it was a minimal, resistant and token incorporation of Roman imperial paraphernalia on the part of the provincials. Signs of ‘syncretism’, then, always need to be interpreted. For example, to understand why most deities in the eastern part of the empire did not merge with Roman counterparts, but retained their individual personalities and characteristics, whereas in the west pre-Roman gods acquired Roman names, or non-Roman and Roman divine names were linked, we need to investigate much more deeply the nature of Roman religion outside Rome; we need also to attend to the agenda of all those groups involved in developing a new Roman imperial world view – throughout the empire and over the centuries.11

    Another theme, central to this chapter, is the ‘imperial cult’ offered to the Roman emperor or his (deified) predecessors, with temples, festivals, prayers and priesthoods in every province of the empire. The historian Cassius Dio in the third century A.D. saw cult of the emperor as one unifying factor in the religions of the vast imperial territory, one aspect of worship that all Roman subjects shared. After noting the establishment of temples to Augustus in Asia and Bithynia, he goes on to say: ‘This practice, beginning under him, has been continued under other emperors, not only among the peoples of Greece, but also among all the others insofar as they are subject to the Romans.’¹² We have chosen to consider various aspects of imperial cult together in section 3 partly because of Dio’s claim of its universality across the Roman empire, and his suggestion that this form of shared religious practice was one aspect of ‘belonging’ to that empire. On the other hand, we do not want to suggest (and Dio comes nowhere near claiming) that there was a single entity, the same throughout the empire, that can be identified as ‘the imperial cult’. There was no such thing as ‘the imperial cult’; rather there was a series of different cults sharing a common focus in the worship of the emperor, his family or predecessors, but (as we shall see) operating quite differently according to a variety of different local circumstances – the Roman status of the communities in which they were found, the pre-existing religious traditions of the area, and the degree of central Roman involvement in establishing the cult. Besides, there was no sharp boundary between imperial cult and other religious forms: the incorporation of the emperor into the traditional cults of provincial communities, his association with other deities, was often just as important as worship which focussed specifically and solely on him. Nor was imperial cult necessarily the most powerful marker of Romanization in religion: in specifically Roman communities abroad (coloniae and municipia), imitations of the transformed system of Augustan Rome were often a far more important aspect of religious Romanization than any direct worship of the emperor.

    The sources for this chapter are different in their emphasis from those we have used before.13 In attempting to reconstruct provincial viewpoints on the processes of Romanization in the provinces, we have ample evidence from one section of the provincial population only – the educated writers from the Greek world in the first two centuries A.D., some of whom (like Plutarch and Lucian) discuss various aspects of Greek and Roman religion and their interaction.¹⁴ Otherwise the bulk of the evidence for Roman influence on religion in the empire comes from inscriptions or from visual images: sculptured reliefs depicting gods or emperors, for example, may provide evidence for the native or Roman characteristics of a particular deity, or for how an emperor is imagined within the divine system; inscriptions reveal the names of the gods, the religious offices and sometimes the particular rites of towns in Italy and the empire. But at the same time these objects may challenge interpretation. How can you tell, for example, if a statue of Jupiter found in a provincial town is the result of Roman imposition or of enthusiastic provincial imitation of Rome? How can you know whether a temple to Roman deities in a distant province was the focus of loyal worship by the provincial community or the focus of their resentment at Rome’s dominance? Besides, it is all too easy to patronize provincial aspirations and ideology. Is a rough, ‘unclassical’, Celtic image of a Roman deity to be written off as a demonstrably naive failure by the local craftsman to reproduce metropolitan style? Or is it motivated by a desire to assert local, ‘tribal’ difference from the dominant, imperial, classical culture?

    This chapter emphasizes the changes wrought under Roman rule; so we pay little attention to the relatively unchanging civic cults of the Greek east that continued throughout the period. Yet especially in the west, but also in parts of the near east, the evidence for pre-Roman religious life is very scanty – with few, if any, surviving pre-Roman inscriptions, and few if any images carved in stone, let alone any trace of literary accounts.¹⁵ It is often impossible, then, to specify precisely the individual changes brought about by Rome and Roman influence. We can, however, deploy the evidence we have to assess the overall impact of Rome on the religious life of the empire, and the factors which intensified or diminished that impact – both east and west, from the classical world of the Greek city states to the tribal societies of Britain and Gaul.

1.   Roman religion outside Rome

Throughout the empire the Roman authorities tended to promote a variety of Roman religious practices. Whatever differences there were in the impact of Rome, it was a general rule that governors and other Roman officials favoured Graeco-Roman (rather than ‘native’) gods in whichever province they were stationed, and the governor’s staff regularly included haruspices for the proper interpretation of sacrifices performed on the Roman model. Even more important was the expectation that governors right across the empire would ensure that the provincials, presumably in the context of the provincial assemblies (which consisted of representatives of the individual towns), performed the annual vota (vows followed by sacrifice) for the emperor and the empire.16 Evidence for this is widespread: the practice was recorded by the Christian writer Tertullian in North Africa; coloniae in southern France and Dacia offered vota; a town in Portugal made a dedication to the emperor as a result of the annual vow.¹⁷ Even rabbis in Palestine noted the prevalence of the practice; and Greek cities too sacrificed annually ‘on behalf of the emperor’ – even though such ‘vows’, in the technical sense of promising a sacrifice if something did (or did not) happen, were a peculiarly Roman practice in the context of public civic sacrifices.¹⁸ These vota were in fact an institution common to all types of provincial community – which is particularly striking given that, as we shall see, communities of different statuses and culture had very different relations to Roman religion.

    On the other hand, Roman provinces were not Rome; and the religious rules governing practice in Rome itself did not apply directly elsewhere in the empire. Instead Roman authority was mediated through the governor, according to similar – but not always exactly the same – principles as operated in the city. So, for example, according to Roman lawyers, land in the provinces could not, strictly speaking, be religious or sacred as it was not consecrated by the authority of Rome; it could only be treated as religious or as sacred.¹⁹ But inevitably such legal rules were not always a clear guide to religious practice. The problem of the two categories of land faced Pliny when he was governor of Pontus-Bithynia. In relation to religious land (used for burials) he asked the emperor, as pontifex maximus, whether he could permit people to rebury the bodies of their relatives which had been disturbed by river erosion. Trajan replied that provincials could not be expected to consult the pontifices, and that local custom should be followed.20 Despite Pliny’s uncertainty, in the provinces emperor and governor filled the role occupied in Italy by the pontifices. On another occasion, Pliny wrote to the emperor about sacred land. He enquired whether it was religiously proper for the town of Nicomedia to move a temple, though, to his surprise, there was no ‘law’ which laid down the location of the temple or other conditions applying to it – in Roman terms, that is, no ‘law of dedication’, laying down the location and other conditions of the temple. Trajan pointed out that only Roman and not foreign territory could receive such a law.²¹ The actual Roman rules did not apply to ordinary provincial land, but governors were told firmly in the instructions (mandata) issued to them by emperors to preserve sacred places.²² The role of the governor included supervision of religious matters along essentially Roman guidelines.

    For the rest of this section, however, we shall be concentrating on the different legal and constitutional statuses that affected how the influence of Rome was felt in different communities. We shall explore Rome’s control over the religious practices of the empire and the adoption of Roman religious practices outside the city in a sequence moving out from Rome: Italy, the army, and provincial communities with Roman status, coloniae and municipia. All these, with the partial exception of municipia, consisted of Roman citizens, and all held some consistent patterns of religious practices in common. At the same time we shall show the hybrid complexities that cut across this relatively simple pattern: the very different forms of accommodation with Rome that were attempted even by communities of the same constitutional type; the different significances that could attach to the ‘same’ religious institutions, rituals and symbols.

    Italy formed the core of the empire. All the free-born population of the peninsula up to the Alps had been Roman citizens since the time of Julius Caesar. Italy was not a ‘province’ (it was not, for example, subject to Roman taxation); but remained, in principle, a collection of self-governing communities. Some towns preserved their religious institutions from pre-Roman days, including practices utterly at variance with Roman traditions – burying their dead within the city limits, for example, which was strictly prohibited at Rome.23 However, at least when it suited them, Roman officials did claim authority over the religious institutions of Italy. This is neatly illustrated by an incident under Tiberius, when the equestrian order in Rome vowed a gift to ‘the temple of Equestrian Fortune’ for the health of Livia, only to discover that there was no such shrine in Rome itself. But such a temple was discovered at Antium, a town fifty kilometres south of Rome, where (according to Tacitus) the senate decided that the gift could be placed, ‘since all rituals, temples and images of the gods in Italian towns fall under Roman law and jurisdiction’. Neither the senate nor any other group of Roman officials did actually exercise day to day control of Italian shrines; but in this case at least it was convenient (and presumably seemed plausible) for them to stake a theoretical claim to Roman power over the religious institutions of the rest of the peninsula. Likewise when the Roman authorities moved to expel undesirables from the city, they normally specified expulsion from both Rome and Italy. The Roman college of pontifices also sometimes gave permissions to Italians to repair tombs or move corpses from one tomb to another, and, soon after the death of Julius Caesar, a Roman law seems to have instructed Italian communities to set up statues of divus Julius. Of course, in practice many Italians must have repaired tombs without the permission of the Roman priests, and we do not know how many obeyed the order to erect statues of Caesar; but in both these cases the parade of Roman authority over the peninsula as a whole is significant.²⁴

    The uniquely close relationship between Rome and the rest of Italy is visible most clearly in a series of documents we have already discussed from different points of view in earlier chapters: the surviving painted and inscribed calendars of festivals. In all, forty-seven such calendars survive (often in small fragments), dating mainly to the reigns of Augustus and Tiberius, and all but one coming from Italy (the exception being a colonia in Sicily).²⁵ Of these forty-six imperial Italian calendars, twenty-six are from Rome itself (many, it seems, having been for the use of private associations in the city); the other twenty come mostly from the towns in the vicinity of Rome – generally on public display in the civic centres of the towns concerned.²⁶ The level of detail given in the Italian calendars varies greatly, but all differ from and apparently replace earlier, pre-imperial, Italian calendars and all are mutually compatible, recognizably versions of the same overall system of religious time-keeping. Strikingly they give practically no festivals peculiar to the local city, but only differing selections from the official festivals of the city of Rome. This raises acutely the question of the relationship between the calendar and religious practice. Would it really have been the case, for example, that such rituals as the Lupercalia, so closely tied to the topography of Rome, would have been celebrated in all the Italian towns that chose to mark it on their calendars? And if it was not celebrated, then what function did those calendars have? Why display in the local forum a series of festivals that your own town did not actually carry out? However we choose to answer such questions, it is clear that some towns in Italy – and this seems not to have been the case in the provinces – chose to parade the official Roman religious calendar as (or as if it were) the framework for their own lives.27

    Some of the religious links between Rome and the Italian towns derived directly from historical links in the distant past. The ancient communities nearest Rome, for example, had been Rome’s ‘Latin’ allies in the early republican period and shared a variety of common cultural forms stretching back almost into prehistory and to Rome’s status as a ‘Latin’ city. Thus Alba Longa, Lavinium, Tibur and other Latin towns had one or more of the following priests: flamen Dialis, Vestal Virgins, rex sacrorum, and Salii.²⁸ The Salii and the rex sacrorum (and, once, the flamen Dialis) are also found in a few towns in northern Italy, but otherwise these offices appear almost nowhere else in the Roman empire, except in Rome itself. Interpretation of this common culture could vary, of course. In those early days the religious influence did not necessarily flow from Rome outwards; and some Italian communities might choose to give themselves (not Rome) priority in the relationship – suggesting that if they shared some of Rome’s most distinctive practices, that was because Rome had adopted them from the Latins, not the other way round. A shared religious history, in other words, could be the focus of rivalry and conflicting interpretations.

    But historical links could also be invented. In the early empire, ancestral ties between the Latin towns and Rome were emphasized by a new flowering of such (allegedly) ancient cults. For example, at Lavinium 30 km south of Rome, where there was no settlement in the late Republic, Italians of equestrian rank from the reign of Claudius on held a priesthood which supposedly continued the cult of the Lavinian Penates (the deities that Aeneas had brought from Troy), participating at ceremonies of the Latin League on the Alban Hill, and, on one occasion at least, renewing the ancient treaty with Rome, first made in the fourth century B.C.29 In the second century A.D., after the renewal of civic life at Lavinium, local men from the town began to hold the office, which is attested until the middle of the third century. This is a case of ancestral similarities between the religious practices of Rome and Italy being re-emphasized in the early empire through what was almost certainly an invented tradition. For some observers and participants, no doubt, it was all a picturesque, quaintly antiquarian show; but such instances of constructive archaism also served as another way of representing the religious links between Rome and its Italian neighbours.

    Outside Italy, the body of men which most clearly stood for Rome was the army. In the professional standing army, established for the first time under the emperor Augustus, Roman citizenship remained a precondition for service in the legions (though it might be granted at recruitment); they were gradually recruited from a wider and wider area, so that, by the early second century A.D., they had only a tiny proportion of men from Italy itself – but the rules of citizenship governing recruitment continued to emphasize that the men were troops in the service of Rome. The other main body of troops, the auxiliaries, were not Roman citizens in the early empire, though they were commanded by officers who were citizens and they themselves might receive citizenship on discharge; later it became not uncommon for those who were already Roman citizens to enlist in the auxiliary forces. The official religious life of both sets of troops was predominantly Roman; though that could mean different things and be interpreted in different ways.

    The specifically Roman character of official religious life in the army was enshrined in an official Roman calendar which specified the year’s religious festivals for both legionaries and auxiliaries; this was different in form from the civic calendars of Italy we have just discussed, but (crucially, as we shall see) shared some of their major celebrations. The archives of an auxiliary cohort, the ‘Twentieth Palmyrene’, stationed at Dura Europus on the eastern Euphrates frontier, included a papyrus copy of this calendar which still survives.³⁰ The calendar was written in Latin, the official language of the army, and in neat capital letters throughout. This particular copy obviously received considerable use before it was discarded; the frequent rolling and unrolling of the papyrus had distorted the original shape of the roll and two patching jobs had been necessary. It was certainly not just an official ordinance kept in the files and ignored.

    The festivals to be celebrated by the cohort demonstrate how the restructured religious system of Augustan Rome was, in a modified form, repeated in the army. First, some of the major festivals of Rome (the Vestalia, for example, or the Neptunalia) were marked on the appropriate day by a sacrifice in the army camp too; so too were the circus games established in Rome by Augustus in 2 B.C. at the dedication of the temple of Mars Ultor. The ‘Birthday of Rome’ also appears, presumably added to the calendar under Hadrian, to replace an earlier celebration of the Parilia (as we saw in chapter 4). A second group of celebrations honours the reigning emperor, his family and predecessors. The deified emperors and empresses whose birthdays were celebrated by the army seem to correspond exactly to those whose birthdays were celebrated at this time by the Arval Brothers at Rome, marked in their inscribed record; that is, the cohort’s calendar was in step with at least one version of official practice in Rome. And on 3 January vows were taken for the well-being of the emperor and the eternity of the Roman empire, with sacrifices to the Capitoline triad – again in accordance with practice at Rome itself.³¹

    The forms of ritual prescribed for the army unit were also the same as those performed in Rome – including (as in state cult) both animal sacrifices and offerings of wine and incense (supplicatio). The rules for animal sacrifices were also for the most part identical to those followed in the capital: male deities were offered male victims, and female ones, female victims; Mars received a bull, as was standard; the genius of the emperor received a bull and the divi received oxen. The overlap with the Arval record is again striking: the military calendar even used the same abbreviations for ‘ox’ (‘b.m.’) and ‘cow’ (‘b.f.’) as the Arval Brothers, abbreviations which are not otherwise attested outside Rome. On the other hand, there were a few differences too. At Dura deified empresses (divae) received only supplicationes, not cows in sacrifice as in Rome.

    Of course, both officers and men also worshipped other gods, apart from those honoured and listed in the calendar. We have already noted, for example, the popularity of Mithras in the Roman army. There was, in fact, a temple of Mithras at Dura which was used by soldiers – although no mention is made of the god in our document; and we shall return below to other examples of the varied religious life of a military unit. Presumably this calendar was not intended to regulate the private religious worship of individual soldiers; rather it formed the basis of the official cycle of ceremonies carried out by (or on behalf of) the cohort as a whole, as a Roman institution. Although the Dura calendar is the only surviving example, it is a fair guess that all army units possessed, and in principle followed, a ritual calendar on this model – which may, in fact, with alterations and adaptations, go back to a calendar first issued to the legions under Augustus himself. Rome, in other words, made a version of its own religious system the basis of the religion of the Roman army.

    This guess is supported by a variety of evidence suggesting a common basis to the official religious activity of army units. There is certainly a good deal of material to show that the gods and festivals recorded in the Dura calendar did recur in other military contexts throughout the army.32 For example, from the Roman fort at Maryport, just south of Hadrian’s Wall, a series of 21 altars and plaques survives officially dedicated to Jupiter Optimus Maximus in the course of the second century A. D. by three different regiments, perhaps on 3 January when vows were made ‘for the well-being of the emperor and the eternity of the empire’. Or again, at another fort near Hadrian’s Wall (the third century A.D. legionary supply base at Corbridge) we can see the traces of the rituals that the legion’s official calendar prescribed, even though we do not possess any written version of the document itself. Inscriptions and carvings from Corbridge attest many of the cults known in the Dura calendar: Jupiter, Victoria, Concordia; and the ‘rose festival of the standards’, which appears twice in the Dura calendar, is depicted there on a decorative relief which probably formed part of the shrine for the standards. The specifically Roman focus of the legion’s official religious activity is further attested by a shrine in the base which clearly evoked the foundation of Rome: a relief carving on the pediment showed the wolf suckling Romulus and Remus, This scene was repeated in another camp halfway across the empire: an early third-century inscription from a fort on the Danube refers to the dedication of a signum originis, that is a statue of the wolf with Romulus and Remus.³³

    In general the arrangement and personnel of army camps throughout the empire conformed to the system implied by the Dura calendar.³⁴ In the centre of the camp, at the rear of the headquarters building, was the shrine which housed the legionary standards and imperial and divine images. This shrine is actually called a Capitolium on one inscription.³⁵ In front of the headquarters building was a platform where the commander took omens from the flight of birds. The army had on its staff specialist religious personnel: the victimarii, who killed the animals, and the haruspices, who took the omens from the animals’ entrails. Scenes of military sacrifice, on exactly the standard Roman pattern and attended by the appropriate specialists, are depicted on reliefs from Trajan’s column in Rome (in scenes from Trajan’s campaigns in Dacia, modern Romania) and from the Antonine Wall in Scotland.36 (Fig. 7.1) This re-enactment of the most central and characteristic ritual of Roman official religion was one of the most striking ways in which the army paraded that religion across the known world; in the case of Trajan’s column, that diffusion was then monumentalized in Rome itself, displaying in the very centre of the city the Roman army ritually enacting their ‘Romanness’ on the frontier.

         Fig. 7.1 A scene from Trajan’s column showing the ritual purification of the Roman army, before an assault on the Dacians. A bull, sheep and pig – to form the sacrifice known as suovetaurilia – are led in procession round a Roman camp and in through a gate; they are followed by the victimarii (shown semi-clad) whose function was to perform the killing of the victims. Within the camp the emperor Trajan, head duly veiled, pours a libation over an altar in front of the standard-bearers. (Height, c. 0.80m.)

    The official prescription of Roman gods did not, however, prevent the worship of other gods. After all, by the second century A.D., most Roman soldiers came from places other than Rome and Italy and may well have wished to maintain their original identity within a Roman framework; their religious interests must have been as cosmopolitan as those of any group of ‘Romans’. For example, an auxiliary cohort from Emesa in Syria, which was raised in the 160s and served for a long period (from the later second to the mid third century A.D.) in Pannonia on the Danube frontier, maintained a dual allegiance to Roman and to Syrian gods, to Jupiter Optimus Maximus and to Elagabalus.37 This was not just a matter of individual devotion. Both gods were worshipped officially by the cohort as a whole as well as by individual officers and men. In contrast, the civilian community outside the camp honoured neither Jupiter Optimus Maximus nor Elagabalus, but a variety of local and other gods.38 By the second century most soldiers served close to their homeland (the distance of the Emesene cohort from their native territory was unusual in this period), and soldiers and even high-ranking officers sometimes made offerings to local deities.³⁹ This must have made it easier, for those soldiers who wished, to maintain all kinds of ‘native’ traditions of worship. But even so the dominant religious system of the army as an institution remained modelled on that of Rome.

    After the army, it was Roman coloniae that mirrored the religious institutions of Rome itself most closely. This is clearly illustrated by the regulations for the colonia founded by Caesar in 44 B.C. at Urso in southern Spain that we noted in chapter 3. The surviving copy of the regulations is a re-inscription of the original rules, dating from the late first century A.D. – effectively reaffirming the peculiarly Roman nature of Urso more than a century after its foundation (perhaps to maintain her superiority over other Spanish towns which had received the ‘Latin right’).⁴⁰ This Roman character is evident in almost all the regulations for the life of the colonia, but is particularly striking in the detailed, and well-preserved, clauses of the document that refer to priesthoods. As we saw, the two main priestly groups were pontifices and augures; and ‘as at Rome’ (the regulations specifically refer to the model of Rome) the priests were to be free from military service and public obligation; they also had the right to wear special clothes at games and sacrifices and to sit at games in the same privileged seats as the town-councillors. Their functions too were similar to those of their Roman prototypes: the augures, for example, were to have jurisdiction over the auspices and all matters concerning them.⁴¹

    Much the same priestly organisation and duties are found in all other coloniae. As late as A.D. 322, some 200 years after it had become a colonia, an embassy from Zama Regia in north Africa to the governor consisted of ten men, of whom the first named were the four pontifices and two augures; similarly, at Timgad (also in North Africa), over 250 years after the foundation of the colonia, the council included the four members of the colleges of pontifices and augures as well as other priests of the cult of the emperor.42 An inscribed altar from the colonia of Salona (on the eastern Adriatic coast) gives a glimpse of a local pontifex at work: the inscription records that at the dedication of the altar a pontifex dictated the words to the local magistrate – just the procedure that was adopted at Rome.⁴³ Of course, local priestly functions could never be exactly identical to those in the capital: the authority of both pontifices and augures, for example, was restricted by that of the governor (who himself had the right to authorize the moving of corpses – in Rome a pontifical responsibility). But overall the symbolic structures of coloniae emphasize their status as ‘mini-Romes’ from the very moment of their foundation, conducted with rites that echoed the mythic foundation rituals of Rome itself: the auspices were taken and – like Romulus in the well-known myth – the founder ploughed a furrow round the site, lifting the plough where the gates were to be;⁴⁴ within this boundary, which replicated the pomerium of Rome, no burials could be made; the land within the pomerium was public land which could not be expropriated even by the local council.⁴⁵

    Much of this similarity is due to direct Roman initiative. The original religious regulations for the coloniae and the form of their foundation rituals were devised by the Roman authorities; and coloniae in the late Republic and early empire may also have received specific instructions, directly from Rome, on the establishment of new Roman practices. Priests of the deified Julius Caesar (flamines divi Julii), who was officially deified at Rome, are found outside Rome only in Roman coloniae, in both the eastern and western parts of the empire. A relief honouring one such flamen in the colonia of Alexandria Troas in north west Turkey even shows the distinctive hat (apex) worn by flamines in Rome.46 The fact that this cult is found only in coloniae may suggest that they were responding to official instructions from Rome, issued perhaps in 42 B.C. when Caesar’s deification was finally ratified. Certainly in A.D. 19, when the senate passed a lengthy decree on the funeral honours for the emperor Tiberius’ son Germanicus, provincial coloniae were explicitly instructed to set up a copy, and their magistrates were barred from transacting public business on the anniversary of the death of Germanicus.⁴⁷ At this period, coloniae were expected to move in step with Rome.

    But even without direct instructions, throughout their history coloniae might themselves choose to follow, and parade, Roman models. This was clearly the case in the inscription from Salona, whose introductory formulae involving pontifex and magistrate we have just mentioned. What follows this introduction are the regulations governing the rituals at the altar (the ‘law of dedication’, which Pliny had wrongly expected to find in a non-Roman town):⁴⁸ some of these are spelled out; for the rest it is stated that they ‘shall be the same as the law pronounced for the altar of Diana on the Aventine’. This ancient set of rules in the Aventine sanctuary in Rome originally governed the relations between Rome and her Latin allies.⁴⁹ Here the colonia of Salona, some 170 years after its foundation, chose to adapt this model to articulate its own ritual rules and at the same time to emphasize its privileged relationship to Rome. Two other similar documents from Roman coloniae, one from Narbo in southern France, the other from Ariminum (Rimini) in Italy, refer to the Aventine model in framing their own cult regulations.⁵⁰

    How exactly the population of more distant coloniae gained access to the Roman ritual knowledge implied by these and other rules is unclear. It is certainly possible that in some cases they had very little access to that knowledge; and that Roman models were more of a display than a rule book to be followed to the letter. On the other hand, the governor’s staff and army units stationed in the provinces included Roman religious experts (haruspices and victimarii), who might have been able to offer advice on points of detail; so too might the governor himself.51 More puzzling, in fact, is the question of what general idea of ‘Roman religion’ (if, by that, we mean the religious institutions and practices of the capital) the population of such places would have had – who, though Roman citizens, might have been resident hundreds of kilometres from Rome for generations. One possible channel is Varro’s Religious Antiquities. This treatise, of the mid first century B.C., remained even under the empire the only general work on the Roman religious system. That provincials did turn to it for inspiration is suggested by the effective (polemical) use made of it by the Christian Tertullian, writing in north Africa.⁵² But even this book will have been hard going, and difficult to apply to particular local issues. The problem is a useful reminder, however, that any such statement as ‘the coloniae imitated the religion of Rome’ is always liable to be a shorthand for ‘the coloniae imitated their own image (or conflicting images) of the religious institutions of the capital’.

    This is not to say that the population of a colonia was always in danger of ‘misunderstanding’ the religion of Rome; but rather that imitation of the religion of the capital must in practice always have been a creative process, involving adaptation and change. Two altars from the colonia of Carthage illustrate different ways in which images of (and derived from) Rome might be represented and constructively reinterpreted in a colonia. One, a grand public altar, of Augustan date, stood on the outskirts of the Augustan colonia. (Fig. 7.2) Two of its large sculpted panels survive. One represents Mars Ultor standing between Venus Genetrix and a figure probably to be identified with divus Julius. The other shows a seated female figure, with children in her arms, her lap full of fruit, animals at her feet. This figure is closely based on the scene of ‘Earth’ on Augustus’ Ara Pacis in Rome, though the personified breezes which flank the figure on the original monument have been replaced with a sea god (a Triton) and a female divinity carrying a torch.⁵³ On the other face, the figures of Mars, Venus and (probably) divus Julius are almost certainly derived from statues (not necessarily cult-statues) from the temple in the Forum of Augustus; Mars and Venus are even represented on statue bases, marking them out as statues, not simply deities. It is likely that the altar was produced locally, in Carthage, though we do not know how knowledge of the Roman monuments was disseminated. That is, the colonia here, in its own religious monuments, was explicitly combining themes from two of the major Augustan buildings at Rome itself; with direct ‘quotations’ (as in the statues), adaptations (as in the new flanking figures for Earth, perhaps deities with a particular local relevance) and, of course, a wholly new juxtaposition of scenes. The colonia was expressing its own version of Roman identity, through a creative imitation of Rome itself.⁵⁴

         Fig. 7.2 Two reliefs from altar in Carthage. (a) Left to right: Venus, Mars; probably divus Julius, as strongly implied by the star on his forehead, rather than an Augustan ‘prince’. Mars must be Mars Ultor, the cult-image from his own temple in the Forum of Augustus (see 4.2a). If, as is now suggested (see Steinby(1993–) II.291–2 for discussion), the temple had only a single cult–image, then Venus and divus Julius must have been located elsewhere in Mars’ temple – as statues, but not cult-images. (Height, 0.98m; width, 1.13m.) (b) ‘Earth’, flanked by deities. For the Roman prototype, see fig.4.6. (Preserved height 0.80m., width 1.13m.)

    The second, and much smaller, altar was found close to an inscription recording the building of a temple to the Augustan family – by a wealthy individual (probably an ex-slave) at his own expense, on private land near the centre of the colonia; the donor (so the inscription also records) became perpetual priest of the new cult. Almost certainly the altar belongs to this temple, and dates to the last decade of Augustus’ reign.55 (Fig. 7.3) The scenes on the four sides of the altar again reflect central Roman themes. On the front is Roma, seated on her armour, with a miniature shield and winged victory on her outstretched right hand. The figure may be a version of the Roma (now very fragmentary) which balanced ‘Earth’ on one side of the Ara Pacis; but, if so, the altar with its globe and cornucopia in front of her must be adaptations of the original design; while the figure of Victory on her hand carries a shield modelled on the ‘Shield of Virtue’ awarded to Augustus by the senate.⁵⁶ The scenes on either side of the altar show Aeneas leading Ascanius and carrying his father Anchises out of Troy (a scene immortalized in Virgil’s Aeneid, and represented in the Forum of Augustus);⁵⁷ and sacrificers, who perform their ritual in the distinctively Roman manner with togas over their heads (perhaps recalling similar scenes on the Lares altars of Rome). Finally, on the rear of the altar, to match the figure of Roma, is Apollo (whose temple Augustus had built on the Palatine), seated in front of a tripod. This altar is also a creative juxtaposition of Augustan themes and images, Roma, Aeneas and Apollo, perhaps again influenced by specific monuments in the capital. In this case, the imagery particularly serves the interest of the donor, ineligible (as an ex-slave) for membership of the local council, but nonetheless here proudly asserting his position within the community and within the Roman imperial world. If indeed we read (as may be intended) the main officiant in the sacrificial scene as the donor himself, we see him displaying his own (Roman) piety within a version of imaginary Rome.

         Fig. 7.3 Altar from Carthage. (a) Roma. (b) Aeneas carrying Anchises and leading Ascanius. (c) Sacrifice at an altar, (d) Apollo. (Total height, 1.18m.; width, 1.16m.; depth 1.03m.)

    Imitation of Roman religion in the coloniae was, then, less rigid than the regulations we started with might have suggested. And although coloniae in general borrowed, sometimes closely, from Rome, there was no immutable blueprint. Different coloniae were Roman in very different ways and made different kinds of accommodation with the central imperial power. This is illustrated very clearly in their choice and layout of temples, and (particularly) in the varied distribution of Capitolia throughout the different coloniae.58 A Capitolium, in the sense of a temple of Jupiter, Juno and Minerva on the model of the Capitoline temple at Rome, provided a very clear link with the capital. Some coloniae certainly built Capitolia immediately at the time of their foundation: there are second-century B.C. coloniae in Spain with their own Capitolia, and the regulations from Urso specify major games in honour of the Capitoline triad (though we do not know if the town actually had a Capitolium).⁵⁹ Other coloniae built Capitolia only later, if at all. The colonia of veteran soldiers at Timgad, for example, established in A.D. 100, did not include a Capitolium in the forum, where there was only a small temple which may have related to the emperor; and, when building of a Capitolium began circa A.D. 160, it was sited outside the original area of the colonia.⁶⁰ The options were even wider when a colonia was not founded completely afresh, but when an existing town received some Roman colonists – and so took on the status of a colonia. For example, at Heliopolis (modern Baalbek in Lebanon), which lay in the territory of the colonia of Berytus, a great new civic temple was begun in the Augustan period when ex-soldiers were settled there. The basic design is Roman (with some local adaptations) and the expense of construction, plus the use of imported red Egyptian granite for the portico, strongly suggests financing from Rome, even from the emperor himself. But the name of the main deity, Jupiter Optimus Maximus Heliopolitanus, shows clearly how even the Capitoline god could absorb and display the influence of local culture and conditions.⁶¹

    Towns with the status of municipia (where local citizens had the so-called ‘Latin right’ and some even full Roman citizenship) shared some of the Roman religious features of coloniae; their principal priesthoods, for example, were named after, and modelled on, Roman institutions –pontifices , augures and haruspices. And from the second century A.D. onwards municipia in north Africa also began to build their own Capitolia; Thubursicu Numidarum in Algeria, for example, which became a municipium after A.D. 100 (perhaps c. A.D. 106), dedicated a Capitolium in A.D. 113. A cult that in the first century A.D.had been confined to coloniae (and Rome itself) was taken over by municipia as part of their display of Roman status. But interestingly, that sequence may also be reversed; and on more than one occasion we can see the building of a Capitolium as part of a claim for Roman status (rather than a boast of Roman status already acquired). At Numluli, which lay in the territory of Carthage, some local citizens who had achieved high status in Carthage dedicated a Capitolium for the local Roman citizens and for the village itself. The building of the temple clearly displayed allegiance to Rome, but it was almost certainly a part of an attempt to gain the status of municipium for Numluli which it certainly later held. Similarly at neighbouring Thugga, which also consisted of a group of Roman citizens and a village in the territory of Carthage, a Capitolium was built at just the time that the group of Roman citizens was granted imperial permission to receive legacies; the building of the Capitolium, with its Roman-style cult of the Capitoline triad, was presumably intended to promote their recognition as an independent community.62 Roman religious institutions in the provinces were not merely reflections, then, of different levels of Romanization; they were also useful counters in the competition for prestige, honour and status that was one of the defining features of provincial culture across the Roman world.

    Roman citizens did not, of course, live only in coloniae and municipia; even in the early empire, when Roman citizenship was a privilege virtually restricted to members of the élite, groups of citizens were found (as at Numluli and Thugga) outside communities with any formal Roman status. In many respects these citizens would have lived and worked indistinguishably from their non-citizen neighbours; but for some purposes they might have formed distinct groups within their non-Roman communities. The religious activity of these groups was, no doubt, one of the ways in which they re-affirmed and displayed their ‘Roman’ status; and at the same time it must have been one of the channels that spread specifically Roman religion more widely through the provinces. At Thinissut, a non-Roman town in north Africa, the Roman citizens who traded there made a dedication to Augustus god’.⁶³ They probably had in mind the living emperor, who (as we have seen) was not usually the recipient of direct dedications in Rome itself; but, in this alien context, worship of the emperor may have served to mark the boundary between Roman citizens and the non-citizen subjects of Rome. The position of the dedication may itself be significant: for the inscription comes from a site overlooking a Punic sanctuary of Baal and Tanit. This juxtaposition is in one way a striking illustration of the varied religious culture of this small African town – but at the same time it must have served to emphasize the difference between the Roman cult of the emperor and the Punic traditions of Baal and Tanit.

    Individual Roman citizens too could adopt similar strategies. At Vaga, another non-Roman town in north Africa where there had long been Italian traders and Roman citizens, one Marcus Titurnius Africanus restored a shrine of Tellus (Earth), dating the record – in conventional Roman form – by the consuls of 2 B.C.; and in the Greek city of Nicaea (in Bithynia) an Italian trader dedicated statues of the Capitoline triad to the local god, albeit with a Greek inscription.64 We also read of celebrations of the festival of Saturnalia by Roman students studying in Athens – or by one retired soldier living in the Egyptian Fayum, who wrote to his son to order ten cocks from the local market for the festival.⁶⁵ Individuals no doubt reminded themselves (as much as the community at large) of their Roman status by making specifically Roman religious gestures.

    This general pattern of Roman religious influence in the empire – with its concentration on those groups and places with some formal Roman status – is cut across by a whole variety of different factors. We conclude this section by exploring two of those complexities: first the spread of different forms of what could count as ‘Roman’ cult among provincial communities; secondly the different impact of Roman deities on the élite and non-élite.

    Throughout the Roman world there were wildly different images of ‘Roman’ religion; as we saw in our discussion of coloniae, different communities in the provinces must have constructed their own versions of what they thought was Roman. Of course in earlier chapters of this book we have raised just this question in relation to Rome itself: what was to count as official Roman religion? The negotiability of that category even at the very centre of the Roman world, the changing definitions of ‘Romanness’, is obviously relevant to the ‘export’ of Roman religion to provincial communities – as is clearly illustrated in the cult of Magna Mater.

    This cult, as we have seen, was ‘officially’ introduced to Rome in the late third century B.C. From there, it became a common feature of the towns of Italy and the provincial coloniae and municipia in North Africa, Spain, the Danube region and especially Gaul;⁶⁶ at first the cult members seem to have been limited to ex-slaves and others of low formal status, but from the mid second century A.D.local dignitaries too are found within it. In other words, a cult of eastern origin spread through the Roman world not from its eastern ‘home’, but from Rome and so as a ‘Roman’ cult. By the mid second century A.D., in fact, the cult in Italy and at least the western provinces was under the general authority of the Roman priests, the quindecimviri, who had originally been responsible for the introduction of the cult to Rome. The priests of Magna Mater in Italy and Gaul are sometimes even given the title ‘quindecimviral priests’; and an inscription from the territory of the colonia at Cumae in southern Italy preserves the text of a letter from the college of quindecimviri in Rome, authorizing the local priest of Magna Mater, whom the town had recently elected, to wear the special armlet and crown (the priestly badge of office) within the territory of the colonia.67

    There is also a striking reference to a direct link with the city of Rome in an inscription from the colonia of Lugdunum (modern Lyons), recording the performance of a taurobolium (the cult’s characteristic sacrifice of a bull). This text refers to the ‘powers’ (‘vires’ – probably the genitals of the sacrificed animal) being ‘transferred’ from the Vatican sanctuary (‘Vaticanum’). It is not exactly clear what that means. It seems unlikely (though not impossible) that the bull’s genitals should have been taken from the Vatican in Rome to Lugdunum. More likely, perhaps, there was a ‘Vatican sanctuary’ in Lugdunum itself – but, if so, its name (and no doubt other aspects of its cult) derived directly from the Roman model.⁶⁸ In either case, the direct or indirect dependence of the Lugdunum cult on Rome was unusual. Unlike, for example, the cults of Isis or Mithras, which had no ‘headquarters’ there, the cult of Magna Mater claimed authority from the centre, which had officially adopted the original cult.

    Significantly the inscription also states that this taurobolium was performed (as many were at Rome) on behalf of the emperor, as well as the local community – another link with the capital.

    The cult of Magna Mater exposes the shifting ambiguities of Roman status, and the expanding definition of what might count as Roman in the provinces: under the general authority of the quindecimviri, she counted both as a ‘foreign’ god (the quindecimviri had, as we have seen, specific responsibilities for cults of Greek origin in the city of Rome) and as a ‘Roman’ god (overseen even in provincial contexts by Roman priests). One of the most striking features of Roman imperialism is that (especially in the west) the spread of Roman religious culture through the empire was marked by the diffusion of cults that in the context of Rome itself claimed a ‘foreign’ origin. It was not only the Capitoline triad, but Magna Mater and Mithras, who could stand for ‘Roman’ religion in the provinces.

    Wealth and power are also factors that we have not so far considered in plotting the patterns of Roman religious influence in provincial communities. By and large, in every kind of community the local élite tended to display less interest in local indigenous cults than in the universal deities associated with the Roman empire.⁶⁹ For example, in the colonia of Timgad, magistrates and priests in the course of the second and third centuries A.D.made a series of dedications in return for their offices to principally Roman deities: Jupiter, Victoria, Mars, Fortuna and so on; one text neatly indicates that the declared purpose of its dedicator was ‘the celebration of public religio and the embellishment of his noble city’.70 In southern France, which had been under Roman control since the late second century B.C., we can detect a significant distinction between (in one aspect at least) the religious practice of the local urban élite and those outside that group. ‘Mars’, the most prominent god of the area, appears in inscribed dedications sometimes with a range of local epithets and sometimes without: the dedications with local epithets are usually associated with dedicators who did not come from old established or distinguished families; Roman functionaries and members of the local élites are much more commonly associated with those without local epithets. In addition, 90% of the dedications to Mars with local epithets are found outside the towns.⁷¹ Even in this area, which seems at first sight to be strongly Roman in tone overall (note, for example, the name ‘Mars’), those of higher local status chose to display their relationship with gods who were more obviously part of the Roman system.

    Religious display must have been central in the competition for status, both inside and outside the local community. We can only guess how different it might have felt to make a dedication to Mars rather than Mars Alator; we can only guess what costs (as well as benefits) there might have been in publicly displaying allegiance to explicitly Roman gods. But it is clear enough that local élites expressed their own status, as against their social and political inferiors, by parading close ties with the gods of Rome.

2.  Controls and integration

The religious impact of Rome on communities without formal Roman status was quite different; it was much less a matter of imitation, much more a question of various forms of control and integration. Roman authorities moved to suppress (or ‘emend’) religious forms that seemed to be a focus of opposition to Roman rule – whenever and wherever they found them. But in other respects there appears to be a clear distinction between their approach to the pre-Roman religions in the west and those in the east. In the Greek east the civic cults seem to have continued, outwardly unaffected by Rome; but in the west local gods were transformed and integrated into the Roman pantheon. This distinction, however, raises the question of how we can identify different degrees of ‘Romanization’; what counts as the transformation of religion under the impact of a conquering power; how far we can assess fundamental changes in religious ideology from its outward forms.

    Priesthood was an area of particular concern. In four different areas, we can see how the Romans restricted the power of native priests in the provinces; in all these cases, the priesthoods were organized quite differently from the traditional model of Graeco-Roman city-states (where priests were civic officials, with strict limitations on their authority, drawn in rotation from the local élites) and, to the Romans at least, represented an alternative system of power capable of rivalling their own; often these priesthoods were based on rich and powerful temple institutions.72

    Egypt was annexed to the Roman empire in 30 B.C.; here temples, with their powerful priesthoods and widespread landholding, had traditionally been a major focus of religious and political authority. After annexation the Romans were faced with the problem of negotiating their relationship with these powerful religious institutions.⁷³ In detail, the pattern of Roman action in Egypt is very varied: some temple lands were confiscated in 24–22 B.C. (the temples either leased these lands back to be cultivated for revenue or accepted a direct state subsidy); but extensive lands were also granted (or confirmed) to a temple of Isis in the south of Egypt, and some ‘sacred land’ continued to be administered directly by religious officials. Overall, however, even where existing religious institutions were not abolished by the Romans, there is a clear trend towards increasing Roman supervision, if not direct control. All people attached to sanctuaries had to be registered from 4 B.C. onwards; from the mid first century A.D.onwards temple property and dues owed to the state by the temples also had to be recorded. The temples fell within the responsibility both of the office of the so-called ‘Idios Logos’ (the ‘Special Account’, which handled financial and administrative matters) and of the Roman governor, aided by a separate Roman official, known as the High Priest of Alexandria and All Egypt. This High Priest vetted requests to circumcise candidates for the Egyptian priesthood,⁷⁴ and adjudicated on the qualifications of those already in the priesthood. The Romans probably did not devise all these regulations entirely themselves; they represent in part at least a development of the practices of earlier rulers of the country, the dynasty of Ptolemaic kings (for after all temple power was likely to be a challenge to any secular authority, not just the Romans). However they illustrate a strong assertion of Roman control of Egyptian religious institutions – not just in general, but right down to the level of individual priests, their qualifications and marks of office.

    The religious organization of two other parts of the eastern Mediterranean also posed problems for Rome and similar methods of control were introduced. When Judaea passed from the rule of a client king to direct Roman administration in A.D. 6, the Roman governor (as earlier King Herod) appointed or dismissed the High Priest of the temple; as a further attempt to restrict its potential power base, the office was made technically an annual one – even though in practice (as a compromise, perhaps, between Roman authority and Jewish tradition) the holder was regularly re-appointed. The Romans oversaw the finances of the temple and restricted the competence of the Jewish council, the Sanhedrin. But otherwise the day to day temple organization was unaffected by Rome (though new sacrifices on behalf of the empire were now offered there); and Jews in other provinces and in Italy were permitted to continue sending a regular tax and gifts to the temple – until, that is, it was violently destroyed by the Romans following the Jewish revolt of A.D. 66–70. At that point a special, humiliating Roman tax replaced these contributions to the temple: Jews now had to pay not towards the temple at Jerusalem, but in perpetuity for the rebuilding of the temple of Jupiter Capitolinus in Rome (burned down the previous year).75

    In Asia Minor too there were temples whose priests wielded considerable secular power. Again Roman responses varied; but all tended towards neutralizing any threat that the priestly organization might represent. When a Roman colonia was established at Pisidian Antioch, for example, the sanctuary of the god Men lost its territories and its sacred slaves, but elsewhere the Romans encouraged a gradual evolution towards priesthood much more on the Graeco-Roman norm. In Cappadocia and Galatia there had been four major temple estates, inhabited by ‘sacred serfs’ (hierodouloi) and ruled by priests. Under Rome, the cults continued to be prestigious, but the communities were transformed into Greek-style city-states and the priesthoods tended to become multiple and to be held (annually) by the local hellenized aristocracy.⁷⁶

    In the west any indigenous priesthoods that predated the conquest by the Romans came under pressure. The Romans attempted actively to suppress the Druids for their ‘magical’ practices and promotion of superstitio, though repression may, in fact, have increased their self-consciousness and cohesion.⁷⁷ In other cases the message to local élites was clear, even without drastic action by the Roman authorities: local styles of priesthoods were transformed into a Roman pattern. Occasionally, the ancient names were preserved, but the ubiquitous Roman titles of flamen and sacerdos in towns in the Latin west sometimes at least were the result of a reinterpretation of indigenous priestly offices. Civic priesthood on the Graeco-Roman pattern was the norm and as a result, with the exception of the Druids, no indigenous priestly group in the west appears to have posed a threat to the Roman order.

    Roman forms of control operated in broadly similar ways, with broadly similar aims everywhere; but in other respects the effects of Roman rule on religious life in the east appear very different from those in the west. In mainland Greece and Asia Minor, where Greek language and culture were dominant and respected by Rome, the religious life of the towns did not, to all appearances, change radically under Roman rule. In Athens, for example, the central cult remained that of Athena Polias on the Acropolis, and the Eleusinian Mysteries continued to have immense prestige.78 Since time immemorial the celebration of the mysteries had involved a grand annual procession between the city of Athens and the sanctuary at Eleusis. In A.D. 220 the people of Athens in fact voted to enhance the grandeur of the procession by increasing the participation of the youth (ephebes) of the city: ‘Since we continue now as in previous periods to perform the mysteries, and since ancestral custom along with the Eumolpidae <sc. the Athenian clan with charge over the mysteries> ordains that care be given that the sacred objects be carried grandly here from Eleusis and back from the city to Eleusis...’⁷⁹ ‘Ancestral custom’ could be the sanction not for fossilization but for the evolution of cults within a traditional framework. Earlier, probably in the Augustan period, there was a major reorganization of sanctuaries throughout Attica (the region around Athens): new leases of sacred properties were drawn up, to place the finances of the cults again on a proper footing. This ‘restoration’ of ancestral cults even involved moving several earlier temples from outlying sites in Attica in to the agora in the centre of the city – presumably allowing worship to continue in a new ‘heritage’ setting.⁸⁰ The ancestral cults formed the core of religious life in Athens (and other Greek cities): Isis and Sarapis were important as they had been in the Hellenistic period, but the other elective cults attested in Rome made little impact on Athens.

    On the other hand, there was inevitably a good deal of adaptation as a consequence of Roman conquest; at the very least, traditional cults would have taken on new meanings in the context of a Roman province. We shall see in the next section how ancestral Greek religion provided the framework for various forms of worship of the Roman emperor. It is also clear that the Roman calendar had an increasing influence in the east – though it was not here a question of simple provincial imitation of the central model. When the province of Asia decided to honour Augustus by creating a new calendar, the assembly chose to start the year not, as in the orthodox Roman calendar, on 1 January, but on another date of Roman significance, 23 September, which was the emperor’s birthday. (According to the inscription recording this change of calendar, the precise date of 23 September came as a suggestion of the Roman governor – a glimpse of the complex background that must lie behind many decisions of this kind.)81 The Roman authorities also sometimes sought to control the finances of civic sanctuaries. In the very early Augustan period a Roman official in the province of Asia issued an extensive regulation to Ephesus on religious finances, a regulation known from its subsequent revision in A.D. 44: for example, priests (who would no longer have to buy their offices) would not receive subventions from the city, and public slaves were not to dedicate to the goddess their own slaves who would then be reared at the expense of the goddess. Whether this reform was driven by Roman desire to ensure the financial stability of local cults, or by their desire to eradicate religious practices that did not conform to their own model of piety, it is a clear case of Roman intervention in a civic cult of the Greek world.⁸²

    In most cases, however, we are not dealing with a straightforward opposition between the continuity of local civic cults and Roman interference. All Roman activity in relation to the cults of the Greek world (from generous subvention to drastic eradication) must have been open to various interpretations. If, for example, a Roman official had paid for the restoration of a Greek temple, would that have counted as Roman respect for traditional civic cults? Or would it have been instead (or at the same time) a mark of Roman take-over, of Rome’s domination of those cults? At the very least, a restoration by Rome must have carried a different significance from a restoration by the local city itself. Even if not outwardly ‘Romanized’, traditional religion was now operating within a context of Roman power and empire. And it was often the local inhabitants themselves who were instrumental in parading the Roman associations of traditional cult. We do not know who placed the statue of the emperor Hadrian that (as Pausanias records) once stood within the Parthenon – but at the very least it must have been authorized by the local Athenian officials. It was, however, certainly they who decided to honour the emperor Nero, emblazoning his name across the architrave of the Parthenon.83 In both these cases we can easily understand how the Roman presence could have seemed to some like an outrageous intrusion of the imperial power into one of the most holy cult places of Greece; or, equally, like the incorporation of Rome within the venerable traditions of Greek religion. ‘Continuity’ or ‘change’ can be matters of interpretation.

    In the west, however, where Latin was the dominant language and where there was no unified and prestigious cultural system when the Romans arrived, the religious position of Rome’s subjects was very different from in the Greek east. There was, unsurprisingly, a range of religious continuities as well as resistances to the cults practised in Rome: a calendar in general use in Gaul in the late second century A.D.perpetuated local traditions, and (to judge from names of adherents that are recorded) the ‘Oriental’ cults were of little importance among the indigenous populations.⁸⁴ However, a crucial aspect of religious change (quite different from anything we saw in the east) was that indigenous gods became widely reinterpreted, by the locals and others, in a Roman form. As we noted at the beginning of this chapter, this process of transformation is difficult to plot – not least, we may now add, because the native deities generally become visible to us only under Roman rule, with increased use of writing on durable surfaces and more iconographic representations in stone. However, the excavation of a sanctuary in the Italian Dolomites, an area conquered by Rome only in the first century B.C., offers a glimpse into the changes in one sanctuary from pre-Roman times onwards.⁸⁵ Among the finds, which run from the third or second century B.C. through to A.D. 340, are bronze dippers used for drinking the sacred waters from a sulphurous spring. They were inscribed with the name of the god (Trumusiatis or Tribusiatis), initially in the local language (Venetic), then with the same name transcribed into Latin characters and, only in the most recent ones, with the Graeco-Roman name Apollo. It would be a crude oversimplification to suggest that under the sign of Apollo the cult lost all trace of its native roots; after all there is no suggestion of any major change in the rituals through this period. But at the same time it would be little short of a romantic fallacy to argue that nothing had really changed, and exactly the same native god lurked behind his new classical name. The change (or not) of language and names is not merely a cosmetic issue. At the very least, to call a god not Trumusiatis, but Apollo, was to relate the local healing god to the broader classical pantheon.

    The reinterpretation of local gods (by both Romans and members of the local communities) was widespread in the western provinces, from the north-west of the Iberian peninsula to the Danube.86 As part of his effort to describe and explain to a Roman readership the culture and society of Gaul, Julius Caesar identified and interpreted in Roman terms a number of gods (for example, Mercury and Vulcan) as characteristic of the area at the time of the Roman conquest; while evidence from the imperial period shows how these gods did indeed become assimilated to Graeco-Roman deities, verbally or iconographically – so becoming part of the mixed religious world of Roman Gaul in the combinations that Caesar himself identified, predicted or invented.⁸⁷ At the same time, the design of local temples and sanctuaries was also changed; some elements of Roman architecture were incorporated into the facades, though the overall groundplan remained basically unchanged and distinctively unRoman.⁸⁸ It is striking that in central and southern Gaul dedications to these reinterpreted local gods were not, by and large, made by local élites, who, as we noted in the case of Mars, tended to parade their connections with specifically Roman gods; nor for that matter were such dedications regularly made by Roman officials, soldiers and others from outside Gaul. The overwhelming majority of the worshippers seem to have been relatively humble people – confirming that (however exactly they are to be interpreted) these hybrid deities appealed to those who related themselves less fully to the Roman order.⁸⁹

    The complexities of the local divine hierarchies, and the changes prompted by Roman influence, can be illustrated by two examples, from Spain and Germany. In the north west of the Iberian peninsula the pattern of religious activity was very different from that on the east and south coasts, which had been a Roman province since the second century B.C. and where religious forms on a strongly Graeco-Roman model were well-established.⁹⁰ In the north and west (conquered only under the emperor Augustus) the religious picture was much more varied. So, for example, instead of the cults of the Lares Augusti, the category of Lares was taken over and joined with a variety of local protective deities; likewise, though Jupiter Optimus Maximus does appear both in official and in indigenous contexts, Jupiter alone or with various other titles was more widely worshipped. In both these cases an earlier local god (or group of gods) may have been reinterpreted in Roman guise. The way Jupiter could be reinterpreted (and then re-placed) within a local divine hierarchy is well illustrated in an inscription of the Danigi, a people in what is now Portugal, which lists the sacrificial animals due to a range of deities: Nabia Corona; Nabia; Jupiter; -urgus; and Ida(?).91 Jupiter, we should note, was in second place to the local Nabia (worshipped in two forms); and he received as sacrifice not his usual Roman ox but a lamb and a suckling calf – emphasizing his different role within the local context. There is a Roman tinge in the style of the altar and in the use of a Roman date (given by the consuls of that year), perhaps because the two local landowners who presided over the sacrifices were Roman citizens; but the complex divine hierarchy of the rural Danigi, with its specifically local and regional gods, was only partly integrated into the Roman divine system.

    The second example concerns ‘Jupiter columns’ – columns up to fifteen metres high dedicated to Jupiter Optimus Maximus, which were a common type of religious monument in parts of eastern France and Germany from the mid second to mid third centuries A.D.These columns also illustrate borrowing from Rome, reinterpreted within a local religious hierarchy. They take two different regional forms. The first type, with Jupiter on horseback trampling a giant, is characteristic of the Rhine land between Mainz and Strasbourg.⁹² (Fig 7.4) The second type, prevalent in lower Germany and Gallia Belgica, features a Jupiter enthroned. The prototype for both was the column erected at Mainz by the local population in honour of Nero in A.D. 60, which was itself probably directly based on a Jupiter column in Rome; hence its largely classical iconography. The subsequent Jupiter columns, all a century and more later, included round the base of the column a more eclectic range of deities, often shown in two tiers: the members of the Capitoline triad were usually depicted, but not usually as a triad; and otherwise the combinations of deities were novel, including a range of much less ‘classical’ gods. Though they may have been inspired by the early Mainz column to Nero, the columns were not simply Roman. Their interpretation has been much debated: are the gods local (German or Celtic), or Roman? And in this case it is much less clear that they are to be connected with a particular social group, as we have argued for the dedication of the Danigi: the columns seem to have been erected in towns, sanctuaries and, especially, on estates by all sorts of people from private individuals to local groups to soldiers and officials. They do, however, demonstrate very clearly the flexibility and shifts of ‘Roman’ religion as it was incorporated into provincial ideology: in these columns Roman and native religious forms were combined, and re-combined, into a new celestial hierarchy that guaranteed the cosmic order.

         Fig. 7.4 Jupiter column from Butzbach (Wetteraukreis, Germany), c. A.D. 230; height c.4m. The bottom four panels depict Juno, Minerva, Hercules and Mercury. The next tier up features the seven gods that gave their names to the days of the week: Saturn, Sol, Luna, Mars, Mercury, Jupiter and Venus. On the top of the column is a rider on horseback standing on a defeated giant.

    Religion, however, was not always an effective buttress of Roman rule, or a flexible integrator of the traditions of conquered and conquering. Roman cults and deities could stand all too clearly for the oppressive demands of Roman imperialism. When a rebellion broke out in Germany one local priest of the imperial cult at the Ara Ubiorum tore off his fillets, the symbol of his office, and went over to join the rebels. In the British revolt, the temple to Claudius, seen as ‘a citadel of eternal domination’ at which the provincial priests wasted their money ‘in the guise of religio’, was totally destroyed. When the Jewish revolt broke out in A.D. 66 the first move of the rebels was to end the sacrifices in the temple on behalf of Rome.93

    Conversely local cultic traditions could become the rallying ground for opposition to Roman rule.⁹⁴ The stories of Alexandrian Greeks protesting against the perceived tyranny of Rome include appeals to the Alexandrian Sarapis. Prophecies, originating in the Hellenistic period but still circulating under the Roman empire, foretold the liberation of Egypt and her gods from the foreign oppressor.⁹⁵ In actual revolts, local religious figures are often claimed to have stimulated or even led the rebels. In an Egyptian rebellion of A.D. 172–173 the leader was a priest.⁹⁶ An incursion into the empire from Thrace was led by a priest of Dionysus, who gained a following by his performance of rites; he was probably acting to recover the sanctuary of Dionysus which had earlier been made over by the Romans to another tribe.⁹⁷ In Gaul, at a time of political chaos in Rome, the Druids allegedly prophesied that the (accidental) burning of the Capitoline temple in Rome signified the end of Roman rule over the Gauls. At around the same time, a revolt on the Rhine frontier started with a feast in a sacred grove and a religious vow and was strongly supported by a local prophetess.98 The Jewish revolts against Rome, which were much more significant in military terms than any we have just mentioned, were also aided by the fact that the Jewish faith could be interpreted to offer a coherent religious basis for revolt. At least some of the rebels in Judaea in A.D. 66–70 and 132–135 were inspired by the principle that their god alone should be master of Israel and, in the revolt of A.D. 116–117 which flared up in Egypt, North Africa and Mesopotamia, the rebels in Cyrene seem to have damaged or destroyed temples of the pagan gods.⁹⁹ The subsequent massacre of the Jews in Egypt was probably due in part to traditional enmity to the Jews as religious enemies of Egypt’s gods and the ‘victory’ was still celebrated in Egypt eighty years later by civic festivals.¹⁰⁰

    Local religion operating as a focus of opposition to Rome is a further reminder of the sheer complexity of religious life in the provinces. Individual gods, whether local or Roman, did not stand for just one thing; cults were not combined according to a single standard blueprint which equated one deity unproblematically with another. Religious forms were constantly re-interpreted and deployed in different combinations for quite different purposes. The god that was joined in worship with Jupiter one day might be leading the rebels the next.

3.  Imperial rituals

Various forms of what we call ‘the imperial cult’ are found right across the empire. The army sacrificed to the Capitoline triad on behalf of the living emperor and also to his officially deified predecessors; provincials performed vota to the gods and sacrificed the taurobolium to Magna Mater on behalf of the emperor; and (in the province of Asia) celebrated Augustus’ birthday as the start of their year. In other words, as we have already seen (both in chapter 4 in relation to Rome itself and again at the start of this chapter), cults of the emperor were not an independent element of religious life: sometimes the emperor was placed under the protection of the Olympian pantheon or linked with the traditional gods (as we shall see in the combination Mars Augustus), sometimes cult was offered directly to him. These forms of cult were rarely a separate export to the provinces from Rome, but developed in different ways in the context of the various forms of Romanized religion that operated there. In this section, we have decided to group together some rather different practices which in a variety of ways across the empire related the emperor to the gods. However, we stress that these are very diverse, because they were located in very different contexts. That is, there is no such thing as ‘the imperial cult’.

    The passage of Cassius Dio cited at the start of this chapter shows the significance of the different contexts of imperial cult. Though Dio sees cults of the emperor as a unifying factor across the empire, he draws a crucial distinction between different forms of imperial cult: cults offered to the living emperor by subjects of Rome and the practices of the centre. ‘For in the capital itself (he writes) and in Italy generally no emperor, however worthy of renown he has been, has dared to do this <i.e. have lifetime cults of himself>; still, even there various divine honours are bestowed after their death upon such emperors as have ruled uprightly, and in fact shrines are built to them.’101 That is, official public cults in the capital were restricted to deceased emperors (and members of their families); for the living emperor vows were offered on his behalf to the Olympian gods.¹⁰² Dio further distinguishes between the cults offered by subjects of Rome (Greeks and others) and those to be performed by Roman citizens resident in the provinces. Whereas the subjects of Rome had cults of the living emperor, Roman citizens had cults of the Roman type.

    We can draw further distinctions. Cults of the emperor are found in the provinces at two different levels. Most provinces had a provincial assembly, consisting of representatives of the towns of that province; the assembly at its annual meeting conducted business of provincial interest (whether, for example, the governor should be prosecuted in Rome for corruption) and celebrated an imperial festival. When Dio talks of Augustus giving permission in 29 B.C. to the Greeks of the two provinces of Asia and Pontus-Bithynia to establish cults to himself, he is referring to the creation of cults organized by the two provincial assemblies. In addition, outside the organization of the province as a whole, individual communities established their own cults of the emperor: Ephesus, for example, had not only the sanctuary for the Roman citizens of the province of Asia, but also its own Greek-style cults of Augustus.

    It is conventional to draw another distinction in analysis of ‘the imperial cult’ – between the eastern and the western parts of the empire: in the east the cult was a voluntary matter, absorbed within pre-existing structures, while in the west it was imposed by Rome. We shall notice various differences between east and west in the course of this section, but we have decided to treat the whole empire together, in order to draw out significant patterns right across the Roman world. In so far as Rome had religious expectations of Roman citizens resident in the provinces, it did not matter whether they were in east or west; and civic cults in the west may have arisen from local dynamics similar to those in Greek cities. Rather than stress the east/west distinction, we shall emphasize again the importance of the status of the community offering cult.

    Cult offered to, or on behalf of the emperor, his family or dead ancestors (‘imperial cult’) is just one part of a wider set of associations between emperor and religion in the empire. A distinctive symbol of the Augustan restructuring of religion was the image of the emperor officiating at sacrifice. As we saw in chapter four, the links between emperor and sacrifice were so emphasized that from the reign of Augustus onwards almost no one other than the emperor (and his immediate family) was depicted at sacrifice in public images. This particular imperial scene is represented in a series of sculptures from Rome and Italy; but one of the most striking examples is found in the colonia of Lepcis Magna in North Africa, on an early third-century commemorative arch which was lavishly decorated with four sculptured friezes. One of these friezes shows the offering of sacrifice by the imperial family, probably to the Capitoline triad (Fig. 7.5). Although the centre of the frieze is very damaged, it almost certainly showed Septimius Severus with his two sons; the figure of his wife, Julia Domna, survives on the right of the central group, her hand stretched out to offer incense; behind, a few fragments remain from what appear to be the Capitoline deities – pictured in receipt of the sacrifice. Flanking the central group on the left stands a group of soldiers; and on the right, figures representing Rome and the senate, and a sequence of men clad in togas and other soldiers – in front of whom, from either end, the sacrificial animals are introduced. Other friezes on the arch offer slightly different versions of the place of the imperial family in relation to the gods (one shows the emperor holding an augur’s staff (lituus), linking hands with his son in front of a group of deities, both ‘Roman’ and local); but they combine to stress the significance of imperial harmony and the piety of the emperor in ensuring the favour of the gods for the empire.¹⁰³

    Scenes of imperial sacrifice were not limited to specifically Roman settlements in the provinces. Often stamped as a design on Roman coins, they were found in pockets and purses all over the Roman world. In Asia Minor one or two temples dedicated to the emperor had cult statues which depicted the emperor in the act of performing a sacrifice. The image even seems to have been stamped on sacrificial cakes distributed to the people at festivals through the empire. A number of curious discoveries made in Britain and Hungary have been identified as the moulds for these cakes – showing the emperor offering sacrifice, among other imperial scenes. The emperor in his religious role was literally imprinted on the ritual food consumed at provincial religious celebrations. The distinctively Roman model of sacrifice became a familiar image almost everywhere.¹⁰⁴

    The emperor was regularly represented as under the particular protection of the gods of the local community. In the Greek east sacrifices were made on behalf of the emperor to the Greek gods of the Olympian pantheon. In the west most often it was specifically Roman gods that were emphasized in this role. At Timgad, for example, some of the deities allude to the emperor (‘Victoria Victrix’, the ‘Conquering Victory’ who was responsible for imperial successes); other gods are described explicitly as his protectors (‘Mars Augustus, protector of our lords’).105 The magistrates of a community of Roman citizens (perhaps army veterans) in Tunisia who made a dedication to ‘the gods of the emperors’ probably had in mind the gods that supported the emperor. Thus the Capitolium at Thugga was dedicated to the Capitoline triad ‘for the well-being’ of Marcus Aurelius and Lucius Verus and its pediment featured a relief of an eagle bearing a man aloft, that is an image of the ascension of the emperor to the heavens, a new god. Jupiter, Juno and Minerva protected the emperor in this life and guaranteed his apotheosis after ‘death’.¹⁰⁶ Even in an area such as the north-west of the Iberian peninsula where strictly Roman cults were rare, dedications on behalf of the emperor were made not to local, but to Roman gods.¹⁰⁷ But occasionally local gods too were enlisted in his support. In north Africa, for example, dedications ‘for the well-being of the emperor’ were also made to Romanized forms of local deities: Saturn, Frugifer and Pluto. The divine associations of the emperor extended beyond the Roman pantheon into local religious traditions.

         Fig. 7.5 Part of a frieze on a commemorative arch in Lepcis Magna, c. A.D. 206–209. On the far left (originally the centre of the frieze, but now very damaged) are traces of a male figure (?one of the imperial sons, possibly Geta); next the empress Julia Domna; a flute-player; then other figures who may include a personification of Rome (holding a globe) and of the senate. In front the sacrificial attendants are in the act of killing the animal. See also 6. 1d. (Height 1.72m.)

    In fact a wide range of deities was invoked as protectors of the emperor or as his equals.108 Numerous dedications were made in all the western provinces in the form ‘Saturnus Augustus’, ‘Silvanus Augustus’, or ‘Mars Augustus’ and so on. The interpretation of the dedications is difficult: are god and emperor equated, or is ‘Augustus’ an epithet merely indicating loyalty to the imperial regime? The most likely explanation is that there is a reciprocal relationship between the two terms. On the one hand, the place of the local deity within the Roman order was assured and, on the other, the local deity was a protector of the emperor. The western subjects of Rome, especially those of lowly status, sought to restructure their existing religious systems in order to relate them more or less closely with the gods and emperor of Rome.

    More direct cults of the emperor are also found, but their forms vary greatly, partly in relation to the statuses of the individuals and communities concerned. The establishment of cults of Roma and Augustus in 29 B.C. by the provinces of Asia and Pontus-Bithynia was a model for other eastern provinces: for example, the provincial assembly of Syria subsequently inaugurated an annual priesthood of Augustus and games which were on the international Greek athletic circuit.¹⁰⁹ These provincial cults of Roman power were generally an innovation of the imperial period, even when there had been a provincial assembly under the Republic.¹¹⁰ They were not normally imposed by Rome, but arose from and enhanced the rivalry between individual cities and the standing of those who served as high priests of the provincial cults. In the west the form of the provincial cults varied. Recently conquered and very ‘unRoman’ areas established cults like those in the Greek east, to the living emperor; and in fact it was part of the Roman élite’s image of such subjects that they did worship the emperor directly. A high-ranking Roman recorded with approbation how he had seen in Germany a barbarian chief cross the Elbe in a dugout canoe to touch the divine person of Tiberius (then Augustus’ heir); the chief in addressing Tiberius referred to the local worship of the divinity of the emperor.¹¹¹ The cult of the emperor alluded to here is quite different from the provincial cults of those western areas with long standing Roman traditions which we shall examine shortly.¹¹² It was one of a series established in the early empire by Roman commanders in barbarian areas which had just (they hoped) been conquered. The first of these is found in the north-west of the Iberian peninsula soon after its subjugation by Augustus; a governor serving in 22–19 B.C. dedicated three altars to Augustus (the Arae Sestianae, named after the governor), which probably served as centres for three peoples in this region. In 12 B.C. the three Gallic provinces conquered by Caesar were united in a single provincial assembly at Lugdunum, at an altar of Roma and Augustus, dedicated by Drusus, Augustus’ stepson. In Germany an altar (the Ara Ubiorum) was built in the last decade B.C. or the first A.D.on the banks of the Rhine near Cologne, as a focus for the new province of Germany. At a time when Roman power reached beyond the Rhine to the Elbe a governor dedicated an altar there in 2 B.C.; the barbarian chief who greeted Tiberius is probably referring to the cult of Augustus at this altar. Cults to the living emperor continued to be established subsequently in areas conquered or reorganized by later emperors.113

    Roman citizens, on the other hand, were expected to offer different forms of cult from those offered by people who were subjects of Rome. In various provinces Roman citizens formed official organizations with a religious function, separate from the organizations of the provincials. In Asia and Pontus-Bithynia, when Augustus gave permission for the Greek provincials to worship Roma and Augustus, he instructed the Roman citizens of each province to set up temples in different cities to Roma and divus Julius. At Ephesus both temple and ritual may have been closely modelled on the cult at Rome, and the association of Roman citizens probably existed at least into the second century A.D.¹¹⁴ In two Greek-speaking cities in north Africa inscriptions record vows on behalf of the emperor in Latin – in phraseology that is very close to the formulae used in Rome by the Arval Brothers. These records may have been put up by the Roman citizens resident in those towns, the use of Latin, combined with the display of religious allegiance to the emperor, offering a way of marking their difference from other members of the local community.¹¹⁵

    In the west too Roman citizens formed their own associations. At Lugdunum in fact we find two separate organizations: first the provincial council of the Three Gauls in which both non-Romans and Roman citizens could and did participate; second, as an inscription of A.D. 220 records, an association explicitly of Roman citizens from each of the provinces of the Three Gauls, with its own officers and funds, also meeting at Lugdunum.¹¹⁶We have no clear record of any religious activity on the part of this second association, but it is tempting to guess that it was modelled on that laid down for the provinces of Asia and Pontus-Bithynia. Much more puzzling is its composition. Although the inscription refers to Roman citizens specifically, from A.D. 212 almost all free inhabitants of the empire had been granted citizenship – and so it is hard to see the logic or purpose by this date of a special association restricted to citizens. We may perhaps guess that this was an organization which united not all citizens, but only those who were of Italian rather than Gallic origin. That is, the idea that a defined group of Roman citizens had a particular relation to the emperor seems to have survived the change that ought to have made that idea redundant.

    Similarly, provinces in the west with a significant proportion of Roman citizens followed in general the model of Rome itself.117 The key episode occurred in A.D. 15, the year after the official deification of Augustus in Rome, when permission was given to the province of Hispania Tarraconensis for a temple to divus Augustus in the colonia of Tarraco. Its priests were drawn not just from Tarraco but the whole province, and Tacitus, reporting the decision of A.D. 15, notes that it set a precedent for other provinces.¹¹⁸ There was also a range of divine honours offered to living emperors, beyond the displays of divine protection that we have already noted. In the provincial cult of Gallia Narbonensis and Africa Proconsularis, the living emperor may have been included in the cult alongside the dead and deified; and at the time of Vespasian the cult at Tarraco may have been extended to include the reigning emperor. These developments were a notable divergence from standard practice in Rome itself, and perhaps reflect the influence of other civic communities in the west. But, by and large, the range of religious honours offered to the emperor in these western provincial assemblies seems to have followed the patterns set in the centre.

    If we move on now to the cults of individual cities rather than of provinces, we find again significant differences in the forms of cults, depending in part on the status of the cities concerned. As we have already seen, civic priests of divus Julius are found outside Rome only in coloniae – in both eastern and western parts of the empire. It would seem to follow from this that coloniae, at least in the early empire, were expected (and perhaps even instructed) to follow the lead of Rome. The cult of the numen (‘divine power’ or ‘nod’) of Augustus is instructive here, even though it was a form of cult that did not prove popular. The cult is attested in two towns: a town in Etruria dedicated an altar to the numen of Augustus, at which sacrifices were to be made on the birthdays of Augustus and Tiberius; and the colonia of Narbo dedicated a similar altar in the forum, at which a series of imperial festivals was celebrated. As the cult of the numen focussed closely on the living emperor, it might seem to be overstepping the mark officially set, but there was a precedent from the city of Rome itself. An altar to the numen of Augustus had been dedicated there in about A.D. 6, and the two towns were probably responding to this lead.119 The cult of the Lares Augusti was a much more widespread response to a Roman innovation. Following the major reorganization of the ward cults in Rome in 7 B.C. cults of the Lares Augusti are found throughout the Roman west. At Pompeii a new organization of ‘ward officials’ took office in the same year as the Roman reorganization and the cult later spread to all of Italy. For example, at Ostia, the port of Rome, the cult was established in A.D. 51 with the building of a shrine to the Lares Augusti in the forum; a purification (lustratio) of the area, probably round the forum, was performed, a ritual also carried out by the wards in Rome. Outside Italy, the cults occur in almost every Latin province.¹²⁰

    At the other extreme, communities lacking Roman status generally determined the forms of their cults without reference to Rome. Those communities which we know best, those in the Greek east, regularly established cults of the living (rather than the posthumously apotheosized) emperor. At Athens a small round temple to Roma and Augustus was built on the Acropolis in the Augustan period; it lay close to, and directly on the long axis of, the Parthenon and its architecture was modelled on that of the neighbouring classical temple known as the Erechtheum. In other respects the degree of explicit deification far outstripped the mere association of emperor and god. Athenians sometimes treated living members of the imperial house as themselves divine. They decreed (in the late 190s A.D.) a series of honours for Julia Domna (wife of the emperor Septimius Severus) which identified her with Athena Polias: there was to be a gold cult statue of Julia Domna in the Parthenon, and various sacrificial rites were to be performed by traditional civic and religious officials, both on ‘Roman’ dates (the birthday of Julia Domna; the first day of the Roman year) and at the principal festival of Athena Polias.¹²¹

    The history of these cults in the western empire shows the initiative coming both from the provincials themselves and, on other occasions, from the central Roman authorities, actively promoting festivals, priesthoods and temples for deified emperors.122 As we have seen, the cult of divus Augustus in Tarraco was the result of a request by the Spaniards.¹²³ On the other hand, the fact that at roughly the same time, under the emperor Vespasian, three long-established provinces (Gallia Narbonensis, Africa Proconsularis and Spanish Baetica) established (or at least reformulated) cults on Roman lines suggests strong pressure from the centre – perhaps to be connected with Vespasian’s desire to promote a focus of loyalty to his new dynasty; the provinces of Mauretania in north Africa may also have established a provincial cult at this point. It is fairly certain too that similarities in the institutions of imperial cult between coloniae or municipia and Rome were not always a matter of voluntary imitation of the centre. A centrally controlled and changed calendar of festivals ‘for the veneration of the imperial house’ seems to have been issued to municipia (and presumably coloniae); and these local communities included the centrally directed imperial festivals alongside their locally determined festivals.¹²⁴

    Questions of initiative are not, however, always so easily settled. Often there is simply no evidence to determine the issue: we do not know, for example, who it was who instigated the building of the great imperial temples at Nîmes and Vienne that still stand (the so-called Maison Carrée of Nîmes, dedicated to Gaius and Lucius Caesar; the temple at Vienne to divus Augustus and diva Livia). In many cases, however, inscriptions on civic temples ascribe the responsibility for them either to prominent locals or to the community as a whole.¹²⁵ But even here there may have been a much more complicated history than the simple opposition between local or central initiative suggests. In the case of Tarraco’s request to build an imperial temple, for example, all kinds of factors may have lain behind their approach – even a prompt from the local governor.¹²⁶ The fact that Tacitus records it as a spontaneous gesture from the province does not mean that that is the only story that could be told; and he himself implies that rivalry between different communities for the emperor’s attention may have played a part, when he says that Tarraco’s gesture ‘set an example’ to other provinces.

    All these different forms of honour and cult required priests and other personnel.127 Priests of the provincial cult were generally drawn from the local élite, but the forms of priesthood varied between the more and the less Romanized provinces. Some of the regulations for the office of high priest of the province of Gallia Narbonensis are preserved, inscribed on a bronze tablet. These show clearly how in this highly Romanized province the priest’s privileges and obligations were partly modelled on Rome: his title was flamen; his wife was known as flaminica (like the flaminica of Jupiter), and she seems to have shared at least some of the flamen’s religious duties; the surviving clause which prevents her taking an oath against her will or touching a dead body matches almost exactly equivalent regulations for the flamen Dialis and his wife at Rome. On the other hand, there are significant differences: at Rome all the major flamines held office for life; here the office is explicitly short-term, and a good part of the surviving regulations are concerned with the honours and privileges of ex-flamines.¹²⁸ The parade of Roman models contrasts with the forms of cult in less Romanized provinces, where there were no Roman-style flamines or temples to the officially deified emperor, but priests whose name (sacerdotes) distinguished them from the ancient priesthoods of Rome and altars for worship of the living emperor.

    But many other social groups were involved in different aspects of imperial cult. The organization of the cult of the Lares Augusti in the provinces, for example, was similar to that at Rome. The colleges responsible for the cult consisted of three or four members, who were generally drawn from ex-slaves (though the freeborn are also found). Gradually, however, slaves too entered the colleges, perhaps because ex-slaves preferred to join the more prestigious group of Augustales– who were also in some way connected with the emperor and his honour.

    Augustales, who were common in coloniae and municipia in Italy and the Latin west, have usually been defined in relation to their performance of cult, as if they were priests of the emperor.¹²⁹ They have been taken to be small groups, often six in number (seviri), rather like the associations responsible for the cult of the Lares Augusti. A picture of their role, on this interpretation, comes in an inscription from Narbo – where, after a period of tension between the people and the council, the people decided to appoint six men, three Roman knights from the people and three ex-slaves to perform a series of sacrifices in honour of Augustus and his numen.130 The problem is that this particular inscription makes no explicit mention of ‘Augustales’.¹³¹ But suppose (as it normally is supposed) that these officials from Narbo did have that title – they still represent only one type of the groups that throughout the Roman empire were known as ‘Augustales’. There is other, and much stronger, evidence to suggest that many Augustales were not members of small associations at all (whether they had specifically priestly duties or not), but possessors of a particular local status.¹³² The (fragmentary) list of Augustales from Herculaneum near Naples has more than 450 names, divided into different units (curiae), at least one of which was for the freeborn.¹³³ Augustales ranked immediately below the local council at public festivals. Their officials had the symbols of a magistrate (special toga, lictors, the bundle of rods (fasces) and so on), and in return for their office were obliged to repair buildings, erect statues and put on games.¹³⁴ They also had, in at least some cases, their own special buildings, with imperial statues.¹³⁵ There is no real case for seeing Augustales of this kind as priests, but Augustales, like other public figures, certainly carried out religious functions in their public role: one of their officials is praised in an inscription for ‘imitating bygone piety’ in his concern for a local cult; others were responsible for the cult of the Lares Augusti.¹³⁶ That is, Augustales performed the conventional range of local religious actions, and there is no reason to think of them as particularly connected by definition with ‘the imperial cult’. Their name ‘Augustales’ may in fact not derive from their presumed cult function, but mark the creation of the status by Augustus.

    In at least the ‘civilised’ parts of both east and west the principal social change which accompanied these religious changes was the role of local élites in the service of Rome. In regions such as Spain, southern France, Greece, Asia Minor, Syria and north Africa, those who were ambitious could hope to see their sons entering the Roman administration as equestrians and their grandsons even entering the Roman senate. For many of the local offices of the imperial cult, the holders received prestige in their local communities, as they did for holding other offices or priesthoods. In the west, ex-slaves with Roman citizenship who formed a significant upwardly mobile group could aspire to some public status which articulated their position in the framework of the Roman empire. But the higher up the social ladder you went, especially to the priesthoods at the provincial level, the higher the stakes became. From the point of view of the priests, these major priesthoods could be a stepping stone to further social and political advancement – even into the world of the city of Rome itself.137 From the point of view of Rome, on the other hand, we might see the imperial cult as one of the major ways that the local élites were suborned to the service of Rome.

    This raises sharply what has always seemed an intriguing puzzle about cults offered to the emperor and his family: were they just a political tool (in the service of Rome, or of local élites)? or did they have some ‘real’ religious significance? We have already seen throughout this book from discussion of very earliest Roman religion that the opposition implied between religion and politics is an inappropriate model for thinking about Roman religion. It is hard now to appreciate that Jesus’ claim in the Gospels that one should give unto Caesar that which is Caesar’s and give unto God that which is God’s was, in the context of the first century A.D., utterly startling.¹³⁸ The idea, in other words, that there was (or should be) a clear delimitation of the political and religious spheres of authority cut across most of the Roman assumptions about the relationship between religious and political life that we have seen so far. The success of this reading of Jesus’ message and the dominance of Christianity in the western political tradition has meant that we come to imperial cults with an inappropriate distinction in mind. Ordinary inhabitants of the Roman empire expected that political power had a religious dimension. The opposite was also true: religious cults might quite properly have a political dimension. If we seek to distinguish between cults that were (really) political and those that had a (genuine) spiritual dimension we are doing little more than engaging illicitly in Christian polemic against an alien religious system.

    Another piece of illicit Christianizing often colours our understanding of the relation between imperial rituals and rituals for the traditional gods. If Christianity eventually triumphed over paganism, what kind of paganism was it? The old model which viewed traditional cults in the Roman empire before Constantine as mere decayed survivals has now been largely abandoned. But some would still argue that the traditional cults had been effectively supplanted by worship of the emperor, so that Christianity’s victory was in fact over the idolatrous worship of a human being. This too is a misunderstanding. In the east throughout the period we have been considering the primary identities of Greek cities continued to be focussed on their ancestral gods. For the citizens of Ephesus in the first centuries A.D. the key to understanding their city lay in the cult of Artemis. The city boasted that it was the birthplace of Artemis, that it possessed an image of the god that had fallen from heaven, and that cults of Artemis had been diffused from Ephesus all over the world. In the words of an Ephesian decree,

… the goddess Artemis, patron of our city, is honoured not only in her native city, which she has made more famous than all other cities through her own divinity, but also by Greeks and barbarians, so that everywhere sanctuaries and precincts are consecrated for her, temples are dedicated and altars are set up for her because of her manifest epiphanies …139

Cults of the emperor, which were modelled on the traditional forms of civic cults of the gods, did not displace traditional cults; they fitted in alongside them. For example, in one Macedonian town, a local citizen volunteered to be priest of Zeus, Roma and Augustus and displayed extraordinary munificence in the monthly sacrifices to Zeus and Augustus and in the feasts and games for the citizens.¹⁴⁰ His activities illustrate clearly how (as we have seen on several occasions) the worship of Augustus could be integrated within local religious and social structures.

    The same is true not only of Rome itself, as we have seen in chapter 4, but also of communities closely modelled on Rome. The army placed the Capitoline triad at the centre of its religious life, had cults of other ancient Roman gods, and of the official divi and divae. There was no opposition between the two types of cult: the army, like many others in the empire, started its year with vota to the Capitoline triad on behalf of the emperor and the eternity of the empire. The ancient cults of Rome were the context (if a modified one) within which the emperor fitted.

    Whereas Greek cities retained a largely stable (if evolving) religious system, and the army and (some) coloniae were artificial creations, the picture is in some ways different for many communities in the Latin west. At first sight it might seem that here the imperial cult was an isolable phenomenon, and hence potentially a more easily identifiable competitor for Christianity, but what we have seen throughout this chapter is that in the west ancient cults across a wide spectrum were transformed, and recentred, at least in part, on Rome. The development of the imperial cult may not be so fully integrated in existing traditions as it was in the east – located in a novel, rather than a stable and familiar symbolic context. But even here it was part of something bigger, and must be seen in a context of wider and more profound religious and political changes. Analyses of the imperial cult in the Latin west which examine only the imperial cult itself suffer from a serious case of scholarly tunnel vision and simply fail to grapple with the problem of the relationship of the new forms of imperial rituals to the local religious context.

    Does it then matter whether rituals treated the emperor like one of the ordinary gods or ‘merely’ placed him under their protection? To modern observers it seems crucial, because in modern world religions there is a uniqueness claimed for divinity, and in Christianity specifically one of the central ‘mysteries’ is precisely the relationship between humanity and divinity, as summed up in the relationship between god and Jesus. The issue had some importance also for Jews and Christians in the Roman empire. Both were (generally) happy to place the emperor under the protection of their god, by sacrifices or prayers.141 Normally, they were not expected to do more than that. Even when Christians faced trials before Roman officials, the principal issue was their relationship to the traditional gods, not to the (divine) emperor: the question was – would they perform sacrifice to the gods? But behind this lay a further concern: if they would not support the traditional pantheon (which upheld the emperor), how could they support the emperor (given that praying to the Christian god did not count)? It is true that sometimes in this context Christians were expected to sacrifice to the gods on behalf of the emperor, and sometimes directly to the gods and to the emperor.¹⁴² But the pressures exerted on Jews and Christians to conform were not motivated by theological concerns about the nature of divinity.

    What was at stake for emperors, governors and members of civic élites was the whole web of social, political and hierarchical assumptions that bound imperial society together. Sacrifices and other religious rituals were concerned with defining and establishing relationships of power.¹⁴³ Not to place oneself within the set of relationships between emperor, gods, élite and people was effectively to place oneself outside the mainstream of the whole world and the shared Roman understanding of humanity’s place within that world. Maintenance of the social order was seen by the Romans to be dependent on maintenance of this agreed set of symbolic structures, which assigned a role to people at all levels. Emperors in Rome needed to play the role of first citizen (not god), an ‘ordinary’ (if unequalled) senator, but they also needed to be assured of their superiority over other groups and areas. Roman citizens in the provinces needed to construct identities for themselves which articulated their superiority over mere subjects of Rome, and so followed the precedent of Rome. As for those mere subjects, the centre might expect that they would and should abase themselves before Rome by worshipping the emperor as a god. One emperor indeed asked a delegation of Jews from Alexandria in Egypt in a pointed and hostile manner why they did not sacrifice to him as to a god.144 There could be no clearer way of articulating the hierarchy of social, political and religious relations that formed the Roman empire. The subjects themselves responded to such pressures or demands in different ways: by accommodating the power of Rome within their traditional symbolic structures, or by changing everything in favour of Rome. From the point of view of status it might make all the difference whether the emperor was treated as a god or only placed under the protection of the gods.

The patterns of the early empire were maintained to some extent into the second century A.D.: new conquests of the first and second centuries were treated much as those of the Augustan period. Internally, there were developments in the statuses of towns: new coloniae were rarely created after the Augustan period, but from the second century A.D. existing towns were granted colonial status. Roman citizenship spread slowly in the course of the first two centuries A.D., but in A.D. 212 was granted dramatically by the emperor Caracalla to almost all the free population of the empire.¹⁴⁵ Most newly enfranchised provincials continued to worship their old gods in the old ways; we have, however, one nice example of what seems to be a religious response to Caracalla’s grant of Roman citizenship.

    Within three years of this grant a civic temple to Zeus Kapitolios is found at Ptolemais Euergetis in the Egyptian Fayum.¹⁴⁶ The earlier scattering of cults of Zeus Kapitolios and of games called Kapitolia in Greece and Asia Minor had been in Roman coloniae and in other cities with especial ties to Rome. At Ptolemais the new cult is very striking. The name ‘Zeus Kapitolios’ refers to the god worshipped on the Capitol at Rome (while, significantly perhaps, refraining from calling him Optimus Maximus – ‘Best and Greatest’). This is almost the only cult in Egypt that refers to a specifically Roman god and the calendar of rituals associated with it consists partly of specifically Roman festivals. There were sacrifices on 1 January; 21 April, the birthday of Rome; on eight occasions for the reigning emperor, twice for his deceased father and twice for his mother. But there were some very different rituals too. In addition to the Roman cults, there were also festivals of four Egyptian gods: the crocodile god Souchos, who was the principal deity of the town, Harpokrates, the Nile and Sarapis. To us, the combination may seem baffling. But it seems that the town was picking up the religious rhetoric of Caracalla’s edict granting Roman citizenship to the empire and creating for itself something to count as a new ‘Roman’ cult. But the way it proclaimed that Roman status was not just by the replication of Roman festivals, but by the integration of local Egyptian cults and standard Roman festivals within the cult of Zeus Kapitolios – already itself a strangely non-Roman ‘Roman’ title.147 In just such ways in communities throughout the empire, distinctively Roman and distinctively local traditions were integrated as a response to (and as an articulation of) the power of Rome. Roman religion came in many ‘foreign’ forms.¹⁴⁸

 

---

    1   See, for example, the temple from Dendur, now on display in the Metropolitan Museum, New York: Aldred (1978) figs. 14–18, 28–33, 38–9. Claudius: E.M. Smallwood, Documents Illustrating the Principates of Gaius, Claudius and Nero (Cambridge 1967) no.370 (trans. Sherk (1988) 83–6).

    2   Egypt: Dunand (1983). Rebellions: below, pp. 347–8.

    3   Issues of similarity and difference between the Greek east and Rome have been central to debates on Roman culture since ancient times – see on Dionysius above, pp. 171–4. Throughout this chapter we use the distinction between east and west, while also arguing that some features are common to different parts of the empire. For a stronger version of that position see Woolf (1994).

    4   Levick (1967); Gargola (1995) 71–101; Fear (1996) 63–104. Above, p.157 for the late republican context.

    5   Fear (1996) 131–69. That municipia were not as closely modelled on Rome as coloniae is stated in Aulus Gelius, Attic Nights XVI.13.4–9.

    6   Cobo (1653/1979) 187–8, 211–12, 241–2, a seventeenth-century Spanish historian. Hostile account: Guamán Pomo de Ayala (1567–1615?/1978) 70. For subtle studies of the reception of Inca cults see S.J. Stern (1982) 20–2 and MacCormack (1991). Contrast the less integrated Aztec empire. The Aztecs removed conquered deities to the centre, without replicating the centre in the provinces: Sahagún (1950) II.168.

    7   Holtom (1943) 153–73. One might be tempted to see this as an aspect of the growth of modernity in Japan, of peculiarly modern nationalism, but it (like the Inca case) may be accounted for in terms of the growing general integration of the Japanese state.

    8   For the British in India, Bayly (1989), Metcalf (1994). Inden (1990) is a critique of western (‘imperialist’) constructions of Indian society.

    9   For the occasional Roman use of evocatio, see above, pp. 34–5; 82–3; 132–4.

  10   See for different forms RIB 307 = 2.9b(i), RIB 218 = 2.9b(ii); below, pp. 344–5.

  11   The term ‘syncretism’ is problematic, not least because it has often been used pejoratively to refer to a meaningless mish-mash of religions (see Berner (1982) and Martin (1983) 134–7 for a history of the term). It has, however, been revived in a neutral sense: Pirenne Delforge (1994); Stewart and Shaw (1994). Interpretatio romana, a phrase taken from Tacitus, Germany 43.3, which is also commonly used, at least places the emphasis on interpretation (Girard (1980)), but it stresses the role of Romans rather than provincials, and assumes that one-for-one identifications were possible between Roman and local gods. For the processes, see documents and discussion in 2.9 and R. L. Gordon (1990c).

  12   Cassius Dio LI.20.7.

  13   Toutain (1907–20) remains a useful synthesis on the west, but it is unfinished and anyway omits iconographic evidence. The range of epigraphic evidence is presented by MacMullen (1981) and Lane Fox (1986).

  14   For an introduction, Swain (1996).

  15   Problems of the changing nature of the evidence are compounded by the divisions between different scholarly traditions. For much of the Latin west scholars study either the pre-Roman or the Roman periods, but not both, thus failing to address issues of continuity and change. See, however, Woolf (1998), ch. 8 on Gaul.

  16   Tacitus, Agricola 21. Haruspices: ILAfr 592 (Africa Proconsularis); cf. ILS 4952a (Lugdunum); ILS 8833 = Inschriften von Ephesus V 1540. Cf. Eck (1992). Imperial ex-slaves were another source of local pressures: e.g. Tacitus, Histories I.76. Vota: Pliny, Letters X.35–6, 101–2.

  17   Tertullian, On the Crown 12.3; Apology 35.4; cf. Gaius in Digest L. 16.233.1; ILS 112 = 10.1b (Narbo); Mărghitan and Petolescu (1976) (Sarmizegethusa); AE (1950) 217 (Ammaia in Portugal). Cf. Meslin (1970) 30–1.

  18   Mishnah, On Alien Worship 1.3 = 12.6g; Plutarch, Cicero 2.1. Cf. Lucian’s assertion of the Roman character of the festival: The Mistaken Critic 8. On vota see above, p. 196; examples of vota by Arval Brothers trans. in Lewis and Reinhold (1990) II.516–19; private vows were common throughout the empire (for inscribed examples see 9.5a–b).

  19   Gaius, Institutes II.7 = 10.4c. Cf. generally E. De Ruggiero, Dizionario epigrafico di antichità romane (Rome 1886–) I.190–200.

  20   Pliny, Letters X.68–9 = 10.4d(iii-iv); cf. Codex Justinianus III.44.1. For the governor and the transport of corpses see Digest XI.7.38 (Ulpian), with Gabba and Tibiletti (1960) = AE (1992) 813; also imperial rulings in Digest XLVII.12.3.4 (Ulpian).

  21   Pliny, Letters X.49–50 = 10.4d(i–ii).

  22   Frontinus (?), in Agennius Urbicus, On Disputes over Land (ed. C. Thulin, Corpus agrimensorum Romanorum), p.48.4–12.

  23   Festus p. 146L s.v. ‘municipalia sacra’; Ulpian in Digest XLVII.12.3.5 on whether such municipal laws should now be overridden by general imperial rulings; cf. above, p. 222 for attitudes towards local peculiarities.

  24   Antium: Tacitus, Annals III.71.1; Map 5. Expulsions: above pp. 230–2. Pontifices: Millar (1977) 359–61. Caesar: ILS 73, 73a, AE (1982) 149, with Alföldy (1991) 305; cf. below pp. 329–30 for instructions to coloniae outside Italy and pp. 337–8 on the quindecimviri.

  25   Above, pp. 5–6 on the early calendar; 3.3; Whatmough (1931); Degrassi (1963); Salzman (1990) 7–8; Rüpke (1995). The only extant earlier calendar, from Antium, dates to the early/mid first century B.C. An extract from the Praeneste calendar is given in 3.3b.

  26   Though one was for the ex-slaves and slaves of an imperial villa who formed an association of worshippers of Augustus: Degrassi (1963) 201–12 (Antium).

  27   For a denial that municipal display of calendars was connected with their religious role, Rüpke (1995) 165–86.

  28   Map 5; see 1.5; Wissowa (1912) 157 n.4, 519–21, 555 n.2; Ladage (1971) 8–10. Most of the surviving evidence is imperial in date.

  29   ILS 5004 (trans. Braund (1985) no.460); Wissowa (1915); Purcell (1983) 167–79; Saulnier (1984). On the creation of the treaty, in 338 B.C., see Dubordieu (1989) 339–61. For Aeneas sacrificing before a shrine of the Penates (in a sculptured relief from the Ara Pacis), see 4.3c.

  30   Fink, Hoey and Snyder (1940) = 3.5 (A.D. 223–7). Nock (1952) 223 denied there was an official desire to see the soldiers worshipping the gods listed in the calendar rather than any other gods, but see Fishwick (1987–) II.1 593–608.

  31   These military vota happened in parallel with those performed by civilians throughout the empire; see above, p. 320.

  32   Ankersdorfer (1973); Helgeland (1978); Birley (1978); Rüpke (1990) 184–98, 250–8.

  33   Maryport: RIB 813–35; cf. the dedications to Jupiter Optimus Maximus on altars dedicated by beneficiarii, adjutants (Schallmayer (1990)). Corbridge: Richmond (1943); E.J. Phillips (1977) 12–13, 34, 55; the shrine with the carving of the wolf may have been dedicated to Dea Roma; a sculptured panel from the shrine showed a rustic scene of a faun, perhaps alluding to the first inhabitants of the later site of Rome. On the rose festival see Hoey (1937) and ILS 4918 = 3.7. Danube: AE (1982) 849 = J. Kolendo et al. (eds), Inscriptions latines de Novae (Poznán 1992) no.28, with Kolendo (1980) and Sarnowski (1989) (A.D. 208).

  34   Petrikovits (1975) 75–8; Johnson (1983) 111–17; Rüpke (1990) 165–83.

  35   AE (1989) 581 (A.D. 208) from Aalen.

  36   Domaszewski (1967) index Al under ‘haruspices’ and ‘victimarii’; victimarii now attested in the legions, J.C. Balty, JRS 78 (1988) 99; Antonine Wall: Henig (1984) 86–7.

  37   Fitz (1972) 177–95. Cf. dedication at Carnuntum (in Austria) by an eques in a Canathene regiment to two Arab gods peculiar to the area of the Hauran in which Canatha lies: ILS 4349, revised in J. Ceška and R. Hošek, Inscriptiones Pannoniae Superioris in Slovacia Transdanubiana Asservatae (Brno 1967) no. 12. See generally Haynes (1993).

  38   A fresco from Dura, which used to be interpreted as a sacrifice by the commander of the Twentieth Palmyrene cohort to three gods from Palmyra, may in fact show a sacrifice before three statues of emperors: Pekáry (1986).

  39   E.g. the ‘Aufaniae Matres’ in Lower Germany: Birley (1978) 1525–7. Cf. Wissowa (1916–19) 21–3; Drexel (1922) 8; Eck (1989b) 44–5; above, p. 230.

  40   ILS 6087 = Crawford (1996) I.393–454, part in 10.2a; translation of more in Lewis and Reinhold (1990) I.453–61 and of all in Crawford (1996) I.421–32. Cf. above, p. 157; D’Ors (1953) 167–280; Mackie (1983) 222–3. It is possible that the original rules were revised for this publication, but for arguments against see Gabba (1988). Cf. Scheid (1991), (1995) on Trier, and Rives (1995a) on Carthage.

  41   ILS 6087 = Crawford (1996) I.400–2 sects 62, 66–8 = 10.2a. Cf. Ladage (1971) 11–14, 18–19, 32–5, 39–41, 51–4, 79–80, 88, 103; Galsterer (1971) 59–61; Canto (1981) on AE (1978) 402 (Italica). Unfortunately, the relevant part of the charters of the Spanish municipia, which probably stood at the beginning of the standard charter, is not preserved in any of the surviving copies.

  42   Zama: CIL VI 1686; Timgad: Chastagnol (1978) 26–31 (the document dates to the mid to late 360s A.D.). See generally Dupuis (1992).

  43   ILS 4907 = 10.1c. For Rome see Varro, Latin Language VI.61; Tacitus, Histories IV.53.

  44   Above, p. 175 for Romulus and the Parilia; relief showing ploughing in 10.2b; Plutarch, Romulus 11.1–4 = 4.8a; Levick (1967) 35–7 and Sammlung von Aulock Index (1981) 224 s.v. ‘Koloniegründer’.

  45   Boundary stones: ILS 6308 (Capua). Burials: ILS 6087 (Urso) = Crawford (1996) I.403–4 sect 73; Frontinus, On Disputes (ed. C. Thulin) p.7.2–5 (the section is misplaced in the text but ancient). After the taking of the auspices, the professional land-surveyors could proceed, laying out the land divisions of the colonia and orienting them in accordance with the direction of the mid-day sun: Hyginus Gromaticus, Disposition of Boundaries (ed. Thulin) p. 135.1–14; also Frontinus pp. 10–11. Cf. Le Gall (1975) 301–8.

  46   Weinstock (1971) 405, 408–10, pl. 31.2, though he does not note the connection with coloniae, M. Walbank (1996) on Corinth. Above, pp. 140–9 for deification of Caesar.

  47   AE (1984) 508 frr. II a and b (Tabula Siarensis), translated in Sherk (1988) no. 36. Note also the birthday of Germanicus (24 May) in the calendar from Dura: Fink, Hoey and Snyder (1940) 45 (and 136–8) = 3.5.

  48   Pliny, above, p. 321.

  49   Inscriptions of these rules in archaic lettering survived in the sanctuary to at least the Augustan period: Dionysius of Halicarnassus IV.26.5; Festus p.164L s.v. ‘nesi’. For the creation of the temple see Livy I.45 = 1.5d.

  50   Salona: ILS 4907 = 10.1c (A.D. 137); Narbo: ILS 112 = 10.1b (A.D. 11); Ariminum: CIL XI 361 (first century A.D.); the procedure was so standard that the terms were given in highly abbreviated forms. Though these documents have been much studied, it seems important to us to stress the ‘colonial’ status of the towns concerned.

  51   A man is shown reading from a book (once) on a military sacrificial relief: Henig (1984) 86–7.

  52   Tertullian, To the Gentiles II. Augustine later made a similar use of Varro in his City of God (e.g. VI.5 = 13.9). Cf. Price (forthcoming).

  53   Find spot: Gsell (1920–8) VIII.117. Wuilleumier (1928) 40 showed that the two reliefs are of exactly the same dimensions. Cf. Zanker (n.d.) 18–20; Fittschen (1976) 187–9; Torelli (1982) 39–42; Kleiner (1992) 100–2. On the Augustan colonia see generally Gros (1990); Rives (1995a). On Mars Ultor see above, pp. 199–201 and 4.2; on Ara Pacis above, Fig. 4.6, and 4.3.

  54   Though art historians have generally treated the two reliefs separately and simply as evidence for Augustan Rome, the real interest of the reliefs lies in the light they cast on Carthage.

  55   ILAfr 353 = Z.B. Ben Abdullah, Catalogue des inscriptions latines païennes du Musée du Bardo (1986) 253 no.7, with Rostovtzeff (n.d.); Poinssot (1929); Rives (1995a) 53–7.

  56   A similar allusion is found on one of the Lares altars from Rome: above, p. 186.

  57   Above, Fig. 1.1 for an early statuette; Forum of Augustus, 4.2.

  58   2c; Bianchi (1950); I.M. Barton (1982); Todd (1985). Cologne: Ristow (1967); Follmann-Schulz (1986) 735–8. Corinth: M. Walbank (1989), with architectural observations of C.K. Williams (1989). Jerusalem: Schürer (1973–87) I.542, 550–1, 554. For increased emphasis on the Capitoline triad at Ostia, an ancient colonia, under Augustus and Hadrian see Meiggs (1973) 352, 380–1.

  59   pain: Keay (1988) 117, 145, 148. Urso: ILS 6087 = Crawford (1996) I.403, sects 70–1.

  60   Timgad: I.M. Barton (1982) 308–10; cf. Xanten: Follmann–Schulz (1986) 766–9.

  61   Seyrig (1954); Liebeschuetz (1977) 485–9; Millar (1990) 10–23 and (1993) 281–5, who notes that the nature of any prior local cults is very obscure, and in particular that the triad postulated by Hajjar (1985) is very hypothetical. We are grateful for the advice of Dr H. Dodge on the construction of the temple. Cf. above, p. 283 and below pp. 384–6 for a sanctuary of the god in Rome.

  62   Bianchi (1950); I.M. Barton (1982); Rives (1995a) 114–32. Thubursicu: Syme (1951). Numluli: CIL VIII 26121; Ferchiou (1984) on decoration. Thugga: CIL VIII 15513 (A.D. 166–9), temple; 26582b (A.D. 168), legacies; below, p. 351.

  63   ILS 9495 = ILAfr 306 = Z.B. Ben Abdullah, Catalogue des inscriptions latines païennes du Musée du Bardo (1986) 73 no. 190. Fishwick (1987–) II.1 452–3 argued that this is a dedication to an Augustan god, of the type discussed below, p. 352. See also a dedication to the Capitoline triad by Roman citizens resident at Troesmis, a non-Roman town near a legionary camp near the mouth of the Danube: CIL III 6167 (A.D. 138–61).

  64   Vaga: CIL VIII 14392, with Smadja (1980) 153. Nicaea: Robert (1978) 275–6.

  65   Aulus Gellius, Attic Nights XVIII.2; Fayûm Towns and their Papyri, ed. B.P. Grenfell (London 1900) 119.28–9 (c. A.D. 100).

  66   Above, pp. 96–8; Wissowa (1912) 320–7; Vermaseren (1977) 60–9, 126–44; Schillinger (1979); Turcan (1989) 61–8.

  67   ILS 4175 = 10.4b (A.D. 289); cf. ILS 4131 = 6.7b (A.D. 160). The quindecimviri also authorized the dendrophori (‘tree-bearers’) of Cumae who served the same cult: ILS 4174 (A.D. 251). Wissowa (1912) 320–1; Schillinger (1979) 358–60.

  68   ILS 4131 = 6.7b; there was certainly a ‘Vatican hill’ at Mainz-Kastel: ILS 3805. If this text does refer to the sanctuary in Rome (Map 2 no.6; below, p. 384), then the original taurobolium must have been performed there.

  69   Février (1976); Le Glay (1984) 156–7; Rives (1995a) 100–72 stresses their interest also in native cults. See below, p. 357 on imperial priests.

  70   AE (1987) 1078; cf. Pavis d’Escurac (1980–1); above, p. 217. Note also the Roman cults of the colonia of Savaria, though they receded in importance in the second and third centuries A.D.: Balla (1967).

  71   Lavagne (1979); also Carré (1981).

  72   See R.L. Gordon (1990c) 240–5.

  73   The conventional view holds that Rome took a series of measures of increasing scope to bring both temple lands and temple personnel under the control of Roman administration: Swarney (1970) 57–9, 83–96; Whitehorne (1980–1); Stead (1981); cf. Thompson (1988) 271–6 on Memphis. For a revisionist view see the forthcoming book by P. Glare.

  74   From Antoninus Pius onwards circumcision was illegal except for Jews and for Egyptian priests.

  75   Schürer (1973–87) I.376–9, II.218–23; Goodman (1989). Sacrifices: Philo, Embassy to Gaius 157 = 12.6c(ii); below p. 361. Gifts to Jerusalem seen as ‘barbarous superstitio’: Cicero, On Behalf of Flaccus 67; cf. Tacitus, Histories V.5.

  76   Debord (1982) 56–61; S. Mitchell (1993) I.81–2. Antioch: Strabo XII.8.14 (p.577C); Levick (1967) 85–7.

  77   Above, pp. 221; 234.

  78   Nilsson (1961–7) II.327–58; Lane Fox (1986) 27–101; Price (1998) ch.8. Above, p. 223 for Roman initiates.

  79   IG II² 1078 = F. Sokolowski, Lois sacrées des cités grecques (1969) no.8.

  80   Sacred properties: IG II² 1035, dated by Culley (1975) to 10/9 – 3/2 B.C. Temples: Camp (1986) 184–7; Alcock (1993) 192–5. The claim that these moved temples also housed the imperial cult is not founded on secure epigraphic evidence. (Gaius Caesar, Augustus’ son, is called ‘new Ares’ not in the agora, but in the Theatre of Dionysus.)

  81   Calendar: Samuel (1972) 171–88; cf. Price (1984) 106. The governor’s suggestion itself followed a request by the province for ideas.

  82   E.M. Smallwood, Documents Illustrating the Principates of Gaius, Claudius and Nero (Cambridge 1967) no.380 = Inschriften von Ephesos la 17–19 (trans. Braund (1985) 213–15). Cf. Debord (1982) 211–12; Price (1984) 69, 103. For Roman control of asylum see above, p. 224.

  83   Hadrian: Pausanias I.24.7; Arafat (1996) 163. Nero: Price (1984) 149; Arafat (1996) 153–4.

  84   Duval and Pinault (1986); Olmsted (1992). ‘Oriental’ cults: Toutain (1907–20) II; Le Glay (1984) 156; Alföldy (1989) 74–5.

  85   Pascal (1964) 140–4; Pauli (1984) 152–5.

  86   Alföldy (1989) 79–82 on preponderance of local deities.

  87   Caesar, Gallic War VI.17 (= 2.9a), 21; c£ inscriptions in 2.9b, Tacitus, Germania 9.

  88   Home (1986).

  89   Toutain (1907–20) I.297–314, 388–92, III.193–467; Clavel-Lévêque (1972); Lavagne (1979); Letta (1984). Wightman (1985) 177–87 and (1986) shows that the pattern is less clear in Belgica, because there were far fewer unambiguously Roman deities. Cf. Derks (1991); Van Andringa (1994) on Aventicum. For cults of pre-Roman Gaul see Brunaux (1988); Roymans (1990); Goudineau et al. (1994) include studies of both pre-Roman and Roman Gaul.

  90   Below, p. 355 n. 120; Lambrino (1965); Pastor Muñoz (1981); Nicols (1987). Cf. Fear (1996) 227–69 on differences within Baetica. For religions in Spain: Mangas (1986); Keay (1988) 145–71; Vázquez y Hoys (1982), testimonia.

  91   AE (1973) 319 (A.D. 147) with Le Roux (1994b). For a nice example from western Gaul see ILS 7053 and AE (1969–70) 405 with Chastagnol (1980) and Scheid (1991) 51. In Africa too one needs to distinguish between Jupiter Optimus Maximus, an official deity, and Jupiter without the epithet: Kallala (1992).

  92   Bauchhenss and Noelke (1981); in English, Schutz (1985) 66–7.

  93   Tacitus, Annals I.39.1, 57.2; XIV.31.4, 32.3. Jews: Josephus, Jewish War II.409–21; Roth (1960); above, p. 341.

  94   Momigliano (1987); Goodman (1991). Pekáry (1987) lists cases of unrest and revolts. See also S.J. Stern (1982) 51–71 for religious millenarian opposition to Spanish rule in Peru.

  95   Musurillo (1954) 4–5, 45 (= 12.6f). Koenen (1970); also (1984).

  96   Cassius Dio LXXII.4.

  97   Cassius Dio LI.25.5 (29 B.C.), LIV.34.5–7 (11 B.C.).

  98   Tacitus, Histories IV.54, 61, 65, V.22, 24. Cf. above, p. 341 on Druids.

  99   Schürer (1973–87) I.531, 544–5, II.598–606; cf. Barnes (1989b) on dating.

  100   Frankfurter (1992).

101  Cassius Dio LI.20.8.

102  Public cults of Italy did not always follow the Roman model, but rather that of towns in the provinces – suggesting perhaps that Italian towns were perceived (and perceived themselves) as subordinate to the power of Rome, rather than partners in it.

103  Strocka (1972), with full reconstruction of the frieze; Kleiner (1992) 340–3; also illustrated 6.1d. For a similar image from Rome see 6.1b.

104  Asia Minor: Price (1984) 185. Moulds: Boon (1958); E.B. Thomas (1980) 184. Cf. in general R.L. Gordon (1990b) 202–19.

105  Above, n. 70. ‘Lords’ refers to the emperors.

106  Tunisia: AE (1977) 855. Thugga: I.M. Barton (1982) 317; above, p. 336. Cf. Smadja (1985) on the emperor and the pantheon.

107  Tranoy (1981) 332–3. Cf. Le Glay (1984) 168–9.

108  Nock (1925) 91–3; Le Glay (1984) 166–9; Price (1984) 91–100 for the east; Fishwick (1987–) II.1, 446–54.

109  Syria: AE (1976) 678; Inschriften von Magnesia 149. On provincial cults see Deininger (1965) 7–98, 158–61; Price (1984) 56, 66–7, 72–3, 75–7, 83, 88, 104–5, 128–31, 226. On Galatia see S. Mitchell (1993) I.100–17.

110  For cults of Dea Roma and of Roman governors, above, pp. 158–60.

111  Velleius Paterculus II.107.2 = 10.6a; cf. address to Nero as god by Tiridates, king of Persia: Cassius Dio LXII.5.2.

112  Fishwick (1987–) I.97–146. Spain: Tranoy (1981) 327–9. Gauls and Germans from across the Rhine were instructed by Rome to offer cult to the memories of Drusus and Germanicus at Mainz: Tabula Siarensis frag.a.29–32, Crawford (1996) I.37; Suetonius, Claudius 1.3; cf. Lebek (1989) 67–72.

113  Fishwick (1987–) I.298, 301–7; Fitz and Fedak (1993) 265–7. There was an altar, or perhaps even a temple, to Claudius in his lifetime: Simpson (1993) contra Fishwick (1987–) I.195–218, (1991). Cf. a temple for the emperor built A.D. 166–9 by a tribe in Arabia on the encouragement of the governor: Bowersock (1975).

114  Cassius Dio LI.20.6; Crawford (1996) I.493–5 (with different interpretation); Inschriften von Ephesos II 409, V 1517, VII.1 3019. Price (1984) 76–7, 169, 254.

115  Cyrene and Ptolemais: Reynolds (1962), (1965), (1990) 71–2.

116  AE (1955) 210 = P. Wuilleumier, Inscriptions latines des Trois Gaules (1963) no.221, a text curiously ignored in recent years.

117  Fishwick (1978); (1987–) I.150–68, 219–94. Much of this account necessarily remains hypothetical: Le Roux (1994a).

118  Tacitus, Annals I.78. The temple complex was completed only in the later first century A.D., when the provincial priests are first attested, though work on it probably began much earlier: Fishwick (1996) 176–82.

119  ILS 154 (A.D. 18) (trans. Braund (1985) 62); ILS 112 = 10.1b (A.D. 11). Cf. above, p. 207. For a dedication of A.D. 11–12 to the numen of Augustus see AE (1948) 8 = Inscriptions of Roman Tripolitania (1952) 324a (Lepcis), with Fishwick (1992c). These provincial dedications are sometimes seen (wrongly) as overstepping the mark established in Rome.

120  Above, pp. 184–6; Vitucci (1946–85) 403–5; Pascal (1964) 71–3; Ladage (1971) 94–5; Silvestrini (1992); Hänlein-Schäfer (1996). Pompeii: ILS 6381. Ostia: Map 5; H. Bloch (1962); Degrassi (1965); Bakker (1994) 118–33. Numerous cults in the Romanized towns of southern and eastern Spain: Alarcão et al. (1969). Also among Italian business men at Alexandria: CIL III Supp. 12047 = F. Kayser, Recueil des inscriptions grecques et latines (nonfunéraires) d’Alexandrie impériale (Cairo, 1994) no. 5.

121  Price (1984) 147 n.40; above, pp. 343-4 for Nero and Hadrian. For Rome and Augustus at the provincial level, see above, p. 352. Julia Domna: Oliver (1940) = 10.5c.

122  There is here a contrast with the provincial cults in the Greek east, which resulted, as we have seen, from local initiatives, though the contrast is not an east/west one but one of status: Augustus instructed (rather than merely permitted) Roman citizens in Asia and Pontus-Bithynia to establish cults of Roma and divus Julius.

123  Above, p. 354. Another Spanish province, Lusitania, follows a similar pattern: a temple of divus Augustus in the colonia of Merida, and a provincial flamen (of Livia) in the Claudian period.

124  AE (1986) 333, paras. 31, 90, 92 (Lex Irnitana) (trans. JRS 76 (1986) 182–99). Cf. above, p. 251; Herz (1975), (1978); Fishwick (1987–) II.l, 482–501; Rüpke (1995) 540–6.

125  Nîmes: Amy and Gros (1979); Gros (1984). Vienne: Pelletier (1982) 446–52; Hänlein-Schäfer (1985) 244–6; André et al. (1991). Inscriptions: Hänlein-Schäfer (1985) 90–3. For the position of imperial temples in the transformation of urban space, Gros (1987), (1991).

126  Tacitus, Annals I.78. Price (1984) 65–75 on the relations between central and local initiatives.

127  Civic priests in the west, of Roma and Augustus, and to lesser extent of other emperors: Geiger (1913); Etienne (1958) 197–250; Fayer (1976) 213–54. For local games see e.g. AE (1992) 374 (territory of Amiternum, A.D. 2).

128  ILS 6964 = 10.4e. The flamines of the Three Gauls capitalized on their position by erecting statues of themselves and their families in the sanctuary at Lugdunum: F. Richard (1992). The priests of the cult at Tarraco were also called flamen, and also wore the special Roman hat (apex).

129  Duthoy (1978); Ostrow (1985); Fishwick (1987–) II.l 609–16.

130  ILS 112 = 10.1b, with Kneissl (1980). The ‘knights from the people’ were probably inhabitants of the colonia who were not colonists but did have the equestrian property qualification. The council could grant honorary membership of the order of Augustales: AE (1987) 239 (Terracina).

131  This is in fact a classic case of the modern tendency to lump together some rather diverse epigraphic evidence under the heading of Augustales.

132  Abramenko (1993a). For a similar problem with ‘haruspices’ see above, p. 20.

133  AE (1978) 119; cf. (1989) 181.

134  Symbols: Petronius, Satyricon 30, 71.9; on tombstones, Schäfer (1989) 55–6, 218–21.

135  R.J.A. Wilson (1990) 111–13, 297; De Franciscis (1991); Patras: Archaeological Reports 1987–88 (1988) 28–9. However, the building at Sarmizegethusa has now been re-identified as a basilica: Etienne et al. (1990). Statues: Trebula Suffenas AE (1972) 154 = Supplementa Italica n.s. 4 (1988) 178 no. 43) and Aquincum (CIL III 3847).

136  Jupiter: AE (1990) 138 (Terracina). Lares Augusti: Etienne (1958) 275–9; Fishwick (1987–) II.1, 614–16; AE (1992) 302; Petronius, Satyricon 65 = 8.6b.

137  So, for example, we must imagine that the office of provincial priest would bring its holder into contact with (and within the patronage of) the local Roman governor.

138  Matthew 22.15–22; Mark 12.13–17; Luke 20.20–6.

139  Inschriften von Ephesos 24. Price (1998) ch.2.

140  Supplementum Epigraphicum Graecum XXXV.744 (A.D. 1). Cf. 10.5, and generally Price (1984); some criticisms in Friesen (1993) 142–68.

141  Philo, Embassy to Gaius 355–7 = 10.6b; above, pp. 284; 341.

142  Price (1984) 220–2; above, pp. 239–41.

143  Above, pp. 36–7; Price (1984) 207–33.

144  Philo, Embassy to Gains 353–7 = 10.6b. Admittedly the emperor was the bizarre Gaius Caligula, but he merely made explicit a generally latent issue.

145  Above, p. 241.

146  L. Mitteis and U. Wilcken, Grunzüge und Chrestomathie der Papyruskunde (Leipzig 1912) I 96 (trans. An Economic Survey of Ancient Rome, ed. T. Frank (Baltimore and London 1936) II.662–8); Sammelbuch 9489 = Papyri della Università degli Studi di Milano IV (1967) 233 (Tebtunis) might refer to this temple. Glare (1994) notes earlier interest in ‘the imperial cult’ at Ptolemais. Cf. Rübsam (1974) 47–52; M. Walbank (1989) 381–3 for Greek cults. The only Egyptian parallel is the Capitolium at Oxyrhynchus, in existence by the late second century A.D.: Oxyrbynchus Papyri 2128 etc.

147  On Roman religion in the third century A.D., Alföldy (1989).

148  For one aspect of this issue, see Price (1984) 234–48.