How exactly religions spread in the Roman Empire between Augustus and Constantine is a deceptively simple question. This article seeks to show that the question is more complex than it might seem at first sight, but also that some patterns can be discerned.Footnote 1 Obviously, many cults are newly attested all over the Empire: Jupiter Optimus Maximus, Isis, Jupiter Dolichenus, Mithras, Judaism and Christianity are just the most obvious. How these cults spread and the preconditions for religious mobility need some thought. Of course, cults do not move on their own, their movements depend on the movements of people. Take, for example, the ambassadors to Rome representing the Alexandrian Greeks in their dispute with the Alexandrian Jewish community. According to a later propagandistic Greek account, each party brought to Rome ‘their own gods’, the Greeks a bust of Sarapis, and the Jews presumably sacred books. The leader of the Greek embassy, Hermaïskos, was standing up valiantly against the hostile emperor Trajan, when suddenly the bust of Sarapis broke into sweat. Trajan was amazed, crowds gathered, and fled to the hilltops.Footnote 2 This fine story is emblematic of the ways that cults formed part of people's literal or metaphorical baggage.
I propose to focus on the movements of people, and the different sorts of cults that they carried with them. Approaching this topic through analysis of one particular cult, for example Isis, Mithras, Judaism, or Christianity, is common, but unsatisfactory, as it runs the risk of failing to present what is distinctive about each cult, or what is common between them. It is also traditional to analyse the pagan cults in terms of their origins, or rather their alleged origins — Rome, Asia Minor, Syria, Egypt and Persia. But distinction on the basis of ‘origins’ is also extremely unhelpful, as it fails to take into account the social contexts of the cults. I want to suggest that it is crucial instead to distinguish between two sorts of cults, ethnic cults and elective cults: that is between, on the one hand, cults which are part of the actual or imagined ancestral heritage of a genos or ethnos, and, on the other, cults which a person chooses to join.Footnote 3 In modern scholarship, the distinction between these two types of cults is mirrored in the divide between the work of Toutain, who emphasized civic (or ethnic) cults in the Latin West, and that of Cumont on ‘Oriental’ (or elective) religions.Footnote 4 Of course, in practice the relationship between the two ideal types of the ethnic and the elective is very complex. Some, perhaps many, cults were both ethnic and elective. They had an ethnic base, but also attracted in outsiders. Nonetheless, the overall typology is useful, and the two sorts of cults involved quite different dynamics.
I shall be exploring the patterns of interaction that often go unexplored under the heading of ‘religious mobility’, and exposing some of the questions left unanswered by that convenient shorthand. What are the different ways in which a religion can ‘move’ or ‘spread’? What conditions in the Roman Empire made religious mobility more or less likely? What kind of connections (within and outside the family) promoted the spread of a new or foreign cult? Why, for example are so many different religions represented in the tiny town of Dura Europos, and comparatively few in the much bigger settlement of Pompeii? Reflection on these issues will, I hope, offer fresh insights into some of the big problems of Roman imperial religion, from the distinctions to be drawn between western and eastern religious traditions to the rise of so-called ‘monotheism’ and the place of Christianity within the Roman Empire and the wider geographic and symbolic world.
I shall be drawing on a wide range of sources: literary, epigraphic and archaeological. And from time to time I will explore the patterns of the spread of cults as they can be reconstructed through material that survives on and in the ground. But first a word of warning, in general, about the use of distribution maps in reconstructing religious mobility. It goes without saying that they can plot only the surviving or archaeologically traceable evidence; and we need always to ask how representative that evidence is of what there once was. Sometimes we can be misled by our traditional assumptions about the physical form that particular cults took. We are used, for example, to thinking of the evidence for the cult of Mithras as very solid, but it has become clear that such solidity is only part of the picture. Of the seventeen sanctuaries of Mithras discovered in the north-western provinces since the Second World War, most are not marked by well-built stone architecture, or by what we think of as the conventional sculptural monuments (Mithras killing the bull, etc.). For example, at Tienen in Belgium the sanctuary, dating to the third century a.d., was conventional in its size (12.5 by 7.5 m) and in the fact that it was sunk 1.2 m below Roman ground-level, presumably to produce the effect of a ‘cave’, but it was not built of stone and had no stone sculpture.Footnote 5 In an age of less careful excavation it would have been missed entirely, with perhaps just a report of the chance discovery of some Mithraic ‘small finds’. To put this another way, if we rely too heavily on the well-known reported remains, we may end up with a map that plots our own preconceptions of Mithraism rather than the cult itself.
I shall begin by outlining some of the key features of what I have termed ethnic cults (I), and elective cults (II) and the relationship between them. I shall then go on to discuss the local (III) and imperial (IV) contexts of those interactions, before sketching out the stages by which just a few religions came to transcend the boundaries of the Roman Empire.
I ETHNIC CULTS
To start with ethnic cults, the most important are Roman cults, too often ignored by people working on religious change in this period. The outlines are clear enough.Footnote 6Coloniae of Roman citizens established for the urban poor of Rome and for veterans of the Roman army in the late Republic and early Empire included specifically Roman rites: the founding of coloniae echoed the foundation rituals of Rome itself, with the taking of auspices and the ploughing of a furrow around the new city, as on the relief from the colonia of Aquileia; pontifices and augures, modelled on the priesthoods at Rome, were found in many, perhaps all, coloniae; Capitolia, shrines to the Capitoline triad of Jupiter, Juno and Minerva, were common in coloniae, though they do not necessarily date from the time of foundation. But there was no exact blueprint, no imposition of a standard model. Instead, new coloniae borrowed from Rome, or from what they imagined Rome was like, sometimes very closely, in their desire to establish themselves as high-prestige mini-Romes in alien lands. Individual Roman citizens living in provincial communities other than coloniae might also adopt similar strategies — at Nicaea in north-west Asia Minor dedicating statues of the Capitoline triad to the local god, or in Egypt arranging for the purchase of cockerels for the festival of the SaturnaliaFootnote 7 — to remind themselves, and the surrounding community, of their prestigious Roman status.
Such borrowings from Rome were rooted in the imagined past of Rome itself. An important part of Roman memories about their past was that Aeneas managed to rescue his household gods, the Penates, from the sack of Troy and brought them with him safely to Italy, establishing their cult at Lavinium, not far from where Rome was to be.Footnote 8 The large series of reliefs of the Homeric stories produced in Rome in the first century a.d., the so-called Tabulae Iliacae, includes Aeneas carrying his father Anchises with the Penates on his lap, first leaving Troy and then boarding the ship that would take them west. The story was depicted in the Forum Augustum, from which it was picked up in the decoration of the Sebasteion at Aphrodisias. It was also well-enough known to be used on the tombstone of an Italian freedwoman, expressing the ideal of devotion between the generations. And it was even parodied in a Campanian wall-painting, with a dog-headed, ithyphallic Anchises carrying not the Penates but a dice box.Footnote 9 Not only was Rome itself founded, indirectly, by Aeneas, but important tokens of this event were visible in the living cults of the late Republic and early Empire. Such memories of the movement of cults that lay behind the foundation of Rome were built on and adapted by those creating new communities that were modelled on Rome.
At the same time as Roman cults were being recreated in the provinces, both east and west, eastern ethnic cults were being brought to the West. Here the picture is in detail extremely complicated, but at least the outlines are clear. By the time of Augustus, according to a contemporary observer, Rome itself had become the home for innumerable ethnic groups, all ‘of absolute necessity worshipping their ancestral gods according to their local customs’.Footnote 10 A vivid example of this general phenomenon is provided by the sanctuary of the Palmyrene gods in Trastevere on the west side of the river Tiber.Footnote 11 The sanctuary contained dedications by immigrants from Palmyra to ‘their ancestral gods’. The languages of the dedications emphasized the dedicators' eastern roots: Greek and Palmyrene were employed, as well as Latin. For example, the latest such monument from Rome depicted Aglibol and Malakbel, with a dedication in Greek and Palmyrene.Footnote 12 The dedication was to ‘the ancestral gods’ by a man who described himself, in the Greek version, as a Palmyrene, and who dated the offering in the Seleucid calendar used at Palmyra. The two gods were worshipped together in Palmyra, and the iconography of the gods shaking hands employed here reflects that used in the homeland, the cypress tree between them alluding to their ‘sacred grove’ at Palmyra.
Diaspora Jewish communities also belong in the context of ethnic cults. At Rome there were at one time or another at least ten synagogues, with a community numbering in the thousands.Footnote 13 Adherence to eastern origins is evident in both language and iconography. In the Jewish catacombs Greek was used in three-quarters of the epitaphs, and imagery was based on a specifically Jewish repertoire (including the menorah). The social dynamics underlying the diaspora differed in different parts of the Empire. In the Greek East, widespread Jewish settlements are found in the Hellenistic period, resulting from voluntary emigration by Jews, presumably for commercial reasons amongst others, much as with the Palmyrenes and many of the other ethnic groups in Rome. In the early Empire, Jews lived in the port cities of Puteoli and Ostia, again for commercial reasons. But in Rome the Jewish diaspora was probably largely the result of Roman enslavement of Jews in Judaea, from the time of Pompey onwards. For example, one of the synagogues in Rome was named after a Volumnius, probably the procurator of Syria under Augustus, and the community may originally have consisted of the family's slaves and freedmen. In the middle of the first century a.d. a well-informed Alexandrian Jew, Philo, believed that most of the Jewish population of Rome under Augustus consisted of freedmen, brought to Italy as war captives and set free by their owners, without having been forced to alter any of their ancestral customs.Footnote 14 The suppression of the major revolts in Judaea in a.d. 70 and 135 resulted in tens of thousands more Jews being sold into slavery, some of whom were brought to the West. This may account for the Jewish community in Carthage, not attested before the second century a.d.Footnote 15
The Palmyrenes and the Jews exemplify a general pattern of continued devotion to their ancestral cults by those from the eastern half of the Roman Empire.Footnote 16 Examples abound of this devotion on the part of people from Greece, Asia Minor, Syria and Egypt. The reasons for migration were very varied — for economic reasons, because of enrolment in the Roman army, or because of enslavement — but the key point, not always sufficiently emphasized, is that evidence for the movement of religions must be set in a socio-political context.
The actual distribution patterns of these mobile cults varied, depending on the specific reasons for migration — eastern traders did not always go to the same places as Roman soldiers — but the principles of transmission were the same. For cults of soldiers, we are fortunate to have a fairly good grip on the movements of groups of legionary soldiers from one posting to another, postings sometimes involving moves of thousands of miles. In the case of auxiliary cohorts we know from their names where the cohort was levied, and so how far a particular group of soldiers moved.
The situation is more murky with the movements of civilians. At least in the major cities, there were regulations about rights of residence and there were the means to enforce those regulations,Footnote 17 but it is also clear that there was a huge amount of movement of people within the Roman Empire, with individuals flowing fairly freely, and goods simply liable to local and Roman taxes. This can all look very random, but of course the movements of people are not random, just extremely complicated to picture, especially on the basis of surviving scraps of evidence. So we might note the man who is probably a Syrian trader at Lyon, whose tombstone noted him ‘bringing to the Celts and the land of the West all that God has fixed to be born by the land of the East, fertile in all products’.Footnote 18 In addition to such individual items, there is the varied evidence for common routes for the transmission of material goods.Footnote 19 Long ago, Cumont realized the importance of trade routes but, at the time he was writing, scholars accepted the dominance of ‘Oriental’ centres of production and exchange. Of course those centres were important (witness our Syrian trader at Lyon), but today we can no longer accept ‘Oriental’ economic dominance. Instead, the modern picture is of a much more multi-centred economic world, with North Africa and Spain also having major centres of production of goods such as grain, wine, olive oil and fish sauce. This new picture no longer provides the neat fit that Cumont saw between trade centres and the spread of ‘Oriental’ cults to Italy and the West. In other words, some, but only some of the mobile traders worshipped their ancestral cults in their new homes.
In particular, the striking absence of western Mediterranean cults in Italy and the Greek East requires explanation. People moved east from Spain or Gaul, but did not obviously take their local ancestral cults with them. The explanation lies in the nature of changes to the religious systems of the Latin West under Roman rule. As a broad generalization, local pre-Roman cults were transformed with the coming of Rome.Footnote 20 The processes of transformation are not easy to plot, because in most cases the local deities become visible to us only under Roman rule, with the greatly increased use of writing on durable surfaces, and iconographic representations on stone. Changes in the nomenclature of the gods are the most obvious point, from Trumusiatis or Tribusiatis to Apollo, or from Mullo to Mars Mullo. Such changes were not merely cosmetic. To call a god not Trumusiatis but Apollo, or not Mullo but Mars Mullo, was to subordinate the local god to the broader Roman pantheon. Such subordination explains the general invisibility of western gods elsewhere in the Empire. Admittedly, soldiers recruited from the Tungri, a tribe living just west of the river Maas in Gallia Belgica, very occasionally made dedications in Britain to their unRomanized local deities (Ricagambeda, Viradecthis, etc.),Footnote 21 but such dedications were rare. The more normal pattern is represented by a dedication in Rome by two members of the Praetorian Guard to ‘the holy, ancestral gods’.Footnote 22 The pair specified their origins (as did other praetorians making similar dedications) as coming from another tribe in Gallia Belgica, in the upper Somme basin, so one might expect that their gods would be local ones, but the long list of their ancestral gods reads as follows: Iupiter Optimus Maximus, (Sol) Invictus, Apollo, Mercury, Diana, Hercules and Mars, all utterly unremarkable in Rome at this time. That this pair of soldiers was completely invested in the transformation of their local religious system illuminates why local western gods almost never formed part of the baggage of migrants from their homelands.Footnote 23 Thus those Treveri who made dedications outside their homeland never made them to Lenus Mars, their principal deity.Footnote 24 However, if the deities ceased to be purely local and became instead regional or ethnic in a broader sense, then they were more obviously acceptable and mobile, as for example with the dedications in Britain to the Italian, German, Gallic and British Matronae, and to the African, Italian and Gallic Matres.Footnote 25
In the case of ethnic cults from the East, the routes of transmission can in some cases be plotted. Take, for example, cults of Isis in the West. As Bricault's excellent map shows, most of the cults lie on or close to the Mediterranean coast.Footnote 26 The cults were brought here by merchants from the eastern Mediterranean. For example, in the Spanish provinces the earliest cult, probably at Emporiae (modern Ampurias), was founded in the first century b.c. by a family from Alexandria, who had business interests at Emporiae.Footnote 27 In Gaul we cannot be so specific, but the pattern is clear.Footnote 28 The cults, dating mostly to the second century a.d., were found mainly near the Mediterranean coast, and it is here alone that we find organized priesthoods and religious associations. Cults were also taken inland up river routes, especially the Rhône, presumably by eastern traders. But there are large blank areas on the map. The cults were not taken further, to inland Iberia, Gaul or Britain. Contrary to impressions given by some studies, cults of eastern origin were not ubiquitous.Footnote 29
II ELECTIVE CULTS
The spread of entirely elective cults was also dependent on people moving, but these cults were more complicated in that they required the creation of new worshipping groups. Take Mithraism as an example. At Aosta our only evidence of the cult is a dedication to Mithras by a circitor, a travelling customs officer, belonging to the Gallic customs region, the Quadragesima Galliarum, which extended from the Rhine to the Alps.Footnote 30 As he may have been on the move, he could make a dedication, but was hardly in a position to found a new cult group at Aosta. But other movers did have that opportunity. For example, one Firmidius Severinus served for twenty-six years in the Roman army, in a vexillation of the Eighth Augustan legion seconded to Lyon, and then retired to Geneva. Still styling himself ‘soldier’ and not ‘veteran’, and so presumably soon after his arrival at Geneva, he dedicated an altar to ‘the Unconquered God, Genius of the place’ in a.d. 201.Footnote 31 As the phrase ‘Genius of the place’ is highly unusual in this context, and as the altar was dedicated as the result of a vow, we should assume that Severinus on settling in Geneva managed to found a Mithraic sanctuary and association. Something similar may have happened also at Tienen, which we noted above. Tienen is notable because it was not a military settlement, which was the usual context for Mithraic worship in the north-western provinces. It is also important because the wonderfully careful excavations have shown that a grand feast for over one hundred people was held here, perhaps at the summer solstice, and perhaps to mark the building or renovation of the sanctuary. Such a feast must have been held outside the building. This is a striking corrective to the standard view that all Mithraic rituals were held away from the public eye inside the cult buildings. In this case, one might hypothesize that a veteran from the Roman army retired to Tienen, creating a new Mithraic sanctuary, and inviting much of the town to a special feast.
In some cases, it is possible conjecturally to reconstruct not the movements of individuals, but routes of transmission of cults. In Syria there is evidence for the cult of Mithras from the mid-second century a.d. onwards.Footnote 32 It is attested at three coastal sites, notably Caesarea Maritima at the south, and at inland sites: three between Bostra and Damascus; two north-east of Apamea; and Dura Europos off to the east. On the model of Isis in Gaul, one might think that the cult was taken inland from the coastal sites. This is probably true for the southerly inland sites, from Caesarea east to Bostra. But the cult at one of the more northerly coastal sites, Sidon, is attested only in the fourth century a.d., and in any case geography — the great ranges of mountains behind the two northerly coastal sites — makes this model of transmission unlikely for the northerly inland sites. The cults probably reached the northerly inland sites as a result of troops travelling along the major Roman road running from Europe through the Cilician Gates to the Euphrates. Certainly, at Caesarea the Danubian-style roundel proves that there was direct connection between the cults of Mithras on the Danube and in Syria, and at Dura the cult was broadly-speaking indebted to cults in Italy and more generally the Rhine-Danube frontier zone.Footnote 33
As the example of Mithraism shows, people who travelled sometimes took cults with them. A Christian carpenter Papos, born at Arados in Phoenicia, worked and died at Nicomedia in north-west Asia Minor, aged forty-one.Footnote 34 He presumably created a Christian family for himself, or at least found a Christian community that knew him as Eumoirios (‘Blessed’) and buried him as he wished, with a discrete but unambiguous cross carved towards the bottom of the tombstone. Or there is the case of Arberkios, Bishop of Phrygian Hierapolis, who travelled west to Rome and east to Syria and beyond to Nisibis in Mesopotamia, apparently meeting Christians everywhere he went.Footnote 35 Aberkios represented himself as travelling within a network of pre-existing Christian communities. At the same time, pagans travelled to healing cults, oracles or Panhellenic festivals, similarly operating within a pre-existing network, not creating new cults.Footnote 36
As part of this travelling world, cults and religious objects could be taken anywhere. Consider just two examples. At the great healing sanctuary at Grand in the Vosges, a sanctuary rooted in local traditions, the finds include splendid ivory astrological tables from Egypt.Footnote 37 We have no idea how they got there, but for them to be useful, there would have to be also a person learned in arcane lore. From the East, in the highlands of Phrygia in central Asia Minor, about as far as one can get from major centres of population, there is rich epigraphic evidence of Christian communities in the second and especially third centuries.Footnote 38 At Eumeneia and its surroundings about twenty funerary texts have Christian formulae. In the territory of Appia, in the upper Tembris valley, more than twenty tombstones inscribed ‘Christians for Christians’ are known, dating between the mid-third and mid-fourth centuries. Or at Temenothyrae, in western Phrygia, we find a community of Montanists (a Christian prophetic movement), which is consistent with the literary view of Montanism as originating in Phrygia. So all nicely remote, or maybe not. After all, Phrygian Christians in the second century were in contact with their fellow Christians in Lyon,Footnote 39 and in turn Christian authorities took action against what they saw as the Montanist heresy. The Roman Empire depended on connections: everyone in the Empire would know of a change of emperor as fast as communications permitted. (There was surely no-one in the Roman Empire like the Japanese soldier who fought in the Philippines during the Second World War, and refused to believe in the Japanese defeat, surrendering only in 1974.) Nowhere was truly isolated in the Roman Empire; everywhere was connected in one way or another.
The elective cults required new religious communities to be formed for them to operate fully.Footnote 40 There arises therefore the major question of how new people were brought into elective cults. Public spectacle and shows are not likely to provide the answer for newly arrived cults. Mithraists and Christians generally did not put on public rituals, and —though the public ceremonials surrounding the martyrdom of Christians may, as is often claimed, have attracted some new adherents — they were hardly in themselves great advertisements for the cult. Nor, I think, is it useful to think of a Market Place of Religions, where individuals could shop around for the cult that suited their needs.Footnote 41 The model of a ‘market’ with choices being made by individual consumers is surely anachronistic. All the cults discussed in this article, ethnic and elective, pagan and non-pagan, were based on groups, large and small. Those groups created religious ideas and meanings, and expected high levels of engagement. There is no sign of a consumer approach to religion. Rather than spectacles and displays, soap-box evangelism, or Market Places, we should think instead, as many have said in the past, of the dissemination of cults by personal contact, within family, professional or social contexts.Footnote 42 That is easy to say, and is commonly assumed, yet it is only a starting point. We need to go further and ask in what contexts such contacts worked and in which others — and why — they did not.
The cult of Mithras offers perhaps our best vignette of contexts and mechanisms for any of the elective cults. A magnificent bronze tablet lists members of the Mithraic sanctuary at Virunum (in modern Austria).Footnote 43 The document was drawn up originally in a.d. 183, with a list of thirty-four names filling the first column and a half, recording those who had paid for the restoration of the ‘temple’ of Mithras, perhaps after a storm. The lay-out shows that the originators expected to add new names as time passed. Indeed, in the first year or so, probably as a result of the plague, no less than five members died, their names being marked with a theta, for the Greek thanon. Eight new members were enrolled in a.d. 184, and, according to the very plausible theory of the original publisher of this text, new names, sixty-four in all, were added annually until a.d. 201, the additions being in very obviously different hands. In this document we have a unique glimpse of the ongoing recruitment over a seventeen-year period to an existing Mithraic community, which had perhaps been founded a generation earlier.Footnote 44 How this recruitment took place, both initially and subsequently, is the issue. Virunum was the capital of the province of Noricum, and remained the centre for the administration of the province's finances, though some elements of administration had recently been moved elsewhere.Footnote 45 Unsurprisingly, all but two of the members were Roman citizens (with one free non-citizen and one slave member). About a quarter of them were freedmen, and very few had local, Celtic names. In general, the members seem modestly successful — rich enough collectively to pay for the rebuilding of the ‘temple’, without the intervention of any named patron. In addition, one member was a local magistrate and imperial priest, another was able to pay for the bronze plaque and the paintings on the ceiling of the ‘temple’, but neither were patres, holders of the highest grade in the association. Some of the members were close kin (father-son, or brothers), others were freedmen of existing members. Ongoing recruitment happened within families (kin and freedmen) and from other local contacts. In addition, we know of two new members who were immediately made patres. They must have brought their Mithraic rank with them when they moved from elsewhere, which is an interesting point about the recognition of a Mithraic pattern of initiation.
We may agree that spread through personal contacts and individual movements is basic, indeed obvious, but we need to think more carefully about exactly what model we might best use. Contemporaries opposed to religious innovation employed the vigorously negative model of the spread of diseases. Pliny, commenting as Roman governor on the spread of Christianity in Pontus in northern Asia Minor, said that ‘it was not only in towns, but also in villages and the countryside that the contagion of this dreadful superstition has spread’.Footnote 46 ‘Contagious disease’ is a misleading metaphor. Not only does it embody a negative view of the spread of cults (which will hardly do); it also leads us to think of a linear spread, along sea- or river-routes, and along the land arteries of the Empire, whereby the new cult infected each place through which it passed. The varied case-studies which I have already presented have suggested that continuous linear diffusion was not the case. If personal contacts are basic, then the movement of individuals and groups, even if along obvious routes, does not entail the creation of new cults all along those routes. It must be much more rooted in community, contact and interaction.
So the question is: what sorts of personal contacts are at issue? At this point, a distinction favoured by modern sociologists is useful. Inspired by the American sociologist Mark Granovetter, they distinguish between ‘strong’ and ‘weak ties’.Footnote 47 Strong ties are links between an individual and family or close friends; weak ties are the links between an individual and acquaintances. We tend to imagine that strong ties are the ones that are most important in our lives. Granovetter does not dispute their significance for each of us, but he has shown that weak ties have crucial strength (hence his catch phrase: ‘the strength of weak ties’). Weak ties enable us to reach out beyond our closely bound network of family and close friends to another, loosely-connected network, in which few of one's acquaintances may know each other. Granovetter's interest was in the ways that people make use of their strong and weak ties to get jobs for themselves. We might hypothesize that those people with a good set of weak ties are in a strong position to bring in members to a new cult; in fact, it would only be possible for an elective cult to gain a substantial number of followers if it did exploit such weak ties. People working together in imperial finances at Virunum would be a good example of a group of such people. Strong ties, say within a particular family, could also be important, as we know was the case in some ancient cults, but would not lead so easily to their wider dissemination. For example, an inscription records the numerous members of the ancestral cult of Dionysus within a senatorial family in Rome, but this was an essentially static context.Footnote 48 This model of connection, approved from the top, was unproblematic for members of the élite, but innovation from below, infiltrating into such families, was very likely to be viewed negatively by them.Footnote 49
Given the importance of personal contacts, and especially of personal contacts involving weak ties, the next question is why people listened to an acquaintance suggesting that they join this new cult. This raises all kinds of unanswerable questions about the day-to-day interaction of ‘ordinary’ men and women in the Roman world, about who could say what to whom, and about different forms of persuasion. But one thing is clear: context is crucial. The listener needs to be receptive for the message (however it is delivered) to be successful. We should not assume, however, that because a particular cult or message did turn out to be successful, the world must therefore have been ‘yearning’ for it. For example, some of the cults constructed death as much more of a ‘problem’ — and at the same time offered a ‘solution’ to it. But that does not mean that the pre-Christian world in general was yearning for a new answer to the ‘problem’ of death.Footnote 50 That would be an entirely circular argument.
Two arguments about context were once very influential. Scholars used to argue that the Greek world after Alexander the Great was essentially deracinated, with populations uprooted from the communal ties that had bound them together in the Archaic and Classical periods. The second argument was that civic cults had lost their meaning for people, leaving them adrift in a vast and potentially meaningless world. For the Greek East, this second argument followed closely from the first. For Rome, it had different roots, in the old orthodoxy of the emptiness of Roman cults in the late Republican and Imperial periods. As Cumont said, ‘Il n'a peut-être jamais existé aucune religion aussi froide, aussi prosaïque que celle des Romains’.Footnote 51 This dramatic statement is an important element in Cumont's explanation of why his oriental cults spread as they did. But today the Hellenistic world is seen not as deracinated, but as firmly rooted in evolving civic structures. And civic cults, both Greek and Roman, are seen as lively focuses of communal and individual energy, and as continuing bearers of meaning for their citizens.Footnote 52
The current argument that has taken over as a major feature in the analysis of context is the alleged rise of monotheism.Footnote 53 Polytheism, it is claimed, with its multiplicity of gods, each with different functions, was intrinsically less attractive than monotheism in offering an over-arching framework of meaning. Such a framework was especially important in a world whose horizons were now so broad. Whereas civic cults, vibrant though they were, offered only fragmented world-views, the new cults, whether of Isis, Mithras, Judaism or Christianity, were alike in offering the idea of one god, which entailed the construction of global and hierarchized world-views. Common though this argument is, it is in my view fundamentally flawed.Footnote 54 We certainly cannot assume that monotheism triumphed because it is inherently more coherent and hence rational a system than polytheism (drawn though we might be to such an assumption by the modern dominance of Judaism, Christianity and Islam). Polytheism was, and is, capable of constructing viable and meaningful world-views, and might indeed be thought of as a more complex and sophisticated system than monotheism. The point is evident from structuralist and later studies of the Greek gods, and from studies of Hinduism in India today.Footnote 55
But we should not simply assert the equality of the two terms, monotheism and polytheism, but rather question the meaningfulness of the alleged opposition between them. After all, the key point about the Greek and Roman gods is that the ancients both recognized their variety and multiplicity, and also conceived of them as forming some sort of unity. The conjunction of these two views is not a relatively late development in the Roman Imperial period, but went back to much earlier times.Footnote 56 For example, Xenophon recounts the following story about his own military exploits: when attacking a city, but finding himself trapped inside, he realized that it was only divine power that had intervened and had enabled him and his men to escape. Xenophon did not see the epiphany of a particular god, but talked of ‘one of the gods (θεῶν τις)’ as providing the means of salvation.Footnote 57 Or to take another example, when Greeks and Romans created anthropomorphic cult statues, they were not committed to the belief that the gods were just like people (a belief which is, and was, absurd to the philosophically minded), but rather to the idea that the human form offered one way of representing the gods.Footnote 58 In other words, polytheism was an attempt to understand and make sense of divine power. As such, the multiplicity of the gods was perfectly compatible with a sense of the oneness of being. It follows that there was no trend towards ‘pagan monotheism’, because there was no perceived lack of monotheism. There was a whole series of changes and developments within paganism, including the idea of declaring one's commitment to a group, and the development of new types of world-views. Perhaps the key development comes with the Jewish and then Christian idea of exclusivity of commitment, of the idea that our god is not just the over-arching and all-encompassing deity, but also a jealous deity to whom alone worship should be given.
III THE LOCAL CONTEXT
In addition to worrying about how we should characterize the broad cultural and conceptual context of the spread of cults in the Roman Empire, we need also to pay proper attention to local contexts. Local contexts offer another important way of thinking about the receptivity of individuals to propositions put to them by their acquaintances. At the level of individual cults, it is often claimed that the existence of a Jewish community in a particular town smoothed the path for those seeking to create a Christian community there. The evidence for this point is not straightforward. The Acts of the Apostles often shows such a situation, but its picture is so highly tendentious in relation to the Jews (who are shown at the end as turning their backs on the teachings of Christianity) that it is hard to take individual anecdotes at face value.
Is it more fruitful to think about the structural implications of variations in the scale of urban complexity? Were the smallest towns the most homogeneous, and the least open to new cults? Certainly, Rome, the largest city by far in the Roman Empire, also had the greatest range of ethnic and elective cults of any city in the empire.Footnote 59 The point was obvious to the jaundiced eyes of Tacitus, who viewed the spread of Christianity from Judaea to Rome as characteristic of the way that ‘all hideous and shameful practices collect and gather [in Rome] from every quarter and are extremely popular’.Footnote 60 The public festivals of the city provided some sort of framework for all those in the city, if they chose to participate, but alongside them flourished all sorts of other cults. And Rome was wholly exceptional, not just for its size, but also for the variety of cults worshipped there. Ephesus and Carthage, among the next largest cities in the Empire, each had a much more restricted range of cults. So too with the more modest coastal towns round the Bay of Naples. At Pompeii and Herculaneum the ancestral civic and domestic cults provided the principal religious framework for most of the population of the towns. There were cult associations — to Isis, Venus, Bacchus and Sabazius — but it seems very likely that those associations were marginal to the life of the towns.Footnote 61 By contrast the Christian communities in tiny towns in Phrygia, at which we have already looked, show us that flourishing new communities can be found in places that we might have expected to be the bedrock of conservatism and commitment to ancestral deities. In other words, variations in urban complexity do not predict much in terms of religious diversity.
Nonetheless, it is important to consider how much variety there was in religious life in towns in the Empire. Were they mainly like Pompeii and Herculaneum, or were they mainly much more diverse? The eastern town whose religious life we know best is Dura Europos in eastern Syria, with its famous Mithraic sanctuary.Footnote 62 Because Dura was sacked by the Sasanians in a.d. 256 or 257 and then abandoned, the third-century town is well preserved. Its state of preservation makes it tempting to treat the town as ‘potentially our best case study for social and religious life in a normal Near Eastern small town under the early and high Empire’.Footnote 63 The underlying problem is that the excavations carried out here under the aegis of Cumont and Rostovtzeff, though written up with brilliant panache, need further work, now being carried out by the Franco-Syrian team under Pierre Leriche. In addition, few have taken the opportunity to look synoptically at the evidence for all the cults found in the town, though Ted Kaizer has now taken on this daunting task. In our present state of knowledge, it is clear that there were numerous religious buildings, fifteen in all, for a settlement covering about 60 hectares, with a total population which I would estimate at about three thousand people in the Hellenistic period.Footnote 64 What do we make of a small town, little more than a village, often assumed to be an isolated outpost of the Roman Empire, with such a variety of religious options?
I would like to raise the question of how typical Dura was, and so how far, and in what ways, we can generalize from it. Dura, meaning ‘fortress’ in Aramaic, lies on a steep bluff overlooking the Euphrates, protected by wadis on two sides. The town was founded as Europos in the third century b.c. by the Seleucid king Seleucus Nicator, was captured by the Parthians towards the end of the second century b.c., remaining in their hands, apart from a brief interlude in a.d. 115, until captured by the Romans in a.d. 165. There was a modest Roman military presence for the next thirty years, but in a.d. 194 Dura became a major Roman military base, first against the Parthians, and then the Sasanians. Fortifications surrounded the town on three sides, with a great gate at the entrance of the route to Palmyra. On the highest point of the site was a large fort, with cliffs below it dropping down to the Euphrates.
The fifteen religious places in Dura are extremely varied, falling into at least seven different categories. Worshipped there were Greek gods, whose priesthoods were still held in the second century a.d. by people claiming Macedonian descent: Zeus, Apollo, Seleucid Ancestors, Seleucus Nicator, and also Artemis; Palmyrene deities: Bel, Iarhibol, Aglibol and Arsu, and the Gaddé (or Tyche) of Dura; deities from the village of Anath, 140 km downstream from Dura, deities called Aphlad and Azzanathkona; Aramaic cults: Artagatis, Adonis, and Zeus Kyrios Baalshamin; military cults: Mithras and Jupiter Dolichenus; a Jewish synagogue; and a Christian church. This variety is quite extraordinary.
Dura provides wonderful evidence for people and cults living side by side. But it is hard to tell what they made of each other. In a town of this size, people must have known of the existence of the range of cults. There is no surviving evidence to suggest hostility towards other cults, but then all or most rituals will have been performed within the walls of the individual religious places, not exacerbating negative attitudes held for other reasons. At the most there was, as Jaś Elsner has argued, passive resistance, as individual cults articulated their identities and expressed their superiority over other cults.Footnote 65
Who brought this great variety of cults to Dura? Some cults were maintained as part of the Macedonian heritage of the town. Others were established by people from Palmyra. Some were brought in by Roman soldiers (legionaries transferred from the Danube and auxiliaries from Palmyra). That much is easy. Beyond that we have to think harder about the economy of the town.Footnote 66 The possible peculiarity of Dura is raised if one looks from the town west, into the Roman Empire. Here Dura is, and was in antiquity, surrounded by desert, or steppe if you like, useful only to nomadic pastoralists. On this side Dura had no agricultural hinterland. Hence perhaps the old idea that Dura was simply a ‘caravan city’. In reaction to that idea, some suggested that Dura in the Roman period was simply a military base, with nothing much going on beyond locals supplying the army. In fact, inhabitants of Dura since the Parthian period had owned vineyards and olive trees and grown grain and legumes on the fertile land beside the river Euphrates. Parapotamia, ‘Land beside the River’, the term used in an official document, extended perhaps 200 km along the Euphrates.Footnote 67 Dura certainly had a significant local economy of its own.
In addition, regional and long-distance trade were important. Trade overland certainly took place. For example, the sarcophagus of a rich merchant from Palmyra now in the local museum displays on the side the animal that made possible this trade, a camel.Footnote 68 A man from Hatra, 200 km to the north-east, made a bilingual dedication in Hatrean and Greek in the sanctuary of Atargatis in Dura, recording his gift of money to the Sun god, the main deity of Hatra.Footnote 69 The dedication is most simply explained if the man from Hatra was in Dura because of trading ties. But it was access to the Euphrates that was crucial. The river was a major trading route down to Mesopotamia, the Persian Gulf and beyond to the Far East. Dura's connections with Mesopotamia are exemplified in the close ties that the Jewish synagogue had with that area. Among the texts from the synagogue are graffiti in Persian, recording the (positive) views of Jewish visitors from the east on the synagogue paintings.Footnote 70 Rabbinic texts of this period assume a steady flow of people between the long-standing Jewish community in Babylonia and Judaea. A few Jews may have been travelling to see particular religious teachers, but they cannot have been travelling in this period to celebrate religious festivals in Judaea, and most, including those who came to Dura, must have been traders. As for trade west from Dura, the town may lie on one of the two routes between the Euphrates and Palmyra, some 300 km to the west. That route must explain why Palmyrene religious interest in Dura is a major feature of the town from the first century b.c. onwards.Footnote 71
The evidence for Dura's richly diverse religious life has to be set in context. Dura's political and religious ecology was very peculiar. The town, despite its size and prosperity, had little in the way of ordinary civic institutions, it shows no signs of benefactions to the city by the local élite, and it lacked any ordinary Greek civic or religious architecture.Footnote 72 Some, and perhaps most, of the religious sites were the preserve of particular groups. And those groups were a complex mixture of the ethnic and the elective. I would suggest that the variety of its religious life was the product of its economic life, which brought people here both from its hinterland and from further afield.Footnote 73 In some respects, Dura should perhaps be seen as more comparable to Mediterranean port towns, like Puteoli or Ostia, than to ordinary small towns in the Roman East.
IV THE IMPERIAL CONTEXT
I turn now from the context of individual towns to the broadest context of all, that of the Roman Empire. The first point to make is that most of the ethnic and elective cults did not spread beyond the Roman Empire. There were no cults of Isis, Jupiter Dolichenus or even Mithras across the frontiers, even though some people, especially traders, travelled far and wide. Outside the Empire, I note only a solitary ‘temple of Augustus’ marked on a Roman world map, the Peutinger Table, in southern India.Footnote 74 And conversely, we find almost no cults from outside the Empire within its bounds. Dura, which was in prime position for such cults, has none; no cults from Mesopotamia or further east are found there. Mithraism might seem to be an exception, because of its claim to Persian origins, but of course that claim is largely fictive, and the cult was to all intents and purposes constructed within the Empire.Footnote 75 The one significant exception was Manichaeism, to which we will return shortly, and it was repressed in North Africa by the emperor Diocletian precisely because of its advance from Persia: Manichaeans were attempting ‘through the accursed customs and perverse laws of the Persians to inject people of a more innocent nature, namely the temperate and peaceful Roman race and our whole world, with (as it were) their malignant drugs’.Footnote 76 Diocletian's rhetoric shows that the power of the state could be made to protect the boundaries of the Roman Empire from foreign religious incursions, a rhetoric that also underpinned his vigorous actions against the Christians.
The important issue here is how the various cults represented themselves in relation to the Empire. The mobile ethnic and elective cults continued to focus on their actual or alleged places of origin: Isis in Egypt; Mithras in Persia; Jupiter Dolichenus in Commagene in eastern Asia Minor; Aglibol and Malakbel in Palmyra; Jahveh in Jerusalem. Indeed a focus on now remote places was part of the point of these cults. But they also provided in different ways their own sacred canopies, generalized and sometimes utopian frames of reference.Footnote 77 These sacred canopies related explicitly to the Roman Empire in different ways. They were all at least compatible with the Roman order, with dedications, sacrifices and prayers being offered ‘for the well-being’ of the emperor. The cult of Jupiter Dolichenus is interesting in going further than that.Footnote 78 In the second century a.d. the common name of the deity was not simply Jupiter Dolichenus, but Jupiter Optimus Maximus Dolichenus, and his female consort Juno Regina. That is, the worshippers borrowed the names of two of the principal deities of the Roman state. In so doing they implicitly asserted the over-arching position of the deities of Doliche, perhaps in competition with the Roman state cult. It was the deities of Doliche that provided a sacred canopy for the whole Roman Empire.Footnote 79
The worshippers of Jupiter Dolichenus could make this grand claim with some degree of safety because their cult was, or seemed to be, an ethnic cult.Footnote 80 Worshipping the gods of one's ancestors could be presented as obviously virtuous and almost entirely unproblematic. On the other hand, adherence to elective cults, or adherence by outsiders to ethnic cults, raised serious problems of religious and social identity. Judaism I have mentioned so far simply as an ethnic religion, but of course it was more than that, attracting new, gentile adherents. Jewish proselytes were condemned by Tacitus for their wickedness in scorning their ancestral religion, and according to a Greek historian, Dio Cassius, the emperor Domitian put to death his own cousin and exiled his wife on charges of ‘atheism’ (or what in Latin would have been called superstitio), because ‘they had drifted into Judaism’.Footnote 81
Christianity faced even greater problems in relation to issues of tradition and adherence to ethnic, ancestral cults. A cult whose founder had been put to death by the Romans, and whose adherents could also be executed simply for being Christians was in a tricky position. In the eyes of some outsiders, it was an unacceptable form of superstitio, and could claim no legitimate authority, having fallen away from the faith of the Jews (whose standing was itself problematic), and was now targeting non-Jews as members. These issues come together in a third-century graffito from the imperial palace on the Palatine in Rome: Alexamenos is mocked for worshipping his god, a crucified donkey.Footnote 82
In response to such views, Christianity was presented by some as an ethnic religion, and hence as socially acceptable. Second-century Greek apologists made Christians themselves another genos or race, with its own history and legitimate religious practices.Footnote 83 Aristides, the author of an Apology for Christianity, reshaped the traditional ethnic divisions of the world to accommodate Christians. According to the version preserved in Syriac, there were four divisions: ‘barbarians and Greeks, Jews and Christians’; the barbarians claimed origins in Rhea and Kronos, the Greeks from Hellen, the Jews from Abraham, while the Christians traced the ‘origins of their religion’ back to Jesus the Messiah. In the alternative version preserved in Greek, the world was divided not into four but three races: ‘the worshippers of those called by you gods’ (itself subdivided into Chaldaeans, Greeks and Egyptians), Jews and Christians. It was the idea of three races that became the most common.Footnote 84
The point may have originated in accusations levelled at Christians by outsiders: Suetonius, for example, described the Christians as ‘a genus of people holding a new and mischievous superstitio’.Footnote 85 The Christian apologist Tertullian in his ad nationes, mocked the common accusation that ‘we are called a third race’, claiming that it was not obvious which were the first and second races, and that in any case they were all Christian now.Footnote 86 Whereas Tertullian simply sought to refute the allegation (as many other allegations made against Christians), Aristides and others sought to spin the accusation to their advantage, in the hope of obtaining socio-political legitimacy.
So the Roman Empire was the context for most of these cults, but it did not necessarily define the ambitions of all of them. Mithraism, as we have noted, claimed its locative centre outside the Empire, in Persia, and its initiates proclaimed ‘Hail [to the Fathers] from East to West under the protection of Saturn’.Footnote 87 The Mithraists' imaginative world did not map onto the Roman Empire, but extended from east to west, with a strong astronomical canopy.
The ambitions of Christianity also were not limited by the Roman Empire. I am not thinking in the first instance of the political loyalty or otherwise of Christians, but of how Christians conceived of the position of their faith in the world. The rhetoric of Eusebius, who argued for the loyalism of Christianity, is instructive. Towards the beginning of his work on the Preparation of the Gospels, he noted the multitude of rulers and how violence and disorder was eliminated by Augustus after the birth of Christ, in fulfilment of the prophets (1.4). And in his History of the Church, he claimed that in the reign of the emperor Tiberius Christianity spread rapidly throughout the whole world, with churches established in every city and village. Indeed Tiberius himself, so Eusebius claimed on the authority of Tertullian, threatened the accusers of Christians with death. ‘Heavenly providence had wisely installed this into his mind in order that the doctrine of the Gospel, unhindered at its beginning, might spread in all directions throughout the world.’ That is how Eusebius ended Book 2 of his History of the Church, with ‘the world’ being by implication ‘the areas under the authority of Tiberius’. But he opened Book 3 on a much broader canvas. After a reference back to the condition of the Jews, Eusebius continued: ‘Meanwhile the holy apostles and disciples of our Saviour were dispersed throughout the world’ (3.1). The first region mentioned is Parthia, off to the east (assigned to Thomas), then Scythia, off to the north (assigned to Andrew), and only then other regions that fell within the Roman Empire. But it is very striking that Eusebius does not specify them as falling within the Empire. By this time, he really was thinking of a global religion.
Manichaeism developed the global ambitions of Christianity. Its founder Mani was born outside the Roman Empire, in Mesopotamia, receiving his calling in the year we call a.d. 240, but which he dated by the achievements of the Sasanian kings. His evangelizing began in the Persian empire (as the Sasanians styled it), converting a local ruler to the new faith, appearing before the Sasanian king Shapur I, and then attempting conversions also in India. From c. a.d. 260, he also sent out apostles to the west, to Egypt, Syria, and perhaps as far as Rome. According to a later account preserved in Middle Persian, ‘Many wonders and miracles were wrought in these lands. The religion of the apostle was advanced in the Roman Empire’.Footnote 88 One such miracle was wrought by Mâr Addâ, one of the earliest missionaries in the Roman Empire, who was said to have cured the sister of Zenobia, Queen of Palmyra; the story offers a neat sidelight on the importance of Palmyra on the route from Mesopotamia to the Mediterranean.Footnote 89 We have already seen how worried Diocletian was by the spread of the cult in North Africa, and we now know that the spread of Manichaeism in the Empire was not just a set of ideas that lingered on to influence the young Augustine, but was indeed a religion with real adherents. The Manichaean texts from the Dakhleh Oasis in Egypt's western desert, 600 km south of Alexandria, reveal a flourishing community here in the fourth century, in contact with Manichaeans in the Nile valley, using Syriac-Coptic glossaries of Manichaean technical terms to aid their mission.Footnote 90
Mani's ambitions for his religion were more universal than those of the other religions that he knew — Zoroastrianism, Buddhism, Judaism and even Christianity. Calling himself ‘the Apostle of Jesus Christ’, he articulated a vision of a church that was, he believed, superior to that of Christianity. To Mani was widely attributed an extraordinary ten-point list of the advantages of Manichaeism over the first churches (i.e. Christianity). The first of the ten points observed how the first eastern and the first western churches were divided from each other. Mani's hope was that his proclamation would be heard in both East and West, in every language, and in all cities.Footnote 91 His mission was to spread his universal message, in every language (a crucial point), throughout the world. For him the ideology of the Roman Empire was a matter of supreme unimportance. His was a vision of a world religion.Footnote 92
