Once upon a time, perhaps toward the end of the eighteenth century, at any rate when Greece was still part of the Ottoman Empire, Athens was suffering from a long drought. The Turks went to their mosques and prayed to Allah, to no avail. The Greeks gathered in their churches and sought help from the Prophet Elijah, yet the Attic skies stayed blue for months on end. Last and—in the eyes of the others—very much least, the “Arabs” or “Ethiopians,” the Turks’ black slaves, gathered at their open-air mosque amid the pillars of the old temple of Olympian Zeus, near the base of the Acropolis. With their women and children all together, they prostrated themselves on the ground. Then three times they cried out to God, beating their breasts and pinching their children till they howled. After the first cry, a few small clouds peeped out from behind Mount Hymettus. After the second, the sky became thoroughly overcast. During the third, sheets of rain swept across the little town’s houses, its mosques, and its churches.1

In the multicultural Ottoman Empire, as in the multicultural US nowadays, religious and ethnic identities often reinforced each other. Sometimes ethnic differences might be even stronger than the bonds of shared faith—hence the need of the Athenian “Arabs” for a place of worship of their own, apart from the Turks. The basic exclusivist model was provided by the most ancient of the monotheisms, Judaism. The continuing history of the Jews has provided a permanent example of the close link between religion and ethnicity.

And yet there are elements of the Torah—the Decalogue, for example—that by no means stand in the way of a more open, inclusive, and universalist understanding of religion. In the Hellenistic period, Jews such as Philo wrote about their faith for the benefit of Greeks, and formerly polytheist converts were to be found in many synagogues. The earliest Christians did not immediately or easily disentangle themselves from the synagogue; but the Gospels of Matthew, Mark, and Luke have Jesus send out his apostles to baptize all peoples, while the Apostle Paul made a decisive opening toward the gentile world. As for God’s third great revelation to mankind, the scripture in which it was set down calls itself repeatedly an “Arabic Koran,” and was addressed in the first instance to Arabs. Yet it did not take long for others to decipher it, as they had the Hebrew of the Torah and the Greek of the Gospels.

It was in a speech delivered on the Areopagus hill, just at the opposite end of the Acropolis from the temple of Olympian Zeus, that Paul first revealed to the Athenians God’s new dispensation, while at the same time manifesting a certain respect for their ancient and prestigious culture, at least for the reverence they accorded the “Unknown God.” He even quoted from the poet Aratus. Three hundred years later, not long after Constantine became the first Roman emperor to accept Christian baptism, Bishop Cyril of Jerusalem proclaimed that the Church is called “catholic” because it extends throughout the whole world, but also because it teaches whatever mankind needs for salvation. The Church is the repository of truth in all its aspects.2 Taking as its model the Children of Israel, who “despoiled the Egyptians” before embarking on their exodus, or (more eirenically) Jesus who came to fulfill, not to destroy, the law and the prophets, the early Church was, it seems, disposed to take over whatever there was of truth and virtue to be found in the gentile as well as the Jewish world, before the Incarnation.

Hence the belief that there had been “Christians before Christ”—including the prophets, of course, but perhaps also the ancestors of Jesus who are so carefully listed in the Gospels of Matthew and Luke. Balkan Christians in late Byzantine times and afterward, taking their cue from an iconographic theme already known in thirteenth-century Italy, often painted Christ’s genealogy onto the walls of their churches in the form of the “Tree of Jesse.” Beside the recumbent, sleeping figure of David’s father, Jesse, springs up a mighty vine, in whose tendrils are painted portraits of Christ’s ancestors. Some of these ancestors, though, are not to be found in any gospel: they include, among others, Homer, Solon, Thucydides, Plato, Aristotle, Josephus, and Plutarch—even Aristophanes. The sages of ancient Greece—along with Plutarch and the Jew Josephus, both of whom had lived after Jesus—are in this way accepted as foretellers or witnesses of the Gospel. Each usually holds in his hand a scroll on which is inscribed a prophecy of Christ’s coming—the idea is taken from late antique compilations such as the Prophecies of the Seven Sages, which attributed to various ancient Greek divinities (notably Apollo), and to mortals as well, oracles of the destruction of the old gods and the coming of Christ, along with theological pronouncements of distinctly monotheist tone.


Even though, in the shape of Josephus and Plutarch, a handful of unbaptized figures from the Christian era slipped into the “Tree of Jesse,” they were exceptions. They could not much dilute the Church’s sense that, as the Gospel spread, those who ignored let alone opposed it did so at their own risk. As a sequel to his Paganism in the Roman Empire (1984), the Yale historian Ramsay MacMullen has now written Christianity and Paganism in the Fourth to Eighth Centuries, a study that traces the deliberate resistance offered to the Gospel by worshipers of the old gods, and other ways in which, even when this phase had passed, certain practices associated with polytheism “continued,” as he repeatedly puts it (e.g., pp. 156-157), within the context of the Church. These persistences and echoes of polytheism, which were possible both despite and because of the Church’s basically assimilationist instinct, fall into various categories, not all of them discussed by MacMullen.

First of all, there were worshipers of the old gods who quite simply carried on worshiping them exactly as they had always done. The Church gradually found ways of making such insouciance more and more difficult and even risky, but MacMullen shows how common these persistent worshipers were, especially in rural regions, at least until the mid-sixth century, when the emperor Justinian cracked down on them.

Secondly, there were polytheists who accepted baptism and all outward signs of Christian allegiance, but continued in secret to serve the many gods just as before. We may call these “crypto-polytheists,” by analogy with the much better documented “crypto-Christianity” practiced, for example, by Armenians and Greeks in Asia Minor in the nineteenth and twentieth centuries when, for whatever reason, it seemed advisable to appear Muslim, even if in the privacy of one’s household one continued to live and worship according to the Christian religion. We have an excellent account of a crypto-polytheist temple of Isis at Menouthis, a small town near Alexandria, just at the moment in the late fifth century when it was discovered by the local Christians. According to Zacharias of Mitylene, the temple was installed in a private house that belonged to a priestess. It was in a sealed-off room that could be reached only through a narrow window—the priest, too, had to squeeze through the window in order to offer sacrifice.

When the Christians got wind of this arrangement, they broke in and found a collection of idols and an altar drenched with blood. Crying, “There is only one God,” a mob set upon the images and made a bonfire of the less valuable ones, including statues in the Greek style, as well as the Egyptians’ usual collection of embalmed dogs, cats, monkeys, and crocodiles. The crowd then demolished the house itself, and the rest of the statues they carried off to the bishop, twenty camel-loads of them. It turned out that the polytheists had been bribing the local Christians to turn a blind eye to their gatherings and rituals.

The crypto-polytheists of Menouthis seem to have been small fry, and once exposed in such abrupt fashion they probably had little courage to carry on. At much the same time, though, a well-connected doctor called Gesius, having been obliged to accept baptism, recited as he emerged from the font Homer’s verse,

Ajax utterly perished when he drank the briny water.

MacMullen does not seem to have put Gesius—whom he would find, I suspect, a kindred spirit—in his book, though I admit I may have missed a passing reference to him. The notes to Christianity and Paganism are deep pools of erudition, which scholars who read less assiduously than MacMullen will fish in for years to come.

These two varieties of deliberate and self-conscious preservation of the polytheist tradition are perhaps the only ones that merit being called “survivals.” The third variety includes polytheists who accepted baptism but failed to conform sufficiently to Christian norms, either out of ignorance or else because they succeeded too well in maintaining earlier social identities even within the bosom of the Church—or in some cases for both reasons. They quickly came to be regarded as “heretics” by other Christians. We know of one such group, which converted out of sheer terror just after the great earthquake that hit Beirut in AD 348. The Church historian Philostorgius records that

the larger part of the town collapsed, with the result that a crowd of pagans came into the church and professed Christianity just like us. But some of them then introduced innovations and left, stripping off as it were the conventions of the Church. Dedicating a place of prayer, they there received the crowd, and in all things imitated the Church, resembling us just as closely as the sect of Samaritans does the Jews, but living like pagans,

a criticism perhaps intended to apply to their morals as well as their manner of worship.


Fourth, there were the echoes of polytheism to which MacMullen devotes his final chapter, entitled “Assimilation.” Here we have to do not so much with conscious polytheist practices as with habits that were not given up largely because they were never questioned. They had probably not been thought about much in the first place, or at least had not been thought about for a very long time. These customs included, notably, burial practices, to which MacMullen devotes much space. Bishops ranted ceaselessly against members of their congregations who danced or got drunk at the graves of kin or the tombs of the martyrs. They saw this as recrudescent polytheism; but the dancing drunks may well have just been doing what they had always done.

Finally, it is worth mentioning an idea that became popular among ethnographers, especially in the nineteenth century in countries, like Greece, that were, after long centuries of Ottoman rule, trying to reinforce an ethnic identity and assert a national one. In such circles it was believed, all too easily and uncritically, that contemporary Thracian fire-walkers or Mytileniot bull-sacrificers were symptoms of an unbroken polytheistic, especially Dionysiac religion that Christianity had never rooted out. For example, a well-known classical scholar wrote in 1910 that “by their acceptance of Christianity they [the Greeks] have increased rather than diminished their number of gods; in their conception of them and attitude towards them they have made little advance since the Homeric Age.”3 By constructing an ethnic pedigree built on such foundations, some Greeks and their supporters hoped to legitimize their claim to nationhood in the eyes of European elites bred on Homer and on Gibbon’s unflattering opinion of the Eastern Roman Empire and Christianity. Greece is today a land where parents christen their babies Achilles, Agamemnon, or Aphrodite, and when they reach school age, take them to admire restored “monuments of the ancient Greeks” which for centuries served what is still the state religion as living places of worship. Although strictly speaking revivals, these practices carry an umistakable implication of survival.

Behind these various modes of polytheist persistence, there were widely fluctuating degrees of awareness, and differing intentions, on the part of people who in one way or another showed respect for the old gods. But what of the Christians? Were they as sensitive as scholars today may be to every last shade of difference or similarity between polytheists and Christians? And what other factors were there, environmental perhaps as well as human?

MacMullen is attracted by the idea that at least some worshipers of the Church’s saints could see no difference at all between them and the gods they supplanted. Of the doctor saints Cosmas and Damian he observes that they “had to explain to their worshipers in Constantinople that they were not, contrary to a wide misunderstanding, Polydeukes and Castor,” the heavenly brothers of polytheism. The story MacMullen refers to, though, concerns a single polytheist who, misled by certain of his fellow religionists into believing that Cosmas and Damian were in fact Castor and Pollux, went to seek their help at a time when he in no sense regarded himself as a Christian and one of the saints’ worshipers. He accepted baptism only when he realized his error, and after calculating that the saints were more likely than the gods to heal him. Of course, unthinking people entertained all sorts of strange notions about the gods and the saints; but it seems likely that thoughtful Christians, while seeing the sense of maintaining, say, a healing tradition where one already existed, would have taken for granted that the forms, the symbols, and the names had to be changed.

More commonly, MacMullen contents himself with emphasizing that “the flow from non-Christian into Christian usage was…unbroken.” He regrets our “lack [of] the same sort of revealing texts from individual worshipers of Christian times that have survived from the pagan,” while surmising that “there is no reason to suppose that much was lost, so to speak, in translation.” Despite the carefully inexplicit language, one senses a certain kinship on MacMullen’s part with the nationalistic ethnographers of the nineteenth century. Like them, MacMullen is preoccupied with particular men and women, individually or in groups, whom he sees as solely responsible for carrying on cultural traditions. In fact, he is so concerned with recapturing the texture of particular cases of Christians’ often stressful and confused contact—physical contact—with the remnants of a past that was still very much alive, that he never lets himself stand back to describe the larger setting. The natural environment has no part in the picture he composes. Each point is illustrated by a cluster of references, for instance a “random sample” of “archaeological data,” as he puts it, drawn indifferently from widely separated parts of both the Mediterranean and the non-Mediterranean regions of the Roman Empire.

The Mediterranean, in particular, should not be treated with such disregard for its variety, and for the patterns in which that variety arranges itself. This is a region in which a long sequence of religious transitions can be studied, from antiquity to the present day. Polytheism gave way to Judaism and Christianity; Judaism and Christianity retreated in turn before Islam. MacMullen’s chosen time span, AD 300 to 800, embraces the first two centuries of Islam—of which he says nothing, despite the fact that by the early eighth century the caliphate had absorbed the eastern, southern, and western shores of that once Roman lake, the Mediterranean.

This layering, century by century, millennium by millennium, of the region’s religious life, can still be traced in the landscape and read off its mute stones. The Parthenon in Athens, for example, once a temple of Athena, then a cathedral church in honor of Mary the Mother of God, and finally a mosque, has now been converted into an archaeologists’ showpiece. But Jews and Muslims still worship near each other, however unamicably, at the tombs of Abraham and Sarah at Hebron; Muslims still visit Saint Thecla’s monastery at Malula near Damascus, in search of healing.

Especially in Asia Minor, where a well-documented process of Christianization was followed, after half a millennium or more, by a well-documented process of Islamization under the aegis of Seljuk and then Ottoman Turks, it is possible to trace parallel phases of assimilation. Saints establish themselves next to gods and gradually eclipse them, inheriting their faithful and some of their specializations, as happened with Thecla and Apollo at Seleucia (Silifke) on Turkey’s southern shore. Later, Muslims seeking healing and enlightenment are attracted to the sanctity of churches and monasteries, and set up their own shrines close by, like the splendid fourteenth-century mosque of Isa Bey at Ephesus, constructed between the ruins of the basilica of Saint John the Evangelist and the site of the Temple of Artemis, one of the Wonders of the Ancient World. In other places, the older shrine is taken over by the new cult—Athens’s temples of Hephaestus and Asclepius as well as Athena are made into churches; the churches of Holy Wisdom in Thessalonica and Constantinople are reoriented, with the coming of the Turks, toward Mecca. Saint Nicholas’s church in Chania, Crete, to this day has both a bell tower and a minaret.

When we observe these waves of assimilation and try to define what they have in common, we are bound to wonder what other factors were in play, beyond the convenience or whim of individuals at particular historical conjunctures. Was it just a happy chance, as MacMullen appears to think, that Christianity emerged “wonderfully dynamic” and “enormously enhanced” from this phase during which it interacted with polytheism and assimilated some of its practices? What of the landscape itself? Did its features encourage a variety of mutual influences between Christians and polytheists? And were there certain basic religious needs that people felt had to be fulfilled by whatever religion came along? MacMullen complains that his fellow historians are happy to define Judaism or Christianity, but rarely say what they mean by “religion” itself. Even if we exclude Islam, as MacMullen does, the kinds of evidence he deals with may yield the beginnings of an answer to such questions.

First, the Mediterranean world (the heart of MacMullen’s concern) is one of great topographical variety within constricted space. Compare, for example, Lycia with the central Anatolian plateau, or Lebanon and Palestine with the Syrian and Jordanian steppe. The characteristic Greek city-state commanded a small, more or less fertile plain isolated from the next city’s territory by mountains that were hard to cross, especially in winter. Its awareness of its own communal identity, as well as its claims on resources coveted by others, encouraged attachment to its distinctive divine patrons.

This attachment was recognized by Constantine’s polytheist nephew, the emperor Julian, when, with an emperor’s topographical sweep and a philosopher’s sense of priorities, he proposed a pantheon of ethnic patron-gods—Ares for the Germans, Athena for the Greeks, and so forth—all under a “common father and king of all peoples.” Today, in Greece, one can see similar tendencies. Each town has its saintly protector and its patriotically observed annual festival. Mary the Mother of God has many different names—for example, “Swift to Hear,” “Golden Breasted,” “Consoler,” “Leader of Hosts.” Other names by which she is known indicate special attachment to a particular place, and reveal her worshipers’ sense that Mary responds to them more flexibly than does God the Father or God the Son.

The divine sanctioning of local identities, virtually impossible in those (mainly Protestant) forms of Christianity that refuse to give saints any prominence, is one structural resemblance—and no doubt, often, a form of continuity, too—between polytheism and Christianity. It points simultaneously to the influence of a distinctive landscape and to a shared appreciation of the divine world’s plural aspect (which is not, therefore, unique to polytheism).

Another resemblance between the functions of polytheism and Christianity can be found in the need people have felt for mediation between man and the divine, especially before the arrival of scriptural religions that provided a mediating text. And after that event, there were still regions where the scriptures were less well known than they have been in some Protestant circles of northwest Europe or North America. Christian folk were often deeply ignorant of their own scripture, whether in the Elizabethan England described in Keith Thomas’s Religion and the Decline of Magic (1971), or the Iberian lands described by George Borrow in The Bible in Spain (1842). When one must address oneself to the One God or to Christ King and Judge, a mediating saint can seem very necessary. There are consolations that fellow human beings, however heroic, however unsullied, can offer more easily than can the Godhead, especially when it is not a specific favor one wants, but just the sense of a presence—in a widow’s lonely bereavement perhaps, or a traveler’s terror of the absolute nighttime darkness our light-polluted age can no longer easily imagine. This is where the saints come truly into their own, and remind us of the gods who were once so close at hand.

Through these continuities between polytheism and Christianity in a particular landscape, and in the way both Christians and polytheists acquire a communal identity and try to satisfy a need for mediation, we come to perceive more clearly what lies beyond or behind MacMullen’s many individual stories. We also see the plural as well as the single—polytheistic as well as monotheistic—aspect of the divine world, whether viewed by polytheist, Christian, or, eventually, Muslim. Composed at the end of antiquity, the Koran rejects that plural aspect as it is manifested in Trinitarian Christianity and calls to a purer monotheism. So it was that the Turks and “Arabs” of Athens implored Allah for rain, while the Christians prayed, not directly to God, but to the Prophet Elijah. Yet Islam was to establish its own cults of saints. Christian writers were inclined to accept, if not always explicitly then at least through the paradoxes of their language, how mysterious and impenetrable to human intellect the relationship between the many and the one in divinity must be—nowhere more so than in the person of Jesus, in whom divine simplicity and the complexity of Man intersect in the incomprehensibility of the Incarnation.

The point of raising such questions about local identity and mediation between the believer and the divine is to suggest that, by standing back a little from the materials offered us in such abundance by MacMullen, one may be able to sketch something of what it means to be religious, rather than specifically Christian or “pagan” or Muslim. When the Turks and the Greeks and the “Arabs” each gathered—separately and exclusively—to pray for rain, it was readily if not explicitly understood that, in a matter of common interest, anybody’s prayers might conceivably be answered. When the “Arabs” finally did the trick, it was not their doctrine or their ethnicity that was given credit, but the simplicity, urgency, and faithfulness of their prayer.

There was, moreover, a continuity they were not aware of, any more than were the Turks, or even the Greeks. The ancient Athenians had reverenced Deucalion’s grave exactly where the “Arabs” now prayed. After escaping Noah-like from a great inundation, Deucalion had settled in Athens and built there the original temple of Olympian Zeus. Within its precinct, “the ground is cloven to a cubit’s width,” according to the second-century traveler Pausanias. “They say that after the deluge which happened in Deucalion’s time the water ran away down this cleft. Every year they throw into it wheaten meal kneaded with honey.”

This Issue

December 16, 1999