nothing of this experience, but the “entering” or “crossing of the threshold” was evidently an extra ceremony which only a few of the clients chose.²² On the evidence to date, those who “entered” were also among those who underwent mystery rites.

If others refused, they could be forgiven, for we have now learnt from the archaeologists what the word implied. By the light of torches, the prophet and thespode and perhaps the secretary stooped into one of the two low tunnels which ran underground to Apollo’s sacred spring. They bent themselves for a journey through a low, narrow corridor which was roofed in marble of a deep shade of midnight blue. The corridor ran for some thirty yards and changed direction seven times before it stopped at the door of two underground chambers. Here was Tacitus’s “cavern,” vaulted suitably in stone. The sides of the first room were fitted with stone benches and housed an “omphalos,” or navel stone, of deep blue marble, like the famous omphalos at Delphi. It signified to visitors that they had reached the oracular centre of the earth.

A narrow Corridor led from the first chamber to the second room, where the god kept his sacred spring. The spring survived to reward its French excavators only thirty years ago, for the water table is high at Claros, and its rise now hampers access to the tunnels. The prophet, we must assume, passed into this inner chamber. Iamblichus states clearly that the prophet, not the thespode, drank the water, and on this point, too, we must follow him.²³ He helps us to make sense of their relationship. The prophet had not eaten for a whole day, and was primed by his rites and his hours of isolation. Whenever he drinks the god’s water, says Iamblichus, he “is not in control of himself and does not follow what he is saying or where he is, so that he finds it hard to recover himself even after uttering his oracle.” Was this inspired utterance really cast immediately in neat iambic verse? Some of the surviving oracles are metrical tours de force and they make this notion impossible. There was, after all, a thespode. First came the incoherent sounds of inspiration, induced by the solemn occasion and the expectations which surrounded the sip of Apollo’s water. Then came a second, ordered voice, the voice of the thespode, or “singer of oracles,” who put into intricate verses the basic message which Apollo had inspired. The thespode had had a day or more in which to reflect and to listen to his questioners’ news. By divine insight, Apollo’s verses neatly matched the problem in hand.

Questioners who had stayed above ground heard these sounds at a distance as they echoed through the underground corridors of stone. If they were sitting in the antechamber, they had the thrill of a closer proximity. Perhaps the secretary sat with them on the benches, taking down the thespode’s version in the recently developed skill of shorthand before the words had slipped from human memory. Together again, the temple staff and their clients branched off down a second tunnel and turned seven times through a similar maze of midnight blue. Then they emerged into the sacred night, the blaze of torches and the lingering smoke of incense.

Such was the consultation which lay behind the words on Oenoanda’s wall. The “questioners,” surely, were people from Oenoanda, not distant questioners whose answer happened to be known to people inside the city. They had gone to Claros armed with an intriguing question: “What is the nature of God?” The prophet muttered, the thespode took up the challenge in verses of the best oracular theology which was known to his age. It is in oracles, too, that we can match it: “unshakeable” is a word known in another oracle’s text, while the nearest parallel to the first two lines are verses in a “Sibylline” oracle, which a Jewish author had composed in Greek, probably around the turn of our era.²⁴ The questioners returned with twenty-one lines of oracular wisdom, which they shortened into a manageable text. They then inscribed six lines where an arched doorway led back to a tower in their outer wall. It was not the neatest of inscriptions and its site was sheer and inaccessible. But it basked on its altar in the morning sun.

Did the city approve their despatch, and if so, when did it send them? The answers remain uncertain. The city’s wall was abandoned in favour of a smaller inner circuit and the date of this change is probably the 260s A.D. Did a private group of questioners inscribe the oracle in the 270s when the outer wall had gone out of public use?²⁵ It is notable, however, that no other text was cut into this wall’s facing and that the oracle was rather special. It was placed carefully for its relation to the rising sun. The shape of the lettering does not establish a firm date, but the years around 200 have been proposed for their form: in the 270s, the text would be a great rarity, one of the only inscriptions to survive from any Greek city in Asia at this period. The text’s language suits either date, but on balance, the earlier is preferable, when it fits with the well-known “epigraphic habit” of the age. Then, other cities displayed in public the results of their journeys to Claros. They put up Apollo’s texts on their public squares and temples or obeyed his orders and put the image of “archer Apollo” above their gates, like a holy icon, to protect them in times of plague.²⁶ Diogenes had already given a huge text of philosophy to Oenoanda: visitors to Apollo then brought the gods’ own view of the nature of God. An oracle raised the tone of the place, and although the inscription was quite modest, it is easier to suppose that the city’s authorities had approved the reworking of a single block, so carefully sited in their city’s outer wall. Later, Chromatis added her lamp below it and inscribed her small altar to the Most High god, in full awareness of the text above: perhaps she identified its “all-seeing god” with the Most High god whom she worshipped in a local pagan cult. By then, perhaps, the wall was no longer the city’s public defence, and she walked to the old perimeter to tend her lamp beside a former tower.

As a text for one city, the inscription is of great interest. Once again, it takes us beyond a city’s cults to beliefs which could animate pagans’ actions. People dreamed of the gods and sensed their presence, but over and above their cult acts, they could ply them with theological questions. In answer, they could receive a text of divine wisdom, echoing the views which their philosophers liked to teach. They could discuss it or contest it, feel proud of it, inscribe it and obey its clear commands. How unusual, though, was this type of advice?

By a happy chance, we can trace Apollo’s influence beyond the wall of this one remote city. In the second and the early third century, his shrine at Claros drew delegations from a far wider network who acquired a fortunate habit on arrival.²⁷ They arranged for their names to be inscribed and their visits dated wherever there was space on the shrine’s blocks of marble. During the second century, the site turned into an archive for any researcher on civic and religious life, for inscriptions ran everywhere, on the statue bases, on four of the steps which led up to the temple, on the very columns of the temple and its Doric façade. Some three hundred texts have now been recovered, and although they are not yet published in full, enough has appeared during the past seventy years to make Claros the centre on which views of pagan worship in the Greek East during the Antonine and the early Severan period can be firmly based. The earliest civic inscription which has so far been reported is a record of envoys from Perinthus on the Sea of Marmara whose visit falls late in the reign of Trajan, perhaps around 110 A.D. The texts tend to date themselves by the years of the shrine’s officers, but their system of dating has been unraveled and we now know that the surviving inscriptions extend to a date around 205 A.D.

The major cities of mainland Greece and old Ionia are almost entirely absent from the client cities who chose to inscribe their names.²⁸ These clients extended widely: west, in a year of crisis, to Stobi in Macedonia, south to certain cities inland in Caria, north to the northern coast of the Black Sea, quite often to Thrace, to the cities of Pontus and, here and there, to places in Phrygia and Cilicia. The delegations came mostly from the lesser or more recent centres of Greek city culture: Oenoanda was keeping appropriate company. Usually, they lacked an oracular cult of their own and any connection with Delphi or Miletus at the time of their foundation, and so they turned to Claros, rather than to these old, alternative shrines. In most years, cities sent envoys who inscribed their names and left us proof of their religious pilgrimage. The picture is even better than that.²⁹ The envoys often came with a choir from their city, a choir of boys or of boys and girls, so that we can watch the musical children of these cities coming again and again to hymn the god. A walk to Apollo’s great temple was a high point in childhood for these groups of singers. Sometimes, they were a batch of seven singers, or nine, or twelve, numbers which were pleasing to the god. Off they set from their city with their distinguished “sacred envoys,” their choirmasters or hymn teachers and their paidonomos (or “tutor”). Sometimes, too, they took their cities’ local poets, minor names in the history of Greek literature who had nonetheless won fame and prizes at the many contests and recitations in the Antonine age, unsuspected talents like Permissos, son of Nothippos, prize-winning poet and teacher of hymns by appointment “for life,” who visited Claros persistently in the mid-second century, a man with the name of a poetic river on the Muses’ mountain. His job passed to his son, Permissos, who reached his city’s council and in the 170s described himself as “poet extraordinary” from Laodicea on the Lycos.³⁰

Within these small choirs for Apollo, we can watch the intricate family relationships, as brothers and sisters travelled together: several choirboys were sons of prominent fathers who travelled, too, as the hymnodes or city delegates. Choir service ran proudly in these families who passed it from father to son, and included men of great position in their home cities. Often, the trip to Claros was a long, lonely walk of many hours’ duration through mountainous country. More distant clients sent envoys, but not a choir, and in the nearer delegations, the fathers were probably glad to be travelling, too, and seeing that their children came to no harm. The “tutors” and attendants would be men over forty years old, who were supposed to know how to control themselves on a lengthy walk with choirboys.

We can also watch the careers of the children who attended Claros’s shrine, beginning as choirboys, then passing on with age and occasionally returning later as choir leaders and delegates. Some of them came again and again, fifteen, twenty-nine, even thirty-three times, according to the inscriptions of two modest Carian cities, from which two fathers came to Claros at least twenty-nine times, their sons at least fifteen times, making their families the best-attested pilgrims in the pagan world. Hymns and oracles were a way of life for these people, men who led their delegations on a long annual walk and needed no “higher” religion to satisfy their needs. Some cities chose a “priest of the children” from among their eligible youth, and the choirboys of our later inscriptions grew into the civic notables of the years of Gordian’s reign: they would not readily desert a past in which they had walked to Apollo, shuffled, cleared their throats and sung him the hymns of their local poet.

Once, we can watch a former choirboy’s progress from Claros to the very edges of the Empire. During the reign of Hadrian, T. Statilius Solon visited Apollo in the choir of his Carian city, Heraclea-by-Salbace, a place which was a keen, almost a yearly, client of Claros. Afterwards, T. Statilius Solon passed into the ranks of the Roman army, and almost certainly he is the same Statilius Solon whom we can detect as a centurion in an army inscription on Hadrian’s Wall in Britain. He is obviously the Statilius Solon who honoured a dead friend in a Greek and Latin inscription from Brigetio on the Danube. He had carried his memories of Apollo from Claros to the borders of Scotland and the banks of the Danube River.³¹ We must allow for the presence of these Clarian “old boys” when we find their Apollo being consulted by distant cities or by an army unit on Hadrian’s Wall. At one stage in life, they had walked and sung: they knew the words of Apollo’s hymns and grew up trusting the shrine where once they had honoured the god.

Apollo at Claros was consulted not only by civic delegations.³² We meet individual questioners in our texts, a nephew of the Emperor, a Cynic philosopher, a merchant from Pontus, the orator Aelius’s helper and, in fiction, the parents of two young lovers, puzzled by the causes of their apparent “sickness”: the Clarian god’s oracle is central to the plot of a novel by Xenophon, composed c. 150–250. A certain Symmachus and his sons consulted Apollo about the health of their crops and inscribed his answer at Yaliniz-Serai, in northeastern Phrygia.³³ Perhaps Claros’s Apollo can now be detected farther south. Just to the west of the Euphrates, in the Kurdish mountains, his advice may have reached a valley of ancient royal monuments. A steep hill overlooks the shrines of the kings of Commagene and a fine Roman bridge: on its slope, one Candidus re-erected and inscribed an “age-old altar” on the advice of “Apollo’s immortal prophecies” and set up a statue of Zeus the king.³⁴ Candidus was a victorious Roman general who served in the Eastern campaigns of 195 A.D., and Apollo, perhaps, was the Clarian, whose advice was so often sought on statues and restorations elsewhere. If so, the site is the wildest and most spectacular of any which has preserved his advice.

At Claros itself, no texts of Apollo’s oracles have been found. We know, however, that there was a record office, and in the burial ground of nearby Notium, we have the epitaph of one Gorgos, honouring him as “elderly and very bookish” and referring to him “culling” the “page of the singers.” The text, datable c. 150–110 B.C., has been well explained as a tribute to a prophet at Claros who had gathered up the “pages” of previous oracular poets at the shrine.³⁵ They were surely kept on the site in some perishable form: in the second or third century A.D., one Cornelius Labeo could quote an oracle’s text in his book On the Oracle of Clarian Apollo.³⁶ He can hardly have toured the client cities in order to find his verses, although it is in the client cities that we still find them on stone, sometimes with their questioner’s name before Apollo’s verses. Durable inscribed texts from Claros derive from the client cities, not the shrine itself: the words on Oenoanda’s wall are true to this epigraphic pattern.

While Christians travelled to the Holy Land and marvelled at God’s wrath against the Jews,³⁷ pagan choirs were travelling yearly to Claros, to sing and to see their delegates “enter” the temple tunnels. However, the ceremonies at Claros were not unique. There were other famous oracles in Asia, but none larger or more famous than the temple of Didyma, which stood outside Miletus, city of the visiting gods. This gigantic shrine had recovered from the disorders of the late Republic and had promptly started a pattern of inscriptions which forms a pair with those at Claros. Not only did the god’s questioners inscribe some of their questions and answers in the precinct and the adjacent territory of the city. Since the early 30s B.C., Apollo’s prophets took to inscribing their names and careers on a newly constructed Doric building which stood in Didyma’s sanctuary.³⁸ Like the visitors to Claros, they cut their names all over its surface, on the walls, the columns and its decorations. While Claros flourished, Didyma persisted, two seats of Apollo whose stones have left a sample of the language of the gods.

We know of no rivalry between the two shrines, but there were certain differences. At Claros, Apollo was made visible in a huge, half-naked statue of late Hellenistic style, urbane, relaxed and at ease with his lyre. At Didyma, he was represented in the most widely travelled holy image of Greek history.³⁹ Back in the archaic age, shortly before 500 B.C., the sculptor Kanachos had cast an Apollo in bronze for the shrine. Naked, the god held a bow in his left hand and a stag in his right, which could be moved by an artful mechanism. The god’s hair and expression had the stiffness of the pre-classical age, and at Didyma, this image had a special veneer of history. In 494 B.C., the Persians had removed it to Susa after their sack of the city: in the wake of Alexander’s conquests, it was kindly returned by King Seleucus, Apollo’s protégé. Five centuries later, in the Imperial period, it still stood at Didyma’s altar beside the temple, an archaic image which had the venerable awe of a travel to Persia and gave a curious, old-fashioned shape to Apollo’s form.

At the two oracles, the patrons of the surviving inscribed texts also vary. At Claros, they are the delegations of cities. At Didyma, in the Imperial period, cities did continue to consult the god,⁴⁰ but all but one of the inscribed oracular texts are answers to individuals, most of whom are known as Milesian priests or members of the prophets’ families. Not all were Milesians by origin. One of the prophets had come from Cyzicus, a city which had consulted the god, while a curious inscription refers to another prophet who was “summoned by the god” and promised and performed many generous duties for Miletus and its cults.⁴¹ The man was a doctor, Pomponius Pollio, a Roman kinsman, evidently, of two namesakes who were governors of Asia in 151/2 and 167/8. How did the god’s “summons” come about? The first move, presumably, was made by Pollio, one of several Roman senators at this period who were drawn to the oracles of Greek Asia. In return for his office, he “promised” valuable work on the site’s buildings and Sacred Way: his offer, perhaps, was referred to Apollo for approval, and so the god “called” this stranger to his service. The duties were very expensive, and at Didyma, too, we find the familiar compound of “volunteering” and a shortage of candidates, attested intermittently in the second and third centuries. Some prophets, therefore, were particularly proud of their service and generosity: their exceptional concern was sometimes approved by Apollo himself, encouraging them all the more to inscribe their names and honours on the site.

Apollo’s language at Didyma did not differ noticeably from his language at Claros. At Didyma, he gave a prominent benefactor of the province of Asia some solemn phrases on holy Dawn and the Supreme god, calling him “leader of Ethereal Fire.”⁴² The text belonged in or before the 140s A.D., and a few decades later, Claros gave the Oenoandans verses which touched on similar themes. The shrines’ agreement was not surprising. Although Miletus no longer staffed Claros, the thespodes and prophets of the two shrines shared the same education. They tended, however, to prefer different metres. At Didyma, almost all Apollo’s surviving answers are cast in hexameters. At Claros, he also used hexameters, but sometimes he broke into iambics or more complex metrical forms.⁴³

The method of consultation also differed. At Didyma, the god inspired a woman, his prophetess, but she was attended by a prophet, whose career usually included civic service in Miletus. Once again, we owe our best description of a consultation to Iamblichus, whose words must be pictured against the ruined site. Questioners could not intrude on Apollo whenever they wished. In the second century A.D., they travelled down the paved Sacred Way from nearby Miletus or arrived on the nearby shore by boat. Before them stood the enormous oracular temple of Didyma, whose thick columns and staircases still amaze us and whose vast inner courtyard defied every plan for its completion. While the questioners waited and sacrificed, they could lodge in the housing which we know to have spread on the second-century precinct, or paradeisos, as it was still touchingly described, in a word of Greco-Persian origin.⁴⁴ The prophetess followed “a rule of complete purity,” presumably sexual purity. She bathed, and as at Claros, so at Didyma she prepared for the god by eating nothing. She fasted not for one day, but three, and “a multitude of sacrifices” preceded her work: famished, she lived meanwhile in the inner shrine and “was already possessed by the divine light, enjoying it for a long while,” at least in Iamblichus’s view.

The method of consultation is uncertain, as the temple’s ground plan is ambiguous. Two staircases lead down into the huge inner court and two staircases lead up onto the roof. A large opening, like a window, stands above ground level and gives onto the fore-temple: perhaps the questioners received their answers here. Apollo delighted in choirs and singing, and at Didyma, as we shall see, he approved them in an oracle of mid-third-century date: here, too, we must imagine choirboys accompanying his consultation.⁴⁵ Perhaps the questioners waited outside in the fore-temple while the prophet and other officials went down the vaulted inner staircases to meet the prophetess in the enormous area of the unroofed inner shrine.⁴⁶ Iamblichus lists various ways in which the prophetess might be inspired, and these ways are not alternatives, but differing aspects of a single ceremony. The priestess held a rod “handed down from the god,” like the rod which Apollo was said to have cut from a bay tree and given to the legendary founder of Didyma’s priesthood. She sat on an axon, a fascinating word which ought to mean a rotating, cylindrical block: in archaic Greek, an axon pointed upwards vertically, like a post, but the priestess must have benefited from a slight shift in its subsequent meaning. The axon was set beside the small sacred spring which welled up at the rear of the inner court. The water wetted her feet or the hem of her prophetic robe and gave off a vapour which she “breathed” when she received the god. Unlike the prophet at Claros, she is not said to have drunk it.

The prophetess made noises of inspiration, and the prophet stood by to turn them into verse. He changed yearly and was perhaps assisted in his poetry by one of the secretaries or temple staff: here, we can only guess. The prophetess needed time to recover after her long fast and her solemn contact with the god, and presumably she stayed in the inner shrine, lost in the anonymity which still surrounds her. Only once do we know a prophetess’s name, and then, c. 200 A.D., she belongs to a very well-born family.⁴⁷ If she is a typical case, the job was not held by simple, unlettered persons: perhaps the prophetess added more to our surviving oracle texts than we imagine. The prophet’s, however, is the name which we know most often. He returned with his companions up the vaulted passageway of stairs: did he then climb higher, ascending the upper staircases to the roof space above the forecourt? Alternatively, he appeared at the huge window which the doors had closed from the forecourt’s view. The god’s response was announced with due ceremonial at one or other site, and copies were made available for questioners at the nearby record house.

At Claros, the prophet tunnelled far below the ground, while at Didyma, the prophetess sat on the axon and made contact with a surface spring.⁴⁸ Both shrines focussed on water in their differing ways, and both set a problem to Greek theorists who wished to explain why prophecy and inspiration were general forces but were connected with these particular local pools. How, from contact with Apollo’s water, did texts of such lofty theology emerge? For an answer, we can look to a third site, Delphi, which is lit up for us by three dialogues by Plutarch, set at dates between the mid-60s and c. 120 A.D. They show us better than any other texts what company Apollo had begun to keep.⁴⁹ In the Imperial period, the site of Delphi was frequented by teachers and men of letters, travellers, philosophers and those eternal funds of vivid and useless information, the Greek guides. Through their persons, Plutarch’s Delphic dialogues allowed a hearing to many philosophic views on the value and nature of oracles. He introduced a Cynic and an Epicurean who dismissed them as frauds,⁵⁰ a Stoic who defended them too uncritically and then, at the end, a speaker with Platonist sympathies who settled the topic in terms with which Plutarch would have sympathized. His characters were a wide-ranging company, from whom Delphi heard many cosmopolitan stories, the tales of people who had travelled in Egypt or visited the distant Siwah oasis, the views of an Alexandrian philosopher on a tour of Greece, the reports, even, of a tour in Britain and the Scottish islands round Mull. Delphi in the 80s had heard from Scribonius Demetrius, a schoolteacher who had travelled to these places, although his home lay in distant Tarsus. He told stories of the old god Cronos, who was said to be chained on a Scottish island: other islands, he said, were haunted by the spirits of great men whose deaths caused appalling storms at sea. We know Demetrius for his dedications to Ocean and a sea goddess which were found on two bronze tablets while the railway station was being built at York; Plutarch knew him for his learning and remote tales.⁵¹ These visitors gave Apollo access to the latest modern geography, while others brought philosophy and speculative theology. Plutarch’s Delphi gave scope to his friend Serapion, a man whom we know as a Stoic and a poet and as the author of advice on the proper ethics for doctors.⁵² Plutarch dedicated a Delphic book to him, as he dedicated another on the gods Isis and Osiris to a Delphic priestess called Clea. They give us a priceless glimpse of the atmosphere which hung round Delphi and help us to understand the tone which Apollo assumed elsewhere.

Clea is a woman who stands for much we might otherwise miss in evidence based on the masculine world.⁵³ Her parents had been friendly with Plutarch and he greatly respected her learning and knowledge. She was an informed worshipper of Isis and a reader who could be interested in Pythagoras’s doings in Egypt. Plutarch’s book to her has been classed as one of the most difficult of surviving works in Greek, yet Clea was thought able to cope with it, a woman, said Plutarch, who must not take the stories of Egypt’s gods too literally. A second work in her honour, The Brave Deeds of Women, was not so heavy. Plutarch and Clea had been lamenting a female friend’s death, and Plutarch had started to tell how the virtues of men and women were identical. He finished off his pleasant conversation by sending her stories in which women showed men’s courage. “Concerning the virtues of woman, Clea, I do not hold the opinions of Thucydides.” He did not cite the usual examples of feminine courage; these, he said, were “well known to a woman who keeps company so solidly with books.” Instead, he sent her some curiosities, as befitted a lady of letters who was familiar with Egyptian theology, books on the nature of the gods and the masculine virtues of Greek and Roman members of her sex.

Lesser versions of Clea, one feels, were among the society women to whom Christianity appealed. It, too, had a theology, tales of female heroism and scope for female endeavour, albeit of a sexless variety. Plutarch’s other Delphic tract was addressed to Serapion, a Stoic, and discussed the difficult meaning of the letter “E” which was inscribed so prominently at Delphi. Plutarch cast the text as his memories of a discussion which had been held at the shrine in his youth, in the mid-60s A.D., when Nero was roaming in Greece. His text is a telling companion to the Apollo who spoke to Oenoanda, and Plutarch arranged his dialogue with skill.⁵⁴ He recorded wryly his own youthful views on the enigmatic “E,” his passion for mathematics and number symbolism, and by contrast, the dry, logical explanations of his elder, the logician Theon. The climax was a speech by Plutarch’s own teacher, Ammonius, who had travelled to Delphi from Athens and brought a higher philosophy to problems posed by Apollo.

This speech is preceded by some revealing details. We hear how residents of Delphi and a priest at the shrine attacked the more fanciful theories of the meaning of their puzzling “E.” A “Chaldaean,” Plutarch hinted, had recently explained it by a theory of the planets: when we find astrology and favourable words for “Chaldaean” wisdom in Apollo’s answers, we must allow for the presence of like-minded visitors within earshot of the god.⁵⁵ Plutarch also alludes to tales of the theologoi which anticipate much in our subsequent oracles. These self-appointed spokesmen called Apollo “indestructible and eternal” and enlarged, too, on his transformations into a fiery substance and the form of the god Dionysus. When we find an Apollo describing his identity with the Sun, Dionysus and the Seasons, we must remember the loose theology of these speakers at his shrines.⁵⁶ They raised these questions for visitors but did not answer them exactly. Visitors then put the problems to the god.

When Ammonius ended the debate, his speech was already concerned with the central question which was to be put to so many later oracles: what is the relationship between the Supreme god Apollo and the others? His answer, perhaps, was individual, but already in the mid-60s, it touched on much which we later hear from the gods themselves.⁵⁷ The Supreme divinity, he said, is one, “eternal, uncreated, undying,” “immovable, timeless, undeviating”: Apollo, for Ammonius, is this Supreme god, higher even than the Sun, with which many confuse him. A lesser power oversees the world and the multiple forms of the gods, but we must praise Apollo the Supreme god with the words “thou art,” while honouring his constant advice to “know ourselves,” our mortal nature and frailties. This Platonist philosopher was a younger contemporary of the Alexandrian Jew Philo, in whose works we find a fascinating match for this elevated way of thought. It had been current among teachers in Philo’s Alexandria: in one late text, Ammonius is said to have come “from Egypt” to Athens, where he taught and held high office. Perhaps he, too, had studied in Alexandria and learned these views there.

The argument is less important than its tone. Here at Delphi were clever men disputing on the “meaning” of an old, meaningless symbol. It was a totally misguided effort, but they revelled in it, airing every option, astrology and number symbolism, logic and a lofty Platonism on the relation of God to Apollo and his fellow divinities. Their belief in the significance of puns and wordplay was applied to the names of the gods, and Plutarch wrote up their discussion in his later years, claiming to have remembered it precisely: Apollo, as we shall see, was equally retentive. In the 60s, said Plutarch, Ammonius seemed to us all to have “proved correctly that the god Apollo is no less a philosopher than a prophet.”⁵⁸ Subsequent oracles all over the Greek-speaking world are a gloss on Ammonius’s proof.

When we next meet the Delphic oracle in literature, the atmosphere around it is hardly less elevated. It was probably in the early third century that Heliodorus gave a brilliant picture of Delphi in his fictional Ethiopic Tales, an aspect of his novel which has been misread as “Delphic propaganda.”⁵⁹ When the Egyptian priest Calasiris arrived at Delphi, he was said to have been hailed spontaneously by the god as his “friend,” an honour which had attached to legendary visitors in the archaic Greek past. When news of its repetition spread, Calasiris was plied with honours and stayed at Delphi to examine life at Apollo’s shrine. Sometimes he busied himself with the “many varied sacrifices which were offered to Apollo by visitors and locals alike,” and sometimes, like Plutarch, he conversed with the philosophers “who flocked to the temple as if to a true temple of the Muses.” Their conversation had not changed, although a century or so had passed since Plutarch’s books. “They began by asking me questions on the gods and Egyptian cults… they wanted explanations of the Pyramids and the Labyrinth. They did not forget any one of the oddities of Egypt; discussing and listening to Egyptian matters is what attracts Greeks most.” Clea would have been in her element.

This impression of Delphi is not pure fiction: we know of one man from Delphi, in the Imperial age, who did visit Egypt’s Valley of the Kings and inscribed his name on the site.⁶⁰ Back home, no doubt, his stories interested Apollo’s shrine. An elevated philosophy also distinguishes the most famous Delphic oracle to have survived from the later third century, a poem which caught the imagination of W. B. Yeats. When the Platonist philosopher Plotinus died in the year 270, the most pious and diligent of his pupils sent to Delphi to ask about the destination of his soul. Apollo excelled himself.⁶¹ Long ago he was believed to have hailed Socrates as the “wisest of men” and his favour for Platonism had not been lost on its exponents. In the mid-second century, we find the head of the Platonic school attending the games at Delphi, and an inscription on stone lists the Delphians’ honours for a cluster of Platonist philosophers who extend across the generations in the second-century schools. With time, names were added to this list, and they suffice to prove that Platonist philosophy was still drawn closely to the shrine.⁶² Did these Platonists serve in some capacity the god who was thought to have praised their master? A eulogy of Plotinus was a natural sequel to this prolonged relationship, and so eloquent is Apollo’s long text that we can only regret the loss of most of his Delphic answers in the previous century and a half. Subtle echoes of Homer kept company with a learned allegory of the soul, based on a scene in the Odyssey: Plotinus himself had used an allegory from the Odyssey in a similar context, and imagery from the Odyssey found a home in the books and funerary art of subsequent pagans and Christians.⁶³ Plotinus, said the god, had enjoyed the gift of divine light. He had aspired beyond this “blood-devouring life” to the “blessed sights” which his soul enjoyed. Buffeted by the seas of life, it had returned to the heaven of Plato and Minos “and all the choir of Love.” This remarkable oracular poem was not a fake, as the scholarly Porphyry cited it with great respect. Had one of Apollo’s prophets attended Plotinus’s classes? The god approved the philosopher whose metaphors he duly borrowed.

At Claros and Didyma, the tone was not different. People travelled no less far to consult Claros, and although we know nothing of the thespodes’ education, some earlier priests and personnel are suggestive.⁶⁴ In the Hellenistic age, Claros was the home of the poet Nicander, who had probably served Apollo’s shrine and was credited with books on “all oracles.” His surviving poems are among the most contorted in Greek literature, but oracles would welcome his “combination of a repulsive style with considerable metrical accomplishment.” The “very bookish” Gorgo was a match for him, a “painstaking” researcher into old oracular poems and a “lover of wisdom.” At Didyma, in the 60s A.D., the prophet Damas had shown great zeal in restoring the old traditional rites in his city; in the mid-third century, another prophet, Ulpianos, was praised by Apollo for attending to the “old sayings,” presumably the texts of an earlier age. In the second century, another Milesian prophet styled himself “successor of Plato”: he would have revelled in Plotinus’s Delphic obituary. Men of such education ensured that Apollo kept abreast of the schools: we will have to ponder and place a remarkable text from Didyma in the third century in which Apollo decried colossal statues and lavish sacrifices because the “gods are not in need of possessions.”⁶⁵ Ultimately, this statement owes a debt to the philosophers’ theology, as do texts in which Apollo used puns and clever wordplay on the “meanings” of the gods’ own names.

Like many shrines, oracles had libraries and archives, while attracting talk of local history and natural wonders from their many clients and travelling visitors. They gave scope to one of the most persistent and distinctive types in Mediterranean civic life: the antiquarian. Apollo, as we shall see, could speak with rare learning, and we can detect a pleasant, close relation between the minor names of learning and the gods. They were the poets who wrote the hymns for Claros: at Ostia, near Rome, inscriptions show us the demigods Castor and Pollux answering three questions about one Septimius Nestor in elegant verse.⁶⁶ “How well has Nestor sung the story of your birth?” “How long will his literary fame endure?” “To which god should his statue be dedicated?” Nestor, from Greek Asia Minor, was a man of letters who enjoyed a wide following in the early third century A.D. He wrote erudite poems on curious stories, odd facts about farming and the habits of hyenas. He rewrote the entire Iliad, avoiding in each “book” the particular letter by which it was numbered. The statue of Nestor has been found, but unfortunately it lacks its head.

The clients of oracles were people who wished to know and argue, to be reassured or guided through their many choices of thought and action. Such a public ought to have interested Artemidorus while he compiled his book on dreams. Apollo, he wrote, had “stood beside” him “very clearly” and had inspired his work. Artemidorus had toured the games and festivals of Mediterranean cities: surely he stopped at Claros, so close to his “second city,” Ephesus? He had personal contacts among the choirs of at least one client city: at Claros, he could revel in “ce public d’informateurs de choix… cette affluence de dévots inquiets et confiants… Quel milieu idéal au point de vue psychologique.”⁶⁷ The gods’ horizons had grown with their prophets’ education, as Philostratus, too, had recognized: he shows the point clearly when he imagines how the pagan wise man Apollonius once forced his way into the oracular shrine of Trophonius in central Greece.⁶⁸ “I wish to descend,” Apollonius said, “on behalf of philosophy.” Dressed in his philosopher’s cloak, he descended into the oracular cave and reappeared “seven days later” at Aulis with a complete book, filled with the questions which he had asked. They concerned one issue: “What, O Trophonius, do you consider to be the most complete and pure philosophy?” Trophonius was said to be delighted with the style of this consultation, and Apollonius came up from the cave with a volume which was “still preserved” at Antium on the coast of Italy. Philostratus said that the people who lived around the oracle had told him the details of this episode, and the book, he believed, had been sent to the Emperor Hadrian, that connoisseur and patron of oracles and oddities. It had been shelved in his palace, he said, and still drew many tourists in the early third century. It befitted the tone of contemporary Didyma and Claros, for “it contained the views of Pythagoras,” wrote Philostratus, “since the oracle was in agreement with this type of wisdom.” This fascinating relic has not had the attention it deserves, although it summed up one aspect of oracular culture and thus attracted visitors for so long.

“What is God?” asked visitors to Claros from Oenoanda: their question was not so unusual for an Apollo who consorted with philosophy, and we can match it to texts which survive from the shrine. At Claros, Apollo once spoke on the nature of the Supreme god Iao: “in winter,” said Apollo, “he is Hades; Zeus, when spring begins; the Sun, in summer; in autumn, delicate Iacchus.” At Didyma, Apollo called the highest god Aion, or Eternity, a versatile concept which was also connected with Ethereal Fire and this same idea of a cycle of the Seasons: again, Apollo was reflecting a theological fashion, for the concept Aion enjoyed a greater prominence in the art and theology of the Imperial period.⁶⁹ Dawn and an east-facing aspect were also proven tastes of the god: Apollo at Claros told Symmachus and his sons to set up his altar, facing east to the sun. It is probably to Didyma that we should trace a fragmentary request and oracular answer which were discovered at Ephesus.⁷⁰ They concerned the siting of a statue of Eros and an altar to the great gods and they also spoke of “life-enhancing Dawn.” The altar on Oenoanda’s wall was true to this wider pattern of advice.

For further questions of this type, we have only to turn to Christians who seized on these oracles of higher theology and claimed that they proved the truth of their own. If they needed a little improvement, Christians were well able to give it. Christian authors gutted the pagan oracular books and saved the better pieces as Christian proof texts which improved, like wine, with the keeping. Yet, as they were verse texts, their metre set limits to their rewriting, and on the whole, their fragments have survived without too much alteration. Some of the fifth century’s most turgid Christian sermons come to life when they cite an oracular gem which plainly goes back to a pagan Apollo. The richest Apolline collection survives in our shortened version of the work On True Belief. Shortly before 500 A.D., its unknown Christian author added four books of pagan oracles and prophecies to the previous seven in which he had set out Christian doctrine. They were only a pendant to his exposition of the truth, but this lowly role allowed some intriguing Apollos to have their say.⁷¹ Old Apollos and pagan divinities are still trying to speak to us in many of the extracts.

The questioners from Oenoanda find their closest company in the pagan questions and answers which our epitome of this Christian book has preserved.⁷² “When somebody asked Apollo if only the Most High is without beginning or end, he replied as follows…” “When somebody asked when he should worship the Ineffable god, Apollo showed that every place is in his power and that he receives worshippers everywhere. His reply was…” “When somebody asked if he could come near to God by careful attention to his life, Apollo answered…” In this oracle, Apollo denied that his questioner could attain “such a godlike privilege,” as it was reserved for the Egyptian Hermes, “Moses of the Hebrews” and the “wise man of the Mazacenes,” none other than Apollonius the pagan philosopher. This reply is not a Christian forgery, as it omits Jesus from its list. Like his prophets, Apollo respected the mirage of Eastern wisdom, the godlike figures of the past and the legendary Apollonius, a pleasant insight into the range of a third-century prophet’s reading.

These oracular answers have a language and balanced phrasing which we can match with other pagan oracles and hymns. Although they are preserved in Christian handbooks, the majority are obviously genuine. On True Belief quoted a splendid text which had stood in the second book of Porphyry’s Philosophy from Oracles and existed, then, by c. 290 A.D.⁷³ Its wording makes Oenoanda’s answer seem rather prosaic, as it distinguished three grades of pagan “angels” at different removes from the pagan Highest god, “the all-powerful, most Kingly and sole Father of mortals and blessed Immortals.” After a while, the theology of higher divinities tends to repeat itself, whatever the religion. Apollo’s words could almost have served as Christian hymns.

The point was not lost on Christian preachers. There has been little interest in a mid-fifth-century sermon “On the Trinity,” wrongly ascribed to Didymus the Blind, but like Books 9 and 10 of On True Belief it quotes in profusion the words of pagan gods, many of which are impeccable in metre and vocabulary.⁷⁴ “Immortal God, supreme, dwelling in Ether, Imperishable, Unshakeable, Eternal, always the same”: these quotations and several others make Oenoanda’s answer seem really quite conventional.⁷⁵ It is clear that questions on God were not the curiosity of one small Lycian city and that Oenoanda was one among several inquirers.

Can we pin any of the texts in Christian sources to an original date and place? A brilliant series of conjectures has shown how excellent were the questions and answers which the author of On True Belief preserved. We only have a brief epitome of his book, but three of its quotations preserve the questioner’s name, and in one case, we can add another from a separate Christian source. “A certain Poplas asked if it was proper to send to the Emperor about monies for a public show…”⁷⁶ The unusual name Poplas has been traced to Aelianus Poplas, prophet of Apollo at Didyma around the year 220. He had served as a priest of the Imperial cult and had asked his god whether he should go on an embassy to request the Emperor’s financial support. He had held every sort of civic magistracy in Miletus and had volunteered for the job of archiprytanis, but nonetheless, he turned to Apollo to guide his ambition. A second oracle shows him at a lower ebb. His property had begun to diminish and his health was failing, so he asked Apollo who could help him. Poplas’s two questions remind us of the costs of civic munificence, the ups and downs of the highest careers, the role of the Emperor and the problems of paying the expenses of shows in the Imperial cult. His family intertwined with Miletus’s great prophetic dynasties, and Poplas’s nephew survived to serve the god. Poplas the prophet suspected no fraud and practised no deception, but when life was hard, he turned to Apollo for advice and ensured that the answers survived for posterity.

“A certain Stratonicus saw a dream and asked Apollo if he should believe it.” The dream concerned Stratonicus’s length of life, a type of dream which Apollo once sent to the orator Aelius. The name Stratonicus is very much more common, but a Stratonicus is known as a prophet at Didyma in the 30s B.C., while another turns up there c. 180 A.D., although he is not known to have been a prophet.⁷⁷ “When somebody asked Apollo about the fate of the soul after death…”: the same oracle was quoted by Lactantius the Christian, who traced the question to “Polites the Milesian.” We know just enough about the top men in Miletus to prove him right.⁷⁸ A Polites had dined there with Mark Antony in the 30s B.C., but our man is surely the later Polites whom we know to have visited Rome in the year 177 A.D. Like Poplas, he was well aware of the hazards of an embassy to the Emperor. He, too, was a man of philotimia who went to plead on his city’s behalf, “asking” permission for their new festival, the “Didymeia Commodeia,” whose games honoured the Emperor Commodus. We also find this Polites’s name on his issues of Miletus’s coins, among whose types stands the image of Apollo Delphinios, implying that Polites had served as his “crown-bearer.” Polites, then, had been honoured on Apollo’s behalf, and when he wanted to know about life after death, he turned to the god in his shrine at Didyma. The prophet that year had a philosophic schooling, and so, therefore, did the god’s answer. “While the soul is still in the body,” said Apollo, “it tolerates the pains which cannot hurt it. When the body fades and dies, the soul ranges free through the air, ageless, forever unwearied.”

This late Christian handbook has preserved four oracles whose questioners can be connected with prominent Milesians and prophets of Apollo at Didyma. When it names a fifth, the chances are high that he should be found at Didyma too: questioners at Didyma inscribed their names and questions on the site or in nearby Miletus, and thence, they could pass easily into the books of Porphyry and others. “Theophilus,” says our Christian epitome, “asked Apollo, ‘Are you God, or is somebody else?’” There is no shortage of men called Theophilus in the inscriptions of Greek cities, but the named questioners in this collection have been men from Didyma, and there, sure enough, we find Julius Theophilus, son of a prophet of Apollo and himself a priest of the Emperors.⁷⁹ His career extended after 215, to judge from his service in the city’s festivals. He was the contemporary of Poplas and the near-contemporary of Polites, and belonged with other third-century Milesians who referred points of religion to their god.⁸⁰

Not only were Oenoanda’s question and answer not untypical: the questioners now turn out to have moved in impeccable company. In the years between c. 170 and 230 A.D., prophets and civic notables at Didyma were asking similar questions. The tone of these higher questions must not be misjudged, for they were not necessarily “anxious,” like the “troubled” practical questions of Poplas. They could also arise from curiosity and a continuing interest in theology among people who had studied philosophy and knew that the philosophers had raised more questions than they could ever hope to answer.⁸¹ The natural counterpart to their inquiries is the young Galen, whose father had sent him round each of the philosophic schools in the mid-second century, but had asked him not to commit himself too hastily to any one view. Oracles could settle these doubts and contradictions.⁸² The opening scenes of that fascinating Christian fiction, The Recognitions, show how the young pagan Clement had been worried by the problem of the fate of the soul after death. After a vain search for the truth, he was said to have met St. Peter and come to rest as a Christian. The main core of The Recognitions was probably written c. 200 A.D.; a few years earlier, Polites had asked his Apollo the same question and found the answer directly from a god. Apollo, however, could not be pressed too hard. We should possibly attach to Theophilus’s answer several lines in which the god admitted that he knew only some of the truths about the highest heaven.⁸³ The Supreme god, he said, lived in isolation, and Apollo could only rely on hearsay about his nature. Although he knew more than he was willing to reveal, he told his questioner to stop bothering him, “asking what is not right and meddling in higher things for the sake of your own wisdom.” When Apollo was at a loss for an answer, he could still put his clients in their place. At Claros, too, his tone was similar, and the style was copied cleverly by a novelist.⁸⁴

How many people, however, does this evidence concern? Priests and prophets are the very people whom we would expect to hold fast to oracles and honour their god by inscribing his texts. Are we only studying a few untypical enthusiasts? It seems not, for Apollo spoke to a much wider audience, as we continue to learn from the Oenoandan text and another type of inscription in his clients’ home cities.⁸⁵ We have a widespread group of inscribed Latin dedications which follow a similar formula, though some of them misspell it: “to the gods and goddesses,” they say, “according to the ‘interpretation’ of Apollo at Claros.” Their spread is remarkable. They have turned up near Nora in Sardinia, where the text is now kept in a small village church, in cities in North Africa, re-used in the wall of a church in Dalmatia and far away on the Empire’s extremities, in Housesteads’ fort on Hadrian’s Wall, where the text was put up by a Tungrian cohort. This British text, at least, must be dated after 140 A.D., because of the cohort’s movements. The others are likely to belong in the second to mid-third century A.D., befitting the pattern of civic life and the history of visits to Claros. Were they all the result of an Emperor’s central initiative? The suggestion has been made and accepted, but it has nothing in its favour.⁸⁶ There is no evidence of such a central edict, nor is it easy to imagine an Emperor with a “religious policy” which required them, or an Empire so willing, in its remote corners, to obey such a curious command. None of the texts mentions the Emperor or his well-being, and it is far from clear that they are all of a similar date. They are better understood as dedications which grew up naturally during the mid-second to the mid-third century A.D., sometimes in answer to a consultation by the place concerned, sometimes, perhaps, on reports of others’ consultations. If the texts had a common motive, it may have been fear of the great epidemics of plague which threatened civic life in this period. The formula, however, is tantalizing: how far did Claros’s “interpretation” go? Presumably, it amounted to a list of the proper honours to divinities who would keep a fort or city safe and healthy. Did Apollo also soar higher and send a touch of his “negative theology” to groups who asked for his “interpretation” of the gods? After Oenoanda’s text, we cannot exclude the possibility. When people feared misfortune or wished to know the truth, they asked Apollo himself, from the borders of Scotland to the highlands of Lycia, from Sardinia to Banasa and Cuicul in Africa. Here, indeed, was a demon whom the Christians rightly feared. Apollo at Claros spoke to the Greek and Latin world alike, and the similarity of these brief texts suggests a certain standardization in the directions which he gave.

In one case, we can pursue such questions on gods beyond Claros and Didyma, and trace them far up the course of the Nile.⁸⁷ It was probably in the early third century that an unknown traveller journeyed to Talmis on the Nile’s Upper Cataract and recorded his moment of divine revelation. He had seen signs of the power of the god who was honoured in the large temple and he had “pondered and busied himself very much with them.” At Talmis, there was no prophet and no sacred spring, but the traveller prepared personally for the god’s revelation and “made himself a stranger to all vice.” He prayed and offered incense, and he watched as the sun rose. At the hour when Oenoanda’s altar was warmed by the dawn, the god Mandulis gave power and breath to his image and his shrine. Even here, the statues of the gods seemed to whisper philosophy, far up the course of the Nile. What he had wanted to know was this: “Are you the Sun god, or not?” Here, the question was born of travel and personal experience, not the indecisions of the philosophical schools. At dawn, in Talmis, the god revealed the truth: “I knew you then, Mandulis, to be the Sun, the all-seeing Master, king of all-powerful Eternity.” Once again, the idea of Eternity featured in an individual’s theological text.

From Egypt to Ionia, we can enjoy a rare opportunity, oracles on matters of philosophy and texts in which gods discuss themselves. When had these theological oracles begun?⁸⁸ Only a tiny, random sample survives, but it belongs exclusively in the Imperial period: already in the first century A.D., Plutarch’s speakers at Delphi had considered asking Apollo about his relation to the other gods. They would not, however, have received an answer in philosophic verse, because that verse style belonged with a clear oracular “revival,” which dates from the reigns of Trajan and Hadrian (98–138 A.D.). Soon afterwards, questions to gods on philosophy were presupposed by the story of Apollonius and the “book” which he composed in Trophonius’s cave: this story was abroad by c. 160, at the latest. At a similar date, jokes about gods and philosophic questions appealed to Lucian’s sense of satire: clearly, then, they had become a contemporary fact of life.⁸⁹

By the 170s, Polites the “crown-bearer” was asking Apollo about the soul. From these philosophical topics, it was a small step to the higher theological questions. They grew naturally out of a simple problem: the title or form of address which a particular god preferred. As gods began to multiply their names and attributes, questions of their identity gained in urgency: we find one such question imagined in The Golden Ass, composed c. 170. In the first and the early second century A.D., we have already remarked a greater tendency to merge gods’ names and relate them to a Supreme god: this idea was not new, but its language does loom larger in our texts. As a consequence, we then find signs of the theological oracles. Theophilus may well belong in the early third century, as may the Oenoandans and their “inquiry.” Other such texts owe their survival to two collections, Porphyry’s, compiled c. 270–300, and Cornelius Labeo’s, of less certain date but perhaps a mid- to late-third-century work. The texts which these authors preserved had existed earlier, but perhaps not very much earlier. It may well be through Porphyry that Eusebius the Christian knew a text in which Apollo discussed himself: “Helius, Horus, Osiris, king, son of Zeus, Apollo…” Like his prophet, inevitably, Apollo combined the names of the gods: he “controlled the reins of the dawn,” he said, “and the starry night,” and he was “king” and “immortal Fire.”

Awareness of these texts can help us to understand several currents in later pagan piety, which have seemed much odder when isolated from the language of conventional gods. Intellectually, the second and third centuries A.D. have been judged to have preferred revelation to reason. This priority was not new, but Porphyry, a thinker, it has been said, “with an incurable weakness for oracles,” has seemed to sum up the decline.⁹⁰ By c. 300 A.D., at the latest, he had gathered a collection of texts from the gods into three or more books, entitled Philosophy from Oracles. They have been dismissed as a youthful excess: “les superstitions qu’il voudrait ennoblir sont par nature trop grossières…” and his contact with the “rationalism” of his great teacher, Plotinus, has been thought to have won him away from this false turning. This dating rests only on guesswork. Philosophy, we now see, was evident in the words of the gods whose texts were inscribed at Didyma and nearby Miletus: perhaps Porphyry engaged in some shrewd epigraphy to collect the Didyma texts for his book. It may well have been a work of his maturity, for he had no doubts about its importance. He thought that the collection gave a means of salvation to its more perceptive readers. Can we justly criticize him? Like their own prophets and thespodes, the “gods” knew the school texts of Plato. If Apollo at Delphi had spoken like a Platonist during Plato’s lifetime, the master would have welcomed his support. In the oracular revival of the second and the early third century, there were philosophers serving as prophets: Platonists and Stoics, perhaps a Pythagorean, even an Epicurean.⁹¹ The Platonist names inscribed at Delphi are evidence of their philosophy’s continuing links with Apollo: we can suspect a similar relationship at Apamea in Syria, where an oracular shrine of Zeus kept company with a circle of Platonists in the third and fourth centuries A.D.⁹² It is not too surprising that people accepted the gods’ own evidence when it seemed to agree with their own.

This rising current of “philosophy in oracles” also helps us to understand an influential fraud. In the second century, a collection of “Chaldaean Oracles” began to circulate and was later to surface noticeably in intellectual life from the later third century onwards. Porphyry wrote a commentary on them and their devotees beguiled the Emperor Julian. The neatest way to account for their origin is to assume that the Oracles first appeared anonymously in the earlier second century. They were supposed to be the Chaldaeans’ Eastern wisdom on the rites and nature of the gods, and as time passed, these obscure, high-flown texts gained a growing authority.⁹³ Nobody knew their precise origin; nobody could fully understand them, and Platonist philosophers struggled to write an explanatory commentary. Their reputation was made. They were a jousting ground for the hyperintelligent, and they purported to describe the arts by which a pure and spiritual man could compel the gods. Now we can add another ingredient. Like the best type of fake, they closely resembled their genuine relations, for high-flown philosophy was also spoken by the established second-century gods. One of the Chaldaean Oracles’ keenest modern students fell into their trap. He credited them with Apollo’s answers to Theophilus and the Oenoandans, before the latter’s text refuted him by turning up on the city wall as words from Apollo himself.

Like Apollo, these “Chaldaean” verses expressed much which was being argued by the philosophers. Part of their theology is close to views which we can find in our few fragments of the Platonist philosopher Numenius, who wrote in the earlier second century and took an interest in Eastern wisdom, including the views of “Moses,” and the means by which gods could be attracted by their own symbols and images. “Some sort of bridge,” concluded a great connoisseur of later Platonism, “must have linked the two systems, but I find it hard to be quite sure which way the traffic ran.”⁹⁴ Perhaps Numenius or a close associate composed the Oracles and published them anonymously. The point of their subsequent success is that we, too, cannot be sure where the fiction begins and its sources end. Nor, perhaps, could Apollo. One Apollo, known to Eusebius and subsequent Christians, ascribed “wisdom” to the “Chaldaeans alone”; another said that only those who had a “sacred token of the gods” could see the gods themselves.⁹⁵ Are these texts so obviously fakes? If Apollo’s prophets read and accepted the Chaldaean Oracles, the god, too, would pick up these views, both of which match the Chaldaean Oracles’ teaching.

The continuing sensitivity of the gods to contemporary teaching helps us to appreciate other aspects of Porphyry’s book. He is not known to have cited “Chaldaean” texts in it, but he did know texts in which Apollo approved Eastern wisdom. He also knew texts in which the gods discussed their relation to astrology and the power of the stars: like their prophets, the oracles had absorbed this system of thought.⁹⁶ The second book of Porphyry’s Philosophy gave space to this topic, and also to the “compelling” of gods to attend a “recipient”: priests and philosophers believed in these arts and so, therefore, did their gods. They also revealed their favourite names and the very words of their hymns:⁹⁷ we can well understand how similar hymns and oracles continued to be composed and passed off as the verses of prophetic “Orpheus”, and why the invocations and higher theology of learned magic were sometimes recorded as “oracles” from a god. The style was easily copied, and the Orphic hymns and the magical papyri drew on widespread expectations. In the early Christian era, “paganism” had thus become more than the sum of the acts performed in its cults: it had acquired its own body of divine wisdom. When Christian Emperors banned cult acts, this doctrine was still able to survive and re-emerge. It is central, once again, to an understanding of the Emperor Julian, a man to whom the gods’ own oracles were as evident as their “manifest” presence.

Julian was aware that the Chaldaean Oracles deserved respect, although his understanding of their meaning is unlikely to have gone very deep. He, too, had experienced the gods’ sudden “presence” in their statues and greatly respected the theurgists among his contemporaries.⁹⁸ His conversion away from Christianity owed much to the sense of immediacy which these pagan philosophers could still elicit from their gods. In Christianity, he had found nothing like it. As Emperor, Julian took up the office of prophet at Didyma, and also stepped into its intellectual legacy. Like Theophilus, he believed that Apollo was the “master-founder of philosophy,” and like other local questioners at Miletus, he accepted that the gods revealed their own hymns and titles. He knew as well as any third-century prophet at Didyma how the god “bore witness” to his pious servants. “Many of the gods’ utterances,” said Julian, confirmed the respect which was owed to their priests, and he quoted a splendid text from Apollo which stated the favour of gods for pious men. They kept their “swift eye” on human conduct, and confirmed their constant, Homeric presence. Like the text for the Oenoandans, these lines combined high-flown adjectives with concepts of Platonist and Pythagorean style. Julian did not only live by the words of the pagan gods: he died to the sound of them.⁹⁹ In his last hours, he was consoled by a charming oracle from the Sun god, or Apollo, on the fate of his soul: it promised him deliverance from the sufferings of his mortal limbs and a place in the “Ethereal Light” of his father’s heavenly court, the place from which his soul had descended to human form. How Polites the “crown-bearer” would have sympathized with this response. If we miss the importance of these oracles for Julian, we mistake his piety for an esoteric intellectual reaction which none of his cities could have been expected to share. Rather, he was the great heir of the third century, when the gods had spoken philosophy to their clients.

Such, then, is one side to the language of the pagans’ “rootless and compromise gods.” In pagan company, it finds its last honours in the writings of Proclus the Platonist, c. 470–480 A.D.¹⁰⁰ When he cites an “oracle” on the soul’s aspirations to return to heaven, he does not identify it as one of the various Chaldaean Oracles cited in his text. The verses belong to another source, perhaps to Apollo at one of his two great shrines. Not every soul, the god says, will realize its wish, nor will a life spent in “sacrificing” assist it: somebody, it seems, had asked if constant offerings to the gods could bring a person’s soul to the heavens. Apollo denied that it could, opposing “wisdom” to external piety, and once again spoke philosophy to a questioner. Philosophy, however, was not the only side to Apollo’s business. It was combined with advice on specific practical problems, the oracles’ staple diet. What, then, was the scope and manner of this standard advice and in what sense, exactly, had oracles “revived” in the early Christian era? There were other puzzles in contemporary theology, not only the nature of divinity but the identity of a new God, Christ. At Claros, people were told to pray to a Most High god, facing east at dawn. In some texts, Apollo uttered no favourable words for blood sacrifice or cult in a temple; elsewhere, his clients were wanting to know how the Supreme god related to other divinities. Had the Christians sharpened these questions’ relevance, and by the third century was Apollo trying to map out common ground between the two sides?

II

To attempt an answer, we must come down from the heights for a while and look at the wider oracular landscape in the second and third centuries. At one end, it is clearly focussed. In the early 80s, Plutarch set a dialogue at Delphi whose speakers assumed that the oracles were in decline.¹ They wondered if the prophetic current at each site was fading or if the shrines had died away because the local populations had dwindled. The sceptics in their company were more forthright, suggesting that oracles were simply a hoax. Perhaps we can connect these discussions with a text from Apollo himself. Porphyry’s book quoted an oracle in which the god had been asked about the status of Claros and Delphi. Other sites, Apollo said, had come and gone, having “gushed out on the back of the earth as springs or swirling breaths of vapour.” Some of them “had been taken by Earth below the bosom of the ground,” while others had expired with the passage of time. Only three major sites survived, he said: Claros and Delphi and also the “inspired water of Didyma.” The remark allows us to place the text: Apollo at Didyma had been asked about his sister shrines at Claros and Delphi, and in reply, he had praised them as the sole survivors beside his own.²

In the early second century, this “decline” of the oracles began to be reversed. By c. 200, Christians still wrote polemically as if the gods had fallen silent, but they were ignoring the contrary facts at the sites to which they referred.³ Even so, the change is still discounted: “What,” we have recently been asked, “can one make of the assertion that oracles, ‘it is true, enjoyed a recovery in popularity in the second century’—for which a single inscription is cited, recording help sought by a city in Sardinia from Apollo in Claros? Such characterizing of the feelings and thoughts of fifty million people on any day out of thirty-six thousand has something ludicrous about it.”⁴ Perhaps the full range of evidence can remove some of the absurdity, not one inscription but more than three hundred civic dedications at Claros alone, beginning c. 110 and witnessing, in the 130s, the rise of the thespode and a change in the way the site worked. At Delphi, Plutarch lived to see his earlier dialogue on “decline” be put out of date. He ends a second dialogue, c. 120 A.D., with words of delight on a recent transformation of the shrine’s amenities, a benefaction which we can connect with the Emperor Hadrian.⁵ His generosity had brought new life to the site, although Plutarch’s text ascribes the revival to Apollo himself.

Delphi was not alone in receiving material favours at this date.⁶ At Didyma, we know of benefactions to the shrine by the Emperors Trajan and Hadrian, and again, they helped the oracle’s prestige. In Alexandria, by c. 130, the great temple of Serapis had recently been rebuilt and its “manifest” god was better placed than ever to impress his worshippers in oracular dreams. In Syrian Apamea, the oracular temple of Bel was constructed during Hadrian’s reign. In Pergamum, benefactors in the Hadrianic age changed the face of the great precinct of Asclepius, a god who also gave oracles to questioners. Around 140, the “old and never-dying” Apollo at Patara began to speak again, and at Claros, the dedication of the great Doric temple bears the name of Hadrian. The oracle’s flowering even induced a rival in the neighbouring city of Ephesus. Between 105 and 120, an enthusiastic Ephesian magistrate, Dionysodorus, spent large sums of his own money on repairing altars throughout the city and honouring the cults of all the gods. In the city’s prytaneium, a central civic building, he installed an oracle, “making a chamber fit for the gods.” In a decree, the people and council of Ephesus then stated that this oracle should remain “in perpetuity” where Dionysodorus had placed it. Its god was Apollo of Ephesus, not Apollo of Claros: the shrine was Ephesus’s retort to nearby Claros’s prowess and it stood in the heart of their great city’s ground plan. Once again, cults flourished on the intercity rivalry of the age.

These texts have been joined by another, a very pleasant surprise.⁷ From Cyzicus, on the Sea of Marmara, we now have an inscribed oracle from Ammon’s shrine in distant Libya, given to the city’s “worshippers of Amnion,” whose names follow its text. The oracle had been brought all the way from the god’s oasis at Siwah, once favoured by Alexander the Great: it was inscribed between 123 and 132 A.D. Some fifty years earlier, one of Plutarch’s speakers had been lamenting the oracle’s sad decline and reporting his journey to its site of former glory. Now, we find it reassuring worshippers as far afield as Cyzicus and sending them skilful hexameter verses. They were told, in turn, to go with offerings to Apollo at Claros, evidence of one shrine’s support for another. We glimpse a whole pattern of oracular interrelationships which Plutarch’s essay on “decline” had never countenanced.

As these oracular sites enjoyed a new life, they were joined by their obverse: forgery. The Chaldaean Oracles were faked in the years of oracular revival; in remote Paphlagonia, as we shall see, a prophet of Asclepius was accused of fraud when his oracle began to draw clients from far and wide during the 160s. The possibilities were summed up in the person of the Sibyl. If Sibyls wished to die, nobody in the Roman period respected their wishes. Sibylline oracles proliferated, extending far beyond the verse collections which still survive. Prophetic Sibyls spoke up for pagans, then Jews, and eventually they would speak for Christians, too.⁸ Literary men with a gift for hexameters composed prophecies which they ascribed to the Sibyl and cast in obscure, allegorical language, drawing on a pool of Sibylline prophecy which focussed on political and collective disasters. There was no knowing where a new Sibyl might suddenly surface. From Perinthus, on the Sea of Marmara, we have iambic verses which members of a Bacchic cult society chose to inscribe in the second century A.D. “Good luck to you. An oracle of the Sibyl: When Bacchus shall be sated, then blood and fire and dust will be mixed together.” We know nothing of the cult officers who put their names to this brusque text, men with names like Spellius Euethes.⁹ Their text is a reminder how Sibylline verses had spread and multiplied: did authors sometimes believe that an ancient Sibyl appeared to them and inspired the texts herself? We will meet this question again in early Christian company.

By the second century, the Sibyls’ ancestry had become a civic and cross-cultural battleground. Nobody had established who was the first Sibylline woman, or when she had lived. Indeed, how many Sibyls were there: ten or more? Which was the oldest: the Cumaean, the Erythraean, the Jewish, or the Babylonian who claimed to have spoken before the Flood? Cities competed for the Sibyl’s birthplace and waged a minor version of their wars for the birthplace of Zeus or Homer. Was the true Sibyl a woman from Marpessus or, more plausibly, from Erythrae? In the year 162, Erythrae staked a definitive claim.¹⁰ Her local “cave of the Sibyl” was visited by the Emperor Verus and was then adorned with statues of the Sibyl and the Sibyl’s mother and equipped with a channel for her spring. “No other is my country,” said the accompanying verses, polemically, “only Erythrae…” Here in Erythrae, the Sibyl was emphatically at home. She had lived for 900 years, said the texts, and her prediction that “Erythrae will blossom again” had at last been proved true. The city had just welcomed a “new Erythros,” a second founder, the Emperor Verus himself. A Sibylline prophecy flattered the city and the Emperor, and very soon the Sibyl’s image appeared firmly on Erythrae’s coins.

Her new monument united much in the oracular culture of the Antonine age. A passing Emperor had taken an interest in yet another ancient centre of prophecy; its site had been embellished, and its new form then distinguished its city above all competing rivals. It had also benefitted from a fraud, a “Sibylline” prophecy which had been composed to predict this “second flowering” of Erythrae. We do not know this prophecy’s date, but the events of 162 were held to have proved its truth. Age and uncertain origins lent respect to all manner of predictions. We have a Sibylline poem which “foretold” the eruption of Vesuvius and God’s vengeance on the Romans: in fact, it was written shortly after the event of 79 A.D., but when Plutarch met its verses out of context only twenty years or so later, he thought they were an old, genuine prophecy in which the ancient Sibyl had proved her foresight.¹¹ If Sibyls could hide their origins, so could anonymous “Chaldaean” oracles or “ancient” oracles like the prophecy of “Hystaspes” which predicted a fiery end to the world. The older the prophecy, the more impressive it seemed to some of its readers in the antiquarian culture of the time. People were aware of a great length of time between themselves and these prophets’ lives, but their words were still relevant: no great technological change had divided their era from much of life in the Roman Empire. Past seers had special credentials: they were children and brides of the gods or pious prophets of the ancient East. Almost nobody had a convincing test to distinguish an old text from a fake. The point was not lost on contemporary Christian authors.

In the second century A.D., the second flowering of oracles was true enough, affecting Apollos, Sibyls and competitive forgeries. Was it, though, a “recovery of popularity”? Once again, the greater prominence of a religious feature has been explained psychologically, in terms of rising credulity or the growing “anxiety” which gnawed at the heart of the cities’ age of peace. From this view, it is a simpler step to explain the rise of Christianity, a faith which cut through pagan doubts and fears. Before, however, we accept this perspective, we should be clear what an oracle was and how, like the Sibyl, it related to competitors.

In the Greek world, observed Sir J. L. Myres, an oracle “has been defined as the conjunction of an uncanny place and a canny person.”¹² The canniness of men like Poplas and Polites should not be set too highly, for they went sincerely and innocently about their business. The uncanniness of place was more relevant, a crowning example of the connection between cults and the landscape. Not for nothing are many of the old Greek oracles the most bewitching sites to survive from antiquity, not only Delphi and the sacred valley of Claros but lesser sites, too: the cavernous passage at Lebadeia or the oracular site of Ephyra, which was recently rediscovered in northwestern Greece beside two rivers, on a hill and inside a maze of passages.¹³ The Greeks themselves had connected inspiration with peculiar places. Participants in Plutarch’s dialogues wondered seriously about the quality of Delphi’s air and its alleged vapours: was there a local “current” of prophecy which came and went at the different sites at different times? What, indeed, was so special about Claros’s spring or Didyma’s pool? Iamblichus, in the early fourth century, was still asserting the importance of place against Porphyry’s doubts.¹⁴

The role of the person also aroused a lively interest among intellectuals. Sceptics took one view, but others were divided: did the god enhance a prophet’s mind, casting a spark, perhaps, or playing on it as an instrument, or did he suspend reason altogether? The views were not mutually exclusive, but enhancement made more sense if you looked at the verse oracles: the style and metre were too feeble to be the god’s own. Suspension made more sense if you considered the prophet or prophetess. At Delphi, Plutarch saw a prophetess go raving mad when consulted at an inauspicious moment, and at Claros and Didyma, said Iamblichus, the prophet (or prophetess) was “enthused” and “inspired.”¹⁵ Oracular inspiration continued to derive from a hungry trance, and debates on its nature were to have a long history. They were picked up by Christian leaders in the 160s and 170s who used them against their own “false” prophets. Later, they were translated into Arabic and interested some of the Muslim philosophers who wished to explain their own Prophet’s gifts.

Major oracles, therefore, were a recognized combination of man, god and place. This combination was elaborate, and there was also scope for one or other part of it: a place without a person or a person in no one place. There were also techniques which relied on neither element. In the second and third centuries the pagan cities enjoyed all the forms of supernatural advice which historians of early modern Europe collect and analyse in later Christian contexts. There was only one exception: witchcraft. Pagan society knew no “Devil” with whom individuals could make a pact, and thus no torture and persecutions of “false” prophets and prophetesses. These features were a consequence of Christianity.

To see the oracular revival in focus, we must do justice to its full range of options. Prophetic places were as abundant as ever in the Imperial period, and many of the most famous were connected with the presence of a deity in dreams. A nest of dream oracles still flourished in southwestern Asia where the old heroes spoke, Sarpedon, Mopsus and Amphilochus: in the 80s, Plutarch’s speakers savoured a story of Mopsus’s dealing with a sceptical Roman governor of Cilicia. When some Epicureans urged the governor to test the shrine, the hero guessed his agent’s question and confounded him.¹⁶ From Gaul to Palmyra, gods and heroes also gave guidance by signs and omens at their shrines and sacred springs. At Memphis, a divine meaning was found in the words of children who played in the court of the temple of Apis.¹⁷ Serapis gave oracles in Alexandria and also appeared in dreams, while the ugly curious god Bes had sprung into fashion in the Imperial period to serve Greek and Roman visitors to the Valley of the Kings in Egypt. He received their written petitions and gave answers, Bes the “entirely true, the giver of dreams and oracles, never-lying, acknowledged throughout the world, heavenly…” On the walls of his temple, visiting athletes inscribed prayers for themselves and their families: shown in their sports costumes, they match the athletes whose questions were known to Philostratus and whose dreams were so keenly researched by Artemidorus. The future mattered to these sportsmen, as it also mattered to the sick, for whom there were also the widespread cults of Asclepius. Almost everywhere, the god Asclepius gave advice to clients in dreams, but sometimes he uttered oracles too, which were distinguished from the guidance of his nightly presence. In a classicizing age, almost any past hero could surface and give advice: at Athens in the Imperial age, we even have a text from Harmodius and Aristogeiton, the heroes of the city’s “liberation” in the late sixth century B.C. They were the proper people to speak in an antiquarian’s dreams.¹⁸

We have only to follow Pausanias again in mid-second-century Greece to appreciate the range of available prophecy. Once a year, in Argos, a bull was sacrificed and the priestesses were made to prophesy by drinking its blood. At the charming shrine near Oropus, clients still wrapped themselves in sheepskins and dreamed of the healing hero Amphiaraus.¹⁹ The most awesome consultation was reserved for the hero Trophonius at Lebadeia in nearby Boeotia, whose terrors far exceeded Claros’s.²⁰ A client lived for several days in a shrine of the Good Daimon and Good Fortune, sacrificing freely, abstaining from hot baths and waiting till the diviners proclaimed a favourable night from their study of the entrails of a ram. Then he was led to be washed and anointed by two young boys. The priests took him to drink of the two springs, Memory and Forgetfulness. He gazed on a secret image, prayed, worshipped, dressed in a linen tunic with ribbons, and put on a pair of local boots. He was taken to the oracle’s entrance, a chasm into the ground, down which he climbed on a thin ladder and then passed, feet first, into the lower darkness, holding cakes of honey. Underground, he was taught about the future, “not always in the same way, for one person sees, another hears.” He returned feet first through the same narrow hole and was revived by the attendant priests, who set him on the Chair of Memory and asked him all he had seen and heard. He had to dedicate his findings on a tablet at the shrine, none of which, unfortunately, has been found. He emerged “gripped with terror and quite unaware of himself or those around him,” but later, his wits returned to him and also, added Pausanias, “the power to laugh.” We know of this remarkable ritual because Pausanias himself had endured it. The shrine was not moribund. By the mid-fifth century Christians had placed a martyr shrine near the cavern’s entrance: if the oracle had not been active at a late date, they would not have troubled to neutralize it with a shrine of their own.

Wherever there was water, indeed, there was a possible source of prophecy.²¹ The major oracular shrines used its powers: Claros and Didyma had their springs, while Delphi’s Cassotis was piped down to the oracular temple, appearing to enter it, although it cannot be linked exactly with the rites of inspiration. At a simpler level, people threw offerings into water to see if they swam or sank, while holy springs everywhere upheld the sanctity of oaths. From Gaul to Sicily, suspects were subjected to hot geysers which scalded perjurers, and in Cilicia, there was a special spring which made them choke. Everywhere, springs were able to heal and cure, helping men and also the animals on whom human life depended. In Thrace, however, certain streams were thought to make horses more than usually uncontrollable.²² The shrine of Demeter at Patrai had a spring into which visitors dipped a mirror on thin cords: then they prayed, burned incense and gazed into the glass, “which showed them a sick person either living or dead.” The oracle answered questions of health and sickness only. There had once been a similar spring at Cape Taenarum “which showed ships and harbours,” but by the time of Pausanias, it was said to have been spoiled for ever by a woman who had washed her dirty clothes in it.²³

The specialization of these oracles is a further sign of their abundance. Uncanny places were everywhere, and without any expensive ceremonies, people turned to them for knowledge of the future. In the early third century, the historian Dio was as intrigued as Pausanias, fifty years earlier, by the properties of these places and went out of his way to investigate them.²⁴ At Apollonia, he described the two jets of flame which were still active above ground and onto which clients threw incense, having made their prayers. If the flame shot up, the prayer was to be granted, but if it died away, the answer was unfavourable. Here, the oracle refused questions on the awkward topics of death and marriage.

Meanwhile, prophetic persons were to be found everywhere, in the cities, the countryside, in every cultural zone of the Empire. If we are right to date the “Chaldaean Oracles” to the Hadrianic age, it was then that a new art, theurgy, had begun to take shape among certain intellectuals. Educated theurgists then combined these rites with an emphasis on pure living which any prophet at Didyma or Claros would have understood. Following up “traces” left by the gods, they insisted that their art was distinct from malevolent magic. Its masters are not to be confused with the lunatic fringe of modern parapsychology or mediums at a séance. They were developing the ideas of prayer and oracular wisdom which were shared by most of their educated contemporaries and the great contemporary shrines.²⁵

At a less esoteric level, street prophets were strongly in evidence. We hear much about prophetic women,²⁶ “pythonesses,” as they were popularly known, but it is not true that oracular “possession” was a distinctively female gift or essentially imagined in the metaphors of man’s “possession” of women.²⁷ There were male “pythons” too, and at shrines like Claros and Patara, women played no part. More generally, possession was believed to attend the “young and the somewhat simple” of either sex: in Plutarch’s day, the Delphic prophetess was still a simple country girl. When the city of Laodicea-by-the-Lycos sent its delegations to Claros, it sent its own local “prophet of Apollo” and “priest of the children,” both of whom were young boys, like the city’s choral singers.²⁸ Childhood and prophecy, these people assumed, were natural allies. The boys looked so innocent and so unspoilt. No barrier of worldly experience blocked them from divinity; no cares and distractions marred their godlike beauty. They were expected to speak, so their tongues ran away with them and their fantasy took wings. Private practitioners were quick to exploit the possibilities, and spells for obtaining a prophecy made ample use of a child’s transparency to God.²⁹

Cynically, we might feel that children were most open to trickery and autosuggestion. The ancients, however, were more trusting. In court, Apuleius reminded his audience how the vulgar and learned agreed on this point, that the powers of divination and prophecy resided especially in small boys.³⁰ Their explanations varied, but personally, Apuleius believed in the young soul’s innate power of divination, which could be lured by music and sweet scents to revert to its primal nature, half-mortal, half-immortal, so that it foretold events to come. Others disagreed. Prophecy, they said, was the gift of the divine intermediaries, or daemones, who moved between gods and men and could enter into a child who was handsome and quickwitted, and then “he would give prophecies when touched on the head and clothed in a fair, white cloak.” Earlier in this century, children, dressed in white, were still being given a prominent role in the drawing of the winners in Italian state lotteries.

These prophetic techniques were matched by others, in which no woman or child was required as an intermediary, because they turned on significant chance and a relevant use of the lot. This service had become freely available in cities during the Antonine age. At Pharai in Achaea, said Pausanias, the image of Hermes was set behind a stone hearth onto which were clamped lamps of bronze.³¹ “Inquirers of the god come to it at evening and burn incense on the hearth. They fill the lamps with oil and light them and lay a local coin called a ‘copper’ on the altar to the right of Hermes. They whisper their question into the god’s ear. Then they stop their own ears, leave the marketplace and, just outside it, unblock their hearing: whatever words they then hear they consider to be an oracle. The Egyptians have a similar custom at the temple of Apis.” The busy client could also consult the gods by throwing dice or knucklebones. Pausanias remarked how sets of dice were left lying on small ledges below statues of Hermes or Heracles, and one of the continuing gains from archaeology is the discovery of ever more stone fragments of these dice oracles and their inscribed lists of answers.³² Examples have been found in Phrygia, in a differing form in Cilicia and also in Pisidia, and there are other examples from Thrace. Their forms varied and two main types are now evident, although the procedure was always similar. The questioner approached, threw the set of five or seven dice and looked up their combination of numbers on the inscribed list of answers. “Five threes: Zeus and Athena. Now is the time for you to attempt what you have come to ask about. The gods will fulfil everything according to your intention. Zeus and Pallas Athena are sending you victory…” Again, these answers were not the result of pure chance. They were attached to particular gods and abstract powers, while the method itself was said to have been conceded by Apollo to the care of the god Hermes.

These oracles were not the hidden interest of a lower-class minority. The better-preserved examples seem to have stood on rectangular monuments in cities and were surmounted by a statue, presumably the statue of a god. The majority were found on stones which belonged closely to a public temple or to a section of their city’s gateways. Several of the answers address visitors or passersby and are well suited to their placing in crowded public thoroughfares. One set, from the Pisidian city of Anabura, stood above a public vote of thanks to its donor, Attalos, a man who “loved his country, and was a Founder and Benefactor.”³³ These titles belong with an immensely grand civic figure. This type of consultation had a near-relation in alphabetic oracles, texts which required a questioner to pick model letters of the alphabet and read off the appropriate answer beside each letter on a list. This type, too, was widespread, and we have one good example from Oenoanda itself.³⁴ Like the dice oracle, it gave answers in verse and favoured a riddling, enigmatic style: “You are kicking against the pricks: you are struggling against contrary waves: you are searching for a fish in the sea.” Answers could only be very general when the questions were so vague and impersonal. Like the dice, these oracular “alpha-to-omegas” were under the care of the gods Apollo and Hermes. Most of the local variants draw on a similar core of answers and suggest that these oracles had had a common origin. Had they spread from a major oracle centre of Apollo, perhaps Delphi itself, which may have imposed the basic core and style and left others to elaborate it?

The texts of these oracles are datable to the second and third centuries by the style of their inscription and make an interesting parallel to the revival of the major inspired shrines. Like cash cards, they made the gods’ resources swiftly available without priests or complex offerings. Although the patterns of answer vary locally and the answers are usually banal, most of the sets share a common concern with men’s prospects for travel and trade.³⁵ They are concerned to predict the shape of the future, not to analyse problems in the past. Each answer in a dice oracle came under the care of a particular divinity, distinct gods who had extended their service as one more “infernal snare” in the cities in which the early Christians were putting down roots.

“Prophets,” too, were alert to the scope for ready-made, practical questions and answers. If there was no dice oracle to hand, a client could turn to an “expert” and his impressive oracular books. We happen to know one example from a sequence of finds on papyrus.³⁶ It first came to light when a papyrus of the later third century revealed parts of a long list of Greek questions. At first, they seemed to be destined for a traditional oracle and their bleaker examples were misread as a proof of the “mood” of their age. Soon afterwards, they were related to a longer and older collection, the so-called Fates of a legendary sage, Astrampsychus. These Fates had only been known in Christian manuscripts of a later date, but the papyri have proved that they already existed among pagans by the later third century A.D., and probably earlier. They have sufficed for a brilliant decoding of the system by which they worked. The questioner chose an appropriate numbered question from an “expert’s” book, whereupon the expert matched the question with a numbered answer and gave his client a response which seemed to be chosen by an act of Providence. We do not know when Astrampsychus’s Greek almanac originated, but its surviving range of questions contains inquiries which only suit a date after the mid-first century A.D. Significantly, its questions suit the various social classes, high and low alike. While matters of marriage and health, love and inheritance interested everyone, others varied with the client’s rank. “Shall I become a councillor?” “Shall I go on an embassy?” “Will I be sold into slavery?” Astrampsychus catered for more than a hundred eventualities and helped expert owners of his almanac to earn a living by following the simple rules. His clients were not only humble and uneducated people, nor was his book shortlived. It passed, with similar Fates, into the Middle Ages, when it was still used.

Astrampsychus’s text is only one symptom of a wider pattern of practice. Throughout the ancient world, it was normal to prefer divination to indecision. Intellectuals did not argue the interesting problem of whether or not the preference was rational. Although a few remained sceptical, most of the others merely argued about the forms which divination took.³⁷ Here, there was a rough distinction between the value of oracles which a god inspired and these lesser techniques which relied on human materials. A few writers denied that the lesser techniques had any merit, but their complaints did not stop men from practising them, any more than attacks on macroeconomics have stopped administrators from trusting its theories. The prizes were too tempting. It was left to lamblichus to include both types in his Platonist philosophy and abolish the old distinction between inspired prophecy and minor divination.

Techniques of this type offered help where help was always welcome: they also attached to a congenial intellectual climate. Like us, the ancients were deceived: just as macro-economics has traded on mathematics, or sociology on science, in antiquity, too, dubious attendants found a home in the company of rational astronomy, mathematics and medicine. In its simplest forms, divination rested on an argument from analogy and the interpretation of “signs.” More generally, this type of thinking was rooted in other arts which flourished in second-century culture, the study of dreams and the stars, the flimsier sorts of medical prognosis and even the study of people’s eyes and faces. Artemidorus’s interest in predictive dreams and their “allegory” is akin to oracular divination and its art of conjecture by signs and analogies. The astrologers, too, claimed divine inspiration for their art, although the better theorists claimed only to study “signs” and tendencies, not determinant causes.³⁸ Like dreams, the stars were grounds for a speculative prediction, although the macro-economy of the heavens was too complex to be caught in any one causal theory. A few thinkers denied there was anything in the theory at all,³⁹ but it took more than a good argument to kill an art which promised so much. Perhaps the summit of divinatory skill was the art of telling people’s nature and prospects from their faces. In the Antonine age, the great sophist Polemo drew together all previous learning on this riveting topic and added examples from his own experience.⁴⁰ By studying people’s faces, he had foreseen all sorts of trouble, problems for Hadrian while hunting, for a bride on her wedding day in Samos, for a woman in Perge who was walking, wearing her veil, to the temple of Artemis. Naturally, also, Polemo could see through his rival sophists by assessing the shape of their hideous faces, a fact which he reported in a torrent of extreme abuse. While the tireless Artemidorus was collecting and analysing dreams, Polemo, the prince of sophists, was judging his fellow men from the style of their walk or the thickness of their hair. Like Artemidorus’s, the art required a use of analogy and wordplay, combined with a strong sense of social categories and “types” of character. Of course, other intellectuals would have mocked it, but the future lay with these types of study. In the Antonine age, Ptolemy compiled his great guide to astrology; Artemidorus amassed books on the art of dreams; Polemo described the predictive power of physiognomy. Their books passed through Syriac into Arabic and fascinated the court societies of Muslim rulers. With the help of Polemo, people could pick a decent slave, long after his book’s composition: slaves, like horses, were judged by their eye, that unerring “mirror of the soul.”

These arts were not new, but their great literary authorities were writing at the height of the oracles’ revival, in the second century A.D., a time when the gods’ “providence” was freely acknowledged. In the past, each of these arts had owed something to Greek culture’s contact with the older civilizations of the Near East. Like a younger, freer economy, the Greek world had pillaged its neighbours’ techniques, refined them into general theories and made prophecy, divination and oracles a major Greek industry. In the early Christian era, “information technology” was blossoming under the care of the pagan gods: was it, however, an entirely free enterprise? In an Imperial age, divination was strictly banned if it concerned the prospects of Emperors: what might not follow from the horoscopes of the Emperor and his rivals? Control, however, could go beyond these awkward subjects, as we have learned from a recently published papyrus which well sums up the continuing range of the technology and the varying views of its worth.⁴¹

In the year 198/9, it shows a prefect of Egypt ordering his district governors in very strong language to ban the “chicanery” of oracles and divination. “Encountering many who believed themselves to be deceived by the practices of divination, I swiftly judged it necessary, in order that no danger should ensue upon their foolishness, herein to enjoin clearly that all people must abstain from this hazardous inquisitiveness.” Officials who failed to carry out the order were themselves made liable for its penalty. “Hazardous inquisitiveness” amounted to “oracles or written scripts given as if from a divinity or the procession of divine statues.” This rite with statues was an old Egyptian practice, but the “written scripts” are more tantalizing. Probably they are the pairs of questions which we find on papyrus and to which a god seemed to “give” an answer.

Why had the prefect been so bothered? His edict cannot be dated exactly, but it falls within the months in 198/9 when the Emperor Severus is known to have visited Egypt. On this visit, Severus is said to have investigated the country’s secret texts and made many of them inaccessible: did the prefect issue his order after he had seen Severus’s interest? However, no word in the text requires this connection and its ban goes wider than divination concerning the Emperor. The prefect, it seems, had been shocked by the swindles of prophets and diviners, and so he struck at their entire undergrowth. He failed, but his edict is excellent proof of their vigour and their continuing capacity for dividing opinions about their merits.

It is necessary to bring the range of these techniques together, not just for their antiquarian interest but also for the wider question of mood. If we set the “decline” and “rise” of the inspired oracles against this mass of evidence, their explanation has to be sought in a different direction. Theories of a psychological change, or a “new mood,” are no longer helpful. There had been no sharp rise in credulity, no pressure from new, unsatisfied anxieties. Almost all of this range of persons, places and techniques had been credited for centuries, flourishing because they met enduring needs. In the second century, inspired oracular shrines were merely establishing themselves once more in the market: the market itself was not new.⁴² The questions, then, are how and why the oracles re-established themselves. What did they offer which other services did not and what was needed for them to be able to offer it impressively?

Oracles were not concerned to defeat the rival services: at one level, they joined them. The temples, too, used techniques of drawing lots and gave simple answers to written petitions: at Claros, there is evidence of an oracular use of dice, and at Didyma, of an early use of knucklebones. No doubt, the clients took their questions to several authorities and sampled a range of advice, but perhaps they felt that the answers from a temple had more authority. Perhaps, too, they were glad of the privacy, a rule which private diviners could not be trusted to observe. In Egypt, during the second and third centuries, we can sample the questions which were put to various gods and temples by a technique which did not involve inspired speech: they were written on papyrus, and in Upper Egypt, papyrus survives.⁴³ The bulk of the surviving questions follow more or less the same form. They address their god, pose their question in a conditional “if” and then add “give me this” or “bring out this” as their final sentence. Some neat papyrology has helped us to understand their phrasing. The questions were often written on one and the same papyrus and then detached into halves. One half phrased the question negatively, “if I should not go,” the other, positively, “if I should go.” The two slips were then sealed and submitted to the god for his choice, and the client phrased his request as “give me” or “give me back” this piece of papyrus. We do not know how the god chose the answer, but perhaps his priest simply drew it from a jar: possibly, the god pointed to it by means of his statue, the old Egyptian rite whereby the god’s statue was borne on his priest’s shoulders and tossed or turned to direct him to one of two alternative answers which lay on the ground before him. Whatever the method of answer, we do have a precious range of questions which extend from the Hellenistic period into the Christian empire. They are very homely:⁴⁴ “Isis: are you the origin of my trouble and will you give me help?” “Am I to stay in the job of tax collector?” and that question of agonizing luxury: “Which of two women shall I marry?” These questions are personal versions of those which we also find in Astrampsychus’s book of Fates. At this level, the gods came naturally, because they had such obvious uses to the anxious and the ambitious alike. They were a marriage bureau and a career service, a medical surgery and a farmers’ bulletin.⁴⁵

Their service could only survive and retain credit at this practical level by setting limits to the suitable forms of a question and answer. The gods were prepared to consider a choice between alternatives, but if mortals asked for too much, they risked provoking a god’s displeasure. If mortals did not pose alternatives, they ought to cast their questions as requests, “whether it was preferable and better” to do this or that.⁴⁶ The god, then, could not be refuted. If he advised action and the result was disastrous, questioners were left to reflect that the alternative would have been even worse. Many of the questions concerned clients’ prospects in love and business, travel and public service, and distinguish their clients from contemporary tribal societies whose diviners are more concerned to analyse the past than to choose between options for the future. Life in antiquity was less uniform, even in small Egyptian townships, where there was a strong emphasis on prospects and choices between possible futures. Among the surviving questions, there is less of the obsessive concern with adultery which has characterized the tribal oracles, fondly described in the modern Sudan. Occasionally, questioners did ask the god specifically about a culprit, and here, perhaps, there were risks. Did the priests of an oracular crocodile god indulge in careful preliminary questioning before they put such questions to their god? Observers of modern oracles have commented on these preliminary discussions and on the gods’ convenient gift for rude, dismissive replies. So, too, in the Greek-speaking world, we have seen how gods could rebuke their questioners’ “insolence” if their questions went too far. Respect for the gods’ anger and temper was a good insurance for the system, and faults, meanwhile, could always be blamed elsewhere. By Porphyry’s time of writing, there were several alternatives. An oracle might have been given by a lesser daemon, who did not know the truth: mundane questions, Porphyry warned very tellingly, were particularly liable to attract an untruthful spirit. Otherwise, the gods’ natural truthfulness might be interrupted while they came down from heaven. The lower, earthly regions fell under the powers of fate and the stars, forces which interfered with the gods’ normal service and distorted clear transmission.⁴⁷ In these and other ways, one false science supported another, preventing users from seeing the culprits: themselves.

III

In this maze of techniques and oracular practice, where did the consultation of great shrines and inspired prophets belong? Were most of their questions significantly different, questions which concerned the world of the gods, perhaps, whereas divination and the other oracular arts concerned the world of men? To test this distinction, we can look at the intriguing sample of oracular consultations which clients inscribed on durable materials in the Imperial age. The surviving texts are a random sample, and insofar as they were inscribed for posterity, they may be one-sided: the more impressive and solemn answers were perhaps more likely to be displayed and preserved. They also raise difficult problems of dating. Sometimes we can find a clue which fixes them more closely, but often we have only the style of their lettering by which to judge, and on this alone it is very difficult to place an inscription within the second or third century. Nonetheless, the texts do illustrate a major side of life at classic oracular shrines. First we can set their concerns beside the questions of God and the soul with which we began. Then we can use them to put a newcomer in perspective.

For a start, the oracles which survive in inscriptions at or around Didyma show the problems which could arise over details of cult. At Didyma, the treasurer, Hermias, noticed that the altar of Fortune had been crowded by newer buildings in Apollo’s paradeisos, and thus was no longer visible: could he please move it to join the other gods’ altars in the circle around Apollo’s own main altar? Apollo agreed, in a single vague hexameter: “It is right to honour all the immortals and worship them all.” This same Hermias recorded another question asking the god what was “dear” to him, and again Apollo answered in a single line. “It is better and finer to act with traditional purpose.” When were these questions posed?¹ Their lettering has been placed both early and late in the third century A.D., but the earlier date now prevails, after the recent discovery of Hermias’s longer text. He was deeply concerned for the details of cult and the wishes of the god whom he served.

A similar problem troubled Damianos the prophet, a man who had come from Cyzicus to serve Miletus’s shrine. He had noticed there was no altar to Kore the Saviour goddess from his native city among the “encircling altars of many gods”: could he, then, put one by the altar of her mother, Demeter? Kore, he claimed, was Apollo’s “most holy sister,” a curious genealogy for the goddess, our Proserpine, but one which Apollo did not rebut: “Perform the honour of joining the encirclement,” the god replied. Damianos described himself as a “lover of the gods” and returned, keenly, with a second question: would Apollo also be “legislator” of the “pious form of address in hymns” for Kore, whose altar was being honoured? Apollo replied, “Let us invoke her as Saviour with fair cries at times of sacrifice, gentle to encounter with her mother Demeter.”²

This question and answer are full of religious interest. Damianos, too, referred the placing of an altar to Apollo himself: this “lover of the gods” had been “troubled,” he said; he besought “master Apollo Helios of Didyma,” an equation of Apollo and the Sun which the god himself had probably approved at the shrine in an oracle and which was in evidence by 200 A.D., at the latest. In reply, Apollo assumed the age-old patterns and fears of a divine epiphany. While offerings were made to Kore, she must be “called” with propitious cries of invocation. She must be hailed as Saviour, her native title in Cyzicus, and be greeted as “gentle to encounter,” a phrase which carefully prescribed her mood in advance of attendance. These old patterns of prayer and divine presence still mattered to the god and his prophet: as in Homer, “the gods are hard to encounter openly.” Damianos’s question has been ascribed to the later third century A.D. on very tenuous grounds, but “master Apollo Helios” is a title known earlier and a date in the first half of the century is now much more likely for his text.

The dating of these two questions is important, because their rites and language have been taken as proofs of Didyma’s declining years. In fact, they belong in the age of Poplas and his associates, when a general loss of vigour was not in evidence. The crowded “encircling” altars, the equation of Didyma’s Apollo and the Sun, the encroachment onto the shrine’s “precinct”: none of these changes was a consequence of the mid-third century and its hardships, let alone of a growing apathy as Christianity then spread. With the passing of time, Didyma had become packed with ever more pagan gods and altars, and like his prophets, Apollo retained a lively sense of the awe and presence of gods when “called.” These oracles were not the products of a growing religious indifference: they belong with Alexandra’s question on the gods’ “visitations” and the praises of Demeter, inscribed on her same stone. Their concern for the gods’ mood and identity recurs in a further text, which was found some distance from Didyma, but had almost certainly been transported from an original home at the shrine. Its upper part is missing, but the rest of it introduces us to a female questioner, who was returning, like Damianos, to put a second question.³ “You told her to propitiate Hera, but where should she do so?” Evidently, there were several shrines of the goddess Hera, presumably in Miletus itself, and Apollo replied in some lofty and puzzling verses: she should worship Hera at the “halls of an effeminate man” near baths for tired, elderly limbs and the dances of virgins to the flute. Perhaps a particular shrine of Hera stood near a set of baths and the shrine of a god which was served by a eunuch priest. Whatever the reference, the question and answer turned on a familiar concern: how exactly to “propitiate” a god, a word which oracles at Didyma favoured, and which temple would have the best effect? Again, the moods of the gods were not to be risked without due care. It is to Didyma that we can trace another text, known to Christian authors, in which Apollo told the people of Rhodes to “propitiate” Attis, Adonis, Dionysus, apparently running the gods’ names into one.⁴

Prayers and sacrifices were not the only activities which continued to involve gods in mortal life. Their presence was also freely invoked in oaths: Tertullian, c. 200, complained of the problems for fellow Christians who wished to borrow money but had to swear oaths by the pagan gods to their creditors.⁵ What was the proper form of an oath? If Apollo could “legislate” on prayers to Kore, he could also define the terms of an oath by the gods. In his book on oracles, Porphyry quoted an elegant text in which Apollo at Didyma listed the various gods and their functions, and showed a precise grasp of their separate identities without any trace of “monotheism.” A second text survives as his answer to one Rufinus, who wished to know how his sea captain should swear an oath. The text is an inflated statement in hexameters, requiring the poor man to stand with one foot in the sea, the other on dry land, and to hold sea and shore in either hand. He must invoke the heaven, the earth and “the life-giving leader of Ethereal Fire”: “this is an oath which the exalted dwellers in heaven do not themselves ever dare to dishonour.” This text is particularly important, because we know this Rufinus as a Roman consul and an immensely rich benefactor of Pergamum, where he built the circular Pantheon. After his death, Apollo answered another problem of cult: he told the men of Pergamum where to bury this magnificent hero “because of the past excellence of his life.” This millionaire and great public figure had the particular obsession of his type. He was anxious that his ship captain should not dare to cheat him, and so he consulted the god on his enterprise. To distinguished men like Rufinus, there was no rigid line between the details of trading and their own social respectability. There was also no doubt about the value and presence of the gods in both.

Vows, too, involved gods in mortals’ willingness to stand by an obligation. We have seen the quantity of stone reliefs and dedications “vowed” in the hope of favour or in obedience to a dream, but did the gods themselves always want such a vow to be fulfilled? In the marketplace at Miletus, one Ulpius Karpas inscribed his question to Apollo: “whether it is ‘dear’ to Serapis for him to perform his vow.”⁶ We know this Karpas as a member of Miletus’s council and a prophet, in his own right, of a Most High god in the city in the early third century. He turned to Apollo, not to the god whom he served, when he wished to know another god’s preference. Then he inscribed the consultation proudly in public.

If Apollo could define the preferences of other gods, he could also speak on his own. It is in this light that we can best make sense of Didyma’s fragmentary text on songs and sacrifice which was inscribed on the reverse of a marble block in the wall of the Prophet’s House. Just enough survives for the drift of the oracle to be clear.⁷ Apollo emphasized that the immortal gods did not need “possessions,” or gifts of hecatombs and offerings of gold and silver colossal statues. What Apollo preferred was song, especially if it was sung just before his oracles were delivered. “I enjoy all singing, but particularly songs which are old.” In the past, he concluded, he “drove away plagues, putting to shame the grievous threads of the Fates.” In the 270s, Apollo at Delphi did refer rather coldly to mortal life as “blood-devouring,” and in texts like the Oenoandans’, blood sacrifice is not an evident “command.”⁸ In the Didyma text, Apollo was protesting more generally at excessive gifts, whether animal victims or works of art. He ended by referring to the very origins of his cult, a time when he was believed to have rid Miletus of plague with the help of a mysterious alphabetic hymn. Its symbols were remembered and still known to Clement, a learned Christian author, at the turn of the second century.⁹ His musical defeat of the plague was relevant to a text on the value of song.

What is the origin of this remarkable text, which shows a true “philosophy in oracles” and survives without its question? The inscriptions of a great family at Didyma may allow us, at last, to place it.¹⁰ From the family of Poplas the prophet, we have a text in honour of a lady who offered a due “hecatomb” and poured libations at the oracular shrine, all “in accordance with the oracles of the god.” The precise degree of her “kinship” with Poplas is not known, but this pious lady is a younger member of his line, active, then, c. 230–250 A.D. It is particularly interesting that the back of this same stone was cut with a long, proud text which gives us our fullest family tree of Poplas and his prophetic relations. It then quotes Apollo’s oracular praises for Flavius Ulpianus, a prophet who was commended for his respect for ancient oracles and his care of “sacrificial business.” This prophet is one of the latest whom we know in Didyma’s inscriptions and probably held office in the 250s.

At this date Apollo’s own words were still praising a prophet who sacrificed keenly, and due hecatombs “in accordance with the oracles” were being paid at the shrine. The oracle text on song cannot already have existed at Didyma on a block of the Prophet’s House; otherwise, these pious servants would have been flouting the explicit words of the god. On other grounds, the text has been dated late, to the end of the third century, where it has been seen as Apollo’s “swan song”: the choirs in his cult, on this view, had been forgotten, and as soon as the god tried to revive them in this text of spiritual piety, the Christians silenced his shrine altogether.¹¹ It may, however, be the opposite, not a swan song but an accompaniment to an oracular restoration.

The pious sacrificing prophet, Flavius Ulpianus, had a prophetic cousin, Aelius Granianus Ambeibios Macer, no less, whose statue was dedicated elsewhere in Miletus by his wife, Agatho. Her inscriptions called him “dear to Phoebus.” His statue also stood near the temple, where its inscribed base was rediscovered, showing Macer to have been a man of notable talent. This member of a great Milesian dynasty was also a victor in one of the city’s great third-century festivals: he won the contest for the best “encomium,” or speech of praise. He had also been lyre player in Miletus’s cult of Apollo Delphinios, “advance lyre player,” on the likeliest translation, who played before the central moment of cult: he seems to have held this honour while a boy, when he had also been a “choir leader,” presumably in Apollo’s cult. As the years passed, Macer married well and, like his father, became prophet of Didyma’s Apollo. In his year of office, we know, his wife donated a roof and stucco to the colonnade round the Prophet’s House. Macer, we have also learned, “restored ancestral practices” in the cults: a notable scatter of inscriptions still honours him and his great forbears at the shrine.¹²

Is not Ambeibios Macer the missing clue to Apollo’s “swan song”? Other members of his dynasty had recently been slaughtering hecatombs and showing expensive care for sacrifice “in accordance with Apollo’s oracles”: they are known, too, for their patronage of large monuments and statues elsewhere at Didyma. Around 200, we know of the lavish work of a daughter of another great Milesian family: she dedicated a statue and a silver relief of the gods, which may have been a scene of the birth of Apollo and Artemis.¹³ This type of costly display was not true piety to Macer’s mind. When he volunteered for the job of prophet “at his own promise,” he undertook a different gift, a part of his “restoration of old customs.” He revived the prominence of choir song, an art which he had practised personally in his youth. Through his musical prophet, Apollo spoke in support and added a mild reproach: the gods did not want statues and hecatombs, the piety of Macer’s rich cousins, but they rejoiced in choirs and enchanting song. Songs, said the god, were to be sung “before” the oracular words were delivered. In his youth, Macer had served Miletus’s other great cult of Apollo as the official lyre player at this very moment, playing for the god “before” the rites began. There may, indeed, be more. Macer’s own father was also a prophet, Granianus Diodorus Phanias, and he, too, had shown an exceptional talent: he had been a victor in one of the grand “circuit” games in the Mediterranean. Competitive sport was not alien to great upper-class families: in Miletus, Poplas’s family included a boxing victor, while other prophets had winning athletes in their line.¹⁴ However, the elder Granianus may have been a literary victor, a public speaker, perhaps, at the great festivals. If so, the performing arts ran in his son’s blood, and so, very possibly, did philosophy: at Didyma, one “Phanias” described himself as “successor of Plato” and also served as Apollo’s prophet. It is quite possible that he belonged earlier in Macer’s family, where the name Phanias recurs.¹⁵ Once again, Apollo proved only as wise as his prophet. Ambeibios Macer was a lifelong musician and a prize-winning man of letters, well able to write hexameter verse: philosophy ran, perhaps, in his family’s past. In his year of office, Apollo spoke to him in favour of song; his words were well chosen and the text wore the stamp of philosophic piety.

This text, we now see, was not a swan song:¹⁶ it pointed to lively concern and rivalry between two branches of a rich prophetic dynasty. One year, the god had received hecatombs “in accordance with his oracles”; in another, he disclaimed hecatombs and praised ancient music instead. Inevitably, Apollo agreed with his yearly prophets, and when Macer “restored” ancient customs, the god endorsed the change. How great, indeed, had this change been? Had the choirs really ceased altogether in the years of Poplas and his fellow enthusiasts? In Miletus’s other great cult of Apollo they had not,¹⁷ for Macer was still leading choirs as a boy and playing a prelude on his lyre. Perhaps Macer merely revived an old, neglected hymn and set it at a new moment, just before the prophetess spoke from her seat. These very points were singled out for approval in Apollo’s oracular text: at the same time, the god belittled cruder piety, the hallmark of Macer’s cousin and kin.

These texts on cult were wholeheartedly conservative; they praised the “old” music and “old ancestral customs”; they endorsed attention to the old oracles and ordinances and to “ancestral purpose” in worship of the gods. They take us through the early third century A.D. into the 240s and 250s, the age of the Emperor Gordian and the first edicts which required sacrifice to the pagan gods. These Milesians needed no religious “revival” from above: Apollo and the details of his honours had not lost their interest for people who led his cult. They questioned their god for oracles of the type which we suspected: answers to details of change and practice in matters concerning the gods. Divination and dice oracles could not handle this business with authority. Nor could they sort out a running problem which even gods could not escape: troubles with the staff.

If Apollo could define cult practice, he could also give the final authority on priesthoods. From Claros, he sent a “testimony” to the purity of Prisca, a priestess of Artemis who had served the goddess for sixty years in her Macedonian city of Stobi.¹⁸ The city consulted him with a small delegation in the mid-160s, the years of plague. Maybe the envoys also drew his attention to their chaste, long-serving priestess whose family was so prominent in the city’s upper class. Then, or separately, Apollo applauded her virtue, perhaps confirming the continuing wisdom of her life appointment. Questions of tenure certainly concerned him. From Claros, he sent a most obscure answer to the people of Aizani which seems to bear little resemblance to their question.¹⁹ They had asked if a particular priest should remain the priest of the Founder, evidently beyond his usual term of office. Perhaps there had been a dispute, but Apollo’s answer sufficed to settle it and was inscribed on a building, perhaps the Founder’s shrine. The text seems to allude to an oracular riddle from Herodotus’s first book and may have been an answer from Claros in the years before the lifelong thespode improved the quality of the oracles.

These questions, once again, concerned affairs of the gods, not the political affairs of men. If tenure is one staff problem, correct selection is always another. At Didyma, in the late second to the early third century, we find Apollo easing the process, again in the highest social circles.²⁰ Saturnilla was a very well-born lady who had been married and borne children, but then, perhaps as a widow, she was appointed to a priesthood of the virgin Athena. Thereupon, some questioners asked Apollo for his opinion of her suitability. The god noted their delay in raising the matter and commented, it seems, on the contrast of Aphrodite, goddess of sexual love, and the virgin Athena. Nonetheless, Athena, it seems, had accepted their choice, and even though Saturnilla was not a virgin, her appointment, said Apollo, could stand. It seems that willing and eligible virgins were in short supply during these years, but Saturnilla’s exceptionally good birth reconciled Apollo to her choice. She was connected with Roman senators and athletic victors, and her sons were so pleased with these sub-Homeric verses and their theological wordplay that they inscribed them as a memorial to their mother’s honour.

If Apollo could approve the servants of other gods, it is not, then, so surprising that he also spoke about his own. Recently, we have learned that he “appointed” a prophetess at Didyma whose name, Tryphosa, is the only one we know in all the years of the office.²¹ In the Imperial period, again, Apollo intervened to endorse a woman’s appointment: the honour was specially recorded in a family inscription, suggesting that it was not the god’s usual practice. What, then, of his male prophets? Here, they could hope for a separate honour: a divine “testimonial.” Cities passed honorary decrees “testifying” to the virtues of their magistrates: the gods caught the habit and “testified” to their servants.²² The habit was probably more frequent than we realize: we find the god of a prophetic spring at Palmyra “testifying” to civic dignitaries, while artistic reliefs in the city show gods making dedications to human benefactors. At Didyma, the prophets’ inscriptions show something similar, Apollo’s good references for his yearly prophets.²³ In the 60s A.D., he bore “witness to the holy soul” of Philodemus, who was the only man to have held the job of prophet and crown-bearer in one and the same year: Philodemus inscribed this honour above carvings of his four crowns of office. In the Imperial period, he “greeted” his prophet Posidonius, who had been chosen by lot “three times over,” perhaps for three different jobs in one year. “Eternity will not forget his glory…” said the accompanying verses, perhaps words of Apollo himself. To the glory of another prophet, said his verse inscription, the “witness” was not a man, but a god: he was “cared for by divine Apollo.” In the mid-third century, Apollo “testified” to the noble forbears of Macer the musician, perhaps in a “testimonial” to his mother, herself a priestess and a cult official: we can follow this “nobility” across four generations from prophet to prophet. His cousin, the prophet Flavius Ulpianus, went one better: the god, said his inscription, bore witness to him “often.” Beneath, a verse text praised Ulpianus’s care for “fleecy sheep” and remarked that Apollo would therefore reply “in turn” to his questions. The answer went on to commend Ulpianus’s sacrificial activity. This “testimonial” was a unity, but it has not been rightly understood. Apollo was not complimenting this great prophetic family, the dynasty of Poplas and others, on its sheep farming and share in the trade of Miletus’s wool. The “fleecy sheep” belonged with Ulpianus’s sacrifices, the very “hecatombs,” we have suspected, which Macer’s Apollo then decried.

Like the oracles on tenure and selection, these testimonials have taken us into the highest social circles at the shrine. And yet, there was no Apollo: in all sincerity, the prophets were congratulating themselves. A public memorial was part of a prophet’s reward: from the walls of the Prophet’s House, we have a long range of inscriptions which commemorate the “pious” prophet of each year. Sometimes, they go further: around 200 A.D., Andreas, son of a famous prophet and civic figure, recorded his ancestry and previous offices and his “doing all else that I could with moderation.”²⁴ Beneath, a set of verses asked Apollo the king to “look on him with gentle eyes.” Did Apollo “testify” to Andreas, too, although his inscription did not say so explicitly? We cannot be sure, but we can say something about the manner and occasion of one type of “testimonial” for those who received it.

In the early to mid-third century, the god gave a very special honour to Ulpius Athenagoras, a prophet and civic magistrate.²⁵ He gave him an oracle “automatically,” in the Greek word automatizein, which meant to “act or speak without prompting.” Spontaneous oracles of this type had not surfaced in our inscriptional evidence since the Delphic legends of the seventh century B.C. One day, off his own bat, Apollo applauded this prophet who had also volunteered for a high civic office in the same year. The event, then, was exceptional, suggesting that Apollo’s “witness” was not usually spontaneous but that it resulted from a specific question. The wording of Posidonius’s testimonial implies that it was given at the prophet’s inauguration. The prophet, it seems, put on his robes and his simple crown and prepared for the heavy expense of office, standing before his contemporaries. If they put a question to the god – “are you pleased,” perhaps, “with our chosen prophet?” – he would then bear “witness” to his merits. These occasions were not the only time for “testimonials”: Flavius Ulpianus had received them “often.” But they are the ones we know best.

Like Apollo’s advice on tenure, these “testimonies” helped traditional pagan religion to run smoothly. They confirmed the choice of a prophet and also gave an honorary return to the very people who were spending great fortunes on the burdens of office. Apollo’s words could also be conciliatory in a wider sense. At Didyma, we have seen him steering a rivalry between sacrifices and ancient songs, conducted, most probably, by cousins within one dynasty. From Delphi, at an identical date, we have recently been made aware of a text for a similar occasion.²⁶ In the mid-third century, the city of Side, in southwestern Asia, received a smart new podium of stone, the site, most probably, of musical contests in the city’s “Pythian” games. In addition to some artistic reliefs, it bore two sets of four hexameter verses which were each addressed to an individual. They gave moral advice on the practice of modesty and approved their addressee’s “moderate thoughts.” The podium was dedicated by two city councillors, father and son, and the evidence of the “Pythian” festival, of Side’s cults and her civic coinage has allowed the referral of these poems to Delphic Apollo himself. Not only are they remarkable evidence for Delphi’s clientele and poetic abilities as late as the 250s, when the shrine has often been thought to have fallen silent. The verses also praise the Apolline virtue of “moderation” in two city councillors, contrasting it with the “glittering robes” which were the dress of a high official, probably the agonothete, who took care of athletic festivals. It is perhaps too adventurous to see an outright quarrel behind these verses, as if the two city councillors had donated a fine monument for the festival, thereby annoying the senior official who would otherwise run the show. Like Didyma’s text on music, the lines do endorse one party’s innovation, but they combine this favour with generalized words of approval, reminding benefactors that they please the gods by their hearts, not the trappings of their office.

It is a small step from these questions on cults and their staff to questions on possible human trespass into objects and areas which concerned the gods. For a start, oracles could identify unnatural oddities, the sort of discoveries which Romans discussed and sometimes agreed to declare a “portent.”²⁷ Between c. 110 and 115 A.D., the people of Perinthus visited Claros because they had found something odd in their territory. Their inscription is not specific, but it may have been an image of a god lost long ago in some distant spot: they “set it up” after their discovery. Like the miraculous icons which are “found” in the country districts of Catholic Europe, these lost statues of gods roused religious concern. To solve the problem, the people of Perinthus sent some very high-ranking citizens to Claros, leading nine little choirboys, six of whom were their own sons. At Didyma, c. 200, a very fragmentary text tells of problems caused by another such “image” and mentions the role of a dream and the need to “propitiate” the result. One of the most famous instances enjoyed a new public life in the second century A.D. Back in the Hellenistic age, the people of Magnesia, near Miletus, had sent to Apollo at Delphi for advice on an image of Dionysus which they had found one day in a thicket. Apollo told them to send for some Maenads and start a cult of the god in their city, among its “great and mighty people.” Good oracles never died, least of all when they honoured a cult and flattered a city’s virtues. In the second century A.D., some four centuries after this text’s delivery, a keen worshipper of Dionysus inscribed it again and displayed it on a pillar in Magnesia.

Lost images were only one of many such problems.²⁸ When an unnamed Roman Emperor ordered the widening of the Syrian river Orontes on its course to Antioch, the workers unearthed the bones of a giant skeleton. Off went envoys to Claros, whose Apollo pronounced the body to be the hero Orontes “of Indian descent.” Philostratus later relished this story: it took a god to make sense of a dinosaur’s bones. Apollo also coped with oddities from the sea: at Didyma, he upheld the existence of Tritons, “creatures of Poseidon, a watery portent,” and his text was cited with respect by a pious man of letters in the later second century A.D. We can see, once again, how Alexandra the priestess had turned naturally to Apollo when beset by that supreme curiosity, the visits of the gods in dreams.

These problems arose from the trespass of men into the sphere of the gods. It is in this light that we can understand another type of trespass, feared by building contractors in Miletus, as we learn from their publicly inscribed oracle, which they set beside an upper flight of steps on the back wall of Miletus’s theatre.²⁹ The building contractors for a part of the building had asked the god “whether they should carry out the vaults and arches” and “whether they should consider work given by their native city or work of some other kind.” The question has been read as an episode in a building strike: could Apollo ever have been more useful than when letting two sides agree without losing face? His involvement, however, had another cause. The work had been under the supervision of one of his prophets, but the man had recently died. Not unnaturally, the workers turned to Apollo to discover whether to go ahead with his prophet’s plans. Apollo answered in the style of an “expert” in arbitration and conciliation. He contented himself with verbose generalities on the value of brains and brawn, the need for a good architect and sacrifices to Athena and muscular Hercules. The inquiry probably belongs in the reign of Hadrian when major work on Miletus’s theatre was taking place.

As this answer shows, there was scope for advice from the god in almost any personal enterprise: clients might wish to be told which god they should honour in order to have the best hope of success. Divination would not tell them, whereas an inspired oracle would, and it is in this light that we can appreciate a marvellous question and answer, given c. 150–200 A.D., which stood on the walls of Miletus’s temple of Serapis.³⁰ “Apphion, also known as Heronas, from Alexandria” asked Apollo about his prospects in the arena. He was beautifully aware of the gods’ constant assistance in his life: he was a theatrical “star” and his ancestral gods, he said, “stood by” him, “as you do too, master, in whatever business he undertakes.” His question was this: would he “acquit himself gloriously at all times in his act of dancing on tiptoe and training bulls and will he be serving [Apollo?] with a fair name?” Apollo told him to pray to various gods and enjoy them as his “protectors”: Apphion, it seems, was a bull tamer who danced on the bulls’ bare backs. He needed the gods’ reassurance on the “glory” of his act, and when he received it, he was so proud that he inscribed it prominently on the shrine of Serapis, the god whose “swift eye” Apollo told him, once again, to “propitiate.” The text has the clumsy style and construction which we would expect from a bull tamer using his own words. Once again, it was not enough to receive an oracle which praised a questioner, a place or a cult. It was as well to inscribe it for others to read and admire for ever: did the priests, perhaps, of Serapis encourage the carving of this pleasant text? Oracles profited from that general connection in antiquity whereby the texts which were inscribed tended to be texts which reflected well on one party, or very badly on another.

These and other questions are very lively, but they are a far cry from the free Greek past, when Delphi’s Cretan priests had helped Sparta with her constitution and were said to have encouraged the first Greek tyrants. At Delphi, Plutarch’s debaters looked wistfully back to the old days when oracles had settled foreign policy and advised their clients on wars. Yet there was still scope for important business, and Plutarch was being too negative. At Didyma, Poplas the prophet asked how best to approach the Emperor for financial aid: in Athens, we have an oracle from Asclepius which was inscribed on a small cult table and told one of his worshippers how best to approach the governor of Achaea and to intercede with him for favours.³¹ We can identify the man as the governor in 209/10 A.D. and see how the gods’ horizons had moved with the times. No longer the crisis of war, but the problem of access to Roman officials, prompted Greek questions to the gods. What, though, of questions from and about the Emperors themselves?

Here, we enter a dangerous, but tantalizing, field. The establishment of the Emperor’s rule at Rome had given a new focus for divination and “information technology”: power now centred on an individual, and there was special profit, therefore, in knowing the course of an individual’s career. At Rome, the older style of state divination had declined with the new requirements of a monarchy:³² in the Greek world, kingship was no novelty, but we hear rather more of oracles for present and future rulers, not least because Roman Emperors ran in less stable dynasties and there was greater scope for an outsider, or a sudden change. At Didyma, Apollo was believed to have given a favourable response to the young Trajan; certainly, Trajan favoured the shrine conspicuously when he became Emperor. According to Dio, the temple of Zeus Belus at Apamea was believed to have uttered some robust Homeric lines to the young Severus when he consulted him while a private citizen. Perhaps Trajan heard something similar at Didyma. Out of context, a set of Homeric verses could be very ambiguous.³³

Publicly, the establishment of new dynasties was indeed linked with signs and reports of approval from the gods, a pattern which runs from Vespasian to the fourth century A.D.: in this light, Constantine’s favour for Christ was only one episode in the long history of usurpers’ religious publicity. However, neither we nor contemporaries could distinguish clearly between real consultations made before the event and stories which an Emperor or his supporters liked to circulate afterwards. Inquiry about an Emperor’s future remained a treasonable offence, punished by severe penalties. The repeated history of laws against the practice suggests that it retained its attractions, but the inquiries, surely, were best conducted in private, not at a shrine with thespodes, prophets and secretaries.³⁴ According to Dio, the shrine of Zeus Belus at Apamea gave an elderly pretender in the Severan period some different Homeric lines, contrasting an old and weary ruler with a victorious new warrior. Zeus Belus was not such a fund of Homeric jokes, especially as the same three lines are later said to have been given to an unfortunate rival of Constantine. They were obviously a witty invention, an art from which Severus later suffered. After his accession, he was said to have consulted Zeus Belus again, only to be answered with a line from Euripides: “Your entire household shall pass away in blood.” Had this line ever been uttered, it was a shrewd bet on any Imperial dynasty, but the story, surely, was a fiction, answering the claim that Zeus Belus had favoured Severus’s appointment.

However, the succession and prospects for it were not the only ambiguous matters of state: what of existing Emperors and their prospects in war? Here, despite Plutarch, there was an old and continuing field for prophecy.³⁵ In the Imperial period, warring rulers asked the shrines for advice; in the Christian Empire, they would ask Christian “holy men.” Trajan is said to have consulted Zeus at Baalbek about his prospects in war in the East and to have mistaken a hint of his death in the answer. The consultation, at least, was plausible: throughout antiquity, generals needed to know which gods to honour in order to fare “as well as possible.” There may also be truth in a story that the Emperor Cams, in the 280s, marched East against Persia with an oracle’s support. Once again, Julian was heir to a long tradition when his Eastern campaign in the 360s was encouraged by words from the gods.

Was there scope, perhaps, for those who feared an Emperor, as well as for those who wished to be Emperor themselves? Here, questions were particularly dangerous, but not every oracular shrine was public, and once, at a private temple, we need have no doubt about this type of consultation.³⁶ We know it from Dio’s own researches into the history of Condianus, a senator who had suffered the worst excesses of Imperial tyranny. During the 180s, the Emperor Commodus arranged the murders of Condianus’s elderly, distinguished uncle and father and the death, too, of his brother. Condianus escaped by a trick and wandered as a marked man through the Empire. On his way, he consulted the hero Amphilochus’s dream-oracle at Mallos and left a drawing of his dream on a tablet at the shrine, which showed a child strangling two serpents and a lion pursuing a fawn. We know about this tablet because Dio himself saw it as a young man when his father was governing Cilicia in the 180s. Condianus, it seems, had enjoyed an apt, if ambiguous, dream, which, as Dio discovered, could be referred to Hercules, a god whom Commodus especially honoured. Just as Hercules had strangled two serpents, so Commodus had already strangled Condianus’s father and uncle and was hounding Condianus, the innocent “fawn”. In a dream-oracle, the initiative lay not with a prophet but with the client’s own imagination. The “meaning,” therefore, could be relevant and more explicit: the tablet is an excellent insight into the terrors of Imperial rule and their impact on an individual’s mind.

It was not, then, entirely just of Plutarch to imply that oracles’ scope had shrunk with the passing of free political life. Monarchy created its own, momentous demands. In general, its questions are not likely to have differed from the type which we find inscribed by other individuals: which gods should be honoured and how best should the honours be paid for matters to go as well as possible? These questions did not only attach to new ventures, a war or the gift of a podium, a bareback dance on a bull or the building of a theatre. They attached to existing doubts and problems, tenure for life, the finding of statues and heroes’ bones, a sudden deluge of dreams of the gods. These questions were unified by one constant presence: awareness of the gods’ potential anger, fear of it and a wish to “placate” it and avert it by correct performance. The newly found oracle from Ammon in Libya begins with a significant reassurance: “do not be afraid, Cyzicus,” if your citizens honour the god in this particular way.³⁷ Ultimately, therefore, epiphany and oracles meet on this shared ground. It helps us to appreciate the oracles’ best-attested role: they performed a service which some have used as a key to religion’s entire function, the “explanation of misfortune.”

When two Asian cities feared raiding barbarians, we have seen how Apollo at Claros equated their troubles with the god Ares and advised them to bind his statue in order to ward them off. This belief in the power of statues has a fascinating counterpart in the higher theology which Claros preached. We have seen, too, how Apollo at Didyma explained the supreme misfortune of the group of woodcutters found dead in a field: Pan, he said, had been responsible for this “chain-saw massacre,” and but for his enemy, Artemis, it would have been very much worse. It is not, then, surprising that less visible enemies were also within the oracles’ competence.³⁸ On the very Acropolis of Athens, beside the Parthenon, stood an image of mother Earth emerging from her block of stone and raising her hands in prayer: Pausanias, who saw it, explained that she was praying to Zeus to rain on her in a time of drought. The inscription on the rock has been dated to the Hadrianic age, c. 120 A.D., and it ascribes the image to an oracle’s command. At a similar date, or perhaps slightly earlier, we have an inscribed question to the oracle of Apollo at Gryneum in Asia Minor, asking about the proper honours to assist the crops. At Nicomedia, in Bithynia (northwestern Turkey), it is evidently to Claros that we should trace a verse text which advises honours to the Moon and the Winds in a time of drought. Here, the god addressed a “ruler of the city”: as we shall see, the great crises of the age caused cities to send very prominent citizens as envoys to Apollo’s shrines. These envoys belonged to families whom we know in happier times for their buildings and speeches and gifts of theatrical shows. In hard times, they feared the gods’ evident anger: they were also people who presided over the persecutions of Christians.

Terrors had always made excellent business for Apollo, and there were never worse terrors than the plagues and earthquakes in the Antonine and Severan age. At Didyma, Apollo’s cult was connected by its very origins to the banishment of plague: in the 250s, Apollo still recalled how he had “shamed” the threads of the Fates and kept off the epidemic. The shrine must have been consulted in these later crises, but the visitors happen not to have recorded the results on stone. Instead, we know about Apollo’s remedies in a fascinating selection of oracles, most of which derive from Claros’s client cities. They carry no date, but the main four probably belong in the great epidemic of the 160s when Roman troops were returning from the East and brought home a disease whose effects read horribly like smallpox.³⁹ The oracles, said observers, were packed with frightened clients, especially Claros, and in these years, our inscriptions by visiting cities include several distant cities who never inscribe their names at Claros again. At Odessos in Thrace, we find a man who had been on an embassy to Apollo and had “driven away disease” from his fellow citizens. Here, it seems, the oracular remedy worked. In Thrace, near modern Edirne, we learn of a group of statues which four women presented “in accordance with the oracles of Clarian Apollo” and which may have stood on a civic fountain to watch over the purity of the water. The most thorough instructions, however, survive elsewhere.

Three cities have left us pieces of Clarian Apollo’s oracles on plague: Pergamum, Callipolis (near modern Gallipoli) and humble little Caesarea Trochetta, a small town in Lydia. The latter two were treated to a metrical and linguistic tour de force.⁴⁰ The god pitched his style very high and lamented the disaster (crying “Woe! Woe!” to each city) while changing from hexameters to three different types of metre in his long, emotional answers. None of these texts explained the plague; they only prescribed remedies. The Caesareans received a vivid verse picture of the plague “brandishing in one hand a sword of vengeance and in the other raising the deeply mournful ghosts” (probably) “of newly stricken mortals.” Only part of the god’s remedy survives, but it runs true to his old-fashioned manner. The “divine law,” he said, required his clients to draw pure water from seven fountains, which they had fumigated carefully. They must then sprinkle their houses with these “nymphs who have become kindly” and must set up an image of Apollo the archer, bow in hand, in the middle of their plain. There, presumably, he would “shoot away” the invading enemy, the plague itself.

Callipolis was ordered to set up a similar statue, “the warder-off of plague,” and was also told to offer blood from sacrificial animals to the “gods below the earth” and to burn all the animals’ meat with spices.⁴¹ The pyre of this holocaust was to be sprinkled with “shining wine and grey sea water”: the victims, a goat and a sheep, must be black. No doubt Caesarea had had similar orders for a sacrifice, but they are missing from the surviving verses.

Of the three texts, the finest is Pergamum’s, which moves in stately hexameters and spends the first nine lines on flattery of the citizens’ ancestry, their closeness to the gods and their especial honour from Zeus himself.⁴² On Pergamum’s steep hill, said Apollo, the infant Zeus had been placed just after childbirth: his statement refuted a host of competing cities which claimed that they, not Pergamum, had received the newly born god. It was no wonder that the people and council of Pergamum decreed that the reply should be inscribed on pillars and displayed “on the agora and the temples.” It also offered advice. Apollo wished to please his son, Asclepius, who resided so conspicuously in the city, and so he told the leader of Pergamum’s delegation to return and divide the city’s youth, or ephebes, into four separate groups. He called them “wearers of the chlamys,” or the military cloak which alluded to their honourable Macedonian origin, and he told each of their groups to sing a hymn to a particular god while their fellow citizens feasted and sacrificed in support. This festivity was to last for seven days, a point which makes Pergamum the most favoured of Claros’s three known consultants. Each libation was to be joined by prayers that the plague might depart “to the distant land of far-off enemies,” perhaps to Parthia, whence it came. The young men’s hymn to Zeus happens to survive in Pergamum, a splendid composition which calls on Zeus to “come propitiously” and honour the city with his presence. His companions were to be gods of civic and social concord, while he himself was hailed with high honours. He was the “dweller on the heights of the Titans,” the realm of the Sun, and was the ultimate master of the crops and seasons. Claros had probably prescribed the exact words of this solemn, archaic hymn in hexameters, a classic text of belief in the gods’ epiphanies.

These remedies were quite useless. They were bad news for their animal victims, better news for sculptors of Apolline statues. They did not, however, impair trust in the god. Like a shrewd prophet, Apollo predicted no easy result: if there was a second outbreak, he told Callipolis, he would promptly send another remedy. The course of divine anger was not predictable, yet in the end, Apollo would be proved right. Eventually, the plague would stop, and if it stopped slowly, the anger of the gods was to blame, not the quality of Apollo’s advice. The survivors, finally, would give him credit, while the dead could express no dissenting opinion.

Although none of these texts gives an explanation of the plague’s arrival, they each remind us how polytheism could account for any outcome: disasters could be blamed on the anger of one of its many gods. For a clear statement of this point, we can look, lastly, to oracles recently found at Hierapolis, in southwestern Turkey. A leading citizen brought them back from Apollo and inscribed them on blocks of stone in his city (the modern Pamukkale). The plague, they explain, was due to the anger of mother Earth at the slaying of her child, Python.⁴³ This explanation is of great interest. It referred the present calamity to a distant episode of myth, a thousand years and more in the Greeks’ prehistory. The “angry displeasure of the gods,” said the text, was paining “many cities and peoples,” and the remedy was to offer “libations and perfect hecatombs” to escape their “generous anger.” The people must kill appropriate victims for Earth, Ether and the gods of the underworld, and around all the city gates, they must set images of Clarian Apollo, “shooting from afar with his holy arrows.” When the “evil powers have been appeased,” the god commanded, “boys and girl musicians [should] go together to Claros, with hecatombs and glad libations.” “Often,” he complained, he had saved the city, but he had not received his share of a meat sacrifice. If the citizens did what he said, in future they would be richer and safer, because they were descended from the god, “through the hero Mopsus and Apollo Kareios.”

This text is very illuminating. Once again, said the god, the plague was best defeated by images of archer Apollo and by sacrifices to the elements, of which Earth’s huge holocaust is described in puzzling, archaic language. Once again, there was a touch of civic flattery: Hierapolis, said Apollo, was descended from a local Apollo and the hero Mopsus, a legend which we can trace on the types of the city’s coinage in the second century A.D.⁴⁴ The text goes on to show us how a delegation to Claros could come about. The god himself had prescribed it, in exactly the form which the temple inscriptions at Claros then show us: similar instructions, we now know, were given by Ammon of Libya to his worshippers in Cyzicus, a city which we would otherwise connect with Delphi and, in one case, Didyma. Claros benefited from the advice of other gods, and at Hierapolis, we may note, Apollo only ordered attendance when the plague had passed. Did he, perhaps, fear infection himself? Not speaking, as yet, for Macer, he insisted that his clients offer him meat: hymns and prayers were not his exclusive taste, and the gods were also to have their hecatombs.

The origin of this oracle is not made plain to us, but it is not the advice of the local Apollo Kareios. On one view, it could be ascribed to the Clarian god, but it plainly distinguishes the Apollo who is speaking from the Clarian Apollo whose images are mentioned in the third person.⁴⁵ Hierapolis is not present among the consultants at Claros whose lists have so far been published, and the city looked elsewhere for its presiding Apollo. It was an early colony of the Seleucid kings, and like other such colonies, it was connected with Pythian Apollo at Delphi. Cults and games of Pythian Apollo were very prominent in the city, while a recently found relief in the city’s theatre shows the town of Hierapolis holding Pythian Apollo himself. The oracle was inscribed by an individual “at the command of Apollo Archagetes,” or “founder and leader.”⁴⁶ At Hierapolis, this title was particularly prominent for Apollo on coins and in civic inscriptions: as elsewhere, it points ultimately to Apollo at Delphi, the great colonizing god. There are two alternatives: either the oracle was given at a local shrine of Apollo the Founder or else it was sought from Delphi herself.

The local alternative has met with general support: the city did have a local oracular temple whose verses were said by Livy (c. 10 B.C.) to be “not unpolished,” and like its fellow colony Laodicea, Hierapolis would have had local prophets in its citizenry.⁴⁷ The existence of a local oracle of Apollo did not stop Laodicea from sending choirs constantly to Claros: perhaps the local god, as in our text, commanded them to go. The god does say, “Attend to Apollo Kareios always: for you are my descendants.” However, we might have expected him to say, “Attend to me,” if Kareios was speaking: the text was from “Apollo the Founder,” not Kareios, according to its preface; there is scope, still, for looking to Delphi instead. We know that another Delphic oracle was sought by the nearby city of Tralles in the Imperial period. It was “given” to a “priest,” a member, once again, of a very grand family in the city.⁴⁸ As it is a verse text, it must belong after Plutarch, and its recipient’s name fits the second century A.D. It ascribed an earthquake to the “thousand-year wrath” of the gods and advised a cult and honours for Poseidon, the god of tremors. Once again, it prescribed the exact adjectives by which a god should be invoked.

At Hierapolis, too, the god spoke of “grievous anger” and “wrath” in heaven. The explanation was similar to Delphi’s, and so, above all, was the accompanying myth.⁴⁹ It was at Delphi that Apollo’s arrows had slain the serpent Python, the event which Earth was “resenting.” Exactly this scene appears on two splendid coin types from Hierapolis, showing Apollo shooting the serpent Python with his bow. They are unique types in all Asia, and the earlier of the two belongs in the co-reign of Marcus and Verus, precisely the time of the great Asian plague. Hierapolis survived and proudly showed the god’s words on her coinage. The scene is Delphic through and through: it has no connection with Kareios’s myth and attributes. At Delphi, Apollo was “founder and leader” of the city, her general “ancestor,” as the god said. It is quite likely that the oracle, with its Delphic explanation, derived from Delphic Apollo in person.

If so, we have found a great rarity, one of the only verse texts from Delphi in the Imperial age. It ought to belong before the earlier third century, because of the coin types and the treatment of the text’s stones: when another shrine of Apollo was rebuilt in the city, they were taken and recut for use in it.⁵⁰ This rebuilding can hardly be later than c. 230, suiting a date for the text in the 160s: builders in Hierapolis were less respectful of their old oracle texts than those in cities elsewhere. To Christians, Hierapolis was the city of St. Philip and his seven prophetic daughters: the recent finds show how rich it also was in inscribed pagan oracles. It had an alphabet oracle, we also know, which gave neat proverbial answers by the usual method; it had these fine Delphic texts and several others known as yet in fragments only, which had been sought and inscribed from the gods. They conform to the pattern we have come to expect.⁵¹ In one, a god laid down the rights of his spokesmen to receive parts of sacrificial meat, and in another, a god approved the wisdom of “nobly born” people, perhaps “kings,” and commented on their benefits from wise second thoughts and a sensible way of government. A fourth response addressed the city as a whole and referred to its “questioning” about “the waters.” The god rebuked any person, on the likeliest rendering, who “seems to themselves to be wiser than the immortals,” and then went on to give advice about the “waters” and the land. Perhaps the city had consulted him in a time of drought.

These recent additions to oracular wisdom in the Imperial age are not unrepresentative. They praised the governors of a city with moral generalities which kept them up to the mark; they advised on details of cult which could cause dispute; they explained and gave cures for drought and plague; meanwhile, the alphabet text offered quick responses to those who wanted instant proverbial wisdom. These further texts’ full wording and origins are not certain, nor do we yet know exactly where and when they were first displayed in the city, but their general tone conforms to other oracular business in the second and early third centuries, adding a fine poetic flourish from Delphi. Oracles told men which gods to honour in order to prosper and which to placate in order to avert misfortune. They did not blame misfortune on mortals’ sins. They named supernatural culprits, and traced their actions to enmities in heaven. Artemis was hostile to Pan, Earth to Apollo, virgin Athena to loving Aphrodite: disruptive barbarians were allotted to the god Ares, plague to the “threads” of the Fates.⁵² In a splendid verse text, recorded as given from Baalbek, Zeus is said to have blamed Poseidon for smashing his temple’s columns. He told a third party, probably the Sun god at the shrine, to threaten Poseidon’s rule over the sea and remind him: “It is wise, Poseidon, to listen to elder brothers…” Poseidon, it seems, had shattered the pillars in an earthquake.⁵³ Even if the text was a fake, it seemed plausible because it made one god pin blame on another, like oracle texts elsewhere.

We can see, then, how inspired oracles had continued to find a market. They belonged with the themes we have studied in the previous two chapters: cults of continuing vitality and the continuing belief in the anger and presence of the gods. Everybody could use them, benefactors, country people or ambitious stars of the arena. They had survived because they were useful, protecting their cities (“often,” as Apollo told Hierapolis),⁵⁴ “propitiating” the gods constantly and helping their cults to move correctly with the changing times. To us, many of these questions can seem overscrupulous, but they show how the small details of practice still mattered to men in the cities’ upper classes, men like the enthusiast Damianos, who called himself “lover of the gods.” Because the gods were “present” and manifest, it was necessary to ask them about changes in cults and about discoveries which might concern them. Otherwise, they might be “unpropitious” and “stand by” men in an angry mood. The old compound of awe and intimacy was still alive.

These questions were not born in a static, indifferent age. Mostly, they arose from the central problem of religious innovation, on which there is a clear division between the Greek and the Roman world. At Rome, innovations were generally supervised by the board of “fifteen men” who had access to the collections of Sibylline books.⁵⁵ These old oracular sayings were consulted for advice on new cults or changes in existing worship, and often their meaning relied on their human consultants’ judgement: in principle, they were a closed “canon” of divine wisdom. In the Greek world, however, these questions were put directly to the gods.

What style, exactly, was thought to befit the gods’ replies? Around the year 120, Plutarch’s debaters at Delphi had lamented the disappearance of oracle texts in verse.⁵⁶ Their absence, they said, had particularly undermined faith in oracular sites. Plutarch also admitted that people objected to the current texts’ “extreme simplicity,” which undermined belief in inspiration. “People,” said his final speaker, “yearn for riddles and allegories and metaphors in prophecy, much as children prefer rainbows and comets to an honest glimpse of the sun and moon.” The age of poetry, he felt, was over and the routine questions of his age no longer invited ambiguity.

Plutarch’s general view was shortsighted. At Didyma, verse oracles seem to have been given in the first century A.D.; at Claros, before the 130s, the “prophet” did give verse answers, although they were probably standardized. After Plutarch’s death, long and elaborate verses sprang back into prominence at the great shrines, at Ammon’s oasis in Siwah (as we have just discovered) and even at Delphi, as we see in her texts for Tralles and Hierapolis, Side and the dead Plotinus. In Plutarch’s day, the Delphic prophetess was a simple country girl, yet by c. 200 A.D., she is known to have been a very well-born lady.⁵⁷ Had this change helped the quality of the oracles? At Claros, the lifelong thespodes clearly made an improvement: perhaps Delphi’s return to verse owes more than we know to the rise of this better class of prophetess.

While Plutarch wrote, the dice and alphabet oracles were already giving puzzling answers, again in smooth hexameter verse. Although they were not quite a rainbow, they were decidedly more complex than a plain, clear sky. “From dusky night there once appeared a ray bringing light…” “Realize the coming purging of body and soul…” When an oracle had to answer general questions in prearranged answers, we can well understand why it used this enigmatic style. To judge from Plutarch, the style would not affect many clients’ belief in the system.

At the shrines, however, Plutarch was correct on one cardinal point: at this date, the inscribed oracular texts do not use ambiguity. In Plutarch’s view, the reason for its decline was, ultimately, political. People no longer asked great questions of state, and on topics like plagues and harvests it was enough for a god to give simple, general replies. The stories of questions by future Emperors took a different line, but perhaps, as we concluded, they were apocryphal. Was Plutarch’s reason really the sole one?

Ambiguity fitted neatly with an age of deep respect for the gods and the limits of their gifts. The gods revealed only so much, and no more: their text then threw the final responsibility back onto man’s intelligence, giving him signs with which he could later share the blame, while essentially blaming himself. In the Imperial period, this respect for the gods was still deep and intact: why, then, was ambiguity rare in its surviving oracles? It was not easily contrived, and even in the age of Herodotus, we can overestimate its use. We have not seen questions which asked inspired oracles to predict the future in any detail: to our eye, the gods were asked questions in forms or on topics in which they could not be proved wrong by events. Even in the classical period, the majority of “ambiguous” oracles may have been fictions, invented or elaborated after the event. Then, too, oracles were not generally a source of detailed prediction, open to disproof. Plutarch, perhaps, was misled by these past fictions and gave a narrow political reason for a feature which had always lain in oracles’ very nature. It helps us to see why other types of prophecy flourished. For a true “prediction,” people would look to one of the so-called Sibylline texts, where warnings of ruin and disaster sheltered under the pretensions of age and fictitious authority. Inspired oracles gave advice on “appeasing” gods, a slow, uncertain business: to use the gods and to profit more immediately from their attendance, people turned to sorcery and spells. Some of the oracles’ business did overlap with the diviners’ and astrologers’ questions: problems of marriage, theft and business were also submitted at their sites, though probably only for answer by drawing lots. The claims of “all-knowing” street prophets and sorcerers were much more extreme than the distinctive core of oracular business, the guarded wisdom which the gods inspired people to speak or dream. Only the oracles had the prestige of official consultation by cities, a point which continued to mark them off from rivals, but their inspired answers were usually quite general.

Archaism and obscurity, not ambiguity, typified the shrines’ oracular verse. They always had typified it, as we can see from Aristophanes’s brilliant parodies performed in the late fifth century B.C.⁵⁸ By the Imperial period, however, these qualities were a pronounced fashion in mortals’ learning and literary style: the language of the gods now magnified a fashion among mortal contemporaries. The gods spoke in archaic phrases, straining for rare words and coining plausible Homeric forms.⁵⁹ Like philosophers, they found meaning in wordplay on the gods’ names and qualities: in the second and third centuries, puns were an art of gods as well as men. To their clients, poetry seemed most inspired when it was most lofty and verbose, a cloud, if not exactly a rainbow. Underneath, however, the message was straightforward: “I bid” you to choose or honour this priestess or that god. In their directness, oracles thus differed from their obvious near neighbours, the messages of gods in dreams. At night, it was only to be expected, said Artemidorus, that the gods should speak indirectly to man and that their veiled meanings should need interpretation.⁶⁰ At Didyma, Stratonicus actually asked Apollo what a dream meant.

Ultimately, Apollo could not help reflecting the prevailing atmosphere. As Yeats came to realize from Mrs. Yeats’s automatic writing, there is a necessary connection between human input and divine output. Apollo knew his Homer and his school philosophy; in an age of keen civic rivalry, he knew his local history and gave clients the compliments which their cities liked to hear. In this taste, too, Apollo was the mirror of his times. He respected cities’ ancestry and prehistory in myth, and sometimes he warmed to these themes with a keenness which exceeded his orders: the men of Pergamum received nine lines of antiquarian allusions to their favour from Zeus and their city’s noble pedigree, “the Telephids.” Hierapolis was reminded of its ancestors Mopsus and the nearby Apollo Kareios, and the oracle inspired the city’s subsequent choice of coin types. At Didyma, Apollo honoured the people of Miletus as the “sons of Neleus who fell prey to an arrow.” These allusions were wonderfully rare and delighted their clients in an age which valued these ancient myths of origins and honorific tales of city foundations. Apollo could take his place beside the most ingenious of the orators who toured and flattered the cities: his diction and wordplay on place names were shared, too, by Sibylline prophecies, composed in the same period. These long verse fictions would have pleased any Clarian thespode, and their compositions remind us how widely these talents were practised among minor men of letters.⁶¹ At Claros, meanwhile, those cities that lacked a cultural history were less fortunate. Caesarea Trochetta could just pass as “Ionian,” but Syedra in Cilicia was more problematic and Apollo escaped by resorting to puns. The people, he said, were Pam-phylians (of all tribes) and therefore pam-migeis (all of a mixture).⁶² Their many component peoples lacked an obvious connection with Greek myth, and Apollo honoured them with a play on words. What, one wonders, had he made of the Oenoandans, whose people combined Lycians, Greeks and Pisidians in a single city?

Who, then, was this Apollo, master of so many complex puns and metres, and gifted with such antiquarian facts about the cities which gratefully inscribed his words? Deep below Claros’s Doric temple sat a thespode who listened to the prophet’s cry. Throughout the second century, he put its meaning into verses and helped the god to express his view. We know the thespodes from inscriptions, where we can follow their infrequent changes after many years in office.⁶³ In the years from the late 130s to the 140s, Apollo was explained by a very solemn figure, Claudius Asclepides, who called himself a Heraclid, no less, “descended from Ardys” through his father’s family. His name looked back to the old Lydian kings, some nine centuries earlier, the predecessor of Croesus, whom Apollo had vainly tried to save. From the 150s to the mid-160s, Regginus Alexander; from 177/8 to 185/6, the job passed back to a Heraclid, Tiberius Claudius Ardys. Did these Heraclids first come from Lydia to Claros in the 130s as the new thespodes and had prophecy always run in their family? It is a fine picture, a thespode who traced his family back to the Lydian kings, advising cities on plagues and drought, sending “interpretations” to Hadrian’s Wall, paying fulsome compliments to “noble Pergamum” and telling the men of Syedra they were “all mixed up.” Who better to know these antiquarian details than poets who believed they were in the Heraclid line? Just as Homer’s gods were still “standing beside” men as “evident helpers,” so the gods were still speaking on the topics of religious practice, correct “propitiation” and the right course to follow in moments of anger of the gods. The age of Herodotus was not yet dead.

IV

The language and business of these oracles allow us, finally, to appreciate the most notorious cult of their age. While Claros and Didyma flourished and the gods still spoke at lesser shrines, they were joined by a prophet whose shrine has seemed wholly preposterous. Emerging in the 150s, it forms a serpentine tailpiece to the age’s oracular culture and says something, too, about the nature of contemporary pagan cult. It arose in Paphlagonia, a place which was way beneath a Heraclid’s dignity. It centred on a man and his model snake.

On the southern coast of the Black Sea, the Paphlagonian town of Abonouteichos had sat obscurely beside its poor, unwelcoming harbour. It had attracted no comment in guidebooks to the coast, except that its anchorage was unpredictable. Visitors relied on the prevailing west wind to take them safely round the capes and small seaside settlements which marked the adjoining stretch of coast: not for nothing did the West Wind continue to appear on Abonouteichos’s coins. Access by road was no easier. A Roman road had begun to be built along the seacoast, but the stretch which connects to Abonouteichos is missing from the road tables and had probably never been built. The town’s hinterland was not easily negotiated. It soon ran out into ravines and the steep forests of the mountain range behind. A Roman road did run inland through the centre of the province, but it was cut off from Abonouteichos by the intervening forest and chains of mountains. The approach route was a local track through rough country.¹

This Paphlagonian township is a resounding proof that it took more than geography to keep a good shrine down. From around 150 to the mid-170s, people flocked to this distant point where Providence seemed to have broken afresh into the world.² Its god gave personal advice to Romans of the highest rank and sent an oracle to the Emperor himself. The cult concerned two figures, a prophet Alexander and a long, coiled snake called Glycon. Together, they rivalled Claros in their power to draw clients from far and wide. The snake Glycon was honoured in two inscriptions from Dacia, a Danubian province, while cities in Galatia and Bithynia showed his portrait on their civic coinage long after the prophet’s death.³ How could “paganism” have come to such a pass, a snake with shaggy hair and two human ears on its head? Recently, he has turned up in two new aspects. He promoted statuary and fathered children

The north and south shores of the Black Sea were linked to each other by networks of trade and exchange and also, we now know, by religious pilgrimage. On the north shore, several cities sent delegates to Claros, and before long, their citizens also called at Abonouteichos.⁴ At Tomi, a carefully buried cache of statues was recently unearthed, containing images of sixteen pagan gods. Among them was a large model of Glycon whose coils would have unravelled to a length of some fourteen feet: Glycon, as a hostile witness told us, was indeed “very bulky.” His cult, we were told, sprouted “paintings, images and carvings, some of bronze, others of silver” and others, we now know, of stone. Another model of Glycon has been reported at Dorylaeum, in Phrygia.⁵ The snake, however, did not stop at self-portraits. When Caesarea Trochetta received its oracle on plague from Claros, it inscribed it on a statue base, presumably a statue of the Apollo whom the god had told it to set up.⁶ The cost was met by the city’s priest of Apollo the Saviour, “Miletus son of Glycon, the Paphlagonian.” Glycon was a common name, but this combination is suggestive. “Miletus” was an excellent name for a man from the coast of Paphlagonia, where claims to a Milesian origin were popular in the cities. A “son of Glycon” is very tantalizing. At the new shrine in Abonouteichos, women are said to have consulted the god Glycon in the hope of conceiving children. The question, we know from Dodona, was quite frequent at oracles, but Lucian, who knew the place, alleged that the new prophet fathered children himself. Might Miletus’s mother have looked to Glycon and told the world that she had borne a child by the shrine’s oracular snake? If so, people in Caesarea believed her and made her son from Paphlagonia their local priest of Apollo. Divinity ran in his veins.

The shrine’s stamina and popularity are no longer in doubt. Not only was it a new growth which coincided with the flowering of Claros: it was a shrine on whose workings we have a sharp eyewitness account.⁷ During the 160s, probably in 165, a literary man, Lucian, travelled north from Cappadocia with his father and family, and while his family forked west, paid a visit to Abonouteichos in the sole company of his manservant. He wrote up the event some fifteen years later and hardly disguised his disgust. The prophet, he claimed, had tried to murder him. He was a dangerous fraud who conned the local “fat and uneducated,” people who had thick accents and were only good for chewing garlic. When the prophet held out his hand for Lucian to kiss it, he claimed to have seized the opportunity. To the fury of the spectators, he bit it instead.

Lucian’s sketch is a reminder that oracles still found their critics and that scepticism was still alive: as he chose to present it, the antagonism between the prophet and his disbelievers amounted almost to war. He sent his story to his friend Celsus as a warning against the tyranny of hope and fear and made such fun of the Paphlagonians and two elderly, credulous Roman governors that it is easy to share his viewpoint and write off the shrine as pure folly. Yet its clients stretched widely and its snake persisted for more than a century in other cities. How can we recover history from a satire of such venom?

Even now, his sketch is read too literally as history without due allowance for his gifts of parody and satire. We should be wary of abuse which reappears in Lucian’s other satirical pamphlets and is plainly his own invention: it tells us more about Lucian and his audience than the nature of the oracle. However, Lucian did observe details at the shrine, although he distorted them by alleging fraudulent motives. If we ignore these motives and the patchwork of his literary jokes, we are left with a hard core of behaviour which was not so very peculiar. Lucian also cited questions and answers from the oracle, and although he had an Aristophanic gift for writing hexameter parodies, his mischief is generally evident. It did not extend to every example: on one occasion, we can check him, and prove that his quotation was honest. He claimed that he had seen some of the god’s answers which a keen devotee had inscribed in gold: if he had wanted, he could have consulted such clients’ copies and noted them precisely. The line is hard to draw exactly: are we merely showing that Lucian’s parody conformed to oracular language and practice elsewhere, or can we also show that the prophet and his shrine were not uncharacteristic? It is certainly not enough to accept the idea that Alexander was a “fraud” or a hoax. He served for some thirty years, long after a hoax would have worn thin. Where we have any control, Lucian is not inventing the entire picture. Here, we can set the actions which he distorts beside their proper contemporaries, not magic or mysticism, but second-century oracles and the cults of Asclepius. If there is a key to the prophet’s success, it lies with what we have just seen.

Lucian tells a brilliant story of the beginnings of the prophet’s fraud. Alexander “the false prophet,” he said, had left home as a young man and drifted through a dubious study of medicine and sham philosophy. When his good looks faded, he fetched up with Cocconas, a wretched songwriter from Byzantium, the first known melodist, then, in that city’s musical history. Together, they tricked a rich Macedonian woman, bought a huge, tame snake at Pella and decided, after a quarrel, that Abonouteichos was the best place for a fraud. They faked oracles at Apollo’s old temple in Chalcedon and sent them to the Paphlagonians, predicting the god’s arrival. They invented prophecies to the same effect from that irrefutable fake, the Sibyl, and left them to spread among Paphlagonian high society. When the crowds in Abonouteichos heard that Asclepius was coming with his father, at once they began to build him a temple.

The earlier scenes in this story are probably pure satire: the Macedonian episode was devised to ridicule the prophet Alexander, namesake of Alexander the Great. The later details are more tantalizing. Other local cities, we know, consulted Apollo’s oracle at Chalcedon about changes in their cults, and there is a hint on Abonouteichos’s coins that her people were not totally deceived.⁸ In the reign of Antoninus, the city of Abonouteichos issued coins which showed Asclepius meeting the goddess Health. Each of the divinities carried a snake, but only later do the coins allude to the prophet’s particular snake, Glycon. In Lucian’s story, Glycon and Alexander were present from the start and the temple was hurriedly built to receive them. But the coins suggest that the city’s concern for Asclepius may have preceded the prophet’s intervention. There were springs and holy places of Asclepius in their surrounding territory and the god was already popular in many Black Sea cities. The temple, it seems, was independent of Alexander’s “fraud.”

Lucian has a fine time with the scenes of Asclepius’s arrival. The Paphlagonians gaped and their thickset jaws dropped ever further as their prophet streaked through the city and claimed that Asclepius was hatching that very moment from a well-placed egg in their temple’s foundations. The god emerged as a snake and “grew” rapidly into Glycon, the pet whom the prophet had concealed. All we can say about this lively satire is that its elements were familiar in other sites’ cults of Asclepius.⁹ The god’s “arrival” to a temple was a highly pitched affair, and several cults were said to have begun with his physical transfer to a new shrine in the form of a travelling snake. These stories were current from Mytilene to Rome’s temple of Asclepius on the Tiber: eggs, too, were honoured in his cult, and marble and granite eggs survive from his shrines in nearby Thrace. The epiphany of any new god lent itself to satire. At Abonouteichos, snakes, eggs and the prophet played a part, but Lucian no longer allows us to see the details straight.

Why, then, was Alexander given charge of the city’s shrine? Lucian claims that the idea was his confidence trick from the start. He also describes him as the pupil of a doctor who had learnt his art from Apollonius of Tyana: “you see, dear Celsus, from what sort of school the fellow came.” Is this only his malicious invention? It is a surprisingly precise claim. Among the many legends about Apollonius the wise man, it seems safe to agree that he had been a follower of Pythagoras and that many of his pupils had gathered in the Cilician temple of Asclepius at Aigai.¹⁰ There, said Philostratus, Apollonius turned the shrine into a “holy Lyceum and Academy, until every type of philosophy echoed in it.” Had Aigai been the seat of Alexander’s teacher and perhaps of Alexander’s studies, too? Lucian goes on to remark that the prophet’s picture appeared in his city’s coins with the attributes of Asclepius and the hero Perseus.¹¹ Here, he is probably accurate, as the coins could otherwise refute him. Alexander, he said, had returned home and amazed the citizens by his dress, his long hair and his claim to be Perseus’s descendant. Why did the prophet bother to stress this ancestry?

Perseus had local connections in Paphlagonia, where his picture stood on the coins of nearby Amisus and Amastris and pointed to the meeting of Greek and Persian culture in the region’s past. Was Alexander claiming a local nobility? Perhaps, but he may also have been alluding to his own training.¹² Perseus was the founding hero of “god-favoured” Aigai, where Apollonius had once taught: respect for this ancestry was still lively in the mid-second century. It could explain the prophet’s double descent, from Perseus the founder of Aigai and Asclepius the god of its temple. It also fits Lucian’s home for his teacher, and if the guess is correct, it puts the prophet’s appointment in a less suspicious light. He had returned to Paphlagonia after several years’ study at one of Asclepius’s major shrines, and when his city was building the god a new temple, who better to run it than the young man who knew Asclepius’s ways? He was a man of culture and Pythagorean philosophy. At Delphi, Apollo kept company with Platonists, and at Claros and Didyma, his prophets and thespodes wrote verses on God and the fate of the soul. Asclepius was no less urbane. At Pergamum, his great temple was thick with sophists and orators, doctors and men of philosophy. How could poor Abonouteichos hope to compete? At Gangra, a major city centre in the south of the province, we find that an order of “ephebes,” for education of the young, was only being established as late as the 180s A.D.¹³ Civic education, it seems, was still rudimentary in Paphlagonia, where parents sent their children abroad for study, yet here was a Pythagorean who knew the proper ways of the god. So far from being a hoax, Alexander was one more contemporary prophet, capable of bringing philosophy to bear on oracles.

At Pergamum, Asclepius sometimes gave his clients oracles, and at Claros, Apollo was sometimes consulted by individuals on their health. At Abonouteichos, the prophet combined the services of Claros and Pergamum, the two most flourishing and cultured shrines of the Aegean seaboard. We must not deny his shrine all novelty. People could distinguish a god with a new genealogy and then worship him as a “new Dionysus” or “new Isis”:¹⁴ Alexander introduced a “new Asclepius” who was manifest in the form of Glycon, the snake. The image was not so peculiar. Snakes had often uncoiled themselves in cults of Asclepius, and if Glycon had human ears, these were no odder than the ears of gods like Isis or Serapis which expressed their role as “listeners,” the standard epithet for so many gods in the Imperial age.

The shrine’s business, too, becomes less peculiar when set against oracles elsewhere. Lucian complained how the god “bore witness” to his prophet through the snake Glycon, yet at Aigai, Philostratus happily recorded how Asclepius had “borne witness” to his hero Apollonius.¹⁵ We have seen how the Apollos at Claros and Didyma were free with these testimonials. At Claros, consultants of Apollo were initiated into the mysteries, and likewise, Abonouteichos installed a longer cycle of mystery rites; Lucian claimed that they honoured the prophet’s divine birth and led up to sex with a well-born priestess, the wife of a Roman official. At Claros and Didyma, the highest class of people asked Apollo who was God: at Abonouteichos, they asked about Glycon’s identity: “I am Glycon,” said the god, “I am come to bring light to mortal men.”¹⁶ Apollo at Claros told client cities to display his portrait and allow it to “shoot away” the plague, yet when Asclepius and Glycon gave the same order, Lucian ridiculed it.

At Hierapolis, one Apollo had spoken well of another and fostered Claros’s cult: in Libya, Ammon at Siwah also referred clients to Claros, home of Apollo the Sun god.¹⁷ In Paphlagonia, it was not so unusual when the new prophet referred clients to the old Aegean shrines, Didyma, Claros and the rest, advice which did not befit an obvious hoax. As his own clients spread, they knew very well what a major oracle offered.¹⁸ Their cities in Thrace and on the Black Sea already sent delegations to Claros, while cults and dedications for Apollo can be traced in other nearby Greek colonies: those founded from Megara honoured Delphic Apollo, while a dedication to “Apollo at Didyma” survives from a Milesian settlement. Asclepius had already spread widely in these prospering centres of Hellenism, so that visitors were not such innocents, ripe for a fraud. At Pergamum and Aigai, Asclepius prescribed the rarest remedies to patients who saw him in dreams. At Abonouteichos, Alexander often dreamed on his clients’ behalf, although his alleged recipes for bear’s grease allowed Lucian much mirth.

Much of the shrine’s daily business was the familiar diet of a local oracle, the sort of requests which were being handled in Egypt by any competent crocodile god. There were questions on marriage and runaway slaves, burglary and bother with the neighbours, although Lucian tells us more about questions on adultery than our other sources for oracles reveal: he alleged indignantly that the god sometimes cited correspondents who did not even exist. At Hierapolis, Apollo approved the wisdom of city magistrates; at Abonouteichos, the Roman governor of Asia was told how best to educate his son. At Didyma, Poplas would ask his god about an embassy to Rome; in Paphlagonia, clients asked whether it was better to go to Italy by land or sea.¹⁹ Polites the crown-bearer at Didyma asked Apollo about the soul; at Abonouteichos, an elderly Roman governor was said to have asked about his past and future lives. It was probably Lucian’s fiction that he was told that he had first been Pelops, then Menander, and would end up as a sunbeam: Rutilianus, said Lucian, “began to imagine he had become one of the heavenly bodies.”

Details of procedure were not too unusual. The prophet charged a fee, like all other shrines, although Lucian alleges that he also paid agents to fan out and increase his business in the cities. He kept theologoi, like Delphi, and sacred exegetes, and like other shrines of Apollo, his shrine used choirboys, allowing Lucian to claim that the prophet corrupted them whenever he fancied. He received many of his clients’ questions as sealed inquiries, like shrines elsewhere. He also charged more for special “self-spoken” oracles in which the god himself replied through Glycon’s person. There was nothing odd about this division of business: it can be found at classical Delphi. If Glycon seemed to speak, that was a clever trick, but other shrines were clever too.²⁰ Eager clients pressed the prophet to pray on their behalf and intercede, like a friend, with his god. This personal role was not so novel. At Pergamum, clients asked a priest to dream on their behalf when they were having a lean time at night. At Aigai, said Philostratus, a person approached Apollonius and asked him to “introduce” him to Asclepius, with whom he stood in such esteem.²¹ Philostratus knew this habit, although he made his hero reject this particular instance. “What need has a man of introductions,” said his Apollonius, “if he be morally good?” At Abonouteichos, as elsewhere, people were not so high-minded.

At Didyma and Claros, the gods spoke the Platonist theology of their prophets. At Abonouteichos, Lucian insists, the prophet was a follower of Pythagoras and was therefore a monstrous fraud. Alexander’s philosophy does seem to be more than Lucian’s invention, and if he had trained at Aigai, its style befitted him. At Delphi, Apollo mixed with the top Platonists, and at Didyma, his prophet might be a “successor” of Plato. At Claros, a thespode like Ardys knew all about middle Platonism. Some of these shrines’ oracles had Pythagorean touches of language, as did the statements of priests and “theophants,” inscribed elsewhere.²² We need only recall Apollonius’s “book,” on show at Antium, in which an oracle had dictated Pythagoras’s views on philosophy. As at Claros, so in Paphlagonia: the god’s oracles were most peculiar when most true to their prophet’s education.

Why, then, did Lucian hate him? The cult was new, whereas Didyma and Pergamum were old and traditional, and it also aspired to an old Ionian pedigree. Lucian had tried to bestow this aura on a temple cult in his own Syria which he described in the ancient style of Herodotus.²³ Alexander the prophet went one better: he petitioned the Emperor and obtained permission for his town to change its awful name. Abonouteichos became “Ionopolis,” a pure Ionian city in an area where many cities aspired to descent from Miletus and where the image of Ionian Homer stood on the coins of nearby Amastris, while a neighbouring town was “Doros,” or Dorian.²⁴ The name survives in the modern Inebolu, for it was through its prophet that this Paphlagonian port became respectable. Lucian had been brought up on Rome’s eastern frontier, where Syriac was spoken in the streets: his satire was the work of one arriviste deeply despising another.

He also hated the prophet personally, and when he tells how Alexander tried to marry the daughter of his grandest Roman client, we do begin to suspect his ambition. He was “prophet for life” and hierophant of the Mysteries at one and the same time, yet the honours were paid to Glycon and his god, not to the prophet’s own person. It was not such a dangerous ambition, to wish one’s city to be given an old Ionian name.

When the prophet died, no successor was adopted. Perhaps he had left funds to pay for his duties in perpetuity, but instead, the shrine was entrusted to a doctor.²⁵ This choice is revealing, a reminder that the shrine’s business was medical, the typical field of Asclepius and the Pythagoreans.²⁶ It combined this art with the giving of oracles in a well-chosen area. The mid-second century saw growing prosperity in many of the Black Sea cities, a region whose rise has often been overlooked. Lucian’s satire on the local “fat and uneducated” disguises people who were rich and willing to send their sons abroad for studies in Alexandria. Did Ionopolis seem a less daunting shrine than Claros for these cities which were not even reddish-brick? At Claros, Apollo told Syedra that her citizens were “all mixed up”: Paphlagonians would have been utterly beyond the pale. But at Ionopolis, the prophet is said to have taken questions in Celtic and Syriac as well as Greek: Celtic for people from nearby Galatia, Syriac for visitors from cities down the roads of the eastern frontier.²⁷

The oracle, in short, succeeded because it flowered in an undersup plied area and combined the features we have followed at such length. Glycon appealed to men’s willing sense of the “presence” of a god. He was a “new” form of an old god in an age which did not unduly respect originality. He spoke philosophy; he answered questions on future prospects; he prescribed cures. He met lasting needs in a style which was slightly livelier than the accepted style of the old Aegean shrines.

If Alexander had been a fraud, he was exactly the fraud which his contemporaries deserved: much of his practice conformed to the general practice of great oracles elsewhere, at a time when those oracles were enjoying a renewed prominence. It is, then, to this prominence that we must return in conclusion. A simple “rise in credulity” is not its explanation. Before the oracles began to revive, many forms of “credulous” divination were already flourishing; when they did revive, educated scepticism was not dead, as Lucian and others demonstrate. It is better to begin an explanation from the business which the inspired prophets handled. They concentrated on questions concerning the gods’ own preferences, and it is from the lasting belief in these gods’ presence that this service could re-emerge. As a constant presence, the gods could both help and hinder, “standing beside” people and their cities or showing a very lively anger. As civic notables revived the gods’ “ceremonies,” oracular shrines revived too. They could guide small details of practice and help to assure the gods’ own favour: the new text from Ammon to Cyzicus is connected with the ceremony of “displaying the crown of Ammon” in the city and evidently assures its clients that they will be safer if they honour the god in this way and send appropriate sacrifices to Apollo, a Sun god too, at Claros. There was a necessary mutual exchange between flourishing cults, flourishing oracles and honours for the gods.

It is clear from Plutarch that this abiding thread in pagan “religiousness” also needed material support. The smarter the shrine, the greater its clientele. At Didyma, in the 60s A.D., we can see a preliminary local revival: Claudius Damas revived “traditional” choirs and ceremonies in the cults of Apollo after a prominent career as a great civil magistrate.²⁸ At Claros, the temple had begun to be rebuilt before Hadrian’s final benefaction. When the Emperors did assist the shrines’ rebuilding, they were not giving funds to a vacuum: before Hadrian, Plutarch and some of his friends had been only too keen to see the oracles revive. Benefactions did not create belief, but they did allow it to spread and flourish. At Miletus, Trajan paid for the cutting and paving of a sacred way to the shrine at Didyma. Early in his reign, he held the office of prophet and “crown-bearer,” the major honours at the site. The festivals to celebrate his office brought visitors from far and wide, and perhaps it is through them that we find the city of Tyre choosing to put up a fine honorary dedication at Didyma.²⁹ Word spread from these occasions that Apollo was back into his stride: at Didyma, Apollo was at one point granted the right by the Romans to receive mortals’ bequests. Roman lawyers cite him as a prime example of a god who enjoyed this privilege, and evidently, the grant was an official ruling, probably from an Emperor.³⁰ Did Trajan or perhaps Hadrian, another benefactor, confirm it at the shrine? Inspired verse oracles needed financing to maintain their staff, and these rights were very valuable. Better buildings and festivals brought more visitors, paying more fees and making more offerings: Lucian considered twelve to fifteen talents a sufficiently monstrous sum for him to attribute it to Alexander’s shrine as its yearly income from fees alone. More visitors and bigger incomes also helped the oracles to improve their style: the “Heraclids” became lifelong thespodes at Claros and verse returned to Delphi. It is evident from Plutarch that this change in oracles’ quality would also increase the clientele.

Why, though, did so many cities process so dutifully to Claros and inscribe their names between c. 110 and 205 A.D.? The earliest of these inscriptions precedes the final dedication of the Clarian temple and the changes in its staff: does it mark the beginning of the cities’ attendance or only the beginning of a new “epigraphic habit”? Even that new “habit” is significant; at the very least, we can conclude that second-century cities chose to publicize their delegations as never before. Within their numbers, we can also see new cities joining the circuit and inscribing, in years when epigraphy was the recognized practice. How can we explain these changes?

At the simplest level, again, attendance on Claros owed an obvious debt to the Roman peace and an age of more settled prosperity for the civic notables. They felt safer in travelling far and had the funds to pay for the choirs and trainers: in Paphlagonia, too, the oracle had flourished on peace and prosperity in its inaccessible site. As the cities’ “ceremonies” and buildings received more funds from local donors, there was a greater need for an outside spokesman on the gods’ own wishes: oracles could solve problems of tenure and appointment and prescribe cult titles, hymns and appeasement of the gods. Once visitors had begun to visit the newer, smarter Claros, the practice would multiply because of the intense intercity rivalry of the age, Hierapolis had an Apollo of its own and felt no need to send a choir to Claros, until Apollo ordered one. But what of a town like Tabai, when nearby Heraclea was known to have gone to the god? Should she not go too, and if Tabai then went, should not Heraclea go more often, thirty, thirty-five, forty times over the period?³¹ Perhaps a particular priest at the shrine first suggested that a civic delegation should inscribe and date its visit on the stones of the site while the temple was beginning to be improved. Civic journeys to oracles were an old and familiar feature of Greek life, but in the second century, there was a particular reason why an “epigraphic habit,” once started, should blossom. Whatever the particular cost or encouragement at Claros itself, one city’s visitors would not wish to lag behind another’s delegation. In turn, the cost of inscriptions increased the shrine’s funds.

Intercity rivalry was joined, most opportunely, by a more basic motive: fear. The improvement of the shrines happened to precede an intense sequence of plagues and earthquakes, perhaps more intense than any in the previous century. We can see this pattern in the texts at Hierapolis: the city consulted Apollo’s oracle in a time of plague and was then ordered to send a mission to Claros. In Paphlagonia, Alexander, too, referred business to the major shrines. The pattern is beautifully clear, now, at Cyzicus, where Ammon from distant Siwah told the city “not to be afraid” but to go with sacrifices to Claros and honour Apollo, “making mention of me, too.”³²

As a result, a city would go obediently and add its name to our Clarian inscriptions, and then, for the sake of its safety, it might come again and again. At Claros, we can see the end product: some cities’ texts say that their choirs had come “in accordance with an oracle.” One good reason, then, for the shrine’s greater host of delegations lies in its own self-advancement: like Alexander the prophet, it commanded people to come. The improvement of the shrines was accompanied by these oracles of “bidding” and perhaps, too, by envoys, encouraging old clients to resume attendance. The great calamities then lent weight to this movement. At Claros, some cities came once only, sending delegates (but no choirs) from further afield. A crisis brought them to the god, and in turn, the shrine’s fame spread ever further, to Sardinia, Africa and Hadrian’s Wall. Meanwhile, former clients and choirboys helped to spread its repute.

We have seen how processions and “traditional” ceremonies did reinforce an image of the cities’ social order, enduring through time. How did this oracular revival relate to this order, at the shrines and in the client cities? “In the Antonine age,” it has recently been suggested, “the oracles had maintained their strong, collective associations. They were valued because they could speak for the city as a whole on questions that affected the city as a whole… they did so largely because they had been studiously maintained for that purpose by local benefactors… By such means, the ruling groups in the city were able to exercise a control of its religious life that went far deeper than mere display. It was they who interpreted the oracles.”³³ Are these ideas of “control” and its setting true to what we have seen?

The range of the evidence does not entirely support them. Didyma flourished, although all but one of the inscribed texts spoke to individuals,³⁴ not to “a city as a whole,” nor, so far as we know, did oracles from sites like Mallos or Patara. At Claros, it was not the upper class of the client cities who interpreted the god’s oracles: the job was done by the thespode, who was appointed for life. His texts were verbose, but they were wholly free of ambiguity and needed no further interpretation, beyond a sense of Homeric Greek and rare compound words. Were clients like the Oenoandans really concerned with “control” and against what, indeed, was “control” needed? Not, indeed, against religious upstarts or individual “holy men”: on a strict definition, there were no pagan “holy men,” for in pagan Greek the word “holy” applied to places, but not to people.³⁵ There were stories of “godlike” seers who could indicate the will of the gods, like travelling oracular prophets. The best known is Apollonius, the legendary wise man who was believed to have advised cities on their cults and ceremonies in the mid-first century A.D. He was not believed to have faced opposition from traditional oracles. He taught at Aigai and his “biographer.” Philostratus, remarked how the shrines of Claros and Didyma honoured him warmly. The Christian book On True Belief has preserved a third-century text in which an oracle praised Apollonius’s wisdom.³⁶ In pagan “religiousness” there was no fear of heresy, no urge to orthodoxy and clerical “control.” The real Apollonius cannot be untangled from the admiring legends, but pagans’ religious fear was a fear of the random anger of gods, not the ambitions of upstart men. It is to this enduring thread that oracular questions and answers attached.

What, too, of the idea of a “studious maintenance” of oracular shrines by benefactors to suit a “purpose,” related to their own power? Like all other culture in this period, an oracle was suspended from the time and money of the upper class. At Didyma, the prophets of Apollo were drawn from a narrow range of interconnected noble families. At Claros, the prophets were no humbler, while two of the thespodes claimed to be blue-blooded Lydians. The choirs and civic envoys to Claros were drawn from their cities’ educated families: they included civic priests and people like Cornelius (Mu)ndicianus Crocus “of the Philopappidai,” who evidently belonged in the highest society of Stobi before he travelled to Claros as the city’s envoy.³⁷ The newly found text of Ammon to Cyzicus was connected with some important figures in the city, a “displayer of the crown of Ammon,” a rich donor of the inscription and “pious” worshippers of Ammon who were people of rank, active in the cult of the Emperors. These persons of quality did not entirely lack a return. They advertised their attendance, at Claros and sometimes, too, at home. At Didyma, as elsewhere, the god did also “testify” to the virtues and family trees of willing prophets. However, we know of no person who used these words as grounds for civic, secular advancement: they were the copingstones on careers which had already been mapped out. In Miletus, Damas, the benefactor and restorer, issued coins in his year of civic office with the types of the Apollo whom he had served.³⁸ Others did likewise, but the god, not the donor, took most of the honour. When Damas was prophet for a second time, he was aged over eighty, acting not for his future career but for motives of correct piety. The “testimonials” at Didyma went to the few great prophetic families who had already held major civic offices or were combining them with “voluntary” service at the shrine. Perhaps the recipients prized them in their rivalry with each other: they did not prize them as a means of “control” of people outside their own small class. Oracular service was not consciously pursued as a device for power. It was one more reinforcement of an order which was not being challenged. The notables went to Claros for their city’s safety and prestige: because they went, they could be felt to be doing their best by the gods. They met the great expense of the priesthoods, and in return, to coax them to give freely, the gods “witnessed” publicly to their merits. The consequent support for their position was incidental: it was not the reason why they undertook these tasks in the first place.

Like “control,” “studious maintenance” is too external a view of oracular activity. The lists of Apollo’s prophets had their interruptions, when nobody “volunteered” for the job: at Didyma, the methods of selection are implied by the prophet’s inscriptions.³⁹ If several candidates stood forward, a choice was made between them by that time-honoured arbiter, the lot. In some years, only one candidate’s name went forward: he was then chosen “without the lot” and sometimes in response to his own preliminary “promise.” Such offers were not always forthcoming. Once, c. 120, the job is known to have passed directly from a distinguished father to a twenty-one-year-old son; once, in the mid-first century, there was no prophet at all. “Studious maintenance” is not quite the term for this delicate balance of “volunteering” and uncontested office: prophets, it seems, were not always easily found. At Claros, the pattern is clearer because the dated records are better: in thirty-eight years out of fifty-four, the god himself held high office at the shrine.⁴⁰ His funds, not the donors’, financed it in the years of its great attendance. At Patara, in Lycia, we can watch the revival of an old, respected oracle and see the relation of benefactor and shrine from the beginning.⁴¹ “After a long time of silence,” the local Apollo began to give oracles once again, yet the primary impulse was an earthquake which had troubled southwestern Asia in or around the year 140. Only when Apollo had already begun to speak did he attract a modest benefaction from the prince of all Lycian donors, Opramoas. This great giver did not “coax the shrine back into life” or try to dominate it. He adorned it, one among his many other buildings, and never held office at the site. The inscriptions in his honour received the gift politely, but discreetly: “perhaps,” said the text warily, “Apollo also took pleasure in this moment because of Opramoas’s piety…”

Oracles, then, were not an explicit source of control or power for individuals: rather, Apollo coaxed benefactors who might otherwise be reluctant, and also helped to contain the mutual rivalries within their class. These tensions were common to all offices which relied on holders’ “love of honour,” and Apollo, here too, reflected the atmosphere in which he lived. At Side, his verses praised moderation and a good disposition in benefactors whose gifts were open to dispute. At Didyma, he confirmed the musical piety of one prophet against his cousin’s extravagance. At Hierapolis, he praised the government of well-disposed leaders. Oracles were not so much a source of civic power as arbiters, so that religious life ran smoothly.

At Claros, the latest-known civic inscription belongs c. 205 A.D., but the shrine did not then cease to draw clients. On a coin of the late 250S, we see the figures of thirteen city envoys standing round a bull as it is sacrificed, before Apollo’s temple: the cities are probably the members of Greek Asia’s distinguished Ionian League. At Didyma, the sacrifices of Ulpianus and the musical piety of Macer, his cousin, belong in the same period; in the third century, we have our one Imperial inscription from a civic delegation, reminding us how much we have lost from Didyma’s own client cities and “colonies.”⁴² In the 250s, Delphi still spoke for Side; in the 240s, the image of the local oracular Apollo still featured on Patara’s coins. The shrines and their client cities were not immune from the great disorders of the 260s and 270s, but even then, as we shall see, those calamities did not undermine the continuing role of their gods.

What, though, of the theological oracles with which we began? The first datable text is the answer to Polites at Didyma given c. 170–180. Those which survive in later Christian books are most likely to derive from Porphyry’s collection: if so, they must have existed by c. 260–270. They show very well how philosophy supplemented cults and conventional piety, adding “beliefs” to ceremonies to make a religio, as the Emperors then described it. They show a lofty, abstract theology in which the Supreme god was defined by negative attributes. Sometimes, they place him beyond the Elements or equate him with Eternity, “which was and is and ever shall be.”⁴³ They also equate Apollo and the Sun god.

These tendencies and modes of thought had already been evident in Plutarch’s dialogues, which were set at Delphi c. 66–85 A.D.: they were not a new development of the third century. As for the equation of Apollo and the Sun,⁴⁴ it was already made by “many people,” as Plutarch’s speakers accepted. The relations of Apollo to Dionysus and to Fire were topics which even the guides and local theologoi at Delphi discussed quite freely. So was the idea of a Supreme god, whose names and forms might vary with the yearly cycle of the seasons. When these questions were put to the gods themselves and inscribed or preserved for collectors, they were not a sign of a new, growing pagan doubt. There were sound, healthy reasons why such questions should occur within paganism itself: Miletus had priests and cults of a Most High god; curiosity is a sign of liveliness; the relation of the Supreme god to other gods was a genuinely interesting question; in reply, Apollo’s answers were not monotheist, recognizing only one God. It has, however, been the essence of this chapter that Apollo was sensitive to atmosphere. These texts, then, cannot be considered without regard to the true monotheists in the pagan cities, the Jews and the Christians who were winning converts meanwhile.

In Miletus, we have a famous third-century inscription from “Jews who are also devout worshippers,” carved on seating in the city theatre. Whether thesejews are Jews by birth or Gentile converts, they are clearly a group of some prominence, willing to share in their city’s culture.⁴⁵ We have another intriguing text, also from the theatre, in which each of the seven archangels is invoked in a separate panel: “archangel, protect the city of the Milesians and those who live in it.”⁴⁶ Its date is uncertain, but not earlier than the third century A.d., and Jewish ideas are surely present, although the text may not originate from a practising Jew. It is, then, very apt that Didyma spoke on the Jewish religion which was evident in the city’s public places.⁴⁷ Lactantius quotes a text in which Didyma’s Apollo praised the Jews’ respect for law and their worship of the “Creator of all.” According to Augustine, Apollo had been asked on this occasion whether “reason” or “law” was preferable; in Greek, perhaps, logos or nomos. In reply, he adduced the Jews to support his argument for “law,” or nomos. This text had probably been found in Porphyry’s collection and existed, therefore, at Didyma by c. 260: the first of Porphyry’s books also quoted an Apollo who approved the wisdom of Eastern peoples, especially the Chaldaeans and the “enviable” Hebrews, who worshipped “in a pure manner” a single God and believed in the seven zones in heaven.

At Didyma, Apollo spoke on God and the Most High; at Claros, on the nature of “Iao,” a Greek version of the Jews’ Yahweh, who was known in popular spells and sorcery. People did ask explicitly about the Jews, and it is hard to believe that none of the theological questions on “God” owed any impulse to the presence of prominent Jewish groups in the second- and third-century cities.⁴⁸ If Greek-speaking Jews attended Miletus’s theatre, they could also discuss and dispute with the pagan citizenry. Apollo’s texts are firmly pagan, but they do discuss God in a language which Jews could endorse. It is quite credible that, at times, the questions arose from the presence of Jewish groups and sympathizers, a presence which is emerging into clearer prominence from the growing evidence of the second- and third-century cities in Greek Asia.

The Jews were not the only alternative presence in these years, for the Christians, too, were winning converts. Here, the oracle for Oenoanda has been seen as Apollo’s attempt at a compromise, answering a question which began from pagans’ disputes with Christian atheists.⁴⁹ Its text, indeed, used language which Christians later applauded; it needed only a small Christian revision: it called the gods “angels” and ordered no sacrifices in their honour. However, the words of oracles elsewhere suggest that any resulting compromise was not deliberate. The Christian author of On True Belief cites a text in which Apollo foretold disaster for “those who have forsaken the ways of their ancestors.” He quotes it as the god’s rebuke to Jewish questioners, but may it not have begun as a rebuke to Christian converts?⁵⁰ Apollo certainly spoke on the subject, and we owe to Porphyry’s book a plausible text in which Apollo gave a degree of praise to Christ. He was a wise man who worked miracles and died a “bitter” death: the miscreants were the Christians, who insisted on worshipping his mortal body when it had been torn and disfigured by nails. Their cult was absurd: God Incarnate, said Apollo, was a myth.⁵¹ Porphyry also knew that a distressed husband had asked an oracle how best to dissuade his wife from Christianity. Apollo held out little hope: it was easier, he said, to write on water or fly like a bird than to shift a woman from such impiety, the worship of a man who had been put to death in irons. Christ’s soul was immortal, but not his body, and the cult was thoroughly misguided.

Nowadays, these views might find sympathy from the Churches’ very bishops, but at the time, they were not grounds for a compromise. Apollo found the Jews’ religion easier company than the Christians’. In their language and remedies, explanations and praises, oracles were thoroughly traditional, and were valued for being so. In Plutarch’s dialogues, it was his friend Serapion, the Stoic and poet, who defended a divine origin for every single word in oracles’ texts.⁵² “We must not fight against the god,” said this literalist, “nor abolish providence and divinity along with prophecy. We must try to find solutions to apparent obstacles and not betray our fathers’ pious belief.” Greeks and Romans agreed alike on this great conservative case for their cults. At Didyma, c. 200, Apollo repeated it as a matter of course to the inquiring treasurer Hermias. In the 250s, he praised prophets like Macer and Ulpianus, people “to whom it was axiomatic that nothing could be both new and true.”⁵³ Oracles united respect for “ancestral practice” with a strong awareness of the awe and potential anger of the gods. These attitudes, as we shall see, underlay the cities’ persecutions of Christian “atheists.” When the oracles diagnosed divine anger, they prescribed archaic rites of sacrifice and hymns for an appeasement by their clients, but Christians, in these same cities, could not participate in these divine “commands.” The origin of cities’ local outbursts against their Christians is usually concealed from us, but it may owe something to the prior advice of an oracle on rites and honours. At Abonouteichos, it seemed plausible to Lucian to allege that the “false” prophet told his crowds to throw stones at “atheist” Christians and Epicureans. By 250, the connection between oracles and persecution is clearly attested.

This oracular activity, as we shall see, extends beyond the mid-third century, and with its help, we can correct one perspective in Gibbon’s unsurpassed account of the rise of Christianity within the Empire. By the third century, “human reason,” he believed, “had already obtained an easy triumph over the folly of Paganism… . Yet the decline of ancient prejudice exposed a very numerous portion of human kind to the danger of a painful and comfortless situation.” They could not bear the fruits of reason’s triumph: rid of their superstition, they were all too ready to find it elsewhere. They were “almost disengaged from their artificial prejudice, but equally susceptible and desirous of a devout attachment…” Christianity met their need: “Those who are inclined to pursue this reflection, instead of viewing with astonishment the rapid progress of Christianity, will perhaps be surprised that its success was not still more rapid and still more universal.”⁵⁴

Gibbon mistook the views of a few unrepresentative thinkers for a “vacuum” which he ascribed to the majority. From the men of Lystra to the people of Oenoanda, we have seen a different world, one where the gods were still “evident,” standing beside their clients in dreams and guiding them with words or signs of their will. There was no “disengaged” majority: there were those to whom pagan worship came naturally and those who expressed a stronger articulate piety. Pagans, too, had their gods as “helpers” and “fellow workers,” an evident company who might accompany a man “always.” Potentially helpful, they had not lost their old Homeric capacity for sudden awe and anger. Their company might reassure a city or an individual, but their continuing “mildness” and “ease of encounter” could not be taken for granted. In their cults and honours, the “traditional purpose” still mattered, as the language of gods themselves proclaimed. Wherever the cities’ benefactors financed it, it helped to reinforce the image of an enduring community, true to its past. It also linked town and country in a mutual exchange of visitors.

Respect for the ideal of “tradition” did not exclude changes in reality, and by the second century A.D., certain currents in pagan cult are prominent in our evidence. More techniques aimed to secure and tame the “presence” of the gods; more techniques tried to elicit the future and ensure the best course through it. The cults of the household and city had been joined by the cults of small societies, a pattern which had a long history of development but which shows a particular intensity in the evidence of the Imperial age. People could now turn to groups for particular religious worship as a specialized activity within their cities’ life. The evidence for religious mysteries in these groups, especially mysteries in the cult societies of Dionysus, is overwhelmingly evidence of the Imperial period. At that time, mystery rites begin to be attested for other cults, for Isis or Serapis, the oracle at Claros, even the worship of Emperors. Some, though not all of them, promised greater ease for the soul after death, reassuring pagans of an anxiety which was widespread, though not universal. There were worshippers, too, who felt protected by a god against the arbitrariness of fortune, the malice of sorcery and envy or the limiting constraints of the power of the stars: newer techniques exploited or charted these forces but did not necessarily weaken the appeal of “protecting” divinities, who were present in dreams and images. While art and the ancient cult statues continued to define people’s sense of the gods, philosophy continued to discuss the concept of a Supreme god, his qualities and relation to the other divinities. Oracles then made this language the language of gods themselves: were they really only the interest of a few enthusiasts and a very small literate elite? Perhaps this position can never be entirely refuted, not even by the public inscriptions of theology, which many people might not read, or consider. It calls, in the end, for an intuition about the concerns and discussions of people in large and small Mediterranean towns where work is not constant, all day, every day, and the spoken word is vastly more influential than the written. Grand epithets for God are not hard to assimilate, wherever people meet and know Greek: providence and a Supreme power, the nature of the heavens and the various types of worship are not questions which only interest professors, and they are not unsuited to a long afternoon between harvests or a wet winter’s day. At the very least, these debates on the higher theology bear witness to a continuing uncertainty: who was to say who the highest God might be? If the Supreme god was unknowable, who was to say which one of the many cults of different peoples was right or wrong? At its heart, therefore, pagan theology could extend a peaceful coexistence to any worship which, in turn, was willing to coexist in peace.⁵⁵

We have dwelt on the lasting traditional “religiousness” and its continuing vigour, at least to the mid-third century. Yet we should not mistake it for “religion” in the strong Christian sense. In the Imperial period, some people in the cities were turning to find this different alternative, in the Jews’ synagogues and the Christian churches. Both these religions continued to attract outsiders, especially the Christians’ faith. Their converts, we have seen, were not abandoning a static or dying religious culture. Rather, they were joining the most extreme option in a period when religious issues were very lively; their chosen option joined cult and philosophy; it gave a clear code of conduct; it promised hope and an absolute triumph over death and Fate. The Supreme god of the Platonists and the oracles is, ultimately, a remote figure, more negative than positive; it has that distance and chilliness which is best sensed, still, in the very style of the Pantheon in Paris, a building put up by its deliberate heirs. The Christians, too, could cite prophecies in a world where past oracles were widely respected; they knew exactly who God was, in an age of discreet uncertainty: their God was a God of history, proven in events; above all, he had sent a Son, to redeem men by actions of total selflessness. It is time, then, to turn to the Christian option, in the years until 250; then we can set it beside the pagan cults and oracles and follow them both through the difficult years to Constantine himself.