Leach: Kingship and divinity

This work is licensed under the Creative Commons | © Edmund R. Leach.

Kingship and divinity

The unpublished Frazer Lecture, Oxford, 28 October 1982*

Sir Edmund Ronald Leach


Mr Vice Chancellor,

I come to praise Caesar, not to bury him. When I accepted your invitation to deliver this Frazer Memorial Lecture, I had planned to offer a favourable reassessment of one of Frazer’s most notorious “conjectures,” but the peculiarities of Frazer’s style make this very difficult.

Frazer used his ethnographic evidence, which he culled from here there and everywhere, to illustrate propositions which he had arrived at in advance by a priori reasoning, but, to a degree which is often quite startling, whenever the evidence did not fit he simply altered the evidence! A single example which is relevant for my later presentation will show you what I mean. In volume nine of the twelve-volume third edition of The golden bough, Frazer cites evidence that as far back as the fifth century BC the Roman Saturnalia was celebrated around the time of the Winter Solstice in late December (Frazer 1911–15, vol. 9: 306). Furthermore, Frazer admits that there is no evidence that it had ever been held at any other date. However, according to Frazer’s theory, Saturnalia ought to have been a Spring Festival held around the end of February. So he simply “conjectured” that at one time this had actually been the case and thereafter proceeds with his argument as if the evidence relating to the Winter Solstice Saturnalia really applied to a Spring Festival.

The golden bough is full of rewritten ethnography of this kind. Even so, despite the adverse comments that I shall feel bound to make, my central purpose is to show that, at least in one notable instance, Frazer’s guessing was closer to the mark than most people might reasonably suppose.

Since time is short, I shall proceed directly to my principal theme. First we need some background. The golden bough ranges over a vast terrain, but for present purposes it will suffice if we accept the précis of the book’s subject matter, which Frazer himself provided in 1905. This is what he then said:

The divine king has been gradually evolved out of the old magician or medicine-man…. The title of the King of the Wood, bestowed on the priest of Nemi, along with his relation to the woodland goddess Diana, suggest that the two together personified the powers of vegetation in general and of the woods in particular, and that their union may, on the principles of homoeopathic magic, have been regarded as a means to ensure the reproduction of that plant life on which, in the last resort, both men and animals depend for their subsistence….

People feared that if they allowed the man-god to die of sickness or old age, his divine spirit might share the weakness of its bodily tabernacle, or perhaps perish altogether, thereby entailing the most serious dangers on the whole body of his worshippers who looked to his as their stay and support. Whereas, by putting him to death while he was still yet in full vigour of body and mind, they hoped to catch his sacred spirit uncorrupted by decay and to transfer it in that state to his successor. Hence it has been customary in some countries, first, to require that kings should be of unblemished body and unimpaired mind, and second, to kill them as soon as they begin to break up through age and infirmity … these principles led in other places to a practice of allowing the divine king or human god to live and reign only for a fixed period, after which he was inexorably put to death…. On the analogy of such customs I conjecture that the King of the Woods at Nemi formerly reigned for a fixed period only, probably for a year, and that he had to slay himself or be slain at a great festival when his term of office was up (Frazer 1905: 160, 291-92).

Well, one guess is as good as another, so why not? It is perfectly possible. But analogy has seldom proved to be a profitable way of filling in gaps in ethnographic knowledge, and in this case even the analogies are problematic. Frazer writes as if we had firmly established records of scores of different regicidal customs of the sort he describes, but the fact is that the evidence is no more substantial than are stories about sightings of the Abominable Showman: “I once met a man who told me that when he was a boy he had seen …” and so on. The actual eyewitness to the slaying of the Divine King is always off-stage. Myths of this general kind have certainly been very common and of world-wide distribution, but just how often the killings actually happened remains a moot point.

Yet the stories are interesting even for the most sceptical. Why should we be asked to believe in the possibility of the killing of a god? What is the relationship between kingship and deity? What is the connection, if any, between the dying god-king of Christianity and the dying god-king of Nemi? It is this third question that will chiefly concern us this evening, but it may help you to follow my argument if I start by offering brief answers to the other two.

The words “god” and “king” are habitually used in very loose fashion. If you were to lump together all the imaginary entities which the literature describes as “gods,” you might perhaps detect a certain vague family resemblance, but that is all. So let me be more specific. For the purposes of our present discussion, “God” is a human construct which represents the non-human Other. God is what we are not but would like to be. in particular, man is mortal and impotent and constrained by the limitations of here and now; God is immortal and omnipotent and unconstrained by time and space. The manifest purpose of religious activity is to establish a bridge between the limited domain of here and now and the unlimited domain of the Other; thereby some of the potency of the Other will flow into the here and now for our benefit. This bridge between the here and now and the Other can be established in either of two ways: we may go there or it may come here.

In the first case, a devotee may kill himself (or a part of himself, through self-mutilation or through the sacrifice of an animal with which he is identified) and thus achieve a kind of absorption of the self into the Other. The devotee is “made sacred” through the operations of the sacrifice. The passion with which certain of the early Christian saints (e.g., Ignatius), sought martyrdom so that they could become reunited with God exemplifies such behaviour. Ignatius reputedly achieved his ambition. Legend declares that he was eaten by wild beasts in the circus at Rome. In the same part of the world at the same period, as is well attested by a variety of authorities cited by Frazer, the Syrian Goddess Astarte (or Ishtar) was served by eunuch priests who were considered to be representatives of the goddess’ lover Attis. Enlistment to the priesthood occurred annually at the spring festival at Hierapolis when devotees in a state of frenzy would castrate themselves in front of the altar of the goddess.

Alternatively, instead of the devotee attempting to achieve unity with the Other, procedures may exist whereby the Other is brought into the here and now. One common example is that in which a deity speaks through the mouth of a spirit medium while the latter is in a condition of trance. In other cases, ritual procedures serve to mark off a sacred space, which is treated as a part of the Other temporarily located in the here and now. The metaphysical visitors who enter this sacred space infect the attendant congregation with their sanctity. But potency from the Other is dangerous stuff. You can easily have too much of it. The bridge to the Other side must be closed as well as opened. The gods may be invited to visit us for a season but then they must be driven away again. This implies that if the divine visitation takes the form of incarnation in a human being, the god-man must be killed, at least in symbol, so that the god can return to the Other to be brought back again in due course by further rituals. In a few relatively rare cases, performances of this latter kind are associated with an ideology of ritual cannibalism. The body of the god-man who has suffered sacrificial death is eaten by the congregation of the faithful who thus ingest into themselves directly the residue of the departed god’s sacredness. The Christian Eucharist is a sacrament of this general type.

As is quite logical, the ritual procedures by which gods are brought into the here and now and then sent away again are the prototypical markers of time in human societies of all kinds. The visitation of the gods is, in Van Gennep’s (1909) terms, a rite de passage, a beginning of new time either for the initiate entering into a new stage of life or of office, or for society as a whole.

Finally, there is the matter of sex and gender. Strictly speaking, sex is not an attribute of the Other. Immortal gods do not have to reproduce themselves. But sexual difference and sexual potency are the dominant concerns of mortal men, and when men invent their gods they create them, in this respect, in the image of man. The symbolic transformations which then occur seem to be remarkably uniform. The divine potency of the Other is seen as male sexual potency and female fecundity. Where human beings are divided into males and females, gods are divided into static aspects and active aspects, the active aspect being commonly regarded as the female consort of her relatively static male partner. If this conjunction of male and female deities is credited with progeny, then the child will usually be male. The characteristics of this “son of god” will largely recapitulate those of his divine father, but the Son will be more nearly human, a mediator who is relatively accessible and active in this world, as is his mother. Christianity follows this pattern but, in the case of Roman Catholicism, it is the Mother Goddess rather than the Son who is the mediator par excelence. But terror of sexuality has sometimes led to an insistence that God is a unity with no divisible aspects (which is the position adopted in Judaism and Islam) or to some alternative kind of non-sexual divisibility. The astonishing theology of the Christian Trinity is an example of this latter solution.

Kings are not gods yet they partake of divinity. Kings are mortal, but the office of the kingship is immortal; a king’s actual potency is circumscribed but, in theory, the power of the kingship is absolute. This latter power derives from God. But the divinity of kings, unlike the divinity of gods, is fragile and subject to pollution. It must be ritually renewed every time a new human individual assumes the office of king, and perhaps more frequently than that. The kingship must repeatedly be cleansed. All this is fully compatible with Frazer’s line of argument. But Frazer believed that the ritual renewal was always directly concerned with the magical regeneration of vegetation and also that it regularly entailed the actual (rather than the symbolic) killing of human kings. Neither of these latter conventions can be justified on the basis of the actual evidence.

I must give you just one more bit of general anthropological theory before we get back to Frazer’s fairy tale of the King of the Wood at Nemi, which he used as his central model for the dying-god-king. In the language of contemporary anthropology, the assertion that a particular story (either oral or literary) is a myth need not imply that it is untrue. Stories are myths if they are used as characters of justifications or precedents for social action, whether secular or religious. Whether the precedent story in question was or was not true as factual history is entirely irrelevant. In the context of the ritual occasion the precedent is treated by the actors “as if” it were factual history. That is what matters. Myth is believed to be true by those who use it. In this sense, all the Christian gospel stories are myth for Christians. Although practising Christians are deeply committed to a belief that the key events recorded in the gospels “really happened,” the fact is that the historicity of the stories (if any) is irrelevant for the religious implication of what the texts contain.

For example, when, following the story of the miraculous feeding of the five thousand in the Wilderness, Jesus is made to say: “He that believeth on me hath everlasting life. I am that bread of life. Your fathers did eat manna in the wilderness, and are dead. This is the bread which cometh down from heaven, that a man may eat thereof, and not die. I am the living bread which came down from heaven … the bread that I will give is my flesh, which I will give for the life of the world…. Whoso eateth my flesh, and drinketh my blood, hath eternal life; and I will raise him up at the last day” (John 6:47-54)**—this is a direct mythical justification for the proceedings of the Eucharist. The story presupposes the coexistence of another myth, namely the story of the Crucifixion, but the text has already transformed the body of the crucified god into the bread of the Eucharist—“the living bread that came down from heaven.” It does not call for any “real” human sacrifice in the way that Frazer supposed.

But now back to Frazer’s model of the dying god. One feature of the story is easy to overlook. The hypothetical, once-upon-a-time ritual involves two priest-king-god figures, not one. King “A” reigns for a period; his powers decay along with the vegetation and at the end of the annual cycle he is sacrificially killed by, or at the behest of, King “B” who then reigns for a period in his stead only to die in turn the next time round. Some of Frazer’s supposed examples appear to be descriptions of a genuine ritual sequence through which real King “B” succeeds to the office of real King “A.” But other cases involve complex role reversal. King “B” is the “real” King while King “A” is a mock King-substitute. When the season for ritual renewal is at hand, the real King “B” temporarily withdraws from office; King “A” is briefly installed in his place, ritually slain, and then succeeded by the real King “B” who resumes his office. In still other cases, as in the one we are to discuss, Frazer appears to postulate a still more complicated sequence of Box and Cox involving two mock kings but only one sacrificial victim. I don’t want to defend any of this, but you need to understand what Frazer was saying.

In the second (1900) edition of The golden bough Frazer quite explicitly fitted Christianity into his general scheme by means of a provocative section about parallels between the Christian story of the Crucifixion, stories relating to the Jewish festival of purim, and supposedly much earlier materials relating to the Babylonian festival of the Sacaea. But Frazer’s argument was vigorously attacked by Christian scholars and Frazer lost his nerve. In the twelve-volume third edition the offending passage is removed from the main text and relegated to a long footnote-appendix. This had the effect that in the one-volume abridged edition, which is the only one which most people have seen, all direct reference to the Crucifixion is eliminated. In this version, the original material survives only in an obscure remark in the preface, where Frazer mentions “the interpretation which I formerly gave of the festival [of Sacaea] as a New Year celebration and the parent of the Jewish festival of Purim” (1922: vi). He gives his readers no hint that the interpretation in question was centrally concerned with the Crucifixion of Christ. Yet despite this tacit recantation it seems evident that Frazer never really abandoned his original “conjecture.” So let us take another look at it. The version of the argument which takes up eleven pages of small print in the third edition of The golden bough lacks conviction (Frazer 1911-15, vol. 9: 412-23). Frazer is too anxious to defend himself from charges of mythologizing the Bible, and this spoils his case.

The quest for the historical Jesus has been pursued with great vigour since Frazer’s day, but with markedly negative results. In what follows, I assume that all such endeavours are vain. The story of the Crucifixion is a myth, a sacred tale which justifies the performance of a sacrament. I am concerned with the origins of that myth in the Jewish background from which early Christianity emerged. I see no reason to believe that there was ever any flesh and blood “founder” of Christianity called Jesus (Joshua), but even if such an individual did once exist that would not effect my argument.

The section of The golden bough, to which the “Note” on “The crucifixion of Christ” (Frazer 1911-15, vol. 9: 412-23) would have belonged if it had not been relegated to an Appendix, is entitled “Saturnalia in Western Asia,” a rubric which is made to include not merely Bacchanalian rites among various peoples of the Ancient Middle East, but also masked dances among Eskimaux, Amazonian Indians and Natives of New Guinea, as well as a surprisingly accurate account of a Borneo Head Feast which appears to be totally irrelevant. Carried to these extremes, the anthropological “Comparative Method” of the Frazerian era becomes entirely ludicrous, but when applied with greater caution something can still be made of it. My own comparisons will be much more confined both in time and space.

Frazer built up his story linking the Babylonian Sacaea, the Jewish Purim and the Crucifixion through a complex mesh of circumstantial evidence. The tie between Purim and the Crucifixion is tenuous in the extreme. In some parts of Europe during the late Middle Ages the celebration of Purim included a street play in which Haman, the persecutor of the Jews in the Book of Esther, was hanged (or crucified) in effigy. Frazer argued that this must have been a “survival” from an earlier practice in which Haman had been represented by a “real” victim. A number of details from the gospel story of the Crucifixion might be fitted to a Purim ritual of this kind. Hence Frazer’s “surmise” that Jesus was crucified in the role of a mock king representing Haman.

The tie up between Purim and the Sacaea is equally devious, but at first appears slightly more substantial. Purim was a “late” introduction into the Jewish ritual calendar though it is mentioned by Josephus. In the Middle Ages it was an occasion of carnival and gift exchange. A sixteenth-century rulebook declares that “On Purim a person should drink until he does not know the difference between ‘Cursed be Haman’ and ‘Blessed be Mordecai’” (Tractate Megillah 7b).1

As specifically ordained in the Book of Esther, Purim commemorates the escape from persecution in Mesopotamia in much the same way as Passover commemorates the escape from persecution in Egypt, hence it would be quite logical if, in this context, Mordecai, the hero of the Esther story, were treated as a god-king-messiah in much the same fashion as the Moses of Exodus is treated as a god-king-messiah. This point was not made by Frazer, though it is very relevant for some of the material that we shall be considering presently. Most modern Jewish scholars seem to view the carnival atmosphere of Purim with some distaste. They imply that it was never of any great religious significance. Purim was celebrated just one month before Passover, that is to say, at the full moon of the last month of the Jewish ritual calendar. Frazer does not spell out his calendrical argument, but he seems to imply that Purim and Passover, in combination, constituted a unified New Year Festival and that the reign of the hypothetical mock king of Purim lasted for one month until Passover. The detail which first set Frazer onto his sacaea hobbyhorse is that the main characters in the Book of Esther all have the names of Mesopotamian deities: Esther is Ishtar or Astarte (the biblical Ashtaroth); her cousin Mordecai, the hero of the story, is Marduk—both names of prominent Babylonian deities. The villain, Haman, is named after the Elamite deity Hamman. vashti, the Queen of Ahasuerus (that is Xerxes), who is replaced by Esther, bears the name of the Elamite goddess Vashti (Frazer 1911-15, vol. 9: 365-66, 401).

On the face of it, this certainly seems to establish a link between Purim and Babylonia. But what about the Sacaea?

The claim that the Babylonian sacaea included a rite of installing and later sacrificing a mock king is one of Frazer’s favourite pieces of fictional ethnography. He returns to it over and over again as if it were an established historical fact. But the only authority on the subject is Dio Chrysostom writing in the latter part of the first century AD. As an authority on ancient Babylonian Dio Chrysostom seems implausible, but since he mentions that a version of the Sacaea was later adopted by the Persians, it is just possible that his story contained features of genuine ethnography derived from contemporary Persia.

There is independent evidence about the Persian Sacaea. Strabo, writing around the beginning of our era, tells us that the patron-deity of the festival was the goddess Anahita, who was a transform of Aphrodite and, more remotely, of the ancient mother-goddess ishtar (i.e., Esther). The iconography of Anahita is now fully established though it was not known to Frazer (Ringbom 1957). As we shall see presently, this provides us with independent evidence that Anahita and the biblical Esther were directly associated.

Anyway, as cited by Frazer, Dio Chrysostom’s story was that during the five days of the Babylonian Sacaea “masters and servants changed places, the servants giving orders and the masters obeying them. A prisoner condemned to death was dressed in the king’s robes, seated on the king’s throne, allowed to issue whatever commands he pleased, to eat, drink, and enjoy himself, and to lie with the king’s concubines. But at the end of the five days he was stripped of his royal robes, scourged, and hanged or crucified” (Frazer 1900, vol. 2: 24).

Frazer speculates (without any evidence at all) that in this sacrificial role the mock king/victim was impersonating the God Tammuz, as corn-spirit and lover of ishtar. if you know your Bible, it will be obvious why Frazer found a correspondence between Dio Chrysostom’s Sacaea and the gospel accounts of the Crucifixion. Frazer assumed that both stories can be accepted as factual descriptions of separate historical events. He did not consider the possibility that both stories might be fictions derived from a common source. However, since Dio Chrysostom was an exact contemporary of the Christian Evangelists, this is an alternative that ought to be considered. Frazer himself inadvertently provides evidence that makes it seem rather probable.2

For those of you who are not diligent Bible readers, let me mark up some of the parallels. In the first place, almost all the references linking Jesus with kingship read as if they referred to mock kingship rather than real kingship. Mark’s account of the behaviour of the Roman soldiers might be mistaken for a quotation from Dio Chrysostom: “And they clothed him with purple, and platted a crown of thorns, and put it about his head, And began to salute him, Hail, King of the Jews! And they smote him on the head with a reed, and did spit upon him, and bowing [their] knees worshipped him. And when they had mocked him, they took the purple from him, and put his own clothes on him, and led him out to crucify him” (Mark 15:17-20). The other gospels include further details of mockery and abuse. After Jesus is disrobed, he is scourged. His cross carries a placard “This is the King of the Jews.” (Luke 23:38). Moreover, Christ as King seems to be directly linked with Christ as sacrificial victim. In John 6:15, immediately after the miracle of the “feeding of the five thousand,” we have: “When Jesus therefore perceived that they would come and take him by force, to make him a king, he departed again into a mountain himself alone.” The rest of this chapter contains material from which I have already quoted, which is, in a very explicit sense, predictive of the sacrament of the Eucharist. Then, at the beginning of the next chapter (John 7:1-2), we have: “After these things the Jews walked in Galilee: for he would not walk in Jewry, because the Jews sought to kill him. Now the Jews’ feast of tabernacles was at hand.”

The references to the forcible making of a king, the killing, the Eucharistic sacrifice, and the Feast of Tabernacles all seem to form a unity.

Frazer does not refer to this passage, though except for the reference to Tabernacles instead of purim it fits his scheme. But even without Frazer we have a puzzle here. What could be meant by saying that “they would come to take him by force to make him a king”? What sort of king would this be? Who are “they”? If you are prepared to accept the generality of my earlier remarks about the fragility of divinity of kings and the use of sacrifice for ritual purification then the notion of “forcibly making a king” is not far from “finding a sacrificial victim.”

Later in the same gospel (John 12:12-15) we have the passage which is now celebrated on palm Sunday, the Sunday before Easter. In the gospel story the context is the approach of the celebration of the passover in Jerusalem: “[They] took branches of palm trees, and went forth to meet him, and cried, Hosanna: Blessed is the King of Israel that cometh in the name of the Lord. And Jesus, when he had found a young ass, sat thereon; as it is written, Fear not, daughter of Sion: behold, thy King cometh, sitting on an ass’s colt.” if we omit the last sentence, which is a quotation from Zechariah, this suggests the procession of a mock king rather than a real one.

An equivalent passage in the Book of Esther (6:8-9) describes the procession in which Mordecai is installed as Viceroy: “Let the royal apparel be brought which the king useth to wear, and the horse that the king rideth upon, and the crown royal which is set upon his head … and bring him on horseback through the street of the city, and proclaim before him, Thus shall be done to the man whom the king delighteth to honour.” This honouring of Mordecai coincides with the disgrace of Haman, the previous Viceroy, and Haman is later hanged on the gallows which he had prepared for Mordecai. But Mordecai rides the real King’s horse, not “a young ass.” Frazer’s reference to this contrast of steeds is oblique and avoids direct reference to the gospel story. With regard to the Esther story, his thesis is that Mordecai and Haman are both mock kings. Haman is King “A” and is sacrificed; Mordecai is King “B” and rules in his stead; they are two aspects of the same mythical persona, two phases in the same annual cycle of regeneration.

Frazer does not explicitly mention the Christian Palm Sunday episode. instead, he compares the lordly procession of Mordecai with the very unlordly procession which is reported to have occurred in a Persian New Year Festival at an unspecified era. in the latter, which Frazer refers to as the “Ride of the Beardless One”: “A beardless and, if possible, one-eyed buffoon was set naked on an ass, a horse or a mule, and conducted in a sort of mock triumph through the streets of the city…. Riding on his ass and attended by all the king’s household … he paraded the streets and extorted contributions” (Frazer 1911-15, vol. 9: 402).

However, the heart of Frazer’s far from consistent argument is his comparison of the role reversal aspects of the Mordecai/-Haman dyad with the gospel story of Jesus and Barabbas. Here the gospel story is that it was customary that, at the time of the Jewish Passover, the Roman Governor should release a Jewish prisoner. When Pilate found Jesus guiltless (presumably of treason), he sought to release Jesus under the “custom.” But the mob demanded that Jesus be crucified and Barabbas (“who was a robber”) be released. But the name Barabbas means “son of the father.” Moreover, one early version of the relevant text in Matthew (27:17) gives him the name Jesus.3

Frazer’s main “conjecture” was that the Jewish celebration of Purim included a ritual drama in which a mock-king Mordecai somehow changed places with a mock-king Haman, with the latter ending up on a sacrificial scaffold. To this guess he added another: “if it be asked why one of these temporary kings should bear the remarkable title of Barabbas or ‘Son of the Father,’ I can only surmise that the title may perhaps be a relic of the time when the real king, the deified man, used to redeem his own life by deputing his son to reign for a short time and to die in his stead” (Frazer 1911-15, vol. 9: 419).

You think this is ludicrous? So do I. In any case, Frazer’s “surmise” inverts the evidence because, in the story, Barabbas does not die but survives. It is true that the sacrifice of firstborn children had at one time been widespread throughout the Semitic Middle East, but, even so, Frazer’s suggestion that the Crucifixion of Christ was an historical event in which the Jesus of history was playing the sacrificial role of Haman in a gory piece of ritual theatre pushes the limits of “conjecture” and surmise far out onto the horizon. Yet again it is not wholly impossible. The Roman practice of public execution by crucifixion was part of a pattern in which theatrical killings were commonplace. Slave gladiators killed one another in the circus. Heretics were devoured in public by wild beasts. Such practices were only one remove from human sacrifice of a purely ritual kind. In later ages Christians burned witches and heretics by the hundred, another form of ritual purification through “scapegoat” human sacrifice.

On the other hand, despite Frazer’s fantasies, there is no evidence that any kind of human sacrifice was ever an element in the Jewish purim ritual. On the contrary, purim seems to have been instituted precisely at a period when all forms of animal sacrifice other than the killing of the passover lambs in the Temple at Jerusalem were going out of fashion not only in palestine but among Jews everywhere.

Nevertheless, I am still puzzled. The similarity between Dio Chrysostom and the Evangelists seems too close to be accidental: what does it signify?

At this point, let me emphasise that the recognition of a link between the death of Haman and the Crucifixion of Christ was not just an aberration which resulted from Frazer’s frenzied search for dying vegetation gods; it was a part of Christian mediaeval tradition. This is shown by the unexpected appearance of a crucified Haman in the Sistine Chapel ceiling. The individual pictorial scenes in Michelangelo’s design all refer, in an immediate sense, to stories from the Old Testament. But they are also all predictive symbols of themes from the New Testament and from Catholic theology. The panel depicting the execution of Haman exhibits this double value with particular clarity. Haman doubles as Judas Iscariot and Christ crucified; Esther doubles as the wife of Ahasuerus and the Queen of Heaven.

But my concern at the moment is not with the way that mediaeval and Renaissance Christians read the stories from the Old Testament as “prefiguring” those of the New, but rather with how these same Old Testament texts were used by first-century Jews to justify their ritual performances. How can we possibly know? In the Bible, as we now have it, the only reference to purim outside the Book of Esther is in 2 Maccabees, where the “day before Mardocheus day” is made into a festival to celebrate the victory of Judas Maccabeus over the Syrian general Nicanor (2 Maccabees 15:36). This was a genuine historical event which occurred around 162 BC. But there is now good evidence, which was not available for Frazer, that at least in some forms of early Rabbinical Judaism, the Mordecai-Esther story was very important.

This new evidence comes from the Dura-Europos synagogue. Dura was a parthian city captured by the Romans in AD 165. It became a Roman military outpost, but was destroyed by the Sassanians in the year 256. It was then abandoned. The synagogue, which belongs to the third century AD, was excavated in the 1930s; the murals are now in Damascus and have been subject of a vast, highly polemical, scholarly literature (e.g., Kraeling 1956; Goodenough 1964). In what I shall now say, I shall, as far as possible, steer clear of the controversies.

Image 1 shows the Michelangelo panel I mentioned just now. I have discussed it in detail elsewhere. I show it here just to convince you that the death of Haman, which it depicts, really is made to double as an image of Christ crucified.

Image 1
Image 1: “The punishment of Haman” pendentive, Sistine Chapel, Michelangelo (1511).

Image 2 shows the West Wall of the Dura Synagogue in its present form in Damascus. It is, as you can see, quite well preserved, though the scenes in the uppermost of the three bands of pictures are badly damaged and therefore open to diverse interpretations. Since the Torah shrine is in the centre of this wall, we may safely assume that the pictures here displayed are of maximum importance. The experts are in reasonable agreement as to what most of the pictures represent. Their disagreements are mainly focussed on how far any particular picture is associated with other pictures as part of a unified programme and about what that programme might be. So far as this West Wall is concerned, some features seem quite apparent. For example, the colouring of the panels in the lowest register seems to imply that those on the right are balanced against those on the left. This hypothesis is borne out by the themes of the individual pictures. On the far right is a scene which shows the rescue of Moses from the bulrushes, which is treated as a resurrection from the dead. On the far left is another scene of miraculous resurrection from the dead: in this case, the story of Elijah restoring life to the widow’s child.

Image 2
Image 2: The West Wall of the Dura-Europos synagogue.

The panel immediately to the right of the Torah shrine introduces the theme of Kingship (Image 3). It carries a title: “The anointing of David by Samuel.”

Image 3
Image 3: “Samuel anoints David” panel on the West Wall of the Dura-Europos synagogue.

That immediately to the left of the Torah shrine is again about Kingship (Image 4). It shows at the right: a debatable scene which includes Ahasuerus and Esther, both of whom are named. Ahasuerus sits on the throne of Solomon, identifiable by its decorations. Notice Queen Esther’s very peculiar crown and the fact that she is seated at a higher level than the King.

Image 4
Image 4: “The Purim triumph” panel on the West Wall of the Dura-Europos synagogue.

At the left: Mordecai, in his triumph, riding a horse. He is named. The experts all seem to agree that the groom leading Mordecai’s horse must be Haman, though he is not named (Tawil 1979: 93). But the point I want to emphasise is that Mordecai is being treated as the counterpart figure to David on the other side of the Torah shrine: he is being made a King. The significance of the figures in toga-style costumes in the centre of the panel is disputed. The other figures are all in Persian court dress and they wear their hair long in a sort of Afro-style, which is here used as a marker of nobility. Mordecai wears the same costume as King Ahasuerus and this is also worn by Pharaoh in the right-hand panel. Mordecai is wearing a “diadem,” a fillet of white cloth around his head, to bind his hair.4

In pagan iconography the wearing of a cloth diadem was a mark of divinity and even the Roman Emperors who expected to be treated as gods did not assume the diadem before the time of Diocletian (Gibbon 1782, vol. 1: Ch. 13). The fact that Mordecai is wearing a diadem thus marks him out as very special indeed. incidentally, the peculiar stance of Mordecai’s horse is to be found elsewhere only in depictions of deities. in short, Mordecai seems to be represented as an example of Frazer’s ideal type, “the divine King.”

This fits with another striking feature of these panels which is that they include representations of pagan mother-goddesses, or rather two versions of the same pagan mother-goddess. In the nativity of Moses scene (Image 5) the naked lady lifting the child from the water, who, from one point of view represents the daughter of pharaoh who adopted Moses as her child, can be positively identified as Anahita/Aphrodite by the Jewel she wears round her neck (Goodenough 1964, vol. 9: 200). She was, as I mentioned earlier, the patron-goddess of the persian Sacaea. Likewise, the very distinctive turreted crown worn by Queen Esther, to which I previously drew your attention, is characteristic of images of her original namesake, the goddess Ishtar/Astare, of whom Anahita was a transform (Goodenough 1964, vol. 10: 50).

Image 5
Image 5: “The infancy of Moses” panel on the West Wall of the Dura-Europos synagogue.

I do not infer that the Jews of Dura worshipped Ishtar/Anahita, but rather that they had made the two Old Testament Queens, Pharaoh’s daughter and Esther, into a near-divine intermediary analogous perhaps to the Virgin Mary in contemporary Catholic Christianity. These female figures only appear in the lowest band of pictures, which is the most earth-bound.

Some contemporary Jewish scholars have found all this very shocking. Yet despite the Second Commandment and despite anything that might be inferred from Josephus, archaeology shows that Jewish synagogues in the early centuries of our era were regularly decorated with pictures. The consensus among the experts seems to be that there are no good grounds for supposing that the practices of this synagogue were specially heretical by the standards of the third century AD.

The panel immediately above the Torah shrine (Image 6) depicts: on the left, three temple symbols: namely, a seven branched lampstand (menorah), a palm-branch (lulab) and a citron (etrog; on the right, the scene of Abraham’s non-sacrifice of Isaac with the “ram caught in a thicket” in the foreground.

Image 6
Image 6: The Torah shrine at the Dura-Europos synagogue.

The design of the reredos area above this panel was changed on several occasions. Image 7 shows a reconstruction of the first version which has been interpreted as the Tree of Life in the Garden of Eden with the empty throne and the empty banqueting table of the future King-Messiah.

Image 7
Image 7: Reconstruction sketch of the first version of the reredos at the Dura-Europos synagogue.

In the first modification, an enthroned King-Messiah with two attendants and a Lion of Judah in front was added. In a final modification (Image 8), the whole of the original design was scrapped but the enthroned King-Messiah is now surrounded by representatives of the thirteen Tribes of Israel, each of the half tribes of Joseph being separately represented. The King-Messiah whose head is now destroyed wears the same persian regalia as the other divine Kings.

Image 8
Image 8: Reconstruction sketch of the final version of the reredos at the Dura-Europos synagogue.

To go back to the West Wall as a whole (Image 9), if we start at either the right or the left-hand corner of the bottom register and move first to the centre and then upwards, we have the same sequence. Death followed by rebirth. Then the initiation of a King-Messiah. Then sacrifice leading to the rule of the King-Messiah in the Last Days. I am not saying that all the pictures in the synagogue fit with this iconographic programme, but that is part of the story. It is a programme which could be fitted to Christianity with little modification.

I am only going to consider one further set of pictures (Images 9 and 10). These are the panels on the lowest register of the North Wall. Here again they need to be read from the corners towards the centre. The panel as a whole falls broadly into three sections: left-centre-right. All commentators agree that the left and centre sections depict the vision of resurrection from the dead in the Last Days and the creation of a unified Kingdom of Israel under David as King-Messiah as described in Ezekiel 37, but there is no corresponding consensus about the right-hand section.

Image 9
Image 9: The left-hand sections of “The cycle of Ezekiel” (“The valley of death,” “The mountain of transition,” “The valley of life”) panel on the West Wall of the Dura-Europos synagogue.

As you can see, image 10 shows an altar above which (or behind which) is a green military tent, which contains unidentifiable objects which might be looted Temple treasures. On the left, a splendidly armed military figure wearing a diadem seizes a figure dressed in court costume who clings to the altar. On the right, this same figure is being beheaded by another individual who might be his double. in the background are four armoured soldiers. One relatively plausible interpretation which I prefer to any of the alternatives that have been proposed is that what is depicted is the destruction and spoliation of the Temple at Jerusalem by Nebuchadnezzar. in that case, the nearest direct Biblical quotation is in 2 Chronicles 36: “[Jehoiakim] did that which was evil in the sight of the LORD his God. Against him came up Nebuchadnezzar king of Babylon, and bound him in fetters, to carry him to Babylon. Nebuchadnezzar also carried off the vessels of the house of the LORD to Babylon … and Jehoiachin his son reigned in his stead” (2 Chronicles 36:5-8). Elsewhere (2 Kings 25:27-30) we are told that after Jehoiachin had been taken captive to Babylon, his royal status was recognised and he lived as a Prince in Babylon for the rest of his life.

Image 10
Image 10: The right-hand section of “The cycle of Ezekiel” panel on the West Wall of the Dura-Europos synagogue.

The similarity of the names of father and son is intriguing. In apocryphal legend, Nebuchadnezzar did not just bind Jehoiakim in fetters but had him killed and his body thrown over the wall. If Frazer had seen this picture, he would surely have decided that the look-alike of victim and executioner was not accidental and that if the victim is Jehoiakim, the executioner is his son Jehoiachin. There is, so far as I know, no textual evidence to support any such “conjecture,” but, in context, it fits very nicely. The wall panel as a whole is about death and resurrection and the coming of the heavenly kingdom of the King-Messiah, and here we have a story about the death of Jehoiakim in Jerusalem who is succeeded by his son Jehoiachin, who rules in comfort in the “other world” of Babylon. But anyway, even if the precise Biblical reference of the picture is open to doubt, its meaning, at a mystical level, seems plain. In this context, the execution of a figure who is first depicted clinging to an altar must have sacrificial scapegoat implications. Before the nation can be reborn as a unified kingdom under a King-Messiah, purification is necessary. The destruction and spoliation of the Temple by a divinely appointed foreign agent, either Nebuchadnezzar or the Romans as the case may be, is here depicted as such a purification by human sacrifice.

There is one more detail of Dura archaeology which I must mention before setting back to my main Frazerian theme. Although the Synagogue is, by chance, much the best preserved, it was only one of a number of rather similar buildings. There were at least nine other religious buildings in Dura at the time of its destruction, including a small Christian baptistry. The principal deities of these sanctuaries included Bel, Zeus, Adonis and Mithras. They all had painted wall decorations and indeed the pillars of the Christian font were given an identical decoration to those of the Torah shrine in the Synagogue, which must have been painted by the same craftsmen. Though differing in ostentation, the general layout of all these religious buildings was rather similar. In particular, the arrangement of the principal shrine included a representation of the cult deity. In the Christian case, this was a picture of Christ as the Good Shepherd. In the case of the synagogue, God appears only in the form of intrusive hands, but the focal picture of the enthroned King-Messiah above the Torah shrine is in the position which would have been occupied by a direct representation of God in the other cases.

So where have we got to? Possibly just a bit further than you may suppose.

Most histories of early Christianity have been written by Christian theologians who tend to assume that the polemical hostility towards Jews and Pagans displayed in the writings of the early Christian fathers was characteristic of Christians in general. This seems to me unlikely. During the first century AD Christianity developed as a sect within Judaism. There were many such sects. In the year 90 AD the Orthodox Jews in Palestine formally excommunicated all sectarians, including Christians. But twenty years later we still find Ignatius complaining to the Magnesians in Asia Minor: “It is monstrous to talk of Jesus Christ and to practice Judaism” (To the Magnesians 10.3, in Lake 1912: 207). It seems to me very likely that some of the Christians of Dura-Europos were also practicing Jews.

In broad outline, Frazer’s belief that all the religious systems of the ancient Middle East had much in common was certainly correct. New cult practices do not emerge by spontaneous generation; they start out as deviant variations of what was there before. And in situations of this sort ordinary people (as distinct from cult professionals) seldom feel that sectarian differences are either radical or permanently hostile. Differences of religious belief and practice are much more matters of interpretation than of objective fact. In modern Sri Lanka a visiting tourist who attends a festival at any of the principal cult centres will be told that the shrine in question is Buddhist, or Hindu, or Christian Catholic as the case may be. But such a traveller has only to use his eyes to see that those who are participants in the rituals are indiscriminately Buddhist, Hindu, Catholic and Moslem. Their interpretations are different. Where one devotee may perceive an image of the Christian Madonna, another sees the Goddess pattini. Rival saints may denounce one another as frauds and heretics but for ordinary people one god is very like another. It is in that kind of climate that early Christianity first developed.

But where does Dura fit in? It is very unlikely that the ritual proceedings in the Dura synagogue included sacrifices of any sort. The synagogue was a place of communal worship and prayer; a place for the preaching and study of the law. Religious leadership was in the hands of scholars rather than priests. This synagogue was built less than 200 years after St. paul wrote his Epistles to the Corinthians, the first of which contains the foundation text (later copied by the Evangelists) justifying the Christian sacrament of the Eucharist. Throughout those Epistles, paul repeatedly reaffirms his claim to be a Jew as well as a Christian. He explicitly treats animal sacrifice as a characteristic of pagan rather than Jewish ritual performance. Among the themes which pauline Christianity takes over from first-century Judaism are those of resurrection from the dead and the Messianic replacement of corruption by incorruption: “So also is the resurrection of the dead. It is sown in corruption; it is raised in incorruption; it is sown in dishonour; it is raised in glory; it is sown in weakness; it is raised in power…. The first man is of the earth, earthy; the second man is the Lord from heaven …” (1 Corinthians 15:42-47). This also is the theme of the Messianic pictures in the Dura synagogue. The mythology entails the repeated replacement of the evil king by the divinely ordained, virtuous Messiah. The replacement of sinful Haman by virtuous Mordecai is one such instance, and indeed, in the pictures, it has pride of place.

But the analogy with the Christian story as perceived by Frazer was back to front. If the Crucifixion had really been an enactment of the Mordecai/Haman mythology, then Barabbas should have died and Jesus should have survived. But, in fact, in the Christian myth, Barabbas is redundant. Jesus fills both roles: he is the son of the father who dies and the son of the father who lives for ever. If he dies as Haman, then he is reborn as Mordecai. That was how Michelangelo seems to have perceived the matter and I think he was right.

But Frazer was not altogether wrong. The Dura paintings demonstrate the strength of Messianic beliefs in a Judaism that was contemporary with early Christianity, but they also show that the Mordecai/Haman mythology had a very prominent place in those beliefs. In the Mediaeval versions of Purim with its echoes of Carnival and Saturnalia, Mordecai and Haman were not just the inverse of each other, they were muddled up. In the Christian story the Christ who dies as sin-loaded scapegoat, as “the full, perfect and sufficient sacrifice” reappears as the King of Heaven in all his glory. Frazer perceived this similarity but confused matters by his insistence that symbolic representations must always be seen as survivals of actual historical events—in this case, Haman sacrifice. But that he should have perceived the similarity at all is remarkable and he deserves our praise on that account.


Frazer, James G. 1900. The golden bough: A study in magic and religion. Second edition in three volumes. London: Macmillan.

———. 1905. Lectures on the early history of the kingship. London: Macmillan & Co.

———. 1911-1915. The golden bough: A study in magic and religion. Third edition in twelve volumes. London: Macmillan.

———. 1922. The golden bough: A study in magic and religion. Abridged edition. London: Macmillan.

Gibbon, Edward. 1845 [1782]. The history of the decline and fall of the Roman Empire. In eight volumes. Revised and annotated by Henry H. Milman. Paris: Baudry’s European Library.

Goodenough, Erwin R. 1964. Jewish symbols in the Greco-Roman period. Volumes 9-11, Symbolism in the Dura synagogue. New York: Pantheon Books.

Kraeling, Carl H. 1956. The synagogue: The excavations at Dura-Europos. New Haven: Yale University Press.

Lake, Kirsopp (trans.). 1912. The apostolic fathers, vol. 1. London: Heinemann.

Ringbom, Lars-Ivar. 1957. Zur Ikonographie der Göttin Ardvi Sura Anahita. Åbo: Åbo Akademi.

Tawil, Dalia. 1979. “The Purim panel in Dura in the light of Parthian and Sasanian art.” Journal of Near Easter Studies 38(2): 93-109.

Van Gennep, Arnold. 1909. The rites of passage. Chicago: University of Chicago Press.

Edmund Leach was Professor of Anthropology at Cambridge University. He published many books, including Political Systems of Highland Burma, Rethinking Anthropology, and Levi-Strauss.


* Publisher’s note: The publication of the 1982 Frazer Lecture by Edmund Leach is part of Hau’s project to unearth previously inedited material of great relevance for the history of ethnographic theory. This unpublished lecture was identified by Giovanni da Col and remained neglected in King’s College Archive in Cambridge. Anastasia Piliavsky kindly offered to act as editor of the original typescript and to contact Leach’s literary executor on behalf of the journal.

Editor’s note: Edmund R. Leach’s previously unpublished Frazer Lecture on “Kingship and divinity” was delivered at the University of Oxford on 28th October 1982. Leach’s remarks in the typescript (held in the archives of King’s College, Cambridge) suggest that he intended to put it into print, but it has remained, now twenty-two years on from his death, unpublished. The piece has the brusque style of writing for verbal presentation, which was nevertheless characteristic of much else that Leach published. The text has been left in its original form, albeit some light proof-reader’s touch-ups, and the addition of references and images to replace the slides that originally accompanied the lecture but which are now missing. I am grateful to Louisa Brown, Leach’s literary executor, for her permission to publish the piece. Anastasia Piliavsky (King’s College, Cambridge).

** All biblical references are to the King James Bible (A.P.).

1. This declaration, which Leach dates to the sixteenth century, can actually be found in a sixth-century Babylonian Talmudic text cited here (A.P.).

2. One of his parallels for the mock king in Dio Chrysostom’s Sacaea comes from an almost identical description of the way the Saturnalia was celebrated by Roman soldiers stationed on the Danube at the beginning of the fourth century ad. Frazer assumes that this is an ancient tradition deriving from the hypothetical time when Saturnalia was celebrated in the Spring (Frazer 1911-15, vol. 9: 308). The story relates to the martyrdom of St. Dasius, but it can hardly be a direct borrowing from the gospels since it has all the details of Dio Chrysostom’s version.

3. Frazer also draws our attention to a passage in Philo Judaeus, who reported that when Herod Agrippa passed through Alexandria on the way to take up his appointment (by the Emperor Claudius) as King of Judaea, he was mocked by the populace in a burlesque street play in which a harmless lunatic called “Carabas” was dressed as a king and proclaimed as “Lord.” Frazer suggests that Carabas is a corrupt version of Barabbas (Frazer 1911-15, vol. 9: 418 seq).

4. In the English text of the Book of Esther this is referred to as “the Crown Royal” and the same term is applied to Esther’s crown. In the Hebrew the term is nezer, which means a mark of separation of consecration which comes from nazir (“to consecrate”). in other parts of the Bible nazir may refer to either a “consecrated prince” such as Joseph (Genesis 49:26) or a priest-hero such as Samuel, but, in general, it is closely associated with the theme of messianic salvation. The special outward mark of a Nazarite was that he wore his hair long.