Magic and Religion

II. STAGES IN MR. FRAZER'S THEORY

Let us now trace the stages of Mr. Frazer's theory that the Sacæan victim is both god and king.

1. First in order of statement comes the description of the Sacæa, combined from Athenæus, who mentions no victim, and Dio Chrysostom, who does. We learn (from Mr. Frazer, not from Dio)205 that the victim 'dies in the king's stead.' But 'we must not forget that the king is slain in his character of a god, his death and resurrection, as the only means of perpetuating the divine life unimpaired, being deemed necessary for the salvation of his people and the world.'

That is Mr. Frazer's theory: we have seen no proof of it, we have remarked that sacrificed victims are not hanged; that kings are not scourged; that there is no evidence beyond conjecture for an earlier Babylonian process of burning; while conjecture also explains whipping and hanging as a 'mitigation,' or alleges that possibly the victim was hanged first and burned afterwards.

Here the king is certainly not,206 on the face of it, a god of vegetation: if anything, he is more like the Chitome in Congo, who was a 'pontiff.' His credulous people believed that the world would end if the Chitome died a natural death, 'so when he seemed likely to die' he was clubbed or strangled. He was sacrificed to no god whom he incarnated.207 He was not clubbed once a year (like the Babylonian king of Mr. Frazer's theory); he was given a rude euthanasia 'when he seemed likely to die.' Does science ask us to believe that each Babylonian king had the cosmic rapport of a Congo savage pontiff, and was sacrificed after a year's reign, because a savage pontiff in Congo is put to death, not annually, but 'when he seems likely to die'?

Here, whatever science may expect us to believe, we are told by Mr. Frazer that the king in Babylon was annually sacrificed, as a god, indeed, but not explicitly as a god of vegetation, who has a resurrection.

2. A Babylonian god of vegetation, and a known god, appears in ii. 123, 124. This god is Tammuz. We hear that 'water was thrown over him at a great mourning ceremony, at which men and women stood round the funeral pyre of Tammuz lamenting… The dead Tammuz was probably represented in effigy, water was poured over him, and he came to life again.' Mr. Frazer does not here plead for a human victim. The festival 'doubtless took place in the month Tammuz (June-July),' or in different places, at different times, from midsummer to autumn, or from June to September, as the late Mr. Robertson Smith calculated. Tammuz, so Mr. Sayce is cited, 'is originally the spring vegetation, which dies in his month, Tammuz or Du'ûzu' (June-July).

Here, then, we have a death and resurrection of Tammuz. It occurs in June-July, or June-September, and Tammuz is undoubtedly the god of spring vegetation. But Mr. Frazer does not here tell us that the king of Babylon is also Tammuz. Tammuz is not whipped and hanged at the Tammuz feast in July. His dead body is 'probably' a dummy.

In vol. ii. 253 Mr. Frazer returns to the victim, the mock-king, of the Sacæa. But he says nothing here about the real king of Babylon. He wishes to show how and why the victim is divine. Now, in ii. 26, we were told that the victim is divine because he 'represents a dying god.' 'For we must not forget that the king dies in his character of a god.'…

Was Mr. Frazer satisfied with this explanation given in ii. 26? Apparently not; for208 he gives a new explanation and a different one. 'It seems worth suggesting that the mock-king who was annually killed at the Babylonian festival of the Sacæa on the sixteenth day of the month Lous may have represented Tammuz himself.' Here the Tammuz dummy or effigy of ii. 123, 124, is, perhaps, discarded. Still, if a real live Tammuz was burned on a funeral pyre209 his ashes might well be represented by a dummy. It has not yet occurred to Mr. Frazer, as it does later, to have the re-arisen god personated by a living human counterpart (Mordecai in a later page) of the dead Tammuz (Haman). The festival of the Sacæa is now a Tammuz festival, a religious feast, and, indeed, is identical with that of ii. 123, 124, for it occurs in the month Lous. Now Lous, says Mr. Robertson Smith, 'answered to the lunar month Tammuz,'210 and the month of Tammuz211 was June-July, or June-September.

There could not surely be two Tammuz feasts in the month Tammuz? We are therefore confronted by the singular facts that Tammuz lay 'on a funeral pyre'212 and also that, as the Sacæan victim, who, Mr. Frazer thinks it 'worth suggesting' personated Tammuz, he was at the same feast, the Sacæa, whipped and hanged.213 Mr. Frazer goes on: 'If this conjecture is right, the view that the mock-king at the Sacæa was slain in the character of a god' (Tammuz) 'would be established.'

But it was established already, was it not on other grounds, to Mr. Frazer's satisfaction, in ii. 26? There the criminal victim died as a king, and as a god, for the king was a god, and so was his proxy. Now, on the other hand, if Mr. Frazer's latest conjecture is right, the victim dies as a real known god, Tammuz. We keep asking, Was the king also an incarnation of Tammuz? May I not be excused for surmising that we have here an hypothesis in the making, an hypothesis resting on two different theories? If Mr. Frazer holds that the king of Babylon was also Tammuz, as the mock-king was, here was the opportunity for saying so, and proving the fact from Babylonian texts.

Mr. Frazer here gives us a Tammuz feast in which Tammuz lies on a funeral pyre, and also a Tammuz feast in which the human representative of that deity is whipped and hanged, while 'the dead Tammuz was probably represented in effigy,' water was poured over him, and he came to life again. How? In the person of Mordecai? These are the results of ii. 123, 124, and of ii. 253, 254.

These things are, confessedly, conjectures. But one thing is quite certain: the Sacæa, wherein Tammuz either lay on a funeral pyre, and afterwards had water poured over him, 'probably in effigy,' or was hanged, was a festival of June-July. Variations of calendars, however, might make the Sacæa fall 'from midsummer to autumn or from June to September' (ii. 123, note of Mr. Robertson Smith). These dates are remote from Eastertide.

To this point Mr. Frazer214 promises 'to return later.' He does so in the most disconcerting manner. For when he returns the Sacæa, which were in the month Tammuz, June-July,215 startle us by being held in March or March-April.216 May I not say that I seem to detect traces of an hypothesis in the making, and of discrepant theories? We have already been rather puzzled by the Tammuz on a funeral pyre, who has cold water poured over him, 'probably in effigy,' and also is honoured by being whipped and hanged in the person of a human representative, a mock-king, at the same festival. But perhaps there were two Tammuz feasts in the month of Tammuz? And possibly the victim was whipped and hanged at one of them, while his mortal remains were burned on the pyre at the other? 'It is quite possible,' says Mr. Frazer, when explaining why a victim of a sacrifice was hanged, not burned as is usual, 'that both forms of execution, or rather of sacrifice, may have been combined by hanging or crucifying the victim first and burning him afterwards;'217 but he neglects the buxom opportunity of corroborating this conjecture, by referring to the Tammuz victim who had both a funeral pyre and a gibbet, in ii. 123, 124, 253, 254.

III. A POSSIBLE RECONCILIATION

There is, perhaps, a mode of reconciling the dates of the Tammuz festivals, at one of which Tammuz was honoured with a pyre, at the other (in the person of his representative, the Sacæan mock-king) with a gibbet. Dr. Jastrow places a Tammuz feast in the fourth month, which, if the Babylonian year begins, as Mr. Frazer says it does, with the month Nisan, means that the fourth month and a Tammuz feast occurred in our June-July. But Dr. Jastrow also writes that in the sixth Babylonian month, our August-September, 'there was celebrated a festival to Tammuz.'218

Thus Tammuz might have his gibbet in June-July, and his pyre in August-September. But alas! this will not do, for the pyre is of June-July.219 Nor can he have his gibbet in August-September, as I had fondly hoped, for he is to be identified with the mock-king of the Sacæa, and the month of his hanging is Tammuz, Lous, or June-July, if Mr. Robertson Smith is right.220 Thus I really fail to believe that Tammuz could have both a burning and a hanging in June-July. I hoped that Dr. Jastrow's two Tammuz feasts had solved the problem, but I hoped in vain.

IV. THE SACÆA SUDDENLY CHANGES ITS DATE

Meanwhile, even though we have allowed for two Tammuz feasts, are we also to admit a third Tammuz feast at the March festival of the Sacæa? For in vol. iii. 151-153, March has become the date of the Sacæa, rather to our surprise, for the date had been June-July.221 Now three Tammuz feasts in six months seem one too many, if not two. Consequently the arguments which in ii. 123, 124, 253, 254, show the Sacæan victim, because he died in the month Tammuz, to represent the god Tammuz fail, perhaps, if the victim really died in March, at the Babylonian Zakmuk, or Zagmuku, a feast in honour, not of Tammuz, but of Bau (a goddess), and later of Marduk.222 Neither Bau nor Marduk is Tammuz; nor does the victim seem likely to represent Tammuz, after his death is shifted from the Tammuz feasts of May-June or June-July, July-August, to March, when the feast was really in honour, not of Tammuz, but of Bau, or later, of Marduk.

All our difficulties, indeed, pale before the fact that the date of the Sacæa, when the possible Tammuz victim was hanged, is fixed twice; once, with much show of reason and 'with unconcealed delight,' in June-July, in the second volume; while, next, it is argued from, in the third volume, as if the date were March-April.

I conjecture, therefore, that the July date was not inconsistent with what is now Mr. Frazer's theory when he revised his second volume. Otherwise he would not have said that Mr. Robertson Smith's decision as to the July date 'supplies so welcome a confirmation of the conjecture in the text,'223 and then, in iii. 152, 153, have proceeded to argue on the presumption that Mr. Robertson Smith's calculations may be, for the purposes of the theory, disregarded. And they are disregarded, as we shall see. If they were dubious, they should never have been welcomed.

V. VARIOUS THEORIES OF THE VICTIM

Meanwhile, for our own argument, as to the precise nature of the Babylonian King's divinity, vegetable or not, I do not think that we have yet found the King of Babylon explicitly identified with a god of vegetation.

The victim, remember, was at first divine, either as proxy of the king, incarnating, I think, a god unknown; or as full of cosmic rapport, as a man-god of the second species.224 Next his divinity was established, if Mr. Frazer rightly conjectured that he 'represented Tammuz himself.'225 Next he was a criminal vicariously sacrificed for 'the saving of the king's life for another year.'226

Next 'it would appear that the Zoganes' (the same old victim) 'during his five days of office personated not merely a king but a god, whether that god was the Elamite Humman, the Babylonian Marduk, or some other deity not yet identified.'227 Next the victim personated 'a god or hero of the type of Tammuz or Adonis, (and) enjoyed the favours of a woman, probably a sacred harlot…' in addition to the caresses of the royal seraglio.228 Next the indefatigable victim represented the king, 'the human god, the Saturn, Zoganes, Tammuz, or whatever he was called,' though all we know of the god Zoganes is that Zoganes was the title of the slave lord of the household at the Persian Sacæa.229

It would thus appear almost as if all gods are one god to Mr. Frazer by a kind of scientific 'Henotheism.' Humman or Saturn, Zoganes or Tammuz, Marduk or Adonis, any one of them, or all of them, will do for the king to incarnate or personate. Any one of them, or all of them, will figure as representatives of vegetable life in company with Zeus and the horses of Virbius! 'We may conjecture that the horses by which Virbius was said to have been slain were really embodiments of him as a deity of vegetation.'230 Now let me too say 'we may conjecture.' Mr. Frazer tells us that 'horses were excluded from the grove and sanctuary' of Virbius.231 Is it putting too great pressure on evidence to conjecture that the horses, while being driven out, were whipped? Now the horses embodied, perhaps, as we are told, a deity of vegetation. They were whipped, and therefore it was usual to whip the representatives of a deity of vegetation. This solves our problem, why was the victim, the divine victim, whipped?

Seriously, have we not in all this book to do with that method of arbitrary conjecture which has ruined so many laborious philosophies of religion?

As to one essential conjecture, that the Babylonian, or rather the Persian, kings represented a deity of vegetation, I can offer only one shadowy testimony. Nebuchadnezzar for a while exhibited a caprice in favour of a purely vegetable diet. This may have been a survival of a royal taboo. As a god of vegetation, a king would not eat vegetables any more than a savage usually eats his totem. But some savages do eat their totems on certain sacred occasions, and that may be the reason why Nebuchadnezzar, for a given period, turned vegetarian.

VII

ZAKMUK, SACÆA, AND PURIM

It is necessary to get the death of the Sacæan victim into touch with Easter. The Sacæa, when he died, had been in June-July, in vol. ii., in Mr. Frazer's first edition, before he evolved his theory. When the theory is evolved, in the second edition and third volume, the Sacæa prefer to occur in March-April, which gets the sufferings of the mock-king into touch with the Jewish Purim, and so within measurable distance of our Passion Week, though the June-July date of the first edition survives in the second volume of the new edition. The change of date of the Sacæa is arranged for by the plan, rejected by Meyer and Jastrow, of identifying the Persian Sacæa and the Jewish Purim with the ancient Babylonian Zagmuk or Zakmuk, a New Year festival of March-April.232 To be sure, if that be the date, we seem bereft of our useful Tammuz, from whom, in ii. 254, it was conjectured that the victim mock-king derived his divinity, an old superstitious belief which 'shed the halo of divinity' on the victim of Calvary. For the Tammuz feast was certainly in June-September. However, perhaps there were three Tammuz feasts, resurrection and all, and Mr. Frazer's last choice of a date, in March-April, has the immense advantage for his theory of getting us near Eastertide.

But did the Sacæa actually desert their old date, June-July? To prove that we must identify the Sacæa, a Persian, with Zakmuk, a Babylonian feast, which really fell in March or April. The old Babylonian feast, Zakmuk, is known to the learned through inscriptions. We have seen that M. Cumont and Herr Meissner inclined to regard Zakmuk as identical with the Sacæa, while the feast Zakmuk-Sacæa is supposed by Mr. Frazer to be the origin of the Jewish Purim. But the Sacæa fell in the Macedonian month Lous, as Athenæus tells us according to Berosus, a Babylonian priest, using the Macedonian Calendar. And Lous, as Mr. Robertson Smith proved, was our July.233 Zakmuk, on the other hand, fell in our March-April, and Purim in our March, neither of which is July, when the Sacæa were held.

Now it is desirable for Mr. Frazer's argument that the Sacæa should fall, not in July, as it did in ii. 254, but in or about Eastertide. Mr. Frazer therefore shifts the Sacæa from July to Eastertide in face of difficulties.

All we know concerning Zakmuk is234 that this feast, originally a feast of Bau, says Dr. Jastrow, fell about the vernal equinox (near the beginning of the old Babylonian year); that, after a certain period, it was held in honour of the chief god of Babylon, named Merodach; that a council of gods was thought to meet in Merodach's temple, under his presidency, and that they determined the fate of the year, 'especially the fate of the king's life.' The festival existed as early as 3000 B.C., whereas the Sacæa, 'so far as appears from our authorities, does not date from before the Persian conquest of Babylon' (536 B.C.).235 But in spite of dates it is desirable for Mr. Frazer's purpose to identify the Persian Sacæa with the Babylonian Zakmuk. For, if he succeeds in this, then Sacæa must fall when Zakmuk fell, and nearly when Purim fell, at – or not so very far from – Eastertide. But236 Sacæa was eagerly welcomed by Mr. Frazer as a July, not a spring, feast, whereas, in iii. 152, Sacæa is identified with Zakmuk, which did fall in spring. Again, we have not even a hint that any mock-king, or Tammuz man, or anybody, was slain at the Babylonian feast of Zakmuk, as a man was slain, says Dio, at the Sacæa. However, Mr. Frazer tries to show that Sacæa and Zakmuk may be the same feast. For Sacæa and Zakmuk are names that resemble Zakmuk and Zoganes.237 We may reply that the word Sacæa also rather closely resembles the name of the tribe of Sacæ, from whom the Perso-Greeks derived the word Sacæa, while the Sacæa were held to commemorate a victory over the Sacæ. Again the word Sacæa, which was a drinking feast, resembles the word Sáki, Persian for a pourer forth of wine. I The word Sáki is Arabic, being the nomen agentis of the verb Saḳi "to water" (abreuver). This root is common to several Semitic languages —e. g. Hebrew and Æthiopic – and if we could prove the word Sacæa to be of native Babylonian origin, it might very probably come from the same root,' Mr. Denison Ross informs me. In any case we cannot build on resemblances in the sound of words. That argument for the identification of Zakmuk and the Sacæa fails.

Next Mr. Frazer contends that since, at Zakmuk, the gods determined the fate of the king's life, it was a critical time for the king. Now 'the central feature of the Sacæa' appears to have been 'the hanging of the mock-king for the saving of the real king's life.'238 Here, then, are two critical hours for the king: one at Zakmuk, when the gods settle his fate; one at the Sacæa when his life is saved by the execution of his proxy. Are not then these two critical periods one period, and is not Sacæa another name for Zakmuk?239 But Mr. Frazer has also told us that the main feature of the Sacæa was the death of a man who represented Tammuz, and was killed after doing sympathetic magic with a sacred harlot.240 Was there, then, in connection with this Tammuz man, a third Tammuz feast in March-April, for there were two, in June-September? Thus, even if we could admit that, because two periods are critical, both are the same period, yet as the victim of the Sacæa was a Tammuz man, slain to do good to the crops, we are unable to concede that he also died 'in the king's stead,' and to save his life, unless the king was Tammuz. Besides, no authority tells us that either, or both, of this victim's deaths occurred at the Babylonian feast of Zakmuk: it occurred at the Persian feast of the Sacæa, if at all.

Indeed, even if Mr. Frazer's two arguments for the identity of Zakmuk and Sacæa were persuasive (and' how persuasive they are we have seen), there would remain a difficulty. For Berosus says, as we saw, that the Sacæa fell on Lous 16, which is July, whereas Zakmuk fell in March-April.

I. HISTORICAL DIFFICULTY

This obstacle seems to be, and really is, insuperable. But Mr. Frazer, undaunted, writes: 'The identification of the months of the Syro-Macedonian Calendar is a matter of some uncertainty; as to the month Lous in particular the evidence of ancient writers appears to be conflicting, and until we have ascertained beyond the reach of doubt when Lous fell at Babylon in the time of Berosus' (say 200 B.C.) 'it would be premature to allow much weight to the seeming discrepancy in the dates of the two festivals' (namely Zakmuk and the Sacæa). Henceforth Mr. Frazer's hypothesis seems to me to proceed on the fancy that Sacæa and Zakmuk are identical, which is impossible, since the Sacæa fall in July or September, and Zakmuk in March-April.

It is absolutely certain, historically, that Sacæa and Zakmuk cannot be identical. They were as remote in date as they well could be. For the conflicting evidence of ancient writers as to the date of the month of the Sacæa, namely the Macedonian month Lous Λῶος Mr. Frazer gives two references. The first is to Mr. Robertson Smith's proof that Lous is July.241 That does him no good. The second is to Smith's 'Dictionary of Greek and Roman Antiquities.'242 In that work I read that the only doubt as to the month Lous is whether it fell in July or September. Smith's 'Dictionary' is a book so common and accessible that I need not inflict on the reader the nature of the conflicting evidence. It is enough to say that the month of the Sacæa, Lous, was almost certainly July, but, if not July, was undeniably September. Now neither July nor September is Eastertide, or near it. So that the effort to make the Sacæa identical with Zakmuk, and therefore more or less coincident with Purim, and with our Easter, is an absolute failure. The Jews, then, could not (as in Mr. Frazer's theory) borrow abroad a July or September mock-king, and attach him to a vernal festival, their Purim. Thus, as Zakmuk is several months remote from the Sacæa, it is not identical with the Sacæa. Mr. Frazer himself says: 'If the Sacæa occurred in July and the Zakmuk in March, the theory of their identity could not be maintained.'243 But he loses, rather than gains, if the Sacæa were in September, and that is the only possible alternative. The game is over; the mock-king of Babylon died, if at all, in July or September, at the Sacæa; not at Zagmuk or Zakmuk, in March-April. There is not a known hint that any mock-king died in Babylon about Eastertide, or earlier, at the feast of Zakmuk.

I confess that when I found Mr. Frazer declining to 'allow much weight to the seeming discrepancy in the dates of the two festivals,' till it was 'ascertained beyond the reach of doubt when Lous fell at Babylon in the time of Berosus,' I presumed that 'the apparently conflicting evidence of ancient writers' meant a difference of opinion as to whether Lous was a spring or a midsummer month. But I looked at Smith's 'Dictionary' and found nothing of the sort! The difference of opinion, the conflict of evidence, is concerned (see Smith) with the question whether Lous was September (as it seems to have been in the time of Philip of Macedon) or whether it was July, as in the time of Plutarch. Neither opinion gives Lous the faintest chance of being a spring month. Therefore the vernal Zakmuk is not the Sacæa; therefore there is not the ghost of a reason for guessing that a mock-king was hanged at Zakmuk; therefore Zakmuk, in April, cannot lend a hanged mock-king to Purim, in March; therefore Purim, having no slain mock-king, cannot hand one on to Eastertide, which, moreover, does not occur at the same date as Purim, but some weeks later, as may happen. Therefore the mock-king, if he had been divine (which he was not), and if he had been sacrificed (which he was not), could not have lent his 'halo of divinity' to gild the Cross at Calvary. But that he did so is Mr. Frazer's hypothesis – sometimes.

II. PERSIANS ARE NOT BABYLONIANS

The Sacæa, according to all our authorities, was a Persian, not a Babylonian, feast. We have not a tittle of evidence to show that the Babylonians, with whom Zakmuk was a feast of old standing, ever heard of the Sacæa before they were conquered by the Persians (B.C. 536). Mr. Frazer admits this: the Babylonian custom, 'so far as appears from our authorities, does not date from before the Persian conquest; but probably it was much older.'244 Why 'probably'? On the strength of this 'probably' Mr. Frazer calls the doings at the Persian Sacæa 'a Babylonian custom.'245 It was a custom of the Persian conquerors of Babylon, if we can believe Dio Chrysostom; but we have no evidence that it was a Babylonian custom. Yet it 'has just got to be' a Babylonian custom that Mr. Frazer may attach it to a vernal Babylonian feast, Zakmuk, and so to Purim, and so to Eastertide.