Merlin and his magic forest must be cut down to size before we can see the shape of the trees, but first let us examine the undergrowth. After graduating from the University of Bucharest in 1928 Mircea Eliade spent three years in Calcutta studying classical texts of Indian mysticism. His special concern was with the ascetic and ecstatic techniques whereby the would-be saint, having achieved a state of psychological dissociation, can persuade himself that he has access to the powers of the other world, being himself neither alive nor dead, neither on earth nor in heaven. All of Eliade’s subsequent writings have been concerned with this central theme, the symbolic modes through which communication is established between the sacred and the profane. His attitude is that of a Jesuit: he is scholar and believer at the same time. Eliade left Romania at the end of the war and later settled in Paris; for the past ten years he has been Professor of the History of Religion in the University of Chicago. The “history” which he pursues is not concerned with chronological sequences or the analysis of the causes and consequences of particular events, but rather with the development of human thought over vast regions of time and space. But this evolution is a very simple two-stage affair: for Eliade modern man stands to archaic man as Christianity to pre-Christianity. The cosmological ideas which characterize archaic religion are everywhere the same and may be exemplified, in Frazerian fashion, by any snippets of exotic ethnography which conveniently come to hand. Modern man is unique because the religious mythology of Judeo-Christianity is set in a matrix of chronological time. Christian time is on-going, it had a beginning and will have an end but it is non-repetitive, it is “historical.” In all other religions, time is a cyclical process. Instead of advancing boldly towards the discovery of a New Jerusalem, archaic man is content to engage in recurrent but imperfect imitation of divinely ordained archetypes fashioned by the ancestral deities in the first beginning.

ELIADE’S DIAGNOSIS may be challenged on many different grounds: bad history—there has never been a radical discordance between Christian cosmology and cyclical notions of time; bad ethnography—it is not true that the cosmologies of “archaic” man always incorporate notions of cyclical time; bad method—comparative ethnography in the style which Eliade employs, can only illustrate by example, it can never properly be used as a basis for generalization; bad psychology—Eliade takes for granted the Lévy-Bruhl fashions of his youth which assumed that ethnographic evidence reflects a pre-logical archaic mentality radically different from that of rational thought (Lévy- Bruhl himself abandoned this theory in his later years); confusion of terms—the most interesting parts of Eliade’s writings become fogged by his failure to distinguish clearly between the content of a set of symbols and its structure. It is only fair to add that in the last chapter of his latest book, Mephistopheles and the Androgyne, Eliade shows himself sensitive to most of these criticisms and seems dimly aware that he may have been maintaining indefensible positions, but he makes no retraction, and so he must be judged.

Whatever may be the faults of method, it is easy to see why the thesis as such should appeal strongly to several kinds of anti-positivist. I observe that the translation of Eliade’s larger works was sponsored by the Bollingen Foundation, always a stalwart patron of Jungian psychology; several others were originally prepared for the Jungian Eranos circle at Ascona; three books carry the imprint of the Catholic publishers Sheed and Ward; the laudatory comments quoted on the dust covers are consistently of Jungian or Catholic origin. One may suspect that harsh objectivity is not one of Eliade’s outstanding virtues.

A difficulty of another kind is bibliographic and linguistic. Down to about 1938 most of Eliade’s writings were published either in Romanian or bad French. They included studies of Asian alchemy (1935), Yoga (1938), and Babylonian cosmology (1937). Most of his current volumes appear to be revamped versions of these early works. Thus The Forge and the Crucible, English edition 1962, is essentially the same as the 1935 and 1937 Romanian items, and although the author claims that “a number of chapters have been added and the whole work rewritten to bring it into line with the most recent views on the subject” it is still entirely naive. Every methodological error of which Sir James Frazer and his contemporaries have ever been accused is here exhibited in its purest form. The translation factor introduces additional problems, sometimes of labyrinthine complexity: Yoga, published in New York in 1958, appeared in French in Paris in 1954; the latter version was a combination of two earlier French publications dated 1948 and 1936, this last being a translation from the Romanian of a work originally composed in English! All the books now available in English were originally composed in some other language and in quite a different order but even here the facts are preposterously difficult to disentangle. The Foreword to Birth and Rebirth says that it was published in 1958 and that it is a rewrite of lectures delivered in Chicago in 1956 which were first written out in French, but page 17 of Mephistopheles (1966) tells us that Birth and Rebirth was first published in 1961 and that it is a rehandling of Naissances Mystiques (Paris, 1959). As a final complication several extra titles have been thrown in just to improve the sales. Cosmos and History (1959) is the same book as The Myth of the Eternal Return (1954), Rites and Symbols of Initiation (1965) is the same as Birth and Rebirth (1958), Mephistopheles and the Androgyne (1966) appeared in England as The Two and the One (1965). The gentle art of selling the same pup twice is well established among academics, but is seldom manifested on quite this scale or in quite so glaring a fashion.


SINCE ELIADE PROFESSES to be an expert on archaic modes of thought he necessarily relies very heavily on anthropological sources and his formidable bibliographies convey the impression of enormous erudition. But here again the proliferation of titles arouses a certain skepticism. A man who publishes a dozen books within fifteen years and appends over a thousand references to at least three of them is probably learned in only a rather superficial sense, but Eliade’s long book lists at least indicate what he has not read and in some cases this test is quite shattering. Eliade’s basio thesis depends on the recognition that the metaphysical polarity represented by our words sacred and profane is tied up with an awareness of temporal alternation. Any secular (profane) period A1 can only be distinguished from its successor A2 if there is an intervening “sacred” period B. Likewise any transition of secular status (e.g., that from maiden to married woman) must be marked by an interval of sacred time in which the initiate (the bride) has no secular status at all. Or again, if there are territories which are distinguished there must be an intervening boundary zone which belongs to neither category. Through this polarization of time and space a certain homology is perceived between such oppositions as Summer and Winter, Life and Death, Child and Adult, Man and Nature, Earth and Heaven, and so on, and the transitions between one side and the other are constantly a matter of religious concern. Now this theme has received great attention from professional anthropologists for many years but all modern work is heavily indebted to three classic sources: Mauss (1960)1 , Hertz (1907)2 , Van Gennep (1909)3 , any one of which is far more illuminating than the whole corpus of Eliade’s writings put together. But the odd thing is that Eliade ignores them. There are references to Van Gennep’s work in the footnotes (but not in the text) of several works published since 1958; the other two items are not mentioned at all. Whatever may be the explanation for this silence it can do Eliade no credit. I am not suggesting that his erudition is wholly fake but that his knowledge of the history of anthropology must be abysmal. This is not a subject which can be understood by reading predigested textbooks and scrabbling through an index to find an appropriate reference. The latest theoretical doctrine of which Eliade shows any understanding is that of the Vienna school of diffusionist culture-history which flourished between 1905 and 1935; functionalism has passed him by and there is no indication that he has ever heard of Lévi-Strauss, who has often concerned himself closely with the kinds of fact which Eliade makes central to his analysis. The musty smell of antiquity is further exaggerated by the translation time-lag. His latest book (1966) has a chapter on Melanesian Cargo Cults but cites no evidence later than 1955. This topic is currently the focus of much lively discussion but Eliade’s comments contribute nothing because all earlier work in this field has been made out of date by the glut of new material published during the past decade.

BUT PERHAPS this doesn’t really matter. “In all Melanesian Cargo Cults,” says Eliade, “the expectation of the catastrophe which will precede the Golden Age is marked by a series of actions expressing absolute detachment from ordinary values and behavior.” Note the word “all.” A writer who is prepared to generalize in this grandiose way is not going to be put out by a mere discordance of evidence or the lack of it. Eliade is a library scholar. What matters for him is that everything he says should be based on what someone else has put in a book. The printed word is authority enough and since there are enough oddities around to write almost everything, it is always quite easy to find “authoritative support” for one’s own opinions. This again was Frazer’s procedure. If Eliade writes:


Recent researches have clearly brought out the “shamanic” elements in the religion of the paleolithic hunters. Horst Kirchner has interpreted the celebrated relief at Lascaux as a representation of a shamanic trance. [Shamanism, p. 503]

most readers will believe him simply because it fits the argument. They will be quite unimpressed by the pedant’s comment that there are in fact no “reliefs” at Lascaux and that no one has the slightest idea why the paintings were made. Again, when Eliade tells us that:

Heine Geldern has established a connection between the human sacrifices and skull hunts that are abundantly attested in Assam and Burma and a matriarchal ideology.

[Yoga, p. 300]

it suffices that he can cite a reference dated 1917. It is futile to point out that Heine Geldern himself had no evidence at all. The technique of citing authorities for everything has the advantage that Eliade is personally committed to no fixed opinion and on some topics he changes his position from book to book without ever saying so. For example in Images and Symbols, p. 23, he says of Robertson-Smith’s totemic-communion theory of sacrifice (dated 1885) that it “no longer enjoys any credit among competent ethnologists and historians of religion,” but in The Sacred and the Profane, pp. 101-2, we meet the following:

According to the myths of the earliest cultivators man became what he is today—mortal, sexualized and condemned to work—in consequence of a primordial murder….

It is at this stage of Culture that we encounter ritual cannibalism. The cannibal’s chief concern would seem to be essentially metaphysical, he must not forget what happened in illo tempore. Volhardt and Jensen have shown this very clearly; the killing and devouring of sows at festivals, eating the first fruits when tubers are harvested, are an eating of the divine body exactly as it is eaten at cannibal feasts.

No mention here of course of incompetent ethnologists or the un-credit-worthy Robertson-Smith, nor even of the Christian Mass. Eliade’s personal mysticism seems to give him a confidence hardly justified by his evidence. He proclaims the truth as an enlightened prophet speaking from a great height. Shamans do not need to be consistent.

But his exposition by citation greatly simplifies my problem—how to find a path through the dark forest. The weight of Eliade’s “scholarship” is probably sufficient to overcome the uncertainty of all but the most obstinate Doubting Thomas, but much of it is just blarney. If we hack all this fat away the residue of hard argument is by no means bulky. What does it amount to?

INDEPENDENTLY of George Dumezil and Ananda Coomaraswamy, two authors whom he now greatly admires, Eliade’s firsthand exploration of Indian mysticism led him to appreciate the importance of notions of polarity. Sin and merit, ascetic inaction and ecstatic violence, this world and the other world, now and not-now, life and death, are not isolated categories which can be discussed and analyzed each by itself, but binary interdependent concepts like the yin and the yang of Chinese cosmogony. The paradox of religious thought and behavior everywhere is that the devotee, having first laboriously sorted out for himself the qualities which distinguish the time and space of things sacred from the time and space of things profane finds himself impelled to sew the two parts of the dichotomy together again. Man puts his heaven in the sky above, but then imagines a rope ladder linking earth and sky and tries to climb the ladder. The Yogin or the Shaman, as the case may be, is a specialist at such ladder climbing; he is the mediator between the powers of the other world and ordinary impotent men.

Eliade makes it all sound terribly complicated and splashes his pages with Sanscrit and theological Latin, but once the anthropological reader sees that the coincidentia oppositorum means no more than the synthesis of opposites, he is on familiar ground. As Lévi-Strauss and others have lately shown,4 the “Hegelian” use of categories is extremely widespread among “primitive” peoples and is one of our principal justifications for insisting (pace Eliade) that there is no radical discontinuity between “archaic” philosophy and that of modern man.5 The cosmological schema in question is not nearly so universal nor homogeneously consistent as Eliade makes out but the ethnographic illustrations which he has assembled are apt enough.

But the trouble with Eliade is that although he stresses his concern with the history of symbols, as distinct from the structure of symbolism, he does not really distinguish one from the other. As part of his Christian-Jungian faith he believes that particular symbols have archetypal and universal significance, so he gives us long essays on the meaning of The Cosmic Tree, of Mystical Light, of the symbolism of knots, and so on. Now it is quite true that the same symbols frequently crop up in entirely different religious contexts and that there is then an “historical” problem as to why this should be so. Does the similarity indicate diffusion from a common source, the recurrence of a universal archetype deeply buried in the human psyche, or just a common human aptitude for resorting to analogy? The answers in most cases are just any man’s guess. On the other hand, because Eliade has recognized that religious symbols occur not singly but as binary pairs, he is really committed to an analysis of structure—not the history of particular symbols but the relations between them. But here he gets into a muddle. At the level of structural analysis his “archaic” system is fairly convincing. The basic religious distinction is between the herenow and the other. The other is the sacred. The here-now is the center of the universe and man constructs it in imitation of a prototype already existing in the other. Man enters the herenow from the other at birth and returns to the other at death. Time is thus a cycle, an eternal return. For Eliade this “denial of history” is the essential characteristic of archaic thought. The time of secular experience (profane time) is cut up into periods by intervals of religious activity (sacred time) which typically take the form of initiations—dramatic performances which imitate the cyclical transition from life to death and death to life. The shaman is the expert who can control the events of secular time because he has one foot in each camp.

MANY ANTHROPOLOGISTS would agree that this general pattern fits a great many sets of ethnographic data but it is the pattern as a whole, the structure, that fits. The symbols in themselves are of only passing interest. Thus, to take an elementary case: If the other world is distinguished from this world, then it must be in another place. The two worlds are separated—by the sky, the sea, a river, a range of mountains—but they can then be joined by a ladder, a tall tree, a boat, a bridge, and so on. In this kind of analysis we attach importance to structural relations rather than to symbols as such; the ladder, the tree, the boat, the bridge are all “the same” because they do the same thing, they link the two worlds. But in Eliade’s Jungian scheme it is the symbol per se that matters, so he tells us about trees and ladders as means of reaching the other world but never gets around to boats or bridges, or tunnels, or rocky cliffs, or heavenly fishing nets, or magic beanstalks, all of which things, and many others besides, can serve the same function in mythical syntax.

All in all, a comprehensive study of Eliade quickly reaches the limits of marginal utility. The Myth of the Eternal Return was the book that established his reputation for Anglo-American readers and it still seems the most worthwhile. Of the others, several are tidied-up essay collections which simply reiterate the same argument with variations in the ethnographic data. Yoga is a substantial and professional work but it is packed with technical Sanskrit jargon, Shamanism and Patterns in Comparative Religion are anthropologically out of date but, like their Frazerian prototypes, full of curious information. Blarney or no, the bibliographies and the footnotes may have their uses. But one theme recurs throughout:

Archaic man acknowledges no act which has not been previously posited…by someone else…What he does has been done before. His life is the ceaseless repetition of gestures initiated by others. [The Myth of the Eternal Return, p. 5]

It is a portrait of Eliade himself! Archaic indeed!

This Issue

October 20, 1966