If we are going to think seriously about shamanism, then we have to confront Mircea Eliade, whose famous book Shamanism: Archaic Techniques of Ecstacy has defined the terms of the discussion for more than fifty years. Yet the book is deeply flawed, and Eliade’s approach has been subjected to compelling anthropological, historical, and feminist criticism.

It is important to note that Eliade never met a shaman, never lived in an indigenous culture where shamans practiced, and never observed a shamanic ceremony. Everything he knew about the topic of his book he learned from the writings of others.

Eliade himself in no way considered this an impediment. Religious scholar Ivan Strenski has pointed to Eliade’s “disdain for the empirical-historical method of investigating religion.” Rather, for Eliade, the truth about religious phenomena was found by a self-authenticating and incorrigible intuition. Eliade is quite clear about this. When a “dialogue with non-European spiritual traditions leads us to rediscover certain negelcted sources of our own spiritual heritage,” he asks, “what is the point of going so far afield and interrogating Indians, Africans, and Oceanians?”

So: we have to look at Eliade’s sources and how he used them. His book was published in French in 1951 and then republished in English, with some additions but no major changes, in 1964. That means that most of the works he relied on were from the 1930s and 1940s, and many were considerably earlier. These sources were of very varied quality. Some were quite good for their time — some field descriptions, for example, often by missionaries; others were affected, to various degrees, by bias, misunderstanding, and hearsay.

Mircea Eliade

But Eliade made no attempt to assess the quality of his sources. He did not distinguish between reliable and unreliable sources, between first-hand and third-hand accounts, or between description and polemic. Importantly, he did not distinguish between primary and secondary sources — that is, between ethnography and interpretation. Every word on a printed page was just as good as any other.

The reason that the quality or provenance of his sources did not matter to him was because his goal was to find little facts — what anthropologist Edmund Leach has called “snippets of exotic ethnography” — that would fit into the true archaic shamanism revealed by his intuition. It did not matter if these little facts came from different cultures or historical periods. As anthropologist John Saliba describes Eliade’s method, “Data from multiple religions and from different historical periods are collected and grouped together … There seems to be an implicit assumption that differences are not very significant.” Graham Harvey, a scholar of indigenous spiritualities, describes this as Eliade’s “insistent fitting of evidence into a pre-existing theological and evolutionary schema.” In fact, what Eliade has done, Harvey says, is to create a previously unknown system — which Eliade calls, confusingly, shamanism — out of bits of ethnographic data from various times and places.

And he often got these facts wrong. Marjorie Mandelstam Balzer, an expert on Siberian cultures, has said that Eliade “is remarkably inaccurate on details about Siberian shamanism.”

We should take particular note of the title of the book. The word archaic had special meaning for Eliade. Following Lévy-Bruhl, who himself later rejected this view, Eliade thought that indigenous peoples think differently than modern people — that there is a prelogical archaic mentality radically different from rational thought. He speaks constantly of the “primitive mind” and its “intuitions” — for example, that for such primitive minds there is “no clear difference between … empirical object and symbol.” Or again: “Primitive man lived in constant terror … For thousands of years men were tortured by the fear that the sun would disappear forever at the winter solstice.” They are just different from us.

Thus, for Eliade, the study of shamanism “launches a dialogue and an interrelationship with the others” — the italics are his. Anthropologist Alice Kehoe notes that this description embraces, in Eliade’s words, “the Siberian hunters, and … the primitive peoples of Australia, the Malay Archipelago, South America, North America, and other regions” — that is, people of color.

Eliade’s work carries a ponderous respectability. He cites massive amounts of literature. But this, of course, raises questions of its own. One might be excused for suspecting that he did not read the hundreds of texts he cites, but rather scoured their indexes for useful snippets. For example, Eliade refers to Evans-Pritchard’s book on Azande witchcraft as a “manual of black magic,” raising the question of whether he was familiar with its contents. And, out of that mass of material, of dizzyingly various quality, there is little doubt that he could pick out enough snippets to prove just about anything. Yet the cumulative weight of such citations conveys the impression of enormous erudition.

Because Eliade simply ignored facts that did not fit his theories, many of his generalizations turn out to be just plain wrong. He viewed shamans as essentially characterized by celestial ascent, ecstasy, soul flight, out-of-body journeys to the spirit realm. His treatise on shamanism is filled with references to the sky, to ascent, to the vertical rather than the horizontal. He saw the sacred in the transcendent, the vertical plane, the center as opposed to the peripheries of things; it is thus at the center, he said, at the tent pole, the mountain, the world axis, that the shaman communicates with the sky, ascends through the central opening, ascends the sacred mountain, ascends to the sky. Eliade was willing to denigrate as decadent or aberrant any shamanism in which the ascent to the sky plays an insufficiently important role. Tungus shamanism today, he said — that is, in the 1930s, when Shirokogoroff produced his important studies — cannot be considered shamanism in its classic form, because, among other things, of “the small role played by the ascent to the sky.”

Eliade also stated that spirit possession played no part in shamanism. He distinguished true, archaic shamanism from spirit possession, which he saw as corrupt, historically more recent, and subject to decline, degeneration, and decadence. At best, he said, such mediumship “may be regarded as an imitation of certain shamanic techniques.”

Now, mestizo shamans in the Upper Amazon do not generally engage in soul flight as described by Eliade; rather, the healing and protective spirits are summoned to the place of ceremony, where they prescribe medicine and direct the shaman where to shake the leaf-bundle rattle, blow tobacco smoke, and suck out the pathogenic objects that have caused the sickness; indeed, even when retrieving a lost or stolen soul, the soul is generally summoned rather than pursued. Occasionally, too, a healer allows one or more spirits — usually the spirit of a powerful deceased shaman — to enter his or her own body, during which time the shaman may remain and engage in conversation with the invited spirit or spirits; or else may vacate the premises and travel to distant planets or brilliant cities, there to gather in convocations of healers or attend at spiritual hospitals, while the now embodied deceased healer undertakes the healing of patients at the ceremony.

Mestizo shamans are not atypical. Anna-Leena Siikala, an expert in Northern Eurasian shamanism, has pointed to evidence for the same three modes of spirit interaction — journey, possession, and summoning — among Siberian shamans. Indeed, these three ways of working could be combined sequentially: one Evenk shaman took his spirits into himself at the opening of the performance, then questioned them, and finally flew with them. The threefold pattern in Siberian shamanism has been confirmed by historian Ronald Hutton in a thorough review of the literature. Similarly, anthropologists Larry Peters and Douglass Price-Williams surveyed forty-two societies from four different culture areas and found that shamans in ten of those societies engaged in out-of-body experience or journeying, eighteen in spirit “incorporation,” eleven in both, and three in some different state.

We have learned a lot about shamanism in the fifty years since Eliade’s book was first published. There is now a much wider range of excellent ethnographies; debates within the field have sharpened an awareness of many of the assumptions that underlay the fieldwork of many decades ago. Indeed, we now know, too, much more about ethnobotany, hallucinations, and the actions of such substances as nicotine and dimethyltryptamine. To get a sense of just how much has changed, you should take a look inside Graham Harvey’s Shamanism: A Reader, published in 2003, for an anthology of key texts.

Why did this book become such a cultural icon? It is, frankly, pretty boring; the tone can perhaps best be described as plonking. But timing is everything. The English translation, published in 1964, was waiting on the shelves when Carlos Castaneda’s first book, The Teachings of Don Juan, came out in 1968. Although Eliade himself denigrated the use of hallucinogens as a degenerate form of shamanism, the very heft of his book, along with its characterization of shamanism as a technique of ecstacy, seemed to legitimate the experimental use of psychoactive substances. A lot of people back then were really, really interested in celestial flight, and they welcomed a grand narrative that approved it.

One advantage of the book is that it is a compendium of sources that are otherwise difficult to obtain. Large chunks of the book are paraphrases — often very close to the original — of descriptions of ceremonies and practices from around the world. But, unlike Eliade, the cautious reader should evaluate the source, and separate description from commentary.