Over 50,000 people were executed as witches between the years 1500 and 1700 and a similar number underwent trial on the charge of witchcraft but were acquitted or died before sentence. This happened not during the Dark Ages or the supposedly superstitious medieval centuries, but in the sixteenth and seventeenth centuries, during the Renaissance, the Reformation, and the Scientific Revolution. There is a naturally dramatic and spectacular side to the subject, and films like The Devils and plays like The Crucible have made the images and dynamics of the early modern witch hunt familiar to a wide audience.

In recent decades historians have devoted much attention to witchcraft persecutions. As part of the post-1960s boom in interest in the popular, the repressed, and the marginal, a very large and impressive body of scholarship has been created which tries to make sense of this chilling aspect of European history. The starting point is the mass of trial records stored in the archives of every region from Scotland to Sicily. Students of European witchcraft must start by being students of the law courts.

Attempts to explain the witchcraft persecutions of the past have pursued two complementary paths: the motives of the accusers and the minds of the judges. Most witchcraft trials had their origin in legal charges made by aggrieved parties, who believed that sickness had struck their cattle, a malady fallen on their children, or their potency been impaired by the actions of a malevolent neighbor who could rely on the help of wicked supernatural powers. Hence one path to an understanding of witchcraft is to explore who accused whom, to analyze patterns of class, sex, and age, and to connect charges of witchcraft with the social tensions and strains of the early modern period.

A classic example of this approach is Keith Thomas’s Religion and the Decline of Magic, published in 1971, which studied the evidence for witchcraft in England and concluded that “charges of witchcraft were a means of expressing deep-felt animosities in acceptable guise.” In particular, he argued that those who made accusations of witchcraft were trying to deal with a sense of guilt or shame that they felt for some breach of the traditional duties of neighborliness and charity. The person to whom they had refused charity became the witch. It does, indeed, seem to be the case that old, poor women, the classic receivers (and demanders) of charity, are disproportionately common among those accused of witchcraft.

Another way to make sense of the trials of the great European witch hunt is to concentrate on the beliefs and expectations of the judges and inquisitors who conducted them. These men were often highly educated. In Catholic Europe they might be university-trained friars, and in both Catholic and Protestant countries among them would be gentlemen trained in the law and clergymen with theological knowledge. They brought with them into the courtroom the baggage of their education, which included a set of beliefs about the Devil and his followers. They knew that there was a Prince of Darkness, lord of this world, who was as seductive as he was powerful. Given these premises, there was nothing incredible in the idea that wicked or foolish human beings might enter into a league with Satan or his demons. As the members of the European intelligentsia elaborated their demonological science, the details were worked out of how witches made their pact, what powers they received, and how they might be recognized. A literature of diabolical studies and handbooks of witch hunting came into existence. Among its better-known examples are The Hammer of Witches of the German Dominican inquisitors Krämer and Sprenger and the Demonologie of James I of England and VI of Scotland.

The mentalities of the judges are important because they were much more than passive sifters of the evidence brought before them. They could put leading questions to the accused and to witnesses, they could make suggestions drawn from their own beliefs about the Devil and his ways, and, to encourage appropriate replies, they could use torture. “Where there is no torture, there can be little witchcraft,” claimed one famous legal historian. A chance survival of a unique and moving document from Bamberg in 1628 shows exactly how torture manufactured witches. The burgomaster of Bamberg, Johannes Junius, had been accused of witchcraft and tortured until he confessed. He managed to smuggle a last letter from prison, addressed to his daughter. “Many hundred thousand goodnights, dearly beloved daughter Veronica,” he begins. “Innocent have I come into prison, innocent have I been tortured, innocent must I die. For whoever comes into the witch prison must become a witch or else be tortured until he invents something out of his head….”

The current consensus of scholarly opinion about witchcraft is that in the villages of late medieval Europe there existed a centuries-old tradition of peasant sorcery, aimed at harming one’s neighbors, using such means as curses or image magic. This occasionally, although rarely, resulted in law suits between parties. In the fifteenth century a major change occurred as this malevolent magic was completely reinterpreted by the learned demonological tradition, which saw witches as a sect in Satan’s sway who met in perverse nocturnal gatherings or Sabbaths to worship their evil deity. Such a sect was clearly a threat to society as a whole. Associating the new ideas of sect and Sabbath with local sorcery was like exposing phosphorus to air. Traditional rural magic was demonized and the mass hunts and burnings began.


Historians and other scholars who explain the witch hunt in such ways are clearly concentrating not on witchcraft, but on the prosecution of witchcraft. They see what needs explanation as the hunt, not the quarry. Indeed, most of them would doubt the very existence of a quarry, if by that we mean active members of a sect of witches, rather than angry old women muttering a curse. The band of witches with its ritual gatherings and nocturnal flights lived in the minds of inquisitors and torturers, not in the villages of early modern Europe. But beyond wild confessions inspired by the bullying of judges or the tortured victim’s desperate eagerness to say whatever would appease the torturer, were there not real secret rituals in the peasant world of the past? Did women and men not actually assemble in ecstatic nocturnal gatherings?

The historian who has been willing to answer “yes” to these questions most emphatically is Carlo Ginzburg, formerly a professor at Bologna and now at the University of California at Los Angeles. In 1966 he published a remarkable study (issued in English translation in 1983 as Night Battles) which revealed, on the basis of the records of the inquisition in Friuli in northeastern Italy, the existence in the years around 1600 of a group of people who claimed that their spirits periodically left their bodies to do battle with evil witches in order to ensure the fertility of the harvest. These benandanti (“those who go well”), as they were called, sharply distinguished themselves from witches, but were, nevertheless, a genuine, self-identifying group of mystic travelers, who also claimed contact with the dead. Ginzburg saw in them members of “a single agrarian cult, which…must have been diffused in an earlier period over a much vaster area.” In his new book Ginzburg turns his sights to this earlier period and vaster area.

It is possible to obtain a fair impression of the kind of book this is by browsing through the illustrations. Starting with a roof tile from Roman Gaul, we move via a mosaic from a twelfth-century cathedral, a sixteenth-century engraving, and a prehistoric bust from Spain to a third-century BC wooden sculpture from Ch’angsha, China. We return briefly to sixteenth-century engravings before moving through Scythian gold combs, Visigothic buckles, and a Roman statue, to culminate in an engraving by Urs Graf (Crippled Devil, 1512) and a fresco from the Villa of the Mysteries at Pompeii. The text of the book is, however, more wide-ranging and eclectic than the illustrations would suggest. Indeed, it appears that only the twentieth century and black Africa are excluded from its main thesis. The thesis is that the Sabbath as it appears in the trial records of witches in the early modern period is only a small part of a much larger phenomenon and that phenomenon is a body of ancient, Eurasian, shamanistic myths and rituals.

Testing Ginzburg’s hypothesis turns out to be a delicate task. In Ecstasies the author is on the track of a “deep mythical and ritual stratum,” in which magical flight, metamorphosis into animal shape, and the journey to the land of the dead are the most salient elements. He can draw his material from any time and place, starting as he does with a temporary “renunciation…of a unilinear and uniform time.” His interpretation of mythical material follows loosely the structuralist (here “morphological”) principles of Lévi-Strauss, his psychology flirts with Jungianism, his cultural analysis has room for the ahistorical and illdefined category “folkloric culture.” Let us examine the method at work.

A typical passage occurs on page 236 in the middle of a discussion of ritual lameness, the wearing of one sandal (charmingly termed “monosandalism”) and their connection with the underworld. Caeculus. who appears in Virgil’s Aeneid as leader of a troop of single-booted soldiers, has been mentioned by Ginzburg as worth attention. He was also the founder of a city, a thought that leads us to the subject of Romulus, founder of Rome. One tradition about Romulus is that he headed a band of young outlaws devoted to cattle theft. Ginzburg proceeds:


In the legendary biography of the young hero, the theft of livestock carried out in league with their [sic] contemporaries was an obligatory stage, virtually an initiation ritual. It repeated a very ancient mythical model, amply documented in the Indo-European cultural milieu: the journey to the beyond to steal the livestock of a monstrous being.

It has been suggested that this myth should be seen as a reelaboration of the ecstatic journeys to the realm of the dead made by shamans to procure game for the community.

After moving from brigand-founder legend to underworld journey to shamanism in this way, Ginzburg goes on to discuss Hercules, presumably because of the legends of his cattle-raiding and his descent to the underworld. He mentions Hercules’ association with the Scythians, an ancient people who are important in his argument as a central source of shamanism, who transmitted shamanistic practices to neighboring peoples. The passage concludes, “The presence in China of a mythical hero to whom feats analogous to those of Hercules are attributed, has been tentatively ascribed to Scythian mediation.”

Of course the interpretation of myths is about as easy as the interpretation of dreams or literature. That does not mean that, in despair, it can be based on free association. The problem with Ginzburg’s kind of interpretation is that controls are absent and the criteria of association of one myth or practice with another are entirely subjective. An equally ingenious array of myths, legends, and rituals could have been mustered to demonstrate an entirely different pattern. What supports the author’s shift from the tale of brigands rustling cattle to the journey to the underworld? The relationship is described as a “repetition.” Who is repeating what? The myth of the underworld journey is supposedly “a re-elaboration” of shamanistic ecstasies. The implication, presumably, is that shamans’ accounts of their journeys in the spirit formed the basis of stories about rustling trips to the underworld that then served as models for stories about rustlers in the here-and-now. What is overlooked in this train of thought is the existence of cattle rustlers in the society that produced the Romulus story.

Instead of seeking the meanings of myths and rituals in the social practices and experiences of those who told them or enacted them, Ginzburg attempts to build a taxonomy and a genealogy for his chosen motifs. On the one hand he searches for morphological similarity through “apparently negligible details which, touched by the magic wand of comparison, suddenly reveal their secret physiognomy.” Lameness or monosandalism is a case in point: any problem with walking or oddity in footwear, in any story anywhere, must be somehow linked and can thus serve legitimately to pose the question, “How is it possible for similar myths and rituals to reemerge, with such insistence, in such extremely heterogeneous cultural environments—from Greece to China?” An extremely lax criterion of identity creates a body of material whose similarities then become an explanatory problem. Such is the power of a magic wand.

The anti-empiricist structuralism of the myth analysis is combined uneasily with a theory of genetic diffusion, that is, the spread of myths and legends over the course of time from one population to another. Ginzburg has a startlingly concrete sense of transmission through specific racial groups. At one point he tries to explain the recurrence of similar stories in Scotland, France, and Romania by their common Celtic tradition; at another he produces the Scythians as the missing link between ancient Greek, Celtic, and Chinese mythology. The low point is reached when the existence of a female cult in sixteenth-century Sicily seems about to be explained by the presence of Celtic mercenaries on the island in the fourth century BC. Eventually, however, the connection is wistfully dismissed. Eurasian shamanism becomes a central force. The absence from black Africa of the story of Cinderella (whose “monosandalism is a distinguishing sign of those who have visited the realm of the dead”) “cannot be accidental. We propose to relate it to the absence in the same area of shamanistic phenomena….”

The building blocks of Ginzburg’s grand theory are not in themselves defective but his treatment of them and his construction methods do not inspire confidence. For instance, the existence of unofficial peasant cults of some kind in the countryside of late medieval Europe is virtually certain. Even though the evidence comes entirely from hostile clerics or judges, it is widespread and weighty enough to demonstrate that villagers did on occasion perform rituals while dressed in animal skins or put out food for female spirits. What may be doubted, however, is whether these ceremonies constituted a coherent body of belief or ritual. Rather than revealing the hidden traces of Eurasian shamanism, they are more likely to be the local rites that one might expect to encounter in any nonliterate agrarian society.

Moreover, there is nothing intrinsically absurd in the hypothesis of the spread of myths, rituals, or symbols from one human community to another. If all written records of Christianity were miraculously destroyed, scholars of the future might well hypothesize the spread of a certain ancient religion wherever their archaeologists found crosses. They would have to try to distinguish decorative or functional crosses from cult objects, but their hunch about the diffusion of a religion would be, in principle, correct. The real problem is that establishing this fact would tell them nothing about the significance or meaning of the cross in the myriad circumstances in which it has served as a Christian symbol: a medieval procession, a Jesuit mission, a twentieth-century gospel church. Even when correct, a model claiming to trace the descent of briefly characterized myths and practices is somewhat vacuous. This explains why Ginzburg’s book, on such a sensational subject, is actually dull to read.

Ginzburg is clearly seduced by his own theme. The depth, the ancientness, the universality of the motifs he is pursuing have drawn him in ever deeper. His references to “a remote Eurasian substratum,” “myths…from the most remote places and periods,” and “the subterranean layer of unitary Eurasian mythology” reveal the allure of what Shakespeare called “the dark backward and abysm of time.” But the first thing to be learned when studying the past is that it wasn’t the past at the time. Ginzburg’s poetic intoxication is a sympathetic trait but it leaves aside the question of what the experience of ecstasy or the participation in the rituals or the hearing of the myths meant in the lives of those who underwent or created them. How did the myths function in the human communities precisely located in space and time who gave them birth or nurtured them? On these questions Ginzburg has very little to say.

It is easy to become impatient with a book that never sits still, that never looks long, coolly, and hard at any piece of evidence but lurches with claims of incessant discovery from one time and place to another. When faced with the statement that “the existence of a connection between the mushroom used by the shamans to attain ecstasy and lameness will at this point not seem inconceivable,” the glassyeyed reader may be willing to agree simply because “this point” has been reached after 306 pages and 990 footnotes. In its main thesis the book recalls the work of the long-discredited Margaret Murray, who argued without convincing evidence that European witches were followers of an ancient pagan fertility cult. Ginzburg’s exploration of hallucinogenic mushrooms and mysterious cults of female deities might lead some readers to wonder whether we are witnessing the birth of a New Age historiography. Such responses would be facile, however. It clearly makes a big difference whether the roots of witchcraft are to be found in the spite and envy of the early modern European peasant village or in the ecstatic cults of prehistory. Unfortunately, despite its enormous accumulation of material, Ginzburg’s book fails to address the first possibility and fails to prove the second.

There is nothing intrinsically impossible in the argument that shamanistic practices from one culture shaped the cults and religions of another. The pioneering study by E.R. Dodds, The Greeks and the Irrational (1951), argued exactly this in the case of the ancient Greeks, seeing the post-Homeric development of ecstatic and pneumatic elements in Greek religion as a result of influence from the very Scythian cultures that loom so large in Ginzsburg’s analysis. Dodds, however, was a careful and discriminating scholar who was not lured into “panshamanism,” as he termed it. He explicitly denied the case that all ecstatic elements in Greek religion derived from a mysterious “Orphic cult” that some scholars had postulated and he did so in words that apply equally well to Ginzburg’s Ecstasies: “The edifice reared by an ingenious scholarship upon these foundations remains for me a house of dreams.”

This Issue

June 13, 1991