University of Pennsylvania Press

The European witch trials that began in the fifteenth century have been explained in many ways, but always assuming that witchcraft was a unified concept. Work on the history of witchcraft has come to a point at which we both can and must rethink this and other basic assumptions about the rise of these trials. We must reconsider how far the concept of witchcraft was consolidated into a single imaginative construct during the fifteenth century, and how the mythology of witchcraft functioned in distinct places. Most basically, we must ask anew how useful standard models of historical explanation are for understanding the early witch trials.

At the turn of the twentieth century, Joseph Hansen traced the evolution of what he called the “collective concept” (Kollektivbegriff ) of witchcraft, corresponding to what we would now call the imaginative world of the Sabbath.1 All the components of this concept had their separate histories: the nocturnal assembly, flight through the air with demonic assistance, the pact with the devil, sexual intercourse with incubi and succubi, and bewitchments or maleficia. In Hansen’s construction, which has never seemed controversial, it was in the fifteenth century that these notions fused into a cohesive notion of witchcraft, to which the Sabbath was of central importance. Whether agreeing or disagreeing with Hansen in other respects, historians have followed his lead in all this, accepting as a given that in the fifteenth century there was a more or less established conception of witchcraft. The single mythology of witchcraft is then explained by tracing its roots in medieval [End Page 79] heresies, in the practices of ritual magic, in the tenacious peasant culture of Europe, or in the effort of Christians to shore up their own uncertain faith by grasping for evidence of the supernatural.2 Historians of various persuasions agree on this point: whatever their sources, and however diverse they may be at first, the elements of witchcraft consolidated quickly into that cumulative construct, into a concept of “the Sabbath” that emerges relatively early in the fifteenth century through the aglutination of elements in a single mythology.

We now have fuller documentation at our disposal than even the great source-compiler Hansen could claim. We are therefore in a position to see that in the fifteenth century there was not a single mythology of witchcraft but multiple mythologies, and that in certain places these mythologies remained intact for several decades, preserving their independent contours with remarkable tenacity. We can now see, too, that when these mythologies did lose their stability, the circumstances leading to change require special attention. We can trace the patterns of both stability and modification with greater precision than Hansen could have hoped to attain. The point is not simply that mythologies of witchcraft are plural rather than singular, and that they come in regional varieties—although this much is true, important, and more obvious now than any of us once knew. The further argument I want to make is that mythologies of witchcraft functioned differently under different circumstances.

Two sources in particular compel us to revise our notions of how the witch trials began at the end of the Middle Ages. First, it would be difficult to imagine a source for the history of witchcraft more important than manuscript 29 in the Archives Cantonales Vaudoises at Lausanne. This manuscript gives us detailed information about the trials of twelve men and seven women from the southern edge of the Pays de Vaud, the region of Lausanne, along the north shore of Lake Geneva, all tried for witchcraft in the years 1438–98.3 [End Page 80] This collection of legal records provides a wealth of information early in the history of the witch trials, from the very heartland of those trials. The accused are called heretics (heretici), but their transgressions are those of conspiratorial witches. The documentation for this series is full and candid enough that we can tell how the trials began, how they progressed, and how interrogation and torture molded the outcomes. The manuscript has been known to scholars since 1908–9, when Maxime Reymond published two articles based on it, and some of us have made use of it in later decades; but it was in only 1995–97 that a team of historians published editions, French translations, and extended commentary in four volumes.4 The second source, less extensive than the first yet still vitally important, is a series of trial records in the Archivio di Stato di Perugia, most especially a set of four trials between 1455 and 1501, published in 1988 by Ugolino Nicolini.5 The trials took place at the Umbrian town of Perugia. The accused here were almost all women, and they were called streghe or witches.

In each of these sources we can see a mythology of witchcraft that remains fundamentally stable over several decades. But it is a different mythology in each case: the witchcraft of the documents from Perugia bears almost no resemblance to that of the manuscript from Lausanne. In a conventional interpretation we might say that they describe different versions of witchcraft, with differing accounts of the Sabbath. But not all contemporary observers would have classed the heretici of Lausanne and the streghe of Perugia in the same category. I will want to examine the implications of these differences. I will then turn to other witch trials of the fifteenth century, in parts of France where the mythic vocabulary was less stable, and where the myths of witchcraft were largely borrowed from western Switzerland, from central Italy, or both. And I will argue that it makes a great deal of difference whether the mythology of witchcraft was indigenous to a region or imported into it. [End Page 81]

First Representative Case: Jordana De Baulmes

Few of the early trials for witchcraft are more poignant than that of Jordana de Baulmes, tried in 1477 in the castle of Ouchy, on the south side of Lausanne.6 Neighbors of hers from the outskirts of Vevey first brought her before the papal and episcopal inquisitors. At the outset of the trial she was asked if she knew a man named Gérard Reymond. Yes, she replied; he was a fellow villager, who two years ago had approached her with the words, “Come on, you false woman, you heretic, you caused my animals to die, and now I have a cow that is sick, and I’ve given up on its life, but if it dies I will make you burn for it, because you are a heretic.”7

Faced with such initial allegations against her, Jordana denied them, but she confessed to a deeply troubled conscience on other grounds.8 Twenty years earlier she had quarreled with her husband and gone to help her parents with the vintage. They took her to the neighboring town of Fribourg, where she had a romantic adventure with a man who “loved her so much that she conceived a daughter by him.” Her lover left her and refused all help; she gave birth secretly and left the infant daughter exposed at a fountain, where the child evidently survived long enough to be baptized but then died. Jordana had confessed this sin and received absolution from a Dominican friar. She added that when she was living at Gruyère she had a second illegitimate child by another father, who again refused his help, and this time she buried the infant without baptism behind some building. But a few days later she retracted the second story, saying she had in fact never lived in or near Gruyère; she told this fabrication thinking she might secure release from prison by giving the inquisitors a false confession. For years, however, she clearly had been haunted by the first experience. It would be too much to expect of the inquisitors the gentle wisdom of a sympathetic counselor, and in fact this was not their strength. Faced with a woman who freely admitted her crushing sense of guilt, they were indifferent. The memories she told were not the sort that interested them.

Eight days into the interrogation, Jordana’s judges found a way to make progress in their proceedings.9 Rather than asking her outright if she had been guilty of witchcraft, they asked what she could tell them about the deeds of heretics, and on that point she had much to relate: in their assemblies the heretics drink and eat, consuming the flesh of children; they lie together [End Page 82] in the manner of brute animals; they fly on brooms anointed with unguents. She had heard all this a good twenty years ago, from sources she could not remember. She did not know at what time the heretics assemble, because she had never heard that told. Did she know anything else? No, except that—and here she took a fateful step—just the previous night a great multitude of candles with blue light came before her in the prison where she was detained. But with this the book of her memory snapped shut: she knew nothing further. She was taken to the place of torture, where at the portal she suggested perhaps a devil was hidden somewhere on her person and was keeping her from confessing.10 Suddenly she lost her power of speech, as if she were being strangled. Soon, however, the floodgates opened, and in further interrogation Jordana’s story conformed to the judges’ expectations. Her tale now began as it had before, but it veered off in a very different direction. She said she had quarreled with her husband and moved to Fribourg, where she had a baby girl by a lover who deserted her. Alone and depressed, she found herself approached one day by a man with the suspicious name Sathanas, dressed in black, like a great lord, who asked the cause of her sadness and said he would give her consolation if she agreed to do as he said. With this she made her entry into the witches’ company. On more than one occasion she attended a nocturnal assembly, called here a “sect.” In it she denied God, the Virgin, and baptism. She gave the devil an obscene kiss on his posterior. She and others at the assembly ate the flesh of children and had sex with each other “in the manner of brute animals.” One of the locations for the assembly was a “mountain” above Blonay, to the north of Vevey.

Jordana was by anyone’s standards an unreliable narrator of her own story, as the inquisitors clearly perceived. Having followed their script fully, she later retracted part of what she had confessed, but only part. She denied, in particular, an earlier confession about her role in a host desecration.11 Evidently [End Page 83] the judges believed that a long period in prison might win her back to a more compliant mood, and when they resumed interrogation seven weeks later she confirmed that her previous confessions had been true, except for the matter of the consecrated host.12 Through the convoluted record of her proceedings, we can see the intertwining of three imaginative realms: her own, in which the number of her lovers and illegitimate children was uncertain and unstable; that of her neighbors, in which she could be the occult cause for illness to livestock and perhaps other misfortune; and that of the inquisitors—confident, fully elaborated, and destined to prevail over the other two imaginative worlds, as well as any real world one might care to posit.

The Lausanne Manuscript and the Mythology of Witchcraft in the Pays De Vaud

The confession that Jordana eventually gave her judges was essentially that of virtually all the accused who occupy the folios of the Lausanne manuscript. One of the most important facts to emerge from these trials is the consistency of the confessions extracted from the accused. The manuscript provides a coherent body of cases; although the defendants were quite different in the backgrounds and in the circumstances leading to their accusation, they were made to conform over several decades to a script that was remarkably consistent from one case to another. The standard script for confession included thirteen elements, which can be clustered for convenience in three groups.

First, there were matters having to do chiefly with the relationship between the heretic and the devil: (1) the circumstances of the heretic’s induction into allegiance with the devil, usually at a moment of despondency; (2) the appearance of the devil or demon to the heretic, in human or bestial form; (3) the homage paid by the heretic to the devil, which included the obscene kiss sub cauda and the giving of a bodily part and other offerings; (4) denial of God, the Virgin, the sacrament of baptism, and sometimes other holy persons and objects; (5) desecration of the sacred, often involving the trampling on a cross; and (6) payment to the heretics of a specified sum of money, which was generally not paid, or when paid turned out illusory. The most dramatic part of the narrative is often the opening. The script called for an account of the witch’s fall into complicity with other heretics, and almost [End Page 84] always this testimony appears in the record. Induction typically presupposed inducement: people were led into the sect as a solution to some problem, sometimes a quarrel or some ongoing enmity. The most often cited temptation was the promise of wealth, whether or not this came as a solution to the pressure of poverty. The spirit typically appeared at first in human form, generally that of a man dressed in black, and then changed into the form of an animal, usually a black cat, to receive the heretic’s obscene kiss beneath its tail.

Second, there were matters having to do with the heretic’s participation in activities along with other heretics: (7) attendance at the assemblies, called either sects or synagogues (never Sabbaths); (8) transport to those assemblies, usually on an anointed stick, and usually instantaneous; (9) eating at the sects of the flesh of infants whom the heretics had previously killed and exhumed; (10) sex with other heretics; (11) illumination of the assemblies by a fire that emitted blue light and by candles. Attendance at the assemblies could be impressively large: at one there were a hundred twenty people, at another nearly three hundred men, women, and demons.

Third, the accused told of misdeeds carried out in the broader society at the command of demons: (12) violation of religious duties, and profanation of the Eucharist; and (13) maleficia, chiefly the killing of persons and animals by use of powders given by the demon. Apart from these thirteen points of confession (which are not explicitly broken down in this manner in the documents), the heretics were expected to give the names of as many accomplices as possible in all these deeds.

Only rarely is there deviation from or even significant addition to the established mythology. Late in the sequence of trials two of the witches were said—in the midst of more typical allegations—to have fallen from the clouds during storms. Precisely how their tumbles related to their other activities remains unclear. And the notion, which figured prominently in literature of the ninth century, seems to have resurfaced unaccountably in this much later setting, as if fallen from a Carolingian cloud that somehow strayed over late medieval skies. In any case, this was a slight addition to the received pattern, which otherwise persisted intact.13

The geographical setting for these trials calls for close attention. In a seminal article of 1967, H. R. Trevor-Roper suggested that witchcraft originated in mountainous regions.14 Perhaps he was thinking of cases such as these from [End Page 85] western Switzerland. But in fact these took place on the urbanized north shore of the Lake of Geneva. Even the mons de Blonay at which some of the witches’ assemblies took place is not a mountain but a hill on the outskirts of Vevey. Jordana de Baulmes is typical: she came from Corsier, not far outside the town of Vevey, in the elevated land within easy walking distance of urban life. Her parents were vineyard workers, probably small rural proprietors. Within Vevey she might have been recognized as the proximate outsider, the sort of villager often met in passing at a town market: a person markedly different in class and culture from the inquisitors and other churchmen of the town. Yet her initial troubles came not from them but from her neighbors in Corsier. If she had lived in a mountain valley, it might never have occurred to anyone to report her to the inquisitors, who rarely penetrated the mountainous hinterland. Jordana’s misfortune was to come from a village close enough to the town that she was caught between two cultures, surrounded by neighbors who could easily walk downhill to Vevey and hear inquisitors preaching against witchcraft, or witness the public spectacle of women and men burned as witches. The record of the Lausanne manuscript must be read against this background.

It might seem that the mythic complex produced by this interrogation is unexceptional: that it is simply what witches anywhere were expected to confess. This is not in fact the case, but there is a handful of fifteenth-century trials outside the diocese of Lausanne that belong recognizably to the same family as these. In 1462 four women and four men were sentenced at Chamonix in Savoy, and the pattern of confessions followed that of Lausanne quite closely.15 Again in 1477 a woman named Antonia was tried in Savoy, and again her confessions follow the general pattern of confessions in the diocese of Lausanne.16 Even the sequence of interrogation leading up to the interlocutory sentence and torture was essentially the same as that used in the diocese of Lausanne. Clearly an interrogatory from Lausanne was being used in these trials. And it comes as no surprise that the inquisitors who presided at these proceedings were deputed by an inquisitor who worked in Lausanne. The connections are less obvious in the trial of Jehanneta Lasne by municipal authorities at Fribourg in 1493. Her trial record does not report on the interrogation leading up to her confession, but still her confession resembled closely that of the women and men tried by inquisitors in the [End Page 86] diocese of Lausanne, closely enough that we may assume the magistrates were following the model set by the inquisitors in the Pays de Vaud.17

Second Representative Case: Filippa Da CittÀ Della Pieve

In 1455 the secular authorities of Perugia tried and executed Filippa da Città della Pieve as, among other things, a strega who drank the blood of children.18 Unlike many of the characters tried in and near Umbria for witchcraft, she is not spoken of as a professional healer, diviner, or specialist in other people’s amorous requirements; she operated in her own interests, not those of clients. The record is explicit and minutely detailed about her magical practice. It tells of the love potion she concocted with semen, her own menstrual blood, and a powerful herb, which she harvested on a Thursday before sunrise while mouthing incantations. Using this and other magical procedures, she managed to compel at least four men to love her with unquenchable fury. Other ingredients, including hairs of a hanged man and body parts of a bat and a raven,19 she placed in a pouch, which she buried near the house of a man with whom she had quarreled, causing him to die. These and other crimes are told in circumstantial detail, including names and dates as well as means. She had been practicing her crimes for the past twenty years, and she had gained a considerable reputation as a formidable enemy. Yet she acknowledged a prior if not better craftswoman named Clarutia Angeli, also of Città della Pieve, to whom she was a disciple.

It was a further category of offense, also learned from this Clarutia, that qualified Filippa as a strega. Being in league with the devil, the two women went about killing young children by sucking and draining their blood. One morning before dawn Filippa stripped herself naked, anointed herself with an unguent, and pronounced the conjuration, “O devil, I give myself to you in soul and in body; carry me where I tell you.” Immediately both she and her teacher were swept away to a neighbor’s house in transfigured form, presumably that of a small animal, and they sucked the blood of a child while his mother slept. The next day, the child died. After his burial, they exhumed the body; they used it and other children’s corpses to make magical powders and unguents. Indeed, she killed more than a hundred infants, drinking their blood as they lay beside their mothers. [End Page 87]

The Perugia Documents and the Mythology of Witchcraft in Umbria

Through much of central Italy, women such as Filippa were tried at least sporadically as streghe. The term has come to be a standard Italian word for “witch,” but in this context it means something rather more specific. It is a term associated with a particular mythology, that of women who go out at night, enter people’s homes, and kill infants. The term is a reference to the screechowl (strix, strigis in Latin). This mythology had very little in common with that of the Pays de Vaud. The streghe were almost all female; even when one male strega was tried, it is clear that he was thought to be acting alongside women who acted the role more fully than he.20 They operated independently or in pairs, rarely in larger groups. They flew, but not usually to assemblies. When they flew, they took the form of animals rather than flying on the backs of animals. They might be accused of invoking and collaborating with demons, but they did not do homage to the devil. They killed children, almost always by sucking their blood, but they did not eat the children’s bodies.

The epicenter of these trials, the place where the notion seems most often to have surfaced in the course of prosecution, was Perugia, yet this concept of witchcraft was not specific to Perugia or even to central Italy. Bernardino of Siena knew the type and reported that he had encountered a strega of this sort in the 1420s at Rome.21 In a trial at Todi in 1428, the phenomenon merged with the notion of a witches’ assembly, although the mythology of that assembly was not fully developed, and it played here at most a marginal role.22 The witch at Todi, Matteuccia di Francesco, was said to have carried out much of her activity in the region of Perugia, but others operated further afield. Maria “the healer” of Vicenza went forth as a strega and attacked (streavit) around three hundred children in various places in and around Brescia; half of these died, while the rest she then cured.23 Alfonso de Spina assumed there were witches of this sort in fifteenth-century Spain, and we will see at [End Page 88] least fragmentary evidence for the concept also in southern France, suggesting its distribution across the Mediterranean coast.24 Ethnographic evidence shows much wider diffusion of the pattern; in the Tlaxcala region of Mexico between 1959 and 1966, Hugo Nutini collected evidence of roughly three hundred cases of blood-sucking witches (called tlahuelpuchis) who take on animal form, insinuate themselves into people’s houses, and kill children, mostly infants.25

Returning to Perugia, once again in this town we find a coherent mythology to which the accused conformed over several decades. But when we compare it to the mythology from the Pays de Vaud, the most striking conclusion is how little the two mythic complexes resemble each other. What I have already suggested for the two case studies applies more broadly: these mythologies have almost nothing in common. Whereas the confessions from the Pays de Vaud focus primarily on the witches’ assemblies, the streghe of central Itally operated independently or in pairs, rarely in larger groups. In the Pays de Vaud, the bewitchments perpetrated among neighbors tended to fall from view once attention focused on the mythology of the witches’ sect; in central Italy, there is far more detail about the bewitchments, and the mythology itself centers far more squarely on the most important of these maleficia: the metamorphosis into the form of small animals, the entry into people’s houses through small openings, and the attack on infant children, killed by sucking their blood. While the witches of the Pays de Vaud killed babies to eat their bodies in their communal assemblies, those of central Italy evidently killed because they themselves need to drink infant blood.

Whereas the Vaudois witches inverted ecclesiastical values, those of central Italy inverted the values of domestic life and particularly of motherhood. The witches in the Pays de Vaud were said to renounce all that was sacred, including their baptismal vows, they took the devil as their master rather than God, and they desecrated the eucharist. The mythology revolves centrally around this flaunting of the Church’s values, the inversion of sacred order, clearly a matter of deep concern to the churchmen who tried these witches. In central Italy, the streghe entered into people’s homes surreptitiously and attacked infants precisely where they should have been most secure, at their parents’ [End Page 89] side. The infants’ mothers would have given them breast milk, recognized in medieval culture as a transformed and nourishing gift of the mothers’ own blood, their own life force.26 The streghe, inverting the order of nature, sucked this vital fluid from the infants’ veins. The values they challenged were those basic to family life, and for the family men who tried them, this was a fundamentally serious threat. One might have expected the inquisitors at Vevey to be equally attentive to these concerns, but even when Jordana de Baulmes exposed herself to them as a quintessentially bad mother that was her script, her reality, not the inquisitors’ mythology. The mythology of the Pays de Vaud centered on the mocking of social values, the beliefs and rituals that defined Christian society, by a sinister inversion of that society. The mythology of central Italy focused rather on the affront to private and domestic values by streghe who operated not as a society but alone or in pairs. And the differences remained over decades.

Filippa’s is the earliest trial of a strega found in the Perugian archives, but another Perugian case ten years earlier, known from other sources, is in its own way revealing. A woman named Santuccia da Gualdo Tadino was burned for having killed fifty children and bewitched people by means of consecrated hosts given to her by a priest.27 The case is mentioned in a contemporary chronicle. It is mentioned also in a sermon by the Franciscan friar Giacomo della Marca, who had been summoned to Perugia by the municipal authorities to lend his moral authority to the prosecution. He had not been the instigator of the trial, or the catalyst by which a figure of folklore became identified in real life, but his presence even as a latecomer makes the case an interesting study in the collaboration of power, principle, and popular pressure. His sermon is further interesting because of one context in which he mentions Santuccia. He is speaking about the association of humans and demons, and the effects of that association. If you doubt what he has to say, he urges looking into the face of a sorceress; you will see that it is diabolical, because the devil dwells in her at all times. This is half a century or so after Raymund of Capua told his readers how he had seen the face of Catherine of Siena suddenly transformed into the face of Christ, who was present in [End Page 90] her.28 Like Catherine’s face, Santuccia’s becomes a gestalt image, and once one viewer shows a new way to see it, others will perceive it likewise, imagination bending easily to the force of suggestion.

Coercion and Collaboration

Comparison between the witch trials of the Pays de Vaud and of central Italy demonstrates key distinctions not only in the mythologies that emerge from the records but also in the ways these mythologies functioned—and at this point we come to the heart of our analysis. The Vaudois trials manifested competition between the imaginative worlds of the accusers, the accused, and the (ultimately coercive) inquisitors, while the Umbrian cases showed a greater degree of consensus and collaboration because the script was familiar to all parties.

As we have already seen, the Vaudois trials were quite fully scripted. The script to which the confessions conformed was supple enough to accommodate some particularities: it allowed for the incidental variation needed to give it a semblance of plausibility. Thus, the name of the devil who approached the witch might be scriptural, folkloric, or simply invented: Satanas, Grabier, Rabiel, Robinet, and other names occur. The theriomorphic apparition too was subject to some variation: the spirit usually changed into the form of a black cat, in one case a horned gray cat, but at times some other horned beast, a calf, a wolf, a black dog, or a horse. Still, the script was fixed in its outlines and in many of its details. The scripted character of the records can be seen in small and apparently inconsequential details. With fair consistency throughout the series of trials was the report of fire and candles illuminating the assembly with strangely blue light. The judges showed some curiosity on this point; whether or not the color blue was intrinsically significant, the consistency of the report showed that these heretics were all members of the same sect and thus a unified conspiracy. In an era without artificial lighting, one can imagine that unnatural illumination would convey a powerful sense of the supernatural and in this case the demonic. Jordana de Baulmes said that when the fire was extinguished, it appeared as though there had never been a fire: further evidence that the illumination was unnatural and the fire illusory.29

Elsewhere Jordana gives particularly telling evidence of how the accused was made to conform to the script. Asked what form the devil took for the [End Page 91] obscene kiss, she said twice that he was a man dressed in black, the form in which he originally appeared to her. But she had departed from the script, which required the devil to receive the kiss in bestial form, and her judges did not allow the departure. Again a second time she gave the wrong response. It was only after several days that she finally gave the correct answer, that she had kissed the devil in the form of a cat.30 How did she know that this was the right answer? Perhaps a jailer gave her the tip she needed. Or she may have remembered this detail at last from an inquisitor’s sermon that she heard (or heard about) when some previous witch was condemned and burned.

The accused were expected to confess that they had been forced to renounce God and all that was sacred.31 A formal denial of this sort is recorded in all the trials except one. In this matter the script was reasonably clear, but renunciation of the sacred was evidently more difficult for some of the accused to admit than adherence to the demonic, and so there was variation in what they were willing to confess even when they clearly knew what was expected of them. Jeannette Barattier said that at first she had denied God, the Virgin, the heavenly court, and baptism, taking a demon as her lord and master, but at another point she agreed to deny all except for the Virgin and baptism. Others were willing to deny God but not the Virgin: eloquent if perverse testimony to the hold of Marian piety in late medieval culture. And some said they had denied God with their mouths only, not with their hearts. Once again, the testimony clearly came in response to an unvarying interrogatory, and the range of variation allowed in the responses was not broad.32 [End Page 92]

The inquisitors induced the accused to follow the script through essentially three means. The most obvious means was torture, and the record is explicit about the use and the efficacy of torture. Jaquet de Panissère, tried in 1477, said he was willing to undergo torture but would confess nothing, and indeed when he was lifted up a bit from the ground on the instrument of torture he persisted at first in his denials, but soon he confessed.33 Torture was sometimes applied to secure fuller confessions than the accused had already made. The mere threat of torture could sometimes secure a confession, even when torture was apparently not in fact used. At times the accused began by refusing to confess their crimes, then broke down, and said they had been unable to confess earlier because their demon masters had in one way or another prevented them.

Short of torture, there was a second way of inducing conformity to the script: carefully patterned interrogation, seen especially well in the trials of Catherine Quicquat, Jaquet de Panissère, and Pierre Menetrey.34 The interrogation focused initially on circumstantial factors, clearly meant to catch the accused off guard. Were the prisoners aware why they had been arrested? Had they ever fallen under suspicion of witchcraft? Did they have enemies? Very little emphasis was placed on questions about actual guilt; the possibility was raised whenever possible indirectly. Rather than asking whether they had cursed anyone, the interrogators asked whether they had ever made threats that were followed by misfortune. Interrogation about what one knew about witches could be useful as a way of moving the trial forward. A defendant who admitted knowing the deeds of witches in some detail might be suspect of learning them at firsthand. This line of interrogation served [End Page 93] further as a sort of rehearsal for confession of guilt. A person who had articulated what the inquisitors wished to hear in the third person was better prepared to give the same information in the first person.

A third means of bringing confession into conformity with script was the pressure placed on the accused by others who had already confessed. Those who had confessed their own crimes were expected then to implicate others, and to confront those they had implicated face-to-face. Thus, two convicted witches testified against Marguerite Diserens, saying she was one of their company, having attended the assembly with them. When she denied it, they besought her to her face, for the love of God, to confess her guilt and receive the mercy of God and of the Church, as they had done. She persisted in her denial, but not for long. When she confessed, Marguerite was among those who accused Pierre Menetrey. When he denied his complicity, it was her turn to beg him to confess and to throw himself on the mercy of God and of the Church, as she and others had done. Although reported in formulaic language, these confrontations gave opportunities for reminding the accused about the sort of confession expected of them.35

All this apparatus of instruction and coercion, however, presupposes a dimension of these trials that is easy to overlook. One of the most important aspects of these trials in the Pays de Vaud is that the mythic complex reported in them was evidently not indigenous to the region. It was imported from elsewhere, and at least during the fifteenth century it was evidently not integrated into local culture. It came from written sources, composed in other regions around the 1430s. The earliest of these sources, the Errores Gazariorum, probably originated in the Aosta valley, perhaps taken from Aosta to Lausanne in 1440 by the newly elected bishop, George de Saluces (Saluzzo), as the latest editors have suggested, although its themes are not in all respects those of the Vaudois witches.36 Related works were written soon afterward in the Dauphiné and in German-speaking Switzerland.37 Some elements of [End Page 94] the mythology contained in these sources may perhaps have come from folkloric sources of those other regions. Be that as it may, this was a mythology that soon became closely identified with the Pays de Vaud, yet it was not native to the local folk culture of that region. What the Lausanne manuscript reveals is the careful cultivation of an invasive species. And because the mythology was not indigenous, it did not come with culturally sanctioned clues that could help identify the sort of person who might plausibly be identified as a witch. For precisely that reason, I propose, the subjects were widely diverse in their backgrounds. One remarkable feature of these trials is how little the accused resembled each other before their interrogation, and how much they did afterward. The imported mythic pattern, systematically imposed by a rigorous tribunal, could be imposed just as effectively on anyone. Given the absence of locally accepted marks of a plausible subject, almost anyone could become accused and convicted.

In precisely this respect, the trials in central Italy were fundamentally different from those of the Pays de Vaud: the mythology disclosed in and around Perugia was indigenous. The myth of the strega was deeply rooted in Italian folklore and well known to poets in antiquity.38 Being integrated into the local culture, it came accompanied by expectations about the sort of person who might plausibly be a strega. Whereas the subjects in the Pays de Vaud were highly diverse, the accused in Perugia were far less so. They were women, regularly if perhaps not professionally engaged in mediation of supernatural powers. The mythology was itself different from that of the Pays de Vaud, but even more importantly it functioned differently, operating within a cultural complex rooted in the soil, a complex that told people not only what a strega did but what a strega looked like. She did not come from an agrarian hinterland; if she seemed strange, it was because within the urban culture of central Italy she was strange: she was meant to be so, and she might support herself or gain a kind of stature within her community by exploiting her image as the uncanny woman, perhaps the hag.

In the Pays de Vaud, incidents of bewitchment are generally given in the [End Page 95] early stages of the trial: the initial stimulus to prosecution was typically someone’s charge of love magic, bewitchment of cattle, or some other maleficium. Once the trial had begun, there was a clear break, and these accusations were quickly submerged in a flood of testimony about the mythic elements of the satanic assembly.39 Bewitchment might figure in this mythology, in the sense that the devil was said to supply the means for afflicting society, but the language at this point is generic. There is relatively little talk about how the witch has afflicted such-and-such a person by particular means in a specific year. In the trials at Perugia, however, the documents are full of such information. There is no caesura separating talk about bewitchment from mythological discourse. The mythology of the strega and her activities is still usually couched in highly particular language, indistinguishable from that about love magic and bodily harm: Filippa da Città della Pieve sucked the blood of a neighbor’s child in March of 1434, perpetrated love magic on a young man in May of 1440, and used sorcery to kill a man she had quarreled with in 1448. The mythology of the blood-sucking witch is as fully integrated into the particularities of daily life as the rest of the charges given in the condemnation. It arises, like the other charges, out of the locally established vocabulary of fear.

This is not to say that the secular judges of Perugia did not torture their subjects or apply other forms of judicial coercion; they may have done so, even if the records are not the sort that would give us that information. The notion of the strega was surely known to all, but torture may have been needed to induce self-identification as such a malefactor. Even if it was used, however, there are two crucial differences: first, the mythology is indigenous; second, it is clearly and directly linked with the bewitchment that aroused popular anxiety and brought the accused into the tribunal in the first place. Rather than revealing a competition of imaginative worlds, these trials disclosed fundamental agreement about the mythology in question.

Excursus: a Third Mythological Pattern in Northern Italy

The stable mythologies of the Pays de Vaud and of Umbria seem to have been without parallel in late medieval Europe, yet there is evidence of a third mythic paradigm, less fully disclosed and developed, that had some staying power. Trial records show a scattering of trials in northern Italy that seem to [End Page 96] fit neither the Vaudois nor the Umbrian paradigm. Between 1384 and 1480 there were cases at Milan, Faido, Brixen, and Brescia, all in northern Italy and Italian Switzerland, but these trials were widely scattered and nowhere frequent.40 One key element in this pattern is the notion of an assembly that is called a “play” (ludus), with a female presider, sometimes a figure named Oriente who gave instruction on healing and other matters, elsewhere a Lady Abunda.41 The record may dutifully identify the presider as a devil in the form of a queen, but even then the theme of female leadership is discernible beneath the inquisitorial bias. A second leitmotif is an emphasis on bounty and beneficence: the devil brings a supply of bread and cheese, and the name Abunda also hints at this emphasis.

At times there may be hints of either the Vaudois or the Umbrian mythology in these trials: the assembly may involve eating human flesh, or the drinking of children’s blood. Still, what we see here are surely elements of a third mythic complex, perhaps authentically folkloric, but overlaid with other notions in ways that make a coherent cultural setting impossible to recreate. I see this northern Italian mythic complex as related tenuously at best to the other two mythologies. Indirectly it may have been a source of folkloric material later taken over in much distorted form, but (pace Carlo Ginzburg, for whom these trials are paradigmatic) it is only incidentally connected with the mainstream of early witch trials.42

Niklaus Schatzmann has recently published a study of trials in the Leventina Valley in the canton of Ticino that shows in detail how this paradigm could become partially assimilated to others.43 Faido, a town in this valley, [End Page 97] had already been the site of relevant proceedings in 1432. Roughly a generation later, in 1457–59, thirty-seven individuals were tried in the valley for witchcraft, and twenty of them burned. While most of the accused were charged with blasting trees and harming humans and livestock, eleven of them in particular were accused of mythological witchcraft. They attended a “play” at which they met other witches, sometimes a hundred fifty or more, and they encountered the devil usually in bestial form. During the first month of the proceedings there were elements of the typical northern Italian pattern with its emphasis on bounty: those gathered at the ludus might feast on bread, meat, and cheese. But before long, certain of the accused said they could not share in the feasting; one of them explained that she was one of the lesser members of the assembly. As the series of trials continued, beneficence yielded to horror and the devil came to be a fire-spewing monster who reacted savagely if his followers did not bring him the offerings he wished. Those tried later in the series described anointing stools or other objects as media for flight to the assembly; one had been warned not to cross herself during flight. If in these respects the ludus became assimilated to the Vaudois paradigm, without ever becoming as detailed in the account of the assembly, one trial bears the clear mark of the central Italian mythology: it was a male witch who told how members of the company went out at night in the form of cats and stole infants from the cradle to kill them by sucking their blood.44 Clearly the paradigms were not hermetically sealed off from each other, yet perhaps just as clearly they began as distinct paradigms and then in varying measure began to converge.

Fusion of Mythologies in the Theological Literature

If the mythologies of witchcraft seen in the records from Lausanne and Perugia are fundamentally distinct, and remained so for generations, was it only modern historians who conflated them into a unitary mythology? No, because theologians and writers with some training in formal theology led the way in the fifteenth century. It was perhaps inevitable that they should do so: writing in Latin, but accounting for vernacular culture, they used Latin terms [End Page 98] as equivalents to various vernacular ones, and in the process they made the Italian strega and the Spanish bruxa become synonyms for sorcière and hexe.

In the third quarter of the fifteenth century, the Italian Dominicans Girolamo Visconti and Giordano da Bergamo both wrote about streghe.45 They both claimed to be writing about beliefs held by common people (vulgares communiter tenent). But Visconti spoke of the streghe as attending a ludus or festival at which a domina ludi presided, and he spoke of obscene and disgraceful behavior at these assemblies, which suggests that he was conflating the streghe of central Italy both with the witches of northern Italy and with those of western Switzerland. And Giordano da Bergamo ascribed to the streghe behavior typical not for the central Italian but for the Vaudois paradigm, such as sexual intercourse with demons. The Spaniard Franciscan Alfonso de Spina began by describing witches in terms of the central Italian paradigm but then spoke of them as meeting at night by candle light and giving the obscene kiss to a wild boar. And he believed they abounded in the Dauphiné and Gas-cony, by which he seems to have meant Languedoc.

For theologians of this period, the question of overriding importance was whether the deeds of the witches were possible, and if so how. The typical answer was that the devil achieved the witches’ effects, either by delusion of the senses or by the locomotion that spirits could exercise over matter. For the theologians, examples of such demonic power could be drawn from the widest variety of sources: from scripture, from classical literature, or from recent judicial testimony. They were hardly concerned to distinguish one body of folklore or mythology from another. In their effort to find as many examples as possible of how demons could delude people and manipulate physical bodies, these writers helped to obscure distinct cultures and reduce multiple phenomena to a single phenomenon.

It was in the demonological literature more than in the courts that the mythology of the Sabbath eventually became consolidated, to some extent in the fifteenth century, more fully in the sixteenth. The process was one of convergence, and this point calls for attention. To speak of the witches’ assemblies in the Pays de Vaud, in northern and central Italy, and throughout France as different forms or versions of “the Sabbath” is to suggest that there was some common root from which these variants emerged. It may be more helpful to speak not of divergence but of convergence: not of variations on some basic type, enshrined perhaps in the Errores Gazariorum, but rather of [End Page 99] regional mythologies which over the course of the fifteenth and following centuries came (perhaps never fully) to be fused.

Diffusion of the Vaudois Model in France

Even in the courts, however, there was at least sporadic diffusion of the Vaudois mythology of witchcraft, particularly in France, where some trials of the mid-fifteenth century gained a notoriety that spawned further literature on witchcraft, and thus further theological blurring of mythologies. The common French term for witchcraft from the mid-fifteenth century was Vauderie, and the witches were known as the Vaudois. These terms were linked with French variants on “Waldensianism” and “Waldensian”; Pope Eugenius IV appears to have made this connection already in 1440.46 Yet the words would surely have resonance also because they conjured an association with the Pays de Vaud. The sources are typically silent about which connection they had in mind, but contemporary writers often cited western Switzerland and adjacent regions as breeding grounds for witchcraft, and association of the Vaudois-witches with the Vaudois-Swiss would have been a simple and understandable crossing of semantic wires.47

In any case, the mythic complex worked out and systematically imposed on the accused on the shores of the Lake of Geneva appeared in only partial and shifting forms when it spread westward. We have here detached fragments of an originally coherent model. This is true even for the nearby mountainous territories of the Dauphiné.48 These fragments might become subject matter for new demonological literature that would impose a new kind of order on the fragments, as happened at Arras when nearly three dozen persons were tried for Vauderie in the years 1459–62.49 But even in [End Page 100] this celebrated case there was little consistency about mythic matters: the records mention at one point dancing at the witches’ assemblies, but the notion does not recur; two of the subjects were said to have killed children, but they were not said to have eaten the flesh; some of the accused, but not all, had made pacts with the devil; the obscene kiss appears in the well-known manuscript illuminations depicting the witches of Arras, but it figures less prominently in the trial records.50

Two other cases of Vauderie occurred not far from Paris in the years 1452–53. At Provens in western Champagne the Vaudois were said to attend an assembly called a mescle, but what they did there other than stripping naked is not clear. The members killed children and brought their bodies to the mescle to be roasted and eaten. Otherwise, links to the Pays de Vaud are few. At Evreux in eastern Normandy the accused, a Carmelite theologian named Guillaume Adeline, confessed to having attended “synagogues” of the Vaudois at which he kissed the hand and then the posterior of the presiding demon.51 He had renounced belief in the Trinity, the Virgin, the cross, holy water, consecrated bread, and adoration of roadside crosses—an odd mixture, combining doctrinal repudiations with cultic renunciations. Mainly what Adeline did at the assemblies was to serve as the theologian in residence, preaching diabolical doctrines in lieu of those he had taught at Paris. His practice is called Vauderie, but it consists of little more than bits and pieces of the witchcraft evidenced in the Pays de Vaud.

Much the same could be said about the cases of Andrée Garaude at Bressuire in 1475 and Martiale Espaze at Boucoiran in 1491.52 In both these trials, [End Page 101] the accused were marginalized women; Espaze had a history of sexual promiscuity. In both cases, the witches’ assembly is called not a synagogue but a Sabbath. In neither is the mythology of the Pays de Vaud preserved intact. But in the trial of Martiale Espaze we encounter an extraordinary mixture of themes. She is inducted into the company of witches and pays homage to the devil in the form of a goat, giving him the obscene kiss on the posterior. The women at the Sabbath hold candles in their hands. She renounces God and the Virgin, and after tracing a cross on the ground she tramples and spits on it. She has extensive dealings with her demon, named Robin, more outside the context of the Sabbath than at it. But then, suddenly, toward the end of the interrogation, we find her behaving more like an Italian strega than like a Vaudois heretic. She goes out and kills infants, sometimes in the company of another woman. She presses upon and evidently strangles the children as they lie in their cradles. She does not suck their blood, but neither does she take the bodies to the Sabbath to be eaten. In this case, then, we have assorted fragments of both the first and the second of our mythic complexes.

The situation in France, then, is entirely different from what we have seen elsewhere. There is no coherent sequence of trials orchestrated by inquisitors over several decades, with confessions forced into the Procrustean bed of a fixed mythology. But neither is there a generally accepted indigenous mythology; rather, we find a mythic complex now twice transplanted. Given neither source of coherence, the result is inevitable: a highly fluid mythology, in which no two trials resemble each other very closely; sporadic rather than sustained prosecution, in which the victims may be disreputable women but may also be respected theologians, because the imported and fragmented mythology is accompanied by no criteria at all for its application.

Were it not for these trials in France, we might wish to rehabilitate some version of Hansen’s Kollektivbegriff: we might wish to say that there was a single cohesive mythology of witchcraft, that found in the literature of the 1430s and in the trials of the Pays de Vaud, and that the cases from central Italy represent simply a secondary phenomenon, integrated only peripherally into this mythological complex. What the French trials suggest is that the Vaudois mythology lost much of its cohesion once it was transplanted, and no longer served so rigorously as a script to which confessions regularly conformed. Once it reached France, the mythology became distinct from both major alternatives and capable of meshing elements of both.

Was it the report of inquisitions in the Pays de Vaud that inspired these trials in France, or was it rather the literature of the 1430s and later witchcraft literature? The distinction may be unresolvable (because the fragmentary mythology in the trials does not allow close comparison with the fully developed [End Page 102] mythological sources), and it may be moot (because the same ideas could have come from either channel). But two points are worth emphasizing in this connection. First, as already proposed, the French term Vauderie suggests association with the Pays de Vaud, and this association is reinforced by contemporary evidence of a link between witchcraft and the western Alpine regions. Second, and more to the point, of all the writings penned in and around the 1430s, the one most closely associated with the trials in the Pays de Vaud was the Errores Gazariorum, which seems not to have had a widespread readership, survives in only two manuscript copies, and is unlikely to have been the direct source for French notions of mythological witchcraft. In all likelihood, the ideas contained in this treatise gained currency in France in large part indirectly, through their influence on prosecution in Franco-phone Switzerland.

In other parts of Europe there were only the faintest echoes of this mythology. From German-speaking lands we have the report of a chronicler, Matthias of Kemnat, about witches (Unholden zauberin) apprehended around 1475 in the Upper Palatinate, particularly at Heidelberg and Zent bei Tilsberg. They were said to ride at night on cats and brooms. Some of them confessed that they rode out on Ember Days to rouse storms. They made people lame, and they also could cure people.53 A trial at Lucerne in the mid-fifteenth century likewise contains elements of the Vaudois paradigm.54 But in most Germanic regions we find trials only for bewitchment, not for mythological witchcraft. The Malleus maleficarum, written in Latin for the use of German judges, focuses mainly on bewitchments (or maleficia), analyzes the collaboration between witches and demons required for their efficacy, and borrows from Romance-language areas the notion that sorcery itself is part of a conspiracy, but it does not develop the mythology of the Sabbath, and in general does not drink deeply from mythological springs.55

Concluding Methodological Postscript: Explaining Why and Explaining How

It follows from what I have been arguing that if we set out to explain why the witch trials of the fifteenth century began we are asking the wrong question. It is wrong, first of all, because it assumes precisely what this article has [End Page 103] argued we cannot assume: that there was a single phenomenon for which we can give an accounting. Finding an explanation that will apply to the trials for mythological witchcraft in the Pays de Vaud, in Umbria, and in France is a pointless exercise. The trials in these regions were of very different sorts, and they came about in significantly different ways. These were phenomena that had some relationships, and in some people’s minds did become conflated, but there is no reason to assume common grounds for prosecution.

Even if we limit ourselves to the trials in any one of these regions, however, it may not be helpful to ask why they occurred. We may be able to give a plausible account of why a particular theologian or other writer promoted belief in mythological witchcraft, as Michael Bailey has done for Johannes Nider,56 but very few of the judges involved in witch-hunting left us written material, and we cannot assume their intent was always the same as Nider’s. If we had a concentration of trials in a particular time and place, perhaps eight or ten in a single town over three years, we would certainly want to know what was happening in that town that led to such an outburst of prosecutorial zeal. But even in the Pays de Vaud, what we have is nineteen trials over sixty years, fewer on average than one every three years, with different inquisitors presiding and different local circumstances in the background. In 1448 there was a cluster of three trials, but very different in character; one of them, in particular, involved a man who presented himself voluntarily before the inquisitors. We can isolate certain factors that in a general way served as causes or at least necessary conditions for witch hunting: the development of inquisitorial procedure and the use of judicial torture, the articulation of theological arguments showing how demons could carry out the effects ascribed to witches, the emergence of learned (often university-trained) elites that took a serious interest in reform of religion and morals at the grassroots level, and the rise of complex societies in which divergent classes and cultures encountered each other, with the inevitable heightening of tension and misunderstanding. But there were places where all these conditions were met and still there was no particular surge of prosecution for sorcery, let along mythological witchcraft. We can say that certain specific notions must have been linked with particular experiences, as fear of the streghe was obviously inspired by the experience of sudden infant death, but should we assume there were no sudden infant deaths in Germany, England, or other places where they were not blamed on blood-sucking or otherwise mythical witches? However relevant all these factors no doubt are, [End Page 104] they do not add up to a cogent explanation why witch-hunting begins at a particular place in a particular time. However much we know about necessary conditions, too much depends on local circumstances we do not know, or know insufficiently. And too much depends on the obvious but unpredictable factor of imitation: one peasant accused a neighbor of witchcraft because another had done so, or one judge listened sympathetically to such charges because another judge whom he respected had heeded them, even if the precise circumstances were quite different.

Thus, while in many other cases it may make perfect sense to explain why historical events and developments occur, the early witch trials (1) were too diverse in character and circumstance to reduce them to any overarching set of causes and (2) presuppose a range of rather obvious conditions that were individually necessary but jointly insufficient to serve as causal explanations. Yet none of this is counsel of despair, because even if we cannot expect to explain why the witch trials occurred, we still can and must explain how they took place. We can examine with increasing precision and sensitivity the various dynamisms that unfolded in the course of prosecution. And explaining how the trials occurred may in a sense bring us as close as we dare hope to come in explaining why they took place. If we can perceive more clearly how judicial coercion and judicial collaboration functioned, sometimes overwhelming and at other times cultivating the community’s imagination, we are likely to be less surprised at the apparently trivial circumstances—the village quarrels, the zealotry of particular judges, the spontaneous consensus in identifying a particular woman as a strega—that serve as catalysts giving a network of conditions the potency of a cause. Explaining the rise of the witch trials is a classic example of a problem that calls for supple and nuanced explanatory models, and explaining how the trials came about is likely to carry us much further than perhaps inevitably flat-footed attempts to explain why they did so.

The further implication is that the witch trials cannot be conceived as simply and straightforwardly the product of demonological literature. Even if the Errores Gazoriorum did much to inform the interrogatory used in the Pays de Vaud, its availability cannot in any simple sense be said to have caused the trials there. Much light has been shed recently on the interests that led the demonologists to write their works.57 The relationship between the literature and judicial practice was broadly reciprocal: if some texts inspired prosecution, [End Page 105] in other cases it was prosecution that provoked the writing of texts.58 In any case, the witchcraft literature represents essentially another of the conditions that made witch-hunting possible but not inevitable, and why this literature affected judicial practice in some places while not in others cannot be accounted for by the literature itself. The main further difference that the literature makes is that it changes somewhat one’s sense of whether witchcraft was in fact a single and cohesive phenomenon that might be explained. The trial records would suggest otherwise. The demonological literature might also be read as suggesting a multiplicity of views, but the writers of this literature far more than contemporary judges at least had a strong interest in creating the mythology of witchcraft as a widespread and uniform threat to Christendom.

In an article entitled “ ‘Many reasons why’: witchcraft and the problem of multiple explanation,” Robin Briggs has argued against monocausal and otherwise simplistic explanations of the witch trials.59 My argument here may seem similar to Briggs’s, and it may thus be useful to highlight three crucial differences. First, I am making a claim not primarily about the reasons cited in explanation of witch-hunting but rather about the phenomena we are trying to explain. My argument is not that the witch trials had multiple causes. Rather, it is that witchcraft—and even mythological witchcraft—was a cluster of distinct phenomena that need not have been seen as related. Not only the explanans but the explanandum was multiple. Second, I am suggesting that the process of convergence—the emergence of a perception that disparate phenomena had something in common—is itself a development calling for historical explanation. It may seem that I am being a splitter rather than a lumper, but if so that is only because I see the process of lumping itself as something interesting and not to be taken for granted. Third, I am proposing that the convergence in question is a convergence in perception, and that perceptions call for a different type of explanation from events. It is usually more productive to examine the process by which a perception arises and takes hold (“explaining how”) than to seek some elusive cause behind it (“explaining why”). [End Page 106]

Causal explanation may well be possible in the case of trials for maleficia, particularly when there is reason to suspect that the accused were in fact guilty, or when there is plausible political motive for the charges. Explaining the judicial concern with mythological witchcraft is a more complex and elusive challenge. Can we hope to explain why some healer-diviners (but not others who were equally deviant) were perceived as blood-sucking streghe? Or why certain village quarrels (but not countless others equally fierce) led to charges before inquisitors who had read the demonological literature? Or why some marginalized crones (but not others who likewise muttered curses and rejoiced at their neighbors’ misfortune) evoked echoes of the Vaudois paradigm? We might just as well ask why some people see a gestalt image as a duck while others see it as a rabbit. There may be a reason, but the evidence required to find it is unlikely to be at hand. Can we, then, explain more generally why these mythologies gained whatever currency they held? No doubt we can offer conjectures, but much depended on the power of suggestion, on the persuasion of charismatic zealots (both international and local), and on the self-perpetuating power of imagined threats, a tissue of contingent and imponderable circumstances, operative in some regions but for no clear reason absent in others. What we can and must do, rather, is investigate the processes that fostered these perceptions, showing how Italian magistrates drew upon regional folklore in prosecution of streghe, how inquisitors in western Switzerland exploited the tools of inquisitorial prosecution, how French judges borrowed and adapted fragments of the Vaudois mythology, and how theologians in particular began to weave these and other strands into a more complex and comprehensive mythology of witchcraft. We are unlikely to arrive at a set of causes under which trials for mythological witchcraft could only be expected. But given that the trials did occur, we can show how they unfolded, and how differing setting and perceptions led to their unfolding quite differently from one region to another. It is fuller investigation of these processes that will lead to ever deeper understanding of the early witch trials. [End Page 107]

Vaudois paradigm Umbrian paradigm
Key source: Archives Cantonales Vaudoises, Lausanne, MS 29: records of 19 trials, 1438–98, in and near Lausanne and Vevey Key source: Archivio di Stato di Perugia: records of four particularly relevant trials, 1455–1501, in Perugia
The accused twelve men, seven women widely varying in character and situation The accused mostly female typically marginalized
Relationship with devil circumstances of induction apparition and homage denial of God, Virgin, baptism desecration of the sacred payment to heretic Relationship with devil invocation and collaboration (without homage)
Association with other witches at assemblies (“sects” or “synagogues”) arrival and departure on stick eating flesh of infants sex with other heretics illumination with blue light Association with other witches operation independently or in pairs, rarely in larger groups
Misdeeds committed within broader society violation of religious duties and profanation of eucharist bewitchments (maleficia), including the killing of children Misdeeds committed within broader society metamorphosis, flight, entry into houses through small openings killing of children by sucking their blood, use of bodies for magic
Function of the mythological complex imported and imposed on populace uneven competition of accused, accuser, and judges operating within different imaginative worlds no established criteria for application subordination of maleficia to mythology Function of the mythological complex indigenous and elicited from populace collaboration of accused, accuser, and judges operating within shared imaginative world established criteria for application integration of maleficia with mythology [End Page 108]
Richard Kieckhefer
Northwestern University


I am grateful to Brian Levack, Michael Bailey, Jessica Roussanov, and Robert E. Lerner for their tremendously helpful suggestions on this article.


1. Joseph Hansen, Zauberwahn, Inquisition und Hexenprozeß im Mittelalter, und die Entstehung der großen Hexenverfolgung (Munich: Oldenbourg, 1900; repr., Aalen: Scientia, 1964), 35.

2. The positions of, respectively, Jeffrey Burton Russell, Witchcraft in the Middle Ages (Ithaca, N.Y.: Cornell University Press, 1972); Norman Cohn, Europe’s Inner Demons: An Enquiry Inspired by the Great Witch-Hunt (London: Chatto, 1975; rev. ed., Pimlico, 1993); Carlo Ginzburg, Ecstasies: Deciphering the Witches’ Sabbath, trans. Raymond Rosenthal (New York: Pantheon, 1991); and Walter Stephens, Demon Lovers: Witchcraft, Sex, and the Crisis of Belief (Chicago: University of Chicago Press, 2001).

3. Because it deals with material beyond the chronological scope of this article, I am leaving out of consideration Pierre-Han Choffat, La sorcellerie comme exutoire: tensions et conflits locaux, Dommartin 1524–28 (Lausanne: Cahiers Lausannois d’Histoire Médiévale, 1989).

4. Martine Ostorero, “Folâtrer avec les démons”: Sabbat et chasse aux sorciers á Vevey (1448) (Lausanne: Cahiers Lausannois d’Histoire Médiévale, 1995); Georg Modestin, Le diable chez l’évêque: Chasse aux sorciers dans le diocèse de Lausanne (vers 1460) (Lausanne: Cahiers Lausannois d’Histoire Médiévale, 1995); Eva Maier, Trente ans avec le diable: Une nouvelle chasse aux sorciers sur la Riviera lémanique (1477–1484) (Lausanne: Cahiers Lausannois d’Histoire Médiévale, 1996); and Laurence Pfister, L’enfer sur terre: Sorcellerie à Dommartin (1498) (Lausanne: Cahiers Lausannois d’Histoire Médiévale, 1997). For a concise overview see Martine Ostorero, “Les chasses aux sorcières dans le Pays de Vaud (1430–1530): bilan des recherches,” Schweizerische Zeitschrift für Geschichte: Revue suisse d’histoire: Rivista storica svizzera 52 (2002): 109–14.

5. Ugolino Nicolini, “La stregoneria a Perugia e in Umbria nel Medioevo: con i testi di sette processi a Perugia e uno a Bologna,” Bollettino della Deputazione di storia patria per l’Umbria 84 [for 1987] (1988): 5–87.

6. Maier, Trente ans, 333–61.

7. Maier, Trente ans, 334; see the similar incident reported on p. 348.

8. Maier, Trente ans, 338, 342–44.

9. Maier, Trente ans, 344.

10. Maier, Trente ans, 346.

11. Maier, Trente ans, 350. For parallel cases see Joseph Hansen, ed., Quellen und Untersuchungen zur Geschichte des Hexenwahns und der Hexenverfolgung im Mittelalter (Bonn: Georgi, 190l; repr. Hildesheim: Olms, 1963), 467–99, 477–84, and 548–51; E. Hoffmann-Krayer, “Luzerner Akten zum Hexen- und Zauberwesen,” Schweizerisches Archiv für Volkskunde 3 (1899): 30–33; René Filhol, “Procès de sorcellerie à Bressuire (Août–Septembre 1475),” Revue historique de droit française et étrange, 4th ser., 42 (1964): 77–83; Abbé Garnodier, Recherches archéologiques sur Saint-Romain-de-Lerp et ses environs (new ed., Valence, 1860), 234–35 n. For the broader tradition, in which Jews are usually seen as the antagonists, see Peter Browe, “Die Hostienschädung der Juden im Mittelalter,” Römische Quartalschrift für christliche Altertumskunde und für Kirchengeschichte 34 (1926): 167–97; Charles Zika, “Hosts, Processions and Pilgrimages: Controlling the Sacred in Fifteenth-Century germany,” Past and Present 118 (February 1988): 25–64; and the “Croxton Play of the Sacrament,” in Non-Cycle Plays and Fragments, ed. Norman Davis (London: Oxford University Press for the Early English Text Society, 1970), lxiii–lxv, 58–89.

12. Maier, Trente ans, 352.

13. Pfister, L’enfer sur terre, 186, 188, 198. On the Carolingian precedent see Russell, Witchcraft in the Middle Ages, 82–83.

14. H. R. Trevor-Roper, The European Witch-Craze of the Sixteenth and Seventeenth Centuries and Other Essays (New York: Harper, 1969), 102.

15. Hansen, Quellen, 477–84.

16. Jh. M. Lavanchy, Sabbats ou synagogues sur les bords du lac d’Annecy: procès inquisitorial à St. Jorioz en 1477, 2nd ed. (Annecy: Abry, 1896); Hansen, Quellen, 467–99.

17. Hansen, Quellen, 590–92.

18. Nicolini, “La stregoneria a Perugia,” app. 1–5, pp. 52–58.

19. Conspicuously absent here is the hoopoe, on which see Richard Kieckhefer, Forbidden Rites: A Necromancer’s Manual of the Fifteenth Century, Magic in History, 1 (Stroud: Sutton, 1997; University Park: Penn State University Press, 1998), 66 n25.

20. Nicolini, “La stregoneria a Perugia,” app. 1–8, pp. 66–73.

21. Franco Mormando, The Preacher’s Demons: Bernardino of Siena and the Social Underworld of Early Renaissance Italy (Chicago: University of Chicago Press, 1999), 52–108, esp. 54–72; Saint Bernardino of Siena, Sermons, ed. Nazareno Orlandi, trans. Helen Josephine Robins (Siena: Tipografia Sociale, 1920), 166–67.

22. Domenico Mammoli, The Record of the Trial and Condemnation of a Witch, Matteuccia di Francesco, at Todi, 20 March 1428 (Rome, 1972).

23. Paolo Guerrini, ed., Chronache bresciane inedite dei sec. XV–XIX (Brescia, 1922), 1:183–85; Giuseppe Bonomo, Caccia alle streghe: La credenza nelle streghe dal secolo XIII al XIX, con particolare riferimento all’Italia (Palermo, 1959.)

24. Alphonsus de Spina, Fortalitium fidei, in Hansen, Quellen, 145–48.

25. Haracio Fabrega and Hugo Nutini, “Witchcraft-Explained: Childhood Tragedies in Tlaxcala, and Their Medical Sequelae,” Social Science & Medicine 36 (1993): 793–805; Hugo G. Nutini and John M. Roberts, Blood-Sucking Witchcraft: An Epistemological Study of Anthropomorphic Supernaturalism in Rural Tlaxcala (Tucson: University of Arizona Press, 1993).

26. See Caroline Walker Bynum, Holy Feast and Holy Fast: The Religious Significance of Food to Medieval Women (Berkeley and Los Angeles: University of California Press, 1987), 179; Thomas Laqueur, Making Sex: Body and Gender from the Greeks to Freud (Harvard University Press, 1990), 104–6; and William F. MacLehose, “Nurturing Danger: High Medieval Medicine and the Problem(s) of the Child,” in Medieval Mothering, ed. John Carmi Parsons and Bonnie Wheeler (New York: Garland, 1996), 11.

27. Nicolini, “La stregoneria a Perugia,” app. 1–4, pp. 50–51.

28. Raymond of Capua, The Life of Catherine of Siena, i.9.90, trans. Conleth Kearns (Wilmington, Dela.: Michael Glazier, 1980), 82.

29. Maier, Trente ans, 360.

30. Maier, Trente ans, 348 (9/18), 352 (9/21), 360 (11/10).

31. Typically the denial was of God and the Virgin, sometimes also the Trinity, at times also baptism, and occasionally the saints (or the heavenly court). Presumably the implication is that introduction into the Christian faith is being foresworn, a renunciation if not an undoing of baptism, and this could be done either explicitly or by a denial of the Trinitarian formulas of baptism. This would explain why the Trinity had to be renounced even after God had been.

32. Ostorero, “Folâtrer avec les démons”, 202 (God, with mouth only, not heart), 204 (God, his face and the cross, the sacrament of the altar, the Trinity, and baptism), 204 (God, the Trinity, baptism, the face and cross of God, and the sacrament of the altar), 216 (God and the heavenly court, the Virgin Mary, and the sacraments), 266 (God and things pertaining to God); Modestin, Le diable chez l’évêque, 198 (God Almighty, the Virgin Mary, and the entire heavenly court), 204 (God almighty, the Virgin Mary, the entire heavenly court, and holy baptism), 218 (God and the Holy Trinity), 228 (God and the Holy Trinity), 302 (God and all that is of God, in word but not in heart), 312 (God almighty, the blessed Virgin Mary, and the entire heavenly court); Maier, Trente ans; 178 (God, the Virgin Mary, the entire Trinity, and baptism), 182 (God, the Blessed Virgin Mary, the entre Trinity, and baptism), 188 (God, the Virgin Mary, and the whole Trinity), 208 (God and the whole Trinity, the Virgin Mary and the whole heavenly court, baptism, and all things divine), 230 (God, the Virgin Mary, the heavenly court, and baptism), 234–36 (God, the Blessed Virgin Mary, baptism, the saints [sanctos et sanctas Dei], and all things that are God’s—she agreed to deny all except the Blessed Virgin Mary and baptism), 264 (God, the Virgin Mary, and the entire Trinity), 308 (God and baptism), 314 (God and baptism), 326 (God, the Virgin Mary, and baptism), 348 (God, the Virgin Mary, and baptism), 355 (God, with mouth but not heart), 361 (God); Pfister, L’enfer sur terre, 204 (God, the Virgin Mary, and the Holy Trinity), 218 (God, baptism, and the Virgin Mary), 220 (God the creator, baptism, and the Virgin Mary—but then God the creator of heaven and earth, and baptism, but not the Virgin Mary), 240 (God and the entire Trinity but not the Virgin Mary), 262 (God the creator but not the Virgin Mary). On the use of an interrogatory see Ostorero, “Les chasses aux sorcières,” 112.

33. Maier, Trente ans, 298.

34. Ostorero, “Folâtrer avec les démons”, 236–57 (Catherine Quicquat); Maier, Trente ans, 287–331 (Jaquet de Panissère); and Pfister, L’enfer sur terre, 250–73 (Pierre Menetrey).

35. Pfister, L’enfer sur terre, 240, 254.

36. Georg Modestin discussed the bishop’s role in “Des Bischofs letzte Tage: Georg von Saluzzo und die Hexenverfolgung im Fürstbistum Lausanne (1458–1461),” at a conference on Hexenverfolgung und Herrschaftspraxis (Wittlich, 2001); see Ostorero, “Les chasses aux sorcières,” 112. While the Errores Gazariorum surely was an important source for the trials in the Pays de Vaud, the proceedings of these trials do not share with the treatise a special and sustained interest in the devastation of the countryside.

37. For all these texts see now Martine Ostorero, Agostino Paravicini Bagliani, and Kathrin Utz Tremp, eds., in collaboration with Catherine Chène, L’imaginaire du sabbat: Edition critique des textes les plus anciens (1430 c.–1440 c.) (Lausanne: Cahiers Lausannois d’Histoire Médiévales, 1999); for the connection between the Errores Gazoriorum and the trials in the Pays de Vaud, see especially Ostorero’s comment on p. 334–37. For a useful comparison (clarified by an accompanying chart), see Michael Bailey, “The medieval concept of the witches’ Sabbath,” Exemplaria 8 (1996): 419–39.

38. See Ovid, Fasti: Roman Holidays, vi.101–68 (an entry for June 1), trans. Betty Rose Nagle (Bloomington: Indiana University Press, 1995), 155–56; Paul Murgatroyd, Mythological and Legendary Narrative in Ovid’s Fasti (Leiden: Brill, 2005), 4–6; and Samuel Grant Oliphant, “The Story of the Strix: Ancient,” Transactions and Proceedings of the American Philological Association 44 (1913): 133–49.

39. See Richard Kieckhefer, European Witch Trials: Their Foundations in Popular and Learned Culture, 1300–1500 (London: Routledge, 1976), where I emphasize this disjunction between maleficia and the mythology of what I there call diabolism.

40. Ettore Vergo, “Intorno a due inediti documenti di stregheria Milanese del secolo XIV,” Rendiconti del Reale Instituto Lombardo si Scienze e Lettere, 2nd ser., 32 (1899): 166–68; “Le streghe nella Leventine nel secolo XV,” Bollettino Storico della Svizzera Italiana 6 (1884): 144–45; Nicolai Cusae Cardinalis Opera, 2 (Paris, 1514), fol. clxxiir; Carl Binz, “Zur Charakteristik des Cusanus,” Archiv für Kultur-Geschichte 7 (1909): 145–53; Guerrini, Chronache bresciane, 185–86; and Pierangelo Frigerio and Carlo Allessandro Pisoni, “Un brogliaccio dell’Inquisizione milanese (1418–1422),” Libri & Documenti: Rivista quadrimestrale (Archivio Storico Civico e Biblioteca Trivulziana) 21, no. 3 (1995): 46–65.

41. For evidence of Lady Abunda in thirteenth-century French culture, see Roman de la Rose, lines 18411–98, ed. Daniel Poirion (Paris: Garnier-Flammarion, 1974), 491–92; Romance of the Rose, trans. Charles Dahlberg (Princeton, N.J.: Princeton University Press, 1971), 305–6.

42. Ginzburg, Ecstasies.

43. Niklaus Schatzmann, “Hexenprozesse in der Leventina und die Anfänge der Hexenverfolgung auf der Alpensüdseite (1431–1459),” Schweizerische Zeitschrift für Geschichte: Revue suisse d’histoire: Rivista storica svizzera 52 (2002): 138–42; and Verdor-rende Bäume und Brote wie Kuhfladen: Hexenprozesse in der Leventina 1431–59 und die Anfänge der Hexenberfolgungen auf der Alpensüdseite (Zürich: Chronos, 2003).

44. Schatzmann, Verdorrende Bäume, 334–35, 377–80; 335, 380–82; 338, 387–91; 340, 391–95 (feasting on bread and meat and cheese); 341, 398–99 (among the lesser members); 344, 405–8 (savage reaction from devil); 344–45, 408–11 (savage reaction from devil, up to 150 and more at the assembly, stealing infants and sucking their blood); 346, 412–14 (devil demands gifts); 349–50, 421–24 (attendance of the ludus in the form of cats); 354–56, 435–40 (warning not to cross self ); 359, 449–51.

45. Jordanes de Bergamo, Quaestio de strigis, in Hansen, Quellen, 195–200; Girolamo Visconti, Lamiarum sive striarum opusculum, in ibid., 200–207.

46. Hansen, Quellen, 18.

47. Ostorero et al., L’imaginaire du sabbat, 511–13.

48. Pierrette Paravy, De la chrétienté romaine à la Réforme en Dauphiné: évêques, fidèles et déviants (vers 1340–vers 1530) (Rome: É cole française de Rome, 1993), 849, “à l’exception de la vision de Marguerite Vincent-Roux, remarquable par le caractère complet et structuré qu’elle présente, les évocations du sabbat restent rapides, imprécise et floues.” But even this account (given ibid., n. 62) is far less complete and nuanced than in the Lausanne manuscript. (The assembly is referred to in the Dauphiné as a fach or synagogue.) It is worth bearing in mind that Claude Tholosan, the judge in the Dauphiné who also wrote a treatise on witchcraft, believed that the assembly was illusory; see Ostorero et al., L’imaginaire du sabbat, 355–438.

49. Gordon A. Singer, “La Vauderie d’Arras, 1459–1491: An Episode of Witchcraft in Later Medieval France” (Ph.D. diss., University of Maryland, 1974); Franck Mercier, La Vauderie d’Arras: Une chasse aux sorcières à l’Automne du Moyen Age (Rennes: Presses Universitaires de Rennes, 2006); Giovanni Gonnet, “La ‘Vauderie d’Arras’,” I valdesi e l’Europa: Collana della società di studi valdesi (Torre Pelice, 1982), 99–113; Franck Mercier, “L’enfer du décor ou la vauderie d’Arras (1459–1491): les enjeux politiques d’un proces d’inquisition à la fin du Moyen Âge,” Heresis: revue semestrielle d’histoire des dissedence médiévales 40 (2004): 95–121; and Jan R. Veenstra, “Les fons d’aulcuns secrets de la théologie: Jean Tinctor’s Contre le Vauderie: historical facts and literary reflections of the Vauderie d’Arras,” in Nine Miedema and Rudolf Suntrup, eds., Literatur—Geschichte—Literaturgeschichte: Beiträge zur mediävistischen Literaturwissenschaft (Frankfurt am Main: Peter Lang, 2003), 429–53 (see pp. 445–46 for a discussion of the term Vauderie).

50. Thus, the account of the mythology in Mercier, La Vauderie d’Arras, 61–85, is drawn almost entirely from the theological writings rather than from the trial record.

51. Hansen, Quellen, 467–72; Martine Ostorero,”Du danger de prêcher que le sabbat est une illusion: autour du cas de Guillaume Adeline,” in Le diable en procès: Littérature démonologique et sorcellerie á la fin du Moyen Âge, ed. Martine Ostorero and Etienne Anheim, Médiévales: langue, textes, histoire, 44 (Saint-Denis: Presses universitaires de Vincennes, 2003), 73–95.

52. Filhol, “Procès de sorcellerie à Bressuire,” 77–83; Bligny-Bondurand, “Procédure contre une sorcière de Boucoiran (Gard), 1491,” Bulletin philologique et historique (1907), 380–405.

53. Hansen, Quellen, 235.

54. Hoffmann-Krayer, “Luzerner Akten zum Hexen- und Zauberwesen,” 30–33.

55. The literature on this work is extensive; see especially Peter Segl, ed., Der Hexenhammer: Entstehung und Umfeld des Malleus maleficarum von 1487 (Cologne and Vienna: Böhlau, 1988).

56. Michael D. Bailey, Battling Demons: Witchcraft, Heresy, and Reform in the Late Middle Ages (University Park: Penn State University Press, 2002).

57. See especially Michael D. Bailey, “From Sorcery to Witchcraft: Clerical Conceptions of Magic in the Late Middle Ages,” Speculum 76 (2001): 960–90.

58. Perhaps most obviously the Malleus maleficarum (largely inspired by Heinrich Kramer’s trials at Innsbruck and elsewhere) and the writings of Johannes Tinctoris (provoked by the case at Arras in 1459–60), on which see now Jean Tinctor, Invectives contre la secte de vauderie, ed. É mile van Balberghe and Frédéric Duval (Tournai: Archives du Chapitre Cathédral; Louvain-la-Neuve: Université Catholique de Louvain, 1999).

59. Robin Briggs, “ ‘Many Reasons Why’: Witchcraft and the Problem of Multiple Explanation,” in Witchcraft in Early Modern Europe, Jonathan Barry, Marianne Hester, and Gareth Roberts (Cambridge: Cambridge University Press, 1996), 49–63.

Additional Information

Print ISSN
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.