There are people who believe themselves to be witches. One can find them without difficulty on the Internet, and on a recent canal trip I was surprised to pass a whole series of narrow-boats (Black Cat, Sorceress) apparently inhabited by practising witches. The modern scholarly literature on the history of witch beliefs and witch trials, however, first took shape in opposition to Margaret Murray’s The Witch-Cult in Western Europe (1921), which claimed that Renaissance witches were worshippers of pagan gods. It has therefore been resolutely agnostic about the existence of actual witches in the period of the great witch-hunt. Even Deborah Willis, who reads the confessions of English witches with close attention, maintains only that ‘they suggest, if not a shared set of practices, at least a shared fantasy life.’
During the Renaissance almost everyone agreed that witchcraft worked. Moreover, being thought to be a witch could have significant practical advantages, at least in England, where those prosecuted had usually been believed by their neighbours to be witches for some years before they finally came to trial, and had in the meantime been treated with due respect. It would be surprising if no one had thought witchcraft worth practising. However, modern studies of Renaissance witchcraft were written in opposition not only to Margaret Murray, but also under the influence of Evans-Pritchard’s Witchcraft, Oracles and Magic among the Azande (1937). The Azande of the Sudan believed witchcraft to be pervasive, indeed held it to be the prime cause of death, yet Evans-Pritchard could find no evidence of any ‘self-conscious’ witches among them. The question ‘Were there people who believed themselves to be witches?’ has thus long been excluded from consideration. But it seems that this self-denying ordinance is becoming increasingly hard to sustain: James Sharpe’s Bewitching of Anne Gunter includes the story of Elizabeth Stile, executed in 1579, told, for what is I think the first time in modern scholarly literature, as if she believed herself to be a witch.
The modern history of the witch-hunt begins with a series of books published nearly thirty years ago by Hugh Trevor-Roper, Alan Macfarlane, Keith Thomas and Erik Midelfort. Until very recently, however, it was the history of trials on the Continent alone that seemed a lively subject, and in 1996 Diane Purkiss could still properly complain of ‘the torpor of English witchcraft studies’. It was as if scholars were mesmerised by Thomas, who appeared to have studied all the surviving printed records, and asked all the appropriate questions. But now the study of the witch-hunt has become truly vigorous, both in England and on the Continent, the last five years have seen almost as many books as the previous 25. In Britain, there are undergraduate modules, special subjects and entire MAs devoted to a subject that had no academic standing 30 years ago. Before 1969, the trials of witches were of interest only to magicians and antiquarians; now they are more intensively studied than Puritanism or the politics of the Civil War.
The subject has obvious attractions. Every aspect of society was reflected in the witches’ trials, so that they can be studied from very different viewpoints: from those of early modern legislators (Bostridge), lawyers (Behringer), theologians and scientists (Clark), as well as from those of both accusers and accused. Because of the nature of the sources, the trials have provided some of the best examples of ‘microhistory’ or ‘the new narrative’. This, pioneered in French by Le Roi Ladurie (Montaillou, 1975), and in English by Natalie Zemon Davis (The Return of Martin Guerre, 1983) and Robert Darnton (The Great Cat Massacre, 1984), involves giving a detailed description of events in the lives of ordinary people and is almost always based on court records – A Trial of Witches is a good, The Bewitching of Anne Gunter a fine, and Shaman of Oberstdorf a superb example of the genre. At the same time, there were enough trials to invite statistical analysis. We know, for example, that in early modern England cows were held to be the victims of witchcraft ten times more often than chickens.
Naturally, witch trials have been a crucial subject for feminists (such as Mary Daly and Marianne Hester), though a feminist analysis is complicated by the fact that women were usually the accusers as well as the accused, and the witch-hunt has not lived up to initial expectations that it would prove to be a holocaust previously hidden from history. Witchcraft also represents an opportunity for New Historicist literary critics such as Deborah Willis to put history to work, and witch narratives invite poststructuralist readings (Diane Purkiss provides a good example). This is a subject now conscious of its own classics – the collection of essays Witchcraft in Early Modern Europe marked the 25th anniversary of Religion and the Decline of Magic – but not yet sufficiently developed to be at the forefront of discussions of theory or method (where it lags behind history of science, for example). And like any boom field it demands a steady supply of new works of synthesis (James Sharpe’s Instruments of Darkness is sensible on England; Robin Briggs subtle and imaginative on Europe).
Before asking what this spate of publication amounts to, we need to put the subject in perspective. In Europe between 1450 and 1750, there were perhaps a hundred thousand trials for witchcraft (not the nine million once claimed); half of those tried were executed; 80 per cent of them were women (but in peripheral areas such as Iceland, Estonia and Finland the majority were men). In England, there were probably two thousand trials and five hundred executions between 1542 and 1736, the period during which witchcraft was illegal. In New England, there were 35 executions between 1620 and 1725, of which 19 were at Salem in 1692. By contrast, some nine hundred were burned in the tiny bishopric of Würzburg alone between 1626 and 1630.
As far as demonology is concerned, England in this period was part of Europe, for the demonologists all read the same Latin authorities; but in other respects England was an exception. On the Continent and in Scotland, Roman Law permitted the use of torture in carefully defined circumstances. Witchcraft cases were regarded as belonging to a small category of exceptional, because invisible, crimes – poisoning was another example – where the normal precautions must be set aside. Thus, torture was freely used to extract confessions to such impossibilities as flying through the air to attend the sabbat. Through torture one suspect could be forced to incriminate large numbers of others, and an enthusiastic judge, a skilled hangman or an anxious ruler (Bishop Philipp Adolf von Ehrenberg of Würzburg was one) could turn any of the scattered, occasional trials which were the norm across Europe into a witch-hunt that ended in the execution of dozens or hundreds of innocents. In England, by contrast, torture was not used, and so (though, as we shall see, the causal connection is not as straightforward as might at first appear) witches did not confess to flying through the air. Instead, they were accused of (and sometimes confessed to) cursing their neighbours all too effectively. Only briefly, in East Anglia between 1645 and 1647, did anything resembling a systematic witch-hunt take place. There, sleep deprivation was used as a substitute for rack and thumbscrew, and at least a hundred were executed.
Many different types of story can be told about witchcraft. But how many? Thomas and Macfarlane were originally inspired by Evans-Pritchard’s conclusion that accusations of witchcraft were a way of handling social tensions. As a result, Evans-Pritchard entered the literature as a social functionalist, and Thomas and Macfarlane were delighted to show that witchcraft accusations in England originated in conflicts over charity between poor widows and their wealthy neighbours.
But Evans-Pritchard’s book had three parts, and historians seem to have read only the first. His discussion of Azande oracles was designed to show how the people’s belief in them was coherent, practical and validated by experience: here Evans-Pritchard presents truth as a cultural artefact, and this section of the book is still admired by philosophers and anthropologists. A quarter of a century ago, Stuart Clark set out to demonstrate that demonology could be approached in these terms, and the successful completion of his vast project adds a whole new dimension to the literature. Such, however, is Evans-Pritchard’s reputation for functionalism that Clark avoids all mention of him, drawing instead on the work of his student and Oxford colleague, Rodney Needham.
Evans-Pritchard’s section on magic was different again: here he showed how witch doctors were trained to fake supernatural events such as psychic surgery. This sceptical approach was based on the conviction that most Azande did not know what was really going on, but that a determined investigator could get at the truth. Early modern witchcraft would appear to invite two sorts of sceptical story. One, obviously is to write of torture, false confessions and misplaced fear; much the best recent study along these lines is Behringer’s book on Bavaria: by studying the practicalities of torture (who did it, under what supervision) he shows how the innocent were transmuted into the guilty while most people continued to believe that justice was being done. Small details, such as the number of cells available to hold prisoners, turn out to be decisive in determining how quickly suspects could be processed and how many were convicted before too many respectable people were caught up in the net and enthusiasm waned.
There is another sceptical story that might be written, however, one closer to the spirit of Evans-Pritchard’s account of psychic surgery. For there are key early modern cases in which contemporaries became convinced that accusations of witchcraft and claims of demonic possession were being deliberately fabricated. James Sharpe now gives us a full-scale account of such a case, that of Anne Gunter, which is well documented because it resulted in a trial in Star Chamber in 1606 in which she and her father were accused of attempting to pervert the course of justice.
Three different approaches – functionalist, relativist and sceptical – were thus available from the beginning, and two new ones have been developed in recent years. First, since John Demos’s work on Salem (1982), some historians have been keen to provide psychological explanations of the fear that gripped whole societies, but in particular those who believed themselves bewitched. Much of the most interesting work done on these lines (Briggs, Purkiss and Willis among the more recent books; while Lyndal Roper’s Oedipus and the Devil is already a classic within five years of its first appearance) has sought to bring psychoanalytic (and particularly Kleinian) categories to bear. That Freudians should want to turn their thoughts to fear, malevolence and witchcraft is hardly surprising – Freud himself took an interest in the first attempt to argue that many of those convicted were innocent, Johann Weyer’s De praestigiis daemonum of 1566, which offered a medical explanation for false confessions. More surprising is the extent to which a Kleinian version of history has, through the study of witchcraft, entered the mainstream of historical culture, despite the marginal status of Freudianism within British culture, and of Kleinianism in particular within American Freudianism (Demos himself was a follower of Erik Erikson).
Second, the founding premise of any conventionally rationalist approach has come under attack from an unexpected quarter. Thirty years ago, Hugh Trevor-Roper (who had himself served as an interrogator for military intelligence) could take it for granted that primitive superstitions were irrelevant to the witch-hunt: witches’ confessions were simply words put into their mouths by their torturers. In Night Battles (1966) Carlo Ginzburg had already demonstrated that in Friuli a peculiar group of people, the benandanti, believed they had the power to leave their bodies and gather together to do battle against evil. Without torture (in Italy and Spain the Inquisition was reluctant to resort to torture and sceptical of the results obtained by it), Inquisitors persuaded them to describe their spiritual journeys and their meetings with fellow adepts. Moreover, the powers they attributed to themselves were at first respected by their society; though within a generation they and their neighbours began to doubt that such magic could be distinguished from witchcraft.
Ginzburg’s approach, particularly as subsequently developed in Ecstasies (1990), has met with considerable scepticism (Perry Anderson was highly critical of it in the LRB), but is triumphantly vindicated in Behringer’s Shaman of Oberstdorf, an extraordinary account of the beliefs of an Alpine horse wrangler executed in 1587. Chonrad Stoeckhlin believed that he travelled outside his body and spoke to angels, and was executed for witchcraft. Ginzburg, Behringer and Purkiss are bringing about a new alliance between history and folklore. Popular beliefs in fairy hordes and armies of the night (often recorded in this century by Nazi historians preoccupied with the history of the Volk) fed the fantasies of medieval Inquisitors and thus lie at the origins of Continental witch-hunting; indeed, it seems that when witches confessed to impossible things, such as flying through the air, they sometimes meant at least part of what they said. If English witches made no such confessions it was not just because they were not tortured: they held very different beliefs about the supernatural. Confessions may be tainted by having been extracted through torture, but the verbatim record of interrogations can still tell us a good deal about the beliefs, and (as Lyndal Roper has shown) the psychic dramas, of the accused.
Of the dozen books discussed here, Clark’s Thinking with Demons is the longest and a tour de force, Behringer’s Shaman of Oberstdorf the shortest and most delightful. Behringer’s versatility in writing two such different books as Shaman and Witchcraft Persecutions in Bavaria, each excellent in its way, is remarkable. Both he and Clark have the capacity to astonish because they take beliefs in ‘impossible’ events straightforwardly and literally. Behringer shows that witchcraft beliefs formed only part of a wide spectrum of supernatural stories, Clark that demonology (properly defined as the study not of the supernatural but of the preternatural, for only God can perform miracles) formed an integral and successful part of intellectual life. It would be difficult to think of two books more different in character: one explores the fate of a single peasant and teases out evidence from fragmentary and unreliable sources; the other (drawing on the vast body of evidence surviving from early modern libraries) maps patterns and structures so large that individuals become merely representatives, inhabitants of fields of intellectual opportunity.
The best of the witchcraft histories have been based on Continental sources partly because English assize court records are much less informative (well-documented cases, such as the 1662 trial at Bury St Edmunds studied in A Trial of Witches, are the exception), and we rarely hear what the accused in England had to say. English courts were interested in maleficium, the curse or spell that led to death or disease; but English villagers had other means of identifying witches. They believed that all witches were accompanied by familiars, demonic animals who fed on their blood by sucking on witches’ ‘teats’ (warts or skin growths, these were very different from the invisible stigma diaboli – patches of skin insensitive to pain – for which Continental torturers prospected). In England, those who confessed to being witches confessed to keeping familiars. It’s impossible to keep familiars out of histories of English witchcraft, but Deborah Willis seems to be the only person to have given them any serious thought. The difficulty they present is that there seems no logic to them: rats and toads may be conventionally evil; but cats and dogs are often benevolent; and it begins to seem as if there is no animal that could not be accused as a manifestation of the devil when one discovers a ferret, a lamb, even a bee suspected of being a familiar.
Willis solves this conundrum by starting with the one thing familiars have in common: they all suckle. Consequently, witchcraft in England has something to do with motherhood, and post-menopausal women – the most frequently accused – have familiars in place of babies. This story seems fragmentary and incomplete, however. It is not the malevolent mothers that puzzle me, but their brutish babies, their ‘imps’. Where did they come from? Where did they go? In what universe were they at home?
Other stories remain to be told. Thomas and Macfarlane were able to describe early modern English witchcraft in terms of conflicts over charity because contemporary intellectuals were keen to stress that this was where relations between neighbours were prone to turn sour; Macfarlane even wrote an article describing one Elizabethan clergyman as a ‘Tudor anthropologist’. In fact, this ‘anthropological’ insight has its origins in the writing of an impoverished Kentish gentleman, Reginald Scot, who was preoccupied with charity because his views had been shaped by the Family of Love, a 16th-century analogue to the Quakers. Thomas and Macfarlane seized on the functionalist moment in Scot’s 1584 attack on witch beliefs (the only systematic attack in 16th-century Europe). But Familists, rather like psychoanalysts, believed that the only devils are within. Moreover, Scot was convinced that magic went hand in hand with fraud, and wrote a how-to manual of conjuring tricks to prove it. He was a determined empiricist: he satisfied himself devils did not exist by trying to conjure them up, and ‘proved’ there were no witches by employing old women as undercover agents. We might as well describe him as a ‘Tudor analyst’ or a ‘Tudor rationalist’ as a ‘Tudor anthropologist’. But, though he was the first to use the words ‘familiar’ and ‘imp’ in print to mean demonic spirit, he had no interest whatever in folklore. And his whole purpose was to show that belief in witchcraft was incoherent and contradictory.
The best of the new books ask questions that had not occurred either to Scot or, almost three hundred years later, to Alan Macfarlane. But we still have no adequate account of the peculiarities of witchcraft beliefs in England, partly because Tudor and Stuart intellectuals assumed that witchcraft was, as the demonologists claimed, the same the world over. Despite thirty years of research, English historians still have to free themselves from this elementary misconception.
Other books referred to in the writing of this piece:
Malevolent Nurture: Witch-Hunting and Maternal Power in Early Modern England by Deborah Willis (Cornell, 264 pp., £13.50, 1995, 0 8014 8194 5)
The Witch in History by Diane Purkiss (Routledge, 296 pp., £15.99, 1996, 0 415 08762 7)
Witchcraft and Its Transformations, c.1650-c.1750 by Ian Bostridge (Oxford, 274 pp., £40, 1997, 0 19 820653 4)
A Trial of Witches: A 17th-Century Witchcraft Prosecution, edited by Gilbert Reis and Ivan Bunn (Routledge, 276 pp., £14.99, 1997, 0 415 17109 1)
Witchcraft in Early Modern Europe, edited by Jonathan Barry, Marianne Hester and Gareth Roberts (Cambridge, 368 pp., £40 and £15.95, 1996, 0 521 55224 9)
Instruments of Darkness: Witchcraft of England 1550-1750 by James Sharpe (Penguin, 384 pp., £9.99, 1997, 0 14 013065 9)
Witches and Neighbours: The Social and Cultural Context of European Witchcraft by Robin Briggs (Fontana, 457 pp., £8.99, 1997, 0 00 686209 8)
Witchcraft Persecutions in Bavaria by Wolgang Behringer, translated by J.C. Grayson and David Lederer (Cambridge, 476 pp., £50, 1997, 0 521 48258 5)
Damned Women: Sinners and Witches in Puritan New England by Elizabeth Reis (Cornell, 212 pp., £32.50, 1997, 0 8014 2834 3)
Send Letters To:
London Review of Books,
28 Little Russell Street
London, WC1A 2HN
Please include name, address, and a telephone number.