Scholarly interest in the early modern European witch-hunts has grown enormously in recent decades, with particular focus on building an accurate profile of the protagonists. It was in the 1970s that Keith Thomas and Alan Macfarlane first developed what has since become the prevailing witch stereotype, as they ‘argued convincingly’ that witches were usually female, elderly, often widowed, and therefore socially and economically marginalised (Gaskill, 1996, p. 258). Supported by strong evidence, the Thomas-Macfarlane hypothesis has since often been accepted uncritically and treated as somehow definitive by many academics. [1] By deconstructing the concept, however, I will demonstrate that the Thomas-Macfarlane archetype is, while reasonably correct in a number of ways, severely flawed in others. Certainly, in imagining their subjects as sharing a particular set of characteristics, both undervalued the significance of geographical variations, as well as the vast complexity and diversity of early modern European people’s lives. A key point I will stress is the importance of listening to contemporary voices themselves, many of which warn that anyone could be a witch.

In the Thomas-Macfarlane model the archetypal witch was female. Thomas explained that James I estimated the ratio of female to male witches at twenty-to-one, while the Macfarlane pointed to the fact that women accounted for 92% of those accused in Essex (Thomas, 1971; Macfarlane, 1970). Writing in 1974, Andrea Dworkin seized on these claims and took them to their maximum. Citing extensively from Heinrich Kramer’s 1487 book, Malleus Maleficarum , Dworkin argued that the witch-craze was a calculated patriarchal assault on women (Dworkin, 1974). Even in its own time, some of Kramer’s comrades considered his treatise radical for the strength of its convictions and, in fairness, it certainly reveals evidence of fierce misogyny and gender bias by 1970s standards too (Kramer, 2015). [2] Over thirty years later Christopher Mackay consequently echoed Dworkin, calling the Malleus a self-conscious attack on the female gender’ (Mackay, 2009, p. 25). Nevertheless, Dworkin’s characterisation of the witch-hunts as ‘gynocide’ and reference to the ‘slaughter’ of nine million women is incredibly extreme. Most scholars broadly agree with Brian Levack’s estimate which puts the total death toll – both men and women – at 45,000 and researchers such as Elspeth Whitney and Clive Holmes found accusations against women were actually often made by members of their own sex (Dworkin, 1974, pp. 118-50; Thomas, 1971; Macfarlane, 1970; Levack, 2006; Whitney, 2001; Holmes, 2001). [3]

Dworkin’s controversial remarks aside, Thomas and Macfarlane’s original claim has some validity. Recent studies have found that women constituted over 70% of all the accused in large parts of Europe, including central regions such as the Holy Roman Empire, Poland, and Hungary, as well as most of Scandinavia, and much of the British Isles. Using collated data from numerous smaller studies, Brian Levack estimated that women made up around 75% of all those accused in Europe during the period (Levack, 2001; Levack, 2006; Roper, 2004; Gaskill, 1996; Rowlands, 2013). [4]

Clearly, more women than men were accused of witchcraft. However, this must not obscure the fact that a significant 25% of the accused were men. The existence of male witches has only really begun to be explored by academics in recent years, for example in Laura Apps and Andrew Gow’s 2003 monograph, Male Witches in Early Modern Europe , and Rowlands’s 2009 volume, Witchcraft and Masculinities in Early Modern Europe (Apps & Gow, 2003; Rowlands, 2009). Evidence of contemporary theological support for the notion of male witches is abundant, as Malcom Gaskill points out: in the minds of early modern Europeans ‘witches and women were never equated’ (Gaskill, 2001, p. 107). One tract warned, for instance, that ‘neither be they all women’ and another proclaimed that ‘men, as well as women, may be subject to this trade’ (Roberts, 1616, p. 5; Cooper, 1617, pp. 180-81). Similarly, one of the earliest descriptions of the illicit Witches Sabbath [5] reported the presence of sorcerers ‘of both sexes’ (Nider, 2015, p. 55).

In fact, when broken down by region, surviving records show that in some areas male witches accounted for a substantial proportion of the accused. Around 40% were men in Aragon, 50% in Finland, 58% in Friaul , 60% in Estonia, 70% in Russia, 75% in Normandy (though the figure was around 50% in France more generally), and an enormous 92% in Iceland (Levack, 2001; Levack, 2006; Kivelson, 2001; Roper, 2004; Schulte, 2009). The idea that witches were almost always women would therefore not at all have matched with the lived experiences of the people residing in these areas. Even in places where women predominated among the accused though, suspicion could still easily fall on men. For instance, in the city of Salzburg and the region of Carinthia, both part of the Holy Roman Empire, men made up a sizeable 59% and 68% of the accused respectively (Schulte, 2009).

In addition to being female, the idea that witches were aged is, according to Rowlands, practically an axiom of early modern witchcraft research (Rowlands, 2001). Macfarlane estimated that as many as 80% of the accused in Essex were over forty and Jonathan Durrant claimed the same was most likely true of Germany too (Durrant, 2007; Macfarlane, 1970) Levack concluded that, across Europe, ‘a solid majority’ of the accused were probably over fifty (Levack, 2006, p. 149). Surviving records rarely noted the age of the accused, making it difficult to be definitive, but their research is underscored by the prevalence of the elderly in popular contemporary portrayals of witches. In England a 1655 fable described one as 'long nos'd, blear ey'd, crooked-neckt, wry-mouth'd, crump-shoulder'd, beettle-browed' (L.P., 1655, p. 7). A Kentish gentleman’s description of a defendant on trial is remarkably similar: '[she was] lame, bleare-eied, pale, fowle, and full of wrinkles' (Scot, 1584, p. 7). [6] In 1646 a minister even complained that people were seeing witches in everyone ‘with a wrinkled face, a furr'd brow, a hairy lip, [and] a gobber tooth' (Gaule, 1656, pp. 4-5).

Some historians have argued that older women were more likely to be accused because of the contemporary belief that their age gave them the propensity to exhibit erratic or unpleasant behaviours (Rowlands, 2001). Early modern Europeans typically subscribed to the view that the menopause fundamentally altered the balance of the bodies’ four regulatory humours. [7] They held that women’s habitually ‘cold’ and ‘moist’ bodies became increasingly ‘dry’ as they stopped menstruating, which adversely affected their moods and made them susceptible to the devil's deception. It also allegedly gave them a desire for moisture that made them jealous of younger fertile women, causing friction within communities (Roper, 2004). Robin Briggs tried to verify this, claiming that accused women tended to first arouse suspicion once they reached menopausal age (Briggs, 1996). In 1949 Sona Burstein offered an alternative view, arguing that these erratic behaviours were actually a result of oncoming senile psychosis . Indeed, the sixteenth-century physician Johann Weyer remarked, for example, that the elderly were ‘by their age not sufficiently settled in their minds', leaving them ‘more subject to the devil's deceits’ (Rowlands, 2001, p. 52).

Again, while there is evidence to support the notion of the witch as stereotypically elderly, some recent scholarship presents a more complex picture. Levack’s Witch-Hunt in Early Modern Europe shows that the majority of witches in Scotland were under the age of fifty, for instance. Lyndal Roper’s article in New Perspectives on Witchcraft, Magic and Demonology similarly demonstrates that older women in Augsburg appeared less often in trials after 1700, when children became the primary target of suspicion. In fact, close to half of all accused witches in Württemberg, Saarland, and Würzburg, as well as a majority of those in Rothenburg, were under fifty years old (Levack, 2006; Roper, 2001). [8]

A number of popular contemporary sources appear to counter the traditional age narrative too. Hans Baldung’s artworks, Three Witches and Witches Sabbath, both include a mix of younger and older women, for example (Roper, 2004). Kramer’s Malleus was also eager to stress that individuals of any age could be ensnared by evil. He gives one example where the devil had sought to corrupt a young virgin girl, while another passage described twelve-year-old girls copulating with incubi (Kramer, 2015). [9] Similarly, French Judge Pierre de Lancre recorded instances of thirteen-year-olds participating in sexual orgies at the Witches Sabbath and a Scottish minister described children accepting the devil’s mark (Lancre, 2015; Hutchinson, 2015). [10]

Contemporaries across Europe clearly did not think of the witch as archetypally elderly. Referring to records of witchcraft trials in Rothenburg, Rowlands points out that they say very little about the haggard appearance of the suspects; it is their actions they focus on. If their bodies are mentioned, it is often simply in reference to the witches’ marks that had helped prove their guilt and, for Rowlands, it was the very physical inconspicuousness of many suspects that led accusers to search them for marks in the first place. Furthermore, the notion that the menopause somehow makes older women targets of suspicion contradicts the fact that contemporaries identified menstrual blood as corrupting. On that basis, post-menopausal women would have been considered less dangerous than pre-menopausal women and girls. It is also highly unlikely that of the estimated 90,000 people accused of witchcraft, anything like a majority were suffering with senile psychosis as Bernstein claimed (Rowlands, 2001; Levack, 2006) . James Sharpe rightly warned against these kinds of arguments, calling them ‘simplistic reductionism’ (Sharpe, 1996, p. 10).

Most importantly, as the Malleus suggested earlier, a number of the elderly women prosecuted as witches might actually have been far younger when they first aroused suspicion. For instance, in some German territories an individual could only be prosecuted for practicing magic after receiving numerous denunciations, meaning suspects might live for years before we get any legal record of their existence. Complainants therefore needed to cultivate support for their accusations from their communities, not least because the defendant could also lodge a slander suit against their detractors as a counterattack (Levack, 2006; Rowlands, 2001; Rowlands, 2003; Kallestrup, 2015). [11] But contemporaries also saw the ability to cast spells as a transferrable skill passed down through the generations. Macfarlane himself acknowledges this, highlighting one case where a sorcerer named Joan Cunny claimed to have first learned witchcraft some twenty years before she was finally caught and prosecuted (Macfarlane, 1970).

This idea that witches were mostly of at least middle-age is likely responsible, at least in part, for the belief that they were also characteristically widowed. Thomas insisted that ‘many of them’ were widows and Macfarlane supported this claim by pointing out that they accounted for 40% of all the accused women in England (Macfarlane, 1970). Modern research has built on this, demonstrating that similar trends prevailed in other parts of the continent. For instance, around half of those accused in Lorraine, Toul, and the Inquisitorial Tribunal held at Siena between 1580-1721 were widowed. They represented a clear majority in Saarland, accounting for 64% of the total (Rowlands, 2001; Rowlands, 2009; Levack, 2006).

This research also shows that geographical variations are, again, profound. To give a few examples, the combined efforts of Levack and Rowlands show that widowed women made up just 7% of the accused in Mainz, 19% in Sweden, 21% in Scotland, 23% in Rothenburg, 26% in Venice, 30% in Horn, 33% in Basel, 34% in Geneva, 36% in the Jura region, 38% in Montbéliard, and 38% in Kent. Levack and Rowlands’s research found that it was married women who generally made up the majority of the accused in these regions, followed by widows, and then single women (Levack, 2006; Rowlands, 2001; Rowlands, 2003).

It is partly because of precarious status of widows in early modern society that historians have seen them as a natural target of accusations. Certainly, the loss of a husband, whose earning potential would almost certainly have exceeded their own, could easily see a woman plunge toward the bottom of the social ladder. (Wrightson, 2004; Gowing, 2012). [12] The Thomas-Macfarlane ‘charity refused’ hypothesis posited that widowed women were often dependant on community aid in order to survive in the harsh economic climate of the period. However, if a neighbour refused to provide such help and then experienced some misfortune in their life, they might suspect the alms-seeker had used magic to exact revenge. The example of Margery Stanton was, according to Thomas, ‘typical’. One neighbour turned Stanton away from his house, whereupon his child became ill; another denied her milk and also became unwell. After a long list of offences, Stanton was eventually tried at Chelmsford in 1579. (Thomas, 1971, p. 662; Macfarlane, 1970).

There are problems with the ‘charity refused’ model devised by Thomas and Macfarlane. Firstly, it fails to acknowledge that the poor probably appear disproportionately vulnerable to suspicion simply because they were by far the largest social group. Moreover, it is also true that those accused of witchcraft were not necessarily economically or socially inferior to their accusers. Malcolm Gaskill highlights the 1586 trial of Joan Cason in Kent, who he notes was probably wealthier than her accuser, [13] whilst records from Scotland list the status of some victims as nobility (Gaskill, 1996; Larner et al., 1977). Likewise, Lyndal Roper discovered that some of those tried in Nördlingen were members of the interrogator’s own social circle (Roper, 2004). In fact, in some cases status, power, and wealth could make an individual the target of persecution. In 1629 a Rothenburg maidservant was banished from the city for spreading rumours that her master’s wife was a witch, in addition to telling tales of a nocturnal dance that had been attended by some of the city’s most distinguished residents and politicians (Rowlands, 2001; Rowlands, 2003). A year earlier another extraordinary case came to light in Bamberg, the centre of one of Europe’s most concentrated witch-hunts. The city Mayor himself was accused, tried, tortured, and executed for practicing magic (Junius, 2015).

Indeed, in many cases the victims of accusations were not especially antagonistic, nor were they socially isolated (Rowlands, 2001; Gaskill, 1996). Many of those tried in Kent during the mid-seventeenth century, for instance, were well-known faces in the community and they could attest to their innocence by drawing on the support of friends. (Gaskill, 1996). Other scholars have found similar examples from places like Scotland, Eichstätt, and Venice (Larner et al., 1977; Durrant, 2007; Seitz, 2011). [14] It may even be the case that the high acquittal rate for accused witches in Europe – estimated at 50% – was precisely a result of the support many managed to muster from their communities (Gaskill, 1996; Levack, 2006).

The fate of the Mayor of Bamberg is indicative not only of the often chaotic nature of the witch-hunts, but of the more diverse social, economic, and political environment that helped shape them. It also suggests that disputes between factions within a particular social group could lead to someone making a formal denunciation against a perceived enemy. In 1610 Joan Cariden verbally assaulted Robert Greenstreet, the Mayor of Faversham in Kent, after becoming increasingly unhappy with his administration. A new Mayor was soon elected in his place, but decades later, when the witch hunts in England were at their height, Greenstreet took up office once more. Within twelve days he had Cariden thrown in gaol and, shortly afterwards, executed as a witch. (Gaskill, 1996). More common were cases like the one Louise Kallestrup records from the town of Skagen in Denmark. Here, Anne Lundtz was convicted of witchcraft by the town court, after numerous witnesses testified to her use of dark magic. In keeping with its legal system, the trial then moved to one of Denmark’s provincial courts, but as no one turned up to accuse Lundtz formally, the case was thrown out. (Kallestrup, 2015).

This paper has set out to challenge the stereotype of the early modern witch, presented by scholars such as Thomas and Macfarlane. The image of the witch as an elderly female widow, of limited social and economic means has some merit as I have demonstrated using a range of sources. The evidence in favour of the Thomas-Macfarlane stereotype is most compelling when examined from a distance, as Levack showed when he claimed that in Europe as a whole 75% of the accused were women. However, while there seems to be, as Rowlands put it, ‘an element of truth’ in the stereotype, in drawing on modern monographs and in-depth regional studies this paper has also exposed some of its more serious flaws (Rowlands, 2001, p. 88). For instance, recent scholarship on male and child witches, as well as on the social and legal dynamics behind accusations, which demonstrate they could be driven by factionalism and potentially take decades to reach court, where they may never even be examined.

How, then, in light of these findings, might we re-define the early modern European witch? Broadly speaking, the accused was usually a married woman. However, her particular situation would inevitably vary enormously according to time and geography. Differences in age, social status, and financial power create a more nuanced picture, making it hard to determine how these characteristics fit as part of an overarching archetype, though we can say that the lower-classes are well-represented among the accused. What becomes clear in this paper is the folly of relying on such stereotypes in the first place. In order to represent the past with real accuracy and sensitivity, it is imperative the historian is attentive to the enormous dynamism of everyday life in early modern Europe. Among the accused were young unmarried women, prominent men, children, the wealthy, the poor, and many others, in addition to those that align more closely with the stereotype. Thomas and Macfarlane were pioneers, writing in the 1970s when scholarly interest in the early modern European witch-hunts was minimal; that the subject inspires so much passion and debate amongst academics today is testament to their achievements. The stereotype they inspired is, however, something of a blunt instrument. Blunt instruments have their uses of course, but they do not encourage precision or accuracy. By bringing together the best modern research, with close reference to a wide range of contemporary sources, this article has promoted a more complex, and I might cautiously argue more fascinating, view of the early modern European witch-hunts and its victims.