By expanding the geographical scope of the history of violence and war, this volume challenges both Western and state-centric narratives of the decline of violence and its relationship to modernity. It highlights instead similarities across early modernity in terms of representations, legitimations, applications of, and motivations for violence. It seeks to integrate methodologies of the study of violence into the history of war, thereby extending the historical significance of both fields of research. Thirteen case studies outline the myriad ways in which large-scale violence was understood and used by states and non-state actors throughout the early modern period across Africa, Asia, the Americas, the Atlantic, and Europe, demonstrating that it was far more complex than would be suggested by simple narratives of conquest and resistance. Moreover, key features of imperial violence apply equally to large-scale violence within societies. As the authors argue, violence was a continuum, ranging from small-scale, local actions to full-blown war. The latter was privileged legally and increasingly associated with states during early modernity, but its legitimacy was frequently contested and many of its violent forms, such as raiding and destruction of buildings and crops, could be found in activities not officially classed as war.
The previous chapters presented challenges to generalisations about male witches. The next two chapters follow a similar approach to conventional perspectives on the demonological treatment of witchcraft and gender. Through the examination of witchcraft theorists’ descriptions of male witches, we aim to show that, just as with the ‘real life’ cases, modern scholars’ views do not take sufficient account of the complexity of early modern learned theories about witches.
The sources for this discussion are demonological treatises published in the fifteenth, sixteenth and seventeenth centuries. The body of witchcraft literature is much too large to permit a complete survey; there is, however, a smaller group of works that could be considered canonical, at least from the perspective of contemporary scholarship. This ‘canon’ includes Jean Bodin’s De la demonomanie des sorciers, Johannes Nider’s Formicarius, and the Malleus maleficarum, which occupies pride of place within the literature as possibly the most (in)famous treatise of them all. In this chapter, we present data compiled from ten of these canonical works, as well as a brief discussion of demonological illustrations.
Demonological literature has received relatively little scholarly attention, especially compared with the number of published studies that focus on non-literary, archival materials such as court records and pamphlets. Although witchcraft scholars refer frequently to the Malleus and its fellow witchcraft treatises, they rarely engage in sustained analyses of these works. One reason for this neglect is the social-historical emphasis on non-elite subjects; another is the generally poor reputation that demonological texts have as documents of barbarity, superstition, and irrationality. H.R. Trevor-Roper, for example, said of them that
To read these encyclopedias of witchcraft is a horrible experience. Each seems to outdo the last in cruelty and absurdity. Together they insist that every grotesque detail of demonology is true, that scepticism must be stifled, that sceptics and lawyers who defend witches are themselves witches, that all witches, ‘good’ or ‘bad’, must be burnt, that no excuse, no extenuation is allowable … When we read these monstrous treatises, we find it difficult to see their authors as human beings.1
The historiographical tide is shifting, and several monograph studies of demonological works and the contexts in which they were produced are now available.2 There are also essay collections and several individual articles dealing with this material.3 In general, however, demonological literature has not attracted much scholarly attention, and within the historiographical genre of demonological studies, research has focused on the links between demonology and the rise and decline of witch-hunting.
This approach often takes the form of summarising and criticising the arguments of individual authors. For example, Christopher Baxter’s essay on Johann Weyer’s De praestigiis daemonum consists mainly of an evaluation of Weyer’s effectiveness in opposing witch-hunting. Baxter argues (against Trevor-Roper) that Weyer’s work is actually ‘a protest against policies of toleration’; if only, Baxter implies, Weyer had been more rational, he could have achieved the proper goal of arguing for tolerance.4 Baxter remarks in his conclusion that ‘Weyer’s writings badly misfired as a defence of witches’, in part because ‘his simplistic theological commitment prevent[ed] him from pushing through to a logical and systematic re-interpretation of traditional attitudes an unprecedentedly diverse, though individually unoriginal, range of arguments for tolerance.’5
Although there is genuine debate over the role of demonology in the development and decline of witch-hunting,6 interpretations of demonological conceptions of gender and witchcraft are remarkably uniform in their concentration on what witchcraft theorists had to say about women. There is some disagreement over how representative or extreme demonological misogyny was (see chapter 1’s discussion of Stuart Clark’s views), but the focus on women itself has not been challenged. This is understandable, to a point. There is no denying that the major demonological treatises of the period, both those that advocated witch-hunting and those that opposed it, accepted that most witches were women. For example, in their explanation of why more women than men were witches, the much-quoted Institoris and Sprenger incorporated many citations of classical, biblical and medieval authorities. One of these, a citation of Valerius’ letter to Rufinus, reads: ‘Chimeram mulierem esse nescis sed scire debes quod monstrum illud triforme insigni venustetur facie leonis olentis maculetur ventre capre virulente cauda vippere armetur. vult dicere. quod est aspectus eius pulcer. tactus fetidus. conuersatio mortifera.’7 Most modern readers will be more familiar with Montague Summers’ translation: ‘You do not know that woman is the Chimaera, but it is good that you should know it; for that monster was of three forms; its face was that of a radiant and noble lion, it had the filthy belly of a goat, and it was armed with the virulent tail of a viper. And he means that a woman is beautiful to look upon, contaminating to the touch, and deadly to keep.’8 When one reads such statements,it is not difficult to see why modern scholars,especially feminists, have been fascinated by demonological treatments of women.
The problem is not that scholars have paid attention to the early modern demonological discourse about women and witchcraft; it is that they have given their attention only to the discourse about women, as if early modern authors said nothing about men as perpetrators of witchcraft. This is not because scholars are unaware of the presence of male witches within demonological texts. Several historians, all of whom were writing from a feminist perspective,have noticed that early modern authors discussed male witches. For example, Anne Barstow mentions that Henri Boguet, chief judge of St. Cloud in Franche-Comté and author of the Discours des sorciers, said that male and female witches were equally addicted to the carnal pleasures offered by the Devil.9 Carol Karlsen remarks on the language used by early modern authors: ‘While authors of theological descriptions of witchcraft sometimes employed female pronouns when speaking generally about witches, more commonly they used the “generic” male.’10 Susanna Burghartz, while arguing that ‘a concentration on women as witches is certainly evident’ in Nider’s Formicarius, says that ‘all the same, Nider’s examples of witches include an astonishing number of men.’11
The critical point here is that this is all that these authors have to say about the inclusion of male witches in early modern texts: they mention it once, then never bring it up again. In Barstow’s case, the reference to Boguet’s ‘even-handedness’ is not even in the main body of the text; it is buried in an appendix. Burghartz’s astonishment at finding a number of male witches in the Formicarius is particularly telling. Burghartz knows from reading Richard Kieckhefer’s study of medieval European witch trials that in the mid-fifteenth century, when the Formicarius was composed, a significant proportion (32 per cent) of witches were male, and she knows that ‘it was always possible to prosecute men for witchcraft’.12 Why, then, the astonishment? Her reaction to the Formicarius indicates a strong degree of conditioning by a historiography and ideology that is always already committed to a particular way of reading demonological texts. Modern readers appear to assume that there will be no surprises in the gendered discourse of these texts, and then seem unable to address them except in deceptively casual, throwaway remarks.
Even Stuart Clark, whom one might expect to engage with male witches in demonology, does not. Indeed, he does not mention their textual existence at all, despite his careful reading of the early modern gendering of witchcraft. Like the other historians discussed above, he is well aware that many witches were men. He evinces no interest in this fact, however, preferring to focus on why witches ‘were conceived to be women.’13 Clark suggests two points that bear on the subject of male witches: first, that there was a ‘lack of conformity between demonological theory and the actual sexual breakdown in those witchcraft prosecutions where men made up a significant minority or even a substantial portion of those accused.’ Second, he states that ‘it was literally unthinkable’, at a demonological level, ‘that witches should be male,’ because early modern theorists were committed to a hierarchically structured binary framework in which femaleness was associated with evil.14 In effect, Clark argues not only that there was a conceptual affinity between women and witchcraft, but also that there was a conceptual barrier between men and witchcraft.
Clark’s conclusions are similar to those of Eric Wilson, whose Cambridge dissertation is the first modern study in English of the Malleus maleficarum. Wilson devotes a chapter to the issue of women in the Malleus, but never once refers to the fact that the text’s authors also wrote about men who were witches, or to their use of both maleficus and malefica. He notes that ‘the text even employs the feminine form of the Latin word for witch … rather than the more commonly employed masculine.’15 Wilson knows full well that many witches were male, and he states his agreement with Christina Larner that witch-hunting was sex-related rather than sex-specific;16 nevertheless,he concludes that there was a ‘necessary relationship between women and witchcraft’.17
Both Clark and Wilson suggest that early modern demonology was sex-specific and thus different from witchcraft prosecutions, which were merely sex-related. This position is actually more extreme than that of most feminist scholars, who at least hint that demonological conceptions of gender and witchcraft were not always tidy and coherent. It is not, perhaps, completely surprising that it is the feminists who have shown more sensitivity. Their primary purpose, generally speaking, is to explain the preponderance of women in witch trials; they examine the contents of demonological treatises for evidence of systematic misogyny, and are less interested in the treatises as worthy objects of study in and of themselves.
In contrast, scholars such as Clark and Wilson are interested in recovering demonological texts from the historiographical neglect into which they had fallen. Clark and Wilson are not unsympathetic to feminist concerns, and treat the issue of women and witchcraft seriously; however, their underlying concern is with demonstrating the logic and coherence of early modern demonology. This leads Clark and Wilson to exclude male witches from their discussions because they are unable to incorporate them without subverting their intellectual-historical agenda of making the texts respectably coherent.18
We have emphasised the exclusion of male witches from analyses of witchcraft literature so strongly because it is simply not credible that scholars working with these texts (at least those working with original-language editions) do not know about them.19 As table 2 shows, it was not at all uncommon for early modern authors from the fifteenth to the seventeenth century to refer to witches in both masculine and feminine terms. Furthermore, certain treatises included illustrations of both male and female witches, and individual male witches are described in many tracts. It is only by deciding, a priori, that male witches are insignificant that one could treat early modern demonology as sex-specific.
When we ‘discovered’ that Heinrich Institoris and Jacob Sprenger referred to witches in both masculine and feminine terms, one of our first questions concerned the relative frequency of the masculine usage. Just how often did these and other authors employ masculine forms when discussing witches, and did they do so more frequently than they used feminine forms? Modern commentary on the language of demonology (in a philological sense) is sketchy, and we found nothing that answered our specific questions about the Malleus satisfactorily. For example, Sigrid Brauner offers a brief discussion of the use of maleficus and malefica in the Malleus, but her assertion that Institoris and Sprenger ‘use the male plural form malefici for sorcerers in general but reserve the female form malefica for the modern witch’ is mistaken.20 In 1.1, Institoris and Sprenger refer to ‘moderni malefici’ specifically: ‘Tertium etiam sane intelligere expedit cum moderni malefici sepius ope demonum transformantur in lupos et alias bestias.’(And third, it is indeed useful to understand how modern [contemporary] witches [male; and possibly female, according to convention] are often transformed by the power of a demon into wolves and other beasts.)21
Our approach to this problem was fairly simple: we counted the number of times the authors used masculine and feminine terms for witches. Our aim in tabulating this information about linguistic gendering is to challenge Clark’s notion that early modern witchcraft theorists were incapable of imagining that witches could be male, on the grounds that language choices are not accidental and that early modern authors must have meant to use both masculine and feminine terms. If they were capable of representing witches as male, it follows that they were also capable of conceptualising male witches – otherwise,their language would make no sense.
For this purpose, an extensive survey of the very large corpus of witchcraft literature is unnecessary. A small sample of major texts is sufficient to show that witchcraft theorists had no difficulty representing witches as male. The texts were selected partly on the basis of availability in original-language editions, but also for their status as well-known treatises with which most historians will have some familiarity. There is a certain degree of uniformity, in that nine of the texts are ‘pro’ witch-hunting, the exception being Friedrich Spee’s Cautio criminalis.
Early modern terminology for magic and magic-users varied considerably, even within elite discourse. The terms maleficus and malefica were quite common, but so were the terms saga, sortilegus, and veneficus, to list only a few. The sample texts employ several different terms. In our tabulation, we were most interested in parallel forms, such as maleficus/malefica and sorcier/sorcière, but we have also enumerated other terms so as to give a reasonable indication of the relative frequency of references to male and female witches. In general, words that appear only once or twice, or have a specialist meaning, such as necromanticus or pythonissa, have not been included in the tabulation.
The Formicarius, Malleus maleficarum, De laniis, and Flagellum haereticorum use maleficus and malefica almost exclusively. Daneau employs a variety of terms in De veneficis: veneficus (24), sortilegus (16), maleficus (4), and sortiarius (140) for male witches; venefica (1), saga (1), and satanae (86) for female witches. In the Tractatus, Binsfeld prefers maleficus and saga (30), but also uses malefica (17). Rémy’s Daemonolatreia uses sortilegus (24) and saga (34) most often for male and female witches respectively, but also maleficus (6) and malefica (5), and refers to a sortilega once. Bodin and de Lancre employ sorcier and sorcière. In Cautio criminalis, Spee uses maleficus for male witches and saga for female witches. The figures given represent the sum totals of both plural and singular forms.
Where possible, we have used first editions of these treatises, as indicated in the endnotes to the table; however, first editions have not always been available, which raises the issue of textual stability. As Adrian Johns has argued, contra Elizabeth Eisenstein, early modern printed texts were not inherently stable, and the fixity of print in the period has been much exaggerated.22 There is no reason to assume that demonological treatises were any less susceptible to piracy and printers’ errors or emendations than, say, Galileo’s Sidereus nuncius, whose later editions featured corrupted images.23
Indeed, comparisons of two editions of the Formicarius24 and twenty-one copies (sixteen different editions) of the Malleus maleficarum25 reveal that demonological treatises were not completely stable texts. For example, the first lines of both treatises contain textual variants. In the 1480 edition of the Formicarius, book 5 begins: ‘ Ultimo loco per libellum quintum sub formicarum proprietatibus de maleficis & eorum decepcionibus concludere restat’.26 The same line of the 1669 edition reads: ‘Nunc per libellum quintum sub formicarum proprietatibus de maleficis & eorum deceptionibus agere placet.’27 The first line of the 1487 Malleus maleficarum reads: ‘Utrum asserere maleficos esse sit a deo catholicum quod eius oppositum pertinaciter defendere omnino sit hereticum.’28 In several later editions, the word defendere is replaced by either asserere or offendere.29
This textual variation suggests a need for caution when assessing the reception of demonological works. Different editions were highly unlikely to be identical. Historians of witchcraft need to develop an awareness of this instability. Certain variations call into question broad assertions about the reception of the Malleus and other texts: for instance, despite some claims to the contrary, not every edition of the Malleus included a copy of the papal bull Summis desiderantes.30 On the other hand, the gendered terminology of the Formicarius and Malleus is remarkably stable. We have detected only one variant, which is found in the Formicarius. In book 5, chapter 3, the 1669 edition has maleficarum where the 1480 edition has maleficorum.31 Granted, our variants search has been limited; however, we are reasonably confident that this particular feature of demonological treatises was not affected significantly by textual instability, and that the textual (grammatical) gendering of witches was consistent during the early modern period.
Counting words may seem to be a strange way of uncovering textual meaning. We are more accustomed to what we might think of as qualitative approaches to language, in which we puzzle over the meanings and intentions of specific language choices; we do not usually quantify those choices.In this case,however,quantification provides a means of introducing some methodological rigour to the discussion of gender and witchcraft. Given the hegemonic status of conventional readings of demonology, it would be extremely difficult to make the case for the inclusion of male witches as significant subjects without some kind of hard data.
The data presented in table 2 are not meant to establish definitively the universal patterns of gendered language usage in early modern demonology. However, even a rough approximation of the complexity of demonological concepts of witchcraft and gender is an important step toward breaking out of our own tendency to essentialise past ideas.
|148032||Nider||Formicarius, Bk. 5||47||13|
|148733||Institoris & Sprenger||Malleus maleficarum||197||453|
|158035||Bodin||De la demonomanie des sorciers||820||399|
|159138||Binsfeld||Tractatus de confessionibus||157||47|
|161340||de Lancre||Tableau de L’Inconstance||335||296|
These figures represent a reduction of complex usage patterns to a simple dichotomy between masculine and feminine references. Speaking broadly, the references to witches in demonological texts may be divided into ‘abstract’ and ‘concrete’ categories. Abstract references concern the characteristics and abilities of witches as a ‘species’. For example, in the first chapter of book 5 of the Formicarius, the Piger asks the Theologian (‘Theologus’; the Formicarius is written as a dialogue between these two speakers) about witches and their relationship with demons: ‘Desidero igitur noscere Primo quot modis & qualiter malefici & supersticiosi sibique similes reguntur equitantur & dementantur per demones.’42 (I desire therefore to know, first, in how many ways and by what right witches and the superstitious and those like them are ruled, ridden and driven mad by demons.) Abstract references are ordinarily plural, and very often masculine. Nider, for instance, always uses the masculine when referring to witches in the abstract. Some abstract references are singular, and are masculine or feminine depending on the author and context. In the Malleus maleficarum, singular abstract references are usually feminine, but the section concerning whether the Devil and witch must cooperate in order to perform maleficium begins: ‘An catholicum sit asserere quod ad effectum maleficialem semper habeat demon cum malefico concurrere vel quod vnus sine altero vt demon sine malefico vel econuerso talem possit producere.’43 (Whether it is orthodox to affirm that in order to achieve an act of sorcery a demon must always work together with a witch, or that one without the other, for instance a demon without a witch or vice versa, is able to produce such an act.) In this passage, the abstract witch is referred to in the masculine ablative singular form malefico.
Concrete references describe actual witches, usually in connection with a trial. Concrete references are either masculine or feminine, singular or plural, depending on the context. Early modern authors never, to our knowledge, confuse the gender of their concrete references. Whether a woman or a man is the subject of discussion, the term used to describe that individual always agrees in grammatical gender with their biological sex. We have found no instance of a particular male witch described in grammatically feminine terms, or vice versa.
It is important to distinguish between abstract and concrete language because such references have different, but related, heuristic values. Abstract references suggest the degree to which early modern witchcraft theorists conceptualised witches as inherently gendered. It is an imperfect measure of such conceptualisation: abstract masculine references were in all likelihood meant to include females, and in any case grammatical gender cannot always be read as indicative of ‘actual’ gender. For example, several masculine Latin nouns with feminine endings – agricola, nauta, and poeta, to list a few – refer to traditionally masculine occupations. On the other hand, the fact that different forms existed for witches suggests that the use of a masculine noun, such as maleficus, was not purely a matter of grammatical convention. Conceptual flexibility was built into the languages of early modern witchcraft theory.
The texts do provide clues as to how their authors were using language. In the Formicarius, Nider uses the masculine plural malefici unless referring specifically to female witches. His statement that Peter of Como burned many witches of both sexes (‘multos vtriusque sexus incineravit maleficos’) suggests that such usage is inclusive of men and women.44 Here, maleficos clearly means both men and women. This is actually a concrete reference, but it is reasonable to assume that his abstract references are also inclusive, since he devotes a later chapter to the special wickedness of women. Nider does not say ‘maleficos et maleficas’; presumably, this would have been considered redundant.
For the most part,Jean Bodin follows Nider’s pattern.He uses masculine plurals and singulars unless female witches are the specific subject of reference. For instance, Bodin begins his first chapter: ‘Sorcier est celuy qui par moyens Diaboliques sciemment s’efforce de paruenir a quelque chose.’45 (A witch is one who knowingly tries to do something by diabolical means.) This use of the masculine singular in a chapter devoted to defining what a witch is constructs the witch as male, not female – at least in a linguistic sense. Similarly, Pierre de Lancre begins his Tableau de L’Inconstance with a chapter on the inconstancy of demons, in which (abstract) witches are referred to in the masculine. He says, for instance, that ‘A la vérité les Démons ont quelque certaine légèreté, laquelle fait qu’ils peuvent aisément et en un moment surnager et enfoncer, et en communiquer les moyens aux Sorciers non pas que de la on doive tirer une preuve certaine et infailable qu’ils sont Sorciers.’46 (In truth, demons have a certain inconstancy, which makes them able easily and in a moment to float and to sink, and to communicate the means of doing so to witches, not that one should draw from this certain and infallible proof that they are witches [i.e. in a swimming test].)
The linguistic patterns of the Malleus maleficarum are far more complex than in the works of Nider, Bodin and de Lancre. Institoris and Sprenger begin with masculine references, but gradually switch to an almost exclusive use of feminine forms for abstract discussions.This linguistic transformation is not smooth. Institoris and Sprenger explain that the heresy of witches ought to be described in feminine, not masculine terms, because most witches are women: ‘Plura hic deduci possent sed intelligentibus satis apparet non mirum quod plures reperiuntur infecti heresi maleficorum mulieres quam viri. Unde et consequenter heresis dicenda est non maleficorum sed maleficarum vt a potiori fiat denominatio.’47 Nevertheless, they do not always apply this rule consistently. At the end of Part I, while explaining that the sins of witches exceed the sins of Adam and Eve, the text switches several times from maleficorum to maleficarum.In Part III,which deals with the practicalities of witch trials, a discussion about witnesses includes the statement ‘Item sicut hereticus contra hereticum ad testificandum admittitur, ita maleficus contra maleficum.’48 (Likewise, just as a heretic is permitted to testify against a heretic, so a [male] witch may testify against a [male] witch.)
The use of masculine references in abstract discussions of witches was probably not intended to suggest that most witches were male,especially in texts such as the Formicarius, Malleus maleficarum, and Tableau de L’Inconstance, which included sections explaining why women were especially attracted to witchcraft.However,it does indicate a readiness to represent witches as male without any need to justify or question such representations.
Concrete references to male witches reinforce this impression of a flexible linguistic and conceptual framework. We saw in chapter 2 that accusers and officials demonstrated no sense of cognitive dissonance when confronted with male witches. Likewise, the authors of the four treatises discussed here show no signs of confusion or need to explain the existence of male witches, despite their views that witches were predominantly female.
In the Formicarius, Malleus maleficarum, Demonomanie, and Tableau de L’Inconstance, at least one individual male witch is described at length, often recurring in the text as a kind of forensic exhibit. In the Formicarius and Malleus, this witch is Staedelin (or Stedelein), the ‘grandis maleficus’ of Poltingen;49 in the Demonomanie, Trois-eschelles appears several times (Staedelin also makes an appearance); and de Lancre devotes several pages to a discussion of Isaac de Queyran.50 There are many other references to male witches. Bodin, for instance, discusses various famous, learned witches, including Cornelius Agrippa, Pietro d’Abano and Guillaume de Line.51 In the Malleus maleficarum, Institoris and Sprenger include a section concerning three types of witchcraft that only men practice.52
What is so striking about these passages describing male witches is that, again, the authors make no effort to justify using the terms maleficus or sorcier. Clearly, male witches were not considered by these authors to belong to a fundamentally different category of evil-doer from female witches; regardless of sex, they were all witches first and foremost.
Images of witches provide additional evidence for the capacity of early modern Europeans to conceive of male witches. Demonological treatises were rarely illustrated,53 but some of those that were offer further proof that witches were not believed to be exclusively or necessarily female. Most early modern images of witches, for instance the famous drawings and paintings of Albrecht Dürer and Hans Baldung Grien, depict female witches. There are, however, a number of images from demonological tracts that depict male witches.
The first examples are contained in Ulrich Molitor’s De laniis et phitonicis mulieribus, which was first published at Constance in 1488.54 Although the work’s title refers specifically to women, two of the six woodcut illustrations are of male witches. In the first such illustration, three witches are flying on a forked stick while simultaneously transforming into animals. It is clear from the clothing that at least one of the figures is male (see figure 6).55 The second of these woodcuts depicts a man riding what appears to be a wolf (see figure 7). Charles Zika has found that in this particular illustration, the witch is a man in all but one of the almost twenty versions published in the 1490s (the female version appears in the Ulm 1490–91 edition of De laniis).56 Olaus Magnus’ Historia de gentibus septentrionalibus (Rome, 1555) also contains illustrations of both male and female witches. In one woodcut, a male witch is using a knotted rope to influence the winds.57
Finally, Francesco Maria Guazzo’s Compendium maleficarum, first published at Milan in 1608,is a veritable treasure trove of images of male witches. The work contains a large number of woodcut illustrations, almost all of which depict both men and women. One group of four woodcuts shows men and women trampling the cross, being baptised by the Devil, giving him clothing, and swearing allegiance to him while standing in a magic circle (see figures 4, 3, 8 and 5 respectively). There is no doubt that these people are witches. In all four images, a well-dressed man is foregrounded, an equally well-dressed woman stands immediately behind him, and a mixed group occupies the background.58 Another illustration depicts a man and a women together roasting an infant while a couple in the background prepares to boil a child in a cauldron (see figure 9).59
To Charles Zika, the De laniis woodcut of the male witch riding a wolf suggests that ‘the gender link in representations of witchcraft has clearly not yet been established in the 1490s’.60 Given that this image was reproduced in sixteenth-century editions, and that the sixteenth- and seventeenth-century works by Olaus Magnus and Guazzo also included depictions of male witches, we ought to push Zika’s insight farther. The (female) gender link in visual representations of witchcraft was strong in the early modern period, as evidenced by the fact that the majority of images depicted witches as women; however, this link could not have been absolute. If early modern artists and printers had been incapable of imagining male witches, it is improbable that they would have chosen to depict them, especially in illustrations of demonological tracts that furnished many descriptions of the activities of female witches.
Conventional assessments of demonological concepts of witchcraft and gender need revision. First, Stuart Clark’s statement that demonological theory about gender and witchcraft did not conform to the patterns of prosecution appears somewhat too broad. One could argue that although patterns of language usage do not conform precisely to patterns of prosecution, they do reflect them, in that both generally include a minority and, at times, a majority, of men. There is, thus, no wide gap between demonological theory and actual prosecutions.
Second, views of early modern demonology as sex-specific, exemplified by Sigrid Brauner’s assertion that ‘by 1500, the sex-specificity of witches was so widely accepted that it was implicitly assumed in texts about witches’61 and Stuart Clark’s remark that it was literally unthinkable that witches should be male, are clearly overdrawn. Witchcraft theorists may have taken it for granted that witches were mostly female, but they did not treat witchcraft as sex-specific. One can interpret early modern witchcraft theory as sex-specific only by ignoring a considerable body of evidence to the contrary.