7

THE EARLY MODERN PATCHWORK

THE EXECUTIONS INSPIRED by the new concept of the satanic witch lasted from those first known examples in the Pyrenees and at Rome in 1424 until the final one in Switzerland in 1782. Four decades of intensive work by experts drawn from virtually every European nation have resulted in a consensual picture with regard to most features of the trials that produced them.¹ Between those two dates between forty and sixty thousand people were legally put to death for the alleged crime of witchcraft, with the true figure more probably in the lower half of that range. This figure is, however, deceptive in two ways, for the trials were concentrated both in space and in time. They were found mostly in a zone extending across Northern Europe from Britain and Iceland to Poland and Hungary, and from the extreme north of Scandinavia to the Alps and Pyrenees. Furthermore, even within the region across which trials were relatively common, the new concept of the demonic witch proved to be a slow-burning fuse. During the fifteenth century it was confined mainly to the western Alps, northern Italy and Spain, the Rhineland, the Netherlands and parts of France; and does not seem to have claimed more than a few thousand victims at most. Between 1500 and 1560 this range did not much expand, and the overall number of trials seems to have decreased, before an explosion in the second half of the century.

Most of the victims claimed by the early modern witch-hunts in fact died in the course of a single long lifetime, between 1560 and 1640. Two factors may account for this. One is that it was the period in which the crisis in European religion ushered in by the Reformation came to a peak, and Catholic and Protestant engaged in a series of all-out contests. This sent the religious temperature to fever level in many places and individuals, and produced a greater willingness to perceive the world as a battleground between the forces of heaven and hell. The typical proponent of witch trials was a pious reformer, the age’s equivalent to the Observant friars of the early fifteenth century, who wanted to purge society of wickedness and ungodliness; to such people the destruction of witches was usually only a single item on a list of measures to achieve an ideal Christian polity. The period was, however, also marked by the nadir of a long climatic downturn producing colder and wetter weather and decreased crop yields. Although this only rarely acted as a direct provocation to witch trials, it probably produced a general atmosphere of heightened vulnerability and insecurity which encouraged them. After the 1640s they diminished in their heartland but spread out to the fringes of Europe instead, and were most numerous in Poland, Hungary, Croatia, the Austrian lands, Sweden, northern Norway, Finland and New England in this later period: in most of those areas, they were produced by the introduction of heightened religious reform and intolerance.

Unlike the earliest proponents of the new stereotype of the demonic witch, those who warned against it in the most intense period of trials no longer claimed that it was a recently appeared menace. Instead, they portrayed it as one known since ancient times, but suddenly swollen to unprecedented proportions as Satan reacted to the opportunities created by religious division and the challenge presented by the extension of Christianity to large areas of the Americas and some of Asia. In their view the crisis represented by a novel superabundance of witches required a proportionately determined response to detect and destroy them.² The results of such a response differed markedly from place to place. It seems likely that the majority of European villages, even in areas of relatively intense witch-hunting, never produced a single arrest for witchcraft, and trials were notably rare in large cities. Averaged out across the Continent, about three-quarters of those tried were women, but this figure conceals major local variations. Likewise, the majority of victims were drawn neither from the wealthier ranks of society nor from the very poor, being ordinary peasants and artisans like their accusers, but, again, local experience threw up exceptions to this rule. If they conformed to a particular human type it was that of the bad neighbour, quarrelsome and inclined to curse and insult; yet very many were generally normal personalities who happened to have the wrong friends or enemies at the wrong moment.

On the whole, trials were most frequent, and execution rates highest, where the people in charge of the criminal justice system were most closely involved in the local fears and hatreds that produced the accusations. These could include very small states, such as many in Germany, which had over two thousand different jurisdictions, or the Swiss Federation, or where a relatively decentralized machinery of justice prevailed, as in Scotland or Norway.³ Where such areas were also characterized during this period by local elites or a local ruler determined to purify religion and morals, and economic, religious and social tensions, the preconditions for witch trials were in place. Religious identities were in general irrelevant to the matter, as the most intense regions of witch-hunting were Calvinist Scotland, Lutheran northern Norway, and some Catholic states in western and central Germany and in the Franco-German borderland. In most places the pressure to prosecute came from below in society, originating among the common people, but in a minority it was imposed by the rulers of the state concerned.⁴ Some of the worst rates and totals of execution were produced by ‘chain-reaction trials’, in which large numbers of people were arrested and forced to denounce yet more; but territories such as Lorraine, where one or two people were accused at a time, could still accumulate large overall death tolls. Ultimately, the incidence of witch-hunting in a particular area, even one with all the necessary preconditions of trials, could be a matter of caprice in which factors of fortune and personality were dominant: in the 1610s the lordship of L’Isle, in the Swiss Pays de Vaud, had all the preconditions identified above, and yet of the four villages in it, only one produced accusations (though those escalated into a particularly savage hunt there).⁵ It seems that across most of Central and Western Europe there were substantial numbers of suspected witches living among their neighbours but who were never denounced to a magistrate.

It is abundantly clear that personal and factional enmities, and political ambitions, often formed a context for accusations. The latter never seem, however, to have been merely a pretext for the resolution of such other tensions: rather, they were generated by very real fears of bewitchment. In most places a legal proceeding was an expensive, difficult and inconvenient solution to those fears, with counter-magic, or the reconciliation or intimidation of the presumed witch, much easier options. Service magicians were sometimes denounced as witches, but seem to have made up a minority of the accused in any area studied. They were often more prominent in the process of accusation as providers of initial magical counter-measures against suspected witches or detectors of them. The period of witch trials may be regarded in many ways as a scientific experiment, fitted for an age of geographical and scientific discovery, and as such it failed. Witch trials did not seem to yield obvious and measurable benefits to communities that engaged in them, and the problem of providing clear evidence of guilt came to seem worse rather than better over time. Their context of religious conflict and intolerant confessional states faded away as European elites came to feel more prosperous and secure, and judicial machinery was subjected to greater central supervision and direction. Suspected witches, Christians from other denominations, and fornicators and adulterers, were all increasingly tolerated together, by central and local rulers inhabiting a more rational and less demonic universe, ordered by a less demanding and more remote God. By the end of the twentieth century it had become common for experts to reject single-cause explanations for the witch trials, and to adopt an approach to accounting for them summed up in the phrase ‘many reasons why’.⁶ As a means of finding the reasons for why trials occurred in particular places at particular times, which was the main business of those who propounded it, this was certainly the best theoretical perspective: the mesh of preconditions and triggers differed significantly between locations. As a means of accounting for the trials as a whole, however, it suffers from a flaw: that the single obvious reason for them was the appearance of the stereotype of the demonic witch in the fifteenth century, and its eventual application to a wide swathe of Europe. As one of the best-known witch-hunters, the French judge Pierre de Lancre, put it, the description of witches’ assemblies ‘which occurs in various lands, seems to be somewhat diverse . . . But, taking everything into consideration, the most important ceremonies are all consistent.’⁷ It is the manner in which local people emphasized certain aspects of the stereotype rather than others, and infused it with their own traditions and preoccupations, that gives regional studies of their trials their especial importance: but the centrality of the stereotype remains, even if it was not in itself a sufficient cause of witch trials. What is proposed in this chapter is to survey the findings of those regional studies, made across Continental Europe mostly during the past five decades. This exercise is undertaken with the particular intention of enquiring into what effect, if any, ancient and folkloric beliefs had in determining the incidence of witch-hunting, the images expressed in it, and the identity of the victims of it.

Such an undertaking has to reckon with two problems of evidence which beset anyone with an interest in the popular component in witchcraft beliefs and witch-prosecution. The first is that evidence for the ideas that propelled trials is found in only a minority of the surviving records of them. The second is that by definition, virtually all those records were made by members of the social and political elite. They were almost always concerned to prove what they were hearing to be either illusory or demonic, rather than asking the kind of questions about belief and identity in which historians would now be most interested. Moreover, both accusers and accused in court cases were operating under conditions of restraint, by which their statements were expected to conform to certain predetermined models to enable a trial. These problems have, however, been obvious to most of the scholars who have carried out the studies on which the synthesis here is based, and they have developed strategies to cope with them. It remains apparent that ordinary people sometimes made statements which the investigating authorities found surprising, disturbing or irrelevant, and which were still incorporated into the record; and that there were distinct differences between early modern images of witchcraft and patterns of prosecution in particular regions, which can be accounted for in terms of local tradition. It is that relationship, between tradition, action and written record, which is at the heart of this chapter.

The Dream Warriors

The proposed survey might well begin with Carlo Ginzburg’s benandanti, the most spectacular example of a magical folk tradition yet uncovered in early modern trial records. In general, they were simply service magicians in the extreme north-eastern Italian province of Friuli, healing, divining and breaking bewitchments like all of their kind. As said before, however, they also sent out their spirits at night (in the ‘Ember’ fast days which fell in each of the four seasons) to fight witches for the fertility of local farmlands. Like those of the witches, their spirits rode horses, cats, hares or other animals to the battlefield, formed into battalions with flags and captains and duelled with plant stalks. If they won, a good harvest resulted, and in any event they returned at the end to their sleeping or entranced bodies. They were not invested with this power but gained it naturally by being born with a membrane, the caul, over their heads at birth, and were called to fight when they grew to adulthood. A few claimed to visit the dead on their spirit-journeys and learn their fate. They battled in the name of God and Christ against witches as servants of the Devil – representing in Ginzburg’s words ‘a Christian peasant army’ – but their identity as magicians drew the attention of local inquisitors from the late sixteenth century. Benandanti began to denounce people to the inquisition as witches, and to be denounced themselves, and their night flights were assimilated into the image of demonic witchcraft. In the mid-eighteenth century they disappear from history.⁸

Immediately to the east of Friuli, in Slovenia and the Istrian Peninsula, the south Slavonic cultural zone began, and in Istria an Italian commentator recorded in the seventeenth century a belief in people called cresnichi or vucodlachi. These were born with a caul, and their souls were believed to go by night, especially in the Ember Days, to fight in bands for the fertility of the coming season. Unlike the benandanti these retained a presence in folklore collected in later centuries, as kresniks or kudlaks. The former were almost identical to the benandanti, save that their spirits went forth at night in animal form rather than riding animals. The latter were malicious magicians, who took the place of witches in some places in the role of fighting the kresniks (like them, taking animal shape), who protected sleeping humans and the farmlands.⁹ In the huge south Slav region to the south-east, comprising Serbia, Bosnia, Herzegovina and Montenegro, the equivalent figure was the zduhač, also born with a caul and able to wage spirit-combats in the clouds to protect their clan territory, although these were fought against the zduhačs of rival clans. They are recorded in relatively modern folklore, as are similar personalities under other names in parts of the same region and of Macedonia, Bulgaria and Croatia. Some of these went in animal form, some waged battle against the spirit-champions of rival communities, and some opposed witches or evil spirits. All, however, were distinguished by a caul or other unusual features at birth, and all fought in spirit form at night, usually at special seasons, to protect their villages, and especially the crops of those. Most of this area was under Muslim rule in the early modern period, of the Ottoman Turks, and so bereft of the sort of records that could throw up references to them in Christian areas: but an inquisitor in Dalmatia, the Croatian coastal region, in 1661, reported a belief in good spirits who chased away bad weather.¹⁰

In other parts of the same huge region, and further east and south-east, bands of young women or men acted out the spirit-battles in physical form, by touring their neighbourhoods in the Whitsun season, when fairies and devils were supposed to be especially menacing, to perform dances, plays or blessings to protect homes and farmlands. This custom was recorded in southern Macedonia as early as 1230, and in modern folklore collections in Croatia, Slavonia, Serbia and Romania, where it persists to the present day. In the Romanian districts, the modern dancers also acted as healers, of people thought to be afflicted by fairies or demons, and their patron, who was also in places the queen of the fairies, was ‘Irodeasa’. This was probably the Herodias of the medieval night rides, and shows how far the tradition of her as a nocturnal spirit had travelled from its apparent Western European source.¹¹ To the north of the south Slav region lies Hungary, and its figure of the táltos, which has been discussed above. It was argued there that neither this figure nor the benandanti could be confidently assigned to a pan-Eurasian shamanic province, and it may be considered here whether the táltos makes a better fit with the Balkan one of spirit-battles. It seems that it makes a remarkably good one. Like the other figures in the Balkan tradition, the táltos was marked out by special signs at birth, and sometimes operated through dream or trance, and fought for the good of local people in spirit form and animal shape or on animal steeds as well as carrying out all the usual functions of a service magician. Táltosok tended to portray themselves as Christian saints fighting satanic opponents, witches and demons. Their battles were also sometimes against each other and against foreign magicians.¹² They are mentioned in medieval Hungarian records, and quite good early modern accounts of their beliefs and claims were provided by witch trials, and indeed other elements of those Hungarian trials seem to derive from the same folk tradition. Early modern Hungarian witches were often thought to send out their souls to work evil, through their mouths, in the form of a small animal such as an insect. They were said to ride on animals or take their form, and fight over the fertility of land, occasionally on behalf of their own villages, although they generally worked harm instead.¹³ One Hungarian scholar has noted from the trial records that a characteristic of Hungary was the ‘duel of magicians: healer, táltos, midwives and witches all fight each other’.¹⁴ People tried as witches in Croatia in the same period spoke of forming into military companies at their gatherings.¹⁵ In modern Romania it was believed that witches were born with a caul, transformed into animals, and went out in spirit guise at night to form bands that fought each other.¹⁶ Modern Serbs thought that witches worked evil by letting their spirits leave their bodies in the shape of insects or birds.¹⁷

There seems to be good evidence here of a compact expanse of territory, with its centre in the south Slav cultural province, characterized by a belief that both service magicians and witches sent out their spirits on special nights, in dreaming or entranced states, to do battle with opponents. Whether this had existed from pre-Christian times, or was a popular tradition that had developed in the Middle Ages, is impossible to tell. In this perspective the benandanti become Italians who learned the idea from Slavonic neighbours, and the táltosok are Magyars who did the same, whatever remote and unproven connection they may have had with Siberian shamans which may have predisposed them towards the idea. The idea that to be born with a caul conferred abilities to communicate with spirits, or send out one’s own spirit, is found across Europe, but only in this area was it associated with the dream or trance battles.¹⁸ Here, therefore, the early modern trials of witches and other magicians did uncover a distinctive regional belief system, and so provide historical records of it; but the system itself neither provoked nor transformed the trials. Those were introduced into the region as part of the Catholic Counter-Reformation movement of spiritual purgation and renewal, and in much of it as part of German (or Austrian) cultural imperialism. The benandanti were not the main business of the inquisition in Friuli, and only twenty-six of 2,275 witch trials recorded in Hungary mention a táltos.¹⁹ Rather, the Balkan ‘dream warriors’ only served to tinge the trials slightly with aspects of their belief system, at the extreme northern and western limits of their range.

The Northern Shamanic Region

Earlier in this book, it was suggested both that the Sámi of northern Scandinavia represented a people who had a ‘classic’ shamanism of the sort found in Siberia, definitely descending from a pagan past, and that Sámi influence created a ‘sub-shamanic’ zone among the Norse, including their settlements in Iceland. Another such zone existed among the Finns, either because of their own ancestral connection with western Siberian peoples or because of Sámi influence. As the Sámi retained a formidable reputation as magicians all through the early modern period – which has also been discussed earlier – it is an obvious question how far they attracted charges of witchcraft when the Scandinavian kingdoms to which they belonged became notable centres of witch-hunting in the seventeenth century. The answer is that they certainly featured in witch trials, a total of 113 being prosecuted, and over thirty of them executed, in both the Norwegian and Swedish provinces of Lapland: the Norwegians tried about half the number that the Swedes did, but put to death three-quarters of them while the Swedes spared the lives of all but a few of those tried. Two other features of these statistics are significant. The first is that they reversed the typical European gender balance, for seventy-three of the seventy-six Sámi tried for magic by the Swedes were male, and so were nineteen of the twenty-seven burned by the Norwegians. This simply reflected the balance within Sámi culture itself, in which the practice of magic, with its shamanic rites, was associated mainly with men. The other feature is that despite their reputation, the Sámi magicians were not the main target of the witch-hunts in the far north. Finnmark, the extreme north-east province of Norway, was the scene of one of the most intense witch-hunts in Europe. It took place, however, principally among the Norse settlers in the fishing villages, and – in conformity with the Norwegian as well as the European norm – its victims were overwhelmingly female. The same is true of the main Swedish witch trials, which claimed hundreds of lives in the heartland of the kingdom. One reason why the Sámi tried by the Swedes normally escaped execution was that they were regarded as savages practising tribal superstitions rather than as recruits to the satanic witch conspiracy. Even the more severe Norwegians tended not to accuse Sámi of consciously worshipping the Devil, which is why about a third of those tried for magic escaped execution; and several of those who did not were found guilty of more routine crimes as well. Though they made up almost half the population of Finnmark, they represented 18 per cent of those accused of magical offences. Nor were the spectacular shamanic rites recorded among the Sámi apparent among those actually put on trial: they feature, if at all, in muted form.²⁰ So, the existence of what has been described before as a genuine shamanic province in northernmost Europe definitely left its distinctive mark on the pattern of witch trials there; but even within this region, it did this as a side-show.

What of the wider ‘sub-shamanic’ zone suggested earlier as embracing much of medieval Fenno-Scandinavia? In Finland its influence on the early modern trials again seems apparent, because there they also bucked the European trend. Overall, about half of those accused of magic-related offences were male, and men formed a clear majority until the late seventeenth century, and male defendants were especially numerous among the native population. The association of witchcraft both with devil-worship and with women was stronger in areas of Swedish influence and settlement, and took a long time to gain purchase on the Finns. This matches the native association of magic mainly with a male practitioner, the tietäjä, even though, as across most of Europe, service magicians themselves were rarely accused of destructive acts.²¹ As among the Sámi, shamanism itself has a muted presence in the Finnish court records. One expert has concluded that ‘the shamanistic witch culture appears not to have played any major role in witch trials during the early modern era’; another, that ‘my sources do not point towards shamanistic remnants, let alone practicing shamans. These materials mention no trance or describe no magical travel to this or another world in any detail.’²² None the less, the impact of native tradition on the patterning and gendering of trials seems clear.

The same may be true of Estonia, to the south in the Baltic ethnic zone, where a similar situation is recorded: the concept of satanic witchcraft was mostly imposed following the conquest of the area by Sweden, and 60 per cent of those accused of harmful magic were male, though slightly more women were executed.²³ One study has emphasized the rooting of local beliefs in a persisting popular paganism, comprising worship at sacred groves and stones, especially on Midsummer Eve, in almost every south Estonian parish.²⁴ It would not, on the face of things, be strange to find such a survival here, on the eastern edge of the European world, where the last state on the Continent to convert formally to Christianity had been neighbouring Lithuania, in the late Middle Ages. On closer inspection, however, the single example provided of such practices is less convincing. It portrays a continuing devotion to St John the Baptist, whose feast was on Midsummer Day, and the stone at the centre of this was an altar to him. The rite was a curing of the sick, with prayers to the saint, and the ‘sacrifices’ were the bandages of those who thought themselves healed, and offerings of wax, of the sort familiar at medieval shrines. Other such rites were also conducted on saints’ days. There are also records of peasants counselling their fellows not to attend church, and to blame their misfortunes on doing so. The author presents this as pagan resistance to Christianity, but the context is that of the first attempt by the Swedish authorities to survey the results of imposing Lutheran Protestantism as the new official religion on a native Baltic populace that had been accustomed to medieval Catholicism. It is therefore not clear that what was going on was a clash between Christianity and paganism, rather than between different kinds of Christianity.²⁵ The early modern Estonian records do, however, connect with a very convincing and well-recorded folk belief system among the Baltic peoples, which may well have pagan roots: twenty-one people (out of a total of 205 in surviving witch trials records) were accused of killing livestock while in the shape of wolves.²⁶ To early modern European demonologists, Livonia, the territory of the Liv people, which extended across much of modern Estonia and Latvia, was the land of werewolves par excellence.²⁷ They tended to tell two stories about them: that once a year at midwinter all Livonian werewolves held a great assembly, or series of them, and that they were locally supposed to be the great foes of witches, and protect communities from them. One included a report of an encounter with a man at Riga who claimed to be one and to send out his spirit in wolf form to fight a witch disguised as a butterfly, while lying entranced. In a now famous trial held in what has become Latvia in 1692, an old service magician also asserted that he was a werewolf, and battled with witches and demons in hell three times a year alongside his fellows, in wolf form, for the fertility of the farmlands and in the name of the Christian God.²⁸ This account caused Carlo Ginzburg to make an understandable connection with the benandanti, and join the two together as survivals in different places of an ancient shamanistic cult extending across Central Europe. His hypothesis remains, as said before, possible but unprovable: it may well be that the Livonian tradition had a completely independent point of origin from the Balkan one.²⁹ What is interesting in the present context is the proven existence of a belief among the Balts that special people possessed a magic that could transform their spirits into wolves, to fight those of witches for the common good. This left its imprint upon local witch trials (though again in a minority of cases). It may help explain the high proportion of men among the accused in general, if men were prominent among such magicians – although most of those actually denounced for being destructive werewolves were female – and may also (perhaps) extend the ‘sub-shamanic’ province of Northern Europe from the northern to the eastern side of the Baltic Sea.³⁰

In a previous chapter, the argument was made that the effects of shamanism could be found in the magical beliefs of medieval Scandinavians in general, but none of those seems to be present in the witch trials of the bulk of the peninsula. Elements of older belief systems not necessarily related to shamanism do feature at times. The idea of the witches’ sabbath was probably digested more easily because of the pre-existing concept of women, human or supernatural, who flew by night.³¹ As said before, magical knowledge, including witchcraft, had been associated with trolls. By the early modern period, educated Scandinavians no longer believed in trolls, or cared much about people who tried to commune with them, but in popular testimony in Norwegian witch trials, demons sometimes took troll form. Furthermore, two service magicians prosecuted in 1689 claimed to have gained their powers from ‘earth trolls’.³² Nature spirits make slightly more frequent appearances in Swedish cases, such as that of the service magician at Söderkoping in 1640 who confessed that he had enjoyed years of sex with a being shaped like a beautiful woman with a foal’s tail who came to him in his boat or in the woods and gave him good luck in fishing and hunting. In a neighbouring district another man claimed to get his own luck in hunting from a forest nymph, who opposed an old, ugly and black being which tried to prevent him from killing the animals. Three other men testified in later decades to having had similar relationships with wood or mountain spirits in the form of lovely women (though two of these had shaggy legs). Vivid and important though these examples are, they represent just five out of many hundreds of trials for witchcraft and magic in early modern Sweden.³³ More often pagan elements in early modern Scandinavian magic seem to exist in the eye of the modern beholder. One historian has described Danish love spells as ‘a mix of pagan and Christian symbols and rituals’; but while her Christian example is an invocation of St Thomas Aquinas, her pagan one is the use of chickens’ eggs.³⁴ In such contexts, ‘pagan’ seems to be shorthand for ‘natural’ or ‘secular’.

There is, however, one very striking form in which a definite element of Norse paganism survived into the early modern period, and that was in ceremonial magic. Just as elsewhere that magic had preserved the names of Egyptian deities as powerful spirits, so the most famed of Scandinavian gods continued to be associated with magical workings, although as devils. It seemed that in the north the Christian tactic of demonizing the divinities of older religions had worked with particular effect. Those divinities certainly remained known to educated Scandinavians throughout the Christian period, as characters in myth (much as Greek and Roman deities were throughout Europe and former Irish deities in Gaelic lands). As demons, however, they, and especially their leader, Oðinn, retained a supposedly ‘real’ presence. A late fourteenth-century Norse rune stick invokes Oðinn as ‘greatest among devils’, as well as calling on Christian powers. In 1484 a man tried for theft in Stockholm confessed to having ‘served Oðinn’ for seven years. Nine years later another thief was executed, for having dedicated himself in a cemetery to ‘the devil Oðinn’ to get rich, and a text from the late 1530s stated that people who suddenly became mysteriously wealthy were suspected of having made a pact with Oðinn. Another Swedish case, from 1632, involved advice to find wealth by going to a crossroads at night to make exactly such a pact, with Oðinn as a devil. A trial in 1693 said that he came to those who invited him with black servants, dogs and coach horses, the latter having flaming eyes.³⁵ From Iceland comes a seventeenth-century book of magic which contains a curse in the names of Lord God the Creator (repeatedly addressed), Christ (‘Saviour’), Oðinn, Thor, Frey, Freya, Satan, Beelzebub and spirits with unknown names: the powers of heaven and hell are thus indiscriminately enlisted.³⁶ All this provides a spectacular example of how ancient gods could be fully assimilated into Christian mythology, though they do not seem to feature in the witch trials themselves.

The crossing has now been made to Iceland, the medieval literature of which provided some of the best material for the argument, made earlier, for a hybrid Norse magical culture that incorporated elements of shamanism. It may readily be expected that this would influence the nature of early modern Icelandic trials; and indeed this has been claimed. Of the 120 people tried for witchcraft in Iceland, only ten were female, and only one woman was among the twenty-two individuals who were burned. One of the first scholars to write of this in English related it directly to medieval tradition, and through that (vaguely) to paganism and shamanism.³⁷ Things are, however, not quite that simple: in the medieval Icelandic literature, destructive magic is as much a female as a male practice, and the most apparently shamanic aspects of magical technique, such as seidr, are much more female. The key to explaining the Icelandic anomaly lies instead in a broader European phenomenon: that ceremonial magic was, throughout the medieval and early modern periods, essentially the preserve of men. In seventeenth-century Danish law, which was enforced in Norway and Iceland, the possession of books of magic was a criminal offence, and the Icelanders took that very seriously. The people charged with magic in their island tended to be possessors of such books, and therefore male, and those executed were the men who were proved to own them and believed to have used them to harm others.³⁸ The concept of a diabolic witch conspiracy, with assemblies and rites, was almost wholly missing. There may however still be an informing link with the ancient past, and that lies in the heavy stress placed in medieval Icelandic texts on the importance of written words and characters – runes – in working magic. It seems to have been the combination of that with the importation of book-centred ceremonial magic that produced the peculiar nature, and gendering, of the Icelandic trials.

Another loose end that could usefully be tied up in this context concerns Russia, and the neighbouring territory of the Ukraine which became part of the Russian state in the seventeenth century. That state bordered on Siberia itself, homeland of classical shamanism, contained large numbers of Sámi in its northernmost parts and linked the ‘dream warrior’ zone of the Balkans and Hungary to the ‘northern shamanic zone’ of the Baltic and Fenno-Scandinavian regions. It should therefore logically play a pivotal role in the mapping of ancient and medieval traditions onto early modern belief systems, as expressed in the witch trials. At first sight the trial evidence supports such an expectation, because in the mainstream European context Russia was anomalous as so much of the Fenno-Baltic and Sámi areas were: it had relatively few witch trials and even fewer executions for a state of its size – around five hundred trials in the seventeenth century with a 15 per cent death rate – and three-quarters of those accused were male.³⁹ On closer inspection, however, a relationship between these results and ancient tradition seems to be missing. In the words of a prominent Western expert, ‘Neither shamanism nor paganism makes the least detectable appearance in Muscovite [early modern Russian] magical practices.’⁴⁰ Once again, the eye of the beholder may be partly at work here, as the same historian summed up the spells which feature in Russian witch trials as mostly poetic nature imagery glossed with Christianity.⁴¹ An older generation of scholars might well have characterized that imagery as pagan, but this would also be an arbitrary and subjective judgement, and it certainly seems as if nothing that can be called shamanism by any definition is present, and as if ancient and folk tradition cannot explain the particular nature of Russian witch trials. Russian folklore collected in the modern period abounds with spirits, of the household and the wild, but they make no more than two appearances in the trials, and few in spell books.⁴²

The reasons for that lie elsewhere, in the peculiar cultural isolation of seventeenth-century Russia, which meant that the new concept of the satanic witch had not got in. The raw materials for it were there, for Russians already had a strong belief in the Devil and lesser demons and thought that humans could make pacts with them; feared witchcraft; and embedded a strong distrust of women within male culture. None the less, these traits were never brought together into the mainstream European idea of a demonic conspiracy, because neither Protestantism nor Catholicism made any inroads into Russia’s Orthodox Christianity, which itself never generated that idea. Russian elites had virtually no contact with foreign cultures. As a result, attitudes to magic simply continued the medieval European norm, individual people being prosecuted for using or attempting to use magical operations to harm others; and indeed throughout the sixteenth and seventeenth centuries this remained a weapon in the factional politics of the imperial court, as it had in those of medieval Europe. The spells alleged to have been used were relatively simple and employed mundane objects and ingredients. As most of the accusations arose from local and personal tensions over power (such as those between social superiors and inferiors), and men were most commonly involved in these, it was natural that men should bear the brunt of them more often. The medieval Russian propensity to turn on presumed witches at times of natural disaster seems to have vanished. Ironically, the early modern European construct of satanic witchcraft was eventually introduced to Russia, in the reforms of Peter the Great, which were intended to bring his nation into conformity with general European norms and are often taken as commencing its ‘modernization’. This was in the early eighteenth century, when witch-hunting had died out in most of the Continent (though not in neighbouring Poland and Hungary). Fortunately, its late introduction was not accompanied by a rise in the religious temperature, which would have ensured large numbers of trials, and so the effect was limited and short-lived.

It may be useful, at this point, to look quickly at the other areas of Europe in which men were usually prominent as victims in the witch trials and see what cultural factors may have underpinned such an outcome there. One of these was western and central France, which remains relatively understudied in this regard.⁴³ An immediately identifiable reason for the high proportion of men among the accused (about half) was the significant presence among them of two groups: churchmen and shepherds. The former may have been vulnerable because of their continued association with learned ceremonial magic, which had been linked to them in the region in trials since the fourteenth century. The latter appear to have been regarded as practising an especially magical occupation in these parts of France, and there is a good study of the prosecution of them as witches in Normandy, where local belief held that shepherds could work harm with toad venom and stolen consecrated hosts.⁴⁴ There seems to be no evidence of the origins and age of this folk tradition. The final area in which men were prominent was most of what is now Austria, and there is even less information as yet available to account for this.⁴⁵ A study of Carinthia, where two-thirds of the accused were male, suggests that this was because witches there were especially associated with bringing bad weather, and weather magic was locally thought of as more of a male interest. This could well reflect a distinctive local tradition with long roots.⁴⁶

On the other hand, the same study also notes that persecution of witches was also linked strongly to a legal drive against begging, and this occurred elsewhere in Austria, and especially in the territories of the archbishop of Salzburg. There, in the same period of the late seventeenth century, a savage hunt was launched against young male beggars in what became one of the last major series of witch trials in German-speaking lands. This was the end product of a sharp reversal of the local tradition of charity, made in response to changing economic circumstances, and the charge that beggars cursed the more fortunate members of society, in jealousy and as vengeance, justified the change.⁴⁷ The hunt for beggar-witches in parts of Austria, which may do much to explain the high profile of men in the trials there, cannot therefore be ascribed to any traditional stereotypes. It seems, rather, a late development propelled by a specific crisis in economic and social relations. All in all, therefore, the witch trials sustain the concept of a ‘shamanic’ and ‘sub-shamanic’ province confined to the far north-east of Europe, which had some impact on the gendering of accusations there, even while archaic elements were rare in the trials themselves.

Bloodsuckers, Wolf-riders and Ladies

In the Alps and Pyrenees, and the lands immediately south of them, a rich crop of folkloric motifs features in prosecutions for witchcraft, one of them certainly very ancient. This is the figure of the strix, the night-flying, child-killing female demon. By the Middle Ages, if not earlier, it was, as said, being merged with that of the human witch, and this composite directly underlay the formation of the early modern stereotype of the satanic witch. It has been shown how it appeared in the first trials which embodied that stereotype, in 1424, and how in the Pyrenees the word for that kind of demoness, bruja, changed into that for a witch; in Italy the same thing happened, as striges, the Latin plural for strix, became the standard learned term for witches, and gave rise to the modern Italian strega, meaning a witch. The persona remained with the name. When an Italian witch-hunter, Gianfrancesco Pico della Mirandola, wrote a book to justify his activities in 1523, he called it simply Strix.⁴⁸ Among the characteristics of the stereotypical witch which he assembled was that of killing babies by pricking them with needles and sucking their blood. His principality of Mirandola was on the northern plain of Italy, near Modena; further south, at Perugia and Siena, women were also tried for this offence in the fifteenth and sixteenth centuries.⁴⁹ In northern Spain, likewise, the initial outbreak of accusations against child-murdering witches in the 1420s was followed by others, over a broader area, in the following hundred years.⁵⁰ In about 1450 a Castilian bishop denounced the new concept of the satanic witch as a fantasy, and especially the belief that these bruxas got through chinks or turned themselves into animals in order to enter houses to suck babies’ blood.⁵¹ This belief underlay accusations of witchcraft in northern Spain until the end of witch trials there: it was the major spur behind the great Basque witch-hunt of 1609–14.⁵²

Across most of early modern Europe, killing infants and young children was one of the most important crimes alleged against witches, and was, as said, fundamental to the development of the new idea of witchcraft; but this vampiric element, derived from the strix, was confined to the northern edge of the Mediterranean basin. It continued eastwards to the limits of Italy, in Friuli, where the witches fought by benandanti were said slowly to consume the flesh or blood of small children, so that they wasted away.⁵³ As ancient Roman rule had done, and as the belief in dream warriors also did, this concept crossed the boundary between Italian and Slavonic linguistic zones, so that in the modern folklore of Serbia, the special crime of witches was to kill babies in this manner. During the witch trials in Croatia, women confessed to eating the hearts of children, and leaving them to die slowly. Croats also believed that witches, in the form of cats, sucked the blood of adults.⁵⁴ The special characteristic of the modern vampire, as a blood-sucker, may indeed have developed from this concept of the witch, as it came to be applied to the restless dead in the eighteenth century.⁵⁵

In a part of the western Alps, a completely different local tradition obtained: that witches rode wolves to go about their nocturnal attacks. This was found in north-west Switzerland, from Basel to Luzern and Konstanz, a region in which wolves represented as great a menace from the natural world as witches did in the human one; so in this sense it was natural that they were twinned. Wolves abounded, however, in other parts of Europe without becoming steeds for witches, so the element of the caprice of local imagination is also at work here. Elsewhere in the regions where wolves and witches were both feared, and associated, the wolves were regarded as being either disguised demons serving or commanding the witches, or the witches themselves, changed by demons into wolf form or given the appearance of one by illusion.⁵⁶ The exception, of course, was Livonia, where some locals at least believed that it was benevolent service magicians who were the werewolves. At any rate, the motif of the wolf-ride occurred in both trials and literary works across this particular expanse of Swiss mountains and valleys, and the roots of it seem lost.⁵⁷

The important medieval belief in the nocturnal travels of the ‘lady’ or ‘ladies’ also played a notable part in the trials, but only in a few locations: the Alps, northern Italy and Sicily. The stereotypical witch portrayed by Pico della Mirandola not only fed on babies but attended ‘the game of the Mistress’ to feast and have sex; in his diabolized view of this event, she offered consecrated hosts to the Mistress, for defilement. Pico’s territory was close to that of Modena, where one woman confessed in 1532 to going to ‘the game of Diana’, where she profaned the Christian cross and danced with demons on the orders of ‘the lady of the game’. Another in 1539 was recorded as saying that she went to a witches’ sabbath over which ‘a certain woman’ presided.⁵⁸ Across the Plain of Lombardy from Modena, at the foot of the Alps, was Brescia, where a woman tried in 1518 said her mistress was a beautiful lady called ‘Signora del Zuogo’ (Lady of the Game), who was served by other human followers and by devils. Up in the mountains of south Tyrol, in the Italian-speaking Val di Fiemme, it is the goddess Venus, or ‘Erodiade’ (Herodias) who features in the confessions taken there in 1504–6. Venus had probably migrated from the German-speaking lands to the north, where the legend of her court in a mountain, the Venusberg, was well established by the end of the fifteenth century: a confession by a man in 1504 directly reflected that legend, by speaking of entering that mountain and finding its most famous inhabitant, the knight Tannhäuser, there, as well as ‘the woman of the good game’ (who was not, apparently, Venus). All were, again, demonized for the trials: Venus was said to travel with a retinue of black horses, and to turn into a snake from the waist down for half of each week, while ‘Erodiade’ was now an ugly black woman in black clothes who travelled on black cats.⁵⁹ The ‘good game’ or ‘good society’, with or without its lady, also featured in trials in Lombardy and the Italian Alps in the late fifteenth and early sixteenth centuries at Como, Mantua, Ferrara, and the Valtellina. Its westernmost occurrence in this context was in the Val di Susa in west Piedmont, and its easternmost involved one of the benandanti of Friuli: it spanned trials for magic in the northern Italian linguistic zone.⁶⁰

North of that, in the German-speaking Alps, nocturnal spirit cavalcades and processions featured much less in the early modern trials, unless described purely as those of witches and demons making for the sabbath. The outstanding exception was Wolfgang Behringer’s ‘Shaman of Oberstdorf’, the service magician condemned for witchcraft who came from Germany’s most southerly village in a remote valley near the border between Bavaria and Austria. In Behringer’s skilful reconstruction of his belief system, he had mixed together mainstream Christian concepts such as angels, heaven and purgatory with a local folk one of the Nachtschar (‘night company’), benevolent night-flying spirits.⁶¹ The only other apparent reference to such phenomena in a judicial process in the northern half of the Alpine zone comes from Interlaken, far to the west, in 1572, when the local Bernese governor reported a woman who claimed to travel with the Nachtvolk (‘night people’).⁶² This paucity of records from trials is the more remarkable in that popular traditions of such spectral night-wanderers were common in this region, as discussed earlier, as were prosecutions for witchcraft. It may be that the lack of any recognized leader for the spirits concerned, north of the Alpine watershed, made them more difficult to assimilate to the stereotypical witches’ assemblies; but the assimilation should still have been easy to make had people wished it. Far south of its northern Italian stronghold, however, the tradition of the roaming nocturnal ladies, with a leading goddess-like figure and selected human adherents, was very much alive, and prominent in trials, during the late sixteenth and early seventeenth centuries. This was in Sicily, where Gustav Henningsen found around seventy cases from that period in which the local inquisitors tried donas di fuera (ladies from outside), the service magicians described earlier, who claimed to have gained their knowledge from the superhuman ‘ladies’ – bearing the same name – with whom they made contact at night.⁶³

One other old, and possibly very ancient, folkloric motif was found in the north Italian trials involving the ‘lady’ and her ‘good game’: a magical rite whereby an animal, normally an ox, which had been consumed in the feast at the ‘game’, was restored to life at the end. This has been extensively studied by Maurizio Bertolotti, and usually involved an alleged process of gathering the bones and hide of the animal together and touching them with a staff or stuffing them with straw. The trick was essentially a deceit, because the animals died permanently soon after or were lastingly enfeebled: it was really portrayed as a means of diverting suspicion from the witches.⁶⁴ It is recorded in confessions of people tried for diabolic magic at Milan, Canavese, the Val di Fiemme, Modena and Bologna between 1390 and 1559, and represents an extension of the convenient medieval belief, commonly found in accounts of visits by spirit hosts to houses, that after they and their human friends feasted there, the food and drink that they consumed was magically replenished, to leave no trace of the theft. It had, however, an independent origin from this more general belief, for it was attested in two successive accounts of the miracles of St Germanus, both dating from the eighth century, in which the saint restores a calf. It is also in Snorri Sturluson’s thirteenth-century stories of the pagan Norse deities, presumably based on older tradition, where it concerns the resurrection by the god Thor of a flock of goats, using his hammer. Bertolotti argued that the saint’s miracle was derived from the story of Thor, and that behind Thor in turn stood a prehistoric hunting myth centred on a superhuman ‘Lord of the Animals’ who caused the prey of hunters to be reborn and so ensured a continuing supply of food.

Wolfgang Behringer has made a further contribution to the study of the motif, by bringing in more miracles attributed to medieval Christian holy men from the Netherlands – St Pharaildis, St Thomas of Cantimpré and Wilhelm Villers – who were all said to have restored animals to life in a similar fashion. He acknowledged that all these stories may have been inspired by the Bible, and specifically by the vision of the valley of dry bones in the Book of Ezekiel, but thought this less likely than Bertolotti’s hypothesis of a pagan hunting myth as the point of origin. In support of this, he cited not just Snorri but Burchard’s condemnation, quoted before, of the enduring popular belief that the cannibal witches of Germanic mythology restored to a brief life the people whom they killed and ate. He also produced ethnographic parallels, of a Caucasian tribe that thought that its god of the hunt revived animals killed by his devotees and the habit of Siberian hunters who left the bones of their kills unbroken to make resurrection possible. He referred to similar beliefs from elsewhere in Asia, and Africa.⁶⁵ All this is entirely plausible, though the absence of the resurrection of an animal from bones and hide in any actual ancient pagan source, Greek or Roman, must give some pause to a conclusive acceptance of it. What the ancient world gives us instead is an idea, expressed most vividly in the Metamorphoses of Apuleius, that witches could magically restore a human being whom they had killed in the night to a brief, but convincing, appearance of full life; this would harmonize with the Germanic tradition quoted by Burchard. As things stand, the European tradition of the use of miracle and magic to resurrect a slaughtered animal from its body parts is a medieval one; and its most obvious function in the witch trials is to represent a fantasy whereby relatively poor people could enjoy huge free meals of beef without payment or retribution. This sense of privilege and gratification lies, as has been suggested before, at the heart of the medieval beliefs in spirit hosts of the night who swept up favoured humans.

Mediterranean Mildness

As was said earlier, the great majority of the early modern executions for witchcraft occurred between 1560 and 1640. This was also the period in which the regional inquisitions that defended the purity of the Roman Catholic religion in the Western Mediterranean basin, and which represented some of the most formidably efficient investigative and punitive machines in Europe, launched a determined attack on magical practices of all kinds. The results, however, have come to be recognized as remarkably mild: several thousand prosecutions for magic yielded at the very most five hundred death sentences.⁶⁶ This was because of a general lack of a sense of danger from a satanic conspiracy, so that charges of collective devil-worship, and of pacts with Satan, were very rare. Torture was seldom used on those arrested, and there was little pressure on them to name accomplices: on the whole, witches were treated as ignorant folk deluded by the Devil, not as dangerous criminals.⁶⁷ At Venice the inquisitors held over six hundred trials concerning magic between 1550 and 1650, about a fifth of which were for witchcraft, but most ended in acquittal and none in execution.⁶⁸ Similarly, no executions are recorded in Sicily, and the notorious Spanish Inquisition managed to try more than five thousand people for using magic between 1610 and 1700, without burning any.⁶⁹ The Portuguese one put one person to death for such an offence, although it regularly tried cases that concerned magic and sometimes prosecution of them rose to peaks.⁷⁰ The inquisitors in Malta not only regularly prosecuted people for using magic but held two mass trials in the seventeenth century, one involving forty women, and yet passed no death sentences.⁷¹When Louise Nyholm Kallestrup compared the sentences passed by seventeenth-century courts for acts of magic in Denmark and the district of Orbetello in the Papal States of Italy, she found that the lightest sentence in the secular Danish system more or less matched the harshest pronounced by the inquisitors at Orbetello.⁷²

Such a pattern calls out for explanation, and at first sight the presence of deep-rooted popular traditions in the Mediterranean lands concerned, which worked against savage witch-hunting, seems a plausible potential answer. Such a factor has in fact been argued for Spain by Gunnar Knutsen, who has drawn attention to the hundreds of executions recorded in the northernmost provinces in the decades around 1600, mostly by secular courts, and the absence of them further south. In a study of Catalonia and Valencia, he has contrasted the situation in the former, where the idea of demonic witchcraft took root readily among a traditionally Christian rural population in close contact with French culture, with that in the latter. There the peasantry consisted largely of recently converted Muslims, who retained from Islam a belief that magicians should be able to control demons rather than become their servants, together with a weak fear of witchcraft and a limited concept of Satan. They transmitted these notions to at least some of their Christian neighbours.⁷³ There must be truth in this picture: after all, it has been shown how existing folkloric concepts enabled the new stereotype of the satanic witch to root itself very early in the Spanish Pyrenees, and spread through that region. There was no Islamic influence in mainland Italy, but in much of the peninsula (as in much of Spain) there was a lively belief in the power of the ‘evil eye’ to blight unintentionally and as a force of nature rather than of the Devil, of the sort which has been identified as dampening down a fear of witches elsewhere in the world.⁷⁴

It is very likely that such cultural factors did act to prevent a ready reception of the image of the demonic witch in much of the Catholic Mediterranean world; and more may be uncovered by local studies.⁷⁵ It is also, however, apparent that in themselves those factors are insufficient to explain the relative absence of witch-hunting in that world during the period when it was most prevalent elsewhere in Europe. After all, northern Spain and Italy had been cradles of the new image of witchcraft and settings for many of the early trials for it, and northern Italy had produced some notable early proponents of the need to hunt witches. The bishop of Brescia burned sixty people in 1510 and sixty-four in 1518.⁷⁶ The Italian Alps and the Plain of Lombardy were indeed probably the area of most regular and lethal witch-hunting during the first century in which the new stereotype of the witch was in existence. Even in the southern Spanish region of La Mancha, six people were executed for offences related to magic between 1491 and 1510.⁷⁷ Moreover, there is plenty of evidence that the concept of the demonic witch was spreading far across the Mediterranean basin, and that there was sufficient public belief in it to create the conditions for savage witch-hunting had the authorities been inclined to that. At Venice, people regularly confessed to making pacts with demons to obtain their desires, and the witches’ sabbath featured in six confessions from rural areas of the republic. Crowds called for women to be burned as witches, and churchmen read the works of demonologists who advocated witch-hunting.⁷⁸ Novara, at the extreme northern end of Piedmont, and the extreme north-western end of Italy, was exactly the sort of Alpine environment that had fostered the new image of satanic witchcraft; yet in 1609–11, at the height of the trials in Northern Europe, the episcopal inquisition there prosecuted eleven people who confessed to full participation in the sabbath, and all escaped with prison sentences.⁷⁹ In Tuscany in 1594 a friar drunk on demonology tortured a midwife into admitting that she worshipped Satan at the sabbath and murdered children on his instructions.⁸⁰ In the Otranto region of south-eastern Italy, it is plain from the legal records that fear and hatred of witchcraft were much stronger among the general populace than among churchmen.⁸¹ In Valencia in 1588 the inquisitors were faced with a teenage girl who claimed to have had sex with the Devil, and in the following century with a woman accused of flying into houses to bewitch the inhabitants, and a man offering his services as a witch-finder.⁸² Sicily produced a woman in 1587 who claimed to ride with others through the air on billy-goats to worship a royal couple of spirits who presided over a feast and orgy.⁸³ Malta contained people who confessed to invoking Satan, and the early modern Portuguese frequently talked of demonic pacts, and occasionally of night flights by witches and of the sabbath.⁸⁴

Clearly, then, popular belief throughout Italy and Iberia, and their attendant islands, could have assimilated the new model of witchcraft, and counter-balancing cultural traits could only have slowed down that assimilation. Another factor must have been at work, and Gunnar Knutsen spotted it in his comparison between Catalonia and Valencia: in the former, the Spanish Inquisition was much weaker than in the latter, and less able to restrain the lay magistrates who responded more avidly to a public fear of witchcraft. Such restraint was the crucial determinant of result: in Valencia the girl who confessed to having sex with Satan was sentenced to religious instruction (after a beating), the woman accused of night flight was acquitted, and the would-be witch-finder was punished. In all the other cases of alleged diabolism cited above, similar moderation was observed and the diabolic elements played down by the inquisitors: the woman who was forced to confess to demonic witchcraft by the Italian friar was then released on the orders of his superior. General studies of the history of the respective inquisitions reveal a gradual formulation of central policy that made such outcomes at first possible and then mandatory.

In 1542 a central tribunal was established in Rome to oversee local Italian inquisitions, and by the 1580s this was advising caution in local trials of witches and enforcing it on some. In 1575 Pope Gregory XIII ruled that nobody could be arrested simply because of a denunciation by somebody already under trial for witchcraft, while in 1594 Pope Clement VIII banished a southern Italian bishop whom he thought had prosecuted it too recklessly. Around 1600 the tribunal accepted a protocol which was sent out to most Italian inquisitors from the 1610s: all alleged deaths from witchcraft were to be investigated by medical experts operating under oath; suspects were to be held in different cells to prevent them from mutually reinforcing their fantasies; and investigators were to avoid any leading questions, identify local hatreds operating in accusations, and consider only objective evidence. This made convictions for witchcraft practically impossible, and after 1630 papal authority effectively ended witch trials in the Italian peninsula.⁸⁵

A parallel process occurred in Spain, where from 1525 the Supreme Council of the national Inquisition began to reduce death sentences imposed on suspected witches by its local representatives, accusing the latter of excessive credulity and use of torture to force confessions. In 1526 it anticipated the later papal decree by decades, ordering that nobody should be arrested simply on the testimony of somebody already accused of witchcraft. It also sought to take over itself the cases of those formally charged with witchcraft who pleaded innocence. The last execution of somebody for witchcraft by a member of the Inquisition in Aragon was in 1535, and the last in Catalonia in 1548. Trials persisted in Navarre, and in 1609 a serious witch-hunt on the French side of the border spilled over into that province, and produced a major panic, with almost two thousand accusations. The first inquisitors to deal with them were persuaded of the reality of some, and burned six people. Subsequently, however, the Supreme Council sent out a more scrupulous representative, Alonso de Salazar y Frias, who became convinced of the patent falsity of most confessions, and the impossibility of clear proof in the case of the remainder. His report convinced his superiors, who were also shocked by the expense of the investigation. Thereafter they adopted a code of rules for trials of alleged satanic witches which demanded such stringent proofs that it rendered convictions virtually impossible. Witch-hunting was now confined to those parts of north-eastern Spain, especially Catalonia, where the authority of the Inquisition was weakest and trials could be conducted by secular courts with relative freedom. Even there, however, the inquisitors did their best to halt the proceedings, reinforced by royal authority from 1620, and by the end of the 1620s Spain was apparently free of trials for diabolic witchcraft.⁸⁶

The great influence of the Papacy and the Spanish upon the western Mediterranean lands in general explains why the other territories in the region followed the same trajectory in the same period. The new oversight and professionalism injected into the inquisitorial process by the foundation of central supervisory bodies seems in itself to have engendered a more rigorous and sceptical attitude towards accusations of demonic witchcraft, and a growing disposition to view even those people who confessed to dealings with the Devil as deluded and in need of redemption. This change then became a factor in regional power politics, as interventions to prevent credulous and destructive witch-hunting enabled the central tribunals to enforce their authority more effectively over the localities. Eventually a cautious attitude to accusations of witchcraft, and a programme of correction and not extermination for those convicted of attempting to work magic, became a matter of ethnic identity. Seventeenth-century Italians, in particular, could be surprised and horrified by the huge body counts being stacked up by witch-hunts in Northern Europe.⁸⁷The Mediterranean inquisitions remained forbiddingly effective machines for the persecution of magical practices, and even moderate punishments such as imprisonment, flogging and public penance would have been traumatic for those who suffered them. None the less, they rescued a region representing about a quarter of Europe from the most concentrated and deadly period of the early modern witch trials. They seem to have done so, moreover, because of political and ideological developments among the religious elite, in which popular beliefs played only a supporting role, in certain places, and not a decisive one.

The Silent Centre

What, however, of the core area of the early modern witch trials, where the majority of their victims perished: the German-speaking lands, the French-speaking parts of the Rhine and Moselle basins to their west and Poland to their east? It has already been noted that these had sprouted rich medieval popular traditions, such as those of the ‘furious army’, Holle and Perchte, which should have meshed easily with the concept of the witches’ sabbath. Modern folklorists, led by Jacob Grimm, uncovered a still flourishing lore of nocturnal spirits of this sort, with strong regional hallmarks. All this testifies to a prolific set of beliefs, grounded in the culture of ordinary people, which should have informed the nature of witch trials in the way in which striges, wolf-riders, superhuman ladies and dream warriors did further south; and yet most of the evidence suggests that it did not.⁸⁸

In saying this, it is important once again not to forget the deeper perspectives. The image of the satanic witch that was transmitted to Northern Europe was based partly on an ancient concept, that of the strix, and the facility with which the Germanic cultural zone picked it up may well have owed much to its own ancient native tradition, of the cannibal witch who attacked all age groups rather than specifically children.⁸⁹ This tradition might also help to explain why the majority of those accused in this region were women. Moreover, the basic concept of witchcraft was itself ancient, as was the spectrum of magic-working which extended from witches to service magicians, and which many early modern people often saw as more of a polarity between the two. The belief that bewitchment could be cured if the witch agreed to recall it was so common, widespread and ingrained that it must also have been very old. It can, moreover, be readily suggested that pre-existing beliefs in spirits that flew or rode by night would make that of the witches’ sabbath easier to adopt. The basic narrative of temptation by a devil (or the Devil) with which most confessions to satanic witchcraft were supposed to commence must have drawn on the common and widespread folk-tale motif of friendly spirits who encounter distressed human beings, usually in a place out of doors, and become their helpers. It is not surprising in view of this that the names given to attendant demons in confessions of witchcraft made in the northern part of Continental Europe were sometimes those attributed to fairies or equivalent beings.⁹⁰ Beyond these historical truisms, however fundamental and important, there is not a lot to record. The offences credited to witches varied slightly between regions: for example, they were commonly charged with raising destructive storms in the Alps, southern Germany and southern France, with sending wolves to kill livestock in Lorraine, and with stealing milk in Scandinavia, Poland and much of northern Germany. These distinctions may well rest on much older traditions, but had probably in addition a functional aspect, reflecting the nature of the local economy. Other specific regional or national characteristics attributed to witchcraft could likewise rest on old folkloric motifs: Danish witches were thought more likely to cause illness than death, and those imagined in northern France and the southern Netherlands were especially given to inflicting impotence.⁹¹ Whilst the demonic pact was central to witch trials in most parts of this heartland, the concept of witches’ assemblies – the sabbath – was rarer in some than others, and it is not clear whether this was because of predispositions of belief based on local notions of the supernatural, or accidents in the importation of the new idea of the witch. The same is true in the portraits of witches’ activities. In the German-speaking districts of Lorraine, they were believed to meet in groups of varying size and attack other humans collectively; while in the French-speaking parts they met in standard-sized assemblies and operated as individuals. It is certainly tempting to see ancient cultural differences behind these variations, but impossible to prove them.⁹²

Specific folk motifs are rarely easier to detect in the trials across the region. Poland had folk traditions of a more harmless and playful Devil than the one generally imagined elsewhere; and these may have been influenced by pre-Christian beliefs in wood and water spirits, and household spirits which could be placated with gifts. Traces of playful demons do appear in the Polish trials, but the historian who has noted them has also pointed out that the relationship between ancient pagan and early modern popular beliefs in nature spirits, and between both and images of devils, remains speculative.⁹³ German trials sometimes threw up folkloric images. In one at Rottenburg, in the south-west, a man was accused of appearing at the sabbath as a mounted hunter, a ghostly figure from local lore; and across southern Germany some of the misdeeds alleged against witches were those associated with malevolent local spirits.⁹⁴ When the villagers of Gebsattel in central Germany asserted in a case in 1627 that witches were especially abroad on Walpurgis Night (30 April), they were echoing a tradition found across Northern Europe that attributed uncanny qualities to this date.⁹⁵ Such details are, however, both relatively rare and incidental. When Edward Bever considered the records of trials in south-western Germany, he acknowledged that the region abounded with traditions of a parallel spirit world, operating largely independently of the orthodox Christian one, in which some people could participate; but this made remarkably little appearance in the actual cases he studied.⁹⁶

A major example of the way in which people in the European heartland constructed stories about satanic witchcraft is provided by their answers to the question of how witches travelled to the sabbath. Presented with it, usually under interrogation and often under torture, individuals came up with a variety of answers, partly reflecting local belief, but also what they could imagine or invent on the spot in response to a Europe-wide stereotype being articulated by the prosecutors. As a result, those methods presented in demonologies multiplied with time. In the Malleus maleficarum of 1486, the same idea as that recorded in the earliest Alpine witch trials was retained: that witches rode on a piece of wood greased with an ointment made partly from human baby fat.⁹⁷ By the time that Jean Bodin wrote, almost a century later and basing his information on trials in southern France and Italy, the ideas had elaborated. Some people were now said to apply the ointment to their own bodies, and then fly, while others, with or without using the grease, rode animals of different kinds, or a broom or a pole.⁹⁸ Older notions were thus surfacing, as the use of an unguent on one’s own body was attributed to the Roman witches in the fictions of Apuleius and Lucian (or whomever was writing in his style), while the hosts of Diana had ridden on beasts in the canon Episcopi. Shortly after Bodin, Nicholas Remy recorded that people he had tried for witchcraft in Lorraine in the mid-1580s had confessed to flying up the chimney to the sabbath, or anointing themselves and putting one foot in a basket, or putting one on an anointed broomstick. Others rode a wicker net or reeds after speaking a spell, or on a pig, bull, black dog or forked stick; or just walked.⁹⁹ On the far side of Central Europe, in the records of Polish trials, the tales told were equally varied: one woman claimed to fly up her chimney, another to ride a normal carriage, another a horse, and a fourth a bewitched labouring man, while a fifth flew after smearing on ointment.¹⁰⁰German records show the same pattern.¹⁰¹ Some peoples had a more restrictive view of the options: in Swedish trials witches were just said to ride either animals or bewitched humans.¹⁰² In 1612 Pierre de Lancre tried to rationalize the bewildering range of testimony available from his own experience as a trial judge. He decided that some people only attended the sabbath in dreams or thoughts, while their bodies stayed in bed. Those who went in physical form did so by walking, or by use of the baby-fat ointment, on their own bodies or on staffs, brooms or animals, which gave the apparent power of flight to any of these – though he himself concluded that this apparent power was always a devilish illusion.¹⁰³

It is difficult amid all this to find any distinctive local formulations of such an important aspect of the construct of satanic witchcraft. What is striking instead is the propagation across Europe of what became a remarkably standard range of options, from which people selected according to local or individual choice. While the options originally arose as a result of particular trials, and some drew on ancient ideas, their propagation was the work of the elites who introduced the construct of the sabbath into area after area.

This conclusion may stand with respect to the general part played by specific folkloric and ancient motifs in witch trials in the central zone of Europe in which most of those trials occurred. The mass of recent research suggests not only that such motifs played an occasional, incidental and marginal role, but that the opposite phenomenon was immensely powerful: the newly developed stereotype of satanic witchcraft developed by late medieval preachers and inquisitors made a considerable impact on the popular imagination, once introduced into an area. To be sure it did so slowly, patchily and with some features emphasized or adopted more in particular places than others, but it none the less became very widely accepted and understood by the people who feature in trials as accusers and accused; indeed, trials represented an especially vivid means of transmitting it. In many parts of Europe, especially outside the central zone, people were prosecuted for alleged acts of harmful magic alone, without any reference to a pact with Satan or organized assemblies. Nevertheless, it can be strongly argued that the readiness of European elites to allow such prosecutions was itself driven by an enhanced consciousness, and fear, of witchcraft produced by the stereotype of a satanic religion.

In this context it is worth asking how much any aspects of such a religion were acted out by any of the people subsequently charged with witchcraft: did any of them actually try to be satanic witches? This question of the ‘reality’ of witchcraft was posed in the global context as part of the first chapter of the present book (a passage which readers may wish at this point to revisit), and it was suggested there that it is very hard to reach any firm conclusions with respect to it even in contexts in which living people could be interviewed by scholars. It is even harder to do so when testimony is refracted through old written texts. This problem is summed up by two statements from recent experts. One is Robin Briggs, who has declared that ‘historical European witchcraft is quite simply a fiction’; the other is Brian Levack, who stated that it ‘has a solid basis in reality, in that certain individuals in virtually all societies do in fact practise harmful or evil magic’.¹⁰⁴ Both are in fact complementary, because they refer to different phenomena. To Briggs witchcraft represented the belief that people made pacts with the Devil to enable them to work genuine magical harm on other humans, and met in assemblies to worship him and engage in murderous and disgusting activities. Levack was speaking solely of attempts to hurt others by means of magic. Both, however, require further interrogation.

There is a great deal of evidence in favour of Levack’s dictum. As he pointed out himself, curse tablets and images stuck with pins are solid evidence of ancient attempts to harm or coerce others, while medieval and early modern books of ceremonial magic contain destructive spells. The court records of early modern Europe are full of proven cases of individuals who attempted to damage or kill by physical means, and who were heard to utter curses against others. It seems unthinkable that some of them would not have used spells aggressively if they believed that they would work. The problem is that of proving it in any individual case. This was the reason why the inquisitors in the republic of Venice never convicted anybody for the specific crime of magical harm: even in cases where material evidence – of suspicious objects like bones, feathers and inscriptions – was produced from the homes of alleged victims, it could all either have been planted or got there by innocent means.¹⁰⁵ Where such trained professionals could not find solutions on the spot, historians cannot hope to do better. The matter ends in a paradox whereby there is virtual certainty in principle that people attempted to work witchcraft across early modern Europe, but no apparent way of demonstrating it conclusively in the case of any named individuals. A similar problem attends Edward Bever’s bold attempt to extend to Europe the insights gained by doctors working in the developing world and discussed in the first chapter of the present book: that somebody who believed with utter certainty that they had been bewitched could fall ill as a result and even die. He has assembled a range of more recent medical and psychological insights to show the manner in which such fear can weaken the immune system and put pressure on vulnerable organs, of both humans and livestock: in this sense, in Europe as elsewhere, witchcraft could ‘work’.¹⁰⁶ At this distance in time, however, it is medically impossible to prove that this actually happened in the case of any of the alleged early modern victims of witches, let alone that any of those accused actually performed the actions needed to create such an effect. Once again a reasonable presumption cannot be grounded in conclusive evidence.

A rather different, but equally considerable, problem attends the diabolic elements of early modern witchcraft, and the whole mental assemblage of the satanic witch cult developed in the fifteenth century. Here in a sense there is no bridging an enormous conceptual gulf, because modern historians completely reject the literal reality of that cult, however much they may attempt to understand and explain belief in it. In this they are simply following the path laid out by early modern scholarly opinion itself, which came, first in practice and then in theory, to abandon that belief. After all, medieval Europeans did not have it either, until the fifteenth century, and in long historical perspective it was a relatively short-lived phenomenon. Its abandonment does, however, mean that there is a point at which every historian of it simply chooses to disbelieve the testimony of those who held it, as an arbitrary decision: there is, after all, ample apparent first-hand evidence on record that people worked witchcraft in partnership with demons whom they worshipped, and no objective means of proving conclusively that all of it is false. We simply decide to reject it as anything other than a fiction or a metaphor. This is an area of enquiry in which no academic investigator ever goes native, as there are no known cases of a professional scholar of the early modern witch-hunts coming to believe in the ideology that underpinned them.¹⁰⁷ Nor are there any agnostics: no academic historian overtly gives the benefit of the doubt to the idea that Satan might have been active in early modern Europe in the ways described by so many alleged confessing witches. We all choose not to believe, because of the grim record of the results of believing.

That still leaves open the question whether there were any actual would-be satanic witches in early modern Europe. In other words, once the demons are out of the picture, except as fictional characters, were there people who assembled to worship them and committed the actions involved in that worship, as described in accounts of the sabbath? Here again, despite so much apparent testimony, all the professional research of the past half century seems to unite behind the conclusion that there were not, and that all the witches’ assemblies described in the records were illusory. Furthermore, as described earlier, there is an equal consensus that accounts of those assemblies were not mistaken or distorted portraits of some other religious tradition, such as a pagan one: they just never happened. It seems therefore that in the case of the attempted use of witchcraft by early modern people we have a strong presumption that something happened without quite being able to prove that it did, while in that of the satanic witch religion, we have ample evidence for the existence of something, which we disregard on the grounds that it is incredible. Using the logic that was applied to the attempted use of witchcraft itself, it is easy to believe that some individuals, in moments of despair and pain, prayed to the Devil, or devils, for aid against their enemies or persecutors, and some may have offered him a pact to gain it. In a post-Reformation era in which large numbers of people changed their allegiance from one form of Christianity to another, stigmatized by their former denomination as a satanic parody, this may not have been such a difficult step for them to take. Furthermore, there is solid evidence that during the period after the witch trials, in which the offence was less lethal, individuals did draw up pacts which they hoped to make with Satan in an effort to obtain their worldly desires: the evidence consists of the written pacts themselves, to which the writers (mostly urban men and soldiers) admitted.¹⁰⁸ The problem is that when most demonic pacts are described in witch-trial records and demonological literature, they are attended by elements that the modern age regards as fantastic; and any attempt to reconstruct what actually happened must be both speculative and depend on an arbitrary and subjective reordering of the source material. All these factors converge on the final problem of how those fantastic elements arrived in the testimony in the first place. In very many cases they were clearly induced by torture, confinement in appalling physical and emotional conditions, browbeating and brainwashing.¹⁰⁹ In others, however, they were not. When the sceptical inquisitor Salazar arrived in Navarre determined to get to the bottom of the panic over witchcraft there, his problem was not in discerning the untruths told by the accusers, but in discerning those told by the accused, who were providing detailed confessions in huge quantity. Only the most patient analysis succeeded in revealing their contradictions and inconsistencies, so that he could report back flatly to his superiors that ‘the witches are not to be believed’.¹¹⁰ Vivid dreaming, trance states, hallucinations, schizophrenia, false memory syndrome and Stockholm syndrome, and a prominence among those making voluntary and detailed confessions of children and young adolescents, may in sum account for this phenomenon; but ultimately they may not. Gustav Henningsen, studying the confessions of the Sicilian women who claimed to interact with ‘ladies from outside’, decided that dreaming could not account for the manner in which these people told of travelling in groups together three nights a week. In his opinion, such regular and purposeful imaginary activity could only be explained by some kind of technique that induced an entranced sleep that enabled collective mental experiences using telepathic communication. He wondered if experiences of the witches’ sabbath had been achieved by the same means.¹¹¹ At that moment, he had stepped over the boundaries of current scientific knowledge; and this is the territory into which the study of early modern witchcraft may ultimately lead us. Such a step would depend, however, on certainty that the Sicilian women concerned were not exaggerating the regularity and coherence of dream experiences as part of defending their reputations as service magicians empowered by good spirits. It is hard to see how to achieve such certainty.

All this, moreover, still leaves a broader problem of whether any sort of group activity, involving any kind of magical rites, lay behind accusations and confessions of satanic witchcraft. Nobody has tried harder to find some in the Continental sources than Edward Bever, working on the records of trials in the German state of Württemberg. He has shown how the idea of diabolic witchcraft reached the elite there through printed works and the local university, and was then spread to the populace, mostly by the local Lutheran Church. He attributes the content of most accusations and confessions of it to dreams, delirium, psychoactive substances, out of body experiences, false memories, lies, self-hypnosis, errors of perception, personality disorders and other forms of cognitive dissociation. He also rules out the possibility that those accused formed anything like a religious or counter-religious sect, but leaves open one that a few people engaged in collective activity in which they shared ideas about magic, and even initiated others into means of working it. The problem with this, again, is one of proof. His prime witness is a would-be service magician tried for witchcraft, who claimed to have learned his magic in the realm of the goddess Venus, hidden in a mountain in the manner in which it is described in the famous late medieval romance of Tannhäuser, which has already made an appearance in the present book. Bever was impressed by the personal tone of the account, and the way in which it conformed to internationally reported experiences defined as shamanic by many scholars: but he wisely acknowledged that it could have been the result of dreams or an active imagination rather than of a local shamanic tradition, and indeed much of it sounds fantastic.¹¹² What is certain is that magicians did sometimes band together for ad hoc operations in which numbers could count, most obviously the detection of buried treasure; but that is all the certainty that we possess.¹¹³

There is thus a curious paradox in the relationship between the early modern witch trials and ancient and folkloric tradition. On the one hand it has repeatedly been emphasized that the construct of the satanic witch that underlay the trials drew on very ancient images and ideas; indeed, to some extent it let loose the fears associated with these after about six centuries in which they had been damped down by the reluctance of Christian churchmen to take them very seriously. On the other hand, the new construct took a long time to develop and an even longer time to spread widely, and was a thoroughly late medieval one based on orthodox Christian ideas and preoccupations. Moreover, the direct contribution of older motifs and traditions to the actual incidence and nature of the trials was minimal. It was most marked around the periphery of the main zone of the trials, in the far north and south-east of Europe and in a belt running along the southern watershed of the Pyrenees and Alps and in the lowland areas below that. In the heartland of the trials it was very small indeed, despite the existence of a flourishing and ample folklore concerning nocturnal spirit worlds, and even in the peripheral areas in which it featured more strongly, it appeared in only a minority, and usually a small minority, of trials. In a belt of territory running from Finnmark through Finland to the Baltic lands, it probably did affect the gendering of accusations, but in general the trial evidence serves incidentally to expose folk beliefs rather than folk beliefs serving to explain much about the trials. The latter were propelled and dominated, instead, by a new, almost pan-European concept of witchcraft propagated by elites and accepted into general culture. Such a conclusion is, however, based on a general survey of the evidence generated by local studies spanning the Continent. There is still a chance that an investigation of issues relating to the witch trials, focused on a particular region, may employ insights taken from ancient and medieval cultures to explain patterns which other perspectives do not. This sort of close reading may yet throw up new answers to existing problems within the subject, and the last part of this book will consist of a series of them focused on Britain: an island at once furnished with the right sort of records, a rich medley of cultures and an ample tradition of existing scholarship, all readily accessible to the author.