9
IT WAS NOTED in the first chapter of this book that, across the world, there have been instances of peoples who either have not believed in witches or have not feared them much. One of the obvious questions to be asked of the European witch-hunts, in this context, is therefore whether such peoples existed in early modern Europe, and, if they did, whether their presence exerted an influence on the incidence of witch trials there. In another chapter, the possible presence of earlier beliefs that mitigated a fear of witchcraft was considered as an explanation for the striking rarity of lethal witch-hunting across the Mediterranean basin in much of the early modern period. In that case, such beliefs were not accorded decisive importance as an explanatory factor; but the British Isles may represent a more promising hunting ground. When one of the leading recent historians of the early modern witch trials, Robin Briggs, published a map of Europe showing the local incidence of prosecution, it depicted a striking line bisecting the archipelago. To the south and east of it were England, Lowland Scotland and the fringe of the Scottish Highlands, areas that had seen a significant number of trials, and in places a very large number. To the north and west, the map was blank. There lay most of the Scottish Highlands, the Western Isles, the Isle of Man, Wales and Ireland: the regions commonly known as the main Celtic areas of the British Isles.1
The word ‘Celtic’ has already made an appearance in this book, in the last chapter where Carlo Ginzburg was quoted as suggesting that the ‘shamanistic’ cult of nocturnal goddesses, which he identified as underlying the early modern construct of the witches’ sabbath, had been a Celtic tradition. The term will, however, be used here in a different sense from that which he intended, and between the two usages lies a major shift in scholarly conventions. Until the 1990s, the Celts were generally regarded as having been an ancient family of peoples, united by a common language group, culture and art as well as ethnic ties, which had extended across the whole area between Ireland and Asia Minor, and Scotland and northern Italy and Spain, in the last century BC. They were thought to have spread outwards across it from an original homeland in Central Europe. This academic convention underpinned Ginzburg’s putative cult, because it suggested strongly that phenomena recorded in northern Italy and northern Scotland drew on the same ancient cultural roots. It was, however, a relatively recent one, which had developed fully only in the early twentieth century, and at the end of the century it collapsed almost completely among British experts in the Iron Age, joined by some colleagues in other nations. It was now recognized that the convention concerned depended upon using the term ‘Celtic’ simultaneously for a group of peoples, a group of languages and a style of art, and that the three of them did not in fact correspond exactly in ancient times.2
The result has been that most British scholars, and some elsewhere, have abandoned the term when referring to ancient history. Those who still try to find a place for it there have likewise rejected the idea that it refers to a racial or cultural group, and argue instead that it could be applied to a set of languages and values, embraced by different ethnic groups, which spread peacefully from the Atlantic seaboard across much of Europe during late prehistory. The debate over the matter is still continuing, but no party to it would have time for the idea of a single ethnic and cultural ancient Celtic province of the sort which had been envisioned when Carlo Ginzburg wrote. Instead, there is an apparently complete consensus that the term ‘Celtic’ can still be legitimately and precisely applied to a group of languages, and so by extension to the ethnic and cultural identities developed around those languages since the Middle Ages, by the Bretons, Cornish, Welsh and Manx, Irish and Scottish Gaels.3 The last four peoples all appear to have been characterized by a remarkably low level of witch-hunting in the early modern period, and therefore the question is worth posing of whether there was anything inherent in their cultures that predisposed them to such a characteristic.
The Debate over Witch Trials in Celtic Societies
From the 1970s, historians began to notice the apparent scarcity of witch trials in areas with Celtic languages, but most preferred explanations for it that disregarded any common factors. In Ireland it was ascribed to reluctance on the part of the native population, who remained overwhelmingly Catholic, to denounce their own people for crimes that would be tried by a legal system dominated by British conquerors who professed a rival, Protestant, religion. It was suggested that these tensions between the two ethnic and religious groups replaced those within communities of the kind that generated accusations of witchcraft.4 Experts in early modern Wales were also inclined to adopt local, and functional, explanations, that Welsh communities were more cohesive and less fractured by economic tensions than those in England, and so neighbours less likely to accuse each other of being witches.5 It was also noted that they held to a customary law that emphasized compensation of victims rather than punishment of criminals, and so was less likely to lead to the execution of people for witchcraft.6 Historians of the Scottish witch trials tended to deny that there was anything to explain. Some pointed out that a couple of Lowland counties also produced few prosecutions, especially in relation to population, making the Highlands seem less exceptional.7 Another came up with a quite large figure for Highland witch trials, by including in it all those in the fringe of the Highland area where indeed many took place.8 They were especially found in the islands and peninsulas at the mouth of the Firth of Clyde and the coastal region of the Moray Firth and the inlets north of that. It was the Moray Firth coast that had produced Isobel Gowdie. These regions were relatively easy of access from the Lowlands and the north-eastern one contained a string of towns characterized by a hybrid Lowland-Gaelic culture.
Over the same period, however, a different approach was more slowly developing. It was presaged by a passing suggestion by a historian of Gaelic Scotland in 1994, both that its lack of witch-hunting was significant, and that it had been due to cultural differences, Gaels conceiving of supernatural activity in ways distinct from those of Lowlanders.9 In 2002 a further suggestion was made, again without any sustained research or argument to support it, that the apparent absence of trials in both Gaelic Scotland and Gaelic Ireland was due to such differences, Gaels tending to blame fairies for the kind of misfortunes which elsewhere were charged against witches.10 Half a decade later, a leading historian of English witchcraft beliefs turned his attention to the Isle of Man, which possessed the best records for any early modern Gaelic society. He found that the island had produced few trials, but had a strong image of the malignant fairy, and agreed that the two might be related.11 Meanwhile further research was under way in the early modern Welsh criminal records, which are also very rich, and proved that neither community solidarity nor customary law had prevented the Welsh from accusing and executing each other on a grand scale for other offences than witchcraft. In particular, late sixteenth-century Wales had seen a ‘thief hunt’, which had claimed about four thousand lives. Furthermore the Devil had a high profile in the Welsh popular culture of the age; and yet witch trials were few and a demonic element even rarer in them. They were, moreover, mostly found in geographical areas, and historical periods, of greater English influence. It was suggested that this pattern might be explained by cultural factors.12
In 2011, I published an article of my own which attempted to draw these converging strands of enquiry together.13 It concluded that the core area of Scottish Gaeldom – the Central and Western Highlands and the Hebrides, representing about a third of Scotland – had barely participated in the early modern Scottish witch-hunts, which were among the worst in Europe. It provided in fact eight known cases, out of 3,837 identified in the whole nation, and these were often of a special kind, in which witchcraft was a subsidiary offence to another, such as cattle-stealing, or they were launched as part of a deliberate attack on native culture. This pattern was sustained on the Isle of Man, which produced just four cases, two of which (a double trial) ended in execution, after which no more were brought. This is the more noteworthy in that across Northern Europe, islands were exactly the kind of small self-contained communities in which witchcraft accusations flourished. This was true in the Baltic and at the north-eastern extreme of Norway, and also in the Channel Islands of the English Crown, where a Norman culture predominated, and the Orkney and Shetland Isles of Scotland, where a Scandinavian one was dominant. Man however seemed almost as disinclined to them as those other Gaelic islands, the Hebrides. Witch trials seem to have been totally absent among the native Irish, and the rich Welsh source material reveals just thirty-four of them, the majority in regions under some English influence, with just eight convictions and five executions. My article proposed that in the case of Gaelic Scotland, the local spirits of land and water were regarded as being especially ferocious and dangerous, perhaps because of the formidable nature of the terrain, and the same exceptional fear was accorded to the local equivalent of elves and fairies, the sithean. These seem to have been dreaded more acutely and persistently than in Lowland Scotland or England, for committing precisely those attacks upon humans and their animals and homes that were credited elsewhere to witches. Gaelic Scots did also have a belief in witches and some fear of their deeds, especially the use of magic to wreck vessels at sea and steal the profits of dairy farming; but this fear was considerably tempered by other factors. One was that to an extent the Gaels regarded curses and spells as a legitimate means of furthering their own designs and thwarting or punishing enemies. Use of them did incur censure if the action concerned seemed disproportionate or unjust, or was conducted with deception or spite, but this was also true of actions involving physical tools or weapons. There was little sense of witchcraft as an inherent force of evil that menaced the whole community. Another factor was that Scottish Gaels also disposed of an unusually wide range of rites and objects – such as pieces of iron or bread, the Bible, special stones, salt, burning embers, specially made and shaped holes, sprigs of juniper or rowan, and an array of prayers, blessings, rhymes, chants and spoken formulae – which were trusted as effective in averting hostile magic. A third factor which would have damped down accusations of witchcraft was a widespread Gaelic belief in the ‘evil eye’, the damage inflicted by which was in many, and perhaps most, cases presumed to be unintentional. The damage concerned was, as in other regions where this belief was held, the blighting of humans, beasts, crops and domestic processes in a manner generally credited to witches, with the difference that the perpetrator could not automatically be held responsible. It was counteracted, instead, by a variety of charms, spoken or material, or by the avoidance of individuals credited with the power, or by expecting them to avoid gazing directly at others, or their property. The ‘evil eye’ is recorded elsewhere in Britain, but much more rarely, and there it was generally regarded as a deliberately wielded weapon, being one of the vehicles of witchcraft. That is why it never features in the defences presented in British witch trials.
My article then turned to the other early modern societies with Celtic social groups and cultures to ascertain whether similar beliefs were found there, and the result was uniformly positive. The study mentioned earlier of witch trials in the Isle of Man found that the early modern Manx were not only characterized by a strong belief in fairies, but also an acceptance of the legitimacy of the formal cursing of people who had wronged the person delivering the curse. After trials for witchcraft were abandoned there, the normal response to an accusation was an attempt to reconcile the people involved, and to make an accused person, if they admitted placing a curse on the accuser, apologize and withdraw the action.14 This was exactly what the new research into early modern Wales also disclosed. That portrayed a society in which the image of the morally depraved witch, an inherent menace to her community, was rarer than in England and Lowland Scotland. The native Welsh attributed uncanny misfortune to the involuntary evil eye, or spells cast by service magicians hired to help prosecute personal feuds. Counter-magic and the blessings of priests and power of prayer were regarded as being effective against both. Likewise, cursing an adversary was regarded as acceptable if employed as retaliation for injustice, and a common response to an accusation of witchcraft was to arbitrate between the parties and win an apology and the retraction of the curse if the person making it was found to be in the wrong. Wales was also revealed to have a fear of the depredations of fairies greater than that in England and perhaps more than that in Lowland Scotland.15
The remaining Celtic society to be considered in the article was Gaelic Ireland, and the same pattern held there. That was revealed to have a pronounced belief in witches, but the main activity charged against these was the magical theft of dairy produce, a belief also found in Gaelic Scotland and Man. This could be a serious matter for subsistence farmers, but was normally regarded as being on the level of an annoyance and irritation, and used as a whispered mechanism to explain why some people prospered and others did not, without apparent reason. Apparently almost missing from Irish culture was an image of the witch as a killer, of humans and livestock, who was motivated by natural malevolence. That role was allotted to the fairies, which were dreaded as the source of uncanny illness and death, to people, crops and beasts, and uncanny misfortune in the home and business, as well as as abductors of vulnerable humans, especially children. Much care was expended on avoiding, propitiating and repelling them; and scholars noted that the deeper into purely Gaelic areas they went, the more the fear of fairies seemed to wax and that of witches to wane. There was also a persistent Irish fear of the evil eye, often used unintentionally, and, just as in Gaelic Scotland, belief in a wide range of magical remedies that were believed to ward it off.
Recent Reflections on Gaelic Witch Beliefs
The article was published with the intention of developing the debate and of testing the ideas proposed within it, and not of concluding matters. At the time of writing the present book, initial responses to it seem to have been favourable. A well-researched essay on witch trials and beliefs in the early modern northern Highlands has accepted its conclusions.16 So has Andrew Sneddon, who has emerged as the leading expert in early modern Irish witch trials, and filled out knowledge of the Irish context considerably. He has confirmed the lack of prosecutions among the Gaelic majority in the island, but also among the medieval English settler population, the ‘Old English’, who continued to adhere to Roman Catholicism. Even the ‘New English’ settlers, who arrived as Protestants in the late sixteenth and the seventeenth centuries, only produced four known trials, with one execution (making a glaring contrast with the usual estimate of around five hundred executions in England, let alone those in Lowland and Scandinavian Scotland, where in relation to population witch-hunting was twelve times as intense as in England). This was, he shows, despite the fact that both Irish Catholics and Irish Protestants believed in witches during the period: Old English aristocrats and the wife of one Gaelic nobleman accused relatives of attempting to use witchcraft in family feuds, and there were many rumours of it among the New English. It was also despite the fact that Ireland had more or less the same laws against it as England, and the same legal machinery with which to take action. To make the contrast still more glaring, it is now clear that Irish Catholics, including Gaels, made regular use of law courts controlled by Protestant English officials, with respect to offences other than witchcraft.17
Sneddon has demonstrated that the lack of trials among the Protestant settler population was largely due to the fact that the bulk of that population arrived in the seventeenth century, when the judges who dominated the law courts were starting to become cautious in accepting accusations of witchcraft. With respect to the native majority of the population, he held to the view that cultural factors among it left it disinclined to make such accusations at all. He accepted that it lacked the concept of the satanic witch, and thought that witches did not normally injure humans or livestock. He agreed with the significance in preventing witch trials of a belief in the evil eye and in fairies, and in the efficacy of counter-magic, and documented those more securely back into the early modern period. He also emphasized the importance of the belief, which the Irish shared with other Gaels, in the chief activity of witches as being the lesser one of stealing milk or butter by magical means. This, he pointed out, had existed since the high Middle Ages, if not before, as in the late twelfth century Gerald of Wales had reported the tradition, which he said was found in Ireland, Wales and Scotland, that ‘old women’ turned into hares to suck milk from cattle.18 Sneddon also documented from folklore studies, some of which had been cited in my own article, the endurance of this belief into modern times.
The prevalence of the concept of the witch as thief of dairy produce by the twelfth century across the regions named by Gerald might indicate that it was specifically a Celtic motif, as the lands concerned are areas in which Celtic languages were found. It must count for more, however, that they were the main areas of the British Isles in which a pastoral economy was predominant, because the same idea was found in such regions across much of Northern Europe. It is condemned as a vain superstition (and one specifically concerning women) in the eleventh-century penitential of Burchard of Worms, in the Rhineland, which drew in turn on earlier texts.19 It was found in modern folklore in much of Scandinavia, including the motif of transformation into a hare, and there it has been documented back to the early fourteenth century in the more general form that witches were believed to steal milk. During the late Middle Ages that belief became a standard component of church paintings in Sweden and Denmark, with outliers in Finland and northern Germany: and again women were specifically the targets.20 In the early modern period, milk-stealing was one of the most common crimes of which women were accused in Polish witch trials.21 The abundance of Polish witch-hunting, and that of trials in the Scandinavian regions named, is proof that an association between women and magical milk-stealing was not in itself a disincentive to witch-hunting. Instead, like the voluntary evil eye, it could be readily incorporated into the construct of satanic and dangerous witchcraft. It may be suggested that the other factors that existed alongside it made Gaelic culture less inclined to witch-hunting; and this exercise in building on Andrew Sneddon’s valuable research may illustrate, again, the potential of the comparative method, across societies and regions, when studying the subject. One further reflection is relevant before leaving the figure of the milk-stealing witch, and that is that, across the Gaelic world, the magical theft of dairy produce was not thought to be confined to certain women: it was also, or became, one of the many injuries and nuisances which were blamed on fairies. At any rate, it features as such in nineteenth-century folklore, as collected in Ireland, Wales and the Scottish Highlands.22 If the same feature was a part of earlier belief systems in those regions, then even a malpractice commonly associated with witchcraft would not necessarily be blamed upon it, thereby reducing further the inclination to hunt witches in these regions. Another striking characteristic of the rich nineteenth-century Irish folklore collections is the overall balance of misfortunes ascribed to witches and fairies. Not only were the latter feared more in general, but they struck at the core of human concerns. The attacks of witches tended to be on farming produce and processes, but fairies killed and injured people, and their animals, and were especially dangerous to children and young adults, the prime targets of witchcraft in both popular and learned belief across most of the Continent and indeed most of Britain.23 If these beliefs perpetuated those held by the same societies in the early modern period – and the societies concerned were little different then, at the level of rural commoners at which the folklore concerned was collected – then much of the absence of apparent animosity towards witches would be explained.
These considerations may well beg a further question: what of the remaining areas that have been the home of surviving Celtic languages? Did they show the same pattern with respect to witch persecution in early modern times? Two such areas may readily be identified – Cornwall and Brittany – and in both cases there appears to be a paucity of relevant research. In both cases also, however, that may itself reflect a paucity of actual trials. At first sight it might seem as if Cornwall might simply be disregarded, as too small and too Anglicized to be included in the sample, its native language reduced by the early modern period to the extreme western districts. On the other hand, it certainly did retain a distinctive ethnic and cultural identity in that period,24 and had few known trials for witchcraft, especially in comparison with most of the West Country. It had, in fact, just twelve, compared with sixty-nine in neighbouring Devon and sixty-seven in Somerset, and all but one of those twelve trials occurred after 1646, when Cornwall had effectively been conquered by an English army at the end of the English Civil War and submitted to an exceptionally Anglicized administration. On the other hand, Dorset, the fourth county of the West Country, had only thirteen known trials, so Cornish particularism may not account for the contrast in itself.25 Brittany is a different matter, being a substantial region and the European Continent’s major centre of Celtic culture. There seem, however, to be no available published studies of witch trials or early modern beliefs concerning witchcraft in it. All that can be said so far is that it does not seem to have been notable for witch-hunting.
A little earlier in this chapter, the issue was raised of the use of modern folklore to extend or interpret the early modern evidence for a predisposition against witch-hunts in Gaelic parts of the British Isles and in Wales. In the fifth chapter of the present book a warning was delivered against the back-projection of folklore collected in the nineteenth and twentieth centuries to fill gaps in the evidence for particular beliefs and customs in earlier periods with the assumption that folk culture was essentially timeless. The argument for a fundamental cultural element in the lack of witch trials in at least the majority of Celtic societies has not quite fallen into that trap. My article in 2011 did draw heavily on nineteenth-century folklore collections to illustrate Gaelic and Welsh beliefs concerning magic, but also emphasized that the collections themselves contained much internal evidence of change and development; and considered that evidence. It furthermore pointed out that much of the material recorded in the nineteenth century was collected from elderly people who learned it when young, pushing it back into the previous century and so effectively to the brink of the early modern period. Moreover, the argument in that article was based on using folklore collections to flesh out and add insight to the scarcer material from early periods. It attempted to avoid the error of treating them as though they represented a body of information about significantly older times, which could be uncritically back-projected into them.26 For example, Martin Martin’s account of the Hebrides at the end of the seventeenth century, which is an important source for early modern Scottish Gaeldom, contains brief accounts of local beliefs in the efficacy of specific forms of counter-magic in allaying magically induced misfortune; of the theft of milk and butter by women using magical means; and of traditions concerning the evil eye.27 It is surely legitimate in such a case to join this information with fuller descriptions of the same beliefs as held in the same communities a couple of generations later, as long as the sources for them are clearly identified and distinguished.
To employ modern folklore in this cautious way is to follow a precedent established by other sorts of recent, highly regarded, research into early modern Scottish belief systems.28 This precedent has been continued in subsequent work on witch trials in Celtic-speaking areas, notably that of Andrew Sneddon, who, when considering Ireland, was careful to cite early modern (and where available, medieval) sources together with more recent folklore, to show that together they represented a continuum. My article in 2011, moreover, proposed a further control mechanism on the use of modern folklore to add depth to early modern material: the employment of medieval Irish literature, which together with modern folklore could allow the early modern data to be reflected both backwards and forwards in time, building up, if this strategy were successful, a very strong case for particular and persistent elements in Celtic societies which could explain the pattern of prosecution for witchcraft.29 There was no space in the article to offer more than a cursory summary of the Irish material concerned, but the present book provides an excellent opportunity to carry out just such an exercise, and extend it to other classes of medieval Irish record and to Welsh sources. Such an investigation of a possible ancient and medieval backdrop for an early modern phenomenon is wholly in keeping with the aims and methods of this book, and the results may either substantiate the argument for a cultural explanation for the apparent reluctance to hunt witches in Celtic societies, or force a major rethinking of it.
The Medieval Context
The written sources for medieval Gaelic Ireland are very rich, comprising for the present purpose devotional works, law codes, secular poetry and secular prose tales of the deeds of heroes. There is unhappily no similar body of evidence for medieval Man and Gaelic Scotland, although careful projection of the Irish material onto both is possible as place names and incidental references show both shared many cultural traits, and in the Scottish case some saints and heroes. Wales, on the other hand, possesses every class of medieval record that is found in Ireland, though in smaller quantity. To commence with the Irish sources, the secular laws, encoded between the sixth and ninth centuries, show little sign of a concern about magic. There are a few specific prohibitions which really reflect other priorities, such as a banning of love magic, as to coerce somebody else into falling in love with the magician could affect marriage alliances and inheritances; a prohibition on casting spells to cause impotence, which could have much the same consequences; and one on the taking of human bones from churches for use in magical recipes, which dishonoured the dead and desecrated the place.30 Similarly, a legal treatise declared that the fine for causing somebody’s death by a spell should be the same as that for murdering somebody and then concealing the corpse: the concern here was with killing by stealth, as dishonourable, instead of disposing of an enemy openly.31 There is no sign of the witch figure in these law codes, and the situation remains similar in the penitentials, composed between the late sixth and eighth centuries by churchmen concerned with imposing punishments to expiate specific sins. One forbids the use of magic in general, or perhaps destructive magic in general (the word used, malifica, could mean either when used by a cleric) in orthodox early medieval Christian fashion, and regards clerics and women as being especially prone to teach it. Another condemns the employment of magical practices for love potions and to induce abortions and (above all) to commit murder; and with that the concern of these documents for such matters seems to run out.32
Moving to the saints’ Lives and heroic literature, it is clear that the medieval Irish had a considerable interest in magic as a literary motif, but the figure at the centre of it is not the witch, but the druid; and this is for the absolutely straightforward reason that druid (drui or drai) was simply the medieval Irish word for somebody who worked magic. This breadth of definition made the category extremely porous, as anybody could be called a druid at the time at which that person was working magic; conversely, full-time specialists in working magic could be considered to be full-time druids, and so members of a distinctive class or order of person.33 To blur boundaries (and perhaps to confuse matters) still further, certain occupations, notably that of high-grade poet, or blacksmith, were often regarded as inherently able to wield arcane powers, and those members of them who did so were credited with such powers because of their trades, and not automatically called druids. In the literature, people who are called druids often carry out acts of destructive or deluding magic of the sort associated with witches all over the world: they curse and blight people and their possessions, raise tempests and fogs, cause delusions, and transform human beings into animal shape or into stone, subdue and bind them to their will, and raise magical barriers to their activities. To increase the resemblance, the pagan druid features as the favourite foe of Christian saints, to be defeated and so converted, humiliated or destroyed by them. So are druids simply the medieval Irish equivalent to witches?
The answer must be negative, for two different reasons. The first is that magic is treated in the heroic literature, at least, as a neutral force. There are therefore plenty of good druids in these tales, especially in the guise of counsellors to kings and defenders of their peoples, who function as admired, wise and benevolent figures. The main role of the druid in this class of literature, indeed, is not as a wielder of harmful magic but as a diviner or prophet. The second reason spins off from the first: that even the bad druids are not regarded as inherently evil in their activities, though some may be unpleasant as people. They are cast in the role of villains, for the most part, because they are opposed to the characters in the tales with whom the audience is expected to empathize. The partial exception to this rule only serves to confirm it: the druids who are pitted against saints in hagiographies, as the prime defenders of the old and wrong religion against the new and correct one. As pagans, they are defending a bad cause, but are still the misguided or self-seeking champions of their native societies, and not self-consciously the foes of humanity. At worst, they exemplify the evils of paganism as a false belief system, and the negative images of them draw heavily on those of magicians at ancient royal courts in the Bible, who are defeated by Hebrew prophets, or of false prophets exposed by Christian apostles.
In both the heroic stories and the saints’ Lives, druids are usually male when their sex is indicated, while at times they are specifically identified as female. Given the hazy linguistic status of the role of ‘druid’ in Ireland, however, this may simply mean that both women and men were expected to work magic, and that in most of the contexts depicted in the literature that concerned magic, the protagonists were men. At times, however, there seem to be references to specifically female forms of magic, or to women as especially feared forms of magician. The reference to them in a penitential has already been cited, and the epic story of the second battle of Mag Tuired portrays an army about to engage in battle, supported by curses placed upon its enemies by druids, but also by four other specialist kinds of magic-workers, including ‘sorceresses’.34 A hymn credited to St Patrick asks for divine protection against the ‘spells of women, smiths and druids’, while a prayer credited to St Columcille claimed that the speaker heeded (at the end of a list of superstitions) neither ‘omens nor women’. 35 None of these sources, however, explains what, if anything, was distinctive about women’s magic. A saint’s Life, that of Berach, describes a group of malevolent female magicians in action, led by a pagan one who was determined to kill her young stepson, who was under the protection of the saint. She summoned her ‘band of women of power . . . to ply druidism, and craft, and paganism, and diabolical science’ to destroy the boy, but Berach’s prayers caused the earth to swallow them all up.36
The Irish texts are replete with technical terms for curses, such as áer and glám dícenn, kinds of poetic malediction recited in verse; corrguinecht, (a particular mode of cursing uttered while standing on one leg with one eye closed and one arm outstretched); congain connail (magical wounding); and tuaithe (a spoken charm). One law text states that such techniques were sometimes used while piercing an image of the person to be harmed.37 Jacqueline Borsje, perhaps the leading recent expert on the treatment of magic in medieval Irish sources, has commented that
when we look at the Irish terms for supernatural verbal power, we are stunned by their variety. Many of these words are translated simply as ‘magic, incantation, charm, spell’, but the variety of terms seems to reflect a variety of meanings. The definitions of what they stood for have been lost.38
Her words have to be heeded by anybody considering the place of destructive magic in the medieval Irish imagination. It is plain, however, that such magic is wielded in the tales in pursuit of private and specific ends, with practical benefits in mind, rather than for the general joy of working harm. When deployed on behalf of characters of whom the storyteller approved, it was regarded with proportionate approval; and indeed down to the early modern period, poets who allegedly possessed the ability to wound or kill adversaries with their verses (when provoked) were greatly admired for their prowess.39
All this needs to be borne in mind when reading accounts of apparently bad magicians in the heroic stories (often called ‘witches’ in English translation). Perhaps the best known of these magicians are those responsible for the death of the greatest hero of the Ulster Cycle of stories, Cú Chulainn. Their identity developed as the Middle Ages progressed. In the earliest version, found in eleventh- or twelfth-century texts, the kin of two warriors killed by the hero decide to avenge them by hiring three old women, each blind in the left eye (often a sign of magical prowess in Irish tradition) to bring about Cú Chulainn’s doom. These do so by trapping him into breaking a prohibition laid upon him against the eating of dog flesh, so ensuring that death must come upon him.40 In later medieval versions, the same trap with the same result is sprung by the children of one of the dead warriors, Calatin. They learn magic of specific kinds, which are carefully itemized but can mostly be translated now only with the imprecision of which Jacqueline Borsje warned: the sons acquire druidecht (druid-craft), coimlecht (hostile spells), admilliud (blighting) and toshúgad(‘bringing forth’), and the daughters dúile (magic which could relate to books, elements or animals) and amaitecht (lethal magic). All sacrifice an eye as part of the process of education in magical abilities. Once proficient, they use enchantments to handicap Cú Chulainn’s people, the Ulaid, as well as to destroy him.41 They are clearly to be regarded as villainous, because they employ an underhand trick to get rid of somebody who is an honourable and admirable man as well as a great hero, and are the enemies of the people with whom the story identifies. Their action, however, has an entirely understandable and legitimate motivation in itself: to avenge a parent.
The taking of a long chronological perspective enables a comparison between these elite sources from the Middle Ages with the folklore collected from Irish commoners in the nineteenth and early twentieth centuries. Those commoners considered it a customary expression of enmity with somebody else in the same community to curse the land farmed by that person. It was done by depositing rotting matter on that land secretly with an imprecation, which would cause the luck of a farm to decay even as the materials concerned did. Such action was countered by discovering and burning the deposit, coupled with prayers for protection and blessing with holy water.42 The same moral economy therefore prevailed as that found among the elites in the medieval stories: that the use of magic as a weapon was a neutral force in itself, which was awarded virtue or vice according both to the manner in which it was wielded and where a person’s sympathies lay in the dispute. Another abiding link between these two bodies of evidence is belief in the evil eye, as a danger to both humans and livestock. In both cases, it was regarded as a force that could be used both intentionally and by accident, and the issue of intent was the vital one in determining whether its use was culpable.43
There is, however, a different way in which the medieval Irish texts may connect with the witch figure, and that is through their inclusion of terrifying and dangerous women with superhuman powers. Whether or how these connect with the Graeco-Roman child-killing demons or the Germanic cannibal witch is a difficult issue to resolve. One penitential, from the late sixth or the seventh century, forbids Christians to believe in lamia or striga, as delusions.44 By these Latin terms it seems clearly to mean either the predatory Mediterranean night demons or else the women who were associated with those demons, but the ruling may simply echo those in contemporary European law codes, without specific reference to native Irish belief. A correspondence with that belief does, however, seem to be established by a glossary of names from heroic tales, composed at any point between 1050 and 1200.45 It tells of how certain women of Munster, the south-west Irish kingdom, had a habit of invading houses to kill new-born boys. Their powers, which seem at least partly magical, were too great to resist, but one hero was saved from them as an infant because one of the women had a fondness for him and hid him under a cauldron. The others detected him there and attacked him, but only blasted one ear, which remained permanently reddened. This sounds authentically Gaelic, but there seems to be no other appearance of such figures in the literature and there are no precise equivalents to them in the modern Irish folklore. Spectral child murderesses are recorded from various regions in that, but are the ghosts of specific evil women, haunting their former neighbourhoods because of their sins in life, and best fought with the exorcisms of Catholic priests.46
None the less, medieval Irish literature is full of violent and scary women, who are usually described in English translations as ‘hags’ or ‘witches’. A typical appearance of these in a tale is represented by the seven who were defeated by the hero Art mac Cuinn when they attacked him at night in an oak forest.47 To describe them as witches in the sense of this book is, however, difficult for two reasons. The first is that there is no sign that they use magic, as such, as their weapons seem to be physical, with piercings and hackings. The second is that it is not absolutely clear that they are human. They are sent against Art by a superhuman queen, dwelling in a parallel otherworld, whose enmity he had incurred. Likewise, in one of the stories which make up the Fenian Cycle, that group of tales concerning the warriors led by Finn mac Cumaill, Finn himself and some of his men are trapped and bound by magic in a cave by three hideous hags, with coarse and dishevelled hair, red and bleary eyes, sharp and crooked teeth, very long arms, and fingernails like the tips of cows’ horns. They intend to kill the heroes with swords, but one of their comrades breaks in and slays two of them with his own sword, before forcing the last to release his friends. This time the hags concerned are definitely not human, but the daughters of a being from a parallel world who wants to punish Finn for having offended him.48
More often the nature of the murderous hags of medieval Irish literature, whether they use magical or physical weapons, is left ambiguous. That literature abounds with images of divine females, apparently former pagan goddesses, who delight in battle, incite it and engage in it, and as such inspire terror among humans. The savage crones of the stories often look more like downsized versions of these than human beings who have learned magic. In another Fenian story, from the high to late medieval period, ‘demonic females of the glen’ join ‘the hounds and the whelps and the crows’ and ‘the powers of the air, and the wolves of the forests’ in ‘howls from every quarter’ to urge armies to destroy each other.49 Likewise, the twelfth-century historical saga entitled The War of the Gaedhil with the Gaill portrays another list of bogey figures as eager for the bloodshed of the great battle which concludes the story, which may be translated as ‘battlefield spirits and goat-like battlefield spirits, and maniacs of the valleys, and destructive witches and shape-shifting supernatural beings and the ancient birds and the destroying demons of the air and the heavens, and the misfortune-giving demonic supernatural host’.50 Once more, however, the term rendered as ‘destructive witches’, amati adgaill, is a relatively rare one for ‘females with destructive supernatural powers’, who might be human or might not, and if human might be ghosts or might not: it might equally be rendered ‘frenzied women of destruction’.51 In appearance the medieval Irish hags bear a clear resemblance to some of the Roman portraits of witches (which passed in turn into early modern and modern European currency), but their nature seems essentially different. Moreover, they represent external enemies to the heroes and communities of which the Irish stories are told, and not a hidden and internal threat to them; and so, again, hardly qualify as witches under the definition adopted in this book.
The same consideration applies to the much rarer appearances in the stories of beautiful and alluring human enchantresses, of whom the most obvious example is the one who seduces a king, Muircertach mac Erca, in order to bring about his death, in a twelfth-century story.52 He and his followers are inclined to regard her as superhuman, but she insists that she is not, but a Christian woman, and at the end she confesses that she was motivated to her deed by a desire to avenge her family and people, whom the king concerned had destroyed. None the less, she disposes of extraordinary powers of deception, such as appearing to turn water into wine and fern leaves into pigs, and stones, clods and stalks into warriors. Such abilities were associated with demons in Christian theology, and an alliance with such beings would explain why she insists that no churchmen be allowed to enter the royal household while she resides there. Despite all this, having secured her revenge she dies a penitent Christian, and the narrative does not entirely censure her for what is, after all, an understandable original reason for her behaviour. Once more in medieval Irish tradition, a witch-like figure is shown to have been propelled into using destructive magic by personal and specific grievances and not a general malevolence.
This discussion has already drawn attention to the importance in this literature of relations between human beings and a superhuman race in human form, possessed of innate magical powers, who inhabit a parallel otherworld with ready access to the apparent world. This otherworld is often situated within hills or prehistoric tumuli, and its inhabitants are commonly termed the Túatha Dé Danann. They have an aristocratic or royal lifestyle and a small number of them were almost certainly pre-Christian native deities. Their relationship with people is very close to that of fairies and elves in early modern and modern Gaelic folklore, and they are themselves fairly clearly ancestral to at least some of those beings. Although their treatment of humans is often more benevolent and patronal than that of the later fairies, they are also very dangerous when provoked, as Art and Finn found out to their cost (together with many other heroes). Spells represent some of their main weapons in dealing with the human race, as also with superhuman enemies and with each other, and both fear and admiration of them run through all the major cycles of medieval Irish stories. The later Gaelic situation, in which serious misfortune is blamed more on non-human than human beings, is already prefigured in these fictional representations.
A very similar race of semi-divine beings in human form, sumptuously dressed and equipped and naturally adroit in magical arts, features prominently in medieval Welsh prose and poetry. It inhabits in particular a parallel otherworld called Annwn or Annwfn. In the early twentieth century, indeed, it was more or less orthodoxy among scholars of medieval romance that the fays that play such a major part in chivalric tales across Western Europe are descended directly from these Irish and Welsh figures.53 If that were indeed the case, then the latter would ultimately be ancestral to the later English and Lowland Scottish ‘royal fairies’ as well. It has, however, now been abandoned by most experts in the field, not because it has been disproved but because it seems impossible to prove, and so to be a sterile quest.54 This problem is of no consequence to the present study, which needs only to note that the later Welsh fear of fairies is like the Irish one presaged by the presence of these beings in the medieval imagination.
Another figure held in common by both medieval Celtic cultures is the murderous hag. A classic group of these appears in the Welsh romance Peredur fab Efrawg, and is often rendered into English as ‘the Nine Witches of Gloucester’. In the Welsh, however, the term for them, gwinodot, is a broader one for fearsome old women, and though they seem to have insight into the future they possess no other apparent magical powers. Instead they use conventional weapons with terrifying strength, and devastate whole districts until they are killed by Arthur and his band. Another of the same kind is disposed of by the same group of heroes in a different story, Culhwch ac Olwen, and, again, she uses physical strength to overcome, wound and cast out warriors who invade her cave, until Arthur himself slays her with the throw of a knife, so avoiding grappling with her.55Once more, it is not clear how much these beings belong to the human race. They lingered in popular Welsh lore as the figure of the gwrach, a hideous old female being who haunts wild and lonely places and terrifies travellers: it is significant that this being never appears in an actual early modern witch trial, but belongs wholly to imagined situations.56
Two historians of early modern Welsh society and culture, Richard Suggett and Lisa Tallis, have recently assembled a great deal of material, dating from between the late fifteenth and the early eighteenth centuries, which is relevant to this investigation and plugs the gap between the medieval literature and the modern folklore in a way that does not at present seem possible for Gaelic Ireland.57 Belief in the evil eye, often involuntarily used, is well attested as is also one in the power of charms and prayers to ward it off. So also is a considerable degree of acceptance of the validity of cursing, usually performed before the Reformation by a priest and used against people who had hurt good parishioners. After the change of religion it became an individual act, commonly in the form of a prayer to the Christian God for retribution, taking the form of the death of the person concerned or the loss of her or his property. It was often a weapon of the socially weak against the strong, and so especially used by women. The respectable remedy against it was to pray for protection, often enlisting the help of saints and employing the water of holy wells, both of which continued to enjoy a greater popularity in Wales than in other parts of Protestant Britain. There was also, however, a flourishing culture of folk magic, much of it provided by service magicians, which could also be used to avert and remove ill fortune and ill wishing. In addition, a fear of fairies, as malicious and predatory beings, is recorded back to the sixteenth century. All these features would act to damp down animosity against witches.
This belief system was, however, also susceptible to remoulding by external influence: in technical scholarly terms, ‘acculturation’. Richard Suggett has shown how from the 1540s the new word wits, a direct borrowing of the English ‘witch’, was starting to make an impact, and with it a new sense of a specific enemy who harmed people deliberately and from malice. By 1600 it could be coupled with a further association, of a communal devil-worshipping sect dedicated to working evil. This figure became quite deeply embedded in Welsh society during the seventeenth century, especially in the more Anglicized southern counties, and began to play a part in the articulation of enmity and suspicion between neighbours. Wales was on its way to becoming a witch-hunting society, even as the fringe of Scottish Gaeldom, especially the eastern coastal strip, actually did by the seventeenth century. That Wales did not really become one may be attributed to the fact that the process of acculturation began too late and was too incomplete to have made sufficient impact before the time that educated opinion in Britain as a whole began to turn against witch trials. This may also be suggested as the reason why the main Gaelic hinterland of Scotland remained immune to them; while the Manx effectively experimented with the execution of alleged witches and then backed off from it, and were thereafter exposed to no further external pressure to change their minds again. It may therefore be concluded that there is now a sufficient accumulation of evidence, gathered from periods spanning the medieval, early modern and modern, to enable a confident proposal that Wales, Man and Gaelic Ireland and Scotland represented a set of societies that traditionally lacked a serious fear of witchcraft, in the manner of others found across the world, as discussed in the first chapter of this book. This lack was sufficient to enable them to resist the adoption of large-scale witch-hunting, and in many areas witch-hunting at all, even though for a time that became a characteristic of the English and Scottish states that dominated them. If accurate, this conclusion represents a striking example of the manner in which ancient and medieval tradition could, at a regional level, play a decisive part in preventing as well as encouraging the persecution of people suspected of witchcraft.