submitted to the Victoria University of Wellington

in fulfilment of the

requirements for the degree of

Master of Arts

in English Literature

Victoria University of Wellington

1999

This thesis is both a summation of the origin and nature of the character of the Witch as she appears in popular literature from the 17^th to the 20^th century, and an examination of how that character has evolved in the works of recent science-fiction and fantasy authors. The first part of the thesis examines what are generally considered the two major sources of the modern popular tradition of the Witch - the Weird Sisters of Shakespeare's Macbeth, and the witches of popular fairytales. The Weird Sisters are examined in the first chapter, and analysed as a negotiation or compromise between several different traditions of the Witch, especially popular rural traditions and elite "demonological traditions". It is argued that their compromise nature leads to a radical indeterminacy in the Weird Sisters, an ambiguity which makes them especially powerful characters - they therefore become universal antagonists, composed of the hostilities and anxieties of all parts of society. The chapter on the witches in fairytales continues this analysis of the Witch as a compromise construction - here, Witches are examined as mediations between the popular rural folk tradition and the pedagogical project of the 19^th century collectors/editors of folktale. Thus, the Witch as universal antagonist becomes a nursery antagonist as well - becoming the embodiment of both the teller's resentment of women who do not keep accustomed place in society, and the child's resentment of the punishing mother. She is a generalised Outsider figure, a combination of all conceptions of the "bad woman", who can be punished and scapegoated without sanction.

The second part of the thesis examines how the Witch has been reclaimed in modern speculative fiction. The third chapter, examining the Lancre Witches in Terry Pratchett's Discworld novels, reinterprets the popular strands of the literary tradition by depicting Witches who are accepted by though separate from their rural society, who combat all forms of tyranny over human freedom, and whose true power comes from complete self-assurance and self-control. Pratchett thus reinvents the popular Witch tradition to portray the Witch as an individualist hero free from society and yet necessary to it. The final chapter emphasises this reading in examining several science-fiction and fantasy novels which are based on the neo-Pagan conception of Witchcraft. The role of this new kind of Witch as "necessary outsider" and individualist hero is re-emphasised - we see how the Witches of these novels use the neo-Pagan conception of "magical training" as a means to strengthen the personality and thus their personal independence, and how their struggles in the novels are against all those who wish to circumscribe the freedom of the communities in which they live. Overall, then, it will be argued that the Witch tradition as it has evolved since Macbeth is of individualist outsiders antagonistic to the community; and that, in line with the general preference of speculative fiction for individualism, modern writers are positively reinventing the Witch as the individualistic outsider necessary, and helpful, to the community.

ABSTRACT *

ACKNOWLEDGEMENTS: WITHOUT WHOM... *

INTRODUCTION: Who is the Witch? *

The Weird Sisters of Macbeth: The Birth of a Literary Tradition *

Introduction: Why Start Here? And Which Witches, Anyway? *

The Social Environment: High Treason and Churn-Cursing in Early Jacobean Britain *

The Construction of the Weird Sisters: A Negotiation, An Exploitation, or a Muddle? *

A Deed Without A Name: The Weird Sisters, Macbeth, And Crimes Against Nature *

Conclusion: Why Do They Live On? *

Old Wives' Tales: Witchcraft in 19^th Century Children's Literature *

Introduction: Folk Tales, Nursery Tales, and Fairy Tales *

The Fairytale Witch as Wicked (Step)Mother *

The Fairytale Witch as Hag of the Woods *

The Fairytale Witch and Behaviour Control over Women and Children *

Conclusion: The Fairytale Witch and the Self-Fashioning of Children *

"What's Real, What's Not, And What's The Difference": Witchcraft, Self-Fashioning and Freedom in Terry Pratchett's Discworld Novels *

Introduction: A Setting on the Edge of Reality *

Rebellion Against Narrative: The Sources of Discworld Witchcraft *

Lancre Witchcraft as Social Contract and Personal Discipline *

Plays, Mirrors and Fairy-Gold: Imposed Identity vs. Free Self-Fashioning *

Conclusion: Magic, Reality, Autonomy and Identity *

Necessary Outsiders: Neo-Pagan Witchcraft and Romantic Individualism in Science Fiction and Fantasy Literature *

Introduction: Neo-Pagan Witchcraft in Four Different Fictional Contexts *

"A Priestess of the Goddess Bows To No Man": Neo-Pagan Magic as a Path to Self-Fashioning And Personal Power *

"Between the Worlds, In All The Worlds": The Multiple-Reality Neo-Pagan World and the Witch as Necessary Outsider *

"Tyrants In Heaven And Earth": Neo-Pagan Ethics and the Construction of the Antagonists *

Conclusion: Witchcraft, Speculative Fiction and the "Romantic Intellectual" *

CONCLUSION: The Witch's Power of Transformation *

REFERENCES *

First and foremost thanks go, naturally, to my supervisors, Geoff Miles and Jane Stafford; for reassuring me that I was capable of doing this, for deftly steering my lines of inquiry into the right directions, for evaluations of my work which did my confidence no end of good, and for not questioning my prolonged absences while I went off and wrestled with the reality of thesis writing. I hope that I have repaid your trust.

Financial support from VUW's scholarship fund, from the J.L. Stewart scholarship and from my ever-helpful grandparents. My mother, Patricia Hobbs, read drafts of some chapters and assured me that my turgid academic prose was accessible to laypersons. Extra proofing help from the incomparable Ana-Thérèse Ward.

Considering the subject matter, it should come as no surprise that many thanks go to the Wiccan Association of New Zealand (WANZ), many of whose members are fans of Terry Pratchett. Invaluable emotional support came from, in the early days, the Little Witch Group (Samantha, Anoushka, John and Nigel); in the latter days, Bonsai Jungle (Jude, Evan, Tim and Jenny); throughout, from all the members of the Free Commonwealth of Penguinea, and from my various long suffering flatmates. All of the above have put up with far too many excesses of both enthusiasm and black despair as I slogged my way through. Consider this my apology, and heartfelt gratitude. Extra honorable mention to Samantha for the reading list, and Catherine for Carpe Jugulum. Speaking of Terry Pratchett, the Karori Dramatic Society's production of Wyrd Sisters was of invaluable help to me in understanding the dramatic significance of the play. My old friend, Catherine Corrigan, made a great Granny Weatherwax.

My deepest gratitude to Kate Bush, Happy Rhodes, Robert Fripp, the thirteen members of Yes, Peter Hammill, the New Model Army and the late great Frank Zappa, whose music kept me sane while I worked. "I thank you for your expressions; your music has set me free." Fictional inspiration came from all the authors I studied, and from The Craft, a movie at which I split my sides laughing; intellectual inspiration, from J. Michael Straczynski and Kate Bornstein, who came into my life at just the right moment; divine inspiration, from She Who Sings In The Heart, and from He Who Dances.

Last but quite the opposite of least, this thesis would never have come to fruition without the extra-special help, sustenance and perspective of Jude Fitzpatrick, Jenny Howard and Tara McIvor (Wellington), Dan Wardlow and Wendy Vig (San Francisco), Carolynne Robertson-Dunn (Canberra) and Evan Gallagher (Brisbane). I love you all. This work is dedicated to everyone who ever wanted to be a witch, and to anyone who might find this work useful in their own research. It is only a beginning, after all. Good luck!

Daphne Lawless

Wellington, March 1999.

Everyone knows what a Witch is. My grandfather, when I explained to him the topic on which my thesis would be based, replied, "Witches? You mean, those jokers on broomsticks?". The image of the Witch is "the shrieking hag" which Diane Purkiss mentions - either the Weird Sisters of Macbeth, filthy hags brewing up diabolical potions from various entrails, or the Wicked Witch of the West, Margaret Hamilton in green blusher in the 1939 MGM film, cackling as she sends the Winged Monkeys after Dorothy and her friends. She wears a black pointy hat, rides a broomstick, has an evil black cat, warts, and a hooked nose, and often lures children into her gingerbread house. Good Witches, although they exist in literature, are in the minority - as Susan Wolstenhome notes (Baum, p. xxiii), fundamentalist Christians object to the depiction of such things in literature, taking the biblical injunction "Thou shalt not suffer a Witch to live" (KJV, Exodus 18:22) seriously. The Witch has become entirely a popular tradition - elite society has not seriously believed in witches for almost two hundred years, but her image continues in children's literature and the iconography of metaphor and festival custom. To obsessively seek scapegoats is called witchhunting, and the hag on the broomstick appears along with Dracula and the Wolfman as staples of the American Hallowe'en celebration. Even though the Witch is no longer taken seriously, she survives for children, for fun, and as the symbol of a credulous, superstitious past.

This image, however, has begun to come under threat in recent decades with certain literary reinventions of the character. To some extent, especially recently, this has been sparked by the neo-Pagan movement, seeking to establish their own brand of feminist Witchcraft as an accepted religion; but the Witch has begun to be reinvented by authors, especially in the fields of science fiction and fantasy, who bear no allegiance to that religion, although they may be aware of it. It is perhaps necessary to declare my interest at the beginning - as a neo-Pagan myself, who in my formative years was inspired by many of the fictional witches which I talk about in this thesis, the question of why and how the Witch has begun to become rehabilitated in the popular literature of the late 20^th century is of special interest to me. Accordingly, this thesis is an attempt to ask why to two interlinked questions - why has the literary Witch survived in popular culture long since she disappeared from the demonology of "respectable" society? And why is this tradition being reinvented to make her a heroic figure?

This thesis is essentially divided into two parts, each part containing two chapters. The first part is designed to trace the growth and establishment of the conception of the Witch which was illustrated in the previous passage. In studying this idea of the Witch, which it would be tempting to call "archetypal" for its place in popular culture if it were not for the Jungian associations that term would raise, we need to trace two major questions. Firstly, we need to establish its origins - how the character came to be - and secondly, to establish the reason for its survival - why it has survived with such a strong grip on the popular imagination. In other words, the chapters on the Weird Sisters and on fairytale Witches seek to answer the twin questions: where did they come from? and why do they still exist? It should be noted here that it is not being proposed in this thesis that these two "streams" of the modern popular conception of Witchcraft are related at their source, or that one explicitly influenced the other; simply that these are the two recognized sources of our popular tradition, which is why they are considered in separate chapters.

I intend to argue firstly, in the chapter on Macbeth, that the Weird Sisters were a negotiation, or a compromise, between several different traditions of witchcraft - including both the popular rural tradition in which Witches cursed churns and children and otherwise disrupted the subsistence agricultural society, and the elite traditions in which they were considered agents of Satan upon earth, enemies of both God and King. It is this essential polyvalency, I will argue, which has led to the longevity of the Witch as a generalised symbol of threat to the prevailing Western social order. In the second chapter, I will demonstrate the parallel evolution of the Witch of fairystories - here, the rural folk traditions of her character were mediated with the pedagogical requirements of the 19^th century collectors/editors of folktales, who wished the stories to impart good moral lessons to the children of the book-buying European middle class. Thus, I will argue, even after the Witch had ceased to be considered a threat in the elite discourse, and almost so in the popular discourse, she retained her power as a symbol of anarchy and disorder in the "civilising" discourse applied to children. I will focus my arguments mainly on the tales of the Brothers Grimm, two tales by Hans Christian Andersen and Baum's The Wizard of Oz.

The second part of this thesis documents an essential break with that tradition by the authors of science fiction and fantasy, who reclaim the Witch, that quintessential "outsider" figure, as part of a project of presenting the self-fashioned, self-reliant individualist as a role model for their "romantic intellectual" readership. The third chapter focusses on the "Lancre" trilogy of Terry Pratchett's Discworld novels, which I suggest present "witchcraft" (small w) as a social function, a negotiation between witch and community, and "Witchcraft" (large W), as a process of radical personal independence - thus, I will suggest, they not only reclaim the early modern popular tradition of the Witch, but represent it as a means of self-fashioning for modern individuals. This aspect of reclaiming the past will be explored further in the final chapter, an examination of the use of neo-Pagan Witchcraft (Wicca) in a selection of science fiction and fantasy novels. Wicca as a belief system places the highest good upon individual self-expression, and presents "magical training" as a process of self-fashioning so as to attain agency over one's environment - I will show how this has been used, in novels of varying genres, to construct heroic figures who are both necessary to their community and outsiders to it. The modern literary Witch, I thus argue, is constructed as an individualist heroine - the outsider-figure, reclaimed from an earlier literary tradition where she is the enemy of society, transformed into the individualist whom society might distrust, but who is necessary to it.

Above all, it must be re-emphasised that this is an attempt to draw the outlines of a history of a popular cultural tradition. Macbeth began as popular entertainment before its academic canonisation; the fairytales of Grimm, Andersen and Baum still have their place in popular culture by their continued popularity as material to be read to children, and as successful films; and science fiction and fantasy definitely belong to the realms of popular literature. My aim in this thesis is, firstly, to trace where the traditional popular literary image of the Witch comes from, how it was synthesised from often quite diverse sources, and why it retains its power; secondly, to examine how the authors of speculative fiction have opened up a dialogue with it by reinventing the concept of the Witch as individualist hero, rather as enemy of the community. It is my contention that this represents an ideological change in the popular tradition, from an ethics of community solidarity to an ethics of libertarian self-fashioning. Further, I will link this with the evolution of the Witch from a sort of "antihousewife" for early modern rural society, to nursery bogey embodying all the wildness which 19^th century pedagogy sought to exclude, to an image of the "necessary outsider" which appeals to those who feel excluded by the Western middle-class social consensus for which the Witch had become the prime symbol of disorder.

A brief note concerning terminology: in this thesis I often capitalise the term "Witch". As I mention several times in the body of the work, one of the major problems with studying witch-characters in literature is separating out the different and conflicting uses of the term, and I often use "witch" and "Witch" to convey importantly different meanings. In the first two chapters, then, when the term "witch" is uncapitalised it can be read as the broader use of the term, applicable to any of the many and conflicting definitions of the witch which come from popular culture, elite discourse or literature. The capitalised term, "Witch", on the other hand, refers to the specific character that we are tracing in this thesis - the character of the Witch which has become an archetype in popular culture, which I will argue has its origins in the Weird Sisters of Macbeth and in the witches of fairy-tales such as the Wicked Witch of the West. This character, I argue, has several features in common in the several tales in which she appears, but is greatly solidified from what I show to be the confusing and contradictory witch-beliefs of early modern times. In the same sense, the term "Stepmother" is capitalised when we are discussing the fairy-tale Wicked Stepmother who is indistinct from the Witch, as opposed to stepmothers in general. In the chapters concerning modern sci-fi and fantasy literature, the capitalised term takes on a different meaning again. In the Pratchett chapter it refers to the Witch constructed in the Lancre trilogy as a radically self-fashioned individual exampled by Granny Weatherwax, rather than the various images of the witch presented in literature and folk-tales and challenged in the novels. In the chapter on neo-Paganism in speculative fiction, accordingly, the capitalised term is used when referring to a neo-Pagan Witch, or member of the Wiccan religion, by the same convention by which the terms "Christian" or "Jew" are capitalised. Overall, then, the uncapitalised "witch" is a radically indistinct term which is capable of wide and differing application and meaning; the capitalised "Witch", however, is the name for a particular personality, of folklore, mythology or the rhetoric of personal empowerment.

In one way, beginning a history of the tradition of the Witch in English literature with the Weird Sisters of Shakespeare’s Macbeth might seem an odd, or even arbitrary, choice. Witches, as defined in the Oxford English Dictionary’s main sense of "a female magician or sorceress", or at least characters who fit the description even if they do not carry the name, appear in both Malory and Spenser, to name just two mediaeval or early Renaissance writers. The OED, moreover, records the word "witch" (and its Old English equivalents, wicca (m.) and wicce (f.)) as extending back all the way to AD 890. And of, course, the concept of female magic-users stretches back before then – to classical, biblical, tribal or even archetypal definitions of the term. So, why should we start with the Weird Sisters, characters who are sometimes denied to be witches at all and are actually only called witches once in the play itself?

Simply put, this thesis is not intended to become a sociological examination of witch beliefs throughout recorded history, nor even a history of what the word "witch" has meant in literature. Instead, this work is designed as a history of a specific character - that literary creation which, whether we wish it to or no, springs to mind when we use the word "witch", and which is still used as a figure in popular literature today. And, as Diane Purkiss points out in The Witch in History, that figure as we know her today has her genesis in Macbeth. "When we say witch,", says Purkiss (p. 180), "we can hardly help thinking of Macbeth’s witches". Thus, it can be argued that "the Scottish play" is the only place where our history could begin. Whatever we think of such a formulation’s political or aesthetic merits, the Witch as a literary character is established as a ragged, warty, ugly old woman, cackling with mischief and boiling up noisome potions in a cauldron, and it is to Macbeth that we owe this picture. For example, as Purkiss states, the modern association of Witches with cauldrons is due almost entirely to that play (p. 212). Thus, if we are to examine the Witch-character as she has changed over the last four hundred years, we must start by examining where she came from, and how she was constructed. In this chapter, therefore, I intend to lay the basis for the development of my argument by addressing three main questions about the Weird Sisters. Where do they come from? What is their function in the play? And why is it that they have become the foundation for a character tradition that is only now becoming challenged at the end of the twentieth century?

I shall begin by examining the differing ways in which the civilised, literate elite and the mainly illiterate peasant class of the late 16^th and early 17^th centuries conceptualised witchcraft. This will lead on to a discussion of the Weird Sisters as a "negotiation", or compromise, between these different elite and popular traditions. Subsequently, I will attempt an examination of the several different ways in which the Sisters can be intepreted – as women, as supernatural beings, as diabolical agents, as psychological projections and symbols – and connect this with the central role of concepts of "equivocation" in the play. I intend to show that the real power of the Sisters, both in the play and as characters, lies in their ability to transgress (or inhabit) boundaries and to confound interpretation with both their appearance and their speech. They tell Macbeth that they do "a deed without a name", and I will connect the breakdown of language in the face of such ambiguity with the Macbeths’ refusal to name either their desires or their crimes, which leads to their madness and downfall. I will conclude by arguing that this is what makes them such potent symbols of evil – they represent a revolt against not only the established orders of family, society, church and state, but against rational conceptions of language itself. In summary, I will argue that the Weird Sisters are constructed as figures marginalised in every conceivable way by the "natural order" of the time, but conversely acquire the power to wreck the foundations of that order by virtue of their very marginalisation and exclusion.

As Purkiss says, when we think of witches today it is hard not to think of Macbeth’s Weird Sisters – but what image of witches would the audience who went to see Macbeth when it was first performed in 1606 have had, to make the Weird Sisters such a popular creation? The most obvious distinction between the audiences of that time and those of the present era is that witchcraft (or, at least, beliefs about witchcraft) would have existed for the audiences of the Jacobean era not solely as a literary or artistic construction, but as a real phenomenon. Even for those skeptical about the actual existence of witches, the fact that a significant proportion of society believed in them would inform how they would interpret witches in stage presentation. I will begin by briefly examining the socio-cultural phenomenon of witch beliefs at the time, concentrating on how the educated elite had a very different idea of what witchcraft entailed from the rural peasantry.

"The term ‘witch’ was labile, sliding across a number of different and competing discourses," (Purkiss, p. 93), and it is generally accepted that these discourses can be broadly divided into "elite" and "popular" traditions. The former could also be described as a "scholarly-clerical" tradition, in that it was based upon both Church teachings and the speculations of learned men, and was circulated in written form. The latter could be called a "rural-peasant" tradition, and it is mainly known to us from the testimonials and depositions of witnesses and suspects in the witch-trials of the time. The great difference between the two can be described as a difference of scope – for, as I intend to show, the "popular" tradition saw the witch as a threat to domestic happiness and order, whereas the "elite" saw the witch as a threat to the authority of both King and God.

The elite ideas about witchcraft of the early Jacobean period could be divided again into sceptical and demonological discourses. The sceptical viewpoint, explaining witch-stories away as peasant superstition or mere trickery, was most ably expressed in Reginald Scot’s Discoverie of Witches (1587) and was widespread during the Elizabethan period. Another point of view spanning the two discourses, expressed by George Gifford in A Dialogue Concerning Witches and Witchcraftes (1593: see Willis, p. 95), did suggest that the crime of a witch was "dealings with the Devil" - however, Gifford opined, such a witch had no real power, the Devil only giving her the illusion of such power to mock her. The demonological discourses, on the other hand, were to a large extent imported, being highly influenced by Continental views on witchcraft (expressed most infamously in the Malleus Maleficarum), some of which came with the new King, James VI of Scotland and I of England. As Harris (p. 4) notes, the Scottish witch-beliefs were much closer to those of their continental neighbours in many ways – for example, the concept of the coven, which was mostly unknown in England, and James’ interest in demonology and the "discovery" of witches led to many of the more "flamboyant" Continental witch-beliefs gaining the greater currency around court (Harris, p. 5). Although scholarly argument of the time ranged widely over several different definitions of witchcraft, sorcery or other forms of magic, the generally accepted definition of witchcraft was "the gaining of occult powers through a formal pact with the devil" (Harris, p. 3), and therefore abrogating one’s baptismal vow of devotion to God . Among other things, this may explain why both Catholic and Protestant churches accused the other of being in league with witches - in addition, Protestants often associated Catholic belief in salvation by works and the special powers of saints with witchcraft (Willis, p. 119). This more traditional religious view of witchcraft was, however, reinforced with a very strong political association. In his time as King of Scotland, James had been subject to several assassination attempts, and many of these allegedly involved witchcraft in some way (Wills, p. 42). For example, a group of witches from North Berwick under the leadership of John Fian were tried in 1590 for attempting to sink the King’s ship on the way back from Norway (Harris, p. 40). The linkage between the religious and political was made stronger by James’ evolving belief in the divine nature of monarchical authority (Turner, p. 118). According to that theory of monarchy, the Bible states that earthly rulers are to be treated as gods in their domain, and since the monarchical order is a reflection of the divine order upon earth, a crime against the King is tantamount to a crime against God. In this way, treason and witchcraft were conflated. It can be said, then, that the elite conception of witchcraft in the Jacobean court was of treason against divine authority, and that the line between this and earthly authority was deliberately blurred why what Purkiss describes as James’ "crusading paranoia" (p. 207).

The English popular traditions of witchcraft, on the other hand, were concerned with a very different realm indeed from the realm of God and the King. This was reflected in the judicial system set up to punish it – in England, witchcraft was tried by the secular rather than the church courts, and it was a civil rather than a criminal offence, that is, an offence against the life and property of one's neighbours rather than against the state. (The exception was witchcraft concerned with assassinating the monarch or forecasting her death – Harris, p. 8). Further, in the popular imagination the Witch was not envisaged as simply a servant of Satan - the "imps" who she was accused of sending against her neighbours were not seen as creatures of Hell, but of a "third world" of the supernatural, the "fairy-realm" neither attached to heaven nor hell (Willis, p. 91). (This concept was severely frowned on by clerics of the Reformation, who insisted that all creatures must be of God or else of the Devil.) Purkiss thus argues that the only direct documentation of rural-peasant witch-beliefs, contained in the transcriptions of depositions at witch-trials, clearly indicates that that community saw witchcraft as a crime against the domestic sphere of society. In rural society, witches were mainly accused of crimes against the accepted female sphere of action and responsibility – of killing infants or domestic animals, or making it impossible to bake bread or churn butter. In a society based on subsistence agriculture, these were indeed crimes against survival itself, or at least against the female-gendered contribution towards it. The Witch, Purkiss argues, was thus constructed as an "antihousewife" or an "antimother […] a powerful fantasy which enabled women to negotiate the fears and anxieties of housekeeping and motherhood" (p. 93). Similarily, Deborah Willis describes the rural idea of the Witch as a "malevolent mother", who was imagined to feed her familiars on her own blood from an extraneous "witch's teat" (Willis, p. 9). These fantasies attached themselves to old women, the most marginalised figures of the peasant social system and thus convenient scapegoats, especially as their reliance on begging engendered guilt, which in turn engendered resentment (see esp. Willis, p. 239f). In the several stories where women have accused other women of cursing their children or their cooking utensils, extending into the mid-twentieth centuries, Purkiss claims that we can see a pattern where the Witch was seen as "usurping" the natural privileges of the housewife of control over her domestic sphere (p. 111) – that is, attempting to take what did not belong to them. Further, Willis associates the power of the Witch with fantasies of maternal omnipotence (p. 8) - the mother or nursemaid has absolute control over the surroundings of the child, in the same way that the Witch is seen to have over the natural world.

This can be seen as the source of the explicit female gendering of the Witch we have come to accept. Newer theories of biology from the 16^th century put the full responsibility for succesful child-raising on the mother (Willis, p. 17) - this put her in a position of power over her children contradictory to her accepted role as domestic servant, and therefore engendered male suspicion. In the elite discourse, both men and women could be traitors to God and/or King and therefore witches, but the popular discourse placed witchcraft squarely in the domestic, and therefore female, area. How, then, did men of the peasant social order interpret witchcraft? Peter Stallybrass argues that by representing a challenge to the traditional function of woman as guardian of the domestic arena, the figure of the Witch was used as an instrument for patriarchal social control. He makes the point that the rebellious associations of witchcraft which dominated the elite discourses on the subject could also be applied to the popular tradition, in that a woman such as a Witch who abrogated her traditional responsibilities over housekeeping and children was herself in revolt against the natural order (Stallybrass, p. 206) – and, therefore, in revolt against patriarchal authority. Thus, in the same way as the elite discourse conflated rebellion against God and King, the popular discourse can be seen as a negotiation of women’s fear of failure and usurpation in their domestic realm and men’s fear of usurpation of their role of head of the family.

We can then see the ways in which the Witch was seen by both the elite and popular discourses in the early Jacobean period as variations upon a common theme – that of a threat to the natural order, a "disturbing of the peace" (Willis, p. 85), whether it be specifically the peace of religion, monarchy, patriarchy or domestic propriety. When the two discourses came together, the conflict was sometimes obvious - this happened increasingly in formal witchtrials, where magistrates took it upon themselves to intervene in such disturbances to the social order of the community (Willis, p. 83). As Purkiss states (pp. 145f), women of the lower social orders would often accept the identity of "witch", even voluntarily, to enable them to carve out an independent sphere of power within the belief systems of their society, but this "self-fashioning" was to an extent undermined by the elite’s insistence in either attempted debunking of witchcraft or interpreting it in terms of Satanic pacts. "Most JPs were either demonologists or sceptics," says Purkiss (p. 153), and she adds that the confessions of village witches in the trials of the Elizabethan and Jacobean eras can be interpreted as a "negotiation", between the desire of the accused to tell her own story according to her belief systems, and the need for the confession to be acceptable to the conception of witchcraft held by the examining authorities. Further, the most important difference between the elite and the popular conception, as Willis points out (p. 15) is that in the discourse of demonology, the Witch was nothing more than "a drudge of Satan", his tool on earth, whereas in the popular rural tradition she was credited with autonomy and power over imps and familiars of her own. The idea of the "mother gone bad", using her power for evil rather than good, thus intersects with the idea of the Devil's handmaiden to construct a composite figure wherein female autonomy and evil are identified, as with mediaeval speculations on all women's responsibility for the sin of Eve. It is my intent to show that the characters of Macbeth’s Weird Sisters can be interpreted as being a product of precisely this kind of negotiation between popular and elite conceptions of Witchcraft.

The purpose of this section is to examine the sources from which the Weird Sisters are constructed, and how they can be related to both elite and popular definitions of the term canvassed in the preceding section. I will contrast this with a description of what literary depictions of witchcraft existed prior to Macbeth and the other plays of the "Jacobean witch-vogue", and thus hope to show that the Weird Sisters owe more to the contemporary discourses about witchcraft than to any previous literary tradition. Willis suggests (p. 160) that the middle-class and mobile nature of playwrights of necessity put them at the intersection of both native and foreign, elite and popular discourses . Accordingly, I aim to point out how the Weird Sisters of Macbeth can be seen in all these ways - as betraying servants of God and King, and as malevolent nurturers.

Diane Purkiss makes the point that the high-point of prosecutions and convictions for witchtrials, the Elizabethan era, was simultaneously a low point for the depiction of witchcraft on stage (p. 181). "The Witch of the Elizabethan stage gradually emerges from a mass of supernatural figures: sorcerers and sorceresses, classical witches, wise women, prophetesses and fairies." (Purkiss, p. 183). To put it another way, the "solidification" of the concept of the fictional Witch only occurred with Macbeth and the other plays of the Jacobean witchcraft vogue. Prior to this, "witch" could be used with little precision to indicate almost any figure, human or nonhuman (although, even at this stage, usually female) identified with the supernatural or magic. The "classical" witches mentioned are the female magic-users of Greco-Roman myth, featured in the works of such authors as Ovid, Lucan, Seneca and Virgil, who had regained currency during the Renaissance (Harris, p. 21). Malory and Spenser make frequent use of Circe-like figures in their works (ibid.), and there were many popular ballads concerning witchcraft. However, it is only with the Elizabethan era that we start to find witches portrayed on the stage. John Lyly’s witch-plays, Endimion and Mother Bombie, are among the more interesting of these in that they work with two distinct witch-traditions, that of classical literature for the former and peasant witch-beliefs of the type canvassed above for the latter (Harris, p. 27). Dipsas, the witch in Endimion, works with the same associations and makes the same boasts of magical power as classical witches such as Circe and Medea (Purkiss, p. 188) – Mother Bombie, on the other hand, inherits the power of fortune-telling, rather than acquiring it via supernatural/Satanic pact, and thus would probably not have been considered a witch within the meaning of the elite discourse. However, both are constructed primarily as harmless, comic characters – there is no sense of the witches in the play as threatening in any real sense. Moreover, dealing with powerful women in a dramatic manner was made more problematic by the need to avoid offending the Queen (Purkiss, p. 189). However, as the general perception of witches among the English elite moved from the sceptical to the demonological with the arrival of King James in London, the tenor of dramatic treatments of witchcraft became far more serious and tragic – as M.C. Bradbrook puts it, "Macbeth was the first play to introduce to the stage in a serious manner the rites and practices of contemporary witchcraft" (quoted in Jorgenson, p. 116). Also, a male monarch made it easier to portray powerful women as evil, especially considering James' resentment of the interference of both his mother, Mary Queen of Scots, and of Elizabeth I in his reign in Scotland (Willis, p. 125). In the remainder of this section, I will analyse what can be discovered about the sources of the Weird Sisters, with reference to previous literary constructions and to the contemporary witch-beliefs already canvassed, and argue that they effectively supersede the "harmless" witches of the Elizabethan literary tradition, to become the exemplars of a new form of writing about witchcraft. Like Turner (p.22), I have chosen to refer to "the Weird Sisters" rather than "the Witches" – it has been the intent of this chapter to show that the meaning in our literary tradition of the second term is based on the first, rather than, as is commonly believed, vice versa.

The section from Holinshead’s Chronicles of England, Scotland and Ireland from which the inspiration for the Weird Sisters derives bears only a vague resemblance to the characters we recognize. Holinshead records that as Macbeth and Banquo

It can be seen, then, that the fundamentally equivocal nature of the Weird Sisters stems from their very origins. Holinshead can only speculate as to whether these women were flesh and blood creatures, the goddess known as the Fates (Norns to the Norse, Parcae to the Greeks), or remnants of "the fairy folk", who were sometimes seen as supernatural and sometimes as the remnants of an older civilisation (Harris, p. 2). The "elder world" can be perhaps be seen as the more diverse spiritual world postulated by the classical or Catholic theologies, rather than that of the Reformation which assigned all spiritual entities to either Heaven or Hell (Willis, p. 91). As I intend to show, this very indeterminacy of the characters, or "polyphony" as Turner and Rosenberg call it, can be seen as the most important part of their nature. The power of the Weird Sisters in the play stems from fundamental ambiguities about what they are – debate even exists over whether they can be called Witches (Turner, p. 14), and what their powers are.

Almost all critics recognize the Weird Sisters as constructed from several different sources, from classical/literary, scholarly and popular traditions of witchcraft. The most obvious sign of classical literary influence is Hecat, the leader of the witches who appears (in sections generally considered to be later, non-Shakespearian interpolations – Hunter, p. 167) in scenes III.5 and IV.1. Her name is a contraction of that of the Greek infernal goddess Hecate, who is the goddess of the witches of the classical writers. The title character in Ovid’s Medea calls upon her, especially, and scene IV.1 especially has many signs of classical influence - although, like Hecat's name itself, slightly distorted. For example, the cauldron scene is like a grotesque reconsideration of similar spells from that play (Harris, p. 36), and Macbeth’s "conjuring" speech (IV.1.49-57) is based on a speech of Medea from Ovid or Seneca (Wills, p. 63). Wills further opines (p. 65) that the significance of the mixture of these classical and popular traditions was a sign that witchcraft was considered a constant in human history – that is, the Weird Sisters are being explicitly linked with the likes of Circe and Medea. Harris (p. 35) also points out that the classical oracles were famed for giving out prophecies as ambiguous (sometimes deliberately so) as those that the Sisters hail Macbeth and Banquo with (scene I.3). The most important way in which the Sisters build on literary tradition, however, is not in what they are, but what they replace. Harris (p. 48) notes that they serve the same function as various ghosts, demons and oracles did in earlier dramatic tradition – intrusions from outside the order established at the beginning of the play to facilitate the plot, whose "supernatural soliciting" cannot be seen to be either good or evil until the plot has played out. The main differences, Harris argues, are that the Sisters are more dramatically interesting – less "one-dimensional" – and also that the entire supernatural element of the play can be connected back to them. So, the Weird Sisters have some connections to previous depictions of witchcraft, but they serve quite a different purpose dramatically. In this sense, as any many others, they represent a new literary tradition of character.

The popular traditional elements of the Weird Sisters come from a far larger range of sources. The distinction referred to in the first section between Scottish and English witch-beliefs, the former being more influenced by Continental ideas about "pact" witchcraft, is partially obscured in the Weird Sisters. As might befit Scottish witches, they are part of a coven – a concept hardly found in English witch-stories, according to Harris (p.3) – but, on the other hand, in scene I.1 they are generally considered to be responding to the call of their familiars, and the witch’s familiar was a concept peculiar to English tradition (Stallybrass, p. 195). Thus, even their national origins are more ambigious than the "traditional Scottish witches" which Hunter (p. 39) describes them as – it could be possible, in light of the cross-border nature of their portrayal, to describe them as British witches. In appearance, on the other hand, they conform broadly to both elite and popular tradition, which emphasised the ugliness and filthiness of Witches (Jorgenson, p. 118). In the popular tradition, this serves to connect them to the most marginalised members of the peasant community, the old and the deformed; in the elite tradition, to express their hideousness as agents for evil. Jorgenson quotes a demonological author as explaining that "the Devil, who is vile, chooses for servants filthy old hags whose age and poverty serve but to enhance their foulness" (ibid). Filth is a persistent element in the characterisation of the Sisters, from the "fog and filthy air" which features in their chant (scene I.1) to the contents of their cauldron, which seem chosen solely for their disgusting nature (IV.1) To complete the picture, one of the symptoms of old age in women is increased growth of facial hair, and the Weird Sisters possess this further grotesquerie. The Weird Sisters conform in appearance, then, to the image of what the popular tradition would claim a Witch to be – a filthy, ugly old woman.

In their behaviour in the "sabbat" scene which begins I.3, they draw much from the tales from the rural witch-trials. The Second Witch reports herself to have been "killing swine" (scene I.3) – murder of domestic animals was a common accusation against witches, tying in with Purkiss’ interpretation of peasant witchcraft as primarily crime against domesticity. The tale of the sailor’s wife which immediately follows (I.3) is an uneasy mixture of peasant tradition and topical allusion. Chestnuts, as Purkiss states (p. 208), were a staple of peasant diet, being ground for flour when wheat was unavailable, and Harris (p.12) notes that most accusations of witchcraft were occasioned as acts of revenge for real or imagined slights. Again this reflects back on the witch’s place among the most marginalised of a rural subsistence society – Rosenberg (p.4) describes those accused of witchcraft and their "helplessness against society except in any power they have – or imagine they have – to injure those who munch on food and give them nothing". So the First Witch is here behaving as a village witch would be expected to behave, with one major exception – her revenge is not directed against the sailor’s wife personally, but against her husband. Moreover, this revenge is couched in specifically sexual terms. "I’ll drain him dry as hay;/ sleep shall neither night nor day/ hang upon his penthouse lid" (I.3), she says, and it is possible also, as Jorgenson (p.120) does, to detect a sexual undercurrent to "Like a rat without a tail/I’ll do, I’ll do, and I’ll do". This is an irruption of the elite discourse of witchcraft into a peasant story. The witch of the demonological concept, having formed a pact with Satan, was often set to seduce men and deprive them of their precious bodily fluids (Rosenberg, p. 19), but this element was absent from the "women’s stories" on which Purkiss bases her account of popular witch-traditions. However, even in this context the true nature of the Weird Sister's craft is radically indeterminate - exactly what she will "do and do and do" remains the Deed without a Name (Willis, p. 214).

Also to be found in this small anecdote are very strong topical influences which might be expected to appeal to King James, who, as previously noted, had been the subject of several witchcraft-related assassination attempts. The Tiger, on which the husband has gone, was a ship recently returned to harbour in England after almost two years at sea (Hunter, p. 141) – it is also reminiscent of the North Berwick case, whose witches were accused of "raising storms" to sink the King’s ship (Harris, p. 40), and having sailed in sieves. This is further emphasised by their mode of speech – octosyllabic couplets such as Gillis Duncan, one of the North Berwick accused, was said to have used (Purkiss, p. 199). Similarly, the earl of Bothwell, who raised rebellion against James in the name of his mother, Mary Queen of Scots, was said to have consulted with witches as Macbeth does (Willis, p. 209), and Macduff, "from his mother's womb/ Untimely ripped" (V.8) might have appealed to James as a symbol of male power divorced from interfering maternal or quasi-maternal interference (Willis, ibid).We can see this particular episode, then, as a characterisation primarily based on peasant witch-beliefs, but with a significant leavening of learned and topical allusions to appeal to the beliefs of the courtly audiences of the time. As Purkiss explains (p. 207) the effect is of "slippage" between the various definitions of what a witch was supposed to be.

The objects associated with the magic of the Weird Sisters draw closer to the literary sources and to the notions of the scholarly than to rural tradition. The First Witch brandishes a pilot’s thumb from a shipwreck (I:3) - witches were supposed in the scholarly tradition to achieve necromantic powers by such acts of corpse-robbing, and such activities had been expressly prohibited by James’ 1604 anti-witchcraft statute (Harris, p. 40). The ingredients which enter the cauldron in IV.1 are, as we might have come to expect, "borrowed from all over the place" (Purkiss, p. 212); Roman dramatic representations of witchcraft and Continental accounts of sabbath-brews are seem to be the main sources. Willis suggests that the cauldron can be seen as a womb-symbol, another indication of the Sisters' maternal nature, especially in that two of the visions it brings forth are of children (Willis, p. 231). However, the ingredients appear to be chosen indiscrimnately from several different traditions, for sensational, grotesque effect rather than consistency with any recognizable witch-belief – a mockery of the trappings of peasant witchcraft for the amusement of their social betters. Purkiss, however, notes a connection with the domestic nature of popular witch-beliefs in the method in which the prophetic spell is put together – it is, she claims, a parodic recipe, with such instructions as "Cool it with a baboon’s blood/ Then the charm is firm and good" (Purkiss, p. 212). As previously stated, before its association with witchcraft via this play the cauldron was simply a peasant cooking vessel, and in its perversion to a loathsome use we catch here again an echo of the tradition of the witch as the enemy of domesticity.

In summary, then, the Weird Sisters of Macbeth are, in the words of Harris, "founded on, but transcend […], ancient and contemporary beliefs about witchcraft" (p. 44). I have shown how they are constructed from a sometimes uneasy mixtures of popular, scholarly and classical versions of witchcraft. In appearance and manner they resemble the old women who were mainly accused of witchcraft in the rural trials. In their speech and in their actions they mainly draw on the stories of what peasant society believed witches to do, but also appeal to the conceptions of witchcraft that were common in the Jacobean court – to use the distinction of the first section, they are mainly "churn-cursers", but have just enough allusion to the witches who were thought to have committed high treason in Scotland to pique the interest of the King's "crusading paranoia" concerning witchcraft. In addition, they bear some resemblance to earlier dramatic representation of witches in cosmetic detail, but are very different in both form and function from the earlier classically-based or comical figures, in that they are strongly influential over the plot even when not on stage. It is possible to argue, moreover, that their dramatic function is as overseers of the unfolding of the plot, a theme emphasised by the way in which they are referred to far more often than they appear, which has been pointed up in productions where the Sisters are a constant shadowy presence in the background of the play (Rosenberg, p. 15). This portrayal, which fits them into one of Holinshead’s suggested identities, that of "goddesses of destinie", is however at odds with their characterisation, which is by turns grotesque, topical and gnomic, but which tends to mark them as human, if human with supernatural power. It cannot be doubted, then, that the nature of the Weird Sisters is identical to the nature of their prophecies – defying interpretation – and it is my intent to argue in the conclusing section of this chapter that this is no accident. I aim to show that it is precisely because of their diverse, confusing characterisation, appealing to so many different and conflicting conceptions of what witchcraft is, that they are so succesful in their function as the abiding and controlling spirits of the plot of Macbeth, a play based on equivocation, confusion and deceit.

Rosenberg (passim) lists seven different ways that the Weird Sisters can be categorised: as women, witches, sorceresses, agents of the Devil, Fates, projections of Macbeth’s own unconscious, and as symbolic representations. Wills (passim) adds one other way – as representations of the Jesuit priests who were thought to have encouraged the Gunpowder plotters in their unsucessful attempted assassination of King and Parliament. Stallybrass describes them as depictions of the threat of female and peasant power to the patriarchal/monarchical order (p. 206), whereas Purkiss regards them as exploitative caricatures of the real stories and concerns of rural women. About the only thing, it seems, that we can truly describe the Weird Sisters as is polysemic – they can be interpreted in many different ways, all of them accurate in some ways and inaccurate in others. In this concluding section I will seek to explain the lasting appeal and success of the Weird Sisters, to the extent that a literary tradition has been founded upon them, by showing how their very equivocation can be interpreted as a fundamental challenge to the accepted order of the Universe as conceived by the ruling discourses of society. I shall focus on the dramatic effect of both the confusion that surrounds their identity and the ambiguity of their prophecies, and examine how this relates to the fatal trap of self-delusion that Macbeth finds himself caught up in.

Perhaps the most succinct description of the Weird Sisters that it is possible to produce is "liminal figures" (Harris, p. 38) – that is, they inhabit boundaries or grey areas, between the supernatural and the physical worlds, between the nobility and the peasantry, between human and non-human, between man and woman. This fundamental ambiguity even encompasses their name. The first word is consistently spelt "Weyard" in the first folio edition, which is thought to represent the pronunciation of Shakespeare’s time (Rosenberg, p. 12) – except in certain utterances by Banquo and Macbeth, where they are called "weyward". Thus, the very name of the characters is an equivocal pun – they are both "wayward", in the sense of capricious, self-willed and nonconforming, and "weird", in the older sense of the term, meaning involved with destiny. (Indeed, our modern use of the word "weird" to mean strange or uncanny derives from Macbeth - Harris, p. 33.) The Sisters are of such a fundamentally ambiguous construction that even the primary signifier of identity in a patriarchal culture – gender – is equivocal in them. "You should be women,/" says Banquo," and yet your beards forbid me to interpret/ That you are so" (scene I.3). Other sources of ambiguity include Rosenberg’s distinction between "witches" and "sorceresses", which is based on the fundamental question of how much power they have over the course of the play – are they merely foretelling Macbeth’s future, or are they causing it to come to pass? Do they hang, as Harris (p. 35) suggests, as "Furies" over the play, tormenting Macbeth on into damnation - malevolent, omnipotent mother-figures, in Willis' analysis (p. 218) - or are the supernatural elements of his descent such as the airdrawn dagger (I.7) figments of his own tortured psyche, for which the Sisters become the scapegoat? How much responsibility do the Sisters have for the events of the play, and can we accurately describe them as evil, in the diabolic sense of the term that the demonologists of the time would have used? Certainly, they appear to have a delight in confusion and mischief (Hunter, p. 39), and Macbeth definitely considers them to be so – he probably means the phrase "night’s black agents" (IV.1) to refer to them, and describes his second visit to them as arising from the need to "know/ by the worst means, the worst" (III.4). However, the Sisters, who state that "fair is foul, and foul is fair" (I.1), can be seen as indifferent to good or evil in general - the most radical example of their ambiguity. A modern psychological analysis might describe Macbeth as "projecting" his own faults onto those he wishes to make responsible. As Rosenberg (p. 16) points out, the Sisters’ delight in mischief is tempered by their apparently favourable interest in Macbeth himself. In short, much as the Sisters do "a deed without a name", they are themselves nameless – or rather, they have many names, none of which is correct. The reason why they provoke such consternation in the other characters, and subsequently in the audience, is that it is as impossible to decipher exactly what they are, or what they are doing.

Stallybrass interprets this radical ambiguity as the main reason why the Sisters are such potent figures of evil in the play – they evade, by their equivocation, the categorisation which is necessary for a patriarchal/monarchical social system, based on distinctions between nobility and peasantry and between man and woman. They do not even explicitly take sides in the battle in I.1 and seem indifferent as to whether Macbeth succeeds or not, as we might expect from members of the lowest class of society (Willis, p. 217). They are thus simultaneously antagonistic figures in both major discourses of Witchcraft – they are simultaneously bad subjects, plotting the downfall of the legitimate monarchy, and bad neighbours, who take violent revenge for petty slights and cook their spells out of everything that is considered foul. In addition, they are bad women, in that they dare to give advice leading to the overthrow of Duncan, the rightful patriarch. Accordingly, Macbeth's treason is gendered feminine in the play – Willis suggests that civil strife in Shakespeare's plays is often linked to a conflict between feminine and masculine values (p. 172), and Macbeth’s cowardice in murdering Duncan under trust is associated with Lady Macbeth’s "unnatural" influence over her husband. As well as being the enemy of everything that is "good" by the prevailing discourses of the time, they are simultaneously the most marginalised figures of society, old peasant women (Rosenberg, p. 4). Thus, as Turner (p. 21) says, their ambiguity becomes a universality – because they are neither one thing nor the other, because they are so diversely constructed as to be simultaneously the antitheses to all the conceptions of good order held by the various sections of society, they are universal symbols of "revolt against the natural order" (Stallybrass, p. 206). The natural order as presented in the play is the God-fearing, patriarchal monarchy of Duncan, which the Sisters are depicted as wishing to overthrow through sheer love of confusion and mischief – and as Turner (p. 18) states, any assertiveness or desire from women which conflicted with the customary demands of subservience and modesty was considered "unnatural". The Sisters are unnatural because they refuse to fit into the marginalised, powerless category defined for them by this natural order, they are "unnatural" – similarly, the thunder and lightening which heralds their entrances, and which they are generally assumed to have provoked, would have been "likely to be perceived by Shakespearean man as ‘unnatural’" (Rosenberg, p. 1). So, the conflation between witchcraft as revolt against God and revolt against the King implicit in the elite tradition is further extended in its negotiation with popular traditions in the person of the Weird Sisters – now, witchcraft is defined as revolt against the very nature of the universe. In this way, although the Weird Sisters never actually commit the crimes of the play, they become perhaps its ultimate villains.

If monarchy is conceived as being a reflection on Earth of God’s rule over the universe, then revolt against the monarchy is itself a crime against nature, and the Witches provoke Macbeth into unnatural acts. This unnaturalness has been pointed up in several productions in which men have played the Sisters, adding strength to Stallybrass’s point (p. 201) that the crime of the Sisters is a crime against patriarchy in particular. Another aspect of the Weird Sister’s crime against nature is their revolt against language. Renaissance thought did not consider language to be a neutral, contingent medium for the transmission of meaning – all things in the universe were considered to have "true names" in some universal language (often assumed to be Hebrew). "To give false names, to pervert language, was a sin against nature," (Wills, p. 95) and the Jesuit’s doctrine that it was permissible to equivocate under oath for Godly purposes was considered blasphemous by Protestants of the time. Therefore, the ambiguous nature of the Sisters’ prophesies, leading Macbeth to interpret them incorrectly and thus lead himself into treason and damnation, would have been considered among their worse crimes. The revolt of the Sisters against the natural order of language is encapsulated by their declaration to Macbeth that they do "a deed without a name" (IV.1) – that is, an action opposed to the natural sphere to which language pertains. Macbeth’s own treason and murder can be considered in itself a nameless deed – not only because of its "unnatural" nature, so that none might dare to name it, but in that the Macbeths themselves refuse to name it. We never get a clear picture of the events leading up to the murder of Duncan, since we only hear about them during the conversation of the Macbeths, and they consistently refuse to explicitly name their desires, their hopes, or the means by which they aim to achieve them. "They speak of it mutedly, obliquely, with euphemism, and with extensive reliance upon it and other pronouns without antecedents" (Jorgenson, p. 47). A modern psychological interpretation could even say that Macbeth is "in denial" – his better nature steadfastly refuses to name what he is thinking and doing, presumably in the belief that what is not named does not exist. Even if the Sisters are not causing Macbeth’s descent in reality, they must receive the blame if Macbeth is to be absolved of responsibility for his own actions. In such a way, the Sisters fulfil the function of scapegoats which witches consistently fulfilled in both elite and popular discourses throughout the Renaissance period (Turner, p. 16), and the ambiguity which the play relies on is enhanced – the inability to name the nature of the Weird Sisters is connected to the inability of Macbeth to face the reality of his crimes.

Purkiss suggests that the difficulty in interpreting what the Sisters "really are" is offered as an appeal to James I’s interest in hunting down and exposing witches (p. 207). She makes the point that unlike, say, Ben Jonson’s Masque of Queens, in which the witches are transparently ridiculous and fake, the Weird Sisters are constructed to as to require both the audience and the characters to interpret them correctly, which none of them succeed in doing. This shows up the inutility of attempting to use the word "witch" in terms of one uncontestable definition. This error, for example, leads Wills (p. 46) to attempt to interpret them solely by the scholarly definition of witchcraft as a pact with the forces of darkness, an approach which short-circuits their primary characteristic, and the feature that makes them so fundamentally important for the process of the play – that is, their ambiguity. "They arouse our horror; they tempt us to call them witches to relieve that horror; and then, because we cannot do so, our horror redoubles." (Turner, p. 20) However, this ironically shows up the fact that the modern literary definition of "witch" is based directly on the Weird Sisters. We can say that because they are such sui generis figures, and yet subsume such a variety of different stories and beliefs about witches into themselves, the literary tradition of the Witch had to redefine itself around them. Thus, the Weird Sisters cannot be unambiguously called "witches", because they are constructed from so many different sources with differening definitions of what a witch was, but ironically they consequently become the new definition of "witch" for literature in English.

Not all commentators agree in praising the ingenious ambiguity of the Weird Sisters and their role in the play, however. Purkiss, who previously describes the construction of the confessions in the witch-trials as a "negotiation", rejects the use of that term for the construction of the Weird Sisters, referring to them instead as "an awkwardly compressed mass of diverse stories… popular culture [buried] under a thick topdressing of exploitative sensationalism" (p. 207). In Purkiss’s account, the Weird Sisters are not so much constructed as jumbled together, and offensively so at that. She regards them as a purely sensationalist construction, "pandering shamelessly to the novelty-hungry news culture of Jacobean London" (ibid.) She takes the sexualisation of the story of the sailor’s wife and the garbled, grotesque nature of the cauldron scene as evidences of male appropriation and exploitation of women’s stories and fantasies, as well as elite mockery of the very real fears of peasant society. Macbeth’s use of witch-stories of the popular tradition, she says, "looks less appealing once the listener is conscious of the female voices silenced" (p. 207) – and presumably the voices of rural society as well. She makes the point that whereas Stallybrass praises Shakespeare for refusing to reduce the Sisters to old village women (Stallybrass, p. 195), this assumes that old village widows are less dramatically interesting than stereotypical evil hags (Purkiss, p. 208) – Purkiss rejects this assumption, and suggests that in that way a possible source of rich dramatic interest is lost for the cause of "cheap sensationalism".

While these complaints are no doubt justified, the question still remains of how such shallow, sensationalist stereotypes could have become the basis for an entire literary tradition of witchcraft. Purkiss, however, possibly answers her own point when she says that the use of stories of village witchcraft in Macbeth "depends on the stories and events half narrated in them remaining utterly unreadable and inscrutable" (p. 208). I believe that I have shown in this section that the abstraction of the Weird Sisters from the traditions from which they were constructed, while it may well have been for sensationalist or elitist reasons, is precisely why they have survived and formed the basis for our modern tradition – that as abstracted figures of revolt against all that is held to be natural, they achieve a dramatic power that transcends the social conditions of both their time, the time of the play and the time of modern readers. The silencing of women’s voices and their exploitation to express patriarchal anxieties that Purkiss complains of may be regrettable, but the Weird Sisters cooking up the destinies of kings in their cauldron on the blasted heath have survived far longer in literature to trouble the complacency of patriarchy than a faithful portrayal of rural women’s domestic anxieties of the time could have. Whether conversion into a new stereotype is an acceptable price to pay for these women’s stories to pay for literary longevity in a patriarchal tradition is a question that can only be decided by each reader for themselves.

My argument to this point can be encapsulated as follows: prior to Macbeth and the other plays of the Jacobean witch-vogue, the literary depiction of witches was either as stock figures from the classical tradition, or as harmless comical peasant figures. This contrasted with a social milieu in which the concept of witchcraft was taken very seriously indeed. Those cultured, literate elite who did not dismiss it out of hand as the ignorant fancies of peasants considered it the worst of crimes, not only a revolt against God but (by the doctrine of divine right of Kings) against the state. Among the peasantry, on the other hand, the witch was the enemy of good housekeeping and domesticity – Purkiss’ "powerful fantasy" for women to ease anxiety about their own sphere of responsibility by means of scapegoating. In both cases, it was the most marginalised members of the community – old women, ugly and deformed, incapable of making an independent living and thus forced to rely on begging – who were imagined by a resentful community to be the enemy of the very social order itself.

The Weird Sisters of Macbeth are, as Purkiss states, fundamentally a sensationalist creation – they have been thrown together from peasant tradition, the paranoid fantasies of the new Jacobean monarchy, and a few half-remembered pieces of classical witch-lore. Why, then, did they go on to become the image that most modern readers call to mind when the word "witch" is mentioned? I have argued in this chapter is that their diverse, inconsistent construction creates (perhaps accidentally) a fundamental ambiguity in their nature – because they are created out of pieces of the deepest fears of every section of society, they effectively become a symbol of universal rebellion against what was considered "the natural order of society". Whatever the politics of their construction, they would not have survived if they did not reflect a need to create a universal scapegoat – a rebel against God, King, house, child, social order and even the very language. Because they are women (if they are women) without names who commit deeds without names, they could have been the most frightening things of all for the literate readers of the time, who based their identity and social position on their ability to explain the universe with language and logic, without recourse to peasant superstition. The Weird Sisters’ influence over Macbeth, then, is a fantasy of what could happen if peasant superstition were to re-emerge from the margins and dethrone the learned, patriarchal, monarchical order of the universe. This anxiety about the power of peasant superstition, and the old women who embody it, will be explored in the fairy-tale analyses of the next chapter.

The Witch-character as exemplified by the Weird Sisters, therefore, shows two faces. To those privileged by the dominant paradigms, the "natural order" of society, she is everything bad condensed into one package – the bad mother, the anarchist saboteur, the agent of Satan, the dirty peasant, the total nihilist who threatens to unleash mindless chaos on all the spheres of human life. Conversely, to those excluded or marginalised by this "natural order", she encapsulates the sources of power available to those on the margins of society. These are the powers associated with dirt, with woman, with poverty, with irrationality, with anti-Christianity and with rebellion against authority - in short, all those forms of power that those in power try to not even name, let alone think about, and thus hope to wipe from existence. The literary Witch-character, therefore, became defined as the person capable of committing "a deed without a name". In future chapters I aim to show how those both sympathetic and unsympathetic to the idea of the Witch have given many different answers to what that deed might be.

In the previous chapter, we explored the way in which the Weird Sisters of Macbeth can be interpreted as a compromise or negotiation between the differing "elite" and "popular" discourses of Jacobean witch-beliefs. In the centuries following the Jacobean era, the conception of witchcraft as a pact with demonic powers, encouraged by the Church and held by a faction of the educated elite in early modern times, succumbed to the spread of Enlightenment ideas, and the "sceptical" discourse which considered the Witch nothing more than peasant superstition prevailed among the educated classes. Thus, by the nineteenth century the figure of the Witch was no longer taken seriously in elite circles, and the "demonological" discourse of Witchcraft became regarded, outside of the Church, as a superstition in itself. However, the other "superstitious" discourse of Witchcraft - that based on rural witch-beliefs - enjoyed something of an afterlife, becoming integrated into aristocratic and bourgeois circles via its transformation into the subject material for children's tales.

This chapter will focus on works by three collectors and/or authors of such tales, whose works have continued in popularity up to the current era . Firstly, the Kinder- und Hausmärchen (Children's and Household Tales) were German folktales collected and edited by the brothers Jacob and Wilhelm Grimm and first published in 1812 - Wilhelm released further editions of the tales up until 1857. This collection includes many of the best known children's tales of today, such as "Snow White", "Hansel and Gretel" and "Rapunzel". The latter two will be considered in some depth in this chapter, along with certain less well-known tales from the collection which are interesting for their portrayal of witches. The analysis for this chapter will be based on Ralph Mannheim's translation of the 1857 edition as Tales for Young and Old (1978) which was described by Maria Tatar as the first "reliable English version" (Tatar, p. xxii). Secondly, the stories of the Danish writer Hans Christian Andersen, written in the middle years of the nineteenth century, include both adaptations of folk material and works of Andersen's own design based on literary sources - we will investigate one tale from the first class ("The Tinderbox") and one from the second ("The Little Mermaid" - Conroy, p. 251, identifies the literary sources for this tale). The third author that will be considered is the American Lyman Frank Baum, who wrote a series of children's books concerning the Land of Oz, of which the first, The Wonderful Wizard of Oz (1900) has become, via the MGM film of 1939 starring Judy Garland, a major cultural icon. Indeed, as Diane Purkiss notes, Margaret Hamilton as the Wicked Witch of the West in the film is one of the two "dominant images of the Witch", along with the Weird Sisters of Macbeth (Purkiss, p. 276). Thus, to trace the image of the witch in popular culture through literature, the fairytale witches must be our second primary source.

This chapter accepts the definition of the "fairy tale" used by Bottigheimer (p.9): a tale based in a setting "isolated in time and place" from the setting in which it is told, which uses magic or magical beings among its major plot devices and without a religious "teleological focus on death with its heavenly or hellish consequences". We can therefore say that the characteristic of the fairytale is that it is based in an Otherworld, whether that of the past or that of "Never-Never Land" - consequently, there is a strong tendency to analyse the fairytale ahistorically, as embodying archetypal (from a Jungian point of view) or psychological (as Freudian commentators such as Bruno Bettelheim would have it) truths. However, commentators such as Zipes and Bottigheimer point out that this approach neglects the mediating effect of the different treatments in the retelling of the traditional material. As Tatar also explains (p. xix), ahistorical analyses of the tales often base their arguments on features that were added to the tales by their literary transcribers rather than those elements from the oral tradition, and thus confuse the two. This chapter, therefore, will examine the fairytales as mediated texts - text based on oral tradition, or drawing inspiration from that tradition, but mediated by their mainly male, middle-class transcribers/authors, in much the same way that we have argued that the Weird Sisters are the result of Shakespeare's mediation between popular and elite witch-beliefs.

For example, Bottigheimer mentions that these tales have been officially associated with female tellers (p. 10), a theme taken up in great detail in Warner's work which will be explored later, but here it is sufficient to note that these "fairy stories" or "Mother Goose tales" were in the 19^th century almost always mediated through male transcribers, editors and publishers. So in this sense, the stories are a mediation through male interpreters of a tradition coded female - and they are also of mixed origin as to their class basis. Purkis (p. 277) notes that "folktales still speak to us in a female voice", and connects this female voice to the female voices of the early modern period whose "fantasies about witches" were explored in the previous chapter as one of the main sources of the Weird Sisters. However, this female voice is also a rural peasant voice, as previously explained, and the nature of the class mediation of the folktale material is also important. Purkiss also mentions (p. 277) "nurses and governesses bringing the concerns of a different social class into the castle" - the folktale thus becomes the nursery tale. Therefore, there must be a mediation to make the tales understandable, explicable or morally appropriate for children. The latter seems to have been the main thrust behind Wilhelm Grimm's progressive editing of the tales - Tatar notes (pp. 19-20) that one of the main aspects of the Märchen that disappeared between the 1812 and 1857 editions was virtually every reference to pregnancy, and speculates that this may have been in response to the slow sales of the first edition, associated with complaints about its suitability to be read to children.

The other important mediation that comes in the transformation to the literary fairy tale, in the hands of the male transcribers/editors, is the pedagogical slant which the tales were given. "Almost all critics", writes Zipes (p.3), "agree that educated writers [of literary fairy tales] purposely appropriated the oral folk tale and converted it into a type of literary discourse about mores, values and manners so that children would become civilised according to the social code of that time" - and further, (p. 9) "the purpose of the tale from the beginning was to instruct and amuse, that is, to make moral lessons and social strictures palatable". However, this often required an ideological adaptation on behalf of the literary transcribers, from concerns of the rural peasantry from whom the tales came to those of the bourgeois parents anxious to raise moral children. The prime example which Purkiss uses (p. 280) is of the Witch's house in "Hansel and Gretel", which was originally made from bread, rather than gingerbread - hence the tale shifts from rural anxiety on how to feed children during famine to middle-class concern about the "greediness" of children. Purkiss explains that as the tales moved from rural subsistence life to the urban bourgeoisie, the anxieties of the parents changed from how the children were to be fed to how they were to be brought up as moral, civilised beings - Zipes mentions (p. 23) that the imposition of strict behavioural controls on adults by the bourgeois code of civilities gave rise to an image of children as "corruptible innocents", and produced the pedagogical obsessions of fairytale collectors/authors such as the Grimms and Andersen. This aspect of the literary fairytale is enthusiastically supported by Bettelheim - the German title of his work translates to "Children Need Fairytales" (Kinder Bräuchen Märchen), and it is his contention (p. 4) that the tales contain valuable psychological lessons to help the child cope with its own subconscious impulses. However, as Zipes points out, the lessons taught by the literary tales are shaped by their adaptors - and as Purkiss notes, these change according to the differing anxieties of the adults who write and read fairytales of each different era. Zipes notes (p. 8) that the world of the older fairytales is "curiously amoral" - it may be possible to argue that morality is a luxury which can be afforded when the question of how to feed one's family is no longer pressing.

This process of adapting morality was consciously engaged in by the Grimm Brothers in particular - as Bottigheimer notes, "in their introduction the Grimms themselves called this cheaply printed little book a childrearing manual, an early hint that it offered paradigms for appropriate behaviour" (p. 4). In addition, the works by Tatar and Bottigheimer referenced track the effects of Wilhelm's progressive editing intensify his own particular moral preferences in the tales (see esp. Bottigheimer, ch.6). As for Hans Christian Andersen, Zipes in particular interprets his tales as strongly supporting the social and economic power structures of his day - in particular, "The Tinderbox" as extolling social Darwinism, and "The Little Mermaid" as warning against the dangers of upward social mobility. This, of course, may be seen a consequence of his strong social analysis of the tales - Bettelheim, for example, interprets royalty in fairytales as mainly representing the absolute power of parents over children (p. 205). However, the social analysis seems more fruitful for an understanding of the tales as historical artefacts. Baum, on the other hand, explicitly eschews this moralistic project - in his introduction to The Wonderful Wizard of Oz, he presents that book as "a new kind of tale" which seeks to entertain only and eschews "all the horrible and blood-curdling incident devised… to point a fearsome moral to each tale" (p. 3). The difference apparent in Wizard by the removal of the "civilising" imperative will be explored in greater detail.

It is the task of this chapter to show how the mediation of the primarily female and rural sources of witch-stories through the Grimms, Andersen and Baum differs from that of Shakespeare, mainly as a function of the different social and ethical positions of the authors. These are all tales dealing with an Otherworld of some description, and of course the Otherworld, or more correctly its intersection with our own, is the traditional place of the Witch, the liminal sphere inhabited by the Weird Sisters. Likewise, this chapter will examine the intersection of the fairy-tale forests with the desires of 19^th century middle-class male writers to "instruct and/or amuse" children. These works that this chapter will examine have continued to be commonly read to and by children up to the present day, and thus the depictions of witches therein can be seen to be formative influences on the figure of the Witch as envisaged in popular culture today. The Kinder- und Hausmärchen contain several tales featuring witches as primary antagonists, and construct a Witch character which, much like the character of the Weird Sisters, is made up from a negotiation between the various types of wicked female who would be recognized by the audience of the time - the wicked stepmother, the woodland hag in league with the dark forces of nature, and the diabolical tempress who offers forbidden knowledge (often sexual) at an insupportable price. Much the same kind of Witch is also presented in the tales of Andersen, particularly "The Little Mermaid", whose Sea Witch is nothing more than the "hag in the wood" transplanted to a submarine domain. This Witch as "omnibus wicked woman" also appears in The Wonderful Wizard of Oz. This chapter, therefore, will be structured as an investigation of the three main traditions from which the antagonistic fairytale Witch can be said to be constructed. However, in connection with Wizard I also aim to show how Baum, uniquely among the authors studied, also constructs a positive inversion of this character - the Good Witches of North and South, thus creating a beneficient concept of female power which can also be applied to Dorothy, the book's protagonist. It is my contention that this reclaiming of the figure of the Witch that can also be detected in the revisionist Witches presented by recent works of science fiction and fantasy that will be explored in the second half of this thesis.

As mentioned in the previous chapter, Purkiss (p. 93) interprets the witch-beliefs of early modern rural women as constructing the Witch as "antihousewife or antimother" - an evil usurper of the woman's role in rural society. Therefore, it would not be surprising if this image of the Witch as the domestic usurper was not carried over into those fairytales based on rural folktales. However, the mediation required to make the tales conform to the needs of tales for the children of the 19^th century urban bourgeoisie adds emphasis to the question of the Witch as the "bad mother" - which, as will be explored further, was often changed in the tales to "wicked stepmother". In this section I will avail myself of two different perspectives from which to analyse this facet of the fairytale witch - the ideas of rural women analysed by Purkiss and reported in the preceding chapter, and Bettelheim's Freudian attempt to explain why a child might need the concept of an evil mother.

Willis suggests (p. 37) that the accusations of cursing in the record of early modern witch trials can be seen as evidence of women's verbal violence towards each other - similarly, Purkiss's epilogue on the fairytale Witch emphasises those features of the character which can be traced to "the female voice whose fantasies about witches are visible in [witch] trial depositions" (p. 277). In a society where feeding the family was not only a female occupation, but one fraught with danger in a subsistence economy, the mother might well be expected to go without food herself if necessary for her children's sake - Purkiss suggests that the cannibalistic Witches of the tales can be seen as the embodiment of the guilt of mothers who put their own needs before those of their children (p. 279). In offering food the Witch also "symbolically assume[s] the role of the child-victim's mother" - in Purkiss' example, "Hansel and Gretel", the cannibal-Witch entices the starving titular children with a house made of "bread and cakes", luxury foods in subsistence rural society. The Witch thus seduces and destroys the children whose parents are unable (or unwilling) to feed them themselves, and thus punishing the "real" parents for their failure of parental care. Thus, we can see the witch as the locus of the mother's anxieties about her own fitness for childcare. Furthermore, this image of the Witch as "antimother" can be connected not only to women's anxieties, but back to children's fantasies connected with parental power, and desire for revenge (Willis, p. 48). Purkiss (p. 119) cites a mother who discovers that, according to her children, "after I put them to bed and kiss them goodnight, I go out of the room and remove my mask and clothes, and reveal myself to be a witch". Bettelheim further suggests that children must envisage the mother who loves them and the mother who punishes them as separate entities (a real mother and a false mother - p. 69) - and thus, the triumph of oppressed children over the Stepmother/Witch in the fairytales can be a "fantasy satisfaction" of children's resentment of their parents (p. 52). Accordingly, Tatar suggests that the witches of Grimms' tales are "so pitiless that they fail to elicit pity when the tables are turned" (p. 182) - they are deliberately constructed so that it is impossible to feel sympathy for them in their demise. In much the same way, Wolstenhome suggests that Dorothy's killing of the Wicked Witches in Wizard of Oz is specificially designed to relieve Dorothy of all blame - her house crushes the Witch of the East, and she is apparently ignorant of the effect that a bucket of water will have on the Witch of the West. Wolstenhome comments (Baum, p. xxxix) that in Oz, "the bad mother will not stay dead" - reappearing as Aunty Em as well as the two Witches - which further absolves Dorothy from guilt for abandonment and murder.

The aspect of the Witch as "bad mother" for children is also expressed in the way in which she often forces her child-victims into domestic servitude - this is the complaint of Little Brother and Sister about their wicked stepmother, it is what the witch in the woods forces Gretel into, and also what the Wicked Witch of the West inflicts upon Dorothy in Oz (Baum, p. 151) - the latter also punishes her slaves the Winkies by beating them with a leather strap (p. 142), a traditional instrument of domestic punishment. The wicked stepmother figure, then, enters the tales as an "evil substitute" for the real mother - just as Bettelheim reports (p. 134) about a girl who was convinced that her loving mother was occasionally replaced by a Martian impersonator who would punish the girl. As Warner points out (p. 206), the high rates of death in childbirth for rural early modern women, and the usual quick remarriage of their widowers, meant that the presence of the stepmother in the rural home was more a rule than an exception. It is important to note that the stepmother, as substitute and "usurper" mother, thus becomes a mother figure that it is safe for the child to hate. the Witch and the "wicked stepmother", in fact, tend to be conflated and are sometimes explicitly identified throughout the tales, and it is possible to argue that in one sense, the Witch is actually the Stepmother in her unmasked form.

The role of the idea of the Deep Woods in the depiction of fairytale witches will be explored in the next section, but some words can be said here concerning tales such as "Hansel and Gretel" or "Little Brother and Little Sister" (Grimms' Tales no. 11, Mannheim p. 41). In these tales, the wicked stepmother abuses the titular children within the home, either starving them or forcing them into domestic servitude - but when the children escape from her into the woods, she reveals her true power as a Witch. The wicked stepmother in the latter tale "had seen the children go away and had crept after them stealthily, as witches do, and had put a spell on all the springs in the forest", a spell which eventually succeeds in transforming Little Brother into a deer. The stepmother who turns Hansel and Gretel out is not explicitly identified with the witch who attempts to consume them, but after Gretel burns the witch in her own oven, the children return home to find that their stepmother has coincidentally also died (Mannheim, p. 62). The Witch in "Hansel and Gretel" is a different kind of bad mother than the callous stepmother - she appears to be a kind old woman, feeding the children and settling them down in clean, white beds, but this is nothing more than bait in a trap. The Witch is not only voraciously cannibalistic, but is able to dissemble and disguise her true nature to snare the unwary - this can be identified with the child's resentment of the stepmother as a stranger usurping the mother's place. The "family drama" aspect of the Stepmother/Witch stories is often intensified by the presence of the biological children of the stepmother, as in "Little Brother and Little Sister" and "White Bride and Black Bride (Grimms' Tales no. 135, Mannheim p. 461). In the first tale, when she learns that Little Sister has become a queen, "envy and jealousy" in the heart of the stepmother prompts her to set her own daughter, "ugly as the night" up in her place (p.44). The stepmother bent on revenge and putting her thoroughly undeserving biological daughter in the rightful place of her stepchildren is also the main plot device of the latter tale - and in both tales, mother and daughter are both severely punished for their crimes at the end. This may have been expected to appeal to the resentful children of remarried fathers.

The other main way that the fairytale witch operates as an "antimother" is that she fosters dependence, rather than independence. As previously mentioned, the fairytale Witch seduces to entrap, much like the classical Circe of the Odyssey. She encourages Hansel and Gretel to eat her house, thus leading to what Bettelheim's psychological reading views as "a regression to gorging dependence" (p. 161), and this dependence becomes literal in the children's transformation into meat animal and kitchen slave respectively. The cannibalistic nature of the Witch can conversely be seen as a trope for maternal possessiveness (Tatar, p. 140) - the "bad mother" will be prepared to cannibalise her own children for her own ends. The witch in "Rapunzel" (Grimms' Tales no. 12, Mannheim p. 46) is literally possessive of her titular de facto stepdaughter, keeping her under lock and key - it can also be argued that the stepmother in "Hansel and Gretel" is figuratively cannibalising her stepchildren by abandoning them, as the Witch literally attempts to consume them. In other situations, the Witch puts herself in a position of power over the protagonists by having something that they need or desire, but demanding it at an extortionate price - Bettleheim (p. 94) sees the Witch in this aspect as a "projection of wishes and anxieties", a synthesis of both all-good and all-bad mother-figures. Often, reflecting the rural-subsistence origin of the sources, the need is simply for food, as in "Hansel and Gretel", or as in "Rapunzel", where the witch demands the daughter as the price for the food the mother craves whilst pregnant. In "Six Swans" (Grimms' Tales no. 49, Mannheim p. 171), the King is forced to marry the daughter of a witch (of course, a witch herself), as the price of being shown the way out of the Deep Woods, thus leading to the typical "stepmother" drama later in the tale. This trope also occurs in the works of the other authors. The Sea Witch in "The Little Mermaid" demands the mermaid's most valued possession (her voice) in return for her heart's desire (human form); later, she demands the beautiful hair of all the Mermaid's sisters as the price for giving them the chance to rescue their sister. As previously mentioned, the Wicked Witch of the West enslaves Dorothy, and her main interest seems to be enslaving as many people as possible - she does not kill the Cowardly Lion precisely because she plans to "harness him like a horse, and make him work" (Baum, p. 148). As Bettelheim explains, the fairytale witch can be seen as a symbol of the "interfering older woman" who does not want her children to grow into adult life, including freedom from dependence and adult sexuality (p. 297) - Rapunzel's "stepmother" punishes her precisely for having begun her sexual life (Mannheim, p. 48). Warner, on the other hand, notes that old women were traditionally the sources of disapproved knowledge, especially in the realms of sexuality (p. 404) - in this, we can perhaps see the resentment of the rural society against a figure they despise, but is necessary to the community, as well as a child's resentment of the parent who, while loving, has absolute control over them and is necessary for their survival.

"[The villainous stepmother] takes on a single well-defined function in fairy tales - one that … magnifies and distorts all the perceived evils associated with mothers… [She] reemerges in the woods as a monster equipped with powers far more formidable than those she exercised at home," concludes Tatar (pp. 142, 146). Tatar also mentions that these are also sometimes not distinct from other types of monstrous women, such as ogresses (p. 137). This strong identification has continued in certain modern adaptations of some other fairytales which feature a wicked stepmother for example, in the Disney film of Snow White (1950), the evil Stepmother/Queen disguises herself as a beggarwoman/witch to deliver her stepdaughter the poisoned apple. The stepmother in the literal sense is, of course, literally a replacement for the "true mother" - but moreover, Warner points out that the French word for "stepmother" also means "mother-in-law" (p. 219). The mother-in-law is another figure which early modern women might have seen as their antagonist in the family home, seducing and alienating children's affection and perhaps attempting to exclude the wife altogether. This resentment can be seen in the story of the Tin Man's origin in Wizard (p. 60), where the jealous mother of his sweetheart pays the Wicked Witch of the East to enchant his axe so that it chops him into pieces - again, we have the identification of the Witch and the jealous older woman. We can see the fairytale witch, then, as a generalised "evil other woman" in a domestic drama - the scapegoat and embodiment of guilt of the rural mother, and also the "false" mother-figure whom it is safe for children to hate. However, as Warner notes (p. 239), in the process of mediation this trope for inter-female rivalry was often usurped by male editors/writers - the rivalry between women is often used in the tales to glorify the male figures, or to absolve them from blame. For example, in the earliest versions of "Hansel and Gretel" the decision to abandon the children is made by both parents together - by the 1857 text, however, it is at the stepmother's insistence, with the father agreeing only reluctantly (Mannheim, p. 56). In later sections we will explore further how the male mediators/authors of the tales constructed the Witch as the embodiment of all the qualities considered unacceptable for the women of 19^th century European society.

As mentioned above, the distinction made in the tales between the wicked Stepmother and the Witch tends to be the distinction between the domestic sphere and the Deep Woods. (Compare the MGM film of The Wizard of Oz, where the interfering old woman back in Kansas is identified with the Wicked Witch of the West.) The Witch can be therefore seen as a combination of personality and place, only attaining her full power in her appropriate place - hence, this section will explore how what we may call "the Deep Woods" are presented in the tales as the domain, and source of power, of the Witch. The relationship between the rural community and nature was far from idyllic - as Warner points out (p. 298f), it is only with the rise of industrialised society in the last few centuries that nature has become considered welcoming to people, or animals considered playmates. The intersection of human society with this unforgiving realm has always been of importance to rural pre-industrial societies - hence the shaman-priests of nomadic society, who take on the character of the "totem animal" of the tribe and act as an intermediary between the two worlds, attempting to ensure the success of the harvest or the hunt. The shaman is a liminal figure, neither belonging to one world or another - and, as mentioned in the previous chapter, the Witch has traditionally been considered another such liminal figure. "The transgressive has access to the unknown and ineffable", as Warner notes (p. 124), and liminal figures, belonging to neither one world nor the other, are certainly transgressive - indeed, Starhawk (1989: 236n.) suggests that the German word for "witch", hexe, and the English "hag" can be derived from common Germanic roots signifying "woman on the hedge", a boundary figure. I will argue in this section that the Witch as "hag of the woods" is considered a figure not entirely human, not entirely animal, but whose power consists in being able to move between the two worlds.

As will be explored in the next section, women's speech, or control thereof, is of vital importance in several of the tales, and one of the aspects of speech is conjuring - the ability to enforce one's will by properly phrased commands upon other people or natural forces. Bottigheimer notes that "the oldest German literature, the pre-Christian Merseburg Spells… bear witness to an early and perhaps continuous belief [in] a peculiarly female ability to control, direct or affect natural powers," - and further, that "the intimate link between women and natural powers may be viewed as part of a tacit pact between ancient Germanic society and women, that natural processes be understood to be under feminine control, while acts of aggression and governing fall to the male sphere" (Bottigheimer, p. 43). Hence, we have a tradition in German culture of the sphere of nature being coded female - Bottigheimer goes on to note that in contrast with male sorcerers, the female witches of the tales do not so much overpower other individuals with magic, but instead evoke the power of magical objects or forces. It is interesting to note, as Tatar does, that the Witches in the Grimms' tales are virtually never heard to cast verbal spells, or conjurations (p. 41). Purkiss points out that, according to records of early modern witch trials, this is exactly what we should expect - "mostly their acts of magic are silent and unseeable, detectable only by half-hidden signs" (p. 277). Thus, the stepmother in "Little Brother and Little Sister" enchants all the brooks in the forest to entrap her stepchildren, although her words are never reported (Mannheim, p. 44). Every example of overt magic cast by Witches throughout the tales is similarly intimately linked with the magical power of place, and of wildlife. The witch gains control of Rapunzel by means of her parents' theft of the herb of that name (Mannheim, p.46) - this principle of "identification" is a basic concept of magic, and the power of the ancient shamanic figures came from precisely their identification with animals and plants. The Witch in "Two Brothers" (Grimms' Tales no. 60, Mannheim, p. 215), in fact, can be seen as the very embodiment of the Deep Woods - when she is killed, the impenetrable forest miraculously opens to let the protagonists through (p. 239). Also, as mentioned previously, the wicked stepmothers of "Little Brother and Little Sister" and "Hansel and Gretel" become, either explicitly or implicitly, the Witch when their stepchildren enter the forest. The Witch as mediator with the powers of the Deep Woods is so strong that even in Andersen's "The Little Mermaid", a tale that takes place mostly on the bottom of the sea, the Witch figure actually lives in an accurate submarine replica of the standard Witch's Wood:

Here Andersen ingeniously substitutes semi-mobile polyps for the traditional animated trees of the Witch's forest (which turn up in The Wizard of Oz: Baum, p. 223); the Sea Witch has toads and snakes for familiars, just as we might expect of a witch on land, and is also a grotesque variation on a mother-figure. It appears necessary that the Witch must be in her wood to have her power - even if that wood must be on the bottom of the ocean. Again, the power of the Wicked Witch of the West in The Wizard of Oz is exercised through mainly through animals - her command of crows, bees, and finally the Winged Monkeys (Baum, pp. 146-50). As mentioned in the previous section, the hostility to the Witch can be seen as the resentment of the rural community of the "necessary outsider" figure who holds power over them - for example, the house of the Witch in "Hansel and Gretel" is full of gold, another way in which she is marked out as unnatural (Bottigheimer, p. 131). As the rise of industrial and urban society progressed through the 19^th century, the strong connection of rural communities to the land faded and, accordingly, the need for shamanic figures - this can possibly be associated with the number of tales which seem to be taking revenge on the Witch.

It is important to note throughout that, much like the shamanic figures referenced, the Witches in their guise as "Hag of the Woods" are identified with wild nature to the extent that they are no longer entirely physically human. For example, the Wicked Witch of the West is "so wicked that all her blood had dried up" (Baum, p. 152). The witch who entraps Hansel and Gretel has "red eyes and a keen sense of smell like an animal" (Mannheim, p. 59); Little Brother and Sister complain that their stepmother treats the dog with greater kindliness than her human stepchildren (Mannheim, p. 41). In "Two Brothers", the old witch who petrifies one of the brothers and his animal companions perches in a tree throughout, and tricks the brother into enchanting his own animals with a stick that she throws down . The way in which the witch "dragged him and the animals to a ditch" amid "peals of laughter" (p. 229) further suggests that the petrification was done out of little but mischief. This suggests another source for the Witch in her aspect as wild woodlands sorceress - this love of mischief is similar to the elves and fairies of traditional European lore, who were also almost, but not entirely, human, had power over animals and loved mischief. Similarly, the titular character in "The Old Witch", a tale left out of the Kinder- und Hausmärchen but republished in Michaelis-Jena (p. 146), requires the protagonist to teach snails to dance - an "impossible task" of the type usually associated with more clearly non-human figures such as Rumpelstiltskin. Warner (p. 181), to emphasise the point, identifies "the forest-dwelling witch or crone" with the werewolf as "the kind of beings associated with marginal knowledge, who possess pagan secrets and are in turn possessed by them" - both being liminal figures between humanity and nature. The forest is the home of old secrets, identified with evil by the peasantry and with the superstition of the past by urbanites - the Witch is she who is in touch with them, and is therefore both outside normal society and dangerous to it.

Fairies or werewolves, however, are generally considered immortal or at least difficult to kill - the witches of the tales have been domesticated to the extent that they can generally be killed like any other human, although burning seems to be the preferred method. This domestication and control of the otherworldly, quasi-shamanic woman of power will be explored further in the next section, but it is sufficient to note here that the witch, being not entirely human or civilised, is generally not granted civilised considerations of mercy. In Andersen's "The Tinderbox", for example, a soldier meets a witch in the woods (as we would expect) who tells him how to get a vast quantity of magical goods from inside a hollow tree in return for him fetching her grandmother's tinderbox . When he emerges from the tree, however, he simply beheads her and takes everything (Conroy, p. 5). The Witch, not being civilised, is not entitled to the decencies of civilisation. In the next section, we will explore what this dehumanisation of the powerful woman figures in the tales, and justification of extreme measures against them, means in terms of the goals of social control set by the 19^th century textual meditators who created the literary fairytale from the oral folktale.

The final sections of this chapter will examine the explicitly pedagogical content of the tales of Grimm and Andersen, and the (partial) rejection of such content in Baum. It will do so primarily by noting exactly what in the tales is coded evil, and to what extent that evil is coded feminine, and relate this to both the actions of the Witches involved and to the punishments that they receive. As mentioned before, the world of fairytales is "curiously amoral", in Zipes' words - or, perhaps more accurately, it has a childlike morality (Baum, p. xxxiii) in which the punishment for being a "bad person" is generally lethal. As Tatar observes (p. 97), in the tales characters are given not what they earn by their deeds, but what they deserve solely by virtue of who they are. Accordingly, the soldier who murders the witch in "The Tinderbox" is not punished - as Zipes puts it, "narrative voice and providence" are on the side of the soldier, and witches are "evil per se" and thus can be killed without sanction (pp. 81-2). This goes some way to explaining the disproportionate nature of the punishments meted out to witches throughout the tales. Usually this takes the traditional form of burning, as happens to the Witches of "Hansel and Gretel", "Little Brother and Little Sister", and "Two Brothers" - in the latter case, the Witch is burnt even though she has undone her petrifaction spell under threat of being burnt (p. 231). Bottigheimer (p. 25) notes that fire throughout the Grimms' tales is associated with male control, and thus "closely associated with gender antagonism". Further, even worse deaths are sometimes the fate of the Witch - the stepmother/Witch in "White Bride and Black Bride", for example, is put in a barrel studded with nails on the inside which is then dragged behind a horse. It is perhaps noteworthy that the witch herself unwittingly came up with this punishment (Mannheim, p. 464) - again, the Witch is being punished not merely for her overt deeds, but for the evil that is inside her. As will be mentioned further later on, The Wizard of Oz is an exception to this trend - the Wicked Witches die for being wicked, not for being witches - but in the other tales the Witch is clearly fair game because of her nature rather than her crimes. This total antipathy to the character of the powerful female character, as Bottigheimer, has explored, has strong implications for the gender content of the tales' pedagogical focus.

The divisions between the good and evil in the tales are never ambiguous, and the evil are punished, and punished severely. Judging by Bottigheimer's analysis of the gender differences in punishments for transgressions in the tales, we can say that Zipes' "curiously amoral" fairytale world is only amoral for the male characters, but oppressively moral for the females. Bottigheimer's central thesis is that the moral and social vision throughout the Grimms' tales shows strong gender bias in its morality - women and girls are punished far more severely for transgressions than male characters. Bottigheimer (p. 88) notes that male characters, usually of the "trickster" kind, can "finesse" their way around moral prohibitions, even those imposed by God himself, and still get their reward - women, however, are punished for even involuntary transgressions. Bottigheimer (p. 171) relates this to the Christian concept that sin came into the world through the temptation of Eve - "the notion that all women share in Eve's sin may account for the necessity to punish female characters". Zipes himself notes (pp. 34-5) that "women had become equated with potential witchlike figures by the end of the seventeenth century so that control of their alleged sexual powers was linked by church and state to control of diabolical forces". We can, therefore, say that femininity itself is the sin which marks the witches out in the tales as deserving of cruel deaths without need of committing overt sins. "Within the 210 tales of the Grimms' collection, a witch-burning notion of eradicating (generally female) evil coexists with an indulgent tolerance of (generally male) malefaction." (Bottigheimer, p. 94). The Witch, then, is held up as a warning to the female auditors of the tales as to what may happen if they fail in strict obedience to the male heads of family. "The fairy tales memorialise the idea that women are subject to a summary despatch that can be justified by declaring them to be witches" (Bottigheimer, p. 102) - innocent women throughout the tales are almost burned as Witches, for example in the tale of "Mary's Child" (Grimms' Tales no. 3, Mannheim p. 8). The Witches in these overtly moral tales are occasionally agents of punishment, rather than its recipients. In one of the shortest of the Grimm tales, "Frau Trude" (Grimms' Tales no. 43, Mannheim p. 151), a disobedient girl refuses to listen to her parents' warnings, and runs off into the woods (always a mistake in the tales) to see Frau Trude for herself - Frau Trude is, of course, a witch and the devil in disguise, who turns the little girl into firewood. This particular mode of punishment is interesting, not solely because of the variation it offers on the theme of the child-consuming witch, but because the little girl is burned - the biblical injunction that "rebellion is as the sin of witchcraft" (KJV, 1 Samuel 15:23) can be seen in this identification of the fate of the Witch as the fate of all disobedient females.

The overt sins of the Witches in these tales are many - "envy and jealousy", as previously mentioned, are common to the wicked stepmothers, as is "unnatural" reluctance to nuture children. Another of the main areas in which women's freedom of action is circumscribed in the tales are speech. As mentioned in the previous section, Purkiss notes that traditional rural witches were rarely accused of using verbal spells. However, the question of cause and effect arises here - were the acts of magic non-verbal because it would be easier for the authorities to take action against verbal conjurations? Certainly there were laws on the statute books in England against cursing by the seventeenth century, whose targets Warner identifies as "not only men who swore, but women who could conjure… old folk who might use swearing and vituperation to retaliate… in default of other means of defence" (Warner, p. 39). Early modern writings by men emphasise women's speech, quite possibly the last power left to her in the patriarchal rural society, as especially dangerous - note the pejorative meanings that have become attached to "gossip" and "old wives' tales" (Warner. p. 13). The women who tended to pass on such transgressive speech were often "prostitutes, midwives and wetnurses [who] occupied no fixed point in the structure of society" (Warner, p. 35) - again, dangerous liminal figures. These anxieties were reinforced by religious speculation that Eve's "guilt in paradise disqualifies her sex from further speech" (Bottigheim, p. 170). Bottigheimer's analysis of Grimms' Tales (especially in chapter 5, pp. 51-56) suggests that the "good women" of the tales are characterised by refraining from speech, usually letting the narrative voice speak for them. When they do speak, it is usually in response to the male voice - Bottigheimer identifies "answered" as the verb most commonly applied to the direct speech of good female characters, whereas "said" or "spoke" is reserved for Witches and other wicked women (p. 54). Further, she shows that Wilhelm Grimm's progressive editing took more and more speech out of the mouths of good women and put it into the mouths of the evil ones. The same phenomenon can be seen in Andersen's tales - the Little Mermaid must give up her beautiful voice for her chance to become human (Conroy, p. 46). As she can only acquire a human soul through earning the love of the prince without her voice, this can be seen as another example of female submission being equated with silence.

. Warner further postulates that the anxiety about women's speech shown in the tales may be an anxiety of ownership - the anxiety of the male mediators attempting to assume ownership of what were, in rural society, quite literally "old wives' tales". Warner traces the image of the "old crone" as storyteller, back to the ancient Roman prophetesses such as the Sibyls (Warner, p. 3). "The immemorial storyteller Mother Goose [is] established, by the early eighteenth century, as a Sibyl-Nurse," writes Warner (p. 79), "who instils morality and knowledge of the world, and forsees the future of her charges and prepares them for it." In the progressive editing of Wilhelm Grimm which made the evils of female speech ever more clear, perhaps we can see an attempted revolution in ownership of the tales. The Old Woman who instils wisdom through tales was an uncomfortably pagan concept, and the mere fact that we refer to the Kinder- und Hausmärchen as "Grimms' Tales" shows how effectively ownership of the tales and the moral tutelage therein had been transferred to the male transcribers/editors - even while in English, at least, they were still often called "Mother Goose stories". So, the old woman as Sibyl-Nurse, imparting her wisdom through storytelling, becomes an image of transgression much like her counterpart, the Witch as shamanic intermediary between human and nature. Thus, the narrative voice of authority in the tales becomes more and more specifically male - and even the speech of females is converted into a crime tantamount to the crime of Witchcraft. The fate of the wicked witch becomes the threatened fate of all transgressive or vocal females.

The fairytale Witch that we have explored is the Witch presented as a distinct character - this is distinct from many fairytales where she functions as little more than a plot device. Into this category come those tales where an unseen "wicked witch" changes a prince into a frog for no apparent reason - or the tale "The Old Witch", in which the titular character meets the good daughter of the story in the woods and, for reasons never adequately explained, explodes, setting in motion the plot of the story. (Interestingly enough, this is the only tale of the Grimm collection which mentions a witch in its title.) This thesis aims to trace the evolution of the Witch as a character, and so we have concentrated on those fairytale witches who have their own personalities. Through the preceding sections I have shown three main ways in which the fairytale Witch is presented as an antagonist to the sensibilities of both the rural society which was the source of such tales and the urban society who appropriated them. The Witch is the "antimother", the false mother who cannibalises rather than nurtures children; this concept is associated with both the anxieties of rural housewives and of their children, and the resentment that children felt towards both biological and step-mothers shapes the tales to such an extent that the Witch and Stepmother become identical. The Witch is also the woodlands hag, a figure on the border between animal and human, distantly related to the shamanic figures of pagan agricultural societies, but resented, feared and hunted throughout the tales. Finally, in the hands of the moralistic mediators of the tales, the Witch becomes the embodiment of the sin thought to be immanent in all women; she curses and uses other forms of forbidden speech, thus rejecting her "rightful" state of submission, and the cruel deaths to which she is put point a moral lesson as to accepted female behaviour.

All in all, then, the Witches throughout the tales told by Grimm and Andersen are the same sort of "omnibus Other" figure that we have identified in Shakespeare's Weird Sisters - as Purkiss (p. 283) notes, this is a figure of the witch formed by "clump[ing] together figures of otherness to make one big Other of disorder". Much as the Weird Sisters are a symbol of rebellion against God, King and Nature, thus the Fairytale witches offend against the principles of motherhood, the "civilised" notion of humanity as distinct from nature, and the general principles of acceptable female behaviour in 19^th century society. This is why, contrary to Bettelheim's approach, a historical analysis of the tales is necessary, because we miss vital facets of the tales if we neglect "their essential historicity… the nineteenth-century origins and editing" (Bottigheimer, p. 15) In any case, it can be seen that in both Shakespearian and fairytale narrative, the Witch stands as a figure against which the protagonists of the stories, and through them, the readership, define themselves. However, there remains one important facet of the fairytales mentioned to be discussed in this chapter - the presence of Good Witches in The Wizard of Oz. It can be argued that the character of Dorothy is defined as much by identification with the Good Witches as her opposition to the Wicked Witches, and for this reason the tale is unique in the current context.

Dorothy's original transportation from Kansas to Oz can be seen as a rebellion against the lifeless, grey nature of Kansas life and her foster-family. Here again we have a stepmother figure - Aunt Em - against which the child must define herself; however, Em is not evil, but merely rendered joyless by the unforgiving nature of Kansas life. "The sun and wind had changed her too. They had taken the sparkle from her eyes and left them a sober gray… She was thin and gaunt, and never smiled, now… Dorothy's merry voice reached her ears; and she still looked at the little girl with wonder that she could find anything to laugh at." (Baum, p. 10). The identification of place and personality mentioned in previous sections also applies in Wizard - Aunt Em has been turned grey by the grey land of Kansas, and Dorothy's escape into the colourful land of Oz is symbolised by a switch from monochrome to colour both in the illustrations of the original edition and in the cinematography of the MGM film (Baum, p. 268n.) Further on this theme of identification, it can be argued that when Dorothy lands in Oz she becomes a sort of "de facto Witch". She is certainly taken as a witch by the Munchkins, because of her (unwitting) destruction of the Wicked Witch of the East and because of her white dress (p. 30); indeed, she literally "steps into the shoes" of the Wicked Witch of the East - that is, her magical silver shoes (ruby slippers in the film). This "mistaken identity" is similar to Oz himself, taken for a wizard himself upon his crash-landing in the c