I

There seems at present to be no history of British fairies. This may at first sight appear a churlish statement, in view of the fact that representations of them, in literature or folklore, have been studied by a relatively large number of scholars and often to an excellent standard. None the less, a case could be made that these writers have generally treated their material in other ways than those associated with what is arguably the central purpose of history: change over time. In the early twentieth century, they were most concerned with retrieving the Celtic origins of representations of fairy beings in medieval literature: Alfred Nutt, Lucy Allen Paton, and Roger Sherman Loomis were outstanding figures in this enterprise.Footnote ¹ By the middle of the century Minor White Latham, C. S. Lewis, and Katharine Briggs were especially prominent in another, of classifying the different types of being subsumed under the label of ‘fairy’ and establishing their characteristics.Footnote ² In the past three decades this has largely been abandoned in turn, for a number of different perspectives on the subject, all of which have yielded excellent results. Jeremy Harte, Lizanne Henderson, and Edward Cowan have considered the nature and function of fairy tales, and their value for an audience.Footnote ³ Literary specialists have concentrated on the use of fairies for motifs and plot devices in British texts of the medieval and early modern periods, and the manner in which this expressed the cultural, political, and social concerns of those ages.Footnote ⁴

Those studies concerned with particular periods – comprising most of those just cited – have, of course, no need to make connections or comparisons with other ages. Some, such as those of Katharine Briggs and Jeremy Harte, have ranged widely across time, but treated fairy lore as something essentially unchanging from the middle ages to modernity. In this, both operated within the assumptions of traditional European folklore studies, of an essentially constant rural popular cosmology that persisted through all the dramatic developments in elite and official belief. Among writers who have indeed taken a historical approach to the subject, three stand out in defining both the achievements and limitations of the existing scholarship. The first is Sir Keith Thomas, who, in a few pages of his great work on early modern English religion and magic, managed to lay out major elements of the relationship of fairy lore with both. He dismissed an enquiry into its origins, however, with a comment that these were various, and any pursuit of them would produce only speculative results.Footnote ⁵ This remains true with regard to the ultimate roots of beliefs, but the medieval and early modern material may, it will be proposed, show some development. The second writer is Diane Purkiss, whose justly famous book on fairy tales did examine the successive forms that they have taken, mainly in Britain, from ancient times to the present. Her approach was essentially a psychological one, to suggest the inner meaning of stories about fairies for those telling and hearing them, and she did so in a series of case-studies grouped within each period. What she was less concerned to do was to chart the process by which the forms of each developed into those of the next, and the relationships between them: in her own words, hers was ‘not a continuous story, but fragments of story, woven into the fabric of lives’.Footnote ⁶ It may be seen here whether a ‘continuous story’ can in fact be told for the period 1100–1640. The third author is Corinne Saunders, who gave an important chapter to fairies or characters similar to them in her study of supernatural elements in medieval English romances. It is a valuable analysis, emphasizing in particular the ambivalence of identity manifested by the magical human-like beings which feature in these texts, and their particular functions in different plots. It does not, however, place them in any sequence of evolutionary development in her period, or between it and others.Footnote ⁷

What seems to be missing so far, from all this admirable body of work, is a sustained sense of how British concepts of fairies developed and mutated from the opening of recorded history to the early modern period, at which time they arguably attracted the most attention and generated the most debate; and which bequeathed most of the key images of them to modernity. This would be a study of the making of a ‘tradition’, defined as a body of ideas and beliefs handed down between generations. In this case, the tradition concerned is largely literary, but (as will be suggested) blended indissolubly with broader culture at all levels of society. Such a study will not depend on new material, for that for the subject has long been well established. An assiduous reader of the secondary literature cited above will notice that the great majority of the primary sources used in works published after 1980 had been identified in earlier publications, and most in those which appeared before 1960. None the less, they have been, and can be, deployed for different ends in different works. The sources are essentially confined to England and Lowland Scotland, with a few from Wales and the Northern Isles, and largely to works produced by social elites. None the less, they are collectively ample enough to sustain arguments, and – as shall be shown – there is some significant evidence for the beliefs of commoners at all points of the period under study. There will be little concern here with the possible social functions of fairy motifs or their meanings within a cultural system, largely because these have been prominent concerns of the existing literature. Instead, the preoccupation is with the manner in which the motifs developed and diffused, which is a subject that has been much less considered, and may be a worthwhile contribution to a historiographical review such as this.

II

All previous scholars have noted that the term ‘fairy’ itself arrived in Britain from France only in the high and later middle ages, and that the beings to whom it was to be applied were known before this, across most of Britain, as ‘elves’. The Anglo-Saxon texts referring to these beings have recently been studied by Alaric Hall, in a manner which may well prove definitive.Footnote ⁸ One consequence of his labours has been to reveal how little we can say with certainty about the Anglo-Saxon beliefs. It is clear that elves were feared, for maliciously afflicting humans and their animals, but there are also strong hints that they were models of seductive female beauty. There is no unequivocal evidence that elves were regarded as sources of supernatural power for people, but some that they were associated with human diviners or prophets. No sense of a coherent tradition emerges from the texts, which may be a reflection of reality or just a consequence of the patchy survival of evidence; and the same may be said of the apparent absence of early English stories about such beings. It may be added here that comparable confusion and ambiguity is reflected in the interpretation of specific linguistic evidence: for example, the Anglo-Saxon personal name Aelfwine, ‘elf-friend’, may have indicated somebody who could befriend elves (so confirming a possibility of profitable interaction with them) or as a gesture of propitiation to protect that person from their malice (so confirming a prevailing relationship of hostility and fear).

The likelihood that no coherent view of elves was in fact held in Anglo-Saxon England is increased by reference to the famous texts of the twelfth and early thirteenth centuries that have become virtually canonical in studies of British fairies, and with which the subject is usually taken to commence: above all Gerald of Wales, Ralph of Coggeshall, Gervase of Tilbury, Walter Map, Ralph of Coggleshall, and William of Newburgh. In their accounts of alleged encounters between humans and non-human beings which could not easily be fitted into conventional Christian concepts of angels and demons, were included several motifs which were to be enduring components of fairy lore. The first is the belief in a parallel world to the human one, with human-like inhabitants who occasionally have their own sovereign, and are longer-lived than, and in some ways superior to, people.Footnote ⁹ The second is the ability of such beings to enter our own world, and sometimes to steal human children away from it, while humans could blunder into their realm.Footnote ¹⁰ The third is a belief in beautiful supernatural women, who dance in secluded areas at night, and who can be wooed or abducted by mortal men, but who almost always eventually forsake the resulting marriage for their own realm. Sometimes this fairy-lover motif takes the form of a man having sex with an apparent woman, only to find her turn into a monster.Footnote ¹¹ The fourth is that such non-human beings are often associated with the colour green, either in their clothing or even their flesh.Footnote ¹² The fifth is that they can give blessings to people who entertain them or otherwise treat them graciously, but also afflict them, notably by leading them astray at night into pits or bogs.Footnote ¹³ Associated with this is the sixth, a tradition of human-like creatures which live in or enter homes, where they can make themselves useful to the human occupants by helping them with tasks, or play mischievous tricks on them.Footnote ¹⁴

On the other hand, what is missing in these accounts is any sense of a coherent belief system to contain and explain the stories being repeated and considered by these medieval authors. The absence from them of a being that was especially feared for inflicting physical ills, so plain in the Anglo-Saxon sources concerning elves, would indicate that only a portion of popular tradition was being represented in them; none the less, there is nothing about any of them that suggests they were strictly the preserve of a social elite. Some were explicitly stories told in local communities. As a number of modern commentators have suggested, the intellectuals who collected them and grouped them together were struggling to create a category for them, specifically because none seemed to exist already either in Christian cosmology or established folk belief. As part of this enterprise they strove to find a language for the beings portrayed in the reports, as the range of terms available did not quite describe the kind of phenomenon being discussed.Footnote ¹⁵

A possible escape route was provided for them by no less an authority than Augustine of Hippo, who reported a Mediterranean and Gaulish tradition that woodland spirits mated with human women. He declared that these were demons, of the predatory erotic kind known as incubi.Footnote ¹⁶ Augustine's formula was repeated by such prominent clerical authors as Cassian of Marseilles, who demonized fauni, who led travellers astray by night, and Burchard of Worms, who did the same for sylvaticae, who coupled with men in the form of beautiful women.Footnote ¹⁷ The British authors of the high middle ages sometimes followed this course, but not always, even when the case might have seemed appropriate such as in stories of supernatural lovers.Footnote ¹⁸ Most of the anecdotes that they recorded did not seem to them definitely to suggest demonic activity, and indeed, some accounts of superhuman woodland beings could be filtered through sanctity. The traditions of Evesham Abbey stated that it had been founded on the place where Egwin, bishop of Worcester, had seen three unearthly and beautiful women singing in a forest, and decided that they were saints.Footnote ¹⁹

A similar lack of definition exists in a parallel stream of literature from the same period between 1100 and 1250, the romances which feature encounters between human characters and human-like beings who have sumptuous lifestyles, mirroring those of the contemporary human social elite, and dispose of apparently superhuman powers. In particular, these beings function as lovers, councillors, and protectors for the human knights and ladies with whom they make relationships, and sometimes in negative roles as predators upon them. Leaving aside the vexed and probably insoluble question of the origins of this tradition, it can be said that by the twelfth century it was represented in literary works composed across north-western Europe from France to Ireland. Whereas the scholarly texts discussed above were dealing with incidents that were believed to have taken place in the real world, the romances were uninhibited works of fiction. They are important to fairy lore because those written in French supplied the genesis of the word ‘fairy’ itself, associated with the term fai, fae, or fay, applied to female representatives of the beings described above. In recent years, the role of these characters as plot devices and mirrors of contemporary social and cultural preoccupations has been studied by Alaric Hall, James Wade, Helen Cooper, Carolyne Larrington, Kathryn Westoby, and Corinne Saunders.Footnote ²⁰ In the process, they have made between them two points of relevance to the current enquiry. The first is that little attempt was made to define these beings within a theological framework, or indeed to explain who they are at all or to explore their motivation: they are simply assumed to be mysterious. The second, which is partly consequent upon the first, is that their status as humans or non-humans is often left in doubt. At times, it is explicitly stated that they are human beings who have learned magic, and so gained powers not accessible to most people, while at others they appear essentially to be superhuman; but in most cases they can be assigned to either category and the matter is not considered in the tale.

None the less, they are important to this investigation. For one thing, they represent, as said, the linguistic root of the whole concept of the fairy. As Noel Williams has pointed out, the word fai or fay itself functioned more often as a verb than a noun, to denote the making of something magical and strange, in both Old French and the Middle English texts into which the French themes were transposed. Its derivation or parallel development ‘faierie’ was evolved to refer to uncanny events and phenomena, rather than creatures, and only began to refer to a type of being in English in the fifteenth century.Footnote ²¹ None the less, it enabled the eventual creation of such a type. Furthermore, among the kinds of ‘fay’ found in the French romances of the twelfth and early thirteenth centuries are some who would later populate the category concerned. In Claris et Laris the enchantress Morgana is the ruler of a forest, who captures humans who wander into her realm. The Battaile Loquifer of Jendeus de Brie has three fays, led by Morgain, who come flying to a grief-stricken hero, to carry him off to their magical realm to be Morgain's lover. In Tydorel, a knight emerges from a lake to become the lover of the queen of Brittany, siring the hero of the tale who himself eventually returns to his father's home in the lake. The hero of Renaud de Beaujeu's Bel inconnu is seduced by a ‘Maiden of the White Hands’ in the ‘Golden Isle’, who has known his destiny beforehand. Yonec has a knight fly into a human woman's chamber in the temporary form of a great bird, coming from a realm entered through a hill. Most influential of all in subsequent periods was Huon de Bordeaux, which introduced readers and listeners to Auberon, the dwarf ruler of a forest kingdom, who is possessed of great magical powers, immense wealth, and a white marble capital city. Great pains are taken to reassure them of his illustrious parentage and his Christian piety. His powers were bestowed by magical beings who attended his birth, and his diminutive stature was likewise the result of a curse by another such being.Footnote ²² Such entities also feature in the work of the priest Layamon, reworking the legendary history of Britain for an English-reading audience. He recounts how Arthur was brought up by them, and returned to their domain of Avalon, ruled by a queen, at the end of his reign. Layamon's use of English enabled him to cross the romance with the vernacular genres, by giving them the native name of alven or elves.Footnote ²³

By the middle of the thirteenth century, therefore, it can be proposed that the surviving English sources contain a native tradition of elves, as blighting and perhaps as healing and seductive beings; an international literary one of beautiful, wealthy and powerful fays; and a third category of diverse human-like creatures, who overlapped with the first two types but did not really fit into either. What did not yet seem to exist was any attempt to combine and systematize most, at least, of these forms: if there was a synthesis made by people at the time, it is not visible to us in the fairly abundant surviving sources.

III

During the late thirteenth and the fourteenth centuries, this inchoate sense of an ill-defined or undefined parallel world of magical beings persisted in some texts. At the end of the period, Thomas of Walsingham's chronicle could still contain a classic story of a demonic forest incubus seducing a Hampshire woman, and another of a Cumberland youth who was left demented by an encounter with strange and ethereally beautiful women served by a little red man.Footnote ²⁴ No attempt was made to explain these incidents or set them in any broader context. Throughout the late medieval and early modern periods, moreover, references continued to supernatural beings, inherent in the British landscape, which had vague, functional identities and no known relationship to each other. The ‘puck’ was known from Anglo-Saxon times as a name for a spirit who led nocturnal wayfarers into mires and pitfalls, while the bug (a term with a variety of related words) featured from the later middle ages onward as another entity of the night, distinguished by its ability to strike terror into people. In the fourteenth century the name ‘goblin’ arrived, probably from French, for a similarly unpleasant and hazily characterized nocturnal sprite, whose activities overlapped with both puck and bug.Footnote ²⁵ In the early fourteenth century, a Middle English romance such as Sir Degarré or Sir Degare could still feature an enigmatic, richly dressed, ‘fairi’ knight who couples with a princess in a forest to engender the tale's hero in classic incubus fashion, without finding any need to explain his kind or indeed whether he is human or not.

None the less, by the end of the thirteenth century moves were being made to put a systematic structure of belief around such figures. The South English legendary, which was completed around that time, defined ‘elven’ as former angels banished to the earth for remaining neutral during the war in heaven which ended in the expulsion of the rebel angels to hell. There, they took the shape of beautiful women who danced and played in secluded places, and men could have sex with them, but at their peril.Footnote ²⁶ The Metrical chronicle of Robert of Gloucester, which is similar enough in style and date to the Legendary to make an overlapping authorship almost certain, spoke of ‘elvene’ as spirits often seen in wild places, in the form of men and women, who seduced humans.Footnote ²⁷ By the early fourteenth century, a preacher's manual, Fasciculus morum, could condemn as a devilish illusion a widespread belief in ‘elves’ who took the form of beautiful women dancing at night with their queen or goddess. The belief concerned, according to the manual, included the detail that these elves could carry off humans to their own land, where heroes of the past dwelt.Footnote ²⁸ Meanwhile, some of the first English romances used classical influences to provide another framework for systematizing the fays. The clearest example is Sir Orfeo, composed around 1300, which retold the ancient myth of Orpheus and Eurydice; but in this version Orfeo has to retrieve his wife, not from Hades but from the land of a nameless ‘King of Fayré’ (or ‘Fare’ or ‘Fairy’), who takes the role of the Roman god Pluto as ruler of a realm of the dead, though, in this case, of those who have met untimely ends: in this context, it is full of deeply disturbing imagery. None the less it is a fair, green land, where the king reigns in state with his queen and sometimes invades the human world with a retinue to hunt beasts or abduct people.Footnote ²⁹ The result is a well-rounded picture of a fairyland, developed out of images of the more regal of the individual fays into a standard and universally applicable concept.

These early steps in creating it prepared the way for the leap taken by the end of the fourteenth century, when Chaucer could speak, famously, of how in the days of King Arthur, Britain was ‘fulfilled of fayerye’, and ‘the elf-queen, with her jolly company, danced full oft in many a green mead’.Footnote ³⁰ He was taking a composite image of a fay, from the high medieval romances (and especially those of Arthur and his knights) and giving it the definite article that established her as an archetype which was becoming a personality in her own right. The process is still under way, because in another tale, where he sends up the whole stereotype of the knight and the fay, he has his chivalric hero first decide that he must win the love of ‘an elf-queene’, and then enter the land ‘of Fairye so wild’ to encounter ‘the Queen of Faierye’.Footnote ³¹ At yet another point, he falls back on the classicizing tradition, to make Pluto ‘King of Faierye’ with his queen Proserpina, who enter the human world with their retinue and use their divine powers to influence human affairs.Footnote ³² Thus, concepts are still fluid, but a set of associations are crystallizing around the words ‘fairy’ and ‘elf’ which are defining an increasingly familiar place and set of characters.

This process is equally visible in a contemporary English romance by Thomas Chestre, Sir Launfal, a reworking of a twelfth-century Old French tale attributed to Marie de France. It is a classic plot of how a mysterious and beautiful fay gives her love and aid to a true knight, and Diane Purkiss has drawn attention to the contrast between the two versions: in the earlier, the nature of the heroine is left undefined, and she dresses in royal purple, while in the later, she is explicitly the daughter of ‘the King of Faërie’, and dressed in the distinctive fairy colour of green.Footnote ³³ At its close, the later version indeed has her take the hero back ‘into the faërie’ with her. Significantly, there is a third version of the story which occurs between the two, Sir Landevale, produced in Middle English around 1300, and that retains the lack of definition of the first: it was in the course of the fourteenth century that the romance concerned became a true fairy tale.Footnote ³⁴

By the fifteenth century, the literary construct of the fairy kingdom was fully formed, and had penetrated varieties of English literature other than romance.Footnote ³⁵ There are signs that theology was accommodating itself to the possible reality of this realm: one sermon suggested that ‘elves’ were ‘fiends’, but of a low rank not threatening to human salvation.Footnote ³⁶ One of the other major developments of that century was the appearance of Scottish romances, which reveal that by this time the new construct of a fairy land was already truly pan-British. The first of them to deal with it, dating from somewhere between 1401 and 1430, is also the most famous: Thomas of Erceldoune, which tells of how its genteel human hero became the lover of a lady from ‘the wild fee’. She takes him to her own land (entered through the side of a hill), where she turns out to be the wife of its king. He returns to the mortal world with gifts, of telling truth and knowledge of the future.Footnote ³⁷ From the mid-fifteenth century comes King Berdok, which has the first known usage in Scots of the word ‘fairy’, and likewise gives the fairy kingdom a king.Footnote ³⁸ By the end of the century, the fairy genre had become a resource on which Scottish poets could draw for imagery and allusions in their compositions. When Robert Henryson produced his own retelling of the tale of Orpheus and Eurydice, he made Proserpine at once ‘goddess infernal’ and ‘queen of fary’, while Eurydice, on dying, is ‘with the fary taken’.Footnote ³⁹ William Dunbar likewise termed Pluto ‘the elrich incubus, in cloak of green’, while he described the mother of another character as ‘ane farie queyne, gottin be sossery’.Footnote ⁴⁰ In the early part of the next century, Sir David Lyndsay made the court of ‘the queen of Fary’ the place to which one of his characters hopes to go on dying, and repeatedly had others refer to the queen or king of ‘Farie’ in passing, with either affection or fear.Footnote ⁴¹

Over the same period, Welsh literature absorbed the same motif. Buchedd Collen, which is fifteenth or sixteenth century and represents a hagiography written in the style of a romance, has its saintly hero encounter the traditional lord of Annwn, the medieval Welsh underworld or otherworld, Gwyn ap Nudd. Gwyn has now become ‘King of the Fairies’ as well as of Annwn, and when the saint sprinkles him and his sumptuous court with holy water, all vanish leaving green mounds behind.Footnote ⁴² By the mid-fifteenth century, also, the concept of the fairy realm had become part of the mental world of English commoners. In 1450, ‘one calling himself Queen of the Fayre’ operated in Kent and Essex, presumably for profit.Footnote ⁴³ The next year a gang of disguised poachers who raided the duke of Buckingham's park in Kent called themselves ‘servants of the queen of the fairies’.Footnote ⁴⁴ Most significant, perhaps, are the cases of women from Somerset and Suffolk, tried in church courts in the mid- and late fifteenth century for claiming to have obtained magical powers, which in the Suffolk case at least covered the general repertoire of ‘cunning folk’ or ‘wise folk’, of healing, divination, detecting sources of misfortune and finding buried treasure. In the Somerset one, these powers were said to have been conferred by ‘spirits of the air which the common people call feyry’, and in the Suffolk one from ‘God and Blessed Mary’ and a ‘certain little people’ called both ‘lez Elvys’ and ‘lez Gracyous ffayry’.Footnote ⁴⁵ The imported French word had already come to signify among the English, apparently in general, the beings that were known in their own language as elves.

As Alaric Hall has noted, this matches the reappearance in English texts, around this time, of the Anglo-Saxon tradition that elves, and now fairies, inflicted illness on humans.Footnote ⁴⁶ There seems no reason to doubt that this tradition had continued during the intervening period, but is invisible because the kinds of source which would represent it, medical manuals and charms embodying popular belief, are largely missing during that time. Likewise, the associated acceptance that elves or fairies could bestow magical powers on favoured humans, including that to heal, is also probably ancient. Other beliefs concerning them were less certainly continued from the past. There seems to be no certain record in any British medieval text, for example, of the tradition well attested in France and Germany during the high medieval period, that woodland spirits stole human children and substituted sickly or difficult offspring of their own for them.Footnote ⁴⁷ This belief does, however, appear unequivocally in a school handbook of model Latin translations published in 1519, and becomes a regular feature of English fairy lore thereafter, found constantly in both literature and records of daily life.Footnote ⁴⁸ The absence of it in medieval texts, when it is so clearly present in equivalent sorts of record on the continent, is striking, and it is absent from the relatively abundant Scottish references to fairies dating from before 1660, while well recorded in later Scottish folklore. Both facts would suggest that it was imported to England around 1500 and spread north from there.

Other apparent innovations were made in British accounts of elf- and fairy-like creatures at this time, of which the most notable was the appearance of Robin Goodfellow as a particular name for one. This is first recorded as used playfully by one of the correspondents of the Paston family in 1489, and in 1531 William Tyndale allotted this character a role, of leading nocturnal travellers astray as the puck had been said to do since Anglo-Saxon times and the goblin since the later medieval.Footnote ⁴⁹ Reginald Scot, writing in 1584, aligned him with another long-established type of magical being, the household spirit who performs helpful practical tasks in exchange for reward: in his case, bread and milk. Scot also, however, referred to Robin Goodfellow in another place as a ‘great bullbeggar’, who was once ‘much feared’, suggesting a more hostile nature for him: the attributes of such characters had not become precisely fixed.Footnote ⁵⁰ Scot claimed to be speaking of beliefs obtaining a century before his time; whether he was correct or not, he can be taken at least as witness for those of his own childhood in the early sixteenth century. Another personality of later fame in British fairy tradition was Oberion, who features in English legal cases between the mid-fifteenth and early sixteenth centuries as a spirit invoked by ritual magicians.Footnote ⁵¹ It is possible that he had an entirely different point of origin, but there is at least a strong suspicion that he was inspired by, or related to, the powerful enchanter Auberon, in Huon de Bordeaux. These cases also have value in illustrating, through the evidence given in them, the manner in which ideas and motifs could easily travel between learned and popular culture in the period, with the lower clergy playing a crucial role as mediators.Footnote ⁵² Perhaps we see here one kind of network through which the concept of a fairy or elven kingdom could rapidly have been spread: others would be represented by the late medieval minstrels and ballad-singers studied by Sir James Holt, and hawkers, chapmen, and wandering tradespeople highlighted by Tessa Watt, Margaret Spufford, and Adam Fox, as agents of cultural transmission.Footnote ⁵³

IV

In the period of the English, Welsh, and Scottish Reformations, and (even more particularly) that immediately succeeding, the late medieval concept of the fairy kingdom, and fairies in general, became the subject of intense interest and debate across most of Britain: indeed, fairy mythology was probably more prominent in British culture between 1560 and 1640 than at any time before or since. That mythology featured frequently in a great range of genres, including stage plays, poetry, works of demonology, theology and ceremonial magic, and legal records, spanning the island's social levels and groups. In part, its apparent popularity was due to the appearance or multiplication of the sources in which it featured, produced by the development of secular drama, the impact of printing, and the greater survival of archival materials. None the less, the references to it in legal cases, in particular, suggest a genuinely heightened popular awareness, and it represented a focus for expressions of differing ideology to an extent that it had never done before, and would not do again. Particular characters within fairy mythology, inherited from the middle ages, were given greater colour and prominence in the process: above all, Auberon, of Huon de Bordeaux, who emerged, Anglicized as Oberon, as the favourite name for the fairy king, and Robin Goodfellow, who developed into the most famous of all English spirits. By 1627, the poet Michael Drayton could quip that some in his society were talking of fairies ‘as if wedded to them’.Footnote ⁵⁴ In brief, fairy mythology became a preoccupation of both learned debate and public interest, and by 1640, the standard characters and associations of fairyland were established in the British literary imagination, in a manner that would not alter greatly for the rest of the early modern period. The preoccupation with fairies remained strong for the rest of the seventeenth century and into the beginning of the eighteenth, and produced in this later period some of the most enduringly famous authors to deal with them, such as Robert Kirk and Thomas Hobbes, and some of the most celebrated cases, such as those of Isobel Gowdie in Scotland and Ann Jeffries in England. None the less, the volume of recorded interest was lessening, and the motifs and ideas did not develop much further: which is why 1640 is chosen as the terminal date.Footnote ⁵⁵

Peter Marshall has made an excellent preliminary survey of the range of debate over the nature of fairies in early modern England, as Lizanne Henderson and Edward J. Cowan have of that in Scotland, while Diane Purkiss has studied a wide selection of the material for both.Footnote ⁵⁶ None the less, more can be said, and space permits here just a broad survey of the range of views held. One cluster of discussion centred on the idea that fairies were evil beings, and the most extreme form of this idea characterized them as demonic. It was a tradition which, after all, went back to Augustine and could tap into the enhanced interest in the Devil which was one feature of post-Reformation spirituality. Its most distinguished exponent in this period was King James VI and I himself, and it was also expressed by other Scottish and English writers.Footnote ⁵⁷ An equal number grouped fairies and devils together, but implied some difference in kind.Footnote ⁵⁸ Meanwhile the ancient fear of elves, as inflicting malicious harm on humans, persisted without any necessary reference to Christian theology, and was clearly a live issue in everyday life.Footnote ⁵⁹ Closely connected to it, there also remained the belief that they, alias fairies, could bestow magical abilities on favoured or canny humans. In the early modern period, they featured as one of the main sources of power – and perhaps the main – for the cunning or wise folk of Britain, from Orkney to the English Channel. By 1600, this function had been assimilated to the late medieval construct of the fairy kingdom, and it was especially the fairy queen with whom people of both sexes preferred to deal. The majority of the recorded cases concern women, and it is not clear from the evidence that this was because they resorted more to fairy belief, or simply ended up in court more often, than men.Footnote ⁶⁰

The most spectacular such court appearances concerned cunning folk who ended up accused of witchcraft, almost all the examples of which are Scottish.Footnote ⁶¹ Margot Todd's study of kirk session and presbytery records provides further Scottish material to prove the close connection perceived between fairies and cunning craft, and also suggests that practitioners were usually safe from prosecution as long as their clients prospered, and normally suffered little even when reported.Footnote ⁶² Fairies were also directly involved in practical magical operations, as proved by the surviving manuscripts of sorcerers from the period between 1560 and 1640. Five contain directions for their invocation and control, and treat them as a distinctive sub-class of spirits, though with the tasks and powers of all other kinds.Footnote ⁶³ This classification of them and specific interest in them does seem new,Footnote ⁶⁴ and records of individual magicians such as Simon Forman and John Dee also suggest a novel excitement about the specific potential of working with fairies as magical servitors.Footnote ⁶⁵ All these reflections suggest that those who classed fairies as demons pure and simple were rare enough almost to count as radical. Those hostile to affection for them, or even belief in them, tended more often to declare them to be non-existent, and the products of a deluded imagination. A subset of this argument was to associate the delusion with the superstitions and impostures of Catholicism.Footnote ⁶⁶ Playwrights hinged comic plots on impersonations of fairies and similar beings, such as goblins, by human beings, or pretended conjurations of them, to deceive fools.Footnote ⁶⁷ These were accompanied by genuine cases of confidence trickery, clustering between 1595 and 1614, in which criminals attempted to part victims from their wealth on pretence of introducing them to the fairy monarchs. It is hard to believe that this ruse would have been adopted, or been initially so successful, had not interest in the fairy kingdom been unusually intense among the English at this time.Footnote ⁶⁸

The reasons for the Scottish emphasis on fairies as components of witch trials have been studied by Diane Purkiss and Emma Wilby, who have pointed to the peculiar English emphasis on the familiar spirit, often in animal form, as a demonic assistant to a witch, as taking the same role as fairies in Scotland.Footnote ⁶⁹ The difference is indeed striking, though one of degree, as animal familiars are occasionally found in Scottish trials and fairies in the English.Footnote ⁷⁰ Both nations, moreover, had the idea that witches could be served by attendant demons, and most Scottish trials of which details survive do not include alliances between witches and fairies. The problem is deepened by the fact that, as said, all over Britain cunning folk claimed to have learned their abilities from fairies, though not necessarily, it seems, to have had a particular fairy to serve them: this is a concept which appears instead in ritual magic and witch trials. Furthermore, nobody has yet found a convincing medieval antecedent for the early modern animal familiar or the English emphasis on it: the only apparent forerunners are in the demonic servants credited to ritual magicians.Footnote ⁷¹ It is possible to fall back lamely on the caprices of regional taste: the eighteenth-century British, after all, universally blessed farms at the New Year, but the form of that blessing was different in the highlands, lowland Scotland and northern England, Wales, and southern England, and there is no functional reason apparent for the divergent customs.Footnote ⁷²

There was also a different way of discussing fairies in the same period, which was positive and even admiring. It was specifically an English phenomenon, seemingly missing in Wales and Scotland, probably because of general decline of secular literary forms in both, though in the Scottish case the hostile attitude of the king may also have counted. By contrast, the English expressions of affection were considerable, and drew largely on the medieval literary tradition of powerful and helpful fays.Footnote ⁷³ The fairies of late medieval romance were, after all, natural monarchs and aristocrats, who led lives of opulence, leisure, and frivolity untrammelled by the ills that afflict mortals. As such, their appeal could extend well beyond the elite: one did not need to possess any of those attributes to dream of a land in which everybody enjoyed them.Footnote ⁷⁴ These traits could also make them obvious counterparts in allegory and fancy to real royalty, above all Elizabeth 1, who became the first monarch to benefit from the heightened interest in fairyland. She was saluted in literary works and in aristocratic entertainments, both as a fairy queen herself (most famously by Spenser) and as the recipient of homage by fairy monarchs.Footnote ⁷⁵ This genre continued into the next reign, most spectacularly when Henry, prince of Wales, himself appeared as Oberon in a court masque.Footnote ⁷⁶

At all levels of English literature between 1550 and 1640, fairyland functioned also as an embodiment of hedonism, summed up in its characteristic activity of song and dance. As such it was a gift to lyrical poets writing English equivalents of classical pastoral, and indeed translations of actual Greek and Roman texts during the period routinely rendered nymphs and equivalent beings into English as fairies.Footnote ⁷⁷ They were also welcome to playwrights who wanted to add a musical interlude, while other writers treated a belief in fairies – or indeed their existence – as one feature of an older, simpler, and happier world, a variation on the theme of Merry England.Footnote ⁷⁸ During the late Elizabethan period, the old tradition of the household helper spirit developed into a literary one that fairies rewarded people who performed their own household tasks well but punished the dirty and lazy. It seems to be unknown in literature before 1600, but rapidly became a fairly frequent motif after then.Footnote ⁷⁹ Robin Goodfellow became an ethical hero, aiding the victims of wrongdoing and punishing the wrongdoers.Footnote ⁸⁰ Whether these themes entered literary works from folk tradition is impossible to determine, but by the 1630s the concept that fairies rewarded and punished household activities was a genuine popular belief.Footnote ⁸¹ A final way in which fairies were rendered attractive was to treat them as diminutive and engage in imaginative explorations of a world peopled with such midgets.Footnote ⁸² This was not itself new, as Gervase of Tilbury and King Berdok, to name two earlier examples, had featured tiny fairy-like beings: but it was new as a genre. However (self-consciously) ridiculous and literary such flights of fancy were, with no counterpart in popular belief, they did make fairies harder to perceive as demonic.

All these images illustrate how much the British took to ‘thinking with fairies’ between 1560 and 1640, and how richly diverse the results were: within the Home Counties of England they could in the same decade provide material for sweet poetic whimsy and for a charge of witchcraft. As King of Scots, James condemned belief in fairies as a diabolic illusion; as King of England, he looked on with apparent equanimity while his son personified one as a noble and admirable being. Shakespeare made them imposing and benevolent in one play, ridiculous in another, a cover story for a fraud in another, and a danger comparable with witches in a fourth. A relatively coherent late medieval concept of a fairy kingdom had not produced a firmer sense of its moral or theological status, despite such fervent discussion. On the other hand, that lack of fixed attitudes prevented any real debate over the subject and hence any conflict: people worked with the concept in different ways, without clashing over it.

V

This review has done more to describe apparent change over time than to explain it, for, given constraints of space, its purpose has been to establish that the changes concerned are indeed apparent: they may represent a shift of form rather than content, but the shift is important. It has attempted to demonstrate, beyond reasonable doubt, that a coherent literary tradition of a fairy kingdom developed in Britain during the late middle ages, the crucial period for this development being the fourteenth century and the process largely complete, across the island, by the fifteenth. It has also been argued that this late medieval concept was then debated and elaborated with considerable energy in British literature, and especially English, during the post-Reformation period.

What is more difficult to prove is how this literary tradition related to a popular one of belief in such beings, using the word ‘popular’ to refer to all levels of society. The sources seem to suggest that the two were directly associated at each stage: the model of the fairy kingdom appears among people at large soon after it has become a literary theme, and interest in it seems to increase among them after the Reformation even as it burgeons and diversifies in books, plays, and poems. This may possibly be an optical illusion created by a proliferation of the relevant kinds of source material, but there are limits to such a possibility. Historians of medieval culture, such as Robert Bartlett, have suggested that the anecdotes concerning apparent spirits recorded by high medieval writers originated in different social groups, and were a sample of those actually circulating in the populace at large.Footnote ⁸³ Likewise, it is hard to imagine that confidence tricksters would have found fairy monarchs such a potent vehicle for their ploys in the years around 1600 had not such monarchs featured prominently in people's minds at that time; nor why playwrights composing for a general audience would have made so much mention of fairies at that time had they not been a subject of proportionate interest.

The appearance of the word ‘fairy’ itself among English commoners by the mid-fifteenth century signals a derivation of key ideas from literature, as it was taken from Old French, and had become attached by poets to the concept of a kingdom of such beings before word and concept feature together in accounts of actual belief and action. This is not, therefore, the mere emergence of a name into texts, unconnected to ideas: both name and ideas – the word ‘fairy’ and the fairy kingdom – were associated from their first attestation by ordinary people. This suggests either an impact of literary forms and ideas on the popular imagination, or that both were evolving together with the same constructs: and of the two possibilities, the former seems much more likely, because the idea of the kingdom was itself based on literary materials. These included distinctively elite forms, such as the fays of chivalric romance and myths from the classical ancient world. If this conclusion is accurate, then it underlines that drawn in general by Adam Fox and his colleagues for early modern Britain; that oral and literate cultures were closely interwoven and constantly interacting.Footnote ⁸⁴ This relationship was clearly nothing new in Western Europe, as (for one example) Matthew Innes has shown the same interplay in the work of a ninth-century monk of St Gallen.Footnote ⁸⁵ What is suggested as likely here is this effect on a grand scale: that a literary model referring to fictional events was constructed from materials which ultimately derived partly if not mostly from oral belief and referred to what was commonly taken to be the ‘real’ world; and that this model was in turn taken into general culture, including oral tradition, as referring to ‘real’ phenomena. If this were the case, then there is a much more famous and widely accepted example of the same process in the figure of Arthur, who appears in the ninth-century Historia Brittonum as a Welsh national hero, with a strong folkloric dimension, and was turned during the rest of the middle ages into a figure in general English and lowland Scottish culture through the medium of written romances and pseudo-histories interacting with oral performance. Here is the construction of another imagined kingdom, with recurrent characters, to parallel the fairy one and overlapping with it.

The reasons for the great interest in fairies between 1560 and 1640 are easy to propose, and seem to have been the result of two interrelated phenomena. The first was an intense interest in the nature of spirits in general, and the attitudes which should be adopted to them, consequent on the complete re-evaluation of religious doctrine produced by the Reformation: those more familiar and central figures in Christian cosmology, angels and demons, were given similar attention, as were ghosts.Footnote ⁸⁶ The Reformation, however, was itself part of a wider European drive to examine, explore, and understand the world with a novel energy, of which humanism and the oceanic discoveries were early manifestations and which was to lead to what is familiarly called the Scientific Revolution. This would naturally have manifested in an impulse to explore and interrogate the concept of a fairy realm evolved in the later middle ages, and consider its benefits, warnings, and uses to an extent never attempted before.

As stated earlier, by adopting a historical perspective on the subject, concerned with dynamism and change, this study has departed from the methodology of folklorists such as Katharine Briggs and Jeremy Harte, who emphasized the continuities in fairy tradition. It also, however, departs from another major emphasis of theirs, on distinct regional beliefs, and from a stress made by Briggs, of making precisely delineated taxonomies of supernatural creatures. In doing so, it does not suggest that any of these methodologies – which have been common in the discipline of folklore studies – was faulty: as has been mentioned, there are abiding themes in depictions of the beings eventually classed as fairies, from medieval to modern times. Modern folklore collections do contain many specific and local varieties of such beings, though these tend to be more profuse in Gaelic areas. Medieval and early modern sources suggest, more hazily, a comparable range of entities, suggested by terms such as ‘bug’, ‘puck’, ‘goblin’, etc., and covering much the same basic kinds of being. The preoccupation of this study has been with those specifically called ‘elves’ or ‘fairies’ before 1640. None the less, it also challenges too great an emphasis on localism in one important respect: that the construct of the fairy kingdom as developed in the fourteenth century was found with a remarkable uniformity across England, Wales, and lowland Scotland by 1500. It may have been applied in different regional ways – hence the greater appearance of fairies in Scottish witch trials – but it remained essentially the same. This would make sense if it were indeed a late medieval development, achieved originally in a literary context, which found a wide and rapid acceptance; and perhaps, in that case, the greater localism and variety of fairy-like beings in the modern collections was actually due in part to a localization and diversification of an original, more homogeneous, belief system.

In suggesting such a sequence of development for such a belief system, it is interesting to note the parallel one in beliefs concerning witches, which is already firmly accepted by historians.Footnote ⁸⁷ Here, likewise, an ancient belief in witchcraft, which was often assimilated to Christian models by the middle ages, was put into a new and much more coherent model in the early fifteenth century, to redefine witches as practitioners of an organized, satanic, religion. This concept drew upon various popular mythologies, but was essentially a literary creation. As in the case of fairies, it represented the drawing together of previously disparate phenomena into a kingdom – another parallel act of spiritual state-building – but this time a clearly demonic one. It eventually reached Britain, and in the period between 1560 and 1640 became the subject of intense debate, from different viewpoints, which was reflected in the same sorts of source material and context as that over fairies and likewise spanned all social groups and most parts of the island. If the model for the evolution of fairy tradition proposed here is correct, then the two may perhaps be regarded as aspects of the same cultural process, in the same period, of a systematization of earlier, looser belief into a late medieval literary model which was then communicated to the British populace and energetically interrogated in the heightened speculative atmosphere following the Reformation. If this parallel is correct, a next step would be to look for other examples of ‘spiritual state-building’ in the period, and draw them together to determine if they form a single pattern. In this manner, not only would the historiography of fairy tradition be advanced, but a much broader advance would be made in that of the medieval and early modern imagination as a whole.