On the west wall of the south transept of Peterborough Cathedral, the last great Norman building of medieval England, is a curious and much-worn limestone Roman relief depicting two dancing figures in flowing robes, wearing pointed hats (Plate 1). The sculpture was reworked in the Middle Ages and long misidentified as a portrayal of two medieval abbots. In reality, the figures represent a man and a woman; they do not hold croziers but spears, and wear not mitres but the Roman pileus (the emblem of freedom), while they carry a bivalve shell between them to evoke their watery identity. Peterborough’s dancing figures are, in all likelihood, ‘a water god and his nymph consort’ once worshipped at the Romano-British shrine that almost certainly once stood on the bank of the River Nene where the medieval cathedral would one day come to be built.Footnote ¹

Plate 1 A Roman relief of dancing godlings, Peterborough Cathedral

As in so many cases where we encounter unique religious iconography from Roman Britain, we know nothing of the dancing godlings of Peterborough; except that by some strange chance their images survived in a Christian church, mistaken for something else, while the cult images of so many other shrines were buried, lost or defaced. Yet these dancing figures are at once strange denizens of an entirely alien religious world and unexpectedly familiar figures. For we know these divine dancers of unearthly beauty, unconstrained by human rules, albeit under another name and accompanied by a different set of cultural associations. All that remains of the divine dance of the nymphs is the ring of mushrooms or lush grass that children call a fairy ring, while the pileus now seems reduced to the red hat of the garden gnome.

This book is about those lesser divinities of Britain who, like the Peterborough pair, dance their way in one way or another through the history of the island: ‘small gods’, to borrow a phrase coined by the fantasy author Terry Pratchett. The ‘small gods’ or godlings are the nymphs, the gods of nature, the fauns and satyrs and the deities of fate and chance. They are a class of beings that while difficult to define, were still known to the inhabitants of this island in 1300 as they had been a thousand years earlier: before Christianity, before England and before the English language. Yet these small gods were by no means a fixed class of beings, and the godlings of 1300 and the godlings of 300 looked very different indeed. Whether any direct lines of descent can be traced between the godlings of medieval England, Wales and Scotland and the small gods of Roman Britain is a difficult question that this book seeks to address. But the story of Britain’s godlings is more interesting than a mere narrative of survival: it is a story of loss, invention, re-invention, imagination, subversion and the re-animation of belief.

Folklorists do not always spend very much time examining the origins of popular beliefs. An earlier generation of scholars was excessively confident in simplistic explanations for the origins of folkloric beings; partly in reaction to that, folklore studies has drifted towards comparative studies less focussed on the question of origins, while few historians have shown an interest in the roots of folklore. The idea that the building blocks of British folklore emerged in the post-Roman twilight of early medieval Britain is, in and of itself, uncontroversial; but it is an assertion often presented as an epistemological ‘black box’. Romano-Celtic and Germanic beliefs went in and, somehow, fairly familiar supernatural and folkloric beings came out. What happened in between is often presented as an irrecoverable mystery. It is the contention of this book that the black box is worth examining, especially in light of new methodologies and perspectives. Questions that seemed not only unanswerable but even unaskable a few decades ago are worth revisiting in light of the most recent scholarship, and among those questions is ‘Where did the supernatural beings of British folklore come from?’

The purpose of this book is to draw on the latest perspectives and methodologies to examine the origins of Britain’s folkloric fauna. It explores Britain’s godlings in the longue durée of the millennium between the Claudian and Norman invasions, and on into the High Middle Ages to the threshold of the early modern era. In doing so, Twilight of the Godlings deliberately transgresses the usual scholarly divide placed between Classical and medieval studies, which has traditionally been a particularly stark one in British history. But it is precisely the fact that folkloric beings seem to bridge the unbridgeable chasm in time between Roman and early medieval Britain that makes them a particular object of interest, and of importance not only for the history of belief but also for understanding the origins of medieval Britain.

Ever since Stuart Clark argued that ‘thinking with demons’ was a key to unlocking some little-understood aspects of the early modern world,Footnote ² historians have been increasingly willing to accept that studying culturally constructed beings – whether demons, angels, saints, or ‘small gods’ – has the capacity to illuminate the past in a unique way. The question of whether supernatural beings ‘exist’ is, of course, beyond the capacity of the historian or the folklorist to answer – but that they exist as cultural artefacts there can be no doubt, and they are more than simply ideas. In societies where they are accepted as real, supernatural beings function as a category of person, and have all the capacity of real people to be embodiments of a society’s preoccupations. The supernatural beings whose existence is accepted (or indeed contested) within a community reveal its self-understanding, its inner tensions, its taboos and its understanding of the familiar and the unfamiliar, the normal and the strange.

To Clark’s ‘thinking with demons’ Simon Ditchfield later added the idea of ‘thinking with saints’,Footnote ³ while others have made a similar case for the historiographical potential of belief in angels.Footnote ⁴ Michael Ostling, meanwhile, has advocated ‘thinking with small gods’ (the enduring godlings of folklore) as a means of engaging with wider questions of ‘continuity and change, tradition and modernity, [and] indigenous religion and its redefinition’.Footnote ⁵ This book takes Ostling’s observation as its inspiration, arguing that understanding the ‘small gods’ of Britain in the longue durée opens hitherto unexplored perspectives on questions of cultural and religious survival, creativity and adaptation in the millennium-long transition from the Romanised Iron Age society of Roman Britain to the medieval Christian world.

‘Folkloric beings’ are non-human supernatural entities of folklore, usually endowed with a human-like personality or living in human-like societies, and called by a great variety of names across cultures (and even within the cultures of the island of Great Britain). As I shall argue in Chapter 1, the names by which these beings are called are usually less important than the cultural ‘niches’ they occupy. Indeed, focusing on names can be a hindrance to historical understanding, cementing stereotypical and limiting notions of what we expect these beings to be. Because they are cultural creations, folkloric beings can change almost unrecognisably over time, and names thus serve as a poor guide to their nature. The ‘demon’ of today’s Christian mythology is quite different from the daimo¯n of ancient Greece; and if we did not know the process by which a name given to godlings and spirits in Greek religion came to be adopted for evil spiritual beings in modern Christianity, the etymological connection between the two words, in and of itself, would be almost entirely useless.

While the existence of folkloric beings undoubtedly helped people in the past to account for events and aspects of the surrounding world that were not otherwise explicable or subject to their control,Footnote ⁶ reductive or functional explanations of such beliefs are ultimately inadequate because the things people believed about godlings and fairies clearly far exceeded any functional social, psychological or ‘pre-scientific’ purpose we might propose. Godlings cannot simply be ‘explained away’ as psychosocial phenomena, because these beings caught people’s imaginations. While speculations as to the functions of popular belief can have value – and this book does not entirely hold back from such speculations – if we are forever seeking ‘rational explanations’ for folkloric narratives, there is a danger that we will be blinded to the significance of those narratives to most people at the time when they were originally told. This book therefore approaches godlings as experiential and cultural realities in the period under discussion, because that was how they were encountered by people at the time.

Supernatural Beings: The Search for Origins

The story of the search for the origins of Britain’s supernatural beings is part of the history of the study of folklore, whose beginnings can perhaps be traced to the development of ethnography in the fifteenth and sixteenth centuries – which was in turn a response to the need to understand unfamiliar cultures left unexamined by ancient ethnographers like Herodotus. Margaret Meserve has linked the rapid appearance of the Ottoman Turks in Asia Minor and the cultural trauma of the Ottoman capture of Constantinople in 1453 with an explosion of learned interest in the Turks, as well as other Asian peoples such as the Mongols and Tatars.Footnote ⁷ I have shown elsewhere that European authors began to take a detailed interest in the pagan peoples of the Baltic at the same time.Footnote ⁸ However, it was the European encounter with the indigenous peoples of the New World that brought true urgency to the ethnographic project, for here were culturally alien peoples without writing, and without a presence in the Classical record, who could be understood only via ethnography and the recording of their stories and customs.Footnote ⁹ Dan Ben-Amos has argued that the encounter with the New World and the ethnographical literature it produced influenced early antiquarians in Britain (such as William Camden) to pay attention to stories and popular customs as an integral part of the antiquarian project.Footnote ¹⁰

If the recording of folklore was part of the early modern antiquarian project from the very beginning, the first British antiquarian to devote a book solely to ‘popular antiquities’ (what would later come to be known as folklore) was John Aubrey. In his Remaines of Gentilisme and Judaisme, compiled in 1687–1688 but not published until the late nineteenth century, Aubrey presented a miscellany of folklore set alongside allusions to Classical literature that seemed to Aubrey to resemble English folk beliefs and customs. Aubrey’s work imitated the structure of Ovid’s Fasti (a series of poetic aetiologies of Roman customs and rituals), and there was nothing new about using the Classical record as a comparative interpretative framework to understand other cultures. However, Aubrey’s decision to engage in ‘othering his own culture’ was unusual;Footnote ¹¹ his approach to fairy lore differs markedly from that of his Scottish contemporary Robert Kirk, for example, whose chief aim was to provide a theologically coherent account of what fairies might be rather than tracing the origins of belief in fairies.Footnote ¹²

In common with his contemporaries, Aubrey’s view of pre-Christian religion (‘gentilisme’) was informed by interpretatio Romana, the tendency to interpret all forms of paganism through the lens of Roman religion (a religious hermeneutic for which the Romans themselves were responsible).Footnote ¹³ Aubrey displayed a specific interest in popular belief in folkloric beings, following his methodology of interpretatio Romana to conclude that the archetypal English fairy Robin Goodfellow could be identified with the Roman god Faunus.Footnote ¹⁴ Aubrey identified the fairies with ‘the nymphes, the ladies of the plaines,/The watchfull nymphs that dance, & fright the swaine’, quoting Theocritus.Footnote ¹⁵ He also identified Pliny the Elder’s report that ‘In the solitudes of Africa a kind of men appear on the road, and vanish in a moment’ as encounters with the fairies.Footnote ¹⁶ Although never articulated, the implied hypothesis behind Aubrey’s speculations was that, at some time in the past – and presumably at the time of the Roman occupation – the religion of Britain was essentially Roman. The ‘Country Gods’ of the Romans degenerated into Robin Goodfellow and the fairies.

Aubrey’s basic thesis that Britain’s folkloric beings were the degenerate remnants of pagan gods remained at the heart of most attempts to understand the origins of folkloric beings into the twentieth century, even if Aubrey’s emphasis on Classical and Roman origins was abandoned in favour of a ‘British’ or ‘Celtic’ mythology, supposedly more ancient than the imported mythology of Greece and Rome.Footnote ¹⁷ Even today, the idea that the supernatural otherworlders of European folklore are gods who have somehow been diminished or demoted and become fairies is a dominant strand of thought about the origins of folklore. But while such demotion can sometimes be argued convincingly in individual cases, the idea that all folkloric beings are diminished gods ignores the fact that ancient pagans, too, had minor spirits as part of their belief systems. The application of Occam’s razor to the problem should guide us to examine the ‘small gods’ of antiquity first, before the formulation of any thesis of ‘demotion’ or diminution becomes necessary.

The idea of ‘Celtic’ mythology largely derived from the twelfth-century imagination of Geoffrey of Monmouth, giving rise to tales of ancient British kings such as Lear, Cymbeline and (most notably) Arthur. The eighteenth- and nineteenth-century rediscovery of the medieval Welsh imaginative literature contained in the White Book of Rhydderch and Red Book of Hergest (known today as the Mabinogion) further transformed understandings of ‘Celtic’ culture, although perceptions of ‘British mythology’ were also distorted by the forgeries perpetrated by Iolo Morganwg (1747–1826).Footnote ¹⁸ Directed interest in folkloric beings first stirred at the turn of the nineteenth century, motivated by a mixture of Romanticism, patriotism and literary-critical interest in earlier writers who made use of the fairies, such as Chaucer and Shakespeare.Footnote ¹⁹

Sir Walter Scott’s 1802 essay ‘On the Fairies of Popular Superstition’ in his Minstrelsy of the Scottish Borders represented an early detailed exploration of the origins of fairy lore. Scott argued that the origins of Britain’s folkloric beings ‘are to be sought in the traditions of the east, in the wreck and confusion of the Gothic mythology, in the tales of chivalry, in the fables of classical antiquity, in the influence of the Christian religion, and finally, in the creative imagination of the 16th century’.Footnote ²⁰ Whatever we may now think of Scott’s interpretation, his basic insight that the origins of folkloric beings are composite and complex remains valid, and represented a significant advance from Aubrey’s simplistic attempt to equate beings across disparate cultures, like Faunus and Robin Goodfellow. In his Letters on Demonology and Witchcraft (1830), however, Scott supplemented his earlier theories with an additional hypothesis that would prove very influential throughout the nineteenth century and beyond:

Scott’s attempt to ‘euhemerise’ folkloric beings (identifying them with historic human populations), freighted as it was with racial and colonial prejudices, was enthusiastically taken up by subsequent authors and became a regrettable cul-de-sac of Victorian speculations about the origins of folkloric beings. The instinct to euhemerise, and to find a ‘scientific’ or historical-realist explanation of fairies in a half-remembered history, was rooted in the Enlightenment. The earliest British writer to suggest that the fairies might be a hidden race of diminutive humans was John Webster, writing in 1677:

After a discussion of the reality or otherwise of pygmy races, Webster concluded that pygmies probably did exist, and that fairies

Although Webster does not say it outright, he implies here that pygmies may once have lived in Britain (in order to account for belief in fairies). This approach reached its apogee in David MacRitchie’s ‘pygmy theory’ of fairy origins, which theorised that fairy lore represented folk memory of diminutive peoples driven to hills and caves by invaders.Footnote ²³ The ‘pygmy theory’ would even inspire an entire subgenre of Victorian literary fiction by authors such as Arthur Machen and John Buchan, in which modern-day Britons unwittingly stumble upon savage races of troglodytes.Footnote ²⁴ Remarkably, even one popular book on folklore published in 2022 could still be found advocating the theory.Footnote ²⁵

However, MacRitchie’s racialised euhemerism was never entirely triumphant in the study of fairy origins. The influential folklorist Thomas Keightley, whose Fairy Mythology (1828) was an early work of serious folklore studies in English, was influenced by the comparative approach of the Brothers Grimm. Keightley’s understanding of fairy lore was rooted in a kind of functionalism that posited universal ‘laws’ by which primitive peoples arrived at belief in both gods and godlings:

Keightley considered it most probable that the word ‘fairy’ derived from the Persian word peri,Footnote ²⁷ and mentioned only in a footnote the Breton antiquary Jacques Cambry’s (broadly correct) view that French fée could be linked to Latin fatua, as articulated in Cambry’s Monumens Celtiques (1805):

Cambry was by no means right about all of this (‘maid’, for example, is a thoroughly Germanic word), and his suggestion that ‘the Romans knew the fairies’ will be discussed at length in this book. However, Cambry’s basic insight that ‘fairy’ can be linked to a cluster of etymologies around the Latin verb fari (from which fatua derives) has stood the test of time.

Keightley’s decision to ignore the possibility of a Classical origin for the fairies may speak to his insistence on seeing fairy lore as a belief of primitive peoples and therefore unconnected with the prestigious civilisations of Greece and Rome. Furthermore, his decision to promote a Persian origin for the word ‘fairy’ contributed to an exoticising tendency that strained to see Eastern origins in European folklore. In the second half of the nineteenth century, the popularity of Spiritualism led some to speculate that Britain’s folkloric beings were best understood as the psychically evolved humans of remote antiquity. However, Spiritualists were also drawn to the idea of fairies as ‘psychic insect life’, and the affair of the ‘Cottingley Fairies’ (in which Sir Arthur Conan Doyle was duped into advocating the reality of a set of fabricated fairy photographs) merely encouraged the idea of insect-like ‘flower fairies’.Footnote ²⁹ Needless to say, insectoid fairies are without any basis in folklore; they are instead a literary creation of seventeenth-century poets like Robert Herrick.Footnote ³⁰

The scholarly impetus to understand fairy lore in the twentieth century emerged from literary criticism, particularly the attempt to understand the origins of the fairies portrayed in medieval romances.Footnote ³¹ Shakespeare studies – particularly interest in the fairies of A Midsummer Night’s Dream – also played a key role, and no-one is more closely associated with the study of Shakespeare’s fairies than the folklorist Katharine Briggs, whose 1952 doctoral thesis dealt with ‘Some Aspects of Folk-Lore in Early Seventeenth-Century Literature’.Footnote ³² Briggs’ 1957 article in the journal Folklore, ‘The English Fairies’, launched a lifetime of scholarship devoted to fairy folklore. In that article Briggs’ opening observation about the study of British fairy beliefs remains as true now as it was then:

Here, as in her subsequent work, Briggs adopted a taxonomic approach to the fairy realm (the very title of her 1959 book The Anatomy of Puck seems to jest with the idea of bringing a scientific level of precision to what might appear the most unscientific of subjects) and her classifications of trooping, solitary, tutelary and nature fairies remain influential. Briggs differed from earlier authors by insisting on the need to engage with folklore on its own terms without the intrusion of functionalist and historical-realist speculations. She advanced the theory that fairy belief was in some way connected with the dead,Footnote ³⁴ which in turn suggested the idea that the fairies were in some way the decayed remnants of a cult of the ancestors associated with prehistoric landscape features like barrows.

In The Fairies in Tradition and Literature (1967) Briggs continued to grapple with the difficult question of fairy origins, concluding that

Briggs’ work continues to form the basic foundation for the study of British fairy lore today, although there are aspects of her approach that might now give us pause. Her preferred theory that fairies are remnants of a cult of ‘half-deified spirits of the dead’ is unaccompanied by much evidence that such a cult existed beyond prehistory. The Romans venerated the di manes but there is no compelling reason to believe that veneration of the dead in Roman Britain was much different from anywhere else in the empire, and the pagan Anglo-Saxons’ preference for establishing cemeteries close to prehistoric features like barrows constitutes insufficient grounds to propose an ongoing cult of the ancient dead. If fairies are associated with barrows and other human-made earthworks, they are equally associated with natural mounds and hills. It is difficult to avoid the conclusion that Briggs, like many folklorists of her generation, was drawn into advocating continuities of belief from remote prehistory that seem rather unlikely.

A further problematic aspect of Briggs’ approach is her liking for taxonomizing supernatural beings, which culminated in her multi-volume Dictionary of Fairies (1976). As Simon Young has observed, in ‘the drive to order British supernatural creatures … dialectal differences have been unknowingly turned into folklore differences’,Footnote ³⁶ which is just one of the problems that can arise when trying to separate folkloric beings like species of real-world fauna. Briggs speculated that ‘the lesser deities’ of the Roman world ‘had descended into being fairies’, but did not offer any explanation of how this process occurred.Footnote ³⁷ It is unclear whether Briggs presumed that such an enquiry was impossible or considered that it lay beyond her expertise – or whether, as a folklorist interested primarily in folklore since the early modern period, questions about the more remote origins of fairies did not particularly interest her. More recent authors have followed in Briggs’ footsteps by setting aside the question of the fairies’ historical origins; Richard Sugg, for example, while thoroughly discussing folkloric explanations for where the fairies came from, makes little attempt to explain where he thinks they really came from.Footnote ³⁸

Keith Thomas reflected the attitude of many scholars of early modern fairy belief in showing more interest in its social and cultural function than in where it came from, observing only that ‘Ancestral spirits, ghosts, sleeping heroes, fertility spirits and pagan gods can all be discerned in the heterogeneous fairy lore of medieval England, and modern enquiries into fairy origins can never be more than speculative’.Footnote ³⁹ Similarly, referring to godlings, Euan Cameron declared that ‘It is … an open question as to whether the raw material of supernatural beliefs actually has a history’.Footnote ⁴⁰ Thomas’ and Cameron’s scepticism has also been echoed by Ronald Hutton, who considers that the ‘ultimate root’ of fairy beliefs is irrecoverable.Footnote ⁴¹ Indeed, Hutton has argued that, in contrast to Olympian gods who survived under the form of cultural allegories, the ‘small gods’ of the ancient world disappeared.Footnote ⁴² However, he is somewhat less pessimistic than Thomas, arguing that late medieval and early modern fairy belief is essentially a literary construct built around a variety of folkloric beings who existed in popular belief within no particular conceptual framework: ‘a late medieval development, achieved originally in a literary context, which found a wide and rapid acceptance’.Footnote ⁴³ This is displayed in works such as Sir Orfeo, a romance composed around 1300 that retold the myth of Orpheus and Eurydice while replacing Hades with the ‘King of Fairy’ and the Greek underworld with a fairy kingdom. In Hutton’s view, the idea of the fairy kingdom then migrated from literature into folklore, becoming an established popular belief by the fifteenth century.Footnote ⁴⁴

Hutton considers the fairies to be the descendants of ‘land spirits’ and ‘rural spirits which had no obvious place in Christianity’. While water spirits and household spirits can be found throughout Europe, the ‘fairies proper’ who constitute a parallel society with its own hierarchy are ‘strictly a northern tradition’, and these fairies can be considered ‘neither personifications of nature nor deities’. Hutton concludes that the fairies

While it is certainly possible for literary tropes to cross over into the realm of folklore, Hutton’s thesis that popular English fairy belief is an essentially literary construct of the late Middle Ages lacks plausibility. Many individual instances of folklore originated in literature, but there is no obvious precedent or parallel for a set of beliefs so widespread and deeply held as belief in the fairy realm crossing over into folklore within a fairly short period. Furthermore, parallels to English, Welsh and Scottish belief in a fairy realm can be found in Ireland and Brittany, suggesting a more ancient and fundamental origin. On the other hand, Hutton’s argument is a helpful corrective to the long-established tendency to assume, uncritically, that folklore always has ancient and immemorial origins. As he has argued recently, ‘Christian Europe, both in the Middle Ages and after, was capable of developing new superhuman figures which operated outside of Christian cosmology’.Footnote ⁴⁶ It is the argument of this book that fairy belief, rather than being constructed from literary sources at a particular moment in time, is the result of sustained interaction between learned and popular culture over an extended period – and the fairies were just as likely to be novel creations as survivals from the immemorial past.

Alaric Hall’s definitive work on elves in Anglo-Saxon England has shown the potential of attentive study of linguistic sources like glosses to clarify the nature of different supernatural beings mentioned in Old English texts. Hall introduced a new standard of rigour to the field that should serve as a model for future study of the origins of folkloric beings, and makes it impossible to return to the vague generalisations of some past scholarship.Footnote ⁴⁷ Hall’s analysis also reveals how little can be stated with certainty about Anglo-Saxon belief in elves, in spite of the fact that the name of the elves migrated into Middle English as a term for the beings later called fairies. On the way, however, Hall highlights the richness and complexity of Anglo-Saxon England’s supernatural fauna, a theme subsequently taken up by scholars such as Sarah Semple and Tim Flight.Footnote ⁴⁸ Hall’s work throws into doubt another traditional view of the origins of England’s fairies as the unproblematic descendants of the elves of Anglo-Saxon England, which seems unsustainable insofar as Hall shows the elves were very different beings from the fairies who later assumed their name.

In addition to Hall, three scholars who have made important contributions to the enquiry into the origins of Britain’s folkloric beings are Emma Wilby, Diane Purkiss and Michael Ostling. Through an analysis of the evidence of Scottish witch trials, Wilby argued that there was evidence of ongoing animistic and shamanistic belief underlying Scottish popular religion, suggesting that fairy belief is perhaps best understood as the persistence of an animistic understanding of a deified nature as an undercurrent to other religious beliefs.Footnote ⁴⁹ While this book is not uncritical of Wilby’s approach, nor of the application of the concept of animism itself, her basic insight that fairy belief may be linked with animism is not one that ought to be set aside entirely – and, once again, it serves as a helpful corrective to the idea of fairies as diminished deities.

Diane Purkiss’ Troublesome Things (2000) is undoubtedly one of the most important books of recent decades on folkloric beings, and while its psychosocial approach is not primarily directed at the question of historical origins (as Purkiss acknowledges), Purkiss was prepared to challenge many of the long-held assumptions (largely derived from Briggs) about fairy origins. Purkiss stressed the parallels between ‘Celtic’ and Near Eastern beliefs and even went so far as to suggest that Celtic ideas about the fairies could have come from the Classical world.Footnote ⁵⁰ While Purkiss’ specific proposal of Greek influence on the pre-Roman Celts seems rather unlikely (and considerably less likely than Roman influence on Britain during the period of occupation), she can be credited with reviving the idea of Classical influence in general as an explanation for fairy belief. Purkiss’ insight has been taken forward by at least one scholar, Angana Moitra, who has sought to trace in detail the origins of one figure, the Fairy King, over the cultural longue durée.Footnote ⁵¹

Beyond the English-speaking world, the French historiography of fairy origins has generally been focussed on explaining the origins of the folkloric themes of the medieval romances. This is a historiography that cannot be neglected in a history of British folkloric beings, because the stories of British and French folklore intersect in the territory of Brittany: an area that is geographically France but historically, culturally and linguistically Brittonic. For Francisca Aramburu, Catherine Despres, Begoña Aguiriano and Javier Benito, medieval fairy belief was a complex amalgam of cultures, yet fairies could nevertheless be said to conflate the Parcae, the Gaulish Deae Matres and the godlings of nature and birth.Footnote ⁵² Claude Lecouteux’s emphasis, by contrast, was on the fairies as embodiments of dream, destiny and fantasy, as well as on the fairy’s role as a psychic double,Footnote ⁵³ while Pierre Gallais eschewed a historical approach altogether, emphasising the universality of the figure of the fairy across all human cultures.Footnote ⁵⁴ Laurence Harf-Lancner echoed Keith Thomas’ scepticism, arguing that since medieval folklore of fairies is now subsumed entirely in literary sources, it is impossible to recover.Footnote ⁵⁵ On the whole, the focus of French historiography is not so much on the historical origins of fairies as on their functions, universality, archetypal character and roles in literature.

Notable recent advances in the understanding of the origins of fairy lore have been made by Michael Ostling, Richard Firth Green and Ronald Hutton. In an important volume that brings together scholarship on belief in godlings across the world, from Estonia to Zambia,Footnote ⁵⁶ Ostling has made the case for reclaiming ‘small gods’ from a realm of cultural studies that tends to take them out of space and time, and returning them firmly to history.Footnote ⁵⁷ By adopting a broad understanding of ‘small gods’ as ‘animistic “survivals” problematically present within a Christianity that attempts to exclude them’,Footnote ⁵⁸ Ostling and his fellow contributors are able to advance the comparative study of folkloric beings in both a European and an international context. Ostling’s firm insistence that ‘small gods’ are a category in their own right, found in both pagan and Christian contexts, is an important corrective to the old idea of folkloric beings as diminished gods. A further significant contribution to the historiography of fairy belief has been made by Richard Firth Green, who has challenged the ‘Celtic fallacy’ that has hampered the study of what is, in reality, a set of beliefs found throughout Europe.Footnote ⁵⁹ While sceptical of attempts to trace the origins of fairy lore or taxonomise the fairies, Green expanded the evidential base for the study of medieval fairy belief by noting that many stories of demons in sources such as pastoral literature and saints’ lives were, in fact, narratives about fairies. Similarly, Ronald Hutton’s insightful recent study of the figure of the Fairy Queen in his book Queens of the Wild builds on his earlier scholarship on the origins and antecedents of fairy belief.Footnote ⁶⁰

Between them, these innovative scholars have brought the study of the origins of folkloric beings out of the ‘Celtic twilight’ in which it had long languished, contributing important insights that, collectively, transform the conversation about folkloric beings. Firstly, the study of folkloric beings should be conducted as rigorously as any other historical investigation, and should make use of all available evidence, including the linguistic and the archaeological. Secondly, folkloric beings should be seen as the product of continuous interactions between oral and learned culture, including literature. Thirdly, folkloric beings may be viewed in the context of underlying animistic worldviews across the longue durée of cultural history, even if the persistence of ‘animism’ should sometimes be treated with a degree of scepticism. Fourthly, there is a strong element of influence from the Classical world in Britain’s folkloric beings. And fifthly, Britain’s ‘small gods’ are best understood not in isolation, but in their European and international context.

Approach of the Book

In 2014 Ronald Hutton observed that there existed no modern history of British fairies, in the sense of a book that approached fairies historically with a focus on ‘change over time’.Footnote ⁶¹ By contrast, ambitious studies of the history and origins of other classes of folkloric being have been undertaken, such as Daniel Ogden’s study of dragons and Simon Young’s work on boggarts.Footnote ⁶² It is this deficit that the present book seeks to address, while avoiding (as far as possible) a teleological approach that merely explains how the fairies of Shakespeare (for example) or the fairies of nineteenth-century folklore came about. Most studies of British fairy belief have hitherto been hampered either by insufficient curiosity about (or willingness to investigate) the question of origins, or by a methodological aversion to a historical approach to folkloric beings. As the Classicist T. P. Wiseman has observed, however, anthropological and comparativist approaches to beliefs and customs have a tendency to argue synchronically, taking inadequate account of change over time. In reality, ‘any community’s dealings with its gods must reflect, at some level, its own needs and preoccupations, and adjust, with whatever time-lag, as those needs and preoccupations change’.Footnote ⁶³ Similarly, belief in ‘an essentially constant rural popular cosmology that persisted through all the dramatic developments in elite and official belief’ is ahistorical.Footnote ⁶⁴ The discipline of folklore on its own lacks the methodological resources to investigate a question – the origins of Britain’s folkloric beings – that requires the analysis of such a variety of sources, over such a long period, that an interdisciplinary approach is required, drawing on the perspectives offered by history, Classics, linguistics and archaeology. At the same time, the present study does not claim to be a complete study of the totality of British lore of supernatural beings, and is fairly narrowly focussed on questions of origins and development.

A history of folkloric beings can be written either from the present to the past or from the past to the present. If it is written from the present to the past, its focus will be on discerning the building blocks of modern belief in the past; but a shortcoming of this approach is that it presumes the character of modern folklore is the final goal towards which history has been moving. Yet the history of popular belief is full of extinct strands of belief that have had very little influence on the present, but are no less historically important for that. However, the alternative approach – telling the story of Britain’s folkloric beings from the past to the present – also presents a difficulty, because it requires us to start off with some idea of the kind of beings we are pursuing through time. Since folkloric beings are cultural constructions rather than empirical realities, this requires us to understand the factors that may have led to particular cultural constructions at particular times – a very challenging demand for remote and poorly evidenced eras of the past.

Tracing the history of belief in ‘small gods’ is difficult, certainly, and it will always be an enterprise that produces only partial results. It is a cliché of historiography that most historical sources are produced by elites, and belief in ‘small gods’, for much of the period under discussion, was characteristic of the secular culture of the unlearned elite and the non-elite lower echelons of society. Yet historians are now well accustomed to the challenge of extracting the traces of non-elite culture from elite sources, and there is no reason to think that a history of fairy belief should be harder to write than, say, a history of Romano-British religion, where the evidence is similarly scattered and problematic. Furthermore, as Richard Firth Green has observed, there is no reason to believe that fairy beliefs were not generated (as well as consumed) by elites, and the assumption that such beliefs always came ‘from below’ is itself a projection of modernity onto the past.Footnote ⁶⁵ It is my contention, therefore, that historians have hitherto been excessively pessimistic about the potential of tracing the origins of Britain’s folkloric beings. While this book keeps in view the question of how late medieval and early modern fairy belief came about, its focus is primarily on examining the development of belief in broadly defined ‘small gods’ in Britain from beginning to end, on the basis that (as I shall argue in Chapter 1) it is possible to trace the outlines of the kind of being we might consider a ‘small god’.

Ferdinand Braudel observed in the 1960s that basic human relationships with nature lie beneath the ever-changing drama of history, as well as the shape of the human mind itself. These are factors that undergo slow processes of change – if, indeed, they change at all within historical time.Footnote ⁶⁶ Since belief in godlings is entwined intimately with humanity’s relationship with nature – to a greater extent, perhaps, than other aspects of religion and belief – its history is an ideal subject for a survey in the longue durée. As Joyce Salisbury has argued, ‘if people believed the same things in two separate periods of time, we might assume a similar belief in the central, though undocumented period’, and the potential of a longue durée examination of belief in godlings from the Roman period to the later Middle Ages has not yet been explored.

Archaeologists of prehistory and anthropologists are accustomed to making use of ‘cable-like’ arguments that contain distinct, separate strands of evidence, and allow for the other strands to cover the gap if there is an evidential lacuna – as often occurs when dealing with the prehistoric world.Footnote ⁶⁷ Some of the ‘cables’ that can be constructed for the study of folkloric beings in Roman, post-Roman and early medieval Britain are significantly stronger than those deployed by archaeologists to account for religious behaviour in prehistory, although the presence of evidential lacunae is no less of a problem. However, the historical and conceptual advances made in the understanding of godlings by the scholars already mentioned in this introduction should dispel the notion that Britain’s folkloric beings cannot be studied historically, as cultural artefacts within time. This study takes those scholars’ determination to study godlings historically as its inspiration.

A study of Britain’s godlings in the longue durée has the capacity to challenge many cherished assumptions about the origins of British folklore, such as the ‘Celtic myth’ that folkloric beings such as fairies can be traced to an imagined ‘Celtic twilight’, either before or after the Roman occupation of Britain – or to the pagan religion of the Anglo-Saxons, of which we know virtually nothing. Related to the ‘Celtic twilight’ myth is the more or less uncritical use of sources from medieval Ireland to support arguments about Britain. While it is true that theonyms (names of deities) in Irish and Welsh are sometimes linguistic cognates, and therefore it would be absurd to deny any sort of religious or cultural connections between the two islands in ancient times, the belief that Irish mythology can illuminate ancient British beliefs rests on assumptions about a pan-Celtic cultural identity that is more of a nineteenth-century construct than a historical or archaeological reality. Even if, for the sake of argument, we imagine that Irish and British beliefs were more or less identical in the pre-Roman Iron Age (which is unlikely), the historical paths taken by Ireland and Britain in the Roman and early medieval periods were so dramatically different that we should not expect much similarity in folk beliefs about the supernatural by the time of the Anglo-Norman invasion in the late twelfth century. That invasion resulted in the eventual imposition of English words like ‘fairy’ on Ireland’s supernatural beings, but Iron Age Ireland did not experience directly the religious influence of Rome, encountering it only secondhand through a Christianity transplanted from late Roman Britain. Ireland’s supernatural world is a unique one, which cannot and should not be imposed as a framework on other cultures.

As well as setting aside the myth of a common ‘Celtic’ identity, this book challenges the idea that buried ancient mythologies can be constructed from medieval literary sources – whether from Britain or Ireland – that are, in fact, works of imaginative fiction. It is the argument of this book that the ‘small gods’ of early medieval England were largely fresh cultural constructions of the period, confected at need in the aftermath of Christianisation and under the influence of Christian learning, against a background of the detritus of Roman, Brittonic and Anglo-Saxon paganisms. The idea that Britain’s godlings were ‘pagan survivals’ should be largely (albeit not entirely) set aside. The ‘small gods’ that later became Britain’s elves, fairies and giants are not Christian, but they are the non-Christian artefacts of a Christian culture. Owing to the influence of learned commentary on popular culture, and the significance of Latin and Classical learning within that commentary, it is the contention of this book that Roman pagan religion was by far the most important cultural background for Britain’s early medieval godlings – albeit rarely directly, through a process of direct survival from Roman Britain. Instead, Roman religion came to influence British folk belief through the writings of the Church Fathers. If the ‘small gods’ are not the children of Rome, they are at least Rome’s grandchildren.

Sources

The sources for the study of a question as large as the origins of Britain’s folkloric beings are very diverse indeed, ranging from the insights of lexicographers to the discoveries of archaeologists. Popular belief in godlings is, by its very nature, very difficult to trace owing to the truism that most written evidence was produced by elite sources who were less likely to mention or discuss such beliefs. In the case of Roman Britain, our direct evidence for godlings is almost entirely archaeological and epigraphic, depending on figurative representations and inscriptions. However, our understanding of that artistic and epigraphic information depends in turn on the written evidence for Roman religion in Rome itself, especially those writers (like Ovid and Varro) who provide the most information about popular religion. The likely distance that existed between Roman belief and practice in Roman Britain and more richly textually evidenced provinces of the Roman world makes the use of such sources problematic but, nevertheless, unavoidable.

The evidence for early medieval belief in godlings is to be found in a diverse range of sources, including saints’ lives, chronicles, glosses, penitentials, imaginative literature and learned commentaries on the Bible and Classical texts. One source of evidence for fairy lore that (as Simon Young has observed) has been overlooked until recently is place names.Footnote ⁶⁸ This is a source extensively explored by Sarah Semple in her analysis of folkloric beings in the English landscape.Footnote ⁶⁹ Literary sources for belief in folkloric beings are difficult to use for a number of reasons, including differences in language between the text and the language used by those who held the beliefs discussed. Since folkloric beings are cultural creations whose conceptual character is closely tied to the words used to name them in specific languages, such linguistic gaps can be very significant indeed. Thus we are often faced with sources in Latin discussing belief in beings by Old English speakers, when those beings’ Old English names are not given. If a Latin source by an Anglo-Saxon author names nymphae, for example, we have no way of being sure what Old English speakers called these beings.

Furthermore, literary sources were produced by people with a much higher level of education than most ordinary people, so it is imperative to be sensitive to the importation of learned assumptions when folkloric beings surface in medieval sources. A learned commentator with a theological education might be more inclined to demonise folkloric beings, for example, than the majority of people for whom those beings were a cultural reality. Literary texts, especially in the early Middle Ages, often date from decades or centuries after the events they describe, and we must always be attentive to the importance of the genre and character of texts. A sceptical stance regarding what can be recovered of popular belief from elite sources is always possible, and often justifiable, but the notion that all early medieval writing was a conversation between the learned that bore little or no relationship to ordinary people’s beliefs is one that stretches credulity. People in early medieval societies who worked for the church and acquired Latin learning were not thereby cut off from their communities or from the societies they grew up in, and if we set up an excessively rigid dichotomy between elite and non-elite culture we are in danger of reviving the old myth of a pagan peasantry co-existing with a Christian elite.Footnote ⁷⁰

The tendency to give undue weight to medieval works of imaginative literature as reliable accounts of folklore has been a persistent hindrance to the study of the origins of fairy lore, although it is also important not to neglect texts of a literary character. However, scholars of the fairy theme in medieval literature such as James Wade now often focus on the ‘internal folklore’ of the text rather than on attempting to draw connections with the putative world of belief that lay beyond it.Footnote ⁷¹ The approach adopted in this book is to treat the fairy theme in the medieval romances as an indication of the importance of fairies in medieval culture, but not as a source from which it is possible to mine or recover folklore or popular belief. In the same way, the idea that the beliefs of distant ages can be reconstructed from folklore collected in the nineteenth and twentieth centuries is not a proposal that the historian should seriously contemplate.

Structure of the Book

The structure of the book is partly chronological and partly thematic, with Chapters 2, 4 and 5 tracing the evolution of belief in godlings between the Roman period and the later Middle Ages while the focus of Chapters 1 and 3 is on questions of definition and the impact of Christianisation, respectively. Establishing what we mean by ‘godlings’ and ‘small gods’ is crucial to studying their origins, and therefore Chapter 1 deals with the issue of identifying godlings. The chapter approaches ancient understandings of minor spirits in the Roman world and considers how godlings can be studied over the longue durée, in this case the millennium or more from the Roman Iron Age to the later Middle Ages. The chapter evaluates the significance and appropriateness of the concepts of animism, ‘shamanism’ and therianthropy to understanding the phenomenon of godlings in ancient Britain, before suggesting certain key characteristics of ‘small gods’: for example, ambiguity as to number and gender. The nature of Britain’s godlings is illustrated through a study of the parallels between the godling, the hag and the witch, while the question of long-term survival is considered through the most promising example of possible genuine survival of veneration of godling-like beings: the spirits associated with water sources throughout Britain.

Chapter 2 turns to a more chronological approach, examining the ‘menagerie of the divine’ that was Roman Britain. Here there is a great deal of archaeological and epigraphic evidence for the veneration of a multiplicity of divine beings – albeit often accompanied with little context that helps us to understand exactly what the significance of such cults was. The chapter introduces the main categories of godlings in Roman Britain, including genii loci, nymphs, mother goddesses and deities of nature, arguing that such cults became much more significant during the fourth-century ‘pagan revival’ that followed the accession of the pagan emperor Julian in 361. In particular, the extraordinary cult of Faunus revealed by the Thetford Treasure – along with other ecstatic nature cults apparently testified by the archaeological record – suggests that the fourth century was an important time for the development of distinctive and inventive Romano-British interpretations of Roman pagan religion. However, the relationship between Romano-British cults and subsequent strains of belief remains unclear, and no definitive link can be established with the early Middle Ages.

Christianity was a growing religion in Britain from the 330s onwards, and Chapter 3 tackles the difficult question of the relationship between Christianity, Christianisation and godlings. The chapter examines the phenomenon of the Christian demonisation of pagan cults, arguing that it was a more complex process than mere condemnation and suppression. Demonisation inadvertently produced the potential for the survival (and even re-invention) of some of the beings it targeted. Through comparisons with the better evidenced Christianisation of other cultures in Europe and further afield, the chapter develops an interpretative framework for the likely changes undergone by popular religion in the lengthy conversion period. The framework includes the likely ‘undemonisation’ of formerly demonised entities and the creative ‘re-personification’ of supernatural forces to account for the survival and reinvention of godlings in a Christianised society – where godlings should not so much be seen as ‘pagan survivals’ but rather as non-Christian artefacts of the Christianisation process.

Chapter 4 examines in detail the early medieval evidence for godlings in Britain, from both Brittonic and Anglo-Saxon sources, dealing in turn with the main categories of folkloric beings such as fauns, elves, the various categories of supernatural women, pygmies and giants. The chapter stresses the interaction between folk belief and learned commentary, identifying biblical commentary and the work of Church Fathers such as Isidore of Seville as the main source of discussions about godlings and, perhaps, as the source of much of the folklore itself. It is the argument of the chapter that by the time of the Norman Conquest, the various elements of fairy lore were present in British popular belief but had yet to be brought together into a single synthesis. These elements included a belief in wild ‘men of the woods’ gifted with prophetic powers; belief in elves; belief in supernatural women, often in a triad, governing the fates of human beings; belief in diminutive otherworlders, sometimes living beneath the earth and belief in heroes who had somehow become supernatural beings.

The book’s final chapter argues that the various elements of fairy belief as we might recognise it, including belief in an underground otherworld inhabited by sometimes pygmy-sized otherworlders, the connection between fairies and fate and fairy sexuality, were brought together as a direct result of the Norman Conquest. The key role played in the Conquest by Breton nobles who felt a cultural affinity with the Cornish and Welsh, combined with the Normans’ desire to escape the English past, resulted in the crafting of a new ‘British’ identity for the whole island of Great Britain by authors with a Brittonic cultural background such as Geoffrey of Monmouth, Gerald of Wales and Walter Map. These authors united elements of English and Brittonic folklore to fashion a new fairy world that was subsequently adopted as the setting for literary romances. The fairies of romance soon took on a life of their own and fed back into popular culture as a source of fairy lore, creating a complex amalgam of belief that was not fully described until the early modern period.