Medieval Fairies:
Now you see them, now you don't

Jeremy Harte

The fairy tradition in literature begins in the 1380s, with Chaucer and Gower. In their eyes, the fairies are already a vanishing race, partly frightening and partly comic. The implication (particularly in the preamble to The Wife of Bath’s Tale) is that people used to believe in fairies, but don’t do so any more. However, the fairy mythology as a consistent set of beliefs (dancing in rings, living in hills, the rule of a queen, and so on) is itself created by the writers who claim to be recording its final echoes. Earlier evidence does not describe these fairies. Instead it details encounters with various supernatural beings who were, in retrospect, treated as if they had been citizens of fairyland.

The otherworldly beings who appear in medieval chronicles are a varied lot. Some of them, such as the barrow revellers in William of Newburgh and the maidens found in a wood by Wild Edric, are deliberately left unidentified; like the ‘maiden in the moor’ of the carol, their non-human status is indicated by allusion and not by direct statement. Others are defined by a single strange character-istic, such as the colour of the Green Children of Woolpit, or the small size of King Herla (a pygmaeus) who rides a goat. The homunculus in an enigmatic encounter story from Thomas Walsingham was both diminutive and dressed in red. The otherworldly race who played with the boy Elidurus had their own language ( a form of Greek) and their own, superior morals. There is nothing in these scattered references to suggest that the beings concerned are of the same type. Moreover, it would be an anachronism to separate these accounts from contemporary reports of diabolical apparitions. All the medieval words for spirits were also used, on occasion, for devils.

The achievement of fairy writers, from Chaucer to Shakespeare, was to expand the hints of an otherworld in the Breton courtly narratives until almost all previous tales of supernatural encounter could be shoehorned into their dominant discourse. Despite Bob Trubshaw’s suggestion in the accompanying article that ‘Broadly speaking, these Middle English accounts conform to the Anglo-Saxon categories of elves, dwarfs and pucks, so seem to represent some continuity of belief’ there is no systematic mythology of fairies before 1380. There are many unrelated motifs – barrow-dwellers, tricksters, small people, household guardians - which we know in hindsight will come together to define the fairy kingdom. But this identity is simply not there in the original references.

Take a word like elf, which Chaucer makes synonymous with fairy. In Old English the aelfs are one amongst many otherworldly communities. The Charm for a Sudden Stitch puts them on the same footing as hags and the Aesir; and they have the same role as the Aesir in name compounds – compare Aelfric and Osric. An Anglo-Saxon vocabulary of 1100 renders dryades etc. as types of elves. As Hilda Ellis Davidson showed in The Road to Hel, the Scandinavian elves are closely assimilated to the Vanir.

By the thirteenth century, the original context of Old English belief had become lost, and people were using the word in various ways. Layamon uses elf to translate the Romance fadas - following a line of thought which was to lead to the elf-fairy equivalence – but other people had other ideas. Robert of Gloucester, explaining what type of being it was that fathered Merlin, says that the sky is full of spiritual beings called elves. Here we are on the verge of the diabolical, as we are in Beowulf when the aelfs are of the seed of Cain. Elves were found in literature, but not in the landscape. They do not appear in southern English placenames: nor, indeed, do fairies - not until the eighteenth century. Instead their place is taken by puca, which appears describing the inhabitants of wells, pits, and barrows. It is tempting to make the medieval pouke as identical with Renaissance Puck, but this is to fall into another retrospective reading. Even in Midsummer Night’s Dream, Puck has the appearance of being transferred into fairyland, a little awkwardly, from some quite separate tradition.

The situation is different in northern England, where aelf is common and puca is absent. This is also the region where elf was retained as the usual word for beings in the modern period, the Romance fairy being rejected. This may well be the result of Scandinavian influence - the fact that aelf is liable to compound with haugr rather than beorg would suggest this.

Scandinavian influence is certainly present in those placenames which refer to dwarfs. The Anglo-Saxons had no concept of the dweorg as a member of a small supernatural race. The word is always glossed as nanus, pygmaeus, and means a short human being. When we meet with clearly mythological dwarfs in North Country placenames, it seems reasonable to suspect Norse influence, as Keightley observed over a century ago.

In short, the origins of the fairy mythology lie not in the remote past, but at the court of Richard II. The creative synthesis which the poets made out of English and French traditions was developed in the Tudor period to include tricksters of the Robin Goodfellow type as well as the familiar spirits of cunning men, and domestic spirits like the brownie. As an English-language tradition, it was able to dominate and then change the native sidhe beliefs of Ireland and the Highlands, introducing alien notions such as small size into their narrative. By the nineteenth century, it was possible for Anglo-Saxon spirits like the grima, scucca and thyrs - who had lived out a quiet rural existence as Church Grims, Black Shucks and Hobthrusts - to find themselves reinterpreted by folklorists (not the folk!) as minor figures in the fairy mythology. This means that we can no longer make out what they were like originally. The fairy glamour of the fully developed tradition has tended to obscure our understanding of the very disparate narratives of supernatural encounters which have been patched into it.

Originally published in At the Edge, No.10, 1998. Copyright 1998.