One of the abiding impressions made by Sir Gawain and the Green Knight (henceforth SGGK) on its readers is that the Green Knight is an embodiment of nature. An earlier generation of scholars saw the poem chiefly as a museum of Celtic folklore, but Celtic paganism and the ‘Green Man’, both associated with a mystical, proto-Romantic reverence for the natural environment, remain a part of its critical heritage. Despite the manifest courtliness and hospitality practised at the Green Knight's castle, where Gawain spends Christmas week before riding to the Green Chapel on New Year's Day, the view persists that the Green Knight presides over a world that is very much ‘natural’ or ‘wild’ in contrast to Gawain's and Arthur's. This is perhaps because both in the form of Bertilak and in the form of the Green Knight, Gawain's host and adversary are represented as a man of the forest. As Bertilak (whose name is revealed to Gawain in 2445), he leads boisterous hunting parties three days in a row, traversing the forest around his castle. As the Green Knight, he maintains the seemingly remote Green Chapel in a rugged landscape that impresses Gawain as ‘wylde’ (2163), the oratory of the devil himself (2190–4), even though it stands somewhere within or very near to the forest where Bertilak hunts – in fact, it is ‘not two myle henne’ (1078), as Bertilak cheerfully informs Gawain.

Gallaphur tant erra que au troisieme jour il monta une montaigne ou il trouva ung temple … Dist l'ancien home …: – Sire, on aoure ceans la Deesse des Songes. – Par ma foy, dist Gallaphur, c'est une fole deesse. – Ne vous en gabbez pas, dist le varlet, car elle est de grand merite, et ne vous conseille point d'entrer ceans au moins que ne faittes vostre paix a elle. – Sire, dist Gallaphur, oncquez plus ne oÿ parler d'elle et le roy Perce-forest en son tamps ne souffroit aourer que le Dieu Souverain. – Sire, dist le varlet, le roy Perceforest aouroit a sa devotion. Mais après sa mort une dame nommee Sarra … donnoit respons aux pucelles de leurs songes …, tellement qu'aprez sa mort les pucelles l'ont nommé la Deesse des Songes et lui ont fait ce temple ou elles l'aourent, car personne ne veille une nuit en ce temple que … il songera aucune chose du tamps advenir dont sur ce pourveoir se pourra. – Vallet, dist Gallaphur, bien sçay que le roy Perce-forest ama moult ceste dame en son tamps, mais on ne doit pas legierement croire sy haulte aventure comme d'une femme mortelle tenir a deesse, combien que ceste nuit demourray ceans pour sçavoir aucun point de sa vertu. … Sy s'endormy … Lors lui sambla que la deesse Sarra lui vint au devant … puis l'emmena sus une tant haulte montaigne qu'il pouoit bien voir tout le païs de Bretaigne, puis lui dist: – Gallaphur, regarde, retiens et mets en memoire ce que voir pues a l'entour de toy. Atant s'en parti et Gallaphur demoura esbahi des merveilles qu'il veoit par la Grant Bretaigne.

On observe aujourd'hui une forte tendance à valoriser les romans tardifs de la fin du quatorzième siècle-milieu quinzième siècle, à deux titres: comme les témoins d'une nostalgie du temps passé, réactivant la devise ‘d'armes et d'amour’, certes, mais aussi comme les précurseurs d'une forme de modernité. Après avoir considéré ces textes tardifs comme décadents, mal construits, la critique les pare désormais de nombre d'atouts, en partie avec raison, en partie parce que nous nous trouvons dans la situation des écrivains de la fin du Moyen Âge, où il faut faire avec ce qui reste, de façon toute pragmatique.

Le Moyen Âge finissant est un grand laboratoire qui expérimente de nouvelles formes à partir des reliques (modèles sanctifiés) et des restes (fragments dépecés) de la littérature antérieure. Démembrement, recyclage, nouvel assemblage sont autant de modalités du processus de récupération et de remploi que nous tenterons d'examiner ici.

Malory's Tale of the Sankgreal (henceforth M) is generally considered the ‘least original’ of his adaptations, conforming in most significant respects to the plot of the Vulgate Queste del Saint Graal (Q). The main alterations Malory appears to have made are thematic; most notably, he clearly downplays the mystical and devotional flavour of Q in his work. However, the precise extent of Malory's manipulation of his materials is hard to quantify, since the version of Q from which he was working does not match any known version of that text. Eugène Vinaver suggested that Malory's exemplar was probably closer to the lost common original of Q than to any of the surviving versions of the text. No French-language manuscript has yet been discovered that seems to represent the version known to Malory, but that does not necessarily mean that Malory's work is the only witness to this particular version of Q. The evidence of the medieval Irish translation of the Vulgate, Lorgaireacht an tSoidigh Naomhtha (L), has, so far, been largely overlooked. L's editor, Sheila Falconer, believed, like Vinaver, that the exemplar for her text ‘ranked high in the MS tradition of the Quest’ and, significantly, there are numerous points at which L and M share details not found in any known version of Q. It seems possible that L was translated from a version of Q close to, or identical with, the version Malory knew and, as such, may provide the best witness we have to the characteristics of his exemplar.

In the prologue to the mid-fifteenth-century adaptation of the ‘rhymed story of Erec, the son of King Lac,’ the prosateur describes his work as a ‘transmutation’: ‘… pour ce que l'en m'a presentee le histoire de Erec le filz du roy Lach en rime, je, au plaisir de Dieu, occuperay mon estude ung petit de tamps a le transmuer de rime en prose …’ (Because I have been presented with the rhymed story of Erec, the son of King Lac, I shall, God willing, devote a little time to transposing it from verse into prose …). The term transmuer (to transform or change) suggests the extent of the modifications the adapter performed as he turned Chrétien's poem into prose. Readers of Arthurian romances will be familiar with the techniques the redactor deployed in appropriating Chrétien's text: abbreviation, amplification and rationalization. However, the full compass of the transformations, which occur on a socio-cultural level but more importantly, inform the literary art, may be surprising to them. Indeed, the modifications affect the structure and the meaning of the romance. While the prose text respects the general outline of the story, it abbreviates or omits many elements, but also embroiders, alters and even adds new material. In addition, it changes countless details in such a way that the adaptation is in fact a completely new text.

Perceforest is the greatest of the unread Arthurian romances. There is still no complete modern edition, and in order to read the last book, you have to use the huge folio volumes produced in Paris in the early sixteenth century, designed to satisfy the enthusiasm of buyers of the new-fangled romances in book form. Neglected until a summary version by Jeanne Lods appeared in 1951, it was only with the appearance of the first volume of the Textes littéraires français edition by Jane Taylor in 1979 that it began to attract wider attention among Arthurian scholars, reinforced by the publication of all but the sixth part of the romance by 2007. Now that Nigel Bryant has produced an English version, its extraordinary riches are available to a much wider audience. This article explores the possible historical context in which it was originally composed.

I say ‘originally composed’ deliberately, because the version that we have is almost certainly a reworking in the fifteenth century of a fourteenth-century prose romance. The language is not that of the mid fourteenth century, and much of the content is similar in style to that of the Burgundian romances of the mid fifteenth century. These were new versions of twelfth- and thirteenth-century texts, by now archaic in their language.

In 1842, The British History of Geoffrey of Monmouth appeared in the series The Monkish Historians of Great Britain. Already published were Bede's Ecclesiastical History, a volume of Gildas's and Nennius's Histories and the Chronicle of Richard of Devizes, along with Richard of Cirencester's description of Britain. Further works of Bede, histories by William of Malmesbury and William of Newburgh, the Saxon Chronicles and Asser's Life of Alfred were all in press. The series page announced that several of the volumes were appearing ‘in an English dress’ for the first time, but this particular book was a revised edition, by the prolific editor and translator J. A. Giles (1808–1884), of Aaron Thompson's 1718 English translation of Geoffrey's Historia regum Britannie. Giles makes it clear in his own preface that both Geoffrey and his first translator should be treated with considerable suspicion. Of Geoffrey, he writes, ‘We do not insert the BRITISH HISTORY in our series of Early English Records as a work containing an authentic narrative, nor do we wish to compare Geoffrey of Monmouth with Bede in point of veracity’. Describing Thompson's preface to the 1718 translation, Giles is blunt with respect to the former's credulity: ‘Prefixed to the work is a long introduction in which the translator endeavours to defend his author from the charge of having inserted the narrative which he professes to have translated from the Old British Tongue. It is now, of course, universally admitted that the whole series of British Kings, from Brutus downwards, is a tissue of fables’.

The term ‘Celtic magic’ has had a long currency in medieval studies, particularly Arthurian studies. being positioned alongside ‘Celtic myth’ as a convenient explanation for elements in vernacular medieval romance whose provenance is not otherwise obvious. Yet both terms. ‘Celtic’ and ‘magic’, are problematic when it comes to definitions, and this is particularly so in relation to two of the most important survivals of Welsh Arthurian literature. Culhwch ac Olwen (Culhwch and Olwen) and Breuddwyd Rhonabwy (The Dream of Rhonabwy). Both tales locate Arthur in the centre of a magic landscape; one that is subject to supernatural events. The figure of Arthur himself is presented quite differently in both texts, and in many ways The Dream of Rhonabwy foreshadows the loss of magic. in the sense of personal charisma and superhuman ability. that accompanies Arthur's appropriation into the French and English traditions. Moreover. particular kinds of literary magic in medieval texts can be related to certain types of narrative discourse. In its most familiar sense. ‘magic’ is associated with narrative agency. that is. with persons or objects who dispense and control the application of magic. whether these are fairy women or kings, or specific objects such as magic rings or potions. This agentive magic. typical of medieval romance. is produced through a discourse of realism which comes close to the modern mode of magic realism. Early Welsh and Irish tales, however, use a different kind of narrative mode; one that foregrounds naturalism rather than realism in its storytelling techniques. This produces a different kind of ‘magic’, an agentless occurrence of wonders that can best be described as the supernatural marvellous. The early Welsh prose tales therefore exemplify a particular narrative strategy which might be called ‘magic naturalism’.