Navigation – Plan du site

AccueilNuméros25Part 1: In the beginning…What makes Breton lays ‘Breton’? ...

Part 1: In the beginning…

What makes Breton lays ‘Breton’? Bretons, Britons and Celtic ‘otherness’ in medieval romance

Leo Carruthers


An exploration of the semantic and cultural fields behind the term ‘Breton’ suggests that the modern word ‘Celtic’ corresponds better to what is implied by the expression ‘Breton lay’. It is commonly supposed that the Breton lays, in both Old French of the 12th century and Middle English of the 14th century, were based on songs originally sung by Breton minstrels. But the word ‘Breton’ is misleading ; while it now refers to the inhabitants of Brittany, in medieval literature ‘Breton’ and ‘Briton’ were undifferentiated, applying to the early, non-Germanic inhabitants of Britain and Armorica, whose language was Old Welsh. For the English and French, these people were ‘other’, their culture a source of mystery. Being neither English, Norman or French, the language of the Bretons, Britons, or British appeared exotic to the dominant political groups. Their poetry, often associated with the Otherworld and a belief in fairy magic, would be labelled ‘Celtic’ later in history, but that term was not available when the ‘Breton lays’ first appeared. The Anglo-Norman poet ‘Marie de France’, writing for the Plantagenet royal court in the 12th century, had at her doorstep the well-attested riches of Welsh literature as a model and inspiration, whereas no evidence exists for any independent songs or tales from Brittany at this time. It would thus be more apt to speak of ‘Celtic lays’ in Old French and Middle English, since the word ‘Celtic’ now conveys to us what medieval French and English poets meant by the word ‘Breton’.

Haut de page

Texte intégral

The Middle English Breton lays of the 14th century

1The ‘Breton lay’ is not easy to pin down because the characteristics of the genre are ill-defined, even within the broader category of ‘romance’ which, in turn, has almost no frontiers. A poem is a Breton lay first of all because it says it is one, or, if it does not actually use the word ‘lay’ (most of them do), it claims to be part of a British/Breton tradition. Why it says so is another matter, together with the question of why Breton, Briton, and British are hard to separate. Indeed the Breton lay genre is ambiguous for many reasons, not the least being what is meant by ‘Breton’ – a term that is far from being as simple as it seems at first sight.

  • 1 Although the name ‘Marie’ occurs in one of the lays attributed to her, the designation ‘Marie de Fr (...)

2Though they claim to be of Breton origin, the lays do not make their first appearance in what we now call Breton, a language which remained entirely oral during the period in question; no Breton sources, if indeed they ever existed, have been preserved for this type of poem. The lays were first written in Old French of the 12th century, especially in that kind of French known as Anglo-Norman because it was spoken in England at the court of the Norman and Angevin kings. The earliest examples of the genre are attributed to a writer called Marie, later known as ‘Marie de France’.1 Some of these texts were translated into English and were written down in the 14th century. Later in that century other poems of the same type were composed directly in English, not translated.

  • 2 See J. R. Clark Hall, A Concise Anglo-Saxon Dictionary, 1894; 4th ed., Cambridge, Cambridge Univers (...)
  • 3 While the ancient Greeks knew of a people they called Keltoi, it is only since the 18th century tha (...)

3The authors of Breton lays normally claim to base their tales on songs sung by Breton minstrels. Many, though not all, are set in the ‘Celtic’ regions of Britain, Brittany, or both. Some of them – even the early ones – do not bother with the ‘Celtic’ criterion but still claim to be Breton lays because of their supposed source or origin, thus showing that the genre was based less on setting than on a tenuous concept of oral transmission by Breton minstrels. The claim is questionable for several reasons, not only because of the absence of evidence in medieval manuscripts, but because of a fundamental misunderstanding of the meaning of the word ‘Breton’ in the context of Anglo-Norman and Middle English poetry. The people referred to by this name were in fact the ancient Britons, that is to say the British Celts, whom the Anglo-Saxons also called ‘Wealas’, or the Welsh.2 It is therefore to Old Welsh literature that we should look for possible sources of inspiration. But even the word ‘Welsh’ does not, in my opinion, express the full meaning of what ‘Breton’ means in the context of the Breton lays. I hope to show that the modern word ‘Celtic’ – which was unknown in the Middle Ages in both French and English – gives a better idea of the atmosphere which the authors of these tales wished to convey.3 Behind it lies a cultural cliché about the nature of the Celts.

Celtic settings in Old French and Middle English lays

  • 4 See Marie de France: Lais, ed. Alfred Ewert, 1944; Oxford, Blackwell, 1960, reissued with introduct (...)

4Let us begin by looking at the Celtic settings of the lays in both Old French and Middle English. The twelve lays attributed to Marie are set in Brittany or Britain or both, with an occasional nod to Normandy.4 They are as follows:

  • 5 Logres is the Old Welsh name for England, so its occurrence in Old French and Middle English romanc (...)

Brittany only: six (Guigemar, Equitan, Le Fresne, Bisclavret, Laüstic, Chaitivel);
Brittany and Logres5 (England): two (Eliduc, Milun partly);
Britain only: three (Yonec in Wales, Lanval in Logres, Chevrefoil in Cornwall);
Britain-Wales-Logres-Normandy-Brittany: one (Milun);
Normandy: two (Les deux amanz, Milun partly).

  • 6 Neustria was a name created for the western part of the kingdom of the Franks, after the death of C (...)

5Brittany is therefore strongly present, in nine cases out of twelve, six of them exclusively. Britain is the setting in four cases, three of them exclusively, especially the Celtic areas of the island – and when England is concerned it is called by the Welsh name, Logres, often found in Arthurian romance. Normandy is mentioned twice, in one case being the exclusive setting, but the fact of calling it by the ancient name of Neustria distances it in history, producing a once-upon-a-time effect.6 Normandy is not only the neighbouring duchy to Brittany, but is of obvious importance to the Anglo-Norman audience, as the Norman and Angevin kings of England were its rulers.

6In comparison, the 14th-century English lays show the following settings:

Britain-England-Wales: Sir Orfeo, Lay le Freine, Sir Landevale, Sir Launfal (also mentions Oléron), Sir Cleges, The Wife of Bath’s Tale.
Brittany or Little Britain, with England mentioned: Sir Degaré, The Franklin’s Tale.
Rome mainly, and a scene in Galicia: Emaré.
Germany/Austria and Rome: Sir Gowther, Erle of Toulous.

7The Middle English lays therefore show a definite move away from Brittany and into the island of Britain, sometimes referred to by that name, sometimes by one of its regions (England or Wales). In fact Marie’s proportion is reversed, since six of the eleven English lays are set in Britain, and England is mentioned in two of the others.

  • 7 See Laskaya and Salisbury, op. cit., p. 1.

8Any connection with either Brittany or Britain is entirely lost in three English cases out of eleven (Emaré, Sir Gowther and Erle of Toulous), the action moving to Germanic and Roman locations. This may partly be due to their religious content and the protagonists’ need to be reconciled with God; for example, God’s will is a major theme in Emaré, and the hero of Sir Gowther has to go to Rome to obtain the Pope’s forgiveness. But over and above any plot requirement, by the 14th century the idea of a ‘Breton lay’ had become conventional, one in which the important feature was a romantic adventure story rather than a Celtic setting. There is usually a marvellous element, but it could just as well be a Christian miracle as a Celtic fairy tale. The evidence shows that when a lay moved away from a Celtic country, it was more likely to be the scene of a Christian miracle. Nevertheless, as the list given above shows, and as Chaucer would confirm in The Franklin’s Tale (see below), the convention was maintained that a ‘Breton lay’ was supposed to be a tale resembling, and perhaps even based on, the kind of stories originally told and sung by Breton minstrels, even where this appears to be patently untrue.7

‘Breton’ and ‘Briton’ as ‘Celtic’

  • 8 On the history of the Celtic languages see W. B. Lockwood, A Panorama of Indo-European Languages, L (...)

9The Breton lays in Old French are strongly marked by their presumed Celtic origins; those in Middle English are somewhat less marked, though the ‘Breton’ convention was upheld. Whether they actually came from, or were inspired by, the region of France now called Brittany raises very complex issues. The medieval word ‘Breton’ is ambiguous in a way that is no longer the case in Modern English. When we say ‘Breton’ now, we automatically think of Brittany, a region of France, but that was not the case in the Middle Ages when ‘Breton’ and ‘Briton’ could be used interchangeably, so that further geographical details had to be supplied in order to make it clear if the people thus named lived in Britain or on the continent. The issues raised involve the birth of the Breton language and its culture, the awareness of insular British origin among speakers of that language, the political relations of the medieval duchy of Brittany with its neighbours, the French perspective on Breton otherness, the analogous English perception of Welsh resistance to English expansion, the Anglo-Norman use of the word ‘Breton’ for the ancient Celtic inhabitants of Britain, and ultimately, the adoption of the terms ‘Briton’ and ‘British’ by the English themselves – in an English kingdom which was by then no longer purely Anglo-Saxon but had become French (Anglo-Norman) as well.8 Such density and complexity can no longer be communicated by the word ‘Breton’ today – but ‘Celtic’ comes close to it, and this point will be dealt with further below.

  • 9 Geffrei Gaimar. Estoire des Engleis – History of the English, ed. and trans. Ian Short, Oxford, Oxf (...)

10An Anglo-Norman named Geoffrey Gaimar wrote, c. 1150 or earlier, a two-part Estoire des Bretons (History of the British, or Welsh) and Estoire des Engleis (History of the Anglo-Saxons in Britain), both reworked from Geoffrey of Monmouth’s Historia Regum Britanniae. His Estoire des Bretons is unfortunately no longer extant, but it is mentioned at the beginning of the Estoire des Engleis.9 The significant thing about his use of the word Bretons in Anglo-Norman is that it means the Welsh, the Celts of Britain, and not the inhabitants of Armorica. When Marie de France speaks therefore of ‘Breton minstrels’ in the 12th century from whom she had heard songs, when medieval chroniclers speak of King Arthur and his knights as ‘Bretons’, or when Middle English poets claim to be writing ‘Breton lays’, we should beware of the trap of supposing them necessarily to mean Bretons from Brittany. It is much more likely that they mean Britons from Britain.

  • 10 Larry D. Benson (ed.), King Arthur’s Death: The Middle English Stanzaic Morte Arthur and Alliterati (...)

11This becomes clear in the English Arthurian poem known as the ‘Alliterative Morte Arthure’, dating from the late 14th century and therefore contemporary with Chaucer. The English writer speaks of the epic battle of Camlan (somewhere in Cornwall) between Arthur and Mordred, which concludes with Mordred’s death and the (ultimately mortal) wounding of Arthur.10 Arthur’s ‘bolde Bretons’ fight against a vastly superior force under Mordred, made up of Picts, Scots, Irish, and giants. The ‘bolde Bretons’ of the alliterative poem can only be Britons, i.e. Romano-British Celts, since none of them had yet fled to Armorica and ‘Bretons’ in the modern sense did not exist.

Celtic otherness: a political perspective

  • 11 Wales was conquered by Edward I of England in 1284 but remained recalcitrant; the principality was (...)

12We need to break away from the ethnic and geographical preconceptions which the modern name ‘Breton’ imposes. In the 12th century and later, ‘Breton’ meant Welsh as well as Armorican. But in addition to the political reality, behind the term lay a cultural construct which enabled the dominant monarchies in England and France to project elements of romance onto a neighbouring and subject people, who were in reality their rivals and enemies.11 The word Breton as used in the Breton lays corresponds less to a particular place than to a perception of the cultural difference marking the peoples living on Europe’s north-western seaboard, in the islands and peninsulas traditionally seen as being distant from the centres of power. A strong element in this culture which attracted the Anglo-Normans and French was its approach to the fantastic through magic and the Otherworld. The perception of Celtic ‘otherness’ is partly based on reality but is also partly imposed. To get a modern equivalent of what this could mean to the inhabitants of the medieval kingdoms of France and England, it is helpful to think of the romantic way the word ‘Celtic’ has been used since the late 18th century. ‘Celtic’ evokes distinctiveness for us in a way that ‘French’ or ‘English’ – taken as the names of political groups as well as of the languages they speak – can never do. Both French and English represent dominant cultural norms; they are national and international languages with heavy implications of political hegemony associated with the creation of the modern nation state. ‘Celtic’ on the other hand, now suggests cultural minorities, difference from the norms, and struggle for survival, expressed in language, poetry, music, song, magic and mythology.

  • 12 See C. Hall, op. cit.
  • 13 Gerald de Barry signs himself in Latin as Giraldus Cambrensis, from Cambria, the old Roman name for (...)

13But the word ‘Celtic’ was not available in the 12th century or at any time in the Middle Ages, indeed not for hundreds of years after that, and so ‘Breton’ was used instead, with much the same kind of cultural overtones. There is no denying that ‘Breton’ does relate to Brittany (Armorica), a duchy of the medieval kingdom of France, but it is much more than that, since it means more generally the inhabitants of Britain who first brought the Briton or Breton language to Armorica. In other words, the close linguistic and ethnic relation between Britons and Bretons went back to a time when the Germanic invasions of Britain drove the native British population first westwards, then across the sea to Armorica. It was Anglo-Saxon settlement that led to the creation of ‘England’ as a region of Britain with a culture different to that of the Britons. And as ‘England’ came into being, so too did ‘Wales’, that part of Britain where the original culture of the British Celts was preserved. Significantly, ‘Wales’ comes from the Old English word ‘wealas’, meaning ‘foreigners, servants, slaves’, which is not a place-name but a description of the status of the native population in the eyes of the Saxon invaders – hardly a term one would expect the ‘Welsh’ to call themselves.12 Indeed, the Latin writer Gerald de Barry (c. 1145-1223), also known as Gerald of Wales, a contemporary of Marie de France and half-Welsh himself, calls it a ‘barbarous’ name.13

  • 14 In the short Prologue (p. 41) she speaks only of ‘lays which I had heard’; the Breton attribution i (...)

14Marie de France says that she wrote these tales usually after hearing a sung version performed by ‘Breton minstrels’ – performances for which there is no other evidence.14 She does not say where she was at the time, or what language the minstrels used. It is all too easy to suppose that she meant she was in France, where she presumably grew up. Even allowing for the possibility that she could have heard Breton minstrels as a child, they would not have sung in Breton but in French, the courtly language. For much of the later Middle Ages, from the 9th century onwards, Brittany was a duchy technically part of the kingdom of France but in fact, like other such territories ruled by powerful feudal families, largely independent of the French king who was based in far-off Paris. As in the other Celtic countries which were occupied by an alien culture and eventually brought into political union with a more powerful kingdom, we find the language of the ruling monarchy (in this case, French) being adopted by the Breton nobility from an early period, beginning in the 9th century when the duchy was constituted, so that by 1200 the Breton tongue was no longer spoken by the upper classes. If Marie de France ever did hear any songs by Breton minstrels, as distinct from Welsh harpists, they would have sung in French.

15But there is no reason in the first place to believe that Marie heard the original songs in France; it seems more likely that she heard them in England, from British (i.e. Welsh) minstrels. In Britain the Welsh language was somewhat better placed than Breton was in Armorica, since there is evidence that some of the leading Anglo-Normans, who married into the Welsh noble and princely families, had children who could speak Welsh. Gerald de Barry, a member of a large and influential Cambro-Norman family which was in the forefront of the Norman invasion of Ireland, is a perfect example of the type.

  • 15 Emily K. Yoder, ‘Chaucer and the “Breton” lay’, The Chaucer Review, 12.1, Summer 1977, p. 74-77.
  • 16 Nearby Normandy – Brittany’s neighbour and sometime rival – is another powerful duchy which was in (...)

16The issue was discussed, and to a large extent settled, by Emily Yoder in an article dating from 1977 in which she equates the ‘Bretons’ of the ‘Breton lays’ with the Celtic British, that is, the people whom the English also called by the derogatory term, Welsh.15 Her conclusion is not as widely known as it deserves to be, and the old idea that ‘Breton lays’ originally came from Brittany dies hard. Yoder does not say anything about actual Welsh literature, except for a brief reference to the Goddoddin; she does not, for instance, mention The Mabinogion, despite its crucial importance for the transmission of Welsh romance to Anglo-Normans.16

Celtic folklore: loss and preservation in the British Isles

17The fairy tales associated with the Breton lays belong to the wider field of Celtic folklore, mythology, and belief associated with the western fringes of European civilisation. From the point of view of the dominant political cultures of medieval England and France, the remote peninsulas and islands on their western frontiers were somewhat strange places, ethnically and linguistically different. The religious culture of Wales, Scotland, Ireland and Brittany occupied an Otherworldly space which had much to do with the survival of Celtic mythology outside the Roman system, a system which itself underwent a radical transformation in the 4th century.

  • 17 See T. W. Moody and F. X. Martin (eds), The Course of Irish History, 1967; 3rd ed., Cork, Mercier, (...)
  • 18 For this reason the Breton tongue in Brittany is closely related to Old Brittonic or Welsh, and not (...)

18The southern part of Britain was an important province of the Roman Empire from the 1st to the early 5th century. Its religious fortunes followed those of the Empire, with the result that Christianity became an official religion there during the course of the 4th century. The Highland regions of Scotland remained pagan. In linguistic terms, Scottish Gaelic was originally a dialect of Irish Gaelic, to which it remained close down to the 12th century. As for Ireland, which had never been invaded by Rome, it was not a province of the Empire. This is the single most important reason for the survival in Ireland of a pagan, Iron Age culture long after it had disappeared from Roman Britain.17 This would prove to have cultural and literary implications, since the Irish preserved ancient Celtic mythology in a fuller form than anywhere else; all other Celtic regions, in Britain and Gaul, had been Romanised, with the consequent loss of pagan mythology, reduced to a dim memory. Following the withdrawal of Roman troops from Britain in the 5th century, much of the island was overrun by the Angles, Saxons, and Jutes, Germanic tribes who were ‘barbarians’ in Romano-British eyes. Under pressure from the invaders, the British Celts were forced to flee to the northern and western regions of the island, while some of them crossed the sea to set up new homes in Armorica, later renamed Brittany, or Little Britain.18

  • 19 According to the dates given by Bede in his Historia Ecclesiastica, the conversion of Anglo-Saxon E (...)

19The survival of Celtic societies in parts of the British Isles, and in Armorica through migration, acted as a cultural counterweight to the dominant influence of Rome, its imperial success, its legal and administrative framework, its powerful army, its towns and roads, its Classical literature, and finally, from the 4th century onwards, its monotheistic religion which sought to eradicate pagan beliefs and practices. Religious conversion and the Germanic invasions of the Empire did not alter the paradigm from the Celtic point of view. The Roman Empire in the West collapsed in 476 under barbarian pressure, at a time when the Irish were in the very midst of their conversion process. The Frankish kings in Gaul soon converted to Christianity (the baptism of Clovis took place in 496 AD); the conversion of Anglo-Saxon England took place in the 7th century, a long process to which Irish monks made a major contribution.19 In this framework Celtic culture tended either to disappear (Gaul) or to retreat to a state of siege (Wales). Only in Ireland did it survive and thrive.

  • 20 See Seán J. Connolly (ed.), The Oxford Companion to Irish History, 1998; 2nd ed., Oxford, Oxford Un (...)

20It is one of the unexpected turns of fate in European history that the Irish, unlike other Celtic peoples, did not jettison their ancient mythology as soon as they had gone over to Christianity.20 The preservation of Druidic lore by the Druids themselves has much to do with this. The fact is that in both Gaul and Britain, the Roman invaders destroyed Druidic cultic centres in the 1st century, but this never happened in Ireland, which was not a Roman colony. Ireland’s conversion in the 5th century took place in a very different political context: the Empire was weak, collapsing in the West, so there was no question of imposing a Roman system in Ireland, nor any question of wiping out Druidism by force. As a result, many Druids who converted became monks, while others became bards employed by the numerous Irish petty-kings. Druidic lore was strictly oral (another reason why it was lost on the Continent and in Britain in the 1st century), but the Irish Druids preserved it, even after their acceptance of Christianity, handing it down from generation to generation.

  • 21 See Robert Welch (ed.), The Oxford Companion to Irish Literature, Oxford, Oxford University Press, (...)

21There are demonstrable links between Irish and Welsh literature going back to the prehistoric period.21 Both peoples spoke Celtic languages, different though related, labelled Goidelic (Gaelic) and Brythonic. But in fact the Irish Sea did not divide them in the prehistoric period in the clear-cut way that we might imagine from the later medieval political scene. There were Brythonic speakers in parts of Ireland and Gaelic communities in parts of Britain, not just in Scotland but also in Wales. In the Christian period, there was much coming and going between the Celtic peoples of these islands, with consequent sharing of traditions and mythologies, of which the Irish had preserved more. When the British Celts crossed over to Armorica, they carried with them much of the common fund of Celtic tradition. In the Golden Age of Irish monasticism (6th to 9th centuries), many Irish monks founded monasteries in Britain, Brittany, the Frankish territories, and beyond. Many Welsh monks also founded monasteries or cells in Brittany, hence the very large number of Breton saints of both Irish and Welsh origin.

22The curious nature of the relationship of the Celtic peoples to their culturally dominant neighbours, whether Anglo-Saxon, Anglo-Norman or French, goes a long way to explain how and why elements of Celtic mythology, magic, fairies and the Otherworld, made their way into ‘Breton lays’ in Old French and Middle English. The otherness of British (Welsh) literature for the Anglo-Saxons and Anglo-Normans is partly based on a material reality – their language and culture really were different – and partly on a political construct, a certain attitude on the part of the more central and politically dominant groups towards what they saw as the ‘fringes.’ In both cases, in Britain and in Gaul, we observe a Celtic language implanted in the west and a Germanic language entering from the east.

The Celtic Otherworld

  • 22 See Peter Berresford Ellis, A Dictionary of Irish Mythology, Oxford, Oxford University Press, 1987.

23Countless generations of European migrants, from the Neolithic peoples down to the Normans, followed the westward movement of the sun, until they finally reached the limits of the Continent and saw the sun sink into the Atlantic Ocean. Here lies the secret of the placing of the Celtic Otherworld in an undefined West, across the sea, beyond the horizon. In the Celtic world view as presented in Irish mythology, the Otherworld is not a spiritual state but a concrete place, part of our material world but normally out of reach to humans. Its physical location may change but it is usually an island lost in the mist, out of sight across the sea. It is an earthly paradise where life is preserved, there is no death, and a person who manages to get there (an extraordinary feat in itself) does not grow old. The Otherworld goes by several names in Irish, including Tír na nÓg, literally the Land of Youth, a fairytale land from which mortals could not return without losing the youth which they had enjoyed there.22 It is truly remarkable that Irish monks in the Christian era should have preserved so much of the ancient pagan mythology, without which our knowledge of Celtic beliefs would be extremely poor.

  • 23 Lebor Gabala is the Old Irish spelling of the name. The Modern Irish spelling is normally Leabhar G (...)

24The Otherworld is a variable concept and although material, it is not always fixed in the space it occupies; nor do its inhabitants remain stable. Some fairies live underground, in a place inaccessible to humans under normal circumstances. According to the legendary Lebor Gabala (Book of the Invasions),23 preserved in Irish manuscripts of the 12th century but containing much older material, in ancient times, before the arrival of the Irish themselves (i.e. the Iron Age Celts), Ireland was occupied by a magical people known as the Tuatha Dé Danann, ‘tribes of the goddess Dana (or Danu).’ When the Irish settled in the island, they defeated these magical tribes who were obliged to disappear or to go underground, to dwell in hills, rivers and lakes. This is the origin of the Irish belief in fairies which remained very strong all through the Middle Ages and down to recent times, lingering perhaps even today. Sometimes mortals can be enticed into the Otherworld, often for love of a fairy (as in Sir Launfal), or for greed of fairy gold, but for a mortal to accept is dangerous, not to say fatal. Time in the Otherworld does not pass as in the human world; a man might never come back, or might return only after many years, thinking he has been away for just a night or two, only to find that all has changed, his family and friends dead and gone.

The Mabinogion and Celtic enchantment

  • 24 The identity of this mysterious poet, the presumed author of the lays, remains a vexed question, bu (...)

25Dame Marie,24 writing in England in the 12th century, had access to a ready fund of British (Welsh) romance as a source and inspiration for her composition of Breton lays. Whether or not she actually heard Breton or Welsh minstrels singing songs which she then developed into stories called ‘lays’, there is a good reason why a 12th-century writer would say this is what had happened. That reason is the otherness of Celtic culture, with its strange tongues, its music and poetry, its non-Classical mythology. This is exemplified in The Mabinogion. Considering that nothing whatsoever is known of any Breton lays in the continental Breton language of the 12th century, it is important to look a little closer to home from the point of view of the English royal court: just on their doorstep, in Wales, lay a strong and vital body of insular Breton (Welsh) literature. This is The Mabinogion, a collection of Welsh stories with many sources which greatly influenced the development of Arthurian romance in French and English. Full of ‘Celtic’ enchantment, the very stuff of the Breton lay, the Welsh tales played a central role in transmitting ‘Celtic’ traditions to English and French poets.

  • 25 See introduction by Gwyn Jones and Thomas Jones (trans), The Mabinogion, London, Dent, Everyman Lib (...)
  • 26 Môn was Britain’s chief Druidic cultic centre, which, for purely political reasons, was destroyed b (...)
  • 27 Gerald of Wales, The History and Topography of Ireland (1188), translated from the Latin by Lewis T (...)

26The Mabinogion is found in the White Book of Rhydderch and the Red Book of Hergest, two 14th- century Welsh manuscripts; some of the tales also occur in older Welsh manuscripts of c. 1200, where linguistic forms and social codes show them to be even earlier than that.25 Scholars therefore estimate that the Four Branches of the Mabinogi were composed orally c. 1050-60, and that one of the other native tales, Culhwch and Olwen, dates from c. 950. An underlying Irish influence can be discerned going back in literary terms at least to the 6th century (the period when Irish literature first entered a written phase), and perhaps to the prehistoric Iron Age as regards oral transmission. The name of the legendary Irish King Lir, for example, is the equivalent of the Welsh Llŷr. Lir is one of the Tuatha Dé Danann, the old gods of the Neolithic period (later transformed into ‘fairies’ by the Iron Age Celts), and he is the father of Manannan, the sea-god whose name is preserved in both the Isle of Man and in Môn, the Welsh name of Anglesey.26 The Irish hero-tales follow a clear pattern of conception, youthful exploits, banishment, and death, which is also found in the Welsh romances. Wales, like Ireland and Scotland, had its medieval bards, who were poets, musicians and story-tellers, long before ‘minstrels’ or ‘troubadours’ were heard of in France.27

  • 28 The definite article in Welsh is y, which can be singular or plural, like English ‘the’, so it is n (...)

27The use of The Mabinogion as a title gained currency only during the 19th century due to the influence of the first complete English translation of all eleven tales, made by Lady Charlotte Guest and published in three volumes (1838-49). The word Mabinogion is not usually translated, but is understood to mean ‘Heroic Tales’ or ‘Heroic Romances.’28 The collection is not, however, a single work. Strictly speaking the name should apply only to the first four tales, known in Welsh as Pedair Cainc y Mabinogi – the Four Branches of the Mabinogi – but seven other medieval tales and romances have been attached to them.

28The Four Branches of the Mabinogi are: Pwyll Prince of Dyfed, Branwen Daughter of Llŷr, Manawydan Son of Llŷr, and Math Son of Mathonwy. Underlying them all is the story of Pryderi, the ‘mab’ or son, the young hero. The Four Independent Native Tales are: The Dream of Macsen Wledig, Lludd and Llefelys, Culhwch and Olwen, and The Dream of Rhonabwy. The Three Romances are: The Lady of the Fountain, Peredur Son of Efrawg, and Gereint Son of Erbin. Jones & Jones describe the Norman-French influence on the last three as being ‘as unfortunate as it was inevitable’ (p. xxv). It may be unfortunate because of the loss of Welsh uniqueness and authenticity, but it is nevertheless most interesting from the point of view of the Breton lays, since it shows that Welsh story-tellers did not exist in a cultural vacuum, outside the Anglo-Norman domain. Nor was the influence all one way – further evidence that Marie de France may have heard stories like these which inspired her Breton lays.

29The eleven tales in The Mabinogion are filled with magic and marvels of all kinds, occurring on practically every page, giving them a strangely enchanted atmosphere, which is reinforced by the narrator’s matter-of-fact attitude. One is struck by how ‘ordinary’ the extraordinary seems to be. It never calls for comment or explanation on the story-teller’s part; it is simply given, a normal part of life, at least in Celtic romance. A few examples will suffice to illustrate this, taken from the Four Branches, omitting most of the narrative details and retaining only the magical or marvellous elements in order to show their importance (some of these episodes are paralleled in the Breton lays).

Pwyll Prince of Dyfed

  • 29 Claire Vial has effectively applied the looking-glass metaphor to Sir Orfeo’s visit to the fairy ki (...)

30In the first section, Pwyll offends Arawn, king of the Otherworld, and in order to placate him, must change places with him for a year. The fairy king operates a double shape-change such that nobody in either one’s home will be aware of the difference. The fairy court is human-like except that everything is more magnificent than Pwyll has ever seen.29 Every night he shares a bed with the beautiful fairy queen but is careful not to touch her.

31In the second section, after his return home and in his own person, Pwyll meets his future wife, Rhiannon, who seeks his help to avoid an unwelcome suitor named Gwawl. At the house of her father, Pwyll makes the suitor a rash promise (a typical folklore motif), thus obliging himself to grant him Rhiannon’s hand. To extricate them from this situation, Rhiannon gives Pwyll a magic bag, which Pwyll tricks Gwawl into entering, thereby rendering him helpless. Pwyll releases his rival from bondage in exchange for his submission and is now free to marry Rhiannon.

32In the third section Rhiannon gives birth to a son. The boy mysteriously disappears during the night and his mother is accused of murdering him. Her guilt cannot be proved but she is obliged to do penance. No explanation is given for the strange disappearance but later events show it to be of fairy origin; it is well known that the fairies steal human babies. He has been carried off, presumably by Arawn the fairy king (for whom no son was mentioned in the earlier section). But he will be returned in complex circumstances, showing signs of his sojourn in the Otherworld.

33In the fourth section, Teyron lord of Gwent finds a baby boy wrapped in silk, left by the fairies at the stable door, perhaps in repayment for his mare which they had been using to produce foals. He and his wife name him Gwri Golden-hair. Unlike a normal human child, he matures and grows very rapidly. But it soon becomes obvious to them that he is the missing son of Pwyll, whom he closely resembles. They bring the boy back to his parents, and Rhiannon names him Pryderi, literally ‘care’, the care from which she is now released. He eventually succeeds his father as ruler of Dyfed (South Wales).

34In addition to the overall marvellous atmosphere, certain events in the story recall similar details in the Breton lays. The fairy king who carries off a human being recalls Sir Orfeo, as does the love between Pwyll and his wife. The baby abandoned at the door and the child’s silk wrapping are analogous to Lay le Freine’s swaddling cloth which leads to her identification when she grows up. The strange origin of Pryderi, human but with Otherworldly powers gained during his fairy sojourn, recalls the demonic birth of Sir Gowther in the lay of that name.

Branwen Daughter of Llŷr

35This is a tale of violent passions, pride and misunderstanding, love and betrayal, death and destruction, in the course of which the lovely Branwen dies of a broken heart because, she says, ‘two good islands have been laid waste because of me.’ The islands are Britain and Ireland, in both of which the action takes place. Branwen is a Welsh princess, the ‘fairest maiden in the world’ whose brother Bendigeidfran gives her in marriage to Matholwch, King of Ireland. But war breaks out between the two kings owing to the jealousy of the lady’s quarrelsome uncle, Efnisien. Peace is briefly restored, and the British ruler gives the Irish king a magic cauldron as a token of friendship.

36This object, the Cauldron of Life, was destined to have a long literary life, though it would be somewhat metamorphosed in the course of time. It came originally from Ireland, having been taken out of a lake by two Irish giants, a man and a woman. It had the desirable property of restoring life to dead warriors who were put into it, who nevertheless remained without speech, a feature which makes them sound rather like zombies. Its first owners had been driven out of Ireland, taking refuge in Britain with King Bendigeidfran and bringing the Cauldron of Life with them.

37In Ireland, Branwen and Matholwch have a son, Gwern, but Branwen is not well treated and she trains a raven to take a message to her brother, who then crosses the sea with an army to rescue her. Bendigeidfran himself is a giant of a man who can fit into neither house nor ship but who is able to wade across the Irish Sea on foot, and then to lie down across the River Liffey in order to make a human bridge for his own men to pass over. A truce is again made, but during the banquet the evil uncle, Efnisien, kills the boy Gwern by throwing him into the fire. The Irish attempt to use the magic cauldron to restore their dead warriors to life, and Efnisien, thinking to gain some advantage from it by following their example, arranges to have himself thrown in alive; but the result is catastrophic, for the cauldron bursts apart and he is killed.

  • 30 One thinks inevitably of the cloth-clothes-mantle-gloves motif which plays a prominent role in seve (...)

38Bendigeidfran is wounded with a poisoned spear and, before dying, gives orders to the seven British survivors, including Pryderi, to cut off his head and take it back to Wales. When they return to Britain they find that a usurper, Caswallan, has taken control and has been crowned king in London. Caswallan has a magic mantle which makes him invisible (though not his sword), thus allowing him to kill his enemies with ease.30

39Pryderi and the other survivors spend seven years feasting in Harlech with the head of their dead king. Then they spend eighty years in a palace overlooking Cornwall in a state of bliss, listening to wondrous birdsong. But a taboo has been laid upon them, not to open a marvellous window which would allow them to look upon Cornwall – why, is not explained, but the window may be interpreted as a portal, a passage into the Otherworld. When one of them insists on opening it, after eighty years, normal memory returns, they lose all joy, are filled with sorrow and grief, and must set out for London to bury the head in the White Mount.

40In Ireland only five pregnant women remain alive after the bursting of the magic cauldron. They and their five sons divide the island between them and become the ancestors of the Irish people. This ‘explains’ how Ireland came to have five provinces ruled by separate kings, which was historically correct in the 11th century when the tale was composed.

  • 31 On the Grail see Norris J. Lacy (ed.), The New Arthurian Encyclopedia, 1991; 2nd ed., New York and (...)

41This complex romance is noteworthy for the introduction of the Cauldron of Life, taken from Irish tradition. The cauldron is just one of the sources of the Grail story, before the Grail was given a Christian interpretation and became the Holy Grail, the life-giving chalice.31

The Third and Fourth Branches of the Mabinogi

42The stories of the Third and Fourth Branches contain similar fantastic scenes and marvellous events. Far from being an incidental adornment, magic is central to the action. But the narrator makes no attempt to interpret or explain it; as a narrative device, magic is simply a given, an integral part of the tale. This is perhaps what makes the Welsh and Irish stories seem so fantastic in the literal meaning of the word.

  • 32 Again, one is reminded of similar elements in the Breton lays, as when Sir Degaré arrives at his gr (...)
  • 33 This male pregnancy fantasy, upsetting to traditional gender roles, might appear to be surprisingly (...)

43Pryderi and Rhiannon return in Manawydan Son of Llŷr, a tale filled with magic. Mysterious thunder is accompanied by a mist which portends strange events. A fairy castle appears suddenly out of nowhere,32 a golden bowl will not unloose the hands of anyone who touches it, and there are charms, enchantments and spells galore. Pryderi is again a central character in Math Son of Mathonwy. There are magic stallions and greyhounds, spells and enchantments (sometimes short term, just lasting for a day), and golden shields made out of toadstools. There are impressive conjuring feats, such as making a ship out of seaweed which is big enough to take human passengers and sail away. Shape-changing can happen by choice (turning oneself into an eagle), but it can also be imposed as a punishment. Sex change is no obstacle. Two men struck by a magic wand are changed into animals three times, male and female by turns, for a year each time (hind and stag, boar and sow, wolf and she-wolf); following their animal nature, they couple and give birth to young; this brings great shame on them when they eventually return to human form, since each of them has had young by the other.33

Culhwch and Olwen

  • 34 On The Mabinogion see N. Lacy (ed.), op. cit., p. 289-290.

44It has long been recognised by Arthurian specialists that The Mabinogion as a whole, and some of the tales in particular, are of special interest to the development of Arthurian tradition.34 One of the independent native tales, Culhwch and Olwen, is the earliest Arthurian romance in Welsh; it is nearly two hundred years older than Geoffrey of Monmouth’s Historia Regum Britanniae, which made King Arthur popular in England and France. Like the tales already mentioned, it is full of magic, shape-shifters, beasts and talking birds. It contains numerous fairy tale motifs such as the jealous stepmother, the asking of a boon, a fearsome giant (Olwen’s father), amazing tasks to be carried out by Culhwch in order to obtain Olwen’s hand (including a magic Irish cauldron recalling the one in Branwen), helping companions, and the hunting of an Otherworld beast. It is set in the court of King Arthur, portrayed as a British (Welsh) ruler. Although parts of it go back to the middle of the 10th century, it is such a complex story in structure and episode that it is questionable if any bard could ever have narrated it entirely in its present form (Jones & Jones, p. xxiii). This is significant from the point of view of the Breton lays, which Marie de France says she based on songs, not on long romances. She may therefore have heard only sections, scenes or vignettes from the long Welsh hero-tales which she described as ‘Breton.’

The Three Romances

45The Three Romances in the final section of The Mabinogion are Arthurian, but they are later, showing Norman French influence in their present form. Given the importance of Chrétien de Troyes, in Britain as well as in France, such influence is not surprising in Arthurian romance of the 14th century, when the manuscripts were copied. But the Celtic elements do not come from French originally, rather from Welsh, so it is hard to disentangle the threads of mutual influence. The point is that in the 12th century, when Marie de France wrote the first Breton lays at the Angevin court, she had at her doorstep in Wales a rich fund of stories which have been well preserved in later manuscripts, whereas in Brittany, conventionally associated with the ‘Breton’ lays, there is nothing in the relevant period.

46Speaking of the Three Romances, a passage from Jones & Jones (p. xxvi) has a direct bearing on Marie de France and her lays:

That the French-speaking Normans are in the direct line of transmission between Welsh traditions on the one hand and both Chrétien and the authors of the Welsh prose romances on the other, seems certain. By the twelfth century it is clear that matters Celtic were the rage in literature, and that for a variety of reasons Arthurian legend was the fashion. Arthur’s dominating position in British story; Geoffrey of Monmouth’s spectacularly popular History; the twofold advantage to the Normans of Arthur as a British, not an Anglo-Saxon, hero, with no unfortunate emotional or political connotations to his story, and the ease with which the imagination played about him rather more than defined figures like Charlemagne and duke William; the skill of the Welsh and Bretons as story-tellers […]: all these were elements of the supremacy.

47If ‘matters Celtic were the rage in literature’ (my emphasis), it is not surprising that Marie de France should have taken inspiration from them. Nevertheless there are two details in this quotation that require comment. One is the use of the word ‘Celtic’, which is anachronistic in the 12th century, since the term was simply not available to people of that time. For them, the equivalent expression was ‘Breton.’ The other is the mention of ‘the skill of the Welsh and Bretons as story-tellers’, which is begging the question, at least as far as the Bretons of Armorica are concerned. Whereas the Welsh in the 12th century had a rich fund of stories which would have appealed to the Anglo-Normans, as The Mabinogion demonstrates, there is not a shred of evidence that it was also true of the continental Bretons. No independent Breton literature has survived, and nowhere outside the lays of Marie de France is there any significant reference to Breton minstrels or storytellers. We may be justified in concluding that it was the Britons (Welsh) who were known for their wonderful stories, and that it was to them that Marie referred when she spoke of hearing ‘Breton’ songs on which she based her lays.

48The Mabinogion contains many references to feasts and entertainments featuring music and story-telling, such that one can almost hear the voices which may have inspired Dame Marie at the Anglo-Norman court. In Math Son of Mathonwy, there is a scene in Pryderi’s hall where Pryderi calls for a tale from his guests, who come in the guise of bards (p. 48). Gwydion, who is ‘the best teller of tales in the world’, reminds him of the custom that on the first night of a visit, the chief bard – Gwydion himself – shall have the say. ‘And that night he entertained the court with pleasant tales and story-telling till he was praised by everyone in the court.’ He does the same thing in a later scene where, at the end of a meal in the hall of the Lady Aranrhod, ‘she discoursed with Gwydion of tales and story-telling. Now Gwydion was a good teller of tales’ (p. 57). It is noteworthy, here, that the Lady Aranrhod not only wishes to hear tales, but wants to talk about them with the bard (‘she discoursed’). No doubt she wants to know more about them, their origin and composition, whether there are others like them, and so on; in fact Aranrhod has intellectual literary tastes, just like Marie de France.

  • 35 See Gerald of Wales, op. cit., Book II, ch. 13.

49It cannot be doubted that this is a realistic reminder of what actually happened in real life in the 12th century, when kings called on bards to tell stories, and ladies sat down with minstrels to talk about their art. And we need not wonder if Welsh bards were capable of learning enough French to tell their tales in the courts of the Anglo-Norman kings and the Marcher lords who controlled the frontier between Wales and England. Gerald of Wales is a good example of a bilingual Cambro-Norman who knew and loved Welsh music, song and story, which he practiced himself as a young man. Furthermore, on his several trips to Ireland in the 1180s, he was enchanted by the music and song of the native people which reminded him so much of his home in Wales where the culture was so similar.35 And with good reason, given the antiquity of Irish mythology and its direct influence on the development of Welsh heroic tales.

Chaucer’s Breton Lays

50The majority of English Breton lays of the later 14th century, those written in tail-rhyme, appear to be native English compositions not having any particular French source, but based more generally on common folk themes. The English poets therefore imitated the idea of a genre, a short romance of love and adventure, usually with a happy end. They were willing enough to call a poem a Breton lay even when it had nothing to do with ancient Britain or Brittany and without having any single, clearly defined source in French or English. The genre seems to be very loosely conceived in the 14th century, and hardly distinguished from romance. The story told is the important thing to these poets.

51In this light, Chaucer’s deliberate placing of The Franklin’s Tale in Armorica shows a determination to prove his awareness of the traditional geography and culture associated with the Breton lay since at least the 12th century. In other ways his tale treats the tradition with irony, giving astronomy and astrology a more important role than fairy magic or religious miracle. This is not to say that Chaucer could not also make use of Breton fairy stories when it suited his purpose, as in The Wife of Bath’s Tale. In fact, the latter’s straight-faced (non-ironical) use of a fairy queen who is a shape-changer gives it some claim to be a more genuine Breton lay than The Franklin’s Tale, where the submerging of the rocks turns out to be a purely natural phenomenon.

52The Franklin’s Tale is a Breton lay, firstly because the character says so in the prologue, and because he gives it many of the features associated with that genre. It is indeed a short romance which shares many of the characteristics of the Breton lays, beyond merely claiming to be one. Like many others it is set in Brittany (57 ‘In Armorik, that called is Britayne’), the characters have British Celtic names associated with the legendary history of Britain, it is a story of love and loyalty leading to a happy end after trial and separation, and finally, it contains surprising effects as well as disturbing references to magic and illusion.

  • 36 For a thorough discussion of this question, see Kathryn Hume, ‘Why Chaucer Calls the Franklin’s Tal (...)

53The Italian source should warn us to expect something new. Chaucer was not using any British or Breton story; nor is it French. There is no evidence that he knew the French lays of the 12th and 13th centuries, so they cannot be looked to as sources of inspiration for The Franklin’s Tale. Yet his determination to call it a Breton lay shows that he was well aware of the Middle English tradition at least. He certainly knew the longer English tail-rhyme romances, as well as the shorter English couplet-lays, but he did not draw directly on the latter for inspiration. Chaucer knew the Auchinleck Manuscript, so he would have been familiar with the older English Breton lays. He provides a brief summary of their characteristics in the Franklin’s prologue, though it is rather inaccurate, paying lip-service to the genre rather than trying to imitate it closely.36 He was aware of the ambiguity of the names Britain/Brittany, taking the trouble to remind readers that England, too, was called by this name (138 ‘In Engelonde, that cleped was eek Briteyne’); the Middle English form, ‘Briteyne’, is identical for both the island and the duchy.

54Chaucer plays upon the expectation raised by the Breton setting which would normally lead the audience to expect supernatural effects, the kind of magical overthrow of the laws of nature that was traditionally associated with the Celtic Otherworld as commonly found in medieval Irish and Welsh texts, as well as in the English Breton lays. At first sight The Franklin’s Tale seems to reinforce the expectation that magical effects are going to be produced by enchantment. Thus a feeling of suspense is raised.

55Clearly Chaucer was very familiar with the elements that made up a Breton lay and understood that this was a literary tradition of long standing in both England and France. It was enough for his purpose to know that there was a ‘Breton lay’ tradition in English poetry, with certain expected characteristics. How far back that tradition went would not have been his concern; in general Chaucer’s approach to literature is not historicist but contemporary; he enjoys questioning convention and playing with commonly received ideas. Vagueness about the meaning of ‘Breton’ already existed in the 12th century and two hundred years later its use as a topos had grown more pronounced in England. As the setting of The Franklin’s Tale shows, even Chaucer was not immune to the force of cultural cliché, since his reference to the ‘olde gentil Britouns’ does not appear to be ironic. But, as usual, he likes to have fun with convention by pushing it beyond its limits, and The Franklin’s Tale aptly does so.

Conclusion: Twilight in the ‘Celtic Fringe’

  • 37 Upon the death of Queen Elizabeth in 1603, James VI, King of Scotland, inherited the crowns of Engl (...)

56There is no doubt that ancient Irish and Welsh mythology preserved stories about the fairy Otherworld which were unknown to the Romans as well as to the Germanic kingdoms of the Franks and the Anglo-Saxons. In modern critical terms, the influence of Celtic literature on the dominant political groups is a post-colonial phenomenon; indeed, the word ‘Celtic’ itself only dates from the 18th century, which is why it is doubtful to what extent it can be used without quotes in the medieval period. At that time, indeed, ‘Breton’ meant more or less what ‘Celtic’ would come to mean in the modern world, especially from the 18th century onwards down to the present day. To speak of a ‘Celtic Fringe’ implies an Anglocentric viewpoint, one in which England and the English are the focal point and the English are imagined as looking northwards and westwards at the neighbouring countries of Scotland, Wales, and Ireland. It also suggests an imperialist attitude which is associated with the ambitions of the Kings of England to conquer and unite under their control all parts of the British Isles, a gradual process that became more marked under the Normans, from 1066 onwards. The Kingdom of England is not coterminous with the island of Britain but it was the English kings who sought to make it so, over a long period of time, by engaging in military expeditions to bring Wales, Scotland, and Ireland under the control of English monarchs. The history of the later Middle Ages is thus marked by their attempt to make ‘English’ and ‘British’ mean the same thing. Though this was never fully achieved, and has always been rejected by the so-called Celtic Fringe, in English usage it became established in the 17th century following the union of the crowns (1603).37

57Because of the way in which the word ‘British’ has been used in modern times, very much associated in the 19th century with the British Empire, it remains constantly necessary to insist on the fact that the further back in time we look, ‘British’ most emphatically does not mean ‘English’ – in fact, quite the contrary. When Marie de France was writing her Breton lays in the 12th century, it was still clear that ‘Britons’ and ‘British’ referred only to the Celtic inhabitants and their language. It was equally clear that in the ‘real’ Arthurian period, traditionally ascribed to the 5th-6th centuries, there was no distinction between Britons and Bretons, the latter being simply an ethnic and linguistic offshoot of the former, settled in Armorica. The spelling is often indifferent to the meaning. In Middle English texts, the word ‘Breton’ invariably means Briton or British, as distinct from English or Anglo-Norman. Medieval writers were not unaware that Britain and Brittany occupied different geographical spaces, but the inhabitants of both places were believed to share a common origin and language. ‘Breton’ or ‘Britoun’ (as Chaucer writes it) meant in Middle English the British ancestors of both the Welsh and the Bretons. Even ‘Welsh’ applied to a territory of greater extent than today: the Old Welsh language was spoken in the north-west and south-west of England, from Cumbria to Cornwall, and in earlier times had been carried across the sea to Armorica. Brittany is also part of the ‘Celtic Fringe’, occupying a position in medieval France similar to that of Wales in relation to England.

  • 38 William Butler Yeats, The Celtic Twilight: Faerie and Folklore, 1893; enlarged ed. 1902, New York: (...)
  • 39 John B. Beston, ‘How Much Was Known of the Breton Lai in Fourteenth-Century England?’, in Larry D. (...)

58In 1893 the Irish poet and folklorist William Butler Yeats entitled one of his earliest books The Celtic Twilight, a collection of supernatural tales based on his own fieldwork and personal beliefs.38 His use of the word ‘twilight’ was intended as a direct reference to Wagner’s Twilight of the Gods (Götterdämmerung), meaning their dwindling into obscurity. The expression ‘Celtic Twilight’ became a way of referring to a kind of romantic, dreamy, and mystical Otherworld, somewhat out of touch with hard-headed English realism, in which the imaginative Celtic peoples, who believed in fairies, were always on the point of stepping through the mist into Tír na nÓg. Such an atmosphere was not new, however; it was already present in the Welsh Mabinogion in the 12th century and was seeping into the Anglo-Saxon and Anglo-Norman consciousness, never to be lost. This contradicts the suggestion by Beston that The Franklin’s Tale was ‘too late in the century to gain from the former association of Brittany with marvels.’39 Far from being ‘former’, this association was still living and vibrant, part and parcel of what the English expected of the Welsh and the Irish; magic and marvels are still present in the tale, and though Chaucer provides a scientific explanation for some of these phenomena, he by no means explains all of them away. To do so, indeed, would be to destroy the ‘Celtic Twilight’ as it was already perceived in the 14th century.

Haut de page


1 Although the name ‘Marie’ occurs in one of the lays attributed to her, the designation ‘Marie de France’ has only been in use since the 16th century. No Anglo-Norman or Middle English readers would have known her in this fashion. It has, nevertheless, become the conventional way of referring to the writer who may have been the author of the group of Old French lays found in British Library MS Harley 978 (early to mid-13th century).

2 See J. R. Clark Hall, A Concise Anglo-Saxon Dictionary, 1894; 4th ed., Cambridge, Cambridge University Press, 1960, p. 398: ‘wealh (gs weales) foreigner, stranger, slave: Briton, Welshman: shameless person. Wealhcynn men of Wales, Britons.’

3 While the ancient Greeks knew of a people they called Keltoi, it is only since the 18th century that it has become customary to identify the modern ‘Celts’ as people who speak the Celtic languages, currently Welsh, Cornish, Breton, Irish Gaelic, Scottish Gaelic, and Manx (the first three are classified as Brythonic, or British Celtic, the other three as Goidelic, or Irish Celtic, a division that went back to the Iron Age or earlier).

4 See Marie de France: Lais, ed. Alfred Ewert, 1944; Oxford, Blackwell, 1960, reissued with introduction by Glyn Burgess, Bristol, Bristol Classical Press, 1995. Based on Ewert’s edition is the translation by Glyn B. Burgess and Keith Busby, The Lais of Marie de France, 1986; 2nd ed., London, Penguin Classics, 1999. The most recent French edition is Lais bretons (XIIe-XIIIe siècles) : Marie de France et ses contemporains, ed. and trans. Nathalie Koble and Mireille Séguy, Paris, Champion « Classiques », 2011.

5 Logres is the Old Welsh name for England, so its occurrence in Old French and Middle English romance shows a Welsh origin and perspective.

6 Neustria was a name created for the western part of the kingdom of the Franks, after the death of Clovis in 511. It covered most of the north of present-day France, including Normandy but excluding Brittany, which was not part of the kingdom. The name fell out of use in the 10th century under the Capetian monarchy.

7 See Laskaya and Salisbury, op. cit., p. 1.

8 On the history of the Celtic languages see W. B. Lockwood, A Panorama of Indo-European Languages, London, Hutchinson, 1972, ch. 6, ‘Celtic’ (for Breton, p. 72-73). On the English / British (Welsh) issue, see for example J. R. R. Tolkien, ‘English and Welsh’, in Angles and Britons: O’Donnell Lectures, Cardiff, University of Wales Press, 1963, p. 1-41, reprinted in The Monsters and the Critics and Other Essays, ed. Christopher Tolkien, London, HarperCollins, 1983, p. 162-197.

9 Geffrei Gaimar. Estoire des Engleis – History of the English, ed. and trans. Ian Short, Oxford, Oxford University Press, 2009.

10 Larry D. Benson (ed.), King Arthur’s Death: The Middle English Stanzaic Morte Arthur and Alliterative Morte Arthure, Indianapolis, 1974; Exeter, University of Exeter, 1986.

11 Wales was conquered by Edward I of England in 1284 but remained recalcitrant; the principality was not fully integrated into the English kingdom until the Act of Union passed by Henry VIII in 1536. Likewise Brittany, united to the French royal domain in the late 15th century by the marriages of two duchesses, Anne and her daughter Claude, to successive French monarchs, but not fully integrated into the kingdom of France until 1532.

12 See C. Hall, op. cit.

13 Gerald de Barry signs himself in Latin as Giraldus Cambrensis, from Cambria, the old Roman name for the region; hence he is often called by the Latin form of his name. Speaking of the Anglo-Normans in 1189, Gerald says that ‘Cambria is called Wales nowadays, that having become its usual name, although it is a foreign word and not really correct’ (p. 220); and ‘To this day our country continues to be called Wales and our people Welsh, but these are barbarous terms’ (p. 232): see The Journey Through Wales and The Description of Wales, translated from the Latin by Lewis Thorpe, Harmondsworth, Penguin Classics, 1978. It comes as no surprise to find that in the Welsh language itself, ‘Wales’ is not the name used for the country, the people or the language, which is Cymru (the country), Cymry (the people).

14 In the short Prologue (p. 41) she speaks only of ‘lays which I had heard’; the Breton attribution is mentioned in the lays themselves, the first one, Guigemar, being typical of the rest: ‘I shall relate briefly to you stories which I know to be true and from which the Bretons have composed their lays’ (p. 43). The pages refer to Lais of Marie de France, ed. and trans. Burgess and Busby.

15 Emily K. Yoder, ‘Chaucer and the “Breton” lay’, The Chaucer Review, 12.1, Summer 1977, p. 74-77.

16 Nearby Normandy – Brittany’s neighbour and sometime rival – is another powerful duchy which was in theory part of the kingdom of France but whose dukes, in reality, ruled independently. Likewise Aquitaine, Eleanor’s duchy which she brought in turn to each of her husbands, Louis VII of France and Henry II of England. As far as the duchy of Brittany is concerned, the Plantagenet episode ended in 1203 with the death of Arthur, Duke of Brittany, a grandson of Henry II and Eleanor of Aquitaine.

17 See T. W. Moody and F. X. Martin (eds), The Course of Irish History, 1967; 3rd ed., Cork, Mercier, 2001, especially ch. 3 (by F. J. Byrne) and ch. 4 (by Tomás O Fiaich). See also Edward James, ‘The Northern World in the Dark Ages, 400-900’, ch. 2 of George Holmes (ed.), The Oxford Illustrated History of Medieval Europe, Oxford, Oxford University Press, 1988, especially p. 74-76 on the conversion of Ireland.

18 For this reason the Breton tongue in Brittany is closely related to Old Brittonic or Welsh, and not to the extinct Gaulish of Gaul, which belonged to the Continental branch of Celtic.

19 According to the dates given by Bede in his Historia Ecclesiastica, the conversion of Anglo-Saxon England took place over the course of nearly a century, between 597 AD, the year of St Augustine’s landing, and 686 AD.

20 See Seán J. Connolly (ed.), The Oxford Companion to Irish History, 1998; 2nd ed., Oxford, Oxford University Press, 2002, p. 93.

21 See Robert Welch (ed.), The Oxford Companion to Irish Literature, Oxford, Oxford University Press, 1996, p. 92.

22 See Peter Berresford Ellis, A Dictionary of Irish Mythology, Oxford, Oxford University Press, 1987.

23 Lebor Gabala is the Old Irish spelling of the name. The Modern Irish spelling is normally Leabhar Gabhála which reflects phonetic evolution, especially of b which has been lenited to bh, pronounced [w]. The text was edited by R. A. Stewart Macalister, Lebor Gabala Erenn, 5 vols., London, 1938-1956 (Irish Texts Society, vols. 34, 35, 39, 41, 44). For commentary see John Carey, ‘Lebor Gabala and the Legendary History of Ireland’, in Helen Fulton (ed), Medieval Celtic Literature and Society, Dublin, Four Courts Press, 2005, p. 32-48.

24 The identity of this mysterious poet, the presumed author of the lays, remains a vexed question, but there was a ‘Dame Marie’, referred to as such by a contemporary writer, Denis Piramus (c.1180); see Glyn B. Burgess and Keith Busby, eds and trans, p. 11-19. She was not called ‘Marie de France’ until the 16th century.

25 See introduction by Gwyn Jones and Thomas Jones (trans), The Mabinogion, London, Dent, Everyman Library, 1949, new revised edition 1993. All references to The Mabinogion are based on this translation.

26 Môn was Britain’s chief Druidic cultic centre, which, for purely political reasons, was destroyed by the Romans in 60 AD. It was also of importance to Gaulish Druids who were trained there, so its loss did not affect Britain alone. See Stewart Piggott, The Druids, Harmondsworth, Penguin, 1968; Ronald Hutton, The Pagan Religions of the Ancient British Isles: Their Nature and Legacy, Oxford, Blackwell, 1991; Kenneth O. Morgan (ed.), The Oxford Illustrated History of Britain, 1984 ; rev. ed., Oxford University Press, 1993, p. 17.

27 Gerald of Wales, The History and Topography of Ireland (1188), translated from the Latin by Lewis Thorpe, Dundalgan Press, 1951; rev. ed., London, Penguin Classics, 2004. This work testifies to the skill of Irish musicians, whom the Welsh and the Scots sought to emulate, with some success: chapter 94, p. 103-104. Gerald was speaking from personal knowledge and experience, since he had been in Ireland more than once, firstly to visit members of his family in 1182, and then to accompany Prince (later King) John in 1185.

28 The definite article in Welsh is y, which can be singular or plural, like English ‘the’, so it is not in itself a good guide as to whether y Mabinogi should be interpreted as a singular or a plural; both are possible, depending on the context. Lady Charlotte Guest, an Englishwoman who, though married to a Welshman, was not a native Welsh speaker herself, thought that Mabinogion was the plural of Mabinogi, a rare medieval word which is difficult to interpret. Her title is not quite accurate, but it stuck, and soon became conventional.

29 Claire Vial has effectively applied the looking-glass metaphor to Sir Orfeo’s visit to the fairy kingdom in her essay on this lay, ‘Sir Orfeo: Through the Looking-Glass, with a Harp’, in her book, ‘There and Back Again’: The Middle English Breton Lays, A Journey Through Uncertainties, Paris, Presses universitaires de France, 2013, p. 150-166.

30 One thinks inevitably of the cloth-clothes-mantle-gloves motif which plays a prominent role in several of the Breton lays, such as Sir Orfeo, Lay le Freine, and Sir Degaré. While the garments in question are not necessarily magical, they operate as either a means of concealment (Orfeo disguises himself three times) or of revelation; rather than making Le Freine or Degaré invisible, they reveal the owner’s identity, both to the hero/heroine and to their parents.

31 On the Grail see Norris J. Lacy (ed.), The New Arthurian Encyclopedia, 1991; 2nd ed., New York and London, Garland, 1996, p. 212-213.

32 Again, one is reminded of similar elements in the Breton lays, as when Sir Degaré arrives at his grandfather’s castle, and later when he comes upon the mysterious island castle where he is fated to fall in love. Sir Launfal, too, is taken by fairies, met unexpectedly in the forest, to Dame Tryamour’s enchanted pavilion which, like its owner, is normally invisible to humans – except when the lady decides otherwise.

33 This male pregnancy fantasy, upsetting to traditional gender roles, might appear to be surprisingly modern, almost post-Freudian; but it is clear that the Welsh author thought of it as a punishment, a humiliation inflicted on the victims.

34 On The Mabinogion see N. Lacy (ed.), op. cit., p. 289-290.

35 See Gerald of Wales, op. cit., Book II, ch. 13.

36 For a thorough discussion of this question, see Kathryn Hume, ‘Why Chaucer Calls the Franklin’s Tale a Breton Lai’, Philological Quarterly, 51.1, 1972, p. 365-379.

37 Upon the death of Queen Elizabeth in 1603, James VI, King of Scotland, inherited the crowns of England and Ireland (the latter country having been constituted as a kingdom under Henry VIII). Thereafter he was known outside Scotland as James I. This was a personal union, not a political one; the three kingdoms remained legally separate, each with its own parliament and administration.

38 William Butler Yeats, The Celtic Twilight: Faerie and Folklore, 1893; enlarged ed. 1902, New York: Dover Publications, 2011.

39 John B. Beston, ‘How Much Was Known of the Breton Lai in Fourteenth-Century England?’, in Larry D. Benson (ed.), The Learned and the Lewed: Studies in Chaucer and Medieval Literature, Cambridge, Mass., Harvard University Press, 1974, p. 319-336, here p. 332.

Haut de page

Pour citer cet article

Référence électronique

Leo Carruthers, « What makes Breton lays ‘Breton’? Bretons, Britons and Celtic ‘otherness’ in medieval romance »Études Épistémè [En ligne], 25 | 2014, mis en ligne le 01 avril 2014, consulté le 04 décembre 2021. URL : ; DOI :

Haut de page


Leo Carruthers

Leo Carruthers has been Professor of Medieval English at the University of Paris–Sorbonne since 1994. Born in Dublin, he graduated from UCD, National University of Ireland, in 1971, and has taught in France since 1977. In 1980 he defended his PhD at the Sorbonne, on the Middle English translations of the Somme le Roi, and in 1987 he obtained the State Doctorate, a study of the prose treatise Jacob’s Well. In 1988 he was elected first President of The International Medieval Sermon Studies Society. Since 2006 he has been President of AMAES, the French Association for Medieval English Studies. His recent publications include Reading Everyman (Atlande, 2008), Reading the Middle English Breton Lays and Chaucer’s Franklin’s Tale (Atlande, 2013) and, as editor, Jeunesse et Vieillesse : images médiévales de l’âge (L’Harmattan, 2005) ; Tolkien et le Moyen Âge (CNRS Éditions, 2007) ; Un espace colonial et ses avatars. Naissances d’identités nationales : Angleterre, France et Irlande (Ve-XVe siècles) (PUPS, 2008) ; Palimpsests and the Literary Imagination of Medieval England : Collected Essays (Palgrave Macmillan, 2011) ; Voix de femmes au Moyen Âge, (AMAES, 2011).

Haut de page
  • DOAJ - Directory of Open Access Journals
  • OpenEdition Journals
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search