Those interested in Celtic mythology, historians of the Welsh nation and students of the Arthurian tradition will all, at one time or another, have found themselves directed to a collection of Middle Welsh prose known by the curious name of the Mabinogion (pronounced Mabin-OGion). Compiled from texts found in two late-medieval manuscripts – the Red Book of Hergest and the White Book of Rhydderch – this collection was initially edited and translated by antiquarians William Pughe and Lady Charlotte Guest in the early nineteenth century. Guest and Pughe applied the term 'Mabinogion' (based on a spurious plural of mabinogi) to their translated compilation. While the Mabinogion collection itself might thus be regarded as a nineteenth century editorial creation, its constituent texts are authentic medieval productions, deriving from originals composed between the eleventh and the fourteenth centuries. They represent a golden age of narrative prose that flourished in Wales over the course of the central middle ages. This distinctive and evolving literary culture forms the context of the Mabinogion, and the focus of our interest in this introductory study.

Eleven separate tales are normally included within the Mabinogion corpus. Within these, two subgroupings - 'the Four Branches of the Mabinogi' and the 'Three Romances' - are traditionally recognised. In chronological order, the texts are as follows:

The Mabinogion texts are concerned with the heroic age or mythological past of the British Isles. They were not original compositions, drawing as they did on pre-existing traditional material, whether from oral or written sources. But these traditions were reworked, often to reflect contemporary concerns. We might read the Mabinogion as both an interpretation of a mythological past and a commentary on the medieval present.

The two and half centuries during which the Mabinogion texts were being composed represent a threshold of critical transition in Welsh history and literature. Here, in this little-known corner of the European Middle Ages, we find the thought-worlds of oral antiquity and literate proto-modernity face-to-face in curious proximity. The transition between the two can be traced as a literary process - which we can observe unfolding on the very pages of the Mabinogion.

The initial experiments with narrative prose are to be found on the pages of Culhwch ac Olwen, representing both a parody and an indulgence of the naïve excesses of the oral-narrative culture out of which this literature was emerging. By the end of the twelfth century, Middle Welsh narrative prose was in its second or third generation, and (along with poetical and triadic material) formed part of an expanding, self-referencing literary tradition. Works such as Llud and Llefelys and the Four Branches of the Mabinogi belong to this period. Vernacular literary self-confidence, as well as foreign influence, accounts for the gilded splendour of thirteenth-century works such as the Three Romances and the Dream of Macsen Wledig. The conclusion of this tradition is marked by the Dream of Rhonabwy, where literary self-consciousness has come full-circle and finally turned in on itself – anticipating the sloughing of the medieval spirit that took place throughout Europe in the following centuries.

Far from being 'a ruin of antiquity' – as Matthew Arnold misunderstood the Mabinogioni – these texts are better understood as constituent parts of a complex and ongoing literary conversation. Within this unfolding tradition, each name, motif and reiterated incident would have formed part of a cumulative constellation of meaning. To understand this intertextual culture, we need to taken in something of the broader historical development of Medieval Welsh literary prose: from the laconic marginalia of the Early British church; to the narrative interlacing of the Four Branches; to the florid fantasy of thirteenth-century Romance.

This introduction will consider a number of themes bearing on the early development of Medieval Welsh literature. We will be looking at the role of the oral tradition – known as the cyfarwyddyd in Medieval Wales – which is especially relevant to the earlier Mabinogion texts. A clarification of the overlapping (but distinct) concepts of myth and storytelling will also be necessary to help us understand the primary elements of this prose ensemble. The social impact of literacy as a general phenomenon, and the specifics of the vernacular literary culture in Wales will also need to be considered, in particular the so-called Triads of Britain, and works of the mythical Taliesin, both of which have close links to the Four Branches and Llud a Llefyls. It will be necessary for us to understand the cultural changes that were taking place in Wales in the thirteenth century – an infusing of Romance and other Continental influences (within and beyond the literary world) which did so much to define the quality and content of the later Mabinogion texts. Finally, we will be looking at the manuscript tradition in thirteenth- and fourteenth-century Wales: a tradition which culminated in the Red and White Books in which the Mabinogion (along with much of the rest of medieval Welsh literature) has survived to this day.

The Cyfarwyddyd – Oral Lore in Medieval Wales

Professor D. A. Binchy once described medieval Irish society as 'rural, tribal, hierarchical and familiar'. The same might well be said to apply to much of western Britain in the early middle ages. Medieval Welsh society was organised around a network of tribal courts, each of which supported a 'household retinue' (teulu) of spear-carrying youths. Cattle-raiding and other forms of low-level conflict between and within these agnatic court communities were the norm rather than the exception. Tribal aristocracies of this type have thrived in a variety of contexts throughout the temperate world: from the Teutonic forests to the plains of the Masai Mara. It might be regarded as the characteristic social form of cattle-based economies at the 'Heroic Age' level of techno-cultural development.

In pre-modern societies such as these, the oral tradition is the medium of collective memory: fluid in its details, but essentially static and conservative in its overall ethos. Its constituent elements might take the form of genealogies, origin legends, hero-tales, topographic lore, wisdom literature (proverbs and gnomic statements), anecdotes and agreements bearing on local law and custom. Early Welsh literature contains examples of all of these genres, and we might assume that much of this material was informed, directly or otherwise, by the ambiant oral tradition. A Welsh term for this body of recitational learning was the cyfarwyddyd (cer-var-with-id), a word which in the modern language simply means 'information' or 'instructions', but in the medieval period probably had a meaning closer to 'lore' or 'testimony'.

Occasionally we find examples of the cyfarwyddyd recorded in writing in a more or less unprocessed form. Some interesting early examples, dateable back to the eighth or ninth centuries, are to be found inscribed onto the margins of an illuminated manuscript known as the Lichfield gospels, also known as the Book of Chad. Despite its association with the West Midlands, this holy book seems to have resided in a Welsh-speaking context at some stage in its history, as indicated by the presence of a number of scrawled notes written into its margins in the Old Welsh language. These marginal notices cover a variety of subjects, but one of the longer examples – the so-called Surexit memorandum – records what would appear to be resolution of a local land dispute:

Tudvwlch, the son of Liwid and son-in-law of Tudri, arose to claim the land of Telich, which was in the possession of Elcu, the son of Gelhi, and the tribe of Judgored : he complained long about it : at last they dispossess the son-in-law of Tudri of his right : the nobles said to one another 'let us make peace' : Elcu gave afterwards a horse, three cows, three newly calved cows, provided only there be no hostility between them from this reconciliation thenceforth till the day of doom : Tudvwlch and his people will require aftewards no title for ever and ever. Witness Teilo, etc. Whoever observes it will be blessed, whoever breaks it will be cursedii

A similar note further on in the manuscript records a list of rents due from various tenants on a local monastic estate. The informants responsible for this summary of rental dues are described as the cimguareit, the Old Welsh form of the word cyfarwyddiad, 'guides' or 'storytellers' (i.e. local individuals schooled in the cyfarwyddyd). Interestingly, we find the term used in more or less exactly the same context in the First Branch of the Mabinogi, when, after his defeat of Hafgan, Pwyll instructs the vanquished king's subjects to 'take a reckoning [kyvarwyd] and find out those who owe me allegiance'.iii

That such matters should be recorded in the blank spaces of a holy book such as the Lichfield Gospels should not surprise us. The book, as a sacred object, would have been an appropriately august repository for the solemn agreement between the two feuding families – bound as it was by the holy and dread names of the local saints – as well as a natural enough place to record the rental dues from the monastic estate. These marginalia form part of a general trend which saw the Early Medieval Church (with its privileged access to writing technologies) employed by local elites as the guarantor of its legal and economic arrangements and a custodian of its communal records. What was originally entrusted to the memory of the cyfarwyddiad became increasingly committed to the vellum archives of the local monastery. An important branch of medieval Welsh literature owes its origins to this process.

Among the most significant components of the cyfarwyddyd was genealogy. Who is related to whom is always the central question in tribal societies. Not only did one's family affiliations determine issues of marriage and inheritance, they informed a far wider range of social arrangements: including who one eats, works or hunts with – who one can and cannot speak to. In a hierarchical society such as that of native Wales, the extended kindred represented not only the primary source of security and companionship for the individual, but also the determinant of his status and relations within the wider community. Kin-affiliations were, nominally at least, the precondition of political opportunity. In traditional Irish law, anyone connected by four degrees or closer to the living king belonged to what was called the derbfine – the 'certain kin'. The office of kingship was in theory open to all adult males within this extended family network. Succession was almost always an internecine issue, if not a family affair. In Wales this hallowed circle was restricted to the kindred of the third generation (offspring of a common great-grandparent), but the basic assumption was the same. Lower degrees of nobility were defined in a similar way. Family connections carried a considerable premium, even across widely extended kinship networks. According to Gerald of Wales, even the most humble native Welshman could trace his ancestry back seven or eight generations – this was the primary means by which individuals could assert an affilition, however remote, within the complex fabric of the tribal aristocracy.iv

Unsurprisingly perhaps, some of the earliest written records from the Welsh-speaking world take the form of genealogical documents. We have a number of texts of this kind from Medieval Wales dating back as early as the ninth century. The most important of these are the so-called Harleian genealogies, drawn up in the mid-tenth century. These include twenty-seven lineage lists, recording the descent and interrelationship of the royal kindreds of Wales and the so-called 'Old North' (Cumbria, Strathclyde, Lothian and the Pennines area – much of which was Celtic-speaking until the seventh century and beyond). While these genealogies cannot be described as accurate in the modern historical sense (particularly at their more remote extremes) they do offer an interesting perspective on the dynastic politics of Medieval Wales – and the way in which these politics were informed not only by genealogy but also (as we will see) by the mythic imagination.

The ruling families of North Wales almost all traced their descent from the Dark Age warlords of the Old North, who in turn were related to semi-historical Romano-British figures such as Coel Hen ('Old King Cole') and Padarn Pesrut ('Paternus Red Cloak'). Further back still, a connection is made to the Belgic overlords of southern Britain in the late Iron Age – Caswallawn ( > Cassivellaunos, the Belgic warlord who fought against Caesar in the first century BC); and Cradawg ( > Caractacos, leader of the resistance to the Claudian invasion in the following century). Caswallawn, in turn, is related to figures such as Afallach and Beli Mawr, who belong to the mythical horizons of British tribal history. This ambiguous zone is explored at close quarters in Mabinogion texts such as Llud and Llefelys and the Four Branches of the Mabinogi, about which we will have more to say below.

The Hen Gogledd (Old North) was an important locus of tradition in Medieval Wales. For the royal dynasties of North Wales in particular, recalling the glories and defeats of these Dark Age warlords constituted a vital aspect of their specific cultural identity. A poignant institution was the recitation of the great elegy known as the Gododdin, which recalled the last great stand of the men of the North against the armies of Northumbria in the years around 600 AD. The opening lines are given here to give a flavour of this epitome of Old Welsh heroic verse:

Men went to Gododdin, laughter-inciting,
Bitter in battle, with blades set for war.
Brief the year they were at peace.
The son of Bodgad, by the deeds of his hand did slaughter.
Though they went to churches to do penance,
The young, the old, the lowly, the strong,
True is the tale, death oer'took them.

The core of the poem is generally believed to be more or less contemporary with the events themselves, composed by a seventh-century poet who would have been personally acquainted with the fallen warriors commemorated therein. The Gododdin would remain part of the bardic canon for another six centuries, no doubt accumulating extra material along the way. Nonetheless, it may be regarded in all essentials as an authentic record of seventh-century events. Another small group of North British poems, attributed to the 'historical Taliesin' (as opposed to the 'mythical Taliesin' - about whom more will be said in due course), might be considered to belong to the same category. The hengerdd, or 'Old Songs', were an important complement to the prose hen chwedlau ('Old Stories') within the oral tradition of that was the cyfarwyddyd.

The historical link between the Old North and Wales was the semi-legendary Cunedda, grandson of Padern Pesrut. The warlord Cunedda was believed to have arrived in Wales with his sons at the end of the Roman era, supposedly establishing a number of small district kingdoms which left their impression on the political geography of Wales. Thus the cantref of Merioneth was held to be named after Cunedda's son Merion, while the district of Dunoding was believed to have acquired its name after another son Dunawd, while a third of these brothers – Ceredic – is held to have been the eponymous founder of the kingdom of Ceredigion.

It is not inconceivable that at least some of these names represent the memory of flesh-and-blood individuals from the fourth-, fifth-, or sixth centuries. But considerable doubt must be cast on the neat genealogical scheme which identifies a sibling relationship between the eponymous founders of the particular northern Welsh districts. Spurious blood relationships of this kind are entirely typical of pre-literate tribal history, and represent a convenient way of expressing contemporary political relationships rather than being genuine record of dynastic realities. We find a clear example of this in the traditional lore of Ghana – an oral record which has many similarities with the cyfarwyddyd of Medieval Wales. Anthropologists record tales of the legendry Jakpa and his sons, founders of the kingdom of Gonja. Like the sons of Cunedda, each of his five sons was associated with one of the territorial subdivisions of this regional hegemony. Some years later, after two additional territories had been incorporated into this West African overkingdom, and sure enough a new variant of the foundation legend was recorded: with seven sons instead of the original five. We might expect that the number of sons of Cunedda may have fluctuated in a similar way with the ebb and flow of regional politics in Early Medieval Wales.

This example illustrates the dynamics of the oral tradition, in which contemporary geopolitical facts on the ground are frequently back-projected into a schematic tribal-historic past. The Welsh genealogies were clearly affected by this kind of process – a testament to their oral origins. However, at their more remote reaches, these lineages also alluded to more generic tribal-historic traditions – recalling names of saga heroes known to the poetry and prose narratives such as we find in the Mabinogion, but also not unheard of in more reputable historical sources as well. We know, for example, that Caswallon ( > Cassivellaunos) and Maxen Wledig ( > Magnus Maximus), both of whom feature prominently within the genealogies, were genuine historical figures, however embellished their profiles might have become in the looking-glass world of the medieval Welsh narrative. The mythical element within this tradition – so often talked about, but so rarely understood – is what we will be considering in the following section.

Sacred Stories: Myth, Custom and Belief in the Brythonic World

Myth, as the philologist K. H. Jackson once wryly remarked, has never lacked definers. Perhaps for this reason, it remains remarkably difficult to establish a workable consensus on what myth actually is, let alone its value or significance. The word comes from the Ancient Greek muthos, meaning 'things recited', and refers primarily to archaic narratives, which typically concern the gods or significant human ancestors and includes a strong supernatural component. During the Hellenistic era muthos was sometimes contrasted unfavourably with logos (analytical reasoning) and thus acquired a wider definition which encompassed all discourse of a primitive or pre-rational nature. Even today, 'myth' retains this pejorative colouring, and is often used to denote a widely-believed fallacy (e.g. 'urban myths'). However, this has not always been the case. Commentators from Plato onwards have regarded the strange, dream-like recitations of antiquity as shimmering with esoteric significance, the visible veil of a higher invisible truth. Something of the same exalted view of myth also informed Renaissance scholars, among whom the science of mythology was born in an effort to extract the inner truth from these ancient narratives through a means of exegetical analysis, just as their scholastic forerunners had worked over the Christian scriptures. More recently, myth was held in highest regard the Romantic folklorists and philologists of the early nineteenth century. A measure of scientific gravitas was leant to this predeliction by the psychologist Carl Gustav Jung, who found in mythic narratives the key to what he described as the 'collective unconscious': a genetically-inherited repository of archetypal images and motifs which continues to inform our thought-processes and cultural productions. More recently Northrop Frye, Kenneth Burke, Fredric Jameson and Camille Paglia have worked from a similar starting point (albeit with rather different destinations): exploring myth as a cultural fountainhead, an eternal paradigm, a set of irreducible formal principles which rework themselves inexorably throughout the history of human thought and expression.

Here, we will confine ourselves mainly to the original definition of myth as simply as narratives of traditional (i.e. anonymous, oral) provenance, concerning the distant past and characteristically involving a supernatural element, some of which may be related to pre-Christian magico-religious beliefs and ritual practices. We will attempt to avoid value-judgements such as those described above, but instead focus on what such material meant to the authors and audiences of the Mabinogion texts, within the cultural context of the Welsh Middle Ages. Inevitably, this will require some consideration of the question of origins. But in line with the strain of 'new criticism' that has emerged in Celtic studies over the last forty years, we will also be adopting the synchronic view – and looking at the function of this mythic material within the extant medieval narratives in which it was used. We will be attempting this latter task in more detail within our studies of the individual texts themselves found elsewhere on these pages.

Myth, it must be emphasised, does not exist in isolation, but rather should be understood as part of a broad cultural continuum that includes proto-legal and political origin stories, usually embedded within tribal-historical tales of the ancestors. Another important (though frequently misunderstood) element of myth was contributed by the narrative ideation underlying magico-religious custom and belief. The pre-literate world did not clearly distinguish these categories. What causes magical and fantastical elements to feature within the memory of actual past events is a complex psychosocial question which lies somewhat beyond the scope of this particular study. It is enough to suggest that the more remote the events in time, space and stature; the greater the extent to they will be assimilated into the magical imagination.

But what is this magical imagination? What, indeed, is magic? Restricting ourselves to one of its less problematic definitions, magic might be defined as a series of actions and modes of thought designed to construct a sense of power which is then deployed for a variety of purposes: primarily to counteract the helplessness experienced by the unstable ego of primitive man. v Magic in this sense is a largely private process, a prototypical form of thought: 'active and creative – pursuing a purpose, finding means, investing power in things, believing in the power, and making use of it.' vi

A parallel process can also be seen at work at a collective level – and this is the magic that interests us here, relating as it does (indirectly at least) to the mythical material of Mabinogion. Communal or ceremonial magic is typically enacted through rituals, dramas or seasonal pageants, often representing the triumph of tutelary ancestral figures against the malignant supernatural forces. The classical Frazerian paradigm of communal magic would be the fertility rite, a seasonal festival to promote the fruitfulness of the land or celebrate (or enable, or participate with) the onset of spring. We might suspect that customs of this kind lay behind stories of 'the king and his prophesied death', the climax of which typically represents the killing of an aging king (representing thanatos, or the departing winter) by a young hero (representing libido, or the incoming spring). A myth of this kind, albeit in a distorted form, can be detected underlying the Fourth Branch of the Mabinogi. Another distinctive Brythonic motif involves the unsettling arrival of a menacing stranger (Pwyll as Pen Annwfn, Gawain's Green Knight, Melwas of the Arthurian tradition) who appears uninvited at the door of the tribal hall on a certain seasonal feast-day, demanding food or sexual favours. These stories, with their chthonic undertones, perhaps contain echoes of the same folkish pageants underlying traditions such as the Mari Llwyd parades, or the frenzied cavorting of the Cornish 'Obbyoss'.

But communal magic could be enacted at different times and for a variety of reasons: not only at significant junctures in the sacred year, but also aperiodically at particular moments of danger or transition. One such ceremony, which was enacted by a coastal Inuit community during the onslaught of a fierce Arctic gale, was described in detail by the Danish explorer K. Rasmussen:

The Shaman was Horqarnaq, a young man, who took some time to enter into a trance. He explained to Rasmussen that he had several helper-spirits at his disposal: the spirit of his late father; the latter's helper-spirit, an imaginary human figure made out of snow; and a red stone that he had found one day when out hunting. He is dubious about his skills, and is encouraged gently by the village women … He enters slowly into an almost frenzied trance and the audience increases, trying to stimulate his frenzy … Finally he no longer recognises the people around him and asks who they are. Then one of the helper spirits enters his body; he no longer has control over his actions; he jumps and dances around, and invokes his father's spirit, an evil spirit. His recently widowed mother is also present, and she tries to calm her son, but others encourage him to greater frenzy. He then names several other spirits of dead people, whom he sees in the hut, among the living. The old women try to guess who it may be, becoming more and more excited as they attempt to solve the mystery. Then one old woman comes forward and calls out the names of the people whom her sisters had not dared to mention: a couple from Nagiutoq who had died quite recently. The shaman cries out that it is they. They have been turned into bad spirits and are the cause of the tempest. The séance goes on for another hour, amid howls and cries and the noise of the storm outside … Then a fearful thing happens. Horqarnaq leaps at old Kigiuna and seizes hold of him; he shakes him brutally and pushes him into the centre of the hut. They struggle and grunt and eventually he, also, is in a trance and follows the shaman docilely until they fall to the floor where they roll around, possessed. The old man seems to be dead and is dragged over the floor like a sack of old rags … The tempest has been killed symbolically. The shaman bites the old man and shakes him like a dog would a rat … The people are silent while Horqarnaq continues his dance … Then he slowly becomes calmer, kneels down by the body and begins to massage life into it. The old man revives and eventually gets to his feet. But he has only just managed to do this, when the whole scene is repeated, and he is again seized by the throat. This happens three times: three times he is “killed” in order to show that man is superior to the tempest. Finally, it is the young Shaman that faints, and the old man rises up and describes the images that are racing before his eyes – naked men and women flying in the air, causing the tempest to swirl before them … they are afraid and they are fleeing … Among them is one whom the wind has filled with holes; the wind blows through these, causing the whistling noise … he is the strongest and will be mastered by the old man's helper-spirit … Then the young shaman recovers and they both begin to chant and sing plaintively, addressing the Mother of marine animals and begging her to send away the evil spirits, to bite them to death … So, the two shamans struggle until the tempest is finished, and the people return, reassured, to their huts, prepared to sleep. vii

This harrowing Arctic episode displays the violent hysteria of the magical consciousness in its raw, unprocessed form. Nonetheless, even within the apparently spontaneous outbursts of Horqarnaq and the old man Kigiuna, we can see the cultural influences at play (e.g. the references to 'the Mother of Marine Animals', a traditional Inuit figure). From a sociological perspective, we might even suspect the structure of the séance itself was more or less stereotypical or culturally-determined. Both the entranced participants and the onlookers saw, more or less, what they expected to see. In this way, the ritual served its purpose 'until the tempest is finished, and the people return, reassured, to their huts, prepared to sleep'. Thus culture prevails over nature.

We will consider a second example, recorded in the eleventh-century Lacununga manuscript from late Anglo-Saxon England. Entitled WiÞ Fæstice 'Against Rheumatism', it includes a dramatic incantation which evokes a vision of both the aetiology and cure of this uncomfortable affliction:

Loud they were, lo loud, when they rode over the mound,
They were fierce when they rode over the land.
Shield yourself now that you may survive their ill-will.
Out little spear, if you are in here!
I stood under the linden-wood, under a light shield,
Where mighty women betrayed their power,
and screaming they sent forth their spears.
I will send them back another,
a flying arrow from in front against them.
Out little spear, if you are in here!
A smith was sitting, forging a little knife,

Out little spear, if you are in here!
Six smiths were sitting, making war-spears.
Out spear, not in, spear!
If there is a particle of iron in here,
the work of hags, it shall melt!

Whether you have been shot in the skin, or shot in the flesh,
or shot in the blood, [or shot in the bone],
or shot in a limb, may your life never be endangered.
If it be the shot of the Aesir, or the shot of the elves,
Or the shot of the hags, I will help you now.
This as your remedy for the shot of the Aesir, this for the shot of the elves,
This for the shot of hags, I will help you.
Fly to the mountain head.

Be whole. May the Lord help you. viii

This recitation would have belonged to a wider ritual-performative context, which we might assume was characterised the same ructions of consciousness associated with the exorcism of the arctic tempest described above. Here we find reference to the fairly widespread traditional belief that certain types of internal pain are caused by invisible arrows fired by supernatural beings (witches or elves are typically implicated in Anglo-Saxon contexts). In the first part of the incantation, the healer describes a past encounter with these monstrous entities, and alternates this description with protective injunctions to the present patient. The ritual climaxes with an extraction and dissolution of the agent of pain, addressed as a magical spear ('Out spear, not in, spear! If there is a particle of iron in here… it shall melt!'). After this, the focus incantation switches back and forth from the source of the afflicting spears (elves, witches, the old pagan gods) to the power of the healer to counteract these agencies. The ritual ends with an evocation of the power of the Christian god, and what would appear to be the dissolution of the extracted arrowheads into the herbal balm. A parallel might be drawn with Native American healing rituals in which magical projectiles were miraculously sucked out of the patients' bodies. ix Culturally-charged psychodramas of this kind, as is now generally recognised, have the potential to unleash a powerful placebo, as well as aiding in a more general sense the subjective restoration of wellbeing.

Both WiÞ Fæstice and the Arctic séance reveal with particular clarity both the performative and the traditional aspect of what we might call 'public magic'. Over time, some of the spontaneous psychodramas that represent magic in its most natural form might acquire canonical status and undertake the status of community rituals. The narrative element within these rituals draws on pre-existing mythic structures of belief – superstitions concerning the spirits of the dead, magical beings such as elves or witches, the powers underlying the cosmos and the forces of nature. As O'Keefe suggests in Stolen Lightning, his survey of the anthropology and sociology of magic 'the witchdoctor who cures does so with verbal incantations and references to myth…stories of primordial magical acts performed by culture heroes'.x But the very act of addressing the entities referenced in the magical incantation also reaffirms their existence, and underscores their significance within the magical universe. Thus the rituals themselves are mythopoeic, revitalising the reservoir of inherited dreams and nightmares that played such an important role in the primitive narrative tradition.

We have rather less direct information on how magic was practiced or understood in Medieval Wales – certainly nothing of the scale or type of the Anglo-Saxon Lacununga survives extant. However, an understanding of the magical ideation and behaviours exemplified in these comparative sources does much to account for some of the more outlandish and 'irrational' elements in the Mabinogion tales. One thinks of Rhiannon's equine penance in the First Branch, or Brân's decapitation in the Second (both of which have specific resonances with documented elements of pre-Christian ritual and belief). Other outlandish (but strangely loaded) gestures in the Four Branches include Manawydan's mock-execution of the captured mouse on the mound at Arberth, or Gwydion's incantatory restoration of the eagle Lleu – all of which might be thought of in terms of primordial magical acts enacted by culture heroes, as described above. Further evidence of magical ideation can be found elsewhere in the Mabinogion, such as in the curious stratagems deployed by the protagonists of Llud a Llefelys (scattering crushed insects over their enemies, digging up dragons from the centre of the island, speaking to one another through a long copper tube). The Romances too show evidence of magical thinking and behaviour, which is deployed in a very specific way in the distinctive literary mode defined by medievalist Anne Wilson as the magical plot. The magical aspect of these thirteenth-century Arthurian prose tales which is clearly apparent in a number sequences, such as the 'Fountain Ritual' at the beginning of Owain, or the banishment of the Mist Hedge Earl's bloodthirsty 'custom' towards the end of Geraint. In the earlier Arthurian tale of Culhwch ac Olwen, we are perhaps closest to the magical thought-world, with the complex series of tasks and injunctions to be accomplished by the hero arguably serving the underlying magical purpose, albeit through an almost carnivalesque form, of liberating the community's life-giving energies.

The Medieval Celtic texts normally described as 'mythical' are characterised by intermittent magical strategies of this kind. But, as we have already suggested, tribal history also features as a predominant concern. These tales are aetiological – origin myths in the most general sense. In many of these accounts, the actions of their larger-than-life protagonists to leave a physical imprint on the appearance on the landscape: plains are cleared, rivers diverted, mountains flattened, entire districts are flooded or drained. More frequently, the imprint of these formative deeds is on the naming of the land: a mountain here is so-called because it was the resting place of mythic hero, or the place where a dispute was settled between two tribal groupings, to give two examples from the Mabinogion. It was during this 'mythic time', as we have seen, that kingdoms or districts often acquired their name from some kind of eponymous founder. Other forms of geo-political development are also projected back into the Foretime: the Four Branches sees the unification of South Wales at the end of the First Branch; the ascendency of the North over the South is sealed at the beginning of the Fourth. Perhaps the most notable of these developments is the accession of Caswallon son of Beli at the end of the Mabinogi of Branwen, which provides the link between the genealogical networks of the Four Branches with those of the historical Welsh princes, as found in the Harleian genealogies and other related sources. Macsen and Llud have a similarly tribal-historic focus. Between them these Mabinogion tales set the scene for the pre-dawn of the Roman invasion, and thus mediate the transition from mythic into historical time.

The mythical narratives are different in scale but not necessarily in kind from the more recent tribal-historical cyfarwyddyd. We have a similar register of language being used in the accounts of contemporary dynastic-political events found in the Brut Y Twysogion. We have the same preoccupation with genealogy, the same dynamics of family conflict, the same accounts of tentative reconciliation. Legal custom and practice, as well as magic, had has its origin in the paradigmatic events depicted in tales such as the Four Branches. Thus, in the Second Branch of the Mabinogi, we find the Irish king Matholwch receiving compensation from the Britons in the form of a rod of silver and a plate of gold for the insult to his honour – the very same 'face price' (wnebwerth) for a prince of Aberffraw that we find stipulated in the lawbooks of Wales. The exchange of livestock that accompanies this treaty resembles nothing so much as the 'horse, three cows, three newly calved cows' given by Elcu to Tudvalch's kin, as recorded in the Surexit memorandum of the Lichfield gospels.

What we think of as myth would have been understood in Medieval Wales simply as one of many aspects of the cyfarwyddyd. As anthropologists and historians now recognise, myths are stories set in the distant past to explain the circumstances of the present: the shape and division of the land, and the relationships between its peoples. These stories may have drawn heavily on supernatural elements, and are punctuated by gestures and actions which seem to belong more to the world of the magical ideation than rational causality. But their concern is as much with law and property claims as with history and religion (again, we must remember that these categories would not have been clearly differentiated in the pre-modern mind). The basic setting was a familiar one – a pastoral landscape, populated by agnatic kindreds who fight and intermarry. Like the protagonists of Ancient Greek myth, the key players described in the Welsh cyfarwyddyd were only too human. They were no less prone to adultery, pride, greed, covetousness, low cunning or momentary stupidity than their latter-day descendants - and much of the understated humour of the Welsh narrative art hangs on these realistic psychological depictions.

The Chwedl: from History to Fable?

Human beings, in all times and places, have enjoyed the experience of listening to stories. Any sequence of events with a defined beginning and end has the potential to be abstracted from the broader matrix of history to be told and re-told as a stand-alone narrative unit. Storytelling has its own set of rules, its own generic conventions. The fact that a story is about to be begin will typically be indicated by certain stereotyped formulae, “Well, let me begin at the beginning…” is one familiar opening device for the modern British anecdote. Among Hausa-speaking peoples in West Africa, the narrator might announce “A story, a story!”, with the audience replying “Let it go, let it come”. In medieval Wales the characteristic opening formula introduces the hero, defining him briefly in terms of lineage, status and geographic provenance (e.g. X son of Y was king of…). The title of the tale is usually found in an equally conventional closing device: “and thus ends the ystoria of X son of Y”, or “this chwedl is called…”

Chwedl (pl. chwedlau) is one of the most common words used by the medieval authors themselves to describe these narrative units in the Mabinogion. The word has a wide semantic field – meaning 'tale' or 'story' on one hand, but also 'history' or 'news' on the other. The distinction between these categories may not have been clearly defined, particularly in the context of traditional material dealing with the remote past.xi The purpose of the chwedl was to inform, to edify, but also to entertain. The Welsh storyteller, like his counterpart elsewhere in the world, used a variety of ruses, structural and stylistic, to prime the imagination the assembled audience. A well-known device of this kind is the so-called 'Law of Threes', the tendency for events to be presented in triplicate series (e.g. the hero meets with three strangers, one after the other), as a way of building tension (the third in the series always being the last and most significant). The stereotypical nature of these devices was mirrored by an equally formulaic quality within the content of the narratives themselves. The heroine was always 'the most beautiful maiden in the world', the hero more often than not an impetuous arriviste. Medieval Welsh narrative also had its own rather more specific conventions. A stag-hunt is usually a prelude to adventure. Giants are typically black-skinned and one-eyed. Noble characters are always clad in 'gold brocaded silk'. The end result was a picturesque but formulaic narrative universe, in which it often feels like a finite number of these motifs are being endlessly repeated and recombined.

Some awareness of this homeostatic quality can be seen in the medieval Welsh catalogue literature known by scholars as the Triads, which (in modern narratological terms) could almost be described as a 'motif index' for native storytelling. The Triads are predicated on the notion of repetition. Historical events or personalities are grouped into threes, defined by type: 'the three Powerful Swineherds of the Island of Britain'; 'the three Fortunate Concealments'; 'the three Chief Courts' etc. These triadic listings are to be found in a number of manuscripts from the thirteenth-century onwards, although there is a case for regarding them as the residue of an oral-mnemonic tradition that probably reached its peak in the previous century, as we will be considering below. Around a hundred such triadic entries have been edited by the great Dr. Rachel Bromwich in her Trioedd Ynys Prydain (1978), from which the following examples have been taken

Triad #21:

Three Battle-Diademed Men of the Island of Britain:

Drystan son of Tallwch.
and Hueil son of Caw,
and Cai son of Cenyr of the Fine Beard

And one of them was diademed above the three of them:
That was Bedwyr son of Bedrawc

Triad #33:

Three Unfortunate Assassinations of the Island of the Britain:

Heidyn son of Enygan, who slew Aneirin of the Flowing Verse, Prince of Poets,
and Llawgad Trwm Bargod Eidyn ('Heavy Battle-Hand of the Border of Eidyn')
and Llofan Llaw Ddifo ('Ll Severing Hand') who slew Urien son of Cynfarch

Triad #35:

Three Levies that departed from this Island, and not one of them came back:

The first went with Elen of the Hosts and Cynan her brother,
The second went Yrp of the Hosts, who came here to ask for assistance in the time of Cadial son of Eryn. And all he asked of each Chief Fortress was twice as many (men) as would come with him to it; and to the first Fortress there came only himself and his servant. (And it proved grievous to have given him that). Nevertheless, that was the most complete that ever went from this Island, and no (man) ever came back. The place where those men remained was on two islands in the Greek sea: those islands are Gals and Avena.
The third levy went with Caswallawn son of Beli, and Gwenwynwyn and Gawnar, sons of Lliaws son of Mwyfre, and Arianrhod daughter of Beli their mother. And those men came from Arllechwedd. They went with Caswallawn their uncle across the sea in pursuit of the men of Caesar. The place where those men are is in Gascony. And the number that went in those Hosts was twenty-one thousand. And those were the Three Silver Hosts: they were so-called because the gold and silver of the Island went with them. And they were picked men.

As can be seen from these examples, there was a generic structure to the Triad, which was repeated with a few minor variations. The triad is introduced in the following way “Three Xs of the Island of Britain” – X can be a type of person, place or category of event. After that, the three exemplars are listed – sometimes with qualifying annotations. As can be seen from the third example, these annotations sometimes expand to the proportions of mini-narratives in their own right. Occasionally, a fourth exemplar is added to the list – as in the case of the first example. In the later tradition, this 'crowning fourth' position is often given to Arthur.

The Triads represent the skeletal remnant what appears to have been an extensive body of oral-narrative tradition current in Wales during the early and central middle ages. Some of these names are known from other sources. The first expedition led by 'Elen of the Hosts and her brother Cynan', for example, is clearly a variant of the Mabinogion tale of Maxen Wledig. Hueil son of Caw is the brother of the cleric Gildas, and is well-known to medieval Welsh hagiography. However, other figures are rather more obscure – Heidyn son of Enygan and his fellow assassins Llawgad and Llofan almost unheard of, aside from a few ambiguous references in the early poetic tradition. However, enough can be seen even from citations of this kind to suggest a vast ocean of narrative material, the majority of which have not survived the passage of time.

The Triads are poised between the worlds of literacy and orality. It has been plausibly suggested their original purpose was to assist a process of oral learning: the apprentice poet having to commit the triads to memory as a means of getting to grips with the narrative cyfarwyddyd, the recitation of which formed an essential part of the bardic curriculum.xii But the very fact that, from the twelfth century onwards, the triads were committed to writing is evidence in itself that this rigorously mnemonic-oral narrative culture was gradually being replaced by a system in which knowledge was retained and transmitted through the medium of the written as well as the spoken word. The increasing appearance of the exegetical notations, qualifying the significance of this or that hero or relaying the circumstances by which they came to earn their position in the Triad, might be seen as further evidence that the knowledge of the hen chwedlau or 'old stories' was not what it had been in previous ages.

Rising levels of literacy in Medieval Welsh society during this time might be seen as both a symptom and a cause of a wider process socio-cultural change. This is not simply a matter of literature enabling an exposure to a wider variety of cultural influence (though this is undoubtedly one of its significant outcomes). Equally important is the change that literacy itself brings to a culture's view of its own past, of history and the passage of time.

In oral cultures, the tendency is to regard time as a cyclic process: defined by the repeating succession of seasons, but also by the stereotypical characteristics of the oral narrative itself, as we have suggested above. As we also saw in the example of the growing number of the sons of Gonja, there is a tendency to assimilate and 'back-project' geo-political change – a tendency that is facilitated by the inherent amnesia of the oral tradition. Thus oral cultures tend towards an eternal present – where the more things change, the more they stay the same. The introduction of writing inevitably fractures this cosy illusion. The very structure of annalistic record – with successive entries one below the last – is merely one of the more concrete manifestations of the process out of which time becomes linear and history is born.

The jagged discontinuities of cultural and political change are revealed in other ways too by the preservation of the ancient written word. Not only do the contradictions in the oral tradition become uncomfortably apparent, the embarrassment of cultural change is also exposed: the anachronism of heroic hyperbole, the muddled credulity of mythic tribal-history. Written records thus reveal a process of permanent and irreversible change – and the ongoing process of cultural evolution. Related to this we find the distinction between fact and fiction, often rather blurred in oral-historical discourse, becomes marked more defined within literary societies.

We can see something of this transition at work in the Mabinogion. There is no doubt that the more 'archaic' native texts such as the Four Branches and Llud a Llefelys would have been regarded as an authentic record of past events, as well as being exercises in narrative entertainment. In the later Arthurian texts, the so-called Three Romances, there may be a nominal claim to historical veracity, but the overwhelming agenda is audience entertainment – storytelling for its own sake. By the time we arrive at the latest Mabinogion text, The Dream of Rhonabwy, we are comfortably within the realms of the self-conscious literary fable. Here we find a full realisation of inter-textual self-consciousness that is the keynote of the literate sensibility, as we see in the ironic attitude towards the bombast of the heroic past displayed in Rhonabwy's encounter with warriors of Arthur.

The Mabinogion tales were the product of a literate culture, albeit one informed by a considerable hinterland of oral tradition. In the earliest of the Mabinogion tales, Culhwch ac Olwen, the traces of this background are most clearly apparent. Not only is it based on the time-honoured scenario of 'The Giant's Daughter' (variants of which are found on every inhabited continent of the world), but this chwedl also includes stylistic features bear the impress of the oral delivery. One thinks of the rich descriptive sequences resembling the crescendos of traditional Welsh areithau (the echo of which can be heard in the chapels and town halls to this day). However, even within this most 'oral' of the Mabinogion narratives, an element of literary self-consciousness appears to be at play. If Joan Radnor's interpretation of Culhwchxiii is indeed correct the author appears to be sending-up the genre, playing on its excesses, as we will consider in more detail elsewhere. Ironic humour of this kind operates partially undercover, appearing on the surface to say one thing, while its true meaning is apparent only to a privileged inner circle of cognoscenti. Such a perspective is wholly characteristic of a literary elite, which probably (in the eleventh century at least) was drawn more or less exclusively from a distinct clerical caste. Read in this light, we can see the target of the parody here is not only the rollicking excesses of the Arthurian topos, but also the (illiterate) lay majority themselves whose enjoyment of such tales was of a rather less sophisticated calibre.

Judging by the range of interests explored in medieval Welsh writing by the late twelfth century, the reading habit seems to have extended beyond the narrow confines of these clerical elites – though what proportion of the lay population were 'readers' of this vernacular literature (rather than just 'listeners') is rather harder to say. In England and on the Continent as well, literacy was also on the rise. But while most of these medieval European readers were taught to read in Church Latin rather than their through their own literary vernaculars, in Wales the opposite seems to have been the case. Most Welsh readers seem to have learned to read and write in their own language, and often remained as 'literary monoglots' – with no direct access to Latin or Old French writing.xiv This had important consequences for the character of the literary culture that was to develop in Wales up until the end of the twelfth century, and would to remain a factor even into the relatively cosmopolitan later middle ages.

The literary language by this stage had stabilised into what we now recognise of classical Middle Welsh, with a well-defined set of syntactical and orthographic principles (which was to endure well into the fourteenth century). Within this, a particular style of narrative prose was to develop: sparce and elegant – one could almost say 'journalistic' – with a slightly different syntactical structure from the natural spoken language. Professor Brynley Roberts describes the texture of this distinctive literary register:

… sentences are made to flow into one another in a logical and harmonious progression. They are frequently linked by conjunctions, or joined phrases, a 'and', sef a wnaeth 'this is what he did', so that the reader is conscious of being moved along by the story. Though coordinate clauses or an elementary pattern of adverbial clause and main clause are the most usual, the syntactic pattern of clauses varies…the word order may be changed from the normal VSO [verb-subject-object] to SVO, OVS…what one is most conscious of is an aesthetic appreciation of the effect of balanced clauses and sentences which contrast with each other. Such smoothness of style is not accidental … it appears to have developed over the years. xv

The Four Branches of the Mabinogi – which epitomised this style of medieval Welsh prose – was a self-consciously 'readerly' work . For instance, the Second Branch at one stage refers to Efnisien as gwr anagneuedus a dywedyssam uchot 'the quarrelsome man whom we referred to above' a clear reference to the literary medium. Its organisation into 'Branches' may reflect the influence of some kind of predecessor of the French prose Romances (although it is not inconceivable that the influence was in the other direction). Either way, there is an orderly, structured quality – including an elaborate system of internal thematic interlace – which one would not expect to find in a more spontaneous oral work.

The Four Branches, along with the early stratum of the Triads and the poetry of the Book of Taliesin represent surviving fragments of what appears to have been an extensive and involved vernacular literature. This literary culture was unusual within the early- and central-medieval European context in that its primary agents were often secular (the cyfarwyddiad and the bardic poets) rather than clerical. A further significant characteristic, as we have seen, was that many of its readers were often literary monoglots, and therefore unexposed to other Latin or European literatures. This state of affairs promoted a tendency towards cultural introspection, as Welsh readers and writers quarried ever more deeply into their own traditions, recombining and interweaving its contents into an intertextual system of astonishing complexity. Something as immersive and intellectually satisfying as the Jewish Talmudic tradition might have evolved out of this vigorous and self-reliant literary culture had certain social and political conditions prevailed. And for a while, during the twelfth century ‘native recovery’, this appeared to be the shape of things to come.

The School of Taliesin

Nowhere do we find this nativist literary culture more characteristically represented than in the pages of the Book of Taliesin. This fourteenth-century codex, consisting entirely of poetic works, appears to have been copied from older manuscript sources, some of which would appear to derive from texts dating as far back as the sixth century AD. As a document it is of particular interest as it is clear that the author of the Four Branches of the Mabinogi was acquainted with a number of the works it contains and evidently regarded them as authoritative sources. Along with the Four Branches themselves, the early group of Triad collections and the Black of Carmarthen, the Book of Taliesin represents one of our most important records of the introspective thoughtworld of the bardic orders.

Combining the roles of poet, propagandist, prophet and historian, the pencerdd or chief bard would have represented an important figure within cultural elite of native Wales, alongside the chief jurist and the tribal abbot. Each of these men represented the summit of one of professional learned classes – whose interests and social ties overlapped to a large extent both with each other and with the wider tribal aristocracy. To a greater or lesser extent, the clergy, the lawmen and the bards had all taken to documenting their key texts in manuscript form during the Early Middle Ages. While the clerical orders were the pioneers of reading and writing in Wales, their efforts were naturally focussed more on the reproduction of key religious and didactic texts, most of which originated from beyond the Brythonic world. Nonetheless, they did produce a number of original works throughout the Early Middle Ages, as well as significant expositions or glosses on existing texts (some written in the Old Welsh vernacular). From the legal classes, the earliest surviving codifications of Welsh tribal law seem to belong to early thirteenth century, but we can imply a tradition of documentation stretching back into the pre-Norman period. For the bardic orders we are slightly better served. As well as the Books of Aneirin and Carmarthen (the former containing the great Gododdin) we have a number of other manuscripts containing works that we might broadly regard as having a pre-Norman ‘bardic’ provenance. The Book of Taliesin appears to have been an attempt to assemble a canonical representation some of the oldest and most paradigmatic expressions from this particular strand of early medieval Welsh literature.

The Book of Taliesin contains an interesting medley of poetic themes – religious, prophetic, eulogistic, historical and mythological. Collectively, this might be seen as spanning the repertoire of one particular group we might refer to for convenience as ‘The School of Taliesin’. Amidst this material, a significant proportion of which seems to date from the tenth and eleventh centuries, we find a much older core, evidently deriving from the same North British context as the archaic Gododdin. It was perhaps around this sixth-century northern material that the later texts accumulated, these older poems evidently being the work of a real-life individual by the name of Taliesin.

According to the biographic profile that his poetry implies,xvi it seems this ‘historical Taliesin’ may have begun his career at the Powysian court in the mid-sixth century, before entering into the service of Urien of Rheged in the North – possibly the most powerful of the British kings of this period.xvii He is remembered by the Historia Brittonum in the following terms:

..Eudeyrn at that time was fighting bravely against the nation of the Angles. Then Talhaearn Tad Awen gained renown in poetry, and Aneirin and Taliesin and Blwchffard and Cian who is called ‘Wheat of Song’ gained renown together at the same time in British poetry.xviii

It is possible that this historical Taliesin did indeed institute some kind of school of bardic training and – rather like the sixth-century founders of the old monastic houses of the British Celtic West – found himself posthumously revered as a figure of mythic proportions, endowed with god-like omniscience and superhuman powers. However, given the lack of contemporary evidence, such a hypothesis must remain within the realms of conjecture. All we know is that at some time in the pre-Norman period the name of ‘Taliesin’ would become associated with sequences of bardic utterances involving the prophesying visions of a shape-changing magician – a protean cult-figure unbounded by space and time. In the Book of Taliesin we find a number of poems written in the first person, from the point of view (so to speak) of this immortal bardic numen. These ‘Taliesinic’ poems, as we will refer to them hereafter, are written in a curiously opaque poetic register which mixes archaic with later forms of Welsh, interspersed with what appear to be bardic neologisms of the poet’s own invention. One is reminded of O’Keefe’s characterisation of ‘magical language’:

Magic speech is not like ordinary speech. It is often extremely odd speech, full of mumbo-jumbo words, archaisms, neologisms and nonsence syllables; it is repetitious, alliterative and full of figures. It may be said in a peculiar tone of voice: chanted or sung or mumbled or sing-songed….Magic speech is neither practical speech of workaday action nor is it the relaxed conversation of human sociability.xix

There is a distinct suspicion that the composers of this Taliesinic verse were indulging in the same kind of obscurantism – a characteristic of esoteric discourse in general, and the language bardic revelation in particular (on this, more below). It is also possible that these already problematic metrical sequences underwent considerable corruption in the process of their textual transmission. It may also be that the linguistic oddities we find in these verses simply reflect the spontaneous idiosyncrasies of the oral-performative sub-culture involved – with little or no attempt to refashion these into a normative 'literary' language. As such, Professor Sarah Higley may be right to describe these works as 'artefact[s] caught between orality and literacy'.xx

The Taliesinic poem, despite its lexical obscurities, tends to follow a fairly well-defined format, of which this excerpt translated by Sir Ifor Williams represents a fairly good example:

A bard here present, I have sung what he will sing,
Let him sing when the wise one has finished;
A lord who refuses me will never afterwards have anything to give…

I know why there is an echo in a hollow;
Why silver gleams; why breath is black; why liver is bloody;
Why a cow has horns; why a woman is affectionate;
Why milk is white; why holly is green;
Why a kid is bearded; why the cow-parsnip is hollow;
Why brine is salt; why ale is bitter;
Why the linnet is green and berries red;
Why a cuckoo complains; why it sings;
I know where the cuckoos of summer are in winter.
How many spears in battle; how many drops in a shower;
Why a river drowned Pharoah’s people;
Why fishes have scales,
Why a white swan has black feet…

I have been a blue salmon,
I have been a dog, a stag, a roebuck on the mountain,
A stock, a spade, an axe in the hand,
A stallion, a bull, a buck,
A grain which grew on a hill,
I was reaped, and placed in an oven,
I fell to the ground when I was being roasted
And a hen swallowed me.
For nine nights I was in her crop.
I have been dead, I have been alive,
I am Taliesin.xxi

These 'aretalogical' sequences frequently begin and end with a formula similar to the opening line of this excerpt. Between those parameters the discourse typically takes the form of a list of "I know.." (gogwn…) and "I was..." (bum..) statements, interspersed with periodic disclosures of the esoteric realms and personalities which are not dissimilar in character to the entranced visions of Inuit or Old English traditional healers discussed above. An example of the latter can be found in the same poem (this time using Sarah Higley’s translation), in which 'Taliesin' (aka Gwion) describes his magical travails:

It is Gwion who speaks,
from the depths it will come (or: “with deep grief”)
It would make from the dying living,
And it is landless (i.e. “foreign”?)
They would make their cauldrons
which boiled without fire.
They would make their webs
for ages upon ages.
Passionately will be brought
poetry out of a poet/prophet.
It is a cruel bondage (or: “hostile alliance”)xxii

From this excerpt, the profound difficulties in handling this material should be apparent, which might excuse or at least explain the reluctance on the part of qualified medievalists to engage with these opaque and virtually untranslatable bardic utterances. The conventional scholarly view was that these declamations were nothing more than histrionic boasting on the part of the bardic poet. On one level at least, this interpretation is perhaps not far from the truth. The poet’s aim was indeed to demonstrate the superiority of his esoteric knowledge – over that of his bardic rivals, and also occasionally over that of other potential challengers such as those from other professions, such as the clerical orders.xxiii This should doubtlessly be seen in the context of the day-to-day struggle for power and influence which typified Medieval Welsh society in general, and the position of the court bard in particular. Verbal aggression was simply part of the bard’s stock-in-trade, and was recognised as such in the well-established convention of the ymryson: a stylised trading of insults, analogues for which can be found across a wide range of cultural contexts.xxiv But in focussing on the more trivial aspects of the Taliesin persona, we risk losing sight of the deeper and in some ways more significant aspects of the bardic function. We may be able to gain some sense of the social dynamics of the chieftain’s court, and catch a glimpse here and there of the recognisably banal motivations of pride and social rivalry; but we should also not understate the considerable difficulties involved in the interpretation of mantic discourse of this kind from a detached modern point of view. It would be a mistake for us to flatter ourselves that we have understood the sum total of what is represented by Taliesinic verse by reducing its significance to ‘nothing more’ than social jockeying and pretentious buffoonery.

There are several salient aspects the Taliesinic declamation which should first be taken on board. The connection between knowledge and experience seen in the passage above (“I was…” “I know…”) is not without significance in itself. The mythical rationale for this supernatural knowledge explained more fully in a late medieval account of Taliesin, which we will summarise below. Something similar may well have been current when the Taliesinic verses were originally performed and transcribed, at some point in the pre-Norman period. Based on the extant late medieval account, the ‘Taliesin myth’ might be summed up as follows:

The sorceress Ceridwen, wife of Tegid Foel, had a son named Morfran who, on account of his great ugliness was known as Afagddu. Ceridwen wished to endow the boy with great knowledge to compensate for his unsightly appearance, so set about gathering ‘certain kinds of the earth’s herbs on certain days and hours’ and boiling these in a magical cauldron. After a year and a day, the drops would be produced from this cauldron containing the essence of all these herbs. On whoever these three drops were to land, knowledge of every art and all things past, future and present would be bestowed.

A blind man was set to stir this cauldron, and a young boy named Gwion Bach was this man’s assistant. After a year and a day, Ceridwen came down to the cauldron with Morfran her son, stationing him close to the cauldron’s edge ready to receive the magical drops. At the crucial time, however, Ceridwen fell asleep. The young boy, Gwion Bach, pushed Morfran out of the way and received the drops for himself. The cauldron uttered a cry and shattered, spilling poison throughout the land.

Gwion Bach, now filled with wisdom, perceives that Ceridwen is now his mortal enemy. He flees in the shape of a hare. Ceridwen, now awake and full of anger pursues him in the shape of a black greyhound. Some stories say that he then transformed himself into the shape of hawk, and she pursued him as an eagle; and so on, through a succession of different forms. Eventually, Taliesin goes inside a barn, where he hides himself as one of the grains of wheat. Ceridwen takes the form of a tufted black hen, pecks up the grains until she swallows Gwion. She carries Gwion in her belly until, nine months later, she gives birth to a baby boy. Not having the heart to cause it any harm herself, she sets the baby adrift in a hide-covered basket. He is eventually discovered by Elphin son of Gwyddno, a noble in the service of Maelgwyn Gwynedd. He is raised by in the household of Elphin, and given the name Taliesin. He later grows up to become a great bard and magicianxxv

So it is his experience of this succession of incarnations – while pursued by the vengeful Ceridwen – that endow Taliesin with his vaunted prodigious omniscience. He understands the secrets of nature, because he has inhabited its multitudinous forms.

While this distinctively Welsh back-story is used to explain the poet’s numerous incarnations, the form of discourse itself represents a tradition that is both older and more universal. A close parallel to Taliesin’s succession of forms can be found in the claims of the mythical Irish poet, Amergin:

I am wind on sea.
I am a storm wave.
I am ocean’s roar.
I am a seven-antlered stag.
I am a hawk on a cliff.
I am a dewdrop.
I am fair body.
I am a boar for valor.
I am a salmon in a pool.
I am a lake on a plain.
I am a word of their poetic art.
I am a word of skill.xxvi

Higley refers to sequences of this kind as 'aretalogies' and notes they are to be found in a surprisingly wide variety of early Indo-European and Middle Eastern hermetic traditions. Clear parallels can be found in Sanskrit, Gnostic and Ancient Egyptian sacred literatures, as well as the vernacular mythologies of Ireland and the Scandinavian North. These aretalogies, invariably declaimed in the first person, offer a dizzying panorama, a kind of 'gods-eye' view of the multiplicity of creation:

These speakers surpass “vision” in their claims and become what they “see”. Sophia, Isis, Visnu, Hari, Odin, WidsiÞ – all speak to us in the form of aretalogies, literally a vaunting of virtue (arÞte) or revelation discourses that reveal the polyvalent nature of their divinity. These discourses take the form of the aforementioned catalogue of different esoteric names, shapes contradictory properties, or impossible places to which the speaker has travelled, and are generally uttered in the first person. Hierarchies and systems are irrelevant; multiplicity is all.xxvii

hat such recitations would have been delivered under the conditions of something like a mediumistic trance (whether real or theatrically simulated) is more likely than not, given the poet’s repeated reference to what he refers to as his awen. This term, which has sometimes been conflated (somewhat inaccurately) with the classical figure of the poetic muse. Deriving from an oblique form of the Indo-European root –*uel (thus cognate with the Modern Welsh awel ‘wind’ and Irish ai ‘inspiration’), awen would appear to designates some kind of impersonal psychic agency associated with the bardic ecstasy. Most strikingly, it gave its name to a class of para-epileptic visionary soothsayers known in Medieval Wales as awenyddion. The Anglo-Norman churchman Gerald of Wales describes these figures in the following terms:

Among the Welsh there are certain individuals called awenydion who behave as if they were possessed by devils… When you consult them about a problem, they immediately go into a trance and lose control of their senses, as if they are possessed. They do not answer the question put to them in any logical way. Words stream forth from their mouths, incoherently and apparently meaningless and without sense at all, but all the same well expressed [ornatus]: and if you listen carefully to what they say you will receive the solution to your problem. When it is all over, they will recover from their trance, as if they were ordinary people waking from a heavy sleep, but you will have to give them a good shake before they regain control of themselves… when they do come round they can remember nothing of what was said in the interval. They seem to receive this gift of divination through visions which they see in their dreams. Some of them have the impression that honey or sugary milk is being smeared on their mouths … It is rather like what Esdras wrote about himself: ‘Behold, a voice called me, saying, Esdras, open they mouth. Then I opened my mouth, and, behold, he reached me a full cup, which was full as if it were with water, but the colour of it was like fire. And when I had drunk of it, my heart uttered understanding, and wisdom grew in my breast’. When going into a trance they invoke the true living God, and the Holy Trinity, and they pray that they may not be prevented by their sins from revealing the truth. xxviii

We are reminded of the words of another bard in Geoffrey’s Vita Merlini, when he describes how he acquired his paragnostic powers through a dislocation of his everyday consciousness:

I was taken out of my true self, I was as a spirit and knew the history of people long past and could foretell the future. I knew then the secrets of nature, bird flight, star wanderings and the way fish glidexxix

We do not know how widespread was the practice of the awenyddion, but it does seem clear that such methods were strongly associated with the particular tradition we might refer to as the School of Taliesin. This is significant for us – given the close intertextual links between the Book of Taliesin, the early Triadic collections and the ‘native’ Mabinogion texts such as the Four Branches and Llud a Llefelys. The names of most of the characters in the Four Branches: Gwydion, Math, Pryderi or Manawydan, for example, are well-attested in the Book of Taliesin and early Triads – but are rarely found beyond these sources. Such names did appear occasionally in the allusions of the court poets (who may or may not have been the same people who declaimed Taliesinic verse) but as Bromwich points out, such allusions become more unusual as we enter the later medieval period. The cultural changes implied by this fading out of the native tradition will be discussed in the following section.

Before we do so, it might be worth attempting to summarise the intellectual world of the bardic schools, as they seem to have existed in Wales in early and central Middle Ages. They represented a prehistoric tradition: bardic poets had been attested among the sixth-century Brythonic kingdoms of the sub-Roman North (the home of Aneirin and the 'historical' Taliesin); as well as among the Iron Age Gallo-Brittonic tribes who confronted the armies of Caesar. One suspects that bardic poets may would have chanted before chariot-riding chieftains of Hallstatt and La Tene. The main bardic stock-in-trade was the court eulogy – praise poems to flatter and glorify their patrons from the warrior aristocracy. (The earliest surviving examples include the lament to the fallen heroes of the Gododdin, as well as a dozen or so praise poems addressed to various sixth-century warlords by the 'historical' Taliesin. These works were imitated in their style and metrical form by the gogynfeirdd court poets in the twelfth and thirteenth centuries). But as well as these stylised propaganda pieces, there were other dimensions to the bardic repertoire which relate significantly to the contents of the Mabinogion. A knowledge of the cyfarwyddyd (historical lore) was probably among the more important secondary functions of the bardic poet: we have the well-known example in the Fourth Branch of Gwydion (in bardic disguise) reciting the cyfarwyddyd at the court of Pryderi, as seems to have been the custom. And beyond this there were other rather more alien aspect to the bardic persona – that of the ecstatic prophet or the frenzied shaman-seer. It is this quality that separates the world of the old bardic schools and their intellectual culture so markedly from our own. And while the hen chwedlau such as the Four Branches of the Mabinogi tend to be written in clear and concise Middle Welsh prose, we cannot lose sight of the fact that it was written by people who regarded (what we would consider to be) the near-gibberish of the Taliesinic verse as authorative revelation. This was a culture within which it was accepted that everyday consciousness was periodically interrupted by profound psychic ructions – embodied and enacted by the entranced bard, but partaken of by the wider community. These are the hallmarks of the 'magical' reality, which in many respects the early Mabinogion texts signal the closing phase.

The Church might be regarded as the midwife of the modern mind, as much as she was the heir to the terrors and ecstasies of the prehistoric experience. Within the intellectual culture of the Middle Ages, the Church stood centre stage, and her influence can even be detected even in a heterodox subculture such as the School of Taliesin. As Marged Haycock (a leading authority on the Book of Taliesin) points out, the influence of the basic monastic education is apparent even in the most obscure aretalogies within this manuscript. The poet’s self-proclaimed expertise in cosmology, computistics and biblical lore may reflect the vestiges of the liberal education taught by monks to the local lay population, or even the early stages of a monastic training undergone by the bardic initiates themselves. The rhetorical sequences in Taliesinic verse ("why silver gleams…why breath is black...why liver is bloody.." etc.) bear a noticeable resemblance to question-and-answer sequences found in widely circulated Medieval Latin textbooks (including the dialogues of Solomon and Saturn, Ioca Monachorum et al.). The influence of other early medieval classics is often apparent in the Book of Taliesin: including Bede’s De Rerum Natura, Isidore’s Etymologica, the histories of Orosius and various well known works of exegesis and biblical aprocrypha. The Taliesinic mind carried within it cargo from these sources cheek-by-jowl with the inherited freight of the native cyfarwyddyd – and both were prone to resurface spontaneously during the altered state of consciousness known in medieval Wales as awen or the bardic trance.

The precise nature of the relationship between the bardic schools and the clerical orders remains somewhat unclear, but perhaps it is less important than some of the other points of opposition within the wider social context. It makes sense to think of the cultural life in Wales as existing on a series of uneven and often overlapping polarities: north vs. south, secular vs. clerical, and (most significantly here) the internationalist vs. nativist orientation. The ‘internationalist’ faction tended to be concentrated in and around the Anglo-Norman Marches – and might be best represented by churchmen such as Gerald of Wales, Walter Map or the Powysian princeling Owain Cyfeiliog (a regular guest of the local Anglo-Norman aristocracy). Culturally, the internationalists would have been more receptive to the fashions (literary and otherwise) derived from Continental and classical sources; politically, they were aligned to the interests of the Anglo-Norman francigenae; in terms of religion, their affiliations were with orthodox Latin Christianity. The nativist faction, on the other hand, were more inward-looking. In religious terms, this group were the ‘old believers’ – adherents of a Celtic Christianity based in the old pre-Norman monastic houses, with close links (religious, literary and political) over the Irish sea. Politically, they would have been represented by the die-hard anti-Norman warlords such as Gruffydd ap Rhys and Owain Gwynedd, about whom we shall say more below. Culturally, we can see this element best represented by works such as the Four Branches of the Mabinogi, the earlier Triadic collections and, perhaps most distinctly, by the poetry in the Book of Taliesin. The heartlands of the nativist faction were in the tribal territories of Gwynedd and the Cardigan Bay coastlands (the focal areas of the Mabinogi).

In the mid-twelfth century, under the leadership of Owain Gwynedd in particular, the nativist party seems to have been in the ascendant, and this political backdrop may well represent the circumstances in which projects such as the Four Branches, or an early version of the Book of Taliesin – might have first been formally commissioned. By the middle of the thirteenth century, the focus of interest seems to have shifted to foreign romances and histories of Anglo-Norman origin. This shift from the nativist to the internationalist outlook in Welsh culture thus had an important political context, which goes some way to explaining the character of the later texts of the Mabinogion. To understand this period of change within the cultural and the political aspects of Medieval Welsh life, we need to focus on the five or six generations between 1150 and the end of Welsh independence in 1282 – the period in which the bulk of the texts of the Mabinogi were almost certainly composed. This is what we will be considering in the following section.

Continental influences in the Age of the Llywelyns

The Triads, those curious bardic catalogues of narrative lore that we considered above, present an interesting barometer of the changing literary interests of medieval Welsh readers over the course of the twelfth and thirteenth centuries. The earliest collections of Triads are populated almost exclusively heroes and villains of a native pedigree. The protagonists listed include heroes and villains from the Old North milieu, as well as various other regional traditions from the wider Brythonic world (including a lost southern Welsh/Cornish/Breton cycle, involving figures such as Tegau Euvron and Caradawg Vreichvras, some of whose fragmentary legends resurface – rather unexpectedly – in the thirteenth-century French prose romances). Mythical Gallo-Brittonic figures such as Mabon, Caswallon and Lleu Law Gyffes also feature significantly in this early stratum of Triadic literature. In the later Triadic listings, however, we find an increasing prevalence of names that owe their origin to romances of foreign derivation.

In this respect, the Triadic tradition follows the general trend in medieval Welsh letters from the thirteenth century onwards: towards a more cosmopolitan literary diet, and away from the inward-looking perspective of the local cyfarwyddyd. Bromwich notes that the mythological allusions in the later court poetry, like the later Triads, suggest a declining knowledge of the oral cyfarwyddyd:

Throughout the thirteenth and fourteenth centuries it becomes increasingly apparent that the bards are drawing on literary sources for their allusions; and the introduction of this material of predominantly foreign origin coincides with the evident decline in knowledge of the native Welsh tales which, as has been shown, is reflected in successive manuscripts of

So what was it that occasioned this change? What caused the Welsh to move away from the self-sufficient nativism reflected in the Book of Taliesin, the earlier Triadic tradition and the Four Branches of the Mabinogi? Did this reflect a wider social change, and if so, what were its causes and its connotations? This question is relevant to us here, as it goes some way to explaining and qualifying the character of some of the later Mabinogion texts – the so-called Three Romances, and the breuddwydion (dream visions) of Maxen and Rhonabwy. The change that took place in thirteenth century Wales runs like a fault-line through the Mabinogion collection, but we need to understand what lies on both sides of this rift to fully appreciate Medieval Welsh narrative prose. To do this we need to understand the so-called Age of the Llywelyns – which began at the start of the thirteenth century with the reign of Llywelyn ap Iorwerth (later known as ‘Llywelyn the Great’) and ended with the death of his grandson Llywelyn ap Gruffydd in 1282 – an event which was shortly followed by the conquest of Wales by Edward I.

To comprehend the dynamics of this formative period, we must first consider the geo-political dichotomies of the previous generation. The second half of the twelfth century in Wales was dominated by two powerful and distinctive regional leaders. The first of these was Owain ap Gruffydd (d.1170), more generally known as ‘Owain Gwynedd’ after the northern Welsh territory that he controlled with such brutal efficiency over the decades of the mid-twelfth century. The second was the prince Rhys ap Gruffydd (d. 1197), more often known simply as ‘Lord Rhys’ – overlord of the southern Welsh lordships and chiefdoms known collectively as Deheubarth – who rose to prominence in the 1150s, and died towards the end of the century in 1197. In terms of the nativist-internationalist spectrum outlined above, it would possible place the northern prince at the radical wing of the nativist faction. Rhys, on the other hand, might be characterised if not as an outright internationalist, then certainly as a centrist-moderate. While he had played an active role in the anti-Norman resistance early in his career in the 1150s, in the later part of twelfth century the so-called Lord Proprietor of the South would adopt a more concessionary internationalist outlook, cultivating favour with the English king.

These two native warlords presented two contrasting geo-political paradigms. Both achieved notable successes during their lifetimes, but neither secured long-term outcomes that were entirely satisfactory for themselves or their native Welsh subjects. It was by reflecting on the careers of these two Welsh princes that the man who would become known as Llywelyn the Great (1172-1240) seems to have developed a set of regional and constitutional strategies which guided the principality through the first three quarters of the thirteenth century, until the death of his grandson ‘Llywelyn the Last’ in 1282.

Llywelyn the Great’s grandfather Owain Gwynedd had emerged from the shadow of his wily and long-lived father, Gruffydd ap Cynan (d. 1136), whose own experiences had been shaped by the collapse of native power following the arrival of the invading Norman warlords at the end of the eleventh century. Owain’s early career saw a vigorous reassertion of native power under the leadership of Owain himself and his brothers in the North, and the dynasty of Gruffydd ap Rhys in the South. What had begun as a guerrilla war of resistance became a fully-fledged rout of the Norman presence in North and West Wales during the late 1130s. Although significant Norman enclaves were to endure in Pembrokeshire and throughout the southern and eastern borders, the Marchers (as they were known) were hereafter on the back foot. England was embroiled in a civil war during 1140s and 1150s, between the factions of Stephen and Matilda, and this circumstance gave the Welsh a freer hand. Around this time an axis of native power was constituted between the Owain’s House of Aberffraw in Gwynedd and the north, and Lord Rhys’ House of Dinefwr in the southwest. It was this informal federation which was to successfully resist the onslaught of Henry II, when the end of the civil war finally gave the English king the chance to attempt to re-assert Anglo-Norman power in the native Welsh heartlands. Their victory against the Angevin king in 1165 confirmed this state of affairs, and led to a state of effective autonomy for what became known as Pura Walia – the native territories of North and West Wales. This was the geopolitical context within which the Four Branches of the Mabinogi took their final shape.

Despite the temporary stalling of the Anglo-Norman military machine, the native Welsh could ill-afford to ignore the powerful kingdom beyond their eastern borders, and the threat to their cultural and political independence presented by the expansive Angevin federation. The two native princes differed significantly in their response to the challenge of Anglo-Norman hegemony. Llywelyn’s grandfather Owain Gwynedd was to adopt a posture of aggressive defiance. The mountain fastness of northwestern Wales remains to this day the most uncompromisingly nationalistic corner of the principality: Owain Gwynedd’s attitude towards English authority remained true to the grain of this distinctive region. The Snowdonian prince’s radical nativism was expressed through culture as much as through military resistance. Not unlike Maelgwyn Gwynedd or Vortiporix (Dark Age warlords described in the diatribes of the sixth-century Gildas) Owain kept numerous concubines, a number of whom were drawn from his close kindred. When the Canterbury Archbishop, Thomas Beckett, challenged the legitimacy of his liaison with first cousin Cristin, Owain responded by expelling the Anglo-Norman Bishop of Bangor and replacing him with a native, Irish-trained candidate. This was a significant move, indicating a defiance not only of the authority of Canterbury, but also of the religious and cultural value-system of French-speaking Christendom as a whole. This return to a neo-Celtic ecclesiastical tradition was of a piece with his consanguineous marriage to his cousin Cristin (who, significantly, was also the sister of his southern ally and kinsman Lord Rhys of Deheubarth). Both actions represented an aggressively localist agenda; a retreat into a native cultural and religious practices; a tribalist favouring of kith and kin over the possibilities of integration within wider trans-regional hegemonies.

Lord Rhys of Deheubarth himself, as we have suggested, was more compromising than his northern brother-in-law in his relations with the English king. Despite joining forces with Owain Gwynedd in the anti-Norman campaigns of the 1150s and 1160s, he showed an increasing willingness to make terms with the English king in latter part of his reign. Unbidden, he offered logistical and operational support to the Angevin invasion of Ireland in 1177. He voluntarily attended sessions at the English court. He accepted the titles of ‘Chief Justice’ and ‘Lord Proprietor of South Wales’ – feudal offices which implied dependence on the English crown. In formal political terms, his status has much in common with that of Pryderi to Caswallon in the Third Branch: a provincial ruler with acknowledged ties with the London-based regional overlord.

At times, Lord Rhys’s enthusiastic fealty attracted derision even from Anglo-Norman observers, as well as within the native community itself. But for all this, he remained the twelfth century’s most successful native ruler, and came as close as anyone in his generation to living up to the title of Rex Brittonum ‘King of the Britons’, the rather fanciful soubriquet that was often used by the native chroniclers in their death notices of powerful Welsh warlords. With his diplomatic relations with England, his dominance of the lesser Welsh chieftains, and his sponsorship of native cultural and ecclesiastical projects, the pragmatic statecraft of this grizzled southern king must have given some food for thought to the young Llywelyn ap Iorwerth, who was coming of age in 1180s.

Llywelyn, in his relations with the English king, seems to have avoided either the sterile isolationism of Owain Gwynedd or the supine appeasement of Lord Rhys. The man who would be remembered as Llywelyn the Great owed his historical reputation to geo-political success borne of strategic adaptability. His political goal was, more than anything else, for an autonomous principality of Wales centred on the lordship of Gwynedd. In pursuit of this aim, there would be times when he would use force against the Marcher lords and the English crown (the latter weakened by baronial rebellion during the reign of King John). But equally, when it was in his advantage to do so, he was happy to cooperate with his neighbours east of the border. In noticeable contrast to his grandfather’s endogamous tribalism, Llywelyn sealed his diplomatic relations with the English crown through a high-profile marriage to Joan, the daughter of the English king. This lead was followed by other members of the Welsh royal tribe, many of whom formed their own marital alliances with powerful Marcher families from the southern and eastern borders: the Mortimers and the De Breoses. Out of these unions, it might be noted, sprang some of the most powerful dynasties of late medieval Britain: the Houses of Lancaster, York, Stuart and Tudor.

Intermarriage with the French-speaking aristocracy may have been one of the instrumental factors in the change we can observe in the political and cultural life of Wales that took place during the thirteenth century. This is apparent on a number of levels. Most strikingly, the native Welsh began building their own fortified castles of mortared stone, similar to those that housed the great Marcher families into which they intermarried. During this time too, it has been suggested that Wales began to assume the characteristics of a unified feudal state. Before the treaty of Aberdyfi in 1216, the relationship between Llywelyn and the other Welsh chieftains was one of a provisional ‘first among equals’. An important change brought about by Llywelyn (which went against the grain of the traditional native political system) was to redefine this loose hegemony into a permanent feudal hierarchy, with the lords of Deheubarth, Powys and chiefdoms of the South now in a state of perpetual vassalage to the ‘Princes of Snowdonia’. In imitation of the monarchies of England and France, Llywelyn also developed his own government bureaucracy, with offices of state issuing sealed writs on behalf of the king. Monetary currency increasingly played a role in Welsh government and society – with taxes in kind, the ‘food rent’ or gwestfa, increasing being replaced by the twnc, or the monetary payment.

Of particular interest here is the fact that the assimilation into the European mainstream can also be seen in the literary culture of thirteenth-century Wales. During the Age of the Llywelyns, considerable effort seems to have been made to translate acclaimed foreign works into the Welsh vernacular. Among these were the famous tales of Charlemagne (often known as the Matière de France, or the Chanson de Geste). Another popular subject for translation or adaptation was the French Arthurian Romances, including those of Chrétien de Troyes, which seem to have resulted in the so-called ‘Three Romances’ (the precise nature of this relationship is something we will be considering in more detail elsewhere). As well as this, we find other medieval European classics such as The Seven Sages of Rome or the philosophical/geographical treatise known as Imago Mundi. A particular favourite seems to have been the Historia Daretis, a narrative account of the fall of Troy. Standard religious works also seem to have been a popular subject for rendition into the local vernacular: a translation of the Athanasian creed (the liturgical staple of medieval Christendom) was commissioned by Efa daughter of Maredudd (the great-grandson of Lord Rhys) in the mid-thirteenth century. Meanwhile her brother Gruffydd was the dedicatee for the translations of Transitus Mariae, a popular piece of medieval apocrypha celebrating the assumption of the Virgin Mary. Both of these texts appear, as we will see alongside the Mabinogion tales in the White Book of Rhydderch and the Red Book of Hergest.

Yet there was more to this avid translation of the classics of European literature than a provincial aping of metropolitan taste. The rendition of these staples of Medieval Christendom in the Welsh language carried political as well as cultural significance. Welsh was being presented as a language of power: one through which the literature of the feudal and ecclesiastical elite might be confidently articulated. This was part and parcel of the same process that saw Wales asserting its position as an autonomous natio among the kingdoms of Europe, a goal that came close to fulfilment during the Age of the Llywelyns. Unlike some provincial elites, Wales did not abandon her vernacular tradition. Instead, this translation of the European canon should be regarded as an assertion rather than a dilution of native cultural identity. It should be remembered that the old native texts continued to be copied recopied, included within the manuscript collections alongside these translations of the European classics. We might conclude that the native Welsh texts were beings presented as equivalent in value to the Continental works – a national literature of comparable stature to the chansons of Francophone Europe, or the gesta of the Ancient Roman world.

So this prolific translation project of the thirteenth century might best be understood as the cultural wing of a broader political and diplomatic movement that was underway in Wales in the thirteenth century. Having rejected the model of the defiant Celtic isolationism (the posture adopted by Llywelyn’s grandfather Owain Gwynedd), Wales was now looking to be taken seriously on European stage, as a sophisticated medieval nation-state with its own literature, architecture and royal bureaucracy.

However, this embrace of the international mainstream was not without its consequences for the local vernacular culture. Certain literary works from the native canon were preserved (notably the Four Branches and the pre-Norman Arthurian Culhwch ac Olwen), but it is also depressingly evident from the mythological allusions in the court poetry from the thirteenth century onwards that the bardic eulogists were losing touch with their home-grown traditions. Instead, they were drawing on more exotic, literary sources, and increasingly focussing on the heroes of Continental Romance – figures that were also found in ever larger numbers in the triadic listings in the later period.

Geoffrey’s Historia Brittonum, which had made so little impression on the inward-looking culture of twelfth-century Wales, was to become one of the most popular subjects for translation in the Age of the Llywelyns. Indeed, such was the popularity of the Brut Y Brenhinedd (the Welsh-language version of Geoffrey’s chronicle) that by the later Middle Ages it would entirely displace native historiography. We might well mourn the loss of the distinctive Brythonic world-view found within the early triadic collections and texts such as the Four Branches of the Mabinogi. The "off-the-shelf" history of Geoffrey of Monmouth, while loosely derived from native Brythonic material, was in large part the handiwork of the Anglo-Norman cleric himself, and informed by an essentially Latinate rather than Celtic historical vision. Yet it was the chronicles of Geoffrey, rather than those of the anonymous cyfarwyddiad, that would become accepted in Wales as the authoritative early history of the British peoples right up until the late-eighteenth century. The complex genealogical-historical traditions of the Old North, the Sons of Beli Mawr, the intimations of deep tribal history within the Mabinogi: all of this would remain virtually forgotten until their antiquarian rediscovery in the early nineteenth century. Only now are that these fragments of native traditions have once again seen the light of day has the process of reassembling their complex superstructure has begun.

However, while regretting the displacement of these ‘genuine’ narrative-historical traditions, we must not make the all-too-common mistake of devaluing what it was that took its place during the Age of Llywelyns. In the Mabinogion collection, the Three Romances typify this era, and one can at times come close to agreeing with the great Welsh playwright and political activist Saunders Lewis that the thirteenth-century Arthurian tales represent some of the finest narrative prose ever articulated in the Welsh language, modern or medieval. While some degree of French influence on these texts is certainly present, this exotic leaven was to fuse potently with the vernacular narrative style, creating a new and distinctive genre of the ‘Welsh Romance’. The hybrid narrative form that resulted retained the concise vividness of Middle Welsh storytelling, but illuminated with this with the tints and highlights of sensuous fantasy, reflecting the influence of Continental Romance. The narrative vision conveyed by the Welsh Romances would go on to inspire the mood and atmosphere of a broader range of late medieval Welsh literary expressions, including the luscious trilling of the cywyd or uchelwyr poets, the most notable of whom was the great Dafydd ap Gwilym. A flavour of the late medieval literary spirit can be found in the famous love poem from the works of this poet, which is worth quoting in full – illustrating as it does how far things had changed by the mid fourteenth century:

The Window (Y Fenestyr)

I walked within leafy enclosures
(my muttering was a frivolous song)
by the side (wild tangled lands),
as I presumed, of the girl's bedchamber.
I was glad to discover (fair valiant maid)
through the grove's branches for a girl's sake
(a powerful love, fierce thief)
a sturdy window on a piece of oak

I sought a kiss (fairest of form)
from the girl through the little oak window,
the fair jewel — it was wrong of her —
refused me, she did not want me;
troublesome was that window of enduring grief,
where it was placed to let in the sunlight.

May I not grow old if there was ever, by way of enchantment,
a window such as this,
apart from the nature of that window (a couple whose predicament was astonishing)
in the fort of Caerllion long ago
through which Melwas, impelled by desire,
came with none of love's trepidations
(extreme pain of boundless passion)
once by the house of Giant Gogfran's daughter.

Although I could stay a while, when it was snowing,
on the wrong side of the window from her,
unlike Melwas I received no reward,
my only favour, by God, was the wasting of my cheeks.
If we were, I and my seamstress, fair jewel,
face to face for nine nights,
with no generous reward, no starlight,
no gain between the two pillars,
as the distress grew ever greater on each side of the whitewashed wall,
lips to lips, I and my proud slender maid,
we could not (golden jewel's courage)
get our two mouths to meet.

At no time are two mouths
through a narrow–pillared wooden window
(my grim death [it is] to be prevented from [realizing my] promise)
able to kiss since it's so confined.
No one by a window
at night between fennel and a row of roses,
without sleep, has ever known such care as I,
in no joyous mood because of a bright pious girl.

May a devil — that lair of a window —
break its pillars with a blunt tool
(sharp edge of wrath), and its broad shutter,
and its lock and key entirely,
and the man (rule of obstruction)
who made such a row of frustrating pillars;
may he slay the bright one which hinders my effort,
and the hand that sawed it,
slay the wicked one which hinders my union,
it impeded me there where the girl was.xxxi

We have come a long way from the archaic poetry of the Gododdin and the Taliesinic aretalogies. In many respects, Dafydd is closer to the modern mind than he was to that of the gogynfeirdd court poets just two generations before his time. The balance between old and new, foreign and native influences in this poem is perhaps what characterises this late-medieval Welsh literary impulse, which flowered most gloriously in the works of Dafydd ap Gwilym but which had been a keynote since the Age of the Llywelyns. The central motif of the lover’s window recalls the classic mis-en-scene of the Provençal troubadour lyric – an important current within the court literature of the central and later middle ages. Yet alongside this we find reference to the traditional Arthurian scenario of the seduction of the Queen Guinevere by the faery king Melwas. Precisely the same blend of courtly literature and native tradition was anticipated, in prose, in the later Mabinogion tales of the thirteenth century.

The infusion of cosmopolitan exoticism with vernacular self-confidence that we find in the poetry of Dafydd ap Gwilym was at work elsewhere within western Europe in the fourteenth century. It was a signal quality in the works of Boccacio, Dante, Chaucer, Deschamps, which also revealed itself in various anonymous works from the same period including the semi-fantastical Travels of Sir John Mandeville (here, the influence of Celtic voyage tales and Anglo-Norman Romance is noticeably evident). These authors stood on the threshold of the Renaissance, and in various important ways might be seen as anticipating the paradigm-shifting genius of Shakespeare, Rabelais and Cervantes. (Looking beyond the world of letters, we might remember that the fifteenth-century pioneers of modernity – men such as Copernicus, Machiavelli, Luther, and Columbus – would have grown up in this late-medieval penumbra.) The important point here is that the rise of secular literacy, which was such a vital characteristic of this fourteenth-century culture, was already well-established in thirteenth-century Wales. In this respect, the thought-world reflected in the Mabinogion was not only a relic of the past but also a harbinger of the future, anticipating developments which would become more general in the centuries to follow.

The Red Book of Hergest and the White Book of Rhyderch

Having completed this survey of Medieval Welsh literature, we are now in a position to make more sense of the manuscript context of the works in question. As we have established, the Mabinogion was a collection: not the work of a single author, but eleven separate texts written over a period of two hundred years, forming part of what we might think of as an ongoing literary conversation. Copies of all the Mabinogion texts were originally selected from the Red Book of Hergest, nowadays housed in Oxford’s Bodleian library. Ten of the eleven texts also appear within a slightly earlier codex known as the White Book of Rhydderch, which now forms part of the Peniarth collection of medieval Welsh manuscripts at the National Library of Wales. Both of these volumes are intriguing documents in their own right, and (in the case of the Red Book in particular) might be seen attempts by their late medieval compilers to represent a cross section of the national literature. Alongside the Mabinogion tales we find scores of other works, covering an astonishingly wide range of provenance and subject matter, including many translated works deriving from French or Latin sources as well as those of native inspiration origins. Later on in this section, we will be considering not only the eclectic content of these ‘one volume libraries’, but also their ordering and structure – how the texts involved were assimilated and categorised by their medieval readers. In this way, we hope to gather some sense the place of the Mabinogion texts within the bibliographic universe of their late medieval readers.

Individual Mabinogion texts, or fragments thereof, are also to be found in a number of thirteenth-century manuscripts. These documents represent an important body of evidence. On the one hand they shed light on the 'pre-White Book' manuscript tradition of the Mabinogion texts themselves. Variations between their different versions offer some valuable clues on problems such as the editing and transmission of Middle Welsh texts. In a more general sense, these palaeographic fragments are also significant as cultural artefacts of the Age of Llywelyns. As such, they give us some sense of the source material out of which the Red and White books would have been assembled, and go some way to qualifying in various ways the diverse content of the great books of the later Middle Ages.

We will start with a manuscript that is little more than a gathering of fragments, known by its bibliographical listing as 'Peniarth 6'. These fragments include two individual pages, which evidently once contained the Mabinogi tales of Branwen and Manawydan respectively. Along with these, we have a couple of individual pages and a folio from the romance of Geraint. The Mabinogion pages have been dated to the second half of the thirteenth century (c.1250-1300), while the copy of Geraint may be dateable to the early 1300s.xxxii Despite its composite nature, Peniarth 6 has been characterised as a southern manuscript on the basis of its distinctive orthographic features.xxxiii While the internal evidence of the Four Branches itself strongly suggests a northern provenancexxxiv , it is interesting to find this trace of a southern manuscript tradition. But it is also remarkable that despite the archaic character of this southern orthography xxxv the text of these Mabinogi fragments remains remarkably close to the equivilant sections found in the Red and White books. These fragments, then, raise some interesting questions about the transmission of the Mabinogi prior to its appearance in the White Book. We might tentatively conclude that its journey from Gwynedd in the late-twelfth century into the Rhydderch’s fourteenth-century collection via a southern scriptorium nonetheless took place in relatively few stages, perhaps involving a sequence of little more than three or four recopyings. This would account for the relative consistency of the textual content of these extant manuscript survivals, as well as their occasional traces of southern orthography.

A slightly different picture is presented by the two versions of the Welsh Romance of Peredur that are found in two thirteenth-century manuscripts, Peniarth 7 and Peniarth 14. The significant variations that exist between these two recensions, and between these and their counterparts in the Red and White books, suggest a rather more prolific and fluid manuscript tradition. No one of these documents would appear to be the direct source of the other: suggesting that Peredur was a popular and widely-copied text in the thirteenth and fourteenth centuries. Moreover, the differences between these various recensions are not only confined to minor details of spelling and orthography, but also in phraseology, content and even plot structure. As well as presenting a considerable headache for modern editors, who have struggled in vain to reconstruct the ‘authorative’ urtext, this prodigous variation also suggests a creative aspect to the scribal culture in medieval Wales, in which copiers felt at liberty to amend or even reinterpret the text, much as an oral storyteller might offer his own, idiosyncratic retellings of a traditional popular tale. While some texts seem to have become ‘fixed’ (and this might explain the relatively stable character of the Mabinogi) it is clear that rather more fluid copying practices also prevailed. (We find something similar in the manuscript tradition of Medieval Ireland, the widely varient recensions of the Tain Bó Cuailgne being a characteristic example.)

Also significant is the range and composition of the other texts found alongside the early Peredur recensions in these thirteenth-century manuscripts. Peniarth 7 contains a number of the Welsh Charlemagne stories as well as ‘The Story of Adam’ and various other translations of popular medieval hagiography. Hagiography and apocrypha also predominates in Peniarth 14 (which also dates from the second half of the thirteenth century), in which a fragment of Peredur is also to be found. A text of Geoffrey of Monmouth’s Prophesies of Merlin is also included in this manuscript, as is a copy of the Welsh translation of Transitus Mariae, which had been commissioned by Gruffydd ap Maredudd (Lord Rhys’ great grandson) around the middle of the thirteenth century. Such texts may be said to typify the medieval Welsh library from the Age of the Llywelyns into the late medieval post-conquest era. Many of the same texts are to be found in both the Red Book of Hergest and the White Book of Rhydderch, as we shall see.

Our final example of a pre-White Book edition of a Mabinogion tale an early version of Breudwyt Macsen ‘The Dream of Macsen Wledig’ which is to be found in Peniarth 16. Like Peniarth 6, this manuscript is in reality an assemblage of previously unrelated fragments of varying age. The section containing the Breudwyt is known as Peniarth 16iii, and is sometimes thought to have been copied by one of the scribes whose hand is found in part of Peniarth 14.xxxvi If this theory is correct, it would seem that this scribe had a particular interest in the Prophesies of Merlin – Peniarth 16iii also contains a Welsh translation of an early thirteenth-century Latin commentary on the Prophetiae. This text sums up as much as any other the paradoxes of the literary culture of thirteenth century Wales: which retained a lively interest in its native tradition, but often preferred to access this through interpretations that derive from Anglo-Norman and Continental authorities (such as the Romances of Chrétien de Troyes or the Historia of Geoffrey of Monmouth). These French or Anglo-Norman Arthurian texts were then retranslated back into the Middle Welsh language and literary style, often garnished with names and other details quarried from the ‘authentic’ native tradition. It was this curious process of cultural re-importation that seems to underlie the phenomenon of the Three Romances (Peredur, Owain and Geraint), as we will discuss more fully elsewhere.

It was from these fragments, or others like them, that the White Book of Rhydderch would have been assembled in the mid-fourteenth century. Rhydderch ap Ieuan (d.1399), after whom the book is named, was the great-great grandson of Maredudd ap Owain, chief of Ceredigion (d.1265), who was himself the grandson of Lord Rhys (d.1197). Rhydderch thus came from what might loosely be described as the ‘royal tribe’ of Wales, and his particular segment of this sprawling progeny seems to have had a long-standing involvement with the literary arts. Rhydderch – apparently some sort of legal official in late thirteenth-century Wales – was also the patron of a number contemporary poets, being described by one as ‘the pope of word-love’. Rhydderch’s own parents themselves were friends and supporters of a number of early fourteenth-century poets, including Dafydd ap Gwilym. His great-grandfather Gruffudd ap Maredudd (son of the aforementioned Cardiganshire chieftain) was the patron of some important of translation projects during the mid- to late-thirteenth century. Among these were Welsh renditions of Turpin’s Carolignian Historia and the Transitus Mariae, a popular piece of medieval apocrypha, both of which were to achieve an well-established position within the literature of Medieval Wales. Gruffudd’s sister Efa herself commissioned the translation of the Anathasian creed (rendered into Welsh as Credo Anathasius Sant). Her great-grandson (a distant of Rhydderch himself) was to continue this tradition, sponsoring the famous ‘Anchorite of Llanddewibefi’, the author of a substantial volume of religious writing produced in the mid-fourteenth century.

Much of this material was to appear in the White Book of Rhydderch, suggesting that the manuscript was to some degree compiled from the domestic archives of this prolific literary family. However, it is also possible if not probable that much of the scribal work involved in the physical production of the White Book would have taken place at the nearby monastery of Strata Floridaxxxvii , with which the family had close connections. It might be possible to see this important Cistercian abbey – founded by none other than Lord Rhys himself – as another gathering point of literary materials which would have found their way into White Book. The abbot of Strata Florida from 1344 to 1380 was Llywelyn Fychan, a noted patron of the arts. It may well have been under his management (implementing the directives of Rhydderch himself) that the White Book was produced in the well-appointed scriptorium of Strata Florida.

The White Book was the work of five different scribes, usually referred to as Hand A, Hand B and Hands C, D, E. Hands A and B have often been linked. The connection between C, D and E is closer still. It has been suggested that Hand D was responsible for some of the rubrication of the White Book as a whole, and may thus have played an ‘editorial’ role in bringing together piecemeal work of the other scribes. He is also of particular interest to us as he appears to have been personally responsible for copying most of the ‘native’ material of the White Book, including the bulk of the Mabinogion tales. All five hands have been dated, with reasonable confidence, to the middle of the fourteenth century.xxxviii

The White Book begins with the so-called Delw Y Byd ‘The Form of the World’, a Welsh rendition of a popular medieval cosmological treatise Imago Mundi written by Honorious of Autun (d.1151). This is followed by a narrative account of the Sibylline Books, representing a popular genre of prophetic literature. This is followed by various works of medieval apocrypha, including the translation of Transitus Mariae that had been originally commissioned by Rhdderch’s great-great grandfather in the third quarter of the thirteenth century. Hereafter, we have a number of texts copied by Hand B. These include one religious work (Purdan Padrig), after which a number of pages appear to be missing. Hand B resumes with a number of translated texts from the Frankish Carolignian cycle. Hand C is found in the sections after this, whose sole (but by no means unsubstantial) contribution is Bown, the Welsh translation of the Anglo-Norman Bevis of Hamptoun.

After this, the script is that of Hand D who (as we have established) was largely responsible for the native material. His contributions begin with the Four Branches of the Mabinogi, which are followed by the popular romance of Peredur, the Romano-British pseudo-history of Maxen Wledig, the ‘expanded triad’ of Llud a Llefris and the chwedl of the Lady of the Fountain (more usually known as the romance of Owain). After these Mabinogion texts we have a series of Triads, some genealogies of Welsh saints and a handful of old native poems, a number of which are associated with the prophet-magician Myrddin (Merlin). After this Hand E takes over, contributing two further Mabinogion tales of South Walian provenance – Geraint and Culhwch ac Olwen, which breaks off after two pages, bringing the White Book (or what we have of it) to an abrupt end.

What can we say about this fourteenth-century codex? The mixture of translated and native works, secular and religious genres is intriguing but (as we have seen) by no means unprecedented in the earlier Welsh manuscript tradition. However, beneath this cheerful eclecticism there is at least some degree of order within the arrangement of these contents. The different ‘Hands’, as we have seen, appear to have distinct if sometimes overlapping domains of interest. Hands A and B, as we have seen, both worked with translated literature, at least some of which may have been drawn directly from Rhydderch’s family archive. Hand A seems to specialise in hagiographic or apocryphal works; while what remains of Hand B’s section of the White Book would suggest a particular interest in the Frankish Charlemagne cycle. Hand C dovetails partly with the latter in his preoccupation with Anglo-Norman Romance. Hands D and E, as we have seen, were preoccupied exclusively with native chwedlau and material deriving from the cyfarwyddyd; although it could be argued they are continuing in the vein of ‘narrative prose’ started by Hands B and C. While it is sometimes unclear to what extent late medieval readers would have distinguished these native works from those translated from a foreign original, the ordering of the White Book at least suggests some kind of differentiation between these various sections of the manuscript, and the spheres of interest (or source materials?) of their respective scribal agents.

As manuscript expert Daniel Huws suggests, the White Book would have been ‘an object of marvel’ in its time. In the mid-fourteenth century, large-scale books of this kind were becoming high-status accessories in the courts of England and France. The White Book may have been Wales’ answer to the Harley lyrics manuscript or the Auchinlek romance book, both of which were compiled in the 1330s. However, despite the existence of these fashionable exemplars, there was something innovatory and experimental about the White Book of Rhydderch, at least within the contemporary Welsh context. It was unusual both in its scale and its content – which was, as we have seen, overwhelmingly preoccupied with narrative prose, ‘Welsh belles-lettres’xxxix . Nothing quite like it had been attempted before, and it signalled a new attitude towards literature – and secular literature in particular – within Wales’ reading culture.

With the Red Book of Hergest, produced some two generations later, we have a more complete realisation of this concept. Huws is probably right to suggest that the Red Book may have been an imitation of the White, even if their textual relationship is rather more complex than one of straight transmission. However, within the Red Book there is a greater range of material and the arrangement of this material has the appearance of being more self-consciously schematic. The collection begins with a series of Latin-derived histories, including a translation of Geoffrey’s Historia Regum Britanniae, onto which is grafted the series of Welsh chronicles of the medieval period known as the Brut y Twysogion, the Chronicles of the Princes. These are followed by translations of the Carolingian works, similar to those found in the White Book. After this, the Red Book includes a series of chronicles and prophesies among which is to be found, for the first time, the Mabinogion text of The Dream of Rhonabwy. This leads into an extensive body of Triadic listings, which (unlike the White Book triads) exemplifies the ‘later’ tradition, with its references to the French-derived romances and other foreign exemplars. After this, we come across what would appear to be a group of romances which begins with a Welsh reworking of the ‘Pilgrimage of Charlemagne’, but continues with the Mabinogion tales of Owain, Peredur, Macsen and Llud a Llefelys (in that order). These are followed directly by the Four Branches, followed by Geraint, Culhwch and Olwen and the translation of Beeves of Hamptoun once again. After this, there are no more Mabinogion texts, but the Book continues for another three hundred columns, on which are found texts of a medical and historical interest, followed by some proverbs gnomic and elergaic verses. The Red Book concludes with a short collection of religious- and court poetry from the later thirteenth century.

Reviewing the Mabinogion texts in situ within the pages of the Red and White Books, one is forced once again to register the extraordinary range of literature with which late medieval Welsh reader seems to have been acquainted – religious and secular works; translated texts and those of a more local pedigree; philosophical, elegaic and comical genres; items of historical record and more imaginative works which in modern terms would be understood as ‘fiction’. Such richness was of course entirely characteristic of Welsh letters from the Age of the Llywelyns, as we have seen. But the extraction of the Mabinogion texts from this dense thicket of medieval literature was the result of an editorial process that took place not in the medieval era itself but in the early nineteenth century, as part of an antiquarian project that culminated in the publication of the Mabinogion volumes by Lady Charlotte Guest between 1838 and 1849. Does it merely represent nothing more than an artificial assemblage based on the tastes and preoccupation of its Romantic nineteenth-century editors? Or are there certain inherent qualities which justify the consideration of these texts as a group, distinguishing them from those around them?

The contents of the White and Red Books do not, on the surface of it, appear to suggest the Mabinogion texts were accorded any special distinction by the late medieval anthologists. They appear to have existed within a mass of narrative literature – some of which was of native derivation, some of which would have been translated or adapted from foreign originals. The late medieval editors – unlike their Victorian counterparts – seem to have made little distinction between these two types of narrative text. Having said this, the Mabinogion texts do seem to appear in distinct clusters within the pages of both the Red and White books. The composition of these clusters – and the position of each in the wider manuscript – may have something to tell us about the late medieval understanding (in classifactory or taxonomic terms) of the Mabinogion texts.

Within the White Book, the Mabinogion texts are not in a continuous series but occur in two main sections, which may reflect the original manuscript sources from which Rhydderch’s codex was assembled. If we are to look to a medieval prototype of the concept of the Mabinogion, its here that we will find it. To quote Daniel Huws once again: “the texts brought together by scribes D and E represent alone a quite unprecedented compilation. The White Book, or a very close antecedent, created the canon which eventually achieved unforeseen fame as ‘the Mabinogion.’” xl

As we have said, within the White Book itself, the texts do not run in a continuous sequence, but are rather found in two separate clusters. Based on the provenance of the surviving pre-White Book manuscripts, we might tentatively define these groupings as the ‘northern’ cluster of texts (consisting of Owain, Peredur, Maxen and Llud a Llefelys) and a smaller ‘southern’ cluster (consisting of Culhwch ac Olwen and Geraint ac Erbin). In the White Book, the Four Branches are placed with northern cluster, although the ordering of the texts in the Red Book of Hergest (as well as the orthography of Peniarth 6i) suggests that this might not have always been the case. The White Book’s northern cluster (beginning with the Four Branches) follows on from Carolingian sagas and the Anglo-Norman romance Beeves of Hamptoun, which is followed in turn by a sequence of Triads, genealogies and old poetry. To the extent that we can read anything into this particular editorial arrangementxli – we might conclude that the ‘northern cluster’ of Mabinogion texts were regarded as belonging, generically, somewhere between the categories of Continental Romance and the native cyfarwyddyd. Significantly, perhaps, the southern cluster is tacked on after this section, perhaps reflecting a sense that it belonged with the latter and was therefore of more native provenance, and perhaps was felt to belong to the category of lore/historia rather than romance/fabula.

In the Red Book of Hergest, the Mabinogion tales (apart from the Dream of Rhonabwy) form a continuous sequence. But they are bookended by texts of Franco-Norman derivation – the Pilgrimage of Charlemagne and Beeves of Hamptoun – suggesting, perhaps, that to readers in fifteenth-century Wales, this entire group of tales (whether of native or foreign derivation) would all have belonged in the general category of ‘romances’. It is perhaps significant that these texts are separated in the manuscript from the Brut and the Latin-derived chronicles – possibly indicating that the late-medieval anthologists regarded these romances as fabula (i.e. fiction) rather than history. This may not have been the case when texts such as the Four Branches were first composed, and it is interesting to reflect on how a piece of pseudo-historical cyfarwyddyd such as this could subsequently become reclassified as a ‘romance’ in the changing world of late medieval Wales.

It is also worth noting that a well-established configuration of modern Mabinogion scholarship, the so-called Tair Rhamant or Three Romances, is not supported in any obvious way by the late medieval manuscripts. In neither version do Owain, Peredur and Geraint appear in a continuous sequence; and there is every suggestion that the three may have quite different manuscript histories. As we will see later on in these pages, these texts do have certain generic modalities in common – as well as sharing a relationship (complex though it may be) with the Arthurian romances of Chrétien de Troyes. However, from the evidence of these manuscripts themselves, it would appear that these shared properties were either invisible or insignificant for their late medieval readers. Geraint was more likely to be associated with the archaic Culhwch ac Olwen (perhaps on the basis of shared provenance or nomenclature). Peredur and Owain appear in a grouping along with Maxen and Llud in both manuscripts, and the rationale of this grouping is rather harder to interpret. It seems likely that these clusterings reflect nothing more than the composition of the original manuscript sources drawn on by the compilers of the Red and White Books.

So to return, finally, to the question of the Mabinogion and its value as an editorial grouping. It has become fashionable in recent decades to deride the work of the first generation of modern translators and editors of the medieval corpus. ‘Romantic’ is the term usually applied (pejoratively) to the motivations of antiquarian pioneers such as Lady Charlotte Guest and her guide and mentor Owen Pughe. This in turn has led some to question the validity of a collection like the Mabinogion, and wonder whether to what extent it reflects the tastes and preoccupations of these nineteenth century romantic nationalists rather than offering an honest representation of the literary culture of the Welsh Middle Ages.

To have achieved the latter, Guest and Pughe would have been better simply translating and publishing the entire body of the Red Book of Hergest – the latter being fairly described as the closest one is likely to get to an entire national literature in a single volume. However, it is doubtful that nineteenth-century readers would have found texts such as the Welsh Transitus Mariae or Dares Phrygius as entertaining or interesting as the Culhwch ac Olwen or the Fourth Branch of the Mabinogi. Naturally enough, the Mabinogion texts were selected on the basis of their appeal to a Victorian reading public. But what is interesting, as we can see, is that an almost identical selection seems to have taken place in the first instance, when these texts were chosen by Hand D for inclusion in the White Book in the mid-fourteenth century. We might liken Hand D’s taste in prose with the poetic tastes of the thirteenth-century compiler of the Book of Carmarthen – whose love for the quirky, the archaic and the mistily-indigenous seems strangely of a piece with the antiquarianism of Guest and Pughe. The suspicion must be that, to reverse the old ironic complaint, nostalgia is still what it always has been. Something like romanticism, it seems, was an active force in the mid-fourteenth century just as it was in the early nineteenth.

A case could certainly be made for greater interest in some of the more neglected works of narrative prose found in the Red and White Books: the Welsh Charlemagne tales, for example, which (judging from the frequency of their appearance in manuscript sources) were at least as popular for readers in thirteenth- and fourteenth-century Wales. Nonetheless, the Mabinogion texts retain a special interest to anyone seeking to understand the specific contribution of Wales to European literature. Collectively, they describe a fascinating historical trajectory: plotting the key moments in the development of a marginal national literature; while also capturing intriguing intimations of some of the wider cultural and intellectual changes taking place in Wales and Western Europe during the centuries leading up to the Renaissance. Individually, each text represents a moment of praxis in the evolving art of Welsh narrative prose: each captures a wealth of contextual detail and artistic fulfilment. Whatever one thinks of the editorial motives of Lady Charlotte Guest and Owen Pughe, we owe them our thanks for bringing these medieval masterpieces to the attention of the modern age.

iOn the study of Celtic literature and other Essays (1910) p.54
iitr. John Rhys in The Text of the Book of Llan Dav: Reproduced from the Gwysaney Manuscript (1893). Reproduced online at
iiiMy translation The Four Branches of the Mabinogi(2005) p.184
ivFrom Descriptio Cambriae trans. Lewis Thorpe (1978) Book I Chapter 17: "The Welsh value distinguished birth and noble descent more than anything else in the world. They would rather marry into a noble family thaninto a rich one. Even common people know their family-tree by heart and can readily recite from memory the list of their grandfathers, great-grandfathers, great-great grandfathers, back to the sixth or seveth generation'
vOne of the most interesting and coherent theories of magic was developed by the social anthropologist Ernesto De Martino, whose work has been rather overlooked in the English-speaking world. The core hypothesis, outlined in Primitive Magic (Prism Press: 1988), is that the stabilisation of the human ego – what De Martino refers to as 'the controlling presence' - is a relatively recent historical phenomenon, which in certain cultural settings even to this day is not a given fact. Magic is most characteristic of the evolutionary phase when the controlling presence was struggling to establish itself, and is under constant risk of being overwhelmed by compulsive instinctuality – a state of identity-less regression which De Martino describes as koinonia (communion): “when the controlling presence [is experiencing the onset of of koinonia]…a typical and easily-recognizable type of anguish is produced. The presence may then become aware of the need to redeem itself, and it does this by creating certain cultural forms. If the presence fades, without any effort or reaction taking place, then the magic world has not yet appeared. If the presence is redeemed and consolidated and is no longer aware of its own fragile nature, then the magic world has already disappeared. It is between these two stages that the magic world makes its appearance – in the struggle and opposition that arise between them – here, the magic world becomes manifest as movement and development: it displays its many cultural forms and is born into the history of mankind.” (Primitive Magic p.70).
viAnne Wilson Plots and Powers: Magical Structures in Medieval Narratives (Florida: 2001) p. 10
viiK. Rasmussen 'Intellectual Culture of the Igluick Eskimos' Report of the 5th Thule Expedition 1921-24 VII, no. 1, Copenhagen 1929. Quoted in translation in de Martino, pp. 108-109
viiiTrans. Gottfried Storms Anglo-Saxon Magic (1948) p.140-143
ixAlaric Hall Elves in Anglo-Saxon England: Matters of Belief, Health, Gender and Identity (Woodbridge: 2007) p.108 ff.
xDaniel O'Keefe Stolen Lightning (Vintage: 1983) pp. 42-43
xiSee Patrick Sims-Williams 'Some functions of Origin Stories in Early Medieval Wales' in History and Heroic Tale: A Symposium ed. T. Nyberg (Odense University Press, 1985)
xiiTYP lxx ff.
xiii'Interpreting Irony in Medieval Celtic Narrative: The Case of Culhwch ac Olwen' Cambridge Medieval Celtic Studies 16 1988 pp 41-59
xivThis conclusion is supported by Bromwich's observation (TYP p. lxxix) that the Bardic classes seem, for the most part, to have been unable to read Latin until the mid-fourteenth century, and that texts such as Geoffrey's Historia only became widely known in native Wales after their translation in Welsh in the early thirteenth century.
xv'From Traditional Tale to Literary Story: Middle Welsh Prose Narratives' in The Craft of Fiction ed. L. A. Arrathoon (Rochester: Solaris Press, 1984)
xviThe Poems of Taliesin edited and annotated by Sir Ifor Williams (English version by J. E. Caerwyn Williams) (1987) pp. lix-lxiii
xviiibid p. xlvii-lv
xviiiHB Ch. 62. Of Blwchffard, Cian and Talhaearn Tad Awn (‘father of the muse’) no more is evident in the medieval record. Aneirin was remembered as the composer of the sixth-century poem Y Gododdin, which (as we have seen) enumerated the fallen British warriors of the Votadini tribe who died at Catraeth (Catterick) in North Yorkshire defending their land against the Deirians of Northumbria
xixO’Keefe, op cit. p.62
xxBetween Languages: The Uncooperative Text in Early Welsh and Old English Nature Poetry (1993) p.211
xxiWilliams, op. cit., p. xv-xvi
xxiiHigley, op. cit., p. 284-285
xxiiiMuch significance has been attached by some commentators of Taliesin’s occasional disparaging of clerical colleagues. However, it would be wrong to read these side-swipes as blanket anticlericism, and more wrong still to see in them evidence of recalcitrant paganism on the part of Taliesin. As we will see, there is a far stronger body of evidence for a social, intellectual and cultural continuum between the bardic poets, law makers and clerical orders in Native Wales in the early and central middle ages. The school of Taliesin are perhaps best understood as representing the extreme ‘nativist’ wing within this cultural establishment, but in no sense were they anti-Christian or cut off from book-derived Latinate sources of knowledge.
xxiv‘…verbal duelling is a widespread tradition, seemingly without racial or national boundaries. One one level, according to this research, the game is a duel of wits and verbal ability. The participant most skilful in using standard sounds and artful in innovating new ones may draw admiration and praise from his listeners, thereby gaining prestige and status.’ (Robert A. Georges and Michael Owen Jones Folkloristics – An Introduction (1995) p.253-254). Rachel Bromwich (TYP, p. 510, n.2) mentions how Taliesin’s ymryson was invoked by the late medieval bard Philip Brydydd, during a contest with a rival as to which of them should present a poem to Rhys Ieuanc at the Christmas festivities at Llanbadarn Fawr. Taliesin’s most memorable ymryson seems to have been with the bards at Elffin’s court (see below), at the end of which he reduced his rivals to a state of infantile incoherence.
xxvThe full text of this legend is to be found in Elis Gruffydd’s Chronicle of the World written around the middle of sixteenth century. Gruffydd claims to have known the story from oral sources, and there is every indication that it was essentially the same tradition referred to by the poets in the Book of Taliesin. Excerpts from Gruffydd’s account have been translated and published by Patrick Ford in The Mabinogi and other Welsh tales (1977) pp.162-181
xxviHigley, op. cit. p.201
xxviiibid p.208
xxviiiDescriptioBK I Ch.16 (p.246-247)
xxixVita Merlini ll. 1162-1168 (Clarke trans., 1973, p.115)
xxxTYP lxxxi
xxxiText and translations of all of Dafydd ap Gwilym's works have been made available online by Swansea university at This translation has been taken from that source.
xxxiiMedieval Welsh Manuscripts (UWP: 2000) pp.60-61.
xxxiiiB. F. Roberts ‘Where Were the Four Branches of the Mabinogi Written?’ in The Individual in Celtic Literature CSANA Yearbook 1 ed. Joseph Nagy (Four Courts Press: 2001)
xxxivWill Parker ‘Gwynedd, Ceredigion and the Political Geography of the Mabinogi’ The National Library of Wales Journal Volume XXXII Number 4, 2002
xxxvCertain orthographic features in Peniarth 6, notably the omission of the prosthetic y-, seem to echo the phonology of the Old Welsh period. This was foremost among the reasons for Sir Ifor Williams assigned an eleventh century dating of the Four Branches, a dating that remains remarkably tenacious, despite having been convincingly refuted by Eric Hamp among others. ‘Mabinogi and Archaism’ Celtica 23, 1999, pp. 96-112.
xxxviRoberts (2001) op. cit. p.70
xxxviiHuws, Medieval Welsh Manuscripts (2000) p.252 ff.
xxxviiiibid p.234
xxxixibid p.246. The few remaining examples of Old Welsh secular literature was, as we have seen, found scrawled in the margins of sacred books – short historical-legal notices like the Surexit memorandum quoted above, or occasional verse ditties like the so-called Juvencus englynion. Well into the Middle Welsh period, the umbilical association between religion and writing would remain, and we may assume that most if not all surviving manuscripts from the thirteenth century would have been physically produced within monastic or ecclesiastical settings. The lay/secular impetus of the White Book signals a distinctive break from this clerical context, even if (as we might suspect) the physical manufacture and inscription of the manuscript itself took place in the scribal hives of Strata Florida.
xlibid p.246
xliAttempting to define an overall editorial scheme for medieval manuscripts – particularly those of an earlier date – is a fraught with a number of problems. Given the rarity of vellum and the other materials of writing, scribes were prone so simply using whatever material was to hand. Most manuscripts (including the White Book) appear to have been the work of a number of different scribes, and one cannot always assume a coherent editorial schema, let alone any taxonomic significance within the ordering of the contents. However, the later medieval manuscripts do give the impression of being consciously assemble works. These books themselves were items of prestige, which (particularly in the case of the Red Book) give some impression of attempting to embody an entire national literature. Thus one might, with some justification, look to the structure of such work and expect to find within it some sense of a taxonomic structure which might have governed the arrangement of its contents. With the White Book, however, special care is required given the incomplete and damaged nature of the manuscript. (see Huws, op.cit, pp.227-263)