The Four Branches of the Mabinogi

'The Four Branches’ or ‘the Mabinogi’ (the terms will be used interchangeably) are among the finest and most unusual pieces of Middle Welsh prose, and are thus accorded pride of place within the Mabinogion collection as a whole. Set in the time of the Sons of the Beli Mawr, the Four Branches might be said to belong to the same mythic pre-Roman cycle as the tale of Llud a Llefelys. But there the similarity ends. Although Caswallon fab Beli is present in the background of the Mabinogi, the focus of interest is not so much in the Sons of Beli Mawr themselves but rather in a group of affiliated client dynasties based in the area of Western Britain now known as Wales. The main action of the Four Branches takes place in two key centres of power within this region: the Cardigan/Teifi valley area of Dyfed in West Wales, and the northwest of Gwynedd centred on the Anglesey and Arfon coastlands. The relationship between these regional power-centres and the crown of London in the Mabinogi bears a strong resemblance to the geo-politics of the late-twelfth century, echoing the relationship between the native Welsh warlords and the Anglo-Norman Angevin kings during centuries between the establishment of the March and the Edwardian conquest of the late 13th century. These parallels were not accidental. It would appear that for the contemporary courtly audiences in Wales, this process of typological insinuation – linking the mythic past with the politics of the present – was an established function of the genre. These typological allusions within the Four Branches (which were no means always flattering to the medieval subjects involved) will be considered in more detail below.

On the surface, the plot of the Mabinogi concerns the births, lives and deaths of the births, lives and deaths of a group of interrelated characters from the mythical horizon of British historiography: the generations before the Roman occupation during the rise of the Sons of Beli Mawr. The origins of these characters and their associated legends are various, as we shall see, but they have been woven together by the medieval author (or the tradition he was working within) into a loose and syncretic panorama of myth, magic and dynastic politics. The family groups involved might be seen to represent the various demographic elements which made up the heterogeneous population of Iron Age Britain, albeit in highly schematic form – with several millennia of prehistoric demography condensed into the space of three or four generations.

The southern dynasty of Pwyll and Rhiannon seem to have been had a closer affiliation with the archaic pre-Celtic substrate – what might be described as the ‘Indigenous Underworld’. The northern dynasties of Llŷr and Dôn, on the other hand, are both linked by marriage with the incoming Sons of Beli Mawr (a mythical reflex, as we have seen, of the Late Iron Age Belgic colonialists). What the Mabinogi as a whole maps out is the gradual transition of geo-political power between these various groupings, charting the archaic tribal history of the island through a complex sequence of hegemonic events – including marriages, wars, alliances and swindlings.

Magic and The Mabinogi

Magic plays an important part in these narratives, even if its precise significance is sometimes hard to grasp. In the First Branch, for example, the Demetian queen Rhiannon takes on the role of a horse – physically carrying visitors to her court from the gate to the mounting block – as ‘penance’ for the crime of cannibalistic infanticide for which she has been wrongly accused. Her enactment of this strange punishment coincides with the mysterious reappearance of the infant Pryderi seventy miles away in the district of Gwent-ys-Coed, at the household of a certain Teyrnon Twryf Lliant. This Teyrnon had been in possession of a prized brood-mare, whose foals had been disappearing mysteriously, just as did Pryderi. The precise relationship between these events is never fully explained, but the sense of an underlying pattern of symbolic interconnections is suggestively developed throughout the text of the Four Branches. At regular intervals throughout the Mabinogi, the rational sequence of events is interrupted by a supernatural event, which often occurs immediately before, after or in conjunction with the outlandish and inexplicable behaviour of one of the protagonists. Closer analysis of these magical ‘moves’ reveals that beneath their savage façade they contain a specific resonance within the underlying archetypal system that extends rhizomically throughout the text as a whole. Much of the Mabinogi is characterised by these strange irruptions, which often reflect decisive shifts of the balance of power between the groups and individuals involved.

Medieval Welsh prose narrative is not given to exegetical disclosure, and these bizarre and occasionally grotesque sequences invariable pass without comment. Naturally, such outbursts do not sit easily with the modern mind, which has an uncomfortable relationship with its pre-rational past. A good example of this is to be found in episode the ‘Feast of the Wonderous Head’ which takes place after the carnage at the end of the Second Branch, in which the wounded king Bendigeidfrân requests that his own head be severed from his body before he dies. This is done, and the head lives on, as the surviving members of his household eat and drink in a supernatural feast in which the boundaries of space and time seem to break down. The concept of the ‘Living Head’ to which reverence is paid and oracular pronouncements are attributed is strongly characteristic of pre-Christian Celtic religious iconography; while the heightened emotion and distortions of time and space which occur during this episode bears the hallmarks of the collective hysteria (as we would understand it today) – the psychological context for magical experience throughout the pre-modern world. Such experiences – in which the boundaries between the objective and the subjective worlds become alarmingly porous – the modern world, for the most part, has left gratefully behind.

Pagan Origins

It is clear, however, that the core of the Four Branches bears the hallmarks of this primitive ‘magical’ consciousness. This may be a legacy of its origins in the pre-Christian past. The rather mysterious term Mabinogi itself has been interpreted by Professor Eric Hamp as deriving from the Common Celtic *Maponaki “(the deeds of) the family of the god Maponus.” The boy god Maponus (known as Mabon in Medieval Welsh literature) seems to have been a popular Gallo-Brittonic deity, who was sometimes equated with the Roman Apollo, about whom we will have more to say in the following section. Building on the work of the prolific W. J. Gruffydd, Hamp detects traces of the myth of Maponus and his family in the First and Third Branches of the Mabinogi. If his interpretation of the word Mabinogi is correct, then this archaic core owes it origin to sacred stories told on the holy feast-days of the deities involved – seasonal pageants which were doubtlessly accompanied by the kind carnivalesque delirium that characterised such occasions throughout the ancient world. After the original significance of the term was forgotten, ‘the Mabinogi’ remained a general repository for tales of pre-Christian origin and atmosphere, which were eventually woven together into the heterogeneous synthesis we find in the Four Branches.

Further traces of these pagan origins are not difficult to find. Many of the protagonists of the Fourth Branch in particular may have been divinities in origin, and we find their cognates elsewhere within the Irish and Icelandic mythical systemsi. The hero Lleu Llaw Gyffes clearly relates to Lugus ‘the Gaulish Mercury’ – a prominent deity in the Gallo-Brittonic west. The extraordinary sequence at the end of the Mabinogi of Math – in which Lleu is found crouching at the top of an oak tree in the form of an eagle – testifies to a rich seam of Indo-European shamanism preserved in this medieval legend. Manawydan fab Llŷr , who appears in the Second and Third Branches, probably owes his origin to a maritime cult prevalent in the Irish Sea area during the early centuries of the common era (Manannán mac Lír, whose divine origins are evident in the early Irish tradition, seems to reflect the memory of this god in the Gaelic world). Rhiannon, as well as being assimilated into the role of the primordial Matrona of the original Maponus mythology, also shows strong affiliations with the Gallo-Roman horse goddess Epona, as we shall see. Other figures such as Teyrnon Twryf Lliant (lit. ‘The Great Lord of the Turbulent Water’), Pendaran Dyfed, Llaesar Llaes Gyfnewid and Blodeuwedd may have their origin in more localised cult legends.

What we know about the religious systems of pre-Christian, Iron Age Britain is largely informed by a fragmentary and often inconsistant range of source materials: including archeological data, classical witnesses such as Posidonius and Julius Caesar, medieval vernacular sources and comparative anthropological models. From this evidence we can reconstruct a picture of a diverse religious landscape, characterised by a myriad of local cult forms (reflecting the ethnically various and tribally decentralised demographics of the British Iron Age). There are some signs that, by the end of the period, a trans-regional druidic priesthood had begun to superimpose a form of religious syncretism which brought some measure of political as well as spiritual unity to the notoriously fractious Celtic world. The precise nature of the druidic religion remains largely a matter of speculation, but we might conject that something like the 'sacred year' of classical paganism, the Hindu religion, or even medieval Catholicism might have been involved: in which a calendrical framework, punctuated by regular feast-days (each of which was associated with a particular sacred figure) united a diversity of cults within a larger system. There is some evidence that, perhaps as part of this syncretising principle, the gods were brought together into a kind of supernatural community, reflecting the tribal and hierarchical social world of their worshippers. A syncretic, medieval mythological text like the Four Branches might be seen as a linear descendant of these habits of mind.

Tribal History

As we have already suggested, alongside the remnants of these pre-Christian sacred tales, the Mabinogi also contained what might be described as ‘tribal historic’ elements – that is to say that it represented a narrative recollection of the significant power struggles and demographic events from the prehistory of the British Isles. As we have seen, this was not ‘history’ as we would understand it today, but rather a series of telescoped, emblematic origin-narratives of the kind found in the early sections of the Old Testament and the more archaic chronicles of the ancient world. However, beneath the myth and hyperbole of these accounts we sometimes find what may be the grain of genuine historic memory. Caswallawn fab Beli Mawr, who looms large in the background of the Second and Third Branches, is clearly a recollection of Cassivellaunos, the Belgic warlord who fought Julius Caesar in the first century BC. Echoes of the Gaulish Sack of Delphi (which took place in 279 BC) have been identified in the Second Branch accounts of the invasion of Irelandii, although this archaic core has doubtlessly been fused with more recent material, including possible recollections of a Romano-British incursion into Ireland under the leadership Bennius or Benignus (recalled in the Irish tradition as Benne Brit)iii. More speculatively still, we might wonder whether the ‘pig wars’ documented in the Fourth Branch might reflect an even more ancient tribal history. The pig, an otherworld symbol in the Welsh tradition, was also central to the agricultural and religious systems of the early Neolithic ‘Grooved Ware’ culture of the British Isles – the originators of the great sacred landscapes Maes Howe, New Grange, Avebury and Stonehenge. Might the seizure of these animals from the Indigenous Underworld represent a short-hand account of the transition of power between the Grooved Ware culture of the early Neolithic to the Bell-Beaker culture of the mercantile, Bronze era Age? Here, however, we are deep within the realms of speculation.

Note: We have very little direct source material for the pre-Christian religious systems of Iron Age Britain. What little evidence we do have suggests a diverse landscape, characterised by a myriad of local cult forms, which mirror the hetrogenous nature of the populations involved. There are some signs that a druidic priesthood had (with some degree of success) created some kind of syncretic magico-religious system, and integrative projects like the Four Branches itself may represent a linear descendent of this tradition. Within such systems, cult figures from a variety of traditions might be integrated into a kind of supernatural community, reflecting the tribal and hierarchical society of their human worshippers.

A Mirror to the Medieval Present?

Any number of such traditions may have been included in the composite brew that was the Mabinogi. It is possible to see that even within the historic era, this synthesis of tradition was far from stable or consistent. New elements might be included, old ones omitted, existing traditions altered and rearranged. While, as we have seen, the learned culture of native Wales was capable of preserving the memory of key mythic- and tribal-historic elements over centuries if not millennia, one has to weigh these habits of conservation against the inherent fluidity of the oral tradition. The recitation of tales of this kind was a social event, in which the storyteller draws from his stock of memorised lore to deliver a narrative which aligns itself with the hopes, fears and expectations of his audience. What was remembered – and the way it was remembered – is determined by the sentiment and preconceptions of the audience as well as the knowledge of the storyteller himself. Within such a context, it is far from unknown for the desire to tell a good story with contemporary significance to trump the conservative instinct of accuracy, literalism and adherence to received learning.

The key point to realise about narratives like the Four Branches was that their social functionality was as much about explaining the present as understanding the past. There is a tendency within many pre-modern cultures to view time as a cyclic process – with the past reflecting the present and the future, and vice versa. A thriving tradition of prophesy and historical allusion would have accustomed the medieval Welsh mind to identifying typological parallels between the mytho-historic narratives of the past and the imminent political realities of the day. In the tangled, violent and often tragic narratives of the Mabinogi, the medieval audience would look for the meaning and significance of their own unsettled times. Specific parallels would be drawn between events and individuals in the Mabinogi and the chaotic arena of twelfth-century dynastic politics. Through narratives like the Four Branches, the storyteller historian was able to offer his own oblique commentary on the present through his narrative interpretation of the past. This would have made the bardic cyfarwydd a figure of considerable power and significance in the public life of Medieval Wales.

This ‘typological’ function of the Four Branches has only recently been recognised by modern literary historians, and is still a matter for some debateiv. One of the first scholars to look at the text in this way was the poet-politician Saunders Lewis, who identified a close parallel between Bendigeidfrân’s invasion of Ireland and that of Henry II in 1172. Lewis also drew attention to the fact that Lord Rhys of Deheubarth attended the court of Henry II in 1177, in a way that was highly reminiscent of meeting between Pryderi and Caswallawn (also at Oxford) at the beginning of the Third Branchv. Following these leads, further parallels have been identified between Lleu, the hero of the Fourth Branch, and the young Llywerth ap Iorwerth (soon to become Llywelyn Fawr, the powerful thirteenth-century prince). This and other geo-political pointers would appear to confirm a late-twelfth century provenance for the final version of the Mabinogi, as it has survived in manuscript form todayvi. Numerous other comparisons can be drawn between the events and protagonists of the Mabinogi, and the ‘headline’ events of twelfth century Welsh public life. An example of this is to be found in the ‘Badger in the Bag’ incident – which may contain at least the hint of an echo of the shaming of the Anglo-Norman Gerald of Windsor, at the hands of the royal lovers Nest daughter Rhys ap Tewdwr and Owain son of Cadwgan in the year 1109. Gerald’s consequent scouring of Dyfed and the exile of Owain and his father, seems to be being obliquely recalled in the fate of Pryderi and Manawydan in the Third Branch.

An even more striking parallel can be discerned between the posthumous fates of Bendigeidfran and the notorious eleventh-century tyrant Gruffydd ap Llywelyn. Though under somewhat different circumstances, both kings ended up being beheaded by their own men, and having their severed head carried to an English royal centre. Through a rather gory piece of synchronicity, the memory of the ‘Wonderous Head’ was reawakened in 1243, when a captured Welsh prince (also, coincidently, called Gruffydd ap Llywelyn) fell to his death while attempting to escape from the Tower of London – the probable location of the ‘White Hill’ in which Bendigeidfran’s head was buried. His body was found at the foot of the Tower, with his head buried in his chest cavity from the impact of the fall. As if this were not enough, Gruffydd’s son was Llywelyn II, the native prince of the Welsh, also suffered a Bendigeidfran-esque posthumous legacy some thirty years later, with his severed head crowned with ivy and displayed – also at the Tower of London – where it remained for a full twelve years. Was this a calculated mockery of the Welsh tradition on the part of the Anglo-Norman authorities, or was something altogether darker and stranger at play? There is more than a slight sense of life imitating art in the destiny of the Welsh nation in its final centuries.

It was this knack of identifying the perennial inflexions of Welsh public life and defining these in the form of mythical antecedents which characterises the discursive method of the author/redactor of the Four Branches. Thus when he represents Gwydion and Gilfaethwy sons of Dôn as impetuous troublemakers, straining at the leash under an aging king, he has in mind not only the recent misbehaviour of the sons of Owain Gwynedd, but also the conduct Owain himself and his brother Cadwalladr in the previous generation. Even as far back as the late 10th century, we have the sons of Idwawl son of Anarawd enacting what was to become a familiar paradigm in Venedotian dynastic politics: an aggressive pair of brothers who would begin as allies, disposing of the older generation and harassing rulers of neighbouring territories before turning on each other and fomenting an internal civil war which would reverberate for several generations. The author was alert to these patterns of recurrence, these habitual fault-lines in the geopolitical landscape of Medieval Wales. One of the most extraordinary things about the Four Branches is the sense that the author has succeeded in synthesising a narrative vision that not only distils and comments on the issues of his day, but does so in a way that is also entirely consistent with his vision of the past. The Mabinogi is best understood as simultaneously serving the functions of myth, tribal history, social commentary and prophetic revelation. In the author’s mind would have been a number of mythical narratives and items of historical lore that he would have grouped under the generic category of the ‘Mabinogi’ (at the core of which was the myth of Maponus and Matrona). Out of these fragments, he would seek to present a composite whole through which the essential rhythms and contours of tribal-political life – both past and present – could be apprehended at a single glance. There is something a little unearthly about a condensed revelatory vision of this kind, which bears the impression of being conceived in the white-heat of visionary inspiration, perhaps under the influence of the subconscious agency known in Medieval Wales as the awen.

Literary Style

It is generally agreed that the Four Branches represents one of the finest expressions of Welsh narrative prose surviving from the Middle Ages. The writer was probably from the second or perhaps third generation of writers of Middle Welsh prose (a style and orthography that seems to have replaced Old Welsh from the early twelfth century onwards) and the ‘smoothness of style’ alluded to above appears in a fully matured form in the Four Branches. The use of dialogue in these tales also shows the artistry of the medieval chwedl at its best, to quote Professor Roberts once again:

Not only does he use dialogue as a narrative device, but he is able to give his characters recognizably individual forms of expression. He reveals oppositions, points to social distinctions, and paints delightfully naturalistic cameos in the conversations of his characters

Techniques such as these exemplify the literary genius of the Four Branches at the microcosmic level. However, his artistry is also evident in the organisation of the narrative superstructure. We have seen how the Mabinogi seems to have emerged from a vernacular literary culture that was relatively well-developed: in which the written prose tale was an established form; and in which a body of narrative, poetic and triadic references was in place to provide a substantial inter-textual hinterland. Welsh prose literature, in other words, had come into its own; and the author of the Four Branches was sufficiently confident with its deployment to begin to play with its characteristic forms and to attempt to extend its potentials. One feature that was to emerge from these experiments seems to have been the application of a complex plot-structure known as the interlace form. In its most basic form, narrative interlacing involves the contrapuntal interweaving of a number of two or more distinct sub-plots. Thus in contrast to the more basic serial format such as we find in Culhwch and the some of the Three Romances (see below), the narrative stream of the Four Branches can alternate between various sub-plots – which tend to converge significantly as the story progresses.

A more sophisticated manifestation of this same literary device is where the interlacing occurs on a thematic as well as a structural level. Here, the sub-plots not only run in tandem but often echo one another in significant ways. A common detail or configuration might appear in each, inviting a reflective comparison between the events or protagonists involved. As French medievalist Eugene Vinaver explains, the discursive method of this literary form consisted ‘less in explaining the action in so many words than in forging significant and tangible links between originally independent episodes … it aimed at establishing, or at least suggesting, relationships between hitherto unrelated themes’vii.

Celticist and previous Mabinogi translator Patrick K. Ford describes these relationships as vertical correspondences, which, he suggests are instrumental to the literary method of the Four Branches:

It is important to insist that [the events in the Mabinogi] must be read vertically as well as horizontally or sequentially … [meaning is] derived from a complex utterances, manifold as it were … a single message stated over and over again in a variety of ways…viii

In other words the articulation of meaning in the Mabinogi is a cumulative process: with objects, characters and places acquiring increasing significance through duplication, inversion or juxtaposition. Understanding this literary method provides the key to the riddle of the interwoven narrative pathways of the Four Branches.

i The parallels between Irish, Welsh and Norse mythologies have been explored by myself (Will Parker) in The Four Branches of the Mabinogi (California: 2005) pp.465 ff. Here the parallels between Lug, Lleu and Loki on one hand and Balor and Bal(d)ur on the other are explored. Further parallels can be discerned between Gwydion and the northern Odin – both of whom occupy the role of the trickster/shaman figure at various points in their respective narratives. These shared elements point to the possibility of a common origin in network of prehistoric cult forms, associated perhaps with the spread of proto-Indo European cultures across northern Europe during the later Neolithic.
ii Koch, John. ‘Bran: Brennos: An instance of Early Gallo-Brittonic History and Mythology’ Cambridge Medieval Celtic Studies 20 (Winter 1990) pp. 1-20
iiiFor more on this see Parker (2005) pp. 283-284
iv Historical typology of this kind is a recognised function of certain genres of medieval Irish writing, as has been recently discussed by Erich Poppe in Reconstructing Medieval Irish Literary Theory: The lesson of ‘Airec Menman Uraird maic Coise’ Cambrian Medieval Celtic Studies 37 (1999); however Sims Williams urges caution with this approach to medieval Celtic literary texts in his article 'Historical need and literary narrative : a caveat from ninth-century Wales'. Welsh History Review , 17 (1994-5), 1-40 highlighting the perils of spurious identification between literary figures and historical personages.
v For both of these parallels, see Lewis, Saunders. Meistri’r Canrifoedd: Ysgrifau gan Saunders Lewis ed. R. Geraint Gruffydd (Cardiff: s. n., 1973)
vi Parker, William. ‘Gwynedd, Ceredigion and the Political Geography of the Mabinogi’ The National Library of Wales Journal Volume XXXII Number 4 (2002) pp. 365-396
viiThe Rise of Romance (Oxford University Press) Oxford: 1971 p.68
viiiP. K. Ford ‘Prolegomena to a Reading of the Mabinogi: “Pwyll” and “Manawydan”’ reprinted in The Mabinogi - A book of essays (ed. C. W. Sullivan III) New York: Garland Press, 1996