The Heroic Age, Issue 1, Spring/Summer 1999

This section is dedicated to essays that stimulate discussion among early medievalists by supporting or attacking prevailing theories on a variety of topics.

This essay is a review of the contextual sections of John Koch's The Gododdin of Aneirin, published in 1997 by the University of Wales Press. The book has been long-awaited and is likely to be regarded as very controversial. For Arthurian studies it offers for the first time a glimpse of how the allegedly earliest literary reference to Arthur may have appeared in the spoken language of the sixth century Britons.

Koch's entire book is, in fact, a very bold undertaking: an attempt by an eminent scholar of Celtic linguistics to reconstruct from a thirteenth-century Welsh text the sixth- or seventh-century precursor which he believes to be the original "ur-Text" of the Gododdin poems. In a recent review of this book, Oliver Padel wrote that "it was right for the attempt to be made, and Koch was the right scholar to do it: linguistically we feel in safe hands" (Padel 1998:45). There is every reason to trust Padel's judgment: a glance at the list of Koch's earlier published work in the bibliography of The Gododdin of Aneirin provides ample testimony of his expertise in this field. Among the list are several key papers which with hindsight may be regarded as the building-blocks for this latest book. Prepublication copies of two of these papers were sent by Koch to A. O. H. Jarman who, in commenting upon their significance in his own edition and translation of Y Gododdin in 1988, neatly summarized the methodology which fashioned the first pieces of the jigsaw: "[Koch] employs the techniques of historical philology to reconstruct representative passages of Y Gododdin in the form in which it may be conjectured that they were composed in the sixth or seventh centuries" (Jarman 1988:xi). Koch's methodology is based on his conviction that it is possible to unravel the layers of linguistic change until the original language of the poet can be identified, reconstructed and restored to the poems. Given his experience in the field of Celtic philology, we should not be in any doubt that the linguistic aspects and elements of The Gododdin of Aneirin are based on scholarship of the highest standard. This review will seek rather to examine the superstructure of reconstructed history which Koch has built upon his "ur-Text" of the poems.

Koch's radical interpretation of the poetry demanded a credible historical context to sustain it. His proposal that the poems' central event, the Battle of Catraeth, should be dated to c.570 rather than to the conventional date of c. 600 necessarily pushes the key personnel backwards by some three decades. These personages are the named heroes of the Brittonic kingdom of Gododdin who, with various warriors drawn from other places, rode out from the kingdom's heartland around Edinburgh to engage an enemy army at a site or district called Catraeth. Apparent references to Anglo-Saxon foes from the Northern English territories of Deira and Bernicia have persuaded generations of scholars that the battle was an archetypal Celts-versus-Germans conflict. In arguing against this interethnic view of early medieval warfare on mainland Britain, Koch is probably right: ethnicity is in any case currently regarded as less of a factor in sixth- and seventh-century Insular wars than issues of overlordship and prestige (Stancliffe 1995:38; James 1989:47-8).

I am less persuaded by his attempt to equate Canu Taliesin's Battle of Gwen Ystrat with the Battle of Catraeth (Koch 1997:xxvii-xxix). The former event is depicted as a victory by Urien Rheged leading the men of Catraeth. The battle is not, however, placed at Catraeth and seems rather to have occurred at a place called Llech Wen in the valley of Gwen, which defies identification. In the poem, Urien is described as an adversary of gwyr Prydein ("the men of Britain"; Williams 1975:31; Pennar 1988:51) but Koch amends Prydein to Prydyn ("Pictland") and identifies Urien's enemies at Gwen Ystrat as a mixed Pictish-Gododdin force. He suggests that this is the same force which the Gododdin poems describe as fighting at Catraeth (Koch 1997:xxix). It is true that there are men in the Gododdin army at Catraeth who appear to hail from Pictland, but no Gododdin warriors appear in Taliesin's poem on Gwen Ystrat. The synthesis of the two battles does not, in my view, stand up to close scrutiny, nor does it add any clarity to our understanding of the relevant poetry. I can see nothing in Koch's argument that would dissuade me from continuing to believe that the two events were largely unconnected.

The Old Welsh heroic poems in any case hardly lend themselves to this type of synthesis, and neither Canu Taliesin nor the Book of Aneirin are able to provide the kind of authentic historical data which such an exercise requires. The pan-British "alliance" which Urien is often assumed to have mustered for an attack upon the northern English seems to have involved the military forces of Rheged, Strathclyde and two unidentifiable kingdoms, one of which could have been Gododdin (Miller 1975:265). There is no warrant for Koch's statement that the alliance contained no forces from the latter kingdom (Koch 1997:xxix).

Beneath the synthesis, however, Koch presents a number of interesting and thought-provoking ideas. He is clearly aware that secular politics and ecclesiastical ceremony all too often rode upon each other's backs. Thus does he wonder whether the alleged baptism of Edwin by a son of Urien Rheged may imply that the British kingdom did not support the Welsh Cadwallon's campaigns against Edwin of Deira (Koch 1997:xxxvi). However, the idea is unfortunately bundled with an unwise denial of Bede's testimony that it was Paulinus, not Rhun ap Urien, who conducted the baptisms of Edwin's people in the River Swale (Bede 2.14; McClure and Collins 1994:98).¹ Koch ignores alternative scenarios, such as the idea of Rhun baptising Edwin while the latter wandered in exile during his youth, or the suggestion that both Rhun and Paulinus may have officiated at the same ceremony.

Koch is on firmer ground when he suggests that, during a long editorial process in Wales, the Gododdin poems may have suffered the excision of references to conflict between the Northern Britons and their Scottish and Pictish neighbors in the interests of anti-English propaganda (Koch 1997:xxxv). Employing a useful analogy, he draws our attention to the tenth-century Old Welsh poem Armes Prydein, which likewise displays antipathy towards the English, and whose creators seemingly "viewed England as the only enemy" (Koch 1997:xxxv). The Armes is a poem of prophecy which looked forward to the expulsion of the English from Britain at a time when the Celtic kingdoms and their Norse neighbors were attempting to resist the domination of powerful West Saxon kings like Athelstan (Williams 1972:v-vi).

It is a pity, in such a detailed and radical examination of the historical background to the poetry, that Koch so often adheres to one particular viewpoint without consideration of alternatives. He accepts as authentic, for example, the Historia Brittonum's claim that the Mercian king Penda accepted a huge payment from Oswiu of Bernicia as the latter stood at bay on the northern extremity of his kingdom. While this claim is not incredible, the alternative scenario offered by Bede—in which the payment is offered but refused—derives from the pen of a writer whose testimony is usually preferred to that of the Welsh pseudo-histories. Koch regards as equally certain the traditional identification of Catraeth as Catterick and makes no reference to the very real doubts which have been cast upon the notion in recent years.² Lack of corroborative evidence likewise renders the idea that the British warlord Gwallawg was the king of Elmet a tenuous hypothesis, yet Koch and many other scholars continue to regard it as a historical fact. The hypothesis rests on two pieces of information: a line in the Taliesin poetry which calls Gwallawg a "judge" in Elmet and a triad which mentions "Ceretic, son of Gwallawg". Ceretic has been equated with Cerdic, a ruler of Elmet mentioned in chapter 63 of the Historia Brittonum (Morris 1980:38). The line of poetry was, however, regarded as doubtful by Kenneth Jackson (1963:31) while it may be noted that the name Ceretic/Cerdic was hardly uncommon in the sixth and seventh centuries.

Koch is incorrect when he says that it is "stated outright" (Koch 1997:xxxi) in one of Taliesin's poems that Urien ruled the territory called Aeron. It is possible from the wording of the poem to draw a slight inference to this effect, but nothing more. It is likewise only a hypothesis—although its constant repetition by generations of scholars has elevated it to something of a consensus—that Rheged extended from Galloway to Catterick.³ The stark truth is that we still cannot define with confidence the heartland of Rheged, still less its frontier provinces, and we cannot therefore weave with unchecked enthusiasm our hypotheses concerning the kingdom's foreign policy.

There are some minor errors or misunderstandings in relation to places and territories. The battle which the Welsh Annals place at Maes Cogwy corresponds to Bede's Maserfelth, not to the place near Hexham which Bede calls Hefenfelth (Koch 1997:xxi, n.4; Stancliffe 1995:84-5). Nor is it accurate to state that Elmet lay in "what is now South Yorkshire" (Koch 1997:xxiii): much of the kingdom, in so far as its extent can be guessed, clearly lay in what is now the county of West Yorkshire (Jones 1975). The heartland of Strathclyde was not Clydesdale, as Koch suggests, but Dumbartonshire, while any detailed physical map will show that the River Carron is anything but "the most direct route from Dal Riata to Manau Guotodin" (Koch 1997:lix).

Koch, like John Morris before him, is keen to fill the gaps in this period's very incomplete record.⁴ Territories that are not mentioned in the surviving poetry or other sources represent such gaps and there is little to be gained by pretending otherwise. Post-Roman Lancashire, for instance, is completely absent from the documentary record and therefore is not "assignable to the Kynferching" (Koch 1997:xxxi) nor to any other North British dynasty.

Koch has a good grasp of wider political issues, such as overlordship, and there is much to commend his idea that the action at Catraeth was an important battle upon whose outcome several matters of overkingship may have rested (Koch 1997:xxxi, n.4). If we could be more certain that the event was historical, we might feel justified in viewing the battle as a clash of ambitious rival kings, from which the Gododdin dynasty perhaps emerged with its military resources depleted and its political status reduced.

Koch is likewise correct in observing that "the sides at Catraeth were already drawn along dynastic and territorial, rather than ethnic and national lines" (Koch 1997:xli). This is a good point and should perhaps be appended to all modern analyses of battles in this period. As noted above, Koch debunks the outmoded "ethnic" view of early medieval Insular warfare and reminds us that it is a fallacy to imagine that "a monolithic Brittonic nation resisted the advance of a comparable monolithic Germanic invader" (Koch 1997:xliv). Much of what Koch writes about the political circumstances around 642 is perceptive and well-grounded. Seldom has there been a more apt description of the defeat-prone Dalriadan ruler Domnall Brecc than Koch's "erratically aggressive king" (Koch 1997:lix). As Koch observes, the coincidence of the deaths in 642 of the Irish king Domnall mac Aed and of the Northumbrian Oswald—two dominant and powerful figures—can only have rekindled Domnall Brecc's desire for gain (Koch 1997:lix). The latter's defeat and death at the hands of Owein, king of the Strathclyde Britons, in December of 642 must have greatly strengthened the British ruler's position: Koch (1997:lviii, n.4) notes that Owein's authority would have increased, with his nearest capable rival—Oswald's younger brother, Oswiu of Bernicia—having been harried and browbeaten by the Mercians. He also points out that it is possible to infer that Owein joined the Welsh kings of Gwynedd and Powys in the Mercians' northern wars between Oswald's death and Oswiu's decisive victory over Mercia in 655 (Koch 1997:xcii). Oswiu must indeed have been, as Koch suggests (Koch 1997:187), a likely participant at the presumed Bernician siege of the Gododdin citadel on Edinburgh Castle Rock in 638: he was in his mid-twenties and would probably have seen military service in his elder brother's campaigns.⁵

Some of the quasi-legendary material associated with the fifth and sixth centuries receives due comment in The Gododdin of Aneirin. The foundation story of Gwynedd and the tale of its founder, Cunedda of Gododdin, is seen by Koch as useful propaganda created by the Welsh during their participation in the Mercians' ravaging of Northumbria in 633-4 and 651-5 (Koch 1997:xcix). Arthur, too, is examined, in the light of the Gododdin verse which mentions him by name. Interestingly, Koch (1997:148) concludes that this verse should be dated to the poems' early period of 570 to 638 and sees it as supportive of Arthur's historicity. In committing himself to so early a date for the reference to Arthur, Koch (1997:148) challenges the sceptics among us to accept either one of two conditions: that the traditional Arthurian chronology of c. 500 is a construct, or that the hero's nonexistence demands that the sixth-century poets must have invented him and placed his activities within living memory.

My overall impression is that Koch's synthesis of the incomplete historical data for this period is interesting and often original but is rather over-elaborate. He accepts too readily various hypotheses as if they are hard historical facts and attaches too much credence and not enough caution to the myriad fragmentary sources. Many of his own ideas consequently rest on shaky foundations. I would, however, recommend that the "context" he suggests should nevertheless be required reading for those whose interest lies in the interaction of early medieval polities in Northern Britain. John Koch has long been an important contributor to our understanding of this troubled period and this book provides the historian or the archaeologist, as well as the linguist, with a major Celticist's interpretation of the background to one of the period's most enigmatic sources. Regardless of an individual reviewer's opinion of the historical value of the Gododdin, or of Koch's edition of it, neither the source nor the editor are of negligible stature within modern scholarship, and neither of them should be ignored.