by Mike Ashley
Dyfnwal’s main opponents, apart from the Picts and the Scots, would have been the British tribes south of Hadrian’s Wall, including the Gododdin and descendants of Coel, the Coelings. In contrast to Dyfnwal’s descendants, the Coelings of that generation were poor. There was Merchiaun (Mark the Lean), his son Cinmarc the Dismal, and Mark’s cousin Sawyl (Samuel) the Humble. However, Sawyl’s cousin Einion was the father of Rhun the Wealthy, suggesting at least some change of fortune. How much of that might have been due to another cousin, Eliffer of the Great Host, is open to speculation.
Eliffer and his sons ruled the eastern Pennines, and thus were on the front line facing the rising menace of the Angles who, under Soemil, had laid claim to Deira, the area between the Humber and the Tees around the middle of the fifth century. Eliffer’s father, Arthwys, must also have been involved in these battles. There would be few reasons or opportunities for the Coelings to venture north of the Wall unless threatened by Dyfnwal’s sons. They had enough to cope with facing the Angles and the Irish. The pattern of battles in Map 7 is too haphazard to reflect any consistent campaign between the Coelings and the Gododdin, and is far more suggestive of battles between the Gododdin and Strathclyde marking Dyfnwal’s territorial gains.
The epicentre of these battles is around the Eildon Hills and the Roman fort of Trimontium. This had ceased to be occupied by the end of the second century, but the Celtic fort at the top of one of the Eildon Hills showed evidence of reoccupation by the end of the fourth century. In Arthur and the Lost Kingdoms, Alistair Moffat has speculated that a Romano-British cavalry unit under Arthur re-established itself in this area, perhaps first at Eildon but then further east at what became Roxburgh. This unit was primarily engaged against the advancing Angles. Moffat does not identify all of the sites from Nennius’s list, suggesting sufficient in the area to presume that all belong there (though surprisingly he concedes that Badon probably was at Bath). There are certainly enough potential sites here to make this area a distinct possibility, whether it relates to Dyfnwal’s expansionism or a cavalry unit holding back the Angles.
These sites mean that the battles were fought primarily on Gododdin soil, but if at this time the area was under the control of Dyfnwal and his sons, there is no room for anyone called Arthur. The only like-named individual in the region, the Coeling Arthwys, could not have managed a campaign this far north.
It is possible, though, to map out an alternative northern frontier. In From Scythia to Camelot, Linda Malcor and C. Scott Littleton propose a sequence that could have been fought by Lucius Artorius Castus against the Picts, around the year 185. They suggest that the campaign could have started at the fort of Ribchester (Breguoin), then under Castus’s command, and from there along the Ribble to its estuary (Tribruit). The pursuit then moved south to a sequence of battles along the Douglas, before the Caledonii headed across the Pennines to York. Castus drove the Picts north to Binchester (Guinnion) and from there back across Hadrian’s Wall where they met again at Yeavering Bell (River Glein). The Picts were now in full retreat, but Castus engaged them again in the Forest of Celidon before pursuing them north to the final victory at Dumbarton (Badon).
Malcor and Littleton don’t offer a site for Bassas, but a site near Stirling would certainly fit their scenario. There are two concerns, however. Firstly, would this campaign be remembered three centuries later, by then somehow attributed to another whose own battles may have echoed those of Castus? And secondly, how does Badon fit into the timetable when it happened in the year of Gildas’s birth? Badon (as Dumbarton) is the one weak link in an otherwise feasible proposal, and it raises again the possibility that Badon was not part of the same campaign, but the culmination of a series of battles against the Saxons.
Malcor and Littleton’s proposal might also fit an alternative northern campaign. A campaign against the Angles could have started at York, and moved south along the Humber to face them in their heartland near Barton at Glanford Bridge (Glein) and then headed north to Binchester (Guinnion) and up to the Wall at Corbridge (one or more of the Dubglas battles). It would not need much of a trespass beyond the Wall to engage the Angles again in the Forest of Celidon. This may have marked the end of one campaign and could have been led by Eliffer of the Great Host from York, perhaps assisted by his father Arthwys. The threat then shifted to the West Coast with the Irish raiders. Battles could have taken place at Ribchester (Bregouin), along the Ribble Valley (Tribruit) and the River Douglas (thus causing a mental link with the previous campaign). This covers all of the sites except Bassas and Badon. Bassas, which we have suggested means “shallows”, might relate to any of the shoreline in Morecambe Bay which was once above sea level but was later engulfed. And if Badon was a separate battle, and not part of this campaign, it need not have been in the north. This second campaign may well have been led by Arthwys in conjunction with Mark of Rheged.
There are other, later, events that may have caused these battles to become embedded in the Arthurian legend. The battle of Arderydd, for instance, which happened eighty years after Badon, in 573, was between Arthwys’s grandsons, Gwenddoleu, Peredur and Gwrgi. The reasons behind this battle are uncertain but it suggests some climactic vendetta between the two sets of families. The fact that it happened north of Hadrian’s Wall, in the territory of Rhydderch Hael, suggests that Gwenddoleu may have betrayed the Coelings and was now in the pay of the sons of Dyfnwal. Gwenddoleu was known for his wealth as he also had one of the Thirteen Treasures of Britain, a gold and silver chessboard. There is another tradition that states that the battle was between Gwenddoleu and Rhydderch, and that the sons of Eliffer were not involved. This may make more sense, as it suggests that one of the Coelings, who had become wealthy, saw an opportunity to gain territory from a young and as yet untried Rhydderch, with fatal consequences. Gwenddoleu was killed and his bard, Myrddin, went insane after the battle and fled into the Forest of Celidon.
Arderydd may have been remembered generations later rather like Camlann, and its closeness to Camboglanna on Hadrian’s Wall may have compounded the error. What’s more, Arderydd is situated in the Forest of Celidon and thus may be the same as Nennius’s seventh battle. The connection with Myrddin/Merlin adds further confusion. Add to this the memory that Dyfnwal was the Pendragon, and the mix could certainly add fuel to a tradition of a northern Arthur.
Gwenddoleu’s cousin was Urien of Rheged who fought against the Angles in Northumbria during the 570s and 580s, with his cousins Gwallawg of Elmet and Morcant of Lothian. This included a major siege at Lindisfarne where, due to betrayal by Morcant, Urien was murdered. The disaster at Catraeth soon followed, the beginning of the end for northern resistance against the Angles. Urien’s battles are commemorated in Taliesin’s battle poem Arise, Reget, and these may include at least one that coincides with Nennius’s list, the eleventh battle of Breguoin, “the halls of Brewyn.” Otherwise Urien’s battles do not overlap with Arthur’s and since the end result is Arthur’s victory as opposed to Urien’s defeat, it suggests that Urien’s campaign is not the source for the battle list. It could still, though, have blurred in the memory, with one sequence of battles ending in a betrayal becoming confused with another battle campaign which also, if you include Camlann, ends in betrayal. Urien features in the later Arthurian legends, but he stands suffiently bold as a possible Arthur himself.
Eastern Frontier
In chapter 5 I discussed Gildas’s reference to “the unhappy partition with the barbarians” in De Excidio, and suggested that this boundary may have been delineated by the Roman Dere / Ermine Street, the modern-day Al. If we plot the battle sites against this road there is a surprising match. It would need to include a few dubious sites, such as those around Stirling, for Tribruit, and that at Stow for Guinnion, but otherwise follows several very feasible locales.
This frontier has the advantage of following Gildas’s partition and of linking together the boundaries of the original Anglo-Saxon settlements in Bernicia, Deira, Lindsey and Eas
t Anglia. Apart from the problem of identifying this boundary around London, it is otherwise supported by archaeological evidence which shows no extensive Saxon settlement west of that frontier until the mid sixth century.
The major problem is that this frontier runs the length of Britain and thus crosses several British kingdoms. Unlike the other frontiers, which are relatively self-contained and thus could be held by one chieftain, perhaps with the help of neighbouring kings, this one would require a significant consortium of kings. Yet this is exactly how Nennius describes it when he says that “Arthur, with the kings of Britain” fought the enemy. The individual best placed to work along this line would have been Arthwys of the Pennines since his territory, probably around the southern Pennines, was midway along this route. We have established that the original battles, under Vortimer, took place along the Humber estuary and this would be the obvious place for the battles to start around Brigg and York, locations also favoured by Geoffrey of Monmouth (see Chapter 9). There would have been a series of battles in the north, where the Angles were most strongly established, before other battles, doubtless encouraged by Aelle, erupted further south. Arthwys would have been reinforced here by the British rulers of Calchvynydd and Rhydychen, the territories around Oxford and the Chilterns, and it’s possible that other battle sites with names now lost ought to be placed in this vicinity.
Badon need not be part of this frontier. The final battle could have marked a last ditch effort by the Saxons to force the British back. It could suggest a pincer movement by Aelle from the south plus the Saxons of Lindsey and East Anglia, perhaps under Octa (who Nennius tells us came south) who came down the Ridgeway or the Fosse Way into the British heartland. That defeat proved to the Saxons that they could not advance so far west and allowed the partition.
Southern Frontier
This frontier could have been controlled by a commander based at any of several major forts, not only those we have already discussed, including Caerleon in Gwent, Solsbury Hill near Bath, or Liddington Castle, near Swindon, but other strengthened hillforts such as South Cadbury, Glastonbury, and Cadbury-Congresbury near Bristol, the first two with close Arthurian associations. To this we must add the Wansdyke and other defensive earthworks in the area which are the only firm signs of a genuine frontier. We can, to a degree, place the battle list along or close to this frontier and with connections to some of the known conflicts with the Saxons.
The first of the battles may well have taken place at Glynde-bourne, near Lewes (Glein), defeating Aelle before moving along the coast to face another Saxon advance (attributed to Cerdic, but probably one of the other eponymous adventurers, such as Port or Wihtgar) at Blackwater (Dubglas) near the Solent estuary. Several engagements could have happened here allowing for a series of Saxon landings. The British may then have retrenched at Old Basing (Bassas), which formed the start of a defence across the south as Saxons now advanced from the north. Another battle at White Hills (Guinnion), Swindon, held the Saxons at bay and this may have ended the first series of assaults.
The second campaign may have started with some daring Saxon or Irish raids up the Severn and into south Wales, leading to battles at Pontardulais, Gellideg (Celidon), Caerleon (City of the Legion) and the Trwc estuary (Tribruit), before pushing the Saxons back across the Severn to Catbrain (Bristol) and the final engagement at Badon (either Bath or Liddington Castle).
This campaign could have started under Ambrosius, being perhaps the end of a much longer campaign by Ambrosius based at Cadbury. A young Arthur may have been in his ranks. Arthur could have taken over as commander of the second campaign. If so, this might suggest we would find some mention of Ambrosius rewarding Arthur as he hands over command. Although Pascent received lands from Ambrosius, as they were probably of the same generation, Pascent is unlikely to have taken over as commander. But Pascent’s son is another matter. This was Riocatus, whom we have encountered before as Faustus’s nephew.
Riocatus’s name, meaning “king of battles”, might imply that he became Ambrosius’s military successor. Riocatus need not have succeeded his father by then; indeed, that may not have happened until around the year 500, after Badon.
Riocatus was the cousin of Cadell of Powys, known as “Gleaming Hilt”, a strange cognomen which seems to suggest some gloriously decked hilt or scabbard to his sword. This is reminiscent of Arthur’s sword Excalibur, because it was the scabbard, rather than the sword, that had magical qualities and would protect Arthur.
One might expect Ambrosius to be succeeded by his own son rather than Cadell, but although we know from Gildas that he had grandchildren, these may have been via a daughter. Alternatively, if Ambrosius was Riothamus, then his children may have lived in Armorica and inherited lands there.
Cadell of the Bright Sword and Riocatus, King of Battles, as joint successors to Ambrosius ruling from the one city in Britain that had not fallen into ruin – Viriconium – could well have become conjoined in later years as a legendary Arthur. If either of these, fighting under the banner of Ambrosius, led an army to Badon, with the final defeat of the Saxons, that would be enough to imprint that memory indelibly into the folklore of the British.
This is all highly conjectural and based on the flimsiest of evidence, but in the Arthurian world there is little else. This suggestion does fit a pattern of battles, and does provide a locale for a possible Arthur-like figure.
There are, of course, many other interpretations of these battle sites, both in terms of new locations or how the battle sequence may have run across the country. In King Arthur, A Military History, Michael Holmes discusses the Anglo-British battles and where Arthur’s campaigns might be located. He generally follows the more traditional locations (Glen in Lincolnshire, Bath for Badon and so on) and allows Arthur free rein across the whole of Britain but in a series of battles spread over several years. He accepts Arthur as the High King successor to Ambrosius, but does not otherwise identify him beyond recognizing him as a great military commander.
Conclusion
Having explored all of the battles and dozens of sites we have been able to make a potential link between some sites and some individuals. Only one of these, Arthwys of the Pennines, has a name which may be resonant with Arthur, but we are, after all, looking as much for the victor of the key battle of Badon, whose memory may have become attached to a later Arthur. This has helped us identify several individuals, especially Dyfnwal, Cadell and Riocatus, who must have been alive at the time of Badon and probably fought there.
Each of the suggested frontiers has its strengths and weaknesses, though the patterns in North Wales and North Britain are not best placed for sustained campaigns against the Saxons. Neither is the western frontier, for all that a campaign could have been masterminded from Wroxeter. My own belief is that only the southern or eastern frontiers provide a plausible explanation for a sustained battle campaign. The eastern frontier has the advantage of linking to a known subsequent “partition” and allows for the likely presence of Arthwys of the Pennines. The southern frontier presents a better explanation for the hill forts in the south and a focused campaign against Aelle as Bretwalda.
Before we take this further, though, we need to remember that the battle list does not cover all of Arthur’s exploits. For a more complete picture we must turn to the Welsh tales.
8
THE WELSH TRADITION – THE OTHER ARTHURS
So far, apart from the Welsh Annals, all of our information about Arthur has come from writings by English or Continental authors. When we turn to the Welsh record, a different Arthur emerges. In this section we will explore the relevant stories in the Mabinogion and other Welsh texts.
1. The Mabinogion
It may seem strange to include discussion of the Mabinogion as part of this “historical” section, and it is true that the stories do cross the divide, being more legend than fact. But, as we shall see throughout these explorations, there are factual elements, and the divide could be drawn almost anywher
e. It is worth reflecting upon a comment by Gwyn Jones and Thomas Jones in their translation of the Mabinogion (Dent, 1949):
. . . when we recall that Arthur was not a French, German or English, but a British king, it is not unreasonable to emphasize the significance of British material relating to him. British material, that is, uncontaminated by the Cycles of Romance, though necessarily affected by the vast complex of Celtic myth and legend.
The Celtic tales of Arthur incorporated into the Mabinogion are amongst the earliest to survive, certainly predating Geoffrey’s History, though not all necessarily surviving in written form from an earlier date.
The Mabinogion is a collection of Celtic tales, edited by Lady Charlotte Guest, with the help of Ioan Tegis who helped transcribe them into English, in 1846. She incorporated twelve in her first edition, although technically only the first four belong to the “Mabinogi”, the stories about the hero Pryderi. It was Lady Charlotte who concocted the phrase “mabinogion” on a misunderstanding of the text. “Mab” means son, and the phrase is generally taken to mean “tales of youth”. It has become a convenient tag for a collection of early Celtic tales, and so it will remain.
The stories incorporated into the Mabinogion come from two ancient collections, Llyfr Gwyn Rhydderch (The White Book of Rhydderch) which was committed to parchment in the early 1300s, and Llyfr Coch Hergest (The Red Book of Hergest), which was written down around 1400. Other versions of these stories survive in manuscript form from at least a century earlier, and were clearly part of an oral tradition long before that. But, as with the sources for Nennius and Geoffrey, since we lack the earliest versions we have no way of knowing how much they have been corrupted in the seven centuries since Arthur’s day.