"And where does the Oxford Arthurian Society think Arthur came from?" In Caerleon on a family visit, I had hardly expected to be subject to an interrogation by a group of keen Welsh Arthurians, and made a non-commital reply on the diversity of interest and belief in the Society, and on my own open mind on the subject. My interrogators had no such doubts. "Arthur existed and he saved the Welsh", said one, while another, with an emotional flourish, proclaimed "When I walk through Caerleon early in the morning I know in my heart that this is Camelot."
It would be nice to have such conviction, but Caerleon is not the only site to have aroused such strong Arthurian feelings. The vast number of sites attributed to Arthur in the regions is almost rivalled by the theories of modern historians, archaeologists and pseudo-scholars. Geoffrey Ashe is a strong advocate of the claims of South Cadbury as Camelot and Glastonbury as Arthur's final resting place, while the authors of Arthur, the True Story prefer Wroxeter and The Berth at Baschurch. Michael Wood suggests that Carlisle, which had a flourishing sub-Roman culture until the seventh century, is the place to look for a real Arthur, and that Camboglanna, now called Birdoswald, on Hadrian's Wall is the original of Camlan. I do not intend to enter the fray and come to any conclusions on where Arthur came from. Indeed, anyone who does come to positive conclusions about such a notoriously obscure period of history is treading on dangerous ground. What I do intend to do is to raise some points which might have some bearing on the question of where, and when, the real King Arthur existed.
Regrettably, even more so than in other periods of history, there are no sources that can be accepted uncritically, and any study of the Arthurian period must start with examining these sources. There are a few written sources which describe the "Dark Ages" between the formal severance of links with Rome, in 410, and the resumption of those links when Papal missionaries came to the Saxons in 597. Those sources that we have are never originals but are copies, or even copies of copies of copies. Anyone who has ever edited a society magazine will know how, in the copying of an article, small mistakes invariably slip in. Usually this does not alter the sense, but in a highly inflected language such as Latin the changing of one letter can materially alter the meaning of a sentence. A possible such error is found in the Annals of Wales, written in the tenth century but quite possibly copied from contemporary reports written in the margins of tables for calculating Easter. Its entry about the Battle of Badon claims that Arthur carried Christ's cross on his shoulder for three days, but it is probable that the word "shoulder" should really be "shield", due to confusion between the Welsh words "scuid" and "scuit".
Another problem of copied sources will also be understood by any editor who has been presented with an article written in what they consider to be bad grammar. It is very difficult in such circumstances to resist the temptation of correcting the original, and this temptation was shared by madiaeval copyists. It has been suggested that at least two Arthurian sources were elaborated by later copyists who, writing when legends about Arthur's exploits had already developed, thought they knew better than the text they were copying. One of these is the Annals of Wales. Its two Arthurian entries, on Badon and Camlan, are uncharacteristically loquacious in a very terse set of annals, whose entries, which are non-existent for many years, are usually one or two words, compared with the four lines on Badon. The other is the battle list in the collection of documents usually called Nennius's History of the Britons, our present copy of which dates from the late tenth century. Arthur has already entered the realm of romance, described as slaying 960 enemies personally in one battle. Neither of these need be later elaborations. Superhuman exploits of heroes are a commonplace of heroic poetry, even those written to describe contemporary events, and there are a few other lengthy entries in the Annals.
However, it must always be borne in mind that seemingly early texts could contain later additions. This is especially the case with Welsh poetry. Poems attributed to Taliesin and his contemporaries contain occasional references to Arthur, such as Y Gododdin, generally dated to around 600, which mentions him in passing as an exemplar of martial valour. These poems, accepted uncritically by Victorian editors, now have their credibility so undermined by modern scholarship that they can not be used as evidence. Some scholars go as far as to suggest that they were the product of a self-conscious literary revival in Wales in response to the Norman Conquest.
Errors in copying are not the only problems with the written evidence. Gildas's De Excidio Britanniae, despite its plainly erroneous historical section, is a key source simply because it is the only one which is definitely contemporary, with the date of its composition, while debatable, usually being put around 540. In it Gildas describes how a powerful ruler summoned Saxon help against his enemies, only to find that the Saxons expanded aggressively. The Britons fought back under Ambrosius Aurelianus, and had a series of victories which culminated in Badon, at around the time of Gildas's birth. Although the peace brought about by Badon still lasts, the current rulers are so sinful that Gildas doubts it will continue. The work is thus not a straight history, but a didactic diatribe against the sins of Gildas's contemporaries. His facts should therefore be treated with caution, although it is reasonable to assume that he would have been accurate, if biased, about events in living memory or he would have had no credibility with the readers he was trying to convince.
In contrast with the documentary sources, it might have been thought that archaeological evidence is relatively reliable. However, it too presents serious problems, mainly due to the very haphazard discovery of artefacts and sites. James Campbell points out that, if archaeological evidence is believed, England had 20 mills at the time of the Norman Conquest, but the Domesday Book shows over 6000. The Dark Ages have no Domesday Book, and so all too often inferences about whole societies have to be made on one piece of pottery or one grave. No known evidence of settlement can not be taken to mean no settlement, when there could be huge amounts of unpreserved or undiscovered evidence. The palatial rebuilding of Wroxeter in the Dark Ages has only been discovered because the site was not later built upon, but there could be many other similar sites unretrievable because of later building. Often all that archaeologists have to go by are graves, which at this time are almost impossible to date, and which prove little about the rule of an area and its political events.
Thus, the sources for the Arthurian period are few and far between and must be treated with scepticism. Out of this tangle, is it possible to come to any conclusions about where Arthur came from? Wherever he came from, he certainly didn't come from "Camelot", which was apparently the creation of Chretien de Troyes in the thirteenth century. Culhwch and Olwen locates his stronghold as Kelli Wig in Cornwall (perhaps Kilbury), which is also mentioned in the Triads as one of the three tribal thrones of Britain. The "Life" of St Carannog has the saint meeting Arthur at his court at Dunster, and the romances in the Mabinogion place him at Caerleon. Malory favoured Winchester, while today sees South Cadbury as the most favoured spot. However, all these attempts to locate "Arthur's Hall" appear to ignore basic realities of kingship. Throughout the Middle Ages, English kings remained peripatetic, taking their courts, scribes and even Parliaments with them as they travelled. One of the first responses to a rebellion would be the personal visit of the king to the troubled area, just as many people with local grievances would take them direct to the king. In an age before telephones, widespread literacy or bureaucracy, rule was personal. In the stressed time of the Saxon invasions, no king could afford to stay in one place but would have to travel around, keeping loyalties and leading defence. In the early Saxon kingdom of Northumbria a vast hall has been found at Yeavering, although this is only 20 miles from the equally large Bamburgh. Both, it is claimed, were seats of King Edwin. And so it would have been with any British king in the Dark Ages. There could be many Camelots across Britain.
While it is impossible to say that Arthur "came from" one particular place, is it possible to limit him to a region, like the man in Caerleon who claimed Arthur was Welsh? To claim Arthur as Welsh, or Cornish, is in fact an anachronism. Modern Cornwall, Wales, southern Scotland, and indeed all non-Saxon England was "Welsh" in that the Brythonic language spoken was the ancestor of Modern Welsh. The dynasty of Gwynedd was reputedly founded around 450 by Cunedda and the Votadini from what is now southern Scotland, and the poem Y Gododdin, describing a battle involving a tribe from around Edinburgh, has been published as part of the Four Ancient Books of Wales. Taliessin wrote poems about the northern king of Rheged, but also one about a king of Powys, and Gildas speaks of "our Countrymen", not "Cornish" or "Welsh". Arthur could have lived all his life in Sussex but still be called "Welsh", both because he was a Briton and because his legend would later be written in what is now Wales. As the Britons were pushed further and further into the west it is inevitable that their stories and legends would go with them. Events which had nothing to do with Wales were thus written down, some centuries later, in Wales, simply because that was then the home of the people who told the stories.
The most promising way to track down a real Arthur is to use logical deduction from the historical situation at the time - a thing many theorists have singularly failed to do when they talk of Saxon mercenaries summoned to Kent to fight Irish raiders on the Welsh coast, or Arthur fighting Saxons in places whose names sound rather like Badon, but are in places the Saxons most definitely were nowhere near at the time. It is therefore necessary to establish precisely what time we are talking about. The Annals of Wales give dates for both Badon and Camlan. Even those who believe that these entries were elaborated can not find any reason to believe that they were entirely fabricated. The historical accuracy of the Annals, when it can be checked, is generally sound, so it is reasonable to assume battles, at Badon and Camlan, did take place when the Annals say they did. For complicated reasons, the dates in the Annals could be 28 years out. Thus Badon was in 490 or 518 and Camlan in 511 or 539. Leslie Alcock argues in favour of the first set of dates, which fit in better wit the archaeological evidence and with Gildas, but a dating to within 28 years is in fact uncommonly accurate in the notoriously hard-to-date Dark Ages, so it suffices to date Badon as 500, give or take several years.
At first sight this dating raises a problem, which is why, if Arthur was active at this time, does Gildas not mention him, and even suggests that there have been no worthy leaders since Ambrosius. However, names are few and far between in Gildas. He names Ambrosius, who had the first victory over the Saxons, and he names the five kings whose sins he castigates. Everyone else remains nameless. He does not name the victor at Badon, although, presumably this name would be common knowledge. As one scholar has said, no politician in the 1850s would feel the need, when referring to Waterloo, to add the name of its victor, Wellington. Moreover, there is evidence that Gildas might not have considered Arthur a hero. Gildas was a monk, and monks had peculiar attitudes to heroes. In the saints' lives written at Llancarfan, in South Wales, Arthur appears as a boorish warlord who is constantly bested by the saints' piety. He tries to steal St Padarn's tunic, appropriates St Carannog's floating altar, and is only just dissuaded by his followers from assaulting St Cadoc's mother, a damsel in distress. Charlemagne's grandfather, Charles Martel, beat the Saracens in one of the most important battles in early mediaeval Christendom, but monks could not find a good word to say about him as he appropriated their land to finance his campaign. It could easily be the same with Arthur. It is entirely possible that he took church land, and any warrior king of that age would have to win followers with luxurious hospitality rather than monkish piety. The Vorteporix Stone, now in Carmarthen, describes Vorteporix as "protector" of Dyfed, possibly a hereditary title, yet Gildas dencounces him as an illegitimate ruler, guilty of adultery, luxury and the like. Thus, Gildas's claim that, since Ambrosius, British leaders have been sadly debased need not be taken as evidence against Arthur being one of Ambrosius's successors. In short, Gildas does not provide any evidence against Arthur as a ruler active in the early years of the sixth century.
Other evidence supports this dating. In 577 the battle of Dyrham separated the British in Wales from the British in the West Country, just as the Battle of Denisesburna in 634 separated Wales from the British of southern Scotland. It is not clear how much these Saxon advances sundered the British, as Saxon settlement was initially very thin in these areas, but local similarity between Arthurian local legends in Cornwall and Wales suggests that at least some of the legends were formed before Dyrham hindered communication. Much depends on the antiquity of local Arthurian legends. Some are undoubtedly modern, such as when a grassed-over seventeenth century knot garden in Scotland is called Arthur's Round Table. Until recently, I had considered most features in the Arthurian landscape to be the product of mediaeval story-telling, after the boom in Arthurian legends in the twelfth century. However, this appears unlikely. Even after Layamon translated the legends into English, few ordinary people would have had access to them. Arthur seldom appears in mediaeval folk-song, and when he does it is in a role which does not resemble his legendary one. Moreover, if they were mediaeval it would be expected that features of the Arthurian landscape would be spread across England, but in fact they are singularly lacking in the east of England. Indeed, their occurence is almost completely confined to the areas of the country that the British still possessed in around 550, a generation after Badon. There is no firm evidence for the antiquity of these sites, but Nennius mentions two in Wales, and two in Cornwall were encountered by a party of clerics from Laon in 1113, before Geoffrey of Monmouth wrote his work.
It is never wise to make sweeping statements about folklore, for much of it is of modern origin, but the incidence of sites attributed to Arthur in folklore does suggest that stories about him were being developed even before 577. It has been pointed out that there are several known British Arthurs after 550, including a prince of Argyll, and none before, apart from King Arthur himself, suggesting that some famous Arthur was possibly being honoured.
So, if King Arthur is taken as being a leader of the British who was active around 500, can anything be said about where he lived? Gildas claims that Badon was a great victory against the Saxons, and it is Arthur who from early tradition has been credited with this victory. Archaeology shows that there was some significant check to the Saxon advance at around this time. In Essex, Hertfordshire and Sussex Saxon pottery from the fifth and seventh centuries has been found, but none from the sixth, and Procupius of Caesarea, writing around 550, records Angles retreating from Britain to the Continent. Between around 500 and 550 the saxon advance had definitely been checked, and hardly extended outside the extreme south-east. Badon must be placed near this area, and it is only sensible to imagine Arthur as being constantly active along this frontier. Clearly, this places no limitations on the location of his strongholds, and archaeology shows that old hillforts around Britain were being fortified at exactly this time. Some 40 refortified sites have been found, from Wiltshire to North Wales, although this has not stopped the over-credulous from claiming any one of them as Camelot simply on the grounds of its refortification. South Cadbury, which appears on road signs as Cadbury-Camelot, had truly massive buildings built around 500, but by 550 its ramparts were collapsing. It was in no way exceptional, and it, and many others, could have been used by Arthur. Similarly, little can be deduced about the location of Camlan, fought, according to the Annals, against Medraut. It was not at all unusual for Britons and Saxons to fight among themselves. In 633 the British king of Gwynedd allied with the Saxon king of Mercia against the Saxon king of Northumbria, and Gildas lamented that, while Badon had put an end to major Saxon problems, the British still suffered from much civil war. all that can be deduced is that Arthur died in one example of such in-fighting, somewhere in Britain.
As the title of this article suggests, I did not set out to come to any conclusions about where Arthur came from, and I have not done so. It is probably safe to assume that Arthur existed. His legend appears to be truly early in its origin, and it is not likely that such a volume of stories arose out of nothing. The Annals, which appear to be a contemporary source, are generally accepted as truthful in most cases, so doubting its Arthur references would be over-scepticism. Sadly, however, over-scepticism is not a characteristic which many Arthurian writers possess. With a few exceptions, such as Leslie Alcock, most adopt an un-historical attitude and find Camelots and Badons at all sorts of places at best unprovable and at worst at odds with the archaeological and historical background. What I have aimed to show is that, above all, it is this historical background which must govern Arthurian research, and not romantic fantasies of, for example, anachronistic nationalism, such as exhibited by the Caerleon Arthurians who inspired me to write this. Sadly, all too often, a historical approach is merely destructive, for it tends to challenge hitherto-accepted sources and truths. We can never say where Arthur lived or fought. The best we can do is to say where he didn't.