Geoffrey Ashe first presented his ideas on the origins of the Arthurian legend and the identity of the historical Arthur in the April, 1981 issue of "Speculum." The following article, which first appeared in the Fall, 1995 issue of "Arthuriana," is re-published with the permission of the author and is a fresh presentation of those ideas.
......................................................................

After prolonged debate, the search for the 'historical Arthur' remains inconclusive, because of the nature of the evidence which historians take into account. Possibilities arise, however, from evidence of another kind. Literary inquiry can lead towards historical insight and identify an Arthur figure who has been noticed at various times, but not adequately considered.

To the question 'Did Arthur exist?' a straight yes-or-no answer cannot he given. More is involved here than historical doubt. With, say, Robin Hood, the straight answer is likewise excluded, but solely by insufficiency of data. A new find might some day make it possible. With Arthur the difficulty cuts deeper. For any ordinary inquirer, the answer 'yes' implies the reality of the Arthur of romance, the idealized medieval monarch, at the centre of a sort of montage that includes Guinevere and Merlin and the Knights of the Round Table. Since Arthur in that sense is a literary creation and didn't exist, the answer 'yes' is wrong. But the answer 'no' is also wrong. It implies that Arthur is fictitious as Don Quixote is fictitious, that he has no factual basis at all. The romancers themselves would never have accepted that, and it cannot be maintained as a definite statement.

Actually, of course, the literary Arthur is a shape-shifter who has taken different forms over the centuries. But all versions presumably derive from a source or prototype earlier than any. There have been numerous attempts to work back to this point, and, more specifically, to pin down a 'historical Arthur' as the starting-point, so that the question of existence can be affirmatively answered. . .on the understanding that this is the Arthur who is meant.

I believe the 'historical Arthur' quest has, in practice, been misguided. Historians in search of him have committed themselves to a certain mode of approach. They have tried to strip away legend and isolate hard evidence. Doing so means dismissing the medieval literature (Geoffrey of Monmouth and everything later), sifting older matter of Welsh provenance, and picking out whatever may be deemed factual or, at least, arguably so. Applied with due objectivity, such a process reduces the data to two Latin documents. They refer to Arthur at no great length as a successful war-leader of Celtic Britons in the fifth or sixth century, embroiled chiefly with encroaching Saxons, ancestors of the English. One of these documents is the "Historia Brittonum," History of the Britons, compiled early in the ninth century, and ascribed dubiously to a monk of Bangor named Nennius. In a single chapter it lists twelve Arthurian battles. The other document is a chronicle, the "Annales Cambriae," Annals of Wales, which is somewhat later and has two Arthurian entries, also about battles. There is a penumbra of Welsh poems and traditions, and support for the Latin texts can be claimed from that quarter, especially from an allusion to Arthur's martial prowess which may be as early as 600. They alone, however, are the documents properly so called (1).

Opinion on them has swung back and forth. One seldom-noted fact is that Edward Gibbon believed in Arthur, on the strength of the 'simple and circumstantial testimony of Nennius.' From the 1930s on, Collingwood's theory of an Arthur who revived the imperial military office of Comes British History Clubrum, and employed Roman-type cavalry to rout pedestrian Saxons, appealed to many including novelists such as Charles Williams and C.S. Lewis. While the cavalry notion faded for lack of evidence, the image of Arthur as a post-Roman commander-in-chief, with or without civilian power as well, and as active in south and south-west Britain, flourished into the seventies. It seemed to have established itself through the work of Kenneth Jackson, Leslie Alcock and John Morris (though Alcock's review of Morris's "The Age of Arthur," which made sweeping claims, was critical; there was never a united front). Some scholars, notably Rachel Bromwich, while accepting an Arthur who was primarily a warrior, dissented as to his homeland and made it northern.

In 1977 an onslaught by David Dumville on all such reconstructions, and on Welsh records generally, set the pendulum swinging the other way (2). Today most historians who consider the 'historical Arthur' at all are sceptical and reluctant to discuss him. An added reason has been the partial discrediting of the topic by the appearance of further 'historical Arthur' books which are mutually contradictory, wildly unscholarly, and sometimes worse.

I would agree with the sceptics, not in giving up the procedure entirely, but in seeing it as a dead end unless it is supplemented in other ways. The Latin texts are too distant in time from Arthur's apparent floruit. The list of battles in Nennius, to use the name for convenience, is probably adapted from an earlier Welsh poem, but there is no telling how much earlier, or what exactly it said. Moreover, even taken at face value, the texts raise other chronological problems. For one thing, they spread Arthur's career over an incredible stretch of time. Two of Nennius's battles can be located with fair confidence, one at Chester and one in southern Scotland; they make sense only in the context of widespread Saxon raiding in alliance with Picts, which is attested by Gildas and Bede; and that phase can hardly have been much later than the 450s. Yet the "Annales" put the last battle of all, the 'strife of Camlann,' in 539 (or 537; there is a slight ambiguity). Was Arthur a centenarian when he fought it? This is no modern quibble. At least two medieval authors seem to have been aware of a crux (3). We might hope at least to locate him in some part of the time-range and then treat everything outside as spurious. Apparently, however, this cannot be done, because the Welsh matter nowhere supplies a chronological fix to calibrate him with known history. We are never told that his first battle took place when X was emperor, or his last when Y was pope.

Even the stripping-away of legend doesn't really work. Reducing the evidence to what is in the "Historia Brittonum" and the "Annales" still fails to get rid of the problem. Thus, both credit Arthur with winning the battle of Badon. It was a real and important victory, mentioned by Gildas somewhere about the 530s when it was within living memory. It may have occurred near Swindon, or farther west, near Bath. But the Historia passage says Arthur slew 960 of the enemy single-handed in one charge. That need not invalidate the whole story of his campaigns, but it means that at least where Badon is concerned, legend-making has entered: a conclusion supported by two Arthurian fables in an appendix. The same may have happened in the "Annales" entry about this battle, which is disproportionately long because of the allusion to Arthur, itself rather curious and perhaps interpolated.

Some accept the 'strife of Camlann' entry as a p"Annales" appears to be real; a completely fictitious Arthur here, with no hint of interpolation, would be anomalous. What it says is that Arthur and Medraut, the original Modred or Mordred, fell at Camlann in 539. The trouble is that to isolate this incident as the sole fact not only upsets almost everything else because of the difficulty over dates, but also suggests that the whole vast cycle grew around a squabble of minor chiefs, otherwise unknown, at an unidentified place certainly far from the Saxon enemy whose repulse was the basis of the Arthurian glory. I, at least, cannot think my way from one to the other. The 'Camlann Arthur' who has been seriously proposed by Michael Wood, for instance, is a reductio ad absurdum of the method, showing that if you push it to its logical limit, the utmost it can offer is a minimal figure who explains nothing.

A more fruitful approach is to ask, not 'Did Arthur exist?' but 'How did the Arthurian Legend originate; what facts is it rooted in?' To do so is to acknowledge that this is a literary problem rather than a historical one, though with a hope that literary investigation may lead to historical insight. Such an approach casts the net wider and introduces a kind of lateral thinking. It allows, for instance, the consideration of Geoffrey of Monmouth, not in the sense of believing what he says about Arthur, but in the sense of asking what his raw materials were: sometimes, plainly, the aforesaid Welsh matter, but maybe not always. The investigation may lead to a real Arthur-figure or it may not. The first requirement is to try.

A crucial question is whether the Legend's roots are as far back as the period it professes to be about; or, to put this another way, whether the bards and story-tellers who created it were using traditions genuinely dating from that period. It is here that archaeology enters. It confirms the story of Saxon incursions into Britain and a phase when the advance more or less halted. However, it is far from confirming the Welsh-derived drama of large-scale warfare triumphantly ended by Arthur's victories. In that respect, it is of little help with the Legend as such. More promising are the results at specific sites.

Three places are outstandingly linked with Arthur. According to Geoffrey he was conceived in a ducal stronghold on the Tintagel headland, and it is generally assumed that he was born there. According to Caradoc of Llancarfan he had dealings with an abbot of Glastonbury, and the monks who exhibited a grave sixty years later said it was his and he was buried there. According to John Leland, citing Somerset lore later again, Arthur's Camelot was the ancient Cadbury hill-fort which can be seen from Glastonbury Tor. The Camelot of romance is fictitious, but a significant point about it is that it is not Britain's capital. It is Arthur's personal headquarters. The possibility of 'Camelot' having a basis in such a headquarters can fairly be entertained.

In all three instances, archaeology has proved occupancy and eminence in the period to which the Legend refers. Tintagel, formerly interpreted as a darkage monastery, has emerged in recent years as a major centre, very likely a regional seat of government, during the fifth century. At Glastonbury a Christian community existed almost or quite as early, if perhaps on the higher ground rather than the site of the Abbey, and may have been the only one in that part of Britain. At Cadbury, excavation in 1966-70 showed that the hill-fort was reoccupied probably during the second half of the fifth century, and fortified with a new stone-and-timber rampart nearly three-quarters of a mile in perimeter, including a gatehouse. Excavation of other hill-forts has since shown reoccupation and refurbishment, but no full parallel for the great Cadbury fortification, with its gatehouse, has turned up anywhere else in post-Roman Britain. It implies a very special occupant with impressive resources of manpower: a king or chief unique (so far as present knowledge goes) in his time (4).

These three places were picked as major locations of Arthur's story, and all three now stand revealed as important and apt in the right period. The implication is clear. At Tintagel the headland would have gone through a long phase of vacancy or near-vacancy before Geoffrey told his tale. He was not spinning a fantasy around famous ruins as he did at Caerleon, he knew some kind of tradition of the place's long-ago appropriateness. At Glastonbury the acceptance of Arthur's grave by Welshmen, against natural inclination, and the non-emergence of any rival grave, go far to establish a similar tradition irrespective of what the monks may have heard (5). At Cadbury, uninhabited for hundreds of years, even a modern archaeologist could not have detected the new fortification by inspection alone, without digging. It is really not to be supposed that the unknown person responsible for the Camelot identification chose the most plausible hill in Britain by a mere guess.

The people who focused on these places knew something about them. A purely accidental three-out-of-three score is beyond serious credence. They drew on traditions originating in the Britain to which they assigned Arthur, the Britain of the century or two after separation from Rome. That is truly where the Legend is rooted. All three places, by the way, are in the West Country, the former Dumnonia. Advocates of a northern Arthur have produced no comparable sites. Obviously Arthur, if he existed, could have been active in that part of Britain and inspired early bardic allusions, but nothing of consequence in the north gives him a birthplace or a headquarters or a grave, and the region's archaeology nowhere links up with any story of him.

While Camlann might etymologically be Camboglanna, a northern Roman fort, the versions of Arthur's last battle never point to this fort or anywhere near it. There are two Camlanns in Wales, still called so, and even the claim of the Somerset river Cam is backed by a report of a mass burial (6).

Given the apparent body of tradition we can venture a little further and glimpse a few individuals embedded in it from whom Arthurian characters are derived. The distant original of Uther's brother Aurelius Ambrosius is a fifth-century British war-leader, Ambrosius Aurelianus; he is mentioned by Gildas and Bede, and a continuity in legend is witnessed by Nennius. The distant original of Mark in the Tristan story is seemingly a certain Marcus, likewise Roman-named, with a father Marcianus who was called after a mid-fifth-century emperor. Romancers gave Marcus a role as King Mark of Cornwall that may be fictitious, but the Marcianus connection shows that he was 'there,' so to speak, from a very early stage (7). Was Arthur there also, perhaps as a real person, perhaps as an imaginary hero? Or was he inserted in the traditions later when they had undergone development?

His name favours the first alternative. 'Arthur' is a Welsh form of the Roman Artorius,' not common, but adequately attested. Arthur falls into place alongside Ambrosius, Marcus, and others in the same category, during a phase when Roman influence lingered. Furthermore, there is a sequel. In the latter part of the sixth century, when Roman names in general had faded out, this hitherto rare one began to enjoy a vogue. Several Arthurs are on record up and down Britain, including a Scottish prince. They are best explained as having been named after a hero established in song and story; and therefore alive or invented earlier, with a long enough interval to carry his bardic fame beyond his own people (8).

As for his historicity or otherwise, two arguments can be dismissed. Critics have urged that because he is credited with fantastic feats, such as his singlehanded slaughter at Badon, he cannot have been real. But fantastic feats were ascribed in America to Davy Crockett, who was real enough. Most were tongue-in-cheek "tall tales," but they were current and popular, and within a year or two of his death at the Alamo, he was seriously alleged to have killed 85 Mexicans during the siege; not 960, admittedly, but still a pretty wild number, and after a much shorter time for exaggeration. It was formerly claimed that anyone said to have slain a dragon must be fictitious. Yet several reputed dragon-slayers, in the Balkans for instance, were undoubtedly real. Far-fetched elements in a story do not discredit the entire story. Far-fetched elements in a career do not disprove its protagonist's reality. Baron Munchausen himself was real.

.............................................................
Origins of the Arthurian Legend:    Part 2   |    Part 3   |    Part 4   |    Part 5