Sub-Roman Britain

For a work of twenty pages to create a new academic consensus almost overnight, forcing upholders of the previous ‘orthodox view’ into hasty retraction or academic oblivion you would think it would need to be, at a minimum, logically argued and devoid of factual errors. This one isn't.

Sub-Roman Britain: History and Legend presents itself as a demand for a more exacting study of those texts concerned with period in question, British history in the fifth and sixth centuries - a period which Dumville defines as 'politically dark', and as a transitional phase. Morris and Alcock, using Celtic texts to write the history of this period of Celtic dominance - and in doing so "breaking with the tradition of twentieth-century English historiography" - have failed to understand the nature of their source material. Study of this material is "still in its infancy". It must be subject to the most rigourous scrutiny, "we need to understand the sources, motives, and technical terminology of each of writers".[1] At the end of Dumville's analysis, however, there's only one writer left to analyse. Having ruled out every other text as too late, too ridiculous, or out of bounds to the historian, the only text remaining from which to write a history of Arthur's period is, in Dumville's view, Gildas' The Ruin of Britain.

Gildas, who doesn't mention Arthur, is the only British text surviving from the period. There are a few Welsh poems which refer to Arthur, but historians may not call upon their witness, Dumville tells us, because they were still in the hands of the philologists, awaiting a ‘secure’ date.[2] This statement assumes historians can contribute nothing from their own discipline to the dating of historical texts. It also tacitly admits the truth of Morris’ assertion, that this period is ‘dark’ because it has not been studied. So how is it possible to reach so firm a conclusion of the question of Arthur, when what may be vital evidence is not yet in the hands of historians? Dumville himself has already reached a verdict on the date of the poems, in advance of the philologists pronouncement: They cannot be earlier than the later sixth century.

The earliest ‘securely dated’ reference to Arthur is the ninth-century Historia Brittonum, a work long attributed to ‘Nennius’ on the strength of a preface which gives the author that name. Dumville regards this as a mistake. The Nennius preface has no claim to be an original part of the document and must be rejected, along with the forger's claim to have 'made a heap of all that I have found' - a claim which deluded incompetent scholars like Alcock and Morris into thinking that the Historia Brittonum preserved unedited sources from a still earlier period. The nameless author of the Historia did have sources, sources which are still extant. Dumville lists them, and concludes "I trust that the mere recital of these sources will suggest their utter flimsiness as records of this obscure century of our history”.[3] This is not an argument. Dumville merely invites us to share his opinion.

And what of Gildas' sources? "Gildas", says Dumville, "is our prime text for the outline history of the period from the end of Roman rule to the mid-sixth-century", because "he alone seems to have had access to contemporary sources for the fifth century and was an eye-witness to the earlier sixth."[4] A prime text is not a primary source. Both words imply ‘chief’, ‘principal’, ‘most important’, but speaking historiographically a primary source is one not derived from any other, generally a contemporary witness. Gildas is not a contemporary witness for the fifth century. Dumville says that he seems to have access to contemporary sources, though in this case he does not list them. If he knew what they were, then surely he would. Dumville, then, does not know for a fact that Gildas derived his account of the fifth century from contemporary sources, it only seems to him that this is so, and on the basis of this subjective judgement we are invited to accept Gildas’ sermon as our prime text for the century. But how do we know the sources Gildas might seem, to David Dumville, to have had are any more reliable than sources the writer of the Historia actually did have?

Of course Gildas’ is nearer in time to the fifth century than ‘Nennius’, but Dumville himself rules that argument out of court in dismissing Bede's contribution to sixth-century history. Bede's History of the English Church and People contains very little information on sixth century history, but the little it does contain is highly significant. This is our prime text for the Augustine mission at the end of the sixth century. "But Bede", Dumville warns, "is not a primary source for later-sixth century history. ... Because his work is a fine piece of scholarship, a mine of information, and written in a clear Latin style, it does not follow that we should necessarily accept his view of centuries for which he is at best a secondary authority as more reliable than that of any modern scholar. The argument that Bede lived much nearer to the fifth and sixth centuries than we do should not be allowed to cut any ice."[5] Bede completed his history in 731, and dated Augustine’s mission to 596-7, 135 years before. About the same time period separates Gildas’ sermon from the end of Roman Britain. So why should Gildas be regarded as more reliable, over that distance, than Bede? Because his Latin is turgid?

And so to Arthur, the real object of Dumville's attack, to whom he devotes a whole paragraph: " We come, last in the fifth century and first in the sixth, to Arthur, a man without position or ancestry in pre-Geoffrey Welsh sources. I think we can dispose of him quite briefly. He owes his place in our history books to a 'no smoke without fire' school of thought. What evidence is there for his existence? Almost twenty years ago the late Professor Thomas Jones gave us an admirably balanced account of the early evolution of the legend of Arthur. Independently, and at almost the same time, Professor K H Jackson published an excellent survey reaching remarkably similar conclusions. The totality of the evidence, and it is remarkably slight until a very late date, shows Arthur as a figure of legend (or even - as Sir John Rhys pointed out last century - of mythology)."[6]

That two scholars who studied the Arthur legend found Arthur to be a figure of legend is not significant, merely inevitable: if Arthur were not a figure of legend they could not have studied his legend. Perhaps David Dumville meant to say these two professors had show Arthur to be purely a figure of legend, but that is not what he actually does say. As for Sir John Rhys, he said something completely different. In Rhys view Arthur was historical. He argued that the legend of Arthur arose from the confounding of a real man with a British deity in consequence of the similarity, or identity, of their names. It is hard to see how Dumville made this mistake, when he gives as his reference the very publication - Studies in the Arthurian Legend, Oxford, 1891 - in which Rhys puts forward his theory that Arthur was the last Roman military leader of Britain, and was for that reason remembered as Yr Amherawdyr Arthur - ‘the Emperor Arthur’.

What David Dumville actually proposes in Sub-Roman Britain, though in his misrepresentation of John Rhys position he doesn't quite admit it, is that Dark Age scholars should tear up all previous academic study right back to the time when Celtic studies first became an academic discipline, and start again from scratch, this time with Arthur ruled out of bounds at the outset. And his fellow historians have agreed to go along with this radical proposal, supposedly on the strength of the arguments put forward in this article. I find this incredible, and would like to suggest an alternative explanation. David Dumville’s article only pronounced the sentence of hereticization against John Morris, it did not provide the reason for it. The real cause of Morris’ denigration is that peculiarity of Dark Age scholarship so sharply revealed in the saga of Arthurian Tintagel.

Heretic Emperor: The Lost History of King Arthur
Copyright © V M Pickin 2009

Return to Home Page


1. David Dumville, Sub-Roman Britain: History and Legend, pp174, 173, 192

2. Ibid. p178

3. Ibid. p177

4. Ibid. p191

5. Ibid. p191

6. Ibid. p187