The once and un-killable king

King Arthur: The Making of a Legend

Nicholas J. Higham, Yale University Press, 2018, 380 pages

LIAM GUILAR marvels to see a sledgehammer being wielded against castles-in-the-air

People in Britain have been telling stories about an ‘Arthur’ since at least the 9th century, possibly earlier. In the Middle Ages, those stories include some of the finest literature ever produced in Europe, culminating in Sir Thomas Malory’s 15th century masterpiece.

Scholarly arguments over the existence of an historical King Arthur, a single figure as point of origin for these stories, are more recent. In the mid-20th century, the idea that there was an historical figure gained ground, but the high-water mark of scholarly attempts to argue the case had passed by the 1980s. Leslie Alcock (Arthur’s Britain) and John Morris (The Age of Arthur) were both respected academics, but both their books, especially the latter’s, received the kind of academic reception about which scholars must have nightmares.[i].

Despite repeated attempts by experts in the field of ‘post-Roman’ or ‘Dark Age’ or ‘Early Medieval’ British history to discredit the various candidates, and despite the lack of evidence to support any of them, the arguments rumble on. Nicholas Higham’s new book is an attempt to demolish the idea that there is an identifiable historical figure who is the real King Arthur. It seems doomed to fail. He is not the first scholar to announce that the historical Arthur did not exist. It’s unlikely he’ll be the last.

In 1977, David Dumville, one of the leading authorities on the sources for early medieval history in Britain, concluded an article that discussed the Welsh evidence for an historical King Arthur:

The fact of the matter is there is no historical evidence about Arthur; we must reject him from our histories and, above all, from the titles of our books

In 2013 an equally exasperated Guy Halsall, an expert on early medieval history, wrote a book offering “a corrective to the shelves of pseudo-Historical ‘Arthurian’ nonsense available in practically every bookshop in Britain”, concluding,

No sane scholar will now argue that there is definitely a “King Arthur” figure in 5th– or 6th-century history about whom anything solid can be said

In 2018 Nicholas Higham, who specialises in what used to be called ‘The Dark Ages’, produced this encyclopedic refutation of the varied and various arguments for an ‘Historical King Arthur’. He lines up the contenders – the Sarmatian Arthur, the Greek Arthur, the list of nominees with names sounding like Arthur or those whose names sound nothing at all like Arthur – the ‘if this, then this, and then that means we’ve found Arthur’ arguments, and one by one he knocks them over.

Higham’s conclusion is that

[…] we can now agree to discount King Arthur as a ‘real’ figure of the past, leaving him and his deeds to the ‘smoke’ and ‘highland mist’ of make-believe and wishful thinking; it is there that he properly belongs

I distrust that first person plural which Higham is fond of using. Reading the book is like being bludgeoned, very thoroughly and very carefully. It should settle the argument. But it won’t. Even the blurb on the cover hedges its bets. Max Adams, identified as the author of In the Land of the Giants, is quoted: “Riveting…brings the historical Arthur to what may be his last decisive battle”. “May be” because, given the nature of the evidence, there is never going to be a final, irrefutable argument.

Candidates for the historical King Arthur have their partisans. But if the experts have become more wary, the field is still held by enthusiasts who fly on a combination of ill-informed speculation and wishful thinking. They simply cherry pick ‘information’ and don’t bother with the usual rules of evidence, source analysis, linguistics or logic. If anyone is arrogant enough to believe that lacking the skills and knowledge required to move through the tangle of evidence puts them in a position to argue with people who have spent their professional careers studying that evidence, then nothing is going to dent their self-confidence.

The question of Arthur’s existence hinges on a very limited number of sources, and the combination of skill, knowledge and training required to assess the reliability of those sources is very rare. There is a world of difference between ‘looking stuff up’ on the internet, or in the library, or in the museum, and doing research. The failure to understand that difference, which is becoming increasingly widespread, lies at the heart of the ‘Arthur Was Real and Eureka I’ve Found Him’ phenomenon.

A scene from John Boorman’s Excalibur

The scarcity and unreliability of the surviving written sources can be hard to grasp. Imagine if 1,000 years from now, you are tasked with writing the history of the Trump Presidency. Your only piece of evidence is a copy of a copy of something from a newspaper. The copy was made in 2320. There’s a name attached to it, but you know nothing else about the journalist. You don’t know which newspaper the text is from, nor do you know if it is from an editorial, factual report, opinion piece or work of fiction. You have no way of checking anything in it against other sources.

And if you think that reconstructing four years from such a fragment sounds difficult if not impossible, then spare a thought for the historian of the 5th and 6th centuries in Britain, who can put the only surviving piece of contemporary insular ‘narrative history’ on a couple of PowerPoint slides. That oldest surviving narrative, written by Gildas within living memory of the battle of, or at, Badon, does not mention anyone called Arthur[ii].

All the other surviving sources were written much later and they all need to be handled with care. The problem with the sources can be demonstrated with one example,

What may be Arthur’s earliest appearance in an insular text comes in a collection of eulogistic stanzas of early Welsh poetry collectively known as y Gododdin

While often referred to as ‘a poem’, Y Gododdin, as Higham rightly states, is a collection of verses commemorating a Northern British raid on the Saxons at Catreath. The raid was a complete disaster and the verses celebrate the men who died. The problem is that Y Gododdin only survives in a 13th century manuscript. The work is generally credited to Aneirin, who is said to have lived in the 6th century. It’s worth pausing to remember that the distance in time between ourselves and Shakespeare is less than this. How much of the material, if any, in the manuscript dates back to the 6th century is a matter of scholarly controversy. Obviously, the date of the ‘Arthur reference’ makes a huge difference to the value of that reference.

Higham is willing to accept an early date, and he quotes the relevant stanza in English. The last four lines read:

He used to bring black crows down in front of the wall 
Of the fortified town – though he was not Arthur – 
among men mighty in deeds 
in front of the barrier of alder wood-Gorddur

Gorddur is the warrior’s name, and he is being praised for his ferocious deeds in battle, although ‘he was not Arthur’. That’s all there is.

The enthusiast says it’s obvious that here we have a reference to a famous Arthur, and this proves stories of King Arthur must have been circulating (off-stage) when this verse was written.

The sceptic asks for evidence that independent stories circulated about a real character called Arthur at the time this verse was composed. The enthusiast points to the poem. Aneirin must have been able to rely on his audience to know the stories, in order for the allusion to work. The allusion proves the existence of the stories and the stories guarantee the validity of the allusion. Dizzying?

Nor does it tell us anything about Arthur except it was a famous name. It doesn’t help us to identify an Arthur, or tell us when or where or if he lived.

A literate person with the necessary patience can follow Higham’s summary of the complicated problems of dating Y Gododdin in general, and of that line in particular. But there are very few people who can read the manuscript, or its facsimile, and the number of people on the planet who have the expertise to negotiate the dating arguments and evaluate the evidence for themselves probably wouldn’t fill a coach for a day trip to Catreath.

And therein lies the real problem. Early British, post-Roman history, is a highly specialized field, but Higham, as did Guy Halsall before him, bemoans the fact that in many ways the specialists have withdrawn from the debate:

Today most specialists distance themselves from the whole issue of Arthur’s reality, citing insufficient evidence to be able to judge his place in history and declaring themselves agnostic on the matter. But their silence leaves the history-reading public with insufficient guidance to the competing claims and without the specialist knowledge to judge between them effectively, for these are highly complex issues

Higham is critical of ‘agnostic’ scholars who refuse to be drawn into a conclusion on the subject. Ironically, Max Harris, whose comment is quoted on the cover of Higham’s book (see above), wrote, in his introduction to In the Land of the Giants:

And then there is Arthur. Historical references to this legendary Romano-British warlord are very few: a list of 12 battles; a great victory recorded at a place called Badon (perhaps Bath in Somerset) a death notice, a possible mention in a battle poem [iii]

This short reference to King Arthur continues with a classic piece of professional ‘agnosticism’: ‘Arthur may be, as many historians have argued, an irrelevance, a distraction’ (p.14). Adams also includes the ‘dates’ from the Annales Cambriae for the battles of both Badon and Camlann in the timeline of the Dark Ages he appends to the book (Adams, Appendix two, p. 429). That ‘may be’ that leaves the door open – just as the inclusion of the Annales Cambriae dates for Badon and Camlann in a timeline with verifiable dates gives them a spurious authenticity [iv].

For all the evidence Higham can marshal, (his bibliography runs for 23 pages of small print) there’s never a knockout punch. If Y Gododdin demonstrates the problems inherent in the evidence, Higham’s detailed discussion of the Historia Brittonum demonstrates why it’s not possible to close the argument.

The 56th ‘chapter’ of a document known as the Historia Brittonum (hereafter HB) is the crucial piece of evidence for enthusiast and scholar alike. HB chapter 56 contains a list of Arthur’s 12 battles. It’s the oldest surviving piece of extended writing about an Arthur. It names him as a victorious war leader against the Saxons.

‘Chapter’ might be misleading. It runs to 23 lines of continuous Latin prose in John Morris’ edition. Though written in Latin, compiled in the early 9th century and ascribed to ‘Nennius’, for most of us, our access to this strange text is through John Morris’ 1980 translation, which is not without its own problems[v].

Higham, like many before him, quotes the ‘chapter’ (p. 185). But I think that misrepresents the HB. The focus on chapter 56 allows people to treat this strange compilation as far more factual than it is. My own interest is in the earlier story of Vortigern (HB 31-49) who takes up much more of the HB than does Arthur. The Venerable Bede, writing at the beginning of the 7th century, following a hint in Gildas, had made Vortigern instrumental in the fall of Roman Britain. In the HB he has become an incestuous, bigamous, drunken fool in a bad folk tale about a beautiful princess. It is not history as understood in the 21st century. Ambrosius Aurelianus, who also appears as an historical character in both Bede and Gildas, has become a vatic child who was born without a father[vi], and St Germanus of Auxerre, who is perhaps the one person in this motley crew who can be established as undeniably historical, has become a spell-working magus who prays Vortigern to a fiery death in his tower[vii]. Reading the whole text does not inspire confidence in its factual accuracy[viii].

Higham, to his credit, takes on the whole of the HB and he’s very good on what it reveals about how different 9th-century attitudes to writing about the past are to our ideas of writing history. His chapter on the HB is worth the price of admission, even if you have no interest in ‘Arthurs’, although he has written about this, at length, before.

However, it’s not possible to dismiss the ‘Battle List’. It’s not enough to point out that no one has identified the battle sites with enough conviction to convince everyone else. (Guy Halsall reported one attempt to do so based on the names of modern pubs (Halsall, 2013, p. 154)); or that if the enemies are Saxons they seem to have been much further north much earlier than any other source suggests; or that the number 12 in a work riddled with biblical echoes seems more than a bit suspect and Arthur’s 12 battles mirror Patrick’s miracles as R.W. Hanning pointed out in 1966 (p. 120): or that some of the battles seem to have been fought by a leader who chose his battlefields because their place names rhymed: or that single handedly killing 960 enemies in a single charge sounds a tad unrealistic. Even with ten hours of daylight, how many deaths is that per minute, every minute, without a break for ten hours[ix]?

There’s more. At least three centuries have passed between the events described and the time of writing. Despite decades of attempts to find one, there is no evidence for an earlier source for the list. A lost poem is the best candidate, but then it would have to be a very strange poem and a list of rhyming battles might still be unconvincing. Anyone who claims that HB 56 is based on accurate oral transmission has to explain how, given that at least 14 generations have passed between writer and event, any oral story could be passed down without alteration. As a rule of thumb, students of oral history accept accuracy is possible in a story passed down for three generations: from your grandparents to you. Not, as Higham points out, a story passed down about your “great-great-great-great-great-great-great-great-great-great-great-great-grandparents (give or take a generation either way)”.

Despite all this, the weight of evidence can only say: ‘it is highly unlikely that this is a reliable source’. Higham’s analysis of the way Nennius uses other names in the HB whom we know to be historical, leads him to the perfectly demonstrable conclusion that: 

This was a scholarly community prepared to manipulate the distant past, shift individuals around and invent characters to make British history fit for purpose. They amended names to better suit their needs, misquoted from and rode roughshod over earlier testimony, fictionalized historical figures and made up others de novo. The harvesting of names from their original setting to be reused in a different context was commonplace

But he can’t prove this is what happened to Arthur. He can only suggest it is likely. He also claims that to understand the HB,

Our treatment of it must depend on understanding why it was written, where the author obtained his information, how he used what he had gleaned and the ways in which he expected his work to be understood

In his discussion of these features, he’s as guilty of speculation as anyone. Granted his speculation is much more well-informed, he’s still straying out of the world of facts. He is attempting to construct not only the context of a text that does not fit any modern genre, but its contemporary purpose and reception when there is no external evidence for these. Without knowing anything about the Real Author of this text, he’s going to draw conclusions about his intentions.

When Higham lists what can be inferred about the author from the text he’s constructing an Implied Author. It’s the reader’s idea of the author. We have no way of knowing who ‘Nennius’ was, let alone why he wrote what he did, if in fact he did write it. The ‘context’ of a written text is always a construct[x]. In the absence of corroborating evidence independent of the text, such a construct is never going to be the final word on the subject.   

I’d back Higham’s informed speculation against most people’s, but there’s no escaping the fact it’s still speculation. He slides from qualified statements, ‘the prologue if accepted as original……in that case he is likely…’ (my italics) to declarative ones:

These were Latin texts written by churchmen tasked with repositioning the Britons within a tradition of European history that centered on Rome

It’s that ‘tasked’ that rings the alarm bells. His reading of the evidence supports his hypothesis, but there’s no way he can prove it.

Sir Bedivere casts Excalibur back into the lake

Higham does need to be applauded for his willingness to accept that medieval authors made stuff up. There’s a peculiar strand in medieval studies, both amongst professionals and enthusiastic amateurs, that works on the assumption that everything that interests us has a prior source. Put like that, it sounds ridiculous. But the unstated assumption is that fiction is a post-medieval invention. So, when Higham surveys the evidence and writes

Wace’s introduction of the Round Table to Arthurian literature was a practical solution to an imagined problem, which there is every likelihood he came up with himself

it’s one of the best moments in the book.

The Tudor invention of Arthur – the “Round Table” in Winchester Cathedral

Given that none of the evidence for an historical Arthur seems convincing, why the persistent arguments? I think people want to believe, and don’t understand or care that the existence of an historical figure, like Alfred the Great or Lady Godiva, is not a question of belief but of provable fact. The arguments over Arthur repeatedly illustrate one peculiarity of early British medieval history. Given the lack of sources for the 5th and 6th centuries in Britain, it’s almost impossible to prove someone didn’t exist. Therefore, says the ‘agnostic’, we have to accept the possibility he did.

Higham is strongly, justifiably against this. He quotes Bertrand Russell:

’Many orthodox people speak as though it were the business of sceptics to disprove received dogmas rather than the business of dogmatists to prove them. This is, of course, a mistake.’ He illustrated the point by supposing the existence of a teapot in orbit around the sun that is too small to be visible through even the most powerful telescope. That this assertion cannot be disproved does not mean that it should be allowed to influence our thinking about the solar system. That way only chaos lies, for such speculations are infinite

Beyond a desire to believe, what reason is there to even entertain the idea of an historical Arthur? The answer to that probably lies in a bad metaphor which should have been dismissed the first time it was used. A long time ago someone advanced the argument that since ‘there is no smoke without fire’ there must be a factual, historical basis for the medieval stories about King Arthur. Higham returns to this metaphor in his final chapter and tries to replace it with another, but it’s time someone got rid of the habit of arguing based on inappropriate metaphors.

There may well be no smoke without a fire, but stories aren’t smoke. They are stories. The metaphor implies that all stories have some kind of factual basis. That’s demonstrably not true. If we throw out the inappropriate metaphor, there’s nothing left but wishful thinking.

Anachronistically-armoured knights feast with Arthur, who is in the Round Table

For all the detail, the knockout blow never arrives. Each chapter has its conclusion which sums up the case against the particular contender/argument, and then everything is summed up again in a concluding chapter. This makes the book laboured and repetitive. While the marshalling of scholarly argument is impressive, as the book progresses and Higham goes after some of the ‘fringe’ dwellers, it starts to sound brittle.

I admit to bemused admiration for Graham Phillips. He has made a career out of finding things Arthurian. He found the Grail. He found Camelot. He identified the ‘Real King Arthur’ as Owain Ddantgwyn using a chain of reasoning that was so circular it makes a spin cycle look linear. He has not let scholarly opposition or derision stop him. Give the man his dues: he’s held his line. Recently he claims to have found Arthur’s grave[xi]. The idea that Arthur’s 5th century grave can be found by reading Sir Thomas Malory’s 15th century text has so little to recommend it that it shouldn’t require pages of detailed refutation. It is a fine example of Russell’s orbiting teapot.

And despite what Higham has written about not being obliged to disprove the existence of orbiting teapots, he’s put himself in the position where he has to do so. If the purpose of the book is to educate the history-reading public, then he has to engage with Phillips’ argument. Reading his three-page explanation of the flaws in a portion of Phillips’ argument (pp. 264-267) is like watching someone trying to swat an annoying but mobile ant with a very large, very heavy hammer. It’s hard not to think that all this erudition could be put to a better use.

Despite all the knowledge, despite the careful explanations, despite the clear statement of intent, it is hard to assess how successful this book is. Who is its target audience? Higham claims,

The purpose of this book, is therefore, to set out the main arguments which are on offer, test each one against the sources on which it relies, and determine which, if any deserve support

But it’s difficult to avoid the feeling that the conclusions were written before the tests had been done.

Anyone interested in Arthurian studies, historical or literary, will benefit from reading the book. It’s an encyclopedic survey of the subject, written by an expert. It gathers together disparate information, and the Sarmatian, Nart and Greek chapters are a welcome summary of those diverse cases. But I wonder if Higham really thinks that someone inspired by the Clive Owens’ 2004 film King Arthur which was advertised as “The untold true story that inspired the legend”, is going to read his detailed, painstaking deconstruction of the argument that Lucius Artorius Castus was the original Arthur (pp 14-39)?

I suspect the people who need to read it probably won’t. And if they do, it probably won’t change their minds[xii]. The growing cult of the self-appointed expert means there is an increasing number of people who think access to the internet puts them in a position to discover what the experts have missed, and to challenge the experts’ arguments. We’ve seen this in the 2020 pandemic; it’s not confined to Arthurian studies.

For experts in the field, they’ve heard most of it before. They’ve read some of it in Higham’s earlier work, especially King Arthur: Myth making and History (Routledge, 2002). I suspect there will be those with recognized expertise in some of the more unusual fields that he has picked his way through who might object to the finer points in some of his arguments, but most of us won’t be in a position to follow the ensuing discussion, let alone play referee.

Those who don’t have the patience to read the book will stay happily deluded. Anyone who honestly thinks Arthur was an Ancient Greek Constellation before he had a career as a medieval king is not going to let the problems of transmission get in the way. Worse, as a contributing factor, there’s someone out there willing to publish your theory, or turn it into a documentary, because people like to watch the little amateur sticking it to the experts, even when the audience has no idea what’s at stake. It’s hard not to love a story that declares the Holy Grail has been discovered in an attic in Coventry, or Excalibur has been found[xiii]. Throw in the idea that there’s an academic conspiracy to hide the truth and there’s a small industry aimed at exploiting those who want to believe. Detailed arguments about manuscript dating, linguistic borrowings, or the problems of editing and dating early Welsh poetry simply do not make great television even if they are being presented by Michael Wood.

While the arguments over the historical Arthur are fascinating for all kinds of reasons, for many who are interested in the stories that accumulated around the name, it has always seemed an interesting irrelevance. Even if it were possible to identify a single person as the point of origin for all these disparate stories, and even if the proof were so convincing Higham publicly retracted everything he’s written here, ‘Arthur’ would be a brutal thug whose claim to fame was his ability to organize the slaughter of other violent thugs. He would have nothing in common with Malory’s Arthur except, perhaps, a shared name.


My thanks to those who read early versions of this essay, and in particular to Peter Hart whose painstaking proofreading saved me from serious embarrassment. All the remaining errors are mine

Works Cited

Adams, M (2015) In the Land of the Giants. Head of Zeus Ltd

Alcock, L (1971) Arthur’s Britain. Harmondsworth, Penguin Books

Dumville, D (1977)  ‘Sub-Roman Britain-History and legend’. History 62:173-92

Halsall, G (2013) Worlds of Arthur: Facts and Fictions of the Dark Ages. Oxford University Press

Hanning, R W (1966) The Vision of History in Early Britain. Columbia University Press

Morris, J (1973) The Age of Arthur. Weidenfield and Nicholson

Morris, J (ed.)  (1980) Nennius – British History and The Welsh Annals. Philimore & Co

Author’s Notes

  1. Despite the comprehensive mauling it received from the experts, Morris’ The Age of Arthur is still on sale, and if the comments on Goodreads are any indication, still encouraging the unwary to believe

2. Gildas The Ruin of Britain. (De Excidio et conquestu Britanniae)

3. Ironically, Higham’s book denies the ‘Historical’ status of everything Adams refers to here

4. The entry before the one for Badon reads: ‘Bishop Edur rests in Christ [i.e dies] he was 350 years old.’ It tends not to be quoted by those who want to believe in the reference to Badon. Higham’s discussion of the Annales is on pp. 222-225. They are included in Morris’ edition of Nennius

5. See Higham p. 178-9 for a discussion of the problems of ‘establishing the text’ in general and with Morris’ edition in particular

6. It’s typical of the HB that in chapter 41 the boy has no known father and in chapter 42 he does

7. The HB faithfully records that there are three stories circulating about Vortigern’s death, one of which involves the ground opening up to swallow him.

8. My own interest is in the way these stories developed, rather than any desire to sort fact from fiction. You can read about the development of Vortigern’s story across time, as well as those of Hengist’s daughter and St Germanus at:

9. I once pointed this out and was told that Arthur would have been using Excalibur and ‘we all know’ Excalibur was an alien artefact

10. I thought this phrase was Peter Barry’s, from Literature in Context (Manchester University Press, 2012) but I can’t find it


12. The repeated use of  ‘probably’ here is nothing more than a conventional stylistic avoidance of declarative statements in an attempt to appear undogmatic. If I were a betting man, I’d bet they won’t.

13. Amateur Sleuth traces ‘Holy Grail’. The Courier Mail, August 14, 1995 p.12