As noted, this change in academic opinion is unknown to the general public. Morris’s Age of Arthur, discredited by reputable historians, is still in print and available in all good bookshops. Books by Dumville are harder to find. Readers with a general interest in King Arthur, spurred on by authors such as Morris, are surprised to find few academic works ready to debate the points.
The ‘evidence’ deserves to be analysed, not simply dismissed. For this reason, I will deal with the written sources in some detail. I will show why they are not used as uncritically as once they were, while re-examining whether they have anything plausible to say about the reign of Arthur.
All Dark Age sources were written to serve particular interests, especially the Catholic Church and the dominant dynasties of Wessex and Gwynedd. They derive from eras when literacy was confined to the elite. There were no sources composed by ‘ordinary’ people. Moreover, the written sources only survived because elite groups had them copied and preserved. This is just as true of the sixth-century ‘admissible’ evidence as it is for the ninth-century ‘inadmissible’ evidence. The work of Gildas survived because it contained a message which it was useful for tenth-century ecclesiastics to perpetuate.
If a history book is written to support particular circumstances, this does not by itself prove it is false. Current circumstances might exist because of those past events. Conversely, knowing that these sources were written and preserved to serve particular interests is also useful. When the material supports those interests, attention should be drawn to it. When it has nothing to do with, or indeed contradicts, them, this provides very useful evidence indeed.
He was not Arthur
Apart from the odd monument inscribed with names such as Voteporix or Drustanus (Sir Tristan?) the sixth-century British left practically no written records. Instead, Gildas tells us, they loved to hear their deeds recited by bards, men with ‘mouths stuffed with lies and liable to bedew bystanders with their foaming phlegm’.
Bardic poems were passed on from generation to generation, surviving to be written down in the Middle Ages. The oldest of the poems, Y Gododdin, gives detailed insight into the lives of the Britons. It includes this verse referring to Arthur himself.
More than three hundred of the finest killed.
In the middle and on the flanks he laid them low
Splendid before the host, most generous willed,
Bestowing horses from his own herd every winter’s snow.
He brought down black crows to feed before the wall
Of the city, though he was no Arthur.
Of men he was amongst the mightiest of all,
Before the fence of alderwood stood Guaurthur.
The poem was written down in the thirteenth century, but scholars agree that many of its verses are much earlier in origin. The most recent work concludes, on grounds of language and content, that this verse is among the oldest, possibly from as early as 570 (Koch 1997). Not only is Arthur the rhyme for the hero’s name, but in the original Welsh all the four last lines rhyme, making it unlikely that Arthur was inserted by a later scribe.
Guaurthur was one of the heroes of the Gododdin, the tribe living around Edinburgh, who took part in an expedition against Catraeth (modern Catterick). This is probably where he provided the crows with carrion, since the word used, Cair, refers specifically to a Roman city, as Catraeth was. The Gododdin were fighting against the Saxons of Deira some time in living memory before 570. In some way, Guaurthur was comparable to Arthur. Arthur was not said to be among the Gododdin. The best explanation is that he was a famous figure the poet expected his audience to recognise.
Keep this in mind when we confront arguments against Arthur being a historical figure. If Y Gododdin were the only source mentioning Arthur, no one would doubt that he was historical, famous as a warrior, from a period sometime before the expedition to Catraeth. No one else in Y Gododdin is a mythical superman, a composite character pieced together from scattered legends. Another verse includes the comparison ‘what Bratwen would do, you would do, you would kill, you would burn’. Bratwen is not a rhyme for the hero’s name, or anything else in the verse. No one, however, has written articles suggesting Bratwen’s name was intruded into the text close to the thirteenth-century date of the manuscript. Bratwen is accepted by all commentators as a genuine character familiar to the listeners.
The Gododdin reference to Arthur ought to be uncontroversial. Unfortunately, the weight of medieval tales is always set in the balance against such simple conclusions. This seems most unfair. If being the stuff of medieval legends is a good enough reason for being banished from genuine history, Alexander, Charlemagne and Richard the Lionheart would be discounted as historical.
It is crucial to know whether Y Gododdin is genuinely early. Like most works from the ancient and early modern world, it survives only in a copy from a much later period. Historians must judge the content, not the physical age of the book. There are three main reasons for dating it to the late sixth century. In its current form, it has verses attributing it to Neirin. The Historia Brittonum says that Neirin was a famous poet soon after Arthur, apparently in the sixth century. This argument is circular, as either of the references could have influenced the other. Moreover, the name Neirin does not appear in the earliest verses.
To fight at Catraeth, the Gododdin would have had to pass through the land of Berneich (Bernicia). Various sources describe a Saxon takeover of Berneich in the mid-sixth century. The earliest version of Y Gododdin only speaks of the Deor (Deirans), the English in the Catraeth area, with no mention of Berneich. It is therefore reasonable to assume that the expedition took place before the Saxon conquest of the more northerly region.
Most compellingly, the language of Y Gododdin is an incredibly old version of Welsh. The manuscript preserves two versions of the text, the first (A) being a more recent and expanded version than the second (B). Many pre-Old Welsh spellings are preserved in both texts, a fact which can be checked from established linguistic theory, place-names and contemporary Irish material. These forms predominate in the B text, which includes the Arthur verse. Koch argues that the B text is itself a composite, with some parts, including the Arthur verse, of sixth-century vintage. His hypothesis is not universally accepted, but all authorities agree that the B text is earlier than the A. The idea that the Arthur reference was inserted when the manuscript was written does not explain why the inserter would put it into the more difficult older text while not carrying it through into the easier A text, where Guaurthur is also named.
Since the nineteenth century, philologists have demonstrated the regular and predictable rules by which languages have evolved. It is now relatively simple to trace how the name ‘Maglocunus’ found in Gildas became ‘Mailcunus’ in Historia Brittonum, later emerging as ‘Malgo’ in Geoffrey of Monmouth and ‘Maelgwn’ in the Triads. However, this process was not understood by Dark Age or medieval writers. It is thus easy for modern historians to deduce the age of the sources by the form of language used, irrespective of the age of the manuscript or any chronological claims within it. At its most simple level, we know that Geoffrey of Monmouth’s source for the exploits of Urien Rheged must be later than that used by the Historia Brittonum, as he calls the king Urianus, while the Historia preserves the name in the earlier form of Urbgen. By studying their language, we can deduce that poems like Y Gododdin are much earlier than the late thirteenth- or fourteenth-century manuscripts in which they survive.
All the evidence suggests that the reference to Arthur was an original part of a mid-to-late sixth-century poem. He was a famous warrior, with whose deeds those of one of the Gododdin men were comparable. To suggest anything other than this straightforward explanation is utterly illogical. The probability is that the reference is to a real warrior of the recent past.
This is a plausible and reasonable inference from the evidence. There is nothing to suggest that Guaurthur was not a real British warrior of the sixth century. There is likew
ise no reason to think any differently of Arthur. Arthur’s existence rests on exactly the same source. If Koch is right, then Arthur was known to have existed before the late sixth century, when the first verses of Y Gododdin were composed. His fame as a warrior made him a fitting subject for comparison to a similarly named Gododdin hero.
This plausible and reasonable hypothesis forms the basis of the rest of the book. There was an Arthur. His deeds were known to a sixth-century poet and his audience. He was comparable to and better than Guaurthur. Like everyone else in the poem, he is not a mythological demi-god. He is not a composite character formed from various stories of men of the same name. For the Gododdin poet there is obviously one recognisable Arthur. We know, in short, that Arthur existed in so far as it is possible to know that any named Briton of the fifth or sixth century existed. There is no reason why he alone should have to demonstrate his existence beyond reasonable doubt, rather than on the balance of probability.
Asserting that Arthur was a real person, however, is not the same as proving that ‘King Arthur’ existed. Though the poem has given us reason to believe that Arthur was real, we will have to look at other Dark Age sources to see what light they shed on the enigmatic comparison. If Guaurthur was not Arthur, then who was?
TWO
Then Arthur fought against them in those days, with the kings of the Britons, but he himself was the leader in the battles. The first battle was towards the mouth of the river which is called Glein. The second and third and fourth and fifth were on another river, which is called Dubglas and is in the Linnuis region. The sixth battle was on the river which is called Bassas. The seventh battle was in the wood of Celidon, that is Cat Coit Celidon. The eighth battle was in Castellum Guinnion, in which Arthur carried the image of Saint Mary ever Virgin upon his shoulders, and the pagans were put to flight on that day and there was great slaughter upon them through the virtue of our Lord Jesus Christ and through the virtue of St. Mary the Virgin His Mother. The ninth battle was waged in the City of the Legion. He waged the tenth battle on the shore of the river which is called Tribuit. The eleventh battle was made on the hill which is called Agned. The twelfth battle was on the hill of Badon, in which 960 men fell in one day in one charge by Arthur. And no-one laid them low save he himself. And in all the battles he emerged the victor.
Historia Brittonum
The source which first gives the military career of Arthur is Historia Brittonum (The History of the Britons). The earliest version is found in Harleian Manuscript 3859, so called because it once belonged to the eighteenth-century collector, Edward Harley, Earl of Oxford. In the Harleian Manuscript, the Historia is anonymous, but other versions give its author as, variously, Nennius, a son of Urbacen, Mark the Anchorite or Gildas, a much earlier writer.
The Arthurian material in the Historia is of vital importance, since it is the earliest record of the actual deeds of Arthur. Whether this material is historical or legendary is crucial to any argument about the reign of Arthur. We must therefore consider what kind of document the Historia is.
Dumville, the current editor of the Historia, is adamant that it is overwhelmingly of a legendary or ‘synthetic’ character. The ninth-century author has heavily edited his sources to fit them into a preconceived framework. Unfortunately, Dumville has yet to complete his publication of the work, meaning that for the past twenty-three years we have simply had to take his word for this. Historians have consciously avoided making any comments but have generally taken his assertions as full permission to ignore the Historia’s Arthurian material. It is academic received wisdom that the Historia is valueless as a historical source for the fifth/sixth centuries.
Popular works on the historical Arthur usually make no reference to this. In them it is generally assumed that the Historia was written by a ninth-century Welsh monk called Nennius and that much of the material is presented at one remove, for instance by translating Welsh poems into Latin, from lost primary sources. This gulf of understanding is compounded by the version of ‘Nennius’ most accessible to amateur historians. This version (Morris 1980) is so inaccurate and inconsistent that it must be used with extreme caution. Its editor, John Morris (author of The Age of Arthur), died before completing his work. What was published was the Harleian Recension, augmented with excerpts from other texts and with no indication as to the criteria used for selection. Other additions, such as a section identifying Badon as Bath, are not found here in any text whatever.
Historia Brittonum was copied numerous times in the Middle Ages. Its disjointed style made it easy for scribes to omit or add sections and update the material. These produced many variants, which we can group together in families called ‘recensions’. The recensions follow, more or less faithfully, a particular exemplar. The Harleian Recension, represented by the oldest surviving text, is generally considered the closest to the original. Whether this is actually true must await Dumville’s full publication. For the purposes of this book, we assume it is.
The Nennian Recension claims to be written by a certain Nennius. Its prologue continues: ‘I have undertaken to write down some of the extracts that the stupidity of the British cast out; for the scholars of the island of Britain had no skill, and set down no record in books. I have therefore made a heap of all that I have found, both from the annals of the Romans and from the chronicles of the Holy Fathers, and from the writings of the Irish and the English and out of the traditions of our elders.’
Unfortunately, it is unlikely that the Nennian prologue was part of the original Historia. If it had been, we might expect it to be reproduced in the other recensions as well. There is no reason to believe that the writer, although he may be passing down a true tradition of authorship, had any genuine knowledge of the author’s sources or intentions. Nevertheless, the name ‘Nennius’ is now conventional for the otherwise anonymous author. I use ‘Nennius’ when discussing the methods and intentions of the author, without committing myself to his actual identity.
Most commentators follow the Nennian prologue in assuming that the author simply gathered together excerpts from various books, mixed them with oral traditions and regurgitated them almost undigested in the Historia. The ancient sources are therefore preserved at only one remove in a sort of historical scrapbook. Dumville, however, is convinced that the writer had worked over his sources in a comprehensive way to fit a chronological framework, leaving little material unaltered. This understanding is crucial to an appreciation of the Historia.
Historia Brittonum covers a broad sweep of time, from the legendary founding of Britain after the Trojan War, through to the seventh century. About two-thirds of the book deal with the most recent 300 years of history. The author provides the approximate date of the book. At the start, he gives the present as AD 831. Later, he calculates that Patrick went to Ireland in AD 405, 421 years before the present (i.e. AD 826). In the same section, he gives Patrick’s date as 438, giving a date for the present of 859. Either an authorial or scribal error has resulted in two different dates for the same event, or the manuscript has been updated. Dumville suggests that Nennius intended a date of AD 829 for the present (IV in Dumville 1990).
Wonders of Britain
After the end of the Historia, there is a gazetteer of Wonders of Britain, the Mirabilia. It is not clear whether this originally formed part of the work. The author seems to be a contemporary of Nennius and to share an interest in the same area of Britain. The wonders have been associated with the Historia from early in the manuscript tradition, passing together into different recensions. I will treat them as the work of the same author, although if they are not, the fact that two ninth-century writers give supporting material on Arthur would strengthen my case.
‘Arthur’s Britain’.
Although the scope of the wonders is national, the fact that most of them are actually to be found in South Wales and the Severn Estuary points firmly to the author’s home area. In Buelt (Builth), he tells us, there is a pile of stones called Carn Cabal bui
lt by Arthur the Soldier. The topmost stone bears the footprint of Arthur’s dog Cabal, made when he was hunting the boar Troynt. In Ercing, 35 miles away in modern Herefordshire, is the wonder of Licat Anir. This Anir was the son of Arthur the Soldier, who killed him and built a tomb there. The author has personally tried to measure the tomb and found it impossible to obtain the same measurement twice. (The name of Arthur’s son is frequently given as Amr. I follow the reading of the current editor.) The only other wonder the author connects to a named individual is a tomb in a church built by St Illtud in Llwynarth, on the Gower peninsula, 50 miles away from the two Arthur wonders.
These wonders are important pieces of information. They tell us Arthur was a soldier, as we might have expected from Y Gododdin, but locate him in South Wales. We know that, as Y Gododdin was transmitted through the early Middle Ages, it acquired verses linking it to Welsh heroes. However, these emphasised the North Welsh Kingdom of Gwynedd, not Builth or Ercing. Moreover, we also know that the verse referring to Arthur pre-dated those interpolations.
Although the wonders are folkloric in character, this gives no reason to doubt that Arthur was real, any more than that Illtud was a real Dark Age cleric. Folklore and legend linked to real characters and events were in the idiom of even the most sober Dark Age historian. The wonders attributed to Arthur are no more than would be expected from a writer of Dark Age Britain. No one doubts Bede’s account of Oswald of Northumbria’s death at the battle of Maserfelth. Yet Bede devotes most of this account to describing wonders such as the cure of a sick horse which rolled onto the spot where Oswald was killed or the man whose house burnt down save for the beam where his cloak, touched by the mud from the site, had hung (EH IV 2). Even tombs of varying length were not considered impossible. The stone sarcophagus made for King Sebbi of the East Saxons was too short. ‘In the presence of the Bishop and of Sighard, son of the monk king . . . and a considerable number of men, the sarcophagus was suddenly found to be the correct length for the body.’ By contrast, although Anir’s tomb and the footprint of Cabal are wonders to the contemporary writer, they derive from less than wonderful events in Arthur’s life, hunting a boar at Builth and killing his son in Ercing.
The Reign of Arthur Page 2
