What We Know about the Historical King Arthur
(and it ain't much)
Much of the allure of the Arthurian saga lies in the possibility of a historical Arthur, the few tantalizing fragments of evidence we have of Arthur as historical figure. Clearly, the idea that the events related in these fantastic stories may have actually occurred in some form has been a large part of the widespread appeal of the Arthurian story dating back to the Medieval period. Since the late 19th century there have been hundreds of attempts to define and distinguish the true historical Arthur from the massive body of story and legend that has grown up around him. Many of these theories have undergone cycles of acceptance and rejection. Theories have been built upon theories, as the question of the identity of a historical Arthur continues to be examined to the present day.
When attempting to identify an original Arthur, historians must confine themselves to sources of evidence that have recognized historicity, that is, justifiable historical merit. Modern historians recognize three such sources: archaeology, historical writings composed prior to Geoffrey of Monmouth’s Historia Regum Britanniae, and what is perhaps the most basic piece of evidence, the name "Arthur" itself, which does not appear in any context prior to the 6th century AD. No matter what theory a given historian may be expounding, his or her arguments must proceed exclusively from these sources of evidence, possibly discounting some of them while building an argument around others.
The Archaeological Evidence
Archaeology is normally one of the primary sources of evidence in almost any historical inquiry. Indeed, researchers have undertaken numerous archaeological studies of alleged Arthurian sites, most notably Tintagel, Glastonbury, and Cadbury, with the idea of authenticating the supposed connections of these sites with Arthur as historical figure, and while numerous theories have resulted both for and against Arthurian associations, almost all archaeological evidence has been entirely inconclusive. The problem is that archaeology deals with sites and objects rather than people. It can give us information about groups of people and societies in general, but almost never about individuals. The sole exceptions are bodies and inscriptions, which, for the most part, have not been forthcoming (with two possible exceptions: The Arthur Stone, and the alleged discovery of the graves of Arthur and Guinevere by monks at Glastonbury in 1191, which supposedly produced both bodies and an inscription, but is widely regarded by modern scholars as an elaborate hoax). For the most part, the closest archeology has come to authenticating a historical Arthur is demonstrating that supposed Arthurian sites were in use during the time period he was supposed to have used them.
The Written Sources
The insufficiency of archaeology to authenticate a historical Arthur causes historians to turn to written sources. Early historians, most notably the influential John Rhys, attempted historical analyses of virtually the entire body of Arthurian literature, often up to and including Malory. The problem with this approach is that much of the literature does not have any historicity. The character of Lancelot, for example, never appears before the works of Chrétien. The famous round table is first introduced in Wace’s Brut. It is clear that many of the elements of Medieval Arthurian literature were introduced many years after the stories originated, and are highly unlikely to represent historical fact. Another problem rises out of the way in which the medieval society viewed history. Whereas the methodology of modern historical inquiry demands the constant independent reevaluation of past historical interpretation, medieval historians saw their role as more of a recasting of previously established historical "fact" with out independent reverification. The result is that historians cannot trust even so-called "histories" unquestioningly. The vast majority of modern historians now recognize historicity only in those works written before Geoffrey of Monmouth’s tremendously influential Historia Regum Britanniae (History of the Kings of Britain), in which Geoffrey synthesized a massive body of preexisting oral and written Arthuriana into a unified whole. In the process Geoffrey introduced numerous elements of which there is no prior record and must therefore be presumed unhistorical in nature (he appears to have entirely fabricated Merlin, for example). The works of Geoffrey and his translators and imitators, most notably Wace and Layamon, were so hugely successful and so widely accepted throughout England and the continent that historians can find no extant "histories" written after 1139 that are not based in part on Geoffrey himself or a work derived from Geoffrey’s work. The result is that historians looking for a "true" Arthur must examine only "pre-Galfridian" works, that is, those works written before Geoffrey (the term "Galfridian" deriving from Galfridus, Geoffrey’s transliteration of his name into Latin).
The pre-Galfridian sources that mention Arthur consist primarily of Saint’s Lives, some transcribed oral folktales, Welsh genealogies, and a few historical chronicles (for a complete collection of translations of references to Arthur in pre-Galfridian sources, see Coe and Young 1995). When historians weed out those sources which are mythological in nature (modern historians tend not to believe in giants and dragons and wizards, although there have been some attempts to view these stories as allegorical references to historical events), they are essentially left with a mere five pieces of written evidence which may be of some historical value: Y Gododdin, a collection of heroic death songs written in old Welsh and thought to be composed in the early 7th century, the Historia Brittonum, a British history composed around 830, the Annales Cambriae (also known as the Annals of Wales) completed in the mid 10th century, four scattered references to Arthurs and Arturs in Welsh genealogies, and, despite its conspicuous omission of Arthur, Gildas’s De Excidio Britanniae, as the only extant source of possible historical value written during the time period Arthur is supposed to have lived.
The Y Gododdin
Y Gododdin is a series of individual elegies for members of a war band of northern Britons of the Gododdin tribe who fell in a battle with the Saxons around the year 600. The consensus among historians is that the work must have been composed prior to 638, but no exact date can be fixed. There is only one reference to Arthur in Y Gododdin, in a passage describing a leader named Gordur (Gwawrddur in old Welsh):
He charged before three hundred of the finest,
He cut down both center and wing,
He excelled in the forefront of the noblest host,
He gave gifts of horses from the herd in winter.
He fed black ravens on the rampart of a fortress
Though he was no Arthur.
Translation by A. O. H. Jarman
This passage establishes that by middle of the 7th century at the latest (if we accept the current dating of Y Gododdin) there existed in Britain a conception of a historical figure named Arthur who was so greatly renowned that even a ruler who would charge 300 men cannot match up to him. The Y Gododdin is considered by many historians to be one of the most convincing pieces of evidence in favor of a historical Arthur, due to its early date, clearly historical conception of Arthur, and historically plausible content overall. Although some historians have questioned whether the reference to Arthur may have been added much later, there is no solid evidence for this view.
The Historia Brittonum
The Historia Brittonum, a 9th century historical chronicle compiling and synthesizing earlier sources no longer extant, contains two sections concerning Arthur. In chapter 56, the author lists twelve battles of highly debated location that Arthur supposedly won against the Saxons, including a certain battle of Mount Badon which is notable because it is also mentioned by Gildas and in the Annales Cambriae. Much has been made of the title the author gives Arthur in this section, dux bellorum. Literally "duke of battles" in Latin, the title is usually translated as "battle leader" or "leader of battles,” and is very similar to the sub-Roman title dux britanniarum given to the commander in charge of the defense of northern Britain. The dux bellorum title has been used to argue that Arthur was not a king but rather a war leader in the service of a king, perhaps Gildas’ Ambrosius Aurelianus. Alternatively, Arthur may have indeed been a king who (perhaps due to his prowess) was elected the war leader of an alliance of multiple petty kings. The second reference to Arthur occurs in chapter 76 of the Historia Brittonum in which the author discusses various wonders including a cairn with a pawprint of "the soldier" Arthur’s dog "Cabal" and the tomb of Arthur’s son "Anir," whom Arthur had slain himself, which changes size every time it is measured. Historians can generally make little of these last two references which seem more mythical than historical in nature.
In recent analyses, the historical accuracy of the Historia Brittonum has fallen under suspicion, especially regarding the list of Arthur’s twelve battles. Historians such has David Dumville have questioned the author’s motives, arguing that the chronicle is essentially a political text historically relevant only to the century in which it was written. One well known hypothesis, based principally on the rhyming battles at Celidon, Guinnon, and Badon, is that the list of Arthur’s battles may be based on an earlier Welsh poem, a theory that could indicate that some of the battles were fabricated, but has also been used to argue an earlier date for this section of the work. Thomas Green has attempted to link some of the battles to specific earlier poems that are clearly mythological in nature, claiming one was a mythical battle against trees and another was a legendary battle against were-wolves. Nevertheless, the Historia Brittonum is the most complete record of Arthur’s military career and has thus become an integral part of many attempts to identify the true Arthur.
The Annales Cambriae
The Annales Cambriae, a yearly record of important events recorded in a document used to calculate the dates of Easter, gives two brief references to Arthur which are important because years are given. The battle of Badon is listed as occurring in 516 and the "battle of Camlann, in which Arthur and Medraut fell" is given as occurring in 537. This is the first extant reference to the battle of Camlann and the "Medraut" mentioned is thought to be the earliest known reference to Mordred. The Annales Cambriae which cover the years from 447 to 957, claim that all the entries were written in the year they describe, which, if true, would make the work the most important evidence for a historical Arthur as the only contemporary source mentioning Arthur. However, there is significant evidence indicating that the earlier entries were written some years after the events occurred, and were most likely borrowed from other histories. Some have even argued that the Annales are based in part on the Historia Brittonum, but there is more convincing evidence that the Annales are completely divergent from the Historia.
The De Excidio Britanniae
Any historian examining the idea of a historical Arthur must account for Gildas’ omission of Arthur from his De Excidio Britanniae (The Ruin of Britain). It is important to note that the work does not claim to be a history. It reads more like a sermon, berating five "tyrants," Constantine, Vortipor, Aurelius Caninus, Maglo-cunus, and Cuneglasus, for abusing their power and transgressing against Christian values, in the process blaming present woes on their various sins. Gildas also praises one Ambrosius Aurelianus and credits him with victory over the Saxon invaders. What makes Gildas’ work so relevant is his mention of a victory at Mount Badon, which is directly credited to Arthur in the Annales Cambriae and the Historia Brittonum. Although Gildas does not explicitly credit the victory to Ambrosius, were there an absence of later evidence for an Arthurian ascription it would seem natural to read the text as awarding the battle to Ambrosius. There are many theories concerning the elusive victor of Gildas’ Badon. For many the omission of Arthur is one of the strongest proofs that the traditional Saxon-fighting Arthur did not exist as a historical figure. Others argue that Gildas merely left Arthur out, perhaps because Arthur did not have virtues Gildas wished to extol or sins he wished to criticize, or less convincingly, that Gildas had some form of personal enmity toward Arthur, and did not wish to credit his victory (yet if Gildas disliked Arthur so much why did he not simply include Arthur among the tyrants? Mark Davis argues that Gildas did so – see below). Alternatively, some theorists have maintained that Arthur and Ambrosius Aurelianus are one and the same, Arthur" possibly serving as a war title or deriving from a Latin clan name Artorius.
The Name "Arthur"
Finally, there are four occurrences of the name Arthur in Welsh genealogical histories beginning in the 7th century. In Senchus Fer nAlban (The History of the Men of Britain), a lengthy genealogy lists one Artur son of Conaing son of Aedan. In Vita Columbae (Life of Columbia) a passage in which King Aedan wonders which of his sons will be his successor describes Artur as Aedan’s son rather than his grandson. De Causis Torche na nDessi (Of the Causes of the Migration of the Dessi) mentions what appears to be a different Artur, son of Petr. Finally, Acallam na Senorach (The Colloquy of the Ancients) describes how a certain Artur, son of the king of the Britons, is humbled by mythical Irish folk hero Fionn and his band of warriors. These four references to Art(h)urs are the earliest know occurrences on anyone bearing that name, a fact which leads to two types arguments regarding a historical Arthur. The first, widely accepted, is that these kings and princes were named after "the" Arthur, who by the mid 7th century had gained significant renown. Second, various cases have been for each of these Arturs as "the" Arthur, despite the fact that they lived two centuries after other sources say Arthur lived. It is possible, however, that some of the deeds of these kings were incorporated into the body of Arthurian legend.
The Artorius Etymology
The third piece of evidence historians use when attempting to identify Arthur, is his name itself, which was not common among any of the groups competing for control of Britain. At present are two main etymologies for the name. The first is a Latin derivation, which theorizes that "Arthur" developed over time from the rare Latin name Artorius, meaning "plowman." Some linguists have tentatively demonstrated that the Brythonic tongue would have transmuted the Latin long "o" sound into more of a "u" sound, and there is some evidence of other words normally spelled with an "o" in Latin, appearing with a “u” in Welsh. If one accepts this argument it becomes easy to see Artorius developing into “Artur,” as Latin endings such as -ius and –ium were regularly dropped in the Welsh vernacular (Londinium to "London" for example). If one accepts this Latin derivation of the name “Arthur,” the natural tendency is to look for a Roman Arthur, or at least an Arthur who would have a Roman name. The obvious candidate is Gildas’ Ambrosius Aurelianus or one of his lieutenants, perhaps a relative. Perhaps Artorius was a third name of Ambrosius, if one sees Ambrosius as "the" Arthur, or the name of one of his generals who won the battle at Badon. Another candidate is the only recorded "Artorius" in historical Britain, a 2nd century Roman soldier named Lucius Artorius Castus who commanded a unit of auxiliary cavalry on a campaign to crush a native uprising against Roman authority. This theory is compelling and gained considerable acceptance earlier this century, because this is one "Arthur" we know for certain existed. There is also some evidence that Castus may have borne the title of dux britanniarum, which could cast new light on the mysterious dux bellorum title given Arthur in the Historia Brittonum. However, in recent studies the case for Lucius Artorius Castus as the original Arthur has been discounted, for lack of evidence linking it to any of the later Arthurian traditions, as being based on the name alone. Furthermore, as Thomas Green puts it, "the Castus theory requires us to see Arthur as a figure who was first of all historical, then became totally absorbed into Celtic folklore and then, at a later point, was historicised into an entirely different era from that of his origins." Green points out that while this is not inconceivable, it is highly unlikely. It has also been suggested that "the" Arthur may have been one of Castus’ descendants, as in Roman society the middle name was often the clan name, and would likely have been passed on, but with no evidence to support it this idea remains pure conjecture.
The "Bear" Etymology
While the Latin Artorius etymology for Arthur’s name is still regarded as perfectly plausible, an equally plausible Celtic derivation has gained wider acceptance in recent years. In the old Celtic language the word art means "bear," causing some theorists to suggest that the name "Arthur" developed from Art-gur meaning "bear-man" (gur is a Brythonic word meaning "man"). In various early Welsh works art serves as a figurative synonym for warrior. Furthermore, a gloss added to a late 12th or early 13th century manuscript of the Historia Brittonum claims that Arthur means Ursus Horribilis, "horrible bear," in Latin. The result of these various observations is that many historians have little trouble accepting the bear etymology of the name "Arthur." One of the most common arguments based on this derivation is that "Arthur" may have been a war title of some sort, a sort of Celtic nom de guerre.
Arthur’s name has become the starting point for many investigations of the possibility of a historical Arthur. Each of the four occurrences of the name "Artur" in Welsh genealogical sources have at some point been argued as the "true" Arthur. Most recently, D. F. Carroll has argued that the Artur son of Aedan mentioned in Vita Columbiae is "the" Arthur, a Scottish king who ruled the region of Gododdin and fought against the Picts rather than the Saxons. Several other theories have been based on three kings named "Arthwys" (pronounced "Artweese"), but only one ruled during the time Arthur is supposed to have lived, one ruling before and one ruling well after. None seem a very strong candidate, and in the end, these theories are essentially based on name only.
Other theories have been built on the assumption that a historical Arthur is substantiated in the histories, but under a different name. Perhaps the most famous of these is the theory put forth by Geoffrey Ashe in his book The Discovery of Arthur, in which he argues that the main prototype for the Arthur of medieval legend was a king who ruled in Brittany known to history only as Riothamus, a title which means "greatest king." Records clearly indicate that a king of Brythonic stock called Riothamus fought against the Visigoths in central France in the year 468. Ashe proceeds to argue that Geoffrey of Monmouth’s account of Arthur’s Gallic campaign was actually based on historical fact. Despite the fact that Ashe’s theory entirely discounts the pre-Galfridian tradition of Arthur fighting against the Saxons at mount Badon, this theory has convinced many.
Several theories attempting to identify Arthur as someone recorded under a different name have focused on the dynasty founded by a king named Cunedda that ruled the region of Gwynedd, located in north-eastern Wales, during the time of Gildas. The most convincing of these theories, proposed by Mark Davis, suggests that Arthur may have been Cuneglasus, one of Gildas’ five tyrants. Davis starts by noting the similarities between Gildas’ tyrants and the kings that immediately follow Arthur in Geoffrey of Monmouth’s History of the Kings of Britain. Gildas’ describes five tyrants: Cuneglasus, Constantine, Vortipor, Aurelius Caninus, and Maglo-cunus. Geoffrey lists four kings very similar in name: Constantine, Vortipore, Aurelius Conan, and Malgo, but does not mention Cuneglasus at all. Davis argues that this is because Arthur and Cuneglasus are one and the same. Davis’ most convincing argument, however, is that Gildas repeatedly calls Cuneglasus "Urse," that is, "bear." Citing the Celtic etymology for Arthur, Davis suggests that Gildas is discussing Arthur, merely translating his name into Latin.
The name Cuneglasus itself is Gildas’ Latinization of Cynglas, the great-grandson of Cunedda and the first cousin of Malgo, Gildas’ Latinization of Maelgwn. After Cunedda’s death his kingdom was divided up into the three smaller kingdoms of Rhos, Gwynedd, and Meirionydd. During the time of Gildas, Cynglas/Cuneglasus ruled Rhos, Maelgwn/Malgo ruled Gwynedd, and their second cousin Cadwaladr ruled Meirionydd. August Hunt argues that Cadwaladr is "the" Arthur based on the fact that his name is composed of the Welsh words Cad, meaning "battle," and gwaladr, meaning "lord" or "leader," a meaning which is identical to that of the dux bellorum title given to Arthur in the Historia Brittonum. Hunt also points out that there are three possible Camlanns in north-eastern Wales, two of which are located in Meirionydd, a hill called Afon Gamlan and two other sites known as Camlan, although this fact could be used to support any number of possible Arthurs. A third theory placing Arthur in the house of Cunedda, suggested by Graham Phillips and Martin Keatman, identifies Arthur as Owain Ddantwyn, Cuneglasus’ father, although this seems less likely.
Although theories abound regarding various alternated identities for a historical Arthur, it should be noted that many people remain convinced of the existence of the Arthur depicted in the Historia Brittonum, the Annales Cambrae, and even to some extent Geoffrey of Monmouth’s Historia Regum Britanniae. Scholars who buy into this traditional theory believe they can say with some assurance that there did exist a Brythonic warrior named Arthur who fought against the Saxons around the turn of the 6th century AD and who played a part in a battle at a place called Badon Hill. Obviously, this view has the weight of 1500 years of tradition. Much of Britain appears to have believed this as early as a century after Arthur’s supposed death. Also, one can always argue that there may be sources that better authenticate Arthur which are now lost or have not yet been found.
Finally, there are those scholars who argue against a historical interpretation of Arthur. It has been justly pointed out that most normal historical inquiries would be dropped in the face of such a lack of evidence as is the case with Arthur. Many are convinced there is no case. David Dumville declares, "The fact is that there is no historical evidence about Arthur; we must reject him from our histories and, above all, from the titles of our books." J. N. L. Myres sums up this position, writing, "The fact is that there is no contemporary or near-contemporary evidence for Arthur playing any decisive part in these events at all. No figure on the borderline of history and mythology has wasted more of the historian's time."
Arthur obviously had to come from somewhere, however. One common theory, argued convincingly by Thomas Green, is that Arthur was not originally a historical figure that was mythologized, but rather a mythical character that was later historicized. Using the bear etymology, some scholars have argued that Arthur was originally a Celtic bear spirit similar to Artaio ("Bear-like one"), a spirit worshipped by an ancient Celtic Bear cult in what is now Switzerland. However, there is no known corroborating evidence of any such cults in the British Isles. Another theory postulates that Arthur was derived from a Celtic woodland spirit, based on pre-Galfridian Welsh sources that portray him as a "woodland adventurer" and locate his court at Kelliwic ("forest grove"). It has even been suggested that Arthur was originally a Raven spirit, based primarily on the mysterious line from Y Gododdin, "He fed black ravens on the rampart of a fortress," and the passing similarity between "Arthur" and the Welsh arrdhu, "very black." Green urges caution, however: "Just as an almost infinite number of historical prototypes for Arthur can be identified with enough enthusiasm, it seems very likely that a similar number of mythical prototypes can also be identified."
So where do all these theories leave the modern scholar in his or her examination of the problem of a historical Arthur? There is a reason why all theories must proceed from the few extant pre-Galfridian sources. While we may harbor suspicions about these sources, they cannot be automatically discounted. However, neither can we simply make the a priori assumption that a historical Arthur did exist and proceed merrily to our various theories. History is rarely an exact science. We must simply make the best conclusion we can with the evidence available, and in this case, there happens to be very little evidence. We must seek the most probable theory, and even in the case of Arthur there are certain probabilities, though fewer in number. If we allow the vast body of evidence to lead us to the most probable conclusion, I believe we can say, with assurance, that there was a historical person called Arthur who won renown as a warrior in the British Isles some time before the mid 7th century. But that is all.
Ashe, Geoffrey. The Discovery of King Arthur. Anchor Press/Doubleday. New York. 1985.
Carroll, D.F. Arturius: A Quest for Camelot. Goxhill. 1996.
Coe, John, and Young, Simon. The Celtic Sources for Arthurian Legend. Llanerch. Felinfach. 1995.
Davis, Mark. “King Arthur and Cuneglasus.” http://www.angelfire.com/md/devere/urse.html. 1998.
Dumville, David. Histories and Pseudo-Histories of the Insular Middle Ages. Aldershot. Great Yarmouth. 1990.
Green, Thomas. “The Historicity and Historicisation of Arthur.”
Hunt, August. “Arthur and Camlann.” http://freespace.virgin.net/david.ford2/ArtCam.html. 1998.
Jarman, A. O. H., trans. Y Goddodin, Britain’s Oldest Heroic Poem. Gomer. Llandysul. 1988.
Myres, J. N. L. The English Settlements. Clarendon Press. Oxford. 1986.
Phillips, Graham, and Keatman, Martin. King Arthur: The True Story. Century Random House. London. 1992.
Rhys, John. Studies in the Arthurian Legend. Clarendon Press. Oxford. 1891.