The Historicity and Historiography of Arthur: A critical review of King Arthur: Myth-Making and History by N. Higham, and The Reign of Arthur: From History to Legend by C. Gidlow
Howard M. Wiseman
School of Science, Griffith University, Australia
©2007 by Howard M. Wiseman. All rights reserved. This edition copyright © 2007 by The Heroic Age. All rights reserved.
Introduction
§1. On the question of whether there was an historical Arthur,1 the academic pendulum never swung further to the affirmative than in the massive work of John Morris, The Age of Arthur (1973). It was quickly recognized by other academics, however, that this work had grave methodological flaws, and a reaction was inevitable. It came with greatest force from David Dumville (1990), who famously wrote: "... there is no historical evidence about Arthur; we must reject him from our histories and, above all, from the titles of our books." This has become the new orthodoxy, with the academic pendulum staying on the negative side to a great degree.
§2. Recently two books have been published which have much to say about the historicity of Arthur, the early writings about him in histories (or at least in texts purporting to be histories), and the development of the Arthurian legends. The books have similar, but significantly different titles; both of them disobey the letter of Dumville's command, but only one disobeys its spirit. King Arthur: Myth Making and History (2002) by established historian Nick Higham of Manchester University, is well-written, scholarly, and, as would be expected, firmly in the tradition of Dumville. The Reign of Arthur: from History to Legend (2004) by Oxford history graduate Chris Gidlow, is not so well-written in places, less scholarly2 and, as might be expected, in favour of an historical Arthur.
§3. However, The Reign of Arthur is not from the usual stable of popular books on this subject, identifying the 'real King Arthur' with reasoning that often does not even deserve to be called tenuous. In particular Gidlow does not make the mistake of John Morris and others, in accepting late (11th century and later) Welsh saints' lives, stories, and poems as legitimate sources for 5th and 6th century history. Rather, Gidlow critically examines early medieval British sources, in particular Gildas' de Excidio Britanniae, the Historia Brittonum (ostensibly by Nennius), the anonymous Annales Cambriae, and Y Gododdin (ostensibly by Neirin). It is on the basis of these texts alone that Gidlow supports the historicity of Arthur.
§4. Higham examines essentially the same corpus of texts as does Gidlow. Unlike Gidlow, he states that "The key question ... is not 'Was Arthur an historical figure?' ... [but rather] 'What role was Arthur intended to perform ... in the texts of the central Middle Ages ...?'" (8). But he recognizes that his second question bears keenly on his first, and he does not shrink from using his answers to the second to imply a negative response to the first. Nor does Higham shrink from disparaging those who have proposed a less negative response; he would, perhaps, include Gidlow with other historians who have used a "synthesizing or reconciling" approach in which different texts are "ransacked for individual facts ...without explicit notice of [their] ultimate purpose ... within the narratives ...." (16).
§5. In the later parts of their books, both Higham (Chapter V) and Gidlow (Part II) discuss the development of the Arthurian legend from the 11th century on, as both recognize that this is best discussed separately from the earlier sources. (This is the less well-written part of Gidlow's book, and contains some minor errors of fact.3) Both authors follow these discussions with a summary analysis of the historicity of Arthur based on the earlier sources. Reading the books in conjunction, one cannot accept Gidlow's affirmation of the reality of Arthur's reign. But one is also led, just as strongly, to reject Higham's confident denial of Arthur's historicity.
§6. The remainder of this article is devoted to this issue of Arthur's historicity, and how it relates to the early historiography of Arthur. Under the headings of the three principle sources, I will critique some of the arguments of both authors, before presenting my own conclusions. But first I should, for the benefit of those who may wish to buy these books, comment on less controversial aspects. Both books are generously illustrated with black and white plates including many Arthurian sites and, in Gidlow's book, documents. Both books are also well-provided with maps, Higham's especially so; in terms of charting the archaeological evidence for Britain in the time of Gildas, Higham's King Arthur compares favourably with any other book I know.
de Excidio Britanniae
§7. This history and social commentary by Gildas was written perhaps a generation after the battle of Badon, and so would be nearly contemporary with the historical Arthur if he were real.4 If Gildas had mentioned Arthur then his historicity would be beyond doubt. That he did not is called by Higham a "conundrum" (31) for those who believe in an historical Arthur. The truth is that it would have been more surprising if Gildas had named Arthur. In his historical account of Britain after Roman rule,5 Gildas names not one single Briton. The sole individual from Britain he does name is Ambrosius Aurelianus, whom he expressly identifies as a Roman.6 Ironically, given Higham's insistence on reading works in context, it is Gidlow who points out (96-7) that to expect to find Arthur's name in the historical (second) section of de Excidio Britanniae would be to ignore its nature: a providential history of the British people in biblical terms, in which individual Britons are largely irrelevant.
§8. Gidlow presents a fairly standard analysis of Gildas' history, although he has some interesting observations to make about the status of Ambrosius' parents, and about the possibility that the five tyrants whom Gildas castigates in the third section of his work did not rule simultaneously. For what Gildas can tell us of Arthur, Gidlow emphasizes the "final victory of our country" at the "siege of Mount Badon" and its aftermath: the peace in which, at first, "kings ... kept to their own station" (Gildas De Excidio 26). This period is what Gidlow calls the reign of Arthur: a time when the British commander at Badon (whatever his name or title) supposedly imposed order upon the kings of the Britons. Gildas tells us that the good kings were succeeded by a generation that "... has experienced only the calm of the present." Of his own time, Gildas observes that "all the controls of truth and justice have been overthrown" so that "external wars may have stopped, but not civil ones." Gidlow also sees Gildas' use of publici (officials) and rectores (governers) as potential evidence for civilian authority at the provincial level in Gildasian Britain, and discusses Arthur's possible relation to any such authority.
§9. Higham, by contrast, has his peculiar reading of Gildas, which he detailed in The English Conquest (1994, 211) and which he reiterates briefly in King Arthur. He interprets Gildas' "final victory" as meaning that, since the battle Badon, the Britons had known only defeat against the Saxons, and that the "calm of the present" was one in which there was a Saxon overlord of Britain. The implications of Higham's interpretation for an historical Arthur are obvious; as he says in The English Conquest, "Not only did Arthur himself not exist but the age which led to his invention was no less fictional." There is no space here to address Higham's reasoning behind these conclusions, but it has already been severely criticised by others (Snyder 1996; Howlett 1996). Higham's deduction that Gildas wrote in 479×484 (two generations earlier than most historians think7) is frankly incredible. For example, Gildas implies that Hadrian's Wall was built in the early 5th century,8 which would have been during the childhood of the elderly still living in 479×484. Thus I think it is safe to say that Gidlow's analysis of Gildas has more to recommend it than Higham's.
Historia Brittonum
§10. The Historia Brittonum (HB) was almost certainly written in Wales, very probably in Gwynedd, in about 830 A.D. Higham argues compellingly that it had two purposes: to re-establish, contrary to Bede, the Britons as a people of the Lord in providential history, and to establish the primacy of the kings of Gwynedd among them. In some versions there is a preface ostensibly written by Nennius, who claims that, in writing the HB, "I ... heaped up everything that I could find." Higham, following Dumville (1986), dismisses this preface, and cites Howlett (1998) to support the idea that the work is actually written in a "sophisticated 'biblical style'" (120). I have criticized Howlett's methodology elsewhere (Wiseman 2003), and Higham himself admits later (128) that at least in one section of the HB, "several source stories are rather crudely interleaved." While Higham is correct to emphasize the "political and ideological concerns" of the HB, its many contradictions, often openly admitted by its author, show that the author did not simply invent his material, but rather made use of diverse pre-existing records.
§11. The HB is the oldest reliably-dated text to mention Arthur, in a section that lists twelve victories in which he led the kings of the Britons, culminating in the battle of Badon. Higham argues that the author of the HB constructed this Arthur to form a doublet with Patrick (the sections devoted to Patrick precede the Arthurian one). Patrick is explicitly compared to Moses in the HB, and Higham sees Arthur as a Joshua figure. Just as Moses, the spiritual leader of the Israelites, was followed by Joshua, their leader in battle, so Saint Patrick was followed by Arthur, the leader in battle (dux bellorum) of the Britons against the Saxons. Thus the Britons are shown by analogy to be people of the Lord.
§12. Higham's analysis here is powerful and novel. But it does not seem to justify his claim that "the extent to which [ideological and political concerns] dominate the text ... is nowhere more clear than in the [Arthurian] passage" (165). If the author of the HB had been free to choose any military leader to follow Patrick, then surely these overwhelming concerns would have indicated a king of Gwynedd. The HB already mentions one king of Gwynedd around this time: Mailcun, who supposedly reigned after Arthur. Mailcun's father (Catgolaun Lawhir according to later genealogies) is thus one obvious candidate to have played Arthur's role. Ambrosius, the high-king of the Britons who precedes Arthur in the HB, is linked with Gwynedd, and Higham suggests that even Patrick may be (141). But the HB posits no connection at all between Arthur and Gwynedd, and in fact links him with locations elsewhere in Wales instead (in a different section of the HB). The simplest explanation for this apparent lack of political acumen on the part of the author of the HB is that his choice was not free, because the Britons in 830 already believed that their forebears at Badon had been led by a man named Arthur.9
§13. Higham claims to the contrary that before 830 Arthur was essentially a "mythological" warrior figure (152), the reason being that in the HB he is associated with a couple of geological marvels, and credited with single-handedly slaying 960 of the enemy in one charge. But as Gidlow points out (47), the most mythological material in the HB is associated with the undoubtedly historical Ambrosius, who appears as a fatherless boy making prophesies on the basis of the actions of magical worms before a council of wizards! Many of the historical figures appearing in the HB have mythic attributes, while Gidlow also points out (19-20) that even the sober historian Bede reports miracles (including a length-changing sarcophagus) in relation to historical English kings. Gidlow goes on to argue that the HB can be read as saying that 960 Saxons were slain by a force directly led by Arthur (rather than by forces attached to the kings of the Britons). However, this argument is not convincing. Neither is Gidlow's discussion of the sources of the Arthurian battle list, predicated as it is on the reliability of the list.
§14. Gidlow's most striking contribution regarding the HB is to identify an apparent double standard among many modern historians (67-70, 126-9), namely of accepting that Gwynedd was ruled by Mailcun but denying that Arthur was the victor at Badon. Gidlow's comparison of the two cases can be summarized as follows.10 The powerful tyrant Maglocunus ( = Mailcunus = Mailcun) is historical; Gildas condemns him at length, but he never says where he ruled. The claim that Mailcun "reigned as a great king among the Britons, that is in the region of Gwynedd" first appears in the HB. Similarly, the battle of Badon is historical; it is one of the key British victories in Gildas' history, but he never says who led the Britons there. The claim that Arthur led the kings of the Britons in the battle of Badon also first appears in the HB. Now the purpose of the HB (to promote the primacy of the kings of Gwynedd among the Britons), and the way the phrase connecting the "great king" Mailcun with Gwynedd appears to be a gloss, suggest that this connection was invented by the author of the HB. Nevertheless almost all historians, Higham included (125), accept Gildas' Maglocunus to be the King of Gwynedd. By contrast, the battle of Badon was almost certainly connected with Arthur prior to the composition of the HB, as I discussed above. Yet many historians, Higham included, insist that Gildas' battle of Badon was not won by Arthur.
Annales Cambriae
§15. Higham makes a strong case that these Welsh Annals were compiled in the court of Owain of Deheubarth in 954 or 955,with purposes distinct from the HB: first, to accommodate the English as fellow Christians, and even as acceptable overlords, of the Britons; and second, to magnify Owain and his kingdom. Given the sparseness of 5th and 6th century entries in the Annals, Arthur, with two entries, emerges as an even more important figure here than in the HB. The first (c. 516) is the battle of Badon, which portrays Arthur as a Christ-figure and, unlike the HB, omits to identify his enemy as the Saxons. The second (c. 537) is the battle of Camlann in which Arthur and Medraut fell and there was plague in Britain (the first of many). This was meant to signify, according to Higham, the beginning of the end of the Age of Saints in Britain.
§16. Arthur's importance in the Annals is, according to Higham (214), because King Owain claimed descent from him. The 13 intervening generations in the genealogies would put this ancestral Arthur in the late 6th century. But this Arthur also appears as the great-grandson of Vortipor, the king of Dyfed whom Gildas names as reigning after Badon. Thus, if Higham is correct, then the compilers of the Annales Cambriae would have succeeded in their purpose (of linking Owain with the Arthur of Badon) only if their intended audience was quite ignorant of Gildas' work—surely a dangerous assumption. Higham is on safer ground when he demonstrates (201) that the author of the Arthurian entries in the Annals was almost certainly drawing upon the Arthurian section of the HB plus the section that immediately follows it. Gidlow's claim (60) that the Annals are independent of the HB thus appears naïve, as do his arguments (64) that the dates of the Arthurian entries may be based on near contemporary traditions. Higham explores a number of hypotheses for these dates, including one proposed by me (Wiseman 2000). None of them point to any genuine evidence predating the Annals themselves.
Conclusion
§17. The Reign of Arthur concludes with a chapter summarizing the evidence for an historical Arthur, and Gidlow's final paragraph states:
The victor of Mount Badon was a real person .... We have ... no reason to think that [the Gododdin and the HB] are wrong in granting him the name Arthur. This man, this Arthur, commanded kings, at a time when private citizens and public officials kept to their allotted positions. In this sense, therefore, it is reasonable to say that the generation which witnessed the siege of Badon did indeed live in the 'reign of Arthur.'
Gidlow is right to emphasize (implicitly here) that the key source for establishing the context of an historical Arthur is Gildas. And he has argued convincingly that historians, who are willing to accept the HB's claim that Mailcun reigned in Gwynedd, have better reason to accept its claim that Arthur was the victor at Badon. But unfortunately, like Morris before him, Gidlow over-steps the bounds of reasonable inference in his statements about the 'reign of Arthur'. Gildas really gives us no idea what, if any, political role the leader of the Britons at Mount Badon played after the victory there, and neither the Gododdin nor the HB hint at any role for Arthur other than a military one.
§18. King Arthur similarly ends with a long answer to the question 'Did King Arthur really exist?' (271). Higham, largely following Padel (1994), explains the Arthur of the HB by positing an extraordinary hybrid figure, "derived ultimately from a Roman-period Artorius"11 mixed with the huntsman-hero type of folklore, associated with the Old Welsh arth (bear), and "perhaps even conflated with some lost pagan deity" (272). As Higham admits, the only candidate for Artorius is the 2nd century prefect Lucius Artorius Castus. Lest the reader think otherwise, it is worth emphasizing that the recorded career of this Roman Artorius, as reconstructed by Linda Malcor (1999), has nothing in common with that of Arthur in the HB beyond being a dux in Britain. Nor is there any evidence that this Artorius was remembered in Britain in the 646 years between when he was recalled to the continent, and when the HB was written. The Roman name Artorius is not even recorded in Britain.
§19. By contrast, the name Arthur (or Arturius or Artuir) was given to at least four high-status individuals in the late 6th or early 7th century (Bromwich, 1975-6). As shown in a recent analysis by Dark (2000b), all four were associated with the interface of British and Irish society. As Dark says, this phenomenon demands an explanation. The one presenting itself most strongly is that they were named after an historical 6th century "prototype" who also had such an association. If, as Higham accepts, Arthur of the HB was derived ultimately from a real person of that name, it seems more reasonable to suppose that the man in question was this 6th century Arthur (as Dark proposes), rather than the 2nd century Roman Artorius. The probable birth dates of the four well-attested Arthurs indicate that this "prototypical" Arthur would have flourished prior to about 570.12 It cannot escape notice that if such an Arthur had a floruit in the first half of the 6th century, he could weel have been the historical Arthur, the victor of Mount Badon.13 One is tempted to conclude that this is why Higham is critical of the "protypical Arthur" argument, calling it "weak, at best unproven" (77).
§20. Higham's closing paragraph serves only to engender further doubts about his objectivity in denying Arthur's historicity. Having entitled his book King Arthur, and having phrased the question of Arthur's existence in those terms, he is then compelled to admit that the HB did not invent Arthur rex at all, only Arthur miles (the soldier). The entitling of the book and the phrasing of the question reveals Higham's implicit attitude: that the HB and other early texts are to be examined from the perspective of the much later sources that invented King Arthur (Geoffrey of Monmouth in particular). The fact that this (undeniably fictional) King Arthur is a "logical development" (273) of Arthur the battle leader seems, in Higham's mind, to confirm that the latter is equally fictional. This shows a bias that Gidlow specifically warns against (324).
§21. In summary, Higham and Gidlow each present some note-worthy, and some flawed, analyses of the historiography of Arthur. On the question of Arthur's historicity, Gidlow errs too far in the affirmative, and Higham too far in the negative. If we seek a balanced answer to this question, we must turn elsewhere. It is easiest also to turn to the time before the 1970s, with its opposite extremes of Morris and Dumville. For example, Sheppard Frere wrote in 1967:14
The evidence is sufficient to allow belief that [Arthur] had a real existence and that he was probably the victor of Mount Badon. It is likely that he succeeded Ambrosius in the leadership [of the Britons]; indeed, he is called dux bellorum in the Historia Brittonum, which suggests a memory of late Roman military titles, and may indicate some sort of unified command arranged between several petty kingdoms.
§22. The qualifiers in this quote (in particular that the evidence allows, not requires belief) seem to balance fairly skepticism with the desire to say something definite about Arthur.15 That said, Frere's interpretation of the title dux Bellorum may have to be reassessed in the light of Higham's analysis. Also, the assumption that petty kingdoms were the only form of political organization among the Britons has been questioned, by Gidlow and others (Dark 2000b). Do difficulties such as these mean that we would do better to follow Dumville's directive, to reject Arthur from our histories? I prefer the view of Snyder (1998, 255), who asks historians to admit that "there may well have been an historical Arthur," to stop griping about quests for the historical Arthur, and to contribute what they can: "a better understanding of the period and place in which Arthur ... may have lived."
Acknowledgments
I would like to thank N. Wiseman for a critical reading of the manuscript.
Notes
1. By 'an historical Arthur' I mean a man named Arthur who led the Britons against the Saxons at the battle of Badon (in the late 5th or early 6th century), which was celebrated by the nearly contemporary historian Gildas. Of course this is not to presume that historical figures not so prescribed did not contribute to the figure of Arthur in pseudo-history or legend. But to allow for the possibility of the conclusion that there was no historical Arthur, it seems to me necessary to define that person in terms of the unique role which (i) is a role attributed to a man named Arthur in the earliest text (the Historia Brittonum) that undoubtedly inspired the figure of Arthur in later pseudo-history and legend, and (ii) is a role that undoubtedly existed historically. [Back]
2. Gidlow's bibliography is considerably shorter than Higham's, and citations are rare in the text. Another source of frustration for the reader is frequent inaccuracies in the index of The Reign of Arthur. [Back]
3. For example, he describes the Patrician Liberius as Justinian's general in southern Gaul (286). Liberius could perhaps be described as Theodoric's general in southern Gaul, or (later in his career) as Justinian's general in southern Spain, but not as Gidlow states it. [Back]
5. That is, a period of perhaps a century and a half following the death of Magnus Maximus in 388. [Back]
6. Moreover, Gildas identifies him as the "last of the Romans", so if Arthur had been the leader at Badon, he would have been, in Gildas' eyes, a Briton. [Back]
7. See for example Dumville's analysis (1984), where he estimates a date of ca. 545 which, he says, agrees with the "usual broad dating" of the second quarter of the 6th century. [Back]
8. Gildas described its building roughly in the middle of the period 388-425×435 according to Higham's chronology (1994, 137). [Back]
9. Of course this does not prove that he was named Arthur. But, as Gidlow points out, if he was not named Arthur we have to suppose that the Britons forgot the name of their leader in this, their greatest victory, even though they did not forget Vortigern and Ambrosius, who lived before him, or Mailcun who lived shortly after him. [Back]
10. One might object that this is an unfair comparison because a region called Gwynedd certainly existed, whereas there is no independent evidence that a man named Arthur existed. However the name Gwynedd (Venedotia) is not attested before the end of the 6th century, while at the time of Mailcun the earlier division of this region between the Ordovices and the Deceangli was probably still in force—see Dark (2000a). Also, there is other evidence, albeit indirect, that there was a famous Arthur in the early or mid 6th century. This is discussed in the conclusion section below. [Back]
11. Artorius would have been rendered as Arthur in Old Welsh. [Back]
12. Dark raises the possibility that one of these four, the Arthur of Dyfed (discussed earlier in the context of the Annales Cambriae), might have been the prototype. However this is doubtful; as Dark notes, following Bromwich, this Arthur was most likely born in the 570s. His birth is thus very unlikely to have been a generation earlier than the earliest of the remaining three, Artuir mac Aedan of Dalrida, whose death in battle was in 590×596. Conversely, Aedan probably died in 604×612, so even if he was 74 at the time (as the annals of Tigernach claim), his son Artuir is unlikely to have been born a generation earlier than Arthur of Dyfed either. [Back]
13. Again, see Note 1. In this context, it should be remembered that high-status Irishmen certainly lived among the Britons in the heart of 5th century Britain, as evidenced by the inscribed stones of Ebicatos and Cunorix, in Silchester and Wroxeter respectively (Snyder 1998; Dark 2000a). It should also be noted that although Dark (2000b) proposes that the prototypical Arthur is the basis for Arthur of the HB, he regards the latter as a legendary figure, and so rejects the idea of "an historical Arthur". [Back]
14. Higham also quotes Frere in King Arthur (26), but only to make light of him. [Back]
15. Note that Frere's first sentence makes a distinction between the existence of Arthur, and his having been the victor at Badon. To make his statement comply with the terminology of the present article, it would be necessary to alter it. Perhaps the following would be a fair rendering: Based on the evidence, it is plausible that there was a renowned "prototypical Arthur." If so, this Arthur was probably the victor of Mount Badon, in which case there really was an "historical Arthur." [Back]
Works Cited
Bromwich, R. 1975-6. Concepts of Arthur. Studia Celtica 10/11:163-81. [Back]
Dark, K. 2000a. Britain and the end of the Roman Empire. Gloucestershire: Tempus. [Back]
— — —. 2000b. A famous Arthur in the sixth century? Reconsidering the origins of the Arthurian legend. Reading Medieval Studies 26:77-95. [Back]
Dumville, D. N. 1984. The chronology of De Excidio Britanniae, Book I. In Gildas: New approaches, ed. M. Lapidge and D. Dumville, 61-84. Woodbridge: Boydell. [Back]
— — —. 1986. The historical value of the Historia Brittonum. Arthurian Literature 6:1-26. [Back]
— — —. 1990. Histories and pseudo-histories of the insular Middle Ages. Aldershot: Variorium. [Back]
Frere, S. 1967. Britannia. Massachusetts: Harvard U. Press. [Back]
Gidlow, C. 2004. The reign of Arthur: From history to legend. Gloucestershire: Sutton. [Back]
Gildas. 1978. Gildas: The Ruin of Britain and other documents. Edited and translated by M. Winterbottom. London: Phillimore. [Back]
Higham, N. 1994. The English conquest: Gildas and Britain in the fifth century. Manchester: Manchester University Press. [Back]
— — —. 2002. King Arthur: Myth making and history. London: Routledge. [Back]
Howlett, D. R. 1996. Review of The English conquest: Gildas and Britain in the fifth century, by N. Higham. Arthuriana 6:371-74. [Back]
— — —. 1998. Cambro-Latin compositions: Their competence and craftsmanship. Dublin: Four Courts Press. [Back]
Malcor, L. A. 1999. Lucius Artorius Castus part 1: An officer and an equestrian. The Heroic Age 1. [Online journal]. Available at http://www.heroicage.org/issues/1/halac.htm. [Back]
Morris, J. 1973. The age of Arthur. London: Weidenfeld & Nicholson. [Back]
Padel, O. 1994. The nature of Arthur. Cambrian Medieval Celtic Studies 27:1-31. [Back]
Snyder, C. A. 1996. Review of The English conquest: Gildas and Britain in the fifth century, by N. Higham. Arthuriana 6.3:69-71. [Back]
— — —. 1998. An age of tyrants: Britain and the Britons A.D. 400-600. Pensylvania: Pennsylvania State U. Press. [Back]
Wiseman, H. M. 2000. The derivation of the date of the Badon entry in the Annales Cambriae from Bede and Gildas. Parergon 17.2:1-10. [Back]
— — —. 2003. Review of Cambro-Latin compositions: Their competence and craftsmanship, by David Howlett. The Heroic Age 6. [Online Journal]. Available at http://www.heroicage.org/issues/6/Wiseman.html. [Back]