Can archaeology help us to find Arthur?

Given that there are no contemporary documents telling us about Arthur, what about archaeological remains? Are then any discoveries that date from the period he is supposed to have lived (some time between the mid fifth century and the mid sixth) that might show that he really existed? While some have sought to document a period at which new Anglo-Saxon settlements ceased to be created in the belief that this is evidence for Arthur’s victories over the Saxons, I am more concerned here with the more positive claims of inscriptions, graves and locations for Camelot. And not all of them are necessarily preposterous and put forward by fringe writers: respectable academics have also been seduced by the blandishments of the Arthurian romances.


The Artognou stone

The Artognou stone: not evidence for Arthur at all


In 1998, an unexpected find was made by archaeologists from the University of Glasgow uner the direction of Chris Morris during their excavation at Tintagel, a site full of Arthurial allusions. Following Courtney Arthur Ralegh Radford’s (1900-1999) excavations there in the 1930s, his interpretation of the site as a “Celtic monastery” had been generally accepted by the academic community. However, following a grass fire on the headland in 1983, new excavations were undertaken by Chris Morris for English Heritage from 1990 onward.

These new excavations showed that Radford’s model was wrong: far from being a community of ascetics living a life of seclusion on an isolated headland, Morris showed that the place had extensive contacts with the outside world and that it was more urban than monastic. This was a shock: there weren’t supposed to be towns in ‘Dark Age’ Britain and the changed interpretation has still not been accepted in some quarters, particularly by advocates of “Celtic Christianity” as a more ‘authentic’ form of the religion than the major modern churches. But this was not the really controversial discovery: it was the finding of an inscription re-used in a seventh-century deposit.

The inscription was made on a piece of slate measuring 350 × 200 × 10 mm, which had been re-used as the cover of a drain. It carried two separate inscriptions. One, in large letters, only survived in part and appears to read …AXE…, which may be part of the Roman name Maxentius or some other name or word. It was the second part of the inscription, though, that caught the attention of the press. Reading + PATERN… COLIAVIFICIT… ARTOGNOV… COL… FICIT…, it presents all sorts of challenges to interpretation, but contains several names: Patern[us], Col (or Coliau) and Artognou. The first is a Roman name, the second and third Celtic. It was the presence of the phoneme Art… in the third name that excited the media.

Although Charles Morris was entirely cautious in his announcement of the discovery, pointing out that the inscription did not name Arthur, others were not so careful. While we might excuse the popular press (and, perhaps, especially the tabloids) for exaggerating and misinterpreting the meaning of the inscription, it is difficult to understand how such august bodies as the Archaeological Institute of America were able to promote it under the heading King Arthur was Real?. The inscription – whatever it actually says (perhaps ‘Patern[us, son of] Coliaw, had this made for Artognou: Col made it’) – does not mention Arthur at all.

The supposed tombstone of Arthur from Mynydd y Gaer

The supposed tombstone of Arthur from Mynydd y Gaer


Amateur historians and authors, Alan Wilson and Baram Blackett have promoted their hypothesis that Arthur could be identified with a King Athrwys of Gwent (specifically Athrwys ap Meurig, known from a number of early medieval genealogies). In fact, the identification was first proposed in the eighteenth century, but no matter. The name Athrwys is no more identical with Arthur than than name Artognou on the Tintagel inscription. However, Wilson and Blackett went further, in that they identified the church of St Peter-super-Montem at Mynydd y Gaer, near Bridgend (Glamorgan, Wales), as the burial place of King Arthur. Having announced their hypothesis, they became concerned about the possibility of looting at the site, which they bought. At an unknown date between 1983 and 1990, they claim to have unearthed the tombstone of Arthur, which reads REX ARTORIVS FILI MAVRICIVS. A subsequent excavation in the church, carried out in 1990 under the supervision of Eric Talbot, at that time with the University of Glasgow, revealed earlier structures and a silver cross, again with an inscription (PRO ANIMA ARTORIVS), was found.

Arthur for the soul?

Arthur for the soul? Illiterate Latin is hardly evidence!

If genuine, these inscriptions would be good evidence for the existence of someone whose name could be written as Artorius in Latin in early medieval Wales. However, neither inscription has been submitted for analysis by acknowledged experts in the field and even to an outsider, they appear curious. For a start, both are ungrammatical, if they are meant to mean what they are claimed to mean. Wilson and Blackett translate the stone as ‘Arthur son of Mauricius’, which to them confirmes the identity of Athrwys ap Meurig with the “King Arthur’ of the inscription. The problem is that the inscription should actually be translated ‘King Arthur Mauricius, of the son’. And while they would like to see the cross inscription as being ‘For the soul of Arthur’, it is actually ‘Arthur for the soul’ (which is probably not as effective as chicken soup). In other words, these inscriptions are, at best, crude forgeries by someone with a very poor knowledge of Latin and certainly poorer than we would expect in early medieval Wales.

Camden’s illustration of a lead cross allegedly found at Glastonbury in 1191

Camden’s illustration of a lead cross allegedly found at Glastonbury in 1191; but there is no evidence that Camden ever saw it!


By far the most famous inscription mentioning ‘King’ Arthur is that said to have been found on a lead cross by the monks of Glastonbury who excavated the purported site of his grave there around 1191. Descriptions of the discovery were written by Ralph of Coggeshall in 1221, Giraldus Cambrensis in 1193 and Adam of Domerham (himself a monk at the abbey) in the 1290s. All give slightly different accounts of the discovery of the body, but it was alleged to have lain in an ancient coffin, hollowed from an oak trunk. They also differ in the wording of the inscription said to have been on a lead cross found above the coffin. Ralph gives it as Hic iacet inclitus rex Arturus in insula Avallonia sepultus (‘here lies the famous King Arthur, buried in the Isle of Avalon’); Giraldus adds the phrase cum Wenneveria uxore sua secunda (’with his second wife Guenevere’) at the end.

In the sixth edition of William Camden’s Britannia, published by Richard Gough in 1607, a drawing of the cross appeared for the first time. It is by no means certain that Camden saw the cross, but Leslie Alcock used the shape of the letters in the drawing to suggest a tenth- or eleventh-century date for it. He was subsequently (and, in my opinion, rightly) criticised for his lack of scepticism regarding the alleged cross, last known to have been owned by William Hughes, a chancellor of Wells cathedral, in the early eighteenth century. However, in the Enfield Advertiser of 17 December 1981, a Derek Mahoney claimed to have rediscovered it in the bed of a lake at Forty Hall near Maidens Brook, Enfield (UK), when it was being drained for dredging, but apart from showing the object to an inexperienced assistant at the British Museum (whom he refused permission for photography), no-one else has ever seen it. In the following year, Enfield Borough Council prosectued him for theft as the cross was allegedly found on its property. Mahoney served a year of a two-year sentence in prison for contempt of court.

It is now thought that Mahoney’s cross was a forgery. He had worked as a lead pattern maker and knew that Richard Gough, who prepared Camden’s Britannia for a new edition in 1607, had lived close to the site of the supposed discovery. Members of the Enfield Archaeological Society, which had overseen the dredging operation on the lake, had seen nothing like the cross removed from the sediments and at least some of the members believed that Mahoney was seeking publicity. Mahoney became ill after his release from prison and was found hanged at home in 1989: a verdict of suicide was recorded. There was no trace of the cross among his possessions and it seems that he had disposed of it in order to avoid its exposure as a fraud.


If there are no inscriptions that can convincingly be shown to mention Arthur, what are the chances of locating Camelot? Although it is not named before the romance of Lancelot, le chevalier de la charette, written by Chrétien de Troyes between 1177 and 1181, this has not stopped enthusiasts for seeking its location. After all, someone as powerful as Arthur must have had a court (or, if we want to think of him as a Late Roman emperor, a bureaucracy) that was housed somewhere. The serious historian John Morris even proposed identifying Camelot with Colchester, the Romano-British Camulodunum on the grounds that the two names contain (almost) the same consonants in the same sequence… So is there really a site that can lay claim to be Camelot?

South Cadbury hillfort

South Cadbury hillfort: as likely to be Camelot as is Milton Keynes

By South Cadbury is that Camelot…

In 1541, the Tudor traveller John Leland (1506-1552) wrote that “At South Cadbyri standeth Camallate, sumtyme a famouse toun or castelle. The people can tell nothing thar but that they have hard say that Arture much resortid to Camalat”. The first archaeologist to attempt an excavation on the hill was Harold St George Gray who found Late Iron Age structures and material near the entrance in 1913. However, it was Courtney Ralegh Radford’s identification of Merovingian glass among finds brought up by ploughing in the 1950s that reawakened interest in matters Arthurian.

After an abortive proposal for excavation in 1959, a Camelot Research Committee was formed in 1965, which entrusted the excavation to Leslie Alcock (1925-2006), then of the University of Wales, Cardiff. It inevitably became known as the Quest for Camelot. Attracting financing from the British Academy, the BBC, the Society of Antiquaries and various other sources, the project began with trial trenching in 1966. Excavation then continued until 1970, with spectacular results spanning much of later prehistory and, excitingly, the early medieval period, ending with the establishment of a short-lived mint around 1010 under Æthelræd II (‘The Unready’, c 968-1016, King of England 978-1013 and 1014-1016).

It was the early medieval occupation of the late fifth century that attracted the most interest. While Leslie Alcock was always rather circumspect about the Arthurian associations of the site, he nevertheless encouraged the popular press to report it as the base of an ‘Arthur-like’ figure who, naturally, was transformed in the popular consciousness to an Arthur of history. Indeed, Alcock’s 1971 overview of the period was called Arthur’s Britain, an acknowledgement that he believed that someone did the things credited to Arthur around the right time, so we might as well call the period ‘Arthurian’. Geoffrey Ashe has gone further, arguing that South Cadbury was nothing less than Arthur’s headquarters. And we may as well call that Camelot…

Unfortunately, while Alcock’s book was not well received by the archaeological community, it made a larger impact on a more general readership. It became popularly believed that Alcock had found evidence for the existence of Arthur, both in terms of what he had excavated at South Cadbury and, more generally, in the pattern of Anglo-Saxon settlement and its apparent slowing down c 500. However, while Alcock’s excavations had uncovered a new class of late fifth-century site – the heavily fortified British hall – it was soon found that it was not unique. If South Cadbury was the base of an ‘Arthur-like’ figure, there were many more of them. If they were all ‘Arthur-like’, which one was the home of the ‘real’ Arthur? Do any of them actually provide evidence for a ‘real’ Arthur? The answer has to be no…

The baths at Wroxeter

The baths at Wroxeter: the Milton Keynes of sixth-century Britain?


One of the great excavations of the twentieth century was that carried out by Phil Barker (1920-2001) at Wroxeter (Shropshire, England) from 1966 to 1990. In a masterpiece of slow, painstaking excavation and recording, he surprised the archaeological community with his demonstration that there was a long sub-Roman occupation of the former Roman city of Viroconium Cornoviorum. Not a place traditionally associated with Arthur, amateur historians Graham Phillips and Martin Keatman built up a case for identifying it with Camelot in King Arthur – the true story (1992). They did so by first equating Arthur with the historical Owain Ddantwyn, an ancestor of the kings of Powys who would have lived around 500. They then argue that Wroxeter was the only place in late fifth-century Powys that could have served as a royal capital. And… well, that’s about it, in a nutshell.

Like all attempts to locate Camelot, it suffers from the fact that there is no evidence that such a place ever existed outside Chrétien de Troyes’s imagination. It starts with an assumption that there was a Camelot to be found and that there was an Arthur to hold court there, then goes out to find the evidence. Without the later stories of ‘King’ Arthur, there is nothing in the archaeology of these places that would lead us to postulate the existence of such a character. We bring our later preconceptions to bear on the interpretation of the data, which is definitely Bad Archaeology.