he King Arthur of Tennyson’s Idylls of the King, written between 1856 and 1885, is the King Arthur of Malory and the French romance tradition before him: Lancelot and Guinevere, Mordred and Merlin, the Round Table and the Holy Grail. In Tennyson’s time the idea that there had once been a historical Arthur, his outline still dimly visible through all the accretions of romance, was effectively dead and had been for three hundred years. Milton thought about writing an Arthurian epic, but it would have been a fantasy, like The Faerie Queene; even Spenser knew that there was a real-world history of Britain, as opposed to Faerie, in which Arthur played no part.

Perhaps the greatest surprise in the fluctuating fortunes of Arthur is the way that he was reinserted into history during the 20th century. It was a development which owed almost everything to a conviction, especially among British writers, that they understood post-imperial history much better than any of their predecessors, and that Arthur made sense as post-imperial. The view of Arthur as a Romano-British war leader fighting a doomed resistance against the incoming Saxons has taken over popular literature, with distinguished examples coming from Rosemary Sutcliff, Henry Treece, C S Lewis and, more recently, Stephen Lawhead and Bernard Cornwell.

For a while he threatened to take over popular history and archaeology as well, but that movement was stopped dead in the late 1970s. Nowadays no serious scholar – Guy Halsall uses the phrase ‘no sane scholar’ – thinks there is anything solid that can be said about King Arthur at all. What energises his new book, Worlds of Arthur, is a feeling that academics ought nevertheless to explain their position, as a service to the ‘interested lay audience’ which pays their salaries. For, he concludes, ‘fact is, after all, stranger than fiction’.

His book falls into four parts. In ‘Old Worlds’ he presents the traditional narrative of post-Roman Arthurian Britain, discussing first the scanty documentary evidence and then the views of archaeologists such as Leslie Alcock, who identified South Cadbury in Somerset with Camelot. In ‘Present Worlds’ he shows just how unreliable the documentary sources are, noting that they have not only been picked over, but cherry-picked over. Read for themselves, rather than to bolster a thesis, they often look markedly different. Proceeding to archaeology, Halsall notes the clash between minimalist and maximalist schools of thought on the coming of the English. According to the silly extreme of the former, there was no Germanic immigration at all, only (as one archaeologist sardonically remarked) ‘an invasion of Germanic life-style advisors’. At the other extreme of the latter is the old fire-sword-and-ethnic-wipe-out thesis, which no one now wants to have anything to do with, since it is overtly racist and logistically implausible.

In ‘Mad Worlds’ Halsall looks mainly at the idea started off by R G Collingwood in 1936 that the Knights of the Round Table are a dim memory of late Roman heavy cavalry – probably Sarmatians once commanded by Lucius Artorius Castus, as in Antoine Fuqua’s ludicrous King Arthur movie of 2004. This ‘has no historical basis whatsoever’.

So, what’s new? And what ought we to think about the fifth and sixth centuries, in which England began to come into being? Where we’ve gone consistently wrong, Halsall argues, is in our insular ‘lack of awareness of the European mainland’. What happened in Britain, he argues, was not so very different from what was happening all over northwest Europe. Changes of burial customs, like the new habit of burying the dead with grave goods, often lavishly, need not mean a sudden takeover by pagan barbarians. It could be a response to the collapse of the villa economy, with a resultant need for lavish display by insecure local gentries. The ‘wave’ theory, according to which Germanic English speakers advanced inexorably from east to west and pushed the British into their present enclaves in Wales and Cornwall, does not fit the archaeology very well. As for the strange absence of proto-Welsh loan words in Old English – which we might have expected to be the result of contact between Germans and Britons – perhaps that could be explained if the fifth-century British spoke Vulgar Latin, like everyone else in the old Western Empire.

Halsall’s most challenging thesis is to suggest that the Vortigern of legend, the tyrant who brought Hengist and Horsa into Britain, could be identified with the fourth-century imperial contender Maximus. The result is that the whole story of the adventus Saxonum, and even Nennius’s famous list of Arthur’s battles, is a confused and anachronistic version of events on the Continent, culminating in the battle of the Frigidus River, referred to in contemporary sources as the battle of the Alps, in 394. (The Alps were known in Latin as vallum alpium, and Halsall suggests that, since vallum is gwawl in the British tongue, it could have been corrupted in Nennius’s Historia Brittonum as ‘Guoloppum’). Maybe ‘a bit contrived’, Halsall concedes, but strongly argued. And as ‘Macsen Wledig’, Maximus certainly left a deep imprint on later Welsh tradition. He was the man who took the troops away and never brought them back.

There is, however, a chink in Halsall’s scholarly armour, something one could well blame on the habit – which he himself notes – that academics have of communicating only to their own coteries. Clearly Anglo-Saxonists live in a different world even from early medieval historians. Halsall thinks Hengist is almost as implausible a figure as Arthur himself. But his name doesn’t mean ‘gelding’, as Halsall claims – that would be unlikely for a fifth-century war leader. It more often means ‘stallion’. And while Hengist is mentioned in one episode of Beowulf, the ‘Fight at Finnsburg’ fragment, which also refers to him, is very different. In view of the scanty remains of Old English heroic poetry, the double appearance of Hengist suggests that he is more than just an invention; and we have much better evidence for the existence of similar Beowulfian characters than we do for anyone connected with Arthur.

Halsall also dismisses the ‘apartheid’ thesis of division between German and English communities presented by some archaeologists to explain features of DNA evidence with the rather lofty remark that its proponents ‘do not seem to have thought this analogy through’. This is because the ‘process of intermarriage between British and English over two or three centuries [which the theory requires] would be explicitly discouraged in an “apartheid-like” social structure’. ‘Explicitly’, maybe! But modern experience shows that even in an openly racist society, such as pre-Civil War Alabama, there was plenty of racial mixing going on – though of course it wasn’t ‘intermarriage’ and it was almost all one way. Halsall also has nothing to say about early Anglo-Saxon laws, which (unlike later ones) make it very clear that although Wealas or the ‘Welsh’ had a fixed place in society, it was a second-class place. This fits neatly with Halsall’s point that while people may not be able to change their genes, they can change their social identity, most notably (as once more in the USA) by ensuring that their children speak the dominant language, not their parents’.

Worlds of Arthur is sensible, well informed and strongly argued, and tries to bridge the regrettable academic–popular divide. I am not sure, though, that Guy Halsall’s fact-based scenario really is, as he claims, ‘stranger than fiction’. For a gripping and by no means ill-informed picture of a collapsing empire, with Arthur on the edge of it, I recommend Alfred Duggan’s 1950s diptych The Little Emperors and Conscience of the King. Duggan knew a lot about collapsing imperial bureaucracies. His stepfather was Lord Curzon, last and in some views least of the Victorian viceroys of India.