King Arthur and the volcanic winter of 536

There is now scientific evidence that the earliest references to Arthur are reliable

Volcanic eruption from 2021 (Creative Commons)

Since the 1970s, many historians have come to the conclusion that King Arthur never existed. Some have even written books where they set out the grounds for their scepticism. Recently discovered (or, rather, rediscovered) evidence, however, shows that this scepticism is entirely misplaced.

King Arthur was hailed as the epitome of a good king in the Middle Ages, but historians have never been sure whether he was a real person or not. He certainly had a lot of things written about him by medieval authors that now look to be mythical. No historians think that he really did have a round table to seat his knights at or that he lives on in a magical cave somewhere waiting to return to save Britain at the time of its darkest hour.

But Arthur is mentioned in two entries in the Welsh annals. Arthur appears in an entry for the year 516 where he is recorded leading the British at the Battle of Mount Badon and in another for 537 that records his death at a place called Camlann. These entries seem to be clear evidence that Arthur was a real person and that we can know when he lived and died. He must have been born in the late fifth century and died in 537.

The entries in the Welsh annals are often treated with suspicion, however. Can we trust them? Is there evidence that they are genuine historical records? The manuscript they are found in preserves entries as late as the tenth century. The annals are written in Latin, but the names they record often feature ninth and tenth-century Welsh spellings. Couldn’t the entries that mention Arthur just have been added to the annals at a date after the sixth century? Couldn’t they just be made up?

Medieval scribes are known to have updated older spellings to more recent ones as they made copies of manuscripts, so the ninth and tenth-century spellings aren’t a good guide as to how old the original entries are. None of the manuscripts of medieval works like annals that have survived are originals – they all reflect copies of copies made before the invention of printing. And a set of annals that ends in the tenth century would be expected to have begun much earlier. Most sets of annals last for several centuries and even ninth and tenth-century ones are usually pretty reliable.

So are the Arthurian entries reliable? Historians accept that the seventh-century entries in the Welsh annals are reliable. The entries from the later sixth century look to record actual events too. Yet the two that mention Arthur are treated sceptically.

Given that the reliability of the seventh-century entries is not disputed, the scepticism regarding the Arthurian entries seems a bit odd. But Arthur is such an important figure that the two entries that mention him have often been scrutinised by sceptics. The ones that follow them generally aren’t, however. The Arthurian entries seem so important, it has often been forgotten how reasonable all the other sixth-century entries preserved in the Welsh annals look to be.

So how reliable are the other early entries in the Welsh annals? Ten years after the death of Arthur, the annals record the passing of a figure who was clearly historical. In 547, the Welsh annals record that Mailcun the king of Gwynedd died in a plague or “great death“ (the Latin description is mortalitas magna). No one doubts that Mailcun lived in the sixth century, but is there evidence that the entry recording his death is genuine?

Historians know quite a bit about Mailcun. The British monk Gildas was a contemporary and evidently didn’t like him very much. Gildas accuses him of having usurped his uncle and having had his nephew put to death so that Maglocun could marry his widow. Gildas’s spelling Maglocun is different from Mailcun, but Gildas’s g would have been pronounced as a y. Maylocun is fairly obviously the same person as Mailcun.

German victims of bubonic plague (Landesamt für Denkmalpflege, Baden-Württemberg).

It’s also quite obvious that Mailcun could have died in a plague in 547. In 541, plague broke out in the Eastern Mediterranean. One of the first places the plague hit was Egypt, then Constantinople. The Emperor Justinian reacted by changing employment law so that workers couldn’t use the scarcity of labour caused by the deaths of so many people to argue that their wages should rise. Ancient writers explain in grisly detail how the pandemic spread. Masters and servants were both mowed down – Justinian is one of the few people known to have contracted the plague and to have survived.

In recent years, archaeologists have surveyed remains from sixth-century graves and determined that the Plague of Justinian was caused by the bacterium Yersinia pestis. It was a form of bubonic plague – the same as the Black Death of the later Middle Ages.

The earliest sign of bubonic plague in Britain is known from the remains of a person whose death radiocarbon analysis has dated to 541. The plague arrived Spain in 543 and it ravaged France and Germany from c. 543-547. The first mention of the plague in an Irish annal dates to 545 where it is called blefed, an Irish term which seems to mean “yellow disease”. The pandemic seems to have struck Britain in two waves, an earlier one that spread along sea lanes, the later one coming overland via France.

But the record of the death of Mailcun has been questioned by historians who seem to think that the entry in the Welsh annals must necessarily be unreliable. Do we really know that Mailcun died in a plague? Where is the supporting evidence?

You would like to think that the fact that the 540s was the time of the Plague of Justinian would be enough. But not if you are a sceptical historian. Just because the 547 entry is ridiculously plausible, doesn’t mean it is genuine.

Then we have the record of Arthur’s death in 537. The entry from 537 records that the year Arthur died was another time of death (Latin mortalitas) in Britain and Ireland. But no plague is known from 537, so the record must be unreliable – or that is what Professor Stuart Piggott CBE, FBA, FSA, FRSE, FSA Scot confidently proclaimed in 1950. All those honorific letters meant that he must have been right.

But Piggott had jumped the gun. The record of 537 doesn’t mention a plague, just that it was a time of death. The cause of the death has now been rediscovered too.

In the 1980s, scientists began to study the influence of past volcanic activity on climate. One year that the climate seemed to be particularly affected by volcanic emissions is 536. Tree rings were studied to see if there was a correlation – and there is. The yearly growth of trees can be analysed by measuring the ‘rings’ of growth you can see in the interiors of their trunks. Very little growth can be detected in tree-rings for 536.

Greek and Roman sources concur. The summer of 536 was strange. The rays of the sun that year were as feeble as moonlight. The abnormally cold weather played havoc with crops and led to the outbreak of famine. Ice-core samples confirm that there were increased levels of volcanic dust in the atmosphere that year. The volcanic winter of 536 had been rediscovered.

Winter landscape with the ruins of Eldena Abbey (Caspar David Friedrich)

The source of the dust veil has been more difficult to identify, with volcanoes in the Americas and Iceland variously suggested as the likely culprits. But the year 536 is now known to have been one of the worst-ever years for European agriculture. A recent study of Norwegian settlement patterns indicates that many villages were abandoned in the year that the summer never came. Scandinavian researchers now link the volcanic winter with the tradition of the Fimbulvetr, the three winters on end that preceded the end of the world in Old Norse myth.

It is also extremely likely that famine broke out in 537 in Britain. Irish annals similarly record a “failure of bread” in 536. The Welsh annals were obviously correct to record that 537 was a year of famine.

If you are a wilful sceptic of course, you will try to find reasons why the entries that mention Arthur are still unreliable. Even after scientists discover evidence that the Welsh annals record real sixth-century events. Perhaps the famine was recorded accurately in the annals, but someone has added in mentions of Arthur at a later date. How sneaky is that? But how likely would that be?

The volcanic winter is new evidence, evidence that shows that the annal for 537 is reliable. But if you’ve made up your mind about something, why would the discovery of new evidence matter? There is still room for you to wriggle about in after all.

But is there a good reason to? If all the other sixth-century entries in the Welsh annals are reliable, why would it be only the ones that mention Arthur that are suspicious – and only in bits? Why not just concede that the scientific evidence should be taken at face value – that it shows that the entries that mention Arthur are reliable? As the economist John Maynard Keynes said “When my information changes, I change my mind. What do you do?”