King Arthur: The Making of the Legend

KING ARTHUR

Nicholas J. Higham is professor emeritus at the University of Manchester. He is the author of many distinguished works including Ecgfrith: King of the Northumbrians, High-King of Britain and The Anglo-Saxon World.

Further praise for King Arthur:

‘A well-produced book by a serious scholar.’

David Miles, Minerva

‘Higham’s readable yet scholarly intervention is to be warmly welcomed . . . He roams widely, touching upon everything from ancient Greek astronomy to modern Caucasian mythography.’

Levi Roach, Literary Review

‘The historical Arthur of the scholars and novelists of the last century is in fact the reflection of a particular style of nationalist storytelling; and the dismantling of this dramatic but misleading model by historians such as Higham enables us to see more clearly the racial and cultural fluidity of Britain in the century and a half after the end of direct Roman rule.’

Rowan Williams, New Statesman

‘A very intelligent book which presents the facts and invites you to draw your own conclusions about this legendary British monarch. If King Arthur didn’t exist, he should have done, and Nicholas Higham’s book shows us why. A superb read: scholarly yet accessible. Highly recommended.’

Francis Pryor, author of Britain B.C.

‘An outstanding and deeply informed overview of the various King Arthurs in history. Accessible and well-written, it is also a significant contribution to the debate around the historical origins of Arthur.’

Anne Lawrence-Mathers, author of The True History of Merlin the Magician

‘It is undoubtedly the case that Professor Higham is extremely well-versed in his subject and that he has a wide-ranging and in-depth familiarity with the literature relevant to his project.’

Mark Jones, Albion

Copyright © 2018 Nicholas J. Higham

First published in paperback in 2021

All rights reserved. This book may not be reproduced in whole or in part, in any form (beyond that copying permitted by Sections 107 and 108 of the U.S. Copyright Law and except by reviewers for the public press) without written permission from the publishers.

For information about this and other Yale University Press publications, please contact:

U.S. Office: [email protected]     yalebooks.com

Europe Office: [email protected]     yalebooks.co.uk

Set in Minion Pro by IDSUK (DataConnection) Ltd

Printed in Great Britain by Clays Ltd, Elcograf S.p.A

Library of Congress Control Number: 2018953721

ISBN 978-0-300-21092-7 (hbk)

ISBN 978-0-300-25498-3 (pbk)

A catalogue record for this book is available from the British Library.

10 9 8 7 6 5 4 3 2 1

For Cheryl

CONTENTS

List of Plates and Maps

Acknowledgements

Introduction: Arthur, History and the Storytellers

PART I  THE ‘FOREIGN’ ARTHURS

1 Lucius Artorius Castus: A Dalmatian King Arthur?

2 The ‘Sarmatian Connection’

3 King Arthur and the Narts

4 King Arthur and the Greeks

PART II  THE ‘BRITISH’ ARTHUR

5 A Dark Age King Arthur

6 Arthur and the Historia Brittonum

7 A British Arthur: Starting the Tradition

8 ‘Fire’, ‘Smoke’ and ‘Highland Mist’

Appendix I The Artorius Inscriptions

Appendix II Arthur’s Battles as Described in the Historia Brittonum and Annales Cambriae

Appendix III The Mirabilia: A List of Wonders and Miracles That Became Attached to the Historia Brittonum

Abbreviations

Endnotes

Bibliography

Index

PLATES AND MAPS

Plates

  1 A wine label from Podstrana, Croatia, featuring the sword in the stone. © The author.

  2 St Martin’s Church, Podstrana. © The author.

  3 The larger ‘Artorius’ inscription, St Martin’s Church, Podstrana. © Drazen Marsic.

  4 One of three pieces of the smaller ‘Artorius’ inscription, Archaeological Museum of Split. © Nenad Cambj.

  5 Drawing of a scene on Trajan’s Column depicting Sarmatian warriors in flight from Roman cavalry. © Maggie Kneen.

  6 ‘Sarmatian’ beads from Chesters Roman fort on Hadrian’s Wall. Courtesy of the Trustees of the Clayton Collection and English Heritage.

  7 A funerary relief carving of a probable Sarmatian cavalryman found at Chester. © West Cheshire Museums.

  8 A lead ‘tamga’ from Ribchester. © Ribchester Roman Museum.

  9 The ‘Apollo’ monument from Ribchester. © Ribchester Roman Museum.

10 Sarmatian burial mounds at Vaskút, Hungary. © Visy Zsolt.

11 Mid-fifth-century Sarmatian pottery from excavations of the brick factory site at Tiszaföldvár, Hungary. © Andrea Vaday.

12 Sarmatian beads of the mid-fifth century from excavations of the brick factory site at Tiszaföldvár, Hungary. © Andrea Vaday.

13 Fourth- and fifth-century levels under excavation at Ribchester in 2016. © Duncan Sayer.

14 A stone footprint at the Dál Riatan stronghold of Dunadd, Argyll, Scotland. © The author.

15 The late medieval crozier of the archbishop of Siena. © Luca Larpi.

16 A mountain village in the Caucasus, illustrated in Douglas Freshfield, Travels in the Central Caucasus and Bashan (London, 1869).

17 The Mote of Mark, Dumfries and Galloway, Scotland. © The author.

18 The Battersea Cauldron, c. 800–700 BCE. © The Trustees of the British Museum.

19 The Antiochos dedication, Chester, England. © West Cheshire Museums.

20 The ‘astronomer’ mosaic, Brading Villa, Isle of Wight, England. © Oglander Roman Trust.

21 Part of the archivolt of the Porta della Pescheria in Modena Cathedral, Italy. © Elizabeth Lapina.

22 The Winchester Round Table, Winchester Castle, England. © John Critchley/Historic England Archive.

23 South Cadbury hillfort, Somerset, England. © Historic England Archive.

24 The farm of Camlan Uchaf in the small township of Camlan, Wales. © The author.

25 Pen-y-Fan viewed from Cribyn, Brecon Beacons, Wales. © Dave Dunford.

26 Arthur’s Stone, a Bronze Age burial site at Llanrhidian Lower, Gower Peninsula, Wales. © The author.

27 ‘King Arthur’s Hall’ near St Breward, Cornwall, England. © The author.

28 Pendragon Castle, Mallerstang, near Kirkby Stephen, Cumbria. © The author.

29 The Maen Huail, Ruthin, Wales. © The author.

30 An explanatory plaque, Maen Huail, Ruthin, Wales. © The author.

31 Illuminated medallion 49 from the fifteenth-century Histoire Universelle. © The University of Manchester.

32 Tintagel viewed from across the causeway with the haven in the foreground. © The author.

Maps

1 The Roman Empire in the second half of the second century CE.

2 The ‘Sarmatian connection’.

3 Roman Britain.

4 Ancient Greece.

5 Britain after Rome.

6 Britain and Ireland.

7 Localising Arthur.

ACKNOWLEDGEMENTS

This book would have been impossible to write without the prodigious amount of scholarly work undertaken by a whole legion of researchers, stretching back to the nineteenth century. It is particularly appropriate in a work dealing with King Arthur to acknowledge the sense in which I have been carried to this point on the shoulders of so many giants of scholarship working in a variety of interrelated disciplines. Arthurian studies has been an extraordinarily prolific area of endeavour over the last century and more, resulting in an exceptional quantity of published work.

Specifically, Martin Ryan very kindly read and commented on most of the chapters presented here in an early draft; as ever, I am extremely grateful to him. My thanks too go to Lindsay Allason-Jones, Frances McIntosh and Paul Bidwell for advice regarding ‘Sarmatian’ finds along Hadrian’s Wall, to Peter Schrijver on the origin of the name Arthur, to Paul Fouracre on all matters Frankish and to Charles Insley for various fruitful discussions regarding the Irish Sea region in the early Middle Ages. Patrick Tosteven very kindly provided photographs of material in Ribchester Roman Museum (Plates 8 and 9), Paul Holder and David Langslow advised on Roman-period inscriptions and Andrew Breeze provided me with invaluable assistance regarding Celtic place-names. Ross Trench-Jellicoe set me straight on the carved stones at Maughold (Isle of Man), Rod Thomson on William of Malmesbury and Fiona Edmonds on Lancashire place-names. John Colorusso very kindly provided me with a draft of his work on the Ossetian Nart sagas prior to publication and corresponded thereon. Anamarija Kurilić gave extraordinarily generously of her time and expertise in discussing the L. Artorius Castus inscriptions, in particular, and their Roman Dalmatian backdrop more generally. Julie Lawton very kindly aided me with academic French, translating passages where precise meanings were critical. Henry Edmunds very helpfully advised me regarding recent genetic research into the Cleveland horse breed, Barry Hobson kindly advised me regarding the vicus at Slack Roman fort, and Duncan Sayer both conducted me around his excavation of Roman Ribchester in 2016 and very kindly provided Plate 13. Donna Sherman gave invaluable help and advice regarding the University of Manchester’s map collection and I am indebted to the University Library more generally for sourcing a wide variety of monographs and other materials for my use. Bill Hanson was instrumental in obtaining Plates 10, 11 and 12, for which my thanks go to Visy Zsolt (Plate 10) and Andrea Vaday (11 and 12). Nenad Cambj very kindly provided me with Plate 4 and Drazen Marsic with Plate 3. I am grateful to John Hodgson for his help with the University of Manchester manuscripts kept at the John Rylands Library, Deansgate, and Nigel Wilkins for his help in accessing photographs from the collection maintained by Historic England. Maggie Kneen drew Plate 5 with her customary skill on the basis of a photographic record made by Roger B. Ulrich. My thanks go to Luca Larpi for all our discussions of Arthur over the years, his advice regarding the sword-in-the-stone story and provision of Plate 15. Many of my students over the last few decades contributed to my thinking about Arthur by asking questions to which I felt bound to seek answers; to all go my thanks. I am grateful also to Heather McCallum, Marika Lysandrou, Rachael Lonsdale and Samantha Cross at Yale University Press for commissioning this book and for all their assistance along the way, to Jacob Blandy for his skills as copy editor and to Martin Brown for his excellent maps.

Notwithstanding the number and scale of these many debts, all errors are my responsibility alone.

It is to my wife, Cheryl, though, that this book is dedicated, for she has borne the brunt of my preoccupation with Arthur over the decades and has found herself walking with me up hill and down dale to view sites that so often present as just odd piles of stones. For her good humour, patience and encouragement along the way my grateful thanks.

1  A wine label from Podstrana, Croatia, featuring the sword in the stone. The stone depicted is the longer of the ‘Artorius’ inscriptions (see Plate 3).

2  St Martin’s Church, Podstrana. One of the replica inscribed slabs is visible in the cemetery wall immediately to the left of the church.

3  The larger ‘Artorius’ inscription, restored and displayed inside St Martin’s Church, Podstrana.

4  One of three pieces of the smaller ‘Artorius’ inscription that was recently re-identified in the Archaeological Museum of Split.

5  A depiction of Sarmatian warriors in flight from Roman cavalry on Trajan’s Column. This scene provides a near-contemporary vision of Sarmatian equipment, but may not have been particularly accurate in its execution.

6  ‘Sarmatian’ beads from Chesters Roman fort on Hadrian’s Wall: a very rare example of material that has been interpreted as being deposited by Sarmatian warriors serving in Britain.

7  A funerary relief carving of a probable Sarmatian cavalryman found at Chester, perhaps a draconarius serving with the legion stationed there.

8  A lead ‘tamga’ from Ribchester, which has so far been the sole artefact of probable Sarmatian manufacture that has been found at this, the only cavalry fort in Roman Britain known to have had a Sarmatian garrison.

9  The ‘Apollo’ monument from Ribchester: the inscription refers to ‘Gordian’s own unit of Sarmatian cavalry’.

10  Three Sarmatian burial mounds at Vaskút, Hungary, reflecting the colonisation of the Carpathian Basin by these Black Sea nomads during the early Principate. No such monuments have ever been found in Britain.

11  Mid-fifth-century Sarmatian pottery from excavations of the brick factory site at Tiszaföldvár, in the Tisza valley, Hungary.

12  Sarmatian beads of the mid-fifth century from excavations of the brick factory site at Tiszaföldvár.

13  Excavation of the fourth- and fifth-century levels at Ribchester in 2016. Both wall-lines and industrial workings are now being uncovered, comparable to many other Roman fort sites in northern Britain.

14  A stone footprint at the Dál Riatan stronghold of Dunadd (Argyll), which may have been used in ceremonies of royal accession by Scottish kings.

15  A sword-in-the-stone motif features on the late medieval crozier of the archbishop of Siena. This derived from the cult of St Galgano, which developed in the late twelfth century at nearby Montesiepi.

16  A mountain village in the Caucasus, with the stone houses and towers characteristic of the area, illustrated in Douglas Freshfield’s Travels in the Central Caucasus and Bashan.

17  The Mote of Mark, a Celtic, Dark Age defensive site near Dalbeattie, Dumfries and Galloway, Scotland. The name derives from the high medieval French motte (‘castle-mound’) combined with King Mark of far-off Dumnonia, as popularised in literature of the thirteenth century.

18  The Battersea Cauldron, a high-status bronze vessel from c. 800–700 BCE found in the River Thames. Such finds illustrate the antiquity of this kind of manufacturing in Atlantic Europe.

19  The Antiochos dedication is a rare instance of Greek lettering on stone in Roman Britain. It is probably significant that this has strong medical associations, as this is a line of work in which knowledge of Greek will have been particularly valued.

20  The ‘astronomer’ mosaic at Brading Villa is thus far the sole discovery of a mosaic image of an astrologer found in Roman Britain.

21  Part of the archivolt of the Porta della Pescheria in Modena Cathedral showing King Arthur (Artus de Bretania) with his supporters attempting the rescue of Winlogee (Guinevere) from a castle defended by Burmaltus.

22  The Winchester Round Table, displayed on the wall of the great hall of Winchester Castle. The table was probably commissioned by Edward I in the thirteenth century (the timbers were likely felled 1250–80) and was refurbished and painted on the orders of Henry VIII early in the 1520s.

23  South Cadbury hillfort, a prominent candidate for the site of King Arthur’s Camelot at least since the sixteenth century. It was an Iron Age hillfort that was refortified in the later fifth century, when it was the residence of a powerful member of the elite.

24  The farm of Camlan Uchaf in the small township of Camlan. The road from Aberangell to Dinas Mawddwy runs across the foreground with the River Dovey behind; several passes led from here northwards and north-eastwards into Gwynedd and Powys.

25  Pen-y-Fan, in the Brecon Beacons, viewed from Cribyn. Pen-y-Fan is the highest of the Brecon Beacons and thought to be the Kaerarthur named by Gerald of Wales in the 1190s.

26  Arthur’s Stone, a Bronze Age burial site at Llanrhidian Lower on the Gower Peninsula. Folklore has it that this was a stone that lodged in Arthur’s shoe on his way to fight the battle of Camlann, which irritated him so much that he threw it 7 miles across the Loughor estuary.

27  ‘King Arthur’s Hall’, an enigmatic 47 × 20 metre rectangular setting of massive stones near St Breward, Cornwall, on the north-western edge of Bodmin Moor, which has been dated variously from the Neolithic to the medieval periods.

28  Pendragon Castle, Mallerstang, near Kirkby Stephen, Cumbria. Despite being named for Arthur’s legendary father, this is a Norman keep built by Ranulph de Meschines, earl of Chester, probably in the reign of William Rufus (1087–1100).

29  The Maen Huail, a badly degraded slab of limestone set against the half-timbered wall of the rebuilt Exmewe Hall, St Peter’s Square, Ruthin, on which Arthur is reputed to have had Huail beheaded.

30  The more recent of the Maen Huail explanatory plaques, put up by Cadw.

31  Arthur killing Mordred in illuminated medallion 49 from the fifteenth-century French manuscript Histoire Universelle. Note the plate armour, which is characteristic of this stage of the Hundred Years War.

32  Tintagel, the Dark Age settlement and later castle site viewed from across the causeway with the haven in the foreground. This has been reputed the site of Arthur’s conception since Geoffrey of Monmouth, but he was in fact recycling a far older folkloric story that had been recorded much earlier in Ireland and in the classical world.

INTRODUCTION

Arthur, History and the Storytellers

When good King Arthur reigned with Guinevere his Queen, there lived, near the Land’s End in Cornwall, a farmer who had only one son called Jack. Now Jack was brisk and ready; of such a lively wit that none nor nothing could worst him.

In those days, the Mount of St Michael in Cornwall was the fastness of a hugeous giant whose name was Cormoran.

— The opening lines of ‘Jack the Giant-Killer’¹

King Arthur today is a truth universally acknowledged, his name far better known than those of the vast majority of historical characters – a fact not lost on entrepreneurs of many different kinds, writers among them. The result has been a whole host of books and films, operas and musicals, plays, TV dramas and websites based upon his story. Arthur’s name adorns country inns and the walls of community schools, ² and promotes such heritage sites as Glastonbury Abbey (where his grave was ‘discovered’ in 1191) – and the live music venue, too, just up the road. He was little short of an obsession for Alfred, Lord Tennyson, poet laureate from 1850 to 1892, ³ and a keen interest of J.R.R. Tolkien. ⁴ Camelot, Arthur’s legendary capital, was commandeered by Jackie Kennedy in 1963 in the week following her husband’s assassination to serve as a metaphor for the White House during his presidency. Equally, for twenty-nine years it was the name of a children’s theme park in Lancashire (it closed in 2012); today it trademarks the UK’s National Lottery.

Nor is Arthur on his own in all of this; his close associates are almost as well known. The game machines used by that same National Lottery are called Arthur, Guinevere, Lancelot, Galahad and Merlin. Way back in 1879 the Ffestiniog Railway named its first steam locomotive Merddin Emrys after the Welsh version of the last of these, Rolls-Royce followed suit with its early aero engines and Merlin’s name has since been used for a long-running BBC drama, an international health charity, project management software, several computer games, and various retail outlets. Guinevere was popular with artists in the Victorian and Edwardian eras,⁵ and attracted film-makers in the twentieth century;⁶ today her name sells bedding and antiques. Lancelot and Galahad have attracted similar attention.⁷ Despite it being a fourteenth-century text in Middle English, far more editions of Sir Gawain and the Green Knight have been published than might have been expected were it not for its strong Arthurian connections. And the ageing Gawain is a central figure in Kazuo Ishiguro’s recent masterpiece, The Buried Giant, the last of Arthur’s knights left guarding the dragon whose breath (adapted for the purpose by Merlin, of course) causes the forgetfulness that ‘darkens’ the age.

Arthur’s prominence is nothing new; for a millennium and more his story has been read and heard, refashioned, reinvented and retold; the process is ongoing, with new versions coming out pretty much every year, mutating from one telling to the next. And Arthur’s story has long drawn in all sorts of unrelated tales – just look at the opening lines of ‘Jack the Giant-Killer’ quoted above,⁸ or the Arthurian setting given to another tale of the same general period, ‘Tom Thumb’.⁹ Neither has been traced back to the Middle Ages, but in each case the connection was made soon after by authors who adopted a style of introduction that recalls the earliest Arthurian romances of the twelfth century.¹⁰

How we view King Arthur has long reflected how we feel about the past, and particularly the deep past – the ancient and early medieval past. Arthur has been pressed into service time and again to support any number of causes. He has represented the Romans and the British (in opposition to the Saxons), the Welsh and the Cornish (in opposition to the English), the Bretons (in opposition to the French), the Campbells (in opposition to various other Scottish clans),¹¹ various claims to the whole island of Britain, and the English, and ultimately the British, in opposition to continental neighbours. He has personified courage and martial prowess, an ideal of Christian leadership, chivalry, the monarchical principle, aristocratic values, conduct befitting a Victorian gentleman, romanticism, the Gothic, medievalism, barbarism, and the ‘so-called Tradition’.¹² And Arthur and his associates have found a home too in the more mystical ‘Celtic’ and ‘new-age’ fringes of modern culture, in legends relating to the foundation of England and the authority of its early church (particularly at Glastonbury),¹³ and as part of wider folk–spiritual responses to the materialism of the present day.¹⁴

Arthur and his friends are everywhere. Their stories, their deeds and the complex of ideas which swirl around them underpin great swathes of our literature and film. The American novelist John Steinbeck was given Malory’s Le Morte d’Arthur as a nine-year-old and adopted it as his ‘magic book’, sparking a life-long love affair with storytelling in general and Arthurian literature in particular, which he sought to share by rewriting it for a twentieth-century audience.¹⁵ Since the 1950s Susan Cooper, Rosemary Sutcliff, Kevin Crossley-Holland and Michael Morpurgo have all repackaged Arthur for children; Mary Stewart, Bernard Cornwell, Count Nikolai Tolstoy and many, many more have done the same for adult readers.¹⁶ When the novelist Kate Atkinson has an ex-glamour-model mother tell her new nanny, ‘Well, Arthur’s called Arthur because his dad was into like Camelot and all that stuff’, she can be sure that we will get it.¹⁷ And when Tom Holt has the sword in the stone turn up in his hero’s bedroom with the words ‘Whoso draweth this sword from this stone shall be rightful king of all England’ engraved upon it, we can relate immediately and smile when he adds, in smaller letters ‘Please dispose of stone tidily after use’.¹⁸ We’ve seen the film,¹⁹ and many have read the book on which Disney based it.²⁰ We know this stuff.

Arthur in and out of history

King Arthur is, however, something of a nightmare from the historian’s perspective. Despite being one of the best-known figures from the Middle Ages – even from the past more generally – there is total disagreement as to when and where he belongs. Was there a ‘real’ Arthur behind the legend, as so many have argued? If so, can we establish what he did, and where and when he did it? Or should we consider him less substantive than that, a figure of folklore or mythology perhaps, a remnant from pagan times? Is Arthur just a figure of legend? Was he a fiction, even, right from the start? Or should we just sit on the fence, unable to make up our minds?

Portrayals of Arthur as a real historical personage stretch back in Wales to the ninth century. Orally transmitted stories passed from the Celtic world to France and northern Italy no later than the eleventh; in England William of Malmesbury knew tales of his deeds by the 1120s. At Oxford, in the 1130s, Geoffrey of Monmouth reframed early British history, with Arthur as his central figure. This ‘historical’ Arthur was translated into French by the Jerseyman cleric and poet Wace, and his story taken up thereafter both in histories and as a backdrop to romances written in the orbit of the first Angevin monarch, Henry II. Arthur’s story was not, though, sufficiently well grounded in near-contemporary evidence to satisfy the greater scrutiny that began to be brought to bear on the past following the Renaissance, so lost his place in history. Even so he remained hugely popular, featuring prominently in the Victorian period, in particular, in both literature and the visual arts. Many still thought of him as ‘real’ in some sense, or at least wanted to, despite most historians urging caution. Then, in the 1890s, scholarly reappraisal of the main ninth-century source for Arthur’s deeds encouraged scholars to think him historical once more. His stock as a figure of British history revived across the first half of the twentieth century, only to be challenged once again in the 1970s, both by historians and specialists in medieval Celtic literature.

Alongside this see-sawing debate regarding the reality of Arthur as a figure of British sub-Roman history, a few scholars have broken out of its confines to argue for very different Arthurs. The figures proposed vary in both space and time. One theory has him as a Roman army officer, associated perhaps with the Roman-period Sarmatians of the Eurasian Steppe, from whom several of his medieval stories derived; another identifies him in star-stories circulating in pre-classical Greece. Each of these lines of reasoning asserts a particular role in history for our hero, based on a unique body of source material. Arthur has yo-yoed, therefore, in and out of Britain as well as in and out of history, across the last century.

Despite the large number of books and articles dedicated to ‘finding’ King Arthur, the questions still remain, for there has been very little agreement. Who was he? Where did his story originate? Is there any reality lurking behind the story-book fantasies which eventually surrounded him? Should his story just be labelled legend? How should we relate Arthur to our understanding of the past?

Bookshops stock numerous volumes which claim to answer these questions, in any number of different ways. So why write another? The reason is this. Today, most specialists distance themselves from the whole issue of Arthur’s reality, citing insufficient evidence to be able to judge his place in history and declaring themselves agnostic on this matter. But their silence leaves the history-reading public with insufficient guidance to the competing claims and without the specialist knowledge to judge between them effectively, for these are highly complex issues. The purpose of this book is, therefore, to set out the main arguments which are on offer, test each one against the sources on which it relies, and determine which, if any, deserve support. This will include both a range of popular ‘British’ solutions and the major ‘foreign’ alternatives that have been proposed across the last century or so. Which (if any) of these Arthurs are sufficiently supported by the evidence to stand up to scrutiny? Which arguments work and which don’t? How should we resolve the historical debate surrounding King Arthur? Should we just throw our hands up in despair, admitting that we cannot know, or is there an answer to the age-old problem of the origins of the King Arthur story? That is what this book will do.

First let us just identify the main theories that we need to consider and outline how this book will be structured so as to engage with them.

The ‘Dalmatian’ Arthur and the Sarmatians

In 1890 it was pointed out that the Welsh name Arthur most probably derived from the uncommon Roman name Artorius. In 1925 the American linguist, Kemp Malone, directed attention to inscriptions discovered in Roman Dalmatia (modern-day Croatia) that commemorate the career of the soldier and administrator L. Artorius Castus. That this Artorius had served in Britain and had commanded ‘British’ troops encouraged Malone to argue that he was the legend’s starting point.²¹

This thesis was left rather hanging for the next half-century or so but interest was rekindled when a new generation of scholars began to think that the mounted knights of the central Middle Ages might owe something to the heavily armoured cavalry of the late Roman army.²² These had included men from outside the Empire, recruited inter alia from various Iranian-speaking peoples on the steppe. We know that Sarmatian cavalry were posted to Roman Britain, and both Alans and Sarmatians were settled in the west in the late Roman period.²³ Could similarities between stories in the Arthurian canon and tales told by classical authors about the Scythians, Sarmatians and Alans reveal debts owed by the Arthurian story cycle to tales told in Antiquity around the Black Sea?²⁴ If L. Artorius Castus could be linked to Sarmatians serving in the Roman army of Britain, then was it possible that the Arthurian tradition was begun by their stories becoming attached to his name?

Importantly, there is a modern witness to the tales told on the Eurasian Steppe in Antiquity. Work has long been under way to collect, translate and publish the folk literature of the Caucasus, where the Ossetians still speak a language descended from Old Iranian. This has raised the profile of a body of traditional stories known as the Nart sagas, which contain various narrative elements of great age. Across the twentieth century this corpus was scrutinised for possible connections with the Arthurian story cycle; parallels were identified and links suggested.

These lines of reasoning came together in the book From Scythia to Camelot by C. Scott Littleton and Linda Malcor, first published in 1994 (then in a revised edition in 2000). This argues for a link between L. Artorius Castus and Sarmatian soldiers serving in Roman Britain and highlights similarities between the Nart sagas and late medieval Arthurian literature. The resulting thesis proposes that many Arthurian legends originated on the Steppe and travelled westwards with Sarmatian soldiers who served under Artorius and then settled in Britain, to be followed later by further colonies established within the Empire. This explanation of the genesis of the Arthurian story cycle challenges the academic community to rethink the origins of King Arthur in dramatically different ways. While it has only attracted limited scholarly support, this theory has been promoted vigorously and gained a toe-hold in wider public perceptions of Arthur’s back-story; it therefore requires our attention.

The ‘Greek’ Arthur

Stimulated in part by this suggestion of an eastern origin for the Arthurian legend, further possibilities have been put forward by the Classicist Graham Anderson, who points to figures with ‘Arthur-like’ names in the myths and legends of pre-classical Greece. These centre largely on stories from ancient cosmology, connected particularly with the star Arcturus (Greek Arktouros) and the constellations Ursa Major and Ursa Minor (the ‘great’ and ‘little she-bear’). It is Anderson’s view that these offer a more convincing starting point for the medieval Arthur than either the ‘Dalmatian’ or ‘Dalmatian–Sarmatian’ theories.²⁵ Again, his case is reinforced by parallels drawn between stories that were circulating in classical Greece and Arthurian literature of the high Middle Ages, suggesting that the Round Table, for example, derived from a tale circulating in the pre-classical world.

This theory centres, therefore, on the proposition that Arthur originated in the literature of the Greek world and passed thence into Latin and Celtic.²⁶ And this ‘Greek’ Arthur was remembered in the central Middle Ages not just in Britain but on the Continent as well, hence the rapidity with which his fame spread across Europe and even beyond. Again, this approach draws in a substantial body of new evidence that we need to consider.

The ‘British’ Arthurs

But such ‘foreign’ solutions make up only a small minority of the reactions to Arthur’s lengthy exile from British history, from the seventeenth century until the nineteenth. Another was to keep him in Britain but consider him a figure of Celtic mythology. This theory enjoyed considerable support in the mid-nineteenth century,²⁷ then found a champion in the distinguished Celticist, Sir John Rhŷs, who argued that there were two originally separate bodies of evidence for Arthur – on the one hand, local place-names and legends, on the other the ‘histories’ of the central Middle Ages. These needed, in his view, separate explanations. There were two Arthurs, therefore, of whom the primary figure was mythological, to be sought in the pre-Christian Celtic world, the secondary a leader of the sub-Roman period.²⁸ Rhŷs’s ‘mythological’ Arthur attracted significant scholarly support between the wars,²⁹ and again more recently.³⁰

The mythological case rests on a body of Welsh folklore and literature written in the central Middle Ages, in which Arthur appears as a heroic figure leading a band of warriors engaged in warfare against a wide variety of opponents, both human and mythological. We find him, for example, raiding the ‘other’ world and bringing back magical items as booty, passing backwards and forwards between the realms of myth and man. Much of the evidence lies in Arthur’s association with local place-names, two examples of which are recorded in the list of ‘marvels’ (mirabilia) which became attached to an early ninth-century Latin work written in Wales, the Historia Brittonum (History of the Britons). If a ‘mythological’ Arthur was already the subject of storytelling in Wales when the Historia was written in 829–30, then might this be the ultimate source of the ‘historical’ figure who first appears therein? In that case Arthur’s emergence into history could be secondary to the figure of pagan mythology.

But the view that Arthur was more probably historical revived in Germany late in the nineteenth century, before being taken up by English-speaking academics keen to distance themselves from the strongly ‘Germanic’ emphasis of England’s early history following the First World War. One such ‘British’ Arthur was championed by R.G. Collingwood, the leading scholar of Roman Britain of the day. Ever more ‘positivist’ treatments were offered by archaeologists trained within this tradition, leading to widespread acceptance of Arthur as a Dark Age figure and the entire period designated the ‘Age of Arthur’. Arthur’s rehabilitation peaked around 1970 with the publication of such well-known works as Arthur’s Britain by Leslie Alcock and The Age of Arthur by John Morris. These provided the foundations for later writers wishing to pursue a ‘positivist’ approach and treat Arthur as a recoverable figure of history, producing any number of different theories regarding his origins. There is today a great mass of British candidates out there, all competing for our approval.

But the books written by both Alcock and Morris were heavily criticised, most particularly by scholars whose specialism was the Welsh texts of the central Middle Ages (on which all claims on Arthur’s historicity as a British figure must ultimately rest). Many academics (but not all) have since sought to cast out Arthur once more, consigning him to the historical touchline (if not the stands). Most are of the opinion that the evidence is just not good enough to say whether or not he existed, declaring themselves agnostic on this thorny issue. There are today four competing schools of thought within the ‘British’ camp, therefore – those supporting a mythological Arthur, advocates of an historical figure of one kind or another, those who believe that the best we can do is be agnostic on this whole matter, and those who see Arthur as legendary or fictional. We will need to assess each of these views as we grapple with our central question: when, where and why did Arthur’s story begin?

Approach and organisation

We need to take account, therefore, of many competing positions regarding Arthur – whether true or false, British or foreign – as we set about testing which, if any, are sufficiently robust to secure our vote. The focus will be less on arguments for and against a particular Arthur than assessment of the different bodies of evidence on which rival theories rest. We will begin by examining the evidence which underpins the various ‘foreign’ theories, starting with the ‘Artorius’ inscriptions from Roman Dalmatia (chapter 1), moving on to explore the potential for links between this Roman soldier and the Roman-period Sarmatians north of the Danube (chapter 2). We will then examine the Nart sagas of the Caucasus in search of material comparable to Arthur’s story (chapter 3), before circling back to engage with the Arthur-like names and the tales in which they occur from Ancient Greece (chapter 4). Part II will then review the (predominantly) literary evidence on which all the ‘British’ theories are based, arranging these texts as far as possible chronologically. We begin where all later commentators have begun, with an examination of potential sources from the period 400–750 that have been called upon by those responsible for one thesis or another (chapter 5), then assess Arthur’s role in the ninth-century Historia Brittonum (chapter 6) and finally turn to other works that refer to Arthur that were written later (as far as one can judge), in Latin or Old Welsh (chapter 7). At the book’s close (chapter 8) we will explore how and why Arthur’s stock rose, fell and rose again as a figure of British history, before finally weighing up the various different approaches to a British Arthur that are current today, offering pointers as to how best to manage the relationship between King Arthur and history in the present.

How, where and when did the legend of King Arthur begin? These are the main questions that we are seeking to address as we test all these different bodies of evidence to establish which can, and which cannot, provide us with a valid answer.

PART I

The ‘Foreign’ Arthurs

1

LUCIUS ARTORIUS CASTUS

A ‘Dalmatian’ King Arthur?

In general, the pseudo-historical Arthur (as distinguished from the Arthur of romance) corresponds with astonishing accuracy to the Artorius of our inscriptions.

— Kemp Malone, ‘Artorius’¹

The earliest suggestion that Arthur was ‘foreign’ to Britain in some meaningful sense dates back almost a century to Kemp Malone’s proposal that the Arthurian legend began not in the British Dark Ages but with the career of a Roman soldier named in two inscriptions found on the Adriatic coast. This L. Artorius Castus has several advantages as a candidate for the ‘original’ Arthur: he certainly existed and is on record as doing some of his soldiering in Britain; he had command of forces that were in some sense ‘British’ and he bore the Roman name from which the Welsh ‘Arthur’ is very likely to have originated. All told, this makes him a contender as the archetypal King Arthur.

The first step in assessing whether or not this individual could have been the ‘original’ King Arthur is to establish as well as we can what the inscriptions actually say – which is less simple and far less certain than one might imagine. The second is to revisit the similarities that have been suggested between this figure of the Roman world and King Arthur. So, how much can we know about L. Artorius Castus and how confident can we be of these ‘facts’? What did he achieve and when? What were the events in which he played a part? How telling are the comparisons that can be made between this historical figure and the Arthur that we come across in the Middle Ages? And how likely is it that his memory survived into post-Roman Britain, to re-emerge in the Arthurian legend?

Before looking in detail at the inscriptions, it will be helpful to establish when this theory first emerged, and then to track something of its later development. It was in 1873 that the longer and more informative of the two ‘Artorius’ inscriptions was published, bringing it to the attention of western scholars for the first time.² In 1890, quite independently, it was suggested on purely linguistic grounds that the Welsh name Arthur originated as the Roman name Artorius (see below, p. 30),³ a view that gained wide acceptance.⁴ When Malone developed his thesis, therefore, he was able to draw together the inscriptional evidence for L. Artorius Castus and a high probability that the name Arthur originated as Artorius. He concluded that Artorius and King Arthur were one and the same.⁵

This identification of King Arthur with a Roman soldier buried in Dalmatia found little support from British scholars between and in the immediate aftermath of the two world wars, when he was generally thought of as a ‘British’ figure fighting to defend his country – and civilisation more widely – against the Germans.⁶ It did, though, attract some interest in post-war Yugoslavia,⁷ where increasing tourism encouraged the Artorius–Arthur connection. The town of Podstrana – where the inscriptions were found – today has an Artorius Society,⁸ tourist shops sell ‘King Arthur’ place mats and plaster replicas of the longer inscription, and a local wine is marketed as ‘Artorius’, its label featuring a sword embedded in the larger of the two inscribed stones (Plate 1).

Malone’s thesis was eventually revived in the United States by C. Scott Littleton and his collaborators,⁹ most particularly Linda Malcor. In a widely read monograph, From Scythia to Camelot, they argue that key elements of Arthur’s story originated in the Roman period around the Black Sea and were carried from there to Britain by Sarmatians who were drafted into the Roman army in 175 CE.¹⁰ Given that L. Artorius Castus’s career included postings on the Danube frontier and in Britain, he was seen as a crucial link between Sarmatians beyond the Danube, the Roman army and stories circulating in Britain that became attached to his name.¹¹ It was Malcor who took responsibility for this strand of the argument, publishing two papers in 1999 alongside a revised edition of the book. She proposes that Castus immersed himself in Sarmatian culture while on the Danube frontier, that he commanded the Sarmatian warriors en route to Britain in 175 and that he then led them to multiple victories in the ‘Caledonian War’ fought on Roman Britain’s northern frontier early in the 180s.¹² In her second paper, Malcor then proposed that the warrior Arthur whose victories were celebrated in the ninth-century Historia Brittonum was a distant memory of this Castus, mistakenly reassigned to the sub-Roman-period war against the Saxons.¹³ Identification of the ‘original’ King Arthur as L. Artorius Castus is therefore an important element within the wider arguments developed by Littleton and Malcor, without which the theory that medieval stories centred on Arthur and his court originated beyond the Danube become problematic.¹⁴ A version of this was then popularised in the film King Arthur, released in 2004, on which Malcor was an adviser. Obviously, we must set aside this (and every other film too) in our discussion of L. Artorius Castus.

The inscriptions

Since he is not mentioned in any other source that has so far come to light, the two inscriptions found at Podstrana, south of Split on the coast of modern-day Croatia (Map 1), provide all the information currently available regarding L. Artorius Castus. Both are from a burial context and both offer some details of his career.

It would obviously be helpful to place this information within a tight chronological framework, to give an immediate context to Castus’s life. We should, though, flag up right from the start that what we do not know about him in this respect far outweighs what we do. In particular, these inscriptions provide no dates. That is not uncommon, for few burial-related inscriptions of the Roman period do, but it does introduce a degree of uncertainty that we can never entirely overcome. Stylistically the inscriptions belong to the late Principate, so the mid-second century through to the late third, with some preference for the first half of this period over the second.¹⁵ Attempts to narrow down the date range all rely on details of the posts listed. Attention has generally focused on his role as a dux (military commander), since this term was rarely used of non-senatorial officers in the second and early third centuries, but it is difficult to tie down the date on that basis with any precision. The alternative is to centre on his appointment as a provincial procurator, but even then we can only establish a likelihood of this occurring in a particular time-frame. There remains, therefore, much informed guesswork regarding the dates of these inscriptions and the career that they celebrate, and risks attach to every attempt to fit L. Artorius Castus to particular historical events.

1 The Roman Empire in the second half of the second century CE: a backdrop to the career of L. Artorius Castrus.

There are two inscriptions under discussion, both damaged. We will begin with the longer and better-known one, even though there are reasons to think that this may actually have been the later of the two to be cut.

The longer inscription

Around 1840, two large inscribed slabs were used in building the cemetery wall of the church of St Martin, which stands on the coast south-east of Podstrana (ancient Pituntini), with the sea lapping the graveyard. The inscriptions were visible on the road-side face of the wall, cut onto a dense limestone quarried locally and widely used for Roman-period inscriptions at and around Salona (modern-day Solin), the capital of the province of Dalmatia.¹⁶ Within a decade these two pieces had been recognised as parts of the same inscription.¹⁷ The two slabs have now been removed from the wall (where they have been replaced with replicas; Plate 2), cleaned, reunited and put on display inside the church (Plate 3).¹⁸ Bringing the two pieces together has provided an opportunity to better gauge the number of letters lost to the break,¹⁹ facilitating reading of the text.

The larger slab was initially thought to belong to the front of a sarcophagus,²⁰ but has now been reinterpreted convincingly as a plaque mounted on the wall of a funerary enclosure or mausoleum.²¹ It provided, therefore, an epitaph commissioned to give a fuller record than