Arthur and the Lost Kingdoms

by Alistair Moffat

  More traditions require to be relied on if the eighth of Arthur’s battles is to be correctly placed. There is precious little else to help with the phrase ‘in the stronghold of Guinnion’, except that Nennius adds that Arthur fought with ‘a likeness of Holy Mary Everlasting Virgin, on his shield and the heathens were turned in flight on that day, and there was a great slaughter upon them, because of the goodness of our Lord Jesus Christ and the goodness of the Holy Virgin Mary, his mother.’¹³⁵

  Two bits of clarity to sit alongside the mystery of the place-name. First a fulsome piece of early Christian propaganda advertising the decisive intervention of Christ and (twice noted) the Virgin Mary. Not surprising, Nennius was a monk and proselytizing part of his early purpose. But why wait for the eighth battle to insert the message? Why not top and tail the whole glorious campaign with God and his Mother? There must have been a particular reason.

  Fort or Castellum Guinnion implies a substantial fortification, perhaps a Roman camp or a hill fort and there are any number of candidates all over Britain, never mind the Borders. Guinnion sounds and is a P-Celtic or Old Welsh word. It means ‘white place’ or more anciently ‘holy place’. This last will clearly fit better with the sudden appearance of the Virgin Mary on Arthur’s shield. He was using Christian iconography to defend a holy place, probably associated closely with the twice-mentioned Virgin, and putting the pagans to flight.

  Yarrow, the Warrior’s Rest and Annan Street

  This narrows the field dramatically, particularly if the pagans in question are Angles or Saxons rather than Picts or Scots (who had in any case been exposed to some missionary activity by this stage). Even further, if Nennius has the sequence right and Fort Guinnion follows the seventh battle in the Forest of Celidon both in geography as well as time, that would place it in southern Scotland.

  Two more bits of toponymy will pin it on the map. The parish church in the village of Stow, which sits prettily on the Gala Water as it winds with the A7 towards Edinburgh, is called St Mary in Wedale. Buried in thick pinewoods a few hundred yards south is a rickle of stones identified on the Ordnance Survey as Our Lady’s Chapel and near it Our Lady’s Well.¹³⁶ In the Old Statistical Account of Scotland (a majestic work dealing with the fabric of the country in encyclopaedic detail), compiled at the end of the eighteenth century, the minister of Stow records ‘a very fine perennial spring, known by the name of the Lady’s Well, and a huge stone, recently removed in forming the new road, but now broken to pieces, used to be pointed out as impressed with the print of the Virgin Mary’s foot.’¹³⁷

  The cartularies of the medieval Border abbeys confirm the existence of a church of St Mary in Wedale before their documents began in the early twelfth century, and there is some evidence that there was a pilgrimage shrine to the Virgin as early as the ninth century. The site is clearly very old indeed but it cannot with certainty have been a going concern when Arthur lived. However, there is some dubious help available. In the late twelfth or early thirteenth century, a monk in the priory at Sawley in North Yorkshire added some glosses to Nennius’s text. This meant he clearly understood Old Welsh for he translated some of it into uncomfortable Latin. He has Arthur go on crusade (he was writing at the time of Richard I’s doomed Third Crusade) to Jerusalem where he makes and has consecrated a replica of the Blessed Cross. This was to ensure Arthur’s victory over the pagans. Then a last sentence which sounds like a repetition of an early tradition and which brings the little narrative back to earth with a local bump. ‘And he himself [Arthur] carried the likeness of Holy Mary, the pieces of which were thus saved in great veneration in Wedale.’ While that is much too specific a detail to have been invented, it proves nothing except the existence of an early tradition of Arthur fighting pagans near the village of Stow.

  More comfort can be had from place-names, if looked at with care, and sense of how time rubs them smooth. Wedale is ancient and it is related to Guinnion. Because Stow is an Anglian name which means ‘holy place’, in itself a memory of the Dark Ages shrine to the Virgin Mary, Wedale is also thought to have been conferred by the incomers. Some Arthurian historians even incorporate its supposed Anglian meaning of ‘dale of woe’ into arguments supporting the placing of the Battle of Fort Guinnion in the valley of the Gala Water.¹³⁸ The Angles called it Dale of Woe because they lost to Arthur in that place. Too neat, I fear, and the Angles were rarely anxious to record a defeat. Wedale is a Celtic name, and a more thorough parsing will show even stronger support for a battle in the valley.

  Around Stow there are a number of P-Celtic names which have been altered to fit into the mouths of the Q-Celtic speakers who settled by the Gala Water half a millennium after Arthur. Tref names, meaning ‘settlement’, became tor names, a Gaelic word for a hill or sometimes a rocky hill. Torsonce Hill is where the remains of the shrine to St Mary stands, hidden by trees, but in 500 it was a Tref name perhaps meaning the ‘settlement of a prince or chief’. Torquhan a mile or so north was also a Tref name, as were several others close by.¹³⁹

  Guinnion in P-Celtic came to be changed into Guidh-dail in Q-Celtic, which was in turn rendered as Gwedale, the valley of prayer. The latter element will bear a second look. Dail can also mean a field and anciently it could equally, just like the English word, stand for ‘battlefield’. Everyone knows that Flodden Field has nothing to do with barley or pasture. With the right name or word alongside, ‘field’ means ‘battlefield’. In the same way I think it likely that Gwedale, and its shortened Anglicized version of Wedale, means ‘the prayer battle’ or ‘the holy battle’ if the first element is pared back to the original P-Celtic gloss of ‘holy’. This would fit well with Nennius’s relative prolixity on the role of the Virgin and Jesus in gaining the victory over the heathens. And also add to the sense of mission that the armies of the Cymry unquestionably had. Not only was Arthur fighting to protect and secure his homelands, he also represented Christian enlightenment in its struggle against the forces of darkness and barbarity.

  Wedale was also the valley of access to the north and to Edinburgh and while it may have suited Arthur to put on the clothes of Christianity to fit the place, his greater strategic purpose was surely to shut down a line of advance for the Angles. In winter the A7 by Gala Water can be a kinder passage to Edinburgh than Dere Street and the A68 up over the windy wastes of Soutra Hill. Arthur fought the invaders on the flat ground of Wedale because he had to, not especially because it saved a holy site from pagan hands.

  There is an old hilltop fort to the north of Stow which Arthur could have used as his castellum as well as a Roman site slightly further away. And Nennius offers a clue to Arthur’s battle tactics in the phrase ‘pagani versi’, the pagans turned or were turned. Without bending the Latin too much to fit the hypothesis it sounds to me like the result of a successful cavalry charge.

  A keen sense of the Roman past stands behind Nennius’s next statement: ‘The ninth battle happened in the city of the Legion.’ Bald but clear. Identify the ‘Urbs Legionis’ and the location of the battle will be there. Several candidates are available: Chester, Caerleon and Colchester were all legionary headquarters.¹⁴⁰ But each had well-established names that appear repeatedly in the Roman record: Deva, Isca and Camelodunum respectively. All these are derived from local P-Celtic roots, and are different and particular. Looking again at Nennius’s Latin, he wrote ‘Urbs Legionis’ or ‘the city of the Legion’. Not ‘the Legions’ but ‘the Legion’. Deva, Isca and Camelodunum were headquarters to several named legions of the colonial garrison, but there was one place, which qualified as a city, that did become closely identified with a particular legion.

  From 122 until the end of the Roman occupation of Britannia, York was the Northern Army Command and the headquarters of the Legio Sextam Victrix. There is no record anywhere, not even in the exhaustive lists of the Notitia Dignitatum, of the Sixth Legion ever being transferred from York. In time the Roman name Eboracum was sometimes replaced with ‘Ad Legionem Sextam’ o
r simply ‘Ad Legionem’. It meant ‘the place of the Legion’ or more likely ‘the city of the Legion’.¹⁴¹

  This close association is not surprising. York was known throughout the west as a great imperial military base. Septimus Severus died there, his sons were proclaimed in the legionary fortress, as was Constantine in 306. Around that time the massive fortifications were augmented with huge polygonal bastions which still survive today. It is therefore clear that when reference was made in Britannia or Britain to the city of the Legion, they meant York.

  Precisely because the city was Army Command North and because the empire valued Britannia so highly that in the fourth century at least six expeditions were launched to deal with the Picts, it is not surprising that Germanic mercenaries were stationed there in the fourth century. In fact one of the notable casualties of the Barbarian Conspiracy of 367 was the Dux Britanniarum, a man with a German name, Fullofaudes. And his colleague, the Count of the Saxon Shore, also bore a German name, Nectaridus. He was killed by the Pictish war-bands.¹⁴²

  It is my contention that by the 490s Anglian settlers had combined with the descendants of German mercenary veterans to establish the embryonic kingdom of Deira based near York. There is evidence that the legionary principia was repaired several times in the fifth century, obviously by men who wanted to preserve it as a stronghold.

  As a focal point, York was served by a network of roads north and south, and, less important, east and west. But geography set the strategic hub a few miles north at what is now the town of Catterick. That is where the Roman road system branched north-west over Stainmore to Carlisle, Galloway and Strathclyde, and almost due north to the Gododdin, the Tweed basin and Edinburgh. If the Anglians at York controlled Catterick then they could move easily against any and all of the northern kingdoms. That is why Mynyddawg of Edinburgh sent his cavalry south in 600 to confront the Angles. Behind the heroics and the florid language lay a cold-eyed military purpose. What was strategically important for the north in 600 was also vital for Arthur in the 490s. But unlike the warriors of Mynyddawg, Arthur defeated the embryonic kingdom of Deira and inhibited its growth for several generations. It was a key victory and possibly the reason why Arthur is so vividly remembered in ‘The Gododdin’, Aneirin’s tale of defeat 100 years later: ‘Gwawrddur would feed black ravens on the wall of a fortress, though he were not Arthur.’ And perhaps the reference to feeding ravens – on the corpses of the defeated – is to the principia at York, Ad Legionem Sextam.

  ‘The tenth battle happened on the bank of a river which is called Tribruit.’ Another river battle at a place suitable for cavalry. Tradition places this in southern Scotland but no clear toponymic case can be made.¹⁴³ Tribruit is not related to Teviot or any other river name in the region. Only tradition supports this, a faint echo of folk memory. All that is certain is that Arthur fought once again on the banks of a river. Ten battles won in the same way. Here is a reasonable conjecture of how these were planned.

  Cavalry forces are good gatherers of intelligence. Long-range scouting parties could quickly and quietly locate an infantry army on the march. Following the Roman road system, still in good repair but probably lacking its wooden bridges, it was possible to occupy a covert position near a river which the foot soldiers would have to ford. At that point Arthur’s herald blew the signal to charge and his cavalry emerged from the cover of the woods at the gallop. And then, about fifty yards from the startled infantrymen clambering out of the water, the troopers would kick for an extra burst of speed to make their impetus irresistible.

  Arthur would choose his moment to the instant. Horses will always try to wheel away from an obstacle of any density, like a force of foot soldiers who have had time to form themselves into a spear-bristling shield wall. Disciplined infantry deployed in geography that will protect their flanks can always resist cavalry. On flat ground, by the banks of a river, it would have been very difficult for an infantry commander to avoid being outflanked even if he had had the time, the training and the presence of mind to get his men into a defensive formation. All the archaeological evidence is that Angles and Saxons were poorly equipped, depending themselves on surprise and ferocity for victory.

  The Nennius list is taken from an oral source, most likely a praise-poem sung in the presence of Arthur or soon after his death. The alliterative ‘Cat Coit Celidon’ is a remnant of the original P-Celtic and three of the battle locations preserve a simple rhyme-scheme: Celidon, Guinnion, Badon. An early gloss on the location of the eleventh battle which ‘was done on the hill which is called Agned’ shows that there was a fourth element. In another version of the Nennius text Agned is replaced by the rhyming Bregion. Which in turn gives away the location. Bregion is a P-Celtic rendition of a Roman name.¹⁴⁴ Bremenium was an important outpost fort north of Hadrian’s Wall and astride the arterial Dere Street.¹⁴⁵ It survives in an astonishing state of good repair close to the village of High Rochester. The plan is nearly square covering almost five acres and in places the stone walls rise fifteen feet from the surrounding ditch. The village green of High Rochester was once located inside the fortress and today there is still a scatter of relatively modern houses, one of which has the principia in its back garden, and a sheltered paddock for ponies in the north-east corner where the walls rise to six feet from the inside.

  The walls would have been in much better repair in the 490s and Arthur could have hidden at least 300 horses in the stables and paddocks of the fort, while his scouts brought him reports of the advance of an Angle raiding party up Dere Street from York. Below the fort runs the River Rede and less than a mile away is Elishaw where the Roman Road used to cross the river by a wooden stilt bridge. Then, as now, there is a shallow ford where men could wade up to mid-thigh in the summer, higher in a spate or after winter rain.

  It is logical to think of the battle at Bremenium as part of the same summer campaign that took Arthur’s cavalry down to Catterick and York. But the order is wrong and Tribruit intrudes. Perhaps for poetic reasons these battles were sung out of sequence, and perhaps Tribruit is a place between Rochester and York where the Cymry scored a minor success, further weakening Deira before destroying the remainder of their army near its base.

  Bremenium was and is an important strategic point. While walking round the ancient walls, I heard the sound of square-bashing echoing in the upland air from the modern army camp by the main road, less than half a mile away. And at Qtterburn, four miles south, there is another, larger camp. Battles were fought at or near Bremenium for millennia. Urien began his campaign against the Angles of Lindisfarne in the 580s by defeating them in this place, and in 1388 in a field by the modern roadside the Scots and the English fought the bloody Battle of Otterburn.

  Arthur’s cavalry wait in ambush

  Tribruit and Agned/Bregion are Celtic names, as is the last of Nennius’s sequence of twelve battles: the great and final victory at Mount Badon. ‘The twelfth battle was at Badon Hill in which Arthur destroyed 960 men in a single charge on one day, and no one rode down as many as he did by himself. And in all these battles he emerged as the victor.’ Add to this the note in the Easter Tables which had Arthur once again carrying ‘the Cross of Our Lord Jesus Christ for three days and three nights on his shield and the Britons were the victors’. Gildas writes of ‘obsessio Montis Badonicis’, a siege that was ‘almost the last slaughter of the rascals’.

  Three sources, much in agreement, for one event is remarkable for the period. It underscores Badon’s importance as the crushing victory of Arthur, although none of the three historians says who the enemy were. But if the appearance of Christian iconography is any guide then they would have been pagans and probably Angles or, most likely, Saxons.

  Badon also sounds like a siege by infantry or cavalry defending a mount or hill. Arthur’s final charge is almost certainly on horseback and much better downhill than up. And such a charge would only succeed when an infantry shield wall was broken and turned and cavalry got behind wa
rriors or forced them to run. That is when a real slaughter occurs, not in the grinding, pushing and shoving of hand-to-hand fighting with spear, sword and axe. Cavalry troopers have a telling height advantage over foot soldiers and can cut at their heads and upper body with tremendous downward force. While 960 is a symbolic/poetic number of three times 300 and three score, it does not smack of arithmetic hyperbole but rather of a large combined army utterly defeated, disabling military activity for a generation.

  Badon was decisive, no doubt, but where was it? As Guledig Arthur led the British kingdoms in what sounds like a coalition. Cunedda’s expedition to Wales shows the reach of the Gwyr Y Gogledd. Therefore when historians insist that Badon Hill was in England’s West Country, possibly near Bath, they may well be right.¹⁴⁶

  Solsbury Hill near Batheaston is an attractive candidate. Standing by itself it is steep-sided and small enough to be defended for three days and three nights by a body of dismounted cavalry troopers. And there is enough room to corral their horses behind them. Despite a total lack of toponymic corroboration, it is a likely location.

  A besieging infantry army would have needed to be very large to surround Solsbury Hill with no gaps in the cordon, and that was precisely the Saxons’ problem. In British territory, near Bath, they would have great problems in feeding that number of men for more than three days. On Badon Hill a cavalry force could graze its horses and live off well-filled saddle bags for some time. Arthur may have waited until the besieging cordon was weakened by the dispatch of large foraging parties before he gave the order to mount and charge. Once broken the Saxon formation was lost and ‘no one rode down as many as he did by himself’.