Arthur and the Lost Kingdoms

by Alistair Moffat

  It is difficult to tell what the strength of the garrison of Britannia was in 367. However the actions of Count Theodosius the following year offer some retrospective sense of the disaster that had overtaken the province. Nectaridus, the Count of the Saxon Shore, had been killed and his small force defeated and disintegrated. A man with a German name, probably a professional soldier from the Rhineland, Fullofaudes, had been Dux Britanniarum and was captured by the Picts.⁷⁴ Many of his soldiers had simply run away and bands of deserters had formed in the countryside.

  When Theodosius landed with four regiments of regular troops at Richborough, he made straight for London so that he could use the provincial capital and the road system that radiated from it to begin the work of re-establishing order. Instead of an invading army of Picts, Scots and Atecotti, Theodosius came across small bands of warriors who had overwintered in the south and had carried on harrying and wasting villas and towns. These were easy to pick off and once he had captured or killed those who did not flee in front of his forces, he issued an amnesty to all deserters. This allowed Theodosius to use experienced men in re-creating the three army commands of the province and to regarrison Hadrian’s Wall.

  Because of their treachery and probably because he did not have the manpower to spare, Theodosius abolished the Areani and also did not reoccupy any of the outpost forts north of the Wall. That necessitated an extension of the frontier policy initiated thirty years before by the Emperor Constans. In essence Theodosius took three of the four original tribal groupings which had grown into the kingdoms of Strathclyde, Gododdin and Rheged as a political, cultural and military base and set professional soldiers in charge of modest cavalry forces as prefects over them.

  This is a vital step in the narrative and it needs to be shown clearly what happened. The northern genealogies provide the first and most important piece of evidence. Around 370 a man called Quintilius Clemens is named as the ruler of Strathclyde.⁷⁵ He is styled in the Roman fashion with nomen and cognomen, or family and personal names in the correct order. Antonius Donatus at the same time assumed command of forces in south-west Scotland, perhaps stretching as far east as Selkirk and south to Carlisle. These were the marches of the embryonic kingdom of Rheged. Catellius Decianus was given authority over the northern Gododdin, that is the Lothians and the place that became known as Manau-Gododdin, around Stirling in the west. His centre was most likely Traprain Law. South of the Lammermuirs were the southern Gododdin whose territory comprised the Tweed basin and the flatlands of modern Northumberland, perhaps down as far as the wall itself. Their commander was a man called Paternus Pesrut, which translates as ‘Paternus of the Red Cloak’. This is a description of a serving Roman officer.

  Their names, their territories and their coincidence in time show that these men were not native kings flattered by fancy Roman titles or names. They were professional soldiers planted by Theodosius in a buffer zone between the Roman walls with a clear brief to prevent a recurrence of what happened in 367.

  There is no doubt that these appointments formed a consistent part of imperial policy. Two years later in North Africa, Theodosius faced very similar problems. This time his solutions were recorded by two written sources. First the Roman historian Ammianus: ‘He sent men experienced in persuasion to the surrounding tribes … to entice them to an alliance, now by fear, now by bribes, sometimes by promising pardon for their impudence.’ And then he ‘put reliable prefects in charge of the peoples he encountered’. Corroboration is supplied by St Augustine of Hippo:⁷⁶ ‘A few years ago a small number of barbarian peoples were pacified and attached to the Roman frontier, so that they no longer had their own kings, but were ruled by prefects appointed by the Roman Empire.’

  Now, even more corroboration is available from recent archaeological work around Traprain Law, the capital of the northern Gododdin. Very significant numbers of fourth-century coins have been found recently in this place and another location. These are not silver or gold coins but small copper pieces only of value in a money economy, the sort of currency that was used by soldiers and accepted in return for goods and services. A full explanation of the greater significance of these finds belongs later in this story but suffice it to say for the moment that they show significant late fourth-century Roman activity in southern Scotland.

  Count Theodosius was a thorough soldier, not only able to defeat the enemies of the empire but also to analyse the implications of victory. The experience of Roman military planners had been that any defeat of barbarians was temporary. After a time more of them came. Theodosius knew that and when he placed his prefects in charge of the nascent intramural kingdoms, he made longer-term recommendations in his report to Rome. Ammianus tells us that he ‘protected the frontier with lookouts and garrisons, recovering a province that had yielded to enemy control, so restoring it that, as his report advised, it should have a legitimate governor, and be henceforth styled “Valentia”.’

  This last was a familiar piece of flattery; the joint emperors at the time were Valens and Valentinian. But more interesting is the creation of a buffer province between the insurgent tribes of the north and the valuable province of Britannia in the south. The correct location of Valentia is important at this point and Roman sources are disappointing in that they assume that a reader will know. Given the appointment of the prefects and given that Theodosius would have wanted to motivate the soldiers stationed between the walls, I believe that Valentia incorporated the old occupied territory of the first and second centuries which lay between Antonine and Hadrian’s walls. In this way he drew the young kingdoms into the empire, made the tribesmen citizens and gave them a stake in its survival and prosperity.

  There is evidence to support this placing of Valentia. First the coins. The new discoveries of 539 Roman coins in the Borders are not hoards of gold and silver cached away during bad times. They are not the stuff of treasure. Almost all are bronze, around half an inch across and some have dates as late as 402. Of no use for bribery and with no value as trophies, these discoveries are of the small change of everyday use, the operation of a money economy in the Scottish Border country towards the end of the fourth century. These coins can only have arrived in the pouches of Roman soldiers and the money chests of their paymasters. Albeit in small numbers and in a new political arrangement, they show that the Romans were back in the Tweed basin for the last three decades of the fourth century. Long enough to have a military impact and long enough to reinforce the old traditions of warfare learned first by the Votadini at Trimontium on the parade ground and in the riding school.

  The tradition of re-Romanization lasted into the memory of St Patrick. He had reason to denounce his fellow P-Celts and in particular the new kingdom of Strathclyde established by the prefect Quintilius Clemens. Patrick attacked his successor Coroticus saying that the behaviour of his soldiers was to be condemned. It meant that they were ‘not citizens of the Holy Romans, but of the devil, living in the enemy ways of the barbarians’.⁷⁷

  More telling is a later memory of the citizens. In his barren attack on the inadequacies of the mid sixth-century values of Britain, the Welsh monk Gildas gave the soldiers of the resistance forces an interesting name. He called them cives or ‘the citizens’.⁷⁸ It is a close lexical cousin to the Romano-British vulgate term combrogi which has a looser meaning of ‘fellow countrymen’.⁷⁹ The P-Celts of Britannia came to call themselves the Cumber and the map is studded with place names that incorporate the word. Comberton, Cumberlow and many others lie in the south-east of England, showing that the Anglo-Saxons did not sweep the humbler sorts of native Britons off the land and drive them relentlessly westwards. The most obvious survival of the name is in the void county of Cumberland which is now incorporated in the larger unit and older form of Cumbria. Historians record that P-Celtic or Old Welsh was spoken in the Eden valley around Carlisle until the early 1500s and in the Lakeland fells it persisted even longer, into the nineteenth century when shepherds in Langdale were kn
own to have used Welsh numbers to account for their flocks.⁸⁰

  But it is in the Welsh word for Wales that the name has endured most vividly. ‘Cymry am byth!’ or ‘Wales for ever!’ is what the passionate roar from the terraces of Cardiff Arms Park, and while not eternal, Cymry is a very old name indeed. And a remarkable remnant of Britannia, a small Celtic nation on the edge of Europe that calls itself the Citizens.

  It was also what the armies of the generals began to call themselves in the early sixth century. Against the barbarians Arthur led the Cymry, the soldiers of the citizens, the last phase of organized resistance to the hordes of invaders who had occupied all of the Western Roman Empire by 500.

  Cymry is powerful in this story of names, not only because of the romance of the thing but also because it rings such a clear note down 1,500 years of experience. It tells us how these men saw themselves, the army of the citizens, heroic in the teeth of relentless barbarian advance. The keepers of light against the onrush of darkness, Christians against pagans, knowledge and order against savagery, ignorance and greed. Perhaps we understand the spirit of the army of the Cymry in our hearts. Without knowing any facts, we instinctively stand with them, even though they lost and even though history has almost forgotten them.

  Arthur’s campaigns, however, will be better understood in a careful context clearly set out. The prefects of southern Scotland are remembered by the northern genealogies as the founders of dynasties and their names were rendered in P-Celtic style. Cinhil for Quintilius, Cluim for Clemens, Padarn for Paternus and Anhun Dunawd for Antonius Donatus. However there is another name entangled with the men in the king lists which also shows a Roman pedigree. He was Macsen Wledig, and his story will take us a step nearer to Arthur.⁸¹

  The victory of Count Theodosius was so complete that the Notitia Dignitatum later records four regiments of Atecotti fighting in Europe in imperial service. So complete that in 370 the prosperity of Britannia had risen to new heights.

  However, there was trouble of a different sort in Rome itself. The army, volatile at the best of times, had begun to resent the privileges granted by the young Emperor Gratian to his barbarian bodyguard. Discontent spread like a cancer through the provincial commands until in 383 the army in Britannia proclaimed a new emperor, a man unconnected with the Roman nobility, a usurper. He was a Spaniard who had fought against the Picts, Scots and Atecotti with Count Theodosius fifteen years before. His name, Magnus Maximus, was bent into Macsen Wledig by the P-Celts who supported him. It simply means ‘Macsen the General’ and it was the first use of Wledig, or Guledig, as a title for a military commander in Britain.

  Gaul deserted the Emperor Gratian without bloodshed and Macsen moved quickly to consolidate his control across the Channel. He welded together a field army out of parts of the British garrison and thereby dangerously weakened the province’s defences, particularly in Wales where he probably removed the remnants of the Twentieth Legion from the fortress at Chester. This allowed the gradual penetration of invaders from Ireland into the Lleyn peninsula in the north and Pembrokeshire in the south.

  Macsen quickly pulled the provinces of the western empire under his banner and within a year Africa, Spain, Gaul, Germany, Italy and Britain acknowledged his imperium. At his instigation the mint at London, now named Augusta, began once more to produce gold and silver coins, sufficient of which have survived to suggest considerable numbers were cast.

  Macsen’s co-emperor in the east, the legitimate Theodosius I, tolerated the usurper for four years until in the Balkans their armies clashed in 388. The battle weht against Macsen and he was beheaded some days later.

  This episode is important to this story in a number of ways. When Macsen was defeated there is evidence, again from the Notitia, that he had recruited native British forces and that after his defeat they retreated into Gaul and settled in Armorica.⁸² The town name of Bretteville remembers this. These men began a tradition of British settlement which continued into the eighth century, established the P-Celtic language which, as Breton, is still spoken in the area of north-west France now called Bretagne, or Brittany, or Little Britain.

  The early British king-lists gave Macsen a founding role in Wales and his prowess is recorded in an early Welsh tale preserved in the collection of stories known as the Mabinogion.⁸³ ‘The Dream of Macsen’ contains much romance and also several accurate historical facts. More pointedly, he was also installed as the founder of the northern dynasty of Rheged, one of the kingdoms of southern Scotland most successful in the sixth century in dealing with the Anglo-Saxon invasions. This is less surprising when the settlement of P-Celtic soldiers in Armorica is taken into account. Because of the lack of a tradition of arms in the south, it is highly likely that he led cavalry warriors from the Gododdin, Rheged arid Strathclyde in his European campaigns.

  This tradition is seated in historical reality. In 384 Macsen returned to Britannia from Gaul to lead the P-Celts of the north against the Picts and Scots.⁸⁴ His campaign was successful and his victory enduring. From around this period a new kingdom based in the Forth valley and on the fortress on Stirling Castle rock came into being. Called Manau or Manau Gododdin, its creation signalled a new policy of aggressive counter-invasion into the territory of the Picts.

  The name itself is difficult to parse. It may be related to the derivation of the Isle of Man and the Welsh name for Anglesey, Ynys Mon, which are both connected to the name of the Celtic god of the sea Mannan or Manannan mac lir in the Irish version.⁸⁵ However that may be, it is a derivation offering little enlightenment. The place-names of the area are much more helpful. In his Ecclesiastical History of the English People, written around 730, Bede described P-Celtic Scotland in a rapid sketch: ‘In the middle of the eastern estuary [the Forth] stands the City of Guidi.’⁸⁶ Bede was usually precise in his Latin and by urbs he meant a substantial settlement and not a few huts on an island. Guidi sounds very much like a reference to Goddeu or Gododdin. The River Forth is tidal at Stirling and the town was almost surrounded by a waterlogged moss until modern times. Urbs Guidi was Bede’s name for Stirling. P-Celtic names survive in the area at Polmaise, ‘the settlement on the plain’ and at Gogar, meaning ‘a height’ or ‘a small hill’. But they are more precise in the east where toponymy strongly suggests the boundary of the new kingdom, a frontier with the Pictish tribes of Fife and the north.

  The first clue is in the old Scottish county town of Clackmannan. Extending no more than a few square miles Clackmannanshire was by far Scotland’s smallest county and that in itself implies an ancient survival, an echo of the P-Celtic kingdom of Manau. Clackmannan is partly a Q-Celtic name which pulls out into Clach na Manan.⁸⁷ With Manau, taking a genitive case by adding a final ‘n’ it means simply ‘the Stone of Manau’. Like the marker of the northern bounds of the kingdom of Strathclyde at Clach nam Breatainn, I believe that Clackmannan sits on a frontier. The only difference is that the stone itself has survived. It stands next to the medieval tolbooth in the centre of the little town, a rough, unhewn rock people call simply ‘the stone’.

  A few miles north lies the steep-sided range of hills known as the Ochils and between them and Clackmannan are two more P-Celtic names that help to define the old kingdom. The Powis Burn runs south from the hills and empties into the upper estuary of the Firth of Forth. It is a name like the Welsh local authority Powys and it has the same derivation. From the Latin pagenses it means ‘provincial land’ or more precisely ‘frontier land’ which, as the eastern-most part of Wales, Powys most certainly is.⁸⁸ The Powis Burn strongly suggests the eastern mark of Manau, but more confirmation is available 800 feet above it.

  The hill that glowers over the coastal plain and its settlements has an odd name. Dumyat is P-Celtic and it means ‘the fort of the Miathi’. Nearby is Myot Hill which shares the derivation.⁸⁹ Both come from Miathi, or Maeatae which was an alternative name for the Picts, probably a sub-tribe based in the Ochil Hills and the up-country to the north of the
m. Clackmannan, Powis Burn and Dumyat are only eight miles apart. The names remember a frontier, a place where ancient enemies stood off from each other. And also a place conquered, probably during Macsen’s campaign of 384, and held by the powerful kingdom of the Gododdin. Their military skills, their ability to mobilize quickly and their political confidence allowed them to occupy and settle flat land with few natural defensive points, overlooked by the Pictish Ochil Hills and hemmed in by the Firth of Forth in the south. The kingdom of Manau lay at the navel of Scotland, holding the land bridge to the Highlands in the north and the fertile grasslands of the south. Able to see the night-time fires of their enemies in the hills and knowing that they could break any peace treaty in a moment, the P-Celts held Manau by force of arms. It was their own buffer state between the Picts and their own southern heartlands. Between the battles of Macsen in 384 and the time of Arthur a hundred years later, the military kingdom of Manau grew in power and in prestige.

  For ten years after the execution of Macsen the wealthy south of Britannia was peaceful. But in 396 there is evidence of renewed Pictish raiding. It seems that they sailed out of their east coast harbours in Fife, the Tay, the Angus coast and the Moray Firth, bypassed the P-Celtic kingdoms of Scotland and struck at the less defended towns and villas of the south. The Roman historian Ammianus records a Pictish raid on London at this time. What is striking is the strategy of the Picts. They had always had a naval capability but their audacity in attacking London and its hinterland with a force limited in size by sea travel is interesting. It showed how vulnerable the south had become after Macsen evacuated most of its garrison in 383.

  In Rome the government of the young Emperor Honorius was managed by a general known to us as Stilicho the Vandal. He launched an expedition to restore imperial order in Britain in 396–8 but could not scrape together more forces to deter future Pictish raids. In fact the armies of Alaric posed such a serious threat to the security of Italy in 401 that Stilicho was forced to withdraw even more soldiers across the Channel.