Arthur and the Lost Kingdoms

by Alistair Moffat

  At Stanwick, near Scotch Corner (in itself a historic reference) where the ancient highway branches west up over Stanemore to Carlisle, the Irish Sea and Galloway, archaeologists have found a reoccupied hill fort which was enlarged hugely to an area of over 800 acres. This is level pasture land including a substantial water supply and it was defended by a perimeter of massive dykes. Here Venutius’s army of northern tribesmen mustered from the Yorkshire Dales, the Redesdale Hills, the Cheviots, the Ettrick Forest and from Cumbria and in the enclosure they fenced and pastured their horses.

  When the legionaries arrived at Stanwick, Venutius discovered that he had created a defensive structure whose great length made it impossible to defend. He was quickly brushed aside and the Roman fighting machine rumbled on, with those who escaped the slaughter at Stanwick fleeing before them.

  There is a pungent archaeological footnote here which is worth pausing for. Down in the ditches of Stanwick, among the debris of fallen ramparts, a series of human skulls were found. When the Brigantes and their Scottish allies realized that the Romans were forming to attack the huge, looping perimeter, their Druids built ghost fences. Placing skulls at important points on the turf walls, they tried to use their magic to protect the tribesmen and their horses and repel the advancing legionaries. Once again their gods failed them.

  The instability of the Brigantes under Cartimandua and Venutius persuaded reluctant Roman strategists that the security of the empire would be best served if the whole island of Britain were conquered and fully absorbed. In order to protect the valuable plains in the south, the Romans began a process of subduing the hill country that bordered them. Four legions were stationed in Britannia, a fighting force of 50,000 men, nearly 10 per cent of the entire imperial army. Clearly Rome considered Britannia valuable.

  The legions paused at York after overcoming Venutius in AD 71, dug in and consolidated their hold on Brigantia. Then they turned west to Wales which they subdued after a dogged resistance. With their flank thus protected, the Romans swung north again towards Scotland knowing something about the horse-riding tribesmen who had come to fight with their allies at Stanwick. Perhaps they were intent on punitive measures.

  Tacitus is silent on this since his attention was caught by the arrival of Agricola, his father-in-law, back in Britain as governor in AD 77 and avid for glory.

  Remembering all the time that Tacitus tells only a Roman story, shining Mediterranean light on a culture that left no written record, no sense of its version of events, it is necessary to look once again at the land and its names, and how it offers another interpretation.

  High up in the Cheviots, where the grey-green hills lie folded like the blankets on a sleeping giant, near the source of the talking Kale Water, where no one lives now, the P-Celts left us a name which marked the moment of the Roman invasion. When the legions climbed through the hills, their scouts came across a well-defended fortress on top of Woden Law. They marched around it and, about a mile away, built a series of camps at a place the P-Celts called Pennymuir. It means ‘at the head of the walls’. The field name Pennymuir Rig straddles the southern wall of what was probably the second camp.

  More pointed is the origin of another name thirty miles to the north-west. Peebles is an oddity, quite unlike other names around it. The market town lies at the foot of another P-Celtic stronghold on Cademuir Hill. When the Romans penetrated up the Tweed valley, they camped near Peebles. Looking down from their fortress, the P-Celts watched the legionaries pitch their leather tents. Pybyll is P-Celtic for tent and the name Peebles remembers something that the Romans did when they first arrived in the Tweed valley in 80.

  Agricola’s first campaign was faultlessly executed. The Ninth Legion left Corbridge and followed a route up Redesdale and over the Cheviots which later became Dere Street and which is now the A68. When Agricola and his scouts gained the top of Redesdale they could see the dominating feature of the landscape of the Tweed valley which lay below them. Rising out of the woods were the three summits of the Eildon Hills. Like an arrow the Romans made straight for them, Agricola instructing his engineers to survey the line of march for a road, much of which is still visible today.

  Broadly, the road and fort system tracks the progress of the Roman conquest of Scotland but the direction it took and the comparative ease with which it was accomplished allow sensible conjecture on the politics. Agricola’s military intelligence had clearly added to his existing knowledge of the tribes of southern Scotland. Tacitus only resorts to the adjectives novae or ignotae when he describes the tribes beyond the Forth and Clyde line or across the Irish Sea. The men of the Borders were neither new nor unknown.

  The Selgovae of the hills of Teviotdale and Tweeddale were the clear object of Roman strategy. They had been with Venutius at Stanwick and, like the hill peoples of northern England, they were certainly hostile to Rome. The line of Roman advance shows that. At the Eildons, Agricola established a huge base at the foot of the great hill fort. It commanded a long view up the Tweed valley, was near the mouths of other Selgovan valleys: Ettrick, Yarrow, Gala Water were the major arteries. And the fort also acted as a road depot, situated on Dere Street, halfway between the Cheviots and the Firth of Forth.

  Agricola then struck north up the Leader valley over Soutra Hill and down to Inveresk, his route close to the eastern borders of the Selgovae. On the west the Twentieth Legion marched north from Carlisle up what is now the A74 planting forts and garrisoning them as they went. Then at Crawford they turned to the north-east over the hills to the Esk valley and went on from there to link with the Ninth Legion on the shores of the Forth in AD 81. It was a clinically executed pincer movement tightly circling the hill country of the Borders and with forts at each valley mouth, Agricola locked the Selgovae into their upland fastness.

  The Romans had already used this method of containment successfully in the Pennines and Cumbria against the Brigantes and also the hill peoples of Wales. On his drive northwards Agricola ignored the east of Northumberland and paid no attention to the lower Tweed valley, the Lammermuir Hills or East Lothian. A more easterly route would have made campaigning much easier; the terrain is relatively flat and if the line of the A1 had been followed then the fleet could have supported and supplied the column, as it was to do later. Also these ignored areas represented the most valuable, fertile ground, far more attractive and productive than the rough bounds of Annandale and Teviotdale. The very likely truth is that Agricola was indifferent to the eastern tribe, the Votadini, because he had done a deal. Either cash was paid or a treaty made, or both. In any case the east is remarkable for an almost total absence of Roman forts or roads. In addition it is very likely that it suited the Votadini to see their aggressive neighbours, the Selgovae, receiving so much attention from the invaders. And if Agricola had explained to them what his overall campaign plan was, then they would have been happy to see northern hostiles also dealt with by their new allies. Which in a historical way is disappointing. If the Votadini had been unfriendly to Rome then Tacitus would have told us more. As ever we will have to depend on names and local knowledge.

  In the southern half of Scotland, the Romans found four tribal groupings in all: the Novantae in Galloway, the Damnonii in the Clyde valley, the Selgovae in the central Southern Uplands and the Votadini in Lothian and the Tweed basin. These names were not conferred on the tribes but derived from what the Romans discovered about them and much must be made of their meaning.

  Novantae first. It compares with the tribe centred in Middlesex called the Trinovantes. The root is transparent for once. The P-Celtic word newydd is related to the Latin novus which is in turn connected to the English ‘new’. But how can a tribe be titled ‘thrice-new’? Better to slacken off the meaning a little to allow ‘vigorous’ or ‘lively’, both possible for the Welsh word newydd. So, the Novantae are the vigorous people, a description the legions might have accepted after the campaign of 83 in Galloway.

  Damnonii is much more difficult. Again
it compares with an English example where there were Dumnonii in the south-east, a name that survives as Devon. Domnan is a Q-Celtic personal name, close to Columba’s biographer Adomnan and a group of warriors in Irish myth history known as the Fir Domnann.

  There is a tradition that they were so called because they were diggers, the men who deepened the earth. However, before the conjecture becomes impossibly strung out, let us remember that modern Gaelic retains the word domhain which means ‘deep’ and old Q-Celtic has dubros for the same thing. Go back to Devon and Cornwall where the tradition of mining for iron ore and tin is millennia old and that should complete a tidy circle around the Irish Sea. The Damnonii of the Clyde valley were known as the diggers or, better, the miners. (Sad to reflect in passing that the closure of the North Lanarkshire coalfield ended a tradition stretching back two or three millennia.)

  While names and places shifted, fell out of use and were got wrong, it is consistency of memory that is most striking. One windy winter Saturday I went up the Yarrow valley to look at some standing stones. About half a mile west of Yarrow Kirk, on the right-hand side of the road to Moffat, there are two massive unhewn stones. A strong local tradition remembers that they are monuments to two warrior chieftains, and further, that their followers were not so honoured, their bodies being disposed of in the marshy pool in the haughland to the north, known still as the Dead Lake.

  Three hundred yards to the west of this place, called the Warriors’ Rest, there is a large flat stone with an inscription on it which is now very weathered and difficult to read. However a nineteenth-century antiquary wrote it down before it faded. It solves part of the puzzle:

  Hic Memoriae et

  Bello Insignisimi Princi

  pes Nudi

  Dumnogeni Hic Iacent

  In Tumulo Duo Filii

  Liberalis

  The stone commemorates two men who fell in battle in that place. For my immediate purpose, it shows the remarkable persistence of a name related to the tribe of the Damnonii and that there were princes who claimed to be sprung from that royal line 500 years after Tacitus notes the name. For a later purpose this memorial to a forgotten battle will be a telling part of a larger picture.

  Immediately to the north of the Dead Lake is a field known as Annan Street. Like the eponymous river in Dumfriesshire, the first part might simply stand for a P-Celtic word for water much in the way that Britain’s several rivers Esk and Usk are corruptions of the Q-Celtic uisge for water. Mind you, another corruption of the same word is a deal more interesting: that is, ‘whisky’. Street is also straightforward. As in Dere Street or Watling Street, it is an English term for a Roman road. So, Water (Roman) Road. Or a Roman road by water. No matter how ingenious the lexical juggling, the topographical facts cannot be bent to fit this name. There is no Roman road in this hilly and windy valley and while Annan may be an old P-Celtic name replaced later by Q-Celtic Yarrow (from garbh meaning ‘wild’), why has it survived in the name of a field 200 yards from the river?

  And, more intriguingly, there is a very long earthwork lying about a mile due west of Annan Street known locally as the Catrail. Nineteenth-century local historians believed that it marked a physical boundary between P-Celtic speakers and Anglian or English-speaking incomers.²⁹ While this has been discredited recently, it seems no more unlikely to me than the walls built by the Emperors Hadrian and Antoninus Pius. In any event the disposition of P-Celtic and Anglian place-names to the west and east respectively of the earthwork is uncannily consistent. And to complete the pattern of competing toponymy, there is a Victorian printing of the Ordnance Survey of the Yarrow valley which was surveyed in 1855 and which describes the Catrail as a Pictish earthwork which is, in turn, what Catrail means in P-Celtic – a fortress, or, sometimes, simply home.

  The Yarrow Stone

  I have taken a long detour away from the derivation of the Damnonii to show how field-walking and intimate local knowledge can put colour on a dreich historical landscape. I shall come back to Annan Street and the Catrail later, but suffice it to say for the moment that Yarrow Kirk was an important place in sixth-century Scotland: a battle fought, princes buried, boundaries set and most remarkably a traditional reading of the ground surviving 1,500 years, and hardly noticed at all in historical accounts.

  Which in a tidy sort of way swings the focus south and east of the Damnonii to the Selgovae, for Yarrow Kirk lies in the centre of their territory. From the angle of Agricola’s advance in AD 80 and his strategic aim of splitting the Novantae from the Selgovae, their western boundary probably lay along the modern A74 and the A702 as it branches north-east towards Edinburgh. The Votadini bordered with them to the north and east, although precisely where is not clear, and to the south the lands of the Selgovae petered out in the worthless wastes of the south-western Cheviots, where their neighbours were the Brigantes.

  That reading of Tacitus places the Selgovae in the Ettrick, Yarrow, Jed and Teviot valleys and in the hill country of upper Tweeddale. And their name came from the way in which that wild habitat forced the tribe to live. It comes from P-Celtic hela, to hunt, more clearly seen in the Q-Celtic cousin-word sealgair for hunter, at least in its spelling.

  Although one name is a fine thread to hang a picture on, the hunters and more particularly where they hunted offer a splash of colour in the landscape if several different pieces are brought sensibly together.

  Let us begin in Rome in 80. It is June, the grand opening of the new theatre known as the Coliseum, and the description of the inaugural games left by the poet Martial. Deep in the dark holding pens under the floor of the arena sits a large brown bear. Frightened and exhausted after its long journey, the animal roars and growls as its handlers hook their long wooden staffs on to its collar. They goad and push the bear along the tunnel to the arena and the baying Roman crowd. Blinking into the blinding Mediterranean sunlight, the bear is let loose and the announcement rings around the Coliseum, ‘a bear from far Caledonia’ sent to Rome by the governor of Britannia Gnaeus Julius Agricola, doubtless to publicize his conquest of southern Scotland and by extension the land of the Selgovae, where bears lived in the forests.

  A fine thread right enough, but two bits of solid information can be drawn: by June 80 Agricola had only just begun to campaign in southern Scotland but the place is known as Caledonia; and although the bear could have been got in any number of ways, its capture signals some sort of relationship (defeat, tribute) with the tribes and also says something about their habitat.

  The Selgovae were hunters who lived up-country, in the wild lands, the less valuable areas. As ever names tell a story where documents are lacking, and pollen archaeology describes an ancient landscape.

  When the legions marched over the Cheviot tops and looked down into the Border valleys, they saw a great forest broken only by hilltops and rivers. The climate was warmer 2,000 years ago and trees were able to grow on ground up to 2,000 feet. The principal species were alder, oak, ash, Scots pine, elm and birch – tall, mostly hardwood trees naturally regenerating, giving cover not only for bears but also wolves, elk, deer and myriad smaller species. The Selgovae had plenty to hunt.

  This tree-covered wilderness is remembered in the names Ettrick Forest, Jedforest and Bowmont Forest, although the meaning of ‘forest’ needs to be looked at twice. It comes from monastic Latin forestis which means a large uncultivated tract of land usually but not always tree-covered. Often in later times it meant ground governed by forest law which protected its use as a royal game reserve. More closely forestis means ‘the outside wood’ as it does for Sherwood Forest or the New Forest. It compares with parcus which means ‘walled-in forest’. There is a good example of this in Selkirk where a large area of rough pasture on the edge of the town, between it and the Ettrick Forest, is called the Deer Park. For completeness the Deer Park is now part of Hartwoodburn Farm, or in another way, the farm by the burn in the Stag’s Wood. It is a neat illustration of primitive forest management with a domestic deer p
ark for breeding and easy hunting, next to the great wood where stags ran wild. A good strong echo from the time when people lived by the hunt.

  Selgovae has not left much of a toponymic trace on the land. Selkirk is thought to be an Anglian word: schelch for wood and chirche for a church or kirk. The last element is beyond doubt but there is obviously a connection between the P-Celtic hela, the Q-Celtic seilg for hunt and the Anglian schelch for wood. Perhaps it is not stretching common sense too far to hear the Selgovae live on in the name Selkirk, the Queen of the Forest as she is known locally? The only other possible descendant is the mysterious place called Segloes included in a list of meeting places for tribesmen licensed by the Roman occupation. But the name is a confusion. It is a corrupted version of Locus Selgovensis. It is one of two places located in southern Scotland which appeared in the Roman list. The first is Locus Maponi in the territory of the Novantes and can clearly be connected to Clochmabenstane, which is a standing stone on the flatlands near Gretna in Dumfriesshire and which remained an important meeting place well into the seventeenth century. A granite boulder, seven feet six inches high, it was associated with Apollo Maponos, the Romanized name of the P-Celtic sun god.

  Locus Selgovensis is much harder to find and archaeology and logistics are a surer guide than place-names in this case. East of Peebles where the Lyne Water runs into the Tweed there is an ancient meeting place known as the Sheriff Muir. Via river valleys it is accessible from the northern hills, from Clydesdale through the Biggar Gap to the west, from the south down the upper Tweed valley and from the east from the Tweed again. For these logistical reasons the Sheriff Muir remained the muster point for the Peebleshire Militia until the late eighteenth century.

  Across the Lyne Water, wedged in a v-shape where the Meldon Burn feeds it, there is archaeological evidence of a massive wooden stockade enclosing an area of twenty acres. The posts were ten to twelve feet high and the area was entered by a wide passageway. It may have been an enclosure for stock gathered there for the cycle of the Celtic festivals. At all events the Lyne ceremonial complex is in the centre of Selgovae territory, accessible to all its outposts and a place sensitive enough for the Romans to build a large fort nearby to keep watch on their old enemies.