Birdoswald (Banna) Roman Fort, Irthing Valley, Hadrian's Wall
As most of my readers knows by now, my favorite candidate for the famous Dark Age Arthur is Ceidio son of Arthwys. But I've bounced Ceidio around a bit trying to settle on his primary court. Three possibilities have been explored: 1) Banna/Birdoswald Roman fort, the birthplace of St. Patrick and site of a sub-Roman and early medieval hall 2) Camboglanna/Castlesteads, where our hero perished along with Modred/Moderatus and 3) Uxellodunum/Stanwix, the largest cavalry fort in all of Britain and the command center of at least the western side of the Wall. It appears to have been known in local tradition as Arthur's Burg.
To try and narrow these options down, I've discussed the significance of the name Arthwys in some detail. This is an eponym for a place, in this instance the Irthing River in Cumbria. Both Banna and Camboglanna are in the Irthing Valley. Arthwys may be compared with Glywys, the eponymous founder of Glwysing in south Wales, from the Romano-British period Glevum used for Gloucester. It is not, therefore, the name of a real chieftain. Nonetheless, Arthwys is claimed as the father of both Ceidio and Eliffer of York (who is elsewhere made the son of Gwrwst Ledlum/Fergus Mor of Dalriada, an intrusion into the early portions of some of the genealogical lines of the Men of the North).
While is thus tempting to simply say, "Well, if Eliffer of York can be son of Arthwys of the Irthing, then why can't Ceidio son of the same man/place be of Stanwix?" But this isn't a simple parallel. For Ceidio (whose hypocoristic name originally meant 'Battle-leader' or 'Battle-ruler' or the like, which is what doubtless produced the dux bellorum title given to Arthur in the HISTORIA BRITTONUM) would have also born the name Arthur (from Artorius). And Arthur was associated by the Britons with their word for bear, 'arth'. So the question we must ask is this: if he were named Arthur because he was born in the Irthing Valley - almost certainly at Banna, where his father's Pendragon title may be a reference to the draco of the Dacians who were garrisoned there - did he also rule from there? Or was his headquarters located outside of the Irthing Valley? And, if so, where exactly?
Ceidio is made the father of Gwenddolau, the lord of Myrddin (Merlin). His name is preserved in Carwinley, i.e. Caer Gwenddolau. However, even here there is a problem, for Gwenddolau means, literally, 'White dales', and is, therefore, almost certainly yet another personified place-name. Still, if we allow for Ceidio having been born at Banna and look towards Carwinley, and then triangulate to Aballava/Avalana/"Avalon" at Burgh-by-Marsh, and imagine traveling via the Roman road from, say, Camboglanna to Stanwix and then north to the Roman fort at Netherby hard by Carwinley, that Stanwix really is at the hub of this grouping of forts.
Stanwix was garrisoned by the Ala Petriana, named for a certain Petra. Either by mistake or as a sort of nickname, the NOTITIA DIGNITATUM calls Uxellodunum 'Petrianis.' Once again, this was the largest cavalry unit in all of Britain. I have before very tentatively put forward the idea that the reason Pedr (Petrus) of Dyfed in southwest Wales names a son Arthur was because the earlier, much more famous one had become linked to the descendants of the Ala Petriana of Stanwix. In considering this possibility a second time, it does not seem so untenable a theory after all. We might assume that a cavalry force under Arthur had assumed the name of the earlier ala or that a newly raised force patterned after that ala inherited the Petriana designation. Had Arthur and his horsemen became famous, a prince of Dyfed named Petrus may well have named his son after this great Northern leader of 'the Petriana.'
[1]
In this section I will be discussing the case that has been recently made by Ken Dark of the University of Reading for the sub-Roman (i.e. 5th-6th century CE) re-use of Hadrian’s Wall, as well as of forts along the Wall and in the adjacent tribal territory of the ancient Brigantian kingdom.
According to Dark, from whose paper I will liberally quote:
“… eight fourth-century fort sites on, or close to, the line of Hadrian’s Wall have produced, albeit sometimes slight, evidence of fifth -sixth-century use. Nor is this simply a reflection of a pattern found father north; for no Roman fort site in what is now Scotland has any plausible evidence of immediately post-Roman use. Thus the situation to the north of the Wall is similar to that found in Wales.
What is more surprising still is the character of the reuse found on the line of the Wall. Two sites, Housesteads and Corbridge, have evidence not only of internal occupation, but of refortification; at Birdoswald there are the well known ‘halls’, while at Chesterholma Class-I inscribed stone of the late fifth or early sixth century come from the immediate vicinity of the fort. At South Shields there is also evidence of re-fortification, and there is an external inhumation cemetery. Another Class-I stone was identified by C.A.R. Radford at Castlesteads [I have rendered the inscription of this stone above in Chapter 3].
At Binchester immediately to the south of the Wall, and at Carvoran, Benwell and Housesteads on its line, there are early Anglo- Saxon burials or finds, while at Chestersand Chesterholm (perhaps sixth century) Anglo - Saxon annular brooches come from within the forts, although these may be somewhat later in date than the other material so far mentioned.
At the western terminal of the Wall, a town-site,
Carlisle, though not necessarily primarily military in the Late Roman period, has also produced substantial evidence of sub-Roman occupation, with continued use of Roman-period buildings into the fifth, if not sixth, century.
Many scholars accept that Carlisle was part of the late fourth-century Wall-system, perhaps even its headquarters, and at Corbridge, the other town-site intimately connected with the
Wall, fifth -and sixth century material has also been found, including, perhaps, evidence of continuing British and Anglo-Saxon use. In the North as a whole, fifth- or sixth-century evidence from what had been Late Roman towns is not common. York, Aldborough, Malton, and Catterick are our only other examples. Two of these sites (York and Malton) were part of the same Late Roman military command as Hadrian’s Wall: that of the Dux Britanniarum.
It is interesting that, of the sites at Manchester and Ribchester– between the Mersey and Carlislethe only fort-sites known to have possible fifth or sixth -century evidence – Ribchester was not only part of the command of the Dux Britanniarum, but also listed as per lineum ualli in the Notitia Dignitatum. It is, therefore, remarkable that out of the twelve fourth-century Roman military sites in northern and western Britain to have produced convincingly datable structural, artefactual, or stratigraphic evidence of fifth-or sixth-century occupation, eleven were, almost certainly, part of the Late Roman military command.
Eight of these were probably within the same part of that command, and eight comprise a linear group (the only regional group) which stretches along the whole line of Hadrian’s Wall from east to west. The two more substantial late fourth- century settlements adjacent to the Wall – Carlisle and Corbridge– have also produced fifth- and sixth-century evidence and two of the other towns with such evidence were also late fourth-century strategic centres under the military command of the Dux.”
After setting forth these facts, and discussing them, Dr. Dark offers a rather revolutionary idea:
“Although it is difficult, therefore, to ascertain whether the military project which I have described was the work of an alliance or a north British kingdom or over-kingdom, there does seem to be reason to suppose that it may have represented a post-Roman form of the command of the Dux Britanniarum…
This archaeological pattern, however it is interpreted, is of the greatest interest not only to the study of the fifth-and sixth- century north of
Britain, but to that of the end of Roman Britain and the end of the Western Roman Empire as a whole. It may provide evidence for the latest functioning military command of Roman derivation in the West, outside the areas of Eastern Imperial control, and could be testimony to the largest Insular Celtic kingdom known to us.”
In another paper, Ken and S.P. Dark rebut P.J. Casey’s argument for a re -interpretation of the reuse and re-fortification of the Wall and its associated forts. His conclusion for this paper reads as follows:
“If one adopts the interpretation that the Wall forts were reused in the later fifth-early sixth century for a series of sub-Roman secular elite settlements, then the associated problems involved in explaining this new evidence of occupation at that time disappear…
So, the interpretation that the Wall became a series of secular elite settlements, discontinuous from the Late Roman activity at the forts within which they were sited, is compatible with the evidence of pollen analysis, while the alternative interpretations are both rendered unlikely by it. This does not, of course, make the suggestion that this reoccupation represents the sub- Roman reconstruction of the Command of the Dux Britanniarum any more likely, but the pattern on which that interpretation is based has been strengthened, rather than weakened, by the new archaeological data, whilst the evidence also hints at a similar reoccupation with regard to the signal stations of the Yorkshire coast and their headquarters at Malton.
Perhaps, then, at last one is able to see answers to many of the most pressing questions regarding what happened in north Britain, and more specifically on Hadrian’s Wall, in the fifth and sixth centuries…
The answer to all of these questions may lie in the rise and fall of a reconstructed Late Roman military command, unique in Britain, which was organized in a sub-Roman fashion reliant upon the loyal warbands of warrior aristocrats (and Anglo-Saxon mercenaries) rather than paid regular soldiers. The organizing authority of this system, probably a king of the sub-Roman Brigantes, assigned a politico-military role to the defended homesteads of these elites, and (as in the location of churches at disused forts, through land-grants?) positioned these at what had been Roman fort sites, but which were (at least substantially) deserted by the time when they were reused in this way. Thus, the ‘Late Roman’ Wall communities dispersed during the first half of the fifth century, but the Wall – andperhaps the north generally – was redefended inthe later fifth and early-mid sixth century onvery different lines, yet not completely without regard for the Late Roman past.”
I would add only that it is my belief this ‘king’ of the sub-Roman Brigantes whom Dr. Dark proposes was none other than the dux bellorum Arthur.
Here is yet another summary of sub-Roman findings at the Wall, prepared by the man who was the director of the archaeological group conducting the Stanwix Primary School dig, Dr. Mike McCarthy. The selection may be found in his book ROMAN CARLISLE AND THE LANDS OF THE SOLWAY:
"At Stanwix, Carlisle, little of the fort, the largest on Hadrian's Wall, has been investigated under modern conditions, and it is certain that much will have been destroyed. Excavations in the school playground, however, have provided tantalising hints that activities continued [past the Roman period], with the discovery of at least two phases of buildings represented by substantial post-pits cutting through earlier Roman deposits...
To summarize, modern investigations at several forts have yielded evidence for sub-Roman activity in key buildings. They include the granaries at Birdoswald, the commanding officers' houses at South Shields and Vindolanda, the bath-house at Binchester and the headquarters building at Carlisle. The conclusion one might draw is that important buildings in important locations (forts) continud to have a function at the point where the old-style Roman military command structure no longer had any real force, and the pay chests needed for the soldiery had ceased to arrive; and we can see this at Carlisle where the barracks fell from use. Nevertheless, the continued use of formerly key buildings, as we can see in several forts, might allow us to infer that this is an element in the archaeology of lordship. If so, it is lordship in transition from a Roman command structure to one of sub-Roman leaders emerging as local chiefs or kings with military titles and authority derived from that of the late fourth century. They doubtless formed small private armies or warbands, and established territoria which could supply their provisions and over which they exercised a quasi-leadership role. They were not yet kings or princes, but neither were they members of the Roman army linked into a wide-ranging command structure. Their authority was derived from the former prestige attached to the place, and their dwelings may, as is hinted in the late phases of the Commanding Officer's house at South Shields, be large and imposing, as the central range location of their buildings at Carlisle and Birdoswald may also suggest."
Dr. Frank Giecco, who was also involved with the same excavation at Stanwix, has informed me that,
“Dating is very hard, but a 5th century date seems likely if I had to choose based on evidence. Stanwix had very large stone post pads. A similar building is recorded at Birdoswald. The Stanwix structure was built over the old Roman barracks.”
