The absence of an English settlement here in the late fifth/early sixth century would suggest a strong military showing which could have resulted in the fame of Arthur among the Gododdin. They would have been going to fight in his old stamping grounds. Of the battle sites, Binchester (Castellum Guinnion) is in the Brigantian area. So is the River Glen known to Bede. The Caledonian Forest and the City of the Legion (if Chester) are within reach of a Brigantian king. The lack of surviving Brigantian material could account for the lack of background on Arthur.
To counter this plausible suggestion, it is not quite clear what he would be doing in the South Welsh or Kentish zones. Unlike the kings of Britannia Prima, clearly benefiting from unity against a common foe, it is hard to see why, apart from sheer patriotism, the King of Brigantia would have made common cause with the southern civitates. His enemies, sea-borne or Humberside Angles, were not the Mercians or Saxons fighting in the south.
The evidence collated by Snyder, and Dark’s own analysis, points to a different source of British resistance in the area – the command of the Dux Britanniarum spanning the civitates of Carvetii and Brigantes. Before looking at the implications of this, we must first consider if the focus on civitas kings has excluded different types of ruling authority in the fifth and sixth centuries.
Britain Has Governors
Dark identifies two different types of rule at the civitas level. He argues forcibly (Dark 2000) that the western tribal rulers are specifically the kings, and that the bureaucratic Roman rulers of the eastern civitates are not, although one, Gwent, actually survived to become a British ‘kingdom’.
I do not believe that Gildas and other contemporary evidence support such specific use of terminology. If any distinction is made by Gildas between western kings and eastern bureaucrats, then a good place to start is with the word rector. Gildas states that Britain has rectores, if not too many of them, at least no fewer than it needs. On first impression, he means there are simply many rulers. Dark, on the other hand, suggests that he is making a distinction between (Roman bureaucratic) rectores and (Celtic British) reges.
This is not correct. Gildas does not contrast rectores and reges. He writes Habet Britannia rectores, habet speculatores ‘– Britain has governors, it has intelligence officers’ – that is, people whose job it is to find out sin and to correct it, the two tasks he now takes on himself. The actual rank of the rectores is not clear. They may include bishops as well as secular rulers, for instance.
Technically, rectors were the civilian governors of the provinces into which Roman Britain was divided. However, there is good reason to suppose that, by the end of the Roman rule, this term was no longer confined to this official usage. Even during the Roman period, the title could be used loosely. One rector in Cirencester called himself both ‘Praeses of Britannia Prima’ and ‘Primae Provinciae Rector’ on his memorial column, although formally these terms were not synonyms (Snyder 1998).
The writer Ammianus used rector for emperors, provincial governors, military officers and barbarian client kings (Snyder 1998). When Gildas says that Britain has, if not too many rectores at least no fewer than it needs, it seems unlikely that he means it has almost too many official Roman provincial governors; he seems to mean a superfluity of feuding petty kings, as he shows later. One of the Gododdin heroes, Tutvwlch, is specifically given the title rector (YG B36) along with the epithets helmsman, rampart and citadel. He is unlikely to be the official holder of a Roman civilian governorship! Nevertheless, there does remain the possibility that rector means something specific to Gildas.
If a rex/rector dichotomy is not supportable, it is still possible that Gildas’s language permits that interpretation that different types of government, royal and bureaucratic, coexisted in Britain.
When Gildas does write of different secular rulers, he contrasts Reges habet Britannia, sed tyrannos; iudices habet, sed impios: ‘Britain has kings, but they are tyrants; judges, but they are impious.’ This presents a better case for two different types of rule. It is a contrast Gildas knew from the Bible, where the judges of Israel are non-royal rulers, distinct from the kings.
It is generally assumed that reges and iudices are synonyms used for rhetorical effect. When Gildas denounces the priests, in similar formulae, he calls them sacerdotes, ministri, clerici and pastores, all synonyms for ‘priest’. This is the simplest reading of the reges/iudices passage. However, for those arguing for different types of rule, there may be some support in this seeming contrast.
First and most simply we should consider whether they are actually judges, in the literal sense. Gildas tells us that the rulers in general go into the seat of judgement but rarely seek out the rules of right justice (DEB 27), that is, that both reges and iudices have judicial functions.
On Dark’s analysis, it could be that the reges are the actual ‘kings’ of the west and the iudices are the Roman bureaucrats of the east. This argument is based on an academic prejudice against the idea of sub-Roman kings. In Civitas to Kingdom, Dark simply used ‘kingdom’ as a blanket term for the sub-Roman polities, but by his later works he has become polemical in asserting that the only reges were the westerners. All we can add is that Vortiporius, called ‘tyrant’ by Gildas, calls himself Protector (strictly, imperial bodyguard) on his tombstone. This demonstrates that contemporaries of Gildas could be called kings or any convenient euphemism, without altering the nature of their power. It is even possible that Gildas uses Rex and Iudex in exactly the opposite way. He calls the legitimate Roman emperors ‘kings’, as he does the Roman usurpers who rule after Magnus Maximus (‘kings were anointed’). Ammianus and the Bible both use Iudex as a title for a non-Roman ruler of kingly status (Dumville 1990), which might imply that they, not the kings, are the tribal rulers. This suggestion is further supported by the use of the title Ut Eidyn (Iudex of Edinburgh) among the tribal Gododdin.
Another possibility is that the iudices rank below the kings, the normal relationship between kings and judges. They could be sub-kings or officials within the same kingdom. On the other hand, it is even possible that they are of higher status! The Roman law code, Codex Iustinianus, is more or less contemporary with Gildas. Justinian, the Roman emperor responsible for its compilation, was in contact with the western states, including Britain. He specifies precisely ‘iudices . . . hoc est provinciarum rectores’ – ‘Iudices, that is provincial governors.’
The named ‘tyrants’ are not all said to be kings. Gildas does not say that Maglocunus is just the greatest king. His kingdom and height set him apart from cunctis paene Brittanniae ducibus – ‘nearly all the leaders of Britain’. These could include both royal and non-royal rulers. Only Vortiporius is said to be a hereditary monarch of a civitas/kingdom. Constantine comes from Dumnonia and we might guess he is the ruler of the civitas/kingdom. Aurelius Caninus, however, is not said to be a king. There is no need to slide him westwards into Glywysing as Dark does. He might equally be the military commander or leader of the bureaucratic rulers of the Dobunni (for argument’s sake) as a tribal ‘king’. Gildas calls all the tyrants kings when he goes on to deliver his prophetic analogies, yet another indication that, if a distinction exists, it is not one Gildas acknowledges terminologically. All this confirms my view that Gildas and other sixth-century Britons used a variety of titles and descriptions for their rulers, without any systematic logic we can detect.
Dark is reluctant to speculate that any forms of authority operate in sub-Roman Britain above the civitas level. Gildas, however, provides evidence that higher levels of rule must exist. The roles of Vortigern and Ambrosius are not explicable if they are simply civitas kings.
First we must reconsider Gildas’s categorical statement ‘habet Britannia rectores’ – that Britain has rectores (provincial governors) – in his own time. Hitherto, we have dismissed this as yet another synonym for the Duces, Reges, Iudices and Tyranni who rule Britain. Although we may balk at taking him so literally, is it possible that he means us to u
nderstand them as different or higher than the kings? The evidence is that ‘Roman provincial governor’ is the only sense in which Gildas uses Rector. He writes of Rectores twice more: rebellious Britons during the Boudiccan revolt butcher the Roman Rectores and Magnus Maximus despoils Britain of her Rectores during his usurpation. So there is a good case for seeing this as meaning something particular. In both cases, low-level rulers, of a British tribal nature, are obviously excluded. They have been taking part in the rebellion in the first case. In the second, usurping and cruel kings will be ruling Britain immediately after. If we only had these two passages, there would be no question but that Gildas was using the word in its technical Roman sense. Only the prejudice that he ‘cannot’ be using it in this sense in his introduction prevents it being understood in this way.
One of the features of the good generation before Gildas’s time is that ‘kings, officials and ordinary citizens served in their allotted roles’ or ‘kept to their appointed station’. Does Gildas mean they were subject to the legitimate governor of the province? The publici could be the officials of non-royal bureaucratic government, again contrasted with the kings.
One indication that, in some sense, provinces continued beyond the end of Roman rule is that the four met different fates in the fifth and sixth centuries. This is demonstrable by the patterns of Saxon settlement. Maxima Caesarienses was settled along the lines of the former civitates, with concentrations in strategic areas. The civitas name of Kent survived. This indicates settlement while the Roman-style administration was still functioning. Flavia Caesariensis had settlements of Saxons with little regard to the Roman patterns, as though carried out in a disorganised free-for-all. The two Britannias were hardly settled at all. Britannia Prima would come to be dominated by warlike kings while Britannia Secunda saw Roman static defences re-employed. It is likely these differences arose from different provincial responses to the crises of the mid-fifth century.
There are basically two models for how provincial authority could still exist. If it is independent from the civitates, then some degree of consent must exist. The civitates would have to provide food, supplies and manpower to the administration, which had no hinterland for itself. Perhaps more plausible is that the provincial ruler is also one of the civitas rulers, using his local power-base to achieve hegemony. This is the way the Saxon Bretwaldas derived their power. In this case, the King of the Dobunni (for instance) might exercise authority over the other civitates in his province.
Another indication that the concept of the Province of Britannia Prima still existed in Gildas’s day is that, as far as we can tell, all the tyrants he denounces come from that area, as Higham points out. We might even suggest that the survival of the name Britannia for the whole island was influenced by the two provinces where sub-Roman/British rule survived being those called Britannia.
Could Arthur be one of these provincial governors, of Britannia Prima perhaps, in the previous generation? Is that the reason he is distinguished as fighting with ‘the Kings of the Britons’? If they are the rulers of civitas kingdoms and he is not, this would make sense of the apparent distinction made by Nennius.
Alternatively, the governor of Britannia Prima might be the coordinating authority, employing Arthur as his Magister Militum to lead the British rulers against the Saxons. This does not preclude any of the other suggestions. Arthur could be a sub-king, king, non-king or non-Briton to occupy this role. He could be chosen by the provincial governor or, as happened elsewhere in the west, he could have dominated the provincial government by virtue of his military power. Commitment to the province as a whole could explain why even Gildas’s tyrants chase thieves energetically all over the country. If Arthur acted in a similar way, it could explain his presence in Ercing and Builth. They would be in different kingdoms but parts of a single province.
Another possibility is that the kingdoms of Britannia Prima are linked by dynastic alliances in the previous generation, explaining why three of the kingdoms seem to have been ruled by lion’s whelps. In Gildas’s time, hegemony over the province might have been imposed by coercive power. Maglocunus, Dragon of the Island, is able to depose or kill other tyrants. In the Badon generation, however, the import of Gildas’s approval of the kings, officials and private persons who kept to their stations is that they consented to a legitimate overlord.
An indication that some kind of provincial responsibility lingered in the minds of the northern successors is the fact that the northern forts were maintained. These included locations already discussed: Cataractonum (Catraeth of the Gododdin), Vinovium (Castellum Guinnion?) and Camboglanna (Castlesteads). These were occupied and fortified by some kind of successor authority.
The forts ran across the territories of the Kings of the Carvetii and Brigantes. The kings must have cooperated with a revived Roman office for mutual defence, possibly a governor of Britannia Secunda, or a new Dux Britanniarum. If Arthur held one of these positions, then we might imagine him fighting at a civil war centred on forts like Camboglanna.
The official responsible for the forts, according to the Notitia Dignitatum, was the Dux Britanniarum. This role, as we have noted, has often been attributed to Arthur by modern historians. There is no sign of the forts having a single administrative centre or a hinterland as distinct from the civitates in which they were found. The Dux must therefore have acted in cooperation with the kings of the Britons. A hypothetical Arthur as Dux Britanniarum would have defended the Wall and adjacent areas against attacks from the north and maybe the south as well.
The Dux Britanniarum controlled the frontier troops stationed in the region of Hadrian’s Wall. By the fifth century, these border troops were often hereditary, making it far more likely that they had been left behind by usurpers like Maximus when they used British forces in their struggles for power. However, all sources’ view of the end of Roman Britain hinge on the idea that, without Roman garrisons, the Walls cannot repel the invading Picts and Scots. It is the clear import of the Honorian rescript, and Zosimus’s account of the Britons’ own response to the barbarians, that there is no major surviving Roman military presence in the island. The whole rationale for the Saxon settlement, to fight the Picts and the Scots, is that there were no continuing garrisons in the Wall area. Dark provides an answer to this conundrum: his analysis of the Dux Britanniarum forts shows that, while they are not continually occupied from the period of Roman rule, they are reoccupied and fortified during the Badon generation, perhaps when a new Dux was set up (Dark 2000).
An invigorated British resistance, combined with a sense of responsibility for the Province of Britannia Secunda or the Diocese, might have seen the British appoint their own Dux Britanniarum. Publici (state officials) served in their correct ranks, as Gildas says. If Nennius is actually making a point about the different nature of Arthur’s power as Dux, against the kings (Roman-derived and military in nature – as opposed to the civil, taxing and quasi-judicial role of a king) then his description does make a lot of sense. Arthur would be appointed by British kings acting responsibly for the good of the wider region. He is fighting with the authority of the kings of the Britons, but he was himself the Dux Bellorum.
It is hard, unfortunately, to see how Arthur’s role can be limited in this way. There is no obvious reason for the frontier-defending Dux Britanniarum to be associated with wars in South Wales or the Chilterns. The only model which explains this is that the whole military structure, with both Dux and Comes Britanniarum, had been re-established. Arthur could even, in this theory, have been the Comes Britanniarum, his northern victories attributable to his support for the Dux. This has brought us back to that old theory of Arthur the cavalry leader, by other means! The existence of the Comes would imply other levels of authority. Even more than the Dux, the Comes required extensive civil support. He had no territorial hinterland to supply his troops, no fortified bases and no widespread supply depots. For a military command of this nature, there would have to be a supportive
civil authority at provincial level.
Furthermore, if the Dux Britanniarum were supported by a mobile reserve, there would have to be a coordinating Magister Militum. It is almost impossible to imagine any organised relief for static garrisons which would not rely on a single strategic commander. Arthur as Magister Militum across the British provinces fits the evidence very well. If we look at the northern archaeological evidence from another perspective, we might suggest that, rather than being the Dux Britanniarum, Arthur was the man who drove out the Saxons and re-established the frontier command to consolidate his victories. We know Saxons must have been in the Wall zone precisely because they were Vortigern’s defence against the Picts, as Gildas tells us. Arthur as Magister Militum, fighting with the kings of the Britons (presumably the Brigantes, Carventii and perhaps the Gododdin), who provided troops and supplies, could have driven out the Saxon frontier troops, re-established control of bases such as Castellum Guinnion and hunted the fugitives into the Caledonian Forest. British garrisons could then have been re-established under a subordinate official, with guarantees from the kings to maintain them. Peace treaties guaranteed by religious sanctions are a feature of the next generation and we can imagine such an agreement to uphold the system. Small wonder that the Gododdin, coming to support the southern Britons by driving Saxons from one of the region’s forts, could draw an analogy with Arthur and his battles.
The idea that Arthur was a Magister Militum, perhaps supporting the command of the Dux Britanniarum in the north and the Comes Britanniarum in the south, for successor provincial authorities makes a lot of sense of the evidence. It explains that his military power was different in nature from ‘the kings of the Britons’ and why he could be found fighting across the Diocese. It may be that Arthur switched voluntarily from one province to another, was wooed by greater rewards or was deployed in different provinces by a residual diocese-wide authority.
The Reign of Arthur Page 16
