The Arthurian Collection is a part of Vortigern Studies

The removal of the incubus of taking Zosimus 6.5.3 literally has important consequences for culture history. It rids us of something which besets and distorts the work of dozens of brilliant historians - the notion that any sort of British anti-Roman nationalism existed as early as 410. Without it, there is no evidence for any nationalistic - in the sense of anti-Roman - party. Constantine III had no desire to break away from the Roman world; he wanted to rule it.  Subsequent British monarchs regarded him as the founder of their line; but this only implies a certain continuity, however crooked, from the throne occupied by the usurper in 407 to that which legend claims for Arthur - and that only means that, in all this period, there was a recognized high king of Britain who was understood as the successor of previous high kings or emperors.  In the 430s, Prosper saw Britain as a "Roman island".  The obsession with Roman law and procedure which we have found plenty of reason to suspect in pre-Saxon Roman Britain sends the same message, and so does St.Patrick's furious refusal to recognize Coroticus as a Roman citizen.  Even Pelagius can be read as the defender of an ancient Roman concept of virtue and goodness against the Christian notion of dependency on God: not even so much Greek, as specifically Roman, in the mould of Cicero and the Catones^[1]. There was no nationalistic element in Britain, no element unwilling to call itself, as sincerely as the Iberian or Anatolian peasantry, Roman.

A century and a half later - and in spite of his formal education and the Ciceronian shape of his masterpiece - nobody could honestly say that Saint Gildas lives in a Roman world of ideas. The decline of Latin secular culture is not more marked in him than in Gregory of Tours, descended from an ancient Roman family and ignorant of almost any literature except the sacred kind; clearly large areas of the Latin West were neglecting it for a new ecclesiastical brand of learning.  But compared with his fellow literary Saints, Gregory of Tours or Gregory the Great, Gildas is notable not so much for the decline of Roman learning, as for the existence of a completely autonomous social ideology based on Celtic ideas of kingship and royal power.  I have called this "resurgent Celticism", a Celticism leavened with Christianity (the legends of A took shape in a fully Christian environment, with no perceptible pagan element) and haunted by the memory of Rome, but essentially alien.  In A, Rome is simply and purely something great in the past, a mighty people who ruled this island long ago, though the greatest British families are of their blood.  The language, the landscape, the inscriptions, still existed, to remind the British of this race of giants; but the empire had passed.  It is even possible to read in the idea of the Romans as a mighty, remote race, making sudden overwhelming expeditions but not staying, a sub-conscious reflection of their distance from the North British, late-fifth-century minds in which the legend formed - real - awesome - but remote, until suddenly brought to mind by some giant memory in the landscape, a road, a villa, a Wall - absence, distance, punctuated by sudden flashes of intense memory and presence.

Resurgent Celticism is seen not only in the literary and political culture of Gildas and his contemporaries, but in such thing as the oft-remarked re-occupation of hill forts deserted since the age of Claudius, not only as military strongholds, but as the poles of the resurrection of a genuine redistributive exchange system that owed nothing to Roman economic practice^[2].  Contemporaries, even where they did not understand what had happened, recognized this new distance between the former Romanam insulam and anything now recognizable as Roman.  Procopius saw British ambassadors as "barbarians", that is definitely not Roman, classing them with Huns and the like.  All that Procopius knew of Britain was what he found in books and the preposterous fables he naively accepted from a Frankish embassy; but the description of British embassies refers to something he, or other members of the Byzantine government, had seen in person.  The British ambassadors they met left them with an impression of alienness no different from that of Slavs or Avars, even though they must surely have come equipped with some Greek and a Latin as good as Gildas'.  The cultural distance of Britain from what was still the heartland of Roman civilization was by now clear and visible, if not to the British, then certainly to the Romans.

Besides - not to the British?  Our only intact major British document, St.Gildas', shows awareness of it with every period, and never stops stressing how far the British are from the Romans.  In Patrick, the distinction between ciues meorum and ciues sanctorum Romanorum is not very sharp; it seems more a matter of shading off from the good (common Western and Christian identity) to the better (Rome’s greatness and the sanctity of the centre of Catholic Christianity) and away from the worse (belonging with the likes of  Coroticus).  But in Gildas, the distinction is absolute.  Ambrosius was almost the last of the Romans, and in all of Britain there were only a few sturdy reliquiae; but there were plenty of ciues - a complementary and separate group.  Whatever Gildas' view of real live Romans, he was quite clear that a great difference existed.  It was visible to anyone who traded with Byzantine ships, who met Greeks or cultured western Romans in any sort of social context, or who managed the long (but hardly impossible) journey to the golden city on the Bosphorus^[3].  At some point in the previous seventy years or so, Britain outside the Saxon areas had undergone a "genetic mutation", from thoroughly and indeed obsessively Roman outpost, to Celtic-Christian "barbarian" culture.

Our one Greek witness, Zosimus, attests - from what I believe to have been a British document, L - that there had been an actual resolution to throw over Roman law wholesale and adopt what he describes as native customs.  I have no doubt that he records - at whatever remove - a historical fact.  It agrees with another fundamental lost document, A, in which Roman ius is described as an unbearable iugum, and the British only achieve their final and desperately needed victory when they adopt, instead, the “easy” yoke of God.  A has no description of a clear political change such as L contained, but foreshadows that change through the ineffectiveness of Roman arms against the Picts and the ideological short-circuit that transfers to God the allegiance previously due to the kings of Rome.

There is no reason to date the writing of A (as opposed to the preservation and development of the historical memories which seem to be its core) very early.  A is to be dated long enough after the end of Roman power to have forgotten the permanent Roman presence on the Wall; and its description of the third Pictish war in A seems influenced by what we can discover of Ambrosius' war against the Saxons. Echoes are in my view to be found in the picture of the Picts, stunningly defeated once in the remote past, yet (by implication) present, active and threatening in the here-and-now: a picture that reminds us of Gildas’ account of Ambrosius’ overwhelming victory and of the Saxons’ continued presence in Britain and of their later resumption of hostilities.  Indeed, the two wars are quite parallel: starting in poverty and chaos, involving the gathering of an army out of the remote scattered strongholds of northern clans and a stunningly successful and apparently brief campaign against previously conquering barbarian enemies, in a country troubled by hunger, but ending - with the Picts as with the Saxons - with no final expulsion of the barbarians from "our" island.  The Saxons are never mentioned, but it seems quite clear that the Pictish war provides a template for the development of the war against them.  And this suggests that A comes not only after the Saxon war, but after the end of Ambrosius’ settlement and the resumption of hostilities.  This is at any rate highly credible in terms of time: if the correspondences we suggested, between A and real late-Roman British history on the Borders, are really there, it must follow that the memory of these events was no longer living, that is that it reached its pre-A form at a minimum of 60 or 70 years after the final victory over the Picts.

A therefore comes at least two or three generations after 407, but early enough before Gildas (561) for him to regard its views as old-fashioned and to modify it in the light of evidence from the Continent; we would not be greatly wrong if we dated it at about 480-530, and that is the date we must provisionally assume for its outburst of British/Celtic, anti-Roman nationalism.

This warns us not to look to what seemed the obvious fault-line, the great Saxon war, which, Gildas claims, inflicted a decisive blow to Roman civilization, driving thousands of refugees across the seas along with a great deal of written scriptorum monumenta.  It is certain that the Roman material culture collapsed in Britain between the fifth and the sixth centuries, with a thoroughness unseen elsewhere in the Roman West: the art of building in stone and the industry of mass-produced pottery, among other things, all but vanish from the picture.  Wood building, and plants much simpler than Roman architecture had been used to, become common, and fortifications in murus gallicus, unseen in Roman Britain since the conquest, reappear.  But this does not suggest a barbarian conquest however ferocious.  Barbarians may have stormed the stone walls of Roman cities, but they then proceeded to settle in them.  And, in general, a swift conquest makes less long-term damage than a long, grinding war.  A barbarian conqueror may shed a lot of blood, but what he wants is to take over the wealth and way of life of his civilized captives.  He strides into their stone buildings, and looks for architects and masons to repair and enlarge them; he looks for, and patronizes, craftsmen and artists to decorate them.  It is not barbarian invasions from outside, but rather prolonged, grinding, vicious wars over a period of decades, that destroy civilizations; in the rest of the Empire, it was not the barbarian invasions, but the Justinianic cycle of wars, that marked the real collapse of Roman culture.

But in the case of fifth-century Britain, another consideration applies.  Which group was more barbarian?  The nature of things suggests that Ambrosius recruited the armies of his war of liberation in the Celtic North.  For more than three centuries - as much time as separates us from Cromwell - the tribes around the Wall had had a close and complex relationship with the Roman empire, while retaining an essentially Celtic culture.  While their most direct bonds were surely with the Army, that empire-wide instrument of power, whose officers (especially the decision-making highest ranks) were apt to come and go every few years, cannot have developed links as solid as those with the Romanized society of the lowlands.  We must imagine connections as close - if perhaps as unfriendly - as those of the great, semi-independent lords of the mediaeval Borders with England, or the Gaelic clans of Scotland with the Scots-English-speaking lowlands.  I do not think it is a coincidence that Paulus Catena's odious butchery of British nobles was followed not only by the collapse of Roman civil order, exploited by Valentinus, but also by the carefully laid treachery, amounting to a large-scale revolt, of the local advance scouts, who reach agreement with the Picts and take over a whole province, whose administrative structure Theodosius the elder has to rebuild from top to bottom: what was done in the lowlands clearly reflected on the highlanders' mood.

It seems therefore likely enough that when Ambrosius began his campaign of re-conquest, he can have counted on ancient family and social links with the tribes of the North.  If I am correct in my reading of the words Vor-tigernos, Vortamo-rix, Emyr and Ameraudur, then his party aimed a propaganda campaign specifically at the Celtic-speaking North.  Where else could he have found an army?  To the Saxons, disarming and claiming the defence of the rich tributary south as their own business would be a natural precaution.  Only where the untameable North abutted on the hostile and dangerous Picts can we imagine British tribes still in possession of weapons and a military structure.

This is not evidence, merely plausible conjecture.  The evidence of Gildasian culture, however, proves that resurgent Celticism was at its core.  And every feature found in our research, except for what is very general, points to the tribes between the Wall and the Picts - present Strathclyde, Lothian, Dumfries/Galloway, and Borders - rather than any other Celtic area.  We have often seen reason to suggest a northern origin for some Gildasian-age cultural artefact or other: think for instance of A or of the continuity of Ninian's see in Whithorn, with its hostility to Patrick's Irish diocese and its abiding claim to both Pictish and Irish missions.  We cannot avoid a strong sense that the cultural roots of Gildas’ Britain lie in the North.

It is not a matter of individual cultural features or fashions, but of the structure and cohesion of a whole culture.  In spite of his admiration for all things Roman, the Celticism of Gildas' masterpiece is consistent.  It extends to all its major social assumptions - such as the centrality of kings, the redistributive exchange system, the impotence of the free classes and their dependency on the sacred and military aristocracies, the importance of the king's comitatus and its military nature, the influence of bards, the political centrality of public banquets, even two or three hints of the Celtic marriage of king and land.  On the other hand, all the features which go back to real Roman ideas - such as Gildas' knowledge of classical agricultural terms, and of Roman military values - are clearly out of books.  At one point, it becomes clear that Gildas was copying - because he made a mistake: in describing a well-arranged army ready for battle, he spoke of the right wing (dextrum cornu) but not of the left (6.2).  What bizarre lopsided disposition is that?  Either he is repeating, without understanding, an account of the peculiar dispositions for a particular battle in which no left wing was needed; or else, well, he has made a copying error.  Clearly, he was not living in Roman institutions.  Where his books fail him, he is helpless: he does not know the right title for Aetius, contenting himself with the vague uirum Romanae potestatis, a [brave] man [endowed with] power from Rome, even though the document itself he was excerpting gave it to him - ter consuli, a consul three times over.  He clearly had no idea what a consul was.

Such a change as this resurgent Celticism does not happen by seepage: what it documents is not the adoption of fads and fashions, that leave the basis of political power largely untouched, but a major change in the cultural assumptions that shape politics.  It is consistent and wide-ranging, altering the very nature of political power - like the French Revolution.  The picture of politics in Gildas is not - as for instance in that mysterious artefact called the British Constitution, which, like God, none have seen but all must worship - the result of several layers of largely unconscious change in the name of dozens of contradictory theories, ending up with a law that is internally inconsistent and hedged about with prohibitions that, no longer having any reason to exist, amount to pointless taboos; it is a massive, solid, consistent, elaborate fact.  We are even given, in A, its analytical view of political power (in narrative form).  Therefore that very passage of Zosimus which gave historians so much trouble when they - following Zosimus himself - misdated the events it describes to 410, becomes centrally important when we realize that it is quoting from a much later, Gildasian-age document, L, the same to which Gildas is reacting throughout his masterpiece: because it is the only document we have to describe, however indirectly, this revolution.

If Zosimus, however, is the only witness to the changeover, St.Patrick is our best background reading on the roots of the process.  Patrick was doubly marginal: first, as an undereducated Roman Christian speaking what seems like a Latin dialect, and second, as someone with far too much experience of a non-Roman society.  His mind lives in two worlds.  He is Roman enough to have a quite Roman idea of culture and its social significance, even though he knows he is short of it; but the background of his clash with Coroticus is in a set of ideas quite alien to Rome even in her decline.

Patrick's description of Coroticus sounds exactly like an anticipation of later times.  It is a picture so familiar to us - backwards into the Celtic past, and forwards into the Irish, Welsh and North British future - that we can hardly remember how thoroughly untypical of anything Roman, even more late-Roman, it is.  His quite different relationship with culture and learning - especially their practical, political dimension - I have covered in Book 4, ch.4; but there is more, which we are in danger of not recognizing unless we date, as we must, Patrick in a time of a still full-blown Roman culture implying Roman, State-centred social relationships.

Patrick himself, who had been living among the barbarians for years, knew their ways.  He understood, even as he hated, what Coroticus did; and he summed up the nature of his opponent's ambitions under two headings: wealth, and royal power.  Coroticus indubitably raided Ireland, and butchered Patrick's converts, to feed what seems to have been a regular trade in slaves with the Picts, part of a larger commercial network which involved several pagan Pictish and Irish (“Scottish”) tribes, from which Coroticus expected gold enough to feed an upper-class lifestyle that included impressive banquets to which Christian ecclesiastics, perhaps bishops - the "holy and humble of heart" whom Patrick tries to warn away - might be invited.  Coroticus had an interest in leaving the Irish unbaptized; as long as they were not part of Christian mankind, he could raid them at leisure, and ship hordes of those often remarkably handsome men and beautiful women to the slave markets - for the Irish are a handsome race, and no doubt a slave-dealer's delight.

But the political dimension, which is clear from the presence of such high ecclesiastics, also shows the other aspect of his activities: Coroticus expects a considerable increase in his political power from these raiding expeditions.  It is not suggested that he aims to annexate or colonize any territory; rather, Patrick seems to be saying that the fact that he takes wealth away from the raided territories ipso facto makes him a greater lord.  He and his men mulierculas baptizatas praemia distribuunt ob miserum regnum temporale[m], quod utique in momento transeat, “distribute poor little baptized women as prizes for the sake of a miserable temporal kingdom which will certainly (utique) vanish away in a moment”.

I would draw attention to that ob miserum regnum temporale.  Patrick does not say that the abduction and sale of the women will help make Coroticus richer and therefore more influential; he does not say that it will reinforce his position; nay, the very fact that he has abducted and sold the women seems to be creating the royal power he desires.  Taliesin praised the king of Powys, whose cattle was never seen in another's pond; Coroticus proves himself powerful by corralling not only other kings' cattle, but their women.  But the miserable shadow of a kingdom by this atrocity will vanish “in a moment”, leaving him as “ashes” under the feet of those he had murdered to achieve his ends.

In a still Roman world, in which conquest meant territorial settlement and extension of jurisdiction, this is an altogether alien concept.  Even the traitor Arvandus, theorizing the end of ius Romanum and the rise of ius gentium in its place in Gaul, saw it in terms of the Visigoths and Burgundian acquiring new territories, and in no other light^[4].  We are in the world of Celtic values examined in previous chapters, in which royal power is enforced not by conquest or colonization, but by tribute, and, where that is not conceded, by plundering raids, so that a successful king is one whose cattle “was never seen in another’s pond”.  Patrick - who probably knew enough to judge - was quite right: Coroticus, whether or not legally a Romano-British citizen, was mentally and spiritually a northern barbarian, a full-blooded Celt, a “fellow-citizen of the demons… [because he] live[s] in the ways of the enemy, [a friend] of the Scots and the apostate Picts”.

Once we understand this, Patrick’s great climax gains another layer of meaning, simple, powerful, and devastatingly impressive.  As he had confiscated the notion of sapientia from the learned gentlemen of the Britains on his own behalf, declaring himself sapiens by God's grace, so he confiscates Coroticus’ ambitions on behalf of his victims, conferring on them the royal majesty in whose pursuit he had killed them.  The vision he summons up to celebrate his dead converts and execrate Coroticus is dense with images of victory, power, rule, royalty: Aeterna regna capietis… conculcabitis iniquos, et erunt cinis sub pedibus uestris… regnabitis cum apostolis et prophetis et martyribus… iudicabunt nationes, et regibus iniquis dominabuntur… "You shall conquer the everlasting kingdoms… You shall tread the wicked down [like a defeated army], and they shall be ashes under your feet… You will reign with the apostles, and the prophets, and the martyrs… They will judge the nations, and will be masters over wicked kings…”^[5]

In Coroticus, the opposite term of the same vision, are gathered the images of looter-pirate-rapist-slaver, and of ambitious politician-king.  He and his fighters are rebellatores Christi, rebels – in the political sense of the word – against their true king, Christ; and Patrick uses vocabulary for them that fits exactly the late-Roman concept of the usurper and the rebel, calling Coroticus' men at once milites - soldiers - and latrunculi - petty thieves.  This was the standard Roman perception of usurpers and their armies^[6], whose rebellion had placed them with organized criminal bands, enemies inside rather than outside the empire.  In fact, once a Roman army had lost its contact with the national ius and state organization, it was hard to see the difference from a particularly large bandit gang; and conversely, antiquity is not without instances of outright bandit bands growing powerful enough to conquer kingdoms.

In the case of Patrick, what this kind of insult tells us is that Patrick saw Coroticus and his men as in some sense regular Roman soldiers^[7] (milites) in revolt against their true king, Christ; that is, in his rather idealized vision of the Roman Empire, it was God who was its true Ruler (it is against this idealized Empire of God that Patrick raises the spectre of an alternative citizenship, ciuis daemoniorum, making the point that religious allegiance and citizenship go together), and soldiers who break the code of religion are the same as pretenders and rebels.  At the same time, Patrick does not doubt that Coroticus is a king as well as an army commander: as Snyder observes, "Patrick asks that this letter be read 'in front of all the people (plebibus) and in the presence of Coroticus himself'.  His language, evoking Old Testament imagery, seems to elevate Coroticus to royal status, where his actions, especially, will be held accountable to God", his true Overlord.

In other words, Coroticus is already a king, but within a hierarchical system in which the British emanation of the Roman empire is above him, and God above all.  His sin is to try and gain even more kingship.  These are Celtic, not Roman, ideas: the hierarchy of monarchies - a living reality, for instance, to the writers of Irish law codes - will be sought for in vain in Roman law^[8], and the excessive ambition of otherwise regularly appointed kings is, exactly, Gildas' bugbear.

Patrick does not trouble to explain these ideas to non-Celts.  This might mean either that he expected those Romano-British church leaders he addressed to understand them, even though they certainly had nothing to do with the Roman view of political power; or that the was so used to them that it did not occur to him that they would be alien to others; or that he wanted the mere description to strike them by its strangeness, underlining that this is no Roman, but a man who "lives in the evil way of the Scots and the apostate Picts".  Whatever the case, this has nothing to do with Roman politics.  It is not only Coroticus’ name that is Celtic – Caractacus, Caradoc, Ceredig, Cerdic – in a still Roman world; his ambition is not a Roman ambition, the kind that drove generals on the Roman borders and senators at the imperial court, but rather the ambition of a Celtic marcher chieftain, eager for tribute and retainers.

By the same token, his court is not a Roman court, but that of a barbarian Celtic lord; it has more in common with the Gododdin than with the household of Constantine, let alone Trajan.  Patrick envisages him displaying and extending the power won by his raid at great public banquets, which might be attended by illustrious guests such as bishops, but which at the same time feast his milites, like Urien feasting his warriors and his bards in the halls of Rheged; and Patrick’s allusion to “words of flattery” are reminiscent of Gildas’ description of the bards of Maglocunus^[9] as criminal flatterers, praising and justifying their master’s murders. These banquets are so politically important that Patrick believes that he will have struck a great blow at Coroticus if he can induce the "holy and humble of heart", the better British bishops, to take no part in them.  Their misguided presence, he seems to think, validates and encourages Coroticus’ behaviour, and one of his bitterest blows - his memorable image of Eve’s “food of death” - is aimed at them.

The difference in atmosphere with Roman civilization can be felt.  Though Roman aristocrats were not slow to feast and revel with their friends and clients, they would not envisage these revels as public confirmation of their political power; they were essentially private matters with no really ceremonial/political aspect.  Nor would it occur to them to use them to reinforce the cohesion of a private army by feasting them; they would be likelier to strike gold or silver coins and distribute them to the soldiers as gratuities.  It would not be at the banqueting table that a Roman would envisage the worst kind of political corruption, but in the private rooms of an Emperor, or in the Senate house, or in the barracks of a pretorian guard, or from the roster of a demagogue.  The whole matter just cries out its essentially barbarian, non-Roman nature.

Yet Coroticus was a Romano-British citizen, and must have felt the value of that strongly enough to smart – or so Patrick thought – when his non-Roman attitudes were thrown in his face.  If he is the Ceredig gwledig of the Strathclyde king-list (the title gwledig, by the way, signified exactly the kind of over-kingship that Patrick thought he was trying to acquire) then he was of Roman descent, the heir of one of those Roman duces remembered by A; and the fact that the claim to Roman descent lasted long enough to become part of A means that it a fortiori was important to Coroticus, who was much closer in time to the Roman origins of his dynasty.  Perhaps Patrick deliberately stressed his Celtic name to separate him from true Romans (if Coroticus was of Roman family, he will still have had a tripartite Roman name, with Coroticus as one term).  Patrick's words suggest an autonomous marcher lord with a great deal of leeway, but subject to pressures from the (Roman) rest of the island.  The bishop-elect of the Irish unmistakably demands that he be excommunicated by a superior, surely all-British, church authority, and clearly feels that such an excommunication would severely damage him.  Nothing in Patrick's treatment suggests that he is wholly independent, and Patrick's challenge that Coroticus is no fellow-citizen either of the British or of the "holy Romans" suggests that Coroticus claimed both these titles and would find it seriously damaging to be denied them.  Incidentally, one witty scholar asked "since when are the Romans 'holy'?"  Answer: since they were, to Patrick, the home and heart of Christian faith and cultural orthodoxy.  He was speaking of the Christian empire; he may not have known that it was dying, and felt its glory the more, the more he lived among barbarians^[10].

Archaeology should be asked the question, which our conclusions from a sadly thin haul of documents cannot altogether answer, how a thoroughly and indeed fanatically Romanized lowland shaded off, whether by degrees or by a clear, visible "highland line", into a highland Celtic world of kings and bards and tribes and endemic wars and vendette; but Patrick's evidence makes it clear that even as a confident, arrogant Roman Britain of dominicati rhetores, decuriones and uillulae existed apparently unchallenged this side of the Wall, the Britanniae also harboured, in parts less often heard about, not just a remnant eking a living on the edge of Roman power, but a whole social organization which, while Christian and contiguous to the Roman south, preserved undiminished, undenied, unchallenged, the whole Celtic ideology of king and land, of tribute and raids, of redistributive exchange systems based on royal fortresses.  Three hundred years of Roman presence, of frowning armed power, of Walls, had altered it in no way we can perceive^[11]; Romanization, which had been a total success in the south, had failed, or perhaps not even been attempted, in the tributary north beyond the Wall.  The Roman army, camped in its stone fortresses and pouring military stipends into the surrounding economy and certainly into subventions to friendly chiefs, had, if anything, reinforced the prestige of the chiefs themselves, now able to distribute beautiful minted Roman gold to their favourites; but had left the land as it had found it.  Except for the Roman garrisons proper, which, however far they extended, can only have covered a fraction of territory and population numbers, the locals lived in a climate and a landscape that encouraged neither Romanization nor stability; the thin little fertile valleys isolated among grim expanses of mountain and moorland, later the stage for the banditry and savage feuds of the Borders.  It is certain that Celtic institutions must have survived almost intact; and that, while Theodosius the elder restored a province and a full-scale Roman apparatus south of the Wall, here there is no evidence for anything except the continuance of native kingdoms, probably under imposed Roman rulers such as we saw reason to read in the memories embodied in the earliest layer of A.

King-lists survive for kingdoms beyond the Wall, stretching to a time before the end of Roman power^[12].  Though a number of Welsh dynasties claim descent from these, no similarly ancient dynasties are known to originate south of the Wall, and the recognizable historical figures among the other Welsh patriarchs are decidedly later - Maxen, Vortigern, Tewdrig, Brychan.

Of these, the use of Maxen in particular shows the late and artificial nature of Welsh dynastic lore (I mean lore from Wales proper, as opposed to the Old North).  Maxen was reckoned as the ancestor of the kings of Dyved, but also as the originator of the colony in Brittany.  A past generation of historians has indulged in the most elaborate attempts to find a historical grounding for this; I myself believe that - unlike the descent of the houses of Gwrtheyrnion and Powys from Vitalinus/Vortigern, or of that of Gwynedd from Maglocunus and perhaps Cunedda - this is a purely legendary claim which cannot be more ancient than the origin of the Legend of the Seven Emperors, in which Maxen is the last Roman emperor in Britain.

I cannot prove it, or at least I do not intend to try, but I feel personally sure that the character of Maximus was introduced into Welsh tradition by Gildas, who drew him from the scarce continental notices he seems to have received from correspondents.  I believe he used him to create a "British guilt" to replace the Second and Third Massacres of A, because the revolt of Maximus is not necessary to the legendary pattern of A, which is concluded with the Third Massacre, the Third Pictish Invasion, and the final British victory; in other words, his presence feels intrusive.  Gildas, I believe, wanted to be rid of the Massacres, because it was not part of his plan to either show British treachery as successful (getting rid of the Romans) or to show the Romans as cowed by anything at all.  The Gildasian picture of Maximus the rebel would then come together with the later misunderstanding of Constantine III as a British king rather than a Roman pretender, and with a more general misunderstanding of the general idea of the Romano-British empire of his successor, to create a picture of Maximus as the last Roman emperor in Britain - the last Romano-British emperor - at the end of a variable but formulaic list of seven names drawn one way or another from historical records; and then reigning royal families proceeded to claim descent from this source of sovereignty.

In other words, the formation of genealogies headed by Maxen happened after Gildas (about 561) but before N (about 635), by which time the Catellids at least are already claiming Maximid ancestry.  N had read Gildas, and his attempt to clear Maximus' reputation was directly based on ideas from Gildas; but he either did not know, or he rejected, the pseudo-history of A, and therefore his use of Maximus cannot be said to have anything to do with it.  This probably relates to the novelty of royal independence in Gildas' time; as we have seen, he sees his obstreperous kinglets as a new and unwelcome phenomenon: breeding in the collapsing British body politic.  As he wrote, they probably were in search of legitimation for their growing power, and subsequent decades may have witnessed a flowering of Maximid claims.  If a sixty-year period seems too short for the formation of an entirely new political mythology, remember what I pointed out, dealing with St.Patrick, about how swiftly a complex mythology can form among a committed social group.  Pace the historians (and they include prestigious names!) who have tried to give it credence, it is wildly unlikely that the historical Magnus Clemens Maximus should have left royal descendants in Britain; on the other hand, it is wildly likely that later, indeed upstart dynasties, with none of the prestigious antiquity of the brenhin of Cunedda's line, should have claimed Maxen as forefather because he was "the last Roman emperor in Britain" (and never mind his character!), thus giving them a link to the most prestigious of all monarchies.

This bespeaks a stage of culture in which the last stages of Roman British history were well and truly forgotten, and in fact Dumville has convincingly argued that the Maxen-based pedigrees are learned Dark Age inventions in the service of local dynasties, quite alien to any real memory of Roman days^[13].  There is nothing to suggest that full-blown Gildasian Celtic kingship, of the type we have been describing, in Wales - or indeed in what is today England - had any origin earlier than the sixth century.

The legend of the Seven Emperors must have been in existence by the time Bede wrote his history.  Bede’s list of seven English Breatwealdas, high kings, is at once learned and legendary.  The seven historical kings it includes, Aelle, Ceaulin, Aethelberht, Redwald, Edwin, Oswald and Oswy, all had strong legendary overtones, and none of them came within living memory; the last of them, Oswy, died over a hundred years before Bede wrote.  That is, it has the same character as the pseudo-learned strings of seven Roman Emperors in Nennius (Caesar, Claudius, Severus, Caritius [Carausius], Constantine, Maximus, Maximianus) and Geoffrey (Claudius, Vespasian, Severus, Geta, Bassianus, Constantine and Maximianus).  And the point is that the idea comes before the list.  Both Nennius and Geoffrey speak of seven and no more than seven Roman emperors, though their lists are different.  This means that, while the idea that Britain was ruled by seven Emperors of Roman race was older than Nennius, nevertheless there was no agreed list.  Nennius knew that the duplication of Maximus and Maximianus is a deliberate artificial genealogical fiction, and yet inserts both in his list; had he had a solid pre-existent list and the knowledge that Maximus and Maximianus were the results of a duplication, he would not have been able to speak of seven emperors, but eight.  In other words, Geoffrey, Nennius and any other author who wanted to use the idea of Seven Emperors could change the list with a clear conscience: a canonic list did not exist.

The point of the list of Seven Emperors is that they represent the monarchy of all Britain, vested in one imperial race, the Romans.  As we have seen, the Gildasian British regarded Britain as a country for Britons to live in but for Romans to rule, and the English may have regarded themselves as superseding the royal group of the island.  Bede comes from Northumbria, one of the last English kingdoms to be formed, where English ethnicity was most strongly opposed to an intensely Celtic identity, and which faced most long and most fiercely a sequel of British kings and priests: the local English may have wanted their own ideological answer to enduring British claims.  They responded to the existence of a list of Seven Emperors with a list of Seven Breatwealdas, asserting in equally pseudo-antiquarian and mythical terms their own right to rule the island; and, as I said, this shows that the legend of the Seven Emperors was in existence by Bede’s time.

On the other hand, there is equally nothing to deny that the Northern pedigrees, which do not depend on it, stretch all the way back to genuine Roman times.  Jones, Morris and Sheppard Frere thought so, Dumville does not dismiss an identification of the Roman-age Coroticus with the Ceredig of the genealogies, and I have given, in my analysis of the formation of A, new reasons to suspect that Roman chieftains, such as are listed in the genealogies, were imposed on North British tribes after 367.  And there is yet another straw in the wind, blowing in the same direction: the possibly historical story that St.Ninian, who is said to have been the son of a Border kinglet, was sent to Rome to complete his education, and then built his new church, his Candida casa or white building, in deliberate imitation of Roman models.  Here we have, again, the association of Border, over-Wall kingship, with Roman identity, brought in from outside and imposed on local realities.  Ninian’s journey to Rome has been doubted – on no good grounds; we should remember that at the exact same time, another well-born Briton with ecclesiastical interests, Pelagius, made the exact same journey for the exact same reasons.

Each reason is thin; all reasons, together, are strong.  The over-Wall tribes had received new kings of Roman birth, meant to keep them under control; so implies an analysis of A.  About the second half of the fourth century – by reckoning the generations – Northern king-lists show a number of Roman names – Tacitus, Aeternus, Paternus, Quintilius, Clemens^[14].  A Northern king’s son whom later generations had good reason to remember, Ninian, imposed a Roman Christian religious model on his over-Wall diocese, and perhaps on the Picts as well; and we remember that we had reason to suspect that A contained a distorted hint or even account of the imposition of Christianity on the over-Wall tribes, parallel with that of Roman dynasties.

This is the missing link to Gildasian Britain, the explanation of the resurgent Celticism so clear in Gildas.  I have said that Gildasian Britain was a barbarian successor state, no less so than any Germanic, Slav or Arabic potentate establishing its military strength among fallen marble columns and broken inscriptions; and this is where the successor barbarians came from.  They came from the old Roman military lands between the Wall and Stirling; they were ruled by kings such as Coroticus; and they were strong enough to sweep south and take Britain all the way to the Channel, establishing, not a renewed Roman state, but the first ancestor of the Welsh nation.

The Celtic name of Coroticus, even if it was only part of a threefold Roman name, shows that the Celtic identity of these northern tribes not only carried on in Patrick's day, but had to a large extent imposed itself on its originally Roman ruling dynasties (the parallel with Old English settlers in Ireland and lowland nobles like the Frasers and Stuarts is irresistible).  Now, there is a peculiar and interesting regularity: both in the pedigree of Strathclyde and in that of Gwynedd, three generations of Roman-sounding names are followed by a Celtic one, Cunedda, Ceredig/Coroticus, which receives special honour as an ancestor.  The importance of Cunedda in legend is all too well known; as for Ceredig, he was apparently obscured by Dyfnwal Hen - which the pedigree makes his grandson - but there must be a reason why the pedigree attributes to him the impressive and coveted title of gwledig, and why he was apparently remembered - in no friendly terms, it seems - in the vernacular tradition of neighbouring Ireland (Dumville: "The form Coirtech found in the capitulum in the 'Book of Armagh' is what would be expected in Old Irish if the name had developped naturally in Irish from the fifth century... Coroticus would have been a natural Latinization of a form common to British and Irish which would become Old Welsh... Ceritic/Ceretic but Old Irish Coirtech".  Saint Patrick op.cit., 115).

This suggests two cases in which, three generations after the original installation of the Roman dynasties, a generation came in which was decidedly, even consciously Celtified, and which was in both cases notable for its conquering activities.  It is perhaps not a coincidence that this generation follows the time I suggested for the (Third) Pictish invasion, in which the border tribes' Celtic plebs and Roman rulers faced and defeated a most terrifying enemy.

No such Celtic name as Coroticus or Cunedda is heard elsewhere in the Roman Empire before about 468.  Then, suddenly, two men with similar and highly Celtic names - king Rigothamus and bishop Rigocatus^[15] - turn up in the heart of Gaul, the former fighting a Roman and Catholic war in defence of the remnants of Roman power south of the Channel, the latter appearing - some time after the defeat of Rigothamus - as what Sidonius Apollinaris calls a "twice exiled" Briton bearing spiritual aid to a group of other Britons.

This is the first appearance of Britain's new Celtic political subjects in European history^[16]; but it is to be noted that, though called by clearly Celtic names, there is no hint in Sidonius, our best source for both, that they are anything but cultured Romans in behaviour and attitude.  Sidonius addresses “Riothamus” as he would a Roman official, cultured and devoted to the public good, as Jeremy Du Quesnay Adams pointed out^[17], the letter is clearly addressed at someone who "...had had the benefit of a rather genteel education, or [else] members of his staff had".  But it is more than that: Sidonius attributes to Riothamus typically Roman values and feelings, such as in particular an earnest concern for the public good.  There is an element of personal feeling about his note that Rigothamus was so sensitive as to actually blush for other people’s misdeeds (uestri pudor inspicimus, cuius haec semper verecundia fuit ut pro culpis erubesceretis alienis); this is not a remark that could be addressed to just any and every Roman or barbarian general from whom justice was being sought.  Sidonius claims to have personally witnessed this personal quality of tender conscience.

Likewise, Sidonius speaks of Rigocatus (in a letter to that other Briton, Faustus) as "that venerable man", bishop and monk, who had taken refuge with Sidonius for a few months, on his way from Faustus himself to "your Britons", probably that community of British refugees in Burgundia, survivors from Rigothamus' defeat.  He was taking to them one or more books from Faustus^[18], which seems to show that Faustus and Rigocatus believed them to be in need of instruction.  This agrees with the fact that Rigotamus' troops seem to have behaved with complete disregard for Roman law and customs.  Sidonius writes to Rigotamus in the tone and manners of one highly-placed Roman nobleman to another; but he complains of the indiscipline and lack of respect for Roman law of his troops.  In fact, Sidonius’ carefully-written letter contrasts, with a subtlety and deliberation one does not expect from him, Rigothamus’ own civilized virtues with the uncontrolled behaviour of his men.  He is careful not to be negative even about them: they are not actually bad, but argutos, armatos, tumultuosos, uirtute, numero, contubernio contumaces: loud, armed, stormy, swollen by their courage, their number and their comradeship-in-arms; but, borne up by these no doubt good things, they have taken over or set free (it is not quite clear) the slaves of an already struggling and penurious Roman landholder of no rank or name, who is so terrified by their obstreperous behaviour that he has not, except under the protection of Sidonius (who, lest we forget, was a great lord) dared to approach the commander for redress.  One has the impression that this new British military community which so suddenly appeared in Gaul, had, as a whole, not much knowledge or respect of Roman religion and Roman ways.  But Sidonius appreciates Rigothamus’ civilized conscientiousness; and in the case of Rigocatus, whom he had good opportunity to get to know – blockaded as they were together for several months - he seems to see no difference between Rigocatus and Faustus, a much earlier British in-comer for whose saintly virtues he had the highest respect (he wrote a poem about them).  In spite of the cataclysmic changes that have taken place in the country between Faustus' arrival in the mid-420s and Rigocatus' in the early 470s, he assumes, in that Faustus regards Rigocatus' British flock as his own (Britannis tuis^[19])

The Ambrosius of Geoffrey is concerned with re-establishing "laws long since in disuse", which, as I said, must mean Roman laws.  I have postulated that L, the lost history of the Saxon war, contained an account of a Celticizing revolution, probably leading to the resurgent Celticism of Gildas’ time; but these are two hints, coming together from widely separate areas of evidence, to suggest that Ambrosius was not the leader of such a revolution, but to the contrary that he and his movement’s leaders - men like Rigothamus and Rigocatus - intended some sort of Roman restoration.  "At the mouth of two or three witnesses shall the truth be established."  And there is a third: G, the polemic from which Gildas drew his account of the Saxon settlement and revolt, was written long after the event.  The animosity directed at the tyrannus suggests that it was written after Ambrosius had killed him  - which, as we have seen, was probably the first act of Ambrosius' war of liberation.  It may well amount to a justification of Ambrosius' killing of his rival, and possibly of his execution of the Saxon leader.  The fact that it is still entirely rooted in the old Romano-British obsession with the details of Roman law, indicates that a writer who intended to justify Ambrosius' actions, did so exclusively by reference to Roman law.  Roman law still had its obsessional and near-sacred role in the minds of the British aristocracy.  The great switch to a Celtic law, recorded by L and preserved by Zosimus, must have come later.

Rigocatus, incidentally, was a monk as well as a bishop.  His obviously aristocratic name ("king of battle") seems related to that of Rigothamus; if they came over together, in temporal and spiritual charge of the 12,000 Britons who fought at Bourg-de-Deols, they may even have been brothers, or father and son (it is perhaps relevant that Ninian, the first known Borders bishop, was said to be the son of a king.).  This in turn suggests that the British monastery from which Rigocatus had come had some aristocratic character, or at least connection; and if the connection with Faustus meant anything more than the bond between two exiled and nostalgic Britons, it may have something to do with Faustus being a former abbot of the aristocratic monastery of Saint Honoratus at Lérins.  There is nothing to prevent us from imagining an imitation Lérins set up in Britain to train well-born clerical intellectuals, and keeping some contact with its prestigious Mediterranean model.  (At least two fifth-century monastic foundations are known, Whithorn and Bardsey Island; and God, literally, only knows how many others were destroyed beyond recovery by the lava flow of English conquest.)  It is quite possible that the thread of Roman mentality and manners that Rigocatus and Rigotamus took to Gaul was acquired at the school of such a monastery.  It is even possible to sense a foreshadowing of the enormous importance that monasteries were to gain in Ireland, Wales and Scotland, largely without towns, often even without villages, as centres of concentration and power for the ecclesiastical class.

Ri(g)ot(h)amus, the British king who leads barbarians to fight for Rome and who seems to Sidonius a Roman at heart, is close to the double nature of Patrick's Coroticus, a barbarian leader hosting mighty public banquets for his soldiers, raiding and plundering into enemy land, who yet is a Roman citizen, receives high-ranking Roman citizens such as bishops at his banquets, and feels the honour of citizenship strongly enough to be affected - so Patrick thinks - if it is denied to him; and like Rigothamus, Coroticus leads troops who seem to have no leaven of Roman discipline or law.  Yet the traitor Arvandus and the Visigoth Euric saw his troops as the only important military obstacle to be removed – while the “Greek Emperor” Anthemius could simply be de-recognized – to impose the ius gentium on Gaul and remove her from Rome for ever; and the Greek Emperor saw them as the only host willing, in a dark and dreadful time, to fight for Roman law and right.  Rigothamus is a Roman with a barbarian name, leading barbarians, for the last time, to die for Rome.

But Jordanes calls Rigothamus "king of the British" clearly and twice over; he had no doubt that he was not a functionary in the Roman fashion, but a head of state.  This implies that the army he brought over was "his" army not in the sense of an Roman army obeying its regularly appointed commander, but of a tribal following loyal to its king^[20].  Of course, Jordanes' Gothic sources had every interest in underlining the non-Roman nature of Rigothamus' office.  This made him no more than another intruder into Gaul, with less right to intervene there than the Visigoths, who had a certain amount of written law and treaties to validate their position; but that does not mean that it was a lie.  In fact, his name - "most royal" - strongly suggests that he either was envisaged as a royal heir from birth, baptized with a kingly name; or that he took it as a title.  If, therefore, Ambrosius was at this point in control of Britain - which we see no reason to doubt - then Rigothamus was at the head of his army, or of one of his armies, not by virtue of imperial^[21] appointment, but as being a king with a mighty following.  It follows that the Romano-British restoration I suggest Ambrosius attempted was limited by the power of those stubborn things, facts.  On the ground, the armies by which Ambrosius held Britain and intervened in Gaul were not Roman forces loyal to the res publica that organized and paid them, but tribal levies ran by their kinglets.

The new Celtic culture of Gildasian Britain is rooted in the lands beyond the Wall, the lands in which Patrick's enemy Coroticus already lived in ways that Patrick regarded as alien and repugnantly un-Roman.  These tribes were part of Britain's system of defence - it is significant that, though the British Roman state must have had to pay them some sort of subsidy, it did not feel the need to mint coins for the purpose: evidently payment in kind or bullion suited them well enough - and every crisis pushed them further to the forefront of political power.  Already by Patrick's time, their political weight was such that Patrick felt that the crime against his own converts risked going unpunished for political convenience; by the 460s, there is no reason not to believe that they formed the bulk of Ambrosius' armies, and their influence will have correspondingly increased.

There is nothing strange about this: everywhere in the Roman West, the fading into nothingness of the organized army apparatus and the attempt to manage and defend the empire with tribal levies led by chieftains and kings is the hallmark of the fifth century^[22], and everywhere the effect was the same: the ultimate establishment of virtually independent Germanic armies swiftly turning into states.  If the armed following of a barbarian leader regarded him as a king, it was little good the emperor making him a legate, a tribune, a comes, dux, magister militum or Consul^[23]; what mattered was the power-generating relationship with the troops that did the actual fighting.  The only difference between Britain and the rest of the West is that Britain grew her own barbarians.

This lay a mighty claim upon the country's future.  The armed, indispensable highland Celts would inevitably tend to impose their consuetudines wherever they went.  The first generation of leaders, whether of Roman origin like Ambrosius, or of Celtic descent like Rigothamus and Rigocatus, were still Romanized, and though they led an army of Northern Britons not at home with Roman civilization and law, they fought a Roman war for a Roman ideal.  But British armies were staffed by Britons from the northern kingdoms; a whole intact political structure flanked by that indispensable appendage of Celtic monarchy, a bardic class; it follows that a Celtic system of laws, equally native, equally connected with the local power structure, and equally elaborate, must have been in place, since one of the recognized roles of bards was to transmit traditional law.  This is the system of British law that came down from the north as the Saxon war went on; the ancestor, in some degree, of mediaeval Welsh law-codes.

The ideological path to this passes through the aggressive tribal Christianity shown in A.  We remember that the permanent removal of the Christian precious substances, wine and oil, to Rome, is to be read as an assertion of Britain’s continuing Christian and Catholic allegiance; and that Britain’s final victory over the Picts is preceded by its act of self-submission to God, Who takes, as I said, the semantic space of the Romans as supreme defender and King of the island.  That A consciously placed God in the semantic space previously occupied by the Romans is confirmed by Muirchu's legend of Coroticus, in which Coroticus represents the uolpeculae, the treacherous lower-order British kings, teyrnedd or tyranni; but the party which subjects it to the deserved punishment which the uolpeculae suffer in Gildas is not Rome, but Patrick and his God.  The semantic circle is completed when we reflect that Patrick was the legatus not only of God, as he himself said, but of the Church of Rome.  This shows that the ideas concerned were still understood in Ireland when the story was written: the British, as they depict themselves in A, have consciously and willingly transferred their earthly allegiance not to the Emperor, but to God.  It is God who now has charge of the royal function among them.  In His name, they destroy the Picts; and, as the myth of A indubitably took form in the years of the struggle against the Saxons, it is in His name that they must be understood to be fighting the Saxons.

A, as we have seen, is an analytical account of the function of royalty; it may be read as a manifesto for a new Celtic conception of political power, marking a genuine severance between the culture of the British tribes of the north and that of the distant Roman world.  However, when Rigothamus and Rigocatus were senior people, a high commander and a bishop, Roman realities still existed.  They come too early to be part of the development of Gildasian culture; for one thing, their generation could still remember the permanent Roman presence on the Wall - something that A explicitly denies.

There is no reason to date the writing of A (as opposed to the preservation and development of the distorted historical memories at its core) very early.  The fact that the Saxon war seems to have entered the realm of cultural truths places it quite far from its earliest beginnings, and the fundamentally Celtic attitudes it shows do not suggest the Roman restoration we have seen reason to suspect in the 460s, when Ambrosius, Rigothamus and Rigocatus all showed continuing allegiance to Roman ideals.  We have seen reason to believe that its picture of the relationship of Briton and Roman, with Roman ius being an intolerable iugum, is neither more nor less than the charter of the revolution that put an end to Roman law in Britain and replaced it with “British” ways, and that justified it by declaring the Britons consecrated to God, and making God their only emperor.  At the same time, we should not forget that the clever use of language we perceive in the deeply significant pun ius/iugum/iuramentum represents quite an advanced kind of Latin learning.

We should also remember that a certain amount of national records existed in independent Britain before the first Saxon war, as we have seen in discussing Gildas' sources, and that they were available to the person who wrote the anti-Vitalinus diatribe G, which, however rancorous and distorted, is closely based on facts; they must also, in some form, have reached the author of the Annales Romanorum - a document of very Dark Age vintage indeed - who gives a perfectly historical account of the origin of the Saxon war in the British refusal to pay annona to the settlers^[24].  Such records would have kept a historical account of the Pictish invasion (as well as of the notorious Saxon raid of 409/411), which would exclude the obviously legendary features of A; and it follows that A reached its written form in an environment that had no contact with whatever state records or Ambrosian accounts survived.  In other words, A is not a product of the house of Ambrosius.

Some time, as we said, after 480, A was the expression of a new, confident Latin-speaking but Celtic culture developping its own voice and consciously establishing its own attitudes.  It is indubitably the product of a learned class, to judge both by its profound articulacy and by its able Latin.  We need not take too narrow a view of what constitutes a learned class; for instance, there is no reason why it should be able to read and write, so long as the sort of things it knows are recognized as knowledge, as learning, by the wider society around them.  Gildas, whose learning was of mandarin level, did not know how to compose a chronology.  We have seen reason to believe that such a class was responsible for the preservation and analytical mythologization of the events north of the border between 360 and the 410s, forming eventually into a national mythology justifying the existence of a national aristocracy of Roman descent; and as A shows a very strong concern with the legitimacy and power of the Roman-descended royal lines, we must take it that it formed in a loyal environment - that is, at their courts.

There is no reason to envisage such a class anywhere south of the Wall.  Patrick, our witness for the attitude and nature of Coroticus, expects the local learned classes to be on the standard pattern of Roman aristocracy.  Even in the sixth century, Gildas could still see the presence of Celtic bards at Maglocunus’ court as something peculiar and repellent, and Cuneglasus, a man very like Maglocunus in many ways, harbours not bards but clergymen - the only learned class Gildas regards as legitimate. For that matter, the most ancient surviving Welsh poet, Taliesin, is from beyond the wall.  (Once we get rid of the fables about Urien leading Gwallawg, Morkent and Rhydderch to besiege the English in Lindisfarne, we can finally accept that his field of action was in north-west Britain, and his court almost undeniably in the village that preserves its name, Dunragit in the far west of Galloway.  The moving poems about the fall of Rheged have to be regarded as legend; the English did not get that far.)

Gildas regarded the "flatterers" of Maglocunus as part of an infernal engine of barbarization and demoralization, fawning on everything that was bad about conquering monarchs; in other words, as powerful promoters of social change in a militant, Celticizing direction.  Nobody who read the poetry of Taliesin will disagree: there clearly is a model of royal behaviour in which he believes very intensely and which he wants to see spread.  The arrival of someone like him at a royal court does not merely mean the addition of a kind of entertainment to royal pleasures, but the adoption of a new kind of learning, with colleagues, pupils and an elaborate ars poetica, promoting that model.  A modern parallel would be university colleges.  Their foundation in Mexico City and Lima (the 1500s), Cambridge, Massachusetts, and Quebec (1636), Calcutta (1856) or Stellenbosch (1871) represents the entrance of Mexico, Peru, British and French North America, India, and South Africa into a kind of learning structure typical of the West.  It is not only the introduction of certain subjects, but of a whole model and style of learning that brings those countries within the culture of the West.  By the same token, a bard did not merely bring songs: he brought a system of values, a very intense training regime, and a way of life and thought so involving and convincing that it lasted for centuries.

It is not only, however, in high literary culture that we see the signs, not just of a collapse of Roman culture into simpler ways of life, but of the conscious adoption of a full-blown Celtic culture, including decidedly elaborate features, as an alternative to Roman ways: it is also in material culture, in particular the return of hill forts such as the Iron Age would not have considered outlandish.  When a mighty king of Dumnonia started rebuilding the high place of South Cadbury, it was as though 500 years had not passed.  Some of it - made with the rubble of Roman building, yet! - has been identified as murus gallicus^[25], a wall-building technique not seen in Britain since Vespasian's artillery disposed of similar fortifications in the same area.  Now murus gallicus is not a primitive or haphazard way of building: it consists of a dry-stone frontage built on a stout frame of timber and backed by solid earth, and requires considerable professional skill.  What is more, there is evidence that details of its building, at least in certain individual specimens, are not functional, but designed to either display the wealth of the builder, or to suit religious or other considerations: in particular, some muri gallici have their timber frames shot through by enormous iron spikes which, according to experts, would do little or nothing to strengthen it against assault.  In general, the pre-Roman Celtic tradition of hill-fort building was heavily shot through with abstract, non-functional, probably religious ideas^[26].  It was also a skilled trade.  In other words, the skilled masons who built muri gallici were a cultural institution of much the same nature, if not the same prestige, as the bards. Its reappearance in Dumnonia shows not the collapse of Roman civilization, but the arrival of a different culture - in other words, the same picture we have been drawing all along.

And it is surely not a coincidence that both these skilled trades, bardism and hill-fort-building, are directly connected with the status, even more than the power, of kings.  A number of hill-forts were not functional in terms of defence; some may have been corrals, market places, or sacred enclosures; but they all represent concentrations of local power and prestige.  They imply the concentration of the land’s resources in one royal hand; in other words, they are consistent with the picture of a redistributive exchange system.  It is no coincidence that the legend of the salvation and reconstruction of the monarchy of Britain – that is, the legend of Vortigern and the dragons – features the building of a hillfort as its central issue.  By the same token, to have a great bard at court avails you less than nothing in terms of brute force, political or military; but it makes all the difference in the world to your prestige.  The insignificant northern caterans Urien and Owein had the most extraordinary post mortem careers, featuring in several separate heroic cycles, due purely to the eternity-grasping genius of their court poet.

There was an ethnic divide between north British barbarians and south British Romans that remained noticeable till the end of the Gildasian age and beyond.  The English used two separate words to describe the inhabitants of Britain: Wealh and Cumber.  Cumber is, of course, nothing else than the transliteration of a familiar word, archaic *Combroges, modern Cymraeg, "fellow-countryman", the Celtic equivalent of Gildas' ciuis, which, as we have seen, stands for the free-born Britons with no political rights or Roman identity, the mere dweller in the land, presumably not a Latin speaker.  Wealh, on the other hand, is the local version of an ethnic name spread across the Germanic world.  Current etymological theory has it that it comes from the name of a Continental Celtic tribe, the Volcae, who abutted on Germanic territories in the centuries before the birth of Christ; be that as it may, its Celtic origin is far less important than the fact that throughout the Continent, from the Rhine to the Vistula and from the Balkans to Iceland, it has one and only one meaning - Latin-speaker, Romance-speaker.  The French-speakers of the Low Countries are Walloons - "little wals".  Tyrolean hillmen know their disliked Italian neighbours as Wälsche (the expression is commonly regarded as an insult)^[27].  Throughout the Balkans^[28], Romance peoples are consistently called Vlachs or Wallachs; hence Wallachia for the heartland of Romance Dacia.  The Poles use the loan-word Wlochy for Italy, the heart of the Romance world.  And in Icelandic poetry, one sometimes comes across a remote but powerful individual, whose halls are full of treasures, called the Kjarr of Valland - that is (somewhat to our surprise) the Caesar of Roman-land^[29]. It is only in Britain that the Welsh are a group of Celtic, rather than Latin, foreigners.

Or are they? The English not only had two different words for defeated native of Britain; they named two different and separate territories for them - Cumberland for Cumber-, and Wales for Wealh. On the map of England itself, there are no place-names in Wal- in the far north, in counties Northumberland, Durham and Cumbria^[30]: these were evidently no place for Wealhas. But elsewhere, from Brighton to York and from Hereford to Norwich, you would be just as apt to meet a Wealh as a Camber. Place-names in Wal- and Cumber-, corresponding to ancient plots of lands and settlements left to the conquered natives in early Saxon land-sharings, are scattered pretty evenly; the neighbourhood of London has both a Walton and a Camberwell. In other words, Wealhs and Cambers could be found anywhere in central and southern Britain. I therefore suggest that Britain south of York was, in English eyes, a patchwork of settlements pertaining to these two groups. And this corresponds to the fact that, to this day, the people the English call Welsh call themselves Cymry; I mean that the ambiguity is systematic. In a territory which the first English identified as the land of the Wealhas, the Combroges identity triumphed, leaving behind as a fossile the English name of the country (or rather of the people, since Wales is a fossilized plural) as out of place as the present-day names of Bohemia, Macedonia or Massachusetts.

I have no doubt that this corresponds to Gildas' distinction between Romani and ciues, so distant from Patrick's subtle shading and near-identity between ciues meorum and ciues sanctorum Romanorum. Gildas never uses the expression ciuis Romanus, even though he must have been familiar with it - if nothing else, for Lord Palmerston's reason: St.Paul used it in his beloved Bible. But in Gildas' Britain, Romani are not ciues, and ciues are not Romani. It is at once an ethnic and a social distinction: Britain was a country for ciues to live in, but, as we have seen, for Romani to rule. And what the wide spread of land-names in Wal- tells us is that, even after the first flush of English conquest, there were plenty of conquered communities which the Germanic English identified not as Combroges, ciues, but as Welsh - Romani. Gildas thundered at contemporary Latin-speaking aristocrats, many with Roman names and Roman pedigrees, that they were not Romans at all, but Britons; and, culturally, he was right. But the Latin he uses is not only that of a virtuoso, but that of a native speaker, attuned to shades of meaning which (speaking as someone who writes English as a second language and is all too painfully conscious of his own occasional lapses) only come naturally to someone brought up in fluent Latin from birth. His family was probably bilingual; the ease with which he throws the insulting lanio fulue at Cuneglasus suggests that he found it just as easy to think in Celtic.

To attempt an exclusive definition of nationality is both impossible and useless; but to interpret what flagrantly glares at you in everything from place-names to high literature is both easy and inevitable. Ciuis always means, in Gildas, someone of lower social rank as well as of Celtic nationality; the two ideas are inextricably linked. And we need not doubt that the Latin and Roman identity was long seen as superior.

Indeed, I rather think that the word ciuis began as a sort of ethnic special pleading, suggesting that though the person so described is not a Romanus by blood and culture, nevertheless s/he shares in the commonwealth, in the identity, in the common interests, in the common legal rights, of Rome. My own people's recent historical experience shows an irresistible parallel: the once-popular dialectal Italian expression paisan or paisá (in book-Italian, compaesano). This was primarily a word of appeal, used by poor emigrant Italians to people recognized as fellow-Italians, especially if well-off or educated; it had something of a pleading tone, calling on the common origin in an alien environment, even as it recognized that, but for the common origin, there was little between the two Italians concerned. Alternatively, it might express relief and joy in finding a fellow-countryman in an alien and unwelcoming environment; and in that case, it could be used for a social equal. The parallel is not complete. The working-class Italian emigrant had nothing to compare to the ancient alternative culture of the Celtic British border realms. His country's culture, though varied, was univocal, and the use of the dominant, book-Italian language, as against pure dialect, would define an educated Italian almost as strictly as the use of book-Latin would have defined a Roman Briton as opposed to a highland, Celtic-speaking Britto. But in both cases, there was a strong element of overcoming cultural and social alienness by appealing to a common birth. Until the last couple of decades and the rise of the Bossi abomination, it never occurred to anyone to deny that the urbanites of Florence and Milan are fellow-countrymen even of the most uneducated, dialect-speaking farm labourers from the wilds of Calabria; before any such state as Italy existed, they all saw themselves as Italians, as against Arabs, or Turks, or French, or Germans, or Spanish, or English. In this context, the original sense of fellow-countryman must have been someone who is part of the same commonwealth or national group, but shares the dominant culture partly or not at all; a marginal person, a recognized "internal" barbarian.

The alien, unwelcoming, perhaps even dangerous environment of lands of immigration stimulated this kind of appeal; it was not uncommon, in the days of the great Italian migrations, for emigrants to be set upon or even murdered, and France, in particular, became notorious for an unpunished pogrom of Italian labourers at Aigues-Mortes. By the same token, the dangerous and violent environment of Britain after the first Saxon revolt would naturally stimulate a sense of mutual belonging and dependency, of common citizenship, between lowland Roman Britons and highland Celts.

It may well be that the real sense of common citizenship and interest was forged in what A regarded as a defining moment, the terrible Pictish invasion that followed the end of Roman rule. What probably happened was that Constantine III stripped the border of all professional troops, and even weakened the border tribes, since we know that in 409 he sent his son Constans, the former monk, to gather a large body of allied barbarians from Britain, the Honoriaci, whom he sent into Spain - a province full of Honorius' relatives - to bring it under his control^[31]. This must have been the opportunity which the Picts, whose defeat by Stilicho was by now more than ten years old, were waiting for. According to our reconstruction, they mauled the border kingdoms almost to the point of complete destruction: faced with overwhelming enmity, and probably deprived of any elite Roman support, the borderers found the destinies of all Britain cast on their shoulders. They won, or at least took part in a British victory, and were from then on willing to regard themselves as the heroic, Christian guardians of the Britanniae.

This will, of course, have reinforced their sense of tribal self-worth, and therefore all their traditions. It may have led to a peculiar new compromise with the Roman world to the south. Though we known that the Romano-British aristocracy was very committed to a Roman identity and to Roman law, we do not know how it related to its ancient pre-Roman past. It certainly must have been conscious that the tribal world of their northern neighbours was an image of the world of their own pre-Claudian forefathers (Ausonius, across the sea, praised some of his highly Roman contemporaries for being "descended from Druids"), and may have heard variants of the Celtic speech of the borders among their own tenants on their estates. There is plenty of evidence to show that Celtic dialects were spoken even in Romanized areas, though a distinction has to be drawn between the language of the north, spoken by poets and heard at the court of kings, and the farmer and labourer patois we have to imagine east of the Fosse Way. I repeat: there is no trace whatever of any native political organization, dynasty or reign, south of the Wall. Highland areas - Cornwall, Wales, the Pennines, Cumbria, north Yorkshire - may have been less Romanized than the lowlands, but there is no reason to presume that they carried a high Celtic culture for all that; the likelihood is that they counted mainly as grazing areas to supply the south-east with beef, lamb and wool, but that they were integrated into a fully Roman trading system in which such places as York and Chester were the final markets for their cattle. They were probably no more independent of Roman organization than the average herder, down the centuries, has been of the empires he lived in.

The way I read the evidence is this: that from the time of Ambrosius’ revolt, before 468, the northern tribes - which had supported him - were regularly tapped for fighting manpower. The obscure movements of Rigothamus’ British force in Gaul during the war of the Loire might give us a template for their development: the fact that they had a bishop of their own suggests as much a tribal group as an army corps, and the fact that they fled from their disaster at Bourg-de-Deols to settle in Burgundian land as a recognizable tribal unit - whose future history, in so far as they had one, was not that of an armed force but of a small ethnic group with its own identity - also suggests that they were in Gaul as much as armed settlers, probably with wives and children, as an army. I think that through the course of the late fifth and early sixth century, Britain and any other territory held by the Ambrosiad dynasty would have seen a development of military Celtic settlements, distinguished from the Romans of the south by their own language and culture, each led by a recognized aristocratic group (perhaps a tigernos and his family) and possibly with its own priest or even bishop.

These settlements, in my view, led not to the mixing of Celtic and Roman identities, but to the formation, in a Romanized and Latin-speaking lowlands, of a network of distinctive and separate tribal settlements, which kept their own culture and law in more or less conscious defiance of the Roman conformity of their surroundings; though it is possible that they may have attracted those among the population - if there were any - that still kept some sort of Celtic dialect among the cultural Latin of the country. (It must however be rmembered that the jargon of serfs without cultural and political institutions of their own, simply subject to a Latin-speaking landowning class, would not be the same thing as the learned Celtic speech of justiciars, bards and royal courts; indeed, given the strong class feeling among Celtic royal families, it might well be more repellent to them than Latin.) This would explain the patchwork of Wealh and Camber names in lowland England. Therefore, when the Celticizing revolution of L came, it could call on the loyalty or at least the interest of a considerable number of armed settlers who had preserved their Celtic identity and their warrior values, and indeed had probably been encouraged to preserve them, in order to defend the lowlands. In the next chapter we will meet evidence for such a social grouping, with a very distinct and separate identity, on the shores of the Loire some time after 508AD.

Notes

^[1]As attractively argued by STORONI MAZZOLARI op.cit. 257-273, pointing out Stoic antecedents to Pelagius' moralism. Stoicism, of course, was the Roman philosophy of choice, incarnating the stern ideals of Roman grauitas and uirtus.

^[2]I would suspect that many of them never altogether lost, during the Roman period, the cultic significance that hill-forts certainly did have in Celtic paganism; and that local noble families, however Romanized, may have been linked to them by family traditions, possibly by hereditary priesthoods. Once the settled Roman world collapsed, it would not be unimaginable for local noble families to resort again to the ancient fortifications of their ancestors.

^[3]Or even to the closer continent. There was still cultural continuity in Gaul, especially in the South, and - not only before the Justinianic catastrophe - in Italy. We know that the British clergy had some contact with, and interest in, the Frankish kingdoms, and I have a vague suspicion that St.Gildas himself may have met at least one Italian.

^[4]Sidonius,op.cit., Letter 1.7.

^[5]It is worth noticing that Patrick is, whether consciously or unconsciously, placing himself in the role of the prophetic, vision-seeing bard praising the triumphs of his king: a Celtic attitude, replicating the Celtic view of the two aristocracies of society, druidic and royal. He himself is the visionary sage, knowing by inspiration the present and future fate of men; his neophytes are the royal group, triumphing in eternity over their enemies and rejoicing in eternal royalty at the court of the ultimate King of all.

^[6]RAMSAY MACMULLEN, The Roman concept Robber-Pretender, in Revue internationale des droits de l'Antiquite' 3rd series, 10, 1963, 321-25, quoted in SNYDER op.cit. 304 n.72.

^[7]SNYDER, op.cit.308, n.12, quotes Bury and E.A.Thompson as arguing that Coroticus was connected with Roman military organization. My conclusions are arrived at independently of theirs, but I certainly agree.

^[8]It is also, in spite of deep similarities, not a feature of Western mediaeval feudalism. The point with feudalism is that for all the hereditary and virtually royal powers of the average duke, count or baron, they were at least in theory functionaries of their king or emperor, swearing an oath of loyalty and accepting the jurisdiction of their sovereign, who could, and on occasion did, take away fiefs, or hand them to someone else, or intervene in their administration. Ireland kicked for centuries against English attempts to introduce similar principles in its law.

^[9]There is a slight possibility that the notion of, specifically, Celtic bards as "parasites" may have an ancient pedigree. Poseidonius of Apamela, the famous Greek witness to Celtic society, described them as parasitoi, though in Greek this can mean no more than "those who sit at the table with you, or who share in the sacrifice". The coincidence with Gildas' term, however, is interesting, especially since Patrick - the only earlier witness we have to the realities of a British Celtic court - also speaks, however vaguely, of "flattery". Most of what was written in antiquity about the Celts is now lost, but it is possible that Patrick and Gildas may both reflect a traditional criticism of Celtic bardic poetry. Certainly, the outrageous flattery is sufficiently evident in surviving Irish and Welsh praise poetry; and the parasitical aspect in the story of the arch-poet Senchan Torpeist. A legend says that, in the seventh century, their claims grew so outrageous that a king of Ireland tried to destroy them all. LADY WILDE, Ancient legend of Ireland, London 1888, 159-166. Both aspects may well have struck outside observes enough to become a locus classicus. Of course, there is no evidence of this; indeed, while overwhelmingly likely, it is not 100% certain that Gildas was even speaking about bards.

^[10]An excellent remark by Snyder (op. cit., 295, n.36): "Patrick is clearly proud of his family's status, defined... in Roman terms, and perhaps he is intentionally using the phrase nobilitatem meam [my good birth] as a play on his name Patricius" [aristocrat]. I am not sure whether "proud" is the right word: but he certainly does call his good birth into account against those who question his learning and rank, implying that he is unfit to be a Bishop.

^[11]As the Romans had obtruded onto the consciousness of British tribes both sides of the Wall for a good three centuries, it is very likely that the highlanders evolved an epic mythology which included Rome's overwhelming presence as part of its world-picture. In other words, among the Celts of north England and Scotland, the Romans must have been a matter of myth and legend even while they ruled Britain.

^[12]According to Nora Chadwick, the names which precede Roman ones in the Northern pedigrees are recognizably Pictish - N.K.CHADWICK, Celtic Britain, London 1962, 40. It is possible that these may represent an actual historical memory, the only one available, of the local dynasties overturned and replaced by Roman lords after the disasters of 367; in which case, we conclude that the border tribes were ethnically close enough to the Picts to use the same names.

^[13]DUMVILLE, Sub-Roman Britain: history and legend, in History (new series) 62 (1977) 173-92.

^[14]The pedigree of Coel has an Urban and a Grat - clearly, Urbanus and Gratus or Gratianus - and Coel himself can easily be read as a late Roman name using the word Caelus, heaven, like that of Pelagius' colleague Caelestius; among his descendants are a Cadrod - Quadratus - of Calchvynydd - Kelso, in Scotland - and a Garbanion - Germanianus.

^[15]Sidonius spells his name Riocatus. I have ventured to correct his spelling, as I have done with that of Rigothamus.

^[16]The presence of a British bishop called Mansuetus at a small local Council at Tours, in 461 could be related. The year seems too early to connect him with the Celtic warriors of Rigothamus, and his Roman name does not suggest a common origin; but he might be an exile from the Roman south of Britain, who found pagan Saxon overlordship incompatible with conscientious Catholicism. If he was an exile in 461, then this gives a tighter date for Ambrosius’ revolt: after 461 and before 468.

^[17]JEREMY DU QUESNAY ADAMS, Sidonius and Riothamus, in Arthurian Literature 12 (1993), 157-164.

^[18]For some reason, he had failed to mention one of these books, written by Faustus himself, to Sidonius; who, when he found out, pursued him, asked to see the book, and copied out some sections, before sending Rigocatus on his way. The background to these strange actions is lost, and I will not attempt a guess; it does not seem to have reduced Sidonius' opinion of Rigocatus – he still calls him ipse uenerabilis, and he tells Faustus that they left each other on the best of terms - but it does remind us of the epic struggle over the copying of a book, which, in Irish legend, preluded to a great war and to the exiling of St.Columba.

^[19]Faustus may have kept in contact with events in Britain. There is a suggestion that he is the author of the Chronicle of 452, which mentions and dates the Saxon revolt for us. If that is the case, then he might have taken a particular interest in an army of fellow-countrymen stranded in his own land of exile, and in a bishop from Home. IAN WOOD, in After Empire, ed.G.AUSENDA, Woodbridge 1995, p.205; quoted in SNYDER op.cit.275 (n.47).

^[20]The huge number involved, however, do not make it likely that this should be the levy from a single tribe: 12,000 men would probably wipe out its whole manhood. It is likelier to amount to drafts from all the various tribes, plus perhaps south British and Christian Irish, under the leadership of one king (at least; we don't know that Rigothamus did not have any other men of royal rank under him. Perhaps Rigothamus himself was of higher rank than the kings of other tribes; in later terms, a gwledig rather than a teyrn).

^[21]The British claim to an imperial Roman title in succession to Constantine III was probably still a living reality.

^[22]The case is excellently set out by that pernickety but often penetrating contrarian, Roger Collins, in his Macmillan history of Europe: Early medieval Europe 300-1000, 80-99.

^[23]In the fourth century, the Alamannic king Fraomar was taken into the Roman army, with his men, with the rank of Tribune (general); by the beginning of the fifth, Alaric was demanding that of magister militum - commander of the whole army; in 507, Clovis received that of Consul. Evidently the ranks available to successful Germanic warlords willing to identify their fortunes with Rome's were getting higher.

^[24]In book 6, ch.9, note 33, I have also tried to show that Nennius inserted another passage from the Annales, describing the surrender of Vitalinus to the Saxons, into his version of O, and that the Annales fragments can be identified both by their content and by two telltale slips of the tongue - the statement that the Saxons intended to destroy Vortigern’s exercitus, army, and a threefold repetition of a phrase - ut facerent - which is typical of the Annales style. In other words, the whole Annales account was historical, without admixture of legend. I wish we had it.

^[25]SNYDER op.cit.182 and note 257, quoting Leslie Alcock's famous excavations. South Cadbury does, according to Alcock, show some Roman-derived features such as a timber gate apparently similar to a simple Roman design; but that does not alter the fact that what we have here is nothing but the mother of all Celtic hill-forts, and cannot be understood except in their context.

^[26]All the data in these two paragraphs come from IAN RALSTON, Fortification and defence, in GREEN, Celtic World 59-78.

^[27]The common etymological consensus is that the Welsh ethnic name simply means foreigners, any sort of foreigners; which is nonsense resting on a groundless presupposition that the ancient Germans were unable to distinguish the ethnicity of anyone who was not one of their people - as if Roman identity and language were not easily recognizable! As a matter of fact, there is a similar word for Slavs: Wendisch (Brandenburg) or Windisch (Austria), with similarly demeaning connotations.

^[28]Although the Balkan world has no ancient native Germanic speech - its only Teutonic languages are the Transylvanian Saxon of mediaeval settlers and the local Yiddish - its various ethnic groups have been strongly marked by Germanic influence, including a long Gothic presence. The Goths first broke into the Roman Balkans in the third century, settled in strength after 376, and probably left behind a considerable residue when their leading clans, organized into Ostrogoths and Visigoths, moved westwards to form great kingdoms.It is surely due to a Gothic substratum that three Balkanic languages from different linguistic groups, Slavonic Bulgarian, Romance Rumanian, and Demotic Greek, share a fossile form of a typical Germanic feature, the use of the verb will (reduced to a prefix or suffix) to form the future tense of other verbs.

^[29]SNORRI STURLUSON, Skaldskaparmal 64, making Kjarr a descendant of a legendary first king Audi (perhaps Augustus?); ANON., Volundarkviða introductory prose and verse 2, where "king Kiarr of Valland" is the father of a Valkyrie, Olrun; Atlakviða 7, where the treasures of the Nibelungs Hogni (Hagen) and Gunnar (Gunther) are said to come from the halls of Kiarr, possibly a memory of the historical role of the Burgundians in the fall of the Western Empire. It must also be remembered that "Kaiser, king of the Greeks", along with Alexander the Great and with the mighty but unknown Hwala (an eponymous hero of the Germanic idea of the Romans?), had proud places among the otherwise wholly Germanic list of ancient kings in the Old English poem Widsið, and that, more interestingly, the first ancestor of the royal house of East Anglia was one Casar Wodning, indubitably Caesar son of Woden. The title of the Roman king of kings had long since entered Germanic legend. His adoption as the ultimate ancestor of the East Anglian Wuffingas may represent some obscure attempt to claim a common origin with folk of Roman descent.

^[30]Names in Wealh- "are distributed widely throughout [England], though there are some gaps, most notably in the extreme north-west and north-east" (underline mine); KENNETH CAMERON, English place-names, London 1996, 45.

^[31]It is possible that these Honoriaci may have shared the unruliness and lack of respect for Roman law later shown by Rigothamus' soldiers. Constantine's negotiations with Honorius - who seems to have been disposed to put up with him as a colleague - were blown sky-high by the news that Honorius' relatives Veranianus and Didimus had been killed; Constantine insisted that they had been put to death against his orders, but from that moment on Honorius was clearly committed to opposing him, by stealth if necessary, by force as soon as possible. The murders were carried out by Constantine's forces in Spain; that is, almost certainly, by the Honoriaci. SNYDER op.cit. 19ff. and notes (270); cf.HANSON op.cit.151, quoting C.E.STEVENS, Marcus, Gratian, Constantine, in Athenaeum 35 (1957), 316-47 esp.325ff.

History of Britain, 407-597 is copyright © 2002, Fabio P. Barbieri. Used with permission.

Comments to: Fabio P. Barbieri