British History, 407-597, by Fabio P. Barbieri

Novels I Classical Literature I (Pseudo)-Scientific I Children's Books I Comics I Music & Movies I Various stuff
What's New I Sitemap I Arthurian articles I History of Britain, 407-597 I View guestbook I Sign guestbook I Poll I About me I Links I Search

  Vortigern Studies > Faces of Arthur > British History > Book VIII > chapter 8.1

Faces of Arthur Index

Vortigen Studies Index

The Arthurian Collection is a part of Vortigern Studies



British History
click here

Chapter 8.1: Preliminary Considerations

Fabio P. Barbieri

As our research went on, we have become increasingly aware of the amount of literate tradition behind all our authors.  Except perhaps for St.Patrick, they all come with their own vast and largely lost literary history.  Gildas, Nennius and Geoffrey all imply bookshelves of earlier written works, read, absorbed, reworked and forgotten; and the past even of minor sources such as Muirchu is not much less diverse and interesting. To analyze their sources and their individual methods of treatment is of overwhelming importance; but nothing can be done until those sources have been identified, sorted and described, as far as the material allows.

We have done so - I hope with some success - with Gildas, Muirchu and Geoffrey, but not yet with Nennius, whose work, much more than the others, is clearly a weave of quotations from otherwise lost written works.  Now, comparison reveals an interesting pattern.  Nennius has clearly read Gildas, and makes infelicitous attempts to imitate what he regards as distinguished Gildasian usages - trans Tithicam vallem for "across the sea"[1], cyulae for "warships"; but he never once excerpts Gildas.  He seems to have read L, making use of his famous solecism magna discrimina; but nothing in his episodic description of the Saxon wars can be led back to such a document.  He has read Bede, but hardly used him: his account of Roman Britain is drawn from earlier sources, both continental and Welsh[2].  He must have read the Discordia Guitolini et Ambrosii, id est Guoloppum; but he tells us nothing whatsoever about it, though he regards it as so important that he places his single mention of it in his summary of the chronology of the world since creation.

Take, besides, the Galfridian source N, clearly another major piece of work.  We have seen reason to suggest that it was written in the second quarter of the seventh century, two full centuries before Nennius; like Nennius, its author was a savant of Vortigernid inclinations - though probably Catellid rather than Pascentiad - working in the shadow of Gwynedd power; it originated in the same geographical area, and, unlike Gildas and some items that Nennius did use - O, the Genealogy of ch.17[3] - was not separated from his time by the catastrophe of the rise of England.  Its author moved in an already Welsh geography, between an already formed Powys and Gwynedd.  In short, if there is a work Nennius is a priori likely to have read, it is this one.  And yet we have to wait for Geoffrey.  And finally, there is another, equally significant kind of omission: there is not a single note on the history of any of the British kingdoms then existing, from Strathclyde to Brittany; no legends, battles, not even pedigrees.  When you think how immensely important - how truly fundamental - genealogy was to the Welsh, this is an oversight of gigantic proportions.

Bede, Gildas, L, Discordia, N, Welsh genealogies: all prominent and significant items, most of them fundamental to Welsh culture, some of them so important that they survived to our day.  We are certain, for various different reasons, that Nennius must have known them all.  He did not use any.

And what materials he did use is equally significant.  He has a marked preference for out-of-the-way material and for Vortigernid-Pascentiad material.  The only Welsh pedigree he does deliver is the Pascentiad one; almost the only really extensive, continuous strands of material in his work - as opposed to summaries, notes and brief quotations - are the Vortigernid-Pascentiad *Gesta Germani and his visibly pro-Vortigernid version of O, which concludes with Pascent being granted land by Ambrosius.  (Nennius did not make up the *Gesta Germani; the fact that its author clearly had not read Bede, while Nennius just as clearly had, speaks for itself.  That he tried to edit it into other Vortigern legends - not quite successfully - also argues that he found it already made and regarded it as authoritative.)  Other than that, his sources are fragmentary (the bizarrely confused summary of A in ch.30), obscure (the Genealogy of ch.17, of which Nennius himself did not know what to make), foreign (the pedigrees of English kings; the Kentish legend of Hengist; the Irish account of the Conquests of Ireland, and of St.Patrick; the Roman-originated notice that only two Emperors, Septimius Severus and Constantius Chlorus, died in Britain) or North British (the "northern memoranda").

Whenever Nennius can, he points out the connection with accounts and personages known to his readers.  He mentions that the original name of his Patrick was Maun; that the Aelfred son of Ealric of the house of Northumbria is the same as Aedlferd Fflessaur; that the servant promoted by St.Germanus is none other than Cadell Ddyrnllug; that the distribution of looted goods back to the kings of the British is the same episode as they know as Atbret Iudaeu.  In fact, he is too eager to connect his obscure or foreign sources with known Welsh accounts, and sometimes he makes big mistakes: as Dumville pointed out, he identifies the Englishman Aeta son of Leodwald with the Welsh hero Eata Glynmawr - wrong both in race and, by several decades, in time.  But what this tells us is that Nennius saw his business as clarifying these foreign or fragmentary accounts by connecting them with items known to his public; which is why he so often quotes a Welsh, rather than Latin, name for an event - he is using the words his listeners would use for it in ordinary speech.

Even when Nennius brings in an extensive and indubitably Welsh account, such as the *Gesta Germani, it is clear from his words and manner that his public does not know it: "I have decided that some miracles wrought by God through him should be written down".  Would any writer refer in those terms to facts known to his readers or listeners?

It is obvious, then, that Nennius is working at the margins of the Welsh historical (or pseudo-historical) tradition.  He is not delivering a summary of all that may be known about Britain in his time, but rather a compilation of foreign and obscure sources for the use of those who do not have the time or resources to hunt them up for themselves.  This may have to do with the reason why his work spread so fast and so far: it was meant, from the beginning, as a standard summary of otherwise obscure sources.

This is not actually a contradiction of Professor Dumville's interpretation of his work as an Irish-style "stitching history", so much as a heavy qualification.  Nennius did indeed stitch and interpret, but his visible and obtrusive, yet often curiously vague chronological time frame, is meant to place all these scraps of information in a comprehensible structure, to supplement, rather than replace, what historical writing existed in his time[4].  When it fades into vagueness - as with the "northern memoranda", where no event is dated even to the century - it is because his sources allowed him neither a date nor a hook on any datable event.  This, in turn, is clearly another reason why he is so keen to find Welsh correspondences, however mistaken - that is, at whatever risk to precision - for his checkable and superficially quite precise English king-lists and for other sixth- and seventh-century events.

The direct and indirect influence of Gildas on the author of the Historia Brittonum was immense, visible not only in the general plan of his work, but also in passing matters of usage such as his constant use of the Gildasian word cyulae for "warships", which - whatever his other debts, to the earlier and widely different writers he copied or epitomized - he used from start to finish.  Nennius must have not only known Gildas, but taken his work as canonic.  For instance, he misunderstood Gildas' statement that Britain had twenty-eight ciuitates to mean, since he probably had never seen a town[5] - urbs, in his writing, always means "royal fortress", for which Gildas uses arx -, twenty-eight royal fortresses; and, as Gildas did not bother to list these twenty-eight ciutates, Nennius simply made one up from all the town and royal fortresses he knew, including a few that never had existed outside legend.  For instance, after Caer Ceint (the royal fortress of Kent, hence Canterbury), he placed the otherwise unknown Caer Gwyrangcon: Gwyrangcon, unknown outside Nennius, was the man from whom Vortigern was supposed to have taken away Kent to give it to Hengist.  Morris, with his typical tin ear for legend, takes the whole Hengist saga as historical, and (in spite of the Welsh name, completely unimaginable in a fifth-century lord of Kent) assumes Gwyrangcon to have been the original lord of Canterbury.  This is obvious nonsense; but it seems a great deal easier to imagine that the dethroned Gwyrangcon, like Vortigern himself after him, went on to take refuge from the Saxons in some other fortress of his own, defending a reduced space; as Gwyrangcon is only a legendary figure, it must follow that Caer Gwyrangcon, this fortress which nobody seems able to find, is also a legendary place. But his dependence on Gildas is as clear his complete absence of anything properly Gildasian: he never excerpts Gildas or sums up any Gildasian item.  In other words, he expected his readers to have Gildas' book on their shelves.

His relationship with L seems exactly similar.  He quotes the peculiar expression magna discrimina from him as he quotes the peculiar expression cyulae from Gildas, though, it must be admitted, to better purpose; but so far as I can tell, there is not one word in him which can help reconstruct L.  Unlike the Nennian version of A, whose peculiarities gave us such an interesting time, there is no Nennian chapter that has anything to say about what Zosimus, more than Gildas, shows to be L's central issues.  Zosimus' statements are few and clumsy, but clear: in the course of an incredibly dangerous struggle against the Saxons, the leaders of British poleis, tribes, resolve to reject Roman law and revert to "British" customs, a movement which spreads to Armorica and unspecified other Gaulish provinces, which overthrow the upholders of Roman law - rhomaios archontas.  Nothing of this may be retrieved in Nennius, even though the author was keen to run down and copy even the most obscure and pointless bits of lore, such as the Genealogy of ch.17.  And if he did not copy L out anywhere, this must mean that - like Gildas, like the history of Gwynedd, like the genealogy of King Mervyn - his audience already knew it all.

In other words, the Welsh educated classes of Nennius' time (which must mean mainly the monasteries) had comparatively common access to what we might call a "Welsh historical collection" consisting, at a minimum, of Gildas, L, N, Saints' lives (the fact that he only excerpted the *Gesta Germani and an Irish *Vita Patricii, the one obscure by his own admission, the other foreign, tells its own story) and royal pedigrees; and he wished, not to integrate this "historical collection" into one single vast narration, but to add a volume of obscure and otherwise unobtainable items to it.  His work's success - Professor Dumville followed a backbreaking paper trail of dozens of manuscripts and variants - and the disappearance of other sources except Bede and Gildas, have muddied the waters in quite a fantastic fashion, leading to the universal scholarly impression that Nennius was typical of what Welshmen knew and thought of their own past in his time; a mistake which begins with the writer of the Preface apologizing for the supposed ignorance of Nennius' predecessors, and carries on even to the most learned and insightful of our own scholars - even though the earliest copyists, at least, seem to have understood Nennius' purpose, supplementing it with equally obscure material of their own, so that some of the most useful Nennian items come from glosses and interpolations.

(From this point of view, his treatment of St.Patrick is of extraordinary interest.  First, the fact that he placed a complete Irish Life of St.Patrick in his work seems to suggest that the heroic Irish view of the Saint was still obscure and exotic in 834AD Wales; and secondly, the fact that he inserted an explanation of Patrick's name Magonus which has no Irish correspondent and which uses a form - Maun - which, according to Dumville's authority, is flagrantly native Welsh - shows that his readers must have been familiar with a different account of Patricius/Maun - as I argued on other grounds.)

The Welsh historical collection excluded A, and Nennius' attitude to A shows us why: while it is still a central account to Gildas, who both accepts and modifies its scheme, Nennius does not even bother to try to understand it, and in fact jerks it to pieces.  It had clearly lost all interest to his age, while the scheme of the Seven Emperors seems to have held complete sway (common to Nennius and Geoffrey, echoed both in Maxen and in Triadic material, and preluded to in the legend of Casswallawn in the Third branch of the Mabinogi).  On the other hand, L is treated as part of the common heritage of Gwrtheyrnion and Gwynedd, probably of all Wales.  And this raises the question: Where did it go?  Why did L altogether vanish in the few centuries that separate Nennius from the bulk of written evidence from Wales?  What happened to so completely erase what was, to both Gildas and Nennius, a capital part of the British/Welsh heritage?

There is an obvious watershed in the time span indicated: the publication of Geoffrey of Monmouth's Historia regum Britanniae, one of the most epoch-making events in the history of British literature.  Nothing in the English, Welsh, indeed European image of the past, was left untouched by this literary triumph.  In Wales, its influence was so enormous that translations and adaptations came to form a new literary genre, the Bruts (from the name of the founding hero, Brutus).  It seems by no means unlikely that this literary earthquake, proposing a striking, frequently novel, nationally flattering, and artistically magnificent picture of the British past, may have removed the hold of L on Welsh culture.  It could not, however, have done so unless it filled its semantic space to some degree.

What Nennius had no intention of doing, Geoffrey had and did.  His compilation includes all the things Nennius had left out: Gildas, Bede, N, famous saints such as Illtyd, Dyvrig and Samson, native mythology about such figures as Brân and Beli, and genealogy galore.  The most important feature of its immense success is its eager and universal reception into Wales, with not a trace of the doubts aired by Anglo-Norman savants such as Gerald "of Wales" and William of Newburgh[6]: from the moment the first Brut enters Wales, the future of Welsh culture belongs to it.  And yet Geoffrey had done incredible things with genealogy, backdating dynastic ancestors by centuries or even millennia, placing the well-known Cunedda or Cunedag (which Nennius had clearly dated after the fall of Maximus) to the sixth century BC, and so on.  If all of his work treated Welsh culture in this hucksterish and dishonest fashion, would not the Welsh reject it?

But in actual fact, not all his dealings with Welsh classics were as careless as that.  He used Gildas with great care, if with the evident purpose of subverting everything the older author had to say (and do we not know modern historians approaching the work of earlier colleagues with pretty similar intentions?).  Whole passages are quoted almost word for word, in their right place in the chronology.  And we remember that, of all historical subjects, genealogy was the most pawed-about, dishonestly handled and obviously variable: by the standards of what not only the Welsh, but all European noble houses did to their own pedigrees, awarding themselves descents from everyone from Priam of Troy down, Geoffrey's treatment is, I will not say nothing special - it has rare breadth and single-mindedness about it, intending to include everything possible and to form a single line of descent - but at least in the mainstream.  To us, it is an absolute barrier to believing anything in the great hoaxer's compilation; to medieval savants, familiar with the vagaries, let alone dishonesties, of pedigree writing, it was nothing of the kind.  Indeed, if they had to choose between an old pedigree attributing descent from all sorts of famous men of old to unremarkable and perhaps despicable contemporary lordlings, and a much grander vision placing the same ancient heroes in a much longer chronological perspective (familiar to them all from their annalistic studies, that regularly took them back to such ancient kings as Ninus of Babylon and Busiris of Egypt) and separating them from the little self-interested men of their day, I think I know which they would instinctively choose.

My point is that the genealogical fictions of Geoffrey only look odd to us because we live in a different world: in medieval Wales, as in the rest of Europe, they would not be surprising except by their thoroughness and brilliance.  He was the ace of aces, the champion at a game they all played; but that does not necessarily disqualify his renderings of other historical material.  And we have seen reason to take his rendering of N and his version of O seriously, reproducing genuine material that Nennius either ignored or suppressed.

In fact, I think I can suggest that Geoffrey's success in Wales, and perhaps elsewhere too, was due largely to a purposeful and successful recasting of the whole "Welsh historical collection" in one stupendous whole.  This, according to Rachel Bromwich, was the universal view of Geoffrey among the first generations of Welsh antiquarians – men such as Robert Vaughan: that he was a transcriber or collector, not an original author, whose sources were good witnesses for the earliest stages of British history.  In spite of his shameless and even whimsical twisting of sources, his work was close enough to known material to be accepted; and – in spite of flaws on which later scholars pounced, such as the relationship of Arthur with Anna, Howel and Gawain - it performed the acrobatic feat of binding everything together, giving that clarity and breadth of vision for which all thinking minds yearn.  It was also brilliantly accessible.  Geoffrey's Latin, except where he is copying, has none of the mandarin quality of Gildas, deliberately forcing the listener or reader to work and impressing them with his rhetorical abilities; Geoffrey's fireworks are of a different kind, in the nature of peacock-like displays of learning rather than of verbal brilliance.  He wrote beautifully, but easily, yet never allowed his readers to forget that he knew a lot more than they did.

As Virgil's version of the Aeneas myth obliterated those of Naevius and Ennius, so Geoffrey's history of Wales obliterated all his predecessors except Bede, Gildas and Nennius, whose work contained information his did not[7].  He did not even have much of a struggle.  Few manuscripts probably existed at all; Geoffrey's work corresponded with a great expansion of book production and reading in Europe, and in particular in England and Wales, when books entered the houses of private persons and began to be read individually and not aloud.  Lewis Thorpe[8]: "From the very beginning Geoffrey makes it clear that he is writing [for] the solitary reader, not to be declaimed aloud in serial form... There are none of the calls for silence so common in medieval vernacular literature, none of the three-fold repetitions felt necessary in the Chansons de geste."  It is by entering British and European culture at this new level (which was soon to produce Aquinas and Dante), the level of the private rather than institutional reader, that Geoffrey's vision of Welsh history triumphed.  It filled the historical horizon of people who had never been acquainted with the bound products of monastic scriptoria, and who therefore could not appreciate how badly it clashed with the most illustrious of his sources - Orosius, Bede, Gildas.  For every educated monk acquainted with the facts - and how many even of the members of the average monastery will have been? - there will have been twenty members of nobility, gentry and urban upper bourgeoisie who read Geoffrey if they read any British history at all; and soon there would be fresh recruits to the monastic educated classes who had grown up with a volume of Geoffrey in their home, and who - exactly because they had a bent towards learning and reading - would have absorbed it before they ever started on a course of higher studies, thus entering college with their minds already full of his particular brand of learning.  (A kind of problem with which modern university teachers are all too well acquainted.)

Geoffrey did not respect his public, but he understood them.  As Neil Wright's analysis clearly shows, his treatment of Gildas and Bede is nothing short of impudent[9]; he evidently felt, like many expert but opinionated scholars who, in today's world, stop contributing to learned publications and start collaborating with newspapers, that he could get away with murder.  But the basis of the "impudent expert" syndrome is real knowledge, manipulated in the service of the author's ego or hobby-horses; and Geoffrey really had read the authors whose work he so carefully twisted.  It follows that where he is our only source for other stories and legends, we are entitled - knowing what he did to those of his predecessors whom we know - to pull a long face, but not to deny their previous existence.

The biggest of Geoffrey's lies is that he could read Welsh - let alone "the ancient language of the British", which, in this context, suggests Old Celtic.  If he could, he would never have made his howler about Ambrius: any Welsh speaker would immediately hear the name Ambrius as cognate to Emrys, and know what to think.  All his sources must have been in Latin; it is significant that he shows no knowledge of those materials which enter the written tradition directly in Welsh, such as the Mabinogi[10] or Welsh poetry.  Compared to his imitator Saxo Grammaticus, it is remarkable that he does not quote a single line of native verse (one thing of which any educated Welshman would have been proud) while Saxo not only quotes extensively but even recognizably, so that under the surface of his cleverly used Latin metres we still hear the Icelandic poetry of the skalds.  Saxo spoke Icelandic/Old Norse; Geoffrey spoke no Welsh.  But the most prominent lost item of the Welsh historical collection, L, was certainly in Latin: even if we could conceive of a written prose work in a Celtic language at so early a date, it opened with a Latin solecism - magna discrimina - that became famous.  We need not doubt that L was regarded as a classic, and, what is more, it was enduring.  Not only did Gildas, in the sixth century, react to it as one master facing an earlier one; but, most tellingly, Nennius did not treat it - as with the Genealogy and A - as an outdated archaic curiosity, but as a widely known contemporary text which he quotes but does not excerpt.

Gildas and Nennius may be full of allusions to and quotations from L that we cannot identify.  We may be certain, from Gildas, that L was at least in part a contemporary, perhaps eyewitness account, and that it concerned the struggles of valiant heroes against terrible odds; in other words, that its picture of the Saxon wars was heavily personalistic, centred on the activities of a number of named individual heroes.  (This is oddly reminiscent of the Gododdin in the emphasis on the enormous odds faced against a numerically superior Saxon enemy, its claim to be eyewitness material, and its praises and descriptions of individual heroes.)  From Zosimus, we gather that the war it described began in Britain and that it coincided with, or was followed by, a genuine political revolution in which Roman laws were rejected and Roman political leaders ejected – the latter being particularly associated with the spread of the war/revolution to Armorica and unspecified other regions of Gaul.  We may be certain that it was not part of the same document as A, the conclusion of whose story was at an earlier point in time, and which, I have argued, must have showed a grasp of Latin more close to that of Gildas than to that of a man who could apply the adjective magnum to a discrimen; also, L only knew one Saxon invasion, which led Zosimus to misattribute his account to the only Saxon raid on Britain that he knew, the famous one of 409-411, and Gildas to speak as though no Saxon had ever invaded the island before Vortigern.

We have seen that what documentary evidence there is tends to attribute to the generation of Ambrosius the character, not of innovators in the name of a "British", Celtic system, but of would-be restorers of Roman law and religion.  While the witness of Sidonius' attitude to Rigothamus and Rigocatus, by itself, proves little, and Geoffrey's notice that Ambrosius worked to restore (presumably Roman) ancient law proves even less, the fact that two such widely separate sources coincide, that they are supported by the still Roman mind shown by G, and that there is no opposite evidence of any kind, does rather strengthen their message.  Living in a fully Celticized Britain, Gildas still saw Ambrosius as the last of the Romans, looking back to the ancient empire (and overseas to the power of Constantinople) rather than forwards to a British present Gildas hated, and to a future as black as pitch.  There are therefore no grounds to attribute Zosimus' political revolution to Ambrosius; on the contrary, it seems likely that he was responsible for the restoration of those rhomaios archontas later driven out - for the attempted reinstatement of a fully Roman system.

The fact that L does not seem to know any Saxons invasion before 432-442 is also suggestive, since it seems to indicate that it regarded that one as the beginning of all things, in the light, that is, of an already ancient past.  This is very unlike other documents we have met. Both the Ambrosian file and the Vortigernid legends look back to a time before 442, are interested in conflicts and questions that mattered in the time of Pelagius, Germanus, Vitalinus and the Mild King; L seems to neither know nor care about any of these things.  He is only interested in "the dangers of most valiant soldiers in grim war" against the Saxons: this is the horizon of his mind.  This separates him not only from Roman Britain, but also from the mental world of the descendants of Ambrosius and Vortigern; for we have seen from some legends - in particular the Vortigernid legend of O, and the bits of possibly genuine Ambrosian lore I have called N2 - that there was a stage in which the Saxons were regarded as defeated for good and for all, and the Picts as merely amusing, clumsy drunken barbarians from the north, easily manipulated by a more sophisticated villain such as Vortigern; while the ancestral claims of Vortigern and his enemies were regarded with great seriousness.  And I cannot emphasize too strongly the fact that the Vortigernid documents represent a late stage of British culture, with N certainly going back to a period when the beginnings of the British state had been forgotten, and O to a literary culture no different from that of Gildas; while the Ambrosian file represents Ambrosiad claims as they were known to Gildas in 561.  In other words, there is nothing to indicate that they were written earlier than A or L; the imaginative language in which they express fifth-century Ambrosiad and Vortigernid claims is clearly not the same in which G had stated Ambrosius’ case against Vitalinus and the Saxons in the fifth century itself.

There is likewise nothing to indicate that the conflicts of Ambrosiads and Vortigernids had any interest for L: his theme seems to be the war against the Saxons, seen, probably, in the light of the heroic exploits of individual warriors; and linked with the rejection of Roman law and ways.  And in turn, neither of these themes is prominent or even identifiable in any document that can confidently be identified as either Ambrosiad or Vortigernid.  The only exception is the unhistorical figure of Vortimer, whose only activity is to fight the Saxons.  Now O is probably contemporary or later than L, dating to a settled Celticized Britain which sounds like it came after Badon Hill, rather than to a long period of uncertainty and Saxon war.  Indeed, Vortimer, no less than the whole business of building the great fortress, seems to represent a Vortigernid attempt to claim the whole theme of the great war, quite possibly designed at a time when L - as Gildas shows - was an unchallenged classic that formed attitudes.  O is written from the perspective of the great royal houses, Ambrosiad and Vortigernid; and it is worth pointing out that L seems to leave no room for what all other sources - Procopius, Gildas, N - imply to have existed - a monarch of all Britain, succeeding in some fashion from Constantine III.  To Zosimus, the poleis act collectively to reverse to "British" customs, with no reference to a superior all-British authority.  And it is L, and L alone, that stresses the multiplicity of the poleis of Britain.  It is as though the idea of Britain as a monarchy with a single identity has faded from sight; only to reappear, at least in a literary guise, by the time of O and Gildas.

A third important feature for the dating and understanding of L is his clear allusion to the British settlement of Brittany.  There is no need to postulate a fifth-century Celtic Brittany.  The character of fifth-century British intervention in Gaul was quite different.  Ambrosius, or a successor to the claim of Constantine III, must have been responsible for Rigothamus' disastrous expedition.  Its purpose was not to conquer Armorica and other parts of Gaul for a non-Roman, British Celtic system, but, quite to the contrary, to re-establish some sort of Roman authority in central Gaul.  Gregory of Tours implies that the British were part of a common Gallo-Roman and Frankish front against the Visigoths.  Sidonius Apollinaris implies that they were there to defend Roman law against ius Gentium.  Jordanes says they were there in the name of Anthemius, or at least with his approval: an ambiguous alliance between two titular emperors, each holding only a shred of the Roman West, and neither possessed of an unchallengeable title.  On the other hand, two characteristics of L's world - Armorica being ruled according to "British" rather than Roman principles, and the existence of a plurality of tribes or poleis - are clearly in charge in Gregory of Tours’ vivid accounts of his own time - 590, more than a century later.  This hiatus does not allow us to suggest that the conquest of Brittany had anything directly to do with Rigothamus and his men.

The Letter to Lovocatus and Catihernus, written in the 510s, points to an intermediate stage in which, while British tribal groups of a wholly Celtic kind were present somewhere in the dioceses of Tours, Angers or Rennes (that is, not necessarily within the bonds of historical Brittany) they had neither political power nor social prestige – to the point where the Bishops could sneer at their cabanas and threaten their priests in the most brutal way and without any fear of political retribution.  This hardly suggests the massive anti-Roman phenomenon hinted at by Zosimus, with Armorica “and other provinces” swept by an anti-Roman, Celticizing movement.  On the other hand, we are not talking of a stranded community like that of Rigothamus and Rigocatus, cut off from its British roots somewhere in Burgundy, but of a group which kept contact with the British mother land and was better informed, through Britain’s contact with the Byzantine East, about Byzantine ritual practices, than the Bishops themselves.  Christian priests do not risk excommunication, let alone the worse things with which Licinius, Melanius and Eustochius threatened them, for no reason: but here we see a conflict which has come to the very edge of a final break, because the Celtic priests were so obstinate in admitting deaconesses to the altar.  In political terms, this suggests a weak but continuing British presence in Gaulish Armorica; armed, probably, since the point of getting tribal Celtic groups to Gaul would no doubt be that they were the fighters of Britain; and not altogether disposed to go down on its knees before the powers of the Gaulish Church, even though they represented spiritual authority and were backed by the victorious Franks.

Between the time of Lovocatus and Catihernus and that of Gregory of Tours stands Procopius, whose picture of the British Isles is notoriously troublesome.  He has two separate ideas in mind: Britannia, a former province of the Western Roman emperor, the greatest island in the world, which he knows only from books and which is mainly in the past; and Brittia, the Franks' large northern neighbour, populated by Angiloi, Frissones, and by those same Brittones whose ambassadors had struck him or someone else at Justinian's court as barbarians.  Internal evidence tells us that Procopius must have had his information from Franks, certainly one of the many Frankish embassies to Justinian, probably from the ambitious Theudebert; and, crucially, they managed to convince him that Brittia and Britannia were separate entities, and that Britannia alone had been the ancient Roman province.  (They seem to have placed Britannia to the west of Brittia - in other words, to have tried to identify it with Ireland.)  The Franks, he was told, were just then allowing members of all the three nations of Brittia, which all suffered from overpopulation, to settle in Frankish-owned land[11].

It is a good question how even Procopius let his informants get away with their explanation for this[12], that they were settling people from Brittia in Frankland so as to gain themselves control over Brittia.  How does the one follow from the other?  However, the trend of the argument is clear: lands in which Angiloi, Frissones and Brittones are settled are lands which belong to the Franks.  Which is rubbish.  One thing that we cannot doubt, even with the scarce sources that we have, is that the relationship of those Britons who settled on the Continent, ancestors of the Bretons, with the Franks, were anything but friendly; the very notion of them settling in Brittany by Frankish permission is nonsensical.  And an even more fantastic lie is the inclusion of the Frissones in the claim of coming from Brittia and being settled in Frankish land: these can only be the Frisians of the North Sea, and the only "continental" land we can be speaking of is Frisia, their own homeland!  There can be no doubt that these Frankish embassies to Constantinople intended to dazzle the Greeks with false claims, that they told as many lies as they thought they could get away with, and that their claim to be overlords of Frisians and Britons was one such lie.

The only group which entertained such relations with the Franks were the Saxons/English.  It seems to be widely accepted that a certain number of Saxons, at first of Continental origin, but soon replaced by islanders, settled in some specified regions of Gaul, namely the hinterland of Boulogne (where John Morris counted an astonishing cluster of a hundred place-names which closely resemble English place-names in their most ancient settlement, East Anglia[13]), Ponthieu, the Bayeux and Caen areas of Normandy, and the mouth of the Loire[14].  This distribution may seem to simply mirror the motions of a seaborne invader, peppering favoured maritime entranceways into Gaul - the pas-de-Calais, Normandy, the Nantes area; but there are features that hint at a different story.  The settlement behind Boulogne is by no means orientated towards the sea, but rather placed across the land from west to east; and it becomes clear that, strategically, it stands between the early-settled Frankish lands of the Low Countries - largely equal to today's Flemish Belgium and southern Netherlands - and the part of Gaul which was still Roman in the third quarter of the fifth century, before the start of Clovis' conquering career.  The fact that most of this territory is today Walloon rather than Flemish, means that the Saxons were settled as a minority, surely armed, among a dense Roman population whose language they eventually absorbed; and the dimension of the settlement - a hundred villages! - shows that this was a major event.  In other words: a great number of Saxons was settled, at some point in the fifth century, in a border area facing the Romans of northern Gaul.  Even before Clovis, the Franks had a very volatile relationship with these people, one moment allies against encroaching Visigoths, the next targets for invasion.  If the account of the Life of Ste. Genevieve of Paris is historically based, then, long before Clovis, his father Childeric was liable to occupy the Paris area militarily, judge and condemn prisoners, and exact onerous dues from the locals - in fact, behave like a conqueror - though he was also open to persuasion by the saintly lady.

The best way to make sense of this is that the Franks took a leaf out of the Romans' book and treated the English as foederati, settling them across a dangerous frontier; ironically, the very same use Vitalinus had thought of finding for them.  But the Boulogne area ceased to be sensitive when Clovis finally became the uncontested master of north Gaul in the 480-90s; there would be no reason to place foreign tribes on it afterwards.  It follows that if any Saxons from Britain were settled there, they were settled before 486.  And the conclusion forces itself on us that the only reason for large numbers of Saxons to leave Britain in this period must have been Ambrosius' victory.  It may have helped that the Franks, at the time, were still pagan, and religion certainly was an issue between Britons and Saxons.  Some tribes may have fled Britain to escape forced baptism or worse; if so, the vast number of settlements - about a hundred villages! - suggests that Ambrosius' victory was a major shock.  For that matter, Morris argues that some more Saxons were sent to the Thuringian border; and we don't even have to believe that all the Saxons who fled the island went to the Franks!  It must have been quite a migration.

This is also the time when the sudden reawakening of Saxon piracy startles the tired and aging Sidonius Apollinaris, reawakening dreadful memories, as he is writing a letter to another Roman who, like him, had had to swallow the bitter pill of Visigothic power and had become their admiral.  It also follows the great war of 468, in which Saxons – it is not clear whether continental or British – came to the mouth of the Loire to join the party opposed to the alliance in which the British were involved.  This suggests a first massive shock - during which a good many Saxons simply flee the conquering British Christians to pagan Frankland - followed by a great recrudescence of conflict of which "Saxon piracy" and the Saxon participation to the war of 468 are two symptoms.  It is impossible to be more precise than this, and it may be that I have already leaned further towards the Morris type of historical fantasy than I would care to; but the various pieces of evidence outlined show, at least, that something happened; and nothing I suggested is in the least unlikely.

Now, unlike the Boulonnais tract, the other Saxon areas of settlement in Frankish Gaul - Ponthieu, Normandy, lower Loire - face the sea in the general direction of Britain and Ireland; and they bracket Brittany.  The Franks themselves, in spite of their connection with the river-borne Rhine trade, were not particularly a maritime people, and most of what trade there was - and in the fifth and sixth centuries there was almost nothing - passed through their neighbours the Frisians.  But their relationship with the Frisians seems to have been quite as tense, ambiguous and yet mutually dependent as with the Romano-Gauls; if they wanted to defend their new conquests from seaborne intruders, they might not want to trust to Frisian settlers - it might make the Frisians too strong.  Procopius' legend of the English princess invading the land of the Varni from the sea with 100,000 soldiers shows that the Franks who told it to him envisaged the English as capable of crossing the ocean in strength; the detail of their ships having no sails but depending exclusively on rowers agrees with Sidonius’ description as well as with archaeological evidence[15]; and if the Franks regarded the English as mighty sea warriors, it would be easy to think of them for a seaborne defence.

The reason why they would want a seaborne defence is hinted by a number of inconsistencies and obvious lies in Procopius and Gregory of Tours.  Procopius was told that the Frankish kingdom originated in an equal fusion of three free peoples, Franks, Romans and Arborychi (obviously, Armoricans), based on a common Catholic faith: the kingdom's unnamed Frankish founder was said to have married an Arborychan princess and then intervened with the forces of both peoples when the Catholic Romans of Gaul were threatened by Arian invaders.  Procopius was told that the integration of the three nations was so complete that Romans units served in the Frankish army with their own panoply of weapons[16].  In point of fact, a good deal of this was nonsense; in particular, the integration of Franks and Romans was at least rather defective, and Frankish law codes attributed to Romans a lower rank and honour price than Franks.  There is no record of an Armorican princess marrying any Frankish lord, and the story seems to be a very curious melange of the role of the still-pagan Franks in driving out the Arian Visigoths in 468 along with a valiant but defeated British army, the marriage of the pagan Clovis with the Catholic - but certainly not Armorican - Chlothilde, and Clovis' invasion of Aquitaine in 507.  As memories of Clovis were apt to be still green, one can only wonder at Procopius' informants.

But this is by the by.  The significant element is the importance given (at some point in the 540-550s) to the political role of the Arborychi.  They are described as equal partners of the Franks in the work of defending the Catholic faith in Gaul in other words, as joint protectors of the native Romans against the Visigoths - and as an equal third in Gaulish ethnic realities along with Franks and Romans.  The only nation this can refer to are the Britons of Armorica; and it follows, one, that their political importance was far greater than what we gather from Gregory of Tours or the Pseudo-Fredegarius, and, two, that it spread over as much of Gaul as the Franks themselves.

Suddenly we are reminded of Zosimus' claim that the anti-Roman revolution spread not only to Armorica, but to vague and unspecified other regions of Gaul (and let us bear in mind that Armorica included not only Brittany, but also the lower Loire and present-day western Normandy; an area larger than Armorica must have meant a considerable share of the whole of Gaul).  We have seen that this refers not, as Zosimus thought, to the early fifth century, but to some generations later; now, let us see what it looks like in the light of these suggestions.  "As they advanced, the barbarians from beyond the Rhine gained control of everything, and brought the men of the British island and some who dwelled among the nations of the Celts [oikontas twn en keltois eqnwn enia] to the necessity of rejecting/revolting against [aposthnai] the Roman principles/government [archs] and live in their own way without submitting to their [the Romans'] laws.  Those in Britain therefore armed themselves and ran terrible danger to free their ciuitates from the menacing barbarians, and all Armorica and other provinces of Gaul imitated the British and likewise made themselves independent, threw out the Roman magistrates [arcontas] and set up their own independent government."

We have already seen that this describes a rejection of Roman ius - archs - in Britain, followed by the defeat of “barbarians dwelling above the Rhine” within the territory of Britain itself; after which the revolution spreads to Gaul, and only then do we hear of actual Roman rulers - arcontas - rather than Roman “rule” or “principles” - archs.  But there is another inference: that the spread of the struggle to Gaul, although it resulted in the expulsion of “the Roman rulers”, was somehow connected with the struggle against “the barbarians from beyond the Rhine”.  And I would add that that curious expression, “the barbarians from beyond the Rhine”, which does not altogether suit the Saxons who reached Britain, is in fact an excellent description of the complex of separate Germanic tribes - Saxons, Franks, Alamans, Burgundians - who, between the mid-fifth century and the sixth, overwhelmed Roman Britain and Gaul.  Far from being a vague and poetic expression, 'hier griekse text!!! seems therefore quite a clear definition for those who “gained control of everything” on both sides of the Channel.  In other words, L contained an account of a British-led, Celtic-minded invasion which competed with Clovis and his successors for control of the fertile plains of northern Gaul and which intended to destroy residual Roman institutions there.

Historians have come to understand that the Frankish conquest of Gaul is really a rather obscure process: even such a capital event as Clovis the Frank’s conversion to Catholic Christianity can be dated with as much as twenty years' margin of error (indeed, Roger Collins thinks he may never have been converted at all![17]).  Geographically, the territory that is now northern France was not part of the Frankish lower Rhine heartland, and we must imagine the territories west of, say, the Somme, as rather a Frankish colony than a familiar land.  Attempted penetration by a rival power is certainly not hard to imagine; in fact, that was what the Visigoths were up to until Clovis crushed them.  And the many place-names Bretteville in north-west France, especially in Normandy, do seem to testify that the people known to the Frankish sixth century as Britti settled in numbers and considerably beyond the border of Celtic Brittany[18].

The majority of Brettevilles are found in the Bayeux region and in Cotentin, one of the areas settled by Saxons; we know that Saxons and Britons were, in the same way, both settled at the mouth of the Loire, which later became part of historical Brittany.  It seems likely enough that the Saxons were placed there to keep local British entities down and outside British powers out. Even so, one event recorded by Gregory makes us curious about their relationship: it seems that, during the great Frankish raid against the British chieftain Waroch in the late 580s, the Saxons settled near Bayeux put on British dress and fought on Waroch’s side, helping to inflict a frightful defeat on the Frankish commander Beppolen, who fell on the field[19].  Gregory attributes the Saxons’ treachery to the Frankish queen Fredegund , whom he hated,, but his testimony is contemporary and as reliable as these things ever are.  Whatever the case, this swift veering to the side of the British by those who should have been their hereditary enemies asks all sorts of questions about the relationship between Saxons, British and Franks; but it does nothing to contradict the suggestion that these Saxon settlements had a particular connection with the British presences in the same area.

The British presence in Armorica is unmistakably Celtic, with no admixture of Roman culture or poetry whatever: it is a direct descendant of the proto-Welsh culture which, I have argued, spread south from the debatable lands beyond the Wall.  Some features of early British culture may have been preserved more faithfully in Britain than in Wales; take for instance the distinction between high kings and teyrnedd, which is implicit in a number of quite late ecclesiastical sources, with the higher lord of Brittany being routinely described as consul and his enemies as tyranni.  The long survival of elaborate versions of the legend of Taliesin, a caste legend of bards - the fables of N'oun-Doaré and Koadalan, collected in the nineteenth century[20] - shows that the class of bards reached and rooted itself in Brittany; Brittany knew the great caste legend of the bards, and it follows that it knew the bards.  In other words, all the typically Breton cultural and political institutions (bardism was both) pertain to an advanced stage of the resurgent Celticism that can only have developed after the death of Ambrosius, and some of which (wars between independent kinglets, cattle raiding, bardism) Gildas saw as unwelcome and recent developments.  I say that all these facts point to one conclusion, and one alone: that whatever the relationship between Armorica and Britain in earlier decades, the British settlement should not be dated earlier than the sixth century.  And as we have seen, while the affair of Lovocatus and Catihernus shows that British tribal entities were present in north Gaul by the 510s, it also shows that they had nothing like the power of the “Counts” of Gregory of Tours’ time, let alone the reach suggested by Procopius and L.  It follows that the events described by L took place, if they took place at all, after Lovocatus and Catihernus, but before Gregory’s account of the British “Counts” of his time.  Perhaps some of the tribal entities of Lovocatus and Catihernus served as a fifth column.

In Gregory's history, the Brittones of Brittany only become visible in 544.  His only item from Brittany or its environs before 544 is the life of St.Friardus, a hermit in the Nantes area, which he includes not in his History but in his collection of Lives of the Fathers; and that has its own ideological point.  To begin with, the name of Friardus is unmistakably Teutonic, probably Frankish; the fact that a member of the conquering northern horde, settled in Roman Gaul, was able to develop the eminent Christian qualities shown in Gregory’s brief Life conveys a strong message about the future of Franco-Roman relationships (another of his Fathers, St.Senoch, was a member of the notorious raiding tribe, the Theifali, who were settled around the modern village of Tiffauges, not far from Nantes).  The whole point of Gregory’s Lives is to show that Gaul and Frankland have had their own Fathers, as high, as saintly, and as much in the bosom of the Lord, as the great names from Christian antiquity; and many of them – such as Friardus and Senoch – were Gregory’s contemporaries.  It also shows a man of Teutonic blood blessing with his presence - for the presence of Saints was regarded as inherently a blessing - an area coveted and partly settled by Britons.  At the same time, Friardus is very humble, a typical man of the people, as far as anyone can be from the values of the warrior aristocracy; as if to say that if you look away from the unpleasant and often monstrous realities of aristocratic conflict in that area, you can find lives of unblemished nobility and saintliness.  (There is also a swipe at Gregory’s least favourite colleague, Bishop Felix of Nantes, who is gently chided by the dying Saint for keeping him waiting.)  In the light of what we are seeing of the conflict between Britons and Franks, I doubt that any of this is casual; Armorica was, at the time, full of British monks, some of whose Lives have come down to us, and at least one of whom – Samson – was well known to the Franco-Gaulish church; but Gregory does not mention a single one of them, except for the wretched Winnoch, who, after showing signs of sanctity, lost his mind and died drunken and insane.

He does, however, mention a British bishop of Vannes; a scandalously usurping one.  In 544 the British lord Macliaw (whose name is the same as that of the later monk and bishop St.Malo), usurps the bishopric of Vannes to protect himself from the murderous attention of his brother Chanao, who has already accounted for three other brothers - 4.4.  There is a wildness, a rootlessness, about the adventures of these Breton lords, Boudic, Conomorus, Macliaw, Chanao, Waroch - ranging all over the peninsula and beyond, killing each other, exchanging lands and subjects - that suggests a first generation of settlers, not yet rooted in the country they have conquered; certainly without any of the signs of ancient and untroubled rule which we should expect if their presence in Armorica went back to the previous century.  The way these local British lords fight each other for control leaves the impression rather of the absence of any superior British authority, with local proconsuls trying to carve out their own territories; it is consistent with a power vacuum; and it is notable that by the time Guntram set up his ill-fated Breton expedition, the factional fighting seems to have died down, and the Bretons face the divided Franks as a united body politic (10.9), to the latter's great discomfiture.

They have, however, been there a while: ...semper Brittani sub Francorum potestatem post obitum regis Chodovechi fuerunt, et comites, non reges, appellati sunt - "after king Clovis’ death [511AD], the British were always under the power of the Franks, and were called counts, not kings".  The word "Counts", in Gregory's post-Roman world, still retained much of its original bureaucratic significance and stood for an office that the king could bestow and remove.  That is, Frankish kings after Clovis claimed that the violent and undisciplined British lords who held Armorica with armed force, indulged in feuds that were unpleasant even by Merovingian standards, and regularly raided Frankish territory, were still functionaries - "counts" - in the post-Roman administration of Frankish Gaul, supposedly reporting to the Frankish King.  Given the undoubted hostility between Franks and Armorican Britons, this can only effectively represent a concession on the Franks' part, legitimizing the British military presence in Armorica.  The Franks would not recognize that the British kinglets were "kings" - legitimate, independent sovereigns on the same level as the Frankish kings, images of the Emperor in their own land - but granted them the title of "counts", direct representatives of the King's sovereign power (the word means "companion", that is companion of the king or emperor) not unlike, indeed, the way in which late Roman emperors granted Roman military titles to men who were in effect independent barbarian kings at the heads of war-bands.

This isolated note, however, comes with the description of wild internal British feuding in 544, and it sounds almost as though Gregory is trying to squash any notion of actual British independence in Brittany by invoking a much older precedent – one from the age of Lovocatus and Catihernus.  We do not doubt, from what we have seen, that any British presence at that time would indeed be sub Francorum potestaem; but the uncontrolled adventures of the Bretons of 544 tell a different story.  Nor should we forget that the tribe of Lovocatus and Catihernus does not seem to have even resided in the borders of historical Brittany, but at some point between Tours, Angers and Rennes; territory that, in their time, indubitably pertained to the Franks.  Licinius’ enthronement in 508 is suspiciously coincidental with the Frankish conquest of Aquitaine after the battle of Vouillé, 507; we have seen that Tours had been held by the Visigoths in the time of his predecessors Volusianus and Verus.  There is no evidence whatever that any arrangement had been reached, or could have been reached, between the Franks, and any British settlers in Brittany proper; or that any such settlers were there at all.

And please note that what Gregory is doing is trying to sneak an answer to an implicit question whether "the Britons" of 544, as a whole, should be called "kings" or "counts"; as if there were no Britons in Armorica except sovereigns.  If "the British" - as a whole! - had been successful against the Franks, they could have been called kings, but, as they were not, they were - again, as a whole - called only counts.  The Armorican Britons known to Gregory were definitely an alien, conquering aristocracy.  Gregory - who is rarely so consistent elsewhere - consistently regards the "British" element in Armorica as being made up purely of the royal or comital houses that had gained power there: dynastic, small and intrusive.  And as he was theoretically the metropolitan of the Armorican British, and as in any case his diocese of Tours was not far from the front line and witnessed the by-products of war with Brittany more than once in his episcopate, we must take his testimony as significant.  He tells us that his contemporary Bishop Regalis of Vannes described them as masters which the people of Vannes, not excluding himself, "have to obey"; for that matter, in the 540s Macliaw had actually usurped the bishopric.  This could not be more at odds with the picture of British status and power implicit in the letter of Licinius, Eustochius and Melanius.

It is therefore possible to smell some fish - quite a lot of fish - about this title of "counts"; both its significance, and its date.  Gregory is of course best informed about his own period (570-590 approx.), in which we need not doubt that the Frankish view of the legal status of the Armorican British lords was what he reports.  But a few decades earlier, Procopius’ supposed dynastic marriage of Frankish king and Armorican princess shows that the Frankish emissaries who made it up saw the "Armoricans" as equally royal as the Franks, and all of Gaul as their field of activity.

Gregory's dating of the title of "counts" is vague and non-committal.  He does not report who gave the British such legal recognition, how, or when, though he knows for certain that it had not been granted, even provisionally, until at least 511.  Now if the British in Armorica were regarded as royal in the 530-540s by the Frankish envoys to Byzantium from whom Procopius gained his information, this means that the doctrine that they were "counts" of the Frankish Kings did not gain the universal currency assumed by Gregory till later; in which case, the vague post-511 date probably amounts to an attempt to back-date a much later political agreement, invoking, perhaps, some vague or partial earlier deal.  To push the argument further, it is possible that some such agreement had been reached with the minor tribal entities testified by the Lovocatus-Catihernus affair, only to be broken and forgotten (until much later) when the Arborychi of Procopius became, in ways yet to be clarified, a much bigger and indubitably royal presence.  Once they were bottled up in Brittany – as they certainly were by Gregory’s time - the Franks had every reason not to preserve any record of a rival claim to a recently conquered territory that was essential to their political power and, under the name of Neustria or Westland, often became the exclusive appanage of one or more of their crowned kings in the successive partitions that cursed their kingdom.

Therefore the Franks did not only lie to Procopius and his fellow Byzantines; they lied all the time, consistently, and as a principle.  The partiality and outright mendacity of their records may be judged by their absolute refusal to mention whatever dealings Clovis and his successors had with the British.  That they did not dare to back-date the title of "counts" to the reign of their great founder, otherwise represented in Merovingian accounts as the ultimate in royal valour and triumph, obviously means that Clovis must have dealt with them not as "counts" but as kings; and his successors two or three decades later treated the Arborychi not only as royal but also as involved with all Gaul.  Therefore, that no record of any such incident, whether peaceful or - much more probably - military, can be found anywhere in Gregory of Tours, the pseudo-Fredegar, or any other source, means that all such information had been deliberately suppressed.  And this means that there was something to hide: if Clovis had been as victorious against the Britons as he was with Goths and Alamans, there would be no reason to conceal the fact.  If Zosimus and other clues strongly hint at wide-ranging British intervention in Gaul some time between the time of Lovocatus and Catihernus (the 510s) and 544, we are not allowed to deny the suggestion on the evidence of clearly deceitful Frankish records; if anything, Frankish untruth means that there is something to hide.

Let us also remember that Procopius' legend of the marriage of a Frankish king and an Armorican princess was Frankish propaganda designed for self-aggrandizement; and what is perhaps more cogent, to warn the aggressive Byzantine power of Justinian off Frankish spheres of interest.  The Frankish claim of fostering the intrusion of this ambitious, troublemaking minority into territories they regard as their own is flagrantly a lie ad usum Constantinopolitanorum.  The importance of what Procopius calls the Arborychi is therefore truly remarkable: they are described as potential equals of the Franks in Gaul, whereas the local Romans are objects of conquest - or rather, "protection" against Arian Goths.  By the 530-40s, the Franks wanted the aggressive Justinian to understand that they had an equal agreement with the Armorican British to defend Gaul from the Arian Visigoths - the sub-text being, clearly, that there was no need for any Byzantines to butt in.

Gregory of Tours never describes the Britons of Armorica as more than squabbling provincial lords.  This argues that by his time, British ambition such as led the Frankish ambassadors to Byzantium to describe the Arborychi as an equal partner in the defence of all Gaul had been well and truly beaten back; but Gregory is amazingly silent about Britain.  Of all neighbouring countries - from Denmark to North Africa and from the Suebi of north-west Spain to Armenia - Britain and Ireland are the only two of which he has nothing, but nothing, to say.  In ten extensive books, he barely refers to the great islands, so close to Tours.  But the History itself contains evidence to suggest that Gregory was in touch with British and Irish matters: he based his account of Joseph of Arimathea (1.67) on the so-called Gospel of Nicodemus, an apocryphal item which, in his time, was known only in the Greek East and in the British Isles (it is said, I don’t know on what authority, to have been condemned by the Papacy and its distribution discouraged[21]).  This proves nothing, for Gregory might as well have received this book from an Eastern as from a British source (an exiled Armenian bishop was his guest for a while); but if the “gospel” had in fact been condemned, a well-educated Easterner would be likelier to know it than a provincial Briton, who, with the best intentions in the world, may have been less well informed.  Indeed, we may be sure that "he" was - simply because the "Gospel" was in fact circulated in the British Isles, with no visible doubt as to its orthodoxy, for long enough to affect the formation of the Arthurian legends, centuries after Gregory.  But whether or not Gregory knew more of Britain and Ireland than he let on, his account of the British colony in Gaul can hardly not have been affected - and distorted - by what is all too clearly a deliberate silence about the mother country.

During the same period, the acts of the Second Council of Tours (567) deal with the scandal of British clergymen crossing the Channel to Armorica be uncanonically ordained bishops.  This is Gildas' scandal of irregular episcopal ordinations, confirmed by an unchallengeable official document; but notice the wording of the law enacted by the Council: "We add also that, in Armorica, neither Briton nor Roman shall be ordained bishop without the consent of the metropolitan and his provincial colleagues". As E.A.Thompson[22] rightly saw, this means that in 567 "British newcomer and Roman native had not yet fused in Armorica".  Twenty years later, in History of the Franks 10.9, Gregory of Tours describes his contemporary Bishop Regalis of Vannes and his people welcoming the troops of the Frankish king Guntram as liberators, and Regalis telling the Franks that "We have to do as the British tell us, and that irks us very much".  That is, Regalis and his people were not British or Breton at all, and the Brittones were in effect a group of conquering trouble-makers whose fundamental alienness was still evident to everyone.[23].  This may be Regalis' diplomatic talk ad usum Francorum; but, at the very least, it cannot be completely divorced from reality, or else the Franks - who, according to Gregory, had suffered much from British depredations and treachery - would be expected to disregard it and take revenge on Vannes.  But they don't; instead, they accept the homage of the Vannetais with every evidence of pleasure.

This does not agree with Zosimus' picture of Gaulish provinces following the British lead against ius Romanum of their own free will.  But then, we would not expect it to.  As I argued that sixth-century British culture starts from the Scottish borders, the long distance from the Wall of Hadrian to the Armorican peninsula can only have been negotiated by conquest and invasion; what would probably be a popular movement north of York would not be apt to be anything but a foreign conquering horde across the Channel.  Even long-time Celtic British settlements such as Lovocatus and Catihernus’ might not necessarily welcome them.  Certainly L would have features of self-excuse, just as, on the other side, Gregory of Tours may have exaggerated Regalis' welcome to the Franks.  L sees the relationship between Britain and large tracts of Gaul undergoing a sea-change, in the direction of much greater closeness, during and on account of the revolution: in other words, and stripped of the revolutionary double-talk which - as we have seen in the twentieth century - is quite capable, indeed eager, to misrepresent a foreign invasion supported by a fifth column as a native and popular decision, this means that Celticizing British forces entered Armorica and other parts of Gaul en masse, enforcing their own political system.  The existence of a smaller British colony in north-west Spain suggests deliberate large-scale action, a British naval policy aimed at controlling the western approaches, looking towards Gibraltar and Byzantium; and if Britain committed itself to an aggressive policy of naval expansion south-westward to secure a sea-link with the heart of the Roman world, this would explain how, in distant Constantinople, Zosimus got to hear of - and garble - it.

The British who colonized Armorica and competed with the still barbarous Franks for the control of northern Gaul were no less barbarian – I mean non-Roman - than they. To quote myself, “Breton cultural… institutions… pertain to an advanced stage of resurgent Celticism”. Everything moves L away from the days of Ambrosius, and from the kind of environment he must have moved in, and into the sixth century. Annals and accounts can and do lie; but a man cannot lie about his basic social assumptions, about the things he does not even think about because they are as natural to him as the air he breathes. And Roman social ideas were obsolete, in Britain, by the time of L. The superior value of Roman citizenship was obvious to Patrick; it was not to L or A. The direct connection between learning and high social position was obvious to Patrick; it was not to Gildas or to Taliesin. L associates the political revolution with the decision of Armorica and other continental provinces to follow the British into war. This cannot be associated with, and can only have come after, the last Romanizing attempt of Ambrosius and his followers: it is to be associated with the triumph of the highland Celticizing forces.

As late as 468 Britain had a Romanizing government disposed to aid the Roman Empire against Arian barbarians, if perhaps for its own purposes. Yet by the time Gildas wrote, the resurrection of Celticism was complete across Britain except for the language. The king of Britain to whom he appeals is not a Roman emperor employing the machinery of the State to crush a rebellious subject, but a Celtic high king going to war (as Gildas hopes) against a recalcitrant tributary king. Gildas even imagines the Roman Empire of old in the same light; which tells us that by his time the idea of Roman Empire had become completely integrated in the purely Celtic view of monarchy. It had surely done so through the incorporation of such ideas in the imperial/royal succession, so that the high kings of Britain kept the claim to be Roman emperors but became effectively Celtic overlords; so that a man who was only familiar with the local "Roman Emperors", would inevitably project their kind of politics on to the Empire of old.

The revolt against Roman ius had completely triumphed. I date Gildas' great work very late, about 561; his visibly unhistorical and Celticizing idea of the Roman Empire, expressed with never a doubt, shows that, as far as he was concerned, resurgent Celticism was fully in charge. The author of L had a sufficiently clear idea of the difference between Roman and British/Celtic ius; Gildas had not, except in a literary sense - his ius is all of one colour. We must assign L to a generation early enough to have influenced Gildas and Zosimus in the sixth century. And finally, I would say - though this is a subjective rather than a historical argument - that L must be dated close enough to Gildas' youth for him to regard it, not so much as a work of the past, with which one does not quarrel, as much as the measure against which a young author of great promise would measure himself - which argues a modern rather than an ancient classic. Young authors do not want to emulate Shakespeare; in our day, they want to emulate Martin Amis, or at most Dylan Thomas.

The dating of the Lovocatus and Catihernus affair shows that we cannot place the revolt as early as the 510s; Gildas’ relationship with L shows that it was recent but established history in 561; Procopius’ tale of the powerful Arborychan princess suggests that it was at its height in the 530-540s. Now, we have seen that while Geoffrey's genealogical history was no better than that of the inventor of the House of Constantine - to which, indeed, he added several freaks of his own - his chronological framework is surprisingly good and suggests the survival of some sort of moderately accurate record. What we have seen of his chronology of Vortigern and Ambrosius tells us that he (or N) had access to a basically sound year-list or account, but to no acceptable king-list. Aware that certain kings - Constantine II(I), Vortigern, Ambrosius - succeeded each other, he had however no idea of their family relationships (and nor, of course, do we); and the chronology we can extract from his adaptation of N has a rather simple and bare air, which was of course never going to satisfy such an extravagant literary talent. Therefore he strung on it all sorts of stories, drawn from all sorts of sources, from N to heroic fables about Eldadus, Eldol and Hengist (unless of course he wrote those himself), to the dindsenchas-style item about the building of Stonehenge. Indeed, it was probably his instinct as a story-teller, his taste for a good tale, that made him prefer one item over another when he came across two alternative and incompatible items. He must have known, at least, Aelle's name and his claim to be the first Breatwealda; but finding himself faced with a vigorous and already ancient narrative account of a quite different leader of the English against the British, he chose Hengist as a suitable villain, granting him even a good deal of martial vigour (Eldol's capture of him is clearly a great deed).

My point is that there is no reason why his sound chronological source should stop at Ambrosius. The period 470-560 was not always without history: sufficient, perhaps even abundant documentation existed - and is lost. L existed. And N does not stop at Uther Pendragon; it would not be possible, because Uther Pendragon, in that as in every subsequent source, is only a prelude. He is there to have children. In N itself, his position in the trifunctional triad of brothers is conditioned by the fact that, unlike them, he is fertile: he has a son and a daughter. And that son is Arthur.


[1]It is particularly significant that, though a number of Celtic Latinists had made use of variations on Tethys as a poetic name for the sea - muro…tethico in the Vita sancti Winvaloei, (Analecta Bollandiana VII); Per meta tythis ignoti orientalis circuli in a hymn to St.Columba, in medio Tethys in the Life of St.Teilo - Nennius seems to be the only one to repeat Gildas’ trans tithicam vallem word for word.

[2]It includes such things as the real name of the legendary Pope who sent legendary missionaries to the legendary king Lucius: “Pope Eucharistus”. Bede regularized this to identify him with the historical Pope Eleutherus, but Nennius ignored his correction. My view, by the way, is that the theory which originated this particular legend in a misreading of Papal Roman archival material relating to “Bithinia” rather than “Britannia” is mistaken: the legend was invented in Britain.

[3]See Appendix XII.

[4]This has a direct bearing on the question of Geoffrey’s sources. Nennius and Geoffrey, we know, share a largely common world chronology. If what Nennius is doing is fitting foreign fragments into a picture of history, then the picture of history in question – the world chronology – must have pre-existed him; in other words, Geoffrey and Nennius took their picture of world history from a common original. I have suggested (above, bk.7, ch.2, note 14) that the mention of Huns alongside Picts as one of Britain’s two enemies came from a notice originating in a brief period in which the Saxons of Britain were tributary to the Hunnish Empire of Attila, misdated by Geoffrey; this makes that suggestion slightly likelier, by reinforcing the theory that Geoffrey really did have a pre-existent chronology on which he fitted other, non-chronological notices such as the one I postulate. And in that case, like Gildas elsewhere, he got it wrong.

[5]Actually, post-Roman British history paid too much attention to the existence and status of urban centres. Given that the things we call hill-forts or royal fortresses could be centres of economic life and industrial production, permanent residences for quite a number of people, and market and administrative centres for the outlying population, I think that the difference between a hillfort and a minor Roman centre would be quite small. Likely enough, the Celticizing party at least would see no distinction. It is only in the case of massive centres of population such as London, where people were counted by many thousands, that a difference may have been perceived.

6]Except perhaps for the anonymous author of Cullhwch and Olwen, as I argue in the next chapter.

[7]In the case of Bede and Gildas there is also the obvious quality of their writing, but that is a far lesser consideration: in the case of many of our lost sources – A, L, N1, N, and O at least - we have been able to suggest considerable literary abilities, which did not save their work from being lost. In the same way, three of the five Welsh poets mentioned by Nennius as classics are lost, without a single line surviving. Literary genius was no guarantee against extinction.

[8]Introduction, in GEOFFREY, op.cit., Harmondsworth 1986, 25.

[9]NEIL WRIGHT, Geoffrey of Monmouth and Gildas, in Arthurian literature 2 (1982), 1-40; Geoffrey of Monmouth and Gildas revisited, ibid. 4 (1984), 155-63; Geoffrey of Monmouth and Bede, ibid. 6 (1986), 27-55.

[10]As his later Vita Merlini has all the hallmarks of a product of the Welsh tradition of hagiography - only applied to a bard rather than to a saint - it is possible that he learned it later; but of course Welsh hagiography would itself be written in Latin. The only reason to suspect a vulgar-language origin for the story of Myrddin is, exactly, that he was a bard and not a monk - and therefore strongly connected with the Welsh language.

[11]PROCOPIUS, The Wars 8.20.9-10; cf.Appendix IV.

[12]There may be a psychological reason for this. This passage is part of his "official" history, which Procopius was writing with an eye on his terrible patron, concealing all the time his innermost rage and loathing; and the way he has of inserting any amount of ethnological notices, whether or not they are any part of the main argument, suggests to me a conscious or subconscious need to say something - anything - rather than the distorted course of half-truths he had ended up having to produce for a man he regarded as Satan. (Of course the business of inserting ethnological notices as parts of a large history goes back to Herodotus: but Herodotus made it all a part of the course of his argument, and it just does not seem to me that Procopius manages that. There is something very irrelevant about, for instance, the whole story of the Varni war.) It is possible that the wild stories of the Franks, true or false, may simply have afforded him mental relief from the pressure of constant lying.

[13]Age of Arthur, 287-291. I disregard Morris' views about the settlement of Saxons on the Thuringian border, not because I reject it, but because I have no time to test its claims, and because, even if it were true, it would not add much to my theory.

[14]STEPHANE LEBECQ, The question of logistics, in RICHARD GAMESON (ed.), Saint Augustine and the conversion of England, Stroud 1999, p.52, and authorities in notes 8-10; MORRIS, Age of Arthur 286-292.

[15]LEBECQ, loc.cit.

[16]History of the wars, 5.12.1-20

[17] - but rather that he was Christian from the beginning. COLLINS, Early Medieval Europe op.cit., 111-115.

[18]MORRIS, Age of Arthur, 89 (map 4), 90, notes 90.3 (p.557) and map 4 (p.624). It goes without saying that I reject Morris’ chronological scheme, though his remarks about British bishops and bishops with British names in Gaul are of enormous interest (note 90.3).

[19]GREGORY, op.cit. X.9.

[20]F.M.LUZEL (ed. and tr. Derek Bryce), Celtic folk-tales from Armorica, Llanerch 1985.

[21]GEOFFREY ASHE, King Arthur’s Avalon: the story of Glastonbury, London 1973, 58. Interestingly, it was not only to become widely known in the West in the twelfth century, at the time of the sudden popularity of Arthurian romance, but to form the backbone of two of the most important Arthurian items, Chretien de Troyes’ Knight of the Cart and Robert de Boron’s Roman de l’Estoire dou Graal; in other words, it seems very likely that this text, until then kept in the narrow area of British and Irish monasteries, had deep and major connections with the Breton national legend.

[22]Procopius on Brittia and Britannia, in Classical Quarterly 30 (1980), 504.

[23]GREGORY, op.cit, IV.4, V.16, 26, 29, 31, IX.18, 24, X.9.

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

Comments to: Fabio P. Barbieri

VortigernStudies is copyright © Robert Vermaat 1999-2007. All rights reserved