I will now risk my neck on the perilous (to me) ground of textual criticism. I submit that, in the first sentence of Nennius ch.30 (tribus uicibus occisi sunt duces Romanorum a Brittannis), a Brittannis must be a blunder. In the rest of the chapter, the British are called Brittones, plural of our old friend Britto, and the natural form for "by the British" would be a Brittonibus. The form Brittannis suggests the first declension ablative plural, Brittanniis, "by, from, the Britains".

Britanniae was the late-Roman word for the four or five provinces of Britain. It was obsolete by the time of Gildas, who always speaks of Britannia in the singular. As for the preposition a, "from, by", it is either a blunder for in or apud, or else a sign that something has dropped out, and that the original said something like tribus uicibus occisi sunt duces Romanorum et pulsi a Brittanniis, "the leaders of the Romans were slain three times and expelled from the Britains". A palaeographer I am not, and those who can judge of these matters may decide that I have seen something that is not there; but I offer this view for their consideration, the more hopefully since it seems to chime in with the most unlikely of parallels. If ch.30 tells us that "the Roman leaders were killed in the Britains", giving no account of the method, but insisting on the place: there is the fact that The dream of Maxen Gwledig, of all things, has Maxen vanish in Britain so that the lords of Rome think he is dead. Roman leaders, rectores, duces, reges, and ultimately an Emperor, go to Britain - and vanish. But this parallel only holds up if we read that opening sentence as tribus uicibus occisi sunt duces Romanorum in Brittanniis, emphasizing the place rather than the people.

If I am right, then, apart from comparative considerations, the most important fact to be drawn from this is that the unknown author was referring to Britain in a way that was already obsolete by Gildas' time. Britanniae is a clear marker for the date of a text or item: fifth-century writers - St.Patrick, Constantius of Lyons, the Gallic Chronicles - all speak of the Britanniae, while no sixth-century writer ever does: Gildas, Gregory the Great, Isidore of Seville, Procopius, Aponius, Cassiodorus, Zosimus, Jordanes (except for once when he is quoting from an earlier source) and Venantius Fortunatus all use the singular, Britannia, Brittia, or any one of a dozen spelling variations.

It is particularly interesting to find that St.Patrick's biographer Muirchu, in the late seventh century, makes the same mistake I postulate as having happened before A reached Nennius: he uses Britannis twice where he was certainly reading an ancient Britanniis.^[1] Writing a century earlier, Gildas uses the singular without exception; yet I have shown that he must have had fifth-century sources, possibly including the Narratio de Imperatoribus..., which must have used the plural. He consciously turned fifth-century Britanniae into sixth-century Britannia; he was aware of the change of terminology, and consciously modified his source material. The same is probably true of Procopius. Muirchu, however, writing when Gildasian Britain was a fading memory and England an imposing and Catholic reality, did not understand the difference, and misunderstood Britanniis as Britannis.

The author of A used Britanniae. The correspondences between Nennius and Gildas show that A was written in Latin; clever Latin, to judge from the ius/iugum/iuramentum pun. A must be older than Gildas, who wrote in 561. On the other hand, though its writer or writers was/were familiar with the fifth-century plural Britanniae, he/they knew nothing of the real and permanent presence of Roman troops in the north of Britain during Roman rule, had no idea that the Romans had ruled in Britain until 407, and thought that the Rescript of Honorius was a response to an appeal from a separate entity Britain. One of the triads that form A’s narrative backbone is "three Roman invasions of Britain", always seen as invasions from outside; both Nennius and Gildas have most of the Romans returning to Italy after setting up rectores or imperator cum ducibus. Living memory of Roman presence on the Wall had been well and truly lost.

Now then: if we date the end of Roman presence to 410, then its living memory must have lasted, even in the north, until 450-460, when those who were adult when it ended died out. We may be sure, from St. Patrick if no-one else, that the plural name Britanniae had outlasted Roman rule proper for some decades; and we can be just as sure that it had ceased to be used at some point before the sixth century - perhaps at the same point in which Roman law, which defined the four or five provinces that were the Britanniae, was rejected in favour of resurgent Celticism. In other words, the author/authors of A had learned to call the Roman part of the island Britanniae from someone to whom it was natural to do so, but his/their picture of Roman history took its shape after about 460. He/they may even have specifically used this nomenclature to allude to an older age, when the country was called Britanniae, in the plural, remembering that that had once been the case but that at some definite point in time the name changed; if Gildas knew and consciously used the change of name, why not his predecessor/s?

A's evident Christianity has something to tell us. The legend of Rome's claim to the sacred substances of the sacraments, wine and oil, does not remove^[2] that other link to a sacred rather than secular Rome, the never-revoked "taking away" of oil and wine. The Christian pattern at the heart of the story, with God taking the role of sovereign/helper and ultimate high king of Britain vacated by the temporal power of the Romans, shows that British identity was defined by the relationship with Catholic and Christian Rome; we British are Christians, the Pagans are that other lot. This means that Britain's pagan past had been thoroughly forsaken: the fight between British and Picts and the fight between Catholic Christians and outright Pagans are one and the same. Yet as late as the 360s, Britain had enough Pagans to finance the sumptuous new temple at Lydney.

The use of Britanniae in the midst of a faded historical background has an attractive partial parallel in Constantius of Lyons, biographer of St.Germanus of Auxerre. When Constantius was writing in the 480s, old and ill, the Saint’s two journeys to Britain, which he described, were barely, just barely within living memory (429 and 437). The country to which Germanus travelled was certainly the Britanniae; but at the same time, Roman and post-Roman social institutions had already faded from memory. I can do no better than quote C.A.Snyder: "Constantius, when speaking of Gaul and Italy, describes the people as populus or plebs belonging to a city (Auxerre, Milan) or a province (Armoricanae, Augustudense); only in Britain are the inhabitants given ethnolinguistic labels (Brittani, Saxones, Picti), the sole exception being the Alani, who were occupying part of Italy (Vita, 4.28.6). Constantius never uses the term Romani. This serves to separate himself from the people of Britain, who are too remote (like the Alans) to speak of in the same [familar terms as] the people of Gaul... he names Gallic cities, whereas in Britain he speaks only of vague regiones... after the debate, Germanus is approached by a uir tribuniciae potestatis ["a man with the office of a tribune"]. This apparently specific piece of information is ambiguous compared to Constantius' precision in Gallic matters: was this man a military tribune? an administrative one? or both?... presumably he is trying to describe for his Gallic audience a man in Britain who exercised the "power" that a tribune would in Gaul... Elafius is described as regionis illus primus and was supported by the provincia tota. [This] is even less precise..."^[3]

Whatever dating scheme we accept, it is certain that between Germanus' journey (429) and Constantius' book (480-490), the Saxon revolt has taken place. Constantius cannot focus on the specific Britanniae, four or five provinces organized in Roman fashion; he finds it easier to think in terms of peoples, Brittani, Saxones, Picti. Therefore, his use of the word Britanniae is a genuine linguistic survival. Constantius loved to write as though the Roman Empire had not collapsed, with a pernickety attention to titles and ranks half a century out of date; he was, after all, an old man. But the outdated term Britanniae is the only service that Constantius can pay to the Roman Britain of fifty years earlier. He is only certain that his hero Germanus had saved the Catholic faith "in those parts", so that "those parts" are Catholic to his day. As he is thinking of contemporary Britain, 480-490, in terms of peoples rather than Roman administrative districts, he must mean, by exclusion, that the Brittani are solidly Catholic, since neither the Picts nor the Saxons can be imagined to be. That, of course, is the picture of A: we Britons are the Christians, that other lot - the Picts within its picture, the Saxons implied though unmentioned - are the pagans.

The conscious alteration imposed by Gildas on fifth-century sources indicates an awareness that what had been the Britanniae was now singular Britannia. This may indicate a later stage in the evolution of political language, but it is just as likely that the author of A, like Constantius, was being consciously archaic and referring to the country in the way he knew it was called in his grandfather's day.

Constantius' relationship to Britain's Roman past has in common with A the fact that the plural name Britanniae is remembered to the almost complete exclusion of any clear idea of its political set-up. Britanniae originally stood for the Roman provinces on the island; but Constantius had next to no idea of how those provinces were actually organized and run - his vagueness about the uir tribuniciae potestatis tells its own story. If he is more aware of Roman realities than A, this is because he is living in a country, Gaul, where most Roman institutions are still in existence (even though, as compared to the realities of his day, he shows a conscious clinging to earlier times); but he sees post-Roman Britain almost exactly as A sees it.

This agrees with my otherwise highly theoretical earlier remarks, that A must date to quite a long time after the arrival of the Saxons, since if its Picts prefigure the Saxons, then the Saxons known to A's author must have been in the island for a long time and survived at least one catastrophic defeat. The likely date for A - until we determine the date of the Celticist revolution whose charter I take it to be - is still very broad: "from a period long after the first Saxon settlement to a period some time before Gildas" would still give us anything from, say, 490, to about 540 (taking Gildas, as I said, to have written his masterpiece about 561). If we take A, as I do, to represent something like the charter of the nativist revolt described by Zosimus, this period emerges as its likeliest setting.

The fact that A is visibly of Northern origin confirms the home of the politico-cultural movement that led to resurgent Celticism. As we would expect, it is the part of Roman Britain, the northern marches, where Romanization had made the least impact and Celtic culture has been preserved, and which also lay further from Saxon battlefields. The cultural hatred it expresses for Picts and Scots does not allow us to place it elsewhere. It may also explain why Nennius treated it as unreliable: it was not only historically unacceptable (both his account of Brutus and his story of Caesar and the Seven Emperors have an anchoring in classical annalistic history that A signally lacks), it was alien. In spite of its vague connections with Roman annalistic tradition, Nennius' account of the Seven Emperors is a genuinely Welsh tradition, tied to at least two geographically quite precise origins. The “Emperor Constantine” was the same as the hero Minmanton of Caer Seint; and the division between Good Maximus and Bad Maximianus was intended to salve the pride of a number of recognizable Welsh dynasties who claimed descent from “Maxen”. A came from a different area: not North Wales, where Nennius worked, nor Gwrtheyrnion, where he^[4] was probably born, but the British "old north", the area around and across the Wall, which, in his time, still had a vigorous, related, but not identical tradition some of whose details eventually filtered south to Wales or were turned into Latin by the hagiographers of Ninian and Kentigern.

Notes

^[1]For all the facts in this paragraph, SNYDER op.cit. 54-57 and notes.

^[2]As does, for instance, the mediaeval Vita Paterni, in which the Welsh Sts.Dewi, Padarn and Teilo are made Bishops by the Patriarch of Jerusalem and divide Britain between their dioceses, thus by-passing Rome and deriving the British episcopate's legitimacy directly from Palestine. This is an evident if clumsy attempt to deny Rome the right to decide on ecclesiastical hierarchy in Britain, certainly related to the bitter passions aroused by the medieval Welsh attempt to make St.David's a recognized archiepiscopal chair independent of Canterbury - an attempt the Pope resisted.

^[3]SNYDER op.cit. 69, 76, 113-114, 291 (n.28 to p.69).

^[4]I mean of course the person who first compiled the Historia Brittonum, whether he was called Nennius or not.

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

Comments to: Fabio P. Barbieri