The Tangled Chronology of Historia Brittonum
The History of the Britons, compiled in Wales around 828 and traditionally ascribed to the monk Nennius (a name I shall use for the author for convenience' sake, though it may be spurious), is one of very few sources for the history of sub-Roman Britain, and the first in which the legendary Arthur is more than a name. It has hence been much pored over by Arthurian enthusiasts, often resulting in somewhat eccentric theories. However, its basic chronology for the sub-Roman era is usually taken for granted, and treated as internally consistent even by theorists who wish to dispute it. On closer examination, however, it is nothing of the sort.
(N.B.: I have used the public domain translation by J. A. Giles.)
Date 1: The end of Roman Britain As early as the sixth century, it is clear that there was significant confusion in British tradition between the rebellions of Magnus Maximus in 383-88, and Constantine III in 406/7-11. This is hardly surprising. Both were military officers, elevated to the imperial purple by legionary mutinies in Britain, who then invaded Gaul, defeated their enemies, were recognised as co-rulers by at least one legitimate Emperor, and enjoyed a short reign before being betrayed and killed; both even began their careers by killing a rival named Gratian. It is hardly surprising, then, that they became partially conflated: and as a result of this, Maximus' revolt was often regarded (quote wrongly) as marking the effective end of Roman rule. Nennius was aware of variant traditions. One, in which Roman rule ended with Maximus' departure for the Continent in 383; and another which was clearly a garbled version of Constantine's rebellion. In the latter version, two more Emperors - Severus and Constantius - ruled in Britain after Maximus, and Roman rule ended with Constantius' murder after a reign of sixteen years. So, for the end of Roman Britain according to Nennius, we have one clear date (383); one indication of a later one (some time after 404, depending on the length of the fictitious Severus' reign); and the possibility that he was actually thinking of a point in the reign of Constantine. (This might be the original mutiny in 406; Constantine's departure in 407/8; the Emperor Honorius' repudiation of responsibility for Britain in 410; or Constantine's death in 411.) But he also gives two variant lengths for Roman rule in Britain - 409 years, or 348. Unfortunately, just as two departures had become conflated in popular tradition, so too had two Roman invasions - the ultimately unsuccessful incursion by Julius Caesar in 55-54 BCE, and Claudius' conquest in 43-51. Counting from Caesar's invasion, both these lengths are far too short to make sense; from Claudius', they give end dates of 452/6 and 391/5 respectively (the variation caused by the fact that Nennius mistakenly dated Claudius' invasion to 47, and might have been counting from then). The former is surely far too late; the latter falls between the reigns of Maximus and Constantine, thus putting it in roughly the indicated area. 409, however, might conceivably be a mistake for saying that the Romans left in 409, which is very close to the accepted dates. Date 2: The accession of Vortigern After Maximus' victory over Gratian, Nennius says the Britons "were in alarm forty years. Vortigern then reigned in Britain." This dates the rise of the "proud tyrant" who would let in the Germanic invaders to c. 423 - except that, as we have seen above, Nennius was extremely confused about when the Romans actually left. He had also referred in the previous sentence to the death of Maximus in 388, suggesting a date of 428 for the rise of Vortigern. However, he elsewhere gives what looks like a precise date, connecting Vortigern's reign with the consulship of Theodosius and Valentinian. Immediately before this, he had referred to these two Consuls defeating and slaying Maximus: this is clearly a confusion. The same pair of Consuls were certainly not in office on occasions forty years apart!
In fact, Nennius has mixed up two pairs of Emperors who happened to have the same names. Valentinian II had been Consul in 388 when Maximus was slain, and his Eastern colleague Theodosius I the year before; they never shared a Consulship, but Theodosius II and Valentinian III did, four times: in 425, 426, 430, and 435. Nennius clearly indicates that he is thinking here of the 425 Consulship, but the possibility of confusion with one of the other occasions cannot be ruled out. (He also slightly misdates Maximus' death to 32 years after the Consulship of Constantinus and Rufus, which would place it in 389.) Finally, he dates Vortigern's reign to 28 years after the consulship of Stilicho, which would put it in 428 - but in the same sentence he associates it with Valentinian III's accession to the Imperial throne in 425.
Date 3: The Adventus Saxonum Vortigern's notoriety, of course, rests chiefly on the story that he was the ruler who first brought in the Germanic peoples who would overrun much of Britain in the succeeding centuries, when he hired the Jutish warlord Hengest as a federate. (The Britons tended to refer to Angles, Jutes, Frisians, etc., under the blanket term "Saxons" once they were in Britain.) This event already had a traditional date by Nennius' time, having been ascribed to 449/50 by the Venerable Bede a century earlier: but Nennius gives it several conflicting dates. a) 428 years before his time of writing, which he dates to the fourth year of King Merfyn Frych's reign in Gwynedd: this would place the Adventus c. 400 C.E. b) When "Gratian Aequantius... reigned in Rome" - a garbled reference to the Emperor Gratian (367-83), and in particular his co-consulship with Equitius in 374. c) 447 years from the Passion of Christ - i.e. c. 480. d) According to a passage added in a tenth century recension, 542 years before the accession of Edmund I of England: i.e. in 397. e) Around the time of St Germanus' first mission to root out the Pelagian heresy in Britain: i.e. in 429. f) In the year 447, suggesting that "447 years from the Passion" may be a mistake for this - but this is associated with a second mention of "Gratian Aequantius"! g) In the fourth year of Vortigern's reign, which he dates in that sentence to 400, but also to the consulship of Felix and Taurus in 428. It has been suggested that this represents the opposite mistake to that surrounding the 447 date, and that he means 400 years from the Passion. h) 69 years before the consulship of Decius and Valerian - two third century Emperors who never shared a consulship! However, several consuls between 444 and 534 were named Decius, including a cluster in the 480s, which might suggest a date in the 410s. (There were no Consuls named Valerian after the third century, but a few named Valerius, including one in 521 who has been suggested as the Consulship intended here. This would give a date of 452: but his colleague was the Emperor Justinian I.) We know from contemporary sources that Britain was "overrun" by a major Germanic advance in 441/2. However, this does not preclude a later date for the Adventus, as Hengest's people could have been hired to fight back against the existing Germanic invaders (rather than against Picts and Scots as tradition indicates). After all, Saxons and other Continental peoples had already invaded Britain unsuccessfully in the fourth century, so no fifth century Adventus could represent their first arrival. It was probably in response to the defeat of 441/2 that the Britons sent the Appeal to Aetius, seeking aid from Rome, in or after 446. (It refers to Flavius Aetius as three times Consul, dating it most probably to the actual term of his third consulship in that year, but - given that he remained a power behind the Imperial throne - hypothetically any time up to 454, the year of his fourth consulship and his death.) It is easy to imagine Germanic federates being hired after this appeal failed, in 447/9; but they could just as plausibly be the very people against whom the Britons sought defence, having arrived in the 420s and later revolted. Date 4: Vortigern's death Other references in Nennius only serve to confuse matters further. Ambrosius Aurelianus, who was famous for coordinating the resistance to the revolted federates, is referred to as already a threat to Vortigern when the latter first came to power, and as fighting a battle at Guoloph in the twelfth year of his rival's reign - but also appears to be identical with the fatherless boy, later identified with Merlin, whom Vortigern attempts to sacrifice when his enemies are closing in upon him, near the end. This is placed earlier by Nennius than by later tradition - it appears to follow swiftly on Hengest's revolt - but it is still an inconsistency. Vortigern's final downfall is associated by Nennius with the second mission of Germanus. Although this used to be conventionally dated to the mid 440s, there are problems with this. For one thing, it has recently been suggested that Germanus may not have died in 448 as traditionally claimed, but perhaps as early as 437. The second mission may in fact belong to the 430s, or may even be a fictitious duplicate of the visit of 429. Nennius appears to associate it with the death of St Palladius and his replacement in Ireland with St Patrick, for which he accepts a conventional dating of 432, but he is somewhat vague - and this date too has been queried, it being suggested that these events in fact belong to the 460s.
What can we make of this? All of this is much messier than the conventional reading, whereby Nennius supposedly dates Vortigern's accession to 425, the Adventus to 428, and the battle of Guoloph to 437. And Nennius' most famous passage, regarding the twelve battles of Arthur, is not explicitly dated at all: he appears to link it to the death of St Patrick, but in terms that make it unclear whether he means that they took place after Patrick's death, or concurrently with his mission in Ireland. Furthermore, Nennius most likely accepted the tradition that Patrick had died in 461: but this date may in fact belong to Palladius, with the real Patrick not dying until 493. In which case, of course, the battles might belong to the latter period - but only might. Unlike the later cobbled-together genealogy which made Ambrosius Constantine's son and Arthur his grandson, he provides no family connections that might help make sense of the chronology here. The way he presents the succession of events seems to imply that there has been no very great leap of time between Vortigern's fall and Arthur's rise, but certainly does not say outright that there has not.
It is true that, while every date is messy, the Adventus is uniquely so. Remove that, and the conventional reading - though imperfect - is the most consistent one available. Even when we examine the wildly inconsistent dating of the Adventus, we find that the number 428 keeps cropping up: not always as the actual date, but if this represented a garbled reading of a tradition in which 428 was the date, it would fit into the conventional reading. However, that conventional reading is also apparently in direct conflict with Bede's earlier and far more internally consistent chronology, while some of Nennius' other dates for the Adventus line up much better with Bede. It's also the case that the battle of Badon, listed by Nennius as the last of Arthur's twelve victories, is agreed by most other early sources (including the most plausible reading of Gildas, who wrote when it was in living memory) to have happened somewhere in the general vicinity of 500, give or take a few years: which can be made to appear consistent with Nennius' account, but only at a stretch, and is certainly many decades later than he would appear to indicate at first reading. Furthermore, Nennius may have pushed a date in the 420s precisely in order to set Vortigern up against St Germanus. (Vortigern was claimed as an ancestor by the dynasty who ruled Powys until 855, and Nennius seems to have been a propagandist for their Gwynedd rivals, with a vested interest in blackening the legendary king's name.)
It is clear that Nennius' use of Roman consulships for synchronisation, authoritative though it seems before compared to the facts, is no more consistent than any of his other methods; and that the existing confusion in British tradition about when Roman rule began and ended, and possibly about when Patrick died, has helped to throw his chronology so completely off that it is very difficult to reconstruct anything from it which makes sense. Furthermore, we do not know where he was getting any of these dates and synchronisations from: many may be pure invention.
However: if we wished to speculate wildly, it is just possible that his confusion about when the Romans left could account for his inconsistency with Bede. If he had information that Vortigern rose to power around forty years after Constantine, and misinterpreted it as meaning forty years after Maximus, the former would be consistent with Bede's Adventus and a conventionally-dated Battle of Badon, the latter with the traditional reading of Nennius' chronology. This makes at least as much sense as any other attempt to account for Nennius' dates, though it does push Germanus out of the picture; but sadly the actual chronology of fifth century Britain must remain tantalisingly obscure.