17 March, 2010

The battle of Arfderydd or Arthuret

The Battle of Arfderydd or Arderydd is one of only four battles listed in Annales Cambriae for the sixth century. Two of the others are Arthur’s battles at Badon and Camlann, so it’s in illustrious company. What do we know about the battle?


Annales Cambriae

573 The battle of Arfderydd ‡between the sons of Eliffer and Gwenddolau son of Ceidio; in which battle Gwenddolau fell; Merlin went mad.‡

580 Gwrgi and Peredur ‡sons of Elifert‡ died
--Annales Cambriae, available online


Gwendoleu a Nud a Chof meibyon Keidyav m Arthwys m Mar m Keneu m Coel
--Bonedd Gwr y Gogledd (Descent of the Men of the North), available online
Gvrgi a Pheredur meibon Eliffer Gosgorduavr m Arthwys m Mar m Keneu m Coel
--Bonedd Gwr y Gogledd (Descent of the Men of the North), available online

Gurci ha Peretur mepion eleuther cascord maur map letlum map Cene├║ map Coylhen.
--Harleian Genealogies, available online

Welsh Triads

Three Horse-Burdens
… Corvan, horse of the sons of Eliffer, bore the second Horse-Burden: he carried on his back Gwrgi and Peredur and Dunawd the Stout and Cynfelyn the Leprous(?), to look upon the battle-fog of (the host of) Gwenddolau (in) Ar(f)derydd

Three Faithful War Bands
… and the War-Band of Gwenddolau son of Ceidiaw at Ar(f)derydd, who continued the battle for a fortnight and a month after their lord was slain. The number of each one of the War-Bands was twenty-one hundred men
--Red Book of Hergest, available online

Myrddin poetry

Merlin, or Myrddin, features in the poems Afallen (Apple Trees, translation available online) and Oianau (Greetings, Piglet, translation available online) in the Black Book of Carmarthen, in which he says he is hiding from Rhydderch and mourns that no-one now honours him after the death of Gwenddolau at Arderydd. This is mystical poetry, not a historical narrative, and should be interpreted with caution, but it may draw on older traditions.

The late medieval manuscript called Lailoken and Kentigern says that the mad prophet Lailoken was also called Merlin, and was living wild in the wilderness following a terrible battle “fought on the plain between Liddel and Carwannok” (source: Wikipedia page on Myrddin Wyllt).


Annales Cambriae is clear that the battle of Arderydd was fought between the sons of Eliffer, named a few entries later as Peredur and Gurci, and Gwenddolau son of Ceidio. All three protagonists and their respective fathers are also listed in the genealogies, and appear associated with the battle in the Triads. This may not be independent confirmation, since it is possible that the sources could have copied each other, but three sources all consistent with each other is impressive.

The Merlin poetry has Merlin hiding from Rhydderch, presumably Rhydderch Hael, who was king of a territory centred on Alt Clut, modern Dumbarton, in the late sixth century. At first glance this might indicate that Rhydderch was the winner of the battle, which would either mean he was allied with Peredur and Gurci or would be inconsistent with the Annales Cambriae. However, the poetry does not explicitly say that Rhydderch fought or won the battle (as far as I can make out), only that Rhydderch is Merlin’s enemy. This is equally consistent with Rhydderch having opportunistically moved in to part or all of Gwenddolau’s former territory after the battle without having fought in it, or with Merlin having wandered into Rhydderch’s territory in his madness.

It is worth noting that although the genealogy brackets Gwenddolau with two brothers, named Nud and Chof, in the same way as Peredur and Gurci are bracketed together in their genealogy, there is no mention of Gwenddolau’s brothers at the battle. It could simply be that no-one remembered to mention them, or that they were dead before the battle took place, or that they were elsewhere at the time, or that for whatever reason they chose not to fight alongside their brother.

All the participants appear in the genealogies of the ‘Men of the North’. Rhydderch was king of the area around modern Dumbarton, and Peredur is traditionally associated with York (see post on Peredur for the rationale). It therefore seems likely that the battle of Arderydd was fought somewhere in the region that is now southern Scotland and northern England.

If the manuscript Lailoken and Kentigern can be trusted (which is a big ‘if’, as it dates from the fifteenth century and is thus nearly a thousand years after the event), it may provide a more detailed clue to the location. It says, “fought on the plain between Liddel and Carwannok”. Liddel must surely be the Liddel Water, which runs south-westwards from the hills at the northern end of the Pennine chain to join the River Esk on the coastal plain at the head of the Solway Firth, and which was later famous, or rather infamous, as the worst haunt of the Border Reivers. As far as I know Carwannock has not been identified, but the first element ‘Car’ is the Brittonic ‘Caer’, meaning a fort. There are plenty of Roman forts in the area, including the chain along Hadrian’s Wall and outpost forts north of the Wall, and there may well have been other fortified places. This would place the battle somewhere on the Solway plain, perhaps at a strategic crossing of the River Esk or the Liddel Water, which is a plausible sort of location for an important battle.

Arthuret House, on the River Esk just south of Longtown, is the traditional site ascribed to the battle. It is on the plain, not far from Liddel Water, and could count as ‘between Liddel and Carwannok’ if Carwannok is to be identified with one of the forts on Hadrian’s Wall. Nearby Carwinley, which stands on the Liddel Water not far from the site of the Roman fort at Netherby, has been suggested to derive from a name something like ‘Caer Gwenddolau’; if correct, this derivation is further support for locating Arderydd somewhere in this approximate area.

The cause of the battle is unknown. The battle was evidently important enough for the compiler of the Annales Cambriae to consider it worth recording. The ‘twenty-one hundred men’ of the Triads is probably a poetic convention, though it may indicate that the forces involved were remembered as being unusually large. Similarly, it is unlikely that the defeated warriors literally continued the battle for six weeks (!), but that may indicate a tradition that the battle was exceptionally hard fought. All of this is consistent with it having been a major battle, either in numbers or because it was politically decisive in some way, or both.

The ‘Men of the North’ genealogies show Peredur and Gurci as first cousins to Gwenddolau, suggesting that the battle was a dispute between two branches of the same family. Perhaps it was a sort of sixth-century Wars of the Roses, with each side claiming the other’s inheritance. It is worth noting that there is no mention whatsoever of Arderydd having been fought against ‘Saxons’ or other invaders; it appears to have been a strictly family affair.

Gwenddolau clearly lost both the battle and his life. Peredur and Gurci evidently survived, since the Annales Cambriae enter their deaths seven years later in 580. So we can reasonably infer that the immediate outcome was that Peredur and Gurci won. Given that the battle was considered worth remembering, it may be that it also had longer-term consequences that altered the political balance in the area. The genealogy ends at Gwenddolau’s generation, suggesting that either he left no heirs or that his descendants did not reclaim their territory.

The area over which Gwenddolau ruled is not known. It is a reasonable inference (though not a certainty) that his territory was somewhere in the region of the battle, in what is now north-west England and/or south-west Scotland. He presumably controlled substantial resources, if he could maintain a large warband, and therefore it can also be reasonably inferred that his territory was large or wealthy or both. If this is the case, the prospect of gaining it as a prize may have contributed to the motivation for the battle.

If Peredur and his brother were indeed based in York (see post on Peredur for the rationale), this raises some interesting questions. What were the kings of York doing fighting a battle on the other side of the country? Was there some sort of regional overlordship in dispute, or were they trying to extend their territories, or were they simply intent on destroying Gwenddolau for personal reasons that had nothing to do with claiming territory? If the latter, it may explain why Rhydderch was present in the area to hound the insane Merlin; he may have quietly moved into a gap left after the York kings went home. If Peredur and Gurci took over Gwenddolau’s kingdom, one wonders how the logistics would have worked to control a territory sprawling from York to the Solway with the Pennines in the way, unless one of the brothers ruled it as a separate sub-kingdom.

Interestingly, a generation or so after the battle of Arderydd the kingdom of Rheged appears to have emerged as a powerful force in much the same area. Its king, Urien, is remembered in Historia Brittonum as having almost destroyed the Anglian kingdom of Bernicia, at a date some time before Aethelric became king of Bernicia in 593. The location of Rheged is not known with any certainty. It was presumably somewhere in the North, since Urien appears in the ‘Descent of the Men of the North’ genealogies and since his recorded enemies and allies are all associated with the region. It was not on the east coast, since that area is accounted for by Bernicia and Deira, so that leaves a gap on the map in the area that is now north-west England and/or south-west Scotland, and it seems logical to place Rheged there. How far its boundaries extended is open to argument, and in any case probably varied according to the military success of its kings relative to their neighbours. This is the same area as the location of the battle of Arderydd. Which could be pure coincidence, or could perhaps indicate that the battle of Arderydd was connected with the later rise of the kingdom of Rheged, either directly or by creating a gap into which a new dynasty could move.

Another interesting observation is that the traditional site of the battle at Arthuret House is less than 20 miles from Birdoswald Roman fort. As discussed in an earlier post, someone in post-Roman Birdoswald built two successive massive timber halls on the site of the north granary, which would be consistent with occupation by a local ruler with control of substantial resources. The halls cannot be precisely dated, and the excavator suggests that occupation may have lasted on the site until around 520 (see earlier post on dating for rationale). If this date is correct, the halls on the north granary site (though not necessarily the whole fort) would have been long abandoned by the time of the battle of Arderydd. However, if the halls lasted longer than estimated, which is possible given the absence of dating evidence, this raises the question of whether Birdoswald and/or its ruler had any connection with the battle of Arderydd. A major chieftain’s hall and a major battle within 20 miles of each other may not be entirely coincidental. Could Birdoswald have been the seat of Gwenddolau or his family? If so, one could speculate that the battle at Arderydd may suggest a context for the end of the sequence at Birdoswald. I need hardly say that this is speculative and that other interpretations are possible.

Annales Cambriae, available online
Bonedd Gwr y Gogledd (Descent of the Men of the North), available online
Harleian Genealogies, available online
Welsh Triads, Red Book of Hergest, available online

Map links
Arthuret House, near Longtown
Liddel Water


Unknown said...

You are absolutely brilliant. I am very fascinated by this period in history that is so shrouded in mystery. I came across your site looking for "explanations" of the Staffordshire horde and for the past week I've been catching up on your writings.

One comment, I would imagine most battles are not clean 1 or 2 day events especially in this period of time. To move 2,000 men and material would take a long time. In terms of actual confrontation I would imagine it would be logistically very difficult to bring a full force to bear on an enemy in terms of force mobility (troops out scouting, gathering food). Also, the surviving troops of a defeated army would take a long time to root out and may fight as guerillas or bandits for years after a battle. Hence maybe the reason for the 6 week notation.

I digress, thanks for the effort and the wonderful site.

Rick said...

Perhaps it was a sort of sixth-century Wars of the Roses

Funny you should say that, because when you mentioned possible opportunism by Rhydderch the first thing that popped into my head was the Stanleys at Bosworth.

I take it that 'Arthuret' is merely a much later anglicization of Arfderydd, and doesn't signify any Arthurian association. Though I imagine antiquarians a thousand years later made one up!

Carla said...

Daniel - Hello and welcome. What a nice thing to say :-)

Good point, it depends how the 'battle' was defined. If it included the logistical build-up and the chasing down of defeated survivors it might have covered a long period. In particular, gathering together lots of loosely allied semi-independent warbands and persuading them to co-operate probably took all summer, or all the preceding winter. Something like that might underlie the mention of the warriors in Y Gododdin feasting for a year before they set out. Similarly, as you say, if some survivors of the defeated side continued to resist, perhaps regrouping in other locations, that could underlie the tradition in the Triad of a warband continuing a battle long after the leader's death.

The actual engagement was probably over within daylight hours in a single day - John Keegan in The Face of Battle comments that single-day battles were the norm in the medieval world. Seen in the context of the wider campaign, longer timescales probably apply.

Rick - some things never change, do they? Wars between cousins always make me think of the Wars of the Roses, or possibly the Stephen-Matilda Anarchy. I daresay early medieval Britain had just as many dysfunctional families and dynastic squabbles as better-documented times.

Good question. If the spelling was Ardderyd (with the double and single 'd' reversed), it would be pronounced phonetically as something like Arthered ('dd' in Welsh is a 'th' sound). In which case Arthuret could just be a phonetic spelling of the original name, not even really an anglicisation. It needn't have anything to do with Arthur, although there might be some sort of common root derived from 'arth', 'bear'. I could imagine 'bear' being a component of a topographical place name.

That said, Camboglanna (Castlesteads) is one of the candidates for Camlann and is only a few miles away, so never say never :-) There might be a genuine connection with the historical Arthur, if he existed, with the place name having been formed from the personal name. I very much doubt that it's possible to tell the difference now. One of the problems is that Cumbria was part of Scotland when William the Conqueror came calling so it isn't in Domesday Book, thus depriving us of one of the most comprehensive sources of early place names. I daresay the local antiquarians played up the possibility of an Arthurian connection for all they were worth :-)

Rick said...

I also wondered if the six weeks of fighting was an echo of a tough mopping up operation. But more likely it is just honest poetic exaggeration, one more way of saying that there were giants in those days.

My impression is that 'battles' were characteristic of older warfare, when armies marched around, periodically encountered each other, and fought. Modern era armies are always in contact, always skirmishing. What we call battles, like Stalingrad, an earlier age would have called campaigns.

Gabriele Campbell said...

The Birdoswald hypothesis might work. We don't know where the Roman Banna comes from, but Carwannock could well be be something like Caer Bhannaic. I won't be surpised if the Roman's had some trouble with the different b - bh - w pronounciations they would hear from the natives. Or the other way round, some 5th century Celt changed the Roman name into something more suitable for his tongue. And some poor scribe got confused further. :)

Carla said...

Rick - it could be either, or a bit of both, stubborn resistance described in terms of poetic convention. 'Giants in those days' is a feature of heroic poetry, and maybe the writers of the Triads went in for it. In the Dream of Rhonabwy a twelfth-century man dreams of going back in time to King Arthur's day, and King Arthur takes one look at him, shakes his head sadly, and says what a pity it is that such little fellows as these now keep Britain, compared with the fine men who kept it before. (Or words to that effect - Google will probably find you an online translation if you want the detail).

John Keegan makes a not dissimilar point, that modern warfare and logistics supports ever bigger battles that cover ever wider areas and last ever longer, though I can't remember how he defines a battle.

Gabriele - I don't know whether Brittonic P-Celtic has something like the 'Bh' of modern Scots Gaelic, or whether it could have transformed a 'B' into a 'W'. But the Romans may well have got 'Banna' from a local name, as they often did, and there's no telling how a harrassed Latin-speaking clerk might have mangled an unfamiliar name, especially if he was trying to write it down in horizontal rain :-) It might have been 'W' all the time for all we know. If Carwannok is intended to be Birdoswald, then that would place the battle somewhere in the low-lying country off the western edge of the Pennines, maybe around the River Lyne. Birdoswald has its modern name, or a recognisable forerunner of it, in medieval documents so it wasn't called Carwannok by then, but the story about Merlin could have been drawing on an older tradition with a different name.

Gabriele Campbell said...

Well, Merlin made it into the family tree of the Campbell Dukes of Argyll, so you can be sure there was more than one legend attached to his name. Maybe there's a Special Scottish edition or something. :) I think he was running around mad in more than one wood as well.

Carla said...

Very possibly more than one Merlin, as well as more than one wood. Like King Arthur, you can make almost anything you like out of the Merlin legends.