Monday, June 19, 2017
Once More Arthur's [Last Four] Battles (a little tribute there to Kenneth H. Jackson's famous Arthurian essay)
Aerial View of Eynsham Park Camp
Readers of my previous posts will recall that I discussed Arthur's last four battles in relationship to the prior engagements, which were all reflections of entries for Cerdic in the Anglo-Saxon Chronicle. I was bothered by the fact that after the Wihtgarasburh battle (= the Castle Guinion of Arthur), the other battles seemed to be tagged on in order to round out the number to a mythological, Zodiacal twelve. These extra battles seemed, superficially, at least, to have nothing to do with any of the other battles listed for the Gewissei in the ASC.
I now have reason to think I may have been mistaken. I had mentioned before that Arthur's City of the Legion battle may well be an attempt at the ASC's Limbury of 571, whose early forms are Lygean-, Liggean- and the like. I discounted the possibility solely because the next battle-site, that of the shore of the Tribruit, was certainly for the Trajectus on the Somerset Avon or over the Severn.
There is a problem, though, with identifying Tribruit with the Avon or Severn Trajectus, viz. there were doubtless many trajecti in Britain! And, indeed, Rivet and Smith (The Place-Names of Roman Britain, p. 178) discuss the term, saying that in some cases "it seems to indicate a ferry or ford..." Furthermore, although the Welsh rendered 'litore' of the Tribruit description in Nennius as 'traeth', demanding a river estuary emptying into the sea, litore (from Latin litus) could also mean simply 'river-bank'. Thus traeth could well be an improper rendering of the word.
If I were to look at Tribruit in this light, and provisionally accepted the City of the Legion as Limbury, and Badon as Bath (which the spelling demands, and which appears in a group of cities captured by Cerdic's father Ceawlin/Maquicoline/Cunedda), then the location of the Tribruit/Trajectus in question may well be determined by the locations of Mounts Agned and Breguoin. These last two battle-sites fall between those of the City of the Legion and Bath, and after that of the Tribruit.
I decided to take a fresh look at Agned, which has continud to vex Arthurian scholars. I noticed that in the ASC 571 entry there was an Egonesham, modern Eynsham. Early forms of this place-name include Egenes-, Egnes-, Eghenes-, Einegs-. According to both Ekwall and Mills, this comes from an Old English personal name *Aegen. Welsh commonly adds -edd to make regular nominative i:-stem plurals of nouns (information courtesy Dr. Simon Rodway, who cites several examples). Personal names could also be made into place-names by adding the -ydd suffix. The genitive of Agnes in Latin is Agnetus, which could have become Agned in Welsh - as long as <d> stands for /d/, which would be exceptional in Old Welsh (normally it stands for what is, in Modern Welsh, spelled as <dd>). I'd long ago shown that it was possible for Welsh to substitute initial /A-/ for /E-/. What this all tells me is that Agned could conceivably be an attempt at the hill-fort named for Aegen.
But what of Mount Breguoin? Well, I had remembered that prior to his later piece on Breguoin ('Arthur's Battle of Breguoin', Antiquity 23 (1949) 48—9), Jackson had argued (in 'Once Again Arthur's Battles') that the place-name might come from a tribal name based on the Welsh word breuan, 'quern.' The idea dropped out of favor when Jackson ended up preferring Brewyn/Bremenium in Northumberland for Breguoin.
So how does seeing breuan in Breguoin help us?
In the 571 ASC entry we find Aylesbury as another town that fell to the Gewessei. This is Aegelesburg in Old English. I would point to Quarrendon, a civil parish and a deserted medieval village on the outskirts of Aylesbury. The name means "hill where mill-tones [querns] were got". Thus if we allow for Breguoin as deriving from the Welsh word for quern, we can identify this hill with Quarrendon at Aylesbury.
All of which brings us back, rather circuitously, to Tribruit. Taking this for a ford, the obvious candidate given Limbury, Aylesbury and Eynsham, is Bedcanforda of 571. This is also found as Biedcanforda and is believed by most to be Bedford (Bedanford, Bydanford, Bedefort, 'Bieda's Ford'). I would not hesitate, therefore, to propose that the Tribruit river-bank is the trajectus at Bedford.
If we accept all this, then we cannot very easily reject Badon as Bath. In truth, with Bath listed in the ASC entry for 577, and made into a town captured by Ceawlin, we simply are no longer justified in trying to make a case for the linguistically impossible Badbury at Liddington.