Plan of Camelot Castle, Somerset
Cadbury Camp, North Somerset
In a previous blog entry (http://mistshadows.blogspot.com/2017/12/note-on-draydon-fort-in-somerset-better.html), I set forth the arguments for Arthur's Dindraithou as a fort at various sites in Somerset and Cornwall. These tentative identifications relied upon -draithou being interpreted either as a Welsh attempt at an English element or as a rendering of a Cornish place-name.
Here I wish to once again emphasize the significance of the odd co-rulership of Arthur and Cado (Cato, Cadwy) at Dindraithou.
The important question here, of course, is why the author of the Life of St. Carannog felt it necessary to have Arthur share rule of this particular fort with the Dumnonian king Cadwy. Obviously, this could simply be a convenient chronological fix. Cadwy was born (according to P.C. Bartram's rough estimate) in c. 500 A.D. But I find that hagiography, like folklore and myth, does not usually bother itself too much with adhering to accuracy in such matters. In addition, as Arthur is indisputably the most famous figure, why risk diminishing his power and reputation by forcing him into a peculiar dual-kingship role with Cadwy?
I cannot help but see in Cadwy's presence at Dindraithou an attempt to indicate that we are dealing with one of the Cadbury forts. In other words, Dindraithou is a Cadbury fort. As it was presumably called after Cadwy, yet Arthur was there (in whatever capacity), the only way to reconcile this apparent contradiction was to make both chieftains ruler of the place. The author of the Vita, in other words, knew that Dindraithou was a designation for a Cadbury fort.
If this is so, we must fall back on Dindraithou as either 1) a 'fort of the traethau' or "fort of the shores" or 2) a Welsh rendering of the Irish Dinn Tradui/Tredui, 'triple fossed fort.' I will consider a third possibility below.
If Dindraithou is No. 1, then only the Cadbury Camp fort in North Somerset works. As I've mentioned before, with the change in sea level at Arthur's time, Cadbury Camp atop its ridge overlooked the often flooded Somerset Levels. Where the ridge met the water could have been called "the traeths" - although the word traeth itself is reserved for a shore associated with an estuary. Cadbury Castle was too far east and was not close enough to the flooded area of the Somerset Levels.
If No. 2, either of these two forts are candidates, as both are multivallate.
So how do we decide between the Cadburys?
Archaeology may have done that for us, fortunately. The famous excavations of Leslie Alcock showed conclusively that Cadbury Castle underwent a massive rebuild at Arthur's time. Because of the River Cam and Camel villages here, local tradition had long identified the place with Camelot. [Camel comes from Cantmael and has nothing whatsoever to do with Camelot, a late French attempt at rendering Campus Elleti in Wales.]
However... and this is a big 'however', if I'm right and Illtud is Arthur's father, I would opt for Cadbury Camp. Why? Because the Llydaw/Letavia/'Brittany' of Illtud's father, Bicanus, is almost certainly the Leadon Valley area of what was Dobunni territory. Dobunni territory extended to Cadbury Camp in North Somerset, and Cadbury Camp is pretty much directly across the Mouth of the Severn from the Dinas Powys where Illtud served as military commander.
Another smaller Cadbury fort is to be found at Cadbury Hill
Alas, there is no archaeological evidence for early medieval reuse of either of these two hillforts.
DINDRAITHOU (CAIR DRAITOU?) AND THE DUROTRIGES
It has occurred to me that if Cadbury Castle in Somerset is Dindraithou, that there might be one more possible explanation for the latter name. This idea is a stretch, but perhaps worth mentioning.
Ilchester just a little ESE of Cadbury Castle (see map below) appears to have belonged to the ancient Durotriges tribe. Here is the entire entry for the Durotriges from Rivet and Smith's THE PLACE-NAMES OF ROMAN BRITAIN:
- Ptolémée, II,3,13 : Dourotriges ( = DUROTRIGES);
- Inscriptions RIB 1672 (pierre provenant de Cawfields sur le Mur d'Hadrien) : C(IVITAS) DUR(O)TR(I)G(UM) (L)ENDIN(I)ESIS.
- Inscription RIB 1673 (pierre provenant de près de Housesteads sur le Mur d'Hadrien) : CI(VITAS) DUROTRAG(UM) LENDINIESI(S).
(Both stones are probably of A.D. 369. For the spelling of the adjectival form of the civitas-name, see LINDINIS)
It is by no means certain that the name bas -i-. though it is traditionally cited in this form. While this is present in all the Ptolemy MSS, it could be an error going back to the archetype or to Marinus. RIB 1672 omits the vowel, but RIB 1673 plainly writes -a- (A) and it is not likely a member of the civitas literate enough to be given the task of cutting the iscription would mistake the spelling of the name of his own people. For a similar (reverse) error in Ptolemy, compare VINDOGARA, which has Ouando- (= Vando-) in all the MSS at II, 3, 7.
DERIVATION. This name is obscure. It be divided Duro-triges or Durot-riges. It is tempting to think it another of the many names in *duro- 'fort ', although Britain we know this element chiefly applied to early Roman forts on low ground, so that it could hardly apply particularly in the name of a tribe whose region contains many spectacular hill-forts (which would have had *-dunon names). However, this *duro- does enter into a few names such as the two British Durocornovium and in Gaul ethnie Durocasses. Holder suggested (I. 1387) that the name might be a reduction of older *Durot(o)-riges, that is *duro- ' fort ' with some kind of suffix or infîx when in composition; but then there is no parallel for this among all the very numerous Dura- names.
The second element is possibly -riges, a plural of *rig- ' king ' which appears in several ethnic names, such as Gaulish Bituriges 'kings of the world'. Watson CPNS 16, note, identifies an element -raige (for older -rige ?) in Irish ethnic names, e.g. Dartraige (dart 'year-old bull or heifer') > Dartry, Cattraige (catt 'wild cat'), Luchraige (luch 'mouse'), and even though these seem somewhat unheroic names, they may have had totemic significance and are acceptable enough; but we have no evidence that in the present name *durot- is an animal-name which would fit into this series. If the second element is an unknown -triges, the British name might be paralleled by the Allotriges (= Allotriges : Strabo III, 4, 7 ; in Ptolemy II, 6, 7 they are the Autrigones (= Autrigones), of Hispania Tarraconensis, with Allo- probably as in Allobroges of Gaul.
If -a- is right, the etymological possibilities are less good. A -rag- element is unknown, but -trag- is said by Dottin LG 193 to be a 'terme de composé' and is related by him to Irish traig 'foot' and Welsh traed (pi.). But this does not relate well semantically to a first element *duro-. The name must be left unresolved.
IDENTIFICATION. A people of southern Britain with their capital at Durnovaria, Dorchester; for the probability that their civitas was at some time subdivided, with a second capital at Ilchester, see Lindinis. Ptolemy attributes to them only Dunum (= Dunium, Hod Hill?), but the distribution of their pre-Roman coins indicates that they occupied Dorset, parts of Wiltshire, Hampshire and somerset, and perhaps the extreme eastern part of Devon.
Proximity of Cadbury Castle to Ilchester/Lindinis of the Durotriges
I can think of a very good word which might be brought into the context of -triges or -trages:
abode Proto-Celtic *trīko-, SEMANTIC CLASS: technology, cf. Old Welsh tricet ‘resides’, Welsh trig ‘stay, a residing, abode, dwelling(-place)’, Breton (milin) tric (Old Breton) ‘gl. permanendi in stupris ‘staying sullied’
From the GPC:
trig1
[bôn y f. trigaf1, trigiaf: trig(i)o; cf. H. Lyd. (milin) tric, gl. permanendi in stupris; ansicr yw union ystyr rhai o’r enghrau. isod]
eg. ?a hefyd gyda grym ansoddeiriol.
Arhosiad, preswyliad, preswylfa, trigfan:
stay, a residing, abode, dwelling(-place).
If this is -triges, then we have simply "Fort-dwellers", a very acceptable name for a tribe known for its many fine forts. We may compare this name to that of the nearby Atrebates, whose name means simply "settlers, inhabitants" (Rivet and Smith; *ad-, to, + *treba, to inhabit).
Again citing Rivet and Smith, "...*duro- 'fort, walled town', apparently usually on low ground (and named in contrast to *duno- 'hill-fort')...". Given the spelling -trag-, and knowing as we do that /t/ (and from t, d) can substitute for /c/, could we suggest that -draithou (or the Irish tredui/tradui) is a corruption of trig/trag and that Din-/Duno- has been substituted for Duro-? In other words, at some point, by some process, Duro-triges/trages became Duno-triges/trages? And that it is this last we happen to have preserved in the form Dindraithou?
Even if this doesn't work linguistically, it is evident that Cadbury Castle was, at one time, in the tribal kingdom of the Durotriges.
