THE LOCATION OF KELLIWIC

A lot of ink has been spilled on the subject of the actual location of Arthur's Kelliwic (or Celliwig) - including by the present author.  But I was never completely satisfied with any of the identifications offered, so have decided to treat of the place once more.

There are many good sources which treat of the several candidates for Kelliwic.  One of the best is Oliver J. Padel's "Some Souther-Western Sites With Arthurian Associations" in THE ARTHUR OF THE WELSH (ed. Rachel Bromwich, A.O.H. Jarman and Brunley F. Roberts).  I refer my readers to that excellent study, as it is not my intention to rehash such material here.

In Triad 1, we are told that the chief bishop of Kellewic is Bytwini/Bitwini/Betwini (modern Bedwenni).  Some have thought this merely W. bedwenni, 'birches', a sort of pun on the meaning of Kelliwic.  But others (including Bromwich) have made a connection with Bodmin, the etymology of which is discussed by Ekwall as deriving from Cornish bod, 'house, dwelling', and either meneich, 'monks' or menehi, 'monsastery.'  The chief elder of Kelliwic in Cornwall is Caradoc Freichfras - which is extremely odd, as this particular Caradoc belong to central and, perhaps, SE Wales.

What I decided to do was to see if there might be any Caradoc place-names at or near Bodmin which could have been wrongly related to the Welsh Caradoc.  There are, in fact, two such places.

Caradoc Place-Names near Castle Canyke and Castle An Dinas

One is Tregardock roughly between Tintagel and Bodmin.  The other is Craddock Moor near Minions on Bodmin Moor.  Either or both could have been fancifully linked to Caradoc Freichfras.

The other clue to Kelliwic's whereabouts comes from Geoffrey of Monmouth's HISTORY OF THE KINGS OF BRITAIN.  He converts Kelliwic into Silchester. The following passage from John Koch's CELTIC CULTURE: A HISTORICAL ENCYCLOPEDIA discusses Silchester in the context of Kelliwic:

Celliwig in Cernyw

(Kernow/Cornwall). The latter name appears to be

a combination of celli ‘wood’ and gwig, probably meaning

‘settlement’ < Latin v¼cus. There have been various

attempts at identifying the place, including proposals

for Calliwith near Bodmin, the hill-fort at Castle

Killibury, the hill-fort near Domellick (Geoffrey of

Monmouth’s Dimilioc), a place in Cornwall called

Cællincg and also Cællwic in Anglo-Saxon sources

(possibly modern Callington). In 1302, two men were

accused of murdering a Thomas de Kellewik in west

Cornwall, but this is the only occurrence of this

Cornish name that seems to correspond exactly to

Welsh Celli Wig. In considering this unresolved

question, it is important to remember that the region

name Cernyw in Culhwch, and in early Welsh tradition

in general, was more extensive than the modern county.

It is also not impossible that early Welsh sources

sometimes mean the old tribal lands of the Romano-

British civitas of the Cornovii in what is now Shropshire

(Welsh swydd Amwythig) and Powys, though a

conclusive example of such a meaning for Cernyw has

yet to be found. A suitably important sub-Roman place

with a philologically workable name would be Calleva

(Silchester), the fortified centre of the civitas of the

Atrebates, which continued to be occupied and free

of Anglo-Saxon settlement into the 5th century, but

Silchester is nowhere near either Cornwall or the

Cornovii.

We have evidence of a sub-Roman Irish presence at Silchester
(http://www.archaeologyinmarlow.org.uk/2011/02/the-enigma-of-silchester%E2%80%99s-ogham-stone/; and see below).

However, Geoffrey gives away what he or his source believed to be the real Kelliwic by placing St. Mawgan (Maucanus) at "Silchester." There are two Mawgan sites in Cornwall.  One is across from Gweek on the Lizard Peninsula, a candidate for Kelliwic.  But there is no significant hillfort or Roman site at this place that could have been thought of as belonging to the great Arthur.  The other is St. Mawgan and the Vale of Mawgan only a dozen kilometers or so west of Bodmin.  

The river at St. Mawgan, as it happens, has a source at the major hillfort of Castle an Dinas, as stated here in the Wikipedia article:

The River Menalhyl (Cornish: Dowr Melynheyl, meaning river of the estuary mill) is a river in Cornwall, England, that flows through the civil parishes of St Columb Major and Mawgan-in-Pydar. Its length is about 12 miles and it flows in a generally north-west direction. The name comes from the Cornish words melyn meaning mill and heyl meaning estuary - estuary mills. The name was recorded as Mellynheyl in the 19th century, but it had been known as Glyvion.

The source of the river's longest branch is near Nine Maidens, about 2 miles to the north of St Columb,50.447°N 4.943°W, at a height of approximately 490 feet above mean sea level. The Menalhyl enters the sea at Mawgan Porth on the north coast of the county, 50.465°N 5.027°W.

Tributaries to the river include one that rises between the village of Tregonetha and Castle an Dinas, and one whose source is south-east of Tregatillian. A smaller river, from Talskiddy, joins the Menalhyl a short distance from its mouth.

The river flows through the settlements of St Columb (the north part of the town, known as Bridge), St Mawgan and Mawgan Porth. Its valley, from St Columb downstream, is called the Vale of Lanherne; it is wooded for much of its length and is popular with walkers.

There are a small number of disused mills along the length of the Menalhyl, and the river feeds some small lakes and ponds. It is also used for agricultural purposes along most of its course.

The river is fished by anglers, especially in its lower course, and fish species such as brown trout, sea trout, lamprey, eel and minnow have been recorded.

The bedrock of the catchment area largely consists of Devonian calcareous slates, grits and limestone.

Castle An Dinas, Cornwall

This is one of only two possible Kelliwics in the Bodmin region.  Here are some good sites describing this monument:

http://www.cornwallheritagetrust.org/our-sites/castle-an-dinas/

http://www.historic-cornwall.org.uk/a2m/iron_age/hillfort/castle_an_dinas/castle_an_dinas.htm

https://www.pastscape.org.uk/hob.aspx?hob_id=430680&sort=4&search=all&criteria=castle-an-dinas&rational=q&recordsperpage=10

Very near the town of Bodmin itself is another remarkable Cornish fortification: Castle Canyke.

Castle Canyke, Cornwall

https://www.pastscape.org.uk/hob.aspx?hob_id=431339&sort=4&search=all&criteria=castle%20canyke&rational=q&recordsperpage=10

https://historicengland.org.uk/listing/the-list/list-entry/1006660

This fort is near some Callywith place-names.  

Castle Canyke Next to Callywith

I would mention that in "The Dialogue of Arthur and the Eagle", Eliwlad son of Madog son of Uther is placed in the wooded Glynn of Cornwall - and it is here that we find the Madoc/Madog place-name nicely displayed.  Eliwlad takes the form of a spectral eagle in an oak and many have remarked that he is, in this sense, very much like the dead god Lleu in the MABINOGION.  I think this observation a very keen one, as we find ancient Madog  names in Nantlle, Gwynedd, where Lleu perched in eagle-form in his oak tree. [I have etymologized Eliwlad as 'grief-lord'; see https://mistshadows.blogspot.com/2020/04/the-etymology-of-eliwlad-final-take.html]. 

Given this contrast and comparison of the Welsh Triad and Geoffrey of Monmouth, I can now say with some confidence that Castle Canyke is almost certainly the Kelliwic of Arthur.  

One caveat, though; as the name Arthur could have been construed as containing W. arth. 'bear', it may be that his being placed in a fort called the 'forest grove' was merely an instance of poetic license.  For where else would the bear be, if not in the woods?

On the other hand, I've shown that Arthur's father descended from the Ui Liathain and, according to the Irish sources, this tribe was in Cornwall as well as being in southern Wales.  

Incidentally, Castle Canyke is quite close to the River Camel, Geoffrey of Monmouth's choice for the site of Camlan.  The etymology of Camel is Cornish Cambull, identical with Welsh cam + pwll, 'crooked stream.'  It is not related to Camlan as either *Cambolanda or *Camboglanna, except in so far as the first elements of these place-names are the same.

In addition, there is Tremodret, 'Modred's farm', in Roche (a Domesday manor) only a dozen kilometers southwest of Castle Canyke.  

IS KELLIWIC IN CORNWALL A LATE, LEGENDARY RELOCATION OF SILCHESTER?

The following selection on Silchester during the sub-Roman period is from An Age of Tyrants: Britain and the Britons, A.D. 400-600 By Christopher A. Snyder (Pennsylvania State University Press, 1998):

This naturally begs the question as to whether Arthur was actually centered at Silchester, and not in far southwestern Cornwall.  Certainly, if we look at his battles as I've laid them out, Calleva Atrebatum would have very much been at the center of things.

Could Geoffrey of Monmouth have (inadvertently) been right about Silchester after all? An interesting possibility, certainly.