The Marwnat Geraint  (Elegy for Geraint) is a poem commemorating a King Geraint who was slain at a Battle of Llongborth. In the poem, the place name is spelt Llogborth and Llogporth in the dative. Its greatest value nowadays is that even the most Arthur-disbelieving historians accept it as evidence of the honouring by Celtic Britons in the eighth century of a past warrior leader called Arthur.
A coherent interpretation is that Llongborth is Langport in Somerset; that the battle is the one the Anglo-Saxon Chronicle records as having been fought by King Ine of Wessex against a British King Geraint in 710; and that this Geraint is the same man as the King Geraint of Dumnonia known to history from correspondence with Aldhelm, the first bishop of Sherborne, dated to 705. (Ultra-sceptics, of course, regard none of these points as certain facts.)
The poem describes Geraint as a “brave man of the region of Devon” (guir deur o odir diwneint), and warriors who fought for him are called “Arthur’s brave men” (y Arthur guir deur). This need not mean that Arthur was present at the battle! Rather, it is saying that men who were heirs to Arthur’s knights fought for Geraint, in much the same way as warriors of Gwynedd might be called ‘Cunedda’s men’ in honour of their kingdom’s founder.
So how might it be that warriors fighting for a king from Devon were called “Arthur’s braves” if Arthur was King of Lindinis? Firstly, it is appropriate to equate Diwneint with Dumnonia. In Roman times Dumnonia was the civitas to the west of Lindinis. It included all of Cornwall and all of modern Devon except the northern part that was in Lindinis. Diwneint (also spelt Dyvneint), the British name that became anglicised to Devon, is the word that was previously Latinised into Dumnonia. The civitas and later kingdom is more commonly called Kernow in its own language, Kernyw in Welsh, Cornwall in English.
The probable explanation is that by the turn of the eighth century much of Arthur’s Lindinis had fallen under English rule. The Anglo-Saxon Chronicle boasts of a West Saxon victory in 658 in which the British were “driven to the Parrett”; and in 661 of the West Saxon king fighting a battle at Posbury, near Crediton. It, significantly, does not say that Kenwal won. But Hoskins (in The Westward Expansion of Wessex) indicates that the “misdating” of charters for church land at Crediton to 680 implies that he did not lose it either. Hoskins shows that by 680 Wessex appears to have gained sovereignty over the southeast corner of Devon and a salient through Pinhoe to Crediton. Glastonbury Abbey records give strong indications that the Abbey came under English control in the late seventh century, also probably during the reign of Kenwal (d. 685).
It is therefore a conjecture consistent with what is known and what is probable that at the turn of the 700s the part of Lindinis which remained in British control was only its districts to the west of the Parrett – in modern terms, approximately the western half of South Somerset, Taunton Deane, Exmoor, and northern Devon north of the River Taw. Somerset was divided between a British-ruled west and an English-ruled east. The remnant of Linidinis, even if it retained its own sub-king, provided its warriors, “Arthur’s braves”, as fighting men arm in arm with the kingdom of King Geraint of Dumnonia. The Marwnaw Geraint poem does not indicate if its author regarded Geraint himself as among Arthur’s brave men.
It is not that the partition of Somerset was negotiated as a treaty or upheld as an agreed border by either side. It is to be thought of as more like a, forever temporary, de-facto ceasefire line which held for about a century after 661 because both kingdoms, British Kernow/Dumnonia and English Wessex/Gewissae, successfully defended their territories when attacked.
On both sides, the balance of military capacity was in favour of the defenders. The location of the battle of Llongborth suggests that King Geraint and “Arthur’s brave men” tried to regain part of the lost kingdom of Lindinis (later dimly remembered as Lyonesse), crossing the River Parrett – the presumed pre-battle boundary – which flows immediately to the west and south of Langport. The Parrett is separated from the town by a now-tranquil meadow, on which it is easy to imagine the battle was fought. Despite their eulogised valour, they lost.
A generation later in 733 (at Battle Farm?), Wessex defeated British forces at Somerton who were attempting to wrest from them control of eastern Somerset in what historical hindsight tells us was their last ever (known) incursion across the Parrett frontier. Equally, though, the English were unable to advance. The alliance or union between Lindinis and Kernow must have been effectively sustained. In the early 700s, Wessex’s attempt to hold a fort west of the Parrett at Taunton failed; and in 722 its attempt to advance westwards along the north coast was defeated at the Battle of Pencon, which I figure was at Quantoxhead (see note 11 of my 9th March 2013 post). As I noted also in an earlier post, according to the Watchet historical website http://www.watchetmuseum.co.uk/saxons_vikings.php, the English only conquered this last part of Lindinis in 815.