The text of Gildas' De excido et conquestu Britanniae is the best contemporary record of sixth-century Britain. As such, historians searching for a real historical Arthur often use this source to validate their interpretations and theories of his lfe, even though Gildas does not mention Arthur by name.
The central controversy of the text lies in the translation of a passage in chapter 26, as this dates Gildas' birth and the battle of Badon Hill:
From that day, sometimes the natives, and sometimes their enemies, prevailed, till the year of the siege of Badon hill, when they made no small slaughter of those invaders, about forty-four years after their arrival in Britain. After this, sometimes our countrymen, sometimes the enemy, won the field, to the end that our Lord might this land try after his accustomed manner these his Israelites, whether they loved him or not, until the year of the siege of Badon Hill, when took place also the last almost, though not the least slaughter of our cruel foes, which was (as I am sure) forty-four years and one month after the landing of the Saxons, and also the time of my own nativity. (Bede, Historia Ecclesiastica, Book 1, chapter 16)
Bede originally translated this passage, interpreting that the battle of Badon Hill came 44 years after the arrival of Saxons. However, this interpretation has been widely discredited by historians. At least two translations are now currently fashionable (although only the Bede version is without copyright resulting in many theories of Arthur on the internet) giving the following outcomes: Translation 1) Gildas wrote the De excido et conquestu Britanniae 44 years after his birth and the battle of Badon Hill. or Translation 2) that Gildas wrote that his birth and Badon Hill came 44 years after a earlier victory over the Saxons.
As the battle of Badon Hill is mentioned in the Annals Cambriae as in the year 516 (sometimes given as 518 AD due to different dating methods), some historians using translation 1 have attempted to place a historical Arthur in a context with the De excido et conquestu Britanniae as being written around 560 AD.
Most historians however agree that the De excido et conquestu Britanniae was written around the 540s. This presents a problem for those using translation 1. Their solution has been to place Gildas' birth and the battle of Badon Hill c.500 AD, ignoring the date of the battle mentioned in Annals Cambriae, passing off the Welsh Annals as unreliable. D.F. Carroll's Arturius: a Quest for Camelot is an example. Carroll concedes a battle of Badon Hill c.500 AD, but disputes it involved Arthur. This leads him to dispute Nennius's 12 battles. He does however assert a battle of Camlann but dated at c.582 AD to fall into his suggestion of the later Arthur mac Aedan as his 'King Arthur'. If this was indeed the case then the Annals Cambriae must indeed be way out; Badon being dated too late and Camlann far too early!
The Annals Cambriae mention the battle of Camlann, and the death of Arthur in 539 as:
The strife of Camlann in which Arthur and Medraut perished; and there was a plague in Britain and Ireland.
This mention of Arthur means that most historians searching for a historical Arthur choose not to ignore the battle. Some use the date and entry even though they had just dismissed the Annals Cambriae as unreliable! The problem then arises that a 37 year gap between Badon and Camlann seems unlikely for a historical Arthur. Nennius' Historia Brittonum has Badon as the last battle of 12, indicating another 11 battles before Badon fought before 500 AD further undermining their case. (Most historians like to use Badon and Nennius as this obviously strengthens the case for a real historical person.Carroll, by naming his Arthur as Arthur mac Aedan, paradoxically has no such worries.) Also the sixth battle of Nennius, the battle of Bassas may be tentatively dated to 508 AD if the Cambuslang placement is accepted. All of this implies that the 500 AD date for the battle of Badon and the birth of Gildas is incorrect.
A novel approach for those wanting to include the Annals Cambriae entry is to assume that the dates of the 'unreliable' Annals are over by about 16 years, thus placing Badon at 500 and hence Camlann at c.520 AD; fitting in with their solution of translation 1. The closer dates also make a historical Arthur more likely. Alistair Moffat, another exponent of a Scottish King Arthur, does exactly this in his Arthur and the Lost Kingdoms, though the book remains a worthwhile contribution to the debate. The Annals entry for 539 has however another point in its favour, it mentions a plague. Dendrochronology of Irish Oak trees has revealed a climate change from about 535, possibly a precursor of the plague in the 540s, which is well documented in other annals. The Annals Cambriae are also confirmed by the Annals of Tigernach in stating Gildas' death in 570 AD; it is strange, then, to find Carroll trusting these Annals instead.
All this confirms the validity of the Annals Cambriae and given that Gildas did write De excido et conquestu Britanniae in the 540s indicates that the correct interpretation is that Gildas was born in 516, the same year as the battle of Badon Hill. This is consistent with translation 2.
Unfortunately the mass of theories that the differing translations of a single passage of De excido et conquestu Britanniae has produced may result in a universally accepted real historical Arthur forever being elusive. I thus hope that my theory is accepted as a valid proposition.