… But the rule set’s title is emblematic of the long-standing idea that the phrase dux bellorum, which appears in the (in)famous chapter 56 of the Historia Brittonum (the so-called Battle-List of Arthur) is a title. ‘War duke’ suggests Dan Mersey in the book and an interview in the glossies.
It isn’t a title; it is a simple descriptive phrase, and the proof of the matter lies in the structure of the passage. The chapter, as I demonstrate at greater length in Worlds of Arthur, is not – in fact it cannot be – a fragment of a lost Old Welsh poem about Arthur. It is a Latin composition dating to 828/9, when the Historia was written. This can be demonstrated by an analysis of the whole chapter’s structure, i.e. including the opening and closing sentences, about the Saxons, which are usually omitted. Those sentences are usually left out because they don’t match the ‘lost poem’ hypothesis, but there is no textual reason for so doing and a close analysis shows that they are an integral part of the whole.
Essentially, the chapter is written in what is called a chiastic structure or chiasmus. That’s to say that it is set out according to the pattern ABCBA, that is to say as two halves that mirror about a crux, where the (literally) crucial message is placed (i.e. at ‘C’ in the ABCBA structure). Here the crux is the battle of Castell Guinnion, the description of which contains as many words as the preceding section about the first four (or seven) battles or the subsequent section about four battles. Sadly, we don’t know where or what the battle of Castell Guinnion was… [I have a hypothesis – not in the book – that it is a mock antiquarian name for Winchester, the royal seat of the West Saxons who had taken over Mercia and were threatening Wales in the very year in which the Historia was written.] The sections are balanced by adding three further battles at Dubglas, the Welsh name of the Caledonian Forest, and so in, in the first half. Apart from the chapter’s structure and the balancing of the sections in the numbers of words they contain (in Latin), there are overall numerological points. The whole chapter is 240 words long, which is 20 times the 12 battles (or alternatively 30 times the eight [in my view] that the author actually knew about…), and 1/4 of the 960 men that Arthur allegedly felled in one charge.
None of this means that the poet didn’t have a poetic source which described the battles of Arthur (albeit only eight, rather than twelve, in my view) but it is equally possible that he drew the names from a number of sources (I think Badon is added to the end of the list on the authority of Gildas and Bede, who said it was ‘nearly’ the last battle in a war that Bede thought was against the Saxons, although actually Gildas never says so) … or indeed (whisper) that he made all of them up, apart from Badon. The facts that the passage is very carefully structured in Latin and that that structure goes beyond Chapter 56 itself to that of the whole work, where ch.56 occupies the hinge between the HB‘s discussion of the South and that of the North, mean, however, that it is impossible to see beyond the HB-author to whatever his sources were and what he’s done to them, not least in rendering them into Latin. They rule out the possibility of the central lines of HB 56 being a fossilised fragment of a lost Welsh poem.
So, where does that leave dux bellorum? A couple of points. One is that it is close to the phrase dux belli (leader in war), which is the phrase used by the Old Testament (Judges 1.1) about Judah’s place within the tribes of Israel. My theory in Worlds of Arthur is that the HB is an appeal to King Merfyn of Gwynedd to unify and lead the Welsh against the English. In that reading Gwynedd would be a figure of Judah. Another is that it is close to the phrase used by Constantius of Lyon to describe Germanus of Auxerre at the ‘Hallelujah Victory’: dux proelii (leader in the battle). Germanus was clearly a political touchstone in ninth-century Welsh politics and he is especially important for the HB. The third point is that, within the chiastic structure of HB 56, dux bellorum has its mirror in the phrase in omnibus bellis victor (victor in all wars/battles), which appears in the mirroring passage of the chapter. Now, dux was a rank in the late Roman army, for sure (as in Dux Britanniarum: ‘The Duke of the Britains’, the name of another recent rules-set, as it happens), but it was also a simple noun – leader, guide, head.
So, dux bellorum in HB 56 simply means that Arthur was the Britons’ leader in the wars, just as he was, later in the passage, the winner in all the wars. Not a title, then. Not ‘war-duke’ (sorry, Dan) – just a plain description with (in Latin) some ninth-century biblical/Christian and political resonances.