- Language and Society among the Insular Celts AD 400-1000
Latin, was seen as a single British tongue: Combrec, 'the local language', was
therefore Bretnas, 'the language of the Britons'. Formally, Cymraeg is derived from
Cymro; semantically, Cymro, meaning 'Briton' rather than 'fellow local', is derived
from Cymraeg. It was Cymraeg 'British' which ensured that Cymro meant 'Briton'.
With this in mind we may turn back to Dyfed, and to the sixth century to the
time of Voteporix, perhaps Gildas's Vortiporius, Demetarum tyrannus.^62 He is
commemorated in a bilingual (Latin/Irish) inscription at Castell Dwyran in the west
of Carmarthenshire in the heartland of early medieval Dyfed.^63 The Latin, in square
capitals, runs as follows: MEMORIA / VOTEPORIGIS / PROTICTORIS. The
other, containing only the name, is in the ogam alphabet and in an early form of Irish,
VOTECORIGAS. In what is, therefore, the commemorative stone of a person of
very high rank,64 even if he is not Gildas's ruler of the Demetae, Irish is admissible
alongside Latin; Welsh does not appear at all.
In discussing the term Cymraeg, I have up to now tacitly assumed that the only
vernacular language spoken alongside Latin was British. For most of Britain for most
of the Roman period, that was true; but along the western seaboard, including Dyfed,
it ceased to be true in the fourth and fifth centuries. A further problem is that it is
not clear for how long Latin was still a spoken language in Britain. On both these
points the inscriptions offer decisive evidence.^65 They show quite unambiguously the
62 Cf. D. Dumville, 'Gildas and Maelgwn', in Gildas: new approaches, ed. M. Lapidge and
D. Dumville (Woodbridge, 198 4), 57.
63 ECMW no. 138.
64 On the term protector see A.H.M. Jones, The Later Roman Empire (Oxford, 196 4), 53-4,
597,636-40; it was used for one of the two corps of elite bodyguard troops (domestici and
protectores) entitled 'to adore the purple'. In effect, the protectores served as staff officers
and were likely to receive high military office as the next step in their careers. Cassiodorus,
Variae Xl.3 I (ed. Th. Mommsen, MGH, Auctores Antiquissimi, XTl.348), is the formula
by which, for Ostrogothic Italy, a primicerius singulariorum (head of the messenger ser-
vice) was advanced to the company of the domestici et protectores, in that he was sacram
purpuram adoraturus. This is an example of the practice whereby deserving veterans or
officials received the honour of adoring the sacred purple on retirement. As the
Ostrogothic example shows, the practice could survive in a barbarian kingdom. If
Voteporix was Gildas's tyrant of the Demetae, it would be surprising to find him being
called a protector since that would imply a recognition of imperial authority of some kind,
such as the purple worn by the parents of Ambrosius Aurelianus, Gildas, De Excidio, 25.3.
On the other hand, it would also be surprising if there were someone called Voteporix
flaunting the right 'to adore the purple' in the heart of Vortiporius's kingdom. He might
conceivably have been a kinsman of Vortiporius (so explaining the close similarity of the
names) of a slightly earlier generation (so explaining the adoration of the purple).
65 On dating see Jackson, Language and History in Early Britain, pp. 158-62; I do not,
however, share his opinion that the Catamanus stone is likely to be later than C.625: square
capitals did not evolve into half-uncials within Britain; rather square capitals were replaced
by half-uncials as a form of lettering appropriate for inscriptions. There is no need to allow
time for a gradual evolution of letter-forms; since the change was a matter of fashion, it may
have happened very fast. It certainly extended to the whole of Wales. Jackson's appeal to the
date of foundation of Llangadwaladr depends on the text of Bonedd y Saint printed in the
Myfyrian Archaiology. The reference there to Llangadwaladr is editorial: see Bartrum's edn,
EWGT,p·5 6.