- Chapter Thirty-Six -
tion, reading upwards; the other, on the right, reading downwards. Only the left-hand
ogam text corresponds to the Latin. The other may be quite separate and may have
been cut at a different time. PVMPEIVS shows, in the first syllable, the Late Latin
confusion of u and o. Like the other Latin name it is in the nominative case. PO PIA is
an Irish form of the same name which has been assimilated both to the phonology and
the morphology of Primitive Irish.^81
In the next major kingdom to the north, Brycheiniog in the centre of South Wales,
two inscriptions, about five miles apart, each of which is bilingual, are of particular
interest.
Trallwng^82 Ogam: CUNACENNI [A]VI ILVVETO
Latin: CVNOCENNI FILIVS / CVNOGENI HIC lAC IT
Trecastle^83 Ogam: MAQITRENI SALICIDUNI
Latin: MACCVTRENI + SALICIDVNIIn the Trallwng ogam inscription the initial CUNA-shows, by his choice of -A-
rather than -0-(namely CUNA-rather than CUNO-), the development of Irish
whereby the link vowel -0-was lowered and unrounded to -a-; this -a-subsequently
lowered the preceding -u-to -0-and was itself lost (':'cuno-> ':'cuna-> 'fcona->
con-).84 This independence of the ogam-cutter vis-a-vis the Latin suggests that Irish
was for him a spoken language. On the other hand, on the Trecastle inscription a few
miles further west, the form of the name in the ogam inscription, MAQITRENI, is
more archaic than the Latin MACCVTRENI, since it preserves the old q, and i as
against the later c(c) and u.^85 Here the Latin form of the name - still, of course, an
Irish name even though in a Latin inscription -is closer to the probable pronunciation
at the time. In this inscription, therefore, we have evidence in the ogam of an Irish
orthographical tradition surviving in Brycheiniog independently both of the Latin
tradition and of the pronunciation of Irish at that period.
In the sixth century, therefore, both Latin and Irish were spoken languages in
Wales. Brycheiniog as well as Dyfed was a kingdom of three languages. Both Latin
and Irish enjoyed a higher social status than did British. By the ninth century,
however, neither Latin nor Irish were normal spoken languages used for all purposes.
Latin was a language of liturgy and learning; Irish was probably not spoken at all
apart from the occasional immigrant or the churchman who had studied in Ireland.
The best indication of the relationship of the three languages is the eighth-and8 I It shows -omp-> -ob-; the last syllable -lAC is probably for -ijah, the primitive Irish
nominative singular of a io-stem.
82 ECMW, no. 70, situated in the church at SN 965 295; Nash-Williams reads CVNACEN-
NIVI, following Macalister, CIIC, no. 342, but the first V is merely a slip for U. Jackson,
Language and History in Early Britain, p. I8s, suggests CUNACENNI [A]VI ILVVETO
and he is followed by D. McManus, A Guide to Ogam (Maynooth, 199 I), § 4. I I (p. 62).
83 ECMW, no. 71; CIIC no. 341. Nash-Williams reads MAQUTRENI but the U is a slip for
I, as shown by the drawing. The stone is now in the British Museum.
84 McManus, A Guide to Ogam, § 5-23.
85 Ibid. § 5.3^2 , 5-33.