Giraldus Cambrensis is cited as the authority
for the tripartite division of Britain (although they do not give a direct
reference to a work or chapter; in fact, it is Descriptio Kambrie i.2,
and his reference is to Kambria, not Britannia); they also claim
to find reference to Tria Regne Britanniae in Gildas and ‘Nennius’.
This is fantasy. Gildas contains no such statement (he merely says that reges
habet britannia – ‘Britain has kings’ – and proceeds to denounce
five of them); the Historia Brittonum Chapter 8 (attributed, rightly or
wrongly, to ‘Nennius’) says that tres magnas insulas habet (‘it has
three large islands’). Unless Blake and Lloyd believe that insulas here
is a mistranslation of Welsh ynys (in its figurative sense of
‘realm’), the context is of islands off the coast of Britannia.
Giraldus talks of the three divisions of Kambria
as Venedotia, Powisia and Sudwallia (the latter with the
alternative Welsh name of Deheubarth, the ‘right-hand part’). These
are clearly identifiable as Gwynedd, Powys and South Wales (interestingly given
a Latinised English name). He describes the origins of these divisions as a
result of the partition of Rhodri Mawr’s kingdom among his three sons in the
ninth century. The implication must be drawn that these divisions of Kambria
did not exist before the partition.
The three realms of Ynys Pridein are
named in the Welsh Brut as Cymru, Alban and Lloegyr.
Cymru is taken to be the Kambria of Giraldus and is certainly the
usual term for Wales down to the present day. Alban occurs as Albania
in Geoffrey of Monmouth and corresponds to the Gaelic Alba, meaning
‘Scotland’, although some Irish writers used the name for the whole of
Britain; it derives from Albion, the ancient name for Britain preserved
in a number of Classical texts, beginning with Pseudo-Aristotle in the first
century AD (who spells it ’Άλβιων). Lloegyr
is the Logres of Geoffrey, and is usually assumed to be England.
Blake and Lloyd, however, use Giraldus to
relocate the three names given in the Brut; in their scheme, Cymry
corresponds to Venedotia, Alban to Powisia and Lloegyr
to Sudwallia. This ignores the strong connections between the first two
names in the Brut and their modern counterparts. Moreover, it does not
explain why Giraldus uses Kambria for the whole of Wales whereas,
according to Blake and Lloyd, the Brut restricts its use to Gwynedd.
Their arguments about the locations of the three
realms depend heavily on the second ‘key’: the names of the principal rivers
and towns.