- Chapter Thirty-Six -
- subsidies, trade, even office within the empire - but they also desired to remain
distinct.^98 The ogam alphabet was created under just these twin impulses of attraction
and repulsion. It was to be an alphabet fit for the commemoration in stone of a
king or a protector. In Wales and the south-west it normally appears alongside
Latin.^99 Sometimes the ogam has the same information as the Latin,IOO sometimes it
has different information,101 sometimes less.^102 What it never has is more information.
Moreover, there are one or two inscriptions in which the Latin seems to have influ-
enced the wording of the Irish.^103 As an inscriptional language, therefore, Irish
remained the junior partner. Yet the contrast with Welsh remains: Welsh was not a
partner at all.
The same relationship of semi-independent and junior status is exemplified by one
of the central texts of early Irish learning, Auraicept na nEces, 'The Primer of the
Poets'.104 It is the business of the Auraicept to raise Irish to the same level as Latin; and
grammar is to be the means by which this elevation of status is to be achieved. Latin
has so many letters; so does Irish. Latin has its declensions; so does Irish. In the
Auraicept, the writer of Irish is always looking over his shoulder at his elder brother,
the Laitne6ir. The only early medieval counterpart to the Auraicept is the grammar of
his native language commissioned by Charlemagne;10S in that instance, too, the
purpose was plain: the language of the Franks was, like Latin, to have the dignity of
grammar. But, unlike the Auraicept, Charlemagne's grammar does not survive.
It is difficult to believe that a principal stimulus to this elevation and cultivation
of Irish as a partner to Latin was not the experience of the Irish settlers in Britain in
the last period of Roman rule. If I had to guess where the ogam alphabet was
invented, I should, without hesitation, opt for South Wales.^106 The example of Gildas
demonstrates that in post-Roman Britain, alongside spoken Latin, there was a gram-
matically correct and highly wrought form of the language - a Latin in which a
prophet might denounce the sins and follies of kings and bishops.107 The gap
98 For examples see the account of the Alamanni in J.F. Matthews, The Roman Empire of
Ammianus (London, 1989), 306-18, and that of the Goths in P. Heather, Goths and
Romans, 3]2-489 (Oxford, 1991), esp. 12I.
99 ECMW, no. 300 is an exception.
100 ECMW, nos. 71, 142.
101 ECMW, no. 70.
102 ECMW, nos. 84, 12 7, 354.
103 ECMW, no. 142 and possibly no. 169.
104 Edited by A. Ahlqvist, The Early Irish Linguist: an edition of the canonical part of the
Auraicept na nEces (Societas Scientiarum Fennica, Commentationes Humanarum
Litterarum, 73; Helsinki, 1982).
105 Einhard, Vita Karoli, c. 29; Einhard's inchoauit perhaps suggests that the plan came to
nothing.
106 Following Jackson, Language and History in Early Britain, p. 156, in preferring Britain to
Ireland; South Wales is chosen simply because it is the part of Roman Britain in which
there was the most extensive contact between Irish and Latin.
107 M. Lapidge, 'Gildas's education and the Latin culture of sub-Roman Britain', in Gildas:
new approaches, ed. M. Lapidge and D. Dumville (Woodbridge, 1984), 27-50, and
F. Kerlouegan, Le De Excidio Britanniae de Gildas: les destinees de la culture latine dans
l'ile de Bretane au VIe siecle (Paris, 1987), ch. I.
722
