- The Early Celts -
spread of the p (> kW) isogloss to Celtic Britain from northern Gaul (spreading
throughout British Celtic), he suggests that it was 'prevented from going further
by the Irish Sea rather as the barrier of the Pyrenees hindered its penetration from
Gaul into Spain'.3! This is a quite risky analysis and appears to stem in part from the
questionable view that we can identify a commonality of language in various areas of
the Celtic terrain in late prehistory or in the early historic period. Comparable with
McCone's approach, but differently argued, is Professor John T. Koch's revision of
his 1983 Oxford Celtic Congress paper in a paper in the Leon Fleuriot Festschrift,32
in which he argues that Celtic speech was established in Ireland and Britain 'very
early in the Iron Age, before the penetration of La Tene influences' (here he relies
heavily on an ingenious reconsideration of the significance of the earliest place and
ethnic names) and assumes a gradual separating out of the languages of Britain and
Gaul. He too claims that the archaeologists' 'La Tene' influences correspond to
what the linguists would term 'Gallicization'. He has now distanced himself from the
view that European mainland iron age proto-languages (later than Proto-Celtic) are
a significant factor in divergences from Gaulish in what he calls 'the Insular Celtic
phenomenon'.
My discussion has in part drawn particular and deliberate attention to some of
the recent proliferation of generally inconclusive hypotheses and syntheses based
on some of the earliest extant evidence concerning the Celtic languages.^33 However, I
am certain that the commitment of so many scholars to contribute towards a more
reliable and comprehensive analysis of this evidence indicates a growing awareness of
its primacy and importance. Those who are well informed about it know that it can
no longer be said to be insubstantial or arid. Recent detailed work on it includes, for
example, discussion of the local and ethnic names of Roman Britain, the language of
early British coin legends, a whole range of aspects of the study of Ogam inscriptions,
the gathering in and careful editing of large corpora of primary sources, detailed
discussion of a host of often very difficult texts, the study of various early epigraphic
traditions, the development of mixed languages and the erosion and demise of Old
Celtic in so many areas. All this work has added much to our knowledge of early
Celtic linguistic patterns. It has shown anew, more often than not in confirmatory
fashion, the wide dispersion of speakers of Celtic languages in antiquity. Increasingly
it has revealed for the first time comparatively early linguistic features which could
not be recovered by processes of linguistic reconstruction from the study of later
sources.
The complex early linguistic evidence is of great importance. It is a vital part of
our knowledge of the early history and culture of the Celts and, treated cautiously
and critically alongside the testimony of so many other sources, it greatly enriches
our understanding of that many-faceted history and culture.
NOTES
I Here I can note only a small selection of references to general surveys relevant for our
discussion: de Bernardo Stempel 1991; Eska and Evans 1993; Evans 1977, 1979, 1983a,
1983b, 1988, 1990, 1992b, 1993; Fleuriot 1981, 1988, 1991; de Hoz 1986, 1988; Kalygin and13