- Chapter Two -
Korolev 1989; Koch 1983a, 1983b, 1985, 1992; Lambert 1994; McCone 199Ia, 1991b, 1992;
McManus 1991; Meid 1989a, 1989b, 1992; 1993; Schmidt 1977a, 1977b, 1979, 1981, 198 3,
1988a, 1990a, 199ob, 1993; Tovar 1986, 1987; Untermann 1983. This highly selective list
has, perforce, been limited to some recent work only.
2 See Ardener 1989: 217.
3 In Ardener 1989: 226.
4 See, for example, Evans 1988, 1992.
5 See, inter alios, Gamkrelidze and Ivanov 1984, 1990; Mallory 1989; Markey and Greppin
1990; Meid 1989b; Zimmer 1990a, 1990b.
6 In Markey and Greppin 1990: 24.
7 In Markey and Greppin 1990: 338. See also Zimmer 1990a and 1990b.
8 On the identification of Celtic, also on the rise of modern Celticism, see Evans 1992b and
1993·
·9 There are similar difficulties with early Insular Celtic. See, for example, Evans 1990;
Schmidt 1993.
10 See especially Evans 1979: 51 1-23 on some of the problems connected with the identifi-
cation and labelling of early Celtic linguistic areas. For Hispano-Celtic see, for example,
Schmidt 1976; Gorrochategui 1991; Evans 1993.
II See Weisgerber 193 I: 173; 1969: Bf.
12 See Lejeune 1971: I22f.; 1972a: 265; 1972b.
13 On fragmentary languages see Untermann 1980.
14 See Zeuss 1871: vi.
15 See especially Fleuriot 1978, 1980, 1981, 1982, 1991; Fleuriot and Giot 1977; Evans 1979:
525; Schmidt 1992: 193f. For the views of Falc'hun see also Jackson 1967: 30-2.
16 The complex question of relationship with non-Indo-European and probably pre-Celtic
languages cannot be faced here. See Wagner 1959, 1969, 1976, 1977, 1982, 1987; Evans
1983b: 951-4; Kalygin and Korolev 1989: 20f.; Shisha-Halevy forthcoming: chapter 4,
section 7 ('Celtic typology and affinities: a personal standpoint').
17 The dating of the continental evidence overall is uncertain. For a general statement see
Lejeune 1972a: 266 (see also 265); Lambert 1985.
18 Here one has to be exceedingly careful as to what one regards as the earliest Insular evi-
dence. For all the challenging work on archaisms in relatively early Insular sources (e.g.
work on early Welsh poetry and on Archaic and Early Old Irish), the difference in date
of these sources from that of so many of the most important Continental Celtic sources
needs to be recognized and respected more than it has been in some recent work in the
quest to set up new linguistic syntheses. See, for example, Evans 1990; Koch 1992; Schmidt
1993·
19 See, for example, Schmidt 1976, 1977a, 1980, 1986a, 1988b.
20 I quote here from Schmidt 1988a: 235-6.
21 See Kalygin and Korolev 1989: 6. I am much indebted to Mr David Howells for his kind
permission to quote from his full English version of Kalygin and Korolev's work.
22 See Kalygin and Korolev 1989: 8-9.
23 See Szemerenyi 1978: 296.
24 See McCone 1991a, 1992.
25 This form, on which there is by now an extensive literature, was first identified by
Professor Oswald Szemerenyi in a brilliant resegmentation of an Old Gaulish formula (see
Szemerenyi 1974).
26 See McCone 1991a: 50; also 1992. McCone has consistently tended to favour the view that
there is a fundamental division between Insular and Continental Celtic. Warren Cowgill
had placed some emphasis on the distinctive character of Insular Celtic in general in his