Non IE words in early Celtic

[Xavier Delamarre]

Reply to Rick Mc Callister :

1/ *abol- (*a:b- with Lex Winter) "apple" : Italic, Celtic, Germanic,
Balto-Slavic, possibly Thracian or Dacian (the gloss dinupula <
*k'un-a:bo:la:). Enough to have IE status. Considered as non-IE because of
-b-, itself considered as non-IE, circular.

2/ *aliso "alder": add Slavic jelicha, Russ. _ol'cha_, lithuanian
_alksnis_, Latin _alnus_ (*alisnos). Enough to have IE status.

3/ Gaulish _bra:ca_ : latin _suffra:go:_ "jarret" (<*bhra:g- "cul"), cf
after Schrader, O. Szemenenyi (An den Quellen des lateinischen
Wortschatzes, 117-18).

4/ Gaulish _bri:ua_ "bridge" (cf place names Caro-briua, Briuo-duron etc.),
germanic *bro:wo: & *bruwwi: (> bridge etc.), OCS _bruvuno_ "poutre,
rondin", Serbian _brv_ "passerelle" (prob. the original meaning).

5/ Germanic _bukkaz_, clearly an expressive gemination of *bhug'o- (Avestic
_buza_) etc.

6/ Gaulish _dunum_ "enclosed place", Germanic *tu:na- (> town, Zaun etc.) ;
C. Watkins, Select. Writ. 2, 751-53, has related Hittite _tuhhusta_
"finish, come to an end, come full circle" (cf Latin _fu:-nes-_ < *dhu:-).
A possibility : I am a bit suspicious because it is a root-etymology.

7/ *gwet- "resin" : not only Celtic-Germanic, (Latin _bitu-men_, may come
also from Osco-Umbrian), but O.Ind. _jatu-_ "Lack, Gummi", Mayrhofer EWAia
1, 565.

8/ Gaulish (& Celtic) _i:sarno_ "iron" : explained convincingly by W.
Cowgill as the regular reflex of IE _*e:sr-no-_ "the bloody (red) metal"
(Idg Gramm. 1,1).

9/ Gaulish canto- "edge, circle" : explained by O. Szemerenyi, Scrip. Min.
4, 2036-38, as _*kmto-_ from a root _*kem-_ 'cover'. A possibility.

10/ "lake" : add Greek _lakkos_ <*_lakwos_ "cisterna", OCS _loky_ "id." ;
for the alternance a / o (as for the word 'sea' *mori/*mari), JE Rasmussen,
Studien zur Morphophonemik der idg Spr. 239-40, has proposed an explanation.

I hope to be able to present a more detailed account of these etymologies
in my "Dictionnaire de la langue gauloise", to appear in the coming years.

I do not believe very much in "Nordwestblock" theories (Hamp, Huld,
recently Beekes, who does not think that *ab- "water" and *teuta: "people"
are  IE), taken from Meillet's "vocabulaire du nord-ouest" (by the way) :
you can prove the antiquity of a designation (by a set of correspondences),
but not draw conclusions in reason of its absence (argumentum e silentio)
in a given dialect. The old problem of dialectology.

 A good example has been given recently : a lot of speculation, with
sociological consequences, had been produced from the fact the Celtic had
no representant of the canonic IE words for son & daughter (*su:nus,
*dhugHte:r). The later word is now attested in the Plomb du Larzac as
_duxtir_ ; by pure chance. the same is true of _lubi_ "love" (no trace of
this IE root in insular Celtic) or _deuoxtonion_ gen. plur. "of Gods & Men"
(no trace of dvandva compositum in ins. C.). etc.

Xavier Delamarre
Ljubljana