Íslenskt mál og almenn málfræði - 01.01.1985, Blaðsíða 104

102
Jörundur Hilmarsson
quent divergent treatment of the dental occlusive, cf. ftn. 4 above). This
development could have taken place in either of two ways.
Either the laryngeal, being in an intervocalic position, was dropped
in the sequence VHnC. That is, I.-E. *H2ueH,ntó-/
*suH2nto- > wentó-/*sunto- > wentó-/*súnto- > winda-/*sunþa-, cf.
Hoffmann (1976:383, ftm 13).
Or, the sequence VHnC was restructured as VHnC as evidenced in
Toch. A want, B yente ‘wind’ < Common Toch. *w'œntœ < Pre- or
Proto-Toch. *wénto- < I.-E. *H2ueH,ntó-, cf. further discussion of the
Tocharian forms below.
Thus restructured, I.-E. *H2ueH,ntó-/*súH2nto- would yield Pre- or
Proto-Gmc. *wéntó-/*sunto-, which through the effect of Osthoffs law
would result in Gmc. *winda-/*sunþa-.
The use of the suffix *-ön for individualizing derivation is widely
attested in Indo-European. Such derivation from thematic stems was
quite common, cf. Lat. catus ‘sharp’ -> Catö (personal name), Gk.
axpapó<; ‘cross-eyed’ -» axpápcov ‘a cross-eyed person’, cf. e.g. Hoff-
mann (1976:379).
Through the addition of a such individualizing ön-suffix, the I.-E.
adjective *suH2nto- ‘sunny’ might have produced the derivative
*suH2ntön ‘the sunny one’, which naturally would have been applied to
5 On the other hand, Scherer (1953:50) thinks that tsunnö is formed through the addi-
tion of the suffix *-ná to the stem *sun-. Beekes (1984:8) merely states that Goth. sunno
is based upon *sh2un-. Hamp (1975:101) does not comment upon the double -nn- of
*sunn-.
6 Similarly, Gmc. *brunnö ‘well’ (Goth. brunna, OHG. brunno/brunna etc.) may
have replaced earlier *brunþö on the analogy of a gen.sg. *brunnez < *brunþnez. How-
ever, this is more hypothetical, for there is no evidence for a basic thematic *bhruHnto-
producing a derivative *bhruHntön as *suH2nto- produced *suH2ntön. ModHG. (dial.)
brunzen ‘to urinate’ might hint at the possibility of such formation, but that is a too
slender basis.
Of course, an athematic n/-formation is found in Gk. (ppéap, gen.sg. (ppéötoi; ‘well’
and possibly in Mlrish tipra ‘well’ < Olrish *tiprar < *to-el<s-bhreur, gen.sg. tiprat <
*to-eks-bhréuntos, cf. (somewhat hesitatingly) Pokomy (1959:144), Vendryes (1978:74,
110).
7 OHG. hunno ’centurion’ is taken by Brugmann (1906:303) as representing Gmc.
*hundö replaced by *hunnö on the basis of a gen.sg. form *hunnez < *hundnez. One
might then wonder whether this form bears witness to a similar process as here postu-