Íslenskt mál og almenn málfræði - 01.01.1985, Blaðsíða 103
101
Reflexes of I.-E. *suH2nto-/-ön
A thus reconstructed I.-E. *suH2nto- would regularly lead to Gmc.
*sunþa- (cf. below), and so would *súH2nto-, if the loss of the inter-
vocalic laryngeal and the subsequent change of -n- to -n- took place
before Vemer’s Law, as assumed in ftn. 4 above.
This explanation of the accentuated zero grade of *sunþa- (->
*sunþra-) makes it possible to interpret the zero grade of *nurþa-/
*nurþra- as analogical, as indicated above.
The phonological development from I.-E. *suH2nto- (or *súH2nto-)
to Gmc. *sunþa- would be paralleled by that of Gmc. *winda- from
I.-E. *H2ueH,ntó- (for the difference in accentuation and the subse-
Also, through *ghimont-, a link with Skt. hemantá- is established without effort: the
latter is simply a regular vrddhi derivation.
Only the geminate -mmof Hitt. gimmant- remains problematic. Meichert (op.cit.,
p. 70, 146—147) derives gimmant- from * gheimnontfhe opts for the root form
*gheim- instead of *ghim-, but admits that this can be contested, for *ghim- is equally
possible) with -mm- through assimilation of -mn-. However, Oettinger (op.cit.) shows
that we here have a í-extension of a stem in -on- rather than a formation in -ont-, i.e. an
fl-stem plus -t-, but not n-stem plus -onl-. For the -mm- he then gives an accentual
explanation. The matter rests unclear.
As hemantá- to *ghimont-, so vasantá- might be a vrddhi formation to the athematic
stem found in the adverbial vasáhjáS the latter is understood as containing a restored
normal grade of the root (for earlier *usont-).
Furthermore, Gmc. *windaz ‘wind’, indicating an I.-E. *H2ueH,raó- (cf. Hoffmann
1976:383 ftn. 13, Bammesberger 1984:134—136), as well as *tindaz ‘tack’ from
*H,dentó- (Bammesberger loc.cit., Darms 1978:402—406), are understandable as
vrddhi formations to I.-E. *H2uH,nt- (cf. Hitt. huuant- < *H,uHint-, weak case stem, or
*H2uH,ont-, strong case stem) and *H,dnt- respectively. Darms’ claim (op.cit., p. 396)
that the Indo-European words for ‘wind’ can be united only under *H2uéH,nto- is
controversial. It may be conceivable (Darms, p. 516 ftn. 21) that Gmc. *windaz repre-
sents I.-E. *H2uéH,nto-, with -d- from -t- because the accent was not on the syllable
immediately preceding it. However, as Darms also points out, the loss of the laryngeal
in intervocalic position and the subsequent change of -n- to -n- must be a very early
procedure. One might assume that it was prior to Verner’s Law (so Bammesberger
1984:135) and implicitly Hoffmann (1976:383 ftn. 13). Therefore a likelier reconstruc-
tion for Gmc. *windaz would be *H2ueH,ntó-. And, of course, if this Germanic word
continues a restructured *H2ueH,nto- with asyllabic -n- (cf. discussions further on in
text), the accent can have been only on the final syllable.
The form responsible for the frequent assumption of an initial accent in this word,
Skt. váta- (RV) < *vaata-, cannot then be taken as a direct correspondence of Gmc.
windaz. Bammesberger (op.cit., p. 136 ftn. 1) suggests váta- might be understood as a
viddhi formation to a fo-participle *va-tá- (with restored normal grade; for other ex-
planations, cf. Mayrhofer 1976:184-186).