Íslenskt mál og almenn málfræði - 01.01.1985, Side 107

105
Rejlexes of I.-E. *suH2nto-/-ön
unless it is assumed that East Tocharian preserved reflexes of both
formations: the athematic nt-formation as found in Hitt. huuant- (cf.
discussions in Kronasser 1962—66:263—264, Darms 1978:396—397,
Tischler 1978:328 — 331 with lit., see also ftn. 4 above) and the the-
matic vrddhi derivative, combined in one paradigm. This would seem
rather unlikely.
The (optionally?) divergent syllabification (VHnC : VHnC) here sug-
gested for A wánt- : want is actually paralleled in another Tocharian
word. The verb A su-, B swá- ‘to rain’ belongs to the athematic Pres.
Class I in East Tocharian (3. pl swihc, pres.ptc. middle sumárn), to
Pres. Class V in West Tocharian (3. pl. suwam, pres.ptc. middle
swámane).
Winter (1965:193) has explained A swihc as deriving from *suwnti
from (in his notation) *swXnti, in which „the laryngeal need not
become vocalized", i.e. *suHirti (with H = H2 or H3 as indicated by B
5wá-)10, 11 > *sunti > Toch. *s(u)wánti > swihc (not I.-E. *suHenti which
would have yielded A *swáhc (if Hf) as pointed out by Winter (loc.cit.,
ftn. 13), or A *swahc (if Hf).
10 Not necessarily H, as suggested by Normier (1980:270); at least Gk. úco rain, send
rain’ does not prove anything concerning the colour of the laryngeal and the nounúexói;
‘rain’, formed like vuperoi; ‘snow-storm’ and nayeTÓ; ’frost’, can easily have an ana-
logical e-vocalism.
11 K.T. Schmidt (1982), basing himself mainly on Winter (1965), assumes the follow-
ing development of the Indo-European laryngeals in ante-consonantic position in Toch-
arian: if preceded by a consonant, all three laryngeals merged in Common Toch. *á (i.e.
CHC < CáCy, if preceded by a post-consonantic resonant or a semi-vowel, the develop-
ment was dependent on the character of the laryngeal: H, caused a lengthening of i, u,
but was dropped after a resonant (i.e. iHh uH, >, ti, but RH, > áR), whereas iH2/3, uH2/3
yielded Common Toch. iyá, uwá and RH2/3 yielded Rá.
In principle, I subscribe to this view. However, I would like to suggest that the devel-
opment was somewhat more uniform: first, H, was lost not only after a resonant, but
also after a semi-vowel, i.e. just as CRH,C yielded Toch. CáRC, so CiH,C/CuH,C
yielded CáyC/CáwC, which regularly became CiyC/CuwC >CiC/CftC with á yielding/
and u before y and w respectively. Second, just as CRH2/3 resulted in Toch. CáRáC, so
CiH2/3C, CuH2/3C resulted in Toch. CáyáC, CáwáC which thereupon changed regular-
ly into CiyáC, CuwáC. This development, as well as such apparent exceptions to
Schmidt’s theory as Toch. B. kárwehe ‘stone’ (I.-E. *gl>rH,uon-) and A párwat ‘oldest’,
B parwe ‘at first’ (usually derived from I.-E. *prH2/3uo-) will be discussed in a forthcoming
article. For A swinc, siimám (I.-E. *suH2/3-), cf. discussion in text.