On 2007-07-05 01:14, stlatos wrote:
> --- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:
>> Apart from the fact that the PIE root had *k rather than *k^
>
> I'm sure it's from kY. The palatalized velars could split into i+k
> or k+s. or combinations in IE branches as I've said before. Among
> others, in Indo-Iranian kY>k between u and front V (so rus'ant- but
> rocis-). This is part of early changes after u varying among the
> branches.
<rús'ant-> is too isolated even within Indo-Aryan to suggest anything
but an aberrant deocclusion of *c. We have <roká-> 'light, brigtness',
where the thematic vowel was certainly *o, and analogical <rocá->
'shining' influenced by the verb, but no **ros'á- anywhere. There's also
rúkmant- 'radiant' and rukmá- 'gold', also against your rule. I can't
see any other evidence supporting the alternation you propose. Arm. loys
shows the Armenian merger of plain velars with palatals after *u, as
also in <dustr> 'daughter' and <luc> 'yoke'.
> Standard reconstructions showing *s added in PIE don't make sense;
> why after specific sounds in different languages?, why KY not K?
> Saying PIE had *leuk+s+mn, doesn't explain why *s would be inserted in
> this word and so many others or why it doesn't appear in Goth lauhmuni.
Germanic has both *léuk-mon- (OE le:oma, OIc. ljómi) and forms with an
*-s- extension: OIc. ljós, OE li:xan, and MHG liehsen (cf. Lat. lu:na,
MIr. luan, Av. raoxs^na, OCS luna, OPr. lauxnos). The "root enlargement"
*-s- is frequent after velars and sometimes seems to produce puzzling
structural changes (*//h2ewg-// vs. *//h2weks-// 'grow', *//h2elg-// vs.
*//h2leks-// 'defend'), suggesting its great age and low productivity in
PIE. But I'm not sure if we have anything as odd as that in the *leuk-
word-family. The most widespread s-form, *le/ouk-snah2, contains the
complex suffix *-snah2 found also in other environments and forming mass
nouns, and some of the s-forms may be derived from (or contaminated
with) the widespread es-neuter *leuk-es-. My suspicion is that forms
like Lat. lu:men are analogical and that the original stem was simply
*léuk-mn., anim. léuk-mo:n (as in Gmc.). If an unexpected *s is
explained as an emanation of *k^, what is it doing e.g. in Ved. áks.i-,
aks.n.áh. 'eye' after an etymological *kW?
Piotr