On 05-01-05 16:29, elmeras2000 wrote:
> That is what we like to believe at this place, and Birgit is pleased
> to be remembered. It would be like Latin albus => albe:re =>
> albidus, only without the middle part: Given the adjective surviving
> in Lith. núogas, OCS nagU, it would be *nogWe-h1-to- 'become naked'
> > *nogWetho- > Italic *nogWeTo-s > pre-Latin *noweDos > *nowedos >
> *noudos > nu:dus.
Just to clarify: do you mean that all "-idus" words are
quasi-participles from *-eh1 statives? In other words, do they go back
to *-etHo- < *-eh1-to-, while there are also *-et-o- adjectives related
more directly to *-es-nouns (the case of *nemes- : *nemeto-, with the
*s/*t alternation having the same historical status as in
*meh1-not-/-nes-)? But if so, why does Latin show no *-eto- words
(correct me if I'm wrong), and why isn't the *-eh1-to- formation
represented in Greek and Indo-Iranian, where we would expect a distinct
reflex of the aspirate? As far as I can see, they have plain *-eto- (and
Gk. -etiko- as an extended variant) corresponding to Lat. -idus. It's
because of this complementary distribution that I have attempted to
combine these formations into a single type.
In other words, I agree that the 'naked' words can be analysed as
related to *nogWo- via the "virtual verb" *nogWe-h1- 'to be naked', but
what I have been contemplating is a more direct relationship,
*nogWe-T-os, where *-T- is an allomorph of stative *-h1-; similarly in
tep-id-us : *tep-es- 'heat' etc. Connected with that, I would predict
the permissibility already in PIE of the type *nogWe-s- 'nakedness' (=
the state of being naked), i.e. abstract s-nouns derived directly from
adjectives, which is what Latin has in <albor> 'whiteness' etc. (beside
the fancier Graeco-Latin type in -e:d-o:(n), quite convincingly
explained by Birgit Olsen).
Piotr