--- In
cybalist@yahoogroups.com, "Sergejus Tarasovas"
<S.Tarasovas@...> wrote:
> While we are on that,
> what do you think of Sl. *z^irU 'generous fare' (< 'life'?) < *gWiH
> (3)-ru- (?) < *gWiH(3)- 'live' and *pirU 'feast' < *peiH-ru- (?) <
> *peiH- (or is it just *pei-, without laryngeal?) 'be nourishing'
or
> *pih3-ru- (?) < *pih3- 'drink'? Does it seem possible they belong
to
> the a.p. c and the u-declension as well? If so, what mysterious is
> about that *-ru- that it causes mobility?
I understand mobility to be a thing that spread in the u-stems. In
Lithuanian all u-stems eventually became mobile, but Dauks^a still
has immobile u-stems. The few cases of type b u-stems are probably
remains of words with initial accent, while some of the type c
examples also have this background (thus medU c : Ved. mádhu, Gk.
méthu). Thus Slavic type c is not diagnostic for the IE accent type
of u-stems. The complete lack of type a in Baltic and Slavic alike
is surprising, but must reflect a common slide in the direction of
mobility.
The failure of Hirt's law to produce initial accent was explained by
Illic^-Svityc^ on the basis of end-stressed cases with a middle -u-
between the root and the ending, as su:numìs, where the root could
not take the accent because it was not on the following syllable. I
find that explanation fully satisfactory.
I would accept *gWíH3-ru- or *gWéyH3-ru- as the substantival variant
of *gwiH3-ró- seen in Skt. ji:rá- 'lively' (and perhaps Latin
vire:re 'be green', viridus 'green' if vowel shortening can be
accepted) any time. For pirU I don't know, because we have no
cognate forms; still, *péyH3-ru would work fine, and if somebody
insists on *piH3-ru I have little to object.
Jens