--- In
cybalist@yahoogroups.com, "alexandru_mg3" <alexandru_mg3@...>
wrote:
>
> --- In cybalist@yahoogroups.com, "Abdullah Konushevci"
> <akonushevci@> wrote:
> >
> > --- In cybalist@yahoogroups.com, "alexandru_mg3" <alexandru_mg3@>
> > wrote:
> However we still have:
>
> Latin api:scor indicating 'directly' h2ep-(or there is also 'an
> aditional' Latin Rule PIE *h1e- > Latin a- that I'm not aware about?)
[AK] I am not aware for any rule that *H2ep- would yield Latin /a/,
niether for *H1e- > Latin /a/. As far as I know, Latin was not
affected by initial laryngeals, see: *H1es-, *H1ed-, *H1epo-, even
*H1ag'- (cf. Albanian jam: Latin sum, Albanian ha: Latin ed-yo, Alb.
hap, prapë: Latin ab etc.). Colored *H2ep- to H2ap- yields simply
Latin /a/.
Pokorny gives three basic form *ap- (richtiger &p-): e:p-. If we
accept last form, we know that it is from contracted form *eH1p- with
zero-grade form *H1p- and metathesized form should be *H1ep-.
>
> and also
>
> Albanian jap (considered with an analogical epetenthic 'j' like in
> jam) indicating also h2ep-
[AK]
Alb. <jap> 'to prodce, give, work' with variant forms <ap>, <nap> is
supposed to come from *H3ep- and has nothing to do with any *H2ep-.
>
> It's also true that an o-grade h1op- will solved the Albanian
> derivation...
[AK]
What do you mean by this assumption?
> Any, hint?
[AK]
No, till you are very much aware for all rules... Sorry that I have
replied in your post!
> Marius
>
Konushevci