Re: Slavic endings

> >> and Gothic gen.sg. -is < *-es(j)o
> >
> > As I suggested, that might be from <noun> es-jo "<noun> its",
> > cf the same construction in Norwegian, Dutch and Jysk, which
> > would explain the e-grade.
>
> But what would this *es(j)o have been attached to?
> A form of the noun stripped of all inflections
> (not yet lost in early Germanic!) and even
> of the thematic vowel?

But that happened in early PIE compounds.

>One could imagine
> haplological reduction like
> *wulfas(a)-esa > *wulfesa,

Not with the "stem" in the genitive, dative would be
more likely (dial. German "dem Peter sein Hut").

> but is it really more economic than the
> direct change of *wulfas --> wulfis in Gothic
> due to the influence of pronominal <is>, <þis>?

The economy is in the reduced number of assumptions
of odd phenomena, which I think 'borrowing from
the pronouns' is. With my proposal, simple composition
suffices.

>In the rest of Germanic we find a reflex of
> *-os(j)o, which reflects the normal IE gen.sg.
> ending of thematic nouns. This ending itself
> is probably of pronominal origin (with the o-timbre
> of the thematic vowel restored in nouns already
> in the protolanguage), but that's a much older story.

Put another way, the composition is pre-ablaut.


Torsten