On 2008-11-02 12:50, tgpedersen wrote:
> But I said 'thematicized', not 'thematicized at some time after PIE'.
The development I was talking about is post-Proto-Germanic. Derivatives
like *derwo-, *deiwo-, etc. occur commonly throughout the family, but
they don't behave like *u-stems at all -- they are declined like
ordinary thematic stems (so are *-wa-stems in OE. There's no *taru- in
Germanic, just *tarwa-. If we assume that *-wa(z) > *-w > -u between
PGmc. and OE, everything is simple and regular.
> > Most instances of *-jo- or *-io- > Gmc. *-ja- have nothing to do
> > with i-stems.
>
> Meaning that they are not documented without the -a- < -o- which
> elsewhere is considered as the thematic vowel. Is that proof they
> never existed? Was there a -io suffix independent of -i and -o?
Most *-io- stems are "double thematics", i.e. thematic derivatives of
already thematic base forms, like *h1ek^wo- --> *hek^wi-o- 'pertaining
to horses' or *newo- --> *newi-o- (no corresponding *i-stem in either
case). Germanic *xarja- comes from *korjos 'war party', derived from
*koro- (Lith. karas 'war', OPer. ka:ra- 'folk, army'). There's no *kori-
anywhere.
Piotr