On 04-09-28 14:10, tgpedersen wrote:
> Since I wanted the thematic stems to be a generalisation of the
> cases with stressed /o/ of the athematic stems (gen.sg. -ós,
> gen.pl. -óm), therefore of the form -ó-[athematic case suffix]
> everywhere, I wasn't too happy with this solution. Until I
> remembered Latin *-os-om; in other words, we would have at some time
> the two forms *-oy-m and *-os-om. Is this the athematic gen.pl.
> suffix glued on to a nom.pl. ('pronominal') and some late stage of
> an acc.pl.?
I wonder if we really should posit *-oj-m to account for thematic
*-o:N(s). If, as is the case with the dual, the plural marker originally
preceded case endings (before the agglutination of *-s accross the
board), let's imagine that the thematic plural looked like this (D
stands for some kind of voiced coronal continuant):
Nom.pl. -o-D-z
Acc.pl. -o-D-m
Then, -oD > -oi, -oDm > -o:m (final lenition, compensatory lengthening)
The first change accounts also for *to-D > *toi, *(m)weD > *wei- (1sg.).
In the nom.pl., *-oDz > *-o:D at a later date, when word-final *D was no
longer lenited but instead changed into whatever underlies PIE *s/*t, so
*-o:D > *-o:c > *-o:s . Note the brand-new analysis of *-o:s, which is
specifically NOT *-o-es. In case anyone wonders, I have a parallel
scenario ready for athematic plurals and for duals.
Any thoughts?
Piotr