Re: PIE i- and u-stems again

> The feminine forms are structured differently. While in the
> masculine forms everything is as if we had a thematic root
> (*te/o-, *ye/o-, *kWe/o-, etc.) followed by the inflected
> forms of *is, the feminine forms seem to be built on *te/o-
> etc. plus the feminine marker *ih2/*jeh2 (not the anaphoric
> pronoun *ih2/*ejeh2). Slavic has preserved the original
> paradigm best:
>
> N *sé-(i)h2 > *sáh2 (OCS ta)
> A *té-(i)h2-m > *táh2m (OCS toN)
> G *te-jéh2-os > *tojãs (OCS tojeN)

Why the nasal vowel *-N?

> DL *te-jéh2-e(i) > *tojãi (OCS toji)
> I *te-jeh2-ét > *tojh2á:(m) (OCS tojoN)

And ditto?

>
> N *té-(i)h2-es > *tãs (Lith. tõs)
> A *té-(i)h2-m-s > *tá:ns (OCS ty)
> G *te-jeh2-s-õm > *tojh2sõm (OCS têxU)
> D *te-jeh2-mós > *tojh2mós (OCS têmU)
> L *te-jeh2-sú > *tojh2sú (OCS têxU)
> I *te-jeh2-mí(:)s > tojh2mí(:)s (OCS têmi)
>
> Vedic (and Germanic) have innovated by adding inflected
> forms of *smih2 "one"(f.) to the oblique sg. forms instead:
>
> N *sé-(i)h2 > *sáh2 (Ved. sá:)
> A *té-(i)h2-m > *táh2m (Ved. tá:m)
> G *te-sm-jéh2-os > *tes(m)jãs (Ved. tásya:s)
> D *te-sm-jéh2-e(i) > *tes(m)jãi (Ved. tásyai)
> L *te-sm-jéh2-e > *tes(m)jã(m) (Ved. tásya:m)
> I *te-jeh2-ét > *tojh2á:(m) (Ved. táya:)

I imagine there must have been a time where the *-ax ending was just
another case ending of the sg (cf the Mordvin indefinite pl). All
reconstructions with fAsg *-xm are not convincing phonetically. This
means there can only have been a full a-stem paradigm, and hence a f.
gender, after the laryngeals were vocalized or whatever, when *-ax ->
*-a:, and a paradigm could be constructed on the newly appearing vowel
stem in analogy with other vowel stems. That's why there's no f. in
Hittite.


Torsten