Re: [tied] IE genitive

On Wed, 16 Apr 2003, Glen Gordon wrote:

>
> >The genitive of aH2-stems is not unproblematic. One expects the
> >gen.sg. to end in *-aH2- + the zero-grade of the gen. morpheme /-os/, i.e.
> >*-s. But Greek and Lithuanian agree on a circumflex
> >long vowel, [...]
>
> By saying "one", of whom does one speak? Is this an indirect way
> of saying "_I_ expect"? I don't see why one should expect this at
> all. From what I understand so far, the feminine *-ax is a very
> late innovation ultimately stemming from the neuter collective
> in *-x.

One speaks of writer, but writer endeavors to be objective. It may cause
you a fatal heart attack, but after much considereation I will have to
agree with you on this point. You may even have spoken the magic word we
have been waiting for. The problem was: Why does the gen.sg. in *-aH2-os
which belongs to the class of "a:"-stems whose stem-forming part consists
of the thematic vowel *-e/o- + the collective marker *-H2 not reduce the
gen. desinence to the zero-grade form /-s/, given the rule that everything
following the thematic vowel appears in its reduced form? Well, perhaps,
as you said, because the form is younger than that rule.

>
> It started in the Late IE period, where the inanimate collective
> *-x was extended to *-&x. Other suffixes have been "thematized"
> in this manner in this period to convert an inanimate suffix to
> an animate one.

What are you talking about? Are you hinting at a rule opposing animate
and inanimate by vowel insertion in the suffixes of the latter to form the
former? I know of no opposition between inanim. -n- vs. anim. -Vn-, or
inanim. -s- vs. anim. -Vs-, or inanim. -w- vs. anim. -Vw-, etc. - do you?

>
> Thus was born the "human collective" suffix marking nouns which
> naturally described a group of people. Abstract words like
> "fatherhood" or "farmer" would employ such a suffix. Evidently,
> because of the vocalism, it was created before the beginnings
> of schwa lengthening before voiced segments which would later
> produce thematic qualitative ablaut (eg: *bHer-e-ti/*bHer-o-mes).

But here the chronological point I would accept is being rejected. I see
no way of reconciliation on this basis.

>
> Now, understanding this, the genitive endings as they were when
> this suffix was created would have been accented athematic *-as,
> unaccented thematic *-&-sy&, and finally the assyllabic *-s for
> stems ending in *i or *u.
>
> The expected genitive form is indeed as we find, *-ax-os (which
> in eLIE would be *-&x-as, accented on the last syllable).
> Afterall, such "feminine" stems did not end in a semivowel. They
> also did not terminate in a thematic vowel. They belonged with
> the athematic animate nouns such as *po:ts "foot" with gen.
> *pedos (eLIE *pa:ts/*pedas).
>
> I hope that all makes sense since I've been up for more than a day
> now.

It looks familiar, that's the extent of the acceptance I will give it.

Jens