[tied] Re: Risoe fo the Feminine (was: -osyo 3)

--- In cybalist@yahoogroups.com, enlil@... wrote:

> Look folks, the irony is that the declension of *to- falls right
> into my hands. If you take a look at the situation, we have a
> few issues like the *tesyo genitive and the other cases that seem
> to be formed on *sm-o-. What's going on? I think it's a lot simpler
> than it first looks and as such, the feminine could have derived
> largely from the masculine endings in no time flat.
>
> The genitive in *tesyexs would ironically show clearly that the
> feminine is NOT ancient because if it were and if deriving it from
> **tesyoz were in any way correct, we'd have to expect **tesyex.
> Lo and behold, another failure. Not only does **z fail to pop up
> but so does **tesyex. So dammit, Jens, stop kicking a dead horse.

On this list we should not be strangers to the idea that languages
can change. I have not combined *tesyo and *tesyaH2s, but perhaps I
should have. I have taken *tesyo to represent the old genitive of
the thematic stem *te-s followed by a relative pronoun that has lost
its inflection. The corresponding feminine should be **taH2s-yo, so
I account for *tesyaH2s in a different way. I have assumed it to be
formed like *tesmo:y, i.e. on a feminine stem corresponding to masc.
*te-sm-e/o-, and I have seen no better solution than to accept *te-s-
yaH2-s as precisely that. That would entail two surprises at the
same time, which may be a little much: The stem of 'one' would be
athematic in the feminine, and the -m- would have been lost in the
feminine. The first surprise is as with devá- : deví:-, so perhaps
there is only one problem, i.e. no more than there will be if you do
it any other way.

If *tesyaH2s is a renewed form, it may have replaced anything -
there is no guarantee that its predecessor was a gender-indifferent
form, that is only a wish entertained by some. The wish could be
right or wrong.

> What we see is what we expect to see if *tesyo is just the way
> it is. Afterall, with masculine *tesyo which we can date

specifically

> to mLIE according to my method, we may trivially derive a feminine
> with *-ex in a stage later than that. (As I said, we have about
> 500 years to do it. Plenty of time.)

There is no valid dating method applying to this. I wouldn't know
what sign there could possibly be that the preform of *tesyo is
itself an innovation stemming from a specifiable period. There is no
trivial way of applying analogy on *tesyo and get a gen.fem.
*tesyaH2s out of that manoeuvre.

> Now if *tesyo- + -ex becomes
> *tesyex, in one easy step, do you honestly think that this is
> impossible. I don't.

If there was such a step, it can hardly be supported by anything for
it does not make much sense by itself. And it does not have to be
impossible to be wrong; however, it's close.

> But wait, it gets even simpler. The form *tesmoi/*tosmoi can be
> confidently derived from *to- + *sm-o-. You're right. However,
> the feminine can yet again be transparently derived from the
> masculine in 1.5 easy steps: *to-smoi => *to-smyexei > *to-syexei.
>
> So what's so hard? What process here takes millenia to develop?
> Nothing. The entire feminine paradigm can be derived from the
> masculine in a mere generation.

This could be true.

> Further, the issue with why we have *to-s(m)yexei instead of
> expected **to-smexei is an issue that you have to raise with
> the stem *sm-o-, not with *to- itself. It's clear that without
> the thematic, *sm- is zero-grade and a defective stem to decline.
> The thematic is a necessary component for an otherwise

undeclinable.

No, athematic does not mean indeclinable.

> So, we find athematic *-ix and not *-ex attached to athematic *sm-
> because all *sm- needs is another syllable to make it declinable.
> We don't need to go from *sm- to *sm-o- to *sm-e-x, when we can
> just go from the base *sm- to *sm-ix-. Thus we find *sm-o-/*sm-ix-.

Yes, but that is not otherwise the system of the language. However,
it may be just that in some very archaic formations. The question is
how old they are. Nothing in the posting sheds any light on the
matter.

Jens