Re: [tied] *pYerkW+

On 2007-05-19 07:24, Sean Whalen wrote:

> I don't believe the -wos/ous distinction came from a
> dif. in light vs. heavy in PIE. The form of the gen.
> was unpredictable from looking at the stem; it came
> from older nom. in either -us or -eus ( > -us).
>
> medhu+s . medhu+'s . sunYeu+s . sunYeu+'s
> medhu+s . medhu+'s . sunYeu+s . sunYéu+s
> medhu+s . medhu+ós . sunYeu+s . sunYóu+s
> medhu+s . medhw+ós . sunYeu+s . sunYóu+s
> medhu+s . medhw+ós . sunYu+s .. sunYóu+s

The orthodox reconstruction of the 'son' word is *suh1nu-s, gen.
*suh1neu-s, just as it should be according to Szemerényi. The same rule
works for i-stems.

> If there had been a PIE rule as such than there
> would be no change in 'river' or 'lynx' as:
>
> daxnus . daxnwos . luugYsnus . luugYsnwos
> daxnus . daxwwos . luugYsnus . luugYswwos
> daxnus . daxuwos . luugYsnus . luugYsuwos
>
> so the forms with and without n wouldn't exist (as
> I've written about before).

I don't think these reconstructions are sufficiently supported by the
evidence, so I'm not convinced of their validity in the first place.

> Also, what timing for the rules would you prefer?
> What would *pl,txus give: *pl,txous or *pl,tx,wos?

Why should this be a problem? Given that the fundamental form of the
root is *pleth2-, with two consonants following the vowel, I'd predict
*pl.th2-éu-s.

> Why would *w be in an r/n-stem when the root you
> suggest is *kYerx+? Shouldn't that be *kYerxr, and so
> be unable to >> keraunós even if it existed?

It's the root of Ved. s'rna:ti 'kill, crush', possibly underlying also
the 'horn' word (an animal's "weapon"). Whatever its origin,
heteroclitic *-wr./*-w(e)n- functioned as a unitary suffix, as e.g. in
*h2arh3-wr./*h2r.h3-wén- > OIr. arbor 'corn', Gk. eidar, -atos 'food' (<
*h1ed-wr./n-) and many others, especially in Hittite.

Piotr