Re: [tied] *pYerkW+

On 2007-05-17 21:21, stlatos wrote:

> PIE had different u- and eu-stems depending on the original ending.
> Most branches have analogy obscuring this but this change is early
> enough to avoid that.

Depending on the structure of the root we have gen. sg. *-w-ós (after
light stems such as *medHu-) or *-éu-s (after heavy stems such as
*pértu-, see Szemerényi). Lat. quercus represents the latter type.

> Definitely not. Some languages have kW>0 between r_u and
> *pYerkWuunos > *kWerkWuunos > *kWeraunos brd> Greek keraunos. In
> whatever language Greek borrowed it from the order of rules shows no
> kWu>ku.

I don't happen to believe in such a change, _especially_ in Greek, where
the delabialisation of labiovelars before in the vicinity of /u, w/ is
well evidenced. And why is there no *teraunos with *te- < *kWe-?

> But *pYerkWn.os > quernus has neither u nor eu after kW;

So what? If your analysis is correct and there's no delabialisation, the
change *p > kW is regular. And of course the adjective may have been
influenced by the noun.

> there's no
> evidence that the exact stem from the root *pYerkW+ that Querquetani
> comes from had eu/u so it doesn't prove analogy of ku / kWeu > kWu /
> kWeu or something similar.

It may have been *perkW-eto- or *perk[W]w-eto. In either case the
Italo-Celtic assimilation of the initial *p is expected.

Piotr