Re: Kluge's Law in Italic? (was: Volcae and Volsci)

W dniu 2012-02-10 23:05, dgkilday57 pisze:

> It looks difficult, but a native PIE-speaker would have had a different
> perspective. The 1966 Guinness Book of World Records claimed that Polish
> <chrzaszcz> was the world's most unpronounceable word. (I doubt the
> McWhorter brothers knew much Polish.)

I don't think they did. And they surely didn't know any Georgian.

> As a *dHlo-denier, does Olsen explain Greek <ge'nethlon> by
> preaspiration of *t by *h1? This would seem to open a real can of worms,
> with <stato's> against Skt. <sthita'->, <platu's> against <pr.thu'->,
> and other examples of non-correspondence of Grk and Skt. unvoiced aspirates.

There is a small can of worms here, though not really in these cases.
Olsen claims that only non-syllabic laryngeals caused the aspiration, so
*st&2-tó- presents no problem, the Skt. *-tHi- < *-th&- < *-t&2- being
branch-specific, like for example the aspiration in <duhitar->. Likewise
in *pl.th2u-, the PIE *h2 followed the stop (though in this particular
word the *-th2- sequence is possibly the outcome of a very ancient
metathesis in the protolanguage itself, maybe older than Olsen's rule).

> And if *h2 preaspirates *t, and she only believes in 3 laryngeals,
> 'father' is a real embarrassment.

Again, it isn't, because the laryngeal in 'father' is syllabic. What
_is_ problematic is the absence of aspiration in 'mother' and 'brother'
which has to be explained either by recourse to analogy ('father',
'daughter') of by refining the condition of the rule (perhaps
sensitivity to the placement of accent). Family terms are tricky. After
all, we have a similar conundrum in Modern English, where <father> and
<mother> (unlike <brother>) are phonologically aberrant (Chaucer still
had <fader> and <mo:der> but <bro:ther> 600+ years ago).

Piotr