Re: [tied] Re: *pYerkW+

On 2007-05-26 16:37, alexandru_mg3 wrote:

> Question: why you wrote Proto-Slavic *xvorstU and not *xvarstU?
> (o>a belong to Balto-Slavic times, a>o is 'recent' in Slavic:
> OCS xvrastU )

So is the development of word-final *-o-s > *-as > *-U and the change of
*w > v. The conventional PSl. reconstruction is *xvorstU < *xwarsta-, if
you want to be really precise.

> Note: Dersksen linked also OE hyrst as cognate with OCS xvrastU (see
> on Leiden)

I know, but Derksen provides no explanation of the problematic details
of that comparison.

> Question: So why 'it's relationship with Slavic *xvarstU is difficult
> to maintain' ? *kWres- could yield *kWers- , that's the whole
> point...

For one thing, *kW gives Slavic *k (or *c^ when palatalised), never *xv.

> Ok, with 'an additional labial segment' : But even so, this new rule
> is not 100% accurate, and a rule that is not accurate...is not quite
> a rule -> this is the fact that 'disturb me' (like kw>p and kw>k in
> Romanian)

I've just shown you that the exceptional development of *kWekWlo- (my
own counterexample) is neither inexplicable nor even entirely
exceptional (OFris. fial does exist). Anyway, sound changes often admit
of exceptions. Nobody questions the regularity of OE a: > ME /O:/ >
Mod.E /oU/ just because a few words (such as <one>) have taken a
different path.

> Why not?
> a) kW > p , from some Pre-Germanic influences (->Dialectal PIE)
> and
> b) kW > xw > h , as rule?

What do you mean, "as a rule"? The normal development of *kW is Gmc. *xw

> hw etc. -- not just /h/. Those hypothetical "pre-Gmc. influences"

don't explain the fact that the reflex *f instead of *xW is restricted
to certain phonological contexts.

Piotr