Pavel A. da Mek wrote:
> So it seems to me that the loss of the laryngeal and the 1st
> vocalisation had to work simultaneously - if the clusters was too
> complex, then it was simplified by deleting of the laryngeal AND by the
> vocalisation of the consonantal /O/, although either of these two
> changes would alone be enough to get an acceptable cluster.
>
>
> *O-perH2-men-é-H2; then the metathesis of O to the end of the C cluster:
> *pOerH2-men-é-H2; then the loss of unaccented vowels:
> *pOrH2-mn-é-H2; then mn > n after a vowelless sequence containing a
> labial:
> *pOrH2-n-é-H2; and the simplification of the cluster:
> *por-n-é-H2; then the stress shift:
> OPIE. *pór-n-e-H2; then the coloring:
> PIE. *pór-n-a:H2
> Greek pórne:
>
> Did I get it correctly?
Jens's original argument was that the *O was syllabified first in heavy
roots (and so was able to attract the accent):
*pÓ.rh2-neh2 vs. *kOr-mós,
but that indeed raises the question why the resulting vowel (or syllabic
liquid, whatever the phonetic interpretation of *[O.]) did not prevent
the loss of *h2. It's also strange that forms like *bHOh2-(m)néh2 did
not lose their laryngeal in what was evidently a heavy cluster. I wonder
if the accent shift is real. Consider the pair *h2wl.h1-náh2 'wool' vs.
*wól-no-s 'curly'. To my simple mind, the former looks like the
collective of the latter. It seems as if the loss of laryngeals caused
by the O-fix (or the alternative -- the loss of *O) in overweight
clusters had been younger than the rise of constrastive accent in such
pairs! I think it's possible that the collective/mass noun never had the
*O- prefix in the first place. Now, pór-nah2 is not a collective or a
mas noun but simply the feminine form of *pór-no-, so it has the same
accent pattern. But in pairs where both forms were common (e.g.
*k^ól-mo- 'stalk' vs. *k^l.h2-máh2 'straw' levelling could be expected,
both in their vocalism and the accentual pattern).
Piotr