----- Original Message -----
From: Piotr Gasiorowski
On 2008-03-18 17:46, Brian M. Scott wrote:
> *welh2- (Watkins) or *huelhn- (Beekes) 'wool', with reflexes
> in Sanskrit, Avestan, Latin, Greek, Balto-Slavic, Celtic,
_And_ Anatolian (Hitt. hulana-, CLuw. *hulani- [SÍG-la-ni-])! The
optimal reconstruction is certainly *h2wl.h1-no/ah2-. George Hinge's
recent proposal (a prothetic vowel in Greek fails to develop before a
closed syllable containing a syllabic consonant) accounts for Gk. le:nos
if one assumes the syllabification *h2wl.h1,no-.
TheTibetan word bal "wool" is intriguing.
It looks closer to the original *b_l that
yielded *-wel- than PIE
They cannot be cognates either
Hulana supposes a velar spirant
which could have been Tibetan hbal
And I won't readily accept as a PIE
cognate a late neolithic word.