On 2006-08-29 02:18, Sean Whalen wrote:
> g^hulnos > paz^ulnus 'slanting, sloping'
>
> So probably *g^hulos formed a new adj. in -no- but
> the ul was kept the same even in the new
> syllabification, just as I posited for sunna/*suln-.
One could equally well argue that BSl. *iR > uR in the context *Cw...n
(absorbing the glide), perhaps also after original labiovelars, as in
*gWHr.-nó- > PSl. *gUrnU 'cauldron, pot, oven'. The only trouble is that
the distribution of zero-grade *iR vs. *uR in Balto-Slavic is messy in
general and so difficult to use as proof of anything.
> Also, for *Lk>Luk (as miluks) OCS alUkati but Lith
> alkti < *uLk- (or whatever form you prefer to rec.)
> 'be hungry' < '*lie ill'.
OCS has alUk- ~ alk- ~ al'k- ~ lak- in this verb, showing different
stages of liquid metathesis (still a change in progress at that time) in
an initial *olC- sequence. We have the same hesitation in al(U)dii ~
ladii 'boat'. There was no *u, PIE or BSl., in the first syllable. Also
Lith. álkti, Latv. al^kt, derives from *álk- (acuted), not from *ulk-,
*uLk- or whatever.
Piotr