> The "counterexamples" mentioned by Collinge aren't counterexamples.How is */o/ in a closed syllable there?
> bharama:na- < *bheromh1no-, with */o/ in a closed syllable.
> apas-, anas- have *h3e-.So the "laryngeal coloring" postdated the lengthening of */o/ in open
> The oblique pada: < *pedeh1 has e-grade, as expected.Could *pótyo:n have come from *poti-on- (with the meaning of "master
> In the cases of avi- and pati-, we have closed syllables in the
> oblique (*h2owy-, *poty-), and in the case of pati-, even in the
> original nominative *pótyo:n, as preserved in Tocharian (petso).