*bhe-, -y, -w

Schmalstieg proposes to analyse the "be" verb as *bhe/o-, plus two
extensions -y and -w. He does the same with "drink": *pe/o + -y,
+ -w.

Orël and Stol'bova has *baw-/*bay- "place; home" and a rare verb
(Chadic only) *bay- "build", the origin of the
common 'beit' "house". Møller has a *bh-u- with Semitic cognates,
eg. Arabic "take up abode".


I know next to nothing about AfroAsiatic, but I recall having seen a
y/w alternation mentioned in several roots, unlike PIE where I
haven't seen it.

Does this mean that an analysis of *bh-w- etc "be (at)" strictly
within PIE is impossible?


BTW I wonder if the pre/postpositon *bhi "at" is related somehow.
Miguel tried to link it as a preverb to Polish 'po-'. The
alternation -u/-i smacks of a locative too?


Sorry for not being very constructive here.


Torsten