> EIEC says so too. But Bomhard thinks *xehs is Nostratic (cognates in
> AfroAsiatic and Finno-Ugric.
Hardly possible. At best you could relate the unextended form *xeh-,
but the extension *-s- should be unique to IE. Considering that the
verb root has nothing but laryngeals in it, it wouldn't be easy at
all to relate it securely to a Uralic form, which would presumably
end up **i- or **a-. That's not much to go on.