>So, now that I have convinced myself that PIE *dheh1- and *doh3- are
>really the same verb
There is a consequence to what you're saying:
Since Latin Greek and Sanskrit consistently distinguish these roots, I
suppose you are arguing:
(a) The root split into two roots at some stage.
(b) This must be before the split-up of Latin-Greek-Armenian-Sanskrit
(unless each of these families independently split the root, which seems
less likely).
(c) Germanic kept the original situation of a single root (well, that is
what you're arguing, isn't it?)
(d) Therefore Germanic split off before Latin-Greek-Armenian-Sanskrit
separated.
Isn't it rather simpler, and better, to acknowlege that Latin, Greek,
Armenian and Sanskrit just happen to be the languages (other than Germanic)
that clearly distinguish PIE *dh and PIE *d ?
Peter