> It should go without saying that the laryngeals were lost in _post_-
> PIE times.
Thanks a lot, Piotr for all these clarifications that help me a
lot 'in this area' where I'm not at all 'confortable'.
I understand that are many unknowns here but I think that 'doesn't
matter' if the resulting model will be the correct one or not at
least to can have one based on which to can derived without
contradictions all the known cases (words)...
Please to post your opinion on the following points (doesn't
matter if your point of view "it is or not 'for sure' the good one")
1. Do you 'prefer' the 'global' lost of laryngeals in initial
position or you 'prefer' their preservation before the consonantic
cluster?
2. Maybe with the exception of the vocalisation of r.,n.,m.,l., do
you 'prefer' the lost 'without trace' of any laryngeal between 2
Cons. or not? (what we will do with *bh2-lo > ballë in this case and
there are other (maybe more clear) examples pro and contra to this
option.
Thanks again,
Marius