Re: *g’(h)- > d as aberrant outcome

--- In cybalist@yahoogroups.com, "Abdullah Konushevci"
<a_konushevci@...> wrote:

What's the non-aberrant outcome supposed to be? dh?

> 1. *g'heim > dimen/dimër `winter', but prefixed form <mër-dhij>

`to

> get cold' < *mër-g'him-yo (cf. mër-dhez `to blush, flush' vs.
> ndez `to light, to kindle' < *n-dogW-ényo, besides *degW- > djeg

`to

> burn')

/rd/ > /rD/ is a regular change (Cybalist 29877), so -rdh- doesn't
actually demonstrate much.

> 2. *g'he:s-ro > dorë `hand', but prefixed form n-g'hes > ngjesh

`to

> press, compress' (cf. Romanian inghesiu `id.')

*g'h > <gj> seems to be an aberrant outcome. Can we eliminate the
possibility of a derivation from *g^el 'make a ball of' (Pokorny
555)? There are many deriviatives in *g^l-, though the closest I
could find was *g^leg^. An -s- extension *g^les is mentioned (as
*gles) in Pokorny (under D. g(e)l-eu-), but the only evidence comes
from Old Norse. -gle- would have simplified to -<ghe>- in Romanian.

> 6. *g'heh1- > g'he:-`to release, let go'. Prefixed forms sh-

g'he:ényo

> > shkonj `go', trashëgoj `inherit'; mër-gonj `to migrate' (cf.

also

> mër-thej `to send off', besides <the qafën!> `move off')

Again, this seems to me to be aberrant. Are we sure these don't
come from *gWah2, *gWem 'come, go', as in Albanian <nga> 'move'
(TBC) or <pregjim>, which Pokorny translates to German as 'Gastmahl
bei der Erstgeburt'. <nga> is derived through an intermediate stage
*ganio:.

Richard.