--- In cybalist@..., João Simões Lopes Filho <jodan99@...> wrote:
> Perhaps the roots was not *mreg^H (with g^H), but *mreg^+H, *mreg^x-
or *mreg^&). So, we can explain the alternance *mreg- / *mregH-
It can be explained without any laryngeals, being an inner Greek
affair. Grk. /k/, /kH/ and /g/ tended to fall together as [N] (spelt
<g>) before suffixes or inflections beginning with /m/. Cf. deigma,
dedeigmai (from /deik-/), dergma, dergmos (/derk-/), tetaragmenos
(/tarakH-/), etc. (the process was grammaticalised in perf. & pperf.
mid. alternations: peplekatai vs. peplegmai, etc.).
Piotr