Re: [Hebrew] dagesh qal - special case question

For whatever it's worth, here is how I remember the mess of forms of words
like "davar" or segolates like "melekh":

"davar" loses both Qamats as soon as they get away from their position
immediately before the stressed syllable: davar->dvaro/dvarim->divreykhem.
The only exception is "divrey" which has no Qamats left although the
stressed syllable follows immediately. It looks as if the "-ey" of
"divrey" is not regarded as stressed but rather the whole smikhut
construct has only one stress at the end of both words (in the Hebrew
Bible, such a construct is often written with a Maqqef and only one taam
hamiqra).

Now with these prerequisites, the presence or absence of spirantisation in
those forms of segolates that have the double consonant (like
melekh->malk-/malkh-) appears to be fairly regular: spirantisation occurs
if and only if the stressed syllable does not immediately follow.

I do not know whether there is a real phonological background to these
observations or whether the connexion is mere coincidence.

Helmut Richter
.