*sr- roots in PIE (2)

[Dennis Philps]

Peter wrote :
>Pokorny lists (s)reigh "klettern" and ten other roots in #sr-

This is just the point I'm making : IE dictionaries such as Pokorny and
Watkins list several *sC- roots (where C= stop or sonorant) which are ofter
related to *sVC- roots with similar or identical 'meanings', e.g. *skei-
'cut'/*sek- 'cut'. But no *(s)r- roots appear in Watkins, even though he
gives *sreu- 'flow, move' along with *rei- 'flow'.

Given that in Watkin's dictionary, *sek- extensions such as *skei- are
listed separately and described as extensions of *sek-', why isn't *sreu-
given as an extension of *ser-?

Regarding *(s)r- roots, only one *(s)r- root, (*(s)reigh- 'klettern') is
listed in Pokorny, as Peter points out, but even this appears questionable,
since the Larousse Dic. of IE roots lists 'klettern' as continuing *gel-,
like Eng. 'climb' (<*gle-m-bh-). All very confusing and unsystematic for
the general linguist wanting to come to terms with PIE root structure.

Pat wrote :
>I suspect the key to this puzzle is that *s + *Herew -> *sreu-.

Now this (if I have understood it) is an interesting idea, which I have
suspected for some time for reasons I won't go into here. If we postulate a
pre-PIE root *(s)Hxer- 'flow, move' (where Hx =  undetermined laryngeal),
then would not this account economically for *ser- (Hx drop-out from the s-
prefixed form of the root), *sreu- (zero grade extension of the s- prefixed
form of the root after Hx drop-out), and *rei- (zero grade of (?)*Hxer-
after Hx drop-out)? Might it not also account for *er- 'move' (Gamkrelidze
& Ivanov) after Hx drop-out from the s- less form? I suppose much would
depend on the relative chronology. Sorry if the specifics are not perfect,
but can any evidence be adduced to support or reject the idea?

Again, many thanks for your thoughts.
Dennis.