n 2008-02-05 09:12, fournet.arnaud wrote:
> The major problem is about all roots have or have not #s-
> it's not just a couple of odd words.
> it's widespread.
No, having or not having "s mobile" is an idiosyncratic property of
individual roots. Lots of roots never show it, and there are also many
roots which have *st-, *sn-, *sw-, *sr- etc., but no forms without *s.
In the Wiki article there are some erroneous claims like *sneigWH- being
a root with "s mobile", which it isn't. The *s- is missing only in those
languages which regularly lose it before nasals (like Latin and Greek).
Piotr
============
Not so sure.
What about Mordvin low "snow" < *n_(gh)w ? (vowel unclear)
No s- because it is Uralic..
And connection with Chinese *ning "to freeze, to congeal" ?
These words have no #s-.
What does "idiosyncratic" means when 50 % exhibit #s- ?
You know PIE had no masculine grammatical gender.
It's just an idiosyncratic feature displayed by less than 50%
of words...
Arnaud
================