On 2008-01-24 03:41, stlatos wrote:
> I do not believe any nouns in -mo- or -no- are derived from *-mn, or
> anything similar. This argument of mine is not about -no- of PIE
> origin, but of a sound change of m>n. after velar stops in Latin (and
> sim. in other IE branches).
That's of course possible, with the obvious exception of K-men wods
(tegmen etc.), where presumably the dissimilatory effect of /n/ blocks
the change. The m/n alternation is what Rasmussen observes in <tormos>
vs. <porne:> (there are also other related phenomena not mentioned in
his postings here but discussed in his publications). He doesn't take
into account other place-of-articulation constraints, but perhaps he
should. I would claim myself (still assuming the *-mn- analysis as a
working hypothesis, pace your scepticism) that after root-final *w (*u)
it's the -m- that regularly remains, contrary to Jens's predictions, see
*s(j)ou[h1]mos, *sroumos, and there is some hard-to-explain dialectal
variation e.g. in the 'foam' word: Lat. spu:ma, Gmc. *faima- vs. Lith.
spáine., Slav. *pe^na, Skt. pHéna-.
Piotr