On 2008-04-15 12:54, fournet.arnaud wrote:
> It was *kuH1on?
The comparative evidence points to *k^won-/*k^un-V-/*k^wn.-C-; the rest
is guesswork. In your case, it's guesswork plus violations of Ockham's
principle by introducing (entirely ad hoc) "diacritic" segments whose
only function is to account for the satem reflexes of *k^. Why don't you
go the whole hog and explain _every_ *k^ as *kh1? Shall we have *dekh1m.
for 'ten' as of today? All you need is plenty of otherwise unmotivated
*h1's in places where no same linguist has put them before.
> Hence zero grade kH1wn > k^wn.
> It's a fourth example of kH1w > k^w.
>
> Thank you, Piotr, for this apt remark.
Since when is the zero grade of *CRHVC realised as *CHRC? If anything,
we have examples of metathesis working in the opposite direction, as the
structure *CRHC is less marked and easier to syllabify.
Piotr