Re: [tied] Re: Eggs from birds and swift horses (was: the palatal s

23-02-2004 15:16, elmeras2000 wrote:

> You are changing the subject: *plH1- is the zero-grade of both
> *pleH1- and *pelH1-, and in both cases it yields Greek /ple:-/; so
> when you find /ple:-/ in Greek you cannot know whether it is from
> *pleH1- as you hope, or from *plH1- as you hope not. Therefore, the
> example is invalid in this context.

One thing that is remarkable, however, is that the full grade *pelh1- is
not attested in Greek at all. *pleh1-, while indistinguishable from
*pl.h1- in most positions, can at least be suspected in those cases when
the full grade is not only expected but securely attested outside Greek
(as in the comparative, which reflects *pléh1-jo:s even in
Indo-Iranian). I find it hard to believe that *pelh1- was eliminated by
analogical substitution in every single instance.

> That will be from *renewed* *gWoyH3-o-s. These must be two layers in
> the treatment of the same morphological class. The other examples I
> have found of the old type are *g^nH3w-o-s (Lat. gna:vus 'well-
> behaved', semantically like French 'sage') and *prH3w-o-s 'first'.
> All three are from roots of the structure *TReH3-. I suppose there
> was some phonetic interference between the infixal consonant and the
> surrounding sounds which changed the sequence *-ROH3- into *-RH3w-
> (or what developed like it). I wouldn't know what a suffix *-wo-
> would be doing here.

Nor would I, and I agree that the *w is most likely a product of
phonetic development. The "phonetic interference" looks very much like
metathesis: *-ROh3o- > *-Rh3Oo- > *-R.h3wo-.

Piotr