On Mon, 18 Dec 2000 23:10:12 +0100, "Piotr Gasiorowski"
<
gpiotr@...> wrote:
>On the other hand, EIEC supports the analysis of sanguen- as *h1sh2an-gW-en-, though neither the *-gW- nor the double nasal extension is justified; nor is any parallel case offered. I suppose the argument would boil down to "Skt. -k in asr.k is a mysery and sanguen is a mystery, so maybe they're the same mystery".
Which "double nasal extension"? The first *-en- (in *h1sh2{en}gw-) is
part of the root, and can only loosely be referred to as the
"heteroclitic *-en-extension". The second *-en- is no more mysterious
than the *-en in <pecten> "comb" from the root *pekt-.
As far as I'm concerned, the case is clear. The -k in Skt. <asrk>
more than likely comes from PIE *gw. The heteroclitic paradigm
must've once been *h1ésh2rgw, *h1sh2ángw-, later reshaped to
*h1esh2r[gw], *h1esh2(a)n-, with loss of the *gw in the oblique cases
(and except in Vedic, in the absolutive too) by analogy with the more
common heteroclitic pattern. Except in Latin, where the oblique form
was used to create a new n-stem paradigm.
Now, the challenging bit might be explaining if and if so how (Pre- ?)
Greek *saimen- (> <haima>) fits in.
=======================
Miguel Carrasquer Vidal
mcv@...