On Tue, 12 Dec 2000 19:02:45 -0200, João Simões Lopes Filho
<
jodan99@...> wrote:
>May there is a n/r alternation in PIE for blood? cf. Sanskrit asrk; Latin
>sanguis (-inis); Greek ear (?) - something like *H1esr-/H1esn- ?
Yes. We have Vedic <ásr.k>, G. <asnás>; Greek <ear>, <eiar>, <e:ar>;
OLat. <as(s)er>; Latv. <asins>; Toch.A. <ysa:r>; Hitt. <eshar>, G.
<eshanas>, which can all be more or less regularly derived from a
heteroclitic r/n-stem *h1ésh2r, *h1sh2nós. Remains the -k in the
Vedic nom.sg. and Latin <sanguis>, G. <sanguinis>, which can be
explained as *h1ésh2ngw and *h1sh2ángw-en-, respectively. The PIE
form can thus be posited as *h1esh2engw- (*h1ésh2ngw- / *h1sh2éngw-).
I've just compared to this the Proto-Kartvelian form *zisxl- "blood"
(Geo. sisxl-, Megr. zisxir-, Laz dicxir-, dincxir-; Svan zisx), if
"reduplicated" from pPK **?isx@... (vs. something like pPIE
**?isxangu).
=======================
Miguel Carrasquer Vidal
mcv@...