On 2008-02-13 21:43, Miguel Carrasquer Vidal wrote:
> As mentioned in LIV, this is a nasal present from the root
> *sweh2d-: *suh2n.d- > suvad- > svad- (still scanned once as
> suvad- in the Rgveda). Cf. also LIV *peh2g^-, n-infix
> *ph2n.g^- > Skt. paj-.
There's also another possibility -- that we are dealing with a mixture
of two different word-families, one connected with the root
*swa:d-/*swad- 'sweet(en)', the other with the compound *h1su-h1(e)d-
(approx. 'good food'), accidentally falling together in Indo-Aryan. If
we can divorce forms like Ved. sú:da- 'cook', su:dáyati, and Lith.
súdyti 'season, salt' from the former, the LIV analysis can be abandoned
and no laryngeal is needed in *swa(:)d-.
Piotr