Re: [tied] More on the crummy sanguis/asrk connection

*wa:c- was generalised in all prevocalic positions. Sanskrit tended to eliminate this kind of allomorphy, since the contrast between a < *o and a < *e was synchronically opaque. Before consonants -k (~ -g) was retained.

Piotr

----- Original Message -----

From: Miguel Carrasquer Vidal

To: cybalist@egroups.com

Sent: Friday, December 15, 2000 11:58 AM

Subject: Re: [tied] More on the crummy sanguis/asrk connection

Speaking of which, is there an accepted explanation for the palatalization of (labio)velars before case endings? We have, e.g. from va:c- (< *wokw-): N. vá:k vá:ca:(u) vá:cas A. vá:cam ,, ,,/va:cás I. va:cá: va:gbhyá:m va:gbhís D. va:cé va:gbhyás G. va:cás va:cá:m L. va:cí