Re: [tied] Methodology

On Sat, 20 Oct 2001 14:19:47 +0200, Miguel Carrasquer Vidal
<mcv@...> wrote:

>About words ending in vocalic /-r./ in Sanskrit: there are none in the
>R.gveda, but -/r./ occurs as a final vowel in the Bra:hman.as (in
>particular for neuters in -tr.).
>
>The situation in Vedic is as follows:
>
> PIE Vedic
>
>1. *-r. -ar
>2. *-r.t -r.t
>3. *-r.s -úr
>4. *-r.gW -r.k
>
>1. *Heg^hr. > áhar "day"
> *h1ouHdhr. > ú:dhar "udder"
>
>2. *yekwr.t > yákr.t "liver"
> *k^ekwr.t > s'ákr.t "excrement"
>
>3. G. *ph2tr.s > pitúr "father's"
> 3pl.pret. *-r. + s > -úr
>
>4. *h1esh2r.gW > ásr.k "blood"

Let me further add that in the other IE languages (but Gathic goes
like Vedic, Armenian seems to have preserved *-rt) the *-r/-n-,
*-rt/*-nt- stems and the single *-rgW/*-ngW- stem merged, usually in
favour of the *-r/*-n- paradigm, in Greek in favour of the *-rt/-nt-
paradigm. Vedic *-r.s is secondary, so the "heteroclitic" soundlaw
can be fully formulated as: /n/ before # or T# (where T is any stop)
-> /r/, except if /m/ or /mV/ preceeds.