Sean Whalen wrote:
>>devá-kr.ta-),
>
>
> Do you agree this comes from an older form with
> accent on the first syllable?
Do you mean the first syllable of the whole compound? No, I don't, since
<devá-kr.ta-> is not likely to be PIE. The _type_ is old, of course, but
this particular _instance_ may have been coined in Vedic times, for all
I know. *deiw-ó- itself is a vr.ddhi derivative, probably with
contrastive accent originally distinguishing the noun from the
adjective. Judging from other parallel formations, the oldest form was
*déiw-o-, alternating with such forms as the hypothetical collective
*diw-áh2 (cf. *bHérh1g^-o-/*bHr.h1-g^-áh2) or the de-thematic adj.
*diw-jó- (Gk. di^os, Skt. divyá-). The complete story of PIE vr.ddhi has
not been written yet.
Piotr