07-07-03 15:46, tgpedersen wrote:
> Is it possible to see the PIE root for Gothic ansu- 'god' as
> composed of two roots: *h2an- "breathe; spirit" and *suH- "force"
> (Savitar, also in the reflexive stem *sw-, and various kinship terms)?
In a recent article, Martin Huld very persuasively argues against
reconstructing the Proto-Germanic word as a u-stem. To be sure, it was
_partly_ attracted into the u-declension in Proto-Norse, but there's
practically no extra-Norse support for *ansu- (though the word is
attested in Old English, Gothic and in OHG onomastics, e.g. in the name
Anselm < *ansa-xelmaz). Huld also refutes the idea that it was a
consonantal stem (*ans-). Most probably it was simply *ansaz with a
heteroclitic i-stem plural (a typical Germanic collective, as in ethnic
names of the Danir/Dene type; gods also formed "tribal" groups!). But if
so, it isn't likely to have anything to do with Indo-Iranian *asura- or
Hitt. hassuwai- 'reign', and may be analysed as *h2onh1-so- 'breath,
spirit' from *h2anh1- 'breathe', like *faxsa- 'head hair' < *pok^t-so-,
from a verb root meaning 'comb' (Lat. pecto:). That would make you 50%
right.
Piotr