On 2008-11-09 22:31, Rick McCallister wrote:
> [*taikuraz 'brother-in-law']
> Is it a contamination from *taikna- in the sense of "pledge, arrhes"?
> Would it have been reanalyzed ss "pledger, one who has paid dowry"?
Unlikely, since we have the same unexpected *k (apparently via *g and
Grimm's Law) in several other words. One possible solution is "Cowgill's
Law", defended by Katz and Ringe (the proposal is that *h2/3 became a
stop between a sonorant and *w, so *daih2w(e)r- > *taikwr-). There is,
however, a more elegant alternative analysis of this "intrusive velar"
by Seebold: *w > pre-Gmc. *g between a sonorant and *u (later than the
loss of *h2 in *daih2we:r-): *daih2-wr.- > *daiwur- > *daigur- >
*taikur-(+ *-a-z) > OE ta:cor.
Piotr