10-12-03 16:54, Harald Hammarstrom wrote:
> I found some old discussions but I am not sure I found the convincing
> examples of unstressed Iranian *a borrowed to Sl. *U?
The verb *xUte^ti 'want' (cf. xo~tI 'willingness, desire') is one of the
suspects -- Iranian *hatya- (< *sn.t-je-) : *hanti-. Then we have the
hydronyms *dUne^strU 'Dniester' and *dUne^prU 'Dnieper', which certainly
contain Iranian *da:nu- 'river' > *dan(u)-.
The rivername Tanew in SE Poland has long been suspected of being
Iranian (with early Germanic mediating between Iraniam and Slavic, cf.
Vasmer 1926): *da:nu- > Gmc. *to:n(w)-o: or *to:n-o:n- > Slavic
*tany/*tanUv- > Tanew. However, it has been pointed out that the early
Polish form of the name was Tnew < *tUny or *tIny (it's hard to say
which is more plausible). That, according to Babik's (2001) study of old
toponyms in Poland, rules out *da:nu- as the source form. It doesn't,
though, if we start with Middle Iranian *d[&]nú- with a reduced vowel
--> Gmc. *u or *i --> Slavic *U or *I. But this etymology is on the
whole very uncertain.
Piotr