---In cybalist@yahoogroups.com, <dgkilday57@...> wrote :
>
[...] This etymology makes a great deal more sense
than the arbitrary deformation of *acu-peter 'fast flier' into _accipiter_ (IEW
19), particularly since such a first element undergoes no alteration in Lat.
_acupedius_ 'swift-footed'.
>
Don't forget the statement which says "each word has its own history", so whatever happened to accipiter can't be contradicted by acupedius. As for the geminate /kk/ in this root, there're traces of it in VL 'water', as shown by the Appendix Probi (aqua non acqua) and Italian acqua.
> DGK: Please don't tease! If you have an alternative to "std PIE theory" which DOES explain the relationship you have proposed, the rest of us need to see it in order to evaluate it!
>
My point is the IE lexicon comes from several sources/protolanguages which in some cases would be more or less close relatives. This stratification is the result of a series of expansions and language replacement processes over several millenia which ultimately led to the historical IE languages. This way, within IE we can have a word A cognate to another word B because they stem from different protolanguages.
In the case of the etymological doublet *aku- ~ *ōk´u-, the former belongs to the Paleo-European substrate reflected in the ancient hydronymy, while the former belongs to a more recent layer, probably the one regarded as native PIE by the std theory.