> Is there any particular reason not to consider paraf a borrowing from
> Latin para:re 'to prepare, provide, procure' (as did Loth)? The
> semantic distance is not that great and we already have W parawd,
> Bret. paret from Lat. para:tus.
If you are referring to Loth's "Les Mots Latins Dans Les Langues
Brittoniques", I have to say that I don't find this book to be
completely reliable. Regardless, Pokorny (IEW) has it as a native
Celtic root and I think that most scholars these days follow him on
this matter (remember that Irish and Welsh words such as
creth/cruth/Crithe and pryd/prydydd are believed to be derivatives of
this root and that there are attested Gaulish names which seem to be
cognate - Pritto, Prittio, et al.).
- Chris Gwinn