--- In
cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:
> If the voiced aspirates passed through an unvoiced stage in Latin,
then
> *all* unvoiced aspirates -> fricatives developed into voiced
sounds in
> Latin medially (including also *s > *z > r).
Thanks, Miguel, that's of course the main argument, so fundamental
that I forgot to mention it: The unconditioned Italic reflexes of IE
*bh, *dh, *gh are voiceless themselves, so the voicing observed in
medial position can have hit also *original* voiceless aspirates -
even by the rules we have already. It is not easy to represent other
scholars' work.
Jens