From: stlatos
Message: 70378
Date: 2012-11-01
>If K>P anyway, no k>x is needed in place of k>p.
> At 6:07:12 PM on Wednesday, October 31, 2012, stlatos wrote:
>
> > --- In cybalist@yahoogroups.com, "Brian M. Scott"
> > <bm.brian@> wrote:
>
> >> At 8:59:53 PM on Tuesday, October 30, 2012, stlatos wrote:
>
> >>> --- In cybalist@yahoogroups.com, Bhrihskwobhloukstroy
> >>> <bhrihstlobhrouzghdhroy@> wrote:
>
> >>>> Does the sequence /kt/ > /tÊ/ due to /kt/ > /xt/ > /çt/
> >>>> > /jt/ > /tj/ (suggested by areal dialectology) really
> >>>> imply palatalization of anteconsonantal *velar* /n/?
>
> >>> The oddity of KC in Romance is easily seen in Rum. kt >
> >>> pt , ks > ps (octo: > opt , coxa > coapsA);
>
> >> A sequence /kt/ > /xt/ > /Ït/ > /pt/ (and similarly for
> >> /ks/ is plausible and not especially odd.
>
> > Then why Nn > mn ?
>
> Why not? It's analogous to the middle step of the other
> sequences, and the first and last steps can't apply.
>