Richard Wordingham wrote:
> As an indication of the independence of paths, you might like to also
> consider the Romance softening /k/ > /c^/ in Romanian with the
> Daco-Romance softening /t/ > /c/ (i.e. '<tz>').
>
> Richard.
maybe Romanian is not the best alleged example here since there are
several linguist who consider the palatalisation in Romanian as being
the result of the Slavic influence.
Anyway, the change of k >c^ is sure as being the first step of
"alteration" of the sound. The "c^" (tsh) itself is a dubious sound
which can be "heard" different enough to allow a further alteration.
thus, the output "c" (ts) and "sh" appears to be logical.
What one knows for sure is that in the IV-V century, the existence of an
"c"(ts) from "ke" is prooved. The examples given by Detschev in Thracian
should speak for this aspect: tsertselos < cercelus as recorded by
Greek. That will imply, the change of "k" to "c^" did happen before IV
century.
The Albanian "s" for Rom. "c^" comparated with Alb. "q" for Latin "k"
will sustain strongly the idea that the change of "k" to "c^" ( before a
front vowel) did happen before Latin influence in Balkan. That is, this
appears to be an old change which did happen BC.
Alex