--- In cybalist@..., guto rhys <gutorhys@...> wrote:1870
>
> The change initial k>p in some IE dialects such as Oscan and some
dialects of Celtic seems to me very irregular as the two sounds have
such different places of articulation. By what means did this mutation
take place or was there an original or intermediate sound such as a
guttural bilabial unvoiced plosive if such a thing is possible.
The change was not k>p,g>b but instead kW>p,gW>b in p-Celtic
(Brythonic, Gaulish), the Osco-Umbrian branch of Italic, and
some ancient Greek dialects (Aeolic). Here kW,gW are so-called
labiovelars. Their exact phonological value can be debated,
but their existence can be verified as distinct from the velars
k,g in several ancient Indo-European scripts, such as Linear B
for Mycenaean Greek and Ogham script for primitive Irish.
I raised the question myself about the phonological sequence
of the transformations kW>p,gW>b some time ago (message #1870)
and my query was amply answered by the experts. You can pick
up the thread of my old message. In any case, the transformation
is less odd than you state, because a labial element of some sort
was surely present in the pronunciation of kW,gW (whatever the exact
value or values they had)
Best - Gregory