k(w)->p

Padraic Brown wrote:
> On Mon, 24 Jan 2000, Rob Nierse wrote:
>
> >Question:
> >is it possible / reasonable to convert 'q' [k_w] to [p] ?
> >and 'hu' [x_w] to [f]  ?
>
> Why not? Some Celts take IE *kwetwer and end up with pedwar.
The IE labalized velar turning into a labial always struck me as
really bizarre.  I think I heard someone on the list here mention that
they thought PIE k(w) was really kp -- a doubly articulated stop.

kp -> k, p, k(w) seems intuitively easier for me to buy than
k(w) -> k, p, k(w) but I have no idea if it's really reasonable.

Does any historical linguist out there know if we have any other,
independent evidence of labialized velars becoming labials, or, on the
other hand, of doubly articulated labial-velars becoming labialized
velars?


---------------------------------------------------------------------
                       edheil@postmark.net
---------------------------------------------------------------------
"Bill Gates is a white Persian cat and a monocle away from becoming
another James Bond villain. 'No Mr Bond, I expect you to upgrade.'"
                                           --Dennis Miller
---------------------------------------------------------------------