--- andrew jarrette <
jarretteandrew@...>
wrote:
I've never understood why in Greek *kw becomes /t/
before *i and *e.
Sean:
Not in every dialect; there's plenty of variation
(such as Boeotian pettares "four" with kW>p before e).
Why did *kw become palatalized and not *k/k' ?
The palatalized velars had probably merged with
plains before any kW>k^>t. A similar "odd" change is
tw>tW>t^>s but not for dw. Uncommon changes sometimes
occur; there were enough dialects of Greek for
something unusual to happen sometimes.
Also in Greek *gw similarly becomes /d/ before *e, but
before *i becomes /b/ -- why the inconsistent
treatment of labiovelars? This reminds me of another
phenomenon I've never understood: why *kw becomes /c/
in Irish, and *gwh becomes /g/, but /gw/ becomes /b/.
Why should *gw be treated so differently from *kw and
*gwh?
Part of the reason is probably a lack of many *b in
PIE; many languages created b from bH, p, gW, m, etc.
in different environments to balance the phonological
system.
Also, I seem to remember gWH > gW and then to g in
Irish (with rounding of some vowels) and to gw in
Welsh.
By the way, doesn't Armenian only palatalize
labiovelars, and not plain velars (or is it
Albanian?)?
Albanian, but it's not universally accepted.
Again I don't understand why labiovelars should be
more prone to palatalization than plain velars.
I don't think they're more likely to change; it just
happens sometimes. The Greek and Albanian sound
change may actually have been an areal change
depending on how long ago it happened.
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com