Re: The pronunciation of post-velars (was Re: [tied] *kW- "?")

Grzegorz lists problematic examples for the *q-theory:

> *k^ad- 'fall': Skt. sad-, Lat. cado 'fall' (a-colour
> even if *k^, not *k, evident)

Why can't *a exist in some cases? It would be odd
if it didn't for the vowel system.

> *keH2- *[kah-] 'desire': Skt. ka:- 'desire', Lat.
> ca:rus 'dear' etc. (yes, -a- here but caused by H2
> rather)

Or rather, both *k and *h2 together colour the
intervening *e to /A/. Still, the phoneme must be
*e here.

> *sk^end- 'shine, appear': Skt. chadáyati, Av.
> sadhaiieiti 'seems', prob. Grk.
> kékasmai 'distinguish oneself'

No problems there.

> *skend- 'leap': Skt. skand-, OIr. scendid/scennid
> 'leap' etc. (with -e- instead of expected -a-)

Oh dear :( This example makes me sad... temporarily,
but I must question the source of the Old Irish
vocalism. Are you sure that this stem doesn't
alternate between *sqend- /sk.And-/ and *sqe:nd-
/sk.e:nd-/? This is yet another possibility that
I suspect the average IE lexicon-maker is not
weeding out of her/his lists.


= gLeN


__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com