Re: [tied] Re: Albanian (1)

On Thu, 22 Jan 2004 14:28:17 +0100, Piotr Gasiorowski
<piotr.gasiorowski@...> wrote:

>It's just a matter of relative plausibility. I realise full well that my
>Proto-Satem is other people's PIE, but it's easier for me to accept a single
>shift of *k : *k_ > *k' : *k in Proto-Satem than the reverse shift
>happening independently in the various Centum groups.

Still, another shift, independently or not, has to be assumed for the
centum languages: the merger of *k^ and *k.

>Any phonetician can
>confirm that for purely physical reasons the affrication of [k'] is
>practically inevitable. (I was once surprised when a Canadian phonologist
>who ran a phonetics class I attended found my Polish [k'] strongly
>affricated, but I had to admit he was right.) The development of [k'] into a
>coronal affricate or fricative is therefore far more natural than its
>retraction. That's my reason for believing that those languages that show
>velar [k] for PIE *k^ never has a palatal stop in the first place.

I agree, although it must be added that retraction of a palatal stop is not
entirely impossible, merely unlikely. The standard counterexample is
(Egyptian) Arabic. PS *g was spontaneously fronted to *d^ in
"Proto-Arabic". This developed to /3^/, /z^/ in the various Arabic
dialects, as expected. In Egyptian Arabic, however, it was backed again to
/g/. Note that uvular /q/ is fronted to /g/ in many Arabic dialects, to
fill the gap left by *g, but that it goes to /?/ in Egyptian Arabic.


=======================
Miguel Carrasquer Vidal
mcv@...