On 2006-05-24 16:26, Mate Kapović wrote:
> You're basically saying that it was still *k'... I don't see the point in
> talking about "non-distinctive affrication" etc.
I simply mean that practically all true palatal stops (I mean those in
which the point of contact is at the back of the hard palate, not just
slightly advanced velars) are affricated no matter if the speakers are
aware of the affrication or not; in most cases they aren't, and the
affrication has no phonological relevance. But it is important from the
_phonetic_ point of view, since it's dramatically increased by any
further fronting, and the passage from /k'/ to any coronal affricate is
both continuous and natural, hence its wide occurrence cross-linguistically.
I'd say that it's quite _unusual_ for palatal stops to remain
diachronically stable, and that they are far more frequently fronted
than retracted; hence my hunch that PIE *k^ was up to a certain point a
plain velar rather than a palatal. The languages in which it became a
palatal underwent parallel evolution towards full "satemisation", but
since its initial fronting was oddly unconstrained (back vowels didn't
prevent it), I think it's more likely to have happened just once rather
than several times independently (unaccountably, in languages forming a
geographical block).
There is more similarity in the development of Balto-Slavic and
Indo-Iranian (where we also have the early and complete merger of the *k
and *kW series and the RUKI rule), while the "satellite" languages,
Armenian and Albanian, show various idiosyncratic developments.
Piotr