Re: ontology of glottalized segments?

On Wed, 11 Aug 2004 01:00:19 -0500, Thomas R. Wier <trwier@...>
wrote:

>Hi all.
>
>Does anyone know any phonetic or phonological facts about how
>glottalized consonants arise?  I've heard that acoustically they
>sometimes are treated as "hypervoiceless".  I'm trying to come
>up with a way for them to arise naturally in Phaleran historical
>phonology without having to make reference to borrowings from
>C'ali, which, of course, has plenty of them, and would thus be too
>easy.
Are you referring to glottalic initiation (ejectives and implosives)
or to laryngealized/creaky(-voiced) sonorants?

IIRC the implosives in Sindhi derived from geminate voiced stops,
and I believe a similar mechanism has occurred in some other lgs.

Ejective fricatives in Palantla Chinantec in Mexico derived
from clusters of /K, s, S/ plus /q/ - /q/ became /?/ everywhere
and the /K?, s?, S?/ clusters became ejectives.

I believe the /C?/ origin is used for most 'glottalized' segments
whether ejective or creaky, though for one Khoisan lg (/Ju, I believe)
I recently read that the ejective stops are actually epiglottalized,
(not sure what the ASCII-fication should be - [t?\], [k?\] ?) giving
the impression of [tX'], [kX' ~ qX'].

I think Paul Fallon published his thesis on ejectives and he might
deal with their origin - I have a copy at home, so I'll try to
check it tonight and see if he addresses this issue.


Bfowol