Hallo conlangers!

On Sunday 08 June 2014 14:08:05 J P wrote:

> > >Articulation in three places simultaneously might be a bit too much
> > >like hard work.
> > 
> > You think?   I wondered about that but dismissed it, given the great
> > amount of carefully-coordinated choreography that has to go on to
> > articulate much of anything.  I mean, to make [w_k] (creaky [w])
> > involves the same three articulators working simultaneously, and
> > there are certainly languages with [w_k] (UPSID knows seventeen).
> > 
> > Alex
> 
> It's probably not an issue of effort. It's that all speech sounds 
consist of
> modulating an airflow generated by various mechanisms, and this concept 
is
> trying to pull off too many things at that all at once.
> 
> Once you have an ejective [p'] arranged, there are no less than two 
closures
> in place: the buccal one enhancing the flow of air in or out of the 
mouth
> cavity altogether, the glottal one enhancing the pulmonic airflow & 
used to
> create a glottalic airstream mechanism to feed into the 1st closure. 
Adding
> a velar closure between them, though, does pretty much nothing. It 
cannot
> be used as a stop mechanism, because the labial closure blocks any way 
of
> releasing it; it cannot be used for an airstream mechanism, since we've
> already mandated that that will be left to the glottal closure.
> 
> It's possible to arrange a situation where the labial closure gets 
released
> first, then the velar one, but this is just a cluster [pk'] with an
> unreleased first member, not really a singular consonant.
> 
> It's also possible to throw in a velaric airstream method, but this 
will
> then be a glottalized bilabial click, not a labial-velar ejective.

As others have said here, ejectives are *technically*
coarticulations of a glottal stop with another speech sound.
A coarticulated ejective would thus be a coarticulation
involving *three* simultaneous articulatory gestures, and
thus rare.

This poses the question, where do ejectives come from?  They
*could* arise from clusters involving the glottal stop or
something similar (how about a future Danish where ejectives
emerge from stød?).  But there are other pathways.  In those
Eastern Armenian dialects which have them, ejectives are the
regular development of voicless unaspirated stops, probably
under influence from neighbouring Georgian.  I happened to
find something in Michael Fortescue's 1998 book, _Language
Relations across Bering Strait_, where he discusses the origin
of ejectives in Itelmen, an indigenous language of Kamchatka
and the only Chukotko-Kamchatkan language with ejectives.  In
Itelmen, ejectives emerged from various clusters.  Examples:

c'e 'go into' < *c@Gæl-
c'it 'bow' < *t@rit@ri
@mc' 'bitter' < *(@)mcj@- < *cemj@-
k'e- 'get used to' < *kæv-
k'- comitative case < *k(C)- < *kænm@-
sic'eN 'left' < *riNceN < *r@Nac@N
c'aq 'four' < *N@raq

(from Fortescue 1998:40-41)

Any discussion of the origin of ejectives also needs to take
into account that back ejectives such as /k'/ and /q'/ are
somewhat more common than front ones such as /t'/, let alone
/p'/ which is missing from many languages that have ejectives
at other places of articulation.

--
... brought to you by the Weeping Elf
http://www.joerg-rhiemeier.de/Conlang/index.html
"Bêsel asa Éam, a Éam atha cvanthal a cvanth atha Éamal." - SiM 1:1