Re: Ejectives, was Re: New H/G lang?

Ed Heil wrote:

>At least according to my book on phonetics (Catford), an ejective is
>produced by closing the glottis, and raising it to produce pressure in
>the vocal cavity.
>
>Now, the nasal passage is not affected by a closed glottis, which is
I'm not sure what you mean by the nasal passage not affected by a=20
closed glottis, but it is definitely possible for a constricted=20
glottis to affect nasal consonants. E.g., creaky voiced nasals.

>the point of being able to close the glottis -- it allows one to
>swallow without getting food in one's lungs.
Strictly speaking, that's not the glottis that does that. Its the=20
epiglottis. The glottis is the chords in the larynx that vibrates=20
when we talk, the epiglottis is the piece of flesh above the larynx=20
that covers the larynx when we swallow.

>Since a nasal is a consonant where air flows out the nasal passage,
>and the glottis is unable to propel air out the nasal passage, by
>definition one cannot have an ejective nasal.
I don't see why the glottis cannot propel air out of the nasal passage=20
when it can propel air out into the vocal cavity. All you have to do=20
is lower the velum to allow air to escape into the nasal passage.=20
Imagine a nasally release ejective stop (e.g. [p_m' t_n' k_N']. In=20
such sounds, the (constricted) glottis forces air into the oral cavity=20
while there is an oral closure. But instead of releasing the oral=20
closure as you would in normal ejectives, you lower the velum allowing=20
the compressed air from behind the oral closure (and above the glottis)=20
to escape into the the nasal passage.

I must add, however, that I don't think its possible for ejectives=20
without an oral constriction (obstruent) element to exist as a speech=20
sound. Sonorants with a glottal airstream mechanism would not produce=20
any auditorily perceptive sounds to become useful as a speech sound.

>However, one could have a nasal quickly followed by a glottal stop,
>and this might form part of a pattern where it played a role analogous
>to glottalics, so for all intents and purposes there might be a nasal
>in a row of ejectives.
This brings us back to the Salishan languages. Indeed, the creaky=20
voiced nasals of some Salish languages can be narrowly transcribed as=20
[m<creaky>? n<creaky>?]. The feature cuts across the obstruent-sonorant=20
paradigm - the feature being [+/-constricted glottis].

-kristian- 8)