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

R. Nierse wrote:

>> Van: Ed Heil <edheil@...>
>> Aan: Multiple recipients of list CONLANG <CONLANG@...>
>> Onderwerp: Ejectives, was Re: New H/G lang?
>> Datum: dinsdag 12 oktober 1999 23:24
>
>> I don't believe ejective nasals are possible, because the glottis
>> cannot be used to produce an airstream in the nasal tract, only the
>> oral cavity.
>They do exist: Sechelt (Salishan) q@m'=E9l 'to become high tide'
>also difficult to pronounce: n@w'os 'face',=20
Actually, the so called 'ejective nasals' of the salishan and other=20
northwest coast indian languages are more appropriately called nasals=20
with creaky voice (i.e., laryngealized nasals). Ladefoged and Maddieson=20
in their book 'Sounds of the Worlds Languages' shows Spectrograms of=20
some Salishan words and the 'ejective nasals' clearly have vocal chord=20
vibration but of a more creaky type. Indeed, Ladefoged and Maddieson=20
labels these as creaky voiced nasals.

I don't think ejective nasals are possible as speech sounds for that=20
would require a sufficient amount of airstream to exit the nose. A=20
glottalic airstream is hardly sufficient in my opinion for it to be=20
auditorily perceptive.

>>  And there can be no voice with an ejective, because the
>> vocal cords are shut completely in order to produce the ejective
>> pressure.  And ejectives must be of fairly short duration, because
>> there is a fairly small volume of air available above the glottis.
>> But beyond that, anything's possible: glottalic fricatives and
>> affricates are certainly common enough.
Ditto! Hence, no ejectives nasal (see above).

>Thomas wrote:
>> So, does this language have a phonemic glottal stop?  If so, wouldn't
>> it be better just to say it's a consonant cluster in which a glottal =
stop
>> is the second element in the cluster?
>Yes it is a consonant cluster. Nevertheless it is a minimal pair in
>Yucatec.
>
>> Besides, "glottalic" (as an adjective)
>> is already most often used in the literature to refer specifically to =
the
>> air stream mechanism, not the place of articulation ("glottal" is =
used
>for
>> that).
>We must keep "glottalic" for air stream mechanism and discard the term
>"glottalized" (and use "ejective"). I even don't know of a natlang that =
has
>[p?] or [t?] instead of [p'] or [t'].
I suspect that there are two types of what many have labeled =
'ejectives'.=20
A 'true' ejective would clearly have a glottalic airstream. So an =
ejective=20
stop would be followed by a brief period of silence as the glottis is=20
constricted and forced upwards. Then there are those 'ejectives'=20
that are followed by a brief period of creaky voicing, because the=20
glottis is not as constricted.

Waveforms shown in Ladefoged and Maddiesons book of ejectives in Navajo=20
and Hausa show that there are clear differences. The Navajo ejectives=20
are released into creaky voicing, while the Hausa ejectives are followed =

by a very brief period of silence. Yet the glottis in the Hausa =
ejectives=20
is release more quickly into normal voicing than it is in Navajo, where=20
the glottis is only released gradually.

One could therefore postulate that the glottis in [p?] would be released =

gradually into a creaky voice (as it is in Navajo), while [p'] would=20
have to be released into a very brief period of silence due to a very=20
constricted glottis.

Its all a matter of interpretation. And as all interpretation goes, may=20
be subject to a little debate.

-kristian- 8)