On Thu, 24 Feb 2005 13:51:21 +0100, Piotr Gasiorowski
>On 05-02-24 12:58, Miguel Carrasquer wrote:
>> Perhaps "slack" is not the most appropriate term. If we
>> start from modal voice and progressively constrict the
>> arytenoids, we get stiff voice, creaky voice and glottal
>> stop. If we progressively widen the arytenoids, we get
>> slack voice, breathy voice, voicelessness and aspiration.
>This is a gross oversimplifaction.
No it isn't. It's a precise description of what happens on
the arytenoid side of things.
>Phonation is a function of _two_
>tensions, adductive and longitudinal, so you can't very well arrange all
>phonation types along a scale where only the adduction or abduction of
>the arytenoids matters.
Why not? We can arrange vowels from back to front.
>It so happens that the high pitch induced by
>voiceless obstruents is caused by the action of the cricothyroids,
>lengthening and thinning the vocal folds and thus increasing their
>_longitudinal_ tension. This has little to do directly with the absence
>of _adductice_ tension (except that both gestures conspire to expand the
>glottis). For modal voice, tension along both dimensions is only
>moderate. With weak longitudinal tension but very strong adductive
>tension you get creaky voice; with minimal adductive tension and medium
>longitudinal tension, breathy voice, etc.
What I was interested in was creating a square or trapezoid
representation of the two factors, in the way the vowels are
The arytenoid dimension is clear. The longitudinal
dimension (raising/lowering of the larynx and/or tension of
the cricothyroids) is not described adequately in e.g.
Ladefoged and Maddieson. We have a downward movement of the
larynx in the case of implosives, and an upward movement in
the case of ejectives, but how the other sounds relate to
this in the longitudinal dimension is not very well
explained (on p. 64 it is suggested that slack voice is
associated with a lowering of the larynx [lowered F1]), and
somewhere else I remember seeing a suggestion that the
larynx is more raised in voiceless aspirates. A preliminary
representation would be:
high close ----------------------- open
| ?/t' tH/h
| 'd <----------> 'd
But I suspect there's more to it than this.
Miguel Carrasquer Vidal