Re: Time to play Identify Those Phones, and a bit of a pharyngeal question

Kristian:
> Eric Christopherson wrote:
> >I thought all the back vowels had
> >about the same degree of "backedness;" at least that's what the IPA chart's
> >arrangement of them would have one believe.
>
> The chart is only an approximate mapping of the vowel space.
> The vowel space is afterall 3-dimensional and it is difficult
> to represent it on 2-dimensional paper. Its like projecting the
> globe on a piece of paper -- distortions are bound to occur.
> I wonder if anyone has created the phonetician's equivalent
> of a cartographer's globe -- a 3-dimensional projection of
> the vowel space. That could be a very useful tool for
> conlangers.
There was a proposal for 3D vowel space conventions in a paper in, I
think, Journal of Linguistics, by, I think, Roger Lass and possibly
someone else, some time around the late 80s early 90s. But the 3rd
dimension was for lip rounding.

Me I think 2D works pretty well for vowels and tectal approximants,
except for laterals and rhotics and rhotacized vowels. One axis for
F1 or distance from larynx to constriction, & the other axis for
F2 and constriction to labial end of the oral tube. Laterals obviously
complicate because of the lateral scrunching of the tongue body, and
rhotics complicate because they add a second constriction (or create
a secondary resonant cavity). Anyway, the IPA chart is misleading not
so much because of its 2-dimensionality but because of its idealization
and schematization.

What I find more annoying with the IPA is the typographical aspect. Many
characters just don't lend themselves to diacritics, because of ascenders
and descenders, and the whole thing is rather pragmatic and unsystematic.
Better would be some kind of analogue of the musical stave; I seem to
recall a phonological notation (gestural phonology?) based on the stave
in a paper by ??Colin Ewen and possibly someone else also in Journal of
Linguistics from c 1990. Or possibly in Phonology (Yearbook). A compromise
would be to have, say, all oral articulations represented by x-height
characters, all laryngeal articulations by under-diacritics, and all
tweaky diacritics over the main character. That's not a full solution, but
it would still make life easier.

--And.