Re: CPA - An ASCII-based phonetic alphabet

> Date:         Mon, 19 Nov 2001 16:12:45 +1100
> From: Tristan Alexander McLeay <anstouh@...>
>
> On Sun, 18 Nov 2001, Lars Henrik Mathiesen wrote:
> > Well, IPA (and thus SAMPA) /w/ is a voiced labial-velar approximant,
> > but I think that Tristan wants a labial-retroflex approximant, without
> > a velar element. So [w`] seems misleading to me too.
>
> Hmm? I didn't want anything! Or maybe I `wanted' a voiced
> labial-velar-retroflex approximate from what I saw, but it was Muke
> who used the [w`] in the first place!
Right, sorry for the misattribution. But the point stands, I think ---
IPA [w] isn't just a labial approximant, so it isn't the right symbol
to start from when transcribing a labial-retroflex one.

I'm not sure if the US English sound in question is labiodental or
bilabial, but I'm guessing the former. Perhaps not coincidentally, a
voiced labiodental approximant happens to be a known idiosyncratic
variant of [r\] in RP; the IPA is script v, X-SAMPA [P] or [v\].

(If bilabial is wanted, the IPA chart has an example for the lowered
diacritic where a beta (voiced bilabial fricative) becomes a voiced
bilabial approximant when lowered (i.e., X-SAMPA [B_o])).

Of course the converter doesn't convert [P`] or [B`_o] either.

> > That aside, the converter page in question seems to balk at any
> > combination of the X-SAMPA rhoticity/retroflexion diacritic with a
> > base letter that doesn't have a corresponding combined form in the
> > Unicode charts. I.e., it accepts things like [r\`], but not [@\`].
>
> There is so a rhoticisied schwa! U+025A according to the PDFs at
> unicode.org, which I think is the closest thing to a definitive guide
> other than those books that I'm never going to get my hands onto :(...
> Anyway, enough rambles from me...
U+025A is X-SAMPA [@`] (and U+025D is X-SAMPA [3`]). When an open-mid
unrounded central vowel ([3]) can be rhotacized, it should be possible
to rhotacize a close-mid unrounded central vowel ([@\]) as well ---
but the converter won't let us.

Lars Mathiesen (U of Copenhagen CS Dep) <thorinn@...> (Humour NOT marked)

Reply