Re: USAGE: YAGPT (was Re: YASPR -- Yet Another Swedish Pronunciation Rant (fuit: THEORY: NATLANGS: Phonology and Phonetics: Tetraphthongs, Triphthongs, Diphthongs))

Hi!

Andreas Johansson writes:
> Citerar Philip Newton <philip.newton@...>:
> >...
> > I presume you mean [o2] for |eu, äu|? That seems wrong for me; I've
> > seen it rendered [OY], though, and that seems to match what I say
> > pretty well.
>
> FWIW, my German textbook writes [ae Ao O2] for |ei au eu|, whereas Duden writes
> |ai au Oy|. I've also seen [OY], but I can't recall seeing it with a mid-high
> initial element before. [o2] would be a flat diphthong, which feels wrong to
> me.
>
> Now, it's probably pronounced as [uQ] in some dialect just to spite me ...
:-)

In the text-to-speech system I helped implementing once, the phones
were written as [aI], [aU] and [OI], but the latter was probably more
like [OY] for the two given speakers.

Duden, unless the 'Ausspracheduden' usually seems to use a strange
mixture of phonemic and phonetic description, so using [i] and [u] and
[y] as the second part instead of the lax counterparts, might be for
simplicity reasons.  If you use the lax variants, you're exactly were
our T2S system was, and I think 'Ausspracheduden' uses those, too.

I find [A] quite strange, actually, since I'd find it at best heavy
dialect to use a back vowel for any /a/ in German, which should be
contralised (so [a_"U)] and [a_"I)] would probably be most precise for
most speakers).  Long /a:/ is often given as [A:] in dictionaries, but
I think mainly for historical reasons and/or to include the quality
distinction made in coastal dialects.

Anyway.

Talking about dialects, for myself I doubt that I really pronounce
[OY] or [aU] (short for [a_"U)]).  My own classification for myself
would be [Oe] (yes, I'm quite sure about missing rounding here) and
[ao] (or better [a_"o]), but I'd probably need a recording and a
spectrum analysis to really be sure.  I still write [OY] and [aU] to
not confuse readers too much. :-)

**Henrik