Re: THEORY: on the teleology of conlanging (was: RE: terminaldialect?)

> Joshua Shinavier wrote:
> > initial r's are pronounced [hr] and r's which precede stops (any stops, not
> > only voiceless ones) and nasals are devoiced (e.g. karn [k&rn] -- is there
> > an ASCII IPA symbol for a devoiced [r]?)
>
> The symbol is just the voiceless diacritic, in most SAMPA, [_0] under
> the [r], thus [r_0] ([r] with a circle underneath in regular IPA).
>
> That said, that's an improbable environment.  If it were preceding
> VOICELESS stops, that would be likely, but for a voiced sound to become
> devoiced when adjacent to other VOICED sounds is unlikely.  [Kurt] -->
> [k&r_0t] is reasonable, but [k&rn] --> [k&r_0n] is extremely
> improbable.  [k&tra] --> [k&tr_0a] is also probable, since there it
> follows a voiceless sound.
Well, I imagine this as a pretty backwards rural dialect, so the improbable
just might happen :)
In Aroven clusters like -rn are usually followed by a vowel, which makes the
/r/ a lot easier to de-voice.
It actually sounds kind of neat, though, if not especially beautiful.
Personally I think glo'al stops in the middle of words are weirder.

Josh