Re: Phono

Aidan Grey wrote:
>
> --- Robert Hailman <robert@...> wrote:
>
> > However, if I define "i" as being [=long], then if I
> > ever come across a
> > long /i/, the program won't use "i" to represent it.
> > If you don't have a
> > letter defined to represent long /i/, it'll probably
> > show a question
> > mark in it's place.
>
>     Just to make sure I understand it:
>
>     So, if at some point /t/ differentiates into
> aspirated and unaspirated, then I would have /t/ as
> [=asp] and /th/ as /T/[#asp]. Then when the output
> phase hits, /th/ must become /T/, because it is always
> aspirated. Working backwards, if the two merged, with
> /T/ > /t/, then I would mark them both with single
> features (+,-), and /T/ would show up highlighted,
> with a [+asp] feature elaboration.
>
>     Do I have it right?
>
>     Thanks a bunch. This explanation made more sense
> (if I got it right, it made a LOT more sense!).
Yeah, that's exactly it. It's fairly straight forward, once you get the
hang of it.

A fun thing to do, is make a sound change whereby vowels become
non-syllabic, just to see what comes out. It's really quite amazing how
Phono tries to cope with things like that. The problem arises because
all the vowels are [#syllabic]...

--
Robert