Re: a request for input.... vocalic inventory

Byron cande:

>i want to go from a system of:
>
>a: e: i: o:
>a i o
>
>to:
>
>a: i: u:
>a i u @
>
>-with a low vs. high tonal distinction, just to make
>things interesting....
>
>so, any ideas how i might derive such a thing from
>existing Cree vocabulary? i realise not having my
>wanted consonant inventory makes it a bit speculative,
>but still... any input would be nice.
   Aha! Easy - very similar to Old Irish development!

   First, where do the high/low tones present? Are long vowels high tone?
I've got other ideas, which I'll get to in a second.

   e: > i: (you might also have shortened e>i show up in certain contexts)
   o, o: > u, u:
   (all of these exactly as in Old irish)

   @ arises in unstressed syllables (from a or orig. o), or from /&/ as a
result of i-umlaut. Or some other way.

   High tone can be an effect of stress - IE supposedly had tone determined
stress, as does Greek. Another way to do it would be that closed syllables
(syllables ending with a consonant not a vowel) become low tone, with high
tone in open syllables. This would happen at an early stage,  that later
sound changes could have low tone open syllables, and high tone closed
syllables. Is there a neutral tone too?

    Examples based on totally random words (stress ', high tone with -
after, low tone unmarked):

    tanopse: > t@-'nu:-psi:- or ta-
    kantora > kan'tu-r@-

    For a properly Amerindian sort of lang, I think you'd need to have
vowels (and clusters) affecting consonants too (tua > twa), and you should
probably allow allophonic variations (so that a>o before u in next syllable
kan'tu-r@- > kon'tu-r@-, and prob i>e/_a, u>o/_a, a>e/_i, and so on).

    Hope this helps,

   Aidan