On Thu, Nov 19, 2009 at 6:58 AM, Paul Bennett <[log in to unmask]> wrote:
> Syllables (and here's where the notion of "syllable" gets a bit abstract,
> and the notion of "vowel" for that matter) can be "glottalized". What this
> means is that they may take an inserted /?/ (written with an apos'phe,
> alphabetically ordered after "y"). Syllables with short vowels may only take
> the /?/ between the onset and the vowel, where it is realized as ejectivity,
> for instance "p'g" /p_>Ag/. Syllables with long vowels may take the /?/
> between the onset and the vowel ("p'g" /p_>A:g/) or "in the middle" of the
> vowel, where it is realised as a full glottal stop ("p'g" /pA?Ag/).
> Similarly, syllables with diphthongs may take the /?/ between the onset and
> vowel ("p'yg" /p_>Ajg/) or "in the middle" of the diphthong, in which case
> the /j/ becomes an /i/, and is likewise written "i" ("p'ig" /pA?ig/).

Hmm. I don't think this analysis makes sense. How are you analyzing
your ejectives as being the same phenomenon as your glottalized
vowels, when at the same time you say they contrast? Instead I'd say
that you have two phenomena here: ejectiveness or glottalization in
onset consonants (is this possible for all onsets, or just the stops?)
and glottalization of long vowels, with a restriction (if I understand
your rules correctly) that glottalized vowels cannot follow an
ejective consonant.

Not sure if any natlangs restrict glottalization to long vowels.

Glottalization of vowels is, AIUI, usually creaky voice, not a full
fledged glottal stop, which would probably be interpreted as a
consonant in its own right. If you really mean the glottal stop,
though, you'd probably want to interpret it as an onset-only
consonant, with a restriction that it cannot begin a word or follow
another consonant, and maybe a rule that it can't follow or precede a
long vowel (which would restrict it to exactly the same environments
as your analysis implies, but may be questionable and is probably
superfluous). Ejectives would remain their own thing.

> That's the sketch as it stands, with the note that noncompound words would
> consist of a single "syllable" m.m. the above deviations from the normal
> usage of that word.

Not sure what you mean by this.

> What I'm thinking of doing with this, is adding /h/ (and /_h/) as a second
> type of glottalization (written "h" and ordered between "g" and "i"). It
> occurred to me very recently that while I had the maintenance hatch open and
> my toolkit out, I could probably add something in the /X/ or /X\/ milieu to
> the mix as well -- though the only sensible remaining letter would be "r",
> ordered between "p" and "s".

You could do the same thing with /h/, with an aspiration distinction
in onset consonants (also contrasting with glottalization), and /h/
proper as an onset consonant that cannot begin a word or be adjacent
to other consonants.