Re: Gzarondan: Phonological Review
|From:||Dirk Elzinga <dirk_elzinga@...>|
|Date:||Wednesday, October 20, 2004, 21:49|
On Oct 17, 2004, at 11:43 PM, Thomas R. Wier wrote:
> From: "Adrian Morgan (aka Flesh-eating Dragon)"
>> Really? This is something that feels very intuitive to me. Certain
>> consonants are characteristically articulated with slightly more
>> energy than others, which is absorbed by increasing the volume of the
>> following vowel. It sounds wholly plausible and sensible.
> Well, I'm not an expert on the phonetic principles underlying
> phonological typology, but I think what you're describing here
> really explains why so many languages prefer all syllables to
> have onsets, rather than why/whether they are heavy. Most of the
> languages I studied with her did not make any distinction between
> the heaviness of phonemically long vowels and syllables with codas
> (Latin, e.g., treats them the same way). Acoustically this might
> be because vowels with codas tend to be slightly relatively lengthened,
> which also might explain why some classes of consonants make
> syllables heavy while others do not. But really we should be
> asking Dirk this, as he is the local expert.
A Long Time Ago I read a paper (by Stuart Davis (?)) which argued that
Arrernte does take onsets into account in determining stress placement.
But it was and remains a controversial claim, and it is generally
accepted that syllable weight is only sensitive to rhymes.
This reminds me of an interesting problem in Shoshoni
(morpho)phonology. For about two dozen disyllabic verbs, the durative
aspect is formed by gemination of the medial consonant; when this
occurs, stress shifts from the first syllable to the second:
kate ['ka41] 'sit (SG.SUBJ)'
katte [kat't1] 'sitting'
wene ['w1n1] 'stand (SG.SUBJ)'
wenne [w1n'n1] 'standing'
hapi ['haBi] 'lie (SG.SUBJ)'
happi [hap'pi] 'lying'
yekwi ['j1G_wi] 'to say/do something (SG.SUBJ); to sit (PL.SUBJ)'
yekkwi [j1k'k_wi] 'saying/doing something; sitting'
(In the absence of Gemination, intervocalic consonants are subject to
lenition -- hence /p/ -> [B], /t/ -> , /k_w/ -> [G_w], etc.)
Coda consonants are not moraic in Shoshoni, but this pattern is still
unexpected from a purely phonological point of view. Either the stress
shift has become one of the markers for durative aspect, and is thus
under morphological control, or the onset is playing a role in stress
placement (or the stress placement requires a good strong onset).
Now that I'm busy again with Shoshoni fieldwork, I may get to find out
about this (and other goodies).
"I believe that phonology is superior to music. It is more variable and
its pecuniary possibilities are far greater." - Erik Satie