Dorsey u circumflex in Biloxi

[Koontz John E]

On Wed, 2 Feb 2005, David Kaufman wrote:
> I guess a lot of my confusion is just from a lack of overall Siouan
> knowledge at this point, but just to make sure I've got this straight:
> you're saying that a stressed schwa can only come about from a
> previously nasalized stressed A.

To be precise - and I apologize if David just omitted this for simplicity
in his restatement - Bob actually said that in Ofo (as opposed to Siouan
in general or Biloxi) he has observed unaccented schwa representing what
comparison suggests is an unaccented /a/ or unaccented /aN/, and accented
schwa representing accented /aN/.  However, I suspect Bob meant this an
encouragement to interpret schwa, accented or not, as a or aN, if other
evidence supported it, or, more simply as a, with nasalization determined
from other evidence.

I mentioned that I have heard unaccented final schwa representing
unaccented aN in Omaha and seen the corresponding words spelled
occasionally in various contexts with final a (not a + raised n), e.g.
/umaNhaN/ as [umaNh<schwa>], egaN as [eg<schwa>] or [ig<schwa>], gdhebaN
as [gdheb<scwa>], and so on.

In a family with vowel systems that typically amount to aeiou aNiNuN or
aeiu aNiN (or aNiNoN?) neglecting length it's reasonable to suspect a
schwa in "preliminary transcription" might represent a or aN, accented or
not.

Dorsey and Swanton worked in innocence of any formal phonemic concept and
probably had no real concept of contextual variation either.  In their
approach the height of linguistic rigor was a narrow phonetic
trancription.  Dorsey uses a broad phonetic transcription for most
purposes, and notes unusual contextual variants more narrowly.  Some sorts
of phonetic detail clearly attract his attention more than others.

He does a lot more specification of phonetic detail in Biloxi, it seems to
me, than elsewhere, perhaps because it was unfamiliar.  Or maybe I'm just
more familiar with the key than the data in this case!  A broad phonetic
transcriptions might in some cases amount to a phonemic transcription, but
not always.  It depends on what guides the instincts in broadening -
merging - things and on how far one goes.

The mapping of vowel phonemes and their allophones to phonetic vowels can
sometimes be surprising.  Check out Marshallese - an example I noticed
recently.

> And I suppose there's no reason to suspect that the underlying Siouan
> sa:pe or sepi (a or e) would have become nasalized in Biloxi, then
> denasalized again to a stressed schwa.  Interesting.  What a puzzle!

Nope on the process in my judgement.  I expect it's just *sapi written
s<schwa>pi.

Siouan nasal vowels are often apparently denasalized in certain contexts
(e.g., after a nasal sonorant or finally or initially).  I've heard it
asserted that some speakers nasalize or denasalize more than others.
Speakers may denasalize sporadically, perhaps more often in certain
contexts, perhaps more often in certain words.  Thus one utterance of a
word might seem to differ from another utterance at another time by the
same speaker.  Denasalization may affect only vowels or also nasal
sonorants.

Nasalization is essentially a supersegmental quantity in most Siouan
languages and can spread across certain consonants (sonorants,
semi-vowels, laryngeals, etc.).  Some languages display regular variations
in nasality of vowels or sonorants determined by morphological processes
that result in adding or removing sources of nasality or barriers to its
spreading.  Nasalization may "occur" scrunched up at one end of a sequence
over which it can spread.  One could account for /umaNhaN/ as
[umaNh<schwa>] in those terms, too.

Nasalization sometimes simply lands off the target vowel, even on an
adjacent vowel across a theoretically "impermeable" consonant, e.g., I've
heard an Omaha speaker say [iNga] (or maybe [iNg<schwa>] in fast speech
for [egaN].