Status of "u" in Omaha-Ponca

[Koontz John E]

Dhegiha, just when you were ready to write *V[a features] > V[a features]
in the list of Siouan sound changes.

On Sat, 13 Aug 2005, Rankin, Robert L wrote:
> Yes, proto-Siouan *u merged with "i" just as you say.  But THEN,
> afterward, "o" began being pronounced as "u".  So nowadays there are
> still only 4 oral vowels, i, e, a, u.  (As recently as the 1970's I was
> still hearing many of the modern "u" sounds still pronounced as "o".

As Bob has explained, we have two changes here.  First a merger of *u and
*i as i, which also occurs in Quapaw.  Second, a shift of *o to u.  I'd
assume that the two changes are less a matter of sequential independent
changes than parallel, associated, and simultaneous shifts in the use of
the vowel space.  However, there are problems with that notion.  First,
there is the failure of *o to shift in Quapaw (per Dorsey, who says it
shifts in OP) and the question of whether Omaha-Ponca /u/ is really [u]
(see below).  Second, we do have a distinct tendency for *u to be [<u">]
(a high rounded front vowel] in Osage and Kaw, without any remarked upon
shift upward in *o.

For what it's worth, I am fairly sure that both Kaw and Osage have in some
cases substituted u [u"] for *i or i for *u in some sets, so that, even
without a merger there seems to have been some confusion or swapping.
Example, Os -ci, OP =di LOC, cf. Dakotan =l ~ =tu, or Kaw -gu- DAT, OP
-gi-, cf. Dakotan -ki-.  I suppose this would have to be a case of
"isoglosses" for some words or morphemes with /u/ or /i/ shifting ahead of
others, though this is an idea that not all linguists happy with.
Anyway, I do not believe this ever happens in accented syllables or
non-grammatical elements.  In short, it happens in unaccented and/or
analogy-prone environments.

Since Dorsey always writes u for PS > MDh *o, I assume that the shift to
the present value was complete by the 1880s.  I don't recall noticing
anything bearing on earlier values in the limited earlier sources.
Maximillian might be the best place to look.

I've always assumed that Bob's "o" examples result from his better
appreciation of what a genuine IPA o would be.  English speakers in
general are ill-prepared to perceive o and many Americans are weak on u,
too, I've noticed.  Also, I may be worse than average at distinguishing
either, due to personal and dialect deficiencies.  In regard to the latter
admission, I was born in the same dialect area as JOD (Tidewater) and
proceeded to Colorado via Detroit, which would be enough to dull anyone's
vocalic sensibilities.  As I recall Tidewater English doesn't have o or u
or <ae> (aesc).  It has e^w, <tense u><lax u>, and e^<ae>.  (I'm not sure
I could do better with IPA.)  It was some years before folks in Colorado
stopped asked me where I was from.  Most thought England.

As to the personal deficiences above and beyond dialect, I am coming to
realize that I have not always risen attentively to the details of Omaha
phonology.

This said, I've never had any problem hearing any and all of the higher
back vowels in Omaha as u (as opposed to a, though not necessarily to
anything else).  It *might* be a lower sort of u, but I'm not able to say.

What is usually written o in Omaha-Ponca (e.g., by LaFlesche and in the
current popular systems) is the /au/ diphthong that arises when the male
speech particle /u/ is added to a final /a/, e.g., dadi-hau < dadi
'father' + ha 'vocative' + u 'male speaker'.  This also occurs in the
particle /hau/ used in male-to-male greetings, male expressions of
approbation, etc.  This is probably not a coincidence, though /hau/ as a
greeting is somewhat widespread in North America, especially Northeastern
North America, but, e.g., it is also found in Hopi, I think.

Just to be clear, what Bob consistantly hears (and usually writes) as o in
Omaha-Ponca is not this /au/ diphthong, but /u/ (i.e., what Dorsey,
LaFlesche, the current popular systems and others, including myself write
as u).

For what it is worth, the three nasal vowels of Proto-Siouan (and, e.g.,
Dakotan), i.e., iN, aN, and uN - the latter more like [oN] in Dakotan -
come out in Dorsey's writing of OP as iN and aN, the latter representing a
merger of aN and uN.  LaFlesche writes oN for this merger.  In fact, he
usually writes moN and noN where Dorsey writes ma and na (with no
indication of nasality) for /maN/ and /naN/, though he retains ma and na
for unnasalized ma and na, e.g., in ama 'the' < *aWa.  One interesting
minimal pair for na vs. naN are the instrumenals na= < *Ra= 'by heat,
spontaneously' vs. naN- 'by foot'.

If you see the merger of aN and oN as a raising to oN, then this parallels
the up and forward rotation of u to u" or i and o to u.  However,
Dhegihanists are not really sure if *aN and *uN have merged.  On alternate
days they think they just have trouble hearing the difference.  Maybe some
aN have merged with uN as something like oN and some have not.  Some
examples seem to turn on adjacent x vs. adjacent gh (gamma).  Some may be
a matter of length of word final position.