Vowel Length in Omaha-Ponca

[Koontz John E]

In OP there are some words that vary in accentuation.  In the previous
examples of 'thick', for example, this variation occurred with s^u'ga ~
s^uga'.  A prominent example is naN'ba ~ naNba' 'two'.  To a certain exten
one can formulate rules like 'accent falls on every other syllable in a
phrase' or 'accent falls on the second syllable when followed by =bi or
some other enclitic.

Alternating:

xdhabe' naNba' 'two trees' 90:320.6
gdhe'ba naN'ba 'twenty' 90:86.9

Moved by enclitic?

gdhe'bahiN'wiNttaN'ga naNba'=bi=ama '2000 they say' 90:88.8
aN'ba waxu'be naNba'=the=di'=hi=kki 'when two Sundays had passed' 90:661.6

But these only work some of the time.

Not alternating:

hiNbe' naN'ba 'two moccasions' 90:297.14
miN'daNbe naN'ba 'two hours' 90:21.27
z^aN'inaN'ge naNba' 'two wagons' 90:642.2
s^aN'ge wa'xe etta'=xti naNba' 'two of the white people's horses' 90:777.8

Not moved by enclitic:

naN'ba=xti=egaN 'about two' 90:247.16

Neither principle at work:

wakkaN'dagi naN'ba=akha 'the two watermonsters' 90:249.1

Here we see waCV'CV naN'ba and waCV'CV naNba':

wasa'be naN'ba=ma 'the two black bears'  90:18.5-6
tti' waxu'be naNba'=the 'the two sacred tents' 90:462.2
nini'ba waxu'be naN'ba=khe 'the two sacred pipes' 90:471.14

There are often doublets or near doublets, as just above or below:

ni'as^iNga naN'ba 'two people' 90:23.2
ni'as^iNga naNba'=dhaNkha 'the two people' 90:356.218

nu'z^iNga naNba'=akha 'the two boys' 90:86.5
nu'z^iNga naN'ba 'two boys' 90:85.14

maN' naN'ba 'two arrows' 90:46.8
maN' naNba'=dhaNdhaN 'two arrows each' JOD 1890:13
naN' naNba'=akha 'the two grown ones' 90:88.14

====

There seem to be several possibilities here, one being that Dorsey
couldn't hear accent very well, and another being that the principles
governing it are too complex for my deductive powers at present.  I'm not
so sure that I believe the former, though the latter could easily be true.

What has occurred to me is that the word 'two' is naN'baa or maybe
naNaN'baa with H(H)LL as its pitch contour or accentual pattern, and a
final long vowel.  I suspect that a final long vowel would occasionally be
salient enough, especially if emphasized by an alternating stress pattern
or perhaps by certain following enclitics, to come across to English
trained ears as accented.

An alternative I'm less comfortable with is that naNaN'ba alone can be
sometimes perceived as finally accented, though it is always HHL, perhaps
simply because it is not reduced to schwa?  Cases like naNba'=akha would
still be analyzable as hearing something like a long vowel as accented:
naN(aN)'ba=akha heard as naNbaa'kha in spite of a H(H)LL(L) contour.
(The final vowel is elided or voiceless after /kh/.)

====

The comparative data doesn't help here.  Dakotan has nuN'pa, reduplicating
as nuN'mnuNpa, i.e., appearing to be monosyllabic with a stem-forming -a.
Ioway-Otoe has nuN(uN?)'we.  Winnebago has nuNuN'p, but all monosyllables
are long and the final vowel being e after a simple stop has been lost.
The length of the first syllable may be indicated by Winnebago nuNuNpi'wi
'a pair', which retains length in a context of a longer word.

JEK