Omaha nasal vowels

[Koontz John E]

On Thu, 24 Aug 2000, Ardis R Eschenberg wrote:
> Okay I will need help with the historical reasonings as the only
> historical notions I have are from comparing UmoNhoN with the Lakhota I
> know (not very in-depth!).

Let me know aide, if you like, what was puzzling.  But basically, to a
large extent m and n in Siouan languages are conditioned variants of w and
r (or dh, in Dheigha), when a nasal vowel follows, i.e., wVN => mVN.
There are certain exceptions in each language, e.g., Dakotan wiNyaN
'woman' vs. OP miN in miNga, s^emi(N)z^iNga, etc., and, another
complication, a few cases of n in particular seem to be "ancient," e.g.,
nie (approx) 'be pained'.

The particular additional complication in OP is that certain other
reconstructed sounds, written *W and *R to distinguish them from *w and
*r, whose exact phonology is not clear, come out in OP as m and n, whereas
in other languages they might be w and l (Dakotan) or p and t ~ c (Osage).
Cases of m and n from these in OP can occur with a following oral vowel,
so you can get words like me 'spring', with m followed by oral e, or ne
'lake' (not common in Omaha, but known in Ponca) 'lake', again with oral
e, or nu 'male, man' (and it's homophone nu 'potato'), with oral u. In
Dakotan these come out (in the cases I remember) ble 'lake', blo(ka)
'male (animal)', blo 'tuber'.  Or consider OP negi 'mother's brother' vs.
Dakotan leks^i.  In all these cases OP has a nasal stop, but an oral
vowel, and the nasal stop originates in what is probably a non-nasal
sonorant of some sort.

(Note that in the clusters *pr, *s^r, and *sr Dakotan and Dhegiha change
*r to *R and sometimes the *p or *s^ is lost, so many cases of n from *R
are actually from *pr or *s^r and their Dakotan cognates will have bl, not
l, or s^l, not l.)

A further complication is that a number of cases of m or n from *W or *R
in OP have a or i following them, and in these cases you get, in principal
a contrast between ma and maN or na and naN or mi and miN or ni and niN.
The contrast of ma and maN is pretty clear, but the others are less clear
and it may be that they have merged so that na and naN are both naN, mi
and miN both miN, and ni and niN both niN.  Or maybe both na and naN have
[aN] but with different degrees or kinds of nasalization, e.g., the former
perhaps doesn't produce intrusive n or m after the vowel before a stop.

This is all an area of OP phonology that hasn't been worked out yet.

Two easy to work with cases of ma and na are the ma instrumental 'by
cutting', and the na instrumental 'by heat', as in ma'=se and na'=kkade.
There is also mu 'by shooting' as in mu'=se.  In all three cases the
instrumental is an outer instrumental that precedes the pronominals:
ma'ase, ma'dhase, ma'se, etc.

One case of *W is *Wi 'luminary (sun, moon)', which seems to come out miN
throughout Dhegiha, and is irregularly niN (niNaNba) in OP.  But in Dakota
it's wi and in Chiwere it's bi.  Mi isn't too surprising in OP, but you'd
expect pi in Osage.  You get miN.  Go figure.

This is the background, anyway.

> Some examples of oN transcribed as uN before a back consonant are:
> iNgdhoNga 'cat' as iNgdthuNga
> ichoNga   'mouse' as ichuNga  (the first 'i' may also be nasal, it seems
> to vary by speaker a bit)

These two are weird cases, probably related, possibly loans into
Proto-(Mississippi Valley) Siouan, but the cognates in other languages
have uN, e.g., igmu(N) 'cat' in Dakotan.  So these could easily be real oN
(or uN, if you want to write it that way).

How about (if you can get it) iNgdhaN as 'eldest son'?  This is pretty
obsolete.  It's kuNnuN in Winnebago.  The matching term for 'eldest
daugher' is winaN, cf. Dakotan winuNna (source of the name Winona).

What about iNc^haN ~ iNthaN 'now'?

Both aN 'do, use' and it's derivative egaN have historical uN.

Note that all the forms above with initial iN might have variants in i.

> Please do send me a list of words to elicit on aN vs. oN.  I'd be happy to
> check as they come up.  I was doing this last year but inconsistently
> which leads to little conclusion.

I'll see what I can come up with beyond the above.

JEK