article

[Koontz John E]

On Tue, 16 Oct 2001, Richard C. Lundy wrote:
> ... My friend referred to the Nakota speaker's sound as not so much
> "N" as such but more "J" sounding. It was reportedly difficult (not
> impossible) to hold a mutually intelligible conversation. He heard
> "Jina waste." rather than the expected "Nina waste." sound. ...

Perhaps the n was somewhat palatalized (more like ny)?  If it was actually
less n-like, and more dy or j (dzh) like, this would not be unexpected on
the assumption that nina (or jina) has a non-nasal vowel in the initial
syllable (cf. lila in Teton).  That is, perhaps for this speaker (or their
community) there is some denasalization of sonorant n before oral vowels.

Of course, one might then expect jida (with the second n also
denasalized), but the diminutive =la of Teton corresponds to =daN (after
an oral vowel) alternating with =na (after a nasal vowel) in Santee, as
far as I know, so the nasality of the diminutive varies across dialects
and we can't assume that nina (or jina)  has an oral vowel in the second
syllable.  Actually, given cases like that we'd have to be careful in
assuming that any vowel in any form was consistently nasal or oral across
the several dialects!

====

Things one might look for further if one were curious and had the
opportunity would be:

What happens with other n that matches l before e or i or iN, e.g.,
in words matching Teton leks^i or le?  Do these also sound like j, or is
it just this one word?

Is there any tendency of n matching other Teton l's before a, o, u, aN,
and uN to appear as something like d?

What about regular t or th or t? before e or i or iN?  I notice that
was^te is not described as was^c^e or was^je.

====

By way of background, it is noted that Dakotan dialects, like other Siouan
languages, contrast oral vowels (aeiou) with nasal ones (aN iN uN), with
uN sounding more like oN to many people.  In addition, most of the
dialects contrast non-nasal and nasal versions of the dental sonorant,
with the non-nasal variant differing across the dialects - some (called
Teton) have l vs. n, while others (called Santee-Sisseton and
Yankton-Yanktonais) have d vs. n.

A number (called Assiniboine and Stoney) have just n, though David Rood
has mentioned that some people report at least some dn for n in Stoney in
contexts I'm not sure were specified.  (I would guess before nasal
vowels or after oral ones?)

In addition, b in y-stem first person inflections and some *pr (or *wr?)
roots like 'lake' cf. Teton ble appears as m in these same dialects
(Assiniboine and Stoney dialects).  And we know that in some limited cases
m appears for b in nasalized y-stems in other dialects, too, e.g., the
famous case of mniN=kte in Teton (not bliN=kte).  These same cases tend to
confirm the assessment that l precedes oral vowels and n precedes nasal
vowels.

To forestall complaints, note that in Stoney dialects mn in the inflection
of y-stems is generally simplified to m, so it's muha 'I have/hold it'
instead of mnuha.

The simple solution is to think something like this:  the distinction of
nasality occurs primarily, or underlyingly, in vowels, and in the Teton,
Santee-Sisseton, and Yankton-Yanktonais dialects nasality spreads forward
from a nasal vowel to nasalize l or d to n and bl or bd to mn.  In
Assiniboine and Stoney the sonorant is already n, however, instead of l or
d, and so the spreading in more or less undetectable.  This doesn't
explain mn (or m), of course.  It also leaves the issue, in fact, the
existence, of things like =daN unexplained.  Why is =daN even possible if
the vowel is nasal?

An alternative that works a bit better is to assume that that nasality is
an underlying property of some vowels, and of all the dental sonorants and
labial-dental sonorant clusters.  In that case, it's orality that spreads
forward from oral vowels in Teton, Santee-Sisseton, and Yankton-Yanktonais
to produce l and bl (or d and bd).  But, in Assiniboine and Stoney it
doesn't spread, and these remain n and mn (or m) in all cases.  Or maybe
it does spread sometimes and you get the occasional d (or j before e, i.
and iN?).  This has the additional advantage of explaining things like the
md (for bd) that Riggs et al. report for Santee.  In that case (i.e., in
the place and time they were working) the orality didn't spread quite far
enough to oralize the m to b.  (I gather that nobody today says mde,
etc.?)

This still doesn't explain =daN, of course.  But if we go a step further
and allow orality to spread outward from the preceding vowel to a
following sonorant n, then that would account for =daN.  And it's
consistent with =na after nasal vowels, too.  But then one would probably
want to claim that in Teton orality spreads so far that the vowel itself
is oralized, producing =la.  The problem is that this won't explain =la
after nasal vowels, where I think it is still =la.

Possibly relevant here is the tendency for a number of other enclitics to
have final nasal vowels in Santee, but not in Teton, e.g., =xti(N), and, I
think =s^i(N).

I haven't addressed another issue, and that is the tendency of n and m to
occur before nasal vowels and y and w to occur before oral ones in cases
where we're pretty sure that l ~ d ~ n or b ~ m aren't historically
involved.  These tendencies also have their notorious exceptions.