Omaha fricative set

Tue Sep 26 03:24:25 UTC 2006

Just to chime in for a brief second --

  > Note that a fair number of American languages west of Siouan distinguish
two s-sounds, one more alveolar and one more dental. >

  Not sure how far west you were thinking, but I can say that Rumsen Ohlone (Penutian) apparently has a 3-way 's' distinction recorded by JP Harrington.  Without taking time to go into too much detail, there's apparently the English-style alveolar 's', a retroflex 's' (which also exists in some Mayan languages), and the 'esh.'  Mutsun, Rumsen's close cousin to the east, however, apparently only distinguished two (no retroflex).

  > In Castillan Spanish and in Portuguese s is quite retracted. >

  I believe in some dialects of Castilian the 's' is apical rather than alveolar, approaching more of a hissing 'sh' sound (which often happens in Greek dialects as well) though not quite 'esh'.  I don't believe this is the case in standard Portuguese among speakers I've heard (although of course Portuguese does maintain the 'esh' that Spanish lost). 

  Dave

Koontz John E <John.Koontz at colorado.edu> wrote:
  On Sun, 24 Sep 2006, Rory M Larson wrote:
> In the past, we've roughly assumed a set of three oral fricative locations,
> each of which may be voiced or unvoiced:
>
> s s^ x^
> z z^ g^

These are pretty standard phonological symbols (in NetSiouan form) for
Siouan usage. People do usually use theta and edh for interdentals and
in linguistic usage some people make a point of writing pharyngeal symbols
for the back series in Stoney.

However, as you're discovering, the phonetic reality may stray somewhat
from the norm for the symbol. For example, Teton [I think!] and Winnebago
have fairly definitely uvular values for the "velar" series.

The tendency of the s set to develop into interdentals (or labiodentals in
Ofo) proably says something about the usual pronunciations of the s set.
The laminal pronunciation you mention or anything more apical and less
alveolar are likely.

Note that a fair number of American languages west of Siouan distinguish
two s-sounds, one more alveolar and one more dental. I'm not clear on how
laminality fits into this distinction. English speakers have a good deal
of trouble with these distinctions because the two variants are both
acoustically acceptable in English, and English speakers actually use a
couple of different s sounds more or less at random, though consistently
for particular speakers. Sometimes one variety or another is regarded as
a speech defect or, going in the other direction, becomes a trademark.
Humphrey Bogart is famous for a very retracted s/z pronunciation. (I
believe I cribbed all of this from my memory of an article by Bill Bright
on s-dot.)

In Castillan Spanish and in Portuguese s is quite retracted. Castillian c
and z are interdental or maybe it's really laminal. New World dialects
generally conflate these two into an apical s/z set. Basque also has a
three way contrast of s, s-dot, and s^, and I think Bob Rankin one told me
that a three-way contrast was quite common in Mediaeval or Middle versions
of European langauges, but has widely disappeared since then.

I'm pretty sure that something like an extreme laminality or apicality
explains the LaFlesche use of c-cedilla for s, and failing a convenient
coresponding voiced symbol he used the same for z, too. He prized his
pronunciation of s/z and used symbols to insist on it. Dorsey mentions
that members of the LaFlesche family had what ammounted to a lisp. I
forget his exact wording - the comment is in one of his manuscripts. If
you look at the work of Fletcher and maybe Hamilton you'll notice that
they write th for s in their transcriptions, and they worked extensively
with the people of Francis LaFlesche's village. I suspect that other
villages' pronunciations dominates modern usage and Dorsey's usage is
probably based on what he encountered among the Ponca.

LaFlesche used his Omaha scheme as the basis for his Osage scheme, and so
c-cedilla appears in Osage, which either doesn't have that kind of s, or
has one not in the form LaFlesche's phonetic key leads one to expect.
Very likely it wasn't really an interdental. That just seemed like the
best approximation of what it was.

> We've also been aware that s and s^ have "muted" alternates s. and s^.,
> which occur regularly before n, and sometimes elsewhere as well.

You'll find many of these marked with "turned s" and "turned c" in
Dorsey's printed texts. Copies of these I've made for people often have
that distinction suppressed as subphonemic, of course!

> After some uncertainty, it seemed everyone agreed that Omaha [s] is made
> with the tip of the tongue against the lower front teeth.

Or perhaps more critically, with the lamina or post-apical blade of the
tongue approaching the upper teeth. Sometimes the apex end might touch
the lower teeth, but it might not be the critical part of the gesture.

I looked ahead to crib the word laminal from Bryan, of course, though Bob
Rankin has also described some of this to me. I was probably asleep when
David Rood went through it still earlier ... (my sincere apologies to
everyone over the years for this problem of mine).

> The difference between the Omaha [s] and the English [s] is hard to
> detect by hearing.

Yes - even knowing the stuff above I can't say that I was struck by it, so
I'm plased to see you folsk looking into it so closely. I did try to
determine whether s^/z^ were pronounced with lips rounded or not, but I
don't recall my conclusions. Perhaps I never really got around to
checking! I wondered about this because Bob Rankin had pointed out that
these sounds are not rounded in some Dhegiha dialects.

> Second, the "muted" versions of s and s^ seem to be more widespread than we
> had supposed. According to one of our speakers, we seem to have a minimal
> triplet of words in the s series:
>
> si 'foot' (> s.i 'seed' (> zi 'yellow' (
It's interesting that you should find muted s in *su words. 'Quail' is
another one, I think, and one that Dorsey is very puzzled about writing,
but I'm not sure about the turkey word. We can probably get you a list of
*su/*zu words if you like. I'm not absolutely convinced that this is the
same thing as the muting. But it is probably a reflex within the s of the
*u. Perhaps the s is more rounded?

> The "muted" form seems to be indifferently voiced. Typically the
> voicing for the following vowel or n begins in the middle of the
> fricative, so it starts out unvoiced and shifts to voiced in the middle
> of producing it. More importantly, I think the traditional "voiceless"
> version is marked by a greater forcefulness in pushing the air through
> the gap.

Producing a more breathy, sharp, or bright effect.

I emphatically support your impression that the distinction between
voiceless and voiced fricatives is more one of "brightness" or, as you put
it, "sharpness" vs. mutedness. I think voicing per se tends to be a bit
secondary to muting or non-sharping (less breathy friction?). In Osage I
think that voicing may not enter into the definition of z/z^/g^ (or gh) as
much even as in OP. I don't remember whether fricatives are more muted
before n is Osage. Osage has many fewer sn/s^n sequences, since some of
the ones in OP represent sR/s^R, which come out st/s^t/sc/s^c in Osage.
The sn/s^n sequences shared by both languages come from *sr/s^r / _VN.

The intermediate muted fricatives in OP before n are definitely somewhere
in between the voiceless and voiced ones, and your early voicing (before
the n) explanation makes perfect sense. I tend to put these intermediate
forms with the voiceless or sharp ones, but this is somewhat arbitrary,
since this is a context where voiceless and voiced or sharp and muted do
not contrast.

What about x before dh? I don't think Dorsey pointed to any peculiar
quality here, but logically x should be muted before gh.

> Finally, we come to our ever problematic x^/g^ sounds. These in fact to
> not seem to be alternates in a single series. They are made at
> different articulation points. [x^] is more forward, I think between
> the top of the tongue and the back of the hard palate. [g^] is farther
> back, I believe between the back of the tongue and the velum or tonsils
> or something.

This is wild - I had the same impression, but reversed! I though g^ was
pretty much a velar (ach not ich) fricative, muted, more or less voiced,
but that x was very bright, voiceless, but uvular fricative.

I am in no position to quibble about which of us is right about position.
You may be right. I definitely remember being puzzled about how to say
what was different apart from the sharpness/mutedness. I wonder if the
really critical feature isn't that sharpness vs. mutedness and the
differences in position, or perceived position, whatever they are due to
our trying to hear the distinction in the wrong terms.

> The complete Omaha fricative set, as I'm conceiving it now, is as follows:
>
> alveolo-
> alveolar palatal palatal velar glottal
>
> forced s s^ x^ h
>
> muted s. s^. g^
>
> voiced z z^

I think this is phonetically correct, barring my uncertainty about the
actual position of x (you write x^). I don't think that the three way
mutedness distinction is necessary in writing. The middle row can be
merged with the top or bottom row as long as people know how to pronounce
particular tokens. The traditional solution would be to write the muted
forms with the voiceless symbols. In the same way we don't make a point
of marking aspiration on ptk in English so we can leave it off in sp st
sk.

> Looking at it this way, the g^ should probably be replaced by another
> symbol, say [x.].

I strongly recommend against this for two reasons:

- g^ or gh (gamma) is the usual symbol for the phonemic element

- the gh spelling (or something approximating gamma) was agreed upon
independently by both the Ponca and Omahas spelling projects and there is
nothing to be gained by flipflopping on this now

I'd say, stick with x and gamma, and write gamma the way it has been.
Just write better pronunciation guides.

A third issue might be that pronunciation is not likely to be so uniform,
as your relatively small sample suggests, and is even less likely to have
been so the past. Moreover, it has undoubtedly changed over time. Dorsey
may nor have been hearing exactly what we hear.

Have you looked at the famous initial gh/medial x words, e.g.,

ghage' 'to cry'

or gaghe vs. gaxe?
bighoN vs. bixoN?

Do you hear waxe 'whiteman' as waxe or waghe? How about 'ice'? Nughe or
nuxe?

---------------------------------
 All-new Yahoo! Mail - Fire up a more powerful email and get things done faster.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/siouan/attachments/20060925/d4b90825/attachment.htm>