Obviate/Proximate and the Omaha verb system

[Koontz John E]

I actually got started loking at Omaha-Ponca trying to figure out what
shu= was and what that =bi= in =bi=ama was.  It would be really
embarassing to have been wrong about either one, but it always pays to
question one's deductions from time to time.  This is all kind of tricky
and maybe I did make a wrong turn somewhere.

On Fri, 3 Aug 2001 rlarson at unlnotes01.unl.edu wrote:
> I wonder if I could get an expansion on the discussion below from John
> or anyone else doing Dhegihan grammar.  I'm a student in Mark's Omaha
> class, and I've been studying the Dorsey texts for much of this past
> year trying to make sense of the language.  Much of it seems clear,
> but the rules and meanings of suffixed -bi or -i, and a- prefixed to
> verbs of motion, are still thwarting me.

Speaking of thwarting, there's nothing in Dorsey's grammar to suggest that
he had a clue what =bi was or what was going on in the third person
singular in general.  The notes in the texts and his files suggest he had
begun to get somewhere on on the two kinds of third person singular by
1890 or so, after 20 years of working with Omaha-Ponca, but was still not
really thinking of bi as an allomorph of i.  He didn't live many more
years after 1890.

For that matter, if I recall correctly, at several points in Boas's
letters to Hahn he encourages her to see if she can't figure out what that
bi is.  I paid less attention to her letters than his, as they were in
German, but I have the impression that an answer was not forthcoming then,
either.

But, to summarize my views, as far as I know, all the Dhegiha languages
distinguish plural from singular using some reflex of the Mississippi
Valley Siouan *=pi plural suffix or enclitic.  It's an enclitic under the
usual MVS rules of repelling stress even when it constitutes the second
syllable of a word. In most of the Dhegiha languages =pi appears primarily
as =pi (or =bi, if there is voicing of unaspirated stops) merged with the
male/female declarative pair a/e as =pa/pe (or =ba/be), with =pi (or =bi)
as an occasional variant when the declarative is missing.  I don't think
there are any dubitative =pi or =bi attested in the other Dhegiha
languages.

In Omaha-Ponca, however, the alternant =bi is found only (a) in personal
names (Ishkada=bi), in songs (as noted by Rory), and before certain other
enclitics in regular text that can follow it, e.g., especially =ama
QUOTATIVE.

As far as other conditioning enclitics, though I haven't worked out the
details, it seems also to occur often before =egaN (I think in the sense
'having', i.e., the conjunct mode marker), sometimes before the
'evidently' evidentials =the/=khe, and sometimes (environment d!) in
quoted material, especially under verbs of thinking.

Otherwise it is reduced to =i, and this, in turn, is usually omitted by
most modern Omaha speakers, except when followed by =the EVIDENTLY, =ga
IMPERATIVEm, and other enclitics.  I gather some folks do keep it finally,
however, and I gather that it is usually retained there by modern Ponca
speakers, since Kathy seems to hear it.

Another context in which =bi is retained is in the negative plural,
somewhat concealed, since =bi=azhi is reduced to =b=azhi regularly.  Here
I mean regularly in the sense of invariably.  I don't see an actual
generalization!  But as to reduction, the -i of =bi was lowered to e in
=bi=ama in the instance I elicited in text, and I suspect this probably
happened in =bi=egaN [=b(e)=egaN'?], too.

I think that Dorsey (and maybe the speakers he worked with, too)
associated the =bi with the =ama as a single morpheme =bi=ama, and that he
took many cases of =bi alone to be reduced versions of =bi=ama, which
explains why he often glosses them 'they said', too.  He doesn't seem to
have associated this bi with the bi in names or songs, though he knew the
one in songs alternated with =i in regular spoken material.

Apart from the use of =bi ~ =i (~ 0) ~ =b... as a plural marker, there is
a pattern in OP of using these markers (all of them, according to the
context) to mark some (in fact, most, but not all) third person singulars.

This parallels a pattern of usage of the animate definite articles as
well.  The animate definite articles come in two sets, the usual subject
set akha and ama (the single, non-moving, and the multiple or moving -
though we're not sure of this), and the usual non-subject (object,
oblique) set dhiNkhe (plural stem dhaNkha), thaN, dhiN, and ma (the
sitting, the standing, the moving, and the whole of).  If the noun third
person singular subject of a verb takes the articles akha or ama, then the
verb will take =bi ~ =i, and vice versa.  If it takes one of the
non-subject articles, then the verb doesn't take =bi ~ =i, and vice versa.
Of course, often enough a third person singular subject won't be a
definite noun, but if it is, this pattern holds.

In short, in just those cases where the verb is marked unexpectedly
plural, the usual subject articles will be used with the subject, and in
just those cases where it is not marked plural, the usual non-subject
articles will be used.  One kind of surprise or another.

I have been calling "proximate" those cases where the third person
singular subject is marked plural (in the verb) and (if definite) takes a
"subject" animate article.  The others I call "obviatives."  This is
because Dorsey or rather his sources describe the latter as situations
where the speaker did not see the event occur, or where the subject did
something at the behest of others.  Rory mentions this, but I believe he
has it backward.  It is the non=bi, non=i cases that are this way.  As it
happens, a very similar characterization is given for so-called obviative
sentences of Kickapoo by native speakers of Kickapoo, making me conclude
that the Omaha-Ponca non=bi/i cases were like these Kickapoo obviatives.
And if the non=bi/i cases were obviatives, then the =bi/i cases were, ipso
facto, proximates.  I hope the Algonquianists aren't too shocked at this.

In fact, subsequently examining cases of non=bi/i subjects in context I
came to the conclusion that I could see a sort of "not in focus, secondary
character"  quality to them in general.  Ardis Eschenberg has looked at
these more extensively since, and concluded that the non=bi/i cases are
associated with subjects that are "off-stage" or not "center-stage" in
narratives, while the =bi/i cases are "center-stage".  The theatrical
metaphor is common in analysis of text these days, and I think it does a
better job of the situation here than my use of proximate/obviative.
Maybe central and peripheral or something like that would work better than
proximate/obviative?  Of course, historically, proximate and obviative
mean something like this - "nearby" and "off the path," but they've come
to be used in a rather specialized way because of their first use in
connection with Algonquian morphology.

Ardis points out that in OP several individuals can be center stage at
once, so this is not very much like a standard Algonquian
proximate/obviative system at all, and it was never really being compared
to the standard "first" obviative, anyway, but only to the special case
semantics of the second obviative in Fox, Kickapoo, and other closely
related languages.

If use of non-subject articles (dhiNKhe, thaN, dhiN, ma) marks "obviative"
of "non-center-stage" references in subjects, then we'd have to say that
all objects are inherently obviative in OP, which is also rather different
from the Algonquian scheme.  There is one oddity here, and that is that
while no object ever seems to take the akha or ama article, there are some
cases of oblique noun phrases that have akha or ama followed by a
postposition.

The category of proximate, or whatever it is, is not restricted to cases
=bi/i as a plural-and/or-proximate marker.  The definite articles
mentioned above happen also to be used as progressive auxiliaries.  In
this capacity they follow the verb, and in this case the verb never takes
a plural or proximate use of =bi ~ =i.  Any plurality is marked by the
article or auxiliary itself.  And proximate/obviative is also marked this
way.

That is, there are some sentences that use non-subject (let's say
proximate) articles with their subjects and these must use proximate
articles as their progressive auxiliary, while other sentences use
non-subject (let's say obviative)  articles with their subject, and these
must use the same non-subject article as their progressive auxiliary.  Of
course, if a subject of a progressive sentences is not a noun or isn't
definite, the article occurs as an auxiliary only.

Incidentally, except for certain special cases, the OP future is marked
progressive, i.e., a definite article follows it.

Dorsey, or rather the people helping him edit the texts he had collected,
noticed some cases of exceptions to the above rules of sentence formation
and characterized them as errors.  They suggested that they could be
corrected by making both the verb and the subject fit either the obviative
or proximate pattern, whichever he wished, not some mixture like he had in
the texts at these points.  He footnotes this situation several tims in
the 1890 and 1891 text collections without attempting to correct the
texts.  I've always been very grateful to him and his consultants for
those notes, as they were the main hint I had that the alternations
mentioned above had something to do with proximate.obviative (or
central/periperhal).

Interestingly, as far as I know, all the exceptions that occur work out to
have obviative marking followed by proximate marking.  Whether the
exception is a verb or a noun depends on which follows which.  I suspdect
that these sentences were initially intended to be obviative, but somehow,
perhaps by dint of repetition as Dorsey wrote them down, changed to
proximate as they went along.  Or maybe it's a kind of speech error that
just tends to occur in OP.  I notice in English that I often end a
sentence somewhat inconsistently with the start.  Sometimes I repeat
things in a corrected form, sometimes I don't.

Examples (made up - I just know I'm going to regret this!):

UmaNhaN=akha PpaNdhiN=dhiNkhe daNba=i 'The Omaha saw the Pawnee'

UmaNhaN=thaN PpaNdhin=dhiNkhe daNbe 'The Omaha (obv) saw the Pawnee'

UmaNhaN=akha PpaNdhiN=dhiNkhe daNbe=akha 'The Omaha is watching the
Pawnee'

UmaNhaN=thaN PpaNdhiN=dhiNkhe daNbe=thaN 'The Omaha (obv) is watching the
Pawnee'

(Somewhere I have some pat examples put together by or for Dorsey, and, of
course, real examples are in the texts.)

So, to sum things up, I see no evidence of two bi markers, only of what
seems to me as a linguist (not a speaker) to be two uses of one bi marker.
One of those uses - marking proximates or centrals - is also done
independently with the proximate (or central) animate articles ama and
akha, and the existence of this separate scheme for marking proximateness
clearly shows that the proximate or central category of nouns is not
associated solely with =bi.

> However, another particle -bi also existed in Omaha, which was perhaps
> an entirely different word, although it fell in a similar position.
> Dorsey has an extensive note describing this, in the story of
> Ishtinike being dropped down a hollow tree, and feigning to be a fat
> raccoon to get some women to let him out (I can't locate the reference
> at the moment.).  This -bi was a dubitive particle which meant: "This
> is my understanding or presumption about the situation I am
> describing, but I am not a witness and hence do not take full
> responsibility for the correctness of what I just said".

I wonder if this was a case with =the (~ =khe ~ dhaN ~ =ge) as an
'evidently' evidential to describe something that must be concluded from
the context, but wasn't actually witnessed?  Anyway, I know the text, and
I'd be interested in the comment.

This story is the part of the Trickster cycle where buzzard, having been
insulted by the Trickster after agreeing to fly him across a stream, drops
him into a hollow tree.  The Trickster gets some women out collecting wood
to cut a hole and let him out by pretending to them to be a fat racoon.
In the Dorsey texts, see pp. 74-78.  (Note that there's a reference to an
Oto version to be published in a projected collection of Chiwere texts.
It is said to have been given by J[oseph] LaFle(s)che.  I don't think I've
ever seen this text in manuscript or otherwise.)

The only note here is one that says that in wedhe t[h]i=bi=ama the bi=ama
refers to the thought of IshtiniNkhe and must not be rendered "it is
said."  The full clause is produced as IshtinNkhe is stuck in the tree and
runs "'Ni'ashiNga we'dhe thi=bi=ama,' edh=egaN=bi=ama".  "'People
seeking-wood they-arrive-QUOTE,' he-thought-QUOTE."  This is certainly
interesting in terms of the semantics of the quotative, but I'm pretty
sure it's not the note Rory has in mind.  From personal experiences of
this sort I'd guess he has the locations of several interesting =bi=ama
notes mixed!

> Now as I understand John's explanation, there is no such thing as
> dubitive -bi, despite Dorsey's note and his consistent glossing of
> singleton -bi as "they say".

I'd have to know the particular note to know what to make of it.  I'm
pretty sure the =bi 'they say' are due to a false association of =bi with
=bi=ama 'they say'.  In that context the =ama is the 'they say'.  Ama just
happens to condition the =bi variant of =i.  There are cases of =ama
without =bi (or =i) before it, because the subjects aren't plural and are
obviative.

> Is the conservatory function of ama phonological in basis?

That is, is there somehting about the phonology of ama that cases bi to
remain bi, and not become i?  I'm not really sure on that one.  At least
two of the situations that preserve b involve following enclitics
of the form (C)VCV'.

> I understand that third person singular and plural have merged.  This
> may be, though my prejudices are against them having merge in favor of
> the plural form.

There are some singulars that lack "plural marking" - the obviatives.
I'd have to say that my prejudices are against the plural replacing the
singular, too, but I can't think of any other way to characterize it.

> My model would predict that one could have pluralizing -i followed by
> dubitive -bi. I have not been able to find a case of this yet, which
> favors the model John presents of only one -bi / -i particle.

I haven't seen any cases of adjacent double plurals or i + bi or bi + i,
either, though there are various cases of more than one plural with
material separating them.  This is common where one of the plurals is in a
negative.

> Now we have the obviate/proximate distinction.  As I understand the
> explanation, a third person verb with -bi / -i is proximate, while
> without one it is obviate.  "Obviate" means something like "off-stage"
> or "out of sight".  How does this work in practical speech?  In
> describing events and situations that are not present to the listener,
> just what does it mean for one actor to be "out of sight" or
> "off-stage" and the other one not to be?

These are all excellent questions, which Ardis might be a bit better
prepared to answer than I am.  It appears that the matter is something of
a judgement call on the part of the speaker, and that we only know the
speaker's mind by what marking pattern they use.  The pattern makes sense
after the fact, but we might sometimes find that we would have guessed
something to be one kind of reference when it was the other.

> This has been long.  Thanks for any thoughts, expansions or
> clarifications anyone can offer!

I hope this has been some help, and I hope it will prove convincing when
compared to the state of things in the texts or with speakers.  If it
doesn't, of course, I'd appreciate counter arguments.  Examples would be
good, if the non-Dhegiha folks will bear with us.

===

Incidentally, I may be a bit out of touch next week and, especially, the
week after, due to travel.

John