Obviative/Proximate and the Omaha verb system (fwd)

[Koontz John E]

---------- Forwarded message ----------
Date: Wed, 22 Aug 2001 01:14:12 -0600 (MDT)
From: Koontz John E <koontz at spot.colorado.edu>
To: rlarson at unlnotes01.unl.edu
Subject: Re: Obviative/Proximate and the Omaha verb system

Please feel free to post any substantive discussion to the list.  I'll
reply offline in this case.  I think I started the offline thing myself,
though intending it mainly as a temporizing apology for not giving you the
full response your excellent points deserve.

On Tue, 21 Aug 2001 rlarson at unlnotes01.unl.edu wrote:
> 1)  If -bi / -i are semantically equivalent alternates of each
> other, then the difference ought to be made by the phonological
> environment.

It is much nicer if the conditioning is phonological, but cases of
morphological conditioning aren't precluded in linguistic analysis if the
data seem to bear them out.

> We suggested that a subsequent ama' or egaN' might preserve -bi, while
> the morpheme would be reduced to -i in most other environments.  If
> -bi and -i are distinct morphemes, however, then their use should
> depend on the semantics of the situation.

Agreed.

> But egaN', "the preceding having occurred", or "because of the
> preceding", is not intrinsically dubitative, and can be used equally
> well for factual as for doubtful clauses.  Therefore, our two models
> differ in their predictions for the ( -bi | -i ) particle preceding
> egaN'. If the dubitative -bi model is valid, then egaN' should
> normally be preceded by -bi in narrative statements, but by -i or
> nothing in most dialogue statements.  But if the phonological
> environment model is correct, then the particle preceding egaN' should
> always be -bi, regardless of whether the statement is narrative or
> dialogue.
>
> Finding compound third-person dialogue statements is difficult,
> but I found two of them in the story, "How the Rabbit Killed a Giant",
> pages 22 - 25.  The first is on page 23, line 10-11.  When the giant
> demands to know which of them had had the audacity to cut up the
> deer they had shot, the two frightened men admit that the Rabbit
> made them do it:
>
>      She' akha' MashtshiN'ge-iN' akha' pa'de wa'gazhi egaN'
>      aNpa'dai ha -- "That one, the Rabbit, bade us cut it up, so
>      that's why we cut it up".
>
> Here the accusation arises from personal experience, and the men
> do not precede egaN' with -bi.
>
> The second is on page 23, line 17-18.  As the giant proceeds to maul
> him, the Rabbit declares the difference between himself and the
> craven men:
>
>      Dhe'ama naN'dhiphai' egaN' a'dhikhi'dha-bazhi'-hnaN'-i;
>      wi' naN'wipha ma'zhi egaN' a'wikhi'bdha ta' miNkhe. --
>      "These ones fear you, so they don't attack you;
>      I fear you not, so I will attack you".
>
> Here again we have no -bi in front of egaN' in either of the two places
> it appears.  The first one has -i, which can be construed as the plural
> particle.  The second has only the first person negator ma'zhi, but
> can't be counted in this test since its subject is not third person.
>
> In these two cases of dialogue, egaN' takes no preceding -bi.  I count
> six other cases of egaN' in narrative statements of the same story,
> each of which does take a preceding -bi (or -b alone, tacked to the
> final vowel of the preceding word).  These are at lines 2, 4, 7, 15, 17
> and 20, all on page 23.

My counter argument here is that there seem to be two egaN conjunctions.
This is something I'd noticed during my last active progress on my
dissertation a few years ago, though I'm not sure I ever associated it
with bi-conditioning and I haven't written anything up on it that I can
recall.  I think I may have mentioned it to Ardis later, but maybe that
was just multiple kinds of ama.

My recollection is that one of the forms was regularly stressed on a
particular syllable (apparently the second). The other was accented in
various ways depending on the accent of the preceding verb (depending on
where the tendency to alternating syllables for secondary accent placed
it).

As far as glosses, one form was usually glossed something like 'so', or
sometimes maybe 'in order that'.  The other was the 'having' conjunction.

It appears from your examples that the first pattern in each of these is
paired, though I'd have to go look at my notes to be sure this is what I
noticed.  Maybe it's more complex and my initial analysis and this one
don't match because I'm completely off base.

Anyway, the first examples you provide have the sense 'so, in order that',
while the group on p. 23 are the other - more common one, "having," what
Bob Rankin recently called (I forget where, maybe just conference comments
or email) the "conjunct mode" after a term used in Algonquian grammar
(maybe elsewhere, too) to refer to generically subordinated clauses (as
opposed to specific kinds of subordinates like conditionals, causals,
temporals, etc.).  These are somewhat like active present or imperfect
participles in Indo-European languages (note tha "having"), but in
Algonquian they take personal inflection.  There may be an implication of
sequencing (not always in Algonquian, I think), so they are somewhat
temporal, but not in marked way like 'when' clauses.  In OP they're just
regular verbs, of course, followed by a kind of generic conjunction.

Now what you need for a counter example would be a 'so' that had a bi or a
'having' that didn't.  The easy way to find this would be to search the
computer file version of the texts (which I assume you have?) for 'so' and
'having'.

I should add that there are also a number of other uses of egaN.
Basically, it is a verb e=...gaN 'to be thus, to be so to something',
inflected e=gimaN, e=gizhaN, e=gaN.  This neat pattern of gi before the
first or second person pronoun or g fused with the third person stem also
occurs with e=ge (alternatively suppletively e=gidhaN from dhaN 'tell'
instead) 'to say to someone', inflected e=giphe (or is it egihe?),
e=gis^e, e=ge.

The underlying verbs (without -g(i)- 'to someone/something') are eaN 'how;
to be so' and, of course, e 'to say' (which has some many alternaive stems
it's difficult to describe except by listing the inflections ehe, e^e, e ~
a=i ~ a=bi).

Anyway, apart from verbal uses, egaN seems to be used after verbs in the
sense of 'sort of'.  Probably in this sense it seems to be a fixed part of
e=dhe 'to think', which is always e=dh=egaN 'to sorta think'.  It's also
part of a pattern or set of related patterns of the shape "eska(naN) ...
gaN=dh=egaN" = 'would that ..., oh that ...' Sort of like Spanish ojala is
how I conceptualize it, though not a loan from Arabic like that!

> Also, I should note that statements or clauses ending in one of the
> words that you call articles, and that I have been calling dispositionals,
> and that Paula Ferris Einaudi calls classificatory verbs in her Grammar
> of Biloxi, never seem to take -bi, even in the narrative.  I would claim
> that to be a standard exception to the rule of dubitative -bi in
> narrative statements.

In these sentences, which I call progressives, though that's not entirely
apt, for various reasons - maybe imperfects is closer to the mark - both i
and bi are missing, I believe, and number and obviation come from the form
of the article.  Basically, the lack of i ~ bi is equivalent to the use of
the obviative articles (dhiNkhe, etc.), while the presence of i ~ bi is
equivalent to the use of the proximate articles (akha and ama).

I'd be interested in your logic with the term dispositional.

> 2) I think I've found a counter-example to our rule that -i and -bi could
> not both occur at the same time.  In "Ishtinike and the Deserted Children",
> page 87, line4, we have:
>
>      E'gidhe shaN'ge i' khaNthaN'i-biama'. --
>      "It happened that the horses mouths were tied, they say".

I'd take this to be a typo or mispeaking, though I'd have to reconsider if
there were many examples.  I've noticed a few other typos here and there,
surprisingly few, though.  There are also definitely some
mistranscriptions and mispeakings, e.g., the errors noted with proximate
and obviative marking (or dubitative) in the footnotes.

> 3) I think I've also found a case of two different -i morphemes
> appearing at the same time.  On page 88 of the same story,
> line 7-8, we refer to "the children who were abandoned" as
>
>      shiN'gazhiN'ga waaN'dhaii ama'

Same reaction. Of course, the only support I can offer to this is that
Dorsey doesn't comment on these somewhat unusual cases, which is negative
evidence, and weak anyway.  Anyone could overlook something.  And, for
that matter, I think Dorsey assumed with you that bi had a dubitative
sense, though he glosses it as quotative.

> These are the only examples I've found of these two cases,
> both in the same story, which is otherwise rather difficult.
> Perhaps NudaN-axa spoke a somewhat different dialect
> than that of the La Fleches.

I think he was Ponca.

> These arguments are tenuous, but so far the dubitative -bi
> model seems to offer the best fit for me.  I'd be interested
> in any counter-examples from Dorsey you could find
> that would support the -bi / -i equivalence model, or
> specifically the obviative/proximate model.  I certainly
> agree with you that the whole issue is very tricky!

In general, why would songs and names be dubitative?