Winnebago Inclusive/Exclusive and Minimal/Augmented Pronominals

[Koontz John E]

On Thu, 15 Dec 2005, Dr. Johannes Helmbrecht wrote:
> I would like to add some clarifications with regard to the inclusive/
> exclusive distinction in Hocank (Winnebago) from my experience.

I'm extremely grateful for this Johannes!

> First of all, Lipkind is basically right with his analysis of forms.
> There is a 1A.DU.INCL hiN- meaning 'you and I'. If this form is
> pluralized with the suffix -wi, it is a 1A.PL.INCL.

I want to draw particular attention to your correction of the plural
marker to =wi from my =i.  That was contamination from Omaha-Ponca, where
it is =i (~ =bi ~ =b etc.).

> The 1SG ha- becomes a 1A.PL.EXCL with the plural marker -wi. Actually,
> there is no real dual form except the 'you and me' form hiN-.

I'm thrilled to have this confirmed!  I should state for the record, by
the way, that I first learned of this pattern from a paper by Jean
Charney.

> I would fully agree with John that we have a very similar pattern in
> Hocank to the ones summarized below in terms of minimal/augmented
> pronouns. There is not much to add to Johns presentation of the data
> form Austronesian and Australian languages.

Probably quite a lot, of course!  For one thing, I realized afterward that
in terms of the examples I could discover this time around, only Siouan
distinguishes the minimal and general augmented form - the "plural" - by
adding a morpheme.  In the Australian and Austronesian cases the minimal
and general augmented (plural) forms are distinguished by suppletion.
There the only cases with a separate augment marking are the duals and
trials, etc.

Another thing is that I have a nagging feeling that there are Austronesian
examples that work like the Rembarrnga one, with the "dual augment"
indicating three individuals, etc.  However, I am not sure of this.

> Perhaps, if I remeber correctly, there are also other morphosyntactic
> properties beside the structure of the paradigm which indicate that the
> 1+2 form is treated as singular in Rembarrnga, namely agreement, but I
> might be wrong.

You're probably right - I'm really not very well informed on the details
of this pattern in Rembarrnga or elsewhere in Australian or Austronesian.
There's no similar pattern in Siouan, except trivially, of course, because
the pronominal marking is in the verb only.

I did wonder if the fact that Hochank has a single independent pronoun ne
for first and second person, per Lipkind, might be connected with this
somehow.  I suppose the same pronoun works for the inclusive, too?

> On the other hand,

As opposed to Dakotan?

> Hocank has the same distinction also in the undergoer series of
> pronominal affixes showing exactly the same pattern. The 1U.DU.INCL is
> waNanNgá- (certainly a historically recent addition to the pronominal
> partadigm) which can be pluralized with -wi resulting in a 1U.PL.INCL
> form.

Can you shed any light on the occurrence of -a- with waNaNga'-?  Is it
always there?

When you call this a recent addition, are you refer to its "outside"
position?

> Interestingly, the 1U.SG form hiN- forms a 1U.PL.EXCL with -wi.

I didn't even think to wonder about this!

> (Note that 1U.SG hiN- and 1A.DU.INCL hiN- are homophonous, but they are
> distinct with regard to the morphological slot of the verb where they
> appear).

This is pretty much in accord with matters in Dhegiha, e.g., Omaha-Ponca,
where both are aN, but in different slots as well.  The aN A12 acquires a
trailing g before the a and u (*o) locatives, but not before the i
locative:

aN-g-a-...   < aN + a-
aN-g-u-...   < aN + u-
aN-dh-aN-... < i + aN

I have my reasons for assuming i before aN in OP, apart from the lack of
g.  I don't think there's any need to assume this outside of Dhegiha,
however.

I don't think the trailing -g- occurs in Hochank or Ioway-Otoe, apart from
the P12 waNaNg-a- for in the former.

> So the system looks like the following :
>
> Actor
> 1SG               ha-
> 1PL.EXCL    ha- ... -wi
> 1DU              hiN-
> 1DU.INCL   hiN- ....- wi
>
> Undergoer
> 1SG               hiN-
> 1PL.INCL     hiN- ... -wi
> 1DU               waNaNgá-
> 1PL.INCL      waNaNgá- .... -wi
>
> The interesting thing about the plural marker -wi is that it may
> pluralize all speech act participants, no matter which syntactic/
> semantic function (actor/ undergoer) they have.

Would you call it a plural, stricting speaking, or an augment?

> In addition, -wi (PL)  can also pluralize both speech act participants
> in a transitve clause.  If there is a first person acting on a second
> person, -wi may pluralize the actor, or the undergoer, or both of them.
> Disambiguation is a matter of the context. In addition, -wi alone may
> represent the 3PL.SUBJ, but these instances are rare in our corpus of
> texts. The default 3PL.SBJ marker is the suffix -ire which has only this
> function.

The -ire plural has cognates with the third person in Ioway-Otoe, Mandan,
and Tutelo, so this is apparently a retention from Proto-Siouan.  IO also
alternates it with =wi in the third person.

...

> In transitive clauses, we have a combination of the actor and the
> undergoer pronouns. Reflexive configurations are avoided. Instead, the
> reflexive prefix ki- is used. Johns is right to point out that
> configurations such as I->us or you->you are forbidden. John is also
> right in assuming that INCL -> second person combinations are not
> possible. The same combinations with EXCL are, however, allowed.

I'm particularly glad to have this confirmed, as I hadn't seen it
suggested anywhere else, though it is implicit in Lipkind.

> Let me add a final remark. It is noteworthy that there are Hocank
> speakers who totally gave up the inclusive/ exclusive distinction. They
> continue to use the different forms listed above but with no difference
> in meaning. They are no longer associated with the incl/excl distinction.

I'm not surprised, though I couldn't have imagined the details of the
direction taken.  Carolyn Quintero has already described the way in which
the Osage inclusive form becomes a straight dual, and the Omaha-Ponca
texts seem to have the simple inclusive form only rarely, while other
Siouan languages have lost it entirely, and sometimes the inclusive form
as well.  Regina Pustet has reported the simple inclusive form as
declining in Lakhota.  Both Dakotan and Dhegiha diverge from the pattern
in Hochank, and I suspect that Hochank represents the original pattern
here.

I don't think that the minimal/augmented pattern is actually unstable, or
it wouldn't be so widespread in the huge, widely distributed Austronesian
family, but it might be difficult to maintain if other languages in an
integrated area lack it, or in the face of a superstratum like English
that lacks it.

One postscript - in tracking down the Austronesian data I needed I was
reminded that English once had a dual.  That is, Indo-European has
widespread traces, and two that remain current in English as late as Old
English are the duals of the SAP pronouns, which in the nominative in OE
were:

      s        d         p
1     ic       wit       we
2     dhu      git       ye

Dh, of course, was written with edh or thorn, and g in git is y and c
in ic is c^.  Dual pronouns governed plural verbs.

No trace of an inclusive, per se, in Indo-European generally, as far as I
am aware, though I wonder if -otros 'others' < Latin alteres in Spanish
nostotros and vosotros might be originally an exclusive marker (inverted
in the second person form).  Does anyone know?   How about Engish
we-all/we-uns?  I'm not from the right dialect area to understand the
implications of those.