Winnebago Inclusive/Exclusive and Minimal/Augmented Pronominals

[Koontz John E]

On Mon, 19 Dec 2005, Rory M Larson wrote:
> It appears from some of the examples given that languages that use the
> minimal/augmented system often have different augments depending on whether
> the number of additional others is one, (two), or many.

Multiple non-minimal forms are pretty common in Australian and
Austronesian, though whole branches lack them and have only simple
minimal/augmented pairs or singular/plural ones with inclusive/exclusive
first person plurals.  The multiple non-minimal forms also occur in Tok
Pisin, which is closely associated with Austronesian languages that have
multiple non-minimal forms.  But, except for Tok Pisin, all of these
Western Pacific examples seem always to involve a suppletive opposition
between a minimal set and an augmented set of pronouns, and the duals,
trials, etc., involve close compounding of mainly the augmented set with
an additional element.

In Rembarrnga - the only Australian example I looked at - this additional
element works like a unit augment, and compounded with the augmented
inclusive form it forms a trial, not a dual.  There is no inclusive dual,
unless you count the minimal inclusive form "you & me only" as a dual.

In Austronesian (and Tok Pisin), the compounded element forming duals and
trials is always a numeral, and all the examples of dual and trial forms I
was able to lay my hands on followed a pattern of pluralization rather
than augmentation.  In this approach, the minimal inclusive form for "you
& me only" is always marked as a dual by compounding with 'two', and the
unmarked augmented form is the "you & me & others" form.  The augmented
inclusive form may combine with 'three' to form a trial form.  We might
call this pattern minimal/augmented/numerated.  The numeration is slightly
out of alignment with the autmentation.

Tok Pisin has something very like this minimal/augmented/numerated
pattern, except there is no consistent suppletion in forming augmented
forms.  The minimal inclusive yumi is marked dual with tupela as in
Austronesian cases.  The trial form yumitripela is 'you, me, and someone
else' in which yumi plays the role of the augmented form.  The unmarked
form yumi plays the role of the unmarked augmented form 'you, me, and an
unspecified number of others', maybe 'you, me, and more than one other' -
it wasn't clear.  The first, second, and third persons behave simply, with
tupela and tripela adding one and two additional others.  These forms take
pela alone to form the general augmented form, except in the third person,
where the suppletive form ol is found.  So here the minimal and augmented
series are:

       min            aug
12     yumi(tupela)   yumi        (no yumipela asserted)
1      mi             mipela
2      yu             yupela
3      em             ol          (no empela?)

The Austronesian and Tok Pisin compounds with numerals are often rather
irregularly reduced.  (I didn't include the Tok Pisin examples of this.)

> Hocank (Winnebago) apparently does not make this distinction
> semantically, but it does have two different augments, =wi (< MVS *pi),
> and =ire.  John points out that =ire has cognates in IO, Mandan, and
> Tutelo, and that both augments occur in the third person in both IO and
> Hocank: ... [So,] I would suggest that in the proto-language, one of
> these augments was a +one augment, and the other a +plural augment.

The Siouan pattern is somewhat different from the Pacific one.  It has
only a minimal set of pronouns, and forms the augmemted set by adding a
"plural" or "augment" marker - not to the pronoun per se, but to the verb.
Logically this might be something like "some" or "many" or "more," though
I tend to think not.

If the Siouan augments distinguished +1, +2 and +many, etc., they might be
expected to correspond to numerals, roughly, even though in the Western
Pacific numerals occur with plural marking rather than with augment
marking.  On that basis, perhaps *pi and its various correspondents can be
compared to *wiN(r)- 'one', though details of the correspondences are
irregular and unusual.  Admittedly the *pi plural marker set is pretty
irregular already, and irregular reductions in enclitic positions are
plausible, cf. the patterns with numerals in Austronesian.

But the *h(i)re plural is restricted to the third person, and I don't see
any evidence of multiple numbers in extant Siouan augment or plural
patterns, even though there are actually quite a number of different
plural markers attested across Siouan.

For myself I'm inclined to see *=pi as some kind of focus marker, from the
various "non pluralizing" uses it encounters.  I've suggested that on the
list.