Winnebago Inclusive/Exclusive and Minimal/Augmented Pronominals

[Koontz John E]

The morphology may be the least interesting thing about Winnebago's
inclusive and "pluralization" of Winnebago pronominals.  We've seen how it
works in Dakotan and in several Dhegiha cases, but here's what I make of
Winnebago, depending mainly on Lipkind.

1) The plural marker =i can co-occur with all four pronominal categories,
including, definitely the first person (h)a, cognate with Da wa, OP a, IO
(h)a, etc.

   12   (h)iN- +/- =i  you+I +/- others  (we-incl. du. vs. pl.)
   1    (h)a-  +/- =i  I     +/- others  (I vs. we-excl. du. or pl.)
   2    ra-    +/- =i  you   +/- others  (you sg. vs. pl.)
   3           +/- =i  s/he  +/- others  (3p sg. vs. pl.)

(I'm ignoring the other third person plural marker!)

2) Most Siouan languages eschew transitive forms in which the inclusive
agent or patient co-occurs with a first person patient or agent.  Thus
there are no we>me or I>us forms.   Winnebago also seems to eschew
combinations of the inclusive with a second person.  Thus there are
(apparently) no we>you or you>us forms.  Anyway, Lipkind didn't seem to
have any examples of them.

I wish Henning were still on the list!

===

The Winnebago pattern with "plural" is a complete version of the sort of
system Bob and Rory have been speaking about, in which the "inclusive" or
"dual" or "inclusive dual" form is one of the primitive "non-plural"
elements of the system, on a par with the first, second, and third persons
singular.

In fact, this is what Dixon and other students of southwest Pacific
languages refer to as a minimal/augmented system.  The minimal terms are
1, 2, 12, 3 or [+speaker -hearer], [-speaker +hearer], [+speaker +hearer]
and [-speaker -hearer], though I'm not sure that the feature analysis is
all that significant an improvement on the numbers.  The augmented terms
indicate that "others" are added to the minimal reference and appear
morphologically as the minimal terms plus a plural enclitic at the end of
the verb.  The plural enclitic is not really a pluralizer per se, but
rather an augment(er), indicating that others are added, not that
multiples of the minimal reference are present.

Outside of Winnebago Mississippi Valley strays from this pattern (or fails
to reach it), by excluding the possibility of pairing the first person
with the augment and throwing that possibility into the scenarios
represented by the inclusive plus the augment.  To the extent that the
inclusive or dual form is eliminated in actual use you get a situation in
which the unaugmented first person and the augmented (or unagumented)
inclusive come to pattern like singular and plural first persons. This
seems to be what has happened in Mandan, where first person singular wa-
opposes first person plural ruN- (nu-), which may be a reflex of *wuNk-.
As I recall there is no plural marker (or augment) in the first person in
Mandan.

Biloxi may take a further step and generalize the inclusive marker to both
first person contexts, though the first person has so many allomorphs it
is hard to be sure they all come from *(w)uNk-.  On the other hand Crow
and Hidatsa seem to lose the inclusive marker and just pluralize the first
person, except with the Crow stative, which seems to use the independent
first person plural pronoun with the third person verb.

In the western Pacific the minimal/augmented pattern is fairly widespread.
It is found in Austronesian, e.g., in the Philippines and New Guinea and
in Australian.  Apparently minimal/augmented systems are rare in the
Papuan language family or families.  It appears that minimaal/augmented
systems have been noticed in Africa, too - not to mention midwestern North
America.

Examples in the Philippines would be Ilocano or Tagalog:

          Ilocano             Tagalog

       min      aug        min        aug
12     ta       tayo       kita       tayo
1      co       mi         ako        kami
2      mo       yo         ikaw       kayo
3      na       da         siya       sila

See http://email.eva.mpg.de/~cysouw/pdf/cysouwPHIL.pdf for a summary of
min/aug behavior in the Phlippines.  I got the Tagalog data from
http://www.copewithcytokines.de/TAGALOG/cope.cgi?002841.

Incidentally, my the Tagalog reference gave this example:  mahal kita
(dear/expense + 12) = 'I love you'.  An implicit reciprocal?

It seems to me that in principle an augment system should indicate the
addition of 1, 2, N, etc., others to the sense of the minimal term, some
of these possibilities being arbitrarily noted as dual, trial, plural,
etc., in descriptions that ignore the pattern.  On the other hand a plural
system should indicate the total number of individuals, 1 (singular), N >
1 (plural), 2 (dual), etc.

Dixon's example of the Rembarrnga dative pronouns follows the augment
scheme quite well:

      min    +1               +N

12    yUkkU  ngakorr-bbarrah  ngakorrU
1     ngUnU  yarr-bbarrah     yarrU
2     kU     nakor-bbarrah    nakorrU
3m    nawU   barr-bbarrah     barrU
3f    ngadU  barr-bbarrah     barrU

(See
http://www.ai.mit.edu/projects/dm/archive/79/Harbour%20Remark%2004.pdf)

I'm using U for barred-u.

Notice that this system uses suppletion, for minimal and non-minimal
terms, and the formant -bbarrah for "only one more."

However, it appears to me that most of the Austronesian approaches to this
sort of thing are actually minimal/plural systems.

Example, Tolai
(see http://amor.rz.hu-berlin.de/~h2816i3x/TokPisinPronouns.pdf)

      sg.        dual    trial   plural
12               dor     datal   dat
1     iau        amir    amital  avet
2     u          amur    amutal  avat
3     ia         dir     dital   diat

In analyzing these, note ura 'two' and utul 'three', so dor is something
like Ilocano ta + (u)r(a), datal is something like that plus (u)tal, and
so on.

It's interesting to see the Tok Pisin take on this, since Tolai is a big
part of the local substrate of Tok Pisin.

     sg.      dual            trial          plural
12            yumitupela      yumitripela    yumi      (*yumipela)
1    mi       mitupela        mitripela      mipela
2    yu       yutupela        yutripela      yupela
3    em       (em)tupela      (em)tripela    ol

As I understand it, yumitupela is speaker + 1 x hearer, while yumi is
speaker, plus 1 x hearer, plus unspecified others.  Presumably the
unspecified others can be either speakers or hearers or even persons out
of the scene.  Multiple speakers are presumably comes down to a question
of solidarity.

I do remember looking at examples of exclusive first person plurals once
in a grammar of Nguna, a Polynesian Outlier language from the
Austronesian.  My recollection is that they weren't so much cases of
people speaking in unison as narrative references reflecting solidarities,
e.g., things like "They said (to someone), 'We (excl., i.e., not you) will
...'"

The distinction between augment and pluralizer is moot if augmentation or
plurality of 1 is indicated by suppletion, and the 12 form indicates
either 1 x 1p + 1x 2p or that plus additional others.  You end up with a
system of the type that is traditionally characterized as singular vs.
plural, with an inclusive vs. exclusive contrast in the first person
plural region.  But if the augmentation is indicated with a separate
morpheme and the basic 12 form can only refer to one speaker + one hearer,
then I think the minimal/augmented analysis looks better.  It seems,
though, that there may be muddy in between systems, though Dakota and
Dhegiha are messy in different ways from Austronesian and particularly
from Tok Pisin.

It appears that inclusive might be the usual term for the basic 12
pronoun.  I'd be OK with dual if we understand that it's not a number, but
a person!