Bruce Ingham's "Nominal and Verbal Status ..."

[Koontz John E]

I just got the current issue of IJAL (67.2, April 01) yesterday and
discovered in it Bruce Ingham's new article "Nominal and Verbal Status in
Lakhota: A Lexicographical Study."  In it Bruce argues that Dakota stems
can be divided into five classes:

1. Active Verbal, with active inflection, incapable of being head of
   subject phrase.

   e.g., lel wathi 'I live here'

2. Stative Verbal, with stative inflection, incapable ditto.

   e.g., mawas^te 'I am good'

3. Mixed Active Verbal/Nominal, with active inflection, can be predicated
   with inflection or with he..c^ha (stative), can be head of subject
   phrase.

   e.g., wauNspe=wakhiye 'I teach, am a teacher'
         wauNspe=khiya hemac^ha 'I teach, am a teacher'
         wauNspe=khiye kiN c^haNzeke 'the teacher got angry'

4. Mixed Nominal/Stative verbal, with stative inflection, can be
   predicated with inflection or with he..c^ha (stative), can be head of
   subject phrase.

   e.g., wimac^has^a 'I am a man'
         wic^has^a hemac^ha 'I am a man'
         wic^has^a kiN he Lakhota hec^ha ;that man is a Lakhota'

5. Nominal, uninflected, can be predicated with he..c^ha (stative), can be
   head of subject phrase.

   e.g., he s^uNka hec^ha 'that is a dog'
         s^uNka kiN xlo 'the dog barked'

There is also what amounts to a sixth class, of perhaps several similar
classes, which take =pi as an impersonal marker and are plural predicates,
can be predicated with he..c^ha (stative), and can be heads of subject
phrases.  A classical example is thipi, which Bruce analyses as 'thing
dwelt in by someone', though, interestinglyy, the Dakotan version of this
stem lacks the locative *o- found in, say, Winnebago or Mandan (I think).

   e.g., ho he wic^hilowaNpi hec^ha olowaN
         Ho, that [song is one which] is a praise-song

This classification deliberately omits transitives, which for purposes of
discussion are essentially a sort of combination of 1 and 2.  I'm not sure
if there is an analogous transitive mix of classes 3 and 4.

Members of class three are sometimes inflected for plural with =pi in
nominal contexts, and sometimes not.  Since members of class 1 can serve
without explicit heads as relative or other nominal clauses, it might be
possible to regard the plural-marked forms as cases of this, the stem
serving in these cases (for whatever reasons) as the predicate of a
relative without a separate head.  Obviously I concede the point that this
is nevertheless a sort of behavior intermediate between being a predicate
and being a nominal.

This general approach is somewhat similar to something I've been feeling
driven to with Omaha-Ponca, to wit, identifying verbal vs. nominal stems
based on their morphology, not their syntactic slot.  If they have verb
derivational prefixes and/or inflection, they're verbs, even though they
may typically occur in a nominal slot.  I believe, however, that for OP
this still leaves us with verb stems that predicate and verb stems that
denominate, with two subcases, those that denominate but are still
inflected (?he teaches me = my teacher), and those that substitute
alternative syntax for inflection - the analog of Dakotan internal
inflection vs. he..c^ha.

I think it would be very interesting to apply Bruce's approach in other
Siouan languages.  I certainly can't do it off the top of my head for
Omaha-Ponca, where I've never quite escaped morphology, I'm afraid!

I do know that analogs of type 4, or, at least, things commonly nominals
in syntactic contexts than can be predicated with statives, are very rare.
The only example I can think of immediately is fairly marginal, since it's
a third person plural kkagha=bi 'they're crows' from the usual noun
kkaghe.

Apart from that, one interesting difference between Dakotan and
Omaha-Ponca is that adding an article in OP precludes plural marking in
the predicate.  Since the articles agree with the subject in predications
(rendered continuative by the addition of the article)  and in
subject-headed relatives, they indicate singularity/plurality themselves
to the degree that they can.  In non-subject headed relatives they
indicate the singularity/plurality of the non-subject head.

The equivalent of Dakotan he..c^ha would be formally something like
?s^e=g=..aN in OP, with he 'that ~ s^e 'that' and the =c^ha 'be such'
being equivalent to (but not cognate with) =g=..aN 'be like (something)'.
But I think that e=g=..aN 'be like that' (the form usually encountered)
would in a predication be more or less 'be like, be similar to', not 'be
one such as, be an instance of'.  The OP stem is also inflected actively
(e-gi=maN, e-gi=z^aN, e-g=aN).  The usual verb for predications is dhiN
'be', which is inflected bdhiN, (s^)niN, ?dhiN.  In fact, the third person
is normally omitted, or, rather, replaced with e.

JEK