Lakota chaNnuNpa

[Koontz John E]

On Mon, 26 Jun 2006, REGINA PUSTET wrote:
> RP> echa-m-uN 'I do' is echuN, and echamiN 'I think" is echiN.

> JEK> In Omaha-Ponca the latter is a'=zaN=miN 'I think', a'=z^a=z^iN 'you
> think', vs. a'=z^=iN 'he thinks'. The a'=z^a= matches e=c^ha= well enough
> to suggest *DEM=ya=.

>   I'm really having trouble with the analysis of echiN 'to think'. 1SG
> is echaNmi, 2SG is echaNni. I can't see any regularity in there, even if
> we say that there is a small class of verbs that (in an internally
> regular way) prefix or infix m- for 1SG and n- for 2SG. For echuN 'to
> do', we said in previous email that the original (Siouan) root is uN 'to
> do', which must have formed a compound with echa-, which could be the
> adverb 'like that', at least from my completely synchronic perspective.
> Under these assumptions, the inflection of echuN (1SG echamuN, 2SG
> echanuN) conforms with the reqúirements for regularity within the
> m/n-paradigm. But what about echiN 'to do'? is there a basic historical
> form iN 'to think' or something like that?

Exactly.  Your perception of ec^ha as 'like that' is reasonable and seems
to work in both cases.  OP az^a uses the OP a= (Os ha=)
"indefinite-interrogative demonstrative", equivalent to Dakotan tV=. So it
is 'like something unspecified' instead of 'like the aforesaid' (or
whatever reading e= should have here).  (Note that Dhegiha has z^- in the
second person of ?-verbs, where Dakotan has n-.  I belive Dakotan has
borrowed the second person in n- from the r-stems.)

> Is the echaN- that we see in the 1SG and 2SG of echiN related to the
> echa- in 1SG and 2SG of echuN?

So the verbs 'to do' and 'to think' are formulated 'to do thus' and 'to
think thus', but the demonstrative used varies.  This isn't without
precedent.  The verbs of speaking have an embedded demonstrative e= that
can alternate with others, including tV=.

OP 'to do for one' is inflected A1 egimaN, A2 egiz^aN, A3 egaN.  The A3
form is used in the sense 'thus', 'like that', 'sort of' in numerous
contexts in OP grammar.   The root aN (cf. Da uN) is the same.  The
enclitic "gaN" 'thus' turns out to be =g(i)-aN, the dative of aN.  There's
no productive reflex of PMV *=yaN (Da =c^haN, OP =z^aN).

The verb 'to say to one' is inflected A1 egiphe, A2 egis^e, A3 ege
(e-g(i)-e), incidentally, so this pattern of DEM=DATIVE=PRO-ROOT occurs
there, too.

> The Omaha-Ponca data now seem to suggest that the case is a lot more
> complicated.

I was hoping they would simplify things, so I've failed in my explanation,
or I'm missing some implication you're seeing.

> And anyway, I still claim that the reader of a Lakota grammar that
> posits a 'regular' m/n-class togther with all the phonological rules
> needed to derive the actual inflectional forms will end up more confused
> than someone who is simply confronted with a list of verbs that do not
> inflect in a carefree way. 'Carefree inflection' in Lakota is what we
> see in verbs that have wa- 1SG and ya- 2SG.

I personally have no problem calling these verbs irregular or a minor
paradigm in the context of Dakotan or Omaha-Ponca.  However, Dakotan does
not have carefree inflection, because it has y-stems, nasal y-stems,
datives, suus-forms, reflexives, and quite a number of little
irregularities that amount to a system much like that in Omaha-Ponca or
Winnebago.  One more minor paradigm would hardly be noticed.

In fact, all of the minor paradigms apply only to a handful of formants
(instrumentals) and roots in each language.  The difference between the
relatively simple situation in Dakotan and the more complex one in
Omaha-Ponca is largely a matter of perception and descriptive choices.
The main thing that Omaha-Ponca has that Dakotan does not is b-, d-, and
g-stems.  All the OP b-stems have the ba- or bi- instrumental.  There is
one d-stem ('to see'), and there are about 5 g-stems (gaghe cf. kagha,
etc.).  There are actually about as many ?-stems in Dakotan or OP as there
are g-stems in OP.  The odd aspirations and kp's and what not in Dakotan
k-stems and p-stems are all reflexes of the stop-stem paradigm.

My favorite minor paradigm, actually evidenced only in Dakotan, is the
*y-stems (c^h-stems in Dakotan).  The only form is epc^he 'I think' (no
other persons exist for this verb, but alternate demonstratives are
possible).  This correspond nicely with OP A1 ebdhe, A2 e(s^)ne, A3 edhe.
However, the OP verb has to have egaN 'sort of' appended to it, i.e., A1
ebdhegaN, etc.  The historical forms seem to be *e=w-ye, etc.

The phonology of the ?-stems is very regular historically, but I'm not
sure that you get much mileage out of

w(a)-uN => muN

which depends on V1 + V2 => V2 and w / __VN => m.

> Even if there are rules which ultimately make it possible to motivate
> the verb forms with m/n, this class, in terms of regularity, is a far
> cry from the wa/ya-class, which is truly regular.

I think it might be fairer to say that wa/ya is transparent than to say
that it is regular.  The y-stems and ?-stems are regular enough.  They're
just not transparently related to wa/ya.

> Do we do the readers of a grammar a favor if we overstuff it with rules?
> Such rules may be very real at the diachronic level, and are definitely
> interesting to discuss, but to what extent should a grammar be burdened
> with these things?

I don't think that a teaching grammar should explain the derivation of
these forms.  But it should recognize the ?-stems as an active paradigm
parallel to the regular actives in wa- and the y-stems in b-.  (Neglecting
the dative paradigms for the moment, because the moment you introduce them
you need to make a special class of the non-dative ka-instrumentals, too.)

I don't particularly recommend y-stem or ?-stem, etc., for use in Dakotan
syncrhonic grammar.  They're no more harmful that a-stems, e-stems, etc.,
or "First Conjugation", "Second Conjugation", etc., or ar-verbs, er-verbs,
etc., in Romance terminology.  However, any terms that cover the
categories and appeal to the author are fine.