double inflection

[Koontz John E]

On Sat, 2 Aug 2003, Pamela Munro wrote:
> I'd like to know the current feeling of Siouan specialists about double
> inflection for subject in verbs like Lakhtoa Ibláble 'I left' (with two
> -bl- subject markers), which is cited in some theoretical literature on
> double inflection.
>
> ...
>
> Is there a compound analysis of the Siouan verbs?

A good reference here might be Allan Taylor's survey of the Siouan motion
verbs, which is in IJAL.  I think the year is (or was) 1976.  There are a
heck of a lot more compound froms than he lists, at least in individual
languages, but he covers the core of the motion verb system.

I believe that everyone considers these verbs to be compound (in this
case, essentially reduplicated) in terms of a historical explanation.  I
don't know that it would be safe to claim that everyone does this in
synchronic descriptions.  I'm lamentably out of touch with the current
(last 20-30 years?) literature on Dakotan morphology, in spite of Trudi
Patterson and others' attempts to correct that deficiency.  I think
Trudi's dissertation would be one place to look.

In general terms, Mississippi Valley languages do a lot of compounding of
motion verbs and positional verbs, both as main verbs and as aspectual
auxililaries, and in such compounds usually both elements are inflected,
unless a causative is added to the mix and preempts the lower level
inflection.  There are also some non-motion lexical verbs that involve
such compounds.  We've been discussing gaN=dha 'to want', for example, and
there are a number of others in Dhegiha languages.  Diachronically these
are often stable, but they do show some tendency to develop into either
infixing or prefixing verbs with a single inflection.  It's instructive to
collect descriptions of the paradigm of hiyu from different souces, for
example.

Apart from essentially lexicalized compound forms like these there are
also other kinds of double inflection.  Less regular (syncopating)
paradigms are often supplemented with a set of regular pronominals in
front of the irregular ones, e.g., modern OP attaNbe, dhas^taNbe, daNba=i,
aNdaNba=i (A1, A2, A3, A12 of 'see') with a-t-, dha-s^- in first and
second person.  IO and Winnebago do this pervasively in some paradigms.
And everywhere the A1P2 portmanteau is almst always added over the A1 form
of irregular stems, e.g., OP wikkaNbdha 'I desire you' < gaNdha 'to
desire, wish'.

Auxiliaries in general are usually separately inflected from the main
verb, e.g., the Crow and Hidatsa future, or in OP that inflected
positional that follows the future enclitic in many contexts.

The OP negative (all Dhegiha negatives, in fact), has a sort of
pseudo-inflection that seems to be made up of a fusions of an old
auxiliary and the plural marker with the negative enclitic.  Certain
postverbal adverbial enclitics in OP also regularly require an inflected
auxiliary to "support them," e.g., a-t-taNbe=m(aN)=az^i=xti=m-aN 'I really
don't see it' in which a-t- and the two m- are first persons.

Finally, in OP and other languages the dative, possessive, and reflexive
paradigms of syncopating (irregular) stems exhibit a pattern involving
inflection of both the underlying stem and the derivational prefix, e.g.,
for gaghe 'to make', the dative is eppaghe, dhes^kaghe, giagha=i,
iNgagha=i, in which e and dhe are from the regular pronominals a and dha
with contracted gi, and the underlying stem is also inflected, cf. the
non-dative forms ppaghe, s^kaghe, gagha=i, aNgagha=i.  In the possessives
and reflexives the "inside" inflection of the first and second person
alternates with an additional -k- in the third person.  What you get is
(reflexive of gaghe, sense 'make for self') akkippaghe, dhakkis^kaghe,
kkikkagha=i, aNkkikkagha=i.

This is relevant to an idea I've been exploring off and on in Omaha-Ponca,
though it works for other Mississippi Valley Siouan langauges like
Dakotan, of seeing verb forms as consisting (potentially, and actually
pretty frequently in fact) of sequences of lower level forms.  The rules
of inflection and derivation apply to these lower level forms, though
derivational processes often compress two lower level forms into a single
one, e.g., by treating a lower level sequence of preverb and root as a
single root when some prefixes are added.  This scheme seems to keep me
from going crazy trying to explain the rules of pronominalization and
derivation, which is not the case if the compounds and other multi-stem
forms are treated as single chunks.  It also seems to help in predicting
accentuation, at least in OP.

I don't know whether this will help with plurals, as plural markers are
the one thing that Siouan languages seem to feel you need only one of, no
matter how many things are pluralized.

JEK