Non-wa Nominalizations

[Koontz John E]

On Tue, 20 Jan 2004 lcumberl at indiana.edu wrote:
> I agree that the terminology is confusing, and the semantics of the individual
> prefixes make it more confusing.  I see the difference between locatives and
> instrumentals as distributional - the locatives can co-occur with each other,
> but the instrumentals are limited to one per stem, and the locatives can occur
> with the instrumentals.

My morphosyntactic definition of locative is something like

INCLUSIVE > LOCATIVE > {first and second person forms}

This really works out to be just the three "one vowel" prefixes in most
cases, though, in OP, the inclusive also precedes maN in maN...dhiN 'to
walk' and maN...naN 'to steal', so, in principle, these two cases of maN
(probably not the same maN, etymologically) could be considered locatives
(or at least movable preverbs).

The idea with movable preverb is that a preverb is something that
precedes the pronominals, while a movable preverb sometimes "moves" to
after a pronoun, specifically the inclusives.  But the logic of both
movable and preverb as terms have failings, and on the whole I don't think
the movable vs. fixed thing works well.

Instrumentals are a bit slipperier as a morphosyntactic class.  For one
thing, they are really two classses - inner ones (that follow pronominals)
and outer ones (that precede them).  The inner instrumentals are basically
anything that comes between the root and the dative/suus/reflexive sort of
prefix.  But in practice there many other things that precede pronominals,
and, on the whole, it is best to consider instrumentals as preverbs (outer
instrumentals) or prefixes (inner instrumentals) that indicate instruments
or causes or manners of actions.

In the same vein one can think of locatives as things that add an
additional object (or an instrument), though I am pretty sure that all of
them actually do either this or pre-empt the principle object status, but
with a directional (or instrumental) sense.

In short, these are categories that any Siouanist can recognize as valid
in a Siouan language, but which are hardly going to appeal to
theoreticians as linguistic primes of some sort.  I think every language
family has such things.  The Algonquian morphological terminology is
certainly rather strange, but works fairly well for Algonquian languages.
One might be able to do better, but so much has been written using the
Bloomfieldian terminology that there is little point in making the painful
shift to "better" terminology.

JEK