Helmbrecht Paper - Status of wa-

[Rankin, Robert L]

> ...  But even in English (or better, in Montague grammar) the expression
"eat-something" or "eat-stuff" is intransitive.  You can't eat-stuff an
apple.  A transitive verb plus an object makes an intransitive
expression.

Uh, no. Not in my vocabulary anyway.  "Stuff" is a noun and the object of
'eat', even with a hyphen as I see it.  Yes, I know I'm being a stick in the
mud insisting on analyzing what's actually there, but I think it's best to
assume that the Montague people were simply naive about the ways the world's
languages can work.  In English a verb may typically be either transitive or
intransitive in many cases.  'Eat' and 'shoot' are good examples (Johannes
pointed out that WI ruje means 'eat it', not 'eat').  In English we require
no morphology in order to intransitivize; we just leave an object off and
that's it.  In Siouan, verbs that are inherently transitive, like 'eat',
require a marker if the speaker uses them intransitively.  That marker is
wa-.  Many languages are of this sort, and I have to believe that Montague
grammarians simply were unaware of such superficial things about various
American and African languages.

> In this sense what we've been calling indefinite object and
detransitivizer or valence reducer are in fact the same.

I agree that some of the problem revolves around terminology.  But I feel
that for Siouan speakers it IS one or the other.

Instances of incorporation (like "eat-stuff", as in *"eat-stuff an
apple")are normally pre-posed to their verbs, as in "babysit", and the
perfectly acceptable "I babysat John's little brother."  Or the even more
generic "I babysat John's new Toyota."  I guess I don't feel we're justified
in assuming that an invisible, imaginary superstructure of 'eat' in English
includes 'stuff' without better evidence.

> How do we know whether something is actually intransitive?  See whether it
can take an(other) object.

Yes indeed, I think I mentioned how peculiar it seemed to me that verbs with
intransitivizing wa- could take an overt (and definite yet) object in WI.
If that's the case, then either:

(a) Wa- is NOT a valence reducer or intransitivizer, or,

(b) There is more than one wa-, and we're dealing with homophony, not
polysemy.

Taka you choice.

Bob