Helmbrecht Paper - Status of wa-

[R. Rankin]

I've been reading with interest the postings on WI wa-
since my earlier note, but I was in Oklahoma and didn't
have time to reply.  Probably most of what can be said
has been said, so I only have a few points.

I don't see a difference between treating wa- in terms
of voice (transitivity) or aktionsart, which is simply
a name for semantic aspect, unless one or the other
clarifies 100% of the cases.  And I don't see ANY
single category doing that successfully.

> I dared to produce and to repeat (Lipkind) some
speculations on the origin of wa- ....

Wherever wa- comes from, it has been there for several
millennia.  It is found with a similar, if not
identical, function (more likely, functionS) in every
Siouan language.  That, plus the fact that WI
'something' seems to lack cognates elsewhere makes
grammaticalization essentially unprovable at this
point.  Also, evidence strongly suggests that there
have been several "waves" of wa-'s.  The initial
consonant clusters beginning with a labial that are
frequent in languages like Dakota certainly appear to
go back to *w-C clusters.  These include bl-, mn-, pt-,
ps-, ph-, p$-.  The b/m/p are reflexes of wV- prefixes
that have undergone syncope.  The first person sg.
prefix allomorphs show that these sound changes
affected wa- sequences from more than one source.  So,
if you want to track the grammatical use(s) of *wa-
over a number of centuries, all you have to do is look
at nouns and verbs beginning with these sequences in
the modern languages (those that haven't simplified the
clusters).  Expect to find instances of "layering",
i.e., wa-bl- words. But of course we should also expect
to find that the functions of wa- have changed in the
meantime.

> Each transitive verb in Hocank may take wa-
indicating a third person plural, ... the P NP is
optional....

I found this interesting.  I don't remember any cases
in my Kaw data of the wa- being used with a nominal
object or patient present in the clause.  If this is
possible, then the analysis of wa- as an
intransitivizer is simply wrong, wouldn't you say?

But if there are "other wa-s" that clearly do reduce
valence, then I think we have more than one wa- prefix.
Homophony (or long/short V near-homophony) could easily
lead to "contamination" of one from the other and we
might expect to find some overlap.

I have a bit of a problem with discussion so far.  We
have sort of been treating "indefinite object" and
"valence reducer" as alternative terms for the same
sets of examples.  These are really different concepts,
and it would be nice to find syntactic evidence for one
or the other.  (Then there are the WI definite object
wa-s also!)

I think that even if wa- (or one of the wa-s) reduces
valence, creating intransitive verbs, that we would
translate the resultant expressions into English with a
transitive but indefinite "something".  We have to
avoid translating Siouan into English and then
analyzing the English!  Mary Haas used to warn about
this.  My earlier attempt to make that point failed, I
think.  So to me, the fact that many of us find that a
translation of wa- as 'something' or 'stuff' generally
works, is irrelevant or even misleading.  'I ate
something' and 'I ate stuff' are still transitive
sentences.  To speakers, they may be totally
intransitive.  How do we know which?

>There are only a few cases - as far as I found out -
where wa- is indeed used with transitive verbs as a
true intransitivizer.

This confirms my suspicion that wa- (if there is only
one wa-) is derivational, not inflectional.

> The next piece in the mosaic of wa- is the fact that
wa- productively derives nouns (or other verbs) from
intransitive verbs only to a very limited degree.

More evidence for its derivational status.

>These are some of the facts about wa- in Hocank. I do
not really know what to do with them.

Johannes has gone a lot farther toward explaining the
facts than most of us!  This has been a really
interesting discussion and the problems have analogs in
virtually every Siouan language.

Bob