Helmbrecht Paper - Status of wa-

[Koontz John E]

On Tue, 9 Jul 2002, Johannes Helmbrecht wrote:
> ... What I know from research within the field of grammaticalization
> is that third person pronouns mostly derive etymologically form
> demonstrative pronouns, but may also occasionally go back to nominal
> sources designating general concepts such as THING, PERSON etc. These
> may be also the source for indefinite pronouns such as something,
> somebody etc.  ...

This is very reasonable and likely, of course, and something of this sort
may well provide the ultimate explanation for wa- 'something' and for wa-
'third person plural'.  It's just that the particular Winnebago form in
question (waz^aN') doesn't work out as an explanation.

> There are only a few cases - as far as I found out - where wa- is indeed
> used with transitive verbs as a true intransitivizer. Such an example is
> the case ruúc 'to eat sth' which happened to be discussed in some of the
> list contributions. Here, the wa- prefix seems to fill the P slot of the
> verb without allowing a free P NP, i.e. there is no anaphoric agreement
> with the NP referring to the food to be eaten. I give the relevant
> examples in (2a-b).

I haven't really looked at the issue of the productivity of wa- in various
capacities, though I have the impression that wa- as a detransitivizer, as
it were, of transitives is absolutely productive in, say, Omaha-Ponca.
Wa-forms are rather rare in dicitonaries, however, since they tend to
occur only in cases where the wa-form has something special about its
meaning or where the verb is salient (like 'eat') but the difference
between the unmarked and derived forms is obscure in the English gloss
(both waru'c^ and ru'c being essentially 'eat', for example).  Wa-forms
are also found in dictionaries in the case Johannes mentions later, where
the wa- is a fixed part of the stem.

> b) wa-ruc^-náNk-s^aNaN
>     ?-eat-prog(sitting)-DECL
>     He is eating (sitting).

This example supports the contention that the stem is short, since accent
is found on the enclitic positional.

> ... The warúc^case suggests that the P participant is put back
> from syntax (case frame) to the lexical meaning of the word. Other verbs
> such as woorák (< wa+horák) 'to tell stories (etym. 'to tell them' ?)'
> and woohí (< wa+hohí) 'to win (etym. 'to defeat them' ?)' work very
> similar.

Although glosses like 'to tell things' or 'to defeat someone' also work.
Of course the ease with which such translations grade into each other is
part of the hypothesis that wa- 'third person plural object' derives from
wa- 'indefinite object'.  On the other hand, Blair's mention of pa- vs.
wi- in Catawba left me with a sudden started realization that the two
senses might just as easily be equally old and of independent origin.

> 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. E.g. the verb s^is^re 'to be broken, to break (intr)' is
> not possible with wa-.

This is the sort of priceless observation that comes out of fieldwork as
opposed to pondering hapax legomena in sparse records.  I do have the
impression that wa- plus intransitive forms are scarecer in Omaha-Ponca
than wa- + transitive forms, and more generally lexicalized.

> There are not so many cases in the Hocank lexicon (cf. Zep's lexicon)
> where an stativ intransitive verb takes wa- to recieve a nominal
> meaning. An example in this direction is wa-c^ék 'young person,
> virgin' (< wa- + ceék 'new'). Lipkind gives other examples, but they
> are not contained in the lexical sources available for Hocank (which I
> consider a bad sign for the correctness of his examples).

It would be my inclination to assume that the examples were valid, and
came from his fieldwork, even if there were not attested elsewhere.  If
there were a pattern of rejecting his exx. by speakers working with later
workers one might wonder.

> In the vast majority of cases, wa- is restricted to the P slot in the
> case frame of a tr. verb. This result is significant because it
> clarifies to some extent the quite general assumption that wa- is
> simply a valence reducing affix or something of that sort. Generally,
> it does not reduce the valence of an intransitive verb and it is
> certainly not the case that wa- is a productive means to derive nouns
> from intransitive verbs.

I do wonder about the wa-intransitives in Dhegiha.  For example, OP
wa...khega 'be sick'.  We could argue that the wa is just an arbitrary
formant on the verb, but this seems to me to beg the question, especially
when comparison with Osage hu...hega 'be sick' suggests that wa in the OP
form is analogous to the incorporated hu (probably 'limb') in the Osage
form.  It's possible the roots khega and hega are unconnected, but I
suspect khega s some sort of dative of hega, historically speaking.

> Now, the reason why I had to discuss the wa- forms in my paper was the
> claim by Lipkind that wa- is a nominalizing affix. The problem with this
> claim is that many of the proposed nominalization in the lexicon can be
> used also as verbs, some are, however, lexicalized in a nominal meaning,
> ...

I agree that the evidence that wa-forms can be used as verbs dismisses the
notion that wa- is essentially nominalizing.  In fact, I doubt it is even
really nominalizing in those wa-stative forms like wac^e'k 'ypoung person,
virgin', though I haven't a ready analysis of what it actually is, other
than a sort of pseudo-nominal 'something' incorporated into the stem.
Wa- is present with various apparently nominal forms, but I think it is
something of an associative relationship, rather than a causal ones.  In
the same way doctors are often around just before people die, for example.

Again it's very useful to find that some locatives underlying
wa-locative-root formations are considered non-forms by Winnebago
speakers.  It is sort of an open question to what extent things like wa-
or locatives, or datives, etc., are productive in particular Siouan
languages (or generally).

JEK