animate wa-

[Koontz John E]

On Sat, 3 Jan 2004, REGINA PUSTET wrote:
> This is my impression too, although the semantic range of wa- also
> covers animals, and the question of whether wa- has to do with
> definiteness, or rather with specificity or referentiality, still needs
> detailed investigation, at least to me.

After making my suggestion on the distribution of wa and wic^ha in
"nonspecific object" use in order to ask Regina if this was what she was
suggesting I rediscovered the place where Regina had earlier said as much.
I think that Linda was saying she found things to be essentially similar
in Assiniboine, though there were also problematic examples with wa where
one might expect wic^ha.

I agree that the issue of whether wic^ha ~ wa in these contexts indicates
indefiniteness or nonreferentiality or nonspecificity or object deletion
is one that remains to be resolved, though nonspecificity looks like the
best bet to me.  The answer might depend on the speech community in
question.

> In other words, the general (and still hypothetical) picture is this:
> wicha- is intruding the domain of non-specific object coding as a
> specialized marker for non-specific animate object. At an earlier stage
> in the development of Lakota, non-specific object coding might have been
> "monopolized" by wa-, which means that the semantic range of wa- might
> must have covered both animate and inanimate non-specific objects back
> then. Today the animate usage of wa- is, by and large, obsolete, and
> merely survives in marginal forms such as w-okiye 'to help people'. In
> most contexts, non-specific animate objects are now coded by wicha-. I'd
> fully subscribe to this analysis -- I'd even like to take the
> argumentation one step further.

In effect, animate third person plural objects in transitive paradigms are
a special case of non-specific object in MVS languages - non-specific
objects used to code animate specific plural objects - and these were
evidently among the original wa forms to be replaced by wic^ha-forms in
Dakotan.  Historical dicitonary data also show lexicalized cases of wic^ha
where an animate non-specific object is meant, along with more numerous
cases of wa in the same capacity.  Contemporary Lakota speakers seem to
prefer wic^ha in this capacity in productive verbal uses.  We're not sure
how long that's been going on.  (Obviously, texts might clarify this
within the last 100 years or so.)

Contemporary speakers still have at least some older forms, presumably
lexicalized, in which wa occurs for a non-specific animate reference.
I'll bet there aren't actually many cases where a form listed in Riggs or
Buechel with animate non-specific wa now substitutes wic^ha.
Constantine's query re. wawiyuNg^api 'inquiry' vs.  wiwic^hayuNg^api
'questions' may be an example, though, of course, this is actually listed
in Buechel.  There might be also be cases in which a new form using wic^ha
with one stem replaces an old form with a different stem that uses wa.

> Over the holidays, I took a close look at the wa-section in Buechel, and
> the results are quite interesting. So far, we have mainly discussed uses
> of wa- as a transitive object marker, but obviously, at least in the
> fixed lexical expressions I found in Buechel, wa- often occurs with
> intransitive bases as well in LAkota. Within these lexical expressions
> -- and that's the crucial point -- the frequency of unequivocally
> ANIMATE wa- is higher than I expected. Here are some examples
> (Constantine has just posted a much more complete list):
>
> wa-kaN-ka 'old woman'  < kaN 'old', -ka 'kind of'
> ...

This is more or less in line with wic^ha not occurring as the third person
plural marker with statives, isn't it?  Of course, once wic^ha occurs as
an animate non-specific object it might occur more generally as an animate
non-specific patient.  Forms like wic^ha'xwa 'drowsiness' (maxwa' 'I am
drowsy') (cited by Constantine) seem to be of that nature.

> Maybe there is an alternative, but as far as I can tell at the moment,
> wa- in the above examples has to be analyzed as referring to the subject
> of the intransitive lexical roots in question.  ... if this analysis is
> correct, and we are dealing with established lexical items here, i.e.
> items that have been around for a pretty long time, such fossilized uses
> of wa- with animate reference could point to an earlier stage in the
> development of Lakota wa- in which this element was totally productive
> with animate reference.

I'd agree with this.  Moreover, such wa + stative forms occur in other
Siouan languages as well.  They might be parallel evolution, but I doubt
it.  Examples, OP wasa'be 'black bear' = 'the one that is black', Wi
wakhaN 'snake' = (?) 'the one that is mysterious', etc.  It is true that
these forms aren't especially numerous in Wi, where the general preference
seems to be waz^aN 'something' + stative, e.g., waz^aNzi 'lemon, orange'.
Or the examples in Wi add -ge (analogous to -ka in wakaNka 'old woman'),
e.g., waxoxge 'cowrie shell', waj^uNsge 'crayfish'.  On the other hand, wa
+ stative i still reasonably productive in OP, e.g. waz^i'de 'tomato
ketchup' < z^i'de 'red'.

I notice that wa is often used with nouns as a sort of indefinite
possessor or whole-namer (as in part of whole), e.g., OP wathaNzi
'cornstalk', wathaN 'squash', wahaba 'corn ear', wamide 'seed', wamuske
'bread', waxiNha 'cloth', wahi 'leg', was^iN 'fat', wamiN 'blood',
wanaNghe 'ghost'. (In Wi wic^aNwas 'corn(plant)', wic^aNwaN 'squash', waha
'fur', wahiNsaN 'down feathers', waniN 'meat', wakere 'faeces', waroic^
'intestines'.)  These are like the possessor cases of wic^ha that
Constantine left out for consistancy, though he left some in, I think,
e.g., wic^haatkuku, wic^hahuNkake, wic^hac^hepa, ...  These are analogous
to the well known cases of animal body-part possession compounds (cf. Boas
& Deloria 1941:70, e.g., wic^hachaNte 'human heart', thaphi 'ruminant
liver', etc., which, of course, have their analogs in OP, e.g., ttez^ega
'buffalo thigh', ttenaNde 'buffalo heart', etc.

> In the meantime, I have tried to elicit additional combinations of
> intransitive bases with wa-, both for animate and inanimate contexts,
> but the output is not very encouraging. I.e., intransitive wa- is not
> very productive these days. ...

> As I said above, in my data (the texts I collected, elicitation, plus
> Buechel, so far), animate readings of wa- appear mostly in the context
> of intransitive wa-. If there is something to this tendency of coupling
> animate wa- with intransitive wa-, then we can conclude, in keeping with
> John's and Bob's view, that animate wa- is of the same age as
> intransitive wa-, i.e. something that has been fragmentarily inherited
> from an earlier stage in the history of Lakota and survives exclusively
> (?) in fossilized forms.

Actually, that seems like a reasonably good argument, though I'm not sure
if it's true that the majority of animate non-specific wa's occur with
stative stems, based on the examples Constantine offered.  It may well be
that the majority of instances of derivations in wa from stative stems are
animate.  In any event, the number of stems doesn't seem to be important
to the argument, but only that this relict set contains such examples.