Adverbs in Lakhota

[Shannon West]

----- Original Message -----
From: ROOD DAVID S <rood at spot.Colorado.EDU>
To: Shannon West <shanwest at uvic.ca>
Cc: <siouan at lists.Colorado.EDU>
Sent: December 8, 1999 1:59 PM
Subject: Re: Adverbs in Lakhota


> DEar Shannon,
> I am a long way from having native competence about complex
> sentences like these, but I'll tell you what I think, and hope that a
> speaker or someone with access to a speaker can give you more details or
> correct my impressions. I am very interested in what you find out, too.
> Where did this sentence come from in the first place?
>

Thank you.  Williamson 1984 pg 214


> On Tue, 7 Dec 1999, Shannon West wrote:
>
> > Hi,
> > I'm a grad student working on Lakhota (and Assiniboine) at the
University of
> > Victoria in BC (not a lot of speakers of it out here), and I'm curious
about
> > something.  Maybe someone can help.
> >
> > owayuz^az^a wan hihani John ophethu ki omakiyake
> > tub         a   yesterday John buy   COMP he.told-me
> > 'He told me that John bought a tub yesterday'
> >
> My first translation of this out of context would have been "a tub told me
> that John bought it yesterday", but that's semantically odd enough to be
> improbable.  For your translation, I would have expected "John owayuz^az^a
> waN hihaNni ophethuN ki omakiyake, and that word order is, as far as I
> know, indeed ambiguous -- John can be subject of either verb or of both.

That is what I would have expected in Assiniboine as well.  The word order
in Assiniboine seems to be more rigidly SOV than Lakhota's.  For example:

s^kos^kobena waNzi hoks^iNna z^e yuta

can only be 'A banana ate the boy', and saying it will get a funny look and
a laugh from the consultant.


> But with "John" in the middle of the sentence, I don't think it's
> ambiguous -- the unit marked with "ki" is self-contained.  In other words,
> the SOV order is rigid and sentence units are self-contained.  The
> first NP in a sentence with two verbs (so it's the first NP in both
> sentences)can be the subject of either one, but once you're "inside" a
> sentence, the NP doesn't construe with a verb outside it except by
> anaphora.

Ahh, so if the NP subject is at the beginning of the sentence it is outside
of the embedded clause.  I.e. this is not correct:  [John a tub this morning
he bought] he-told-me.  If I understand you correctly, this would be better
represented as  John [a tub this morning he bought] he-told-me.  This is why
I asked about extraction.  If hihaNi and owayuz^az^a were moved to the front
of this sentence, it would yield the original sentence.  But that would mean
that the original sentence would have to also have the same reading with
perhaps a different topic / focus distinction.

> > First, could this Lakhota sentence be ambiguous with respect to
'yesterday'?
> > Can it also read "he told me yesterday that John bought a tub"?
> >
> I don't think so; to modify "tell", the hihaNni would have to come
> before that verb, i.e. after the "ki": "tub a John bought ki yesterday
> he-told-me.

Great!  This is what I'm looking for!  If hihaNni had moved, it should have
scope over 'tell' as well.  I think.

> > In this sentence 'he' and 'John' can refer to the same person, a Binding
> > Condition C violation if everything before the final verb is a single
> > embedded constituent.
> I think that's exactly right.  I prefer to think in terms of the
> word order: if "John" occurs early in the sentence, it can be the
> antecedent of a later pronoun, just as it could be if it were in a
> separate sentence.  ("John bought a tub yesterday; he told me that")  I
> think the problem is with the notion of "embedded constituent".  Linearity
> is more important than subordination.

Hmm.  I'll have to think that one through for a while.

>   Is it possible that 'John' is the subject of
> > 'omakiyake' that was extracted out of the clause?  i.e.  it would read
'John
> > yesterday he told me that he bought a tub'.  And if could be, hihani
would
> > also have to move.  The question then is _why_ do these elements move?
And
> > can 2 elements be extracted out of a complement clause in Lakhota?
>
> I have composed this answer in a non-linear fashion, so some of
> what I say in the next paragraph is redundant, even though it was written
> first; please bear with my mental gymnastics.
> At the risk of betraying my ignorance of GB theory, I would
> suggest that you're working backwards: verb last languages do have
> left-to-right interpretation rules even if they don't fit neatly into the
> trees for verb-early languages.  If you start with John at the beginning
> of the sentence, it can be the subject of either verb, i.e. the "ki"
> clause can be embedded between the subject and "tell me" because the "ki"
> clause is the object of "tell".  I don't think it's extracted from the
> "tell" sentence; I think it starts out in the embedded sentence and is
> represented by a pronoun in subsequent mentions, as I suggested above.

Okay, I think this is what I was saying above.  But if in [ [owayuz^az^a waN
hihani John ophethu ki] omakiyake] the ki clause is a CP and John is the
subject of the verb ophethu and also the 3rd person in omakiyake, then
Binding Condition C is violated -- the pronoun refers to an NP it
c-commands.  I know you said that linear order is more important that
subordination, but I still don't see a way around that.

> >
> > I'm more inclined to believe that Binding Condition C doesn't hold in
this
> > language, but I have to be able to give some evidence for this.
> >
> I'm not sure what this means.

Well, the way I see it, this binding condition is a theoretical construct,
not a law.  If the data shows that it Lakhota can have these violations in
grammatical sentences, and there's evidence against the movement of two
elements out of clauses, it makes more sense to say that the condition
doesn't hold for the Lakhota language, rather than trying to manipulate the
data to make it fit.  I.e. in Lakhota, constructions that should be
ungrammatical, aren't.  But in order to say that, I have to be able to
justify it.

Shannon West
University of Victoria
shanwest at uvic.ca