argument structure k'u etc.

[Rankin, Robert L]

> On page 71 Boas&Deloria actually list wo- ('food') among those
"nominal prefixes" you mention above. This was where I was coming from
but couldn't remember the source until I made a search for it.

I see where Jan's argument originates now, but, like Regina and others,
I still question whether wo- could be a reduced woyute.  Alfred's *wol
or perhaps woyul might be a better bet.  While Deloria's data have to be
considered correct, I don't think her and Boas' _interpretations_ of her
data are always right.

While writing about positional verbs, I ran across the claim in B&D that
the irregular verb forms maNka 'I sat', naNka 'you sat' (from yaNkA 'be
sitting') have pronominal allomorphs that show these verbs have become
stative (or 'neutral'/'passive', whatever).  But this is a mistake.  The
pronominals m- and n- here are not allomorphs of ma- and ni- but
historically regular variants of *w(a)- and *y(a)- respectively.  It's
easy to see how a native speaker could reanalyze these as stative
allomorphs, but historically it's inaccurate.

One of the oldest examples of an incorporated noun must surely be
?uN(k)- '1st person inclusive'.  Given various forms of this in other
Siouan languages like waNk- (Tutelo), wa:Ng- (Hochunk), aN(k)-
(Dhegiha), etc., it almost certainly represents an incorporated form of
proto-Siouan *wa:NkE or *wu:NkE 'man, person' (there are nearly exact
parallels with modern French "on" 'we').  But even here the root-final
-k is preserved contextually in most of the languages.  So I think most
nouns don't undergo all that much phonological truncation when
incorporated.  Mostly they only lose -E or -A and then the root-final
consonant undergoes some mutation.

Bob