Inflecting 'to paint'

[R. Rankin]

I'd have to say at the outset that analogy
probably IS the mechanism that has produced these
interesting verb forms.  But, that said, let me
repeat what I think I've said before on the list
(and if I haven't, I've said it elsewhere) about
/?uN/.  The comparative evidence shows very
clearly that this verb means 'do' (and in some
contexts 'be', or maybe they're homophones).  This
is true in every language that has it except
Dakotan.  'Do' is routinely conjugated /m-uN/ in
the 1st person with irregular forms in the other
persons as well.

Then the instrument prefix, /i-/ logically makes
the verb */i-?uN/ mean 'to do with', in other
words 'to use'.  In an ideal linguistic universe
it too would be irregular in exactly the same ways
as 'to do, be'.  In some languages this
conservative situation holds.  But Dakotan (alone)
has really "screwed the kitty" on this pair of
verbs.  It seems to have lost the instrumental
/i-/ and made the difference between 'do' and
'use' a matter of pronominal choice, so we have,
in the 1st person, wa?uN versus m-uN.  Loss of
/i-/ is inexplicable in Dakotan, but spread of the
regular pronominal set is analogical. Other
Mississippi Valley Siouan languages have kept the
conservative forms and not lost instrumental /i-/

Dakotan seems to have REapplied the instrumental
/i-/ to derive an entirely new verb with the
narrowed meaning 'to paint'.  But in Kaw the
verb(s) 'to paint' seem all to be reflexive
/íkkik?oN, ikkinoN/ 'paint'.  Clearly, they are
formed on the basis of */?oN/ 'do', and, equally
clearly, one verb above has the expected glottal
stop while the other has an epenthetic /-n-/, the
expected correspondent to the Dakotan /-y-/.  So
Dhegiha has this same problem.  Can any of the
Dakotan /-i-/ phonemes be considered a variant of
'reflexive'?  If so, Dakotan would match the
Dhegiha forms.  Beyond that, I haven't had a
chance to try to figure out what's what.

Bob

> In the course of our discussions of the wa- that
Regina has just written
> about, I noticed some strange behavior in the
morphology of this verb 'to
> paint, to spread on, to annoint' and wondered
whether other Siouanists
> have either parallel examples or some more
insightful explanation of the
> facts than "analogy".
> The verb may or may not have the initial w-, as
Regina points out.
> Otherwise, there seem to be two pronunciations:
i'uN and iyuN.  Buechel
> records only the variant with -y-, but Regina's
speaker alternates freely.
> At this point, we seem to have a possible
epenthetic /y/
> optionally replacing a glottal stop, or a
glottal stop replacing a /y/ --
> something that doesn't bother me much, though I
can't think of any other
> places where that happens.
> The problem is that the first person form is
iwayuN or iwa'uN,
> plural uNkiyuNpi according to Buechel, but when
the wiyuN form is used,
> Neva (our speaker) says it has to be wiyuNk'uNpi
or wi'uNk'uNpi.  Now the
> y/' alternation is transferred to BEFORE the
pronominal affix, and the
> root consonant seems to be unambiguously a
glottal stop.
> Questions: is the etymology of this word i
'instrument' plus 'uN
> 'use'?  Evidence for yes: the organic glottal
evidenced by the first
> person plural inflected forms, and the fact that
the first person plural
> goes in front of the verb if there is no w-.
Evidence for no: the first
> person singular prefix is -wa-, not -m-.
> Alternative: the /y/ is organic, and the verb
has nothing to do
> with 'use'. Evidence for: Buechel's consistent
transcription with /y/;
> evidence against: the first person plural forms
of the w- verb, where the
> y/' alternation occurs between different
morphemes, albeit in the same
> phonemic (not phonetic, not necessarily
phonological, but phonemic)
> environment (i_uN).  I think the apparent
uniqueness of the glide
> alternation here is also evidence against this.
Moreover, if the /y/ were
> part of the verb stem, then the correct first
person singular inflection
> (before the nasal vowel) should be either -m-
or -mn- replacing the /y/.
> Alternative: the verb stem is i'uN, but 'uN is
not the 'use' verb.
> There is a verb 'uN that inflects wa'uN,
uNk'uNpi, but it means 'live,
> exist', and I rule it out on semantic grounds.
So this theory would
> imply that there is a third -'uN root, perhaps
attested in only this
> verb (with the instrumental prefix, supposedly).
This accounts
> for everything, I think, except the phonetically
plausible y/'
> alternation.
> Question: what does anyone else think is the
phonological UR of
> this verb stem?  Is there etymological data that
might give us a clue?
>
> David
>
> David S. Rood
> Dept. of Linguistics
> Univ. of Colorado
> 295 UCB
> Boulder, CO 80309-0295
> USA
> rood at colorado.edu
>