Motion Verbs (Re: OP: coming and going)

[Koontz John E]

On Fri, 19 May 2006, Koontz John E wrote:
> PMVS    *(h)u            *ku              *hi   (vert *hki)

Correction!                                       (vert *khi)

> Da          u             ku                i   (vert khi)
> Ks         hu (?)         gu               hi   (vert khi)
>
> stem        i 'come'      gi 'come back'   hi 'arrive there

You can check this out easily in Allan Taylor's 1976 article on Siouan
motion verbs in IJAL.  A couple of footnotes help with that:

- The u in Ks and Os is u-umlaut, not i-bar.  Allan misunderstood
something Bob told him.

- The proposed *rh set occurs in a fair number of words, not just *rhi 'to
arrive here'.  In fact, I think this is usually regarded as *th these
days, though some cases of it may arrive from *r(V) + h, e.g., 'to arrive
here' < *re (or *ru?) + *hi 'to arrive (there)' (Bob's suggestion).

- Allan barely scratches the surface of the compounding issue.

I've debated the initial of OP 'to come' with Bob Rankin a number of
times.  I've never been able to find a clear case of hi (without a- and
=i) that seems to be come.  There are some i, I think.  However, as far as
I can remember hu appears in Ks and Os and hi in Qu.  That makes the third
person of *(h)u 'come' and *hi 'arrive there' homophonous if the vowel u
shifts to i, as it does in OP.  That may be one reason for not having *hi
in the third person of 'come'.  Siouan languages have a high tolerance for
homophony, but only if it can be cleverly distinguished somehow in the
cases that would be ambiguous.

Bearing in mind that these are "raw" stems and that "motion toward" tends
to become "start toward," in some languages.

      motion toward                        arriving
      here               there             here            there
      regular vert       regular vert      regular vert    regular vert

PMV  *(h)u   *k-u       *rE    *k -rE    *t-hi   *k -ri   *hi     *k-hi
Para    ?-    k-         r-     reg       reg     reg      h-      reg

Te       u    k-u        yA     g -lA      -hi    g -li     i      k-hi

OP    (h)i    g-i       dhE     g-dhE     t-hi    g-dhi    hi      k-hi
Ks      hu    g-u        yE       -lE    c^-hi      -li    hi      k-hi
Os      hu    k-u       dhE       -lE     c-hi      -li    hi      k-hi
Qu      hi    k-i        dE     k- dE     t-hi    k- di    hi      k-hi

IO      hu    g-u        rE     g- rE      -j^i   g-  ri   hi       -gi
Wi      huu   g-uu       rEE    kE-rE      -j^ii  ki- ri   hii      -gii

Ma      hur   k-uh       rEh    kE-rEh    t- i    ki- ri   hi      k- ih

I've notionally treated initial *t in *thi as a prefix, though this is
purely a matter of internal reconstruction.  Only the vertitive suggests
it within MVS, i.e., perhaps *k-ri < *k(i)-r(e)-hi, with reduction of
*krhi to *kri.  Based on the pronouns one would expect *kathi or maybe
just *kithi for this vertitive.  I've written initial dash on forms where
a change in consonantism is now used to reflect a previous prefix (*t- and
*k-).  This is a common development in Siouan.

The motion verbs are a hotbed of "syncopated" or "irregular" inflection.
The stem *(h)u may originally have been inflected A1 *W-u, A2 *y-u as a
glottal stop stem, but appears except in Dakotan hiyu as an h-stem or
regular.  The stem *hi appears as an h-stem (A1 *phi, A2 s^i) or as a
regular.  The stem *ku appears as a k-stem (A1 *hpu, A2 *s^ku) or as a
regular; *re is, of course, an r-stem (A1 *p-rE, A2 *s^re).  The rest
begin with clusters (or aspirates) and are regular (e.g., *wa-thi, A2
*ra-thi).  Or maybe "regular" means "requiring an epenthetic vowel -a- to
break up a three element cluster due to prefixation."

Dakota has regularized a lot, as has Mandan.  IO and Winnebago haven't so
much regularized as been driven to supplementing the irregular paradigms
with regular prefixes because sound changes have erased the syncopating
pronouns.

The final consonants in the Mandan forms reflect a consonant that appears
before certain endings, e.g., -o?s^ 'declarative' and disappears finally
or before other consonants.  So 'I came' would be wahuro?s^.  Taken
together it is hard not to believe that final -h corresponds to stem
initial aspiration elsewhere.  The disappearing finals in Mandan are -r,
-h, and -?, sometimes -?r.  We know that some Mandan dialects had
different endings for some words (a list from Maximilian analyzed by Dick
Carter), so it appears that "final" behavior has been subject to some
degree of analogical restructuring of these transitions.  So, for example,
a -r form may represent something else.