Dhegiha Plurals & the microfilms.

[Koontz John E]

This is certainly an interesting hypothesis, though ideally we'd have to
understand the introduction of i into -ire (cf. IO -ne - pronounced with
enye before i).  I'm sorry, I think the -hne in my last letter was
spurious.  It's just -ne.  I can't seem to locate Whitman at the moment,
but I believe that's correct.

Apart from that we'd have to assume that this source of i was generalized
from the third person to the inclusive and second persons from the third,
which is where this marker occurs in the Siouan languages that have it.

On Sun, 15 Jun 2003, Rory M Larson wrote:
> Alright, then how would this work as a phonological
> sequence for =ire, supposing it were present in MVS,
> and supposing it were to make it down intact into
> the Dhegihan languages?
>
>                       MVS: =ire
...
>       |                   |           |           |
> HC: =ire            OP: =idhe   Os. =idhe   Ks. =iye

Hc (not Ho?) being Hochank, I assume.

> I'm assuming that OP at least went through an
> intermediate stage of [y] between [r] and [dh].
> I'm basing that assumption on the fact that
> i- verbs take an epenthetic [dh] between the
> i- and a- morphemes in the I-form:

Actually, it seems likely that the -y- stage was in Proto-Mississippi
Valley, since all of the MV languages have reflexes of *r for epenthetic
situations.

PMV (PS)   Te   OP   Os   Ks   IO      Wi

*y         c^h  z^   z^   z^   r ~ y*  r

*r         y*   dh*  dh*  y*   r*      r*

*Note:  Rather complicated mergers of *r with *R in clusters,
        which sometimes look like allophony in r reflex, cf.
        Osage orthographic dh but br vs. OP dh and bdh.
        Whitman (not knowing the historical phonology at all)
        argues that y in IO is old *z^, if I recall.

*ira       iya-    idha-   idha-    iya-    ira-    (h)ira
*iro       iyo-    udhu-   odho-    oyo-    ora-    (h)iro- (roo-?)
*ra        ya-     dha-    dha-     ya-     ra-     ra-
*hirE      =yA     =dhE    =dhE     =yE     =hi     =hi
DAT *hire  =khiyA  =khidhe =ks^idhe =khidhe ?       =gigi

These are (1) the combination of i and a locatives, (2) the combination of
i and o locatives, (3) the regular second person active pronominal, (4)
the simple causative, and (5) the dative of the causative.  I'm not sure
why the pronominal's initial behaves as if it were epenthetic, since it
probably isn't.  Note that the pronominals precede the Dakotan and Dhegiha
causative, but follow the IO and Wi causitive, so the formula was
originally something like *=hi=PRO-(r)a.  Hidatsa and Biloxi have evidence
of the same pattern.  The datives involve considerable innovation in
morphosyntax and are often not really datives.  Wi g < *k-h is regular,
but probably the underlying form there is something like *ki-k-hi.  Osage
ks^ is orthographic practice for /kh/ [kx ~ ks^], with ks^ before i and e
and u.

There are other contexts for the epenthesis of *r in Dhegiha in
particular, e.g., the inflection of regulars with the *i locative, and
generally regularly between *i and the first persons and inclusives of the
form a and aN (or oN).  Also idhadi 'his father' (but not dhiadi 'your
father'), to cite OP forms.

> Since the shift from [i] to any other vowel can
> optionally be interpreted as [y], i-ya and i-a
> sound the same, and both can be understood as
> i-ya.

There are two complicating factors here.  One is that the rhoticization of
*y looks like it dates to Proto-Siouan and was complete before Dhegiha
existed as a separate branch, though, of course, it may have been a
dialect cluster in Proto-Mississippi Valley within Proto-Siouan (if we
assume no substantial displacements of components of PMV and PS, as
displacements (migration) tend to mess up dialect structures - sort of
like trying to move your house of cards to another table).

The other factor is that once it is established in a language that r (or
n) is the appropriate segment to separate some kinds of vowels from other
kinds, and especially if those language lack a contrastive y (and maybe
tend to give r near i a bit of palatalization), then r (or n) becomes the
logical thing to insert between i and other vowels.  Quite a few Native
American languages use epenthetic r (or n) to separate one vowel from
another.

I guess a third factor here is that we might want to consider "the
epenthetic sonorant" whatever it is (y, r, n, dh, ...) as being
effectively a separate segment from whatever it happens to sound like at
the moment.  For example, I have the distinct impression that speakers of
Omaha know which dh's are real and which are epenthetic, at least to the
extent of knowing that the latter are far more elidable in fast speech, so
that you get variants like Is^tinikhe ~ Is^tidhiNkhe or maNdhiNkka ~
maNiNkka or dhiNkhe ~ iNkhe rather more frequently than similar losses
with initial dh of dh-stems, and so on.

I realize that this is relatively more - OK, heretically more -
morphologized than we ideally like phonology to be and that I may have to
do penance for this in the afterworld.

Of course, and dh is fair game in fast speech and I spent a good deal of
time agonizing over the sentence introducing particle ege before I finally
concluded it was just egidhe 'finally, as expected, naturally' said even
faster than usual.  I think that's ultimately from 'to say' - a sort of
more regular version of the dative, vs., well, ege (e=gi-e), though I
don't think this latter ege accounts for the former one.  I suppose
English analogs might be "quotha" or "says'e" or "'n('en) he goes."  Just
a guess.