PMS Correspondences [was Re: Winnebago *p > w (long)]

[Koontz John E]

On Fri, 17 Oct 2003, Rory M Larson wrote:
> MVS      Da       OP        Os        IO        Winn
>
> *p       p        b         p                    w
  *p       p        b         p         w          w

(I'm glossing over OP use of p in Sp clusters, and so on below for t and
k.)

> *t       t        d         t                    j^
  *t       t        d ~ j^    t ~ c     d ~ j^     j^ ~ c^

OP has j^ in diminutives.  Os has c before eiu and maybe some
diminutives.  IO has j^ before ei.  Wi has c^ finally and in CC
clusters.

> *k       k        g         k                    g
  *k       k ~ c^   g         k         g          g ~ k

Da has c^ after i and some e.  Wi has k finally and in CC clusters.

> *hp      ph       pp        hp                   ph
  *hp      ph       pp        hp        p          p

> *ht      th       tt        ht                   th
  *ht      th       tt ~ c^c^ ht ~ hc   t ~ c^     c^

OP, Os, and IO conditioning as for *t.

> *hk      kh       kk        hk                   kh
  *hk      kh ~ c^h kk        hk        k          k

Da conditioning as for *k.

> *ph      ph       ph        ph                   w
  *ph      ph       ph ~ h    ph ~ h    w ~ h      w ~ h

Most of the languages have h in some forms, especially first person of 'to
say'.  OP also in *-phe 'travel' (uhe).  Os ph is ps^/eiu ~ px.

> *th      h        th        th                   j^
  *th      h ~ th   th ~ c^h  th ~ ch   d ~ j^     j^

Da has th in *waNtho 'grizzly', but mostly h.  OP, Os, IO palatal variants
as for *t.  Os th is tx (tends to become kx).

> *kh      kh       kh        kh                   g
  *kh      kh ~ c^h kh        kh        g          g

Da conditioning as for *k.  Os kh is ks^ ~ kx, conditioning as for *ph.

  *p?      p?       p?        p?        p?         p?
  *t?      t?       t? ~ c^?  t? ~ c?   t? ~ c^?   c^?
  *k?      k? ~ c^? ?         k?        k?         k?

Palatalization, again, as above.  I seem to recall that one of the
languages loses p in *p?, but ????

  *s       s        s         s         t^ ~ h     s
  *s^      s^/0     s^/h/0    s^/s      s          s^
  *x       x        x         x         x          x

Here the /... variants occur in the second person of *r-stems and
sometimes a few other forms (as in OP =s^na HABITUAL (or maybe better,
EXCLUSIVELY).  T^ is theta.  IO has conservative variants that match
Winnebago.  The h alternant occurs in some clusters, e.g., sk can be sk,
t^k, or hk.

There is a voiced fricative set that parallels the voiceless one, and is
often considered to derive from it.  (See *y below for IO.)

  *s?      s?       s?        c?        t^?        s?
  *s^?     s^?      s^?       s^?       s?         s^?
  *x?      x?       ?         k?        x?         x?

  *h       h ~ 0    h         h         h          h
  *?       ?        ?         ?         ?          ?

0 is zero, of course!

Da has 0 in at least the motion verbs, h mostly otherwise - nouns and
statives.

The ? set really refers to the ?-stem verbs, and is rather a vexed issue.
You don't find ?-initial nouns or ? in medial or final position (except
sporadically in Mandan).  Usually ? otherwise in a sunchronic sense is a
transition with some vowel sequences and an epenthetic initial before
vowels. The ?-stem verbs usually lack ?, but sometimes have k? or s^?
where k or s^ figure in some languages.  They don't get epenthetic h in
Wi.  For the most part, however, ?-stems behave suspiciously like
*V-initial stems, cf. *?uN leading to OP aN 'do':  A1 m-aN < *w(a)-aN or
*b-?aN or *p-?aN (?), A2 z^aN < *y(a)-aN or *s^-?aN (?).  Wi is the only
language that actually does have s^? in second persons.  Other languages
often borrow the *r-stem form and show n-.

Interestingly enough, the only clear oral ?-verb known is ?o 'to shoot and
wound' (or something like that) in older Wi.  But *(h)u 'come' looks like
one in the second part of older Da wahibu, yahinu, hiyu (A123).  OP has
phi, s^i, i (not *hi)  (A123) even though it keeps h in verbs.

  [unk]    0        0         h         h          h

This unknown initial occurs initially with the *[unk]aNp 'day(light)' set
and at least one demonstrative *[unk]a- INDEFINITE.  The first person
agent behaves just a bit like this, if you consider only OP (a), IO (ha),
Wi (ha).  Os and Da break the pattern, with a and wa.

  *R       l/...    n         t ~ c     r          d
  *W       b/...    m         w         w          w

Recall that *pr (or *wr or *br) often behaves as *R or *pR.  So does *r in
*sr and *s^r clusters, though the pattern do not correlate.

Da l/.. refers to the dialect differences l ~ d ~ n, which get a bit more
complex in conditioning around nasalization.  Note that things like the
diminutive =la are =daN in other dialects:  presence of nasality varies
with the dialect in certain morphemes.  Da b/... seems to be b in Santee,
but w in Teton.  However, a few stems show m, e.g., me.  Not sure what's
up there.  The stem *pi 'moon, sun' appears nasalized (miN) in Dhegiha.

  *r       y        dh        dh        r          r
  *y       c^h      z^        z^        r ~ y      r
  *w       w        w/0       w/0       w/h        w/h

Here the the /... alternates occur in the first person agent.  Winnebago
adds epenthetic h in the #V[short] context generally, but IO does not.

Nobody can explain why IO has z^ in a few forms.  For what it's worth, the
one I remember is 'penis' PMV *ye.  I've suggested that these are either
loans from PDh (though they're very basic words, as can be seen), or that
PCh was on the border between PDh and PWi and some words went one way and
some another.  Note that IO y and Dh z^ are essentially the same thing.
I believe IO has y for z^ from *z^, too, but I'm not sure.

I won't try to go into nasalized variants of *r (n) and *w (m).  This is a
rather vexed area.  Basically, you get n and m with some unexpected oral
reflexes, or n and m in nouns and oral reflexes (at least for *r)  in
verbs.  I summarized this a while back to some extent with *r and *R and
*pr.

One example of a weird nasal set is *wiNh- 'female'.  OP has miN (in
names) and miNga 'female animal'.  It doesn't sound especially nasal, at
least in the latter.  Da has wiN- (compounds) and wiNyaN.  I take the
latter to have the -a noun formant attached, with epenthetic y (from
epenthetic *r, perhaps) and nasal spreading, i.e., *wiN-ra > wiNyaN.
One possible explanation for the failure to nasalize w to m is the -h that
is attested in some of the non-MV languages.  I'd assume something like
*wiNh-ra > *whiN-raN > wiN-yaN, but at least some reanalysis seems to be
needed.

By the way, Bob Rankin has been able to explain the Da -kha, OP -kka,
IO/Wi -ke formant as what happens when you add *-ka to *h-final stems.
So, you have to suppose that nouns like Da maNkha ~ OP maNkka are from
*waNh-ka.  But note that in this case you don't get *waNka in Da.  You get
-ke in IO and Wi because *a > e after a velar.  Of course, you get -s-ke
in *-s-ka nouns, too.  But when you see apparent V-ke, it is from *Vh-ka.

I couldn't seem to fit this in above without interrupting something:  Os
students like to write br (not bdh) for *pr in verb stems and verb
inflection.

I haven't dealt with clusters or vowels to any great extent in this
summary, though they have been mentioned in passing.

> One minor correction (I think):
>
> > *pro-ka 'male' > -dok 'augmentative'
>     [...]
> > With *-a after velars this loss via *-e occurs with
> > clear *-Ca forms like those in *-ka, e.g., the *pro-ka
> > example, which is just nu in OP (no *-ka), but is -dok
> > in Winnebago.
>
> Actually, 'male' in OP is nu'ga, which matches *pro-ka.
> OP nu means 'man'.

I omitted to say that OP has doublets for this form, with and without
*-ka.  Other languages don't.  There's variation on *wiNh- 'female', too,
with Dhegiha having doublets.  I'm not sure that the qualifier reading of
*-ka (applies in Dakotan, too) doesn't arise from its use as a nominalizer
with what amount to relative clauses 'that which is a ...'.  In any event,
the glosses offered are the sort of generic glosses one might deduce for
PMVS from the range of glosses in the individual languages.  The CSD is
more correct or assiduous with glosses.