Crow Citation Form and MVS Theme Form (Re: "WOUND")

[Koontz John E]

On Sat, 13 May 2006 Rgraczyk at aol.com wrote:
> This is interesting.  John, what are the Mississippi Valley suffixes that
> behave the same as the Crow suffixes that follow the citation form?

Warning - I now proceed to beat this topic to death!

Comparing Crow Citation Forms with Moribund Mississippi Valley
Theme Formation Morphology; Cognates and Functional Similarities
Between Crow Suffixes Conditioning Citation Forms and MVS Enclitics

Conditioning Thematic Forms

Crow citation form discussion after Graczyk draft grammar, 2.5.12.

All Crow stems end in vowels.  There are no nasal vowels at present, and

all possible short and long oral vowels occur as stem finals, as well as

the two diphthongs ia and ua.

The relation of stem form to citation form, Table 6:

Stem Ending : Citation Form Ending : Formation of Citation Form

i           : e (= ee)             : lowering
ii          : iia                  : schwa off-glide

u           : o (= oo)             : lowering
uu          : uua                  : schwa off-glide

e (= ee)    : e ( = ee)            : invariant short (mid)
ee          : ee                   : invariant long (mid)

o (= oo)    : o (= oo)             : invariant short (mid)
oo          : oo                   : invariant long (mid)

a           : e ( = ee)            : short ablauting
aa          : ee                   : long ablauting

ia          : ia                   : invariant diphthong
ua          : ua                   : invariant diphthong

The annotation V (= VV) for o and e refers to RG's comment that all mid

vowels are essentially long, though the popular orthography writes some

as short.  Notice that "short e" and "short o" alternate with themselves

(invariant short mid pattern), with short a (short ablauting pattern),

and with short i or u (lowering pattern).

Final e and o utterance-final or in isolation are pronounced [<aesc>]
and [<open o>].  Except for ablauting stems then, one might say that

final i and u lowered to e and o, similarly, in citation forms.

Lowering

e => <ae>
o => <open o>
i => e
u => o
a not lowered, but ablauted

Perhaps short e and o are sometimes just contextual variants of short i

and u, and so truely are short?  Not exactly a pure phonemic analysis,

but students of vowel harmony and Eskimo schwa have often found

similarly awkward neutralizations.  (JEK, not RG, waxing heretical.)

The stem is the base for addition of most suffixes.  The citation form

is the un-suffixed form, and the base for a few suffixes.  The citation

form is also the form used to answer "What's the word for ...?"  The

citation form of nouns is common, but verbs tend to occur with post
stem suffixes.

Suffixes that attach to the citation form are:

-sh definite article (also with names?  JEK) e.g. 'the man'
-m indefinite, non-specific article          e.g. 'a horse'
-n locative                                  e.g. 'in, on'
-taa path                                    e.g. 'along'
-ta 'seem, resemble'

Ignoring the invariant cases, it appears that the behaviors associated

with citation forms are lowering of high short vowels, schwa off-glides

of long high vowels, and ablauting of a and aa to e and ee,

respectively.

Nothing so far said addresses the question of accentual shifts in

connection with citation form or adding of these suffixes.  Randy has

discussed some of this with Bob on the list recently.  So, note

Stem             : Citation Form : Suffixed Citation Form

bili'    'water' : (bilee' ?)    : bile'esh 'the water'
bi'tchii 'knife' : bi'tchiia

short u?
awuu'    'inside': awuua'        : awu'uan 'inside'

e'hche   'know'  : e'hche
bachee'  'man'   : bachee'       : bache'esh 'the man'

awa'     'earth' : (awe' ?)      : awe'en 'on the ground'
long a?

bi'a     'woman' : bi'a          : (bi'ash 'the woman' ?)
bua'     'fish'  : bua'          : (bu'ash 'the fish' ?)

I am not sure the forms listed exhaust the accentual patterns, but it

looks a bit as though the citation forms here were exhibiting V1(V1)' +

V2 + suffix, where the deleted vowel V2 merges with V1(V1)' as low mora

of a simplified V1'V1.  (This is more or less [=la] what Randy and Bob

were discussing.)  So perhaps stems besides the long high final vowel

stems add schwa in their citation forms, but it is absorbed phonetically

by the preceding vowel under some conditions.

Mississippi Valley Cognates of Suffixes to Citation Forms

-sh def art     < *=s^  (cf. Mandan =s)

                Maybe an irregular reduction of *=ki via *=k<raised y>?

                OP animate proximate articles are
                =akha +/- singular and
                =ama ~/- plural.
                The OP obviative proximate article
                =ma collective
                shows that a- is added in =ama.
                Proto-Dhegiha forms would be
                *=a-kha
                *=a-Wa
                *=Wa
                Here we see noun + a + article, but -a- is
                reanalyzed as part of the article.

                Note also the Winnebago article
                =ra ?< *=ya, comparable to the Dakotan postvocalic
                -ya variant of the -a "stem forming" vowel,
                as in wiNyaN 'woman' ~ wiN- (combining form)
                iNyaN 'stone' ~ iN- (combining form)
                heya 'louse' ~ he- (combining form).
                Here we see again noun + a forming the definite or
                independent form.
                Winnebago may generalize *=ya to replace *-a ~ -ya,
                and substitute noun + a for noun + a + article.

-m indef art    Cognate presumably with *wiN-a-, *waN 'a, one'.
                No evidence on thematic form with MVS indefinites.

-n locative     < *-r.

                May be an irregular reduction of PS *=tu, cf., Dakotan
                pattern for *tu.

                For *=tu consider:

                OP =di locative: 'in, at' ('to' with motion verbs)
                Os =ci
                Ks =j^i

                On the whole these suggest
                PDh *=ti (?),

                but I think that

                PDh *=tu is correct, taking Dakotan into account.
                The Os and Ks forms may simply unround in final

       unstressed position?

                Te =tu ~ =l
                Sa =tu ~ =d
                etc.
                suggest
                PDa *=tu ~ *=R, where *R is regular from *t in final
                position.

-taa path       < *taa

                Seems a good match for MVS *=(k)-ta 'to(ward)',
                though it might be *thaN 'from', too!

                OP =tta 'to' ('toward' with motion verbs)
                Os =tta
                Ks =tta
                PDh *=tta < *=hta

                Te =ta, but =k-ta after e 'the aforesaid'
                                        he 'that'

                The extra -k- with demonstratives explains the
                *=hta form in Dhegiha.  The source of this *-k-
                is not clear.  Perhaps *ki 'the'?

                OP =di=thaN 'from' < 'in' + thaN
                   =tta=thaN 'from' < 'to' + thaN

                Te =taNhaN < *=ta=thaN  (*th > h in Dakotan)

                The interesting thing about the Dh and Da locative
                forms is that they condition inserted -a- after
                some nouns,
                e.g.
                OP tti' 'dwelling' + di => tti'a=di
                but Te thi'=l
                (I don't think we ever find *-ya=l or *-ya=tu.)

                OP tti' 'dwelling' + tta => tti'a=tta
                Te thi- + ta => thiya'=ta

                Dakotan makes up for lacking -ya- with *tu
                by using it with -takiya 'toward' and -taNhaN
                'on that side, from'.

                Te thi + takiya => thiya'=takiya, etc.

                Dakotan =tu conditions the e-grade of ablaut in
                nouns.


-ta 'seem,      Compare the Dakotan diminutive.
     resemble'

                Te =laka 'consider as' (dependent verb)
                Te =la diminutive

                I think the Dakotan diminutive derives from
                readings like 'a so-called X', 'a sort of X',
                'something that would pass for an X', 'x-like'.
                English use of -kin:  panikin, manikin,
                munchkin, etc., seems similar.

                The Dakotan diminutive conditions the e-grade
                of ablaut in some nouns including the -ye variant
                of the stem forming -e e.g.
                s^uNkwiN'ye=la 'mare'
                s^ahi'ye=la 'Cheyenne'

So, to summarize:

Crow has -a (including perhaps some covert cases) or a => e ablaut to

mark the citation form, and that is used in "independent" (unsuffixed)

contexts and with certain suffixes in the class of articles,

postpositions, and 'seem, resemble'.

MVS has -a ~ -e to mark what I've been called themes, i.e., nouns in

independent contexts (not compounded).  Thematic forms in *-(y)a are

fossilized before certain locatives, and extra a- prefixes some classes

of articles in Dhegiha.  The e-ablaut grade occurs in Dakotan before one

locative (=tu) and before the diminutive which seems to have a 'seem,

resemble' origin.

The most likely cognates for the actual conditioning suffixes (Crow) or

enclitics (MVS) are

PS *wiN-a- 'one' : Cr -m indef art : Te waN indef art : OP wiN indef art

PS *=tu : Cr -n :  Te =tu ~ =l : OP =di : Os =ci : Ks =j^i loc

PS *=thaN path : Cr -taa path : Te =(taN)haN 'side, from' : OP =thaN

extent, from-compounds

and/or

PS *=k-ta 'to' : Cr -taa : Te =(k)-ta- 'to' : OP =tta 'to'


Less secure is

PS *ki def art : Cr -s^ def art : Te ki(N) def art : OP ...

A certain amount of irregularity is involved in the sound

correspondences, and often patterns are somewhat variable, but I think

the general pattern is clear enough.