Crow Citation Forms (Re: WINN TERM "FRENCH")

[Koontz John E]

On Sat, 4 Jun 2005 Rgraczyk at aol.com wrote:

> The final i in hachki is unexpected, and I don't have a good explanation
> for it at the moment.  It's possible that the distributive pl is ita
> rather than ta.

Randy, I wondered about that, too, though on general principles, rather
than on any specifically Crow (Corvine?) considerations.  However, the
only example of such a thing per se in the draft of your grammar is hachka
itself:

ishii'o ha'chki-t-uuhtaa 'their hair is generally long, but'

Versus:

bachi'a-t-uuk 'they fought'

I'm not sure if this counts, since ia is not usually in alignment with a
in behavior.

That was as far as I got while trying to rise to the challenge of Jimm's
examples.

Looking further in Randy's grammar, I guess it's a bit like Low Vowel
Ablaut in which "Stem-final a ablauts to i before the modals immaachi
'will', ishdaachi 'should', and ih 'may, might', and before the habitual
suffix i, ..."  I should note for the benefit of those not looking at the
examples, which I won't quote, that the a > i shift results in two i
vowels in a row, but it isn't simply assimilation, because it also occurs
before immaachi's plural form ommaachi.  Anyway, the problem with invoking
that rule here might be that there are exceptions (e.g., bachi'a ?) with
-ta, but not with Low Vowel Ablaut.  Also, the LVA conditioners all start
with i in at least their singular forms.

> The following is a key for converting stems to citation forms:
>
> Stem         Citation Form
>
> i            e      (ee?)
> u            o      (oo?)
> a            e      (ee?)
> ee           ee
> oo           oo
> aa           ee
> ii           iia
> uu           uua
> ia           ia
> ua           ua

I've reorganized the list somewhat and added a query as to whether it
might be appropriate to indicate e and o as long, on the grounds argued In
Randy's grammar that length is not distinctive for e and o, and that e and
o are primarily orthogaphic variants of ee and oo.

It appears that short vowel stems i/u/a lengthen the vowel and centralize
it; i and a become ee, u becomes oo.

Long vowel stems ee/oo/aa have the centralizing effect only:  ee and oo
remain ee and oo, but aa becomes ee.

Long vowel stems ii and uu add a, which might in the context be thought of
as a centralizing effect, though it doesn't work out as it does with ee
and oo.

Diphthong stems in ia and ua (*ih- and *uh-, I think) remain ia and ua.

Relating this to Proto-Siouan is relatively easy.  What is probably
happening is that thematic vowels *-a or *-e are being added to form the
citation form.  The lengthening and centralzing effects result from this
added vowel.

There has probably been some leveling, but the a and aa => ee represent
alternation of -a and -e in C-final stems and assimilation of ae to ee,
and i and u to ee and oo represent contractions of ia/ie and ua/ue.
Actually, we wouldn't need to consider separate a- and e-themes in Crow
were it not that a- and aa-stems have their citation form in -ee.  For all
other purposes it suffices to think in terms of adding -a.

The behavior of the ii and uu forms reflects adding -a.  It's not clear
why with these stems there is no contraction to ee and oo as there is with
i and u stems.  Perhaps they are *iih and *uuh stems?  I haven't tried to
look at cognates yet to resolve this.

The ia and ua stems (from *ih-a and *-uh-a) behave a bit differently from
the i and a stems because of the lost stem-final h.  In effect, these
stems always occur in citation form.

One detail that tends to confirm that the adding of a thematic vowel is
involved arises from a consideration of the environments in which stems
and citation forms occur.  The stem is the form that occurs when some
additional suffixes occur (or in incorporations).  Stems lose a
final short vowel under various conditions.  The citation form is used
in independent citation and when no other suffixes are added, a condition
difficult to achieve with verbs.

However, there are a few contexts in which the citation form, not the
stem, serves as the base for additional derivation.  In particular, this
occurs with various determiners, like the definite article -sh and with
the postpositions -n LOCATIVE and -taa PATH.  The interesting thing is
that these are just the contexts in which Mississippi Valley languages
tend to insert an extra -a- between their stem forms and their affixes.

For example:

OP nu akha 'the man'
OP nu ama 'the men'

In these cases the a- is an obligatory initial of the article.

Also:

OP tti-a-di 'in the house'
OP tti-a-tta 'to the house'

In these cases an extra -a- is added after certain nouns when the-di
LOCATIVE and -tta PATH are added.  With preceding forms that end in -e,
the e changes to a in conservative usage in the Dorsey texts, e.g.,

OP maNthe 'under, inside'
OP maNtha-di 'within, into'

OP xade 'grass'
OP xada-di 'on the grass'

I suspect Crow -n and -taa are cognate with -di and -tta.

The use of a citation form as the base for some derivations might relate
to the hachka/hachki case.  I notice that Randy lists two -ta morphemes.
The distributive plural which he (and I) identified in isbitchiihachkite
'French(men)' attaches to stems.  The suffix -ta 'appears to, seems' is
attached to citation forms.  While the citation form of hachka would
hachke, perhaps that is reduced to hachki in this context?

I suppose that might account for Hu Matthews' interesting gloss of
'generally their knives are long'.  He's perhaps going with the 'seems,
appears' morpheme, but has an alternative gloss?

The glossing issues with -ta reminds me of the Dakotan diminutive, e.g,
Teton =la, which seems to be derived from the dependent verb laka 'to
consider'.  In the same vein Buechel lists -laka 'rather' and -laka as an
inferential marker (a bit like Omaha-Ponca =the ?).  These senses seem
rather in line with 'appears to, seems' and 'generally'.  And since that
conception leads in Dakotan to a diminutive sense, perhaps the -ta in Crow
isbitchiihachkita 'French' has something of that sense (or translates an
affix like =la).  In other words, this form is effectively 'little (or
lesser) longknives' (= 'those who might be considered longknives though
they're not quite the same' or 'those whose knives are generally, but not
quite consistantly long' or 'those whose knives seem to be long, but may
not be the genuine article').