Ablaut (RE: Obviative/Proximate and the Omaha verb system)

[Rankin, Robert L]

>> In Crow, noun stems ending in i and a have an e-grade that occurs as
the citation form and when the the noun occurs with no further
suffixation. E.g., bili' 'water' (stem), bile' (citation form).  When the
definite article is added to a noun, it is always added to the e-grade, eg,
bile'e-sh 'the water'.  This would support some kind of connection between
e-grade and definiteness.

I suspect this is a sound change. Short e and o raise in final position but
are retained preceding the suffix/enclitic. I suspect that's the source of
Paul's Winnebago -i before affixes too.

>It occurs to me to mention that Mandan has a kind of noun-marker e that
is omitted in some contexts, e.g., in compounding, but also some
independent contexts.  Hollow treats it as an epenthetic vowel, if I
remember, but Kennard thought it was morphological.  Nouns often display a
final -r- or -h- or -?- when the e is added, e.g., ko ~ kore.  The -r- could
be something of a default, as -r- is the usual separator of two vowels, one
of them high in Plains languages where y ands r have merged or for some
other reason there is no y.  This would be my assumption.  However, s is
also not inconceivable that some Proto-Siouan nouns ended in -r (*-r or
(*-y) and some Siouanists have favored this analysis at least at time.

I don't have time to summarize all the discussion Dick, Wes, John and I have
had over Mandan -re.  It is extensive. Suffice it to say that some like a
phonological solution by which -e is epenthetic and then -r- is epenthetic
to break up the VV cluster (if I remember correctly). My own analysis is
more in line with Kennards and is morphological and looks upon -re as
meaning something like "it's a X", where X is the noun. There are also
verbal suffixes -re and -he that recur often but without easily specifiable
meaning at this point.

Bob