double inflection

[Pamela Munro]

This is actually quite exciting, though not yet (I believe) immediately
relevant to the UA case (though perhaps Jason will see a connection).

What the reduplication of person marking in short verbs like ya reminds
me more of is a case from Muskogean. The Muskogean languages, as Bob and
perhaps many others of you know, have a series of verbal ablauts called
grades which are used to mark aspect, e.g. the hn-grade of Chickasaw
basha 'to be operated on' (I'll write nasalized vowels with capital
letters):

basha 'he is operated on'
bahÁsha 'he gets operated on a lot'

sabasha 'I am operated on'
sabahÁsha 'I get operated on a lot'

What reminds me of the Ibláble case is what happens with a verb like
isso 'to hit':

ihÍsso 'he hits him a lot'
sahÁsso 'he hits me a lot'

Grade formation generally operates on the penultimate vowel of the verb
stem. But with a verb like sa-sso 'he hists me', that penultimate vowel
is an inflectional prefix, so grade formation operates on it. Thus what
seems to me to be the parallel to the reduplication of person-marked ya:
normally these morphological rules want to operate on stems, but if they
are dealing with a short stem they may target an inflectional marker.

Pam

Koontz John E wrote:

>On Sun, 3 Aug 2003, Pamela Munro wrote:
>
>
>>This is actually interesting in terms of my question, since although
>>there are two inflections in this compound, as you explain it, the two
>>inflections both occur in the ya part, not the i part with which it is
>>compounded. But normally we do not see double inflection in reduplicated
>>verbs, even monosyllabic ones. (In Jason's and my data, the unexpected
>>double marking is in fact reduplication, but I don't see any other
>>immediate connections.)
>>
>>
>
>David and Pam make a very good point here, and I am embarassed to have
>overlooked it.  I do know one instance of reduplicated inflection in
>Omaha-Ponca, which involves the verb 'to say often'.  Unfortunately, the
>only available form is a second person, which is es^e's^e (cf. 'to say' A1
>ehe', A2 es^e', A3 a=i, A12 aNdhaN=i).
>
>Hypothetically, the first person might be *ehe'he.  I'm not sure how a
>third person would be handled without suppletion.  In fact, most 'say
>often' examples are based on the habitual enclitic =s^na ~ =hnaN ~ =na.
>
>This is a complex verb, morphologically, with suppletive stems, but the
>first and second persons seem to go back to *e=...he, so this is something
>like e=s^-he-s^-he.  Since the h-stems are few and of limited productivity
>(this being an exception) and involve complicating preverbs in cases like
>this, it's hard to conceive that speakers handle inflection generatively.
>I think this very complexity is what allows the inflection to be included
>in the reduplication, so that we have es^e's^e rather than, perhaps
>*es^ehe.
>
>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/siouan/attachments/20030803/9cb4830b/attachment.html>