"irregular" muN, nuN (fwd)

REGINA PUSTET pustetrm at yahoo.com
Wed Jun 14 06:56:39 UTC 2006 

> I’d remove chaNnuNpA ‘to smoke’ which has reduplicated forms for 1 
  and 2
  > SG from the list, this is an idiosyncratic pattern.
   
              I don't understand.  'I smoke' is, as far as I remember,
  chanumuNpa, chanu - m -uNpa; how is that reduplicated?   The second 
  person
  looks reduplicated (chanu-nuNpa), but I think that's just the "n" 
  pronoun
  appearing where it's supposed to go.  
   
  
  The third person of ‘to smoke’ is chaNnuNpa. If the verb were a regular m-/n-verb, the third person would have to be chaNnu’uNpa to get first and second person chaNnumuNpa and chaNnunuNpa. We could actually posit chaNnu’uNpa as basic root and analyze the third person as a contracted form, but still, we need that contraction rule which moves this verb a little farther away from being a ‘regular’ m-/n-verb. The transitive version uNpa ‘to smoke (a pipe etc.)’ has first person muNpa, second person nuNpa and is therefore a ‘regular’ m/n-verb.
   
  Thinking about this a little more, I found that echiN ‘to think’ is not a particularly regular m/n-verb either (1SG echaNmi, 2SG echaNni). I did with these m-/n- verbs exactly what Linda did with the Assiniboine correspondences: I listed them separately in the grammar, each with its full paradigm.
   
  In the 
  English
  cases you provide, you have to memorize the whole paradigm, even though
  you can find more examples here and there.  So you do get a pattern in 
  the
  past of brought/taught/fought, but there's no predictability of any 
  other
  form in any paradigm.  The present of "taught" is teach, not *ting, and
  while "fight" goes with "fought", "write" does not become "wrought".  
  You
  don't even get to count on the vowels all the time: "saw" is past, but
  "draw" is present.  So I think there's a difference in quality between 
  the
  minor pattern in Lakota and the verbs which I, too, would call 
  irregular
  in English.
   
  My English examples were not particularly illustrative because, as you say, there are no predictable patterns emerging. But I’m sure these are there. For instance, there is a mini-class which contains lead-led, feed-fed, bleed-bled, meet-met, etc., where the vowel changes regularly in analogous environments.
   
  Regina
   


ROOD DAVID S <rood at spot.Colorado.EDU> wrote:  
On Tue, 13 Jun 2006, REGINA PUSTET wrote:

> I've just checked on my Lakota grammar files and found that the rule “m-
> ‘1SG’/n- ‘2SG’ with vowel- or y-initial verbs with subsequent nasalized
> vowel” holds up, all in all. But we should say that the vowel or the y
> do not have to be verb-initial, they just have to be where the
> affixation slot for m-/n- is, cf. iyuNg°A 'ask', a verb that you quoted.

I didn't edit that rule down carefully enough -- you're right. I
meant to take out the "vowel initial" and leave "when the pronoun precedes
a nasalized vowel (with or without an intervening "y"); I got the second
half in but forgot to do the delete.

> I’d remove chaNnuNpA ‘to smoke’ which has reduplicated forms for 1 and 2
> SG from the list, this is an idiosyncratic pattern.

I don't understand. 'I smoke' is, as far as I remember,
chanumuNpa, chanu - m -uNpa; how is that reduplicated? The second person
looks reduplicated (chanu-nuNpa), but I think that's just the "n" pronoun
appearing where it's supposed to go. Many of these verbs have a vowel
before the "m" or "n" that is absent in the third person (echuN,
ech_a_mu); Allan always wanted to claim the vowel was in the UR and
deleted in the third person (so our pedagogical stuff has 'to do' as
echa'uN). That's clearly more abstract that is currently fashionable.


The real exception
> to the rule I found is waNyaNkA ‘to see’ which satisfies the rule but
> has bl-/l- for 1./2. SG. Plus lack of nasalization of the second A.
>
WaNyaNka is an example of something Trudi Patterson described as
"nasal leaking" (I think that was her term). There is a tendency for
nasalization to "leak" across /y/ in either direction, unless the /y/ is
in an affix. You see this often in first person non-singular verbs; 'we
go' is often uNyaNpi rather than uNyapi; that holds for all the
causatives, too -- theb'uNyaNpi from thebyapi.. The second "a" in 'see'is
not an underlying nasal vowel, as demonstrated by the fact that it does
not take nasalized pronouns, and it loses its nasal whenever something
separates it from the initial waN- syllable. Weirdly, however, I seem to
recall other dialects or languages in which the cognate of this verb had
the inherent nasal on the second vowel and none on the first.


> On the irregularity issue in general I would say that irregularity is
> a matter of degree, of course, when the number of class members which
> inflect a certain way is the defining criterion. The extreme case would
> be limiting 'irregularity' to a class that has one member, as in the
> case of Lakota eyA 'to say'. Fine with me. But then, the most widespread
> usage of the term 'irregular inflection' in traditional grammar writing
> seems to refer simply to a class that is smaller than the largest
> competing class. Example: past tense in English. -ed forms constitute
> the major class, compared to which stuff like bring-brought,
> teach-taught, fight-fought forms a phonemically uniform minor class. And
> such forms are normally called 'irregular' in the literature, to my
> knowledge. Of course there are additional 'minor' ways of forming
> 'strong' past tenses in English, as in see-saw, which is phonemically
> divergent from the -ought/-aught pattern above. Its a matter of taste if
> we posit many
> minor or 'irregular' classes here or one big one, on the basis of
> whether we have -ed or something else, soemthing that's less
> predictable.
> In a synchronic description of a language, establishing inflectional
> classes on the basis of historical criteria, like possible epenthetic
> y´s in Lakota, may be problematic. And this is particulalry true of
> indigenous languages where historical and/or reconstructed data might be
> hard to come by. In my Lakota grammar draft I didn't even try to do that
> because I don't get enough descriptive mileage out of it. The main
> thing, to me, is providing a more or less complete classification of
> inflectional verb types, no matter what the exact arrangement of these
> types/classes in the script may be.
>
> Anyway, I never fight over terminology. :-)

I'm in agreement with the refusal to fight over terminology, but
there's something inherently scary about the term "irregular"; to me it
means you have to memorize more parts of the paradigm than usual, one item
at a time. In the Lak. case under discussion, everything follows from
knowing the first and third singular forms -- and that is also true for
all the other conjugation classes except for 'to say'. (in some cases
you need the "uN(k)" form too, to be complete, of course.) In the English
cases you provide, you have to memorize the whole paradigm, even though
you can find more examples here and there. So you do get a pattern in the
past of brought/taught/fought, but there's no predictability of any other
form in any paradigm. The present of "taught" is teach, not *ting, and
while "fight" goes with "fought", "write" does not become "wrought". You
don't even get to count on the vowels all the time: "saw" is past, but
"draw" is present. So I think there's a difference in quality between the
minor pattern in Lakota and the verbs which I, too, would call irregular
in English.

I just wish I really did understand the reasons behind all the
-mn- first person forms. I don't have the foggiest idea where 'run' comes
from, for example.

Best,
David



>
> Regina
>
>
>
> ROOD DAVID S wrote:
> I object to calling the conjugation pattern here "irregular". It is
> consistent and regular within Lakota for vowel or y-initial verbs when the
> pronoun precedes a nasalized vowel (with or without an intervening "y");
> it's rare because there just aren't very many roots that fit the criteria,
> but it's not irregular. echiN (echami) 'think', make 'I sit' muke "I
> lie", chanumupa 'I smoke", imuge 'I asked her a question' and the various
> derivations of uN 'to use' which were just cited. The alternative, mn-,
> n- in e.g. wa'i_mn_ake 'I run" is a secondarily nasalized version of the
> "bl" "l" set. I grant you that I can't predict which "-yVN" stems will
> take m- and which take mn-, but I suspect it has to do with whether the
> "y" is historically epenthetic or not.
> I would like to advocate distinguishing between a minor paradigm
> (minor because the required environment for the rule is rare) and an
> irregular one (which would mix elements from more than one pattern,
> perhaps, or which is unique to a single verb, like ephe 'I say').
>
>
> David S. Rood
> Dept. of Linguistics
> Univ. of Colorado
> 295 UCB
> Boulder, CO 80309-0295
> USA
> rood at colorado.edu
>
> On Tue, 13 Jun 2006, REGINA PUSTET wrote:
>
> > Thanks much, Bob -- this will be absorbed into the paper soon.
> >
> > > (a) how noNpa 'two' has developed into a comitative marker; what I'd need is a complete clause that shows the syntactic structures involved. A numeral as the source of an adposition is quite sensational to document since this represents a very infrequent grammaticalization channel for adpositions.
> >
> > There may be some Muskogean influence here. Choctaw has a construction that, although basically a DUAL is often translated with a comitative. They use the expression /itta-toklo/, approximately 'the two (of us) together' with verbs to signal dual participants. Toklo is 'two'.
> >
> > Is this pattern common in Muskogean? If it isn't, it is possible that Siouan or at least Biloxi has influenced Choctaw in this respect.
> >
> > There is a pan-Siouan verb /i-?uN/ transparently meaning 'to do with' (?uN 'do', i- 'with', right?) Among Siouan languages Dakota is the only one, as far as I know, that has dropped the instrumentive prefix i- but kept the meaning 'use' ( = do with).
> >
> > Does the apparently basic 'do' meaning of uN surface in Lakota echuN
> > 'to do'? The etymology might then be echa 'thus, like that, such' plus
> > uN 'to do', and we'd be in a good position to account for the irregular
> > first and second person of echuN (echamuN, echanuN), which correspond
> > nicely with the equally irregular and extremely rare first and second
> > person pattern for uN 'to use' (muN, nuN).
> >
> > Regina
> >
> >
> > "Rankin, Robert L" wrote:
> > May I add one or two comments too?
> >
> > > (a) how noNpa 'two' has developed into a comitative marker; what I'd need is a complete clause that shows the syntactic structures involved. A numeral as the source of an adposition is quite sensational to document since this represents a very infrequent grammaticalization channel for adpositions.
> >
> > There may be some Muskogean influence here. Choctaw has a construction that, although basically a DUAL is often translated with a comitative. They use the expression /itta-toklo/, approximately 'the two (of us) together' with verbs to signal dual participants. Toklo is 'two'. At the moment I can't be more specific than that. Pam probably has a better handle on this than I do. All my Muskog. reference materials are in boxes in my garage at the moment.
> >
> > (b) how oN(ha) developed into an instrumental marker. The very same process is indeed going on with Lakota uN 'to use'. again, I'd appreciate clauses showing the usage of the marker. Do you have any idea what the -ha is doing here?
> >
> > There is a pan-Siouan verb /i-?uN/ transparently meaning 'to do with' (?uN 'do', i- 'with', right?) Among Siouan languages Dakota is the only one, as far as I know, that has dropped the instrumentive prefix i- but kept the meaning 'use' ( = do with). That's all I can add, but it's at least suggestive of a trajectory for the grammaticalization.
> >
> > (c) if saNhiN is a noun -- this is what your translation seems to imply. Could this element function as an adverb as well? And again, if you happen to have examples of the usage of saNhiN that illustrate its development into a case marker, that would be great.
> >
> > Reminds me of the uses of Turkish /taraf/ 'side', borrowed from Arabic.
> >
> > Bob
> >
> >
> > __________________________________________________
> > Do You Yahoo!?
> > Tired of spam? Yahoo! Mail has the best spam protection around
> > http://mail.yahoo.com
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com



 __________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/siouan/attachments/20060613/bd8ecea8/attachment.html>

More information about the Siouan mailing list