chaNnu(N)pa

["Alfred W. Tüting"]

This all actually sounds pretty well and convincing (were not the 
strange -u- appearing out of nowhere).

There's still one thing to be clarified:
Buechel gives unpa [uN'pa] -> munpa, nunpa, unkunpapi ... (what would be 
okay so far), but what about Bruce Ingham's uNpa -> uNmuNpa, uNnuNpa, 
uNkuNpapi(sic!) (my transcription)?? Is this a typo, given the 
inconsistency with 1st pl.? Or does this conjugation derive from the -u- 
mentioned above, being kind of reminiscent to it (-u- -> -uN-) with the 
nasalization triggered by the following -uN-?

Alfred



> Buechel suggests that chaNnuNpa derives from chaNli 'tobacco' plus uNpa
'to smoke'. Many speakers change "l" to "n" after a nasal vowel when the
consonant closes the syllable (e.g. akaNl is often pronounced akaNn), so
the development chaNli > chaNl > chaNn is probably regular, and the
etymology would be chaNli-uNpa.

Regina makes that point that there is vowel deletion in the third person
forms of these verbs if the underlying sequence includes VV; chaNnu-m-uNpa
in the first person is chaNn-uNpa in the third, echa-m-uN 'I do' is echuN,
and echamiN 'I think" is echiN.  It doesn't happen in iyuNga/imuNge
because of the /y/.  Again, for my taste, this process is completely
regular as an application of the phonological rules of the language: there
are no VV sequences anywhere; either a vowel is deleted or a glide is
inserted to prevent them. So there is no need for a "special"  statement
about irregularities -- the problem is not with the conjugation, but with
permitted phonological sequences.  The sequence echa-uN is doing exactly
what it is supposed to do.

Now back to 'smoke'.  If the synchronic stem is chaNn-uNpa, first person
should be *chaNn-m-uNpa.  Here, then, we have a genuine irregularity in
development; an extra /u/ has appeared from somewhere (the ever-powerful
tool of the diachronist, "analogy", comes to mind), and this word does
seem to be "irregular" in that its behavior is not predicted by rules
otherwise needed in the language.

So I return to my original objection to the overuse of the notion
"irregular".  All of the behavior of these verbs except the extra -u- in
'smoke' can be stated by rules that apply wherever the environment is
right.  That makes them rule-governed, even if they are few in number
(recall that "irregular" means "not rule-governed").  I think the REAL
irregularity is the occasional use of -mn- instead of -m- for the first
person of some -yVN- stems.

 >From a pedagogical point of view, you may want to relax the definition of
"regular" and make it mean "most common", but I don't think that's good
descriptive linguistic practice.

David <<