yumi/mipela

[Lise M. Dobrin]

I wonder whether anyone has any insight into yumi vs mipela, which, 
according to all official descriptions I know of (not least of which 
Mihalic's), are supposed to encode an inclusive/exclusive 
distinction. I have been thinking about this since my fieldwork in 
the East Sepik "west coast"/Arapesh-speaking area, where the use of 
these pronouns often fails to correspond to the written descriptions. 
So I would hear things like "Yumi bai wokim nupela hauskuk gen", 
where the speaker was telling me about her family's plans to build a 
new hauskuk when the old one rotted and there was no way (under even 
the laxest criteria) in which I should be figured a part of her 
family, or even part of the surrounding community, since I would be 
long gone from the village by the time the building would take place. 
At the same time, there are situations where the speaker's choice of 
1st pl pronoun is best interpreted as some sort of focusing mechanism 
that depends on the inclusive/exclusive distinction while not being 
determined by it. So while walking with my hostess in an unfamiliar 
village, she pointed out the path we were about to turn onto by 
saying "em ia em i rot bilong mipela", meaning essentially "this is 
the path that goes to our camp (as opposed to other people's)". But 
she might also have said "em i rot bilong yumi" meaning something 
like "this is the path we're going to take". The latter form would 
have been especially appropriate if there were others walking along 
with us who would NOT be taking that path, i.e., with the "inclusive" 
implying an exclusion of another sort.

I know that shifter pronouns often show behaviors like these, but 
since certain uses explicitly contradict the official description, 
might there be a place for mentioning them in a dictionary? The more 
immediate question is, how widespread is this pattern? Have people 
heard unofficial uses that conform to different patterns than the 
ones I'm describing here? Suzanne Romaine discusses children's use of 
such "communal" we in her 1992 book on change in TP, and that 
discussion also appears in a 'Language and Linguistics in Melanesia' 
volume from around the same time. She notes some tendency for 
communal we to occur in samples from areas where the indigenous lg 
doesn't have an inclusive/exclusive distinction. Arapesh doesn't have 
the distinction, which is consistent with this. Romaine doesn't 
really interpret the pattens she records, although she does appear to 
assume that the official description is accurate for Tok Pisin 
historically, which may or may not be true. Somewhere in her 
'Mountain Arapesh'--which is based on research done in the 
1930s--Margaret Mead mentions that her informant would "incorrectly" 
use yumi as an object of bilong when he should have been using mipela.

--Lise Dobrin