Wednesday, June 29, 2011

Why It's OK to Use Learner English

Because I'm studying people's reactions to whether it's correct.

This is a great weight lifted, in some ways. I am not linguistically analyzing these texts. Not really. I am analyzing what other people think of them -- that is, "acceptability." Which, according to Bamgbose, trumps nearly all else.

Big gap that needs to be filled though -- like, today -- is what features to focus on. Shall I use Xu or other lx analyses as a guide? (Don't want to be overly deterministic though.) I can't leave it totally open, because then people will just say "well, this looks weird." (Then again that kind of data is useful too.) I can't be totally like "please identify the morposyntactic features which do not conform to your understanding of Standard Written English for Academic Purposes" because that is lame. So we need something.

Also: Just came across He & Zhang (2010). Right smack dab where I'm headed in terms of looking at norms, but with matched-guise (spoken) tests. Looking forward to reading!


update: There are some problems with this. As I look at the corpus, I become quite  clear that it is indeed a learner corpus. Let's say that I find, as I'm looking at it right now, that the term "human being" is frequently used without an article, yet in a situation that seems to call for a plural. All of my participants have indicated that this is unacceptable, yet it is a common feature of these texts. Let's say I then take that to the Chinese teachers/students, and they all agree that it is unacceptable. What have I proven? That nobody accepts this as an innovation, yet it is a common feature of Chinese L2 writing students' texts.

I guess we can say that's worth doing. I have a hunch that almost none of the features that NSs would reject would not also be rejected by the Chinese teachers. I suppose this would show that CE is indeed exonormative and that these variations are being rejected as potential innovations.

Also, another piece of shaky ground I'm wondering about is the choice to include English students. Presumably, English majors in China are supposed to graduate with "near-native" proficiency (citation?) in English. This is interesting in that it tacitly acknowledges the difficulty of the NS-based model. But it also gives us a chance to test this, if we do the quiz with 4th-year English majors. (I do sort of see some reasons to do this at the end of the schoolyear though, if that is what it's being based on.)

Still,  I wonder if what I'm proposing really differs from Hinkel's study of the features of L2 texts. Is it just that I'm thinking about it differently, as 'potential innovations,' that makes the difference? I guess so.

"Potential innovations" -- remember that. That seems like an important hook to hang some of this on.


UPDATE 2: Remember:

This is a study about acceptability.
It is not a study which actually purports to identify features of CE.
It assumes certain things about CE based on WEs theories.
It aims to test certain aspects of WE theorizing by using (potential) CE.
Norms, standards, and acceptability are all part of WEs theory.
Features is another big part.
Is there an empirical basis for saying a variety of English has certain feature? Yes, but I don't know what.
Will my study provide an empirical basis for what some possible features of CE are? Maybe.
But that isn't necessarily the point.

If NSs and NNSs reject something, it's widely considered an error.
If NSs reject something and NNSs don't, this seems like evidence for an 'innovation.'

No comments: