And so we came to examine interaction in ELF - one of my chief research areas. To me, empirical (data-based) findings should both generate and help us refine our theories on ELF and, indeed, on language and communication.
A key point I was trying to introduce and develop in the 1996 J of Pragmatics article was the notion that ELF interactions can and should be subjected to *unprejudiced description* - in '96, few such descriptions of L2/ELF interactions could be found in the literature.
As I said in class, I strongly encourage you to seriously consider collecting your own data and basing your assignment on analysis of the data. Data will be naturally-occurring ELF talk, so it won't be scripted or rehearsed forms of interaction (e.g. those forms evident in tv shows such as 'Friends' and in movies). What we're talking about, then, is talk that would have occurred even if the researcher had not been interested, present, or involved. At least ideally. You're lucky the internet provides a wealth of possible sources - YouTube not least is a potential goldmine for data, but there are hundreds of radio stations online, too, broadcasting non-scripted interviews, discussion and/or debates - much of which can be extracted - by you - for research ends. Otherwise you should rely on your ingenuity, and the world around you, beyond the pull of good old YouTube. So friends talking - in halls of residence, in their (or your) rooms, on the phone, sitting around in cafes, etc, etc, are all potential data sources. All you need is a small audio-recording device, and these can be borrowed from ECLS (Chris Letts - google him at Newcastle University - runs the AV lab at ECLS and will freely and happily loan out recording devices you can use to grab data from the myriad sources available ... Email Chris at chris.letts@ncl.ac.uk
Once you have some data, listen, listen again, and select 5 minutes to transcribe. Transcribe something and things that are interesting, or puzzling, or intriguing, or that simply catches your eye, or ear. This could form the basis of your data source. Don't use more than 5 minutes of the recordings. Two minutes might be enough. One minute might be fine. Five 15-second extracts could do the trick. It's the quality of your analysis that counts, not quantity.
My introduction to Conversation Analysis (CA) was brief, and I'm going to be relying on you to do your own background reading on CA - but at least there's a lot of decent stuff available online (I especially like
Charles Antaki's webpages on CA - see also the
'ELF Resources' page on the module's website). You also might want to
join the MARG data sessions here in ECLS, which meets - very informally - once a week to do data analysis. It's a great way to learn CA's methods.
Otherwise there are now a few book-length descriptions of CA's methods, not least Paul ten Have's excellent
Doing Conversation Analysis. The journal
Research on Language and Social Interaction (
ROLSI) is probably the best source for examples of high-quality articles that deploy CA methods. See what you can find, and get reading!
Next week we're discussing in detail two of my articles, so do prepare!
Comments welcome!
