What a great week it’s been! I think that’s been down to the fact I’ve had four days on the PhD, and only one day faffing about with moderations and meetings, etc. Having more time in the week on the PhD makes me feel able to tackle more than one thing – I think that’s the key.
So, the start of my week saw me on a mission to read Liz Statham’s book – Scattered in the Mainstream – in a day! Which I did. Hurrah for books written for a lay audience!
Then I turned back to the Baker bilingualism tome and made copious notes for my British Council project lit review. I was quite pleased to come to the end of this massive textbook, I felt a sense of liberation, I admit! So, then I decided to finish Jean Conteh’s book, based on her PhD research (Succeeding in Diversity), as it’s on repeated 7-day loans and, frankly, that’s getting a bit dull. It’s a great record of her two-year ethnographic study, which although small-scale (only 5 children, and interviews and observations of around 6 staff) captured a huge amount of data. She has a really nice written style – informal, clear and very honest. I bet she got incredibly bored of transcribing, though…
I took that back, and found myself getting another four books out – including yet another seven-day loan – fool! I’ve got 33 books on my desk at the moment. If I finish them and get another four out each time, we could be on a rather never-ending mission here. Oh yes, it’s called a PhD…
So, I decided that I wouldn’t read any of them next, but turned to a research methodology article my supervisor emailed me – it’s a 2010 article, by Steven Talmy. It’s really got me thinking about interviews as research instruments. Or not, as he prefers ‘research interviews as social practice’. I’m not convinced at all yet for the need for so much dense terminology in this field (social sciences, applied linguistics, education, etc.) but I do think what he’s saying is bang on. Interviews are such a part of our everyday lives that researchers don’t really give them enough thought when using them in research. It’s just an easy way to get some quotes and real-life ‘voices’ for your publication. But he problematises a few areas (oh dear, not sure I like that word either, but it’s everywhere…).
These were some of the notes I made in Mendeley (oh, that’s the other big news – I’ve migrated to Mendeley from Zotero!):
There is an inconsistency over how the interview in applied linguistics research has been theorized. Talmy suggests that this is due, in part, to the ubiquity of the interview – it is everywhere, so perhaps there is simply an assumption that it is the common-sense research approach. The fact that interviews are everywhere has given researchers an unfounded sense of security, stability and authenticity.
There has been a tendency to take participants at face value and not to problematize the data or the roles of those involved. Interviews can just become ‘black boxes…so widely accepted that [researchers] can just feed in questions and get quotations for [their] publications without worrying about the complex pragmatics that make them work’ (Briggs, 2007, p. 555, cited in Talmy, 2010).
Richards (2009) in Talmy (2010) says that the methodological and analytical aspects of interviewing needed to be clearer in published papers.
Talmy distinguishes between ‘interviews as research instrument’ and ‘research interviews as social practice’ to make his point about how the former claims to reveal the inner voice of participants and make it public, with data from interviews ontologically being simply ‘reports’ of the participants’ worlds.
The difference is that conventionally, interviews have been about the whats – what’s being said. Interviews as a social practice also take into account the hows – “the interactional and narrative procedures of knowledge production” (Holstein and Gubrium, 2003, p. 68, in Talmy, 2010). It is about the difference between process and product.
It makes a key difference at the analysis stage – instead of thematically grouping as interview data usually is, it is more about analysing the assembly of the interview itself – deconstructing the event that was co-constructed.
Talmy reviewed a range of studies that incorporated interview data. He was surprised at how few studies really synthesised the data they found for triangulation purposes.
He found that generally speaking, interview data did not sit amongst the rest of the data but rather was ‘foregrounded’, with most of the research data coming from that source.
He also notes that it can sometimes be difficult to tell the difference between ethnographies, case studies, narratives and life-history studies and queries if this may be due to slippage in the conceptualisation of these different types of study.
Cheung (2005) came in for some praise in terms of explaining methodological and analystical reasonings in this study, but generally, Talmy was disappointing with the background given in the studies he reviewed.
Talmy goes on to discuss in more depth some of the issues as he sees them, starting with the question of ‘interviews as reports”. He talks of the typical manner of presenting participants’ speech as quotes, isolated from the questions. As CA will tell us, answers don’t exist in isolation, they are designed to answer a specific question, so the question is important too, as it’s part of the co-constructed “conversation”.
It is also vital to consider the role of the researcher more closely and to consider that in the analysis of the data – often issues of conflict are mentioned in passing but the data not then analysed with that in mind.
His next area of focus is power. Power relations are sometimes discussed in terms of roles, but often not analysed in terms of the interview data – ie. who has control of the discussion.
The focus of researchers on the power of the voice is also problematic. Assumptions here are that people have just one voice (?), that they are expressing their true self and that the researcher gives freedom to these voices through the interview process. “voice as present, stable, authentic and self-reflective” (Mazzei and Jackson, 2009, p. 1-2, cited in Talmy, 2010).
In the data anaylsis section, Talmy calls on Pavlenko to support his problematising. she says that the weakness in thematic analysis of interview data are:
-a ‘lack of theoretical premise, which makes it unclear where conceptual categories come from’
-a lack of ‘established procedure for matching instances to categories’
-over-reliance on repetition to form the themes
-exclusive focus on what is said, rather than –considering what was not said
-the analysis of the storytelling process itself, maybe the links between the content is more telling.
By re-constructing the interview as a social practice, we rid ourselves of worries of resercher bias, as we acknowledge that there is a ‘fundamentally reflexive orientation to the collaborative character of knowledge production and data generation’ (Talmy, 2010, p. 13).
OK, something to think about, anyway. My husband, talking this through a bit with him last night, gave me the title for this post, from Nagel, whose article I’ve now imported into Mendeley, as that’s exactly the aim, I suppose. A researcher needs to take an invisible standpoint.
Contemplations on identity 