Interviews as Discourse Data

Interviews as Discourse Data

The term discourse analysis (DA) has a wide reference and is often characterized as an umbrella designator for a growing range of different theoretical approaches, analytic emphasis, and typical or preferred data types (e.g., Phillips & Hardy, 2002; Wetherell, Taylor, & Yates, 2001a, 2001b; Wodak & Meyer, 2009). Despite the range of approaches and disciplinary locations of DA, the common ground for variants within the DA enterprise consists of social constructionist epistemology; an interest in the dynamics of interaction, talk, and texts; and systematic analysis on recurrent elements of discursive, cultural meaning making (e.g., Koro-Ljungberg, 2008; Nikander, 2008). Research materials in DA, in the broadest sense of the term, consist of all forms of talk transcribed into a written format from audio or video recordings, and a wide variety of written and visual documents. These may vary from dyadic or multiparty interaction in everyday and institutional settings to (group) interviews and focus groups; the analysis of documents, records, phone and online conversations, diaries and newspaper items, media products, political gatherings, speeches, TV interviews or talk on radio; and, increasingly, the analysis of visual materials and the semiotics of place.

In recent years, however, the status of qualitative interviews as a means of data generation has been a topic of live debate, and the discursive social-scientific field is of two minds when it comes to using research interviews and their relative advantages and disadvantages. The key question in this debate, raised particularly within discursive psychology, is this: Does rendering a topic analyzable necessarily require interview data, or should researchers increasingly or perhaps solely turn to naturally occurring data? This chapter takes stock of this live debate concerning the epistemological status of interview data. It provides an overview of some of the key points of the so-called natural versus contrived debate (e.g., Speer, 2002a, 2002b, 2008) and discusses the analytic mileage and the problems and possibilities of interviews as qualitative and discourse data (see Gubrium and Holstein's “Narrative Practice and the Transformation of Interview Subjectivity,” Potter and Hepburn's “Eight Challenges for Interview Researchers,” and [Page 398]Tim Rapley's “The (Extra)Ordinary Practices of Qualitative Interviewing,” this volume, for parallel discussions). The treatment of the topic and the perspective adopted here is marked by my own background in discursive psychology and by the fact that my empirical work draws on both interview and naturally occurring or naturalistic data (Nikander, 2000, 2002, 2003, 2007, 2009a).

The chapter is divided into three sections. The first discusses the debate concerning natural versus contrived data and the key arguments typically voiced for and against the use of interviews (Griffin, 2007a, 2007b; Have, 2002; Henwood, 2007; Holstein & Gubrium, 1995, 1997; Nikander, 2007, 2008; Rapley, 2001; Silverman, 1993, 1998, 2006; Speer, 2002a, 2002b, 2008). The second section, devoted to classic and more recent discourse studies, illustrates how the analysis of interviews as pieces of interaction in their own right does away with notions of bias and contamination and provides ample examples of the analytic force and richness of findings that originate from these data. The empirical examples show that when analyzing interviews from the point of view of discourse, notions of contrived or researcher-provoked data have and can be turned into an asset, a tool, and a starting point for analysis and that this has resulted in perhaps some of the most influential analysis to date. In the final section of the chapter, I revisit the natural versus contrived debate and discuss the hows and the whats of interviews as well as the cross-fertilization between different analytic traditions, data sets, and potential future developments of DA.

...