Other studies, therefore, aim to analyze not the internal structure of participant narratives, but rather the contingent turn-taking that builds such stories in situ. Many of these studies consequently draw from Garfinkel’s (1967) ethnomethodological paradigm – a paradigm that counterbalances macro-level, over-determined sociological explanations for social behavior with micro-level, contingent, co-constructed analyses of social action (what Goodwin [2018] calls co-operative action).
For example, Ochs and Capps (2001) challenge prevailing structuralist paradigms of narrative inquiry by analyzing recorded conversations, finding that everyday narratives do not often follow a predictable structure, but rather take shape only through accumulated and sedimented turn-taking between interactants. Put otherwise, ethnomethodological analyses of storytelling put more focus onto audience as a driving condition for storytelling – especially if the audience is an intimate or familiar interlocutor. “Familiar conversational partners… are likely to recall relevant background events, ask questions, raise issues and possibilities, and otherwise provoke further thinking about an incident. Such sustained collaborative reflection can destabilize a person’s prevailing interpretation of the self and others” (Ochs & Capps, 2001, p. 17). This stands in stark contrast to storytelling as discussed within research interview contexts or within the fields of scientific publishing or science communication where audiences are either presumed to be (1) naïve (i.e. a researcher hesitant to lead an interview with their own preconceptions or an audience assumed to “know nothing about one’s topic”) or (2) a deindividuated social category (i.e. how writers commonly define a prospective audience). Put differently, akin to Latour’s (1987) analysis of scientists’ packing and unpacking of facts through intertextual discourse, “conversation lays bare the actual dialogic activity through which different versions of experience are aired, judged, synthesized, or eliminated” (Ochs & Capps, 2001, p. 7). It also reveals a paradox for ethnomethodology-driven analyses of discourse among familiar participants as (1) such analyses often prioritize present interactional moves while harboring skepticism for unsaid discourses while (2) dialogue between familiar participants carry a wealth of background assumptions, (i.e. “local narrative formats, recognizable types of situations and people, and prevailing moral frameworks, which inevitably constrain representation and interpretation” [Ochs & Capps, 2001, p. 55]).
While length requirements preclude a full accounting of these narrative analyses of co-constructed storytelling sequences, several are worth mentioning. Michaels (1981), in an analysis of teachers and students’ storytelling practices in a first-grade classroom. She uses conversation analytic methods to tease out contingent meaning-making but also operationalizes racial differences to argue how white teachers often struggle to orient to, and thereby affirm, the storytelling practices of black students in comparison to white students. In a similar design, Blum-Kulka (1993) distinguishes between narrative scales including the present conversation, the telling, the teller, and the tale in order to compare the collaborative storytelling practices in U.S. and Israeli family dinnertime conversations. She argues that while American families “claim access to story ownership through familiarity with the tale, celebrating monologic performances…in Israeli families, ownership is achievable through polyphonic participation in the telling” (Blum-Kulka, 1993, p. 361). Prior (2016), on the other hand, shifts this polyphonic production of stories to the research interview context. In an analysis of semi-structured interviews, Prior demonstrates that participant narratives are not neutrally elicited windows into experience, but rather an interactional exchange driven by (1) the questions initiated by the researcher (2) the turn-taking sequences of which the interviewer plays a major role and (3) impression management on behalf of the interviewee. All contribute to elicited narratives as interactional accomplishments rather than pre-figured segments of discourse (see also de Fina & Perrino, 2011). This thereby parallels Latour’s (1987) unboxing of science’s black box by unpacking the interactional moves that produce interview excerpts within narrative inquiry research reports.
Ontogenetic Reflection
When I listen to this back, I wonder why this approach to data analysis seems to be so far overshadowed by practices of thematic analysis, particularly within applied linguistics. I’ll save my comments on thematic analysis for next week. But one thing this makes me think of is the way that applied linguists seem to cling so desperately to the practice of playing scientist in order to produce qualitative research with rigor, to both openly discuss reflexivity while still attempting to adopt a distanced viewpoint. At least that distanced viewpoint can be made visible through practices of transparency. But it makes me wonder why, as narrative analytic researchers, we try so hard to prepare research reports that read like STEM papers. Isn’t the purpose of adopting narrative inquiry, at least in part, to depart from the IMRAD structure of a research article and to play with conventions? I still believe that this comes from an unwillingness for applied linguistics researchers to assume the mantle of a storyteller. We want to be seen as scientists because we have inherited an ideology that shows scientists to be serious academics and storytelling to be an activity for children.
I think that one potential third path is to return to this analysis not of an “objective reality” captured through the triangulation of themes, but by telling the story not by zooming out to generalize, but by zooming in. We can zoom in to see the ways that stories as discourse unfold and refold in real time. This is what Ochs and Capps were arguing back in 2001, and it seems to be a perspective that has lost a lot of favor in the field. Conversation analysis, on the other hand, seems to understand this. The rigor of conversation analysis is not achieved through a theoretical saturation of codes, nor through a generalizable conclusion, but rather from zooming into a careful attention to detail in the production of transcripts — transcripts that go beyond the semantics of words — and ultimately revealing profound insights about the co-production of stories and knowledge and social action, not by fully setting assumptions aside and adopting unmotivated looking.
I think all looking is motivated by a perspective. But what I dream of are analyses that produce more surprise, that suspend or upend the common assumptions that we make about language, politics, and the formation of knowledge. These practices of research are not inherent. They are inherited. And if we want to produce more readable, exciting analyses, it would benefit us to also find the surprise — or the “wow moment,” as David Boje would call it. This is what I miss.
Leave a Reply