Interaction at the Boundaries of a World Known in Common: Initiating Repair with “What Do You Mean?”

ABSTRACT

A recurrent feature of Garfinkel’s famous breaching experiments in which student confederates were instructed to engage an unsuspecting subject in conversation and subsequently insist that they “clarify the sense of (their) commonplace remarks” is the student experimenter’s use, in attempting to realize such “insistence,” of a turn composed of “what do you mean” plus a repetition of some part of the prior talk. Garfinkel suggested that such utterances tended to provoke moral outrage. The analysis presented here aims to explicate just how such utterances work, how they intersect with assumptions about the distribution of knowledge between participants to interaction, and why they might elicit such strong reactions from those to whom they are directed. Data are in American and Canadian English.

Notes

¹ In using the terms “background knowledge” and “common sense knowledge” in the following analysis, we draw on Garfinkel’s (1963, p. 221) usage and his observation that in conversation, “much that is being talked about is not mentioned, although each expects that the adequate sense of the matter being talked about is settled.” Garfinkel (1967, p. 27) elaborates this point in his discussion of the conversation clarification experiment:

Because they were reporting the actual conversation of particular persons, they would look for these further contents in what the conversationalists had ‘in mind,’ or what they were ‘thinking’ or what they ‘believed,’ or what they ‘intended.’ … All of which is to say that students would invoke their knowledge of … what the parties understood in common.

As these quotes suggest, Garfinkel uses “background knowledge” to refer to more than simply “knowledge” in the narrowly defined sense of, for example, “true justified belief” or some such. Instead, the term glosses the entire range of matters that members draw on in making sense of specific utterances in their sequential contexts, including knowledge, information, and/or beliefs that are (ostensibly) shared and accountably known in common by members of a community, linked to the participants’ shared biographies, or that emerge in, and as, the course of the currently unfolding conversation.

² Schegloff (1997) identifies two basic practices that incorporate WDYM: There is a simple, and presumably basic, form (i.e., “What do you mean”) and an elaborated form, WDYM + X. According to Schegloff, while WDYM is typically used to initiate repair, WDYM+ X can be used either to initiate repair (when X is a repeat of all or part of the trouble source turn) or as a practice for challenging a prior turn.

³ One reviewer asked why we do not retain transcription symbols when referring to utterances in the main text of the article. Adopting such a practice would not address the thorny problems associated with the unavoidable reductions apparent in even the most detailed transcripts and analyses. Because referring unavoidably entails simplification and reduction, participants’ practices appear oriented to adequacy and recognition—rather than fidelity to detail. We use standard orthography (and line numbers) to enhance the text’s legibility and encourage readers to refer to the transcribed extracts (and the recordings from which they are excerpted) to appreciate the features of utterances in their sequential context.

⁴ Permissions for the data used in this study are as follows. Example 1: Cited here from Sacks (1984). Examples 2 and 6: Drawn from a corpus of recordings collected by Sidnell, participants’ consent to use and publish these data in their present form having been acquired at the time of collection. Example 3: Drawn from a corpus of recordings collected by Raymond, participants’ consent to use and publish these data in their present form having been acquired at the time of collection. Examples 5, 7, 8, 9, and 10: Participant’s permission was gathered by the researcher who collected these data, which includes our right to use and publish these data in their present form.

⁵ As the called party, Shelley can’t claim a similarly urgent or pressing agenda, suggesting instead that her reciprocal identification playfully mocks Debbie’s prior turn and by extension, the “serious” action it projects.

⁶ Although “what happened today” (in Excerpt 1) turned out to be a genuine query, because this form can make an explanation or other remedial action relevant next, it can be used as a practice for complaining. In this respect, Priscilla’s use of WDYM may also pick up on the query’s status as a “possible” accusation or complaint (see Schegloff, 2005).

⁷ There is already a hint of laugher in Linda’s repeat of “who’s this.” See Jefferson (1979).

⁸ A particularly clear case comes from subsequent talk in one of the breaching experiments. After the experimenter asks, “What do you mean, you had a flat tire?” the subject is described as “momentarily stunned.” The experimenter’s report continues: “Then she answered in a hostile way: ‘What do you mean, “What do you mean?” A flat tire is a flat tire. That is what I meant. Nothing special. What a crazy question!’”

⁹ By virtue of its alternative formulation, “what does that mean” also appears to be fitted to the details of this context, in which were Rubin to say “what do you mean,” he could easily be heard as merely restating, and insisting upon, what he said in line 08.