Managing Trouble Responsibility and Relationships During Conversational Repair

Abstract

Using conversation analysis, this article focuses on other-initiation of repair (e.g., What?, I'm sorry?) of trouble speaking, hearing, and understanding. This article shows that the act of managing relationships is an essential feature of other-initiation of repair, and that different practices of repair-initiation can constitute different relational events that have different behavioral outcomes. Specifically, this article: (1) argues that context-free structures of interaction bias practices of repair such that other-initiated repair is vulnerable to communicating a stance that responsibility for trouble belongs to the speaker of the talk that inspired the repair-initiation; (2) discusses the implications of trouble responsibility for interpersonal disalignment and the organization of subsequent interaction; and (3) focuses on open-class (Drew, 1997) practices of repair initiation and argues that the apology-based format (I'm sorry? or Sorry?) communicates a stance that trouble responsibility belongs to repair-initiators, rather than to their addressees.

Keywords:

• Conversation Analysis
• Intersubjectivity
• Repair
• Relationship
• Responsibility
• Practice

The author thanks Elizabeth Boyd and Rita Mangione-Smith for sharing data. A draft of this article was presented at the 2005 National Communication Association, and as a talk invited by the Language, Interaction, and Social Organization (LISO) unit at the University of California at Santa Barbara (October 2005), where Gene Lerner and Geoff Raymond provided valuable feedback. The author also thanks Makoto Hayashi, John Heritage, Irene Koshik, Jenny Mandelbaum, Alan Sillars, and two anonymous reviewers for comments on previous drafts.

Notes

1. Note that a conversation-analytic approach does admit that relational structures can transcend individual interactions (Millar & Rogers, 1976; Sigman, 1991), and that variables exogenous to interaction (e.g., sex, race, emotion, and relational history) can affect both the process of communication and its understanding by participants. However, conversation analysis imposes particular methodological conditions for proving such claims (Schegloff, 1987a).

2. Trouble hearing is accountable as a virtual offense at least because the rules for turn taking are designed to motivate listening (Sacks et al., 1974). Additionally, trouble understanding is similarly accountable at least because actors share, rely on sharing, and (all things being equal) trust one another to implement common practices of meaning-making (Garfinkel, 1967); this trust includes the fact that communication is guided by the principle of recipient design: “The multitude of respects in which the talk by a party in a conversation is constructed or designed in ways which display an orientation and sensitivity to the particular other(s) who are the coparticipants” (Sacks et al., 1974, p. 727).

3. That these types of apologies do not clearly index a particular offense is highlighted by two additional observations. First, these types of apologies occur with apparent equal frequency in non-face-to-face settings, where there is no embodied evidence of offense, and they appear to operate in the same fashion. For example, see Extract A, which is drawn from a phone call between two physicians. Here, D2's prospective apology “I'm sorry.” (2→), which precedes the OIR “eight two o:ne,” is produced immediately adjacent to the trouble source (i.e., without delay; 1→).Second, prospectively apologizing, in this type of sequential position, is not reserved for the action of OIR, but for myriad possibly offensive actions. For example, see Extract B, which is drawn from an American phone call. Here, Ann prospectively apologizes, “I'm sorry”, for having to refuse Kammy's request to speak to Sally Murdock.These extracts, along with Extracts 2, 6, and 7 (in text), indicate that, although there are cases in which prior (often visible) conduct may provide participants with ways of determining that a prospective apology specifically indexes an OIR-related offense, this is by no means always the case.

4. Regarding Extract 6, one might argue that D1's “>(speak up) < ” (line 7) and “y’< jus’ have to speak up a little bit.” (lines 11–12)—which arguably accuse/blame the trouble-source producer D2—are counter evidence to the present argument because they claim that trouble responsibility belongs to the trouble-source producer (vs. repair initiator). However, note that, in each case (i.e., lines 7 and 11–12), the first/initial account for the apology, and the account most proximate to the apology, is a formulation of the repair-initiator's personal failure to hear: “ > I = couldn't- < ” (line 7) and “I can't hear >you” (line 11). The management of trouble responsibility within a turn of talk is not static, nor is it dichotomous in terms of whether trouble responsibility is claimed to reside with “self” or “other.” The possibility that prospectively apologizing (for OIR or other actions; see Note 3, Extract B) is a practice for claiming personal responsibility for a possible offense does not preclude apologizers from subsequently, and thus contingently, “spreading” or redirecting responsibility to other participants, as seems to happen in Extract 6.

5. In particular cases, verbatim repeats involved the omission of unit-initial and unit-final objects (frequently, turn-initial and turn-final objects) from trouble sources, such as discourse particles (e.g., So, Now, Well, Oh, Actually, etc.) and tag questions. Schegloff (2006) observed that the function of these objects frequently depends on their sequential positioning relative to immediately prior talk/action, and that the same objects frequently do not serve the same function outside of such locations. Because, relative to trouble sources, responses to open-class OIRs are in different sequential locations (e.g., unit-initial objects are now adjacent to OIRs), unit-initial and unit-final objects in trouble sources are frequently omitted in responses to OIRs because they are not longer relevant in the same way.

6. Three types of cases were excluded. First, there were cases where the trouble source (1→) was not hearable/transcribable, and thus where it was not possible to determine if the repair (3→) was a verbatim repeat. Second, there were cases where apology-based OIRs were Nth (e.g., second) repair-initiation attempts in a series of attempts to repair a same trouble source. For example, see Extract C, which is drawn from an American pediatric visit; at line 1, the physician's “you two” refers to the mother and the father.Here, there is a bias for the mother's second OIR attempt, “I'm sorry?”, to be understood specifically as not displaying hearing trouble because the physician treated the mother's first attempt, “Hm?”, as such by repeating (1’ > ) the trouble source (1→). These types of cases were excluded because the apology-based OIR is not understood on its own, but relative to responses to first-OIR attempts.Third, there were cases where the apology-based OIR was not the OIR ultimately responded to. That is, the apology-based OIR is the first OIR attempt, but it gets pursued, prior to turn transfer, with another type of OIR. For example, see Extract D.The physician initiates repair with, “I'm ° > sor'y,<°” (2→), but, prior to turn transfer (and after the long silence at line 8 where the mother “fails” to respond), he reinitiates repair with “What?” (2’ > ). These types of cases were excluded because the sequentially implicative action was no longer an apology-based OIR.

Additional information

Notes on contributors

Jeffrey D. Robinson

Dr. Jeffrey D. Robinson (BA, Department of Communication, University of California at Santa Barbara; MA, Department of Communication Arts & Sciences, University of Southern California; PhD, Sociology, University of California at Los Angeles) is an associate professor in the Department of Communication at Rutgers University