Exploring essentially three-turn courses of action: An institutional case study with implications for ordinary talk

Abstract

This article describes an adjacency-pair organized course of action in the institutional context of customers calling an electronics repair facility to request the status of equipment they have previously sent in for repair. Relative to the majority of adjacency-pair sequences described in previous research, this course of action is rare in that it is essentially (vs contingently) composed of three (vs two) turns, including status solicitation, status response, and acceptance/rejection of status response. After defending this finding, we situate and discuss its significance relative to prior research – in both ordinary and institutional contexts – on adjacency-pair sequence organization, including implications for sequence-based relevance rules, such as preference organization. Finally, we outline a possible general explanation for why some initiating actions set in motion essentially three-turn (vs two-turn) courses of action, and offer a candidate example in ordinary conversation.

Keywords

Conversation analysis customer service institution intersubjectivity preference relevance sequence

One of the most common and important social structures of action is the adjacency-pair sequence (Schegloff, 2007). The various relevance rules associated with adjacency-pair sequences (Schegloff, 2007) provide much of the foundation for the accountability (Garfinkel, 1967) of social action, the explication of which is a core aim of Conversation Analysis (hereafter, CA). This article uses CA to address the virtually unexplored phenomenon of essentially three-turn courses of action, that is, ones organized by an adjacency-pair-based sequence whose minimal, unexpanded form consists of three (vs two) turns. After a brief introduction, this article describes (and defends the description of) one such course of action in the institutional context of customers calling an electronics repair facility to request the status of equipment they have previously sent in for repair. Subsequently, we situate and discuss the significance of our finding relative to prior research – in both ordinary and institutional contexts – on adjacency-pair sequence organization, including implications for sequence-based relevance rules, such as preference organization (re preference organization, see Schegloff, 2007). Finally, we outline a possible general explanation for why some initiating actions set in motion essentially three-turn (vs two-turn) courses of action, and offer a candidate example in ordinary conversation.

Although periodically contested (e.g. Tsui, 1985), there is an overwhelming amount of evidence demonstrating that, in its minimal, basic, unexpanded form, the adjacency-pair sequence consists of two turns, each of which constitutes a basic part of the sequence (re first and second pair parts, see Schegloff, 2007; Schegloff and Sacks, 1973). For example, in Extract 1, book-ended by lapses of silence in their dinner conversation (at lines 4 and 7; re lapse, see Sacks et al., 1974), Mark initiates an adjacency-pair sequence by making an offer to his wife: ‘Want some more milk?’ (line 5), which she rejects: ‘Mm mm’ (line 6).

Extract 1:

As noted by Schegloff (2007), adjacency-pair sequences ‘can come to have more than two turns (though still two basic parts)’ (p. 14, emphasis added). In cases where adjacency pair sequences come to have three turns, and when the third turn does not itself initiate an adjacency-pair sequence, the third turn is commonly an instance of what has been termed minimal expansion (Schegloff, 2007). Examples of minimal expansions include: 1) the change-of-state marker Oh, which embodies a claim to have been informed by the second-pair part action (Heritage, 1984a); 2) assessments (positive or negative) of the second-pair part action (Schegloff, 2007); and 3) the token Okay, which embodies a claim to accept the second-pair part action (Beach, 1993; Davidson, 1984). For example, see Extract 2 (which has been analyzed by both Davidson, 1984: 127, and Schegloff, 2007: 121). Ali initiates an adjacency-pair sequence by making an offer to Beth: ‘You wan’ me bring you anything?’ (line 1). Note that Extract 2 is comparable to Extract 1 in that both first parts are offers, and both offers are rejected.

Extract 2:

After Beth rejects the offer, ‘No: no: nothing.’ (line 3), Ali claims to accept the rejection with ‘AW:kay.’ (line 4; Davidson, 1984), which proposes closure of the adjacency-pair sequence of action (Davidson, 1984; Schegloff, 2007).

A comparison of Extracts 1–2 highlights the fact that similar types of first-part actions (e.g. offers) that get responded to in similar types of ways (e.g. rejected) sometimes get minimally expanded (as in Extract 2), but sometimes do not (as in Extract 1). Thus, at least some adjacency-pair based courses of action do not essentially involve three turns, and thus these third-turn actions are not always required to achieve course-of-action closure. This is apparently so even when second-part actions are dispreferred, as is the rejection in Extract 1 (line 6).¹ In sum, the production of a third-turn minimal expansion, which implements its own action, is frequently a contingent occurrence, for example, contingent upon the nature of the second-part action and/or its relationship to the first-part action.

Within CA, there is a bulk of research that describes the social organization of adjacency-pair based courses of action composed of two turns (Schegloff, 2007; Schegloff and Sacks, 1973), as well as that of third-turn actions (e.g. Oh, Good, and Okay) that can be characterized as minimal expansions of the basic, two-turn course of action (Schegloff, 2007). The present article attempts to enhance our understanding of a slightly different type of social organization, that of adjacency-pair based courses of action composed of three (vs two) turns, where the third-turn action is not a contingent expansion, but rather an essential part of the course of action. Following our main analysis, which is based on an institutional case study, we review our findings relative to the small amount of prior research on essentially three-turn courses of action in both ordinary and institutional contexts (Berry, 1981; Heritage, 1984a; Mehan, 1979). We ultimately argue that an understanding of essentially three-turn courses of action has the potential to expand that of central CA concepts involving sequence-based relevant rules, such as preference organization.

Data, method, and background

Data and method

Data were collected from the in-house repair facility of a mid-sized organization located in the northeastern United States that sells and repairs cameras and other electronic goods. We refer to this organization by the pseudonym Jack Camera. Data are audiotapes of 193 calls between one of five customer service representatives and customers calling to check on the status of equipment that they have previously sent in for repair. Our analytic method is CA (Atkinson and Heritage, 1984). All calls were transcribed in their entirety using Jefferson’s transcription system (Atkinson and Heritage, 1984).

Background

As background to the data, see Extract 3, which involves an entire call.

Extract 3 (Status 94):

The customer solicits status with ‘How do I check thuh status of that.’ (line 6). The representative’s status response is contingent on her locating the customer’s order in the computer system and, as is routine in these calls, her status response is delayed by an interrogative series (Zimmerman, 1992) in which she solicits and locates the customer’s repair authorization number (lines 8–23). The representative’s complete status response is at lines 26–30.

Data were collected as part of a five-month field observation of Jack Camera. Ethnographically, we know that, when customers send equipment in for repair, they are informed about the organization’s website and encouraged to check the progress of their repair online. This ethnographic claim is supported by the fact that customers sometimes orient to the website in their status-solicitation turns. For example, in Extract 5 (later), the customer says ‘I’m callin:g- ’cause I wanted tuh check on thuh status online . . .’ (lines 17–18). The information on the website contains one type of status – what we are calling current-location status – which involves information about the current location of the equipment in the repair process. This status includes a number of organizationally technical categories, such as ‘logged in’ (i.e. into the organization’s computer system), ‘in line’ (i.e. assigned to a technician), ‘on bench’ (i.e. on the technician’s workbench), ‘on hold’ (for parts), ‘quality control’, ‘shipping ready’, and optionally ‘manufacturer’ (i.e. the equipment was sent to its manufacturer to be repaired). After locating customers’ work orders, representatives’ computers display, among other things, current-location status, as seen in Figure 1 (which is a computer screen shot). In Extract 3, at lines 26–27, the representative produces current-location status: ‘It’s returned from thuh manufacturer as of t’day.’

Figure 1.

Computer screen shot of current repair status.

Upon calling, customers may (but only may) know about current-location status. Customers sometimes orient to knowing current-location status in their status-solicitation turns. For example, in Extract 3, as background to his request, the customer displays that he knows the web-based status of his laptop computer: ‘My- uhm laptop was se:nt onta thuh: Toshiba factory from- uh:m other r’pairs.’ (lines 4–5). Much more frequently, though, status-solicitation turns do not reveal whether or not customers know about current-location status. For example, customers’ entire status-solicitation turns can consist of: ‘Check on a repai:r plea:se’ (Extract 10, later, line 5).

There is a well-documented interactional rule that one should not ask for information that one already has (Heritage and Robinson, 2006; Terasaki, 2004). In the present context, customers are likely to know about current-location status from the website, and thus are likely to be soliciting some type of additional information. Due to this interactional rule, customers’ status solicitations are produced and understood as making conditionally relevant another type of status (i.e. beyond current-location status). As will be demonstrated, this additional status involves an estimated timeframe for when repairs will be completed or returned, or what we are calling repair-completion status. The provision of repair-completion status is thoroughly institutionalized, and this institutionalization is made evident at least by a whiteboard, which is prominently displayed to, and utilized by, all representatives. This whiteboard lists estimated repair-completion times for various pieces of equipment, and is pictured in Figure 2. In Extract 3, at lines 28–30, the representative produces repair-completion status, in this case by informing the customer when equipment will be returned: ‘So: it actually should be going back to you: .h uhm within thuh next two business days.’

Figure 2.

Whiteboard listing repair completion time for the equipment.

In sum, customers’ status solicitations, as first parts of adjacency-pair sequences, make conditionally relevant responses that contain two ordered types of status: 1) current-location status; and 2) repair-completion status (this claim is evident in Extracts 4–10). The possible completion of repair-completion status constitutes the possible completion of representatives’ responses to customers’ status solicitations.

Analysis

We argue that customers’ status inquiries initiate an essentially three-turn course of action that can be schematized as follows, and that is exemplified in Extract 3 (see ‘1->,’ 2->,’ and ‘3->’ in the left margin of the transcript):

Turn 1 CUS: Status solicitation

Turn 2 REP: Possibly complete response to status solicitation

Turn 3 CUS: Acceptance/rejection of status (acceptance preferred)

Our analytic focus is not on the sequential relationship between customers’ status solicitations (i.e. Turn 1) and representatives’ responses (i.e. Turn 2) (which is not unimportant, but which is supported by over 40 years of research on the adjacency-pair sequence (Schegloff, 2007), including extensive work in service interactions (Zimmerman, 1992)); rather, our analytic focus is on the sequential relationship between the possible completion of the adjacency-pair sequence represented by Turns 1 and 2, and customers’ acceptance/rejection of status represented by Turn 3. The following analysis contains three types of evidence for our argument: 1) cases in which, in Turn 3, customers do initially accept status; 2) cases in which, in Turn 3, customers do not initially accept status; and 3) a comparison of status solicitation, as a course of action, with others to demonstrate that Turn 3 is not a feature of all courses of action in the present customer-service context, but rather particular to that of status solicitation.

When customers initially accept status responses

In this section, we demonstrate that, when customers initially claim in Turn 3 to accept (or index their acceptance of) complete status responses, and do so in a relatively contiguous manner (Sacks, 1987), representatives orient to the possible completion of status solicitation as a course of action. Although these cases do not themselves prove our argument that participants orient to the third-turn action of acceptance/rejection as an essential component of status solicitation as a course of action, they represent one necessary type of evidence. For example, return to Extract 3. The representative produces a possibly complete status response across lines 26–30, ending with ‘days.’ (line 30). After a normal transition space (Jefferson, 1984), the customer produces, ‘Oh. okay.’ (line 31), where ‘Oh.’ claims a change in state from uninformed to informed (Heritage, 1984), and ‘okay.’ (line 31) claims to accept the status (Beach, 1993; Schegloff, 2007). The representative immediately orients to the relevance of possibly closing the call by initiating a pre-closing sequence: ‘Okay?’ (line 33; Schegloff and Sacks, 1973). In doing so, the representative treats the customer’s claim to have accepted the status response as indicating possible completion of the status-solicitation course of action.

For a second example, see Extract 4. The customer solicits status at lines 4–5: ‘Can I check thuh sta:tus on my repair.’

Extract 4 (Status 45):

The representative produces a possibly complete status response across lines 23–33 (including both equipment-location status and repair-completion status), ending with ‘ti:me.’ (line 33). After a normal transition space, the customer begins to speak at line 34. The nature of this talk is unclear (symbolized in the transcript by parentheses), and it is quickly cut off (symbolized in the transcript by the hyphen), which projects self correction (Schegloff et al., 1977). After a brief silence, the customer claims to accept the status response with ‘Okay.’ (line 34). Similar to Extract 3, the representative immediately orients to the relevance of possibly closing the call by initiating a pre-closing sequence: ‘Okay?’ (line 35; Schegloff and Sacks, 1973), which again treats the customer’s claim to have accepted the status response as indicating possible completion of the status-solicitation course of action.

There are other practices by which customers can index acceptance of status responses, such as positively assessing them. For example, see Extract 5.

Extract 5 (Status 183):

The representative produces a possibly complete status response across lines 37–40, ending with ‘da:ys.’ After a normal transition space, the customer produces ‘Oh:’ (line 41), which claims a change in state from uninformed to informed (Heritage, 1984a), and then indexes her acceptance of the status response by positively assessing it: ‘Coo:l.’ (line 41; Schegloff, 2007). As in Extracts 3–4, the representative immediately orients to the relevance of possibly closing the call by initiating a pre-closing sequence, ‘Oka:y?’ (line 42; Schegloff and Sacks, 1973), which treats the customer’s positive assessment as projecting possible completion of the status-solicitation course of action.

In contrast to Extracts 3 and 4, rather than moving to close the call, the customer expands (Schegloff, 2007) the status-solicitation course of action by seeking confirmation that her equipment has, in fact, been repaired (vs merely being returned). Specifically, the customer produces, for confirmation, a reformulation of an upshot (using the turn-initial So; Raymond, 2004) of the representative’s status: ‘So it’s- e- it’s fixed_ … then’ (lines 43–47). After a 0.2-second silence (line 45), which can be accounted for by the fact that the customer produces ‘fixed_’ (line 44) with final-level intonation (symbolized in the transcript by a black underscore), which can project turn incompletion (Ford and Thompson, 1996), the representative confirms with: ‘Yes.’ (line 46). This response constitutes another place where the status-solicitation course of action is possibly complete. The customer claims to accept the confirmed status with ‘Okay.’ (line 47), and then (re)assesses the confirmed status positively: ‘awe:some.’ (line 47). The representative immediately orients to the relevance of possibly closing the call by initiating a pre-closing sequence, ‘Okay?’ (line 48; Schegloff and Sacks, 1973), which treats the customer’s claim to have accepted the status as projecting possible completion of the status-solicitation course of action.

When customers do not immediately accept status responses

We argue that, upon possible completion of representatives’ status responses, customers’ third-turn claims of (non)acceptance are accountably necessary (vs optional), with acceptance being preferred. This argument was partially supported by Extracts 3–5 where, after possibly complete status responses, customers relatively quickly (i.e. contiguously; Sacks, 1987) claimed to accept status, and doing so was produced and understood as constituting possible closure of the status-solicitation course of action. Our argument is further supported by Extracts 6 and 7 (following) where, after representatives’ possibly complete status responses, customers produce third-turn actions that index customers’ lack of (complete) acceptance of the status response. As is typical of dispreferred responses, such actions make relevant sequence-expansion (Schegloff, 2007). As will be described in each extract, customers ultimately (i.e. after sequence expansion) claim to accept the status response, which is understood as constituting possible closure of the status-solicitation course of action. For a first example, see Extract 6.

Extract 6 (Status 57):

The representative produces a possibly complete status response across lines 15–20, ending with ‘unit.’ (line 20).² Unlike Extracts 3–5, in which customers proceeded to index their acceptance of the status response, here the customer produces ‘I see:.=h’ (line 22), which merely claims to have registered the status. Rather than initiating a pre-closing sequence, as was the case in Extracts 3–5, the representative treats the customer’s I see as not accepting the status response by expanding her response and clarifying the estimated timeframe for when the repair will be completed. This claim is elaborated immediately below.

Customers tend to erroneously assume that representatives’ turn-around-time estimates are to be calculated starting from the day of the status-solicitation call. However, from the institution’s perspective, turn-around-time estimates are to be calculated starting from the date that equipment was logged into the computer system, which is almost always prior to the day of the call. In Extract 6 (above), the representative clarifies this potential misunderstanding by informing the customer of the start date for the 10–20 day turn-around time: ‘(It) was from thuh day that it was entered in:, . . . <which was> o:n Monday.’ (lines 23–25). Insofar as this clarification indicates that several turn-around-time days have already elapsed, it slightly mitigates the bad news implicated by the status response (i.e. this customer will not receive his camera for three to four more weeks). In sum, the representative’s good-news clarification (at lines 23–25): 1) displays her orientation to the customer’s ‘I see:.=h’ (line 22) as indexing his lack of acceptance of the status response; 2) re-completes her status response, which has the effect of re-implicating its sequential relevance; and thus 3) is a practice for pursuing the customer’s acceptance of the status (Davidson, 1984; Pomerantz, 1984). In doing these things, the representative’s clarification displays her orientation to acceptance being due by the customer.

The representative’s clarification is possibly complete after ‘Monday.’ (line 25). At this point, the customer claims to accept the status with ‘O:kay.’ (line 26). Intonationally, this Okay is drenched with resignation, which is further evidence that it is being used to claim acceptance of the bad-news-implicative status. Similar to Extracts 3–5, the representative subsequently orients to the relevance of possibly closing the call by initiating a pre-closing sequence, ‘Oka:y?’ (line 27; Schegloff and Sacks, 1973). Extract 6 begins to highlight our argument that participants orient to the necessity of a third turn, and to it being organized in terms of a preference for acceptance (vs rejection).

For a second example, see Extract 7.

Extract 7 (Status 4):

The representative possibly completes her status response across lines 18–21.³ The customer indexes his lack of acceptance of the status by quickly and negatively assessing it: ‘Sh::ew Go:d.’ (line 22). After 0.4-seconds of silence (line 23), the representative recompletes her status response as an upshot (using the turn-initial So; Raymond, 2004): ‘So it’s still in li:ne.’ (line 24). Similar to the representative’s recompletion in Extract 6 (earlier, lines 23–25), here it re-implicates the sequential relevance of the status response and is a practice for pursuing the customer’s acceptance of the status. Space prohibits a detailed examination of the customer’s status-related complaints (and the representative’s responses) across lines 25–35. The customer ultimately produces an idiomatic expression of resignation regarding the equipment’s arrival date, ‘I guess it did what it did.’ (lines 36–37), which is a practice for proposing termination of ‘a topic that has involved a complaint or in which speakers have not fully aligned’ (Drew and Holt, 1988: 406; see also Drew and Holt, 1998).

Although, according to the rules for turn taking (Sacks et al., 1974), the enormous 3.2 seconds of silence at line 38 is a place where the representative has the right to speak, we argue that, according to the three-turn sequence organization of status solicitation as a course of action (i.e. its essential three-turn nature), this silence is analyzable as the representative waiting for the customer to claim acceptance of the status, which is due and preferred.⁴ The customer ultimately produces a version of Okay, ‘O::ka:y’ (line 39), which prosodically mitigates its claim to accept the status by being stretched (symbolized in the transcript by colons), and by being produced with final-level intonation, which is a practice for projecting turn incompletion (Ford and Thompson, 1996). Along these lines, the customer follows this Okay with another idiomatic expression involving resignation, ‘guess that’s (.) all we can do.’ (line 39), where ‘that’s’ refers to repair-completion status, and ‘guess’ mitigates his acceptance. The representative treats this idiomatic expression as a request for confirmation, and confirms with: ‘That’s all thuh status that we have as of right now.’ (line 40). After another short silence at line 41, where the representative is arguably waiting for the customer to claim (more wholehearted) acceptance of the status, the customer does so with: ‘O:kay.’ (line 42). The customer orients to this Okay as closing the status-solicitation course of action when he proceeds to thank the representative, ‘Thank you.’ (line 42), which initiates a pre-closing sequence (Zimmerman, 2006).

In Extracts 3–7, customers claims to accept (Extracts 3–5) or reject (Extracts 6 and 7) status responses were produced relatively contiguously (Sacks, 1987) to representatives’ status responses (i.e. within 0.2-seconds). If our argument is correct – that is if customers’ third-turn actions of acceptance/rejection are accountably necessary – then according to prior research on preference organization (Heritage, 1984b; Schegloff, 2007), marked breaks in such contiguity produced by, for example, relatively long silences, should project dispreferred responses, which in this case involve customers’ lack of acceptance of status. This argument is supported by Extracts 8–10. First, see Extract 8:

Extract 8 (Status 24):

The representative possibly completes her status response across lines 36–39, ending with ‘days’. Afterward, unlike Extracts 3–7, there occurs a marked (i.e. relatively long) 0.5-second silence at line 40. Prior to the customer beginning to speak – that is, simultaneous with the customer’s negative assessment ‘Wo:w.’ (line 42) – the representative produces an increment: ‘From thuh time that it’s logged into our system.’ (line 41). Very similar to the observations made about lines 23–25 in Extract 6, this increment clarifies the start date for the 10–20 day turn-around time and mitigates the bad-news implications of the status (i.e. this customer will not receive his computer for two more weeks). This increment: 1) re-completes the status response; which 2) re-implicates its sequential relevance and pursues the customer’s acceptance of status; and thus 3) displays the representative’s orientation to the preceding silence (at line 40) as projecting the customer’s lack of acceptance. This analysis is further supported by the fact that, simultaneous with the representative’s increment, the customer indexes his lack of acceptance of the status response by negatively assessing it: ‘Wo:w.’ (line 42).⁵

After the representative determines the log-in date of the customer’s computer (data not shown), she reformulates the repair-completion status, ‘they should be finishin’ up with it within: (.) two weeks.’ (lines 67–68), which the customer again negatively evaluates, ‘Wow. that’s a long time.’ (line 72), which again indexes his lack of acceptance of the status response. After another short silence at line 73, where the representative is arguably waiting for the customer to claim acceptance of the status, the customer does so with: ‘O(h)k(h)ay.’ (line 74). Although this version of Okay is infused with laughter, which arguably mitigates its claimed level of acceptance, the representative orients to it as possibly completing the status-solicitation course of action by initiating a pre-closing sequence (Schegloff and Sacks, 1973): ‘Okay?’ (line 75). Extract 8 provides evidence that delayed (i.e. non-contiguous; Sacks, 1987) third-turn actions are projectable as being rejections (and thus that rejections are dispreferred), that third turn actions are essential (vs contingent) parts of status solicitation as a course of action, and that the (preferred) third-turn action of acceptance generates the relevance of course-of-action closure.

For a second example, see Extract 9.

Extract 9 (Status 20):

The representative possibly completes her status response across lines 13–16, ending with ‘system’. Similar to Extract 8, there ensues a marked (i.e. long) 1.1-second silence at line 17, which we argue projects the customer’s lack of acceptance of status. Along these lines, rather than accepting status, the customer expands the status-solicitation course of action by producing, ‘So::: fro:m:.’ (line 18), which solicits a reformulation of the representative’s ‘from thuh date that it’s logged into our system’ (lines 15–16) in terms of the equipment’s log-in date, with which the customer can calculate an estimated return date.

Although the customer produces an extended in-breath at line 24, which can be a practice for projecting a turn of talk (Schegloff, 1996), there ensues an extremely long 5.5-second silence. Similar to the 3.2-second silence in Extract 7 (line 38), we argue that, due to the three-turn sequence organization of status solicitation as a course of action (i.e. its essential three-turn nature), this 5.5-second silence is analyzable as the representative waiting for the customer to claim acceptance of the status, which is due and preferred. This argument is again supported by the fact that the representative eventually re-issues (and thus re-completes) a version of her original repair-completion status: ‘That’s ten tuh twenny °uhm business days.°’ (line 25); doing so has the effect of re-implicating the relevance of acceptance of status, which the customer subsequently claims with: ‘Oka:y.’ (line 27). In this case, it is the customer who orients to his acceptance as closing the status solicitation course of action by progressing (Schegloff, 2007) to a new, future-oriented course of action involving whether or not his status will be updated on the website: ‘So: (are)=s there gunna be any upda:te?’ (line 29).

For a third example, see Extract 10:

Extract 10 (Status 13):

The representative completes her status response across lines 11–14, ending with ‘days’. As in Extracts 8–9, there ensues a marked 1.3-second silence at line 15, which we argue projects the customer’s lack of acceptance of status. This argument is supported by the fact that the customer proceeds to initiate repair (Schegloff et al., 1977). Specifically, he uses a partial-repeat-based practice, ‘Ten tuh twenny da:ys?’ (line 16; Schegloff et al., 1977), which is commonly used to challenge the ‘appropriateness’ of trouble sources (Jefferson, 1972), in this case the repair-completion status. The representative’s response, ‘Mm hm, yes.’ (line 17), which treats the customer’s practice of repair initiation as a mere request for confirmation, possibly closes the repair-initiation sequence, which in turn possibly closes the status solicitation course of action. We argue that the long, 1.4-second silence at line 18 is a place where the representative is again waiting for the customer to claim acceptance of status.

Space prohibits a detailed analysis of lines 19–27, where, rather than claiming acceptance of status, the customer continues to request confirmation of several additional pieces of repair-related information. The representative’s confirmation at line 28, ‘Mm hm,’ possibly closes the prior sequence, which in turn possibly closes the status solicitation course of action. The marked 1.2-second silence at line 29 is analogous to those at lines 15 and 18, and is a place where the representative is again waiting for the customer to claim acceptance of status. Relative to his previous initiation of repair, ‘Ten tuh twenny da:ys?’ (line 16), at line 30 the customer re-registers his receipt of the repair-completion status: ‘It takes ten tuh twenny da:ys.’ (line 30). This action is produced in a sequentially third-turn position to merely claim registration of the repair-completion status (Schegloff, 1996) and is neither produced nor understood as soliciting a response from the representative. This claim is supported by the fact that the representative does not produce a response through the brief silence at line 31, and that the customer goes on to claim acceptance of the status: ‘Okay,’ (line 32). As in Extract 7 (line 42), the customer orients to this Okay as closing the status-solicitation course of action when he proceeds to thank the representative, ‘Thank you.’ (line 34), which initiates a pre-closing sequence (Zimmerman, 2006).

Comparison of status-solicitation with other types of courses of action

Two observations emerge when status solicitation, as a course of action, is compared to others in the present institutional context. First, it is not the case that customers’ third-turn claims of acceptance (e.g. Okay), per se, are essential components of all adjacency-pair-based courses of action in this context. That is, for other courses of action, the third-turn action of acceptance/rejection is not necessary, per se, and alternative third-turn actions can suffice. For example, take the case of the course of action initiated by customers’ solicitations of store hours (vs repair status), as seen in Extract 11: ‘What’re your ho:urs,’ (line 4).

Extract 11:

At the possible completion of the representative’s response across lines 6–8 (i.e. after ‘two.’), the customer relatively quietly (symbolized in the transcript by degree signs) repeats the last portion of the reported time frame, and completes it with final-falling intonation, ‘°Nine thirty tuh two.°’ (line 10), which is a type of minimal expansion (Schegloff, 2007) that is a practice for claiming registration of information (Schegloff, 1996). The customer then immediately orients to the possible completion of the hour-solicitation course of action by thanking the representative: ‘thank you very much.’ (line 10; Zimmerman, 2006).

Second, there is also evidence that not all adjacency-pair-based courses of action in this context are essentially composed of three turns. For example, see Extract 12, wherein the customer initiates a course of action by asking: ‘What time do you close toda:y?’ (line 53).

Extract 12:

At the possible completion of the representative’s response, ‘Six pee em.’ (i.e. 6:00 pm, line 54), rather than producing any type of minimal expansion – such as Okay, Cool, I see, or some claim of information registration (as we saw in Extract 11, line 10) – the customer initiates a pre-call-closing sequence with: ‘Thank you very much.’ (line 55). Note that this is treated as such by the representative, who responds to it with ‘No problem.’ (line 56), after which the participants close the call (lines 57–58).⁶ This pattern is evident in other customer-service contexts, such as calls to emergency assistance (Zimmerman, 2006).⁷

Discussion

A description of the social structures that organize action is the raison d’etre of CA (Heritage, 1984b). One of the most common and important structures is the adjacency-pair sequence (Schegloff, 2007), the associated rules of which provide much of the foundation for the accountability (Garfinkel, 1967) of social action. This article extends research on the basic organization of the adjacency-pair sequence by exploring the social organization of a course of action that is essentially (vs contingently) composed of three (vs two) turns. The bulk of this article describes (and defends the description of) one such course of action in the institutional context of customers calling an electronics repair facility to request the status of equipment they have previously sent in for repair. The first two turns of this course of action involve an adjacency-pair sequence initiated by customers’ solicitations of status (Turn 1), and answered by representatives’ status responses (Turn 2). We argue that, upon the possible completion of the adjacency-pair sequence represented by Turns 1 and 2, a third-turn action by customers becomes accountably necessary, by which we mean that if it does not occur, it is ‘relevantly absent’ (Schegloff, 2007: 20) and its absence is accountable (Garfinkel, 1967). In other words, the non-occurrence of the third-turn action ‘is as much an event as its occurrence would have been. It is, so to speak, noticeably, officially, consequentially, absent’ such that resulting silences can ‘be characterized (and, in the first instance, heard) specifically for who is not talking, and what kind of talk they are not doing’ (Schegloff, 2007: 20, emphasis in original).⁸ The relevant third-turn action involves two alternative and asymmetrical types of action, including customers’ claims to either accept or reject representatives’ status responses. Acceptance and rejection embody alternative alignments toward representatives’ status responses, and we argue that acceptance is preferred (re preference organization, see Schegloff, 2007).

Before situating and explaining our findings, we want to address a potential counter-argument, which is that a customer’s third-turn action of acceptance/rejection is not an essential feature of the course of action as a whole (which is initiated by the customer’s status inquiry), but rather a contingent ‘response’ to the representative’s second-turn action of delivering (or giving) information (i.e. equipment status) that was previously unknown to customers (i.e. that representatives’ second-part responses are simultaneously sequence-initiating actions). A first observation is that the representative’s second-part action is indeed one of giving novel information. This observation is supported by the fact that some customers do initially respond in third turn with the token Oh, which primarily functions to index the receipt of news (e.g. Extract 3, line 31; Extract 5, line 41; and Extract 8, line 44; Heritage, 1984a). A second observation, though, is that customers orient to a distinction between the action of indexing the receipt of news and that of indexing acceptance/rejection of such news. For example, in Extracts 3, 5, and 8, the customer immediately proceeds from Oh to either Okay or a positive assessment (i.e. Oh okay or Oh cool), which index their acceptance of status. Thus, customers’ third-turn acceptances do not appear to be produced as responses to representatives’ second-turn deliveries of news, per se. A third and final observation is that, as a third-turn action, indexing the receipt of news is not itself treated as a sufficient response to the second-turn action of status delivery. For example, in Extract 6, when the customer initially responds only with ‘I see:’ (line 22), the representative pursues the customer’s acceptance of status, which the customer ultimately produces, ‘O:kay’ (line 26), and only then does the course of action get closed (see earlier analysis).

Toward the goal of situating our finding, prior research corroborates the existence of essentially three-turn, adjacency-pair-organized, courses of action. Perhaps the best (i.e. most well defended) candidate is the 1) known-information-question; 2) answer; 3) acceptance/rejection of answer sequence, referred to by Mehan (1979) as the question-answer-evaluation (QAE) sequence between teachers and students during classroom interaction (see also Sinclair and Coulthard, 1975; Tsui, 1985). Two examples of QAE sequences are provided in Extract 13 (which was analyzed by Mehan, 1979: 290).⁹

Extract 13:

Although not specifically formulated by Mehan (1979) in the following terms, he provided evidence suggesting that the third-turn ‘evaluation’ action (see arrow 3, above) is an accountably necessary feature of the QAE course of action. That is, regardless of whether students’ second-part answers are correct or incorrect, teachers are strongly accountable for somehow claiming acceptance/rejection of such answers. Furthermore, Mehan provided evidence suggesting that teachers’ third-turn claims of acceptance of students’ answers (e.g. Very good and Okay) are preferred over those of rejection (e.g. No, try again).

Mehan’s findings emerged from a particular form of institutional interaction, and defensible examples of essentially three-turn courses of action in ordinary conversation are virtually absent (but see Berry, 1981; Jefferson and Schenkein, 1978). One exception is found in Heritage’s (1984a) conclusion to his article involving the third-turn action Oh where, based on substantial evidence, he proposed:

. . . that the action of questioning is not only or fully accomplished within the span of a single utterance. Rather . . . the action of questioning is, even in the simplest cases, the reciprocal achievement of two turns in a sequence having, at the minimum, a prototypical Q-A-‘Oh’ structure. Moreover, questions and answers are themselves the media through which a variety of activities are transacted in conversation, and choices among the (third-turn) receipt objects discussed in this chapter play a considerable role in the determination of what these activities have been, or will come to. Included within this scope of determination will be whether a syntactically formed question was produced so as to accomplish questioning. (p. 336, emphasis added)

Based on Heritage’s (1984a) observation reflected in the above quote, our discussion now begins to outline a possible general explanation – one that applies to more than one particular course of action, and one that applies to both institutional and ordinary contexts – for why some first-part actions initiate essentially three-turn (vs two-turn) courses of action. Based on this explanation, we offer a candidate example of an essentially three-turn course of action in ordinary conversation.

There are reasons why, in its minimal, basic, unexpanded form, the adjacency-pair sequence can consist of two turns. The rules associated with many initiating (i.e. first-part) actions – for example, rules of action formation, conditional relevance, and preference organization – provide participants with resources for clearly understanding the nature of the first-part action and what any given second-part action will ‘mean’ relative to the first-part action. Because these resources are social, upon possible completion of second-part actions participants typically share a common understanding (i.e. intersubjectivity) regarding ‘what has happened’, and how those ‘happenings’ are likely to be interpreted by participants. For example, in Extract 1, Mark’s ‘want some more milk?’ (line 5) provides Kim with resources for understanding that an offer is in play, that an accepting/declining response will align/disalign with the offer, and that an acceptance/declination is preferred/dispreferred, respectively. In sum, the rules associated with many first-part actions virtually ensure intersubjectivity (Heritage, 1984b; Schegloff, 1992), and this explains why many adjacency-pair-based courses of action can be essentially two-turn (vs three-turn) sequences. As Heritage and Atkinson (1984) noted over 25 years ago, it is because of these rules/resources that any third turn (i.e. the turn after the response) that progresses interaction forward (Schegloff, 2007) by initiating some new course of action ‘confirms the adequacy of the displayed understandings in the sequence so far. By means of this framework, speakers are released from what would otherwise be the endless task of explicitly confirming and reconfirming their understandings of one another’s actions’ (p. 10, emphasis added). In Extract 1, although Kim declines Mark’s offer, she understands what a declination ‘means’ for Mark; stated negatively, there is virtually no ambiguity regarding ‘what just happened’, and what that ‘means’ for Mark.

However, some first-part actions do not – or at least do not as clearly – provide respondents with resources for understanding the (full or complete) nature of the first-part action and what any given second-part action will ‘mean’ relative to the first-part action. For example, return to Extract 13 and consider the teachers’ known-answer question: What does this word say? This question does not merely make conditionally relevant a particular type of answer (i.e. a word), but one that is valenced in terms of its ‘correctness’ or ‘incorrectness’. By design, this questioning action does not provide the student respondent (i.e. Beth) with resources for understanding what her answer means for the teacher in terms of its correctness/incorrectness. For another example, as initially suggested by Heritage (1984a; see extended quote above), syntactically formed interrogatives are often not clearly projectable as mere (i.e. genuine, or neutral, or unmotivated) request-for-information actions. (This issue was reviewed by Steensig and Drew, 2008, who noted that syntactically formed interrogatives are commonly vehicles for other/additional actions.) Although a syntactically formed interrogative makes conditionally relevant a particular type of answer (e.g. Yes/No, or a Who- When-, Where-based answer), and may further grammatically prefer a certain class of answer (e.g. Yes/No; Raymond, 2003), it does not always provide respondents with resources for clearly understanding the (full or complete) nature of the initiating action, and thus what any given second-part action will ‘mean’ relative to the first-part action.

In sum, compared to some actions, for example Mark’s offer in Extract 1, upon completion of other actions – such as teachers’ known-answer questions and some syntactically formed interrogatives – intersubjectivity appears to be less guaranteed. For example, relative to Extract 1, where Kim can be fairly confident about what her declination means relative to Mark’s offer, Beth in Extract 13 cannot be (as) confident about what her answer means in terms of its correctness/incorrectness. It appears then, that, compared to some first-part actions, other such actions, such as teachers’ known-answer questions, are projectable from their outset as posing relatively more potential barriers to intersubjectivity upon the completion of their second-part actions. Insofar as interaction is built to promote intersubjectivity (Schegloff, 1992), this may explain why some first-part actions set in motion essentially three-turn (vs two-turn) courses of action, where the first-part action makes conditionally relevant a second-part action, and the possible completion of the second-part action makes a third-turn action accountably necessary. This third-turn action fulfills the function, among others, of ratifying intersubjectivity regarding the nature of the relationship between the second-part action and the first-part action. To borrow Heritage and Atkinson’s (1984) language, in the case of essentially three-turn courses of action, upon completion of second-part actions, speakers of first-part actions are not ‘released’ from the task of confirming respondents’ understandings revealed by second-part actions.

We can now return to customers’ first-part status solicitations, which are similar to teachers’ known-answer questions in the following way: customers’ status solicitations do not merely solicit a ‘neutral’ piece of information, but rather one that is possibly valenced in terms of its good or bad news for customers, and thus one that is possibly ‘unacceptable’ to customers. This claim is supported by the facts that: 1) customers’ third-turn actions frequently treat representatives’ second-part deliveries of repair status as constituting either good news (e.g. ‘Oh: coo:l.’; Extract 5, line 41) or bad news (e.g. Extract 7, line 22: ‘Sh::ew Go:d;’ Extract 8, line 72: ‘Wow that’s a long time’); and 2) representatives sometimes explicitly treat their deliveries of repair-status in terms of it being good or bad news for customers (e.g. Extract 5, line 42: ‘Oka:y? good ne:ws’). In a similar way that teachers’ known-answer questions typically do not provide student-respondents with resources for (definitively) understanding whether or not their answers are correct/incorrect, customers’ first-part status solicitations typically do not provide representatives with resources for (definitively) understanding whether or not their second-part status responses will be acceptable/non-acceptable to customers. Thus, similar to teachers’ third-turn evaluations, customers’ third-turn claims of acceptance/non-acceptance function to ratify intersubjectivity regarding the ‘meaning’ of the second-part action.

First-part actions that set in motion essentially three-turn courses of action can be found in ordinary (i.e. non-institutional) contexts. For example, take situations where two people are attempting to arrange a time/date to meet. An initial solicitation by Speaker A of a candidate time/date that ‘works’ for Speaker B does not provide Speaker B with resources for understanding (at least with full confidence) whether or not their proposed time/date will ‘work’ for Speaker A. There is evidence that these types of solicitations set in motion essentially three-turn courses of action, including Speaker A’s solicitation, Speaker B’s response, and Speaker A’s claim to accept/reject the response (with acceptance preferred). For instance, see Extract 14, which is drawn from a British telephone call between Dana and her boyfriend Gordon:

Extract 14:

After Dana’s offer to visit Gordon at his home is accepted (lines 1–2), Dana inquires into an acceptable visitation time: ‘’bout what time.’ (line 4). Dana’s solicitation initiates an arrangement-making action that appears to require, for its ‘completion’, not just a time, but one that ‘works’ for both parties’ schedules. Dana’s first-part solicitation does not provide Gordon with resources for understanding whether or not his second-part response, ‘In the afternoon.’ (lines 5–6), ‘works’ for Dana. Along these lines, after Gordon’s proposed time (lines 5–6), Dana claims to accept it with: ‘Okay,’ (line 8).

Perhaps the closest institutional parallel to our finding involves the relationship between primary care patients’ establishment of new medical problems (i.e. Turn 1; Robinson, 2006), which sets in motion a project of activity that is possibly completed by physicians’ deliveries of treatment recommendations (i.e. Turn 2; Robinson, 2003), which make accountably necessary patients’ (non)acceptance of such recommendations (i.e. Turn 3; Stivers, 2005; Ten Have, 2006). On the one hand, there is not (yet) a clear, and certainly not a direct, adjacency-pair relationship between the establishment of a new problem and a treatment recommendation, and thus it is unclear if this parallel finding constitutes an appropriate comparison (e.g. it is possible that physicians’ treatment recommendations are first-part actions, and thus that patients’ indices of (non)acceptance are second-part actions). On the other hand, it is interesting to note that, as an initiating action, the establishment of a new medical problem is similar to customers’ status requests and some arrangement-making actions (as in Extract 14, above) in that it solicits a response (i.e. a treatment recommendation) that is possibly ‘unacceptable’ to the producer of the action (i.e. the patient), yet simultaneously does not provide recipients (e.g. physicians) with resources for clearly determining such acceptability. It is this type of situation in which we might expect the accountable necessity of a third-turn index of (non)acceptance, which would function to ratify participants’ intersubjectivity on the matter.

The present article suggests a relatively new agenda for CA involving the description of the social structure of essentially three-turn courses of action. An important part of this agenda involves the relevance rules that apply to the relationship between the possible completion of an adjacency-pair sequence and a third-turn action, which are likely to reflexively shape the production and understanding of first-turn and second-turn actions.

Footnotes

Notes

References

Atkinson

Heritage

(eds) (1984) Structures of Social Action. Cambridge: Cambridge University Press.

Beach

(1993) Transitional regularities for ‘casual’ ‘Okay’ usages. Journal of Pragmatics 19: 325–352.

Berry

(1981) Systemic linguistics and discourse analysis: A multi-layered approach to exchange structure. In: Coulthard

Montgomery

(eds) Studies in Discourse Analysis. London: Routledge & Kegan Paul, pp. 120–145.

Davidson

(1984) Subsequent versions of invitations, offers, requests, and proposals dealing with potential or actual rejection. In: Atkinson

Heritage

(eds) Structures of Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press, pp. 102–128.

Drew

Holt

(1988) Complainable matters: The use of idiomatic expressions in making complaints. Social Problems 35(4): 398–417.

Drew

Holt

(1998) Figures of speech: Figurative expressions and the management of topic transition in conversation. Language in Society 27: 495–522.

Ford

Thompson

(1988) Interactional units in conversation: Syntatic, intonational, and pragmatic resources for the management of turns. In: Tannen

(ed.) Advances in Discourse Processes, Vol. XXIX. Norwood, NJ: Ablex, pp. 134–183.

Ford

Thompson

(1996) Interactional units in conversation: Syntactic, intonational, and pragmatic resources for the management of turns. In: Ochs

Schegloff

Thompson

(eds) Interaction and Grammar. Cambridge: Cambridge University Press, pp. 134–184.

Garfinkel

(1967) Studies in Ethnomethodolgy. Cambridge: Polity Press.

10.

Heritage

(1984a) A change-of-state token and aspects of its sequential placement. In: Atkinson

Heritage

(eds) Structures of Social Action. Cambridge: Cambridge University Press, pp. 299–345.

11.

Heritage

(1984b) Garfinkel and Ethnomethodology. Cambridge: Polity Press.

12.

Heritage

Atkinson

(1984) Introduction. In: Atkinson

Heritage

(eds) Structures of Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press, pp. 1–15.

13.

Heritage

Robinson

(2006) The structure of patients’ presenting concerns: Physicians’ opening questions. Health Communication 19: 89–102.

14.

Jefferson

(1972) Side sequences. In: Sudnow

(ed.) Studies in Social Interaction. New York: Free Press, pp. 233–294.

15.

Jefferson

(1984) Notes on some orderlinesses of overlap onset. In: D’Urso

Leonardi

(eds) Discourse Analysis and Natural Rhetorics. Padua: Cleup Editore, 11–38.

16.

Jefferson

Schenkein

(1978) Some sequential negotiations in conversation: Unexpanded and expanded versions of projected action sequences. In: Schenkein

(ed.) Studies in the Organization of Conversational Interaction. New York: Academic Press, pp. 155–172.

17.

Mehan

(1979) Learning Lessons: Social Organization in the Classroom. Cambridge, MA: Harvard University Press.

18.

Pomerantz

(1984) Pursuing a response. In: Atkinson

Heritage

(eds) Structures of Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press, pp. 51–101.

19.

Raymond

(2003) Grammar and social organization: Yes/no interrogatives and the structure of responding. American Sociological Review 68(6): 939–967.

20.

Raymond

(2004) Prompting action: The stand-alone ‘so’ in sequences of talk-in-interaction. Research on Language and Social Interaction 37(2): 185–218.

21.

Robinson

(2003) An interactional structure of medical activities during acute visits and its implications for patients’ participation. Health Communication 15: 27–59.

22.

Robinson

(2006) Soliciting patients’ presenting concerns. In: Heritage

Maynard

(eds) Communication in Medical Care: Interaction between Primary Care Physicians and Patients. Cambridge: Cambridge University Press, pp. 22–47.

23.

Sacks

(1987) On the preferences for agreement and contiguity in sequences in conversation. In: Button

Lee

JRE

(eds) Talk and Social Organization. Philadelphia, PA: Multilingual Matters.

24.

Sacks

Schegloff

Jefferson

(1974) A simplest systematics for the organization of turn-taking for conversation. Language 50(4): 696–735.

25.

Schegloff

(1992) Repair after next turn: The last structurally provided defense of intersubjectivity in conversation. American Journal of Sociology 97(5): 1295–1345.

26.

Schegloff

(1996) Confirming allusions: Toward an empirical account of action. American Journal of Sociology 102(1): 161–216.

27.

Schegloff

(2007) Sequence Organization in Interaction: A Primer in Conversation Analysis. Cambridge: Cambridge University Press.

28.

Schegloff

Sacks

(1973) Opening up closings. Semiotica VIII(4): 289–327.

29.

Schegloff

Jefferson

Sacks

(1977) The preference for self-correction in the organization of repair in conversation. Language 53: 361–382.

30.

Sinclair

Coulthard

(1975) Towards an Analysis of Discourse: The English Used by Teachers and Pupils. London: Oxford University Press.

31.

Steensig

Drew

(2008) Introduction: Questioning and affiliation/disaffiliation in interaction. Discourse Studies 10: 5–15.

32.

Stivers

(2005) Parent resistance to physicians’ treatment recommendations: One resource for initiating a negotiation of the treatment decision. Health Communication 181(1): 41–74.

33.

Stivers

(2008) Stance, alignment and affiliation during storytelling: When nodding is a token of affiliation. Research on Language and Social Interaction 41(1): 31–57.

34.

ten Have

(2006) On the interactive constitution of medical encounters. Revue Française de Linquistique Appliquée 11(2): 85–95.

35.

Terasaki

(2004) Pre-announcement sequences in conversation. In: Lerner

(ed.) Conversation Analysis: Studies from the First Generation. Amsterdam and Philadelphia, PA: John Benjamins, pp. 174–223.

36.

Tsui

ABM

(1985) Beyond the adjacency pair. Language in Society 18(4): 545–564.

37.

Zimmerman

(1992) The interactional organization of calls for emergency assistance. In: Drew

Heritage

(eds) Talk at Work: Interaction in Institutional Settings. Cambridge: Cambridge University Press, pp. 418–469.

38.

Zimmerman

(2006) How closing matters in emergency calls. Paper presented at the annual meeting of the American Sociological Association, Montreal, Quebec, Canada.