Environmentally coupled repairs and remedies in the airline cockpit: Repair practices of talk and action in interaction

Introduction

This article explores the ways in which talk and action maintain the intersubjective sense of action being consequential for flying aircraft. The article aims to extend the analysis of repair from troubles in the production and comprehension of talk to troubles in the coordination of talk and action, which also contributes to knowledge of prevention of flight incidents and accidents (Arminen et al., 2010; Auvinen, 2009; Nevile, 2007; Nevile and Walker, 2005). Faulty actions and the loss of shared awareness between flight crew members cause a risk of aviation mishaps or hazards (Dekker, 2006; Reason, 1997). Pilots employ several modalities simultaneously in locating and managing problems in the course of the cockpit interaction, including verbal repair and reminder, pointing gestures, gaze movements, upper body orientations and object manipulations in the local environments.

Traditional research on the organization of repair has been concerned with problems in speaking, hearing and understanding talk (Jefferson, 1987; Schegloff, 1979, 1992; Schegloff et al., 1977). Multimodal conversation analysis has also explored how corrective actions target the participants’ physical courses of action, their interactions with the material environment and their uses of multimodal semiotic facilities (Goodwin, 1994; Koschmann et al., 2007; Nevile, 2004a; Suchman, 1987; Zemel et al., 2011). We continue broadening the scope of multimodal conversation analysis by pointing out that the repair practices are used to manage not only the linguistic problems of participants, but also their problems in positioning a move, utterance or action in interaction, which also makes salient to re-examine the notion of action employed in conversation analytical studies and to reconsider the still implicitly prominent role of speech act theory.

We show that the problems pilots encounter in cockpit interaction are not simply related to speaking, hearing or understanding cockpit talk, but to the sequential coordination of talk and action. The problems are thus visible and analyzable in the practical accomplishment of verbal and non-verbal tasks and activities or in their positioning. For instance, a pilot may use the wrong term due to an irrelevant source of data or s/he may initiate a particular flight task prematurely (Auvinen, 2009; see also Arminen, 2005b). The various problems of coordination are located and managed with the repairs of talk and remedies of action that are environmentally coupled (Goodwin, 1994, 2007), that is, understandable and intelligible relative to each other and to the sequential and physical environment in which they occur.

Environmental coupling of action is also the backbone for Nevile’s paradoxical repairs in which there is no repairable talk:

It seems in the examples here that there is something being repaired, though in conversation analysis terms there is no trouble-source turn, no actual problem of speaking or hearing or understanding an utterance: Nobody is producing or receiving problematic talk. (Nevile, 2007: 247)

Nevile (p. 247) points out that ‘what is corrected here is not the structure and meaning of an emerging turn or adjacency pair sequence, but the structure of an extended and jointly produced agenda-based sequence of actions’. Remedies can concern larger collaborative projects of sequenced actions even when talk in interaction does not include repairables.

Environmental coupling is generically relevant for the organization of repairs and remedies. Let us have a look at the following example (Schegloff, 1992: 1321 [EAS, FN]). Marty, a visiting researcher, makes a request to Loes, who is the receptionist and keeps the supplies.

1 Marty: Loes, do you have a calendar,

2 Loes: Yeah ((reaches for her desk calendar))

3 Marty: Do you have one that hangs on the wall?

4 Loes: Oh, you want one.

5 Marty: Yeah

In her verbal response (line 2), Loes shows compliance with the request, but her embodied action displays her misunderstanding for Marty, who initiates a remedy at the third position (line 3). Although Loes’s verbal response is not repairable, her embodied response shows that she had misconstrued Marty’s action as a pre for an information-seeking question instead of a request. This is parallel to our example 3, in which the co-pilot’s (COP) embodied response indicated that he was orienting to a different activity phase than the commander (CDR). For Schegloff, the example is problematic as he has decoupled talk and embodied action and is unable to address remedial work as embodied action in its activity context. Schegloff (1997) has since discussed repairs on a sequence’s proper development or progressivity, which might have opened up the right track here.

Social action and materiality

In the mid-1960s, before what became ethnomethodology (EM) and conversation analysis had split, Harvey Sacks (1992a, 1992b), the co-founder of conversation analysis (CA), envisioned a science of social life (Arminen, 2008). Sacks’s idea of the science of social life was to reconstruct and analyze the practices that permit members of society to see and grasp things the way they do (Silverman, 1998). For Sacks, the key was to grasp how people see the world: ‘We know that society has power over our minds. It is an invitation to us, social scientists and others, to understand any way of life by our knowledge of our ways’ (Sacks, 1992a: 486). This invitation to work out what the world is made up of has potential for a general theory that provides a ‘praxiological’ view of societal life and social action. Whatever humans do can be examined to discover and describe the way it is carried out. Sacks’s ultimate goal was to build a science that could deal with the actual details of actions in order to reverse engineer the constitutive elements of the phenomena in society.

Sacks’s vision of the science of social life is compatible with the current emerging material social science; they are both strictly praxiological and finitist programs. For both of them, there is no external cause of social actions. The actual situated day-to-day social practices form the foundation for all institutionalized practices and societal life. At the root of the finitist program is an analysis of the way in which people classify activities, entities and states of all kinds: objects, living beings, processes, circumstances, situations, and so on (Hatherly et al., 2008), resembling Sacks’s profound interest in categories and categorizations (e.g. lectures in 1965–1967, in Sacks, 1992a). Sacks also follows the finitist cause of reconstructing the elementary details of how the mind works on the basis of what can be observed. Both approaches end up reverse engineering the constitutive elements of the phenomena explored.

Currently, the multimodal conversation analysis comes closest to systematic exploration of situated social action in its socio-material context. The interactive organization of embodied action, in which multiple parties are building courses of action together while attending to each other, the details of the talk, and the structure of their environment forms a primordial site for the social organization and the development of knowledge and cognition (Goodwin, 2000; see also Heath and Luff, 2000; Streeck et al., 2011). We utilize the multimodal conversation analysis embedded in material social science to develop a context-sensitive research strategy that amounts to rigorous, detailed findings on social actions in their socio-material context.

Repairs and remedies: Environmental coupling of talk and action

Repairs are an organized set of practices for dealing with problems in speaking, hearing and understanding talk in a conversation that solve communicative problems and restore intersubjective understanding (Drew, 1997; Jefferson, 1987; Schegloff, 1979). The organization of repair falls into two basic dimensions. The first concerns who initiates repair, either the producer of the trouble-source or a recipient – self-initiation and other-initiation respectively. The second dimension concerns the sequential location of repair initiation. Virtually all repair initiations are launched in windows of opportunity following the trouble-source (Schegloff et al., 1977). Repair is composed of the initiation marking possible disjunction with the previous talk and the production of repair, solving or abandoning the problem. The trouble which the repair addresses is referred to as the trouble-source or repairable. In all, there have been a massive number of CA studies on various kinds of repair practices (e.g. Bolden, 2012; Drew, 1997; Kurhila, 2001; Macbeth, 2004; McHoul, 1990).

The studies on repairs were initially concerned with problems in speaking, hearing and understanding talk, but the area of investigation has started to expand. As soon as visual aspects of interaction were taken into consideration, it was noticed that gaze and repair are interconnected. Goodwin (1980, 1981, 1986) has studied the ways in which a speaker monitors the recipients’ attention to talk, and established that there is a constellation between gazes and certain types of self-repair. The recipients may indicate attention by gazing at the speaker. When a recipient’s gaze is lost, the speaker may employ remedial actions to re-establish gaze orientation. Egbert (1996) demonstrated a systematic tie between gaze orientations and certain types of repair, concluding that ‘repair is a social action that can link talk and actions of individual persons in a joint focus of attention’ (Egbert, 1996: 609).

Repairs are often also accompanied by gestures. Streeck (1984, 2009) has studied contextualization of talk through embodiment and postural configurations. Gestures are linked both to gazes and repairs. They can invite attention, thereby drawing gazes and initiating repairs. Further, embodied actions are continuously monitored just like talk, and the positioning of gestures is finely tuned to the sequential orchestration of other modalities of interaction in remedial exchanges as well. Gestures and other embodied activities can highlight the trouble source without necessarily interrupting the ongoing verbal action, by allowing the recipient of the gesture to identify and resolve the trouble (Greiffenhagen and Watson, 2009; Nevile, 2007).

Corrective practices may also target the physical actions of those interacting with or without verbal activities. Goodwin has studied corrective practices from an educational interaction on an archaeological site to young girls playing hopscotch (Goodwin, 1994, 2000). At an archeological site, an archeologist can be shown to monitor her co-worker’s movements and tool use as a part of their collaborative activity, and launch a verbal repair immediately when her co-worker’s physical actions seem to be directed to breaching the shared horizon of their activity. The correction deters the coworker from undertaking a physical course of action considered inappropriate (Goodwin, 2000). Similarly, McHoul (1990), in his study of classroom repairs, analyzes an instance in a geography lesson where a large-scale hand gesture indexing the shape of a territorial boundary occasioned a remedy.

Human–machine interactions and computer supported co-operative work often seem prone to troubles that have to be remedied for the action to continue. Suchman (1987) pioneered this field with her study on the intersubjective problems in users’ interactions with a photocopier. She noticed that the users took the unresponsiveness of the machine as evidence of trouble in their performance of the action taken. To resolve the problem, the users re-inspected the instructions on the machine’s display, re-identified the object of the instruction and corrected their action. Spagnolli et al. (2002) have studied failures in user interactions with the virtual library, analyzing the particular breakdown episodes where the user was forced to abandon a projected course of action and mobilize resources to obtain a more effective one. They argue that the breakdowns in interaction with an artefact reveal the aspects of a system-in-use that hamper the successful completion of an action. Recently, there have been numerous studies concerning problems in human–machine interactions and computer supported co-operative work (e.g. Arminen, 2005b; Arminen and Poikus, 2009; Gardner and Levy, 2010; Greiffenhagen and Watson, 2009; Koskela and Arminen, 2012; Meddeb and Frenz-Belkin, 2010; Nevile, 2009).

Physical actions can also be remedial or corrective. Martin’s (2004) study of physiotherapy pointed out that the therapist could launch a remedy for the patient’s posture and movements by tapping him on the shoulder. The patient has learned a meaning for a particular physical touch as part of the physiotherapeutic process. Similarly, Nevile (personal communication) has described a case in the airline cockpit where a commander launched a remedy just by touching the co-pilot on the arm, and saying nothing. Multimedia augmented communication and collaborative computing are the currently expanding fields in which new repair practices are emerging. Augmentative multimedia can be employed to achieve repair for a speaker who exhibits limitations in producing verbal talk (Aaltonen et al., forthcoming; Wilkinson et al., 2011).

Finally, Nevile (2004a) has opened airline cockpit interaction and its repair practices up for research. His studies touch upon the problematic understandings in routine flights, revealing cases in which airline pilots were having trouble in the production of verbal and/or physical conduct during actual flights, and pointing out their repair practices (Nevile, 2004a; see also Nevile and Walker, 2005). Nevile (2007) also systematically extends the analysis of repair from trouble sources in talk to troubles in timing of action. In the cases considered, a cockpit action is timely and relevant, but not initiated by the pilot responsible for doing so. The pilot not responsible for initiating an action repairs an action’s untimeliness through talk, and the pilot responsible subsequently produces the talk that officially initiates the action. These cases address collaborative actions that are not talk-in-interaction, but part of the broader sequential course of action. As Nevile (2007: 248) argues: ‘Here we have something like a ‘‘repair of a sequence’s proper development’’ (Schegloff, 1997: 510, citing Jefferson, 1981), and ‘‘a repair of a sequence’s progressivity’’ (Schegloff, 1997: 512), but for a larger collaborative project of sequenced actions for tasks.’

The notion of ‘environmentally coupled repair’ is used in the analysis of the organization of talk-in-interaction, that is, pilots having problems in speaking, hearing or understanding the cockpit talk. The term ‘environmentally coupled remedy’ is correspondingly used in analyzing the organization of action-in-interaction (see e.g. Arminen, 2005b), that is, the pilots having problems in accomplishing or positioning their sequentially structured actions in the cockpit. We suggest that repairs of talk and remedies of actions are identifiable types of activity that tend to co-occur and form the systematics of action in a given setting by mutually elaborating each other within their sequential and material environment.

Data and methods

The airline cockpit is a complex socio-technical setting in which the pilots coordinate their talk and non-talk activities to perform the tasks necessary to fly the plane. Flight crew members make their reasoning and understanding accountable and observable through talk and other available resources like gestures, gaze movements, bodily orientations and handling objects and artifacts. The salient non-talk activities in an airline cockpit include rotating and pushing knobs; monitoring displays; moving levers; reading checklists, charts and manuals; writing notes; looking out of the window, and so on. The flight crew needs to accomplish the talk and non-talk activities required for tasks in a strict sequence. The airline cockpit is a multi-task setting in which the performance of a particular task becomes relevant and appropriate only after other tasks have been completed and certain circumstances prevail (see Nevile, 2002, 2004b, 2006).

Flight safety depends greatly on strictly following standard operating procedures and the ‘scripts’ provided by the flight operating manuals. These include formal descriptions of i) what the pilots should say/do during the flight, ii) when exactly to say/do it, and iii) who is responsible for saying/doing it. The pilots’ duties and responsibilities are determined according to two formal roles. The first comprises their official rank or status as either commander (CDR) or co-pilot (COP), of whom the former is usually the more experienced and highly trained crew member. Since an individual pilot’s rank corresponds with a particular level of qualification and skill, it cannot change from flight to flight. The second role is to be either pilot-flying (PF) or pilot-not-flying (PNF). The PF controls the aircraft, making the immediate inputs to control the performance of the plane and taking responsibility for routine planning and decisions for the flight. The PNF typically assists the PF by setting up instruments, reading charts, communicating with air traffic control (ATC) and monitoring the PF’s performance. An individual pilot may be PF and PNF on separate flights made during the day, because these roles are not dependent on the rank of the crew member (Jentsch et al., 1999; Nevile, 2004a).

The research data used includes videotapes of official Finnair pilot training sessions conducted in the Airbus A320 flight simulator at the Finnair Flight Training Center (see Auvinen, 2009). The already experienced, licensed pilots are obliged to attend this particular kind of flight training twice a year. The overall amount of data is over 60 hours of recorded materials, of which nearly half (approximately 26 hours) was used in the analysis. In the videotaped sessions, the crews practice a training technique called Line-Oriented Flight Training (LOFT), in which they must handle a variety of scripted real-time scenarios including routine, abnormal and emergency situations such as fire in a cabin, technical malfunctions, etc. There are three participants in the sessions: the commander, the co-pilot and the instructor, who controls and manipulates the environmental and physical conditions of the scenario and maintains appropriate communication with the pilots by role-playing the air traffic controller, mechanic, cabin crew member, etc.

In conversation analysis, the initial and important stage of analysis is data transcription (see e.g. Arminen, 2005a; Hutchby and Wooffitt, 1998; Jefferson, 1983, 1985). The transcriptions used in this article describe talk, embodied conduct, and their sequential relationship in cockpit interaction. The descriptions of verbal activity illuminate the dynamics of turn-taking and the characteristics of speech delivery. The former includes details about the turn initiations, turn closings, overlapping turns and the gaps/pauses between turns, while the latter contains the essential features of stress, intonation, pitch, loudness and speed (Hutchby and Wooffitt, 1998). The temporal relationship between verbal and non-verbal activity is described in detail. The visible actions performed during the periods of non-talk are described as, for example, ((COP starts to lift his left hand to the speed/mach switch (1.3))). This indicates that the co-pilot initiates his hand movement at some point within a period of 1.3 seconds. The visual actions accomplished simultaneously with talk are marked with the interlocking left brackets. The transcriptions also include still pictures taken from the video-recordings. The timing of the picture/visual event is marked in the transcription by the symbols *, § and £.

The data analysis is based on ethnography and video-recordings of cockpit (inter-) actions. The data sources are not treated separately, but complement each other. The ethnography included ‘data-sessions’ with the professional pilots and other aviation experts, brief fieldwork in operational settings, participation in a ‘Multi-Crew Co-operation’ (MCC) course, and the like. The ethnography offers a background understanding of the coordination of the work and action in the cockpit. The videotaped data on cockpit practices and activities are analyzed in relation to their ethnographic context. The audio-visual data are crucial in providing access to a large array of verbal, visual and material practices through which pilots not only perform their institutional tasks, but also process intersubjective understandings of the flight.

Repair practices in cockpit interaction

Our analysis discerns the sequential organization of repairs and remedies and their relationship to the larger flight activities in progress. The analyzed cases have been selected out of 20 cases in our data. Since cockpit interaction is embedded in the sequential progression of the flight, order in talk-in-interaction is inextricably tied to the stage of mobility of the aircraft and its material and technical features (Koskela et al., forthcoming). The flight tasks consist of the simultaneous or overlapping accomplishment of verbal and non-verbal flight activities and orientation to the consequent sequential order of action made accountable in/through cockpit talk. Repair practices do not escape this inextricable embeddedness of talk, action and socio-material environment. Consequently, the repairs have to be analyzed via the environmental coupling of talk and action in their socio-material context. The empirical part is composed of five cases that are organized according to two kinds of coordination problems: pilots have trouble in coordinating their talk and physical action to perform a particular flight task, that is, problems in talk-in-interaction relate and create problems in the multimodal management of flight tasks observable (cases 1–3), or in the sequential positioning of tasks and activities revealed in and through their cockpit talk, that is, problems in the sequential progression of flight become accountable via talk (cases 4 and 5). The study shows how pilots use repairs of talk and remedies of action simultaneously to resolve their problems and achieve intersubjectively effective collaboration.

The pilots studied use English and Finnish for their flight tasks and activities. When they speak Finnish, the transcriptions follow a two-line format including, first, an original Finnish version and second, an idiomatic or vernacular English translation. In order to pinpoint the occurrence of code-switching (see e.g. Auer, 1984, 1998), the Finnish words with English translations are in bold. ¹ Prior to each transcription, we define the situation of action and specify the coordination problem the pilots are dealing with in their interaction. The transcription numbers, the flight crew ‘identifiers’, the starting times of the episodes and the formal roles of the participant pilots are also described. Lines of analytical interest are marked with arrows (→). Further, the still pictures taken from the video-recordings are used to visualize and highlight the aspects of interaction relevant to the analysis.

Managing problems in accomplishing tasks

All cases analyzed (1–5) demonstrate how repairs of talk and remedies of action intertwine in the collaborative courses of cockpit interaction. The flight crew uses talk and non-verbal activities as a resource in producing, recognizing and making sense of an evolving action in the cockpit. Both talk and action require shared understanding, and can be subjected to repair practices that may amend their course. In cases 1–3, the trouble in performing a particular flight task becomes observable through pilots’ talk. The proper task performance requires a shared understanding of cockpit talk that is achieved through repairs, the task being completed through the pilot’s remedial action in a socio-material interaction.

In case 1, the pilots are returning to Arlanda Airport, Stockholm, because of hydraulic failure in the aircraft’s technical systems. The commander initiates the task of setting the speed (line 5). The co-pilot has problems in hearing or understanding the talk completely, which becomes demonstrable in his own talk and action. This case is shown in two transcriptions: 1(a) simply concerns the conversation between the pilots, showing how they identify and deal with the problem of talk-in-interaction. In addition to talk, 1(b) includes the co-pilot’s non-verbal actions described both verbally and visually in still pictures. 1(b) reveals how the problem of talk-in-interaction appears and is resolved in the co-pilot’s practical task completion, simultaneously addressing the mutual relationship between talk and action.

OPEN IN VIEWER

In lines 1–3, the pilots are accomplishing the task of extending the flaps to position ‘one’. ² After a 9.0-second pause, the CDR initiates another task by ordering the COP to set the speed to 195 (lines 4–5). Instead of doing this, the COP initiates a repair with ‘one?’ (lines 6–7). The COP indicates a problem in hearing or understanding the speed value completely; with the slightly rising intonation in ‘one?’, the COP is asking for the CDR to complement the value. The CDR repairs her order by augmenting the COP’s turn with ‘nine five’ at line 9. After the repair, the COP performs the physical activity necessary for the task in hand (the activity of rotating the switch is not shown on the transcription); the task is completed with the COP’s verbal repetition of the speed value and confirmation that it is ‘set’ (line 11).

Transcription 1(b) shows how the previous sequence of talk is embedded within the non-verbal actions by the co-pilot.

OPEN IN VIEWER

The COP responds to the CDR’s order by moving his hand towards the speed/mach switch (lines 5–6; Still 1). The COP’s repair initiation brings his trouble in hearing or understanding the value completely to the surface of the interaction before his hand has touched the appropriate switch (lines 7–8; Still 2). The COP holds his hand stationary at the switch until the CDR completes the self-repair by repeating the missing digits of the speed value (‘nine five’; lines 9–10). The COP’s repair initiation (‘one’) is environmentally coupled with the suspended hand movement, which together constitute the sense of a retracted action. Only after the completion of the repair does the COP remedy the suspended activity by rotating the speed/mach switch (lines 9–11; Still 3); the COP’s verbal announcement and his hand movement down to his knees confirms that the task of setting the speed is finished (lines 12–13).

In the second case, the crew is going through the ‘taxi’ checklist at Stockholm-Arlanda Airport before taking off. The co-pilot’s task in performing the ‘briefing’ item on the checklist is to describe the departure route and the altitude assigned for the current flight. The route description turns out to be problematic because of an inappropriate orientation to action, that is, the co-pilot is using the wrong notes as an information source in the production of talk. Once repair operations on both talk and action are done, the flight crew members are able to achieve a socially shared understanding about the upcoming arrangements for take-off.

OPEN IN VIEWER

The ‘briefing’ item on the taxi checklist the COP reads aloud refers to a short confirmation of the more thorough take-off briefing done earlier in the preparatory phase of the flight (line 1). Following a pause, the COP takes on the responsibility as a PF to perform a briefing (‘it was my job’) and initiates the briefing with ‘that is’ (lines 2–3). The actual briefing is composed of the departure route ‘Runen four Delta’ read from the notes on the COP’s right, ³ and the assigned altitude of ‘five thousand’ feet (lines 3–4; Still 1). The CDR repairs the COP’s talk by immediately indicating that instead of ‘Runen four Delta’, the departure route cleared for the current flight is ‘A:ros two Charlie’ (lines 4–5). ⁴ After a pause in talk, the COP responds with the hesitation sound ‘uhm’, and a 1.4-second silence ensues (lines 6–8). The COP’s verbal and non-verbal conduct at lines 9–10 demonstrates his reorientation, and in line with the physical activity, putting away his old notes, he announces his reorientation. The CDR acknowledges the COP’s announcement and action with ‘yeah’ (lines 11–12). The COP’s following solo laughter seems to acknowledge and handle his embarrassment over his faulty action (giving incorrect coordinates due to incorrect/old notes; lines 13–14). Next, the COP continues remedying his line of action by orienting to the notes on his sliding table (line 15; Still 2). He receives the information on those notes with ‘yeah.’ and delivers the corrected take-off briefing ‘Ars two Charlie and five thousand blue’ (line 16). ⁵ The CDR finally confirms the briefing and ensures it is congruent with the delivery clearance (lines 17–18).

In case 3, the crew is conducting an emergency landing at Stockholm-Arlanda Airport because of fire in a cabin lavatory. As a safety procedure, the pilots have checklists for normal, abnormal and emergency situations. During the emergency, they have to carry out the appropriate checklists to restore the situation as soon as possible. In this case, the co-pilot’s misunderstanding of the commander’s talk becomes notable in his practical action. At the same time, more implicitly, the case addresses the pilots’ different understandings of the sequential phase of action – what to do next in the cockpit setting.

OPEN IN VIEWER

OPEN IN VIEWER

While the COP is entering data onto the Flight Management System (FMS) the CDR orients to the performance of another task, with his request ‘will you give me the checklist’ (lines 1–7). The COP provides an understanding check of the request: ‘the checklist’ (line 8). During the following 0.7 second, the COP brings the normal checklist into view (line 9). In this way, the COP demonstrates his comprehension of the referent of talk and the sequential phase of tasks. The COP’s non-verbal action displays his understanding that the CDR meant the current phase of tasks requiring the execution of the normal checklist (for an analysis of the example in a broader activity context, see Arminen et al., 2010). At line 10, the CDR engages in third position repair (Schegloff, 1992) by making a distinction between the emergency checklist and the normal checklist (line 10; Still 1). The COP still does not remedy his action, and a gap of 1.5 seconds ensues at line 11. The CDR then reformulates his repair by revising the description of the referent: ‘QRH’ (line 12). ⁶ After a 2.2-second pause, the COP delivers a new understanding check ‘QRH’, which he then repeats with the acknowledgment ‘yeah’ at line 14; at this point, it appears that the pilots establish the mutual intelligibility of the referent of talk and the sequential phase of flight tasks, for which a particular checklist is indexed. Environmentally coupled with his talk, the COP moves the normal checklist away (lines 14–15). He then looks for the QRH and gives it to the CDR concurrently with the area controller’s call (lines 16–17; Still 2). ⁷ Through this sequentially accomplished remedial action, the COP specifies a particular material artefact, the QRH, as crucial for establishing effective collaboration.

The previous cases (1–3) include two interrelated problems, in which the pilots are having trouble in the production and comprehension of cockpit talk and in orientation to action. The cases reveal how problems in interaction are consequential for both talk and action. The troubles in talk-in-interaction can momentarily delay the completion of the flight activity (the task of setting the speed being suspended because of a hearing or understanding problem; see case 1). These troubles can also stem from the participants’ orientation to a faulty line of action (using an irrelevant information source and therefore the wrong term in briefing; see case 2) and they can direct the participants’ orientation to the faulty phase of action (orienting to the execution of the normal checklist instead of the emergency checklist because of an insufficient verbal order; see case 3). The sequential trajectory of the problem-solving process was structured so that the repairs of talk ensued with the remedies of physical action in a socio-material interaction. The proper and safe conduct of flight tasks may thus require an elaborate coordination and mastery of the repair practices in terms of both talk and physical action. In socio-materially embodied interaction, talk and action as well as their repairs and remedies are environmentally coupled.

In the cases analyzed so far, the pilots manage problems in accomplishing flight tasks by initiating remedies of action through repairs of talk. In the final two cases, the timing of the task accomplishment proves to be problematic, the performance of the current task is suspended and some other activity is oriented to instead as timely within the sequential course of cockpit interaction.

Managing problems in positioning tasks

The pilots coordinate talk and action not only to accomplish their duties, but also to demonstrate to one another their situated and moment-to-moment understandings of the sequential organization of flying activities, that is, what flight tasks are completed, in progress and to be done next. In cases 4–5, the pilots specifically face intersubjective problems in the course of producing and coordinating their sequentially organized tasks and activities. The class of troubles discussed emerges in and through the pilots’ premature orientations to particular flight activity. The repair operations are used to remedy the lines of action that are produced and mobilized by mistimed (i.e. premature) talk.

In case 4, the flight crew is returning to Stockholm-Arlanda Airport because of hydraulic failure in the aircraft’s technical systems. While the aircraft is turning right to a heading of 270 degrees, the commander orients to the task of extending the ‘flaps’ to position ‘one’. At this point, however, the commander’s orientation appears premature relative to the ongoing activity of turning in the required direction.

OPEN IN VIEWER

At lines 1–3, the CDR initiates a flight task by ordering the COP to set the ‘flaps’ to position ‘one’. The CDR uses the connector ‘and’ to present the flaps extension as part of a larger course of action (Nevile, 2006; see also Heritage and Sorjonen, 1994). The COP’s response of ‘speed checked’ implies that the current speed is appropriate for the flaps extension (lines 4–5). With her turn at lines 7, 9 and 12, the CDR self-repairs her talk by ordering the COP to refrain from flap extension until the aircraft has reached the assigned direction. ⁸ In this way, the CDR remedies the line of action that she had ordered in her previous turn. The CDR’s talk includes particular word-search markers: stretching the ‘vielä:’ ‘still:’, a micro-second pause within the turn, and the sound of hesitation ‘tota’ ‘er’. The COP proffers ‘direction’ as a solution to the CDR’s word search, which the CDR accepts with ‘yes↓’ (lines 9–12). The COP’s overlapping ‘yeah okay’ displays his understanding and acceptance of the remedy, that is, of the order to temporarily suspend the flap extension until the right direction is achieved (lines 12–13). The remedial action and the re-establishment of sequential order are accounted for by the CDR: ‘so that there isn’t any banking’ (line 14); ⁹ the COP acknowledges and accepts this account with ‘yeah’ and ‘okay’ (lines 15–16). ¹⁰

In the final case (5), the crew is at Helsinki-Vantaa Airport before take-off. Here, the pilots develop and demonstrate their differently situated understandings of the sequential phase of the flight tasks, that is, which normal checklist to perform at this particular moment. The crew has already received a taxi clearance from Air Traffic Control (ATC) and started to taxi ¹¹ to a particular holding point of the runway ‘04 right’; however, the pilots have not performed the ‘taxi’ checklist yet. In his talk produced after the ATC take-off clearance, the commander skips the ‘taxi’ checklist, orienting prematurely to the completion of the ‘take-off’ checklist.

OPEN IN VIEWER

The COP displays his understanding of and compliance with the ATC take-off clearance (lines 1–4) by reading it back at lines 6–7. ¹² The CDR initiates the performance of the ‘take-off’ checklist while the COP is writing notes on the previous clearance (lines 8–10). The CDR’s situated understanding of the phase of the flight tasks seems to be based on the prior conversation between ATC and COP. After completing the writing activity and the talk related to it (‘Zulu Kilo’), the COP replaces the ‘take-off’ checklist with the ‘taxi’ checklist (lines 11–12). ¹³ The COP’s turn not only corrects the CDR’s previous talk but remedies the course of action suggested by the CDR. The CDR acknowledges the repair with the overlapping ‘ya’ (lines 11–13); his ‘taxiing check’ indicates that the pilots have a socially shared understanding of the checklist to be accomplished next (line 14). The COP’s temporal specification ‘first’ implies that the taxi checklist will be prior to the take-off checklist (lines 14–15). An apology at line 14 shows the CDR’s orientation to professional responsibility for skipping the checklist; he further acknowledges the remedy with ‘yeah’ (line 14). During the following 0.7 seconds, the COP brings the taxi checklist into view (line 16); he adheres to the sequential order of the normal checklists by reading aloud the first item on the taxi checklist (‘flight controls’, line 17). In his response, produced after a pause, the CDR confirms that the item ‘was checked’ (lines 18–19). ¹⁴

In cases 4 and 5, the pilots’ premature orientations to a particular phase of action become apparent in their talk. In case 4, the remedial action is launched via self-repair, and in case 5 via other-correction. Repairs mobilize a remedial process through which the party’s orientation shifts to the remedied phase of action, and the activity continues accordingly. In this way, the repairs of talk and remedies of action are intertwined with each other and with the sequential and physical environment in which they occur.

Discussion

Goodwin (2000: 1491) coined an ambitious view that a theory of action ‘must come to terms with both the details of language use and the way in which the social, cultural, material and sequential structure of the environment where action occurs, figure into its organization’. From this perspective, social action is a situated accomplishment that emerges from its practical management through language, social configuration and material resources. The approach to the study of the organization of talk and action in interaction aims to yield the actual details of historically shaped frameworks for action in order to reverse engineer the constitutive elements of phenomena in society. This perspective is consistent with material sociology in that it aims to re-establish the materiality of social action (MacKenzie, 2009).

Since Erving Goffman, there has been sustained interest in the social gatherings of co-present parties who, in addition to talking, engage simultaneously in other activities, such as manipulating material objects (Couper-Kuhlen, 2010). In social interchanges, participant actions consist of successive doings in which actions are linked together and are responsive to each other. Focused activities such as dancing, boxing or car driving are intricately orchestrated, situated and deeply interactional but do not require or allow sustained talk by the participants. In social life, there are number of activities that may allow but do not require sustained talk: ‘making hamburgers, eating dessert, clearing the table, tasting ravioli, making plastic flowers, fixing dinner, eating lunch, going through mail, etc. . . . On this view, situations of “just talking” are very special cases indeed’ (Couper-Kuhlen, 2010: 35).

Until recent years, much conversation analysis has focused on the sequence structure that concerns the relationship between turns at talk, that is, the organization of talk-in-interaction (Schegloff, 2007). Workplace studies (Heath et al., 2000) and other activity contexts bring up the question of the sequential order which constitutes a broader configuration concerning the relative positioning of any kind of move, utterance or action, that is, the organization of talk-and-action-in-interaction (Arminen, 2005b). CA studies have not yet built a systematic view of the relationship between sequential order and sequence structure, and have resolved it with ad hoc auxiliary assumptions (Couper-Kuhlen, 2010).

The conceptualization of social action has been an Achilles’ heel for studies on social interaction. For one, Schegloff has relentlessly pursued an account of social action. In his ‘Confirming allusions’ (1996: 174) he sets out to formulate an ‘empirically grounded theory of action(s)’ that should be ‘a central area of inquiry for sociological theory and analysis’. He develops ‘an account of action in interaction grounded in the observable details of conduct in naturally occurring social settings’ (p. 167) that identifies the fabric of social life and the foundations of sociality (p. 174). The goal is to form a generic account of actions covering ‘the moment-to-moment flow of daily life in- and outside-interaction’ (p. 164, note 3). Schegloff’s empirically grounded account of action has at least three distinct elements:

Although Schegloff’s account of actions is elaborate and generic, some reservations are appropriate. His stipulations on actions are limited ‘to explore something that speakers of language in interaction do observably do with it’ (p. 167). Though other resources than talk are mentioned, the focus is on speaking: ‘we need to press inquiries into what speakers can do – do do – with language and the other resources deployed in interaction’ (p. 167). Consequently, the empirical grounding concerns talk only: ‘accounts of actions’ are ‘the outcome of specifiable practices of talking’ (p. 171). The domain of inquiry is restricted to the practices of talk implicitly following speech act theory and an individualistic theory of social action.

The properties of talk are the vehicles of social actions on the top of which a multimodal layer may be added. Guided by implicit speech act theory, CA is concerned with ‘what action some form of talk or its component is doing’ and ‘what action is that utterance doing or being used to do?’. CA is primarily concerned with the ways in which utterances accomplish particular actions via their position and participation within sequences of actions. Its domain of inquiry has also remained consistent over the years (Heritage, 1984; Heritage and Clayman, 2010; Schegloff, 1992, 1996, 2007), taking sequences and turns within sequences as its primary units (Heritage, 1984). The claimed extension of the program to the study of social action (Heritage, 1984) falls short, as talk is in practice decoupled from other practices.

In practice, much of CA is limited by an implicit tie to speech act theory. Turns of talk are read as units of speech acts. The promise of multimodality also falls short, remaining just a layer over an analysis of talk. CA might benefit from closer acquaintance with material social science that aims at getting beyond the individualistic speech act view by considering performativity as a material process in which talk, action and technical resources are interwoven. Social action is embedded in materiality that makes us to do things via ‘agencements, combinations of human beings, material objects, technical systems, texts, algorithms, and so on’ (MacKenzie, 2009: 4). Multimodal conversation analysis can play a significant role in providing tools for analyzing the materiality of social action and address the dynamic interplay between sets of language and material structures that frame and reframe the social production of action and meaning, as well as catch the materiality of social actions as being embodied and based on distributed cognition (Goodwin, 2000, 2007; Mondada, 2012). These studies may amount to strict empirical scrutiny of agencement, an assemblage, arrangement, configuration or lay-out that is made up of human bodies as well as prostheses, tools, equipment, technical devices, algorithms, etc. in real-time actual social practices (Aaltonen et al., forthcoming).

Aviation is one context in which interaction has a solid material basis as it is anchored to the movements of aircraft and other operators in real time (Koskela and Arminen, forthcoming). The reflexive loop between mobility and situated interaction is called mobile contextual configuration (Arminen and Weilenmann, 2009; Koskela et al., forthcoming). In the work of pilots, mobile contextual configuration enables a reflexive loop between their work and the state of the aircraft in that pilots reflexively adapt, shape and make sense of their activities to maintain the mobility of the aircraft they control (see Koskela et al., forthcoming). Remedial work does not escape this inextricable embeddedness of talk, action and socio-material environment.

Conclusion

Harvey Sacks formulated a vision of culture as ‘an apparatus for generating recognizable actions’ (Sacks, 1992a: 226). This enables us to discover and describe the way in which whatever humans do is carried out, and amounts to a science of social life that works out the this-and-that of what the world is made up. Ethnomethodology, and in particular conversation analysis, has since tried to meet these ideals. Heritage (1984) proposed that conversation analysis is a vast extension of the ethnomethodological account of social action:

The organization of talk thus participates in a dialectical relationship between agency and structure in social life and in a cognitive-moral way. Without a detailed texture of institutionalized methods of talking to orient to, social actors would inevitably lose their cognitive bearings. (Heritage, 1984: 292)

Schegloff (1996) has meticulously continued to develop research policies that would offer a strict analysis of sequential aspects of social actions in real time and in real settings by identifying, specifying and comparing interactional patterns that constitute or contribute to establishing the social world as perceived.

Nevertheless, the delineation of the research domain of CA to talk-in-interaction has meant erecting and strengthening the analytic boundaries between language, cognitive processes, and structures in the material world. The separation of talk from other activities performed via various other media involves the risk of missing details of the parties’ ongoing orientation to action. To explore a socially shared sense of ongoing activity, we have to take the multimodality of social action into account and discern the ways in which social, cultural, material and sequential structures figure together in the organization of action. This perspective is also critical for scrutiny of cockpit interaction and its repairs and remedies based on complex multimodal activities in which talk, action and technical resources are interwoven. The resulting approach can amount to studies on action in interaction that explore the actual practices that constitute ongoing situated day-to-day social practices, increasing our understanding by re-specifying their interactional substratum (Arminen, 2008). Eventually, it may shape, broaden, detail and even correct our understanding of those practices (Peräkylä and Vehviläinen, 2003).