Achievement of Intersubjectivity in Airline Cockpit Interaction_Diss

7/22/2019 Achievement of Intersubjectivity in Airline Cockpit Interaction_Diss.

1/191

PETRA AUVINEN

Achievement of Intersubjectivityin Airline Cockpit Interaction

ACADEMIC DISSERTATIONTo be presented, with the permission of

the Faculty of Social Sciences of the University of Tampere,for public discussion in the Vin Linna-Auditorium K104,

Kalevantie 5, Tampere,on December 5th, 2009, at 12 oclock.

UNIVERSITY OF TAMPERE


2/191

ACADEMIC DISSERAIONUniversity o ampereDepartment o Social ResearchFinland

DistributionBookshop AJUP.O. Box 61733014 University o ampere

Finland

Cover design byJuha Siro

LayoutMarita Alanko

Acta Universitatis amperensis 1465ISBN 978-951-44-7879-6 (print)ISSN-L 1455-1616ISSN 1455-1616

ampereen Yliopistopaino Oy Juvenes Printampere 2009

el. +358 3 3551 6055Fax +358 3 3551 [email protected]/taju

http://granum.uta.fi

Acta Electronica Universitatis amperensis 900ISBN 978-951-44-7880-2 (pd )ISSN 1456-954Xhttp://acta.uta.fi


3/191

CONENS

Acknowledgements 5Abstract 7Glossary o Aeronautical erms and Abbreviations 9

C 1 Introduction 15

1.1 Cockpit interaction on actual flights 17 1.2 Distributed cognition in an airline cockpit 19 1.3 Overview o Chapters 2-7 24

C 2 Teory and methodology 26

2.1 Ethnomethodological conversation analysis 27 2.1.1 Te methodological basis o conversation analysis 29 2.1.2 Adjacency pairs and intersubjectivity in conversation 30 2.1.3 Ordinary as against institutional interaction 33 2.1.4 Summary 35

2.2 Workplace studies 37 2.2.1 Centres or the coordination o human activity 38 2.2.1.1 ool-mediated work in a hi-tech environment 39 2.2.1.2 Production o multiple local perspectives 42 2.2.1.3 Coordination o located and distributed activities 45 2.2.1.4 Sequential organization o cockpit activities 48 2.2.2 Summary 49

C 3 Data and methods 52

3.1 ask perormance in the cockpit 53 3.2 echnical eatures o the cockpit 54 3.3 Cockpit roles and language 57 3.4 Standard operating procedures and checklists 59 3.5 Te data 61 3.6 Analyzing the data 64 3.7 Understanding the data 67


4/191

C 4 Achievement o intersubjectivity in cockpit talk-in-interaction 69

4.1 Organization o repair 70 4.2 Repair practices in cockpit talk-in-interaction 74 4.2.1 Sel-initiated sel-repairs 74 4.2.2 Other-initiated sel-repairs 78 4.2.3 Other-initiated other-repairs 85 4.2.4 Tird position repairs 89 4.3 Conclusions 94

C 5 Achievement o intersubjectivity in cockpit talk-and-action-in-interaction 97

5.1 Interactive organization o social action 99

5.2 Repair practices in cockpit talk-and-action-in-interaction 105 5.2.1 Sel- and other-initiated sel-repairs 106 5.2.2 Other-initiated other-repairs 116 5.2.3 Remainders by the other 125 5.3 Conclusions 131

C 6 Gestures in the achievement o intersubjectivity in cockpit interaction 136

6.1 Study o visual phenomena 139 6.2 Gestures and repair sequences in cockpit interaction 142

6.2.1 Pointing gestures directed to the other crew member 145 6.2.2 Pointing gestures directed to the material object 150 6.2.3 Gaze movement directed to the other crew member 157 6.3 Conclusions 159

C 7 Summing up and conclusions 163

7.1 Problem dimensions 164 7.1.1 Comparative analysis o cockpit talk-in-interaction 165

7.1.2 Comparative analysis o cockpit talk-and-action-in-interaction 167 7.1.3 Te role o gestures in cockpit interaction 171 7.2 Research contribution 172 7.2.1 Te importance o intervention in cockpit interaction 173 7.2.2 Who is doing the intervention? 174 7.2.3 A glimpse o the uture and the LISA model 176

Reerences 178Appendix 1: Flight Hours o Pilots 189Appendix 2: ranscription Conventions 190


5/191

ACKNOWLEDGEMENTS

A great many people and institutions contributed to the production o this work. It is now mypleasure to acknowledge and give my warmest regards to the ollowing individuals and organi-zations or supporting and encouraging me during all these years:

I am most grateul to my supervisor, Proessor Ilkka Arminen, or his guidance and unail-

ing support throughout the process. My sincere thanks also go to Doctor Hannele Palukkaor opening me the doors to the world o civil commercial aviation and being there as a seniorcolleague and a riend.

I also wish to express my gratitude to Proessor Anssi Perkyl or encouraging me to startdoing this work and giving insightul comments as a preliminary examiner and Doctor Mau-rice Nevile or the valuable observations and remarks he has provided as an aviation researchexpert and a preliminary examiner.

My special thanks go to Finnair Captain and Flight Instructor Arto Helovuo, whose prac-tical contribution and support I greatly appreciate.

I am particularly thankul or the financial support rom the UCI graduate school,the Finnish Work Environment Fund and the University o ampere Foundation. I am alsoindebted to the Multimodal Activities and Situated Agencies research project. Special thanksgo to Te Department o Social Research in the University o ampere or providing acilitiesor the research, the Alred Kordelin Foundation or giving the grant or the revision o theEnglish, and the City o ampere or unding the printing o the publication.

I am deeply grateul to the Finnair pilots and flight instructors whose participation madethis research possible. Tank you to the company representatives and managers who showedinterest and support or the study. Special thanks to Hannes Bjustrm, Pekka Perkyl, HeikkiSaloheimo and Juha Siivonen. I hope that the research results are ound to be inspiring and

applicable in practice.I sincerely thank the members o the post-graduate seminar on cultural and interactional

studies in the Department o Social Research or their insightul comments and criticism othis work. I am also grateul to the members o data sessions or their patience and open-mind-edness in trying to make sense o and analyze the complex interaction data.

I would like to acknowledge Rod McConchie or careul and instructive revision o theEnglish o the dissertation. Tanks also to Virginia Mattila or revising the language o theSummary and Acknowledgements, to Pivi Hannila or transcribing one videotape and toMarita Alanko or preparing the layout.


6/191

Appreciation and thanks to all my riends and colleagues or keeping me sane andundaunted at various stages o the work.

I would like to express my heartelt gratitude to my amily my mother Sirpa, my brother

Perttu and his amily, my unofficial grandmother Elsa and my soulmate Ville. Tis dissertationis dedicated to these people, who have been a constant source o support and understanding inall my academic and non-academic endeavors. Special thanks to my golden retriever Hanskior reminding me to take a break and go or a walk.

ampere, October 2009

Petra Auvinen


7/191

A I A C I 7

ABSTRACT

Achievement of Intersubjectivity in Airline Cockpit Interaction

Flight saety and efficiency requires that airline pilots have a shared understanding o whatis going on both inside and outside the cockpit during the flight. Te joint awareness o theaction-in-progress is created and sustained through multiple interactions, such as talk, gestures

and bodily orientations. Tis study concerns the interactive achievement and maintenance omutual awareness between Commander and Co-pilot in the cockpit environment. Te analy-sis ocuses on those cases in which intersubjectivity, i.e., the socially shared understanding oongoing talk and action temporarily breaks down, or threatens to do so, between the two pilotsin flight. Te instances describe situations in which the pilots have difficulty in terms o speak-ing, hearing or understanding the cockpit talk and in terms o maintaining the sequential ordero flight tasks and activities. Te study highlights the different types o repair practices thepilots use to identiy and resolve problematic understandings in order to achieve intersubjec-tivity and optimize flight saety.

Te research data was collected by videotaping Finnair pilot training sessions in the AirbusA320 flight simulator at the Finnair Flight raining Center, Vantaa, Finland. In the sessions,the pilots practiced a training technique called Line-Oriented Flight raining (LOF), inwhich they must process a variety o scripted real-time scenarios including routine, abnormaland emergency situations. Experienced, licensed pilots are required to undergo this particularkind o training twice a year. Te overall amount o data increased to over sixty (60) hours orecorded material, o which about twenty-six (26) hours was used in analysis. Te primaryresearch method was ethnomethodological conversation analysis, which considers talk andinteraction as structurally organized social action. Although a secondary method only, eth-

nography was o great importance in providing the necessary background understanding orthe detailed analysis o videotaped interactions. Te ethnography included, among others, theresearcher participating in a pilot training course on Multi-Crew Cooperation (MCC), observ-ing the cockpit interaction in real operational settings and amiliarizing hersel with the workdone in approach control and in the aerodrome control tower at Helsinki-Vantaa Airport.

Te data analysis was based on comparisons between cockpit interaction and ordinaryconversation (the non-institutional orm o talk). Te participants in ordinary conversationtend to avoid any potential or conflict and disagreement by mitigating and delaying other-repairs, i.e., the activities o correcting the other speaker. In everyday settings, other-repairsperormed straightorwardly and without delay are largely detrimental to social solidarity and


8/191

8 A I A C I

possibly threatening to the ace o the interlocutor. Direct, overt other-repairs can be accom-plished without conflict or discord in cockpit interaction. In contrast to ordinary talk, thepilots main orientation is not towards ace work in the activity o correcting one another.

Trough direct other-repairs, they rather orient to the rapid achievement o intersubjectivityin order to ensure the saety o the flight. Te normative procedures o interaction in ordinaryconversation are thus overridden by the institutional norms o flight saety.

Te flight crew must carry out the tasks and actions in a strict sequence. Te pilots maymomentarily have different understandings o what actions are complete, in progress, and tobe done next and by whom. In light o this study, the most common problem in maintainingthe sequential order o action is so-called premature actions, in which the pilot orients prema-turely to some particular flight action. Te premature orientations resolved by the pilots withconversational repairs are more typical in abnormal and emergency situations than in normalconditions. Te unpredictable and time-critical nature o emergency operations may contrib-ute to the occurrence o premature actions in cockpit interaction. Te task sequence also tem-porarily breaks down in so-called absent actions, in which the pilot ails to perorm the actions/he is responsible or. Te unperormed action is carried out aer one flight crew member hasreminded the other to perorm it.

Te pilots use not only talk, but also different kinds o visual activity as resources orprocessing their problematic understandings in cockpit interaction. Te analysis ocused spe-cifically on the role o pointing gestures in the achievement o intersubjectivity. What is doneor accomplished through gesticulation is at least partly dependent on the target o the gesture(human vs. material), the timing o the gesture relative to talk (gesture produced prior to vs.

simultaneously with talk) and the organization o participation in the course o interaction(recipient vs. producer o gesture). Te pilots use pointing gestures to make their orientationand engagement in processing o intersubjectivity visible. Te visual pointing toward, or exam-ple, the cockpit instrument allows the pilot to direct the other flight crew members attentionto that object and thereore to establish mutual orientation in interaction. Te spatial-temporalcoordination o talk and gesture enables pilots to allocate flight tasks and speciy the inorma-tion sources employed in the production o talk.

Te pilots not only identiy episodes o problematic understandings, but also demonstratetheir proessional competence and know-how by their actions or repairing and reminding.

Te interventions accomplished with these practices can also be seen as evidence o a well-unctioning interactive back-up system eliminating risks to flight saety. In ocusing on theprocesses o cooperation and interaction, the research is closely related to the subject domainso pilot proessional training called Crew Resource Management (CRM). Te study makes apractical contribution to CRM training methods by offering new insights or flight instructorsand other training personnel to utilize in teaching trainees non-technical skills. Te generalguidelines provided by the LISA model can be used as a starting point in helping to achieveintersubjectivity not only in cockpit interaction but also in other environments where saety iscritical: Listen to what the other is saying. Intervene in action when necessary. State your inten-tions. Ask i you do not hear or understand.


9/191

A I A C I 9

GLOSSARY OF AERONAUTICAL TERMS AND ABBREVIATIONS

Abnormal situation: one in which it is no longer possible to continue the flight using normalprocedures, but the saety o the aircra or persons on board or on the ground is notin danger.

Acknowledgement: notification that a given communication has been correctly received and

understood.Air raffic Clearance: approval by Air raffic Control or an aircra to taxi, take off, climb,enter controlled airspace, descend or land.

Altimeter: a cockpit display showing the aircras altitude.Altitude:the height o an aircra as shown on an altimeter adjusted to local barometric pres-

sure.AP: an abbreviation or autopilot.AC: (Air raffic Control): asystem o directing all aircra operating within designated air-

space by radio. Divided into sectors such as ower (aerodrome control or take-offsand landings), Departures, Control (en route aircra), and Approach.

Attitude: the lateral and longitudinal relationship o the aircra to the horizon.CAVOK: Te term(meaning ceiling and visibility OK and pronounced kav-okay) may be

used in place o visibility, weather and cloud, provided that visibility is 10 kilometersor more; there is no cloud below 5,000 eet above aerodrome level, or below the mini-mum sector altitude, whichever is higher, and no cumulo-nimbus clouds (i.e., thun-derstorms); no precipitation reaching the ground, no thunderstorms, and no shallowog or low driing snow.

CDR (Commander): a ormal rank held by a pilot. Te Commander is the member o theflight crew who has the ultimate authority and responsibility or the conduct o the

flight.Checklist: a tool used as a human actors aid in aviation saety to ensure that items on a long

list are not orgotten. It is needed because o the limitations o human memory.Clean aircraf: an aircra in normal cruising configuration, with high li devices (flaps and

slats) and landing gear retracted.Climb (or climb phase): the period during which the aircra climbs to a pre-determined

cruising altitude aer take-off. Depending on the aircra, the altitudes involved, andother actors, this phase may last rom a minute or two to hal an hour or more. Teopposite o a climb is a descent.


10/191

10 A I A C I

COP (Co-pilot): a ormal rank held by a pilot. Te Co-pilot is the member o the flight crewwith accompanying Commander. Te Co-pilot is usually the junior and less experi-enced member o the flight crew.

CRM (Crew or Cockpit Resource Management): part o pilot training which seeks todevelop the ability o pilots to work together as a team and use all the resources avail-able to them to effectively; or example, perorm tasks, assess situations, make deci-sions, and identiy and resolve problems.

Descent: the phase o flight in which an aircra decreases altitude; the opposite o a climb.Descents are an essential component o the approach to landing.

DME (Distance Measuring Equipment): a radio navigation aid providing a constant readouto the aircras distance rom a selected radio beacon.

ECAM (Electronic Centralized Aircraf Monitoring, Fin. lentokoneen elektroninen kes-kitetty valvontajrjestelm): a system that monitors aircra unctions and relaysthem to the pilots. It also produces messages detailing ailures and in certain caseslists procedures to undertake to correct the problem.

EFIS (Electronic Flight Instrument System, Fin. elektroninen lennonvalvontamittaristo):a flight deck instrument display system which shows all inormation regarding theaircras situation, position and progress. It primarily covers horizontal and verticalposition and also indicates time and speed.

Emergency landing: an unplanned landing made by an aircra in response to a crisis whicheither intereres with the operation o the aircra or involves sudden medical emer-gencies necessitating diversion to the nearest airport.

Emergency situation: one in which the saety o the aircra or o persons on board or on theground is endangered or any reason.

E/WD: Engine/Warning DisplayFCOM (Flight Crew Operating Manual): a technical publication written or a specific air-

cra which is used to operate that aircra and to explain its technical specifications.Te manual includes a wide range o inormation, such as the procedures or abnor-mal and emergency operations (in Vol. III).

FCU: Flight Control UnitFlaps (Fin. laipat): adjustable suraces on the trailing edge o an aircras wing. When low-

ered, flaps increase the li o the wing, thereby reducing stalling speed, and increasingdrag, steepening aircras glide angle.Flight Director (FD): a computer-controlled flying instrument combining inputs o other

flying and radio navigation instruments in a single large display located directly inront o each pilot.

Flight Level (FL, Fin. lentopinta): an expression o height in hundreds o eet, based on thestandard barometric altimeter setting o 1013.2 millibars. For instance, 12,000 eeton the standard altimeter setting would be FL120. FL differs rom altitude in thatthe latter is based on the actual barometric altimeter setting or a particular area orairport.


11/191

A I A C I 11

FMS: (Flight management system, Fin. lennonhallintajrjestelm):an aircra computersystem that uses a large database to allow routes to be programmed and ed into thesystem by means o a data loader. Te system is constantly updated with regard to

position and accuracy by reerence to conventional navigation aids.Heading (Fin. ohjaussuunta):a direction described in compass degrees.Human actors: the multidisciplinary field devoted to optimizing human perormance and

reducing human error, which incorporates the methods and principles o the behav-ioral and social sciences, engineering, and physiology. It is the applied science con-cerned with people working together in concert with machines.

IFR (Instrument Flight Rules, Fin. mittarilentosnnt): stipulated procedures or navigat-ing aircra by reerence to cockpit instruments and radio navigation aids alone toenable flight regardless o visibility. Tis is the normal operating procedure or airlineflights.

ILS (Instrument Landing System, Fin. mittarilhestymismenetelm): an electronicapproach aid which enables a pilot to carry out an approach or landing when weatherconditions preclude visual contact with the ground.

IMC (Instrument Meteorological Conditions, Fin. mittarisolosuhteet):weather condi-tions in which visibility is less than specified or visual flying, and in which flight islegally possible only under IFR.

Knot (K): one nautical mile per hour, equivalent to 1.853 km/h.Landing: the phase o flight in which a flying aircra returns to the ground.Landing gear (also called undercarriage): the aircras wheels and associated assemblies. Te

landing gear is used during take-off and landing and to taxi on the ground. Mostplanes use what is called a tricycle landing gear arrangement. Tis system has twolarge main gear units located near the middle o the plane and a single smaller nosegear unit near the nose o the aircra.

Localizer (Fin. suuntalhetin): the component o an instrument landing system (ILS) thatprovides lateral guidance with respect to the runway centreline.

LOF (Line-Oriented Flight raining):reers to aircrew training that involves a ull mis-sion simulation o situations that are representative o line operations, with specialemphasis on situations involving communications, management and leadership. In

short, LOF means realistic, real-time, ull mission training.Nautical mile (nm): a measure o distance used or navigation in the air and at sea. An nm isequal to one minute o an arc o latitude on the earths surace. A nm is 800 eet longerthan a statute mile and equivalent to 1.853 km.

ND: Navigation DisplayNDB (Non-Directional Beacon): a ground-based radio transmitter sending continuous sig-

nals in all directions or use by aircra fitted with Automatic Direction Finder (radiocompass).

PF (Pilot-flying): the pilot in control o the aircra and generally responsible or making mostroutine decisions about the conduct o the flight.


12/191

12 A I A C I

PFD: Primary Flight DisplayPNF (Pilot-not-flying):assists the PF and is responsible or most radio communications.QNH: thecode expression designating the altimeter setting in millibars when set on the sub-

scale o the altimeter, the instrument reads the aircras height above mean sea level.QRH (uick Reerence Handbook): a handbook containing extracts rom the AeroplaneFlight Manual (AFM) or Operations Manual (OM) which may need to be reerred toquickly and/or requently, usually including emergency and abnormal procedures.

Radial: a bearing to or rom a VOR radio range navigation beacon.SD: System DisplaySID (Standard Instrument Departure, Fin. vakiolhtoreitti): a standard and published

departure route rom an airport, i.e., speciying runway, headings, navigation pointsand altitudes to be used during departure, and other critical inormation.

Slats: an aerodynamic device fitted to the leading edge o the wings to delay the onset o stall-ing.

Spoilers: suraces on an aircras wings designed to spoil the airflow over the wings and soreduce li (e.g., during descent or aer touchdown during the landing ).

Standard callouts: used to convey vital inormation with a minimum number o words thathave an exact meaning or all crew members. Proper adherence to standard calloutswill stimulate more meaningul and standardized crew communications and provideor early detection o crew member incapacitation during critical phases o flight.

Standard operating procedures (SOPs): the guidelines as to who-does-what-and-when.Tese are designed to enhance saety, assist the flight crews to manage risks and

ensure consistency in the cockpit.SAR (Standard Arrival Route, Fin. vakiotuloreitti): a standard published arrival route

to an airport, i.e., speciying runway, headings, navigation points and altitudes to beused during the approach, and other critical inormation.

ake-off: the phase o flight in which an aircra goes through a transition rom moving alongthe ground (taxiing) to flying in the air. ake-off is the opposite o landing.

axi (also called ground-taxi): the movement o the aircra on the surace o the aerodromeunder its own power, excluding take-off and landing.

CAS (raffic Collision Alert System): a type o airborne collision avoidance system (ACAS)

based on a amily o airborne equipment that unctions independently o the ground-based AC system to detect potentially conflicting aircra that are equipped withsecondary surveillance radar (SSR) transponders.

ransponder: a radio device fitted to aircra which, when triggered off by certain radar wave-lengths, emits a signal visible on ground radar screens. Te signal usually includesadditional inormation such as the altitude o the aircra.

UC:Co-ordinated Universal imeV1: Decision speed during take-off. An aircra is committed to fly when this speed is passed.

(Fin. Lentoonlhdn ptsnopeus eli nopeus jota ennen lht voidaan keskeyttturvallisesti kiitoradalle.)


13/191

A I A C I 13

Vr: Rotation speed. Speed at which an aircra is rotated into li-off attitude by raising thenosewheel off the runway.

V2:ake-off saety speed. Minimum control speed plus saety margin to allow or engine ail-

ure and other contingencies. (Fin. Lentoonlhdn pienin turvallinen nopeus miklimoottori hajoaa lentoonlhdn jlkeen.)Vector (also called: radar vectoring): A heading given by a controller to a pilot on the basis

o radar-derived inormation to provide navigational guidance.VFR (Visual Flight Rules, Fin. nklentosnnt): stipulated flight procedures or navigat-

ing aircra visually, clear o cloud, in Visual Meteorological Conditions.VMC (Visual Meteorological Conditions, Fin. nksolosuhteet): weather conditions

providing a specified range o visibility, making it possible or pilots to use visualmeans to avoid terrain and other aircra.

VOR (Very High Frequency Omni directional Radio Range): a radio range navigation bea-con.


14/191


15/191

I 15

C 1

Introduction

Flight saety and efficiency requires that the airline pilots have a shared understanding o whatis going on both inside and outside the cockpit during the flight. Te joint awareness o theongoing action and activity is created and sustained through a large variety o multiple inter-actions, such as talk, gestures and bodily orientations in the local surroundings. Tis studyconcerns the interactive achievement and maintenance o mutual understandings betweenCommander and Co-pilot in a cockpit environment. Along with the talk, the study highlightsthe role o gestures and other orms o embodiment in achieving the intersubjective sense oan unolding course o action and interaction. Te analysis ocuses specifically on those casesin which the shared understanding temporarily breaks down, or threatens to do so, betweenthe two pilots on the flight. Te instances describe situations in which, firstly, the pilots have

difficulty in terms o speaking, hearing or understanding the cockpit talk in and through whichtasks and goals are accomplished on a flight deck. In these cases, or example, the airline pilotdoes not hear or understand the reerent o the talk, or s/he misunderstands the purpose o thetalk. Secondly, these instances include situations in which the pilots have problems maintain-ing the sequential order o flight tasks and activities; here, or example, the pilot orients prema-turely to a particular task or (re-) orients to accomplish an activity already done.

In conversation analytic terms, this study concerns the different types o repair practicesthe pilots use to locate and resolve problematic understandings in cockpit interaction (see Sche-gloff, Jefferson & Sacks 1977; Schegloff 1992a; Sorjonen 1997). Te term repair reers to the

organized ways o dealing with the various kinds o trouble in the process o interaction (tenHave 1999, 116). On the flight deck, as will be shown, the airline pilots have troubles related tolanguage use or the maintenance o the sequential courses o action and activities. Tese prob-lems can challenge intersubjectivity, i.e., the socially shared understanding o ongoing talk andaction. Te repair activity is a key resource in achieving intersubjectivity in situations where itis momentarily lost between the pilots. In the conversation analytical approach, the repair istreated as a sequentially structured phenomenon. Te episodes o repair activity consist o arepair initiation marking possible disjunction with the prior talk and a repair outcome solvingor abandoning the problem; the problematic talk which the repair addresses is reerred to as


16/191

16 A I A C I

the trouble-source or repairable. Te repair sequences used in a cockpit interaction fleetinglysuspend the perormance o the task the flight crew members are engaged in. Te detection andresolution o the problem through the reparative operation is a necessary precondition or the

continuation o the task; when the problem is resolved and intersubjective sense o talk andaction achieved, the pilots can resume the suspended flight activity.Te crew members not only use verbal activities in achieving shared understandings on

the flight. Within the repair sequences, the pilots employ and demonstrate their orientationto a large number o different visual and material phenomena, such as their bodies and theeatures o the surrounding physical environment. Te simultaneous deployment o talk, thebody and the material surroundings provides pilots with the resources or locating and deal-ing with problematic understandings in the course o cockpit interaction. Te processing ointersubjectivity is thus analyzed as a multimodal activity in which the pilots juxtapose verbal,visual and material resources to make sense o and display their understandings o the ongoingaction in the cockpit setting. Concentrating on the interconnections between talk, body andmaterial surroundings in the production o meaningul action, the study strongly resonateswith the theoretical notions o C. Goodwin (1994; 1996; 2000; 2003a), who points out thatwithin a situated human interaction, talk, gestures and bodily orientations are deployed in waysdesigned to mutually elaborate each other. In a cockpit interaction, the achievement o inter-subjectivity is not simply a verbal phenomenon but a multilayered activity, inseparable romthe complex configuration o different kinds o meaning-making practices.

Using video-recordings rom the simulator training sessions as the data, the study providessystematic and reliable knowledge o the social-temporal organization o cockpit interaction

in general and the intersubjective problems emerging in the sequential courses o that interac-tion in particular. Te main aim is to analyze and explore the various repair practices the airlinepilots employ to collaboratively manage problematic understandings in the cockpit interac-tion. Te study also addresses and examines the role o the visual conduct o pilots, especiallypointing gestures and gaze movement, in dealing with possible losses o intersubjectivity onthe flight deck. In centering on the processes o co-operation and interaction, the research isclosely related to the subject domains o pilot occupational training called Crew ResourceManagement (CRM). CRM training ocuses on teaching pilots the concepts and skills relatedto such areas as communication, situational awareness, problem-solving, decision-making and

teamwork, thus improving pilots perormance as the members o the team. Te ultimate goalo training is to promote flight saety and the efficiency o flight operations by reducing thepossibility o human error (Advisory Circular 2004a; CAP 720, 2002; Orlady & Orlady 1999,268-294; Helmreich et al. 1999). Te study is thus intended to make a practical contributionto CRM training practices and methods by offering new insights or flight instructors andother training personnel to use in teaching the non-technical skills to the trainees.

In aviation research, the cockpit communication is generally investigated out o its con-text. Te studies avor large-scale quantitative analysis in which particular types o utterance,or speech act, such as questions, requests, instructions, are encoded and counted to answer


17/191

I 17

the predetermined research questions (see, e.g., Kanki et al. 1991; Bowers et al. 1995; Sexton& Helmreich 2000; see also Krifa et al. 2004). In this study, however, the communication isanalyzed in its context, embedded within the sequentially organized interaction between the

pilots (see also Nevile 2004a). Examining the talk in its interactional context reveals that thepilots use not only verbal utterances, but also different kinds o embodied activity (pointinggestures, head movements, upper body orientations, etc.) as resources in organizing their col-laborative courses o action. Tis study shows in detail how pilots orient to these verbal, visualand material modes o expression as relevant while processing the problems o intersubjectivityin cockpit interaction. Tus, as regards traditional research on aviation, the study extends thescope o analysis by taking the multiple communicative modalities the pilots use in producing,recognizing and coordinating their action on a flight into account. Tis research provides amuch deeper understanding o how pilots display their orientation to, and awareness o, theunolding courses o action and activities in the cockpit setting by addressing the multimodalnature o communication.

1.1 COCKPIT INTERACTIO N ON ACTUAL FLIGHTS

Te Australian Maurice Nevile (2004a) introduced airline cockpit interaction as a subject oresearch. Drawing on the theoretical perspectives o ethnomethodological conversation analy-sis and institutional discourse analysis, he addresses routine cockpit talk and interaction in realoperational settings. Tis talk is that which occurs when nothing seriously wrong happens,

the talk though which normal flight tasks and activities are perormed. By careully investi-gating the video-recordings made on actual scheduled passenger flights in Australia, Nevileshows how pilots develop and make their situated and evolving understandings available toeach other as they work together as a flight crew to fly the plane. Te analysis highlights theprocesses o talk and interaction though which flight crew members maintain a joint awarenesso what is going on around them, who knows what, who is doing what, and what they are to donext. Reasoning and understanding is made visible and interpretable through talk and otherresources available in the cockpit interaction, such as gestural movements, handling materialobjects, bodily orientations in a local space, etc. (Ibid.)

Neviles study consists o three empirical parts. Te first part ocuses on the interactivesignificance o pronominal choices o pilots in routine cockpit talk and interaction (Nevile2004a, 31-79; see also 2001a). Te analysis o the recorded interactions reveals how variouspronouns are used and oriented to by the pilots as a socially significant resource to accom-plish relevant cockpit identities in the ongoing course o interaction. Te analysis demonstrateshow selecting the wording o I- me- my, you- your- yours or we- our- ours, or example,is a characteristic eature o pilots communicative practice in establishing and maintainingthe intersubjective sense o who is doing what and what is going on as the flight proceeds toits destination. Te analysis shows specifically how pilots invoke and make salient individual


18/191

18 A I A C I

identities o the Captain and First Officer or Pilot-flying and Pilot-not-flying as well as ashared identity o the crew member or member o flight X through the flexible use o differ-ent pronominal choices in the cockpit interaction.

In the second part o the study, Nevile demonstrates how pilots temporally coordinatetalk and non-talk activities as they perorm the routine tasks required to fly the plane (2004a,81-144). Te ordinary non-talk activities conducted in the airline cockpit consist o pressingthe buttons, moving levers, turning dials, entering data onto the aircra computer and lookingat the displays. Te analysis illuminates how the precise and timely coordination o talk andembodied action contributes to the pilots shared understanding o the progress o the flight.Trough the sequential coordination o talk and visual conduct, the pilots make themselvesaccountable or what they say and do, and exactly when, as members o a flight crew jointlyresponsible or the sae conduct o the flight. (See also Nevile 2002.) Te third part broadensthe ocus to examine the ways in which pilots integrate their mutual talk within the cockpitwith their talk with the air traffic controllers beyond the cockpit. Te analysis concerns howthe pilots radio talk with the distant controller is fitted with the talk and non-talk activitywithin the cockpit between the pilots. Te findings reveal how, through talk and non-talkactivity, pilots establish the shared understandings o who has heard what, and who knowswhat, in terms o the significance o particular radio talk or their joint conduct o the flight(Nevile 2004a, 145-196; see also case example in Chapter 2: 2.2.1.4).

o summarize, Nevile (2004a; see also 2001a & b; 2002; 2004b; 2005a & b; 2006;2007b & c) examines the pilots daily talk and non-talk activities as they work together as ateam to perorm their routine flying tasks. Neviles (ibid.) studies, based on the detailed anal-

ysis o the videotapes made on real operational flights, ocus on the unproblematic, routineprocesses o talk and interaction and the ways in which flight tasks and activities are perormedsmoothly, without interruptions caused by the loss o intersubjectivity. In his article publishedin 2007a, Nevile concentrates on the specific problems the pilots encounter in maintaining thesequential order o action. In line with my research (see Chap. 5: 5.2.3), the article concernsthe cases where a cockpit action is absent, i.e., an action is timely and relevant, but not yetinitiated by the pilot responsible or doing so (see also Nevile 2008). In a ew other studies,Nevile (2004a; Nevile & Walker 2005) touches upon the phenomenon o problematic under-standings between pilots, revealing some cases in which they have trouble in the production o

verbal and/or physical courses o conduct, and pointing out the repair work done to resolve thetrouble. In the video-recorded materials used in my study, the pilots train to fly in both normaland emergency conditions in a simulated environment. Te research thus provides an extensiveand systematic analysis o intersubjective problems o pilots, not only in routine flight situa-tions, but also during crises (e.g., fire in the cabin, technical ailure, etc.).


19/191

I 19

1.2 DISTRIBUTED COGNITION IN AN AIRLINE COCKPIT

Distributed cognition is a theoretical approach developed by Hutchins and his colleagues in

1980s as a paradigm or re-conceptualising cognition (Hutchins 1995a). Teoretically andmethodologically, the approach is based on cognitive sciences, cognitive anthropology and thesocial sciences. Rather than centering simply on an individuals internal cognitive processes, astraditional cognitive approaches do, distributed cognition ocuses on the processes that occurin extended cognitive systems. Within these larger systems, there are multiple people interact-ing with each other and a large body o technical tools and artiacts with which to perormtheir work activities. Tese individuals have overlapping and shared access to knowledge thatenables them not merely to become aware o what others are doing, but also to coordinatethe action collaboratively. Analyzing the properties and dynamics o a particular cognitive sys-tem, such as an airplane cockpit (see Hutchins 1995b; Hutchins & Klausen 1996; Hutchins& Palen 1997; see also Holder 1999) or air traffic control (see, e.g., Fields et al. 1998) involvescarrying out an in-depth ethnographic study o the setting, paying close attention to the activi-ties o the personnel, their communications with each other and their interactions with vari-ous media. Tese processes are conceptualized in terms o the propagation o representationalstate across media. Te propagation o a representational state reers to how inormation istransormed during the accomplishment o an activity. Since the media amount to both inter-nal (e.g., an individuals memory) and external representations (e.g., computer displays, papernotes, etc.), the analysis is ocused on the specific transormations o inormation between themedia in revealing how the various representational states are propagated. (Hutchins 1995a;

Rogers 2006.)Hutchins and Klausen (1996) have analyzed the various cognitive properties o the simu-

lated cockpit system, which is composed o the pilots and their inormational environment.As a method o analysis, the researchers generated various representations o the events inthe cockpit: (I)the audio and video-recordings and (II)transcriptions o the verbal and otherbehavior in the cockpit; and (III)the descriptions and (IV)the interpretations o the actionsthat took place. Te aim is to interweave the data, the actions, the interpretations, and theethnographic groundings as they are needed in a narrative that seeks to present a theoreticalaccount o the observed events (Hutchins & Klausen 1996, 19). In the ollowing, we will see

the first hal o the transcription in their study. Te transcription is taken rom the recording oa simulated flight rom Sacramento to Los Angeles. Te aircra is climbing towards its cruisealtitude o 33,000 eet. Te aircra type (Boeing 727-200) requires a crew o three people: theCaptain (Capt), the First officer (F/O) and the Second Officer (S/O). Te Captain is replacingthe departure chart in his airway manual. Te First Officer is flying the plane, monitoring thevarious flight instruments and handling the controls. Te Second Officer begins a departurereport by radio to the company offices on the ground.


20/191

20 A I A C I

ranscription 1(a)

Note: Te number sequence on the le signifies the Coordinated Universal time (UC); the

descriptions within the { } markers illustrate the pilots non-verbal actions in the cockpit.0216 S/O xxx NASA nine hundred.0224 S/O Departure report. S/O NASA nine hundred rom Sacramento to Los Angeles Inter- national we haveuel on board twenty seven point eight uel boarded is not available out time is one six our five up time is one six five five.0247 Capt Oakland center NASA nine hundred request higher. {F/O reaches to vicinity o altitude alert setting knob when Air raffic Control, AC, begins transmission.}0254 OAK24L NASA nine hundredroger contact Oakland center one thirty two point eight. {F/O pulls his hand back rom the altitude alert knob when AC says contact Oakland center. 2.5 seconds aer the end o AC transmission, F/O looks at Capt} {Capt looks at F/O.}0300 F/O Tirty two eight. Capt Tirty two eight?

F/O Yeah. Capt OK0303 S/O Tats correct, NASA nine hundred. Capt One three two eight, NASA nine hundred. {Capt twists knob on radio console} {F/O looks in direction o Capt}

continues

(Hutchins & Klausen 1996, 15-16.)

Te flight crew is now approaching the altitude which they were last cleared to, i.e., 23,000 eet.Tis means that without the air traffic controllers clearance to a higher altitude, the aircracannot legally climb above that height. However, the flight plan filed or the current flightrequires a cruise altitude o 33,000 eet. In his opening turn, the Captain is thus calling the low-altitude controller and requesting clearance to a higher altitude (see 0247). Instead o clearingthe aircra higher, the controller hands the plane off to a high-altitude controller by asking thecrew to contact Oakland center one thirty two point eight (see 0254). Te number sequencein question is the radio requency o a high-altitude controller: 132.8 MHz. According to the


21/191

I 21

procedures in aviation, the pilot should acknowledge all inormation rom Air raffic Control(AC) immediately. Note how in this case, the Captain transmits delayed acknowledgementto the low-altitude controller (see Captains turn aer 0303). In the second hal o the tran-

scription, the one not shown here, the Captain contacts the high-altitude controller, who givesa clearance or the planned cruising altitude.In analyzing various cognitive processes within the cockpit system, Hutchins and Klausen

(1996) ocus on how inormation about the radio requency o 132.8 MHz moves throughthe system as a sequence o representational states in different media, i.e., rom speech channels(see transcription 1b: 0254) to the internal memory o the F/O, back to the speech channels(see 0300), to the internal memory o the Capt., back to the speech channels again (see Capt.s verbal activity aer 0303) to the physical setting o a device (see Capt. s non-verbal activityaer 0303). Next, however, I will show in more detail how the researchers examine intersub-jectivity as a basis or communication in a cognitive cockpit system. In order to do so, theyconcentrate on the particular verbal and non-verbal activities marked with arrows ()on theright o the transcription 1(b).

ranscription 1(b)

0254 OAK24L NASA nine hundredroger contact Oakland center one thirty two point eight. {F/O pulls his hand back rom the altitude alert knob when AC says contact Oakland center. 2.5 seconds aer the end

o AC transmission, F/O looks at Capt} {Capt looks at F/O.} 0300 F/O Tirty two eight. Capt Tirty two eight? F/O Yeah. Capt OK0303 S/O Tats correct, NASA nine hundred. Capt One three two eight, NASA nine hundred. {Capt twists knob on radio console}

(Hutchins & Klausen 1996, 16.)

Hutchins and Klausen (1996, 22-25) start the interaction analysis rom the Captains lookingactivity at the First Officer; subsequently, the First Officer delivers thirty twoeight; the Cap-tain then asks thirty two eight? Firstly, what is going on in this interaction? Te First Officersutterance thirty two eight assigns a meaning to the Captains staring, i.e., it classifies the Cap-tains looking as a question about the radio requency to be used. Te assignment o mean-ing to the prior look is available or negotiation. Te Captain could, or example, dispute theclassification and claim that he already knew the requency. But this is not the case. Repeating


22/191

22 A I A C I

the radio requency with the rising intonation, the Captains utterance is in line with the FirstOfficers classification o the looking behavior as a request or specific inormation. Secondly,how did the Captain succeed in getting the requency rom the First Officer merely through

his silent look, without using words? In this situation, the pilots share an extensive amounto prior knowledge and expectations o how things should go or, how they typically go in anairline cockpit (or instance, both pilots know, without verbalizing it, that the AC call is sup-posed to be responded to by the Captain). Tis shared knowledge is used by the participantsas a resource or constructing an intersubjective understanding o the particular situation (seeWertsch 1985; 1993; DAndrade 1980). Te intersubjectivity in turn permits efficient kindso communication. Te idea that the Captain was able to communicate his need or specificinormation merely looking at the First Officer was because the glance occurred in a context ointersubjectively shared understandings about the nature o the current situation (Hutchins& Klausen 1996, 24). In this case, importantly, the grounds or the construction o the sharedunderstanding depend on a special distribution o knowledge within the pilot community.

From distributed cognition to the sequential order o action

According to the theory o distributed cognition (Hutchins 1995a), to summarize, the indi-viduals in a cognitive system have shared access to the prior knowledge and expectations thatenables them to become aware o each others conduct and to coordinate their action. In thecockpit case, the pilots were able to construct an intersubjective understanding o their situa-tion based on their access to shared knowledge about how things are supposed to go. Within

this context o shared understandings, the Captain succeeded in getting the radio requencyinormation rom the First Officer by visual rather than verbal means.

Finally, how would the CA approach deal with the present data? Would there be somemajor theoretical and analytical divergences between the theory o distributed cognitionand conversation analysis? o begin with, the conversation analytic researchers seem to havea somewhat different notion o intersubjectivity than the proponents o distributed cogni-tion. Instead o highlighting the meaning o implicit resources (i.e., stored knowledge andexperience), conversation analysis concentrates on the participants use o, and orientation to,explicit resources (i.e., verbal, visual and material practices) in achieving an intersubjective

sense o action-in-progress (see Heritage 1984a, 254-260); consequently, in contrast to thedistributed cognition approach in which intersubjectivity can be treated as a basis or commu-nication (Hutchins & Klausen 1996, 22-25), the conversation analytic perspective considersintersubjective understanding primarily as an outcome o communication, continuously pro-duced and reproduced through the details o the participants talk and visual conduct.

Secondly, the previous case does not necessarily or directly exempliy intersubjectivity asa particular precondition or effective cockpit communication. Instead, this case illustrates thephenomenon o problematic understandings in the sequential course o a cockpit interaction;the pilots locate and resolve the trouble through the temporal coordination o talk and embod-


23/191

I 23

ied action in the cockpit; the processing o the problem momentarily interrupts the currenttask perormance, i.e., the pilot-controller communication, on the flight; the resumption othe suspended activity requires the achievement o intersubjectivity between Captain and First

Officer. Te more detailed sequential analysis could ocus on the 2.5-second pause in cockpittalk, aer the AC message, and the ensuing activities (see arrows marked by me on the tran-scription 1c). Tirdly, the sequentially appropriate environment to start the interaction analy-sis is not the Captains looking activity, as Hutchins and Klausen (1996) suggest, but rather thepreceding activities or moves in the course o cockpit interaction.

ranscription 1(c)

0254 OAK24L NASA nine hundredroger contact Oakland center one thirty two point eight. {F/O pulls his hand back rom the altitude alert knob when AC says contact Oakland center. 2.5 seconds aer the end o AC transmission, F/O looks at Capt} {Capt looks at F/O.} 0300 F/O Tirty two eight. Capt Tirty two eight? F/O Yeah. Capt OK

0303 S/O Tats correct, NASA nine hundred. Capt One three two eight, NASA nine hundred. {Capt twists knob on radio console}

(Hutchins & Klausen 1996, 16.)

At the beginning, the AC instructs the flight crew to contact Oakland center whose theradio requency is one thirty two point eight (see 0254). Aer the AC instruction, there is

a 2.5-second pause in the cockpit talk. What is happening here? At this particular sequen-tial place, the Captain is ailing to respond to the AC message despite the act that it is hisresponsibility to do so in the role o Pilot-not-flying. Aer the 2.5 seconds o non-talk, theFirst Officer shis his gaze towards the Captain; using this looking activity, the First Officervisually addresses the task o responding to the Captain. Te Captain then responds to theFirst Officers gaze by looking back. Te First Officers utterance, thirty two eight (see 0300),displays his understanding o the Captains looking activity as the request or the requencyinormation. By repeating the radio requency with a rising intonation (thirty twoeight?), theCaptain demonstrates his candidate understanding o the prior talk. Once the First Officer has


24/191

24 A I A C I

confirmed this understanding with yeah and the Captain has acknowledged the confirmationwith OK, the Captain starts acting in line with his official duties by responding to, or readingback, the AC instruction (One three two eight, NASA nine hundred.)

In lieu o emphasizing the individual pilots tacit knowledge and belies in the particularsituation, the conversation analysis directs its attention to the sequentially organized courseso talk and social interaction. In line with this, ourthly, the Captains gaze and any other activ-ity accomplished in the cockpit becomes understandable and interpretable in and through itsrelation to the previous turn(s) or move(s); at the same time, importantly, any current actionorms the context or some next action in a sequence, thereore contributing to how the nextaction will be understood (Heritage 1984a, 242). Fihly, rather than the researcher interpret-ing o what is going on in peoples heads, the conversation analyst aims to figure out how par-ticipants themselves interpret and make sense o one anothers conduct. Tese mutual under-standings are displayed in the sequentially organized details o interaction and, because theyare publicly produced they are also available or analysis (Heritage & Atkinson 1984, 11). Ata more practical level, lastly, the reliability o conversation analytic research calls or completeand detailed transcriptions o audio and video recordings, among other things (see Perkyl1997b, 203-207). Within a conversation analytic ramework, the transcription above shouldthus be refined and elaborated to include the exact descriptions o the temporal coordinationo talk and action (such as the duration o pauses within and between the turns at talk, theinitiation o gestural movements relative to talk, etc.).

1.3 OVERVIEW OF CHAPTERS 27

Te aim o Chapter 2 is to set out and describe the methodological and theoretical oundationo the study. Ethnomethodology and conversation analysis (EM/CA) provide the main meth-odological tools or the empirical analysis o the cockpit interaction (see 2.1). EM and CAseek to explain how the orderliness o human activities their regular, patterned or structurednature is produced and made recognizable by the participants acting within local situations(see, e.g., Clayman & Maynard 1995). Te theoretical roots o the current study lie in theanthropology o science and technology and the body o research known as workplace studies

(Suchman 1987; C. Goodwin 1995; Heath & Luff 2000; Luff, Hindmarsh & Heath 2000a).Tese traditions concern the ways in which tools and technologies eature in work and interac-tion in complex organizational environments. Tey ocus on both ace-to-ace interactions, inwhich participants share the same physical space, and the production o activities which arecoordinated between personnel within different domains (see 2.2). Chapter 3 has two basicobjectives: to present and describe, firstly, the social and technical characteristics o the cockpitsetting and, secondly, the research data used in this study (see 3.1-3.7). Te empirical analysiso videotaped cockpit interactions has been conducted along two problem dimensions (seeArminen 2005b; Schegloff 2007). Te first dimension, called cockpit talk-in-interaction, cov-


25/191

I 25

ers the sequence structure, the relationship between turns at talk. Te second dimension, calledcockpit talk-and-action-in-interaction, covers the sequential order, the relative positioning omoves, utterances or actions. Te three analytical chapters (4-6) are constructed around the

problems as ollows: Chapter 4 analyzes the problems the pilots have in speaking, hearing andunderstanding the cockpit talk (see 1stdimension); Chapter 5 extends the scope o analysis byconcentrating on the problems the pilots have in establishing the sequential order o action andactivities during the flight (see 2nddimension); Chapter 6 refines the analysis by consideringthe role and meaning o the gestures o the pilots in processing problematic understandings incockpit interaction. Te main research results and findings are summarized and discussed inChapter 7. Te aeronautical terms and abbreviations relevant to this study are listed in Glos-sary aer Abstract.


26/191

26 A I A C I

C 2

Teory and methodology

Teoretically, this research is based on the anthropology o science and technology and work-place studies (Suchman 1987; C. Goodwin 1995; Heath & Luff 2000; Luff et al. 2000a). Tesetraditions orm a naturalistic approach committed to the detailed study o social and workpractices in complex organizational settings. Te analytic attention is directed towards thetacit body o reasoning and procedures through which the participants produce, make sense oand coordinate their activities with each other. Drawing on the combination o the methods oethnomethodology (EM), conversation analysis (CA) and ethnography, the approach analyzesthe production and coordination o tasks in real-time interaction through talk and visual con-duct. Tis chapter consists o two main sections. Te first Section (2.1) includes a descriptiono the primary method used in this study: conversation analysis with its roots in ethnomethod-

ology. Te second Section (2.2) presents ethnomethodologically inormed research in work-place studies. Te case studies provide a body o empirical observations and findings on thesituated and contingent use o technology in various work environments. Although a second-ary method only, ethnography is o great importance in providing background understandingo the institutionally distinct setting o an airline cockpit. Te ethnographic fieldwork typicallyinvolves the ethnographer participating in peoples daily lives or an extended period o time,watching what is going on, listening to people and asking questions to clariy the issues thatare the ocus o the study (Hammersley & Atkinson 1983, 2). In this study, the knowledgegained through ethnographic strategies is a necessary precondition or the detailed analysis o

videotaped cockpit interactions. Te role o ethnography will be urther discussed in Chapter3; section 3.7 includes a thorough description o the ethnographic work and other related tech-niques that turned out to be useul in enhancing my understanding o the interactions betweenpilots on a flight.

Conversation analysis (CA) offers the most appropriate tools or a detailed scrutiny o thevideotaped cockpit interactions. Te development o this research tradition is closely associ-ated with the ideas o American sociologist Harvey Sacks and his colleagues. Sackss theoreticalthinking was strongly influenced by two contemporary perspectives (see Hutchby & Wooffitt1998, 24-37). One was Ervin Goffmans attempt to promote the acceptance o the interac-tion order o ace-to-ace communication (1959; 1983). Te other was Harold Garfinkels


27/191

M 27

ethnomethodology, which had developed in the empirical studies on practical reasoning andcommon-sense knowledge in everyday lie (Garfinkel 1967). As an introduction, I will brieflysketch the oundations o ethnomethodology and the origins o conversation analysis. Aer

an overall account o the basic principles o conversation analytic methodology (2.1.1), I willintroduce two related concepts which are o special significance or my research: adjacency pairsand intersubjectivity (2.1.2). In doing conversation analysis, particular comparative operationscan be used; section 2.1.3 considers the role o comparisons between ordinary conversationand institutional settings. At the end o Section 2.1, I will recapitulate the affiliations betweenethnomethodology and conversation analysis and the basic assumptions o the methodologyused (see 2.1.4).

Section 2.2 on workplace studies summarizes relevant earlier research done in a range otechnological work environments, including the ground operations control room o an airport,the control room o the London Underground and an airline cockpit. In these complex settings,centres o coordination, the personnel collaborate through various technologies to coordinateco-located and distributed activities and deal with normal, natural troubles in maintainingschedules. An introduction to Section 2.2 concerns the oundations and background o work-place studies. Aer introducing Suchmans (1993) idea o the centres or the coordination ohuman activity (2.2.1), I will turn to more detailed discussion o the empirical studies con-ducted in the field (see 2.2.1.1-2.2.1.4). Te case studies illuminate the social and interactionalorganization o technology in complex organizational settings. Te analyses reveal the ways inwhich work activities are accomplished and routine problems managed in settings saturatedwith technology. Te issues and phenomena to be discussed include the sequential organiza-

tion o human-computer interaction, the interpretative and inerential work by the personnelusing the technical equipment and the various orms o co-participation in collaborative work.Section 2.2 ends with a Summary o key insights and the contributions o ethnomethodologi-cally oriented workplace studies (2.2.2).1

2.1 ETHNOMETHODOLOGI CAL CONVERSATION ANALYSIS

Ethnomethodology, both as a term and as a research orientation, was ounded by the American

sociologist Harold Garfinkel in 1950s. His seminal book Studies inEthnomethodology, whichwas published in 1967, brought this distinctive perspective into the public domain. Garfinkeldid empirical research on how ordinary people use tacit knowledge and reasoning proceduresto produce and recognize intelligible courses o action. In this view, the social order is seen asan emergent achievement resulting rom the joint efforts o members o society acting within

1 Along with ethnomethodology, workplace studies can be approached rom a variety o other perspectives, such as course-o-action analysis, activity theory and distributed cognition (see Heath, Knoblauch & Luff 2000, 305-308; Heath &Luff 2000, 15-19).


28/191

28 A I A C I

local situations (Maynard & Clayman 2003, 174). Te stress on the knowledge-ability o actorsshows how they analyze their circumstances and maintain intersubjective understanding othem (Heritage 1987, 226).

Ethnomethodological studies ocus on the ways in which societal members produce, rec-ognize, and render their actions accountable in context. Te studies emphasize the practicesthat ensure the accountability o actions, i.e., the detailed, collaborative ways in which membersmanage their conduct and circumstances to achieve order in everyday activities. (Zimmerman& Boden 1991, 7.) Te subject o ethnomethodological studies is the ethno-methods thelay practices, knowledge and reasoning procedures through which members o society makesense o and simultaneously accomplish their practical activities. From an ethnomethodologi-cal perspective, then, the social reality is not a preexisting entity but is constantly built up bythe actors. (Coulon 1995, 2, 15-16.)

According to Garfinkel (1967), the practical actions demonstrate the norms the ordinarypeople use in organizing the situation o action. Garfinkels idea o the norms in social actiondiffers drastically rom that propounded by Parsons (1937). Parsons theorizes that the rulesand maxims o conduct are internalized need positions. Te role o a norm is essentially guid-ing or determining conduct in pre-defined scenes o action. From Garfinkels standpoint, thenormative conventions are considered as resources by which participants in interaction rendertheir circumstances both intelligible and morally accountable. Te norms o conduct are con-stitutive o the activities and circumstances to which they are reflexively applied by actors.(Heritage 1984a, 103-134; 1987, 240-248.)

Te lay methodology (i.e., the ethno-methods o people) is uncovered in and through

members use o language and action. For Garfinkel, understanding language amounts tounderstanding utterances or actions carried out in a particular context. Any utterance or actiongains its meaning in the context o its production, i.e., by reerence to the time and place o talk,etc. Te practical, everyday language and action is thus indexical or a given setting. (Heritage1984a, 135-157; Coulon 1995, 17-20.) Te act that language is analyzable only in relation toits context becomes obvious in indexical or deictic expressions (see Levinson 1983, 54-96): thesense o the reerent o this, that, here and there, or example, is contextually determinedwithout exception.

Ethnomethodological studies concern the body o common-sense knowledge and reason-

ing procedures through which ordinary members o society make sense o and act upon theirlocal circumstances (Heritage 1984a, 4).2Te common-sense knowledge in question is specifi-cally that o members. Member as a term does not allude to any social category but to masteryo natural language, where language encompasses both grammar and its use (Garfinkel 1972,304-305). Fundamentally, natural language with its indexical properties provides a windowthrough which to see that the orderliness, rationality, accountability o everyday lie is . . . acontingent, ongoing accomplishment, a kind o work, or doing. (Garfinkel 1972, 304).

2 On the problems o studying commonsense practices, see ten Have 2004, 154-161.


29/191

M 29

Inspired by ethnomethodology, conversation analysis developed rom the collaborationbetween Harvey Sacks and his colleagues Emanuel Schegloff and Gail Jefferson (see Sacks etal. 1974; Schegloff & Sacks 1973). Initially, CA originated in the insights o Sacks, who lec-

tured in the sociology departments o the University o Caliornia at Los Angeles and Irvine in1964-1972. Te tape-recorded lectures were edited by Gail Jefferson and published as a book(see Sacks 1992a & b).3Like Garfinkel, Sacks was intrigued by the levels o social order thatcould be uncovered in the details o language use. Sackss notion o order at all points holdsthat social interaction is a structurally organized phenomenon; not even the smallest detailsshoulda prioribe seen as trivial or uninteresting (Hutchby & Wooffitt 1998, 21-22; Arminen2005a, 12).

o examine the orderly eatures o social action, Sacks began to use recorded data onnaturally occurring talk. Te tape-recordings enabled him to replay, transcribe and study theresearch materials in detail; importantly, the recordings could be observed and analyzed byothers as well (see Sacks 1984, 26). By means o recorded data, Sacks was able to observe themembers sense-making the establishment and maintenance o shared understanding as itoccurred in a situation o action (Hutchby & Wooffitt 1998, 33). Sackss concerns with orderlystructures o everyday human conduct with the help o real-world data are the main charac-teristics o his work and o conversation analytic method. Te methodological principles oconversation analysis are discussed urther next.

2.1.1 Te methodological basis of conversation analysis

Conversation analysis (CA) is the study o talk. o put it more precisely, it is the systematicanalysis o talk-in-interaction (Schegloff 1987a) the term encompassing talk and otherinteractional activity such as physical activities, gestures and the paralinguistic eatures o talk.Te aim o CA is to uncover the tacit, organized reasoning procedures which inorm the pro-duction and interpretation o naturally occurring talk. CA seeks to explicate these procedureson which participants rely in producing utterances and making sense o one anothers talk.(Hutchby & Wooffitt 1998, 1.) Te ocus is on how particular social actions are organized andlocally produced through talk. Te issue o what a given utterance is doing in the service o

some identifiable action such as requesting, joking, complaining or closing the conversation is relevant. (Schegloff & Sacks 1973; Schegloff 1987b.)CA is based on three primary assumptions (see Heritage 1984a, 241-243). Starting rom

the most undamental one, social action and interaction is structurally organized. All aspectso social interaction can be ound to demonstrate stable organizational patterns o action towhich participants orient. Like other social institutions or conventions, these organizationsare independent o psychological, sociological or any other characteristics o the participants.

3 Tis publication is posthumous (Sacks died in a car accident in 1975).


30/191

30 A I A C I

Knowledge o these organizations is part o the competencies the participants use in organ-izing their conduct and interpreting the conduct o others. Te participants orientation tostructural organizations is relevant or the ways they design utterances in the course o interac-

tion (Schegloff 1991).Te second assumption holds that the utterances are simultaneously context-shaped andcontext-renewing. Here, the meaning o context is twoold, reerring both to the local con-figuration o prior activity in which an utterance occurs and to the larger sequence o activitywithin which that configuration occurs. (Heritage 1984a, 242; 1989, 22.) How any actionscontribution to the ongoing sequence o actions is understood depends on the context inwhich it occurs. In this sense, utterances are shaped by the context. While every current actionwill itsel orm the context or some succeeding action, it will contribute to the contextualramework within which this action will be understood. Any current action can thus also beseen to renew the context o a succeeding action. (Heritage 1984a, 242.)

Te third assumption maintains that no order o detail in interaction can be dismissedas trivial or uninteresting beore it is subjected to analysis a significant argument or howconversation analytic research is done (Heritage 1984a, 242-243). An empirical approach tothe study o social interaction is avoured instead o prior theory construction. Te idealizationo data is seen as hindering the development o appropriate analysis, and is thereore shunned.Te analysis is data-driven, i.e., developed rom phenomena which are evidenced in the detailso interaction data. Te social interaction in all its details is best approached through the analy-sis o recorded, naturally occurring data (Heritage 1989, 22; Hutchby & Wooffitt 1998, 13).

Conversation analytical studies describe regular orms o organization in materials pro-

duced by different speakers. Te participants produce and orient to these regularities as norma-tive grounds or action. In other words, the participants hold themselves morally accountableor departures rom such regularities. At this point, the analysis o deviant cases in whichsome proposed regular procedure is not realised becomes relevant (Heritage 1984a, 243-244.) Te deviant cases strengthen the analysis, as the researcher may either modiy the originalaccount so that deviant eatures can be included in it, or, preserve the original analysis and pro-duce a separate analysis o deviant cases (Arminen 2005a, 70-71; Clayman & Maynard 1995,8-9).4

2.1.2 Adjacency pairs and intersubjectivity in conversation

CA describes the procedures and expectations through which participants produce and makesense o ordinary conversational conduct. Conversation analytic studies examine the organiza-tional eatures o talk as they are displayed and understood in the actual events o interaction.

4 For the analysis o deviant cases, see Schegloff (1972) on telephone interaction; Heritage & Greatbatch (1991) andHeritage (1998) on news interviews; Arminen (1998) on AA (Alcoholics Anonymous) meetings.


31/191

M 31

Te units o analysis in CA are larger than the individual utterances alone. Te analysis con-cerns the ways in which utterances accomplish particular actions in terms o their placementand participation within sequences o actions. Consequently, the scope o these studies covers

action sequences and their component unit turns that are conceived as turns-within-sequences.(Heritage 1984a, 245.)o begin with the description o sequentially organized actions, the production o some

current turn proposes a local definition o the situation to which subsequent talk will be ori-ented (Heritage & Atkinson 1984, 5). More specifically, the phenomenon o the sequentialimplicativeness o a turns talk (Schegloff & Sacks 1973, 296) takes place when some currentturn projects a relevant next action, or range o actions, to be perormed by another speakerin the next turn. Tis projection o the relevant next action is conventionally accomplishedthrough the identifiable pair o actions known as an adjacency pair (Schegloff 1972; Schegloff& Sacks 1973, 295; see also Heritage 1984a, 245-253).

Schegloff and Sacks (1973, 295-296) characterize the adjacency pair as (1)a sequence otwo utterances which are (2)adjacent, (3)produced by different speakers, (4)ordered as a firstpair part and second pair part, and (5)typed, so that a first part requires a particular secondpart(s). Instances o pair types include greetings and return greetings, questions and answers,invitations and acceptances/declinations and so on. Te adjacency pair structure is normativein character. o exempliy that point, questioners orient to the act that their questions areramed within specific normative expectations. Tese expectations have sequential implica-tions or the next speakers to perorm a particular kind o action namely, to answer the ques-tion. (Heritage 1984a, 249.)

Providing the next speaker ails to respond, his/her behaviour becomes accountable. Tefirst speaker may iner that the recipient has some trouble in responding. By repeating the ques-tion, the first speaker displays that the answer to the original question was appropriate and itis officially absent (Schegloff 1972, 364).5Te regular occurrence o certain paired actions isexplained by the property o conditional relevance (Schegloff 1972), which stipulates thatthe production o the first pair part makes a corresponding second pair part both relevantand expectable (see also Schegloff & Sacks 1973). Tis constraint allows the speakers (and theanalysts) to recognize whether some conversational events, such as answers to questions, arenoticeably absent (Heritage 1984a, 249).6

Whilst the apparatus o adjacency pair is important in many respects (see Heritage 1989),only a ew points are discussed here. As mentioned, the paired action sequences offer a system-atic basis on which the speakers, and the analysts, can determine whether relevant next actionsare noticeably and specifically absent. Te organization o adjacency pair is also significantin terms o how intersubjectivity the mutual understanding o ongoing (inter)action isaccomplished and displayed in talk. As Schegloff and Sacks (1973) describe:

5 On the location and management o official absences in cockpit talk, see Chapter 5: section 5.2.3.6 On the procedures through which the speakers deal with no responses to their assertions, see Pomerantz (1984).


32/191

32 A I A C I

by an adjacently positioned second, a speaker can show that he understood whata prior aimed at, and that he is willing to go along with that. Also, by virtue othe occurrence o an adjacently produced second, the doer o a first can see that

what he intended was indeed understood, and that it was or was not accepted.Also, o course, a second can assert his ailure to understand, or disagreement,and inspection o a second by a first can allow the first speaker to see that whilethe second thought he understood, indeed he misunderstood. (Schegloff &Sacks 1973, 297-298.)

Te production o a first pair part, such as an invitation, makes relevant a particular actionto be done next, or a limited set o such actions. By acceptance, among other alternatives,the second speaker not only complies with the constraints o the adjacency pair structure butalso displays an understanding o what the prior utterance was doing (see Sacks et al. 1974,728-729).7Te producer o the initial turn may comment on or correct the second speakersunderstanding in the third turn o the sequence. Essentially, the adjacent positioning enablesco-participants to display their understandings o the ongoing talk and to recognize possiblemisunderstandings in conversation (see Schegloff & Sacks 1973, 297-298).

Te structural organization o conversation is managed on a turn-about basis. Teseresources provided by the mechanisms o talk-in-interaction enable the co-participants topublicly display and continuously update their intersubjective understandings (Heritage &Atkinson 1984, 11). In this regard, mutual understanding is a methodical achievement (Gar-finkel 1967, 38-42; 1972, 315-321; Psathas 1977, 89-96; Heritage 1984a, 254-260). Since

public understandings are a kind o by-product o sequentially organized action, the issue ounderstanding itsel is only rarely discussed by the participants. In other words, turn-by-turnorganization o action enables participants to rerain rom explicitly or literally confirmingtheir understandings to one another. (Heritage 1984a, 259.)

Te activity eature o adjacency pair structure is central in terms o action interpreta-tion. Te action used by the first speaker orms a basis or interpretation o what the secondspeaker will say. In the succeeding action, the second speaker displays an interpretation o prioraction. Tis interpretation is publicly available to the first speakers to determine whether theywere understood. (Heritage 1984a, 254-256.) A mutual understanding entailing the courses o

interpretation is thus operationally structured within ongoing interaction (Garfinkel 1967, 31;1972, 321). Because o their public nature, these understandings are also available or analyti-cal scrutiny by social scientists (see, e.g., Sacks et al. 1974, 728-729; Heritage & Atkinson 1984,11; Hutchby & Wooffitt 1998, 15-17.)

7 Te grasp o what the utterance is doing may differ between overhearers or analysts o talk and co-participants in theactual course o conversation: the sense o what is going on in talk may seem ambiguous or the ormer but not necessar-ily or the latter. (See Schegloff 1984b.)


33/191

M 33

2.1.3 Ordinary as against institutional interaction

Conversation analysis started rom the study o the mundane, everyday conversation that is

the predominant medium o interaction in the social world: we use it to raise our children,maintain amily relationships, riendships, love affairs, and so on. It is suggested, then, thatthe basic orms o mundane conversation constitute a kind o bedrock against which moreormal or institutional types o interaction are distinguished. A comparative analysis betweeninstitutional interaction and normative procedures o interaction in ordinary talk provides oneimportant way o doing conversation analytic research. (See Sacks et al. 1974, 729-731; Herit-age 1984a, 238-240; Drew & Heritage 1992a; Hutchby & Wooffitt 1998, 146-149; Arminen2005a, 43-47.)

Te term institutional interaction reers to the talk through which participants conducttheir particular institutional tasks and goals (Perkyl 1997a, 177; Drew & Sorjonen 1997,92).8Interaction is institutional to the extent that participants orient to their institutional orproessional identities to perorm particular work activities, an orientation evidenced in thedetails o language through which the participants manage their institutional tasks (Drew &Heritage 1992b, 3-4; Drew & Sorjonen 1997, 97). Importantly, thereore, the institutionalcontext is not an external constraint causing certain orms o interaction to occur, but ratherthe ongoing accomplishment o the participants in interaction (Hutchby & Wooffitt 1998,171; Arminen 2005a, 19).

Te first systematic analysis o institutional interaction was done by Atkinson and Drew(1979) who compared interactions in the courtroom and everyday settings. Some o their later

studies also concern interaction in the legal environments (Drew 1985; 1992; Atkinson 1992;see also Pomerantz & Atkinson 2003; Pomerantz 1987). Other research areas on institutionalinteraction include, or example, classroom or instructional settings (see, e.g., Mehan 1979;1985; McHoul 1990; Psathas 1992), news interviews (see, e.g., Heritage 1985; Greatbatch1988; Clayman 1988; Heritage & Greatbatch 1991), counseling and social welare (see, e.g.,Linell & Fredin 1995; Silverman 1997; Sarangi 2000), medical settings (see, e.g., Sharrock &Anderson 1987; ten Have 1991; Maynard 1992; Heath 1992a; Ruusuvuori 2000) and archi-tectural practices (Heath & Luff 2000, 155-178; Mondada 2006).

Tese studies show how participants, through the design o turns and sequences o turns,

orient to particular institutional identities, thereby managing their practical tasks in a giveninstitutional setting (Drew & Sorjonen 1997). Te participants may be either proessionalmembers o the institutions or their clients (students, patients, etc.). Te studies on institu-tional interaction concern the ways in which institutions and organizations, such as hospitalsor schools, are talked into being (Heritage 1984a, 290; see also Arminen 2005a). Te aim isto describe the patterns o interaction in institutional settings and to indicate, i possible, how

8 Drew and Sorjonen (1997) use the term institutional dialogue.


34/191

34 A I A C I

they differ rom the characteristics o ordinary talk (Heritage 1989, 33; Drew & Sorjonen1997, 106).

Te institutional eatures o interaction are oen maniested in differences rom ordinary

interaction. Tese differences tend to involve 1) specific reductions o the range o options oraction that are available in mundane interaction and 2) systematic specializations o the inter-actional unctions o the remaining activities (Heritage 1984a, 239-240; Drew & Heritage1992b, 26-27). Te unique fingerprint o any orm o institutional interaction rests on a set oconversational practices that are differentiated rom ordinary talk and rom other institutionalinteractions. Te institutionalized reductions and specializations o conversational options areconventional and exposed to processes o social change. (Heritage & Greatbatch 1991, 95-96;Drew & Heritage 1992b, 26.)

In contrast to ordinary conversations, many kinds o institutional encounter exhibit astandard shape or order o phases. Institutional activities are conducted in a repetitive andfixed sequence, which lends a distinctive structure to such encounters. (Drew & Sorjonen1997, 109-110; see also Zimmerman & Boden 1991, 13.) However, the boundaries betweeninstitutional and ordinary talk are not rigid. For example, the activity o cross-examination isnot restricted simply to legal settings, but may also occur in private homes. Similarly, peoplein workplaces may talk about things unrelated to their work. A single interactional exchangecan thus involve both institutional and conversational eatures o talk. (Heritage 1984a, 240;Drew & Sorjonen 1997, 92-94.)

Te basic comparison between institutional and ordinary talk can begin rom the analysiso turn-taking. Sacks, Schegloff and Jefferson (1974) in act introduced the idea o doing com-

parative research on different turn-taking systems in their early paper on turn-taking, a systemwhich allocates turns between participants. A turn consists o one or more turn-constructionalunits (CU). Te unit types involve single words (such as hello,yes), phrases, clauses and sen-tences. When the current speaker completes any such unit, s/he reaches a possible transitionrelevance place (RP); i.e., a point at which a potential next speaker may start a turn. In ordi-nary conversations, turn-taking is managed locally, without any predetermined arrangements.In more ormal settings, by contrast, allocation o turns is based on a fixed order o who takesthe floor and when. (See Sacks et al. 1974.)

Heritage and Greatbatch (1991) compared the turn-taking procedures between news

interview interaction and ordinary conversation. In the news interviews, unlike conversation,the types o turns are pre-specified to permit the interviewer (IR) to ask questions and theinterviewee (IE) to answer them. Trough the question-answer sequences, the parties not onlyconstitute themselves as IR and IE, but also collaboratively maintain the interview charactero the interaction. Moreover, both IR and IE may use a number o turn constructional units inproducing their talk. Te entitlement or long turns in news interviews departs rom ordinaryconversation where turn-taking system tends to minimize turn size (Heritage & Greatbatch1991, 97-102; see also Sacks et al. 1974).


35/191

M 35

In everyday conversation, however, the speakers may use multiple turn constructionalunits, among other things, in telling stories. Te listeners actively participate in storytellingwith particular response tokens such as continuers (uh huh,yeah, etc., see Schegloff 1982) or

objects treating the prior talk as inormative or news (oh, really, etc., see Heritage 1984b). Innews interviews, the IE may produce long responses to the IR questions. In this case, however,the IR rerains rom producing response tokens in the course o IE talk. Te systematic absenceo tokens is a means by which news audience is maintained as the primary addressee o IE talk.9By sticking to questioning activities only, the IR adheres to the provisions o turn-taking in thenews interview context. (Heritage & Greatbatch 1991, 107-113.)

Te social world o news interviews, as well as o any other institutional orm o interac-tion, is invoked and made actionable in and through talk. Te institutional realities are notonly talked into being but they also exist in and as documents, buildings, official arrange-ments, and so on, as Heritage (2004, 222-223) points out. Te conversatio

Documents

Achievement of Intersubjectivity in Airline Cockpit Interaction_Diss