II.1. The issue

4 The analysis of gestures in work situations, which has been the object of a recent review by Leplat (2013), is an important field of study in French literature on ergonomics and work psychology.

5 Chassaing (2006) has contributed to the definition and understanding of professional gestures by developing a hierarchical model that describes the organisation and structure of gestures learned on the job in the automotive sector. This work has made it possible to establish a clear distinction between the concepts of movement and gesture. In particular, she has provided a categorisation of professional gestures into four groups: (1) the gesture is mixed, as it “requires a sensory, cognitive and motor activity”; (2) the gesture is invested, as it is “intentional and oriented toward different purposes: the system, oneself and others”; (3) the gesture is situated; and (4) the gesture is constructed, as it is “the product of a history, a past and is continuously developing”.

6 Gesture involves several dimensions: psychological, organisational, economic, biomechanical and cognitive (Bourgeois & Hubault, 2005). The study of musculoskeletal disorders, because they are often caused by repeated gestures, has led to a detailed study of the cinematic and motoric aspects of gestures (Vézina, Prévost, Lajoie, & Beauchamp, 1999). In the aviation sector, Aubert (2000) sought to understand what exactly makes up the specificity of skills linked to the gestures of aircraft painters by analysing work situations, for a better understanding and recognition of this profession during an industrial process of reorganisation.

7 In the food industry, observation of meat boners for vocational training purposes (Ouellet, 2012) showed how the choice of variants of a gesture is linked to the nature of the equipment as well as to training, for example, the quality of the tool, or variants in the material (in this case, whether it was a left or a right back quarter that was to be deboned). Authier, Lortie, and Gagnon (1996) described a similar variability in the work of expert handlers: how they adapted their posture to the size, weight, or placing of the objects handled. These authors insist that expertise involves this capacity of adapting an activity to the context, and therefore also emphasise the necessity of examining gesture as an integral part of activity, as opposed to conducting limited musculoskeletal analyses.

8 Clot and Leplat’s (2005) clinic of activity uses the crossed self- confrontation method as a conceptual tool for studying the development of professional gestures and individual and collective practices through reflexivity. This method has been applied to various cases of professional gestures, such as, amongst others, the gesture of braking a train (Fernandez, 2001), the gestures of grave diggers (Simonet, Caroly, & Clot, 2011), mainly for the prevention of musculoskeletal disorders. Those authors have observed the difficulty in accessing, analysing and transmitting the knowledge underlying such professionalised gestures. Thus, the multi-dimensional character – situated (Suchman, 1987), distributed (Brassac, 2008; Hutchins, 1995) and embodied (Varela, Thompson, & Rosch, 1991) – of the knowledge required for carrying out a professional activity would appear to require no further demonstration. The exact nature of the experience is at the core of the issue, as it is individual, unique and shaped by each person over a period of time and depends on the people and situations encountered and experienced. It is this subjectivity, specific to each person, which makes the task of passing on professional experience so difficult. Furthermore, such experience is also considered to be almost ineffable and, because of this strong tacit dimension, not easily transferable (Artemieva, 1980; Kornilov, 2014). Nevertheless, even if experience cannot be literally transmitted, it can be shared to some extent. That is why the tradition of socialisation practice (Nonaka & Takeuchi, 1995), such as professional mentoring (Castéra, 2008; Schön, 1983), is an option frequently used by the manual work sector to enable experts and novices to engage in face-to-face intellectual, social, and experiential sharing directly in the workplace. Unfortunately, the kind of training environment that makes it possible to provide a long period of overlap between the departing expert and the new arrival is less and less viable as a result of increasing time pressures and the mass retirement of experts. This demographical transition requires the replacement of the experts involved in the first industrial development, those whose careers began in the 1960s and 1970s. The challenge for organisations is two-fold: firstly, to safeguard the intangible cultural heritage present in the minds of its experts, and, secondly, to pass this knowledge on to newcomers. Thus, seeking and/or developing novel alternatives for addressing this twofold issue of knowledge management for professional mass-training purposes is an urgent requirement.

II.2. Verbalisation of work activity

9 Analyzing the possibilities for creating or developing training systems for the transfer of professional knowledge, Shadrikov (1974) identifies three categories of objects in terms of their features: those that can be verbalised, those that can be represented in the form of visual images, and those with the so-called sensory features that are difficult to verbalise or represent graphically. According to Kornilov (2014), in the case of the transfer of professional experience, practical problems occur in different forms that are more or less easy to transfer by means of speech: tacit knowledge, individualised knowledge (which contains not only the features of the object but the particularities of the actions performed on it by a specific professional), unconscious stereotyped knowledge related to skills, heuristics, evaluation strategies, knowledge related to specific context, and so on. Kornilov (2014) showed that the problems that occur in professional interaction are determined, in some cases, by the presence of individualised knowledge, knowledge that, when identified with a single word or phrase, has different meanings for different workers.

10 In Europe, as in Russia, a range of techniques of verbalisation has been developed. They can be used to facilitate the verbal transfer of subjective experience at different moments of the action, and according to different types of instructions and settings. Among the best known are, the thinking-aloud protocol (Ericsson & Simon, 1980), or post-hoc verbalisations, including the self-confrontation interview (Theureau, 1992), the crossed self-confrontation method (Clot, Faïta, Fernandez, & Scheller, 2001), the “explicitation” interview (Vermersch, 1990), or the subjective re-situ interview (Rix & Biache, 2004). The method of “instruction to one’s double” (Oddone, Rey, & Briante, 1981) is another technique that involves describing one’s approach to work and instructing another operator as if he/she were a double. For further methods, the reader can refer to the book by Bisseret, Sebillotte, & Falzon (1999). Garrigou, Carballeda, & Daniellou (1998) also list commonalities in verbalizing know-how techniques:

11 –  “They avoid formulations involving explanation;

12 –  They avoid rational and causal-type questions, and therefore the use of ‘why’ which will require justifications from the subjects rather than descriptions of their actions;

13 –  They use interview methods that encourage descriptions of the action process;

14 –  They produce questions that promote the systematic description of the sensory modalities and their different components (sight, hearing and taste, as well as kinaesthetic and olfactory aspects).”

15 In Russia, some other types of verbalisation protocols have been developed, as, for example, (1) giving instructions asking individuals to explain where and how objects can be used whilst prohibiting certain types of verbal descriptions, for example, those which refer to the size, color or material of objects,(2) setting a task to create a plan for certain activities in a verbal format and within a time limit,(3) facilitating the updating of instrumental experience by presenting a list of functional features to be ranked, compared or classified and so on (Kornilov, 2014). The problem of the verbalisation of knowledge encountered by professionals in their practical activity is also investigated through the identification of personality traits related to the verbalisation taking place in practical thinking, which is thinking related to action.

II.3. Methodological gap

16 The psychological verbalisation techniques cited above seek to investigate work activity, but they were not primarily developed for the practical issue of know-how transmission. They do not link the content of collected verbal data and the way in which they can be used for practical purposes. In the field of knowledge management (Earl, 2001; Ermine, 2010; Nonaka & Takeuchi, 1995), methods have developed over the past twenty years, essentially focusing on providing tools and techniques for structuring and presenting knowledge, but the use of video as a medium to convey expert knowledge remains a challenge. Some homemade videos are produced by trainers or field operators in the production industry to meet their needs for innovative teaching tools, but they generally use a prescriptive approach that shows “what-must-be-done” according to rules, thus missing out on the variability of actual practices that go beyond the prescribed. Furthermore, novices also need to share, feel and even “live” the experience in order to be able to fully grasp an expert’s cognition.

17 To sum up, European literature proposes few practical methods enabling video-based data collection for transmission purposes in educational settings. The same assessment applies to Russian literature. Knowledge transfer in the workplace is investigated in psychology in terms of the specific psychological particularities of practical thinking and the verbalisation of its procedural characteristics, rather than in terms of an elaboration of a concrete methodology for studying professional gestures for educational transmission purposes.

18 On the basis of our empirical research within the context of the main French energy producer, we make the argument that using real cases of gestures, collected, analyzed and transmitted through video from the direct point of view of the expert, can provide an efficient alternative for both filling this methodological gap and answering the twofold requirements of the organisation. We present below a qualitative research method centered on the expert’s perspective, which we have designed to render explicit the expert’s tacit knowledge. Based on embodied (Varela et al., 1991), situated (Suchman, 1987), and distributed (Hutchins, 1995) cognition models, this method uses and combines principles of the Activity Theory, the Perceived Quality approach, visual ethnography, and verbalisation methods in order to analyse expert professional gestures for transmission purposes. The next section presents our main theoretical and methodological principles.

III.1. Activity theories

19 In parallel to the French development of ergonomics centered on work activity (Leplat & Cuny, 1974; Ombredane & Faverge, 1955), activity theories centered on the subject appeared in the USSR in the early twentieth century.

20 Since then, many versions of activity theory have been developed worldwide (Bedny & Karwowski, 2004; Engeström, 2000; Nardi, 1996; Stetsenko, 2005; Von Cranach, Kalbermatten, Indermühle, & Gugler, 1982). Engeström’s activity theory is particularly popular, having been developed to take into account organisational characteristics of the activity (such as the division of labour, tooling, and rules).

21 While theoretical work on Activity Theory is sophisticated, methodological approaches to its application in studying work activity lag behind. Our own research is based on the psychological structure of the activity developed by the Russian school of Activity Theory (Leontiev, 1978; Rubinstein, 1922, 1940) but makes little use of its philosophical aspects linked to personality, conscience, or emotions. [1]

22 Activity Theory in the Russian tradition has been developed primarily by Leontiev (1978) and Rubinstein (1922, 1940), who elaborated two somewhat similar versions of the structure of human activity. Generally, Activity Theory aims to render explicit the relationships between the different components of an activity, namely, motives, goals, tasks, actions, and operations.

23 In Leontiev’s version, motives, interpreted not as a kind of experience of need but as a material or ideal object of need, determine the whole activity, which is carried out by means of actions directed by conscious goals. The expected result of activity, which is the goal, defines the actions necessary to reach it. However, actions are ultimately determined by the motive of activity. Indeed one activity can be completed by means of different actions, and one action can be a component of different activities. Operations are concrete ways to realise an action and correspond to conditions of activity. Thus, in the Leontiev’s theory, motives refer strictly to the whole activity, goals refer to actions and operations refer to concrete situations.

24 Rubinstein’s conception of activity structure is similar to the one proposed by Leontiev in the sense that the same activity components are outlined. However, according to Rubinstein, there is no strict correspondence between the motive and activity or between the goal and action. Activity-related motives and goals, unlike action-related motives and goals, usually have an integral nature, express a subject’s general orientation and, thus, are called initial motives and final goals. At different stages of activity, activity-related motives and goals generate different specific motives and goals, which characterise actions.

25 Both activity theories provide an efficient conceptual framework for investigating “human and environment” interactions, while Rubinstein’s version is focused on the philosophical and theoretical developments of an activity linked to concepts of conscience, personality and the life of the subject.

26 In the context of both analyzing and giving a psychological description of a practical work activity, the components being evaluated are as follows: the motives that induce activity; the goals – conscious representations of future results to be reached; the subgoals – intermediate stages in reaching the goal; and the actions and operations carried out by a subject in order to reach the goals. To these should be added the tools used by the person or the organisation to transform the external environment.

27 Some of the components, such as operations, can be observed and recorded, while others, such as motives or goals, are invisible. These invisible components can be revealed and characterised only by means of techniques developed to analyze and evaluate subjective representations of individuals. The modes of investigation that we used to trace relationships between externally observable parameters of activity (movements) and their internal components (thoughts and so on) are presented below.

III.2. Subjective evidence-based ethnography (SEBE)

28 One of the ways of collecting observable components of an activity is by using video ethnography. This naturalistic approach enables the capture and study of flows of activity during real-life practice in the field. Over the past ten years, visual ethnographic methods and techniques have been expanding. They have been used by social scientists in various disciplines (sociology, anthropology, psychology, ergonomics), in order to observe and study human activity in a qualitative way (Goldman, Pea, Barron, & Derry, 2007; Heath, Hindmarsh, & Luff, 2010; Lahlou, 1999; Mondada, 2003; Pink, 2007).

29 As activity is distributed, the workers use not only their embodied competence but also a series of physical tools and instruments to transform material objects (Rabardel, 1997), mediating structures to operate cognitive operations (Hutchins, 1995). Other colleagues and various non- human entities are also involved in activity. Furthermore, action emerges in context: it is situated (Lave, 1988; Suchman, 1987). Consequently, when we try to reconstruct the action and understand what the human actor did, we need to account for all the contextual elements which contributed to the action. These elements cannot be recovered easily from the operator’s memory alone. Therefore, we use video to bring the complete situation back to the operator at the time of recall; this enables the operator to point to the relevant elements that supported the action while it was happening. We use a first-person perspective capture to reconstruct the situation as it was seen by the operator at the moment of action, because, in complex situations, what was relevant for immediate action is usually what the operator paid attention to.

30 First-person perspective is provided by subcams (subjective cameras) (Lahlou, 1999), which are miniature video cameras worn at eye level by the subject while he/she is performing his/her activity. In a second stage called “replay interview”, the subject is asked to describe and explain her/his covert behaviour, as he/she remembers it, to the researcher while watching her/his own subcam recording. The technique of the replay interview is similar, in its aim, to well-known ex-post verbalisation techniques (see II.2), as they all seek to immerse the operator ex-post in his/her own activity, by confronting him/her with a track (not necessarily video) of his/her activity in order to make him/her verbalise her/his cognitive processes corresponding to earlier actions. The specificity of our approach is that it integrates the principles of Activity Theory and Perceived Quality Theory, and, furthermore, it uses first-person perspective recordings of the expert’s activity in order to stimulate his/her memory, so as to take advantage fully of all of the elements of the techniques mentioned above. Thus, this technique, which entails the use of first person perspective digital recording as a basis for analytic replay interviews with participants, is called Subjective Evidence Based Ethnography (SEBE) (Lahlou, Le Bellu, & Boesen-Mariani, 2015). Several research teams now use or have developed customised first- person perspective recording tools similar to the subcam for studying activity in extreme settings and in training contexts (Myrvang Brown, Dilley, & Marshall, 2008; Omodei & McLennan, 1994; Rix & Biache, 2004). Such wearable devices for operators obviously provide practical advantages for practitioners and researchers who analyze work, by facilitating the task of data collection. Indeed, the subcam makes it possible to capture activities on film even while the subject is moving. Regardless of the location or the bodily movements, the device inevitably follows the operator’s focus of attention and action, enabling to make a continuously situated recording of his activity. In this way, data can be collected without any outside observer; the operator is free to organise her/his activities and movements. But beyond this convenient aspect, we have noticed that the “subfilms” empower operators to recall very precisely and accurately what they did, thought and felt at the time of action, even some weeks after having experienced the real-life activity, probably because the subfilms feed the viewer with the exact context of action, attention focus and motor cues. One key advantage of SEBE is, therefore, its introspective “power” for triggering recall of lived experience. Another important advantage is its ability to provide the analyst with a detailed step-by-step understanding of the constituents of activity: goals, subgoals, determinants of actions, decision-making processes, and so on.

III.3. The perceived quality approach

31 The Perceived Quality approach (Nosulenko & Samoylenko, 2001) can be used to analyze both observable and subjective components of acti­vity. This framework is useful in the analysis of practical activities taking place in real-world contexts, and in the identification of which elements of the environment are considered by individuals to be the most valuable in the course of achieving their aims. From this perspective, the definition of perceived quality can be formulated as follows: “A set of subjectively relevant characteristics of the world and the activity, which comes about in the subject with the objective of attaining her/his goals” (Nosulenko, 2008).

32 It is impossible to define a priori the number of so-called “objective” components to be measured in the course of observations. We can only take an interest in those that are pertinent for the subject, the features that manifest themselves in the “perceived quality” of the gesture. The goal, then, consists in matching those components of perceived quality (in our case, the oral explanations given of the gesture by the operator– see section IV) with the observed components of the events (the professional gesture performed within a technical context).

33 The perceived quality approach begins with identifying the aspects of the object or system that are subjectively valuable for certain individuals in the course of the given activity. The strategy is to elaborate, and then empirically prove, a number of hypotheses concerning the parameters of actions. This assumes the inclusion of these aspects in the individual’s perceived quality, and the setting up of an evaluation system based on them as they appear in open-ended individual evaluations. Which aspects of the artifact are responsible for determining these subjective characteristics will only appear in the course of careful examination of an activity. It is by confronting these evaluations with moments of actual activity and their description, using activity theory, that this will happen. For a detailed description of how this is done in practice, please refer to Lahlou et al., (2012), sections 10.2, 10. 3 and 12.3, and Le Bellu, (2011).

IV.1. Preparation stage

36 This stage involves organisation of the study and familiarisation with the activity to be examined, as well as the collection of a range of documents (procedures, handbooks, diagrams of the installation, training documents, and so on), selection of the expert by the organisation and informal interviews with the expert(s). This stage enables the analyst to become acquainted with the aim and content of the activity to be studied. Selection of the place and time of data collection also takes place during this stage.

IV.2. Data collection protocol

IV.2.2. An example of application

42 Figure 1 illustrates the data collection protocol described above by providing screenshots of the video recordings and transcriptions of commentary extracted from the recordings of the gesture “plugging in a 380V valve power supply cell”. This figure provides a view of the same scene, at the same moment, from an external camera (Figure 1A), from the first-person perspective provided by the subcam worn by the expert (Figure 1B), and commentary provided according to the instructions of the goal-oriented thinking-aloud protocol (Figure 1C).

43 It should be noted here that this example does not involve specific manual skills, like those that are used in some professional gestures such as carving or painting, or gestures involved in specific attentional or diagnostic skills, such as driving or hunting. In this case, the movements are simple (pressing a button, turning a handle, pulling out a drawer) and any difficulty resides in the proper sequencing of operations. The experts focus, therefore, specifically on the cognitive aspects of the task, since explaining how to press a button would not appear problematic. In other settings, fine control of the movement can have implications for the final result (Vézina et al., 1999) or may lead to musculoskeletal trauma (Authier et al., 1996). In such settings, fine control is then the main focus of operators and analysts. This is a reminder that every activity is different, and that descriptions should focus on what is especially critical, otherwise they may appear lengthy or irrelevant to operators.

Figure 1: Extract of video and verbal data collected through the application of the data collection protocol. Figure 1: Extrait de données verbales et vidéo obtenues par l’application du protocole de recueil de données.

Figure 1: Extract of video and verbal data collected through the application of the data collection protocol. Figure 1: Extrait de données verbales et vidéo obtenues par l’application du protocole de recueil de données.

IV.3. Analysis of collected verbal and visual data

IV.4. Replay interview

IV.4.2. An example of application

69 After the hypothesis of cognitive analysis and the pre-video editing have been presented to the expert during the replay interview, the model is adjusted in accordance with the expert’s comments. The cognitive model below was built for the application of the “plug in” gesture (Figure 2).

Figure 2: The cognitive model (“tree of goals”) of the gesture “plugging in a 380V valve power supply cell”. Figure 2 : Modèle cognitif (« arbre de buts ») correspondant au geste « brancher une prise 380 V à clapet ».

Figure 2: The cognitive model (“tree of goals”) of the gesture “plugging in a 380V valve power supply cell”. Figure 2 : Modèle cognitif (« arbre de buts ») correspondant au geste « brancher une prise 380 V à clapet ».

IV.5. Creating an educational real-experience based resource

IV.5.2. An example of application

75 All the raw data collected and analysed provided the basic material for building a MAP organised around the cognitive structure of the gesture (Figure 3).

Figure 3: The “tree of goals” chapter in the Multimedia Platform for Apprenticeship (MAP). Figure 3 : Le chapitre « arbre des buts » présenté sur la plateforme multimedia d’apprentissage (MAP).

Figure 3: The “tree of goals” chapter in the Multimedia Platform for Apprenticeship (MAP). Figure 3 : Le chapitre « arbre des buts » présenté sur la plateforme multimedia d’apprentissage (MAP).