Power and group work in physical education

Introduction

The impetus for this article comes from critique we received on an earlier paper (Barker et al., 2015a). In that text, we suggested that during group work, relations between group members develop based on their knowledge levels vis-à-vis their task. Some members ‘become’ experts while others become novices, relative to one another based on their task proficiency. The critique that we received involved the claim that relationships between pupils exist prior to the task and that a range of factors influence how pupils interact. Specifically, the commentator claimed that ‘student interactions, and how they take up the expert or novice role, are never neutral or dislocated from the broader cultural context’ (personal communication).

In our view, this is an excellent point and one that generates important questions. For example, how are cultural factors present during group work? Or more specifically, how are cultural factors connected to performances of knowledge and power in practical settings? As far as we are aware, although connections between knowledge, ability and culture have been explored (Evans, 2004; Siedentop, 2002; Wright and Burrows, 2006), only one paper has dealt with this topic directly in relation to physical education (PE) group work. Brock et al. (2009) looked specifically at how perceived student status influences student interactions during Sport Education. They found that, firstly, status in PE lessons typically depended on students’ socio-economic status, attractiveness, athletic ability and popularity. Secondly, the investigation showed that students that were rated high on some or all of these aspects by peers tended to make more decisions during group work.

Given the prominence of group work in PE (Lafont, 2012; Ward and Lee, 2005), its relative taken-for-grantedness (Barker et al., 2015b), and the limited attention paid to power relations within group work, we believe the topic deserves more consideration. PE scholars have employed sophisticated theoretical tools in socio-cultural examinations of pedagogy (see for example, Burrows and McCormack, 2011; Wright et al., 2012) and it is fruitful to use some of these tools to examine group work. The purpose of this paper is consequently to explore power relations in group work and thus generate a better understanding of why group work unfolds as it does in PE practice. To do this, we use Michel Foucault’s notion of power as action-on-action to identify different types of power relations in group work.

Background

Power relations have seldom been an explicit focus of research dealing with group work in education (see Cowie et al., 1994; Reynolds, 2013, for examples of research that addresses issues of power). At the same time, practically all scholarship on this topic involves implicit assumptions about individuals’ capacities to shape the nature of actions occurring within a given context. Most investigations within PE tend towards one of two broad positions. One is that power exists independently of knowledge and that power parity between members is the ideal situation. Lafont (2012) provides an apt example of this perspective with her explanation of socio-cognitive conflict. From this perspective, group work involves people that have opposing ideas. As disagreement between individuals occurs, individuals are prompted to consider how their existing knowledge matches, or fails to match the world. Through reflection individuals eventually reach some kind of cognitive resolution. For this process to occur, power must be evenly distributed, or members ‘must be at the same level’ (Lafont, 2012: 137). If this condition is not satisfied – if one person has more authority to say how things should proceed – neither person will reflect on their own thinking and learning is unlikely to take place.

Various studies have been grounded on an assumption of equal standing between group members (e.g. Ensergueix and Lafont, 2010). Lafont et al. (2007) have looked at reciprocal turn taking, examining the content and frequency of discussions taking place between group members. Hennings et al.’s (2010) analysis was based on the idea that ‘two learners are provided equal opportunity to adjust the behaviour interactions taking place’ (p. 2). Importantly, they noted tensions, observing that during the course of the activity the ‘integrity of power allocated to the peer observer’ could be challenged (Hennings et al., 2010: 14).

In a recent investigation, Darnis and Lafont (2015) examined both symmetrical and asymmetrical dyads, suggesting that slightly dissymmetrical relationships were more effective in facilitating motor and tactical development for the ‘initially low-skilled students’ (p. 12). They reasoned that ‘initially good’ players had to explain their solutions and this led the initially low-skilled to restructure their cognitive schema. This view is part of the second broad position: that knowledge is attendant to power and a knowledge/power differential will enable the more knowledgeable individual to steer the action. This idea can be traced back to Vygotsky’s (1978) work, which has influenced thinking in the field of PE (d’Arripe-Longueville et al., 2002; Rovegno, 2006) and education generally (Roth and Lee, 2007). In practice, asymmetric power/knowledge relations can be created intentionally as part of peer tutoring-type situations where teachers assign roles to students (this typically happens in Sport Education programs, for example, Siedentop et al., 2011). Roles, though, do not necessarily have to be assigned and teachers may rely on asymmetric relations emerging spontaneously (see for example, Barker et al., 2015b). In our work cited at the outset of this paper (Barker et al., 2015a), we pointed to the fluidity of expertness and provided examples where knowledge of the task did not necessarily result in an increased capacity to make decisions. At the same time, our analysis did not account for why or how this occurred.

Assumptions about power and knowledge are necessarily connected to assumptions about learning. A Vygotskian approach relies on a kind of osmosis model where knowledge ‘seeps’ from high concentration to low. In contrast, the idea that power should be equally distributed is based on a competitive logic where different knowledge claims compete and the ‘winning claim’ will be integrated into the cognitive schemas of the group members. This contention is clearly articulated in Gréhaigne et al.’s (2001) ‘debate-of-ideas’ concept (see also, Darnis and Lafont, 2015).

Yet, if we think about how these assumptions might work in practice, it becomes evident that they have moral dimensions too. To assume that knowledge will flow from one person to another is often to assume that one will help or assist the other. ¹ In other words, a degree of ‘benevolence’ is built into the theory (cf. Larsson et al., 2011). If, on the other hand, we assume that the best ideas will emerge from a ‘fair fight’ between group members’ ideas, we assume that group members are rational and that when they see a more successful course of action, they will integrate it into their own schematic framework. Neither perspective, however, makes overt claims about how power and knowledge interact in group work situations.

In short, our proposal is that learning theories have built-in assumptions about how people interact and how power and knowledge work (cf. also Quennerstedt et al., 2014a). If we want to understand more about how group work unfolds and the role that cultural factors play in the process, it is useful to consider the assumptions made about the nature of power and culture in group work more explicitly.

Power in group work

Understanding power as action-on-action is aligned with the methodological aspects of Foucault’s later works (Foucault, 1980, 1982, 1982/1988, 1984/2000) and involves re-understanding power as something that always exists in actions and relations. The question is not if power exists, but rather how power functions in different situations (Foucault, 1982). Accordingly, power does not necessarily have to be a visible and intentional act of influence. If, for example, basketballs are used in PE practice, students will relate to the materiality of the ball, including the cultural norms of sport and of basketball. We can then talk about a power relation between the basketball on one hand and the student on the other. In this sense, cultural norms affect the course of events and can be explored as power relations. However, and this is important, in the event students can relate to these cultural norms by, for example, following, contradicting or ignoring, and it is through focusing on actions upon other actions that power relations can be understood (Foucault, 1982).

Power then is both about limitations and possibilities where some actions are made possible and others not. As Foucault argues: ‘far from preventing knowledge, power produces it’ (1980: 59). In this way power relations can be seen as a ‘guide’, or something that points towards different directions of action. Because of this, a Foucauldian perspective focuses on how power is practised in relations in different situations, hence power relations (Foucault, 1984/2000).

According to Foucault (1982), power relations have two central aspects. Firstly, power relations imply that a myriad of possible actions, reactions and results open up, and secondly, people involved in the power relation are acknowledged as acting subjects. In this way power is always related to action. Here, Foucault (1982) argues that power relations:

… operate on the field of possibilities in which the behaviour of active subjects is able to inscribe itself. It is a set of actions on possible actions; it incites, it induces, it seduces, it makes easier or more difficult; it releases or contrives, makes more probable or less; in the extreme, it constrains or forbids absolutely, but it is always a way of acting upon one or more acting subjects by virtue of their acting or being capable of action. A set of actions upon other actions. (p. 789)

Methods

Data for this study were collected in connection to a larger research project on learning in PE in Sweden (Barker et al., 2015a, 2015b; Quennerstedt et al., 2014b). In the project, eight lower and upper secondary schools in Sweden participated. Video recorded data using two cameras at each lesson were collected from four lessons at each school. Schools were distributed across four geographic locations with varying sizes and types of communities. Also pre- and post- interviews with teachers and so-called didactic moments interviews (cf. Quennerstedt et al., 2014b) with teachers and between three and five students at each school were conducted.

For this particular study we have in a first step scanned the whole data set of 32 lessons to identify lessons and events within lessons where students work in groups on a particular task without the direct presence of a teacher. From this initial scan, events from five lessons were selected for in-depth analysis, since they all included longer segments of what we define as group work within the lesson. The other lessons and parts of lessons included teacher-led or teacher-supervised activities. The final sample contained three events when students were working on a task related to creating a dance sequence and two events where students were involved in fitness activities.

In the second step, the five lessons or parts of lessons were reviewed more closely in order to select events of particular interest for our purpose – to explore power relations in group work in PE practice. This review revealed that the fitness activities were closely related to discourses of fitness and that students’ actions followed these actions habitually. The dance lessons, however, involved more open tasks and the group work also continued for a longer period of time. To explore a more varied practice in-depth with the possibility of identifying a wide variety of power relations, one of the dance lessons was chosen for in-depth analysis and presentation of results. In relation to the two other dance lessons, the selected lesson could, according to our review, be seen as an illustration of PE practice in terms of power relations in group work. In the lesson we followed three groups of students up-close, and these groups are used as illustrations in our results section.

In this paper, we also used the interviews in the project to collect additional data about students’ previous experiences and also their view of some of the events explored. Interviews were transcribed verbatim. The original language is Swedish and English translations were made with a view to staying as close as possible to the original meaning, rather than the literal translation. With regards to filming, lessons were filmed by one researcher with a handheld camera and one with a stationary video camera. The introduction of cameras is likely to have affected the participants, although neither the participants nor the teacher appeared to be particularly conscious of our presence. The stationary camera was not effective in capturing student dialogue so recording from the handheld camera is relied upon for the descriptions below. Finally, the researchers adhered to the ethical guidelines set out by Swedish law as well as the Swedish Research Council in all stages of the project. Details that could render the teacher or pupils identifiable have been omitted from the descriptions below.

With inspiration from Gore (1995, 2001) and Öhman (Öhman, 2010; Öhman and Quennerstedt, 2008) we look at the micro dynamics of power and how power is put into action. While Gore and Öhman used Foucault’s methodological tools or techniques of power (Gore, 1995, 2001) and power in terms of governance and socialisation (Öhman, 2010), we instead turn more explicitly to his suggestion regarding analysis of power relations. To identify what Foucault called power relations, we focus closely on the selected events in order to explore the functions of a particular action or activity in the event. These actions can be students acting, the task at hand, the purpose of the lesson, artefacts used or the cultural norms of the particular activity.

There is, of course, always a myriad of ongoing actions and activities in PE practice, so we identified particular acts or activities that in some way changed the course of the event, or acts or activities that in some way reinforced the continuation of the course of event. Foucault (1982) further argues that in an analysis of power relations, attention should be paid to five different aspects, as follows.

‘The system of differentiations which permits one to act upon the actions of others’ (p. 792). This aspect is handled in the study through paying attention to students’ differences in competence, knowledge and ability, and also to differences in culture and tradition. These differences are, according to Foucault, both the condition and the result of the functions of power relations.

After exploring these aspects in our selected events in-depth, both authors categorised and described how the power relations function in group work through following and exploring the course of the event using a common matrix building on the five aspects above. The two authors then compared their analyses and worked with what Krefting (1991) calls a peer examination strategy in order to strengthen the credibility of the analysis. The authors then agreed on a final categorisation and the different categories in the results section thus represent different ways of how power relations function in PE group work practice.

However, according to Foucault (1982), an analysis of power relations presents certain problems. He argues that there is always the risk of overemphasising the reproductive aspects of the institution. In our analysis we have handled Foucault’s warnings through our peer examination strategy, together making sure that we do not only focus on the cultural and institutional aspects of the power relations. Indeed, the actions of the students and the productive aspects of students’ actions constitute an important focus of the study and we have tried to make the issue an empirical question rather than a theoretical one. Power is accordingly not just a function of the institution or the individual.

Results

The lesson we use as an illustration is part of a sequence of lessons where students are supposed to create a dance to music in groups of four or five students. At the beginning of the lesson the teacher gives instructions about the task for the lesson.

What you should think about when you work in groups is that you really practice … so you don’t just stand around and talk about other things. Focus should be to work on the dance from the music you have chosen … and that you work from your own ability … your own ability in the dance … and that you contribute to the group … and that it is not only one of you that works hard and the others just tag along. Try to be active in the whole group and work on it the whole time. (Teacher)

Dance project – practice the dance – make the analysis of the music – group name.

Purpose/goal:

- to practice your dance to music

The instructions from the teacher and the design of the assignment accordingly construct a field of possible actions. After the initial instruction the students spread out in the gym into five groups, and the teacher walked around overseeing but not interfering in the students’ work. In this way, the teacher is enabling the students to create a dance but at the same time constraining their actions through his gaze in what Öhman (2010) calls governance through self-governance. The students chose the groups themselves; however, student interviews revealed that groups were changed after discussion with the teacher when the cooperation did not go as they expected. In the results below we followed three groups up close, one all-boys group, one all-girls group and one mixed group with two girls and three boys.

We have identified four ways that power relations function in PE group work: (1) negotiating the task demands; (2) relating to other cultures; (3) performing gender; and (4) acting on and with one another.

Negotiating task demands

Negotiation of the task is a key feature of power relations in PE group work. Each group negotiated their task in a different way representing potential institutionalised ways of how ‘things are done’ in PE practice. These were: (i) dance as contemporary dance; (ii) dance as a music video; and (iii) dance as a comic spectacle, all representing different cultural norms of dance.

In the all-girls group the identified objective they pursued was to follow the teacher’s instructions and create a dance that was co-ordinated, synchronised and looks like contemporary Jazz dance. The girls’ actions are in line with a show for an audience. This group does not have their own music and are working with music being played by the other groups in the gym. From the girls’ actions, the objectives appear to be as follows: to involve everyone; to include synchronised movements; and to incorporate some recognisable dance movements such as the Waltz. In addition, combining verbal and motoric interaction appears to be desirable and the girls constantly propose, copy, modify and negotiate before moving forward. By deliberating rather than one of the students taking charge, power relations can be understood as a process of consensus reaching.

In the mixed group, the objective can be understood as creating something that looks like a music video. Here, more attention is paid to the aesthetic nature of movement. The group has chosen the song, ‘When I grow up’ from the Pussycat Dolls. ² In the dance, they use similar gestures and moves to the dancers in the official music video (cf. www.youtube.com/watch?v=K0K46C82v9o) and perform with a similar kind of cool audacity. It is as if they are supposed to look like they are not trying too hard. The dance also includes moves with sexual connotations, which appear reasonable in a music video discourse, but less so in contemporary dance or in PE practice. In the situation, the students make sexually connoted moves part of how things are done in PE. In a post-observation interview, one of the girls in the group commented that the PE teacher found the dance really good because they used such advanced moves and asked them to perform in front of the whole school. In this way the discourse of music videos transforms the boundaries of the traditional PE discourse.

In the all-boys group, the objective pursued seemed to be to create a ‘comic dance’. At the beginning of the lesson, they start before the other groups and attract the attention and laughter of the other pupils. Tarzan-like chest beating and Pulp Fiction eye-v-ing in their creation adds to the impression of a non-serious spectacle. Throughout their dance they jump up and down and move to amuse rather than display aesthetic skill. Further, the structure of the dance that the boys have agreed upon ensures that all members have to perform ‘funny’ moves individually. In terms of power relations, ‘mock dancing’ seems to be institutionalised in PE in a way that ‘mock football’ or ‘mock track and field’ is not. Interestingly, one of the boys in the post-lesson interview, when looking at the video, stated: ‘It looks more ridiculous than it felt doing it’.

Relating to other cultures

A second significant factor explaining why things happen as they happen in the observed sequences is that group members brought practices that are institutionalised in other cultures into PE. In this particular lesson, the influence of music videos was relatively clear. From the use of the songs that helped to determine the tempo and style of the performances, to the lyrics that provided further inspiration for movements, to actual dance moves – traces of culture outside the walls of the gym can be found in the students’ performances, which shaped how the students interacted with one another. Two points are important here. Firstly, the students did not simply borrow but innovatively modified and elaborated the moves that they incorporated. For example, the mixed group that selected the track from the Pussycat Dolls began by touching their chests at a point in the song where it sounds like the artists are singing ‘I wanna have boobies’. Midway through the lesson, the boys decided that they will touch their crotches instead of their chests according to a more male gendered norm of moving. What is interesting in this case is that the artists actually sing, ‘I wanna have groupies’, thus the moves that are enacted appear to be the result of translation and misrecognition and thus construct a quite specific field of possible actions.

Secondly and related, the incorporation of these cultural artefacts into the PE lesson is not a direct translation. The students may be dancing but it is dance as it is done in PE and the habits and institutionalised practices of PE continue to influence how things are done. As Quennerstedt (2013) has shown, co-ordination and synchronicity are much more important in the PE discourse than aesthetic quality or motoric control. The all-boys group provide an aerobics-type performance and it is difficult to imagine the sequence passing as a ‘real’ dance in any dance culture outside of school PE. Still, there is no intent to perform in real dance cultures and their movements ‘work’ in PE – in this sense, the performances are situated. Further, in this group as in the mixed group, there are signs of embarrassment and humour. On one hand, these signs reveal tensions in the logics of practice between the cultures but on the other, humour and even embarrassment become part of the logic of PE dance practice. It became acceptable for students to make light of their performances by introducing moves that will amuse others. In these respects, the institutionalised habits of different cultures open up and close down possibilities for the students as they interact with one another.

Performing gender

Another manner in which power relations function in group work in PE is in terms of gender. Two different ways in which gender works in PE practices are (i) gendered abilities and (ii) masculine and feminine ways of moving.

In the explored lessons, being able as a boy means something different to being able as a girl and these different meanings affect the course of events. In the mixed group, for example, the dance is performed in two parts. The first part is in line with the music video, while the second part includes more traditional couples dancing with Salsa-like moves. In the first part, the two girls perform the dance while the three boys try to follow as best they can. Gestures and moves indicate an objective in line with the Pussycat Dolls’ music video. In the context of this orthodox feminine dance, the three boys are relatively incompetent and the two girls are relatively competent.

In terms of power relations, the girls’ actions set certain standards. In PE, standards are often set by the teacher and/or in relation to established practices such as team sports (also, of course, gendered). In choreographing a dance though, the students – in this case the girls – have opportunities to define competence. They accordingly create their own space of privilege so seldom awarded girls in PE (cf. Larsson et al., 2011). In the sequence, the two girls hardly acknowledge the boys with eye contact or orientation and could actually be doing the dance without the boys. They focus on themselves in the mirror and have an ongoing dialogue with one another. They only turn their attention to the boys when the boys ask for help and, even then, instructions are curt. The boys are then included in the dance but excluded from its creation. That the girls do not acknowledge the boys and that the boys try to follow is an apt example of a power relation where difference in competence makes students act on the actions of others. The boys could refuse, ignore or resist but they do not. They instead act in line with what Larsson et al. (2011) have discussed as the benevolence in PE-practice.

In the second part of the dance, however, one of the boys introduces Salsa-like moves. In this situation, the members pair up in boy/girl pairs, in line with heterosexual norms of dancing. Here, the boy that has introduced the moves is not in a pair and stands alone as the able one. He takes charge of the process, modelling for and instructing the others. At one point he says: ‘stop for a while’ when he thinks they are not doing as expected, and instructs; ‘the foot behind, not in front’. The boy also shows one of the pairs what to do by manipulating, giving instructions, gesturing with his foot, taking the girl’s hand and twirling her around in a rather classic act of male dominance reinforced in ballroom dancing (Hanna, 1988).

Related to gender in terms of differentiation of abilities are specific masculine and feminine ways to move. In the all-boys group, for instance, masculine coded moves according to traditional cultural norms are performed. They stand in line military-style, and come forward to do individual moves like a dance battle (Lamotte, 2014). The boys also perform ‘Cossack dancing with arms crossed’, ‘Tarzan like chest beating’ and ‘pistol shooting’ moves, all movements with masculine connotations. The performance as a whole is a mix of an aerobic type of running and jumping around together.

As an institutionalised way of moving and negotiating the task at hand, the actions resist a potentially feminising discourse of dance, and it would look strange within the group if one of them tried to talk about precision or wanted to do a pirouette (unless pirouette was done in jest). Gard (2003) has referred to ‘gendered investment’ in terms of who and what students become when publicly performing a dance in front of their peers. Being a boy here is delineated by a gendered discourse in PE (Larsson et al., 2011), and by doing a comic spectacle, the position of a male heterosexual is not necessarily threatened (Larsson et al., 2011). In contrast, the all-girls group fulfilled all requirements of the task, something research indicates is often attributed to girls in school (Myhill, 2002). They are on-task, something that can be seen as a gendered way of how dance is done in PE that is institutionalised in its habits and customs (Quennerstedt, 2013).

Acting on and with one another

In terms of actions on actions, differentiation of competence can be identified in the types of actions and the way that the actions occur across the groups. In the all-boys group, the means in which power relations come into being involve no instances of corrective physical contact, relatively few instances of spoken or gestural directives but many instances of observation and movement imitation. One of the boys, in particular, makes his movements slightly ahead of the others and the others follow him. When the members of the group are talking, the three address their comments to him rather than to one another. It is difficult to grasp the basis for the differentiation taking place. The boy is taller and more mesomorphic than the others, a body that Tinning (2010) suggests has been most suited within a traditional PE discourse. He is wearing a sleeveless vest, and he seems to have what Brock and colleagues (2009) would call status. At the same time, the post-lesson comments of one of the group members are rather oblique. When asked if the group has a leader, the boy says ‘well it’s mainly one guy that does what he wants’. When asked if this is ok, he replied, ‘I don’t know, I’ve never done dance’.

In the mixed group, the members have democratically divided the responsibility for choreography in the previous lesson. This division defines the field of possible actions. One of the girls appears to know the moves slightly better and guides the others. The girls also start and stop the music several times to practise certain parts. The boys say very little and benevolently remain close to the girls and continue to orient their actions to the girls’ actions. In the second part, the boys have determined that the two girls will pair up and dance with two of the boys. Power is put into action as the third boy that has experience with modern dancing outside of school talks more, touches the other pairs to move them into position, demonstrates movements, and controls the music. The other four follow his instructions and ask him for clarification on certain points. The difference between his way of steering the group’s action and the girls’ way of steering the action is quite marked and, perhaps unsurprisingly, leads to different forms of responses. The members cannot simply follow and imitate, for example when the boy has his hand on their shoulders or is twirling them around.

In the third group the actions upon actions are different again. The overall sequence appears as a series of sub-sequences punctuated by moments of discussion. In each of these cases, discussion helps the girls to decide on not so much how but who will move the dance performance forward. Midway through the sequence, for example, the girls stop dancing, unsure of their next dance move. First one and then another of the girls turns to a third girl, essentially providing her with an opportunity to structure the possible field of action for the others. The third girl spontaneously demonstrates a long stride in imitation of the girl in front of her. The others follow her and laugh and the move is incorporated into the performance. The sequence contains typical teaching actions, as means in which power is put into action, where group members demonstrate, explain and physically manipulate in an effort to get the others to move in particular ways. Smaller actions in terms of listening, watching, copying, standing in the ‘right’ place and being quiet and in-active also help to move the performance forward.

Discussion

The purpose of this paper was to explore power relations in group work and understand why action unfolds in the ways that it does. Group work literature in PE has tended to point to the significance of task knowledge (d’Arripe-Longueville et al., 2002) in structuring interactive sequences. Our analysis suggests that there are a number of factors that deserve consideration.

To begin with, power relations do not merely exist between group members. Instead Foucauldian theory (1982, 1982/1988) encourages us to consider how artefacts such as mirrors, mats and other kinds of material objects are institutionalised in the customs of PE, and that they are part of the means in which power relations are brought into being in PE contexts. Prior or historical actions, such as teachers’ instructions or students’ actions from earlier lessons, must also be regarded in terms of power relations. These continue to shape and structure group members’ actions, sometimes in prominent ways. As the results above highlighted, cultural ideas and practices provide resources such as movements and ways of acting that are appropriated and modified. At the same time, they also offer general guidelines, defining appropriate ways of acting. In this way power relations are elaborated and transformed in PE practice. Gendered ways of doing things is a prime example in terms of gendered abilities, ways of moving and benevolent attitudes. In addition, the possibilities to define what ability can be also opens up the possibility for, in this case, girls to act in different ways in PE, a subject that rarely offers the privileged space. From a Foucauldian perspective, individual actions are also important. Instead of being seen as representative of sets of separate, internal intentions or objectives, they are seen as a constitutive part of ongoing, reflexive practices. From here, actions may or may not be coupled with intentions but influence the ways things happen nonetheless.

Further and entirely related, power is seen as a set of relations. Unlike in some earlier work where power was seen as an unequally distributed resource that can either be leveraged for the benefit of less knowledgeable individuals (see Rovegno, 2006, for a discussion of situated perspectives) or evened out in order for logic to prevail (Gréhaigne et al., 2001), power is enacted in practice through relations with the actions of others in relation to the task at hand. This offers an interesting counter-position to claims that popular, athletic or otherwise able students might abuse their power to make more decisions and gain advantage during group work (Brock et al., 2009).

This is also an important point for pedagogues. If power is acted through relations, and relations not just between group members but between members and instructions, traditions, equipment and so forth, then strategies for ensuring fair and inclusive participation during group work must involve more than attempts to motivate or empower group members that are standing back or not participating. Instead of seeing other group members as ‘problematically powerful’, we should focus on the entire situation and ask, ‘why is this activity, at this time and place, with these students, leading to these kinds of practices?’ Why did the girls and boys in the mixed group, for example, act so differently in the first half of the sequence to the second? The music stayed the same and the task was still to choreograph a performance yet the actions of the students changed quite dramatically. The value of a Foucauldian perspective lies in its capacity to open up a range of factors for attention. Teachers might consider changing equipment, location, music, activity, group size or constellation, instructions, performance format or a combination of these aspects, for example, in order to disrupt or change power relations. We realise that recognising increased complexity does not always make action easier in practical settings. We would propose, however, that equitable practices can only be undertaken if this complexity regarding power relations is acknowledged and managed.

A final point relating to power relations in group work concerns learning. The examples above show how the students in the three groups all appear to be learning something about choreographing and performing dance, albeit with slight differences. By exploring their actions in ongoing practice in relation to the task at hand we can see how institutionalised habits and customs create the space for what can be learned. Besides having fun, being active, and working as a team, the boys learn that comic dance is acceptable through the approval of their peers, the mixed group learn that music video moves are acceptable through praise from the teacher, and the girls learn from each other that dialogue and consensus reaching is an appropriate way to choreograph a dance. In some respects one could argue that the lesson is successful: student learning fits with the teacher’s general aim and the goals the teacher presented on the white board. One could also argue that power relations that extend beyond the classroom support the teacher’s objective and to greater or lesser degrees facilitate anticipated sequences of action. In other respects, the students’ actions are somewhat habitual and routine. There are few examples of creative or novel actions and few instances where it seems like the students are really troubled or are unsure how to go on. For us, this suggests that the students are benevolently reinforcing habits and identities rather than creating the possibilities of new ones.

Concluding remarks

We started this paper by citing a critique involving the claim that ‘student interactions, and how they take up the expert or novice role, are never neutral or dislocated from the broader cultural context’. As a reply this paper has used Foucauldian theory to illustrate how cultural context permeates group work in PE lessons and affects (1) how students interact and (2) what students learn, in decisive ways. We would like to finish with a recapitulation and brief elaboration of some thoughts on the relevance of our analysis to practice.

Firstly, we would suggest that teachers do not have to simply accept that popular or sporty kids are going to dominate group work situations in the absence of a teacher. There are more aspects of power at play than just the perceived status of individuals or their skill levels relative to one another. With sensitivity to power relations operating within and around lessons, teachers may find useful ways to encourage equal participation that go beyond putting the ‘right’ pupils together. For example, teachers often design tasks so that all pupils do something comparable. Teachers could, however, provide quite different tasks and/or equipment for each group, perhaps based on the pupils in the group. Further, teachers could introduce new challenges or conditions as group work proceeds. This strategy would result in dynamic tasks that demand different kinds of interactions between the same group members.

Secondly, we might recognise that even if the domination of some students by others is not an inevitable part of group work, teachers are still not going to have the authority to control everything that happens during group work – there will always be an element of unpredictability concerning the pupils’ behaviour. This is only a problem for teachers if they expect to be able to control interactions in the first place. Accepting that learning can be affected by a number of external factors (the popularity of a new song, for example) may reduce teachers’ attempts to steer action and be beneficial to both the teacher (with potentially less stress) and pupils (with the possibilities of making mistakes and learning from them) in the long term.

Thirdly and related, teachers might try to introduce tasks where expertise is ambiguous, or at least, less clear than in traditional PE activities. With open tasks where objectives are not predetermined, the need for communication and deliberation increases and the ways that people interact are less ‘given’, for want of a better word. In some ways, dance provides a relatively good example of an activity in which the objectives are not predetermined (in comparison with say, a game of volleyball). At the same time, dance is still steeped in culture. There is still a challenge, we believe, to devise ways that PE can be done differently.

Finally, we would add that although this paper has focused on group work and we have looked specifically at instances of students working with dance, we believe that we can learn something about PE more generally from this analysis. Questions relating to power, learning and why things happen as they happen are of interest to pedagogues and researchers alike and warrant further attention.