School bullying from a sociocultural perspective

School bullying is an important concern. Whilst there is growing knowledge about the nature, extent and effects of school bullying, areas of complexity in research findings remain. In this paper we develop our thinking on school bullying using a sociocultural theoretical framework. We review existing literature around three main themes: 1) The conceptualisation and interpretation of bullying; 2) The relational aspects of bullying 3) Bullying as part of someone's life trajectory. For each theme, empirical findings are discussed to highlight key issues, and arguments presented from relevant sociocultural theories to provide insight in each case. During the paper, we show how varying strands of research into bullying can be integrated, and how areas of complexity can be explained. Adopting a sociocultural view of school bullying presents implications for both research and practice. Bullying is contextual, and attention should be given to the situated relationships and multiple settings surrounding the behaviour.


Introduction
This paper examines research on school bullying through the lens of sociocultural theorising. Whilst there is a proliferation of empirical research on bullying that has helped us to understand its nature, characteristics, and experiences of those involved (Smith, 2015), we argue that there has been less emphasis in the literature on providing a coherent theoretical underpinning to help integrate and explain complex research findings. Sociocultural approaches have much to offer in this respect. We will argue that an integrative approach, using sociocultural theorising, provides a useful foundation for exploring bullying in context, and challenges some of the taken-for-granted assumptions that can accompany studies with other approaches.
The roots of sociocultural approaches are found in Vygotsky's (1978) work and whilst we recognise this is not a unified theory without contentions, there is a shared understanding of development as shaped by the contexts in which individuals are based, and the social and interactional relations that exist between them. More recently it has been argued that sociocultural theory not only provides a mechanism for understanding cognitive development in interaction, but also social and emotional learning through shared cultural school spaces with peers and teachers (Morcom, 2015). Understanding or meaning-making is mediated through socialisation, interactions and guidance with others (Vygotsky, 1978). Mediation was a central tenet of Vygotsky's work and speaks to a process whereby the individual and the social mutually shape each other (Daniels, 2015). One aspect of mediation is the use of cultural artifacts, such as language or a physical object. Such artifacts were said to "provide the link or bridge between the concrete actions carried out by individuals and groups, on the one hand, and cultural, institutional, and historical settings, on the other." (Wertsch, Rio, & Alvarez, 1995, p. 21). Another important facet of mediation developed by proponents of sociocultural theory is the role of psychological tools which focus on social interactions. de Abreu and Elbers (2005) also point to the ways that social and institutional structures mediate people's psychological understanding of cultural tools through practice.
Sociocultural theories allow bullying to be examined through studying peoples' engagement in various cultural contexts whilst also treating those contexts as dynamic and contested spaces (de Abreu, 2000). Cultural contexts (such as family, schools, social groups) each have their own histories, social norms and conventions which have been established over time and are reproduced through participation of members (Wenger, 1998). Through individuals interacting with and becoming integrated into communities, clues about cultural rules and behaviours are transmitted. Details about acceptable norms and practices are conveyed through various means, such as less experienced members observing established members; forms of apprenticeship where individuals participate in culturally-organised activities and become embedded members; and established members 'scaffolding' learning through guided participation (Lave & Wenger, 1991;Rogoff, 2008).
The potential of sociocultural approaches has already been touched on in some bullying literature through a focus on ecological frameworks (e.g. Espelage, 2014;Monks et al., 2009). Bronfenbrenner's Ecological System (1979) shares common ground with sociocultural theorising, and has received recent interest in the school bullying literature because it accounts for the multiple social and cultural systems in which an individual is situated, and which mutually influence their development. The dialectical relationship between the face-to-face contact of the mircosystem and the broader sociocultural milieu represented by the macrosystem, present a promising site of empirical and theoretical analysis for school bullying. The model emphasises the bi-directional connections between individuals and their environment, meaning that the child both affects and is affected by the settings in which they are situated. However, the existing reviews of bullying literature using the ecological model (see Hong & Espelage, 2012;Huang, Hong, & Espelage, 2013;Patton, Hong, Williams, & Allen-Meares, 2013) tend to focus on the multiple levels that risk and protective factors might operate on, as they contribute to bullying involvement. These papers highlight the value of studying bullying in relation to the social contextual environments that young people are part of, and demonstrate the potential to inform bullying intervention. This is an important theoretical shift which we are keen to build on, but we feel that sociocultural theorising has further potential in this arena. In particular, these papers pay less attention to meaning making, and how thoughts, actions and interpretations are socially mediated. For this reason, the incorporation of sociocultural frameworks which look at how meaning is developed, negotiated, and resolved would be an additional resource to the study of school bullying.
To illustrate our premise, we take three main themes within the research literature on school bullying: (i) the conceptualisation and interpretation of bullying, (ii) the relational aspects of bullying, and (iii) bullying as part of someone's life trajectory. In each case we describe findings from research in the area and illustrate how sociocultural perspectives can assist in understanding these reported findings. To conclude, we will propose some implications for addressing bullying in schools which emerge from this analysis.

The conceptualisation and interpretation of bullying
Debates about how best to define bullying have been a perennial problem within the field (Smith & Monks, 2008). Whilst the widely agreed definition of bullying in the academic literature is that it involves repeated aggressive behaviour directed towards someone with less power, with the intention of causing them harm (Rigby & Smith, 2011), research has highlighted that this definition is not shared, or does not necessarily suit, the complex setting of the school community. For example, Naylor, Cowie, Cossin, de Bettencourt, and Lemme (2006) used an open-ended questionnaire to explore how pupils and teachers defined bullying by asking them to respond to the question 'Say what you think bullying is'. The authors applied a categorical analysis to this data which was the basis for statistical analysis. They report quite distinctive meanings associated with bullying when comparing pupils and teachers' conceptualisations. Quite surprisingly, only 9% of pupils included repetition in their definition of bullying compared to 17.8% of the teachers. Only 3.9% of pupils included the bully's intention to cause harm compared to 24.9% of teachers. In comparison to teachers, pupils also tended to limit their definitions of bullying to verbal or physical forms -neglecting indirect behaviours such as social exclusion. Maunder, Harrop, and Tattersall (2010) compared perceptions of bullying behaviours between secondary school pupils, teachers and support staff using a series of rating scales in response to short scenarios describing a range of different bullying and ambiguous behaviours. Teachers and support staff rated behaviours more seriously and defined a broader range of behaviours as bullying than pupils. In contrast, Menesini, Fonzi, and Smith (2002) found that teachers were less likely to label written descriptions of indirect and verbal behaviours as bullying when compared to pupil ratings of stick-figure cartoons depicting the same behaviours. Using a similar methodology, pupils of different ages have also been found to interpret the meaning of bullying in varied ways (Smith, Cowie, Olfasson & Liefooghe, 2002), with younger pupils tending to adopt a broader definition than older pupils (Smith & Levan, 1995). Furthermore, at a wider contextual level, international research has revealed that the English term 'bullying' does not have an exact translation in other languages, and that different meanings are attributed to terms associated with it (Smith, Cowie, Olafsson, & Liefooghe, 2002;Smith & Monks, 2008).
It is not unusual, nor surprising, that academic research and school policy aim to reach a unifying definition of bullying in school. However, the studies discussed above are valuable for showing that definitions are not straightforward and that attention needs to be given to people's interpretations of bullying when planning interventions. It is important to note that some of the work described above adopt a positivistic epistemological approach, either by way of data collection or analysis, involving pre-defined variables (such as the academic definition of bullying, and different types of bullying behaviours) and comparison of differences between pre-defined groups (such as gender, age, pupil/ teacher etc). These methods have enabled us to see that defining bullying is not clear-cut, but do not necessarily allow us to understand what meanings and experiences are attached to definitions and how these ideas intersect with each other and/or the institutional and historical settings. It is here that we suggest sociocultural approaches could allow us to get to the heart of these complexities.
Sociocultural perspectives would suggest that our understandings and interpretations of an experience are sometimes at odds with how we might define something. A key aspect of sociocultural approaches is the focus on the situated nature of behaviour, and how it is socially constituted. Cultural practices, when framed as the everyday actions or activities that people do, are value-laden, linked to representations of morality and deeply tied to our sense of belonging and identity (Miller & Goodenow, 1995). Taking this viewpoint, what is classed as bullying is based on an interpretation of the situation in which the activity occurs rather than being an objective phenomenon that translates equally across settings. In support of this, there is a body of work on perceptions of bullying which has sought to look at context. Whilst this research has not explicitly adopted a sociocultural position, the reported findings are aligned with the principles of this approach. For example, in qualitative interviews about managing bullying in schools, teachers have indicated that what is identified as bullying depends on the situation, and that there are organisational factors in the school such as relationships between colleagues, relationships with pupils, and institutional procedures which shape how incidents are dealt with . There are also contextual factors about incidents that determine whether they are perceived as bullying, and whether or not teachers will intervene. Quantitative survey research utilising bullying scenarios with a series of scale responses to measure teachers' views of the situations found that teachers rated physical and verbal forms of bullying more seriously than socially orientated bullying, and that their judgements about the seriousness of an incident predicted their likelihood to intervene (Ellis & Shute, 2007). This reflected findings by Yoon (2004) where teachers' decisions about intervention in hypothetical bullying scenarios were influenced by their ratings of seriousness of the incident, how empathic they felt towards the child being bullied, and their perceived self-efficacy to deal with the behaviour (measured via a selfefficacy scale). Teacher responses to hypothetical aggressive scenarios with manipulated content also varied depending on information provided about the popularity of the child and their 'typical' behaviour, and the extent to which teachers identified with the class (Nesdale & Pickering, 2006). These factors suggest that teachers make subtle contextual judgements when faced with an incident, and deciding how to respond. Nesdale and Pickering (2006) use Social Schema Theory (Fiske & Taylor, 1991) to underpin their study, arguing that teachers have schemas about good and bad children which affect their treatment towards pupils. They also draw on Social Identity Theory (Tajfel & Turner, 1979) to suggest that teachers' sense of connection to their class group can lead to bias, favouritism and more negative reactions towards behaviour displayed by members that are seen to go against established group norms. Whilst these theories are useful they downplay or neglect the structural, institutional and historical aspects that would be evidence in say, the 'cultural-historical activity theory' (CHAT) branch of sociocultural theorising. CHAT fundamentally proposes that persons are shaped, but also shape, their social contexts. Moreover, proponents of CHAT would highlight the importance of linking culture, cognition and learning with the historical conditions in which they are taking place (Roth & Lee, 2007).
Research has also shown that the same bullying incident may be interpreted and responded to differently based on the amount of harm caused, and the setting where it occurred. Gentry and Pickel (2014) conducted an experimental study whereby adults acting as a disciplinary committee made judgments about bullying cases. Information presented about the incidents was manipulated in order to examine how it would affect responses. Results showed that the same scenario was judged more seriously when it occurred at school compared to university. The same incident was also judged more harshly when the individual being bullied experienced more harm from it (Gentry & Pickel, 2014). The authors suggested that people have 'bullying schemas' which they use to interpret incidents based on their own conceptualizations of bullying. Alternatively, rather than people having a fixed definition of bullying which is used as a template to evaluate incidents, we suggest that they make meaning from each incident based on their interpretation of the particular circumstances. The turn towards meaning-making has been a central feature of sociocultural theory (Bruner, 1990;Vygotsky, 1978;Wertsch, 1991) wherein individuals are actively producing meanings rather than focusing on specific characteristics of measurement at any given time (Valsiner, 2008). This is supported by the aforementioned research with teachers which revealed their 'on the spot' assessments of bullying situations based on situational factors including the type of behaviour, perceived seriousness, and who is involved (e.g. Ellis & Shute, 2007;Nesdale & Pickering, 2006;Yoon, 2004).
In addition to whether or not a situation is labelled as bullying and how it is interpreted, research has also shown that there are inconsistencies across groups in the terminology applied. Some of these studies are quite atheoretical in their positioning, and despite adopting methodologies that would be sensitive to the central tenets of sociocultural theory (by enabling the study of context, meanings and change), they do not explicitly state what theoretical framework they use to develop their research or interpret their findings. For example, a paper by Lee, Smith, and Monks (2012) point to the problems of applying the English word 'bullying' to other language contexts. In Korean, there are several words to indicate 'bullying-like' phenomena such as 'wang-ta' (severe social exclusion), 'gipdan-ttadolim (collective isolation) or hakkyo-pokryuk (school violence). The authors show how popularity and usage of terms is linked to generational shifts, cultural trends, and different age groups. The authors suggest that terminology follows trends, being more popular at some times than others. They also argue that new terms are created by groups to distinguish behaviours in their culture from those in other groups. For instance, they noted how 10-18 year old Korean pupils had adapted an existing term 'jjin-ta' and put their own meaning on it to reflect the bullying behaviours in their peer group. Linked to this, a review paper examining definitions of bullying by Smith and Monks (2008) discussed evidence showing that i) the academic definition of bullying is not necessarily shared by others in the population ii) there are age-related trends in how bullying is understood iii) terms used to describe bullying differ across cultures, and iv) the type of bullying behaviour displayed may vary between cultural groups. They therefore argue that the meaning of bullying is socially constructed, and has "cultural and temporal variability" (p110). In effect, their work has significant resonances with sociocultural theorising, particularly around language as a culturally and historically mediated artifact/tool. Teachers, when faced with a potential bullying situation, will interpret it through a dynamic process which could include previous experiences of similar situations, their knowledge of and relationship with the children involved, what they think bullying is, organisational climate in the school and many other factors Mishna, Scarcello, Pepler, & Wiener, 2005;Tucker & Maunder, 2015). Similarly, children also have complex ways of both defining and responding to the bullying based on multiple considerations including perceptions about the seriousness of the situation, their relationship with those involved, felt emotions, and social hierarchy in the peer group (Forsberg, Thornberg, & Samuelsson, 2014). Familiarity with the circumstances, and how much knowledge they have of the history leading up to the situation, also shapes interpretations (Forsberg et al., 2014). These findings all illustrate the role of context, and how individuals make sense of a given situation. Using a sociocultural lens, we see these research findings as examples of meaning making. Meaning making is complex and embedded within the communities individuals are part of. The process of making meaning occurs between and within individualsexplaining why definitions vary in different contexts.
The role of context, and indeed history, is a fundamental part of looking at a phenomena using sociocultural theorising (de Abreu, 2000). Cultural contexts are evolving, and are reproduced and developed by its members over time. Therefore how bullying is conceptualised by, and manifested within various groups, will shift. Taking a historical view, we can see how conceptualisations of bullying, and discourses around it, have evolved over time. For example, early work on bullying tended to focus on physical and verbal bullying (see Olweus, 1978) and it was only in the 1990s where indirect and relational forms of bullying were fully recognised . The growth of technology has since opened up new typologies of bullying using cyberspace (Law, Shapka, Hymel, Olson, & Waterhouse, 2012;Williams & Guerra, 2007), and a new label of 'cyberbullying' attached to particular manifestations of bullying behaviours occurring online. This shows how changes at the macro level of the ecological system (Bronfenbrenner, 1979) -such as the introduction of new cultural tools in the form of technological shifts in society -could impact on an individual's experience at school and with their peers through the phenomenon of cyberbullying.
Early conceptions of cyberbullying focussed primarily on use of email and mobile phones but as technology has continued to develop, new forms of cyberbullying are being identified based on the media used or type of action (Slonje, Smith, & Frisén, 2013). Increase in social media use, along with instant messaging and chat rooms have opened up new mediums for bullying. Similarly, discourses have developed new meanings to refer to specific forms of bullying, such as the term 'punking' to refer to male-male verbal and physical violence and humiliation (Phillips, 2007), 'frape' which relates to somebody's online profile being amended without their permission (Collins English Dictionary, 2015), trolling (bombarding with insults and threats), flaming (argumentative communication online), and slamming (groups joining in with online harassment) (Chisholm & Day, 2013).
As well as providing new forms of bullying, technology has paved the way for greater blurring of boundaries for the engagement of bullying across settings such as school, home and cyberspaces (Whittaker & Kowalski, 2015). At one time the home might have been a protective space from which to escape school bullying, but cyberspace transcends these geographical demarcations. These boundary crossings (Akkerman & Bakker, 2011) challenge existing definitions and require a re-conceptualisation of what is meant by bullying. For example, notions of repetition and power imbalance, which are widely accepted as core aspects of a definition of traditional bullying are more challenging to apply to cyberbullying (Slonje et al., 2013). A single incident (such as a picture being uploaded online) can snowball due to others sharing it. It is also more difficult to ascertain origins of power by the perpetrator in cyberbullying. In traditional bullying, physical strength, psychological strength or power in numbers are important determinantsbut these are not as easily translated online (Slonje et al., 2013).
The nature of bullying has therefore not remained static. As society has changed and developed, new forms of bullying have been identified and appropriate labels constructed to describe them. These new trends mean we need different ways of thinking about bullying in light of the changing environment and how this impacts on the individual. The process of meaning making across contexts to which we belong is ongoing and multifaceted. History is important, but much research only captures a moment in time, even when it is methodologically designed to capture complex phenomena. In addition, a challenge faced by methodological approaches akin to a positivistic epistemology is that they require pre-defined variables that may hide or mask some of the complexities that a sociocultural approach can unveil.

The relational aspects of bullying
There has been recent interest in psychological literature exploring bullying as a group-based phenomenon where bullying behaviours can be explained by peer-group norms around aggressive behaviour (e.g. Jones, Manstead, & Livingstone, 2009;Nesdale, Durkin, Maass, Kiesner, & Griffiths, 2008). These studies typically adopt positivistic approaches by attempting to pre-establish what these norms might be and then testing their relational value experimentally (ibid). Other studies have used qualitative interviews and survey-based peer nomination inventories to show that bullying can occur within friendship groups (Mishna, Wiener, & Pepler, 2008;Wei & Jonson-Reid, 2011) -challenging traditional notions that bully-victim relationships are distinct from friendships. Rather than bullies and victims associating with separate peer groups, with contact between them mainly involving the bullying behaviour itself (Wei & Jonson-Reid, 2011), research suggests that bullying is part of a continuum of interpersonal relationships that exist within the peer group where individuals may assume different roles at different times.
In support of this, research studying children's behaviour during incidents illustrate that bullying is a situated, collective behaviour involving the peer group as a whole. Bullying involves more individuals than the child being bullied and the bully, particularly because the majority of incidents occur in social situations in front of witnesses. Seminal work by Salmivalli, Lagerspetz, Björkqvist, Österman, and Kaukiainen (1996) identified four 'participant roles' which are occupied by bystanders in the bullying process. 'Assistants' join the bullies and engage in bullying behaviour (such as 'ganging up'); 'Reinforcers' encourage the bully by watching, laughing, and providing positive feedback; 'Outsiders' move away from situations and withdraw, whereas 'Defenders' actively intervene to try and support the child being bullied. Salmivalli et al.'s (1996) early work on participant roles positioned bullying as collective social action. The focus on the wider peer group shifted attention away from the bully-victim dyad and recognised that bullying was far more complex.
The concept of participant roles has been previously interpreted using a social learning perspective, and also desensitization, whereby exposure to negative behaviours over time result in a disinhibiting effect and potential for modelling the behaviours in the peer group (O'Connell, Pepler, & Craig, 1999). A sociocultural position would suggest greater fluidity within and between these roles dependent on the situation. For example, an individual might flow between 'reinforcer' and 'defender' depending on context. The importance of context was acknowledged by Salmivalli et al. (1996) whereby the particular role children assumed would be determined by a range of factors. However, the common approach taken to study participant roles is for children to complete a self-report and/or peer nomination inventory with descriptions of different behaviours in bullying situations.
Individuals are assigned to a particular role based on their score for the different behavioural descriptions in comparison to others, which assumes that the children involved have a relatively stable role in the peer group. Research into the participant roles that children adopt has shown 'moderate' stability in roles over time, indicating that although there is some consistency, there is also movement between roles (Salmivalli, Lappalainen, & Lagerspetz, 1998). In addition, it has been shown that children can assume multiple roles at the same time, illustrated through them obtaining similar or equal scores on the different participant role categories measured by self-report (Salmivalli et al., 1998). This means there can be flexibility in how they are assigned to particular roles (Sutton & Smith, 1999).
The aforementioned studies, and much of subsequent participant role research which adopt a positivistic epistemology, often address this issue by developing tools of assessment, scoring procedures and statistical approaches to ascertain how best to ascertain individual roles. This is indeed necessary in order to facilitate robust measurement of variables and between-group comparisons. However, a sociocultural position would suggest that measuring participant roles in this way could risk losing the sense of dialogicality between peers in bullying situations. The point we wish to emphasise here is that the underlining epistemology of a study leads the research in a particular direction and shapes where attention is given. There can therefore be a benefit for introducing an alternative lens in order to highlight different aspects and open up new lines of enquiry. The findings suggest that children's behaviour in bullying incidents may fluctuate depending on the situation, and sociocultural approaches recognise there is always a temporal dimension to our actions (Cole, 1995;Hendry & Kloep, 2002;O'Toole & de Abreu, 2005) From this perspective, social relationships are transactional and dialogical, and our interest is on the shifts that might occur in the relationships between people rather than remaining static. Forsberg et al. (2014) examined motives behind bystander behaviour and why children might choose to intervene or not in bullying incidents that they witness. Whilst they did not explicitly position their work in terms of sociocultural theory, their focus, approach and findings are aligned to this perspective. They adopted a qualitative semistructured interview approach using grounded theory to enable them to understand how young people make meaning in bullying situations. A range of considerations were reported, including their relationship to the child being bullied, who was involved, perceived responsibility, feelings of empathy or distressing emotions, and how serious the incident was judged to be. Children were making complex decisions about if and how to respond based on an array of personal and situational factors. The participant role that group members adopt therefore is part of the interaction between their individual characteristics (such as emotions, attitudes and motivations) and environmental factors (Salmivalli, 2009), resonating strongly with the Ecological System approach (Bronfenbrenner, 1979).
In a peer group community, there are common values and understandings, and shared cultural practices (Rogoff, 2003;Wenger, 1998). Relationships within communities are complex and participants will each have different roles and responsibilities (Rogoff, 2003). Relations can be supportive, or conflictual, and may have agreed processes for resolving disagreements. There is a sense of organization and shared routines, and also scope for adaption and negotiation. Viewing the peer group in this manner can therefore help to explain the existence of participant roles in bullying situations, and bullying between friends, because they represent existing practices operating within the group as a whole (such as norms around non-intervention; or norms around collective negative action towards others). Furthermore, the practices individuals participate in and do not engage in form their sense of identity about who they are and who they are not (Wenger, 1998). Engaging in the practice of bullying and playing a particular role in the peer group serves to establish identity through membership of a particular group. Children are powerfully influenced by their peers and may engage in bullying as a way of establishing allegiance to their group, or because of peer pressure from fellow group members who are more powerful (Rigby, 2004). For example, experiments involving children being assigned as members of a group where particular group norms (such as out-group like or dislike, being kind or unkind to others) are manipulated have shown how altering in-group norms affects children's bullying intentions and responses (e.g. Jones et al., 2009;Nesdale et al., 2008). Such findings have typically been explained using mainstream social psychological theories such as Social Identity Theory (Tajfel & Turner, 1979) and Social Identity Development Theory (Nesdale, 2004) which examine membership, identification, and conformity within groups. Exploring results in this way sheds light on the powerful influence of groups, and encourages reflection on the role of the peers in bullying incidents and responses. The important role of context is less prominent in such theories though, and we suggest that drawing on additional approaches, such as sociocultural perspectives, can add further insight and open up new possibilities for exploration. For example, research by Morcom (2015) uses a sociocultural framework to scaffold pupils' social and emotional learning about bullying within the peer group through 'shared affective spaces'. She chose a range of methods including interviews, observations, and reflective logs. Classroom discussions and activities were used to build collective peer knowledge about consequences of their words and actions, promote prosocial behaviour and facilitate the development of positive relationships. Carried out over the course of a school year, the work involved active social participation of peers, the development of shared knowledge and established new traditions and practices within the class group. In this way, the focus was on actions, participation and shared responsibility within the group -providing a subtle change of emphasis from group norms and allowing particular classroom and peer group contexts to be considered.
Sociocultural perspectives may be used to explore how the peer group learns and negotiates their own norms, practices and rituals (Wenger, 1998). New members will learn established patterns of interacting, and what characteristics are highly valued. Participating in the community will involve learning the practices in order to facilitate and reinforce a sense of belonging. Therefore, bullying will be more common in peer groups where aggression, dominance and negative forms of interaction are normalised. For example, Hamarus and Kaikkonen's (2008) qualitative analysis of pupils' bullying accounts revealed how bullying had ritualistic elements that strengthened peer group belonging. It also served to establish and recreate culturally valued ideas and characteristics by targeting individuals who were identified as 'different' to the peer group. The authors argue that bullying has a cultural dimension to it, shaped through established group norms which serve to 'other' those who are not seen to belong.
In a similar way, non-participation (such as not joining in with a bullying situation, or not intervening in a particular incident) also contributes to identity by signalling non-membership of a group (either through active choice or exclusion); or a trajectory of participation within a community (either through peripherality or marginality) (Wenger, 1998). Within sociocultural theoretical work, others have described similar phenomena as 'self-imposed' withdrawal where a child is present but absents themselves from engaging in the practices of the context (de Abreu & Hale, 2009;Hedegaard, 2005). For example, a child witnessing a bullying incident but not participating could signify they do not belong to that particular group, so their non-participation establishes boundaries between different peer group communities. Or, it could be because they do not have a strong enough position within that group (i.e. they assume a marginal position), or they are a newer member who is slowly becoming integrated (.i.e. they assume a peripheral position). Indeed, Forsberg et al.'s (2014) study into bystander behaviour reported that relative status in the peer group was one of the factors influencing decisions to intervene.
In order to address multiple roles operating in bullying situations, interventions have been proposed which serve to increase pupil's individual responsibility in their response to bullying incidents they may witness, and help peers to "withstand the dynamics of the peer group" (O'Connell et al., 1999, p450). This assumes that individuals have agency to separate themselves from the dynamics of the group they are part of. However, Rogoff (2003) describes people and cultural communities as "mutually creating each other" (p37), meaning that rather than seeing individuals as separate to, yet influenced by, their cultural surroundings, both are mutually entwined and interdependent. The essence of having a shared enterprise (in this case, responses to bullying) brings the community together, but also presents an opportunity for members to negotiate ways of being. Therefore in order to promote positive responses to bullying and support for children involved, the key would be to challenge the practices of the peer group as a whole and encourage a collective re-appraisal of group norms. An intervention such as the established KiVa program is a good example of how this can be manifested in practice. KiVa treats bullying as a group problem, and raises pupils' awareness of their collective responsibility to addressing it through promoting group defender behaviour and developing change strategies for supporting children who are bullied (Salmivalli, Kärnä, & Poskiparta, 2011). Whilst this approach is not explicitly sociocultural in nature, it makes important headway in addressing the role of the peer group in bullying. It is our premise however that there may be further potential in this area with an increased focus on context, meaning, cultural practices, community participation and situated relationships.

Bullying as part of someone's life trajectory
Research on children involved in bullying reports evidence suggesting that certain children are more likely to engage in bullying behaviours, or experience being victimized. In other words, characteristics of a child make it more likely that bullying will form part of their life course. This largely comes from approaches focusing on establishing a range of characteristics and outcome measures, with the underlying assumption of stability of traits that remain an integral part of the self over long periods of time. For example, research using 'personality traits' perspectives have noted links between certain traits and participant roles that children adopt in bullying situations. Using a measure of the Big Five personality model, Tani, Greenman, Schneider, and Fregoso (2003) reported that children classified as bullies scored higher on extraversion but lower on agreeableness, defenders scored highest on agreeableness, whereas children who were bullied scored highest on emotional instability. There are also reported links with emotional intelligence, with adolescents who bully others scoring lower on the ability to understand the emotions of others (Lomas, Stough, Hansen, & Downey, 2012). Children who get bullied have been rated by peers, teachers and themselves as having social skills problems, particularly in areas such as appearing weak, looking unhappy, and giving in easily thereby creating the impression of being an 'easy target' which may attract bullying behaviour (Fox & Boulton, 2005).
One problem with a trait approach is that research tends to focus on either 'risk factors,' which may increase children's susceptibility to bullying or 'protective factors,' which may protect children from bullying involvement (e.g. Boel-Studt & Renner, 2013;Ttofi & Farrington, 2012). Whilst this recognises that there are environmental factors that interact to determine whether these vulnerabilities or protective factors are put into practice, it does not consider the dialogicality of identity positions that may be played out by children in any given situation (Bakhtin, 1973;Hermans, 2001). Sociocultural perspectives would talk about identity positioning rather than personal characteristics because the framework supposes that people show changeable identity positions across different contexts and over time. This notion could also be described as a 'nexus of multimembership' (Wenger, 1998, p158) to signify how identities in one context do not necessarily cease to exist in another but speak to each other in an ongoing dialogue. However, our engagement in different practices might alter how the context influences our identity positioning.
We suggest that a sociocultural perspective enables an understanding of bullying as a boundary crossing activity. For example, there is a body of work that shows a link between bullying in school and children's relationships with key figures such as mothers (in terms of responsiveness and over-protectiveness) (Georgiou, 2008), the level of father involvement in the family (Flouri & Buchanan, 2003) and family cultures where there has been discord, harsh discipline, violence, and abuse (Duong, Schwartz, Chang, Kelly, & Tom, 2009;Olweus, 1993;Stevens, De Bourdeaudhij, & Van Oost, 2002). The majority of this research employed survey designs within a positivist framework to analyse relationships between variables and develop statistical models based on predictor variables and outcomes. This work has provided some really useful evidence to understand how aspects of relationships might link to behaviours in other contexts, and has identified important areas of potential vulnerability to bullying involvement. What is less clear from this approach however, is how individuals make meanings from the situations they are in, what meanings they might take from the relationships they have with others, and how they make sense of their experiences and negotiate links between settings. We might propose from the above findings that children living in families where certain forms of interaction and aggressive behaviours are commonplace and normalised, learn these relational styles through their belonging in family life, and play them out in other settings such as school. Adopting a sociocultural position would enable exploration of these processes and relationships as they mediate each other. There is evidence in the bullying literature to suggest that through participating in a peer group community, a child learns certain mechanisms of interaction; and forms expectations about how to behave and how others will respond to them. They become who they are through playing out familiar and established patterns in their community, which makes them view and interpret the world in particular ways. For example, secondary school students who experienced bullying by their peers have talked about 'double victimization' whereby their victim role is constantly reconfirmed both through the treatment of their peers (external victimization) and the internalisation of a victim image by the individual (internal victimization) (Thornberg, Halldin, Balmsjö, & Peterson, 2013). In this qualitative grounded theory interview study, students explained how they assumed a victim identity and acted accordingly through changes in their behaviour including selfdoubt, distrusting others, and self-blaming. Some became resigned to their role, and began to expect it (Thornberg et al., 2013). The authors used social interactionist theories looking at stigma and labelling to underpin their work, but we also see opportunities for sociocultural theorising, which has long been interested in culturally-mediated practices of the past, present and future (O'Toole & de Abreu, 2005;Valsiner, 2008). From this perspective, people draw on the past, intentionally and unintentionally, to inform their current and future practices (O'Toole & de Abreu, 2005). Children who have been bullied or bullied others have been exposed to and learnt particular forms of interacting with others, and this internalisation of knowledge and meanings from past experiences could be used to mediate new encounters. Firstly, this might explain how and why someone's bullying experiences in one context (such as at school) can be reproduced in another (such as a different school -Schäfer, Korn, Brodbeck, Wolke & Schulz, 2005;or the workplace -Smith, Singer, Hoel, & Cooper, 2003). Secondly, this helps us to understand how and why the effects of bullying on psychological and behavioural processes can be longstanding for an individual.
Bullying can be prolonged and may leave a 'mark' on the individual, and shapes their future trajectories both in terms of their self-perceptions and ways of relating to others. For children who are bullied at school, there is an increased risk for experiencing bullying in later life (Smith et al., 2003;Widom, Czaja, & Dutton, 2008). As such, their 'victim role' is reproduced in future settings they participate in. Participation in bullying can be viewed as a 'developmental entity' (Beach, 1999, p124) which has significance for ongoing developmental trajectories. The importance of various activities for subsequent development variessome can be more significant than others at various times. This depends on when events happen in the person's history, and the characteristics of a society that determine typical sequences of activities, and their relative importance (Beach, 1995;Wenger, 1998). Bullying has historical significance as a societal activity in terms of the attention it gets in media and literature; policy in schools and popular discourse. Recognising this can help us to understand the ways in which this can influence individuals. If someone is bullied at a significant period in their development, the experience may be particularly profound for them and shape their identities in the longer term.
There is increasing interest in the literature about victim resilience, and understanding how and why some children seem to be able to cope better and experience fewer effects than others. Research by Sapouna and Wolke (2013) found that qualities such as high self-esteem, and positive relationships with parents and siblings were associated with fewer negative outcomes following bullying. Sociocultural theory might be more inclined to look at the different social and cultural resources people draw on to help them manage (or not) the situation (Crafter & Maunder, 2012;Hendry & Kloep, 2002). So, someone who assumes a particular 'victim' or 'bully' status at certain points will not necessarily remain so (see evidence from Wolke, Woods, & Samara, 2009). People can move in and out of identity positions as bullies or victims, and it is not a simple dyadic relationship. Identities are constantly re-negotiated, which may lead to previous identities being reproduced in new settings, but also allows for re-creations to take place along the way. If we view bullying through this lens it opens up the possibility that the 'bully' or 'victim' identity assumed by an individual at a particular time can change, and does not have to inevitably lead to long term effects. In other words, not all children experiencing bullying will go on to be bullied again in the future, and may not all be affected negatively by it in the future.

Implications and conclusions
We have illustrated that a range of empirical research into school bullying can be interpreted using a sociocultural framework. This approach is beneficial because it recognises, and helps to explain, the complexity of bullying in schools. We have argued that bullying is contextual, mediated by situated relationships between individuals in social contexts that have their own norms and practices. The multiple settings and groups in which individuals interact mutually creates their behaviour, attitudes, social relationships and identities. Therefore it is understandable that bullying is interpreted differently; changes as society shifts; occurs in some groups more than others; and is experienced and responded to differently by individuals.
A key implication for work on school bullying is that sociocultural perspectives force us to move our attention away from the individual bully-victim dyad, and focus more on the contextual, historical and institutional influences that surround the behaviour. Part of this involves assigning greater responsibility to others with whom children interact, and challenging community-based norms which may normalise certain forms of aggressive behaviours or interactive styles. As Monks et al. (2009) argue, "bullying rarely takes place between two individuals in isolation" (p154), and the behaviour and attitudes of other members of the community contribute to its existence and continuation (Wenger, 1998). Adopting a sociocultural interpretive framework encourages us to recognise the different spheres of influence operating on individuals, and reinforces the importance of tackling bullying at all levels of the system (Bronfenbrenner, 1979). We need to move from relying heavily on individualised, blame-based or punitive approaches for addressing bullying, and consider the contextual factors operating around individuals which result in particular patterns of behaviour. This means shifting attention away from seeking to isolate particular variables that may lead someone to engage in bullying or be bullied, and instead recognising the constellation of wider influences that interact together to create bullying in a particular situation (Patton et al., 2013). We need to understand what triggers might be important for initiating particular patterns of behaviour, interactions or responses in specific circumstances, and what conditions might inhibit such practices. We also need to study the reciprocal interactions between people as part of their membership of multiple and intersecting communities, and recognise the important role of identity through participation and ongoing meaning making.
For research into bullying, sociocultural approaches require methodologies which address context, enable the exploration of meaning making, study interactions and relationships, and look across time. We recognise the challenges here, and suggest that adopting multiple methods to study a topic or question may afford the flexibility and sensitivity that is required to capture the relevant nuances. Whilst some of the work we've reported here does address context in this manner, and highlights issues akin to our key arguments, quite a portion of this work does not frame that around any kind of theoretical stance. It is interesting to note that papers explicitly looking at theory on school bullying are relatively few, and a high proportion of empirical work that we have encountered present themselves as atheoretical through their lack of specific reference to theory. This is not a criticism, because no theoretical stance can provide all of the answers and it is therefore important to consider what works and what is appropriate in a particular context (Rigby, 2004). That being said, we do feel that the potential of sociocultural perspectives for understanding school bullying has not yet been fully realised, and that much additional insight can be gained from applying this framework to what is known (and still unknown) about bullying.
For researchers undertaking empirical work in this field, we encourage reflection on the methodological approaches adopted, questions posed and assumptions made. We encourage consideration of additional interpretations and potential alternative explanations for trends in data obtained, and advocate the adoption of tools that allow issues to be looked at from multiple perspectives. For example, an overreliance on self-reports may run the risk of confusing what people say they do with what they actually do. Being mindful not to assume too much from data which is gathering representations leaves open the possibility that people will act differently in different situations depending on context. Taking a theoretical stance like sociocultural theory can overcome some of the potential pitfalls around what is a representation and what is an action because it lends itself to looking at activity and meaning. Sociocultural perspectives challenge the notion of stability in behaviours, characteristics and interactionsrecognising that roles and actions are more complex, involving subtle judgements and multifaceted relationships.
With regard to bullying intervention, there is increasing interest in studying the contexts which influence bullying (e.g. Espelage, 2014;Kochenderfer-Ladd & Troop-Gordon, 2010). Cowie and Jennifer's (2008) whole school approach involves considering contextual influencesreferred to as 'risk factors' -operating at different levels which impact on children's engagement in bullying behaviours, and resonates notably with the Ecological System theory (Bronfenbrenner, 1979). For example, at the individual level they talk about personal characteristics and elements of somebody's biological and personal history that might make them more vulnerable to bullying (reflecting the earlier discussion on trajectories). At an interpersonal level they discuss an individual's relationship with family and their peers, and the types of behaviours that are normalised and valued (reflecting our discussion on relational aspects, and normalised behaviour). At a community level, they refer to factors operating in the school or local neighbourhood, such as presence of gangs; school ethos; and problems with unemployment. Finally, at the societal level they include prevailing social and cultural norms which might normalise aggression as an acceptable means form of interaction. Hamarus and Kaikkonen (2008) note that "bullying occurs in a society where there are norms that underpin it as an ongoing process" (p342), meaning that if we want to successfully address it we need to adopt approaches which operate at a societal as well as school level. This means schools working in partnership with parents and communities, alongside anti-bullying efforts within the school (Espelage, 2014).
Literature on bullying intervention has consistently favoured wholeschool approaches with outside involvement (e.g. Cowie & Jennifer, 2008;DfE, 2014;Olweus, 1993). The most effective school-based interventions are those which adopt approaches operating at a wholeschool level, classroom level, and individual level, whilst also involving parents and the wider community (Cowie, 2011). One of the reasons why such sophisticated programs have been the most successful are likely due to the way they address the multi-layered influences operating in children's experience, and recognise the various contexts in which they participate (Espelage, 2014). O'Connell et al. (1999) argues strongly that bullying interventions orientated towards the peer group need to be strengthened by concurrent initiatives which "challenge existing social conditions that tolerate, and inadvertently promote, bullying and victimization within the peer context" (p450). As Cowie (2011) notes, the culture of the school exerts a strong link on the amount of bullying that occurs between members of that community (including pupils, teachers and within staff teams). Therefore, "bullying needs to be seen embedded within the culture of the organization where it is taking place and in order to reduce its prevalence, we need to focus on changing the system rather than the individuals within it" (Monks et al., 2009, p154). This, in our view, is best achieved when a sociocultural perspective is adopted in order to understand, explain and address school bullying.

Funding statement
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.