• 1 Introduction

  Many theorists view the restorative justice meeting as a transition ritual, bringing about rapprochement between the participants. This process is often interpreted in terms of emotional bonding: discussing the distressing event and its consequences generates empathy, mutual understanding and the intention to reaccept the offender. Some interpret this process in terms of ‘empathic resonance’ and ‘collective vulnerability’ (Abramson & Moore, 2002; Moore, 1993), others in terms of intensified emotions, a shared mood and collective effervescence (Rossner, 2013). However, other researchers have assessed that emotional communication in restorative justice practices is often flat, cautious or reserved, especially in youth conferences (Riley & Hayes, 2018; Snow & Sanger, 2015). In that context, emotional bonding is generally at odds with conferencing realities. A recent book about language use in these meetings notices the following:

  Having surveyed the restorative justice literature and theories of reintegrative shaming and emotional transformation, we were puzzled by the relative absence of sustained emotional language and behaviour in the conferences. Why didn’t the Young Person (YP) cry? Why was the apology often prompted by the Convenor rather than fervently offered by a visibly contrite offender? Where was the ‘passion play’? In fact, what we observed seemed fairly procedural (Zappavigna & Martin, 2018: 269).

  The suggestion that the conference dynamics will typically result in emotional bonding and will have beneficial impacts on the participants seems to be overstated. Further into the text, I will refer extensively to research findings of youth justice conferencing to make this clear. Moreover, the question is whether an approximation of viewpoints of the participants is brought about by emotional dynamics themselves. Presumably, narrative re-storying is of more importance to explain a congruence of viewpoints.
  To explain the rather low emotional temperature of many restorative justice conferences, I will place emphasis on emotion management theory, thereby suggesting that participants’ reservations and discomfort are related to rather demanding display rules. According to sociologist Arlie Hochschild (1979) – the well-known exponent of this theory – display rules are (often latent and informal) social guidelines that direct what we ought to feel in specific social contexts and how we ought to show it. Emotion management is the work individuals do while ‘inducing or inhibiting feelings so as to render them “appropriate” to a situation’ (Hochschild, 1979: 551). Thus, whereas ritual theories focus on the group process – mutual focussed attention, thereby mirroring other people’s emotions (Rossner & Meher, 2014) – emotion management theory explains how people try to mediate the gap between social norms and what is experienced individually.¹xFor a clarification of the characteristics of these (and other) emotion theories see Turner and Stets (2006).
  Although originally applied to occupational groups and organisations, Hochschild’s analysis of emotion work is highly relevant for restorative justice. I will point out that emotion work has the function to manage tensions between individual affects and conference expectations. I will also outline that the display rules which participants would have to comply with form an integral part of a relatively compelling ‘emotional regime’, a specific set of affective behavioural norms which define the ‘manners’ during the meeting. In this regime, there is considerable social pressure to conform to norms and standards concerning how to (try to) express emotions. For example, a participant may enact empathic or respectful feelings, but in fact have reservations. This demands managing the participant’s own emotions, for instance incomprehension or dislike. This perspective may explain why many participants – especially young people – show a resigned attitude: they seem to be unable to meet the high demands of emotion work, managing their inner feelings of alienation or disconcertedness. It is quite clear that this perspective contradicts the restorative justice rhetoric of informal and spontaneous dialogue between autonomous participants. To date, the aspects of emotional regime and display rules – elaborated in emotion management theory²xSee Bergman Blix and Wettergren (2016), Sieben and Wettergren (2011) and Wettergren (2019). Hochschild (2003) does not use the term ‘emotional regime’. – have not been integrated in restorative justice research.³xHowever, there are some beginnings, see Asmussen (2015, 2018). This study aims to contribute to theory innovation, by exploring ‘emotion work’ in restorative justice.
  This does not mean that ritual theories are less relevant to interpret restorative justice meetings. I will argue that the perspective of ritual bonding needs to be complemented with an analysis of emotion work and emotion management. This contribution has two specific research goals. First, I will argue that ritual theories – at least those which are dominant in restorative justice literature – place too much emphasis on the potential positive impacts of emotional bonding. I will discuss some critical issues with respect to emotional bonding and point out that mutual understanding is rather the result of narrative re-appraising and re-assessing. Second, I will outline restorative justice’s emotional regime and specify which typical display rules are enacted by participants in restorative meetings. I will also identify reasons why (young) participants cannot get a grip on these rules and resort to a rather resigned performance.
  Explaining ‘high’ or ‘low’ emotional communication is in need of serious reservations. Emotion and discourse constitute an extremely complex conceptual field and a ‘deeply murky territory’ (Wetherell, 2012: 52). Terms such as affect, discourse and narrative, each have many incongruent meanings. The complexity of emotion, affect and related terms is evident; not only that, the term ‘discourse’ can also be understood in many divergent ways, ranging from language practice to social meaning-making. Because emotions and cognition are inseparable, it is extremely difficult to assess causal relationships between affects, discourse and narrative meaning-making.
  Moreover, studying emotional dynamics depends largely on which formats, settings and social contexts are the object of research. Similarly, it depends on the type of participants (juveniles or adults; distressed or non-distressed victims; first offenders or recidivists; etc.) and settings (mediation or conferencing etc.). I will mainly focus on the mainstream format of juvenile justice. There is much reason to keep in mind the nuanced view of Harris, Walgrave and Braithwaite (2004: 199), which they worded as follows:

  Conferences do not happen in a social vacuum. In every conference, the emotional dynamics are different, due to the styles in which they are facilitated; the social positions, relationships, personalities and the roles of the participants (not only of the victim and the offender); the nature and circumstances of the offence and its consequences; and other favourable or unfavourable conditions.

  The same authors point out that conferences do not unfold along simple patterns. There are no single concrete precepts applicable to all restorative conferences. Moreover, the specific unfolding of conferences depends in large part on cultural contexts.
  Section 2 offers an overview of two important ritual theories, sketching ‘salutary’ emotional bonding processes. First, I will focus on David Moore’s and Retzinger and Scheff’s theories, both formulated around shame management and empathic resonance, and subsequently Rossner’s perspective, based on Collins’ interaction ritual theory. Both theoretical approaches will be critically discussed insofar they suggest that the emotional dynamics of successful conferences inherently bring about change in the participants’ points of view. It seems more obvious to study their changed views by analysing discursive reframing and sense-making responses. In Section 3, I will point out that poor emotional communication can be explained – paradoxically – by the rather compelling ‘emotional regime’ and its ‘display rules’ – such as enacting cooperativeness and sincere commitment – which are typical of a conferencing culture. Many (young) participants feel alienated and uncomfortable in this context and do not succeed in complying with the norms to show the appropriate emotions and to deliver an authentic apology. In the final section, I point out we need both ritual and emotion management perspectives to explain emotional communication in restorative justice. It is argued that complying with prescriptions as to how to express emotions runs counter to the intuitions of many restorative justice protagonists.

• 2 Ritual theories of emotional bonding

  The two ritual theories dealt with here conceive the meeting as a transition ritual, leading to emotional bonding. The first theory – in which shame management and affective resonance are the central notions – is developed by David Moore and related authors and is largely based on Tomkins’ affect script theory. This explicit normative theory has had (and still has) a huge influence in the restorative justice movement, especially in Australia and the United States (Kelly, 2014; Thorsborne, 2016). I will discuss the well-known study of Retzinger and Scheff (1996) as a variant of this first theory.⁴xI will not deal with John Braithwaite’s classical notion of reintegrative shaming, which can be viewed – admittedly – as a ritual theory (Braithwaite & Mugford, 1994), but does not focus on the emotion dynamics and its stages and sequences. Braithwaite is primarily interested in the social effects of this ‘reintegrative shaming’ process, such as prevention, crime control and rehabilitation of the offender. The second theory – elaborated by Meredith Rossner (2013), pursuing Collins’ interaction ritual chain theory – is characterised by emerging collective effervescence and group solidarity. This theory is based on systematic research findings and is not the result of practitioners’ views as is the case with Moore and his co-authors. I will first elucidate the two theories and subsequently address some critical issues in order to complement and refine them.

  2.1 Empathic resonance

  David Moore was one of the initiators of Australian restorative justice conferences. Since 1993, he has published numerous works focusing on restorative justice dynamics (Abramson & Moore, 2002; McDonald & Moore, 2001; Moore, 1993; see also Van Stokkom, 2002). Moore and co-authors point out that restorative justice conferences have the potential to elicit hurt-revealing emotions such as sorrow and grief, revolving around empathic resonance. The authors differentiate between four (ideal-type) stages (Abramson & Moore, 2002: 135). In the first stage, when the distressful event is discussed, emotions such as contempt, anger and fear are visible. In the next stage, revelations about the past and its consequences bring forward ‘a powerful silence’. It engenders a sense that ‘we are all in this together’. The third stage is characterised by a sense of ‘collective vulnerability’, a physiological manifestation of collective ‘deflation’. When the offender is feeling ashamed or when the victim recounts his or her pain and sorrow, the other participants respond with ‘empathic resonance’: they share another’s distress (Moore, 1993). This stage is ‘the fulcrum of the conference’, ‘the point at which the general tone is poised to shift from negative to positive’ (Abramson & Moore, 2002: 135). In the final stages of the conference, the participants show clear signs of relief on their faces. Abramson and Moore describe this whole process as ‘connected learning’, ‘an understanding-beyond-cognition which can happen only when feelings are shared’ (2002: 136). Although each participant takes a risk when sharing and connecting, the authors believe that the process ‘brings about a sense of satisfaction, excitement, and a sense of renewal in those who participate’ (2002: 136).
  Moore contends that the concept of ‘empathic resonance’ captures precisely the ‘powerful experience of shared emotions’ in the conferences. Offenders observe the distress of victims and begin to grasp their point of view, whereas victims observe helpless offenders, thereby lightening the burden of their anger. The mutative force is empathy, not shame (McDonald & Moore, 2001; Moore, 1997).
  Building upon the theories of experimental psychologist Silvan Tomkins and psychiatrist Donald Nathanson, Abramson and Moore point out that emotions and the facial expressions that convey those emotions are the primary source of human motivation. The affect system and its physiological programs – for example, the response of care when people are upset or sad – work as an amplifier and draw attention to whatever signal that is deemed important by the participants (Abramson & Moore, 2002).
  In 1996, Retzinger and Scheff published a profound study of emotion dynamics in restorative justice conferences, based on nine observations in Australia. Like Moore’s analysis, their study is focussed on all relevant emotions which might play a role in restorative justice, not only shame. They, too, stress that the most significant information in these conferences is conveyed not with words but with facial expressions, gestures and physical posture.
  According to Retzinger and Scheff, moral indignation is the most visible emotion during conferences. They call indignation ‘helpless anger’ because victims and their supporters are often incapable of describing the enormity of the infringement. The authors argue that uncontrolled and repeated expressions of indignation are the largest obstacle to symbolic restoration, because it hinders social bonding and identification. The transformation of emotions that mask suffering into emotions that reveal suffering, including shame, seems to be the key to successful conferences, because – in the words of Retzinger and Scheff (1996) – it makes identification, and therefore reacceptance, possible between the parties. If the victim expresses painful emotions, this may resonate with the offender; subsequently, he or she may show regret and remorse and – as part of a chain reaction – the victim may feel an urge to forgive. Offering an apology and expressing forgiving words constitute the ‘core sequence’ of restorative meeting. The authors add that even if the emotional exchange is short-lived – lasting perhaps only a few seconds – this exchange is the key to restoring the victim’s peace of mind and inducing a sense of reacceptance in the offender (Van Stokkom, 2002).
  The theories dealt with in the preceding paragraphs have some common assumptions. First, they highlight the potential destructive effects of shaming and try to identify the specific conditions under which shame feelings can be diverted in constructive ways. Second, empathic resonance – the sharing of vulnerable feelings – produces emotional bonding and therewith possibilities to recognise each other’s standpoints.

  2.2 Interaction ritual and emotional energy

  Drawing from research findings of Australian and UK restorative justice practices, Rossner aims to gain a deeper understanding of what makes restorative justice a potentially powerful experience for its participants (2011: 97). Her theoretical background is considerably different from those sketched previously, as she uses an interaction ritual chain theory, developed by sociologist Randall Collins (2004). Collins has developed a set of principles for predicting when an interaction will ‘work’, and make us feel good, or ‘fail’, and make us feel bad. He maintains that in successful rituals, a conversational and bodily rhythm develops over time. This rhythm results in a distinct feeling of solidarity and group membership (Rossner, 2011: 96).
  An interaction ritual is a social encounter in which participants have a shared focus of attention and a shared mood. When people become ‘caught up in the rhythm and mood of the talk’ (Collins, 2004: 48), it may culminate in a sense of collective effervescence and corresponding feelings of social solidarity. If this vital dynamic can be achieved, Rossner emphasises, it may bring about longer-term impacts on one’s emotional energy, including initiative, feelings of confidence, elation and a recommitment to the group’s standards of morality (Rossner, 2017).
  This model functions as a Durkheimian religious ritual, in which the synchrony in participants’ conversation, bodily movements and emotions leads to feelings of strength, hope and trust among participants. Using Collins’ framework, Rossner accentuates that ‘it is the dynamic process of building rhythm and a shared focus over time in the conference that culminates in expressions of group solidarity’ (2013: 11). She refers to these moments as ‘emotional turning points’ (2013: 60-65): the rhythm changes from conflict and reservation to mutual solidarity. Participants in ‘high energy’ conferences are hugging, crying, laughing or touching each other, and this leaves them hungry for more positive interactions, ‘thus motivating them to engage in pro-social behaviour (2013: 36). This group solidarity has a far-reaching impact: ‘It … is translated into long-term emotional well-being and the potential for reduced offending (2013: 11). Rossner points out that ‘high solidarity’ conferences may result in significantly less offending than ‘low solidarity’ conferences.
  The facilitators she interviewed in the London study categorised their best conferences as being the most emotionally intense for all parties. Therefore, it is important to engage participants in conversational rhythm and balance. In failed rituals, there is no rhythm; the participants are uncomfortable the entire time, and there is no dramatic turning point, leaving participants flat, deflated or angry (Rossner, 2017).
  Rossner believes that the micro-level production of solidarity and shared emotion provides restorative justice – in theory – with the unique power to achieve its standards and goals. She tends to see the interaction ritual dynamics as malleable and controllable. ‘A good facilitator can ensure the right ingredients are there to create a successful conference’ (Rossner, 2013: 142). They should ‘maximize the expression of emotional displays’ (2013: 80). The focus should be on bringing emotions to the surface: ‘the stronger the emotions, whether they be positive or negative, the better the conference’ (2013: 81).

  2.3 Critical comments

  Moore’s idea that the most relevant information is conveyed with facial expressions, gestures and physical posture, is also endorsed by Rossner. However, whereas the former argues that releasing the underlying painful emotions allows space for recognition of others’ feelings, Rossner focusses more on the positive feelings of a shared mood and group solidarity, thereby pushing to the background the shame concept that has dominated restorative justice for so long. Restorative conferences are not simply about ‘shame management’ or rechannelling certain emotions. Successful conferences depend on the ritual outcome: group solidarity and cohesion.
  Although Moore and co-authors have the intention to describe the dynamics empirically, they are fully convinced of the merits of restorative justice. They are not really prepared to assess and describe less fruitful aspects or unintended consequences. For example, these authors stick to the apology–forgiveness script: when the offender apologises, the victim would be keen to forgive. However, it is the question whether forgiveness plays a preponderant role in conferences.⁵xForgiveness is rarely reported in youth conferences in Northern Ireland where more than 17,000 conferences were held until 2016 (Chapman & Chapman, 2016). The studies by Shapland and colleagues (see Shapland, 2016) into mediation with adult perpetrators in England show that in approximately two-thirds of the cases the perpetrator expressed an apology. In response, however, victims did not grant forgiveness, although they were moved by the words of the perpetrator. It is logical that forgiving words are so often missing. Forgiveness is usually the result of a process of processing pain, sadness and loss, and that process does not have to be completed when the meeting takes place.
  Likewise, Rossner’s approach contains high expectations, related to the assumption that a successful conference should bring forward emotional energy. For example, she suggests that conferences can act as a turning point in the offenders’ life, ‘providing them with the long term emotional energy needed to stop offending’ (Rossner, 2013: 145). This seems to be questionable. Even when the offender is involved in emotional energy, it is doubtful whether this will lead to sustained behavioural change. As Rossner recognises, after the conference has taken place, the accumulated emotional energy will die out soon. Another questionable issue relates to the apology. Rossner argues that an apology is likely to come only after a build-up of shared focus and emotional intensity between victim and offender as they express their emotions and tell their stories (Rossner, 2017). Nonetheless, offenders can feel remorseful prior to the conference and express their regret during the meeting, even if there is no shared mood or conversational synchrony.
  Although both theories may explain ritualistic unfolding of emotional bonding well, it is doubtful whether ‘shared mood’ and ‘social bonding’ are always representative for conference realities.⁶xIn an important article – written together with Jasmine Bruce – Rossner acknowledges that only a minority of adult conferences have characteristics of shared mood, shared understanding and meaningful connection between the participants (Rossner & Bruce, 2018). So-called civil conferences – characterised by consensus but lacking intense displays of solidarity and emotional energy – are more frequent. Other conferences fall flat, are tiring, lack mutual goodwill, and still others are (partly) divisive. This study focuses explicitly on overlapping narratives of accountability and harm. In this way, the study offers an integrated approach, including storytelling, performance and emotional communication. The authors also put into perspective the idea that only conferences with emotional energy can be termed successful. Interaction rituals do not simply produce ‘ups and downs’ of emotional energy; they also frequently result in potent feelings of disaffiliation and animosity. Dominating persons may feel energised and hopeful, whereas dominated actors may feel de-energised, resigned, frightened or de-moralised. Disengaged interaction drains energy and may cause participants to experience indifference or fatigue, which reduces the level of agency (Boyns & Luery, 2015).
  In the following, I will take a closer look at two other issues, first the neglected role of narration and meaning-making in the two ritual theories, and second, the tendency to maximise emotions.

  2.3.1 Narration and meaning-making: the importance of cognitive reappraisals

  The ritual theories discussed previously suggest that in successful conferences, shared emotions bring about a change in the points of view of the participants. However, in order to understand perspective change, it is important to focus on speech acts and narratives: the participants’ adjusted views on the distressing event and its aftermath rely on storytelling and meaning-making. For example, appraisal theories of emotion make clear that changed viewpoints are rather the result of cognitive reframing and reappraisal and that shared emotions function as an amplifier, supporting motives and identification. These theories exemplify that emotions result from persons’ interpretations and explanations of events or acts, even in the absence of physiological arousal.
  In a critical discussion of Collins’ theory, Christian von Scheve (2012) contends that emotional bonding may emerge on the condition that appraisals – and their representative and motivational contents – are shared. Only an alignment of these appraisals can lead to the elicitation of similar emotions. In simple terms: ‘you can’t make something out of nothing’; that is, without a minimal set of shared appraisals, the effervescence and enhanced solidarity are unlikely to occur. He concludes that collective emotions are not necessarily able to produce solidarity, but rather they enhance solidarity.
  In a different context, social psychologist Bernard Rimé points out that negative experiences – which disconfirm our expectations – mobilise attention to the production of meaning. Rimé makes a distinction between two types of reactions to a negative event (2009: 75). Socio-affective reactions offer the narrator emotional support like attention, comfort, consolation and empathy; these reactions are often nonverbal and have the potential to reduce distress. On the other hand, cognitive reactions stimulate the narrator’s sense-making, which helps them reframe and reappraise the emotionally distressing episode and, thus, reorganise his or her motives and goals. Research evidence shows that social sharing, which develops along the socio-affective route, is well-suited to alleviate narrators’ insecurity and helplessness but does not contribute to overcoming a past negative experience. By contrast, cognitive articulation and re-storying can contribute to a progressive distancing from the distressing event and to the adoption of a realigned point of view.
  Rimé’s findings show that a narrative reconstruction of reality is helpful for victims (and other participants) to reinterpret distressing events and reconsider the grounds on which their beliefs rested. In this process, emotion functions as a response through which attention is redirected to the production of meaning (Bruner, 1990).⁷xRe-storying is the core idea of narrative mediation. Its goal is to destabilise the initial personal conflict story and open the door for new interpretations (Winslade & Monk, 2000). It focuses on identifying opportunities for redirecting the viewpoints of the participants. However, the main thrust in this approach seems to be cognitive; the participants’ emotional experiences fade into the background. This seems also the case with (re)positioning theory (Langdon, 2016).
  Summarising, the emotional bonding approaches dealt with in the previous sections tend to overlook the question how the participants develop reappraised views on the distressing event and its aftermath. Ideally, ritual theories should focus on the interplay of narrative reappraisals and emotional communication. Re-storying and emotional bonding depend upon each other and each is an indispensable element in a transition ritual.⁸xIconisation theory (Zappavigna & Martin, 2018) offers an interesting perspective in that respect. It deals with ritual interaction, nonverbal communication and the production of bonding symbols (‘icons’) in a balanced way.

  2.3.2 Maximising emotions?

  This subsection focuses on the ambition to ‘maximise’ emotions, respectively body affects and emotional energy. This ambition is derived from the two underlying theories, namely, Tomkins’ affect script theory and Collins’ interaction ritual theory. I will point out that each theory contains somewhat questionable assumptions, which reverberate in the emotion bonding perspectives dealt with earlier.
  Tomkins and Tomkins-inspired theorists insist that affects are innate, visceral forces and intensities that influence our thinking, but are separate of cognitive functions. They must be viewed as nonsignifying, autonomic processes that take place below the threshold of conscious awareness and meaning, independent of and prior to intentions. Whereas emotions are learned and come pre-packaged to individuals, affects are pre-personal, innate programs, which reflect bodily ‘intensity’ (Leys, 2011: 442; Wetherell, 2015: 145).⁹xFor critical reviews of affect script theory, see Leys (2011) and Wetherell (2015). Affect script theory is struggling with many other problems, such as conceptual imprecision and subjectivism (see Leys, 2011).
  Affect script theory¹⁰xThe term ‘script’ refers to an acquired affect management mechanism (Thorsborne, 2016: 30). has had (and still has) a huge influence on restorative justice scholars and practitioners concerned with transforming conflicts in institutions as families, schools and workplaces. Kelly (2014) and other psychologists who have popularised Tomkins’ theory point out that a simple framework, called the Central Blueprint, enhances our survival. It contains principles as ‘maximise positive affect’, ‘minimise negative affect’ and ‘minimise the inhibition of affect’. In the restorative process, these rules can be worded as ‘we should come together to share and maximize positive feelings’ (Thorsborne, 2013: 35). According to Kelly (2014: 30), ‘effective restorative practices restore the ability of individuals and communities to live and function in emotionally balanced ways consistent with the biological directives of the Central Blueprint’. If participants are motivated to minimise negative affect, it would be ‘biologically inevitable’ that undesirable feelings will diminish (Kelly, 2014: 70). Evidently, those rules and directives do not reflect the complex realities of conference practices and seem to bind participants to biologically determined ‘affect programs’.¹¹xFor a similar critique, see Knight, Phillips and Chapman (2016).
  Collins sketches human individuals as emotional energy (EE) seekers, trying to maximise EE. They are looking forward to those situations in which the EE payoff is highest: they maximise interactions that make them feel good and avoid those that make them feel bad.¹²xIn Collins’ terms, ‘human behaviour may be characterized as emotional energy tropism. Social sources of EE directly energize behaviour; the strongest energizing situation exerts the strongest pull’ (2004: 181/2). Basically, human beings are ‘emotional energy seekers, thereby linked to those interactions and their derivative symbols that give the greatest EE in the opportunities presented by each person’s social networks’ (2004: 373). This radical conception of man has been the subject of much criticism. Some critics point out that the EE concept is reductive and one-sided and has an imperialist character. Every human motive must fit the model of EE. If people commit themselves to other concerns and goals, these are redefined as instruments in the pursuit of EE (Baehr, 2005; Salmela, 2014).¹³xOthers fear that energy-oriented ‘choices’ seem to incite people to become myopic: propelled by impulses we would be incapable of anticipating the costs involved when we take part in rituals (Iagulli, 2016: 424). According to emotion theorist Jack Barbalet, Collins dismisses the value of symbols in choice determination. Moreover, he adheres to the quantitative measure of EE when a qualitative factor is all the more necessary: the quality of intentionality, the goal-setting propensity of emotions (2006: 450).¹⁴xBarbalet points out that there are two problems here: ‘it is simply inconceivable that emotion can have an only energetic form without affective content; second, what symbols convey is invariably the affect not the energy. EE is denatured emotion: it is only responsible for the energy in the actors and their interaction, not the direction that interaction takes. Presumably the latter comes from the ritual itself’ (Barbalet, 2006: 450).
  In the same spirit as Collins, Rossner remarks that ‘the stronger the emotions, whether they be positive or negative, the better the conference’ (2013: 81). She emphasises that intense emotions – also negative ones such as anger – may contribute to focus on the interaction and develop the rhythm and entrainment needed for successful interaction rituals. However, the suggestion that emotional energy is the most important factor for a ‘successful’ conference is questionable. Presumably, the quality of emotional communication is more important. Sincere and fair expression of emotions may elicit understanding and compassion. Likewise, a reasonable expression of experiences is important. Anger accompanied by an accurate reference to relevant facts is likely to be interpreted as convincing. The participants’ emotions have more impact when they are worded in acceptable and appropriate ways.¹⁵xRossner and Bruce (2018) recognise the importance of the way in which participants present stories, arguments and claims. They outline some types of talking which cause divergence between the participants, such as talking in generalities rather than specific harm, overstating or understating the harm and aggressively defending the offenders’ position. In sum, viewing restorative justice as a ritual of maximal outpouring of emotions is beside the point. It would be pointless to make emotions such as compassion and remorse the object of quantifying. These emotions reveal important moral values. We cannot reduce these emotions to physical energy, zest, urges, appetite or desires which – indeed – may be maximised.
  These critical comments are not meant to downgrade the relevance of emotional bonding theories per se. Although the theories discussed previously overlook re-storying experiences and tend to overstate the positive aspects of absorption in the group process, they are well-equipped to explain commitment, enthusiasm and elation in a subgroup of vital conferences.

• 3 The ‘managed heart’ in restorative justice

  The ritual theories dealt with in Section 2 suggest that emotional bonding will lead to mutual recognition and value restoration. However, as stated in the introduction, the image of a passion play – and its assumed positive outcomes – seems to be at odds with many conferences and mediation sessions. Youth justice conferences are not the scene of shared emotional outpourings. Actually, conferences in this category do not achieve what restorative justice proponents claim, in particular offering a convincing apology.¹⁶xMany victims do not receive an apology or the apology is deemed insincere (Blecher, 2011). Various research findings learn that dialogue and common ground are often absent and that participants believe that they are misunderstood by the other party (Daly, 2006; for an overview, see Suzuki & Wood, 2017). Rossner points out that conferences with juvenile offenders are less successful – compared with conferences with adult participants – because they tend to have less productive emotions and involve less powerful reintegration rituals (Rossner, 2012).
  Ritual theories assume that irritated feelings, reservations or dissatisfaction are attributable to a lack of shared mood or empathic resonance. However, there are many other factors which might explain disengaged or reluctant attitudes. Maybe the most important one is that the conference in itself can be experienced as an intimidating setting, in which participants must comply with demanding display rules.
  Whereas ritual theories assume – implicitly – that emotional expressions are reflections of the participants’ needs and interests, emotion management theory conceives dialogues as a matter of presenting a self that fits into a specific cultural context. Cultural norms determine what emotions can be expressed and which expressions are acceptable (Asmussen, 2018). This implies that people must manage emotional displays when there is a discontinuity between what they feel and what they must express to others in the presence of an audience. In her book The managed heart, Arlie Hochschild (1983) pointed out that people are often caught in a conflict between display rules¹⁷xHochschild uses the terms ‘feeling rules’ and ‘display rules’ simultaneously. The first term does not seem appropriate because large parts of our feelings are not visible. – the culturally desired emotional expressions – and their actual feelings. As stated in the introduction, these discrepancies require people to engage in what Hochschild calls ‘emotion work’, applying emotion management strategies, thereby giving simultaneous attention to one’s own as well as others’ emotions.
  In this section, I will point out that these display rules belong to a relatively compelling emotional regime, which is typical for restorative justice conferences. This means that participants are supposed to be engaged in performances directed by typical conferencing norms and ideals such as apologising. The question is in which respects participants may succeed to apply these display rules. I will point out that ‘emotion work’ is often so demanding that many young offenders resort to an ‘offstage performance’.
  Of course, poor emotional communication is attributable to many factors, such as language deficiencies, social disadvantages and identification with tough and deviant behaviour that is common within peer groups, including use of alcohol and drugs. Juveniles fifteen years old or younger are in many respects still emotionally immature. This manifests itself in nervousness and being unable to deal with feelings of shame. They do not yet have sufficient empathic care needed to be able to understand the needs of the victim.¹⁸xHayes refers to research suggesting that approximately 50 per cent of young people in the youth justice system have a clinically significant but undiagnosed language deficit (Hayes, 2017: 411). ‘Young people on the autism spectrum and those who have experienced maltreatment (known to be overrepresented in the histories of young offenders) are particularly prone to difficulties in this domain and may in fact display alexithymia – a lack of words for emotions’ (Hayes & Snow, 2013: 3). Participants from socioeconomically disadvantaged groups are consistently found to be more likely to struggle with restorative styles of speech (Cook, 2006; Cunneen & Goldson, 2015; Willis, 2018). Although all these factors are important, the conferencing ‘regime’ itself may nevertheless exert pressure on the participants, possibly leading to poor forms of emotional communication.

  3.1 Restorative justice as an emotional regime

  Here I define an ‘emotional regime’ as a specific set of prescriptions or norms about how to express emotions, which can be sanctioned by gossip, reprimands or exclusion.¹⁹xIn sociological literature, there is considerable confusion about what constitutes an emotional regime. William Reddy (2001) introduced this concept in discussing political orders and their propensity to separate highly valued emotions from deviant ones. Reddy’s perspective is very broad and encompasses national or international (sub)cultural trends such as romanticism and consumerism. This set of norms reflect practical knowledge about emotional communication, embodied in a social institution.²⁰xWettergren (2019: 35) defines an emotional regime as ‘a set of tacit and explicit knowledge about emotion, including when certain emotions are appropriate and how they should be displayed, embedded in behavioural norms pertinent to given social groups’. Hochschild does not use the term ‘emotional regime’. Her perspective on emotion management is micro-oriented, based on a Goffmanian tradition. Therefore, she cannot make clear how culturally desired emotional expressions are (re)produced on the meso-level of social institutions.²¹xSee Tonkens’ criticism (2012). She identifies a gap in Hochschild’s work between micro-level concepts as display rules and macro-level phenomena such as commercialisation. For a study of emotional prescriptions in organisations and social institutions, see Sieben & Wettergren (2011).
  I postulate that any social institution – an academy, a religious congregation or the judiciary – contains an emotional regime.²²xTo widen the scope of this definition issue, I assume that each social institution has a cultural regime and a set of social norms that regulate its interactions (often called ‘scripts’ or ‘codes’). This cultural regime can be divided into an ideational regime and an emotional regime which refer, respectively, to beliefs (culturally desired ideals and goals) and display rules (culturally desired emotional expressions). Emotion researchers Stina Bergman Blix and Åsa Wettergren (2016) point out that the judiciary has an overarching emotional regime of ‘legal dispassion’ in which there is barely room for emotional communication. Very differently, in the absence of adequate definitions, we can term the emotional regime of restorative justice as centred around apologising and ‘authentic confession’. Ida Asmussen (2015: 38) refers in this context to a ‘confessional ethos’, including expressions of regret and reacceptance. In this respect, related aims and ideals of conferencing are important, such as enhancing the opportunity to express authentically what happened and requesting explanations for the harm done (Cook, 2006: 110).²³xThese aims and ideals belong to restorative justice’s ideational regime (see the previous note). In sum, the emotional regime seems to be geared towards speaking sincerely and authentically and showing mutual respect.
  I assume that restorative justice’s emotional regime can be experienced as strict and compelling, even so much that participants remain silent, not so much because they feel embarrassed, but rather because they feel uncertain about how to enact sincere and authentic emotions. The emotional regime gives less space to informal and free expression of emotions than might be expected from a procedure wherein participants are invited to express what bothers them. A contributing factor in this is that the scripted restorative justice conference is rather formal and procedural and involves a high degree of repetition. The conference has strict deliberate turn-taking structures, and the convenor uses appropriate prompts.²⁴xSee Zappavigna and Martin (2018: 290). They point out that the enactment of the conference does involve a hyper-charged axiological process (2018: 291). This leaves the participants – to some extent – uncertain about what to say and how to meet the expectations.²⁵xOn the contrary, mediators/facilitators are fully embodied in this regime and are able to background their emotions, not becoming the object of cognitive reflection. Wettergren (2019) discusses backgrounding emotion in the professional setting of the judiciary.
  In this respect, Hayes and Snow (2013) speak about the youth conference as ‘a high-stakes and highly verbal interchange’. ‘Young offenders need to listen to complex and emotionally charged accounts of the victims’ perspectives and formulate their own ideas into coherent narratives that are judged to be adequate and authentic by victims and other participants in the conference’ (Hayes, 2017: 410). Basically, all participants have to express their emotions in a highly conversational and morally charged process. In this context, participants may feel themselves uncomfortable and alienated.
  While involved in this emotional regime, participants must comply with requirements to manage emotions, and they are not familiar with how to regulate these emotions accordingly. However, they cannot avoid taking up emotion work if they want to uphold the impression that they are committed to the moral purposes of the conference. This means that the participants have to ‘tune in’ to another’s affective state and to use appropriate words to describe their own (Snow & Sanger, 2015: 330).
  As stated in the introduction, display rules can be defined as social guidelines that direct what we ought to feel in specific social contexts and how we ought to show it. The participants are expected to enact the prescribed emotional expressions. There are at least two important display rules which are characteristic for social interaction in restorative justice. They can be described as follows.

  1. Enact a sincere and authentic role. If participants want to summon consciously emotions like sorrow and grief, to help create the desired actions (e.g. expressing an apology or show comprehension), their ‘real’ feelings should be controlled. The communication around apologising is an exacting moral exercise: the offender must convince the victim of his worthiness. He must display regret with genuineness and authenticity. That task demands close attention to the mode of expression and is complicated by fear of rejection and feelings of helplessness (Hayes, 2017: 413). To apologise is to display a humble stance, but this demands that feelings of dissatisfaction or anger are managed. These feelings would be disruptive of the action focus and if these feelings pop up, they should be managed consciously.

  2. Enact cooperativeness. Often participants do not understand what others mean and they may get irritated; sometimes they want to tackle others’ opinions and claims forcefully. There are many other factors which might endanger mutual respect, including the participants’ urge to behave indifferently, defiantly or arrogantly. These affects would also thwart the action focus and have to be contained by displaying sympathy and goodwill, suggesting cooperation and commitment.

  Obviously, there are other display rules, such as the enactment of attentiveness and the willingness to go into conversation.²⁶xEvidently, these display rules are characteristic for many types of social deliberation. These rules relate to all participants. In the following, I will argue that, in particular, young offenders will not be able to fulfil adequate emotion work and that they tend to minimise their contribution in the discussion.

  3.2 Explaining the offender’s ‘offstage performance’

  A Danish study concludes that youngsters demonstrate a recurring, resigned pattern, limiting themselves to an ‘offstage performance’ (Asmussen, 2015).²⁷x‘Offstage performance’ is an ambiguous ‘Goffmanesque’ term: although disengaged offenders are not involved in performance, they are physically present and called to account. ‘Offstage’ behaviour might be interpreted as a defensive strategy: escaping interaction and creating an internal locus of control, which can be situated on the other end of the continuum from high emotional energy (Summers-Effler, 2004; see also Ybema & Horvers, 2017). Simply being in a room full of adults might be an important reason for youngsters resorting to an offstage role. Bolitho (2012: 69) reported the following about the youth conference sphere:

  It may well be that the storytelling process is challenging for young offenders in an environment they may find intimidating and frightening; a victim is present, the process is court-sanctioned, and the room is filled with adults.

  Many young offenders are nervous, worrying and brooding. They often do not know how to handle the alienating setting in which they find themselves. The display rules are unconsciously avoided or obeyed half-heartedly because they do not correspond with inner feelings. In this context, their ‘low key’ performance is not surprising.²⁸xInterestingly, when the victim was not present, young offenders were less affected by the conference (Riley & Hayes, 2018: 108).
  Hennessey Hayes illustrates how challenging some young offenders may find having to give an effective description of the circumstances surrounding their offending. For example, young offenders used lots of one-word or one-sentence answers as ‘I don’t know’ or ‘I wasn’t thinking’ (Hayes, 2017: 414). The conference convenor had to assist these young people in telling their story and suggest possible responses to which the young people would agree. Moreover, they have difficulties in identifying and describing their emotions, whether their own or those of others. Riley and Hayes (2018: 100) bring forward the following findings:

  Most young offenders did not understand what they had been asked by the conference facilitators, they had difficulty grasping the seriousness of what they had done, and had difficulty articulating their remorse. They often answered ‘yes’ to questions to hide their lack of understanding and to ‘get out’ of the conference as quickly as possible. For the majority of young offenders, the language was ‘over their heads’.

  Typical for their communication are minimal verbal responses, shoulder shrugs and poor eye contact. In this context, many young offenders tend to minimise their commitment and enact a ‘small target persona’: adopting physical postures of absence (leaning forward etc.), attempting to minimise the extent to which they come under negative scrutiny (Zappavigna & Martin, 2018: 212).²⁹xSnow compares this communication challenge with speaking in a weak second language. ‘Young people taking part in an RJ conference are effectively in a ‘second language’ environment if their receptive and expressive skills fall many years behind expected levels’ (Snow, 2013: 19).
  Some researchers point out that young offenders, when communicating poorly, run the risk of appearing lazy, rude or unmotivated. Shoulder shrugs, poor eye contact and very brief or one-word replies to questions are misinterpreted by adult participants as indicative of disinterest and/or disrespect (Riley & Hayes, 2018). Not engaging with the process evokes labels as ‘bored’, ‘emotionally unaffected’ and ‘lacking in empathy’ (Snow & Sanger, 2015). Many adult participants assume that offenders chose to be incommunicative and thus lack commitment or sincerity. Similarly, adult participants judged silent parents to be less concerned than active parents who could engage in these processes (Cook, 2006; Willis, 2018).³⁰xVictimologists have found that victims interpret the invisibility of signs of shame or remorse as a lack of such emotions, even if perpetrators experience these (Lens et al., 2015; see also Karstedt, 2019: 106-108). Moreover, this labelling process is reinforced by status inequality: if young offenders are not similar in terms of social status, empathising with them is hampered (Cuff et al., 2016).
  These biased perceptions may demoralise the young participants and reinforce their reluctance to face the victims and explain their actions. Some of them – often unconsciously – revolt against their marginalisation in the process (Riley & Hayes, 2018: 101).³¹xSee also some American research findings, such as Presser & Hamilton (2006) and Gerkin (2009). These defiant reactions become manifest when the victim or another adult participant addresses them in paternalistic ways, provokes discussion or wants to teach them a lesson.³²xIn Blecher’s study, it was found that approximately one-third of the offenders behaved stubbornly or unresponsively (Blecher, 2011). Regularly, offenders are belittled, and the way they are approached is more stigmatising than being reintegrative. They oppose this and feel ‘too intimidated to speak’ and ‘pushed around by others with power’ (Blecher, 2011: 105). Some resist against codes of restorative justice and position themselves as a ‘victim’ of unfair treatment and misunderstandings on the part of the authorities, his/her family and/or others (Asmussen, 2015: 37; Kenney & Clairmont, 2009).

• 4 Conclusions and discussion

  The ritual theories dealt with in Section 2 aim to explain the ‘magic’ of the conference, conjuring up mutual understanding and solidarity. These theories assume that the participants’ efforts to adapt their views on the initial conflict is the result of a process of emotional bonding. I have argued that it is more plausible to interpret emotional bonding as an amplifying process, enhancing feelings of sympathy, consolation or relief. It’s not this emotional ‘glue’ but cognitive responses to a distressing event which initiate a process of re-storying, thereby reappraising the initial conflict story. I have pointed out that emotional bonding and narrative re-storying depend upon each other and that each is an indispensable element in the ritual group process. I also pointed out that the idea of ‘maximising’ emotions is fraught with difficulties and that the quality of emotional communication – including the value of emotional symbols – is of more importance.
  Ritual theories of emotional bonding often assume that poor emotional communication is the result of a lack of rhythm and shared focus. In Section 3, I pointed out that this poor communication is rather the result of restorative justice’s compelling emotional regime which prescribes how the participants should express their emotions. Young participants feel uncomfortable, anxious and uncertain and thus minimise their efforts and assume a resigned ‘offstage’ role. Whereas in ritual theories personal discomfort and confusion seem to fall off the radar, emotion management theory pays attention to the ambivalent interests of participants and their recurring doubts and uncertainties. This theory focuses on the work it takes to cope with display rules and the tensions involved. However, emotion management theory – specifying the kinds of cultural codes which determine how emotions are used – has its own deficiencies. Because of the focus on reasoning about feelings, these theories neglect the power of the pre-reflexive and physical aspects of emotions (Theodosius, 2006; Turner & Stets, 2006).
  Ritual theories and emotion management theory are each unable to explain in detail how emotional communication in restorative justice meetings unfolds. However, both are indispensable. We need an analysis of ceremonial bonding and sense-making as well as emotion management requirements. We have to combine both theories, regardless how different they may be. Taking this into account, we may distinguish between two different modes that could be the object of studying participation. On the one hand, we have a ‘detached mode’ in which self-interest plays a dominant role, for example, young offenders who want to get off from the embarrassing meeting as soon as possible. On the other, we have a ‘committed mode’, in which the participants show focus and devotion, often unaware of their entrainment in collective emotions. Participants may – in part unconsciously – switch between these modes.
  To date, authors studying restorative justice did not say much about emotional management. One of the reasons for that is the influence of restorative justice orthodoxy, namely that the participants are involved voluntarily, expressing emotions spontaneously and authentically. Thanks to the setting’s informality, the participants can have ample room to speak out. Emotion management theory provides a different perspective, emphasising tensions between the required display rules and the participants’ inner feelings. This perspective assumes that many (young) participants are overwhelmed by restorative justice’s emotional regime and resort to an ‘offstage performance’. They experience the required performances exacting and onerous and, subsequently, minimise their efforts to fulfil the requirements of a ‘confessional ethos’. This may also explain why offering a convincing apology is so hard to achieve. In this context, Zappavigna and Martin discuss a conundrum (2018: 248):

  How can YPs sound sincere when they can’t be themselves (or at least not the selves that adults may expect them to be based on their interactions with them in social processes of other kinds)? How can a YP employ, for instance, the kind of talk conference participants might associate with teenagers, when this discourse is at odds with the heavily scaffolded nature of the interaction?

  Evidently, the assumption that restorative justice has a voluntary social climate needs readjustment. In some respects, the social pressure to comply with display rules can be interpreted as ‘compulsory compassion’ (Acorn, 2004). However, this should not be understood as intentionally encouraging participants to apologise or reconcile. As pointed out, participants feel implicit pressure to comply with norms about how to express emotions.
  Some researchers suggest that young offenders strategically conceal their ‘real’ affective states. For example, young offenders would be very well capable to show empathetic facial expressions and hide their masculine self-identifications.³³xIn an observational study, Boran Ali Mercan (2018) points out that emotional cover-ups of anxiety and fear can be caught up between the lines of young male offenders’ discourse. He assumes that offenders’ anxiety feelings are enhanced by the uncertainty of what will happen with them and which punishment they will get. After all, conferences are still the settings of establishing justice. However, as indicated before, many young participants are rather reluctant and clumsy role players. Moreover, restorative justice’s emotional regime does not provide much scope to engage in strategic ploys. Anybody’s body language is thoroughly scrutinised, and strategic intentions would be detected quickly.
  The assumption that offenders act strategically prompts the conclusion that young offenders are insincere and insensitive to the needs of the victim. Yet what is interpreted as insincerity of the participants is in many cases merely a sign of uncertainty about how to express themselves that is engendered by the thorny setting in which they are situated. Moreover, they can feel concerned and remorseful without expressing it. If it is this setting which hampers communication, it would be erroneous as well to attribute the offender’s reticence and timidity solely to shame feelings, as many researchers do. Likewise, one may question the assumption that disengaged attitudes of young offenders are solely the result of emotional immaturity and, hence, the inability to empathise with victims. I would suggest that emotional immaturity is rather an indirect problem: it militates against an understanding of the display rules inherent in restorative justice conferences, rather than impeding empathy per se.³⁴xThanks to an anonymous reviewer who brought this to my attention. Whatever the truth of these assumptions, young persons’ reticence raises the question of what to do with their low apparent engagement. Presumably, better preparation of young persons, especially about what will take place in the meeting and which emotion work is expected to occur, might have a positive impact.
  The assumptions made in this article, as well as many unfolded argumentations, especially those in the context of restorative justice’s emotional regime, are in dire need of further exploration and refinement. Given the theoretical deficiencies and underexplored paths I mentioned, the conclusion is warranted that we are still at a relatively early stage of understanding emotional communication in conferences. As John Braithwaite (2016) pointed out, we are still awaiting a mature and full-fledged restorative justice theory.

References

• Abramson, L. & Moore D.B. (2002). The psychology of community conferencing. In J. Perry (ed.), Restorative justice: repairing communities through restorative justice (pp. 123-140). Lanham: American Correctional Association.

• Acorn, A. (2004). Compulsory compassion: a critique of restorative justice. Vancouver: University of British Columbia Press.

• Asmussen, I.H. (2015). Performing absolution narratives in restorative justice. Restorative Justice International Journal, 3(1), 28-48. doi: 10.1080/20504721.2015.1049870.

• Asmussen, I.H. (2018). Mediation in light of modern identity. In A. Nylund, K. Ervasti & L. Adrian (eds.), Nordic Mediation Research (pp. 133-144). Cham: Springer Open.

• Baehr, W.P. & Collins, R. (2005). Review forum: the sociology of almost everything: four questions to Randall Collins about interaction ritual chains. Canadian Journal of Sociology, January-February, 1-11.

• Barbalet, J. (2006). Emotional payoffs of ritual. European Journal of Sociology/Archives Européennes de Sociologie Europäisches Archiv für Soziologie, 47(3), 446-451. doi: 10.1017/s0003975606000191.

• Bergman Blix, S. & Wettergren, A. (2016). A sociological perspective on emotions in the judiciary. Emotion Review, 8(1), 32-37. doi: 10.1177/1754073915601226.

• Blecher, N.J. (2011). Sorry justice: apology in Australian family group conferencing. Psychiatry, Psychology and Law, 18(1), 95-116. doi: 10.1080/13218719.2010.509038.

• Bolitho, J.J. (2012). Restorative justice: the ideals and realities of conferencing for young people. Critical Criminology, 20(1), 61-78. doi: 10.1007/s10612-011-9150-z.

• Boyns, D. & Luery, S. (2015). Negative emotional energy: a theory of the ‘dark-side’ of interaction ritual chains. Social Science, 4(1), 148-170. doi: 10.3390/socsci4010148.

• Braithwaite, J. (2016). Redeeming the ‘F’ word in restorative justice. Oxford Journal of Law and Religion, 5(1), 79-93. doi: 10.1093/ojlr/rwv049.

• Braithwaite, J. & Mugford, S. (1994). Conditions of successful reintegration ceremonies: dealing with juvenile offenders. British Journal of Criminology, 34(2), 139-171. doi: 10.1093/oxfordjournals.bjc.a048400.

• Bruner, J. (1990). Acts of meaning. Cambridge: Harvard University Press.

• Chapman, T. & Chapman, A. (2016). Forgiveness in restorative justice: experienced but not heard? Oxford Journal of Law and Religion, 5(1), 135-152. doi: 10.1093/ojlr/rwv066.

• Collins, R. (2004). Interaction ritual chains. Princeton: Princeton University Press.

• Cook, K.J. (2006). Doing difference and accountability in restorative justice conferences. Theoretical Criminology, 10(1), 107-124. doi: 10.1177/1362480606059987.

• Cuff, B., Brown, S.J., Taylor, L. & Howat, D. (2016). Empathy: a review of the concept. Emotion Review, 8(2), 144-153. doi: 10.1177/1754073914558466.

• Cunneen, C. & Goldson, B. (2015). Restorative justice: a critical analysis. In B. Goldson & J. Muncie (eds.), Youth, crime and justice (2nd ed., pp. 137-156). London: Sage.

• Daly, K. (2006). The limits of restorative justice. In D. Sullivan & L. Tift (eds.), Handbook of restorative justice: a global perspective (pp. 134-145). New York: Routledge.

• Gerkin, P.M. (2009). Participation in victim-offender mediation: lessons learned from observations. Criminal Justice Review, 34(2), 226-247. doi: 10.1177/0734016808325058.

• Harris, N., Walgrave, L. & Braithwaite, J. (2004). Emotional dynamics in restorative conferences. Theoretical Criminology, 8(2), 191-210. doi: 10.1177/1362480604042243.

• Hayes, H. (2017). Emotion and language in restorative youth justice. In The Palgrave handbook of Australian and New Zealand criminology, crime and justice (pp. 407-419). Cham: Palgrave Macmillan.

• Hayes, H. & Snow, P. (2013). Oral language competence and restorative justice processes: refining preparation and the measurement of conference outcomes. Trends & Issues in Crime and Criminal Justice, 463, 1-7.

• Hochschild, A.R. (1979). Emotion work, feeling rules, and social structure. American Journal of Sociology, 85(3), 551-575. doi: 10.1086/227049.

• Hochschild, A.R. (1983). The managed heart: commercialization of human feeling. Berkeley: University of California Press.

• Iagulli, P. (2016). Randall Collins and the sociology of emotions. Italian Sociological Review, 6(3), 411-429.

• Karstedt, S. (2019). Emotions and criminal law: new perspectives on an enduring presence. In R. Patulny, R. Bellocchi & A. Olson (eds.), Emotions in late modernity (pp. 101-113). Abingdon: Routledge.

• Kelly, V. (2014). Caring, restorative practice and the biology of Emotion. In V. Kelly & M. Thorsborne (eds.), The psychology of emotion in restorative justice: how affect script psychology explains how and why restorative practice works (pp. 26-53). London: Jessica Kinsley Publishers.

• Kenney, J.S. & Clairmont, D. (2009). Using the victim role as both sword and shield: the interactional dynamics of restorative justice sessions. Journal of Contemporary Ethnography, 38(3), 279-307. doi: 10.1177/0891241608322814.

• Knight, C., Phillips, J. & Chapman, T. (2016). Bringing the feelings back: returning emotions to criminal justice practice. British Journal of Community Justice, 14(1), 45-58.

• Langdon, J.J. (2016). Talk it out: toward a narrative theory of community conferencing. Contemporary Justice Review, 19(1), 19-30. doi: 10.1080/10282580.2016.1101932.

• Lens, K., Pemberton, A., Brans, K., Braeken, J., Bogaerts, S. & Lahla, E. (2015). Delivering a victim impact statement: emotionally effective or counterproductive? European Journal of Criminology, 12(1), 17-34. doi: 10.1177/1477370814538778.

• Leys, R. (2011). The turn to affect: a critique. Critical Inquiry, 37(3), 434-472. doi: 10.1086/659353.

• McDonald, J. & Moore, D. (2001). Community conferencing as a special case of conflict transformation. In H. Strang & J. Braithwaite (eds.), Restorative justice and civil society (pp. 130-148). Cambridge: Cambridge University Press.

• Mercan, B.A. (2018). Doing criminological research: affective states versus emotional reactions. Theoretical Criminology. doi: 10.1177/1362480618779399.

• Moore, D. (1993). Shame, forgiveness, and juvenile justice. Criminal Justice Ethics, 12(1), 3-25. doi: 10.1080/0731129x.1993.9991934.

• Moore, D. (1997). Pride, shame and empathy among peers: community conferencing as transformative justice in education. In K. Rigby & P.T. Slee (eds.), Children’s peer relations (pp. 254-271). London: Routledge.

• Presser, L. & Hamilton, C.A. (2006). The micropolitics of victim-offender mediation. Sociological Inquiry, 76(3), 316-342.

• Reddy, W.M. (2001). The navigation of feeling: a framework for the history of emotions. Cambridge: Cambridge University Press.

• Retzinger, S.M. & Scheff, T.J. (1996). Strategy for community conferences: emotions and social bonds. In B. Galaway & J. Hudson (eds.), Restorative justice: international perspectives (pp. 315-336). Monsey: Criminal Justice Press.

• Riley, M. & Hayes, H. (2018). Youth restorative justice conferencing: facilitator’s language—help or hindrance? Contemporary Justice Review, 21(1), 99-113. doi: 10.1080/10282580.2017.1413358.

• Rimé, B. (2009). Emotion elicits the social sharing of emotion: theory and empirical review. Emotion Review, 1, 60-85. doi: 10.1177/1754073908097189.

• Rossner, M. (2011). Emotions and interaction ritual: a micro analysis of restorative justice. British Journal of Criminology, 51(1), 95-119. doi: 10.1093/bjc/azq075.

• Rossner, M. (2012). Emotion across the life-course: the case for restorative justice with adults. In J. Bolitho, J. Bruce & G. Mason (eds.), Restorative justice: adults and emerging practice (pp. 214-230). Sydney: Sydney Institute of Criminology.

• Rossner, M. (2013). Just emotions: rituals of restorative justice. Oxford: Oxford University Press.

• Rossner, M. (2017). Restorative justice and victims of crime: directions and developments. In S. Walklate (ed.), Handbook of victims and victimology (2nd ed., pp. 229-246). London: Routledge.

• Rossner, M. & Meher, M. (2014). Emotions in ritual theories. In J. Stets & J. Turner (eds.), Handbook of the sociology of emotions (Vol. II, pp. 199-220). New York: Springer.

• Rossner, M. & Bruce, J. (2018). Trajectories and typologies of pre-sentence restorative justice rituals. Australian & New Zealand Journal of Criminology, 51(4), 502-518. doi: 10.1177/0004865817749263.

• Salmela, M. (2014). Collective emotions as ‘the glue’ of group solidarity. In A. Laitinen & A.B. Pessi (eds.), Solidarity: theory and practice (pp. 55-87). Lanham: Lexington Books.

• Shapland, J. (2016). Forgiveness and restorative justice: is it necessary? Is it helpful? Oxford Journal of Law and Religion, 5(1), 94-112. doi: 10.1093/ojlr/rwv038.

• Sieben, B. & Wettergren, A. (2011). Emotionalizing organizations and organizing emotions. London: Palgrave Macmillan.

• Snow, P. (2013). Restorative justice conferencing, oral language competence, and young offenders: are these high-risk conversations? The Prevention Researcher, 20(1), 18-21.

• Snow, P.C. & Sanger, D.D. (2015). Restorative justice conferencing and the youth offender: exploring the role of oral language competence. International Journal of Language & Communication Disorders, 46, 324-333.

• Summers-Effler, E. (2004). Defensive strategies: the formation and social implications of patterned self-destructive behavior. In J. Turner (ed.), Theory and research on human emotions. Advances in group processes (Vol. 21, pp. 309-325). Bingley: Emerald Group Publishing Limited.

• Suzuki, M. & Wood, W.R. (2017). Is restorative justice conferencing appropriate for youth offenders? Criminology & Criminal Justice, 18(4). doi: 10.1177/1748895817722188.

• Theodosius, C. (2006). Recovering emotion from emotion management. Sociology, 40(5), 893-910. doi: 10.1177/0038038506067512.

• Thorsborne, M. (2013). Affect and script psychology. In B. Hopkins (ed.), Restorative theory in practice (pp. 25-45). London: Jessica Kingsley Publishers.

• Turner, J. & Stets, J. (2006). Sociological theories of human emotion. Annual Review of Sociology, 32, 25-52. doi: 10.1146/annurev.soc.32.061604.123130.

• Van Stokkom, B. (2002). Moral emotions in restorative justice conferences: managing shame, designing empathy, Theoretical criminology, 6(3), 339-360.

• Von Scheve, C. (2012). Collective emotions in rituals: elicitation, transmission, and a ‘Matthew-effect’. In A. Michaels & C. Wulf (eds.), Emotions in rituals and performances (pp. 55-77). New Delhi: Routledge.

• Wetherell, M. (2012). Affecting discourse: investigating the psychosocial. London: Sage.

• Wetherell, M. (2015). Trends in the turn to affect: a social psychological critique. Body & Society, 21(2), 139-166.

• Wettergren, A. (2019). Emotive-cognitive rationality, background emotions and emotion work. In R. Patulny, A. Bellocchi, R. Olson, S. Khorana, J. McKenzie & M. Peterie (eds.), Emotions in late modernity (pp. 27-40). London: Routledge.

• Willis, R. (2018). ‘Let’s talk about it’: why social class matters to restorative justice. Criminology and Criminal Justice. doi: 10.1177/1748895818804307.

• Winslade, J. & Monk, G. (2000). Narrative mediation: a new approach to conflict resolution. San Francisco: Jossey-Bass.

• Ybema, S. & Horvers, M. (2017). Resistance through compliance: the strategic and subversive potential of frontstage and backstage resistance. Organization Studies, 38(9), 1233-1251. doi: 10.1177/0170840617709305.

• Zappavigna, M. & Martin, J.R. (2018). Discourse and diversionary justice: an analysis of youth justice conferencing. Cham: Palgrave MacMillan.

Noten

• * Contact author: b.vanstokkom@jur.ru.nl.Acknowledgements: The author owes many thanks to Meredith Rossner, Susanne Karstedt and two anonymous reviewers.

• 1 For a clarification of the characteristics of these (and other) emotion theories see Turner and Stets (2006).

• 2 See Bergman Blix and Wettergren (2016), Sieben and Wettergren (2011) and Wettergren (2019). Hochschild (2003) does not use the term ‘emotional regime’.

• 3 However, there are some beginnings, see Asmussen (2015, 2018).

• 4 I will not deal with John Braithwaite’s classical notion of reintegrative shaming, which can be viewed – admittedly – as a ritual theory (Braithwaite & Mugford, 1994), but does not focus on the emotion dynamics and its stages and sequences. Braithwaite is primarily interested in the social effects of this ‘reintegrative shaming’ process, such as prevention, crime control and rehabilitation of the offender.

• 5 Forgiveness is rarely reported in youth conferences in Northern Ireland where more than 17,000 conferences were held until 2016 (Chapman & Chapman, 2016). The studies by Shapland and colleagues (see Shapland, 2016) into mediation with adult perpetrators in England show that in approximately two-thirds of the cases the perpetrator expressed an apology. In response, however, victims did not grant forgiveness, although they were moved by the words of the perpetrator. It is logical that forgiving words are so often missing. Forgiveness is usually the result of a process of processing pain, sadness and loss, and that process does not have to be completed when the meeting takes place.

• 6 In an important article – written together with Jasmine Bruce – Rossner acknowledges that only a minority of adult conferences have characteristics of shared mood, shared understanding and meaningful connection between the participants (Rossner & Bruce, 2018). So-called civil conferences – characterised by consensus but lacking intense displays of solidarity and emotional energy – are more frequent. Other conferences fall flat, are tiring, lack mutual goodwill, and still others are (partly) divisive. This study focuses explicitly on overlapping narratives of accountability and harm. In this way, the study offers an integrated approach, including storytelling, performance and emotional communication. The authors also put into perspective the idea that only conferences with emotional energy can be termed successful.

• 7 Re-storying is the core idea of narrative mediation. Its goal is to destabilise the initial personal conflict story and open the door for new interpretations (Winslade & Monk, 2000). It focuses on identifying opportunities for redirecting the viewpoints of the participants. However, the main thrust in this approach seems to be cognitive; the participants’ emotional experiences fade into the background. This seems also the case with (re)positioning theory (Langdon, 2016).

• 8 Iconisation theory (Zappavigna & Martin, 2018) offers an interesting perspective in that respect. It deals with ritual interaction, nonverbal communication and the production of bonding symbols (‘icons’) in a balanced way.

• 9 For critical reviews of affect script theory, see Leys (2011) and Wetherell (2015). Affect script theory is struggling with many other problems, such as conceptual imprecision and subjectivism (see Leys, 2011).

• 10 The term ‘script’ refers to an acquired affect management mechanism (Thorsborne, 2016: 30).

• 11 For a similar critique, see Knight, Phillips and Chapman (2016).

• 12 In Collins’ terms, ‘human behaviour may be characterized as emotional energy tropism. Social sources of EE directly energize behaviour; the strongest energizing situation exerts the strongest pull’ (2004: 181/2). Basically, human beings are ‘emotional energy seekers, thereby linked to those interactions and their derivative symbols that give the greatest EE in the opportunities presented by each person’s social networks’ (2004: 373).

• 13 Others fear that energy-oriented ‘choices’ seem to incite people to become myopic: propelled by impulses we would be incapable of anticipating the costs involved when we take part in rituals (Iagulli, 2016: 424).

• 14 Barbalet points out that there are two problems here: ‘it is simply inconceivable that emotion can have an only energetic form without affective content; second, what symbols convey is invariably the affect not the energy. EE is denatured emotion: it is only responsible for the energy in the actors and their interaction, not the direction that interaction takes. Presumably the latter comes from the ritual itself’ (Barbalet, 2006: 450).

• 15 Rossner and Bruce (2018) recognise the importance of the way in which participants present stories, arguments and claims. They outline some types of talking which cause divergence between the participants, such as talking in generalities rather than specific harm, overstating or understating the harm and aggressively defending the offenders’ position.

• 16 Many victims do not receive an apology or the apology is deemed insincere (Blecher, 2011). Various research findings learn that dialogue and common ground are often absent and that participants believe that they are misunderstood by the other party (Daly, 2006; for an overview, see Suzuki & Wood, 2017). Rossner points out that conferences with juvenile offenders are less successful – compared with conferences with adult participants – because they tend to have less productive emotions and involve less powerful reintegration rituals (Rossner, 2012).

• 17 Hochschild uses the terms ‘feeling rules’ and ‘display rules’ simultaneously. The first term does not seem appropriate because large parts of our feelings are not visible.

• 18 Hayes refers to research suggesting that approximately 50 per cent of young people in the youth justice system have a clinically significant but undiagnosed language deficit (Hayes, 2017: 411). ‘Young people on the autism spectrum and those who have experienced maltreatment (known to be overrepresented in the histories of young offenders) are particularly prone to difficulties in this domain and may in fact display alexithymia – a lack of words for emotions’ (Hayes & Snow, 2013: 3).

• 19 In sociological literature, there is considerable confusion about what constitutes an emotional regime. William Reddy (2001) introduced this concept in discussing political orders and their propensity to separate highly valued emotions from deviant ones. Reddy’s perspective is very broad and encompasses national or international (sub)cultural trends such as romanticism and consumerism.

• 20 Wettergren (2019: 35) defines an emotional regime as ‘a set of tacit and explicit knowledge about emotion, including when certain emotions are appropriate and how they should be displayed, embedded in behavioural norms pertinent to given social groups’.

• 21 See Tonkens’ criticism (2012). She identifies a gap in Hochschild’s work between micro-level concepts as display rules and macro-level phenomena such as commercialisation. For a study of emotional prescriptions in organisations and social institutions, see Sieben & Wettergren (2011).

• 22 To widen the scope of this definition issue, I assume that each social institution has a cultural regime and a set of social norms that regulate its interactions (often called ‘scripts’ or ‘codes’). This cultural regime can be divided into an ideational regime and an emotional regime which refer, respectively, to beliefs (culturally desired ideals and goals) and display rules (culturally desired emotional expressions).

• 23 These aims and ideals belong to restorative justice’s ideational regime (see the previous note).

• 24 See Zappavigna and Martin (2018: 290). They point out that the enactment of the conference does involve a hyper-charged axiological process (2018: 291).

• 25 On the contrary, mediators/facilitators are fully embodied in this regime and are able to background their emotions, not becoming the object of cognitive reflection. Wettergren (2019) discusses backgrounding emotion in the professional setting of the judiciary.

• 26 Evidently, these display rules are characteristic for many types of social deliberation.

• 27 ‘Offstage performance’ is an ambiguous ‘Goffmanesque’ term: although disengaged offenders are not involved in performance, they are physically present and called to account. ‘Offstage’ behaviour might be interpreted as a defensive strategy: escaping interaction and creating an internal locus of control, which can be situated on the other end of the continuum from high emotional energy (Summers-Effler, 2004; see also Ybema & Horvers, 2017).

• 28 Interestingly, when the victim was not present, young offenders were less affected by the conference (Riley & Hayes, 2018: 108).

• 29 Snow compares this communication challenge with speaking in a weak second language. ‘Young people taking part in an RJ conference are effectively in a ‘second language’ environment if their receptive and expressive skills fall many years behind expected levels’ (Snow, 2013: 19).

• 30 Victimologists have found that victims interpret the invisibility of signs of shame or remorse as a lack of such emotions, even if perpetrators experience these (Lens et al., 2015; see also Karstedt, 2019: 106-108).

• 31 See also some American research findings, such as Presser & Hamilton (2006) and Gerkin (2009).

• 32 In Blecher’s study, it was found that approximately one-third of the offenders behaved stubbornly or unresponsively (Blecher, 2011). Regularly, offenders are belittled, and the way they are approached is more stigmatising than being reintegrative. They oppose this and feel ‘too intimidated to speak’ and ‘pushed around by others with power’ (Blecher, 2011: 105).

• 33 In an observational study, Boran Ali Mercan (2018) points out that emotional cover-ups of anxiety and fear can be caught up between the lines of young male offenders’ discourse. He assumes that offenders’ anxiety feelings are enhanced by the uncertainty of what will happen with them and which punishment they will get. After all, conferences are still the settings of establishing justice.

• 34 Thanks to an anonymous reviewer who brought this to my attention.