In contrast with the popular stereotype of bullies as lacking in social skills and understanding, recent studies have underlined that some bullies will need good social cognition and theory of mind skills in order to perpetrate negative actions against victims and to maintain a dominant status within the group. If bullies are cold, manipulative experts in social contexts, the problem of their behaviour can be related to emotions surrounding moral transgression, or to the relation between moral cognition, empathy and moral behaviour. Do bullies understand other children's emotions but just not share them and therefore lack empathy? Or do they simply not recognise suffering and pain in the victims? How do they feel during a bullying episode? Do they simply feel great and proud for having been tough or do they feel guilty and responsible for what they have done?
This contribution aims at analysing those studies which try to investigate emotions understanding, attributions and coping strategies related to a moral transgression. To this end empirical and theoretical studies are reviewed, and a critical analysis of methods is presented.
In the main, there are two different approaches in the study of the relationship between social cognition and social behaviour, and specifically, between emotion and bullying. The first is represented by an information-processing model, that sees aggressive behaviour as resulting from processing biases in one or more steps in a 5 stages social information process (Dodge, Pettit, McClaskey and Brown, 1986; Dodge and Feldman, 1990; Crick and Dodge, 1994). Applied to bullies and victims, the social skill deficit model would suggest that bullies have similar deficits to aggressive children. Victims, on the other hand, would lack social skills of assertiveness and group entry; furthermore, they have less experience in social interaction, in the teasing and play fighting which (both in family and peer relationships) may enhance interpretation processes of emotional expression, social skills, sense of control and self-efficacy (Smith, Bowers, Binney and Cowie, 1993).
On the contrary, the second perspective stresses the role of adaptive motivation to explain bullying behaviour (Smith, 1991; Smith, Bowers, Binney and Cowie, 1993; Sutton, Smith and Swettenham, 1999a). These authors considered the deficit label inappropriate for bullies. Rather than misinterpreting social cues, or having a limited range of response, some bullies may simply be choosing goals of dominance and status reputation by aggressive means. Dodge himself (1991), although his relevant studies have been in the area of social skills deficit, warns against trying to explain all types of aggression in terms of a deficit. Especially proactive aggression, of which bullying is part, can be more related to goal selection and decision making. Thus, this second approach sustained that bullies are skilled manipulators who use their psychological skills to manipulate the minds of others and cause them distress. Sutton, Smith and Swettenham (1999a) found that bullies demonstrated higher abilities than controls and victims in a second order theory of mind test and in other measures of social cognition. These authors spoke about "Machiavellism" to interpret this attitude of bullies' who are often aware of others' feelings but unable or unwilling to share these feelings with others.
If bullies are cold, manipulative experts in social contexts, the problem of their behaviour can be related to emotions surrounding moral transgression such as guilt and shame, or to the relation between moral cognition, empathy and moral behaviour. Do bullies understand other children's emotions but just not share them and therefore lack empathy? Or do they simply not recognise suffering and pain in the victims and therefore lack social skills? How do they feel during a bullying episode? Do they simply feel great for having been tough or do they feel guilty and responsible for what they have done?
And what about the victims? How do they feel in this relationship? What kind of emotions do they experience? Studies on victims' emotional reactions have pointed both feelings of self-pity and self-blame, and also feelings of revengefulness and anger towards their bullies (Borg 1998). Often the victim's answer to a bullying episode is characterised by a sense of helplessness and difficulties to react and to establish a more balanced relationship with the partner. What kind of emotion blocks the victims' reaction and impedes them in being more assertive?
The work of our group aims at analysing those studies which try to investigate emotions, understanding, attributions and coping strategies related to a moral transgression on the side of the bullies and of the victims. This is a socio-cognitive area at the interface of cognitive strategies, moral values, social relationships and emotions. After having defined the area of research our effort will be directed towards giving a critical review of measures and instruments within each dimension.
In the definition of emotions, some authors stress the social nature of emotions and the consequences of emotional states on personal and social relationships. The emotional experience is developmentally embedded in social experience and the two are reciprocally influential. In a functionalistic perspective, emotion is defined as "the person's attempt or readiness to establish, maintain or change the relation between the person and the environment on matters of significance to that person" (Saarni, Mumme and Campos, 1998). Among other dimensions, Saarni (1999) mentioned the following skills as part of emotional competence:
This area has been the ground of the discussion between scholars who tried to have empirical support for the social skill deficit model and those who didn't report any deficits or delays in aggressive and maladaptive children. Some investigators have gathered data that suggest that children with emotional problems or who have been abused show deficits in discrimination of other's expressions (Camras, Grow and Ribordy, 1983; Feldman, White and Lobato, 1982; Walker, 1981). Some studies gave support for a positive link between social competence and emotion recognition, measured by accuracy in identification of emotional facial expression (Denham, McKinley, Couchoud and Holt, 1990; Edwards, Manstead and McDonald, 1984), but also by accuracy in knowledge of situational antecedents and social consequences of emotions (Cassidy, Parke, Butkovsky and Braungart, 1992).
Other studies showed that children rejected and named as aggressive by the peer group were less accurate in emotions' recognition and in understanding situational antecedents than popular children (Goldman, Corsini e De Urioste, 1980; Vosk, Forehand e Figueroa, 1983; Spence, 1987; Ciucci, Tomada, De Domini and Tassi, 1993; Bilello, Casiglia, Lo Coco and Miceli, 1995). Caprara and colleagues (Caprara, Borgogni, Mazzotti e Pastorelli, 1988; Caprara e Pastorelli, 1989; Caprara, 1992) found that children "at risk" (named aggressive and emotionally unstable), compared with "normally-adjusted" children, showed lesser mastery of recognition of facial emotional expressions, of inference of emotions and of choice of effective form of social influence. Lochman and Dodge (1994), examining social-cognitive processes (i.e., affect labelling...) of aggressive and non aggressive boys at preadolescent and early adolescent age levels, found that severely violent boys at both age levels had difficulties with cues recall and presented a pattern of endorsing unusually positive emotions that they may experience in different settings. Moderately aggressive boys displayed indications that their aggression may be more planfully aimed to achieve expected outcomes.
Particularly studies on bullies and victims reported that both bullies and victims tend to misinterpret ambiguous situations; they perceived, for example, physical fighting or rough and tumble play as bullying episodes and failed to recognise feelings and emotions induced by bullying attacks (Smith et al. 1993). Bullies were different from victims and control groups in evaluating the feelings of the episode's participants: they attributed happiness to the aggressor and few of them described the victim as feeling unhappy. The attribution of happiness to the aggressor reveals that in the evaluation of negative episode bullies are more concerned about the gains they may reach than on possible effects their behaviour can have on the victims. The same cognitive schema, called "happy victimizer expectancy" has been reported on normative sample at ages younger than 8 years old (Nunner-Winkler, Sodian, 1988; Arsenio and Kramer, 1992). Focusing on individual differences, Asendorpf and Nunner-Winkler (1992) in an observational study on the relation between moral motivation and behaviour in primary school children found that emotion attribution to the participants in a deceptive and conflictual setting is the best predictor of children's moral behaviour.
The similarity between bullies and younger children may bring us again to the question: do bullies lack social skills or present some kind of impairment in terms of emotion attributions, or do they develop values and beliefs in order to legitimate their negative behaviour?
Furthermore, most bullies thought that the victim provoked bullying episodes (Boulton and Underwood, 1992). Slee (1993) found that most bullies usually overestimate situational factors as compared to disposition variables to explain aggressive episodes. They have negative expectations towards others and perceive other partners' as hostile (Dodge, 1980). These results could suggest that bullies minimize others' feelings, possibly for their own advantage.. Boulton and Underwood (1992) and Salmivalli, Lagerspetz, Bjorkqvist, Osterman and Kaukianien, 1996a) found that in some cases children bully their peers to reach a dominant position in the group or for lack of empathy for their victims.
A study by Ciucci and Fonzi's study (1999) examined the relation between bullying and facial recognition of emotions. To assess facial recognition of emotion, Ekman and Friesen's task (1975) was used on a sample of 62 bullies, 65 victims and 140 control children: children were asked to view photographs of 2 unfamiliar adults (a male and a female) expressing 6 emotions (anger, sadness, fear, happiness, disgust, surprise); for each photograph they were asked to indicate the corresponding label. Victims tended consistently to misinterpret others' emotions more often than bullies and controls; bullies didn't differ from the control group; bullies and victims in particular made more errors than control groups in recognition of happiness.
Nevertheless, the evidence for the social skill deficit is mixed. Silvern and Wortman (1980) reported that among fourth grade boys, more severe antisocial behavior was associated with high scores in perspective taking. Waterman et al. (1981) found that, among emotionally disturbed boys, higher affective perspective taking was significantly correlated with higher antisocial behaviour. Gottman, Gonso and Rasmussen (1975) did not find status group (high- and low-friendship status) differences in the accuracy of labelling emotions from facial expressions. Other results support the view that bullies posses a good social understanding, often higher than their victims. Sutton et al. (1999b) investigated the ability to understand mental states/cognitions and emotions in 7-10 yr. olds and found that bully score (level of ringleader type bullying) was positively correlated with total social cognition score, whereas victim score was negatively correlated with total social cognition score. Bullies roles were also positively correlated with a "cold cognition" (the mean score per cognitive story minus the mean score per emotion story). The authors pointed out the role of socio-cognitive deficits in the case of the victims. The difficulties in understanding others' theory of mind may leave them open to manipulation by others. In the case of bullies a balanced interpretation between the two models (deficits vs. manipulation) has been proposed: bullies may perceive and interpret social cues accurately but they may differ from prosocial or control children in goal selection and decision making because their values and expectations are different from those of other children.
The most used method to measure movements of facial expressions is FACS (Facial Action Coding System by Ekman and Friesen, 1975). The authors, after long training, learned how to control their own facial movements and produced a series of photos where the action of one or more muscles were presented. Ekman and Friesen calculated 46 AU (Action units) related to facial expressions and 12 AU related to changes in gaze direction and head orientation.
Izard (1978) developed a different coding system defined MAX (Maximally Discriminative Affect Coding System). It codifies only those facial movements related to the expression of the 8 basic emotions. This system measures apparent changes in facial expressions and it has been used mainly with infants. The MAX system comprises 29 facial AU located in three areas of the face (forehead/eyes, cheeks/ mouth, lips/chin). The coding system consists in videotaping the three areas separately, decoding the expressions for three times and writing on a grid the movements in the three areas, the onset and offset time. This measure, like FACS (Ekman and Friesen, 1975), is elaborated from the anatomic movements of the face.
The most used method for emotion attribution are projective measures where children are presented hypothetical situations described by narrative stories, scenarios or cartoons. Children, individually or sometimes collectively, are requested to explain their beliefs regarding what they might feel or think if they were the story main character by means of an interview or a questionnaire (Nunner-Wringler and Sodian, 1988; Crick and Dodge, 1994; Saarni, 1999). Related methods are those involving problems in understanding mental states and emotions (Theory of mind paradigm) (Happé and Frith 1996; Sutton, Smith and Swettenham, 1999b).
An important area of research for bully/victim relationships is that related to self- conscious emotions, often defined "moral emotion". These emotions require that an objective self has developed; that is, children can refer to themselves and have conscious awareness of themselves as distinct from others . The cognitive appraisal involved includes: a) recognition that there are standards to be met; b) evaluation of the self's performance relative to these standards; and c) attribution of responsibility to the self upon success or failure in meeting the standard. These kinds of emotions comprise pride, shame, guilt, empathy and envy.
One of the main contributions to understanding possible differences between shame and guilt is Lewis's (1971) reconceptualization. According to this author, shame and guilt differ in focus on self versus behaviour. The experience of shame is directly about the self, which is the focus of evaluation. In guilt, the self is not the central object of negative evaluation, but rather the thing done or undone is the focus. In guilt the self is negatively evaluated in connection with something but is not itself the focus of the experience. According to Lewis (1971), shame is an acutely painful emotion that is typically accompanied by a sense of shrinking or of "being small" and a sense of worthlessness and powerlessness. Guilt is generally a less painful and devastating experience than shame because its primary concern is with a particular behavior, something apart from self. Guilt involves a sense of tension, remorse and regret over the bad thing done.
Shame often motivates an avoidance response, a desire to flee from the shame inducing situation, to disappear. Alternatively, ashamed individuals frequently become angry and blame others for the shame-inducing event (Lewis,1971; Tagney et al., 1992).
Guilt appears to motivate us in a more moral direction. Guilt keeps people constructively engaged in interpersonal situations. Numerous studies have shown a special link between guilt and empathy (Hoffman, 1998; Baumesteir , Stillwell and Heatherton, 1994).
The functionalist approach gives relevant insights to understand the role of self conscious emotions for social adjustment. Saarni et al. (1998) argue that any emotion can be adaptive or not, depending on circumstances. It is not negativity but rather ineffective emotion regulation that makes an emotion dysfunctional. Emotion disregulation may occur when individuals do not have access to an emotion that is typical and appropriate for the situation or when one emotion becomes dominant to the exclusion of others. Therefore the lack of guilt and shame as well as their overdominance are both maladaptive.
There is a strong support about possible relations between pervasive feelings of shame and depression. Ferguson and Stegge (1998) speak also of a possible relation between ambiguous situations eliciting guilt or shame and maladaptive outcomes. On the same line Bybee and Quiles (1998) elaborated a theory where two different variants of guilt exist: predispositional guilt = personality proclivity for experiencing guilt in response to specific, circumscribed situations and chronic guilt = an ongoing condition of guiltness, regret and remorse unattached to an immediate precipitating event.
Guilt is related to less aggression
Some studies found that children, adolescents and young adults who were higher on
guilt are less aggressive and less likely to engage in acting out behaviour or showing
externalising symptoms (Quiles and Bybee, 1997). Others' reports (teachers, peers,
roommates) as well as self reports are all in agreement. Guilt may inhibit disobedient,
unruly and defiant behaviour in the class.
Alternatively, children who repeatedly commit acts of aggression may become inured to this type of behaviour, exhibiting dulled emotional responses. This seems the case of bullies who generally don't feel guilty for their own acts but, on the contrary are pleased with these and find difficulty in interpreting victims' feelings (Olweus, 1993).
Guilt and empathy
Some authors found a relation between guilt and empathy. Both emotions request an
internalization of responsibility. The relation between the two emotions has been
conceptualised by Hoffman (1998) in the following terms: 1) guilt focuses on a
specific behaviour, implies the ability to understand other's sufferings and emotions
and a need for repair; 2) a sense of internal responsibility is relevant for both
emotions. Specifically, two abilities are necessary to experience the emotion of
guilt: awareness of being the cause of other's sufferings, and answering to this
emotion with empathy.
Shame
A deep sense of shame can represent one mechanism to explain the history
of peer abuse, rejection and silent suffering in the victims. There seems to be a link
between shame and negative attribution to self often reported for victims. Silver and
Wortman (1980) and Graham and Juvonen (1998) found that victims feel responsible
for what happened to them, this can increase the sense of helplessness and the sense
of social inadequacy. Furthermore, victims were reported to dissimulate feelings of
shame and pain using an unexpressive face (Lewis, 1992; Berdondini and Dondi,
1999).
Pride
An emotion which has an opposite value as compared to guilt is that of pride (Lewis,
1992). Pride can be elicited by meeting a challenging moral standard, by fulfilling
moral or personal expectations in spite of circumstances which make it difficult to do
so. For children, pride is most clearly felt when they can share with someone else
what they have accomplished (Saarni, 1999). A feature of bullying is that it often takes place in the context of a group of peers and most children in the group are aware of it (Salmivalli, et al. 1996a). Within this context a motivation factor can be related to dominance and to reaching a relevant status in the peer group (Boulton and Underwood, 1992; Olweus, 1993) Therefore, bullies can experience pride everytime they get attention and respect from others or simply when they make other children laugh.
Measures of guilt and Shame in Children:
A useful description of different measures on self conscious emotions can be found in
the chapter by Ferguson and Stegge (1998).
Self report scenario-based measures
(Tagney et al. 1995) "TOSCA-C" Test of Social Conscious Affect for Children:
This is intended for use with 8-12 years old children. Children read 10 negative and 5
positive scenarios. For each scenario children rate on a 5 point scale the extent to
which they would think or feel in a way that reflect precoded representation of
guilt, shame and externalisation.
(Stegge and Ferguson, 1994) SCEMAS This consists of the 8 situations of the C- CARS plus additional five situations rated not only for precoded guilt, shame and externalisation but also for pride. The new situations are more ambiguous than the original eight and seem to represent the maladaptive components of both guilt and shame.
Semiprojective measures
Children interpretations of interpersonal distress and conflict CIIDC
Zahn-Waxler, Kochanska, Krupnick and Mayfield (1988) developed the narrative
measure CIIDC, in which children respond to a series of questions after viewing each
of four photographs depicting ambiguous situations.
Hoffman (1976) semiprojective measure
This consists of short stories describing wrongful acts. Hoffman used them with
7th graders and 5th graders. Children responded to 2 stories in which they
pretended to be the protagonist. In one story the protagonist cheats in the swimming
pool, and in the second fails to help a lost child who later is killed by a car. Scores for
the maximum and terminal guilt expressed by children can be calculated.
An other report instrument is: My child (Kochanska 1993) - A parents' scale on children's guilt.
Emotion regulation is a central aspect in recent theories on emotional development. It can be viewed as a core feature of normative and individual development and of central significance to the development of relationships and quality of socioemotional functioning. Consequently, researchers have become more interested at inspecting the processes involved in emotion and its regulation among normal as well as among at- risk (Cicchetti, Ganiban and Barnett, 1991) and clinically diagnosed children (Zahn- Waxler, Cole, Welsh & Fox, 1995; Eisenberg et al., 1997).
In general, there is growing evidence that emotional regulatory processes contribute to a sense of well-being, a sense of self-efficacy and a sense of connectedness to others (Saarni, Mumme & Campos, 1998), namely through enhancing problem-solving strategies and an appropriate appraisal of context, critically components of social competence and personal adjustment.
Definition of emotion regulation.
When discussing the emotional regulation construct, some authors (Cicchetti et al.
1991; Walden & Smith , 1997) claim that the emotion system is a dynamic
organisation of physiological, cognitive and behavioural responses that occur inside
and outside conscious awareness, and that the regulation of such a system may
involve adaptations in one or more of its areas (i.e., physiological, cognitive and
behavioural). Instead of circumscribing the regulation process to any of the
designated areas, it is hypothesised that emotion regulation takes places in three
sequential levels: input regulation (sensory receptors), central regulation (information
processing), and output regulation (response selection). Applying a metaphor of the
systems approach, it can be anticipated that any failure of regulation can result in
communication problems between systems. Like in a chain of interrelated events,
emotion regulation can be conceptualised as a sequence of regulatory processes.
In a recent review of literature, Walden & Smith (1997) distinguish four main areas of emotional regulation:
Regulation involves appraisals of one's resources for coping with an encounter and includes both problem-focused strategies and emotion-focused strategies (Fabes and Eisenberg, 1992). These mechanisms modulate the degree of emotional arousal and change the situation causing it. Coping behaviour that is specifically emotion-focused involves cognitive activity that influences either the deployment of attention (e.g., avoidance) or the meaning of an encounter, such as when a person engages in a denial (Walden & Smith, 1997). A problem-focused approach involves cognitive restructuring of the situation.
The construction of emotion regulation seems to be particularly relevant for the victims of bullying. Silvern and Wortman (1980) suggested that victims do not react to others' attacks because they use an emotion-focused coping to change the meaning of a stress situation, so that it seems less threatening and consequently elicits less distress; for example, they place the blame for the teasing in themselves so that they can feel some control over the situation. Some authors note that probability, intensity and duration of aggression will increase if victims are seen as provocative, have a history of acquiescing and show signs of pain and suffering (Patterson, Littman and Bricker, 1967; Perry, Williard and Perry, 1990). Other studies point out how victims can react in a counteraggressive way or show nonchalance (Salmivalli, Karhunen and Lagerspetz, 1996b). Using MAX by Izard to codify children's facial expressions, Berdondini and Dondi (1999) found that victims often showed a "blank face" during a videotaped interview. This finding confirmed the hypothesis that victims may have difficulties in overtly expressing and communicating their own emotions. Other studies about prolonged victimization (such as sexual abuse, trauma, etc.) show that such victims tend to verbally deny the episodes themselves and their actual feelings, with a clear tendency towards emotional dissociation, (Herman, 1992; Shengold, 1989).
One common way to assess outcome behaviours of emotional regulation is to examine coping strategies in response to an emotionally arousing event or situation (direct observational data); (Underwood and Schockner, 1999; Eisenberg et al. 1997; Karp, Mahady-Wilton and Craig, 1999).
Indirect methods usually try to understand the role of both problem-focused and emotion-focused strategies in regulating children's responses. Some of them are:In relation to our aims, which were 1) investigating different dimensions of emotional competence related to bully/victims problems; 2) comparing different methodologies to study the problem, the methods we presented have advantages and disadvantages:
Since few studies have addressed the relation of emotion and bully/victim status in children, there is a need for context specific measures. In this perspective, we decided to start a preliminary study by using a more qualitative and contextual instrument like the "Narrative script cartoon" or SCAN developed by the TMR working party (see this website), focusing the interview on emotions attribution and coping strategies.
In particular our contribution as a group on emotions and bullying was related to improve the section on self-conscious emotions in the script of SCAN bullying. The complete section on emotion is reported below.
GUILT: One child told me that the story made him/her feeling guilty.
SHAME: One child told me that she felt ashamed for what happened in this story.
INDIFFERENCE: One child told me that he/she would feel indifferent for what has been done.
PRIDE: One child told me that he/she would feel great for what has been done.
Specify on prompting about different bullying situations, such as teasing, social exclusion, physical aggression, threatening.