Scholarly article on topic 'Antecedents and Consequences of Satisfaction and Guilt Following Ingroup Aggression'

Antecedents and Consequences of Satisfaction and Guilt Following Ingroup Aggression Academic research paper on "Psychology"

0
0
Share paper
OECD Field of science
Keywords
{""}

Academic research paper on topic "Antecedents and Consequences of Satisfaction and Guilt Following Ingroup Aggression"

Group Processes äf -p* Intergroup Relations 1

2007 Vol 10(2) 223-237 j

Antecedents and Consequences of Satisfaction and Guilt Following Ingroup Aggression

Angela T. Maitner and Diane M. Mackie

University of California, Santa Barbara

Eliot R. Smith

Indiana University

Three studies investigated the role of intergroup satisfaction in intergroup conflict. After reading about real acts of aggression committed by an ingroup, participants reported how those actions made them feel and how much they would support similar aggression in the future. In all three studies, experiencing intergroup satisfaction increased support for similar aggression, whereas experiencing intergroup guilt decreased support for similar aggression. Study 2 showed that ingroup identification increased justification appraisals, which increased satisfaction and decreased guilt, and thus increased support for future aggression. Study 3 provided an experimental test of the model: when justification appraisals were manipulated, emotion and support for further aggression changed accordingly. These findings demonstrate conditions under which intergroup satisfaction can facilitate and sustain intergroup conflict.

keywords guilt, identification, intergroup emotion, satisfaction

In 1973, the United States covertly supported a military led coup against the democratically elected socialist government of President Salvador Allende in Chile. The coup was a bloody one, but CIA backing allowed General Augusto Pinochet to begin a 17-year dictatorship that left more than 3000 Chileans dead or missing. Some commentators have suggested that despite the brutal repression that kept Pinochet in power, satisfaction with the outcome of the event at the time encouraged US intrusion throughout Central and South America for the next two decades (Bancroft-Hinchley, 2001). In this article, we explore how such satisfaction with ingroup aggression can foster support for continued conflict.

The roots of intergroup conflict are complex, residing in competition both for material and psychological gain and embedded in patterns of conflict and entrenched mutual views of antagonists as evil and untrustworthy (Brewer & Brown, 1998; Sherif, 1958; Tajfel & Turner, 1986). We are concerned with the role that emotions might play in exacerbating intergroup conflict once an aggressive act has been committed.

Author's note

Address correspondence to Angela Maitner, Department of Psychology, University of California, Santa Barbara, Santa Barbara, CA 93106, USA [email: maitner@psych.ucsb.edu]

Copyright © 2007 SAGE Publications (Los Angeles, London, New Delhi and Singapore) 10:2; 223-237; DOI: 10.1177/1368430207075154

Although emotion and its regulatory role on behavior have long been considered individually based phenomena, recent theories have claimed emotion as also a group level phenomenon, with implications for intergroup behavior (Alexander, Brewer, & Hermann, 1999; Fiske, Cuddy, Glick, & Xu, 2002; Mackie, Devos, & Smith, 2000; Neuberg & Cottrell, 2002). Intergroup Emotions Theory (IET; Mackie, et al., 2000; Mackie & Smith, 1998; Smith, 1993), for example, postulates that when people identify with a group, they experience emotion in response to events that either help or harm group goals, regardless of whether those events are performed by or targeted at them personally. Specifically, individuals will make group-level appraisals which incite group-level emotions and group-level action tendencies. Thus the experience of these group emotions is fundamental to intergroup relations because they translate interpretations of what is good or bad for the ingroup into group-based action tendencies. Evidence now suggests that identification with a group is key in the experience of group-based emotions (Mackie, Silver, & Smith,

2004), that people experience emotions in response to group outcomes even when those outcomes are not personally relevant (Mackie & Smith, 2002), and that group members' experience of intergroup emotion mediates their desires to confront or avoid the outgroup (Dumont, Yzerbyt, Wigboldus, & Gordyn, 2003; Mackie et al., 2000; Silver, Miller, Mackie, & Smith, 2001; Yzerbyt, Dumont, Wigboldus, & Gordijn, 2003).

Research is increasingly demonstrating a role for emotion in intergroup action tendencies. Collective guilt, for example, has been attributed a role notjust in promoting more positive intergroup attitudes (Powell, Branscombe, & Schmitt,

2005) but in actually inhibiting the desire for aggressive intergroup behavior (Branscombe & Doosje, 2004; Branscombe, Doosje, & McGarty, 2002). Specifically, collective guilt is related to a desire to make apologies or reparations to the offended group (Doosje, Branscombe, Spears, & Manstead, 1998; Iyer, Leach, & Crosby, 2003; Mallett & Swim, 2004; Schmitt, Branscombe, & Brehm, 2004; Swim & Miller, 1999; see also Schmitt, Behner, Montada, Muller, & Muller-Fohrbrodt, 2000). There is even evidence that the more

collective guilt individuals feel for their in-groups' wrongdoings, the more likely they are to forgive the outgroup its own indiscretions (Hewstone et al., 2004). This research demonstrates the importance of collective guilt in regulating intergroup attitudes and behavior (but see also Iyer et al., 2003, for limitations). With its tendency to correct inappropriate behavior, collective guilt clearly plays a pivotal role in alleviating group conflict.

Because of its role in inhibiting intergroup aggression, researchers have investigated the conditions under which individuals will experience collective guilt. Although personal responsibility for an immoral behavior has been found to be central to the experience of guilt on an interpersonal level (Weiner, 1995), research has now shown that people can experience intergroup guilt even when the individual has no direct responsibility for the group's action (e.g. Branscombe et al., 2002). For example, Australians feel collective guilt for the harsh treatment of Aborigines, Dutch feel collective guilt for the colonization of Indonesia, and white Americans feel collective guilt for the historically harsh treatment of African Americans (Branscombe et al., 2002; Iyer et al., 2003; Powell et al., 2005; Swim & Miller, 1999). However, both self-categorization in terms of a group membership and an acknowledgement of ingroup responsibility are necessary precursors to collective guilt (Branscombe et al., 2002). Apparently these conditions are not always easily met. As Branscombe and Miron conclude, 'collective guilt appears to be a fragile emotional response that can be rather easily disrupted' (2004, p. 331).

What happens when groups do not feel badly about acts of intergroup aggression? Especially when aggression is goal directed, when the ingroup is helped in some intended way, group members may instead feel intergroup satisfaction following intergroup aggression. At the interpersonal level, satisfaction is defined as a positive emotional response to obtaining some desired goal or event (Ortony, Clore, & Collins, 1988). Given that aggression is often goal-directed, perceived advancement of those goals through such behavior elicits satisfaction,

and such satisfaction increases the likelihood of further aggression (Berkowitz, 1990). We assume satisfaction has a similar function at the intergroup level. Identification with a group translates appraisals from concerns about how events impact the self personally to concerns about whether events promote or hurt the group. Thus intergroup aggression—aggression carried out by the group or in the name of the group—may produce intergroup satisfaction if it is seen as advancing the group's position. Such satisfaction in turn may increase the desire for further similar behavior. Elsewhere we have shown that successfully executing a desired intergroup behavior leads to satisfaction, accompanied by a strengthened desire to engage in similar behavior in the future (Maitner, Mackie, & Smith, 2006). Taking a regulatory perspective on intergroup emotion thus suggests that inter-group satisfaction might act as an emotional reinforcer of aggressive intergroup behavior, thereby promoting support for continued conflict. Accordingly, we hypothesized that any satisfaction group members felt in response to ingroup aggression would predict their support for future aggression. As collective guilt is related to a desire to apologize or make reparations, we expected intergroup guilt to predict reduced support for continued aggressive action. We tested these hypotheses in three studies.

Study 1

Our first study was designed to demonstrate a relation between any satisfaction and guilt experienced following ingroup aggression and the extent to which group members supported further aggression. Paralleling the interpersonal literature, we expected satisfaction and guilt to be conceptually independent (Ortony et al., 1988). Depending on how they appraised the events, we expected group members to react with feelings of satisfaction and/or guilt after learning or being reminded of negative ingroup behavior. Confirming previous research, we expected these emotions to be elicited even if individual group members were not personally responsible for the group's action, as long as they recognized the aggressing group as an

ingroup. We further expected that satisfaction for group actions would relate to support for similar aggression in the future, whereas feeling guilty for group actions would undermine support for similar action in the future. To test these ideas, we had participants read short descriptions of ingroup aggression and report their emotional reactions and support for future group-level actions.

Method

Participants Participants were 33 students at the University of California, Santa Barbara (UCSB). Only students who identified themselves as American in pretesting were invited to participate. Students received course credit for their participation. They participated up to six at a time, individually responding to questionnaires in cubicle spaces.

Procedure

Description of ingroup aggression Written instructions informed participants that the study was concerned with people's reactions to group action. Participants then read one of two minimal, historically accurate, specific, descriptions of aggressive actions taken by the United States against another non-specified nation. Participants read either: 'The United States used missiles to destroy a factory in an African country; many people died'; or 'The United States supported a coup in a South American country that resulted in many deaths'.1

Dependent variables Participants were then asked to indicate how the action made them feel. Participants used 7-point scales (1 = not at all to 7 = very) to indicate how much they were feeling each of six emotion terms. Terms were intended to assess feelings of satisfaction (satisfied, pleased, content; a = .86) or guilt (guilty, regretful, sorry; a = .81).

Participants next reported their support for future aggression by responding to the two questions: 'In the future, do you think the United States should engage in similar action?' and 'In the future, do you think the United States should avoid similar action?' Participants again used 7-point scales (1 = definitely

not; 7 = definitely yes) to indicate their support. These two items correlated at r = -.76 (p < .001) and were combined to reflect how much the participant supported similar ingroup aggression in the future. When participants completed these items, they were debriefed and thanked for their participation.

Results and discussion

No significant differences in responses were found between the two descriptions of ingroup aggression. All further analyses are collapsed across stimulus replication.

Intergroup emotions and support for intergroup action Overall, participants reported feeling minimal satisfaction (M = 1.81, SD = 0.96) and moderate guilt (M = 4.43, SD = 1.46) in response to the aggressive ingroup action. A repeated-measures analysis of variance (ANOVA) confirmed that participants felt more guilt than satisfaction for the negative ingroup action (F(1, 32) = 69.18, p < .001). The fact that they played no part in this historical action, yet experienced guilt and satisfaction in response to it, confirmed our assumption that participants who identified themselves as Americans considered the United States an ingroup. As expected, the experience of satisfaction and guilt were uncorrelated (r = -.085, ns). Overall, our participants reported being moderately opposed to similar aggression in the future (M = 3.00, SD = 1.49).

Relation between intergroup emotion and support for intergroup action Our hypotheses focused primarily on the relations between intergroup emotions and support for future aggression. We predicted that exposure to information about negative ingroup behavior would result in feelings of satisfaction and/or guilt, and that these emotions would in turn affect group members' behavioral intentions. We correlated participants' emotion scores with their support for intergroup action. Both satisfaction and guilt predicted behavioral intentions as expected. Although participants experienced only minimal satisfaction, the satisfaction they did experience was strongly and significantly related to support for future aggression (r = .62,

p < .001).2 Guilt, on the other hand, was related to reduced support for similar aggressive action (r = -.36, p = .040).

When satisfaction and guilt were simultaneously entered into a regression equation predicting desire for similar intergroup aggression, they predicted nearly 48% of the variance in support for future aggression (F(2, 30) = 13.58, p < .001). Satisfaction uniquely accounted for 35% of this variance while guilt uniquely accounted for 10%.3 Although participants reported very little satisfaction for their groups' aggressive action, the satisfaction they did report significantly predicted support for future aggression. Although group members reported more guilt at the ingroup's action overall, this guilt was less strongly predictive of support for future intergroup aggression.

Study 2

Study 1 showed that the satisfaction people experienced following intergroup aggression was predictive of increased support for further aggression, whereas the guilt they felt was predictive of decreased support. Given the important role satisfaction plays in facilitating and guilt in inhibiting intergroup aggression, we turned to the question of how each of these emotions might arise. According to IET, distinctive intergroup appraisals give rise to distinct emotions. In this regard, appraisals of justifiability seem most relevant to the intergroup emotions of concern. First, perceptions of justifiability clearly minimize the experience of intergroup guilt. For instance, several researchers have shown that perceptions of unfairness, illegitimacy, or injustice evoke collective guilt (Branscombe & Miron, 2004; Mallett & Swim, 2004), whereas thoughts about exonerating conditions minimize collective guilt (Roccas, Klar, & Liviatan, 2004). Second, distributive outcomes that are perceived as just or fair lead to relatively high levels of satisfaction (Adams, 1965; Crosby, 1976; Walster, Walster, & Berscheid, 1978). Thus people appear to experience collective guilt to the extent that they perceive behavior as unjustified and satisfaction to the extent that they view behavior asjustified. We thus assessed appraisals

of justification of intergroup aggression as precursors of intergroup satisfaction and guilt. These emotions were expected in turn to influence support for ingroup action, with satisfaction promoting support for further intergroup aggression and guilt depressing it, as found in Study 1.

Given the important role that identification with a group has been shown to play in the experience of intergroup emotion in general (Silver et al., 2001), and in the experience of intergroup guilt specifically (Doosje et al., 1998), Study 2 also investigated how identification with the aggressive ingroup might influence the relations between emotion and support for ingroup aggression found in Study 1. IET predicts that identification with the group influences the appraisals people make in intergroup contexts. These appraisals should in turn influence intergroup emotions (Mackie et al., 2004; Silver et al., 2001; Yzerbyt et al., 2003). In this study, we expected identification to influence appraisals of the justifiability of ingroup aggression, which would in turn influence intergroup satisfaction and guilt and thus support for future aggression.

We expected identification to increase perceptions of the justifiability of ingroup aggression because when identity is based on group membership, the desire for a positive self-view results in group-serving explanations for ingroup actions (Branscombe & Doosje, 2004; Branscombe & Miron, 2004; Doosje & Branscombe, 2003; Mummendey, Klink, Mielke, Wenzel, & Blanz, 1999). Highly identified individuals are least likely to accept threatening information about the group's past (Doosje et al., 1998), are most able to minimize perception of harm done by the ingroup, and are most able to focus on out-group behaviors or qualities thatjustify the aggression (Branscombe, 2004; Branscombe & Miron, 2004). Thus, high identifiers appear more likely to spontaneously justify their ingroup's aggressive behavior.

This pattern of responses creates conditions that allow us to predict a positive relation between group identification and support for future aggression (since identification leads to justification which leads to satisfaction which

leads to support for further aggression). Such a prediction is consistent with other evidence that highly identified group members are often those most likely to show ingroup bias (Castano, Yzerbyt, Paladino, & Sacchi, 2002; Verkuyten & Nekuee, 1999), especially under conditions of social identity threat (Branscombe, N'gbala, Kobrynowicz, & Wann, 1997; Branscombe & Wann, 1994). We expected this relation between identification and support for future aggression to be mediated by intergroup appraisals and intergroup emotions.

Thus, in Study 2, we hypothesized that: (1) identification would influence support for continued ingroup aggression via intergroup justification appraisals, and (2) justification appraisals would influence support for intergroup action via increasing satisfaction and decreasing guilt.

Method

Participants and design Participants were 68 American students at UCSB. Students were either paid or received course credit for their participation. Participants responded to the dependent variables in individual cubicle spaces after completing an unrelated computer task.

Procedure

Identification assessment American identification was assessed four to eight weeks prior to data collection using the identity subscale of Luhtanen and Crocker's (1992) Collective Self-Esteem Scale. This scale was modified so that the four items referred to the group Americans, rather than to social groups in general (i.e. The American group is an important reflection of who I am). Participants indicated their level of agreement with each statement using a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree).

Description of ingroup aggression Participants read one of the three historically accurate descriptions of negative American actions. The new description read, 'The United States conducted air raids on an Asian country, many people died'. This item was pretested at the same time as the others and was rated as similarly negative (M = 5.47, SD = 1.07).

Dependent variables Participants responded to the item, 'To what extent do you think the United States' action was justified?' on a 7-point scale from 1 (very unjustified) to 7 (very justified).

Participants next responded in the same way to the same emotion terms, and support for future behavior questions as in Study 1. As in Study 1, the satisfaction items were internally consistent (a = .90). Despite their consistency in Study 1, internal consistency for the guilt items was much lower (a = .66). Given their conceptual derivation, their high reliability in Study 1, and the fact that the means produced in this study were similar to those in Study 1, we maintained the same composite score. After completing these measures, participants were debriefed and thanked for their participation.

Results and discussion

Identification and justification Reported identification as Americans yielded a symmetric distribution of scores with average identification at the midpoint of the scale (M = 3.92, SD = 1.37). No systematic differences were found across stimulus replication (F(2, 65) = .83, ns), indicating no failures of randomization.

Participants reported that the US actions were moderatelyjustified (M = 3.84, SD = 1.22). Neither stimulus replication nor its interaction effects (with American identification) exerted any impact on justification scores.

Intergroup emotions and support for intergroup action Satisfaction and guilt were again uncorrelated; (r = -.16, ns). A repeated measures ANOVA revealed that participants reported feeling more guilt (M = 3.86, SD = 1.29) than satisfaction (M = 2.13, SD = 1.32), (F (1, 65) = 49.46, p < .001). Overall, participants reported being moderately unsupportive of similar aggression in the future (M = 3.29, SD = 1.49).

Relations among identification, justification, emotion, and support for intergroup action We

tested our hypotheses using Baron and Kenny's (1986) mediated regression procedure. We predicted that (1) identification would influence support for intergroup action via justification,

and (2)justification would influence support for intergroup action via satisfaction and guilt.

First we examinedjustification as a mediator of the relation between identification and support for intergroup action. We found the hypothesized direct relation between identification and support for similar aggression. As expected, the more participants identified with the group, the more likely they were to support similar aggression in the future (fi = .31, p = .011). Second, we found that identification predicted justification as expected. The more strongly participants identified with the group, the more they spontaneously justified their groups' aggression (fi= .28, p = .022). Finally, when identification and justification were simultaneously entered into a regression analysis, identification no longer predicted support for future aggression (fi= .16, ns; the reduction in the fivalue was significant, z = 2.14, p = .033). However,justification still strongly predicted support for future aggression (fi = .53, p < .001). Thus, the relation between identification and support for future aggression was fully mediated by justification (see Figure 1a).

Next we investigated whether intergroup emotions mediated the relation between justification and support for intergroup action, first investigating satisfaction as a mediator. Consistent with the previous analysis, justification strongly predicted support for future aggression. The more participants saw intergroup aggression asjustified, the more they supported repeating it (fi = .58, p < .001). Justification was also strongly related to intergroup satisfaction. The more participants appraised acts of intergroup aggression as justified, the more satisfied they were (fi= .39, p = .001). When both justification and satisfaction were entered into a simultaneous equation predicting support for future aggression, the relation between justification and support for future aggression was significantly reduced, although not completely eliminated (fi = .39, p < .001, the reduction in the fi value was significant, z = 2.83, p = .005). Satisfaction also strongly predicted support for future aggression (fi = .47, p < .001). Thus intergroup satisfaction significantly, but not completely, mediated the relation between

Identification

| Justification |

P = .16 (p = .31*)

P = .53**

Support for future aggression

Satisfaction

p = ,39*J/

p = .47**

p = .39** (p = .58**)

Support for future aggression

| Justification \

*p< .05 **p< .01

Figure 1(a). Model showing the relation between identification and support for future aggression, mediated by justification. (b) Model showing the relation between justification and support for future aggression, mediated by satisfaction.

the appraisal of the aggression as justified and support for future aggression (see Figure 1b).

Justification was only marginally related to intergroup guilt. The more participants perceived acts of intergroup aggression as justified, the less intergroup guilt they experienced (¡= -.22, p = .068). The inclusion of guilt in a regression equation predicting support for future aggression failed to significantly reduce the relation between justification and support for future aggression (3 = .54, p < .001, the reduction in the ¡3 value was nonsignificant, z = 1.32, p = .186). However, guilt was marginally related to reduced support for future aggression (¡= -.19, p = .062). Because our measure of guilt was less reliable than our measure of satisfaction, we do not know whether our failure to find mediation is a consequence of poor reliability, or because of true lack of relation. However, since Study 1 also found satisfaction to be more closely related to support for future aggression than guilt, it is possible that guilt is simply less related to support for future aggression than is satisfaction.

Although correlational, our results support our hypothesized model:justification was important in increasing satisfaction and marginally decreasing guilt, which, in turn, were predictive of support for future aggression. Also, identification with the group influenced the extent to which participants justified ingroup aggression.

Study 3

Study 2 provided further correlational support for our model of the way in which intergroup satisfaction and guilt can facilitate and inhibit support for intergroup aggression. The extent to which individuals justified ingroup aggression influenced their emotions, which predicted support for future aggression. Study 3 was designed to provide an experimental test of the causal relation among intergroup appraisals, emotion, and support for future ingroup action. We manipulated the appraisal context of ingroup aggression with the intent of changing participants' emotions and subsequent support for future aggression.

Branscombe and Miron (2004) point out that individuals can legitimize their ingroup's immoral behavior in multiple, sometimes simultaneous, ways. As our concern in this experiment was to accomplish a strong manipulation ofjustification (one that might overcome any natural proclivities among those differentially identified with the group), we used a multifaceted manipulation that described the ingroup aggression as either carefully considered, reached by consensus with allies, and successful or as carelessly executed, reached without consultation with other countries, and unsuccessful. We then measured partici-pants'justifications, emotions, and support for future aggression. We expected aggressive actions that were described as carefully thought out, consensual, and successful would be perceived as morejustified than aggressive actions described as carelessly executed, unilateral, and unsuccessful. We expectedjustified acts would provoke more satisfaction and less guilt than unjustified actions. In turn, satisfaction was expected to increase support, and guilt to decrease support, for similar aggression in the future.

Method

Participants and design Participants were 60 students at UCSB run under identical conditions to those reported in the previous two studies. Only American students were invited to participate. Participants were randomly assigned to the cells of a 2 (scenario replication) x 2 (justification condition) between-subjects design.

Procedure

Description of ingroup aggression Instructions and stimuli were identical to those used in Study 1.

Justification manipulation Participants in the justified condition were told: 'The United States took this action after careful consideration of alternatives and consultation with allies. This action succeeded in achieving the United States' objectives'. Participants in the unjustified condition were told: 'The United States took this action without considering all possible alternatives, and having failed to consult with allies. This action failed to achieve the United States' objectives'. This manipulation included multiple types of information—any one of which might have been crucial to the effects obtained—in an attempt to ensure participants' appraisals were influenced.

Dependent variables Participants next responded to the samejustification, emotion, and support for future behavior questions as in Study 2. Internal consistency was sufficient for the two emotion scales (satisfaction, a = .93; guilt, a = .74). After completing these measures, participants were fully debriefed, thanked, and dismissed.

Results and discussion

Intergroup appraisals We subjected participants' justification scores to a 2 (justification condition) x 2 (stimulus replication) ANOVA. Analysis revealed a significant effect of justification condition (F (1, 56) = 8.83, p = .004). As expected participants in the justified condition perceived ingroup aggression as significantly more justified (M = 4.10, SD = 1.21) than participants in the unjustified condition (M = 3.20, SD = 1.13). These results indicate that we successfully influenced participants' justification appraisals as intended.

Intergroup emotions Perhaps because in this study the experimental manipulation created justification differences, guilt and satisfaction were negatively correlated (r = -.27, p = .040).

A repeated-measures ANOVA revealed that again, more guilt (M = 4.51, SD = 1.23) than satisfaction (M = 2.32, SD = 1.22) was experienced

overall (F(1, 57) = 98.73, p < .001). However, we also found the predicted emotion by appraisal condition interaction (F(1, 56) = 23.42, p < .001). Participants reported significantly more satisfaction (M = 2.97) in thejustified condition than in the unjustified condition (M = 1.65, p < .05), and less guilt (M = 4.12) in thejustified condition than in the unjustified condition (M = 4.89, p < .05) indicating that the manipulation influenced emotions as intended.

Support for intergroup action Analysis revealed a significant effect ofjustification condition on support for future aggression (F(1, 56) = 16.10, p < .001). Participants in the justified condition were significantly more (although not highly) likely to want to engage in similar aggression in the future (M = 3.62) than participants in the unjustified condition (M = 2.40).

Relations among condition, emotion, and support for intergroup action To assess the relations among the key variables we conducted mediational analyses. We expected condition (dummy coded; unjustified condition = 0, justified condition = 1) to influence emotions, which, in turn would influence support for future aggression. Even though participants reported low amounts of satisfaction, we expected that satisfaction would account for the relation between justification and support for future aggression as in Study 2.

We first investigated satisfaction as a mediator. We found the hypothesized direct relation between justification condition and support for similar aggression. Specifically, participants in the justification condition more strongly supported repeating similar aggression in the future (fi= .46, p < .001). Second, we found that justification condition predicted satisfaction. Participants in the justification condition were more satisfied than participants in the unjustified condition (fi = .53, p < .001). Finally, when experienced satisfaction was included as a mediator, justification condition no longer predicted the desire to engage in similar action in the future (fi = .19, ns; the reduction in the fi value was significant, z = 3.38, p < .001). However, satisfaction still strongly predicted support

Satisfaction

P = .53**/

p = .56**

Justification condition p = .16 (p = .46**) ^ Support for future aggression

P = -.31*

) = -.26**

Justification condition p = .38** (p = .46**^ Support for future aggression

*p<. 05 **/><.01

Figure 2(a). Model showing the relation between justification condition and support for future aggression, mediated by satisfaction. (b) Model showing the relation between justification condition and support for future aggression, mediated by guilt.

for future aggression (fi = .55, p < .001). Thus, the relation between justification condition and support for future aggression was completely mediated by satisfaction. The justification condition increased support for similar behavior, and this was because such aggression was more satisfying (see Figure 2a).

Next we investigated guilt as a mediator. Again, we found that justification condition predicted guilt (fi= -.31, p = .015). However, when bothjusti-fication condition and guilt were entered into a regression equation bothjustification condition (fi= .38, p = .002) and guilt (fi= -.26, p = .030) remained significant predictors of support for future aggression. A Sobel test indicated there was no significant reduction in the predictive power of the justification condition (z = -1.67, ns). Thus, guilt did not mediate the relation between justification condition and a desire for future aggression, but remained a significant independent predictor (see Figure 2b).4

A regression model includingjustification condition, guilt, and satisfaction accounted for 48% of the variance in action tendencies (F(3, 56) = 18.62 p < .001). With all three predictors included in the model, satisfaction uniquely accounted for 19% of the variance while guilt accounted for 5%. Thus, not only was satisfaction the only mediator between justification condition and support for future aggression, but satisfaction

also uniquely accounted for more variance in the desire for future aggression than did guilt.

General discussion

The data from these three studies provide support for our claim that intergroup satisfaction is an emotion that can play an exacerbating role in intergroup conflict. Using correlational data, Studies 1 and 2 showed that the intergroup satisfaction people felt following ingroup aggression predicted their support for further intergroup aggression. Collective guilt played a consistent, although weaker, role in discouraging future aggression. Study 2 showed that justifications for intergroup aggression were associated with increased satisfaction and increased support for continued aggression, and that those who identified highly with their group were more likely tojustify its aggression. Using experimental data, Study 3 confirmed the role thatjustification played in increasing intergroup satisfaction and decreasing intergroup guilt, and the role that satisfaction played in increasing support for continued aggression. While generally confirming the role that intergroup guilt can play in inhibiting intergroup aggression, these three studies provide evidence that intergroup satisfaction can facilitate support for future aggression.

Although very little satisfaction was experienced by our participants overall, any experience of this emotion was strongly related to the desire to engage in similar aggression in the future (even though this too was quite minimal). We did not succeed in eliciting more than marginal satisfaction about intergroup aggression, nor more than minimal desire to continue intergroup aggression. What we did find, however, is that any satisfaction that is evoked by intergroup aggression will have a powerful effect on the desire to aggress again. This suggests, perhaps discouragingly, that any aspect of intergroup aggression that increases satisfaction will have a dramatic impact on support for continued ingroup aggression, a response likely to exacerbate cycles of intergroup violence. In this regard we were able to show that justification is one crucial precursor to the experience of satisfaction. In addition, we demonstrated

that those highly identified with a group may be more likely to support ingroup aggression because they are (as earlier research suggested) more willing or able to justify such action and thus experience more satisfaction about it.

Further research is clearly required to systematically assess conditions that increase justification, as well as alternative mechanisms that might evoke intergroup satisfaction. Our multifaceted approach to obtaining a strong manipulation of justification included information about the antecedents and the consequences of the ingroup's aggressive action. Although successful, further research might usefully tease apart the effects of consideration, consensus, and success of the aggressive actions, as well as explore some of the the other conditions previously noted as producing justification for immoral ingroup action (e.g. Branscombe & Miron, 2004). More importantly, just as researchers have begun to delineate the preconditions of experiencing collective guilt, more needs to be known about other precursors of satisfaction. Results from Studies 2 and 3 indicate that justification can make acts of ingroup aggression more satisfying. However, in Study 3, although satisfaction fully mediated the relation between justification condition and support for future aggression, justification appraisals did not fully mediate the relation between our manipulation and participants' reported intergroup satisfaction. Just as there are many ways to escape collective guilt, there may be many ways to experience intergroup satisfaction. For example, perceiving an aggressive act as attaining a desired outcome (either in terms of material resources or improvements in group status) may alone make acts of aggression more satisfying, in addition to or despite other appraisals made about the behavior. Our manipulation may have tapped into some of these other possibilities in addition to creating justification.

The exact mechanism that links satisfaction to support for future aggression is also currently unknown. In this study, we used satisfaction to refer to a cluster of emotions including pleasure and contentment. Satisfaction is not typically regarded as a basic emotion closely linked to a distinctive action tendency (Ekman, 1999).

Nevertheless, satisfaction and gratitude have been used to label emotional reactions to hoped for events (Ortony et al., 1988). Thus defined, satisfaction should be related to, but empirically distinguishable from, measures of desire for, support of, or intention to commit emotion-related behaviors. Our data suggest that this is the case. Although strongly predictive of the support measure, reported satisfaction was distinct from both general attitudes about US policies (see Note 3) and support for continued aggression. In Studies 1 and 2, significant variance in the support measure is independent of satisfaction, as shown by the fact that guilt negatively predicts support but is completely independent of satisfaction. In Study 2, the relation between justification appraisals and the support measure was only partially mediated by satisfaction, suggesting again that satisfaction and support are empirically distinguishable. As measured, satisfaction may act as a reward for succeeding and thus increase the likelihood of those behaviors being performed again ('if it feels good, do it again'). Alternatively, satisfaction may reinforce the notion that the ingroup has enough strength and status to aggress, which, in turn, may increase the likelihood of future aggression. Sachdev and Bourhis (1991) reported that members of high status and powerful groups were more discriminatory than subordinate low status group members. Thus such reinforcement of the group's position may make aggression more likely. Finally, much evidence suggests that people in positive states engage in heuristic processing of social information (Forgas, 2000). Satisfied participants may engage in heuristic decision making, using their positive affect as a cue to their support for similar action in the future. Understanding the mechanism through which satisfaction leads to a desire for future aggression may suggest new ways to undermine its relation with support for continued aggression.

To date, most empirical research investigating collective guilt about negative ingroup behavior has focused participants on past action or long standing inequality. Branscombe (2004) asks 'whether perceived responsibility for correcting or solving existing inequality also has potential to evoke guilt' (p. 330). Our data certainly suggest

that group members can feel guilt for more recent ingroup actions, although we did not assess perceived responsibility. Confirming earlier research, our data show collective guilt to be influenced by group identification and perceptions of actionjustifiability. Across studies, however, guilt had a weaker effect on dampening support for future aggression than satisfaction had on increasing it. Thus there may be differences in how people react to guilt elicited by past ingroup aggression and more current ingroup aggression, but this remains an empirical question.

Our findings confirm the benefit of approaches to emotion that maintain the independence of distinct emotions. As the appraisal conditions that are thought to elicit guilt and satisfaction are quite different, we had theoretical reasons to treat the two emotions as independent. This assumption received empirical support in Studies 1 and 2, where we had no control over the spontaneous appraisals that participants made. In Study 3, however, in which we attempted to experimentally manipulate appraisals, we apparently succeeded in creating conditions that were relevant to both emotions. Nevertheless, even when appraisal conditions increased satisfaction and decreased guilt, the value of retaining these emotions as distinct can be seen in the very different relations they had with other assessed variables. The justification appraisals that were made in Study 3 apparently affected satisfaction and guilt to a different extent as well as in opposite directions. Moreover, self-reported justification appraisals completely mediated the relation between our manipulation and guilt but only partially mediated that with satisfaction. Thus the manipulation seems to have affected guilt fully through its impact on justification, whereas it likely affected satisfaction by affecting both justification and other appraisals. Similarly guilt and satisfaction were differentially related to support for future aggression and differentially mediated the impact of justification on such support, all findings that speak to the value of considering them conceptually distinct.

In the quest to understand intergroup conflict, our results are limited by the fact that we assessed only self-reported support for ingroup

action, rather than actual support for such aggression. Whereas the links between emotional experiences and the readiness to engage in certain types of action are fairly well established, the link between emotions and actual behaviors is more tenuous (Roseman, Wiest, & Swartz, 1994). One reason for this is that actual behaviors are more constrained by situational factors than are impulses or intentions. Although group members may feel like confronting or aggressing against the outgroup, acting in line with these desires requires appropriate resources, no other pressing concerns, and so forth. On the other hand, affective responses have been shown to predict intergroup behaviors better than do some more cognitive assessments (Fiske, 1998) and normative pressures can sometimes facilitate rather than inhibit intergroup aggression (Insko et al., 1998). All else being equal though, the activation of a specific emotion and the action tendency or intention associated with it makes the execution of relevant behavior more likely.

Unfortunately, instances of ingroup bias and outgroup derogation are common in the literature (Tajfel & Turner, 1986; Turner, Brown, & Tajfel, 1979). If aggression against an outgroup is normal, then the ingroup may rarely be deemed morally responsible (a necessary precursor to guilt). Thus, it may be the case that guilt is rarely experienced in intergroup conflict situations without specific instructions to focus on the ingroup's lack ofjustification for its actions. Satisfaction, on the other hand, may be rife, especially if the goal-directedness of intergroup conflict is enough to elicit it. If undermining potential justifications for negative group behavior can inhibit this satisfaction, which in turn prevents future aggression, then preventing people from feeling good about the ingroup's action may be helpful in quelling conflict before it spirals out of control.

1. One-hundred and eighteen pretest students read several descriptions of intergroup aggression and made appraisals of the negativity of and of attributions of responsibility for the behavior (using 7-point scales), in addition to reporting

whether or not they recognized the events. Participants rated these two events as highly negative (M = 5.39, SD = .96, M = 5.72, SD = .96 for the South American and African action, respectively). No participants correctly recognized either event, and fewer than 10% of participants reported recognizing the event incorrectly.

2. Because our measure of satisfaction was moderately positively skewed (sk = 1.24) which may distort the magnitude of the relation between satisfaction and support for future aggression, we analyzed our data in several different ways. First, we looked at the correlation between satisfaction and support for future aggression when participants who reported no satisfaction (satisfaction = 1.0) were removed from analyses (increases the relation to

r = .70, p < .001). Second, when satisfaction was dichotomized into a variable differentiating participants who felt some level of satisfaction (satisfaction > 1) from those who were at floor (satisfaction = 1.0), the correlation was r = .40, p = .022. Regardless of skew, participants' level of satisfaction strongly predicted support for continued aggression. Satisfaction was similarly skewed in Study 2, but similar analyses yielded similar results.

3. To ensure that participants' reported satisfaction was not a simple proxy for their general attitude toward an aggressive foreign policy, we conducted an independent pilot study (n = 76) investigating the relation between a general attitude toward an aggressive foreign policy, satisfaction with specific instances of aggression, and support for specific similar aggression in the future. Results indicated that although general attitude predicted support for specific similar aggression in the future, satisfaction uniquely predicted an additional amount of variance in support for future aggression. Thus satisfaction and general attitude exerted independent effects on support for similar aggression and cannot be considered simple proxies for one another.

4. Because our manipulation ofjustification was multifaceted, we wanted to ensure that it was influencing emotion via justification appraisals as intended. Thus, we examined participants' self-reported justification scores as a mediator of the relation between condition and intergroup emotion. We first examined satisfaction as a dependent variable. Appraisal condition was

significantly related to reported intergroup satisfaction. Participants in the justified condition were more satisfied than participants in the unjustified condition (в = .53, p < .001). Additionally, participants in the justified condition believed the action was more justified than participants in the unjustified condition (в = .36, p = .004). Finally, when satisfaction was regressed on both condition and justification ratings, the strength of the relation between condition and satisfaction was significantly reduced, although not completely eliminated (в = .38, p = .001, the reduction in the в value was significant, z = 2.35, p = .019). Additionally, the more participants appraised the ingroup's action as justified, the more satisfied they were with it (в = .41, p < .001). Thus justification appraisals significantly mediated the relation between appraisal condition and intergroup satisfaction as intended. There remained, however, a significant independent link from condition to satisfaction, indicating that appraisals other than justification may have been affected by our multifaceted manipulation that independently influenced reported satisfaction.

We next examined guilt as a dependent variable. Again, appraisal condition was significantly related to reported intergroup guilt. Participants in the justified condition felt less guilt than participants in the unjustified condition (в = —.31, p = .015). When guilt was regressed on both appraisal condition and justification ratings, the strength of the relation between condition and guilt was virtually eliminated (в = .15, p = .221, the reduction in the в value was significant, z = 2.32, p = .020). Additionally, the more justified participants appraised the ingroup's action, the less guilty they felt about it (в= -.45, p < .001). Thus the relation between justification condition and intergroup guilt was completely determined by justification appraisals.

Acknowledgments

The research reported in this article was partly funded by PHS grant MH 63762 to Diane M. Mackie and Eliot R. Smith. Angela Maitner was supported by a National Science Foundation Graduate Research Fellowship. We would like to thank Victoria Springer, Shay Levy, and Nicole Jorenson for helping with the collection of data.

References

Adams, J. S. (1965). Inequity is social exchange. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2, pp. 267-299). New York: Academic Press.

Alexander, M. G., Brewer, M. B., & Hermann, R. K. (1999). Images and affect: A functional analysis of out-group stereotypes. Journal of Personality and Social Psychology, 77, 78-93.

Bancroft-Hinchley, T. (2001) 11th September, Allende and Berlusconi. Pravda On-Line, September 29.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

Berkowitz, L. (1990). On the formation and

regulation of anger and aggression: A cognitive-neoassociationistic analysis. American Psychologist, 45, 494-503.

Branscombe, N. R. (2004). A social psychological process perspective on collective guilt. In N. R. Branscombe & B. Doosje (Eds.), Collective guilt: International perspectives (pp. 320-334). New York: Cambridge University Press.

Branscombe, N. R., & Doosje, B. (2004).

International perspectives on the experience of collective guilt. In N. R. Branscombe & B. Doosje (Eds.), Collective guilt: International perspectives (pp. 3-15). New York: Cambridge University Press.

Branscombe, N. R., Doosje, B., & McGarty, C. (2002). Antecedents and consequences of collective guilt. In D. M. Mackie & E. R. Smith (Eds.), From prejudice to intergroup emotions: Differentiated reactions to social groups (pp. 49-66). Philadelphia: Psychology Press.

Branscombe, N. R., & Miron, A. M. (2004). Interpreting the ingroup's negative actions toward another group: Emotional reactions to appraised harm. In L. Z. Tiedens & C.W. Leach (Eds.), The social life of emotions (pp. 314-335). New York: Cambridge University Press.

Branscombe, N. R., N'gbala, A., Kobrynowicz, D., & Wann, D. L. (1997). Self and group protection concerns influence attributions but they are not determinants of counterfactual mutation focus. British Journal of Social Psychology, 36, 387-404.

Branscombe, N. R., & Wann, D. L. (1994).

Collective self-esteem consequences of outgroup derogation when a valued social identity is on

trial. European Journal of Social Psychology, 24, 641-657.

Brewer, M. B., & Brown, R. J. (1998). Intergroup relations. In D. T. Gilbert et al. (Eds.), The handbook of social psychology (Vol. 2, 4th ed., pp. 554-594). New York: McGraw-Hill.

Castano, E., Yzerbyt, V., & Paladino, M-P., & Sacchi, S. (2002). I belong, therefore, I exist: Ingroup identification, ingroup entitativity, and ingroup bias. Personality and Social Psychology Bulletin, 28, 135-143.

Crosby, F. (1976). A model of egoistical relative deprivation. Psychological Review, 83, 85-113.

Doosje, B., & Branscombe, N. R. (2003).

Attributions for the negative historical actions of a group. European Journal of Social Psychology, 33, 235-248.

Doosje, B., Branscombe, N. R., Spears, R, &

Manstead, A. S. R. (1998). Guilty by association: When one's group has a negative history. Journal of Personality and Social Psychology, 75, 872-886.

Doosje, B., Branscombe, N. R., Spears, R, & Manstead, A. S. R. (2004). Consequences of national ingroup identification for responses to immoral historical events. In N. R. Branscombe & B. Doosje (Eds.), Collective guilt: International perspectives (pp. 95-111). New York: Cambridge University Press.

Dumont, M., Yzerbyt, V., Wigboldus, D., & Gordijn,

E. H. (2003). Social categorization and fear reactions to the September 11th terrorist attacks. Personality and Social Psychology Bulletin, 29, 1509-1520.

Ekman, P. (1999). Basic emotions. In T. Dalgleish & M. J. Power (Eds.), Handbook of cognition and emotion (pp. 45-60). New York: Wiley.

Fiske, S. T. (1998). Stereotyping, prejudice, and discrimination. In D. T. Gilbert et al. (Eds.), The handbook of social psychology (Vol. 2, 4th ed., pp. 357-411). New York: McGraw-Hill.

Fiske, S. T., Cuddy, A. J. C., Glick, P., & Xu, J. (2002). A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition. Journal of Personality and Social Psychology, 82, 878-902.

Forgas, J. P. (2000). Affect and information

processing strategies: An interactive relationship. In J. P. Forgas (Ed.), Feeling and thinking: The role of affect in social cognition (pp. 253-280). New York: Cambridge University Press.

Hewstone, M., Cairns, E., Voci, A., McLernon,

F., Niens, U., & Noor, M. (2004). Intergroup

forgiveness and guilt in Northern Ireland: Social psychological dimensions of 'the troubles'. In N. R. Branscombe & B. Doosje (Eds.), Collective guilt: International perspectives (pp. 193—215). New York: Cambridge University Press.

Insko, C. A., Schopler, J., Pemberton, M. B., Wieselquist, J., McIlraith, S. A., Currey, D. P., et al. (1998). Long-term outcome maximization and the reduction of interindividual-intergroup discontinuity. Journal of Personality and Social Psychology, 75, 695-711.

Iyer, A., Leach, C. W., & Crosby, F. J. (2003). White guilt and racial compensation: The benefits and limits of self-focus. Personality and Social Psychology Bulletin, 29, 117-129.

Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: Self-evaluation of one's social identity. Personality and Social Psychology Bulletin, 18, 302-318.

Mackie, D. M., Devos, T., & Smith, E. R. (2000). Intergroup emotions: Explaining offensive action tendencies in an intergroup context. Journal of Personality and Social Psychology, 79, 602-616.

Mackie, D. M., Silver, L. A., & Smith, E. (2004). Intergroup emotion: Emotion as an intergroup phenomenon. In L. Z. Tiedens & C.W. Leach (Eds.), The social life of emotions (pp. 227-245). New York: Cambridge University Press.

Mackie, D. M., & Smith, E. R. (1998). Intergroup relations: Insights from a theoretically integrative approach. Psychological Review, 105, 499-529.

Mackie, D. M., & Smith, E. R. (2002). Intergroup emotions and the social self: Prejudice reconceptualized as differentiated reactions to outgroups. In J. P. Forgas & K. D. Williams (Eds.), The social self: Cognitive, interpersonal, and intergroup perspectives (pp. 309-326). Philadelphia: Psychology Press.

Maitner, A. T., Mackie, D. M., & Smith, E. R.

(2006). Evidence for the regulatory function of intergroup emotion: Emotional consequences of implemented or impeded intergroup action tendencies. Journal of Experimental Social Psychology, 42, 720-728.

Mallett, R. K., & Swim, J. K. (2004). Collective guilt in the United States: Predicting support for social policies that alleviate social injustice. In N. R. Branscombe & B. Doosje (Eds.), Collective guilt: International perspectives (pp. 56-74). New York: Cambridge University Press.

Mummendey, A., Klink, A., Mielke, R. Wenzel, M., & Blanz, M. (1999). Socio-structural

characteristics of intergroup relations and identity management strategies: Results from a field study in East Germany. European Journal of Social Psychology, 29, 259-285.

Neuberg, S. L., & Cottrell, C. A. (2002).

Intergroup emotions: A biocultural approach. In D. M. Mackie & E. R. Smith (Eds.), From prejudice to intergroup emotions: Differentiated reactions to social groups (pp. 265-283). Philadelphia: Psychology Press.

Ortony, A., Clore, G. L., & Collins, A. (1988). The cognitive structure of emotions. New York: Cambridge University Press.

Powell, A. A., Branscombe, N. R., & Schmitt, M. T. (2005). Inequality as ingroup privilege or outgroup disadvantage: The impact of group focus on collective guilt and interracial attitudes. Personality and Social Psychology Bulletin, 31,

508-521.

Roccas, S., Klar, Y., & Lioviatan, I. (2004).

Exonerating cognitions, group identification, and personal values as predictors of collective guilt among Jewish-Israelis. In N. R. Branscombe & B. Doosje (Eds.), Collective guilt: International perspectives. (pp. 130-147). New York: Cambridge University Press.

Roseman, I. J., Wiest, C., & Swartz, T. S. (1994). Phenomenology, behaviors, and goals differentiate discrete emotions. Journal of Personality and Social Psychology, 67, 206-221.

Sachdev, I., & Bourhis, R. Y. (1991). Power and status differentials in minority and majority group relations. European Journal of Social Psychology, 21, 1-24.

Schmitt, M., Behner, R., Montada, L., Muller, L., & Muller-Fohrbrodt, G. (2000). Gender, ethnicity, and education as privileges: Exploring the generalizability of the existential guilt reaction. Social Justice Research, 13, 313-337.

Schmitt, M. T., Branscombe, N. R., & Brehm, J. W. (2004). Gender inequality and the intensity of men's collective guilt. In N. R. Branscombe & B. Doosje (Eds.), Collective guilt: International perspectives (pp. 75-94). New York: Cambridge University Press.

Sherif, M. (1958). Superordinate goals in the reduction of intergroup conflict. American Journal of Sociology, 63, 349-356.

Silver, L. A., Miller, D. A., Mackie, D. M., & Smith, E. R. (2001). The nature of intergroup emotions. Unpublished manuscript. University of California at Santa Barbara.

Smith, E. R. (1993). Social identity and social emotions: Toward new conceptualizations of

prejudice. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 297—315). San Diego, CA: Academic Press.

Swim, J. K., & Miller, D. L. (1999). White guilt: Its antecedents and consequences for attitudes toward affirmative action. Personality and Social Psychology Bulletin, 25, 500-514.

Tajfel, H., & Turner J. C. (1986). The social identity theory of intergroup conflict. In S. Worchel & W. G. Austin (Eds.), Psychology of intergroup relations (pp. 7-24). Chicago: Nelson-Hall.

Turner, J. C., Brown, R. J., & Tajfel H. (1979).

Social comparison and group interest in ingroup favouritism. European Journal of Social Psychology, 9, 187-204.

Verkuyten, M., & Nekuee, S. (1999). Ingroup bias: The effect of self-stereotyping, identification and group threat. European Journal of Social Psychology, 29, 411-418.

Walster, E., Walster, G. W., & Berscheid, E. (1978). Equity: Theory and research. Boston: Allyn & Bacon.

Weiner, B. (1995). Judgments of responsibility: A foundation for a theory of social conduct. New York: Guilford.

Yzerbyt, V., Dumont, M., Wigboldus, D., &

Gordijn, E. (2003). I feel for us: The impact of categorization and identification on emotions

and action tendencies. British Journal of Social Psychology, 42, 533-549.

Paper received 2 November 2004; revised version accepted 2 July 2005.

Biographical notes

angela t. maitner is a doctoral candidate in

social psychology at the University of California, Santa Barbara. Her research investigates the cognitive and affective consequences of self-categorization and social identification.

diane m. mackie is professor of psychology and communication at the University of California, Santa Barbara. She received her PhD from Princeton University. Her research interests include the affective antecedents and consequences of intergroup relations, as well as social influence and persuasion.

eliot r. smith is professor in the department of psychological and brain sciences at Indiana University, Bloomington, and he also holds a Special Chair at the Free University, Amsterdam. His PhD is from Harvard University. His work centers on the role of emotions in prejudice and intergroup behavior, and socially situated cognition.