Third-party intervention in intergroup reconciliation: The role of neutrality and common identity with the other conflict party

Abstract

Third parties, particularly if neutral, have been found to promote instrumental conflict resolution. Using the needs-based model’s theoretical framework, we investigated whether third parties can also promote socioemotional reconciliation. Study 1 (N = 124) revealed that in the context of fraud between universities, conciliatory messages from either the other conflict party or a third party sharing common identity with it increased group members’ willingness to reconcile more than equivalent messages from a neutral third party. Replicating and extending this pattern, Study 2 (N = 177) exposed Israeli Jewish participants to texts which reminded them of historical transgressions conducted by Palestinians or against them. We found that compared to a control condition, messages supposedly conveyed by either Palestinians or Jordanians, but not by the UN, increased Israeli Jews’ willingness to reconcile with Palestinians. These effects were mediated by the extent to which the official conveying these messages was perceived as representing the other conflict party.

Keywords

common ingroup identity neutrality reconciliation representativeness third parties

The field of intergroup relations has recently shifted focus from reducing prejudice and hostility to promoting positive intergroup relations (Tropp & Mallett, 2011). This shift is reflected in the growing interest in reconciliation (Nadler, Malloy, & Fisher, 2008). Reconciliation can be defined as the process of removing the psychological barriers that block the path to healing discordant relationships and restoring the adversaries’ positive identities (Kelman, 2008; Shnabel & Nadler, 2008). Conceptualized as a relationship-oriented process (Subašic & Reynolds, 2009), reconciliation operates at the socioemotional level (Nadler & Shnabel, 2008) and involves a changed psychological orientation toward the other (Staub, 2008), and a shift in the way former adversaries relate to each other and view their future relationship (Bar-Tal, 2000). These processes are distinguished from conflict resolution, which operates on the instrumental level and may be defined as the cessation of conflict by way of resolving tangible, pragmatic issues (e.g., reaching agreed-upon formulas for distributing resources such as money or land; Nadler & Shnabel, 2008).

Because the concept of reconciliation is relatively new to scientific discourse (Nadler et al., 2008), its related processes are still understudied. One such understudied issue is the role of third-party intervention. Conflict resolution studies have indicated that third parties, particularly if perceived as neutral and unbiased (e.g., Black & Baumgartner, 1983; Stulberg, 1987; Young, 1967) can facilitate conflict resolution. In fact, third parties’ ability to promote conflict resolution is often greater than that of the other conflict party because they may be perceived as fairer (Carnevale & Pruitt, 1992), encourage a “problem-solving” orientation (Fisher & Ury, 1981), allow the conflict parties to make concessions without losing face (Rubin, 1980), and the solutions they propose are not devalued to the same extent as identical solutions proposed by the other conflict party (Ross, 1995).

However, few studies have explored whether third parties can similarly promote reconciliation. Knowles (1958) suggested that third party interventions may improve factual-objective but not emotional-subjective communication between adversaries. If so, third parties may not be able to effectively promote reconciliation since this requires addressing the conflict parties’ emotional needs (Shnabel & Nadler, 2008). At the same time, because the effectiveness of third-party interventions is not necessarily all or nothing, it is important to study the influence of third-party characteristics; in particular whether neutrality is an asset and association with the other conflict party a drawback, as is the case in conflict resolution. The goal of the present research, consisting of two studies, is to provide initial answers to these queries.

Our theoretical framework is guided by the needs-based model of reconciliation (NBM; Nadler & Shnabel, 2008; Shnabel & Nadler, 2008). This model suggests that following transgressions, victimized groups experience loss of power, control, and respect; consequently, their members experience an enhanced need for power and wish to restore their identity as strong and agentic. Conversely, because social exclusion is the sanction imposed upon those who violate moral standards (Tavuchis, 1991), members of perpetrator groups experience an enhanced need for social and moral acceptance which would restore their positive moral identity. When the conflicting parties convey to each other messages that address their unique emotional needs, that is, empowering messages to the victim group and accepting messages to the perpetrator group, this enhances their readiness to reconcile. To illustrate, in one study which focused on the context of the Holocaust (Shnabel, Nadler, Ullrich, Dovidio, & Carmi, 2009, Study 2) Jews showed greater willingness to reconcile with Germans following empowering (compared to accepting) messages by a German official. Conversely, Germans showed greater willingness to reconcile with Jews following accepting rather than empowering messages from a Jewish official. These findings suggest that intergroup reconciliation may be promoted through reciprocal exchange of empowering and accepting messages.

While this previous research pointed to the content of messages as critical for reconciliation, it has overlooked the role of message source. In the experiments conducted so far within the NBM framework the source conveying acceptance and empowerment messages was always the other conflicting group (see Shnabel et al., 2009; Shnabel, Ullrich, Nadler, Dovidio, & Aydin, 2013) but never third parties outside the victim–perpetrator dyad. Yet understanding the potential of third parties to promote reconciliation, as done in the present research, is crucial because the parties directly involved in the conflict are often reluctant to convey empowering or accepting messages to each other due to their concern that such positive gestures would not be reciprocated or be even used against them (Shnabel & Noor, 2012).

The Present Research

In the literature on conflict resolution, third parties are usually conceptualized as neutral entities that facilitate rather than direct the process, thereby helping the conflict parties reach a settlement (e.g., Carnevale & Arad, 1996). Yet, conflict resolution and reconciliation operate on different levels (instrumental vs. social-emotional; Kelman, 2008; Nadler et al., 2008). Therefore, third-party interventions were theorized to be relatively incapable of improving emotional-subjective communication (Knowles, 1958) of the type conveyed in empowering and accepting messages. Consequently, we hypothesized that third parties would be generally less effective in bringing about reconciliation compared to the other conflict party (Hypothesis 1).

Importantly, however, third parties are not a monolithic category (Carnevale & Arad, 1996). We therefore examined two different third-party stances: a neutral position, or siding with the other conflict party. In particular, the third party and the other conflict party can either represent entirely separate groups or share a common identity (e.g., Gaertner & Dovidio, 2000). We hypothesized that third parties sharing a common identity with the other conflict party would be more effective in promoting reconciliation than neutral third parties (Hypothesis 2). While there is no direct empirical evidence to support this hypothesis, it is generally consistent with Volkan’s (1985) theorizing that groups tend to identify other groups as either their “enemies” or their “allies.” If so, groups that share a common identity with the other conflict party may be likewise perceived as “enemies,” following Benjamin Franklin’s famous quote: “the friend of my enemy is my enemy.” Consequently, the other conflict party and third parties sharing a common identity with it may elicit similar emotional responses—in our case, increased readiness to reconcile following conciliatory messages.

Moreover, a large body of work within the social identity approach (Tajfel & Turner, 1979) shows that the perception of a common identity between an ingroup and an outgroup creates the impression of common fate (Brewer, 2000), has positive effects on intergroup attitudes (Dovidio, Gaertner, & Loux, 2000), and fosters intergroup forgiveness (Wohl & Branscombe, 2005). This work reveals that if members of different groups are perceived as belonging to a more inclusive category and consequently share certain identity features, the processes that lead to the responses towards one group (e.g., positive attribution of traits) would be directed also toward members of the other group (see the common ingroup identity model; Gaertner & Dovidio, 2000;2012). While this research has focused mainly on ingroup–outgroup relations, its logic can be extrapolated to group members’ responses towards two different outgroups: to the extent that these outgroups are perceived as sharing a common identity the responses towards one group should be extended to the other group as well. Applying this rationale to the present research, group members’ response to messages conveyed by a third party that shares a common identity with the other conflict party should be similar to their response to messages conveyed by the other conflict party itself. By contrast, neutral third parties not sharing any core identity features with the other conflict party are less likely to elicit emotional responses similar to the responses elicited by the other conflict party; their conciliatory messages are thus less likely to facilitate reconciliation.

Study 1 tested our hypotheses through examining participants’ responses to conciliatory messages from the other conflict party compared to a neutral third party or a third party sharing common identity features with the other conflict party following a single transgression episode, which involved academic fraud. Study 2 replicated and extended Study 1 by testing our hypotheses in the natural context of the prolonged, violent conflict between Jews and Palestinians and adding a comparison of third parties’ messages effectiveness to a no-message, control condition. In addition, Study 2 investigated whether perceived representativeness serves as one essential underlying mechanism that explains the observed effects of message source. Specifically, as opposed to contexts of interpersonal transgressions, conciliatory messages following intergroup transgressions are typically conveyed via group representatives, rather than through the direct victims or perpetrators (Blatz & Philpot, 2010). We expected the group membership of the officials who conveyed the conciliatory messages to determine their perceived representativeness such that officials who belong to the conflicting outgroup would be perceived as representing it to a greater extent than officials who belong to the common identity third party, who would in turn be perceived as representing the conflicting outgroup more than officials who belong to a neutral third party. Moreover, building on research on intergroup contact and the factors determining its effectiveness in promoting positive intergroup relations, such as typicality of group members (Brown, Eller, Leeds, & Stace, 2007; Brown, Vivian, & Hewstone, 1999; Wilder, 1984), we predicted the official’s perceived representativeness to determine the messages’ effectiveness: the more an official conveying a conciliatory message is perceived to represent the conflicting outgroup, the more effective the message would be in promoting the recipient group members’ willingness to reconcile with the conflicting outgroup. Taken together, the present studies were designed to advance our understanding of whether and how interventions by third parties can contribute to reconciliation.

Study 1

Study 1 compared the effectiveness of messages within the victim–perpetrator dyad (i.e., messages from the other conflict party) to messages conveyed by third parties that either shared a common identity with the other conflict party or were neutral and uninvolved. Participants in Study 1 were first led to believe that their ingroup was in conflict with another group. Depending on the experimental condition, participants’ ingroup was either the victim or the perpetrator group. They were then randomly assigned to receive conciliatory messages from the other conflict party, a third party that shared a common identity with it, or a neutral third party.

To maintain theoretical clarity and avoid design complexity we focused on messages previously identified as effective in facilitating victims’ and perpetrators’ readiness to reconcile. Previous studies that compared different contents of message using the full design (Shnabel & Nadler, 2008; Shnabel et al., 2009; Shnabel et al., 2013) revealed that perpetrators’ willingness to reconcile depended mainly on the acceptance message (i.e., empowering messages from victims were relatively ineffective), whereas the empowerment message was critical for victims (i.e., accepting messages by perpetrators failed to promote victims’ willingness to reconcile). Thus, in the present study participants assigned to the victim condition were exposed only to empowering messages whereas participants assigned to the perpetrator condition were exposed only to accepting messages, resulting in a 2 (role: victim, perpetrator) × 3 (message source: other conflict party, neutral third party, common-identity third party) design.¹ We predicted that messages from the other conflict party would facilitate reconciliation more than messages from third parties. Also, messages from a third party sharing a common identity with the other conflict party were predicted to promote reconciliation more effectively than messages from a neutral third party.

Method

Participants

One-hundred and twenty-four German university students (64 women, 59 men, one unspecified; M_age = 23, range: 19–59) participated in the study in exchange for chocolate bars.

Procedure

Participants were told that the goal of the study was to investigate the impact of information provided by specific media (e.g., video vs. audio vs. printed text) on cognitive information processing. Then they were led to believe that they were in the “printed text” condition and asked to read a sham newspaper article allegedly taken from a national student magazine. The article described a national contest for university students organized by the Hochschulrektorenkomission (HRK), an association of university headmasters. Participants learned that because of a fraud, the cash prize was not awarded to the legitimate winners. In the victim condition, the participants’ ingroup (University A) unjustly lost the prize through trickery on the part of the outgroup (University B). In the perpetrator condition, these roles were reversed. The article continued by saying that although the fraud was discovered, the rightful winners could not be awarded the prize money retroactively. Then, depending on the condition, participants read a message conveyed by either the other conflict party (from University B) or a third party.

To ensure high internal validity, the same group—Dresden University of Technology (“TU Dresden”)—was used as the message source in both third-party conditions. However, in the neutral third-party condition it was explained that TU Dresden was an impartial arbitrator in the competition, whereas in the common-identity condition TU Dresden was said to be a partner university of the other conflict party. In the victim condition the message expressed empowerment (i.e., acknowledgment of their ingroup’s competence and right to be proud) whereas in the perpetrator condition the message expressed acceptance (i.e., empathy for the distress felt by the participants’ ingroup as a result of the fraud and expression of social approval despite their actions). The manipulation of empowerment and acceptance was consistent with previous theorizing and empirical practice (e.g., Shnabel et al., 2009).

Measures

After reading the article participants filled out a series of measures, using a 5-point scale (from 1 = not at all to 5 = very much).

Social role manipulation check

Participants indicated the extent to which they agreed with two statements assessing their ingroup’s victimhood (e.g., “Students from University A were victims of a fraud”; r = .77, p < .001) and two statements assessing their ingroup’s perpetration (e.g., “Students from University A deliberately committed fraud”; r = .87, p < .001).

Message content manipulation check

Four items measured the messages’ perceived empowerment (e.g., “Your ingroup has the right to be proud”; α = .86) and another four measured perceived acceptance (“We understand your ingroup’s feelings”; α = .63).

Third-party neutrality versus common-identity manipulation check

For participants assigned to the third-party conditions, two items checked the perception of the third party as neutral (e.g., “TU Dresden students are neutral with respect to both universities”) and two items measured the perception of the third party as sharing a common identity with the other conflict party (e.g., “TU Dresden students have a lot in common with University B,” reverse coded). The four items were combined into a 7-point measure of third-party neutrality scale (α = .63; from 1 = not at all to 7 = very much).

Willingness to reconcile

Participants indicated their agreement with 10 items (α = .86) used in previous studies to measure reconciliation (e.g., Shnabel et al., 2009) and slightly adjusted to the context of the present study (e.g., “The message increases my willingness to promote reconciliation between students of both universities”; “The message reduces the tension between students of the two universities”; “The message brings students of both universities closer together”).

Results

Table 1 presents the means and standard deviations of all the following analyses.

Table 1.

Means (SD) of the manipulation checks and dependent variable for Study 1.

	Social role: Perpetrator			Social role: Victims
	Other conflict party	Common-identity third party	Neutral third party	Other conflict party	Common-identity third party	Neutral third party
Manipulation check for role: Victim	1.78	1.46	1.50	3.55	3.82	3.84
Manipulation check for role: Victim	(0.66)	(0.67)	(0.42)	(0.71)	(0.85)	(0.63)
Manipulation check for role: Perpetrator	3.82	3.89	4.06	1.22	1.50	1.27
Manipulation check for role: Perpetrator	(0.76)	(0.98)	(0.66)	(0.34)	(0.74)	(0.44)
Manipulation check for message content: Empowerment	2.48	2.61	2.70	4.45	4.42	4.64
Manipulation check for message content: Empowerment	(0.71)	(0.94)	(0.85)	(0.57)	(0.72)	(0.42)
Manipulation check for message content: Acceptance	3.52	3.61	3.67	3.18	3.47	3.39
Manipulation check for message content: Acceptance	(0.69)	(0.73)	(0.65)	(1.02)	(0.95)	(0.90)
Manipulation check: Third-party neutrality	–	3.91	4.65	–	3.67	5.04
Manipulation check: Third-party neutrality		(1.16)	(1.27)		(0.73)	(0.98)
Willingness to reconcile	3.46	3.00	2.75	2.61	2.45	2.02
	(0.72)	(0.55)	(0.56)	(0.56)	(0.55)	(0.43)

Social Role Manipulation Check

We conducted a 2 (role: victims, perpetrators) × 3 (message source: other conflict party, neutral third party, common-identity third party) MANOVA with perceived victimhood and perceived role as perpetrator as the two dependent variables. The analysis revealed a significant multivariate effect of role, F(2, 117) = 276.59, p < .001, η² = .83. No other multivariate effects were significant, Fs < 1.92, ps > .11.

Between-subjects analyses showed that as intended, participants in the victim condition perceived their ingroup as the victim more than participants in the perpetrator condition, F(1, 118) = 319.73, p < .001, η² = .73. In contrast, participants in the perpetrator condition perceived their ingroup as the perpetrator more than participants in the victim condition, F(1, 118) = 424.90, p < .001, η² = .78. Neither message source nor the interaction of both independent variables had a significant effect, Fs < 1.92, ps > .11. In line with this, post hoc tests for the three message source conditions revealed no significant differences.

Message Content Manipulation Check

A 2 (role: victims, perpetrators) × 3 (message source: other conflict party, neutral third party, common-identity third party) MANOVA with perceived empowerment and perceived acceptance expressed in the message content as the two dependent variables revealed a significant main effect for role on perceived message content, F(2, 117) = 126.68 p < .001, η² = .68. No other multivariate effects were significant, Fs < 1.

As intended, participants in the victim condition perceived the content of the message as expressing greater empowerment compared to participants in the perpetrator condition, F(1, 118) = 209.56, p < .001, η² = .64. Likewise, participants in the perpetrator condition tended to perceive the message as expressing greater social acceptance compared to participants in the victim condition, F(1, 118) = 2.85, p = .09, η² = .03. All other results, including the post hoc comparisons, were not significant, Fs < 1.

Third-Party Neutrality Versus Common-Identity Manipulation Check

For participants in the third-party conditions, a 2 (role: victims, perpetrators) × 2 (message source: neutral third party, common-identity third party) ANOVA tested whether our neutrality versus common-identity manipulation was successful. The analysis revealed a significant main effect of message source, F(1, 83) = 20.68, p < .001, η² = .20, but no other effects, Fs < 2, ps > 2. As intended, participants in the neutral condition rated the third party as significantly more neutral than participants in the common-identity condition.

Willingness to Reconcile

Consistent with previous findings (Shnabel et al., 2009), a 2 (role: victims, perpetrators) × 3 (message source: other conflict party, neutral third party, common-identity third party) ANOVA revealed a significant main effect of role, indicating that perpetrators were more willing to reconcile than victims, F(1, 118) = 48.08, p < .001, η² = .29. More importantly for the purpose of the present study, there was a main effect for message source, F(2, 118) = 13.14, p < .001, η² = .18. Multiple comparisons (least significant difference [LSD]) revealed that as expected, participants in the other conflict party condition showed significantly greater willingness to reconcile than those in the common-identity (p = .02, 95% CI: [−.62, −.09] or neutral (p < .001, 95% CI: [ −.95, −.42] third-party conditions. Further supporting our hypothesis, participants in the common-identity condition were more willing to reconcile with the other conflict party than those in the neutral third-party condition (p = .006, 95% CI: [−.58, −.07]. The interaction was not significant, F < 1.²

Discussion

The results of Study 1 supported our predictions. Participants in both social roles were more willing to reconcile with the other conflict party when the conciliatory message was conveyed by the other conflict party, followed by the third party sharing a common identity. Participants were least willing to reconcile when the message was provided by a neutral third party. These findings theoretically extend the NBM, which has focused so far almost exclusively on message content (see Shnabel, Nadler, & Dovidio, 2014, for the only exception), by pointing to the critical importance of message source. Specifically, although empowering and accepting messages from neutral third parties were perceived as intended, they failed to translate into increased willingness to reconcile compared to similar messages by groups related to the conflict either directly (i.e., the other conflict party) or by association (i.e., the common-identity third party). Thus, in addition to direct communication, the healing of relationships between adversarial groups benefits from the involvement of a party (message source) related to the victim–perpetrator dyad.

Study 2

Study 2 aimed to extend the findings of Study 1 in several ways. First, to extend the generalizability of Study 1’s findings by replicating them in a different context, participants in Study 2 were Israeli Jews who were randomly assigned to read about either the 1954 Ma’ale Akrabim massacre, in which members of their ingroup were victimized by Israeli Palestinians,³ or the 1956 Kafr Qasim massacre, in which members of their ingroup victimized Israeli Palestinians. These massacres were “symmetrical” in the sense that they both took place in the 1950s and involved the shooting of innocent civilians. Despite the substantially different research contexts, we expected to replicate the findings of Study 1 such that messages from the other conflict party would be more effective in promoting reconciliation than messages from a third party sharing a common identity with it, which in turn would be more effective than messages from a neutral third party.

Second, whereas Study 1 revealed that neutral third parties were less effective in facilitating reconciliation compared to groups involved in the conflict either directly or by association, it did not examine their effectiveness in absolute terms. Such examination is important because in cases where the neutral third party is the only available option, its intervention, if effective to some degree, may be better than nothing. Study 2 therefore compared the effectiveness of the different message sources not only to each other, but also to a no-message control condition.

Finally, the third goal of Study 2 was to shed light on the mechanism underlying the effects of message source on group members’ willingness to reconcile. Theorizing within the field of interpersonal transgressions suggests that empowering messages conveyed by the perpetrators (e.g., messages acknowledging victims’ value and the injustice perpetrated against them) imply that they regret the transgression and intend not to repeat it in the future (Okimoto & Wenzel, 2008). Correspondingly, accepting messages conveyed by the victims imply that they are ready to move on and let the transgression be a thing of the past (e.g., abandon anger and revenge; McCullough, Worthington, & Rachal, 1997). However, a key difference between conciliatory messages conveyed between groups, as opposed to individuals, is that the messages are not expressed by the actual sufferers or wrongdoers but rather by officials (e.g., group leaders) speaking on behalf of the victims’ or perpetrators’ ingroup (Benziman, 2009; Blatz & Philpot, 2010; Tavuchis, 1991). Consequently, the implications of empowering or accepting messages for intergroup relations are determined by the extent to which the officials conveying them actually represent the other conflict party. We therefore theorized that members of the recipient group engage in evaluating the extent to which the official conveying conciliatory message represents the other conflict party, and that perceived representativeness, in turn, determines the message’s effectiveness in promoting reconciliation.

Within contexts that highlight instrumental goals, representing a group means serving as this group’s advocate or agent with the purpose of maximizing its profits (see Teixeira, Demoulin, & Yzerbyt, 2011). Conversely, within contexts that highlight identity-related goals, representing a group means being a typical member of it such that the representative’s feelings, values, and attitudes reflect those of group members in general (Teixeira et al., 2011). Literature within the field of conflict resolution, which involves negotiation over instrumental interests, informs us that third parties’ effectiveness is driven by their neutrality, that is, the extent to which they do not represent the other conflict party (Arad & Carnevale, 1994; Conlon & Ross, 1993; Welton & Pruitt, 1987).

However, reconciliation involves “negotiation” over identity-related as opposed to instrumental issues (Kelman, 2008; Nadler & Shnabel, 2008). Hence, we based our predictions on research within the social identity approach. In particular, research on intergroup contact (Pettigrew, 1998) reveals that contact with a single member of an outgroup can generalize into more positive attitudes toward the outgroup with which the ingroup is in conflict (Allport, 1954; Pettigrew & Tropp, 2006). However, the benefits of contact depend upon the degree to which this outgroup member is perceived to represent the outgroup as a whole, that is, the extent to which he or she is perceived as a typical member of the outgroup, identifies with it, and possesses traits that generally characterize its members (Brown et al., 2007; Brown et al., 1999; Wilder, 1984). Because research on intergroup contact has focused primarily on the ingroup–outgroup dyad without considering external third parties, representativeness has been conceptualized mainly as characterizing members within a group. To illustrate, a within-group approach to representativeness would suggest that a Muslim Palestinian should be judged as representing the Palestinian group more than a Christian Palestinian, because Muslims are the religious majority among Palestinians and therefore constitute more typical members of this group. We suggest, however, that representativeness may also be conceptualized as operating across groups. That is, once a third party is introduced into the ingroup–outgroup dyad, group members may compare the groups at the level of the next inclusive category (see Turner, Hogg, Oakes, Reicher, & Wetherell, 1987) with regard to the dimension in question (e.g., religion). To illustrate, as they share common-identity features, a resident of an Arab country (even if not a Palestinian) is likely to be judged as representing the Palestinians (e.g., identifying with and displaying characteristics typical of Palestinians in general) more than a resident of a European or American country.

Based on this theorizing, we predicted that officials who belong to a third party that shares a common identity with the other conflict party would be perceived as more representative of it than officials belonging to a neutral third party. Perceived representativeness, in turn, was predicted to determine the effectiveness of the conciliatory message, in line with findings regarding the influence of perceived typicality on the generalization of positive attitudes toward the outgroup as a whole (Brown et al., 2007; Brown et al., 1999; Wilder, 1984). More specifically, we predicted that a message from a UN official that was—given our pilot studies (see endnotes)—not expected to be perceived as representing the Palestinians, would fail to facilitate reconciliation compared to the no-message condition. By contrast, messages from both an Israeli Palestinian and a Jordanian official, who were expected to be perceived as representing the Palestinians at least to some degree, were predicted to promote reconciliation compared to the no-message condition.

In order to test these hypotheses, Study 2 used a 2 (role: victims, perpetrators) × 4 (message source: other conflict party, neutral third party, common-identity third party, no-message) design in which following the assignment to a social role, participants were exposed to conciliatory messages from officials belonging to the other conflict party (Israeli Palestinians), a third party who shared common identity with it (Jordanians, who share an Arab identity with Israeli Palestinians), a neutral third party (the UN), or no message.

Method

Participants

Israeli Jewish participants were recruited through snowball sampling to participate in an online study. Seven participants were excluded from the sample because they failed to correctly identify the official’s group membership (see Results); the final sample included 111 women and 66 men (M_age = 27, range: 18–61) who identified themselves as Israeli Jews.

Procedure

Using the same cover story as in Study 1, participants were first assigned to read about either the Ma’ale Akrabim (victim condition) or the Kafr Qasim (perpetrator condition) massacres. This manipulation is based on previous findings showing that although the Israeli–Palestinian conflict is generally marked by a duality of social roles, in the sense that both parties transgress against each other and often compete over the role of the “true” victim of the conflict (Shnabel & Noor, 2012), it is possible to experimentally induce members of the involved groups with feelings of either victimization or perpetration (see SimanTov-Nachlieli & Shnabel, 2014).

Next, participants in all conditions (except for control) read a short excerpt allegedly summarizing the main message of a speech held on the occasion of the 55th anniversary of the massacre (for more information see the supplementary materials). As in Study 1, the content of the message was designed to address the emotional needs of its receivers. Thus, the Ma’ale Akrabim speech conveyed an empowering message (e.g., “The Jews in Israel are entitled to security and self-determination; we must fully respect their right to live with their heads up”); and the Kafr Qasim speech conveyed an accepting message (e.g., “We need to accept the Jews and remember that it is not easy for them to cope with their emotions after the massacre”). Depending on the experimental condition, the official was identified as the head of the High Follow-Up Committee for Arab Citizens of Israel (i.e., other conflict party), the Jordanian ambassador to Israel (i.e., common-identity third party), or the UN envoy to the Middle East (i.e., neutral third party). There was no speech in the control condition.

Our choice of third parties was guided by our wish to maximize ecological validity. We used the UN as the neutral third party because it has been involved in various peacekeeping efforts and has traditionally adopted neutrality as one of its core principles (Kranso, 2004).⁴ We selected Jordan as a third party that shares a common identity with the Israeli Palestinian group because both share an Arab identity. Also, Jordan is the only Arab country, except for Egypt, that maintains full diplomatic relations with Israel and it would therefore be reasonably expected to deliver a statement on the massacres’ anniversaries.

After reading the messages (except for the no-message condition) participants completed the manipulation checks and dependent measure.

Measures

All measures (except for perception of message source) were adapted from Shnabel et al. (2009), using 7-point scales ranging from 1 = completely disagree to 7 = completely agree.⁵ See the supplementary materials for more information.

Social role manipulation check

Immediately following the assignment to social roles, participants indicated the extent to which the massacre they had read about impaired their ingroup’s (a) sense of power, and (b) moral image.

Message source manipulation check

Participants in the three message conditions were asked to identify the group to which the official belonged (Israeli Palestinian, Jordanian, or other).

Message content manipulation check

Four items examined participants’ perception of the extent to which the speech expressed empowerment of their ingroup (e.g., “Jews have the right to have political power”; α = .86); five additional items measured the extent to which the speech expressed social acceptance (e.g., “We can feel empathy towards the Jews”; α = .87). Participants in the control condition were not exposed to any speech, so they were not asked about message source or content.

Perceived representativeness

Three items assessed the extent to which the official conveying the message was perceived as representing the outgroup (e.g. “The official represents Israeli Palestinians”; “The official identifies with Israeli Palestinians”; α = .84).

Willingness to reconcile

Modelled after Study 1, 10 items measured the participants’ willingness to reconcile with Israeli Palestinians (e.g., “I feel more optimistic regarding the future relations between Jews and Arabs in Israel”; “I feel increased willingness to act for achieving peaceful coexistence between Jews and Arabs”; “I feel that the atmosphere between Jews and Arabs in Israel has improved”; α = .93).

Results

Table 2 presents the means and standard deviations of all the following analyses, except for those of willingness to reconcile, which are presented in Figure 1.

Table 2.

Means (SD) of the manipulation checks and perceived representativeness for Study 2.

	Social role: Perpetrator				Social role: Victim
	Other conflict party	Common-identity third party	Neutral third party	No message	Other conflict party	Common-identity third party	Neutral third party	No message
Manipulation check for role: Sense of power	4.16	3.70	4.15	4.18	3.21	3.04	2.59	3.52
Manipulation check for role: Sense of power	(1.49)	(1.06)	(1.28)	(1.23)	(1.11)	(0.96)	(1.22)	(1.07)
Manipulation check for role: Moral image	1.88	2.26	1.77	1.82	4.59	4.61	4.71	4.62
Manipulation check for role: Moral image	(0.92)	(1.16)	(0.92)	(1.18)	(0.78)	(0.78)	(1.26)	(1.20)
Manipulation check for message content: Empowerment	4.09	4.55	4.23	–	5.10	5.35	5.67	–
Manipulation check for message content: Empowerment	(1.65)	(1.79)	(1.55)		(1.12)	(1.30)	(1.14)
Manipulation check for message content: Acceptance	6.14	5.65	5.61	–	4.67	4.45	4.78	–
Manipulation check for message content: Acceptance	(0.70)	(1.09)	(0.93)		(1.23)	(1.29)	(1.68)
Perceived representativeness	5.18	3.93	2.79	–	4.57	3.42	2.54	–
Perceived representativeness	(1.29)	(1.32)	(1.23)		(1.14)	(1.05)	(1.21)

Figure 1.

Willingness to reconcile as a function of social role and message source.

Social Role Manipulation Checks

As expected, a 2 (role: victim, perpetrator) × 4 (message source: Palestinian, Jordanian, UN, control) MANOVA with sense of power and moral image as the two dependent variables revealed a significant multivariate main effect of role, F(2, 168) = 182.68, p < .001, η² = .69; no other multivariate effects emerged, Fs < 1.70.

The between-subjects analysis confirmed that participants in the victim condition reported a lower sense of power than participants in the perpetrator condition, F(1, 169) = 27.31, p < .001, η² = .14. In contrast, participants in the perpetrator condition reported a lower moral image than participants in the victim condition, F(1, 169) = 285.58, p < .001, η² = .63. No other effects were significant, Fs < 2.3, ps < .17.

Message Content Manipulation Check

A 2 (role: victim, perpetrator) × 3 (message source: Palestinian, Jordanian, UN) MANOVA with perceived empowerment and perceived acceptance expressed in the message content as dependent variables revealed that the manipulation of role had a significant effect on participants’ perception of message content, F(2, 132) = 41.59, p < .001, η² = .39; no other multivariate effects emerged, Fs < 1.9.

As intended, participants in the victim condition perceived the message as expressing more empowerment than participants in the perpetrator condition, F(1, 133) = 16.76, p < .001, η² = .11. In contrast, participants in the perpetrator condition perceived the message as expressing more acceptance than participants in the victim condition, F(1, 132) = 30.76, p < .001, η² = .19. The analysis revealed no further significant effects, Fs < 1.2.

Message Source Manipulation Checks

Seven participants mistakenly identified the official’s group membership and were therefore excluded from all analyses (including those reported before).

Perceived Representativeness

We next conducted a 2 (role: victim, perpetrator) × 3 (message source: Palestinian, Jordanian, UN) ANOVA on perceived representativeness. An unexpected main effect for role emerged, F(1, 133) = 4.53, p = .035, η² = .03, such that participants in the perpetrator condition perceived the official as representing Israeli Palestinians more than did participants in the victim condition. One possible explanation is that this result may reflect the fact that the acceptance message (conveyed in the perpetrator condition) was perceived as representing the Israeli Palestinians’ stance more accurately than the empowerment message (which acknowledged the Jew’s right for security and self-determination). Of primary interest for the present study, the effect of message source was significant, F(2, 133) = 33.76, p < .001, η² = .34. Multiple comparisons (LSD) indicated that each condition significantly differed from the other two (ps < .001). As expected, an Israeli Palestinian official was perceived as representing the other conflict party more than a Jordanian official, F(1, 133) = 25.25, p < .001; 95% CI: [.33, 1.95] which was in turn perceived as representing the other conflict party more than a UN official, F(1, 133) = 13.89, p < .001; 95% CI: [1.57, 3.20], Ms = 4.88, 3.68, and 2.67, respectively. Finally, the interaction between role and message source was not significant, F < 1.

Willingness to Reconcile

Consistent with Study 1, a 2 (role: victim, perpetrator) × 4 (message source: Palestinian, Jordanian, UN, control) ANOVA revealed a main effect for role, F(1, 169) = 6.18, p = .014, η² = .04, such that perpetrators exhibited greater willingness to reconcile than victims. Additionally, the predicted main effect for message source was found, F(3, 169) = 11.95, p < .001, η² =. 18, whereas the two-way interaction failed to reach significance, F < 1.

Multiple comparisons revealed that compared to the no-message condition, messages from the Palestinian (p < .001, 95% CI: [.81, 2.54]) and Jordanian (p = .001, 95% CI: [.32, 2.03]) officials, but not from the UN official (p = .43, 95% CI: [−.45, 1.41]) significantly increased Israeli Jews’ willingness to reconcile with Israeli Palestinians. Moreover, whereas the messages from the Jordanian and Palestinian officials did not differ from each other (p = .27) they both significantly improved group members’ willingness to reconcile, compared to messages from the neutral UN official (ps < .001; see Figure 1).⁶

Mediation Analysis

Using the MEDIATE macro developed by Hayes and Preacher (2013), we conducted a mediation analysis to examine the indirect effects of the various message sources on willingness to reconcile through perceived representativeness. To conduct the analysis, the independent variable, message source, was dummy-coded such that the neutral third-party condition (i.e., the UN) was the reference category, producing two vectors (i.e., independent variables) representing the contrasts between the other conflict party (i.e., Palestinian) and the neutral third party, and between the common-identity third party (i.e., Jordanian) and the neutral third party. The control condition included no message, and hence no measure of perceived representativeness. It was therefore not included in the mediation analysis.

For the sake of concision and bolstering statistical power (Schimmack, 2012), and in the absence of any interaction between social role and message source (see previous lines), we collapsed the data across the two social roles (importantly, the pattern of results persisted when testing for each role separately; i.e., the effects of both independent variables (IVs) on the mediator, the effect of the mediator on the dependent variable (DV), and the two indirect effects, were significant among both victims and perpetrators). As seen in Table 3, both the Palestinian and the Jordanian conditions significantly increased participants’ perception of representativeness compared to the UN condition, and perceived representativeness, in turn, predicted greater willingness to reconcile. Also, as predicted, the two indirect effects—that is the path from message from a Palestinian official to willingness to reconcile through perceived representativeness, and the path from message from a Jordanian official to willingness to reconcile through perceived representativeness—were significant (as seen in the table, zero was not included in the 95% confidence intervals).

Table 3.

The indirect effect of message source on willingness to reconcile via perceived representativeness (Study 2).

	B	SE	t	p
IV-1 on mediator: Palestinian source on perceived representativeness	2.20	.28	7.91	.000
IV-2 on mediator: Jordanian source on perceived representativeness	1.02	.28	3.69	.000
Mediator on DV: Perceived representativeness on willingness to reconcile	.34	.09	3.86	.000
IV-1 on DV: Direct effect: Palestinian source on willingness to reconcile	.68	.35	1.98	.049
IV-2 on DV: Direct effect: Jordanian source on willingness to reconcile	.84	.30	2.81	.006
95% CI	Effect	SE	LL	UL
Indirect effects for IV-1
Palestinian source on willingness to reconcile through perceived representativeness	.75	.23	.33	1.22
Indirect effects for IV-2
Jordanian source on willingness to reconcile through perceived representativeness	.35	.13	.12	.63

Note. N = 139 Israeli Jews. Participants in the control condition were not included in this analysis, because perceived representativeness was measured only in conditions that included messages. The two social role conditions (victims, perpetrators) are collapsed. The three source conditions are coded such that the neutral (UN) condition is the reference group. Thus, the Palestinian source (IV-1) represents the contrast between the Palestinian and UN conditions (coded Palestinian = 1, Jordanian and UN = 0); while the contrast between the Jordanian and UN conditions (coded Jordanian = 1, Palestinian and UN = 0) is controlled for. Correspondingly, the Jordanian source (IV-2) represents the contrast between the Jordanian and UN conditions, while the contrast between the Palestinian and UN conditions is controlled for. Bootstrap sample size = 5,000. CI = confidence interval; LL = lower limit; UL = upper limit. When zero is not included in the 95% confidence interval, this indicates a significant indirect effect.

Discussion

In line with our predictions, Study 2 found that compared to a control, no-message condition, messages from either Israeli Palestinians or Jordanians successfully promoted reconciliation among Israeli Jews. By contrast, consistent with Study 1, messages from a neutral third party (i.e., the UN) were less effective than identical messages from the other conflict party or from a common-identity third party. Moreover, compared to the no-message condition, messages from a neutral third party did not bring about any positive change in group members’ willingness to reconcile. These findings point to the limited ability of neutral third parties to promote reconciliation.

Unexpectedly, the effectiveness of messages from Jordanians did not differ from that of messages from the other conflict party. Although our participants clearly distinguished between Israeli Palestinian and Jordanian officials, as was evident in their different perceived representativeness (an Israeli Palestinian official was perceived to be more representative of the Israeli Palestinian group than a Jordanian official), messages from both sources brought about similar emotional responses. Nevertheless, these findings are consistent with our theorizing that third parties that share a common identity with the other conflict party should evoke similar emotional responses as the conflict party itself and give rise to optimism regarding common-identity third parties’ ability to promote reconciliation.

The results of the mediation analysis were in line with our theorizing, revealing that the differences in the effects of the different message sources stemmed from the extent to which the official who conveyed the conciliatory message was perceived as representing the other conflict party. These results highlight the unique nature of intergroup as opposed to interpersonal reconciliation processes. Specifically, reconciliation following interpersonal transgressions is expected to be carried out through direct communication between victim and perpetrator (Benziman, 2009). Supporting this reasoning, intervention by third parties following interpersonal transgressions was found to be generally ineffective, even when they shared a common identity with the other conflict party (e.g., family members; Shnabel et al., 2014). By contrast, because groups cannot conduct direct communication in the same way individuals do, their members may expect any form of conciliatory message to be delivered through official representatives (see Tavuchis, 1991). Apparently, this expectation opens the door to interventions by third parties, as long as they are perceived to represent the other conflict party at least to some degree.

One potential limitation of Study 2 is that the neutral third party was the UN, an international organization with almost 200 member states, which may have been perceived as belonging to a higher level, superordinate category (Mummendey & Wenzel, 1999) compared to the single state representing the common-identity third party. Our choice of this particular operationalization stemmed from our wish to enhance ecological validity. A third-party state that is completely uninvolved in the Middle East conflict was unlikely to convey the emotional-subjective messages that were the focus of the present study, compromising the cover story. Importantly, Study 2’s findings regarding the relative ineffectiveness of neutral third parties replicated the results of Study 1, where the neutral third party did not belong to a higher level, superordinate category.

General Discussion

According to Rubin (1980), “The history of third party intervention is as old as the history of human conflict, which in turn is probably as old as the history of humankind itself” (p. 379). The present research explored whether intervention by third parties can facilitate reconciliation. We conducted two studies using substantially different contexts in terms of transgression severity (violent vs. nonviolent), time, and relevance to an ongoing or single-episode conflict. In support of our hypotheses, both studies found that third parties perceived as neutral were relatively ineffective in fostering reconciliation. Importantly, however, we showed that third-party intervention contributed to group members’ willingness to reconcile insofar as the third parties shared critical identity features with the other conflict party and were perceived as representing it, at least to some degree. Future research might investigate further conditional factors, such as how approachable and generally cooperative the third party appears.

Our findings stand in stark contrast to previous findings within the conflict resolution domain, which highlight the importance of third-party neutrality (e.g., Stulberg, 1987). Indeed, neutrality is an asset when third parties use strategies such as conciliation, mediation, and consultation (Rubin, Pruitt, & Kim, 1994) to promote practical agreements between adversarial groups. Moreover, neutral third parties may set the stage for reconciliation by facilitating communication between the parties to a conflict (Kolb, 1987) and creating an open atmosphere in which they would feel free to express their feelings (Fisher, 1983). Reconciliation itself, however, requires something beyond that: it requires “the reparation of a broken relationship,” which can be obtained through the exchange of “emotionally charged communication” (see the analysis of the South African Truth and Reconciliation Commissions by Gobodo-Madikizela, 2008). Our findings that such emotional communication is effective only through parties whose social identity is directly related to the conflict support the tenet that reconciliation is a socioemotional process that is fundamentally different from conflict resolution (see Kelman, 2008; Nadler et al., 2008; Nadler & Shnabel, 2008). The socioemotional path seems especially important in cases in which both groups have an interest in a future relationship.

By empirically testing the impact of third-party intervention on reconciliation, the present research offers several theoretical and practical insights. Conflict parties often avoid the risk involved in exchanging conciliatory messages out of fear that the other party might not reciprocate, or even exploit their goodwill gestures (Shnabel & Noor, 2012). Moreover, even if groups are willing to take the risk, their conciliatory messages might lead to emotional backlash (e.g., anger) should they fail to bring about the hoped-for positive response (Harth, Hornsey, & Barlow, 2011). For third parties, the risk involved in conveying conciliatory messages is considerably lower; they may therefore be more willing to convey conciliatory messages. Identifying the conditions under which third-party interventions are effective, as in the present research, may guide third parties’ efforts to overcome the emotional barriers between conflict parties that often find themselves unable to untangle the proverbial Gordian knot on their own.

In particular, our findings point to the identity-related characteristics of third parties that may bolster or hinder their interventions’ effectiveness. Future research may examine whether neutral third parties can become more effective by stressing shared identity features associating them with a common group with the other conflict party. For example, shared humanity was found to induce a sense of common identity among groups that did not share salient identity features such as race or ethnicity (Wohl & Branscombe, 2005). Our findings also suggest that the effectiveness of messages conveyed by officials belonging to a third party should increase if they stress their identification with and representativeness of the other conflict party. Future research should explore how groups in conflict may motivate third parties to engage in the process of reconciliation and deliver conciliatory messages on their behalf. Possibly, highlighting aspects of a shared identity might help the directly involved conflict parties recruit exactly the type of third parties whose intervention is likely to be most effective; namely, third parties who share common identity with them.

From a broader theoretical perspective, much of the research conducted within the social identity perspective has focused almost exclusively on the ingroup–outgroup dyad. For instance, the common ingroup identity model (CIIM; Gaertner, & Dovidio, 2000;2012) has focused on how recategorization of an ingroup and a given outgroup as belonging to a common superordinate group (i.e., as sharing a common identity) may improve relations between them. However, intergroup relations, and intergroup reconciliation in particular, often involve more than two social actors. For example, international sanctions against South Africa may have facilitated the fall of the apartheid regime and the beginning of reconciliation between Blacks and Whites. Our research suggests that this narrow focus on dyadic intergroup relations may limit our ability to understand and possibly contribute to the promotion of positive intergroup relations.

Optimistically, however, our research also suggests that the insights gained in such ingroup–outgroup dyad studies are invaluable to the understanding of intergroup relations more broadly. Thus, in the present research we borrowed CIIM’s general logic that perceiving groups as sharing a common identity should elicit similar responses towards these groups, and successfully applied it to predict responses towards third parties. Similarly, we extended the concept of typicality, traditionally viewed as the degree to which a single outgroup member possesses traits that characterize his or her own outgroup (e.g., Wilder, 1984), to denote the degree to which a single outgroup member represents another outgroup, to which he or she does not belong. This extension allowed us to draw on findings in the intergroup contact literature (i.e., Brown et al., 2007; Brown et al., 1999; Wilder, 1984) to successfully predict that perceived representativeness should mediate third-party effectiveness. Future research may continue this trend of applying existing knowledge about social-psychological processes to more complex social contexts which go beyond the “ingroup” or “outgroup.” Hopefully, formulating more sophisticated social psychological models that include more than two social actors (and may therefore correspond more closely to real-world structures) would contribute to the development of novel approaches to managing intergroup relations. The present research may be viewed as a first step in this direction.

Footnotes

Funding

The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under Grant Agreement Number 2934602 (PCIG09-GA-2011-293602).

Notes

References

Allport

(1954). The nature of prejudice. Boston, MA: Beacon Press.

Arad

Carnevale

P. J.

(1994). Partisanship effects in judgments of fairness and trust in third parties in the Palestinian–Israeli conflict. Journal of Conflict Resolution, 38, 423–451. doi:10.1177/0022002794038003004

Bar-Tal

(2000). From intractable conflict through conflict resolution to reconciliation: Psychological analysis. Political Psychology, 21, 351–365. doi:10.1111/0162–895X.00192

Benziman

(2009). Forgive and not forget: The ethics of forgiveness. Jerusalem, Israel: Van Leer. [in Hebrew]

Black

Baumgartner

M. P.

(1983). Toward a theory of the third party. In Boyum

K. O.

Mather

(Eds.), Empirical theories about courts (pp. 84–114). New York, NY: Longman.

Blatz

C. W.

Philpot

(2010). On the outcomes of intergroup apologies: A review. Social and Personality Psychology Compass, 4, 995–1007. doi:10.1111/j.1751-9004.2010.00318.x

Brewer

(2000). Superodinate goals versus superordinate identity as bases of intergroup cooperation. In Capozza

Brown

(Eds.), Social identity processes: Trends in theory and research (pp. 117–132). Thousand Oaks, CA: Sage.

Brown

Eller

Leeds

Stace

(2007). Intergroup contact and intergroup attitudes: A longitudinal study. European Journal of Social Psychology, 37, 692–703. doi:10.1002/ejsp.384

Brown

Vivian

Hewstone

(1999). Changing attitudes through intergroup contact: The effects of group membership salience. European Journal of Social Psychology, 29, 741–764. doi:10.1002/(SICI)1099–0992(199908/09) 29:5/6<741::AID-EJSP972>3.0.CO;2–8

10.

Carnevale

P. J.

Arad

(1996). Bias and impartiality in international mediation. In Bercovitch

(Ed.), Resolving international conflicts: The theory and practice of mediation (pp. 39–53). Boulder, CO: Lynne Rienner.

11.

Carnevale

P. J.

Pruitt

D. G.

(1992). Negotiation and mediation. Annual Review of Psychology, 43, 531–582. doi:10.1146/annurev.ps.43.020192.002531

12.

Conlon

D. E.

Ross

W. H.

(1993). The effects of partisan third parties on negotiator behavior and outcome perceptions. Journal of Applied Psychology, 78, 280–290. doi:10.1037/0021-9010.78.2.280

13.

Dovidio

J. F

Gaertner

S. L.

Loux

(2000). Subjective experiences and intergroup relations: The role of positive affect. In Bless

Forgas

(Eds.), The message within: The role of subjective states in social cognition and behavior (pp. 340–371). Philadelphia, PA: Psychology Press.

14.

Fisher

R. J.

(1983). Third party consultation as a method of intergroup conflict resolution: A review of studies. Journal of Conflict Resolution, 27, 301–334. doi:10.1177/0022002783027002005

15.

Fisher

Ury

(1981). Getting to yes: Negotiating agreement without giving in. New York, NY: Penguin Books.

16.

Gaertner

S. L.

Dovidio

J. F.

(2000). Reducing intergroup bias: The common ingroup identity model. Philadelphia, PA: Psychology Press.

17.

Gaertner

S. L.

Dovidio

J. F.

(2012). Reducing intergroup bias: The common ingroup identity model. In van Lange

P. A. M.

Kruglanski

A. W.

Higgins

E. T.

(Eds.), Handbook of theories of social psychology. Thousand Oaks, CA: Sage.

18.

Gobodo-Madikizela

(2008). Transforming trauma in the aftermath of gross human rights abuses: Making public spaces intimate through the South African Truth and Reconciliation Commission. In Nadler

Malloy

T. E.

Fisher

J. D.

(Eds.), Social psychology of intergroup reconciliation (pp. 57–75). New York, NY: Oxford University Press.

19.

Harth

N. S.

Hornsey

M. J.

Barlow

F. K.

(2011). Emotional responses to rejection of gestures of intergroup reconciliation. Personality and Social Psychology Bulletin, 37, 815–829. doi:10.1037/pac0000064

20.

Hayes

A. F.

Preacher

K. J.

(2013). Statistical mediation analysis with a multicategorical independent variable. British Journal of Mathematical & Statistical Psychology, 67, 451–470. doi:10.1111/bmsp.12028

21.

Kelman

H. C.

(2008). Reconciliation from a social-psychological perspective. In Nadler

Malloy

Fisher

J. D.

(Eds.), Social psychology of intergroup reconciliation (pp. 15–32). New York, NY: Oxford University Press.

22.

Knowles

(1958). Mediation and the psychology of small groups. Law Review, 9, 780–784.

23.

Kolb

D. M.

(1987). Who are organizational third parties and what do they do? In Bazerman

M. H.

Lewicki

R. J.

Sheppard

(Eds.), Research on negotiation in organizations (Vol. 1, pp. 207–221). Greenwich, CT: JAI Press.

24.

Kranso

J. E.

(2004). The United Nations: Confronting the challenge of a global society. Boulder, CO: Lynne Rienner.

25.

McCullough

M. E.

Worthington

E. L.

Jr Rachal

K. C.

(1997). Interpersonal forgiving in close relationships. Journal of Personality and Social Psychology, 73, 321–336. doi:10.1037/0022–3514.75.6.1586

26.

Mummendey

Wenzel

(1999). Social discrimination and tolerance in intergroup relations: Reactions to intergroup difference. Personality and Social Psychology Review, 3, 158–174. doi:10.1207/s15327957pspr0302_4

27.

Nadler

Liviatan

(2006). Intergroup reconciliation: Effects of adversary’s expressions of empathy, responsibility, and recipients’ trust. Personality and Social Psychology Bulletin, 32, 459–470. doi:10.1177/0146167205276431

28.

Nadler

Malloy

T. E.

Fisher

J. D.

(2008). Social psychology of intergroup reconciliation. New York, NY: Oxford University Press.

29.

Nadler

Shnabel

(2008). Intergroup reconciliation: The instrumental and socio-emotional paths and the need-based model of socio-emotional reconciliation. In Nadler

Malloy

T. E.

Fisher

J. D.

(Eds.), Social psychology of intergroup reconciliation (pp. 37–56). New York, NY: Oxford University Press.

30.

Okimoto

T. G.

Wenzel

(2008). The symbolic meaning of transgressions: Towards a unifying framework of justice restoration. In Hegtvedt

K. A.

Clay-Warner

(Eds.), Advances in group processes: Justice (Vol. 25, pp. 291–326). Bingley, UK: Emerald.

31.

Pettigrew

T. F.

(1998). Intergroup contact theory. Annual Review of Psychology, 49, 65–85. doi:10.1146/annurev.psych.49.1.65

32.

Pettigrew

T. F.

Tropp

(2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90, 751–783. doi:10.1037/0022–3514.90.5.751

33.

Ross

(1995). Reactive devaluation in negotiation and conflict resolution. In Arrow

K. J.

Mnookin

R. H.

Ross

Tversky

Wilson

R. B.

(Eds.), Barriers to conflict resolution (pp. 26–42). New York, NY: Norton.

34.

Rubin

J. Z.

(1980). Experimental research on third-party intervention in conflict: Toward some generalizations. Psychological Bulletin, 87, 379–391. doi:10.1037/0033-2909.87.2.379

35.

Rubin

J. Z.

Pruitt

D. G.

Kim

S. H.

(1994). Social conflict: Escalation, stalemate, and settlement (2nd ed.). New York, UK: McGraw-Hill.

36.

Schimmack

(2012). The ironic effect of significant results on the credibility of multiple-study articles. Psychological Methods, 17, 551–566. doi:10.1037/a0029487

37.

Shnabel

Nadler

(2008). A needs-based model of reconciliation: Satisfying the differential emotional needs of victim and perpetrator as a key to promoting reconciliation. Journal of Personality and Social Psychology, 94, 116–132. doi:10.1027/1864–9335/a000148

38.

Shnabel

Nadler

Dovidio

J. F.

(2014). Beyond need satisfaction: Empowering and accepting messages from third parties ineffectively restore trust and consequent reconciliation. European Journal of Social Psychology, 44, 126–140. doi:10.1002/ejsp.2002

39.

Shnabel

Nadler

Ullrich

Dovidio

J. F.

Carmi

(2009). Promoting reconciliation through the satisfaction of the emotional needs of victimized and perpetrating group members: The needs-based model of reconciliation. Personality and Social Psychology Bulletin, 35, 1021–1030. doi:10.1177/0146167209336610

40.

Shnabel

Noor

(2012). Competitive victimhood among Jewish and Palestinian Israelis reflects differential threats to their identities: The perspective of the needs-based model. In Jonas

K. J.

Morton

(Eds.), Restoring civil societies: The psychology of intervention and engagement following crisis (pp. 192–207). Malden, MA: Wiley-Blackwell.

41.

Shnabel

Ullrich

Nadler

Dovidio

J. F.

Aydin

A. L.

(2013). Warm or competent? Improving intergroup relations by addressing threatened identities of advantaged and disadvantaged groups. European Journal of Social Psychology, 43, 482–492. doi:10.1002/ejsp.1975

42.

SimanTov-Nachlieli

Shnabel

(2014). Feeling both victim and perpetrator: Investigating duality within the needs-based model. Personality and Social Psychology Bulletin, 40, 301–314. doi10.1177/0146167213510746

43.

Staub

(2008). Promoting reconciliation after genocide and mass killing in Rwanda and other post-conflict settings. In Nadler

Malloy

T. E.

Fisher

J. D.

(Eds.), Social psychology of intergroup reconciliation (pp. 1–18). New York, NY: Oxford University Press.

44.

Stulberg

J. B.

(1987). Taking charge: Managing conflict. Lexington, MA: Heath.

45.

Subašic

Reynolds

(2009). Beyond “practical” reconciliation: Intergroup inequality and the meaning of non-indigenous identity. Political Psychology, 30, 243–267. doi:10.1111/j.1467-9221.2008.00690.x

46.

Tajfel

Turner

(1979). An integrative theory of intergroup conflict. In Austin

W. G.

Worchel

(Eds.), The social psychology of intergroup relations (pp. 33–47). Monterey, CA: Brooks/Cole.

47.

Tavuchis

(1991). Mea culpa: A sociology of apology and reconciliation. Stanford, CA: Stanford University Press.

48.

Teixeira

C. P.

Demoulin

Yzerbyt

(2011). Choosing the best means to an end: The influence of ingroup goals on the selection of representatives in intergroup negotiations. Journal of Experimental Social Psychology, 47, 228–234. doi:10.1002/ejsp.1909

49.

Tropp

L. R.

Mallett

R. K.

(2011). Moving beyond prejudice reduction: Pathways to positive intergroup relations. Washington, DC: APA.

50.

Turner

J. C.

Hogg

M. A.

Oakes

P. J.

Reicher

S. D.

Wetherell

M. S.

(1987). Rediscovering the social group: A self-categorization theory. Oxford, UK: Blackwell.

51.

Volkan

V. D.

(1985). The need to have enemies and allies: A developmental approach. Political Psychology, 6, 219–247. doi:10.2307/3790902

52.

Welton

G. L.

Pruitt

D. G.

(1987). The mediation process: The effects of mediator bias and disputant power. Personality and Social Psychology Bulletin, 13, 123–133. doi:10.1177/0146167287131012

53.

Wilder

D. A.

(1984). Intergroup contact: The typical member and the exception to the rule. Journal of Experimental Social Psychology, 20, 177–194. doi:10.1016/0022-1031(84)90019-2

54.

Wohl

M. J. A.

Branscombe

N. R.

(2005). Forgiveness and collective guilt assignment to historical perpetrator groups depend on level of social category inclusiveness. Journal of Personality and Social Psychology, 88, 288–303. doi:10.1037/pac0000064

55.

Young

O. R.

(1967). The intermediaries: Third parties in international crises. Princeton, NJ: Princeton University Press.