Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

One of the core concepts in Bar-Tal’s paradigm of understanding groups in intractable conflicts is that of collective narratives (Bar-Tal, 2007). This article presents a theoretical approach and an empirical measure to examine perceptions of collective narratives during a decade in which the political situation moved from peace talks to violence. We looked at the sociopolitical context as a significant factor in the development of empathy, anger, and readiness to legitimate collective narratives of both the “other” and one’s own group.

During the past decades, different theoretical concepts have been developed relating to multidisciplinary aspects of collective narratives: social knowledge, culture, ethos, social representations, historical narratives, social identity, collective memory, and communicative memory. The concepts deal with the way members of a group share a common world of knowledge, beliefs, actions, and emotional patterns, providing a basis for their sense of shared identity (Assmann, 1992; Bar-Tal, 2000; Dougherty, 1985; Griswold, 1994; Liu & Hilton, 2005; Moscovici, 1988; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987).

In social psychology , the collective narrative is defined as a social construct that creates clear connections on a continuum between historical and current events (Bruner, 1990). Collective narratives enable the individual members of the group to understand the social reality and the relationships between their group and other groups and provide personal interpretations that help them to behave and function. These individual interpretations are developed to a great extent by social agents—sometimes overtly and sometimes covertly—that direct the thoughts, the feelings, and the behaviors of each of the group members (Bar-Tal, 2000; Reicher & Hopkins, 2001; Turner et al., 1987; Turner, Oakes, Haslam, & McGarty, 1994). On the macro level, collective narratives are a type of lens through which members of the group perceive past and current events and imbue them with justification and meaning for their social, ideological, and political acts (Bar-Tal, 2000; Sidanius & Pratto, 1999).

The relevant literature differentiates between approaches that focus on the content of group members’ shared narratives and models that center on the processes of narrative structuring by educators, politicians, thinkers, media, and ideologues (Ahonen, 1999). The models that focus on identifying group members’ shared content usually deal with defining the narrative and clarifying the social representation of an event or a particular phenomenon. Thus, for example, some researchers have examined specific representations of the concept of “war” or “peace” among groups in conflict as opposed to groups that are not (Covell, 1999). Other studies have examined representations of leaders or specific historical events (Liu et al., 2005; Von Borries, 1995). In this approach, the collective narratives were studied with an emphasis on content; each study related to a specific group and to a specific conflict. A unique effort at generalizing and developing an applicable measure for various conflicts appeared in an international study, “Youth and History” (Angvik & Von Borries, 1997), which examined perceptions of general historical events among youth in Europe and in the Middle East.

Other approaches center on the processes that form the basis of the narratives, such as social representation theory (Moscovici, 1988) which described the construct of the narrative as a process during which the social dialogue among the group members and the dialogue with the “other” are crystallized. In other words, historical narratives are the same shared social representations that enable group members to understand their shared past and to use it to interpret the present and future, as well as enabling communication between group members and the development of a sense of shared collective identity. Research that has focused on collective narratives has dealt with questions of organizing the system of collective representations, placement of the narratives (e.g., center vs. periphery), and the relationship to social beliefs and behaviors (Liu, 2005; Liu, Sibley, & Huang, 2014).

In this chapter, we will describe a model for research about perceptions of collective narratives (Sagy, Adwan, & Kaplan, 2002) which integrates both approaches described above. The model is directed towards identifying specific content, both cognitive and emotional, in the perceptions of collective narratives of national intractable conflicts. Moreover, the model makes it possible to empirically investigate collective narratives and identify possible changes that can occur over time (Sagy, Ayalon, & Diab, 2011).

We will describe a longitudinal study that has employed this model and an empirical measure to examine perceptions of collective narratives in different historical and political contexts of the intractable Israeli–Palestinian conflict among Jewish and Arab adolescents in Israel. The model examines the relations of the perceptions of “in” and “out” groups and focuses on the levels of the perception of the collective narrative using measures of legitimization, empathy, and anger towards the collective narratives of both groups. Thus, the model is proposed as a measure to explore intergroup relationships, one which is sensitive to changes in the level of perceived intractability of the conflict through different historical periods.

Cognition and Emotion in Perceptions of Narratives

The sociopsychological foundation of perceptions of societal collective narratives includes both cognitive and emotional elements (Moscovici, 1988), which crystallize into a complete integrated perception of the “in” and “out” groups. In accord with this basic assumption, our study examines not only the cognitive perception of the narratives but a variety of emotions that the narratives provoke (empathy, shame, guilt, anger) as well. The model examines these perceptions as expressed among members of the group as individuals. However, the picture we get from this examination allows a diagnosis of trends that are common within the groups at different historical periods.

Although differentiating cognitive and emotional components is sometimes arbitrary and may cause injustice to the description of the real situation (Parkes, 1971), we hope that it may contribute to a more reliable study of the complicated concept of narratives. Knowledge and legitimization of the collective narrative are the two cognitive elements in the model. Members of the social group tend to build their worldview on the basis of their shared social knowledge, coming from shared experiences in the history unique to them (Griswold, 1994). The model’s empirical measure evaluates the level of social knowledge as related to the collective narratives of the in-group and the “out”-group. The reference to knowledge and other aspects of narrative perception are the focus for many political science researchers (e.g., Dolan & Holbrook, 2001). Most of these studies have examined the extent to which individuals do or do not have knowledge of a range of political or historical issues (Von Borries, 1995) and changes in the level of this knowledge over time. Beyond that, as far as we know, the relationship between knowledge and other aspects of narrative perception has not been studied.

The second cognitive component, which is more central to the model, is the level of readiness to legitimize the narrative of the “in” group and of the “out” group. The willingness to legitimize the in-group narrative is an important part of building a national identity and a positive in-group collective image (Kaplovitz, 1990). In contrast, delegitimizing the “other’s” narrative plays a central role in violent conflicts (Bar-Tal & Teichman, 2005). The empirical measure in our model enables assessment of the levels of legitimization and delegitimization in the perception of one’s own group narrative and that of the “other” group in different situations.

Besides its emphasis on the cognitive perception of the collective narratives, the model relates to the emotional repertoire that accompanies these perceptions. The emotional component has been found to be significant in political perception and in particular, in its interaction with cognitive knowledge (Dolan & Holbrook, 2001). Both individuals in society and the society as a collective are likely to develop a collective emotional orientation that lasts over time due to exposure to the same narratives (Bar-Tal, 2007).

The collective emotional orientation is likely to characterize a society or culture when it is integrated with the collective narrative as well as when it is self-preserved by bonding processes that are characteristic of that same society (Markus & Kitayama, 1994). This emotional learning allows for interpretation of the narrative of the in- and out-groups and is passed on via the family and/or by educational, cultural, and political framework (Averill, 1980). Emotional orientation has a special meaning in the context of conflict situations, as, for example, in the collective feelings of fear and hatred in Israeli society or in Northern Ireland (Halperin, Russel, Dweck, & Gross, 2011). Despite the role of emotions as significant components in perceptions of collective narratives, only a few studies have examined particular emotions (Halperin, Sharvit, & Gross, 2011). In the model described here, we attempt to broaden and deepen existing knowledge on the connection between the cognitive perceptions of collective narratives and particular emotions, including both negative (e.g., anger) and positive (e.g., empathy) ones.

We expect that higher correlations between cognitive and emotional elements of collective narrative perception reflect a more coherent attitude. These correlations would be observed in periods with higher tensions between the groups and, mainly, higher perceived intractability of the conflict. Supporting results have been presented in a previous study (Sagy et al., 2011) that examined 8 years of violent intergroup relations.

Development of the Model and the Empirical Measure

In their model, Sagy et al. (2002) focused on the cognitive and emotional elements of the perceptions of one’s own group narrative and the delegitimization and dehumanization of the “other’s” narrative. The model and the measure were developed in the context of the Israeli–Palestinian conflict by a group of Palestinian and Israeli researchers (Sagy et al., 2002), in order to enable a deep understanding of the social processes that take place between members of the two national groups who are living in persistent and violent conflict. The research developed by Sagy and her colleagues (2002) allows empirical examination of the concept in other conflictual contexts such as religious, national, and ethnic groups. Up to now, this model has been studied among national collectives (e.g., Israelis and Palestinian, Sagy et al., 2002): groups that are divided by regime (Palestinian citizens of Israel and Palestinians living in the West Bank, Mana, Sagy, Srour, & Mjally-Knani, 2015), and minority and majority national groups living in Israeli society (Arabs and Jews who are citizens of Israel, Sagy et al., 2011). Two studies focused on religious groups in conflict: One related to Palestinian religious groups of Muslims and Christians in Israel (Mana, Sagy, Srour, & Mjally-Knani, 2012; Srour, Sagy, Mana, & Mjally-Knani, 2013) and the other examined the Israeli Jewish religious conflict between ultraorthodox and national religious communities (Kalagy & Sagy, 2015).

The contextual approach requires that the measure be constructed according to the unique characteristics of the collectives and groups being studied. Moreover, collective contents cannot reflect the shared narrative of all the members of the society studied. Thus, the model suggests examining central contents in the shared social identity which are well recognized by most members of the group. Of course, members of the society accord different degrees of importance or centrality to any specific content, and it is possible to find individual differences among members of the group. Due to the complexity of working with collective narratives, the measure enables assessment of legitimization and accompanying emotions towards specific content in the general collective narrative (Srour et al., 2013).

Based on the DeVellis (2003) model of scale development, the process of developing the questionnaire in each conflict context includes several stages. It begins with focus groups of both parties in the conflict whose participants are asked to describe the relationship between the two collectives from their points of view. Three criteria are used in the selection of the collective narratives that are raised in the focus groups: (1) the narratives are related to a significant historical/political/social events in the past as well as in the present, (2) there is a high level of familiarity of the narrative to the research population, and (3) two contradictory narratives are provided for the same event/issue reflecting the different points of view of each group.

Based on the contents obtained in the focus groups on relevant narratives, the research teams (who are usually members of the examined groups) adjust the themes to specific narrative items. As a result, the questionnaire includes pairs of narratives of each group, each pair concerning a historical, social, or political event. In the questionnaire, two narratives for each of the above events are presented to the participants. They are asked to rate their readiness to legitimize the narrative and their feelings of empathy and anger regarding the narrative of their own group and the narrative of the “other” group on a five-point Likert-style scale (1, don’t agree at all; 5, agree very much).

Here are some examples from the questionnaire in the context of Israeli–Palestinian conflict which includes events like the Balfour Declaration, the Six-Day War, Rabin’s assassination, the First Intifada, the Second Lebanon War, the 2000 Oslo Accords, the Al-Aqsa Intifada, and events in the Arab sector in Israel in 2000, which were presented from the point of view of the two collective narratives (Jewish and Arab narratives). One example is the item about the Balfour declaration—the Israeli narrative is: “Many Israelis (Jewish-Israelis) view the Balfour Declaration as the first international diplomatic recognition of the right of the Jews for a state in the Holy Land.” The Palestinian narrative is: “Many Palestinians view the Balfour Declaration as an unfair and illegitimate promise of the British to the Jews.” Another example is the narrative of 1948. The Israeli narrative is: “Many Jewish Israelis view the 1948 war as an important event marking their survival and independence,” and the Palestinian narrative is: “Many Palestinians view the 1948 war as a disaster/catastrophe.”

The Israeli–Palestinian Intractable Conflict

According to the conceptual paradigm of Bar-Tal (2007), intractable conflicts are defined as protracted, irreconcilable, violent, of zero-sum nature, total, and central; parties involved in such conflicts have an interest in their continuation (see Azar, Jureidini, & McLaurin, 1978; Bar-Tal, 1998, 2013; Goertz & Diehl, 1993; Kriesberg, 1995). Intractable conflicts undermine the security and well-being of societies worldwide (International Crisis Group, 2010).

Bar-Tal (2007) claims that societies involved in an intractable conflict over a long period develop a collective fear orientation because of the ongoing threats and dangers to individuals who are involved. In situations of ongoing conflict, fear leads to perceptual biases about the conflict and about the other side, creating cognitive stagnation and a tendency to avoid risks and bringing about justification of the existing policy (Bar-Tal, 2001; Huddy, Feldman, Capelos, & Provost, 2002). Fear and threat at the individual and collective levels lead to cognitive biases so that events are likely to be processed and interpreted incorrectly in ways that strengthen suspicion, delegitimization, and lack of trust towards the “other” (Bar-Tal, 2001).

Rouhana and Bar-Tal (1998) argued that the Israeli–Palestinian conflict is so resistant to resolution because of the unique characteristics of the opposing narratives, the central one being a perception of exclusive legitimacy. The narratives of the conflict, among both Israelis and Palestinians, have been central to the national identities, are celebrated through myths, monuments, and national ceremonies and are strongly represented in literature and songs as well as in school textbooks (Bar-Tal, 2013; Ben-Amos & Beth-El, 1999).

During the decade investigated in this study (1999–2009), social representations of one’s “own” group and the “other” group significantly changed among cboth groups, due to the changes in the political and social context. For example, the other group moved from “peace partner” to “no partner for peace,” from “choosing peace” to “no choice, war.” We expected that these changes in perceiving the “in” and the “out” groups would be reflected in the perception of both groups of the collective narratives of “in” and “out” groups that expressed their social representations.

In addition to the first stage in 1999–2000 (Sagy et al., 2002), administered when the optimism following the Oslo agreement was strong, three additional stages were conducted: in 2002, during the Second (Al-Aqsa) Intifada when suicide bombers exploded in public areas in Israel and hundreds of Palestinians were killed or injured in military actions all over the West Bank and Gaza; in 2004, after the Israeli army invasion of Palestinian towns in West Bank and massive military operations in the Gaza Strip; and in 2009, after Operation Cast Lead in Gaza, which caused more than a thousand fatalities and thousands of injuries and was initiated as a response to intensive missile attacks on communities and cities in southern Israel.

In the current study, we examined levels of cognitive and emotional perceptions of collective narratives in these different historical–social periods of conflict. We suggest three measures related to narrative perceptions as indicators of the intractability or violence of the conflict:

  1. (a)

    The level of readiness to legitimize the narrative of the “in” and the “out” groups and the levels of empathy and anger towards the narratives of the “in” and the “out” groups are expected to reflect conflict intensity.

  2. (b)

    The gap between in-group and out-group narrative perceptions is also expected to be wider in more violent periods.

  3. (c)

    The correlation between emotional and cognitive ccomponents of the perception of the narratives: the more violent the stage of the conflict, the stronger the correlation as reflecting the coherence of attitudes towards both narratives (Sagy et al., 2011).

Method

Sample

The study was conducted in four stages, during which the questionnaires were distributed to a sample of 5627 high school students, of which 3748 were Jews and 1849 were Arabs. Of the respondents, 2950 were in the 10th grade and 2638 in the 12th grade (see Table 1). The samples in each stage were different and ranged between 545 participants and 1188 in the Jewish samples and between 365 participants and 575 in the Israeli Arab samples. Ages ranged from 15 to 18 among both groups. Careful consideration was taken to provide representation of different types of schools and various types of locations.

Table 1 Demographic characteristics of Israeli Jews and Arabs at the four stages of the study
Full size table

Measures

The Perceptions of Collective Narratives Questionnaire

The questionnaire was formulated by a team of Israeli Jewish and Palestinian researchers.Footnote 1 It included structured questions that presented the mainstream views of the Israeli Zionist and Palestinian narratives concerning particular historical events in the Israeli–Palestinian conflict. Three criteria were followed in the selection of the tested narratives:

  1. 1.

    Level of contradiction—the two research groups hold different narratives regarding the same event.

  2. 2.

    Level of familiarity—the narratives are well known among the research group members.

  3. 3.

    Timeline—the narratives related to significant historical/political/social events in the past (e.g., the Holocaust, the 1948 war) as well as in the present (e.g., the Gaza War in the 2009 study).

Two narratives for each of the chosen events were presented to the participants, one attributed to each group. They were asked to rate their readiness to legitimate and their feelings of empathy and anger regarding the narrative of their own group, and regarding the narrative of the “other” group, on a five-point Likert-style scale (1, don’t agree at all; 5, agree very much).

Different events were selected in each study, taking into consideration the recent events and the length of the questionnaire. Some events appear more frequently in the questionnaire and others less often, based on the team’s evaluation of the narrative as central to the conflict. For example, the Balfour Declaration, which appeared frequently in the first versions of the questionnaire, was eliminated from later versions, while the 1948 war and Holocaust narratives were kept in all versions as very important narratives.

We computed six scores by averaging the respondents’ answers to all statements representing the score (legitimization of all the in-group narratives, legitimization of all the out-group narratives, empathy towards all the in-group narratives, empathy towards all the out-group narratives, anger towards all the in-group narratives, anger towards all the out-group narratives). Reliability of the scales was found to be satisfactory. Cronbach’s alpha of the measures ranged from .60 to .87 for the legitimacy measure in the Jewish samples and from .71 to .89 in the Arab samples. For the empathy measure, Cronbach’s alpha ranged from .59 to .86 in the Jewish sample and from .80 to .89 in the Arab sample. For the anger measure, Cronbach’s alpha ranged from .57 to .88 in the Jewish sample and from .78 to .91 in the Arab samples.

The data was analyzed only after standardizing the scores in each research stage for each presented narrative. This procedure enabled us to draw conclusions about the general tendency of weighing the collective narratives of the in- and “other” groups, regardless of the number and specific narratives that were used in the measure at each stage.

Procedures

The questionnaires were distributed in Hebrew and in Arabic. They were administered to the students in their classrooms during a normal class period and were answered anonymously.

Results

The scores were standardized by calculating a z-score for each study among each national group regarding each element: legitimacy, empathy, and anger towards in-group narratives and out-group narratives. An average of the three elements was calculated after multiplying the anger score with −1, and the gap between in-group and out-group narrative perceptions was calculated for each group by detracting the out-group narrative score from the in-group narrative score.

To test our hypothesis regarding the differences in collective narrative perception during different periods, we conducted a four-way mixed-design analysis of variance (4×2×2×3). Group (Jews or Arabs) and study (2000, 2002, 2004, or 2009) were the between-subject factors, while perception elements (legitimacy, empathy, and anger) and narrative attribution (Israeli or Palestinian narrative) were the within-subject factors. The findings revealed significant main and interaction effects between all factors.

The significant main effect of group (F(1,5410) = 40.91, p < 0.001) refers to the tendency of Jews to assign more legitimacy and feel more empathy and less anger towards collective narratives in general, both of the in-group and the out-group. Moreover, we found a general tendency, among both groups, to perceive the in-group narrative as more legitimate and feel more empathy and less anger towards it rather than the out-group narrative as reflected by the in- vs. out-group narrative main effect (F(1,5410) = 312.32, p < 0.001). The interaction between narrative attribution and group was also significant (F(1,5410) = 3598.79, p < 0.001). This result reveals that the tendency to prefer the in-group narrative is stronger among Arabs than among Jews. The averages of the perception of the three elements among Jews were 0.14 and −0.18 for in- and out-group narratives, respectively, and 0.35 and −0.29 among Arabs (the same order).

The main effect of study period was also significant (F(3,5410) = 37.41, p < 0.001), indicating that narrative perception is affected by the political/social contexts. More interesting is the interaction between this factor and group (F(3,5410) = 24.59, p < 0.001), indicating that both groups did not report similar changes along the four stages. The following presents the results regarding changes along the studies for each element separately (see Table 2 and Figs. 1, 2, and 3).

Table 2 Means of standardized scores of perceptions of collective narrative of in- and out-groups among Jews and Arabs along the four stages
Full size table
Fig. 1
figure 1

Means of z-scores of legitimacy given to in- and out-group narratives among Jews and Arabs. The solid line is the in-group narratives and dashed line is the out-group narratives

Full size image
Fig. 2
figure 2

Means of z-scores of empathy towards in- and out-group narratives among Jews and Arabs. The solid line is the in-group narratives and dashed line is the out-group narratives

Full size image
Fig. 3
figure 3

Means of z-scores of anger towards in- and out-group narrative among Jews and Arabs. The solid line is in-group narrative and dashed line is the out-group narratives

Full size image

Legitimacy

Concerning in-group narratives, Arabs reported a sharp increase in legitimizing the in-group narratives in 2002, and then this tendency was moderated. Jews tended to perceive their in-group narratives as less legitimate in 2002, but they later continued to give more and more legitimacy to their in-group narratives. Concerning out-group narratives, both groups showed a tendency to decrease legitimization over time.

Empathy

Concerning in-group narratives, both groups felt more empathy towards in-group narratives in 2002 than in 2000, but this tendency changed to feel less empathy in the next two studies. Arabs continued to feel low empathy towards the out-group narrative with minimal changes in the consecutive studies, but Jews started with high levels of empathy towards the out-group narrative but reported a continuous decline over time, reaching the level of Arabs’ empathy towards the out-group narratives in 2009.

Anger

Concerning in-group narratives, Arabs reported less anger towards their in-group narratives than Jews in 2000 and 2002, but they reported similar levels in 2004 and 2009. Jews showed a continuous tendency to feel less anger towards in-group narratives in each of the four consecutive studies. On the other hand, while Jews reported a mild increase in feeling anger towards out-group narratives over the years, Arabs reported mixed strong tendencies. Arabs showed an increase in anger towards out-group narratives between 2000 and 2002, but then a decline occurred in 2004 and again an increase in 2009.

In general, both groups showed a tendency to more strongly reject the “others’” group narrative in the recent stages. The only exception was a slight decrease in 2004 among Arabs, but coming back to the 2002 level in 2009. Moreover, both groups showed an increase in adhering to their in-group narratives in comparison with the first study in 2000; the Jewish group showed the highest level of adherence to in-group narratives in 2004 while the Arab group did so in 2002.

Moving to the second measure, gaps between in-group and out-group narrative perceptions could be a good indicator for conflict intensity. Arabs reported significantly higher gaps in all the four stages, whereas the study in 2004 showed the highest gaps in both groups (see Table 2). The strongest change among Arabs occurred between 2000 and 2002, but among Jews the strongest change took place between 2002 and 2004.

The Relationship between Cognitive and Emotional Components of Perception of “Other” Group Narrative

Our third measure, which we hypothesized would change according to different levels of violence in the conflict, was the relationship between cognitive and emotional aspects of the attitude towards the other narrative. Table 3 presents the correlations between z-scores of legitimacy and both empathy and anger towards other group narratives among both groups for the four stages. First, we found stronger correlations between legitimacy and empathy rather than legitimacy and anger among both groups over the four study stages. Moreover, the results suggest that according to these measures, the relations between cognitive and emotional components also became stronger in the more recent stages among both groups, indicating greater coherence between cognitive and emotional aspects of collective narrative perception as the conflict became more violent. The only exception was a mild decrease in the correlation between legitimacy and empathy among Arabs in 2009 as compared with 2004.

Table 3 Pearson correlations of perceived legitimacy to empathy and anger towards the out-group collective narratives among Jews and Arabs along the four study stages
Full size table

Discussion

Our paper presents results from a longitudinal study which took place over 10 years, dealing with the Israeli–Palestinian conflict as it is perceived by the Jewish majority and the Arab minority in Israel. The study examined young people’s perceptions of their own group and the “other” group’s collective narratives. Living in a conflicted area has resulted in different interpretations by each side, of recent as well as of historical events which relate to their group’s belief systems. It is commonly agreed that the sociopsychological repertoire that accompanies intractable conflicts serves as a prism through which society members absorb information about the conflict and interpret their experiences (Bar-Tal, 2007), and that these can lead to a stubborn rigidity of repertoire (Kelman, 2007). In this longitudinal study, we asked if and how this repertoire of interpretations of past and present events, held by adolescents from both sides, changed in accordance with changes in the reality of the conflictual situation between peace talks and violent events. Following the failure of the Oslo Accords in 2000, on the Israeli side, there was an escalation of violence, snipers, and checkpoints in 2002, tanks and invasions in 2004, and air bombing in 2009 and on the Palestinian side, shootings in 2002, suicide bombs in 2004, and missile attacks in 2009.

Our examination concerned three different aspects of collective narrative perceptions. First, we examined interpretations of past and present events as reflected by three measures: readiness to legitimate, feelings of empathy, and feelings of anger towards the narrative of one’s own group and towards that of the “other” group. Second, we investigated the discrepancy between perceptions of the in-group and out-group collective narratives as measured by the gap between the mean of standardized scores of legitimacy, empathy, and anger of both narratives. Third, we explored the relations between cognitive (legitimacy) and emotional (empathy and anger) elements of narrative perceptions as measured by the correlation between legitimacy and anger towards collective narratives of the “other” group. These aspects were examined at four stages of the longitudinal study which, in the political context, moved from peaceful times to violent events.

First, the results reveal a tendency by members of both groups to give more legitimacy and feel more empathy and less anger towards their in-group collective narrative and to give less legitimacy and feel less empathy and more anger towards the out-group collective narratives. This tendency, however, increased over the 10 years of the study among both groups. In general, this tendency was stronger among Arab youth as compared to the Jewish youth. Arabs also showed significantly higher gaps between the perceptions of their in-group and out-group narratives in all four stages. Actually, the differences between Arabs and Jews lessened through the research stages. Although Arabs have continued to show more adherence to the in-group narratives and have rejected those of the out-group in comparison to Jews, these differences have become smaller as time went on, while Israeli Jewish youngsters have gradually become more radical in their perceptions of the two collective narratives and more similar to their Arab counterparts.

It appears that this decade of research, which started with the failure of the peace process and continued with increasing violence, has impacted individuals’ perception of in-group and out-group collective narratives, so that they adhere to the former and reject the latter. The adherence to in-group narratives seems to be an effort to depend on collective belonging as a source of meaning for the challenges and existential fears faced on the national level (Mana et al., 2012; Srour, 2014).On the other hand, the rejection of the other group’s collective narratives might express the social representation of the “other” group as the enemy, inhuman, and seeking to wipe out one’s own group (Moscovici, 1988).

As mentioned, Arab participants presented weaker reactions to changes in the political situation and remained more stable in their perceptions over the decade. This finding could be attributed to their starting point. They were more radical in their collective narrative perception from the early stage during the Oslo talks. In spite of the peace process, Israeli Arabs felt a threat to their identities and existence as a discriminated national minority in Israel (Smooha, 2010). The dual loyalty of this minority both to the Palestinian people and to the Israeli state has become almost impossible in politically violent periods. The collective narrative questionnaire seems to be a sensitive measure in this complex situation.

Examining the gap between the in-group and the out-group perceptions of collective narratives could be another significant reflection of conflict intensity and intractability. We found the highest polarization among Jews in 2004 as they strongly adhered to their own narratives and strongly rejected Palestinian narratives. These results can be understood against the backdrop of the loss of hope for peace, on the one hand, and the intensified violent events and existential fears, on the other. The Israeli Arab minority adolescents, however, reported the greatest adherence to their in-group narrative and the greatest rejection of the out-group’s narratives in the 2002 stage. It seems that the October events in 2000 and the beginning of the Second Intifada are events that can explain this polarization. From the Israeli Arab point of view, the traumatic events of “Land Day” in 1975, during which nine Palestinian Israeli citizens were killed by Israeli security forces in demonstrations against land expropriation in the Galilee, repeated themselves in October 2000; therefore, the social representations of “us” and “them” were reproduced and confirmed by the more recent events, causing a new phase of polarization of perceptions of collective narratives.

These results, however, were changed as a decrease in the gap between perceptions of narratives was found in 2009 among Arab students. This reflects less adherence to the Palestinian narrative rather than more rejection of the Israeli Jewish narrative. It seems that the Arabs have become more critical of their in-group collective narrative rather than accepting the Israeli Jewish majority collective narrative. The change in this stage can be attributed to the Arab minority feelings of shared stress with the Jewish majority during the Lebanon War, when Israeli Arab villages and cities were also attacked by missiles fired from Lebanon (Sagy & Braun-Lewensohn, 2009). We may cautiously suggest that, in that period, the Arab minority experienced less identification with the Palestinian component of their identity but no change in their Israeli one (Smooha, 2010). Overall, it appears that among the Arab adolescents in Israel, in-group narrative perceptions are related to identity construction within complex circles of belonging, from the political situation regarding the Israeli–Palestinian conflict and the regional situation to the minority–majority relations within Israel.

The third measure of perceptions of collective narratives included in the scope of this study is the relationship between the cognitive and emotional elements of the “other” group narrative. This measure seems to reflect the level of coherence of the attitudes towards the “other” (Sagy et al., 2011). We hypothesized that there would be a higher need for coherence in times of threat and stress as a means of achieving a certain closure for the individuals as well as for the collective (Kruglanski, 2004). In wartime, a high level of personal sense of coherence has been found to be a significant resource in coping with stress (Sagy & Braun-Lewensohn, 2009). Thus, we expected higher correlations among the cognitive and emotional elements to appear in violent periods. This hypothesis received support throughout the four stages of the study. We found levels of coherence to be quite different in the two groups of Israeli Jewish and Arab students. The level of coherence in the attitude towards the out-group increased over the years with the escalation of violence. In both groups the lowest measure of coherence was found in the first period during the Oslo talks. It seems that this period of a relative peace, or hopes for peace, “allowed” for a lower level of consistency and more openness to different perspectives in both groups (Sagy et al., 2011). The results of lower coherence during this stage seem to reflect less rigidity of the sociopsychological repertoire and provide the opportunity for social changes needed to proceed with the peace process at the social level. During the following stages, however, which were characterized by more violent events of war and terror, the coherence of the attitudes increased among Israeli Jews. Their coherent perceptions could be a significant resource in coping with the stressful situation. At the same time, this coherent pattern of perceptions could explain the high resistance to change among these youths (Halperin, Russel et al., 2011). As suggested by Sagy et al. (2002), these results also indicate the influence of top–down processes: changing reality and actual social events (peace talks vs. war and terror) lead to changes of perceptions of the in- and out-group among youth who are living in conflicted areas.

Another important issue that requires attention is the differentiation between the two emotions: empathy and anger. Empathy is described in the literature as an emotional-cognitive process (Strayer, 1987) that results in understanding and “feeling with” others, while anger is usually described as a more physiologically related emotion (Harris, Schoenfeld, Gwynne, & Weissler, 1964). Furthermore, anger is evoked by threat against the self, especially when the threat is perceived as an unjust assault that interferes with one’s attempts to realize specific personal goals (Lazarus, 1991). We found that the Israeli Arab adolescents felt less empathy and more anger towards the “other” group narratives than did their Jewish counterparts. This result could be part of the tendency of the neglected and offended minority to reject the majority narrative (Rouhana & Bar-Tal, 1998; see also Nahhas, Chap. 5).

Another differentiation is noteworthy. Over the stages of the longitudinal study, there were relatively higher correlations between legitimacy and empathy towards the “other” collective narratives among Arabs and higher correlation between legitimacy and anger towards the “other” collective narratives among Jews. In other words, it seems that for Arab adolescents, the barrier towards legitimization of the Israeli Jewish collective narratives was associated with a lack of empathy towards those narratives, while the main hindrance of the Israeli Jewish youth towards legitimization of the Palestinian collective narratives was associated with a feeling of anger towards these narratives. The role of empathy and anger in the ability to listen to the “other” life story has been described in several studies relating to Palestinian–Israeli encounters (Sagy, 2000). The ability to empathize with the suffering of the “enemy” and to overcome the anger which arises while listening to the “other” narratives was found to be a contributing factor in enhancing readiness to reconcile (Sagy, 2015). However, as far as we know, the differentiation between these two emotions with relation to the cognitive component of legitimization has not been examined. Our results hint at a different pattern of emotional-cognitive connection among the Israeli Arab and the Israeli Jewish adolescents which can be understood by both cultural and contextual (majority vs. minority) factors. It seems that further research is needed to investigate the different meanings of anger, empathy, and legitimization for these two conflicted groups as well as for other groups in conflict.

Despite these meaningful results, the generalizability of our study conclusions is limited. The limitations are mainly in methodological concerns. The first relates to the question of representativeness of the samples. Although we attempted to carefully consider this question by randomly selecting the schools during the first stage, some of the schools refused or could not participate in the later stages of the research. Second, our sample is limited to high school students whose experience with the “other” group is very limited, and their identity construction process is still in its early stages.

In spite of the methodological limitations of this longitudinal study, it may still indicate some important directions to explore in order to deepen our understanding of perceptions of collective narratives in intergroup relations within the context of intractable conflict.