The Official, The Empathetic and The Critical: Three Approaches to History Teaching and Reconciliation in Israel

Studies of social representations and intergroup conflict often stress the role of collective narratives and historical accounts in prolonging and legitimizing conflict.

In spite of the importance attributed to historical narratives and the acknowledgment of institutionalized history teaching, few studies actually explored the effects of teaching and curricula on learners' intergroup attitudes in conflict-ridden societies. A notable exception is Barton and McCully's (2010) work on the effects of a dual-perspective critical inquiry history curriculum on Protestant and Catholic Northern Irish youth. It appears this curriculum promoted students complex understanding of the other's perspective through an internally persuasive dialogue. Perhaps naturally, while they showed understanding to both sides, learners used curricular contents mainly to enhance and legitimize their in-group standpoint.

History Teaching and Intergroup Attitudes in the Israeli Context
In the Israeli context, few studies were conducted about the effects of history teaching on intergroup relations. Of these, the majority documented the (very rarely implemented) empathetic Dual-Narrative suggested by Bar-On and Adwan (2006). Eid (2010) showed Israeli-Arab students found the Jewish narrative emotionally unacceptable, while Eini ElHadaf (2011) reported that Israeli-Jewish learners appreciated the opportunity to engage with the Palestinian perspective, but they also tended to dismiss it as "emotional and unobjective." On a parallel trajectory, Cohen (2013) claimed that Jewish adolescents studying about the holocaust increased their awareness of minority rights. Arab students and teachers who studied about the holocaust demonstrated increased empathy toward Jews (Abu-Ria 2014; Shoham et al. 2003). Findings, which seem to contradict impressions that holocaust education (especially in its informal activities), promoted xenophobic attitudes (Feldman 2002).
As for other teaching approaches, Pollack (2009, 2015) showed that critical work with conflicting historical sources enabled productive intergroup encounter during online co-construction of historical accounts. Jewish participants managed to contain the threat posed through their Arab participants' assertions by adopting an impartial "academic" role afforded by the critical inquiry approach. With reference to the conventional teaching approach, Peled-Elhanan (2012) made the claim that Israeli-authorized textbooks desensitize young Jewish-Israelis to Palestinian suffering. Analyzing the one-sided, neutralized representation of Israel's role in the conflict, she assumes it leads Jewish-Israeli soldiers to uncompassionate behavior, though she does not supply empirical evidence for the claim. None of the studies compared the effects of competing teaching approaches on intergroup attitudes and interaction in a systematic empirical way. The study described below sought to fill this lacuna. I will present findings and conclusions from the various publications which emanated from it (Goldberg 2014a, b;Goldberg and Gerwin 2013;Goldberg and Ron 2014).

The Curricular Pendulum and Competing Teaching Approaches
During the first decade of the new millennium (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010), history curriculum in Israel has shown contrasting features of innovation and regression, leading to the production of varied and competing curricular materials about the Israeli-Palestinian conflict (Goldberg and Gerwin 2013;Goldberg and Ron 2014). On the one hand, this is the period in which Sami Adwan and Dan Bar-On, along with a group of Jewish-Israeli and Palestinian teachers, formulated a dual-narrative textbook (Adwan and Bar-On 2004). Teaching with this curriculum was based on mutual acknowledgment and affirmation, nonjudgmental listening and perspective-taking (Bar-On and Adwan 2006). In the same decade, the higher-order thinking reform in Israeli education called for promotion of critical thinking and disciplinary practices (Zohar 2009). In history subject, a new curriculum appeared, calling for engagement in historiographical controversies (Israeli Ministry of Education 2008a). Curriculum introduced new and sensitive topics such as the debate on the responsibility for the Palestinian refugee problem (Domke et al. 2009;Israeli Ministry of Education 2008b;Stern et al. 2007).
On the other hand, these innovations elicited strong reactions and a conservative backlash. A new education minister attempted to ban the teaching of the Palestinian perspective in Israeli schools (Kashti 2009(Kashti , 2010. The history subject superintendent issued guidelines to present a "clear explanation of the Palestinian exodus" stressing "Palestinian and Arab leaders' responsibility." While noting the existence of debate on the causes and responsibility for the refugee question, the superintendent's site offered a set of sources stressing Israeli righteousness as a basis for teaching the historical controversy on the topic (Yaron 2009(Yaron , 2010. Officials and conservative politicians issued vehement publicized denunciations of multiple perspective teaching. However, it appears that quite a few Jewish-Israeli teachers still insist on teaching the Palestinian narrative along the Israeli one (Blumenfeld 2015;Goldberg, submitted). Their commitment to helping their students forms an informed and complex understanding of the conflict in the face of students and officials' hostility situating such teachers as "risk takers" (Kitson and McCully 2005).
These vacillations created in fact at least three competing (though not simultaneous or equally supported by authorities) curricular approaches for teaching the history of the Israeli-Palestinian conflict: the later official approach, which stressed a single clear narrative, with a conventional textbook-oriented teaching, and a stress on in-group (Israeli) righteousness; the empathetic dual-narrative approach, which stressed perspectivetaking and nonjudgmental acknowledgment of both sides' narratives; and the educational reform for higher-order thinking approach, which stressed critical disciplinary thinking and engagement in historical controversy evaluating and synthesizing conflicting historical accounts of both sides.
The above curricula demonstrate different features of engagement with the past, which should lead to different effects according to research on collective narratives and intergroup conflict. The official single narrative appears to replicate tendencies for self-justifying and exonerating cognitions and for intergroup attribution bias, which protract and normalize conflict (Bar-Tal and Halperin 2011; Doosje and Branscombe 2003;Roccas and Berlin 2015). The empathetic dual-narrative approach creates conditions for mutual affirmation and for intergroup empathy, reducing competitive victimhood and rejection of threatening out-group perspectives and promote reconciliatory attitudes (Čehajić-Clancy et al. 2011;Vollhardt 2013). The critical disciplinary thinking approach is assumed to curb bias and exonerating cognitions (Roccas et al. 2006), help learners take a critical stance to self-legitimizing narratives, and promote ability to contain complexity and disagreement (McCully 2011).

Comparing the Effects of Competing Curricula of Conflict
This variety of teaching approaches offered an opportunity for tracking the effects of teaching the history of intergroup conflict in a comparative empirical method. Using the competing curricula, three parallel teaching units were created, focusing on the Jewish-Arab war of 1948 and the birth of the Palestinian refugee problem ("Independence War" in Israeli terminology and the "Naqba" [catastrophe] in Palestinian terms). Hundred and seven Jewish and 82 Arab-Israeli high school students (aged 16-18) were randomly allocated to study the topic using one of the three teaching units (genders, ethnicities and political affiliations 1 distributed equally between conditions). Learning in all three approaches lasted about 45 min, consisting of a preparatory presentation, reading aloud of text and individual assignments. Participants studied the unit in their hometowns, guided by a research assistant who was a native speaker of their mother tongue and all materials were adapted to learners' native language. Two weeks prior to and immediately after this learning intervention, participants wrote short compositions, in response to questions about the causes of the war, of the ensuing Palestinian exodus and about responsibility for the Palestinian refugee problem. The narrative participants wrote allowed us to track their preconceptions and changes in the perceived responsibility of their in-group for the harmful outcomes. 2 Along with the compositions, learners filled a mode of social identification questionnaire (Roccas and Berlin 2015), which taps individuals' level of chauvinistic glorification and patriotic attachment to their nation. 3 Social identification is assumed to impact acceptance of collective responsibility. Conservative policy makers also claimed that encounter with out-group narratives would undermine national identification. Learners also filled questionnaires about intergroup attitudes such as interest in the other's perspective of the conflict (Bar-Tal and Halperin 2011) 4 and defense of in-group narratives (Klar and Baram 2016). 5 For detailed description of procedure, materials and measures, see Goldberg and Ron (2014) and Goldberg (2014a, b).
We shall now outline the various effects that history teaching approaches had on intergroup attitudes, perceptions of the conflict (mainly in-group responsibility for conflict-related harm) and intergroup interaction.
rejeCtioN, iNterest aNd respoNsibility: effeCts of History teaCHiNg oN iNtergroup attitudes

Official Rejection and Empathetic Interest: Effects on Interest in the Other's Perspective
History teaching approach had a significant impact on learners' interest in the other side's perspective (See Table 1). As Goldberg (2014b, p. 459) shows, repeated-measures ANOVA revealed an interaction effect of time and condition (F(2.163) = 6.33, p = 0.02, η 2 = 0.05). In the conventional single-narrative teaching approach, learners' interest in the other's perspective decreased, while in the empathetic dual-narrative condition it increased. In the critical condition, interest in the other's perspective remained comparatively stable. An interaction effect was also found for time, condition and national group (F(2.163) = 4.79, p = 0.03, η 2 = 0.04). It showed that the effect of approach on interest in the other's perspective was more pronounced among Arab participants. This may be due to the fact their perspective was not represented in the conventional single-narrative approach, which was based on Israeli official narrative (Goldberg 2014b, p. 460).
These results show that history teaching approach can increase (or decrease) the motivation to take out-group perspectives, an aspect of intergroup empathy be predictive of conflict resolution (Gehlbach 2004). Empathetic engagement with both in-group and out-group narratives had significant positive effect on minority members, perhaps due to their stronger need for acknowledgment and affirmation (Shnabel et al. 2009). Minority members studying the conventional single (majority) narrative experienced a pronounced decline in interest in the majority perspective, apparently in defensive reaction to the silencing of their voice (Yonah 2008).
Repeated-measures ANOVAs revealed no significant interaction effects of time and condition (or time, condition and national group) on modes of social identification and defense of in-group narratives (F's(2.173) = 0.08−0.65, p's > 0.15). None of the teaching approaches caused a significant change in learners' glorification and attachment modes of social identification or their defense of in-group narratives. Nor did the effects of teaching approaches differ significantly. Thus, we can see that, regardless of teaching method, studying the other's perspective on a major historical issue in the conflict did not undermine individuals' identification with their group (whether in the form of patriotic attachment or chauvinistic glorification). It also showed that general commitment to in-group narrative did not falter due to encounter with out-group narrative.

Accepting Responsibility and Curbing Bias? History Teaching Effects on Perception of In-Group Responsibility
Perceived in-group responsibility (and the frequently accompanying collective guilt) is associated with reconciliatory intergroup attitudes. While the conventional single-narrative approach had no effect on learners' perception of in-group responsibility, the other two alternative history teaching approaches had contradictory effects on Arab and Jewish learners. In the empathetic dual-narrative approach, perceived in-group responsibility decreased among Jewish and increased among Arab participants. In the critical condition, perceived in-group responsibility increased among Jewish participants, a pronounced difference in direction and degree from the change occurring in the empathetic dual-narrative condition. We should note that change within each condition was not significant (Goldberg 2014b). The effect on Arab participants may show the power of affirmation in answering the needs of a weaker party in a conflict, as proposed above (Shnabel et al. 2009). However, the inverse effect on Jewish participants is yet to be explained. Nonjudgmental, mutually affirmative exposure to the Palestinian narrative, which stressed Jewish responsibility, should lead Jewish learners to accept, rather than reject, responsibility as it did with their Arab peers. The comparatively increased acceptance of responsibility by majority members in the critical disciplinary approach contradicts normal assumptions about "confirmation bias," which should have led participants to reject the information. However, results align with Roccas et al. (2006) and McCully's (2011) assumptions. It also hints that "impartial" academic practice, as a path for intergroup dialogue, is more accessible to majority members. A finding parallels to Kolikant and Pollack's (Kolikant and Pollack 2009) work on Jewish and Arab learners' online dialogue.
What were the factors that facilitated or impeded acceptance of ingroup responsibility. A bivariate correlation was computed with all relevant factors, and two factors were found to have a significant correlation with acceptance of responsibility (See Table 2). Learners' interest in the other side's perspective was associated with their perception of ingroup responsibility. Teaching approach moderated this relation, which was found to be strongest in the empathetic dual-narrative approach and negligible in the critical disciplinary approach (Goldberg 2014a). This may be related to the stress of the empathetic dual-narrative approach on taking the other's perspective. An undertaking assumed to be highly dependent on individuals' interest in the other's perspective.
Teaching approach also moderated the impact of political affiliation on responsibility. In general, political partisanship and polarization cause selective adoption of information and entrenchment, thwarting the effect of engagement with new information or with challenging perspectives (Bar-Tal and Halperin 2011; Bennett and Iyengar 2008). However, looking at the effect of political affiliation on in-group responsibility within each teaching approach, we find wide variations. Following the learning intervention, a more liberal political affiliation was associated with higher perceived in-group responsibility only in the conventional and empathetic conditions (r = 0.60, p < 0.001, r = 0.31, p < 0.05, respectively), while in the critical condition the relation was nonsignificant (r = 0.10, p = 0.48). To ascertain moderation effect, a structural equation modeling AMOS 21 software was used to compare a model, in which the association of political affiliation with perceived responsibility differed across conditions, to a model in which a cross-condition equality constraint was imposed over the regression weights (Kline 2011, p. 286;Rigdon 1998). Bootstrapping was performed over the model using 1000 iterations. The first model showed good fit indices (NFI = 0.98, CFI = 0.99, RMSEA = 0.03), while an alternative model in which critical disciplinary and conventional single-narrative conditions were constrained to be equal gave a significantly lower fit (NFI = 0.77, CFI = 0.72, RMSEA = 0.17; ΔNFI = 0.23, χ 2 = 6.91, p < 0.01).
In the conventional teaching approach the effect of political affiliation on acceptance of responsibility increased following learning, while in the critical disciplinary condition it decreased (prior to the learning intervention the relation of political affiliation to responsibility in the conventional condition was r = 0.21, p = 0.14; while in the critical condition r = 0.28, p = 0.04). We may infer that conventional teaching enhances the political bias while the critical approach curbs it.
As we have shown, history teaching approach affected (and moderated the associations of) interest in the other side's perspective and perceived in-group responsibility, both of which are assumed to promote reconciliatory attitudes. Reconciliatory attitudes should influence intergroup interactions. Consequently, we found teaching approach has indeed affected actual intergroup interaction as represented by Jewish and Arab learners' deliberation of the conflict's history and resolution.

How We Learn and How We Talk: Effects on Intergroup Interaction
Following the first, individual learning study, participants were invited to participate in a follow-up study about the same topic, involving intergroup encounter and dialogue. Some 130 of the participants of the individual learning study proceeded to engage in dyadic intergroup discussion about the Jewish-Arab conflict. Participants were matched by teaching approach, supplied with the materials they studied in the individual learning study and instructed to discuss and reach joint decisions as to the responsibility and solution for the Palestinian refugee problem. Decisions, or points of disagreement in cases of impasse, were to be recorded in writing, to promote commitment to the task and approximate a negotiation situation. Discussions were conducted in Hebrew (a language both groups speak and understand but Jews speak considerably more fluently) facilitated and recorded by participants, transcribed and analyzed. For a detailed description of procedure, materials and measures, see Goldberg and Ron (2014) and Goldberg (2014a).
Transcripts were analyzed to track intergroup equality of status or dominance in discussion, a precondition of intergroup encounter success (Pettigrew 1998) and the general atmosphere of discussion in terms of opposition and collaboration, as a measure of intergroup behavior, rather than simply intergroup attitudes (Pettigrew 2008).
Dominance was analyzed along the lines adopted by Maoz (2001). We analyzed dominance in the use of time and in the control of discussion. For dominance over time, we computed for Jewish and Arab participants in each pair the percentage of their words out of the total number of words uttered in discussion. For control of discussion, we coded all instances in which a participant gave instructions, changed the topic, initiated procedures or asked intrusive questions. Discussion style or atmosphere was analyzed using a shortened version of Bales' (1976) Interaction Process Analysis to assess discussion style. We coded each discussant's utterance in relation to the other discussant's previous utterance as Rejection, Opposition, Compliance or Elaborative agreement. Discussion outcome was assessed on the basis of discussants agreement (or impasse) on a joint answer as to each of the two questions they discussed.
An MANOVA performed over domination of discussion time and control of discussion with teaching approach as between-subjects factor revealed a small multivariate effect for teaching approach (F(6) = 2.48, p = 0.028, η 2 = 0.12) (Goldberg and Ron 2014, p. 14). As Table 3 shows, discussions carried out among participants who studied in the empathetic dual-narrative condition featured a significantly lower Jewish dominance of discussion time than a control and the conventionalauthoritative conditions. The critical disciplinary condition featured a significantly lower Jewish dominating behavior in discussion than the control and the conventional-authoritative conditions (Goldberg and Ron 2014). In both cases, it appears the exposure to both sides' perspectives promoted a more egalitarian discussion atmosphere. A condition considered essential for successful intergroup encounter (Pettigrew 1998).
This atmosphere apparently led to more collaborative deliberation of the conflict, both in terms of process and in terms of outcome. The proportion of elaborative (in contrast to oppositional) utterances was higher among groups of learners who studied in the two multi-perspective teaching approaches (see Table 4). Collaborative discussion atmosphere, as indicated by the proportion of agreement to opposition utterances, predicted the frequency of achieving a joint decision on historical responsibility (Estimate(B) = 3.22, β(S.E.) = 1.17 (0.42), Wald = 7.76, p = 0.005). Consequently, critical disciplinary teaching had a significant positive effect on the frequency of joint decisions on historical responsibility. The conventional single-narrative teaching approach had a significantly negative effect on the frequency of finding joint solutions to refugee problem as compared to the critical disciplinary approach and to a control group (Goldberg and Ron 2014;Goldberg in press).
Perceived in-group responsibility (which was affected, as mentioned above, by teaching approach) also promoted more collaborative deliberation atmosphere. Having calculated each discussants proportion of agreement and opposition utterances, we could check the relation of a discussants perceived in-group responsibility for the harsh outcomes of the conflict with the acknowledgment of such responsibility in discussion and with out-group peer opposition and agreement in discussion (see Table 5 for means and bivariate correlations). Jewish participants' acknowledgement of in-group responsibility was inversely correlated with Arab peers' opposition (r = −0.35, p < 0.01).
Jewish participants' perceived responsibility was associated with more frequent agreement utterances, and acknowledgment of responsibility among Jewish discussants, which led in turn to more collaborative reactions from Arab participants (see Fig. 1). The relation was not symmetrical (Arab participants did not increase in-group responsibility due to encounter with historical perspectives, nor did they impact Jewish participants collaboration). This actor-partner interaction aligns with the assumptions as to the effect of the stronger party's acknowledgment of Table 4 Mean percentage of agreement and opposition utterances by condition and frequency of agreement on solution for the refugee problem, by condition Adapted from Goldberg and Ron (2014, p. 14), Goldberg (2016)     responsibility on weaker party's reconciliatory attitudes (Shnabel et al. 2009).

disCussioN
We now attempt to sum up the various findings as to the effects of history teaching approaches on intergroup attitude, perception of in-group responsibility and actual intergroup interaction. Conventional singlenarrative teaching reduced interest in the other's perspective, while empathetic dual-narrative teaching increased it, especially among Arab learners. Teaching approaches also moderated the effect of interest in the other's perspective and political affiliation on perceived in-group responsibility. Critical disciplinary teaching curbed these biasing influences, while conventional single narrative, for example, enhanced political affiliation's hold on acceptance of responsibility. We should note teaching approaches did not undermine or even affect learners' social identification or frequency of acceptance of threatening out-group perspectives. Consequently, conventional teaching also had significant negative effect on intergroup deliberation of the conflict and its history, in terms of egalitarian discourse, collaborative atmosphere and joint decision making. Teaching approaches also moderated the effects of individual attitudes on intergroup discussion. The (negative) effect of in-group glorification and defense of in-group narratives on intergroup interaction were markedly stronger in the conventional single-narrative teaching.
These findings align to a large degree with Liu et al.'s (2014a, b) and Barton and McCullys (2010) claims as to the effects of exclusive historical narratives on intergroup attitudes and relations. It appears that conventional history teaching curriculum's representation of the conflict through a single narrative, even when it contains some self-critical information, might be detrimental to intergroup relations. The way learners engage with the information is apparently just as important as the information itself. Such findings align with social psychological research on the effects of critical thinking prompts and affirmation (Čehajić-Clancy et al. 2011;Roccas et al. 2006;Vollhardt 2013). In this respect, both the critical and the empathetic teaching methods harbor greater promise for improving intergroup relations if used systematically. It is therefore quite alarming that both Israeli and Palestinian sides attempt to discourage such teaching methods (Goldberg and Gerwin 2013;Rohde 2014).
However, as, at least on the Israeli side, "risk takers" go on teaching both sides' perspectives (Goldberg, submitted) whether empathetically or critically, it is worth reflecting on their complex effects. The findings on the positive effects of empathetic dual-narrative teaching on Israeli-Arab learners' perspective-taking motivation are reassuring and stand in contrast to former findings (Eid 2010;Rohde 2013). Arab learners may have perceived the teaching of both narrative as Jewish acknowledgment of the Palestinian narrative and the Israeli responsibility it stresses. An acknowledgment assumed to answer the unique needs of the weaker party in asymmetric power relations (Shnabel et al. 2009). Jewish participants' tendency to reduce perceived in-group responsibility may be a reaction to the fact the Palestinian narrative contained no expressions of empathy with Jewish suffering or humanizing views of Jews, assumed necessary by Schnabel et al.'s needs based on model of reconciliation. The empathetic approach also curbed the negative effects of Jewish participants' glorification mode of social identification and defense of ingroup narratives on actual intergroup collaborative discussion. It may be that perspective-taking helps the dominant groups' members overcome the push of in-glorification sentiments toward dominating discussion and antagonizing out-group members.
The effects of the critical disciplinary on Jewish participants in terms of perceived in-group responsibility, and lowered domination of discussion, align to some degree with Kolikant and Pollack's (2009) findings that Jewish participants collaborated better with Arab partners when assuming a more detached academic role. Findings also substantiate Barton and McCully's (2010) claims as to the positive effects of the critical inquiry curriculum and extend them from the realm of internal dialogue to the realm of actual intergroup interaction. As we noted‚ the critical disciplinary approach appears to curb the influence of political affiliation on in-group responsibility. This offers a hope of breaking through the entrenchment and rejection of information, caused by political polarization.
We should take the above conclusions cautiously. First, the sample is quite small, and the intervention was short and extracurricular. Another factor limiting the generalizability of findings is the voluntary nature of the sample, which could cause a self-selection bias, hinted by a higher proportion of liberal affiliated learners compared to national average. However, we should bear in mind that self-selection occurred in most intergroup encounters studies. In favor of the method, we should note this study is the only one currently known to the author in Israel (and actually for that matter also abroad) in which students were randomly allocated to teaching conditions, allowing for empirical comparisons with the conventional teaching approach. Allocation to groups was also performed within each school, thereby curbing to some degree the strong effect of school culture and values.
The implications of the studies seem quite straightforward. Teaching multiple perspectives has a potential to serve the goals of increasing intergroup perspective-taking motivation and improving intergroup deliberation of conflict. This outcome does not risk learners' national identification and esteem, a risk, which apparently may have detained educators and decision makers from engaging in such teaching. However, the prospects for such initiative current political climate in Israel do not seem promising. It may actually be that current decision makers find reduced motivation for out-group perspective-taking a positive outcome of the conventional teaching approach. autHor biograpHy Tsafrir Goldberg is a lecturer in the Faculty of Education in University of Haifa, Israel. His research focuses on the relations of history teaching, Iidentity and intergroup relations. He led research projects on the teaching of the Jewish-Arab conflict and its impact on leaners identity, intergroup attitudes and intergroup interaction, on history teaching and Islamophobia and on teachers' attitudes towards difficult histories. He is involved in teacher education and curricular reform.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.