“It's in My Veins”: Identity and Disciplinary Practice in Students' Discussions of a Historical Issue

Abstract

Learners' identity is considered a resource, but is also assumed to conflict with impartial history learning practices. This empirical study explores the relationship between learners' social identity and their historical practices and understanding. Sixty-four Jewish-Israeli 12th-grade students of Mizrahi and Ashkenazi ethnicities studied a historical controversy concerning the relations between the two ethnic groups. Participants' discussions were analyzed to trace the impact of social identity on historical learning and the application of disciplinary practice. Findings attest to the impact of social identity. Participants frequently approached evidence and applied empathy in ways benefitting in-group image. However, social identity and intergroup interaction also motivated elaboration of arguments and disciplinary practices. Implications for engaging diverse students' identities in history teaching are discussed.

Keywords:

• historical understanding
• social cognition
• social identity
• ethnic identity
• collaborative learning

Studies of history education have often acknowledged the impact of identity and emotion on learning. However, they seldom apply the expansive body of relevant knowledge on social identity and social cognition. Social cognition research has utilized a variety of historical accounts to prompt identity related reactions but has seldom analyzed these reactions as learning situations. This study seeks to remedy this disconnect by integrating social cognition and social identity theories into an analysis of history learning. Tracking the way identity and cognition interact furnishes the basis for exploring how their mutual influence may both bias learning and facilitate it.

DISCIPLINARY STANDARDS, RATIONAL INQUIRY, AND SOCIAL IDENTITY

In recent decades, growing attention has been given to the discipline-oriented aspects of history teaching in Western democratic societies (Israeli Ministry of Education, 2010; National Center for History in the Schools, 2004; National Curriculum Online, 2004). Students are expected to practice “critical reading of texts … analysis of text within its context … understanding the other's perspective and evaluating people by action regardless of their affiliation” (Israeli Ministry of Education, 2010, pp. 8–9). These practices are viewed as primarily rational and seem to demand some emotional restraint and impartiality. For example, Barton and Levstik (2004, pp. 69–70) noted that the “analytic stance,” which dominates current history curriculum reforms, is often seen as “the standard for best history education” because of its stress on rational elements. They contrasted it with the more emotive stances that scholars view as running the risk of “abuse of history” due to identification and partisan loyalties.

Foster (2001) claimed, for example, that the practice of historical perspective taking should not involve emotional identification or imagination. Wineburg, Mosborg, Porat, and Duncan (2007) showed how current emotional-judgmental attitudes impede the use and understanding of historical evidence. VanSledright (2001) gently criticized the way a teacher's practice of empathy was motivated by her feeling of collective guilt, an emotion related to social identity (Roccas, Klar, & Liviatan, 2006). Finally, Goldberg, Schwarz, and Porat (2008) showed that students' ethnic identities biased their evidence evaluation when studying charged topics that were relevant to their ethnic group.

It would seem, then, that commitment to one's social group would be an anathema to rational historical inquiry and disciplinary standards, especially when dealing with charged topics. The following sections provide a brief outline of the literature on the impact of identity on learning history and the notion of social identity and social cognition as sources of bias and then—by way of transition to the current study—explore how social identity and historical controversies bearing on students' identity may actually offer the potential for enhancing cognitive development, rather than only harming or inhibiting it.

Social Identity and “Difficult Histories”

Controversial social issues are considered promising starting points for argumentation, debate, and learning through social interaction (Hess & Posselt, 2002; Schwarz, 2009). As collaborative learning becomes a more prominent item on the educational agenda (Waggoner, Chinn, Yi, & Anderson, 1995), such motivations for interaction are sought in order to enliven learning and promote students' reasoning. However, when controversies are relevant to students' identity, issues become charged, and learning may be impacted by identity and emotion (King, 2009).

Learning about charged historical issues in an ethnically diverse setting, as most public education systems in developed countries have become, may be expected to arouse even further identity awareness and emotion. Studies in multicultural settings, such as the United States and Ireland, show that history teachers tend to avoid controversial issues for fear of the emotional excesses that could be triggered (Barton & McCully, 2005; Gregg & Leinhardt, 2002; Levstik, 2000). As Sheppard (2010) showed, discussing “difficult histories” may trigger strong negative feelings among students of diverse ethnicities, creating barriers to students' engagement and historical understanding. King (2009) documented how even students who voluntarily joined intergroup encounters tended to evade or denigrate out-group peers' accounts of contentious issues.

Social identity theory (Tajfel & Turner, 1986) would also predict that studying the history of interethnic relations in an ethnically diverse setting could have strong (negative) socio-emotional implications. The theory holds that individuals constantly engage in the cognitive process of social categorization (Arab versus Jew, Black versus White, Red Sox fans versus Yankees fans)—usually following salient characteristics or situational cues—a process that arouses the emotional process of social identification. That is, we tend to view ourselves as belonging to one of the accessible categories or groups (referred to in social identity terminology as the individual's “in-group”).

Social identification connects the feeling of self-esteem to in-group esteem. My team's win is a boost to my self-esteem and mood; a negative news item about my nation may be taken as an insult. This motivates social comparison in which the individual, many times unconsciously, seeks points of comparison and evidence that will result in higher status and esteem for the in-group compared to the “out-group.” Such a comparison enables maintaining what is termed as “positive identity”—a feeling that “we” are at least as good (if not better) than “they” are (Tajfel & Turner, 1986).

The theory acknowledges that individuals belong to more than one social category or group and that esteem may be shared by groups, but it holds that, while social categorization and identification may be situational, they tend to be quite dichotomous in each situation. Discussing an issue of interethnic relations in an ethnically diverse setting would make the ethnic category more salient, promoting categorization and identification along these lines. Research in social identity also shows that social comparison operates with relative group status to a large degree as a zero-sum game—the gain of one at the expense of the other (Johnston & Hewstone, 1990). In this analysis, heightened in-group identification and the need for positive identity motivates representation of in-group historical members and action in a better light, leading learners' discussion of historical events and evidence to manifest social cognition biases.

Social Identity and Cognition

Social cognition theory assumes that cognition is motivated and constrained by identity and mechanisms related to it (Fiske & Taylor, 2008). Intergroup bias research shows that individuals processing information about in-group or out-group members almost invariably favor the in-group. Bias will more frequently occur in indirect forms, such as in-group favoritism in the attribution of emotions and application of empathy, but may also guide actual derogation of the out-group (Fiske & Taylor, 2008; Hewstone, Rubin, & Willis, 2002). Identification with one's group and the need to protect in-group esteem seem to motivate the tendency to assess evidence emanating from or favorable to in-group members as more reliable (Voss, Wiley, Ciarrochi, Foltz, & Silfies, 1996).

Furthermore, such social identity needs seem to result in selective sampling—the tendency not to choose evidence unfavorable to the in-group or to portray it as a non-representative “sample” by stressing in-group variability (Doosje, Branscombe, Spears, & Manstead, 1998). In addition, identity needs and emotion are seen to govern our theories of mind (Batson, 2009), resulting in intergroup attributional bias. The common “attribution error” is the tendency to attribute negative intention to out-group members' harmful actions but attribute the negative actions of in-group members to external or contextual factors (Fiske & Taylor, 2008; Hewstone, 1990).

We have seen the influence of identity and the social context of intergroup relations on perception discussed in the literature to date mainly as impediments. Since historical accounts of intergroup relations frequently carry implications for group esteem, we might, on the basis of previous literature, assume that social identity would also negatively impact the way individuals perceive and process these accounts. Consequently identity may impede learning. However, it is worth exploring an alternative possibility.

Controversy and Identity as Resources for Cognitive Development

Escalation of bias and even emotions may not be the only expected outcome of discussing controversial issues among peers. As Reznitskaya and Anderson (2006) showed, peer debate can also mediate and promote the use of productive academic practices. Collaborative learning studies present evidence of dissemination among peers of argumentative stratagems, such as reference to evidence and use of multiple perspectives. Under certain emotional conditions, especially when one's esteem is affirmed, even threatening information can be accepted and used (Sherman & Cohen, 2002).

In the field of history education, encounters with peers' opposite opinions and evidence may also cause cognitive conflict, which can trigger conceptual change (Limon, 2001). Furthermore, discussions of socially charged issues can foster interest (Del Favero, Boscolo, Vidotto, & Vicentini, 2007) that promotes engagement in disciplinary inquiry through such practices as evidence use, evaluation, and historical perspective taking. Since academic debate is an essential feature of historians' community of practice, participation in group debates about historical issues serves as a kind of induction into that community (DiCamillo & Pace, 2010), fostering higher-order thinking and more rigorous—and thus less biased—practice (Wineburg & Martin, 2004). For example, focus on historical disciplinary practices is a strategy utilized by Northern Irish teachers to overcome obstacles and biases accompanying controversial intergroup issues (King, 2009).

Identity and emotion do not necessarily harm historical understanding. Barton (2009) pointed to identity as an essential motivator and claimed that evading charged identity issues denies learners' needs and renders history teaching meaningless. Gadamer (1975) advocated a merging of horizons between the interpreter's positional identity and the historical source or agent. Endacott (2010) claimed that affective empathy does not impede cognitive empathy and at times may serve to mediate it. Students in conflicted societies show interest in encountering the other's perspective and utilize it to construct their identities (Barton & McCully, 2005). Furthermore, Johnson and Johnson (2007), in their “creative controversy” approach, claimed that conflict in desegregated multicultural settings may be socially and cognitively productive. King (2009) reported that intergroup encounters in Northern Ireland aroused group awareness and loyalty but also promoted students' ability to take new perspectives and challenge their own understanding of intergroup conflict (on the condition that caring relationships were fostered). DiCamillo and Pace (2010) demonstrated how encountering the voice of marginalized groups in an ethnically diverse history class harnessed the motivations of students and led to a more robust disciplinary stance.

Collaborative learning of the history of intergroup relations may clearly attune to the prerequisite conditions for productive encounter, which intergroup contact studies identify (Pettigrew, 1998). These include, for example, a joint task, the promotion of an over-arching “disciplinary” identity, and an atmosphere of equality stemming from exposure to varied perspectives. However, research to date has focused on the outcomes of these preconditions with respect to prejudice reduction not on their relationship to the learning processes. Despite the advancement of theories about identity and intergroup contact as resources for education, their influence in actual learning contexts is still underexplored (La Belle, 1994; Ogbu, 1992; Sfard & Prusak, 2005). These issues should be given special attention in cases where the historical processes that shape students' ethnic identities are studied and discussed by the students themselves (Epstein, 1998; Seixas, 1997). As we shall see, Israeli history and the Israeli high school history curriculum offer such an exemplary case.

Diversity and Identity in Israel's History and in the Israeli History Curriculum

In the Israeli 12th-grade history curriculum, students learn about the shaping of their own and others' ethnic identities through the topic of the Israeli “Melting Pot” policy. Promotion of Jewish immigration (“aliyah,” or ascent, its literal meaning hinting at its positive significance) has been a national goal of the state of Israel since its establishment (Israeli Declaration of Independence, 1948). Israel received the highest proportion of immigrants in the world during its first 50 years, with immigrant students at times constituting triple the number of native Israelis (Israeli Central Bureau of Statistics, 2004). One of the main instruments for absorbing immigrants has been the educational system (Resh & Kfir, 2004), in which the Melting Pot policy of cultural absorption was implemented during the state's mass immigration era. The most prominent example of the Melting Pot policy were the “uniform education” institutes established in the vast immigration camps (Zameret, 2002).

Under the Melting Pot policy, educators directed Jewish immigrants to break away from their traditional cultures and adapt to the ideal type of the “new Jew.” This ideal type, a secular Hebrew-speaking productive pioneer/settler, capable of taking arms to protect his homeland, stood in contrast to many of the immigrants' traditions. Immigrant children were urged to change their names, learn new dances and pastimes, and adopt modernized/secularized Hebrew culture. The policy led both to some great achievements and to deep resentments that have resulted in enduring social and political conflicts (Zameret, 2002).

Because immigration absorption policies, such as the Melting Pot policy, affected Jewish immigrants from Europe and from the Middle East differently, they often became the foci of charged interethnic relations. Tense relations developed between Ashkenazi-European Jews, who constituted the majority of Israel's “founding fathers,” and Mizrahi-Oriental Jews, who were a substantial minority of the new immigrants (Kimmerling, 2001). These tensions both shaped the formation of Jewish ethnicities and fueled political rifts for decades—some say until the present (Lissak, 1999).

Mizrahi-Ashkenazi socio-economic and cultural differences also persist, as evidenced by differing collective memories, educational achievement, and even marriage “markets” (Ya'ar, 2005). During the first decades of Israel's existence, socio-economic gaps and settlement patterns created a de facto Mizrahi-Ashkenazi segregation in schooling as well, although since the 1970s, a massive desegregation initiative was implemented, encompassing more than 70% of the Jewish non-religious public schools (Resh & Kfir, 2004). In spite of the generally declared support for interethnic Jewish integration among both the general public and policy makers (Smooha, 2004), Mizrahi-Ashkenazi ethnic tension still persists, albeit at a simmer, erupting periodically. One of its evident expressions is in the fact that references to historical ethnic stereotypes or discriminatory practices arouse strong emotions among both groups (Smooha, 2004; Tsafati, 1999).

As is usually the case with conflict-ridden issues, historical accounts and interpretations of events vary (Schwartz, 1996). Conflicting perspectives about the Melting Pot policy make a rich context for debate and learning, a condition assumed to be beneficial for the development of skills for argumentation (Schwarz, 2009). In the last decade, the Melting Pot policy has itself become a mandatory topic in the state school history curriculum, and knowledge of it is assessed by matriculation exams (Israeli Ministry of Education, 1999). Thus, due to the desegregation of state schools, students of Ashkenazi and Mizrahi ethnicities together learn about and examine a historical controversy relevant to their social identities. The learning of this socially charged and identity-relevant curricular topic provides the context for the present study, enabling the exploration of the impact of social identity on the learning of history.

Goals of Research

This study sought to trace the impact of social identity on historical understanding. Based on the assumption that identity and bias might be revealed implicitly (Fiske, 2002; Hewstone, 1990), the goal of the research was to map and categorize instances in which social identity could be said to influence cognitive moves and disciplinary practices.

RESEARCH QUESTIONS

The questions addressed in this study follow.

1.	Is social identity expressed in students' discussions of a charged, social identity-relevant historical issue, and if so, how?
2.	Does social identity impact evidence evaluation, argumentation, and historical empathy in such discussions?

METHOD

Participants

Participants were 64 12th-grade students from two consecutive cohorts. Thirty-two participants were Ashkenazi (descendants of Jews from Western countries) and 32 were Mizrahi (descendants of Jews from Muslim or Asian countries). ¹ Thirty-eight were male and 26 female, with gender approximately evenly distributed between ethnicities. Participation was voluntary. I invited the students (I was teaching in their school but not in their grade) to participate in an “interesting learning activity for a university research project” about a topic that was included in their matriculation curriculum.

Research was performed in a desegregated school situated in a low socio-economic status neighborhood of an Israeli city. Beyond the local inhabitants, the school also attracts students from outside the neighborhood due to its vision of integration and community values and its comparatively high academic standards. Most of the students from the school neighborhood come from low-income, politically and religiously conservative Mizrahi families. Their parents or grandparents immigrated to Israel from Muslim countries during the mass immigration of the 1950s, the era of the Melting Pot policy. Many of the students from outside the neighborhood come from middle class, liberal/left, secular Ashkenazi families. In many cases, their families were affiliated with Israel's founding elites and the labor movement, which initiated the Melting Pot policy. To some extent, then, the study participants coming together in the school setting may be seen as embodying the drama of mass immigration and of the Melting Pot policy, and issues of culture and identity related to the topic could be assumed to carry special significance, particularly in this school environment.

Although cases of intergroup tension occasionally occurred at school and were even referred to in some of the participants' discussion, they were apparently comparatively rare. Intergroup contact in the school was substantial and protracted, and it included many opportunities for collaboration and friendship. However, ethnicity and intergroup contact was seldom the topic of explicit discussion or debate. In the context of this study, participants of both Jewish ethnicities, Mizrahi and Ashkenazi, who had already studied together for years, now focused on a topic that made ethnic categories salient and ethnic relations controversial.

Procedure

The discussion excerpts and data presented here were taken from a research project about students' historical thinking and collective memory. The project centered on learning tasks designed to promote argumentation and historical understanding (Rouet, Perfetti, Favart, & Marron, 1998; Voss & Wiley, 1999) and serve as an induction into a community of practice (Lave & Wenger, 1991). Learning tasks followed a structured protocol described next.

At the outset of the educational intervention, participants were asked to write down their opinion about the Melting Pot policy. Then, following a coaching exercise in evidence evaluation delivered by an experimenter, participants read and evaluated a series of eight conflicting sources about the Melting Pot policy from which they were asked to extract information about the goals and outcomes of the policy (see Appendix). Then, in groups of two or three, they engaged in student-led discussions, which were recorded and transcribed, and from which the excerpts analyzed in this article are taken. For a full description of the procedure, see Goldberg, Schwarz, and Porat (2011).

Students participated in discussions at times compatible with their school schedules and were initially paired to create ethnically mixed dyads or triads. Due to the inconsistencies of high school (and adolescent) life, unplanned attendance or absence resulted in 17 of these groups being ethnically mixed and 9 homogenous (involving 42 and 21 students, respectively). Students received a discussion instruction sheet containing the two opinion questions that they had already addressed individually in writing:

1.	Did the immigrants progress and adapt with the help of the Melting Pot policy and the “standard education” institutions, or were they damaged and discriminated against because of them?
2.	Were the Melting Pot policy and the “standard education” institutions essential steps that contributed to the construction of the State of Israel, or were they a destructive political takeover?

These were “either/or” questions designed to promote debate and argument (Schwarz, 2009).

The discussion instruction sheet directed participants to try to convince each other of their points of view and to summarize in writing their points of agreement and differences at the end of their discussion. Students were instructed explicitly to back up their arguments and points of view with reasons and evidence from the sources they studied. The conflicting sources that participants had evaluated in the previous phase were at their disposal. Students were told that their discussions would remain confidential and that if excerpts were published, anonymity would be fully maintained. Discussions were self-facilitated and without intervention by experimenters or other adults. At the beginning of each discussion, the experimenter switched on the audiotape and left the room; at the end of their discussion, participants switched off the tape and returned it to the experimenter.

Analysis

Impact of Identity on Historical Disciplinary Practices

This category of analysis identified students' utterances about the historical topic that appear to be influenced by the speaker's identity. These utterances were assessed regarding the implications they had for Ashkenazi or Mizrahi group esteem and moral status. If the utterance impacted the speaker's in-group image in a positive way or out-group image in a negative way, it was considered to be driven by social identity needs (Hewstone, 1990; Johnston & Hewstone, 1990).

Utterances in which Ashkenazi participants presented the Melting Pot policy or its Ashkenazi initiators in positive terms or the Mizrahi immigrants in negative terms were identified as expressions of the impact of participants' identity, since both could be seen as serving Ashkenazi group esteem. Utterances in which Mizrahi participants presented the Melting Pot policy or its initiators in negative terms or presented the Mizrahi immigrants in positive terms were also identified as expressions of the impact of participants' identity, since both could be seen as serving Mizrahi group esteem.

As the literature on rational inquiry, social identity, and social cognition implies, identity is assumed to impact historical disciplinary practices. Therefore, our analysis focused on expressions of the impact of participants' identity that occurred while students were engaged in disciplinary practices, such as working with evidence and applying empathy.

Historical Disciplinary Practices

Historical practices were conceptualized as cognitive moves that are congruent with academic disciplinary practice (Wineburg & Fournier, 1994) and enable more elaborate problem solving (Lee & Ashby, 2000). As such, the level of practice was assessed based on congruence with disciplinary norms or elaboration of the problem/solution.

In the following pages, the focus of analysis will be on the impact of students' social identity on the practices of evidence evaluation, evidence use, and empathy (or perspective taking). The choice to concentrate particularly on these practices stems first from their centrality in current curricular reforms and goals (Barton, 2009; Davis, 2001). Additionally, these practices appear to be quite prone to social identity influence according to social cognition studies. The coding categories of evidence use, evaluation, and empathy were predefined based on current history education research. The fourth practice analyzed, termed here as historical distinction, was as an emergent theme identified in the process of analyzing discussion.

Each instance of historical practice was coded as an independent event. Two independent coders identified cases of historical empathy, evidence use, evaluation, and (in a second round of analysis) historical distinction. Initial inter-coder agreement was above 85%. Differences were settled through discussion and categories refined to achieve full agreement.

Evidence use

This practice refers to reasoning on the apparent basis of evidence. Under this focus of analysis, all citations and explicit references were included to evidence, such as “it says here,” “it is written,” and “they say,” but also more vague allusions, such as “I remember reading” and “historians say,” when these could be clearly traced to specific content in the sources leaflet. Evidence can be used with differing levels of sophistication. At lower levels, it is perceived and referred to as direct access to and true information about the past, while at the higher levels, evidence is seen as motivated testimony, as pieces in a puzzle, and eventually, as the basis for interpretation and synthesis (Lee & Ashby, 2000).

Evidence evaluation

Evidence evaluation is the process of “sourcing” by which the learner determines in what way and to what degree the information contained in a source might be biased or flawed (Britt, Wiemer-Hastings, Larson, & Perfetti, 2004; Wineburg & Fournier, 1994). References or allusion to the credibility of a source, an interpretation, or a historical agent's account were coded as evidence evaluation. These could consist of explicit evaluation (“this looks like denial” or “you can hear what he [really] thinks”) or implicit praise or criticism, such as “everything has an excuse.” Evidence evaluation also reveals the level of sophistication in students' conceptions of historical evidence (Lee & Ashby, 2000; Wineburg & Fournier, 1994), from an assumption that all evidence is true, through a skeptical approach to all evidence as biased testimony, to a nuanced understanding that takes into account the context of the evidence's creation and to other pieces of evidence.

Historical empathy

This category refers to the ability to put yourself in “the other's shoes” and to reconstruct historical agency in terms of mentality, context, and intentionality (Barton & Levstik, 2004; Davis, 2001). Attributions of intention, opinion, or perception to historical figures or groups may all be seen as expressions of empathy. They may feature different levels of disciplinary proficiency, ranging from an assumption of the moral or intellectual deficiency of agents through an assertion of agents' attributes based on contextualized evidence-based inference of their perceptions (Lee & Ashby, 2000). Any references to historical agents' feelings, intentions, or perceptions were thus coded. This included direct and schematic attribution of intentions (such as “they had a wish to erase culture”), inferential mental reconstruction (such as “imagine you were an immigrant” and “let's say I [as an immigrant] don't want my son to dance”), and use of evidence about agents' perceptions (“they believed dance was essential” and “they say here that folkdance was all that helped them hold on”).

Historical distinction

This fourth practice emerged as a separate analytic category, as coders repeatedly noticed instances in which participants articulated historical-moral judgments of the Melting Pot policy's outcomes. These instances were grouped together and their definition refined and redefined along the lines of grounded theory analysis (Glaser & Strauss, 1967). Historical distinction was coded as an instance in which a participant distinguished between intentions and results, goals and means, long- and short-term consequences, subgroups within social groups or organizations, or between differing historical perspectives. These differentiations parallel Shemilt's (2000)concept of narrative frameworks, according to which as learners' narrative understanding develops, they can process and produce more complex narratives. At the higher level, narrative frameworks become more multi-dimensional and polythetic. Developed frameworks encompass the possible divergence of intentions and consequences and the differing interpretations espoused by members of different groups or eras.

FINDINGS

The quantitative analysis features mainly descriptive statistics presenting the frequency of phenomena of identity impact. Wherever relevant, in ferential statistics were added that pointed to the significance of their proportions. Qualitative analysis will focus on particularly telling examples and on contrasting extremes within the previously discussed learning practices. They are arranged as instances of identity-motivated versus impartial utterances and low-level versus high-level practices. Examples of progress or intermediate levels (where those existed) accompany these vignettes.

Quantitative Results

The quantitative analysis points to a marginal impact of identity. Identity-serving instances outnumber instances of more impartial practice, but not to a very large extent. Twenty-eight of the participants made a total of 55 references to sources. Evidence was used in an identity-serving way in 32 (58%) of the cases. Of 89 applications of empathy, 53 (60%) were coded as in-group serving. In both cases, the difference from a normal distribution was non-significant (χ²(1, 55) = 1.47, p = .22 and χ²(1, 89) = 3.25, p = .07, respectively), indicating that identity did not clearly dominate participants' practices. Of a total of 92 instances of historical distinction, 61 (65%) were coded as identity serving. Here the difference in frequency between identity-serving and impartial distinctions was found to be significant (χ²(1, 92) = 8.52, p = .004), attesting to a frequent impact of identity needs.

As mentioned in the Method section, while most discussion groups were ethnically mixed, about one-third of the participants ended up performing the discussion in homogenous groups. Social identity theory leads us to assume that encounters with out-group members heighten in-group awareness and loyalty, as it makes the ethnic social category more salient. This means a stronger impact of identity on learners' practices should be expected in the ethnically mixed groups. Comparing the ethnically mixed and homogenous groups can thus be seen as an additional way of exploring the effect of identity and emotion on students' practice. Groups' heterogeneity or homogeneity had an effect on historical distinction. The mean difference between the number of identity-serving and impartial distinctions used by each participant was significantly lower in the mixed groups (M(SD) = .34 (1.45)) than in the homogenous groups (M(SD) = 1.66 (1.87); t(39) = 2.39, p = .022). Participants in the mixed groups used identity-serving distinctions almost to the same degree as impartial distinctions, while in homogeneous groups, identity-serving distinctions clearly outnumbered impartial distinctions. This finding can be interpreted as evidence that working in ethnically mixed groups curbed the tendency for social identity bias. This finding is quite surprising, because participants were expected to present more identity motivated practices in ethnically mixed groups. Moreover, historical distinction was the practice in which the prevalence of identity-serving instances was most pronounced.

Qualitative Analysis

Evidence Use: Multiple Accounts, Uncertainty, Reliance on Identity

Although engagement with evidence constituted a significant proportion of the learning activity preceding the discussion phase, evidence does not feature frequently in student discussions. Less than half (29) of the 64 participants referred explicitly to knowledge as taken from sources in ways that conform to the category of evidence use.

The infrequent use of evidence may result from the fact that sources contradicted each other and featured different levels of reliability. As Omer, a grandson of Iraqi immigrants, said, “When I read the sources I got terribly confused … because some said it was helpful and some stressed the negative sides. … ” Tom, the grandson of a prominent Ashkenazi Labor Party leader, expressed the uncertainty sources aroused: “You have these excerpts saying this and the opposite, like, you can't exactly rely on them … maybe the sources are reliable and maybe the sources are unreliable.”

However, those participants expressing uncertainty about sources did, after all, refer to evidence. On the other hand, it seems that in some cases, students ignored the need to rely on the evidence provided because of a sense of certainty in their knowledge—a certainty rooted in identity and group allegiance. Thus Leah, a granddaughter of Mizrahi (Kurdish) immigrants, declared to her peer: “You really found who to talk to about these things. It runs in my veins!” However, not once in her quite expansive condemnation of the Melting Pot policy (430 words in 16 turns of conversation) did she refer to a historical or informative source, whether those at her hand or others. It might be that the feeling of close ties to the topic and agents, mediated through social identity, made reference to actual evidence seem redundant. Conversely, Niv, a son of a well-to-do Ashkenazi family, participating in another discussion group, distrusts his knowledge due to lacking the “right” identity: “Say, me, as perhaps those who didn't … like, those who didn't sit in the immigrant camps … you know … European Jews and that, then, maybe I don't know.” This qualification and expression of lack of authority expresses a similar conviction: that knowledge of the past is transmitted largely through group identity rather than gained through analysis of sources.

Students frequently referred to sources in ways corresponding to the lower levels of understanding the use of evidence, that is, as providing direct access to the past or as information. Most discussants who referred to sources (17 of 28) used only one source, treating a piece of evidence in isolation in spite of being clearly aware of the existence of multiple conflicting sources. This may have to do with the fact that evidence was frequently used to support an in-group apologetic stand, a case in which it would seem counter-effective to complement a source with a piece of conflicting evidence.

However, defending in-group esteem did, in some cases, lead to a more complicated juxtaposition of sources. Raya, a granddaughter of Iraqi immigrants, offers a persistent critique of the Melting Pot policy during her conversation with her Ashkenazi peer, Hanah. Still, she referred to four different sources to encompass the outcomes and motives of Ashkenazi officials' actions, starting out with what she sees as a direct account of events (Source 3): “There's this committee report were they like really say the immigrants' complaints, what hurt them … even if, in the eyes of those who did it, they're speaking only about folkdance.” It should be noted that even when Raya refers to evidence as direct information, she notices the varied significance historical agents attributed to the evidence. Now Raya proceeds to rely on a historian's interpretation of the motive for action (Source 7): “Then there's this source, Tsameret, who says they [Ashkenazi Labor movement leaders] feared the religious block will want a theocracy.” Not relying solely on the historian's interpretation, she seeks to support it with a primary source (Source 4): “ … there's the political aspect, like where is it? … [R]ight, that he [director of education] says religious guys represent the forces of darkness and you can't let them teach the immigrants. … ” While this evidence could suffice to make her claim of anti-religious political motivation, she still turns to corroborate it with another source derived from a political opponent of the policy (Source 2): “On the other hand it doesn't seem to me it was so extreme as Pinkas [a religious politician] says; comparing it to the inquisition and murder. … ” Although the religious politician's speech aligns with Raya's general stance, she is surprisingly cautious, qualifying its validity as she tries to incorporate the pieces of evidence together.

This use of evidence in an interconnected approach, corroborating and duly qualifying information from various pieces, reflects a high-level stance to evidence (Lee & Ashby, 2000). However, it is noticeable that Raya does little to explicitly assess the reliability of evidence. Indeed, evidence evaluation was absent from most cases of evidence use, and when it occurred, evaluation tended to reflect learners' identity needs.

Evidence Evaluation: Between Identity Needs and Disciplinary Practice

Of the participants referring to evidence at all, only a few made explicit reference to its credibility. When students referred to evidence credibility, it was usually in order to reject information that might bear negatively on in-group esteem. This happened both when evidence was directly unflattering to in-group members and when it had favorable implications for the out-group.

Evidence evaluation sometimes takes very crude forms, such as derogation of the author of a source favorable to the out-group. Gadi, an Ashkenazi student and son of a retired regiment commander, is preoccupied with the notion of military threat as the reason for a consolidating cultural policy. He tries to introduce into the discussion a historian presenting a parallel, moderately favorable view of the Melting Pot policy (Source 8 in Appendix) by saying “But the question is whether it was the right time to change the [dominant, established Ashkenazi] culture … like that historian here at the end says. … ” Yakov, his Mizrahi peer (who claimed Israeli culture should have been changed to suit the Mizrahi immigrants) immediately interrupted him: “That one [the historian] is dumb!”

However, Yakov later proceeds to a somewhat more advanced epistemological stance: a skeptical attitude that aligns to some degree with the view of all evidence as biased testimony (Lee & Ashby, 2000). When Gadi refers to Ashkenazi educators' testimonies about the motives of their actions (Source 3, investigation committee report), Yakov discredits them, saying “Everything has an excuse, you know.” This is an overgeneralization about the unreliability and dishonesty of public utterances but one that approaches critical thinking assumptions. Gadi follows up with the logical extension of Yakov's epistemological skepticism, claiming “You can't know what happened there.” Here we find that students' identity-driven dispute facilitates a mutual elaboration of their stance toward evidence, albeit to a point that somewhat annuls the utility of evidence use.

Ironically, Yakov then backs away from his relativist-critical approach by referring to evidence from the very same source (Source 3) as speaking for itself. A Mizrahi immigrant child's testimony is regarded as a means of unbiased direct access to the past. Yakov seems to feel he can converse with historical agents and receive full and accurate account of events: “Listen. You can ask the children. The children say: [quotes the source]. Teacher said everybody who wishes to go on the trip must cut his side locks [which would be worn by boys from traditionalist communities]. … ” This depiction of the Ashkenazi educator as manipulating immigrant children into parting with their traditional religious attributes definitely sheds a negative light on the Melting Pot policy initiators.

In response, Gadi, the Ashkenazi discussant, does not cling to his former skeptical epistemological stance (regarding the impossibility of knowing what happened in the past). Instead, he again immediately attunes to Yakov's epistemological stance, attempting to refute Yakov's argument with piece of evidence also taken as self-evidently true. Gadi quotes Source 5, a personal reminiscence of an immigrant looking back at his childhood experience from the distance of 40 years: “On the other hand, Itsik says … just a second … [flipping pages]. … There was no alternative … you can't go through this process of turning into an Israeli without total-Israelization… .”

As they discuss the question whether harsh treatment of the immigrants was justified, these two participants ignore the fact that each source may in some way be biased. A child's testimony and certainly an adult's reminiscence may be seen as self-legitimating accounts. Students' use of verbatim quotations without any qualification or evaluation represents a regression in the stance toward evidence. It is seen how evidence evaluation and the epistemological stances behind it are negotiated. Discussants switch stances to evidence and evaluate its credibility in response to each other and in the service of identity needs.

The evaluation of evidence, however, tells not only a story of bias. Students' identities may have motivated them to find flaws in the reliability of evidence threatening in-group esteem. On the other hand, through debate with peers, they also developed their disciplinary practices, using them in more impartial ways, as seen from the following example.

Tamar, a female Ashkenazi student, discusses with two other Ashkenazi male peers, Nathan and Gil, a source bearing highly negative implications for in-group esteem (Source 4). In this source, the Ashkenazi director of education's public exclamations are contradicted by a transcript of a closed Labor Party forum where he describes his own work as a “war for political influence over immigration.” Presentation of the Melting Pot policy as self-serving partisan politics would definitely harm the Ashkenazi founding fathers' image and Ashkenazi moral esteem.

Tamar refers to this possible interpretation, resisting the notion: “As for political takeover- I SO don't think it's true.” Both boys support her. Nathan echoes “I too don't think it's true,” as does Gil. These reactions seem to denote group loyalty and attempts to maintain group esteem. Still, Tamar demonstrates a critical disciplinary practice, rating the evidence based on the cues taught in the evidence evaluation coaching exercise. “Like maybe his motives were [political] … as he says there in the closed forum transcript? There's this one from the Labor Party that in a closed forum you can hear like what he thinks.”

The context of creation (closed-party forum at the time of events) and the type of document (transcript) serve as bases for a high reliability rating. Tamar mentions explicitly the characteristics of the document and infers from them that it is credible: “You can hear what he thinks.” Such a line of reasoning seems to express acquisition of the disciplinary practice of “sourcing” (Britt et al., 2004; Wineburg & Fournier, 1994). It also overcomes the tendency to downgrade the reliability of evidence that threatens in-group esteem. Tamar's engagement in discussion of an issue that is relevant to her social identity promoted a disciplinary practice alongside the arousal of group loyalty.

As we have seen, Tamar tries to “hear what he [the education officer] thinks.” Indeed, in many cases, discussants attempted to reconstruct historical agents' perceptions and intentions. These attempts will be discussed in the next section.

Attribution of Intention, Attributional Bias, and Historical Empathy

In historical inquiry, since most of the actors are inaccessible to direct interrogation, we can only attribute motives and intentions to them. As noted in the introduction, the attribution process is prone to many biases (Fiske & Taylor, 2008; Hewstone, 1990). As will be seen, in fact, in many cases students' attempts at reconstructing historical agents' intentions and consciousness feature intergroup attributional biases—a bias characterized by the tendency to attribute negative intentions to out-group members' actions.

Students frequently attributed intentionality to historical actions of out-group members. Shiri, granddaughter of Iraqi Jewish immigrants, has just criticized her Ashkenazi peer Yo'el for his approach to her ancestors traditions: “These things you say outrage me … it may be a sign of primitivism in your opinion, but for me it's a thing you can't look down upon.” Now she stresses that the Melting Pot policy's Ashkenazi initiators “wish to erase these [Mizrahi] traditions.” Beyond the clearly negative depiction of educational outcome, she stresses that outcome was intended, using the word “wish,” implicitly downgrading the out-group's moral status. Mordi, a Mizrahi participant and grandson of Moroccan immigrants, explains the Melting Pot initiators' actions on the basis of their being “kind of ethnocentric people who preferred their own culture.” In both cases, the historical agents are implicitly depicted as morally deficient, and their mentality is reconstructed using general attributes (“ethnocentric,” “wish to erase”). This seems to fit with what Lee and Shemilt (2011) viewed as low-level empathy, which offers little insight into agents' consciousness.

This level of empathy is also reflected in a general collective narrative of progress identified in other Western democracies, such as the United States (Barton, 2009). Eras and figures of the past, even in one's in-group, are judged as tainted with deficiencies from which we in our current enlightened society are mercifully emancipated. Progressive narratives' criticism of the past offers members of dominant groups a way of distancing themselves from their morally tainted ancestors.

Tara is an Ashkenazi student and member of a liberal-left youth movement; she depicts Israel's first prime minister (from the same ethnic and sociopolitical milieu) in harsh terms: “What he really meant was: They [the immigrants] are coming … let's wipe out their personality.” Such perspective taking can be deemed critical and impartial to some degree, but it does not facilitate her progression into elaborate perspective taking or disciplinary thinking. In some cases, it was actually participants attempting to protect their in-group's image who applied empathy in more articulate ways.

When his Mizrahi peer Yakov criticizes educators for teaching immigrant children Ashkenazi folkdance, Gadi, an Ashkenazi participant, offers him to join on a quite elaborate attempt at perspective taking:

I think [the reason educators did] it was in order for them [the immigrant children] to integrate. What's this thing about folkdance … think you were a newcomer? They'll be putting you in school with all kinds of people and then everybody goes dancing and only you stand aside.

Here, Gadi tries to reconstruct not just the educators' intentions but also their perceptions of how the Mizrahi immigrant students may have felt. He creates here a nested theory of mind, inviting Yakov to share the leap into the founding fathers' mental image of the immigrants' emotions.

It is notable that Yakov, somewhat later in the discussion, does indeed take up the invitation to join the leap into historical mentalities. As he attempts to reconstruct his in-group members' (the Mizrahi immigrants') consciousness, the close relationship between identity and identification is noticeable: “The victim was their parents who were traditional and their kids were going dancing with girls who exposed their body. That's what people are now coming to … let's say I don't want my son to see naked girls.” Here what seems like an example of Gadamer's (1975) “merging of horizons” can be seen between past events and present perceiver. Yakov starts at a certain distance from the immigrant parents evidenced by use of third-person possessive “their kids.” However, through a connection to the present (“are now coming to”), Yakov identifies with these parents to the extent that he takes their perspective using the first person (“I … my son”).

Both Yakov and Gadi's perspective taking is more sophisticated than a generalized explanation of motivation. It is not based on assumption of moral deficiency of historical agents but rather on assumption of their similarity to the learner—an empathy that is centered on identification with the agents. Gadi's identification with the Ashkenazi educators and Yakov's identification with the immigrants furnishes the motivation to depict them positively.

At the same time, both learners rely on their emotional identification with the agents (born of their identity) to reconstruct their perceptions. This trajectory into perspective taking is still quite problematic in terms of historical understanding. Historical perspective taking should ideally be based neither on the presentist false assumption of similarity, nor on emotional reaction, but rather on acknowledgment of difference and evidence-based contextualization (Wineburg & Fournier, 1994). But, as will be seen, the identity-driven confrontation between the two learners also leads to a higher level of perspective taking.

Yakov criticizes the Ashkenazi educators. He claims, “They should have taught them Yemenite dance steps!” Now, Gadi, his Ashkenazi peer, creates another mental reconstruction, this time based on evidence (Source 5, the official's testimony); he cites: “There's someone [an Ashkenazi educator] here saying that at the times when they [as young pioneers] didn't have food and were dying of starvation and stuff, all that helped them hold on was folkdance. The spirit of Israeli Zionist existence.” Here the application of empathy toward the historical agents is not based on a presentist assumption of their similarity to the student. Instead, the learner reconstructs the unique way these historical agents, the Ashkenazi educators, perceived folkdance, a reconstruction based on the agents' reference to the role of dance in their personal experience as immigrant pioneers.

Both Gadi and Yakov apply empathy to historical figures from their own identity group. The tendency to apply empathy more willingly to in-group than to out-group members is a phenomenon already noted in intergroup relations studies (Hewstone et al., 2002). It is worth noting, then, the cases in which students attempt to take the perspective of out-group members.

Eli, an Ashkenazi student, tries to explain the failure of the Melting Pot policy. He begins with a rather generalized attempt to reconstruct an in-group member's mentality: the first Prime Minister David Ben Gurion's “aspiration that everybody will be integrated.” Eli goes on to elaborate on the immigrants' perception and reactions: “The moment you're told: ‘Well, listen … forget you're Yemenite'—then you'll come and think like: ‘No way! I don't want to!' and you'll be more anti and have more strength to love what you belong to [the ethnic community].”

Such a mental reconstruction features a high level of historical empathy and an interactive understanding of agency. Eli's attempt at reconstructing the consolidation of ethnic identity parallels theoreticians' conceptualization of Mizrahi ethnicity as a reaction to discrimination and cultural pressure (Smooha, 2004). Furthermore, Eli manages to empathize with the position of out-group historical agents, the Mizrahi immigrants. He partakes in the rare practice of giving voice to historical actors and social groups who filled the subordinate role. Such a depiction is notable for an out-group member, for these agents' perspective is usually under-represented in both official narrative and students' narratives (DiCamillo & Pace, 2010).

However, not all students' attempts at perspective taking aimed at attributing intentions and perceptions. In some cases, they actually struggled to convince peers that historical agents “had no intentions” or at least that not all of the members of the same group shared the same outlook. Such cases, in which students attempt to complicate the historical representation, will now be examined.

From Selection Bias to Historical Distinction

Consider the following example, which returns to Tamar, an Askenazi participant, in her conversation with Nathan and Gil (an excerpt ending the section on evidence evaluation). Having just referred to an official's transcribed words exposing the political motivation behind the Melting Pot policy as credible, she now struggles to reject their negative implication. Presumably, if all teachers involved in the educational policy (which were overwhelmingly Ashkenazi) were interested only in political power, it would taint the image of their (and thus her) cultural group.

Tamar apparently sets out to protect that image by claiming the evidence is non-representative: “I still don't believe really most of the teachers worked on the assumption what they're doing now is helping the Labor Party or like … that man, he represents only what he thinks … the minister of education or something. … ” Interestingly, the two boys contradict her. Though they have already declared agreement with her general view and should be susceptible to the same social identity bias, they maintain what seems like a more impartial stance. Nathan confronts Tamar's move. He emphasizes the significance of the speaker's top-ranking position: “He was responsible for the education in the immigrants' camps.” To this reminder, Gil adds the foregone conclusion “ … so he set the values of uniform education … ” (he implies that the official must therefore be seen as exemplifying the educators' motivations). Still, Tamar insists: “The screw inside the big system [the individual teacher]—that's not what he thought he was doing! O.K.?”

On one hand, this short dialogue seems to highlight Tamar's identity-driven resistance to her more impartial friends. In face of their critical, contextualized inferences, she attempts an apologetic reframing of a prominent example as a non-representative sample. On the other hand, during this confrontation, Tamar elaborates a more complex image of the historical scene. Tamar distinguishes two levels of thought and agency: that of “the big system” and that of “the screw” or the grassroots. What started out as an apparently biased selection and reframing of an example seems related to an elaboration of a more complex and differentiated representation of historical reality. It is to this theme in students' discourse that this analytic focus now turns.

One of the biases anticipated by social cognition theories is the tendency toward selective sampling in explaining or choosing examples. This can mean, for example, viewing negative actions of a single out-group member as representative of the whole out-group, while those of in-group members are seen as exceptional (the “black sheep” phenomenon). Selective sampling can also refer to the negative consequences of in-group members' actions by claiming, for example, that only a limited subset of the out-group was harmed or affected by these actions. These cognitive moves may seem manipulative and identity serving (Doosje et al., 1998), but they can also signal a more complex view of historical reality.

Here, beyond reflecting bias, Tamar's distinction also indicates the capacity to differentiate between aspects and levels of society, as between government and grassroots motivations (Barton, 2010). Barton and Levstik (2004) stressed that such a differentiation between individual and group is essential for students' understanding of the complexity of historical phenomena. Tamar hints at this capacity through her differentiation between the official and his subordinates or the individual and the group.

A similar differentiation takes place in contrasting the state and individual citizens, as when Ily, son of a Mizrahi immigrant from Iraq who was educated in the Melting Pot institutions, remarks, “It contributed to the state but it hurt the [immigrant children's] parents.” A related distinction some students made was between the effect of the policy on immigrant parents and its effect on their children. This is a differentiation between subsets of the seemingly homogenous generalized group of “immigrants” and is based on the more complex assumption that the effect of or damage done by the policy was dependent on the specific (and different) cultural consciousness and perceptions of those affected by it. As Gil, one of Tamar's male peers of Ashkenazi origin, noted, “It hurt the [immigrant] grownups, those that were already used to their culture, but the children, coming as yet without all that culture—them it [the policy] helped.”

From a social cognition perspective, the above distinctions may be seen as another kind of identity-motivated attempt at shifting the focus of analysis (the “sample” to be selected). Discussants note there is more than one level or aspect to understanding a historical phenomenon or action but try to focus discussion on the aspect that would benefit their group's moral image. Ily, the Mizrahi student, ends by focusing on the price his in-group members, the immigrants, paid for the benefits the state gained, giving them the moral ground of the victimized. Gil, the Ashkenazi student, ends by turning the lens to those of the immigrants who benefitted from the policy, thereby implying the benevolence of its Ashkenazi initiators, members of his in-group. Nevertheless, distinction refers to the complexity and multi-faceted nature of society and of reality, and as such, it may be seen as an elaboration of historical thinking.

The distinction between different time spans—in terms of factors, outcomes, or perspectives—is also considered a characteristic of higher levels of causal and narrative thinking (Lee, 2005; Shemilt, 2000). In this study, reference to long-term versus short-term effects constituted about one-third of the overall cases of historical distinction. This distinction, which may have been influenced by the terminology of an excerpt from an historian criticizing the policy (Source 7), was more frequently used to criticize the Ashkenazi initiators of the policy. As Yoram, a Mizrahi student arguing with his Ashkenazi peer, states, “All of this [actions taken according to the policy] damaged them [the immigrants]. Could be in the short run it could really have been an easy solution and it helped, but in the long run it damaged!” Although this distinction criticizes the short-sightedness of the Ashkenazi initiators of the policy, Ashkenazi discussants also employ it, such as Tova, who may be taken to make a more impartial move as she claims “Like, it was a terribly short-term solution … ‘let's go, altogether now’ … but in the course of years it came out as the opposite … it caused frustration and discrimination that we see today.”

Long- and short-term distinctions rely on the wisdom of hindsight. But hindsight itself can also impede the elaboration of historical understanding. It seems the present furnishes an advantaged viewpoint over the past, since we see more of the past and of its consequences. Such an assumed advantage can lead to the anachronistic projection of present-day sensibilities onto historical agents (Wineburg et al., 2007).

Returning to Tamar, Gil, and Nathan, the Ashkenazi discussants, it is indeed seen that when Tamar articulates such an anachronism, she bases it on her privileged, present vantage point. Criticizing the Ashkenazi leaders, Tamar demands “They should have shown pluralism.” In response, Nathan points to the anachronistic nature of her claim, as he refers to the historical context: “This seems to me out of touch, because the time, in the world at large, it isn't a time of pluralism. … It's not like today.” However, Tamar justifies her demand precisely on the basis of her later and therefore improved perspective: “Right, but I can say these things because I'm from a point in time of the future.” Now Nathan drives the point home. He stresses that the distinction Tamar should make is not between moral and immoral ideologies but between her own ideology and the perceptions and perspective of historical agents: “The question is whether standard education … benefitted the immigrants or damaged them back then! Not like from … er [your?] … perspective. From their perspective!”

This complex meta-cognitive move expresses quite a developed historical understanding. Nathan and Tamar agree to distinguish between their own beliefs and the horizons of possibility and ideology that were open to historical agents and not to judge the latter according to the current zeitgeist. Tamar's appropriation of this notion is evident when some 39 turns later in the discussion, she asserts, “And even to all those people who say it was possible to make a pluralistic education at the time—it was impossible. … ” However, even this sophisticated understanding can be related back to social identity (and may in part be motivated by backsliding to a more comfortable position, not just by more complex thinking). Both Nathan and Tamar are Ashkenazi, and their differentiation bears on in-group esteem, as the Ashkenazi Israeli founding elite are now exempted from contemporary moralistic criticism.

Shemilt (2000), in his outline of narrative framework development, suggested that at its higher levels, learners will embrace multidimensional and polythetic narratives. These include the ability to contain the “tragedy of good intentions” or the distinction between intentions and consequences. In the present study, this contrast between intentions and outcomes was the aspect of differentiation appearing most frequently in discussions (by a total of 20 participants in 12 out of the 17 discussions in which differentiation occurred), particularly in what could at first glance seem only like identity-relevant apologetics.

In the overwhelming majority of these cases, the goals or motives were depicted as different from and more positive than the outcomes. Thirteen out of 20 students putting forth such a differentiation were Ashkenazi, as is the case of Gil, one of Tamar's Ashkenazi peers in the aforementioned conversation. Referring to the policy outcomes, he says, “It seems the idea was good but in the end it … didn't come out as ‘really something.' It quite … instead of making everyone the same … it very much drove them apart.” Similarly, Tanya, an Ashkenazi student who was pressed hard by her Mizrahi peer, Leah, to declare the Ashkenazi establishment's malevolence, resists: “ … I think it wasn't their intention from the beginning [to erase immigrants' identity]. Like I don't think they saw … like with hindsight we can say … but for some reason I insist their initial purpose was to promote equality. … ”

The repeated emphasis on good intentions underlines the apologetic function this distinction can serve. Denying negative intentionality behind in-group members' harmful actions may seem like a classical attribution bias, mitigating in-group historical responsibility and the implied damage to group esteem. However, such a distinction also expresses an understanding of complexity with reference to both the uncertain relationship between intentions and consequences and the problematic nature of judgmental hindsight.

DISCUSSION

The examination of students' discussions of an identity-relevant historical topic in an ethnically diverse context reveals disciplinary practices were not just biased and impaired by identity and intergroup interaction. It also seems to have been facilitated by them. Using the lens of social cognition theory, most of the instances in which students applied disciplinary practices could be interpreted as demonstrating the impact of their identity. However, many of them go beyond one-sided biased understanding.

This study's findings do attest to a clear degree of influence of identity needs on learners' engagement in historical reasoning and understanding. Identity-serving instances outnumber more neutral or self-critical uses of evidence and applications of empathy. Historical distinctions were clearly used more frequently in apologetic than in impartial ways. Participants tended to evaluate evidence threatening in-group esteem as less credible, as has been found in prior social psychological research (Voss et al., 1996). They differentiated between aspects of events, frequently benefitting in-group image or minimizing harm to it, in ways that can be seen as “selective sampling” (Doosje et al., 1998). Participants attributed malevolent intentions to historical agents who were out-group members, presenting intergroup attribution bias (Fiske, 2002; Hewstone et al., 2002).

The motivation to employ historical learning practices in such a way can be explained through social identity perspective as a convergence of cognitive and emotional processes (Tajfel & Turner, 1986). The cognitive process is that of social categorization, activating Ashkenazi-Mizrahi ethnic criteria through their salience in the topic studied and in the social context. The emotional process was that of social identification whereby students tended to feel they belonged to one group or the other, shared its status, and were impacted by its historic image.

This identification motivated students to interpret historical agents' perceptions and many times motivated the valence of students' interpretation. Emotional reactions, such as collective guilt, indignation, or sympathy for in-group members, also seem to guide students' understanding of historical agents. Interpretations directed by the motivation to maintain in-group esteem could be seen as an “abuse” of history and as an example of the problematic aspects of an emotive stance to history. Emotions arising from social identity clearly can promote identification with specific historical agents, with the unconscious assumption that shared identity means shared feelings. Such identification may render the use of evidence and context as redundant, since “it runs in my veins.”

However, this study also demonstrates how students' social identification fueled debate and caused them to question and challenge interpretations threatening in-group esteem, leading to more sophisticated performance of historical practices. As students' interpretations of historical agents' actions or intentions were challenged by peers, emotional needs prompted by social identification promoted learning, not just inhibited it. The fact that participants in ethnically mixed groups demonstrated a lower tendency to use identity-serving distinctions than participants in homogenous groups may also reflect this process. Social identity theory assumes that encounter heightens group awareness and loyalty. But in these dyads and triads, where identity awareness is assumed to have been heightened, identity-serving use of historical distinction was curbed and the frequency of impartial practice was higher.

Social psychology's tendency to detect bias as a sign of cognitive failure should not distract from noting cognitive developments that can accompany it. In quite a few cases, students' attempts to maintain identity needs and protect in-group relative status seemed to prompt more elaborate disciplinary practices and cognitive stratagems. Faced with peer challenges, learners' attempts to simply dismiss all in-group unfavorable evidence as “dumb” were insufficient. Students proceeded to rely on source characteristics or epistemological problematics to evaluate (and discredit) the reliability of threatening evidence, approximating the disciplinary practices of sourcing and contextualization (Lee, 2005; Wineburg & Fournier, 1994). Elaboration of distinctions, selective sampling, or an emphasis on in-group variability may be seen as apologetic rhetoric arising from identification, but these “biases” also help build more complex, nuanced understandings of the relationship between intention, outcomes, and perceptions of diverse agents, which seems much needed in historical understanding (Barton, 2010; Shemilt, 2000).

Even the “fundamental error” of attributing inner motivations to out-group members apparently fostered historical empathy and facilitated mental reconstructions. Whether in order to attribute negative moral responsibility or to counter it, discussants formed some quite sophisticated theories of mind about historical agents. In some of the more elaborate cases, they even took the perspective of the underprivileged historical agents. Applying empathy or giving voice and agency to excluded others is a valued historical educational outcome (DiCamillo & Pace, 2010). Indeed, taking the perspective of the subordinate agents, the immigrants, went “against the grain” of what seemed to be a prevailing tendency.

In some cases, Ashkenazi students reconstructed Ashkenazi historical agents' mentalities in ways that were not favorable to in-group image and acknowledged the veracity of unflattering evidence. This acknowledgment of negative aspects of the in-group goes against social identity assumptions. It could be that affirmation through the process of discussion itself served to permit or promote the acknowledgment of potentially threatening perspectives (Sherman & Cohen, 2002). The context of desegregation (Resh & Kfir, 2004), prolonged intergroup contact, and possibly the “disciplinary identity” offered by the learning task itself, may have helped such affirmation.

While some of this observed effect may be in part due to the current low vehemence of ethnic conflict in Israel (Tsafati, 1999), or to the emphasis on solidarity in the face of an external threat (Kimmerling, 2001), it still demonstrates how identity itself may be used positively in a diverse setting to facilitate cognitive elaboration and not just conflict.

It is worth adding that although throughout this article “identity-serving” practice was contrasted with “neutral” disciplinary practice, the latter may also be explained in terms of identity. First, acquisition of disciplinary practice expresses the move toward quasi-, proto-, or emerging academic identity and inclusion in a community of practice (Lave & Wenger, 1991; Sfard & Prusak, 2005). Second, criticism of one's in-group could also be seen as a form of self-distancing from the implicit negative in-group moral image. This was explained above as an individual, not a collective, identity move. However, it arises from the burdens of one's collective identity (Abrams & Hogg, 1990; Goldberg, Porat, & Schwarz, 2006).

EDUCATIONAL IMPLICATIONS

The educational implications of this study are several. First, it seems that, contrary to the assumptions of many educators and curricular policy makers, students are eager to discuss and capable of discussing charged identity-relevant topics (Barton, 2009; Del Favero et al., 2007; Hess & Posselt, 2002; Levstik, 2000). The motivational boost that identity may add to students' learning is accompanied by new challenges to educators. Teachers should be aware of social identity biases and use sophisticated ways of engaging learners in a meta-cognitive process of self-awareness about such tendencies. But teachers should also trust the ability of students to develop without much direct facilitation (as in this study) through engagement with such contentious topics.

Second, social studies teachers should note the possibilities that social identity influences offer to the elaboration and development of cognition and analysis. It seems that bias can be utilized constructively to promote critical thinking. The practice of critical evidence evaluation may be facilitated through use of evidence that is unfavorable to the in-group, especially when such evidence is identified as originating from the out-group. From this essentially defensive starting point, students' more universal, reflective and self-critical capacity can gradually be developed.

Third, in a similar vein, it may also be useful to capitalize on identification with one's group and the attribution bias that accompanies it to foster empathy and perspective taking. Indeed, if shared identity is on a trajectory with identification and empathy, there is also reason to arouse it through choice of identity-relevant topics. As we have seen, dealing with such topics may lead both to giving voice to marginalized groups and to elaborating theories of mind.

However, identity should not be perceived as an objective stable trait. It is in many ways situational, enacted, and relational. Students' awareness of identity can be aroused and constructed through encounter, and educators should consider structuring identity-relevant encounters into learning design, through both curricular materials and engagement with peers' identities.

Fourth, there is also some reason to think that structuring group composition to maximize diversity may produce more elaborate learning. This may raise some questions as to the propriety of identity-based structuring and manipulation. However, it should be noted that encounter and heterogeneity, as well as promoting learning, are in fact deliberate goals of desegregation in schools (Resh & Kfir, 2004). It should also be noted that participating in small groups and working toward joint goals are two conditions that are assumed to be favorable for intergroup encounter (Pettigrew, 1998, 2008). A study using a similar educational intervention and documenting Jewish and Arab adolescents' discussion of a charged intergroup topic has revealed that learning in a critical-disciplinary approach promoted collaboration and agreement (Goldberg, 2012).

Further research is needed to assess whether a causal or consistent development can be traced throughout students' discussion. Can we assume that enacting identity-protective cognitive moves precedes or prompts the use of more impartial disciplinary practices? There is also a need to examine further whether and how the composition of discussion groups affects learning. Are mixed or homogenous identity groups more beneficial to certain learning outcomes rather than others? Hopefully, future quantitative analysis of the current data, as well as other future research efforts, will afford some more specific insights into these questions.

Notes

¹Participants were asked to name the countries of origin of their grandparents as a basis for this categorization. Dual ethnicity participants made up a small proportion of the sample (11%, N = 7) and were asked to indicate the ethnicity they thought represented them best.

APPENDIX

Stand, Type, Reliability, and Content of the Eight Sources Students Analyzed and Used in Discussion

	Pro Melting Pot policy		Against Melting Pot policy
	Reliable	Unreliable	Reliable	Unreliable
Primary	Source 1: Standard education curriculum	Source 4: Educational director's defense in front of investigation committee contrasted with his account in a closed party forum	Source 3: Investigation committee's report on education in immigrant camps	Source 2: Member of Parliament's public provocative speech against governmental educational policy
	Examples: kindergarten and middle school practices, typical of the collectivist secularized patriotism of the melting pot	Examples: Exaggerated presentation of benevolence contrasted with political guile	Examples: Religious Yemenite Jews hurt by cultural activities and alleged forced haircut; educators justify their actions	Examples: The cultural absorption policy compared to Catholic Inquisition; allegations of forced secularization
Secondary	Source 8: Historical research about absorption of Mizrahi immigrants	Source 5: Personal reminiscences of immigrant children's educational experience	Source 7: Historical research about the Melting Pot policy	Source 6: Protest novel about immigrant child educational experience
	Examples: Melting Pot policy was successful against all odds; immigrants reached influential positions	Examples: Tensions between children and their immigrant parents were a justified price to pay for progress	Examples: Seemingly a success in the short run, the policy harmed generations of immigrants and their children in the long run	Examples: Teacher degrading immigrant children and treating their parents as primitive; author prophesying ethnic rift and remorse