1. Introduction
Schools rarely appear in the foreground of conflict analysis, yet they occupy a strategic position in the social reproduction of group boundaries. In societies marked by protracted intergroup conflict, the question of who studies with whom, in what language, under whose curriculum, and toward whose collective memory is never merely administrative. It is constitutive of the political order itself. Education shapes not only the cognitive content that young people acquire but also the social categories through which they come to understand themselves and others (Bar-Tal, 2007).
The proposition that education matters for conflict dynamics has long been recognized. Bush and Saltarelli (2000) influentially argued that education has “two faces”: it can mitigate conflict by fostering tolerance, critical thinking, and intergroup understanding, but it can also exacerbate conflict by reproducing exclusionary identities, biased historical narratives, and structural inequalities. Subsequent scholarship has elaborated this dual potential across a range of cases (Davies, 2003; Smith, 2005; Salomon, 2004).
Despite this growing literature, comparative analysis of the structural designs through which educational systems organize intergroup relations remains underdeveloped. Much existing work focuses on a single case or evaluates a particular pedagogical intervention, leaving the macro-architecture of educational systems insufficiently theorized. This study addresses that gap. It asks how the formal structure of an educational system, defined by the degree of separation or integration between groups in conflict, shapes the system’s contribution to conflict maintenance or transformation.
Four cases anchor the analysis: Northern Ireland, Bosnia and Herzegovina, Cyprus, and Israel. These cases share the feature of protracted intergroup conflict with a salient ethno-national or ethno-religious dimension, yet they have arrived at strikingly different educational arrangements. Comparing them allows for the construction of a typology that is both empirically grounded and analytically generalizable. Briefer reference is made to Rwanda and North Macedonia where these cases illuminate particular dimensions of the argument.
The argument proceeds in five steps. The first section develops the theoretical framework, drawing on contact theory, social identity theory, and scholarship on collective memory and educational reproduction. The second proposes a typology of four educational models in divided societies. The third examines each ideal type through the relevant case material. The fourth synthesizes findings across the cases, identifying mechanisms through which different designs influence conflict dynamics. The conclusion considers theoretical and policy implications, including the limits of structural reform absent broader political transformation.
2. Theoretical Framework
2.1 Education and Social Reproduction
Modern mass schooling emerged in close historical association with the consolidation of nation-states, and one of its principal functions has been the cultivation of shared collective identity (Gellner, 1983; Anderson, 1983). In societies where the nation-state project is contested by competing collectivities, this identity-shaping function becomes a site of struggle. Schools transmit not only knowledge but also what Bourdieu and Passeron (1990) termed cultural capital, including the symbolic codes, historical narratives, and dispositions that mark insiders and outsiders within a given social order.
In divided societies, the reproductive function of education takes on a particular character. Educational systems can serve to reproduce group boundaries directly, by separating children into ethnically or religiously homogeneous institutions, or indirectly, by transmitting curricular content that codes the in-group as legitimate and the out-group as adversarial or marginal (Bekerman, 2009; Cole, 2007). The cumulative effect is that even ostensibly neutral curricula often carry what Apple (1993) called an ideological subtext favoring dominant groups.
2.2 Contact Theory and Its Limits
The most influential theoretical framework for thinking about the role of intergroup contact in education is Allport’s (1954) contact hypothesis, which proposed that prejudice between groups can be reduced through contact under specified conditions: equal status, common goals, intergroup cooperation, and institutional support. Decades of research have largely confirmed that contact, when these conditions are approximated, reduces prejudice (Pettigrew & Tropp, 2006).
Yet the application of contact theory to protracted conflict settings has revealed important limitations. Maoz (2011) showed that the conditions for productive contact are difficult to achieve when underlying power asymmetries are pronounced and when contact is structured by the hegemonic narrative of the dominant group. Dixon, Durrheim, and Tredoux (2005) argued that prejudice reduction at the individual level does not necessarily translate into structural change, and may even legitimize unjust arrangements by creating the appearance of harmony. These critiques are particularly relevant to educational interventions, where well-intentioned encounter programs can leave structural inequalities untouched while producing what Salomon (2004) termed a “false sense of resolution.”
2.3 Social Identity, Narrative, and Collective Memory
Social identity theory (Tajfel & Turner, 1986) emphasizes that group membership is a fundamental source of self-understanding and that intergroup comparison generates motivational pressures toward in-group favoritism. In conflict settings, these pressures interact with what Bar-Tal (2000) described as an ethos of conflict, a coherent set of societal beliefs that frames the in-group as just, the out-group as threatening, and the conflict itself as existential. Schools are central institutional carriers of such ethos.
Collective memory provides a closely related lens. Halbwachs’s (1992) foundational insight that memory is socially constructed has been extended by scholars such as Wertsch (2002), who showed how official curricula shape what young people remember about national pasts. In divided societies, history education is often a battleground because each group seeks to inscribe its own narrative as authoritative (Cole, 2007; Paulson, 2015). The educational structure determines whether competing narratives are kept apart, allowed to confront one another, or subsumed under a single official account.
3. A Typology of Educational Designs in Divided Societies
Building on the theoretical framework, this section proposes four ideal-type designs through which educational systems in divided societies organize the relationship between groups. The typology is analytical rather than descriptive: actual systems often combine elements of more than one type, and within any country there may be considerable internal variation. The value of the typology lies in clarifying the structural logics that distinct designs embody.
The four ideal types are: (1) full structural separation, in which different groups attend institutionally separate school systems with distinct administrative authorities, languages of instruction, and curricula; (2) parallel systems within shared institutions, in which groups occupy the same physical buildings but follow separate curricula, teachers, and sometimes schedules; (3) voluntary integration, in which integrated schools coexist alongside the dominant separated arrangement and are chosen by a minority of families; and (4) unified common schooling, in which all children attend a single system with a shared curriculum, regardless of group affiliation.
Each design embodies a particular theory of what education should accomplish in a divided society. Full separation expresses a logic of cultural autonomy and identity preservation. Parallel systems represent a compromise between integration and group autonomy, often emerging from political settlements after acute conflict. Voluntary integration reflects a liberal commitment to free choice combined with reformist aspiration. Unified common schooling embodies an assimilationist or civic-republican vision in which the school constructs a common citizenry.
4. Comparative Case Analysis
4.1 Israel: Full Structural Separation
The Israeli educational system is among the most extensively separated in the democratic world. Jewish and Arab citizens attend almost entirely distinct school systems, with separate administrative streams, languages of instruction, and curricula (Al-Haj, 2002). Within the Jewish sector itself, further separation exists among state-secular, state-religious, and ultra-Orthodox streams, but the separation between Jewish and Arab education is the most consequential for the analysis of conflict dynamics.
This separation is not the product of a single deliberate policy but the cumulative outcome of historical, political, and demographic processes that began before the establishment of the state and were institutionalized in its early decades (Al-Haj, 2012). The official rationale has emphasized the recognition of cultural and linguistic distinctiveness, particularly the right of the Arab minority to education in Arabic. Yet the practical consequences extend well beyond linguistic accommodation.
Research has documented persistent inequalities between the two systems in funding, infrastructure, and educational outcomes (Arar & Ibrahim, 2016). Beyond resource gaps, the curricular content of the Arab system has historically been subject to significant constraints, with limited space for the articulation of Palestinian national history and identity (Al-Haj, 2005). The Jewish curriculum, in turn, has been shown to construct an in-group narrative in which the Arab citizenry appears marginally if at all (Bar-Tal, 1998; Peled-Elhanan, 2013).
The cumulative effect of this design is a system that minimizes intergroup contact during the formative years and reproduces distinct, largely incompatible national narratives. Bekerman and Zembylas (2011) have argued that the structural separation of the Israeli system functions as a form of what they call “identity politics in education,” in which the maintenance of group boundaries takes precedence over the cultivation of shared civic understanding. Bilingual integrated schools, established by the Hand in Hand network beginning in the late 1990s, have demonstrated that integration is possible, but they remain a marginal exception serving a small fraction of students (Bekerman, 2016).
4.2 Bosnia and Herzegovina: Parallel Systems within Shared Institutions
The post-Dayton educational settlement in Bosnia and Herzegovina produced one of the most striking institutional designs in contemporary divided societies: the so-called “two schools under one roof” (dvije škole pod jednim krovom) arrangement. Under this model, Bosniak and Croat children attend the same school building but are taught separate curricula, often by separate teachers, and sometimes in shifts that prevent contact altogether (Hromadžić, 2008). The arrangement was introduced in the early 2000s, ostensibly as a transitional measure to facilitate the return of displaced families, but it has proven highly durable.
Each ethnic group teaches its own version of national subjects, particularly history, language, and religion, and the resulting narratives diverge substantially on questions central to the recent war (Torsti, 2009). The structural co-presence of the two systems in the same building produces a peculiar phenomenology in which children encounter members of the other group in corridors and yards but rarely in classrooms. Hromadžić’s (2008) ethnographic work showed that this design teaches children to navigate ethnic difference through avoidance, normalizing segregation rather than challenging it.
The arrangement has been criticized by domestic civil society, the Council of Europe, and the OSCE as discriminatory and as an obstacle to reconciliation (Magill, 2010). Yet political resistance to consolidation remains strong because each group’s elites view the parallel system as a guarantee of cultural reproduction. The Bosnian case illustrates how educational designs that emerge from peace agreements can institutionalize wartime divisions even as they formally prevent their violent expression.
4.3 Cyprus: Separation Across a Political Boundary
The Cypriot case differs from the previous two in that educational separation is reinforced by a de facto political division of the island. Greek Cypriot and Turkish Cypriot children attend educational systems administered by separate political authorities, with the Republic of Cyprus governing the south and the unrecognized Turkish Republic of Northern Cyprus administering the north. Direct contact between students from the two communities is minimal.
Curricular analyses have shown that history textbooks on each side construct mirror-image narratives of victimhood, attributing responsibility for the conflict to the other community while minimizing the in-group’s role (Papadakis, 2008; Zembylas, Charalambous, & Charalambous, 2016). Reform efforts in the Greek Cypriot system in the late 2000s sought to introduce more pluralistic perspectives, but encountered significant political backlash and were partly reversed (Philippou, 2009).
Bicommunal educational initiatives have operated at the margins, often supported by international actors, but they reach a small minority of students and operate against the grain of the dominant systems (Hadjipavlou, 2007). The Cypriot case underscores how educational separation, when reinforced by political partition, can produce two parallel processes of identity formation in which the other community appears primarily as an absence or a threat rather than as a present interlocutor.
4.4 Northern Ireland: Voluntary Integration alongside a Separated Norm
Northern Ireland presents a distinct configuration. The dominant pattern remains separation between the predominantly Protestant Controlled sector and the predominantly Catholic Maintained sector, a separation rooted in the historical settlement of the early twentieth century (Gallagher, 2004). However, since the 1980s a parallel track of integrated schools has developed, in which Protestant and Catholic children are deliberately educated together, with policies in place to maintain numerical balance (Hayes, McAllister, & Dowds, 2007).
Research on integrated schooling in Northern Ireland has produced cautiously positive findings. Students who attend integrated schools tend to display more positive intergroup attitudes, more cross-community friendships, and a greater tolerance for ambiguity in identity than peers in separated schools, though selection effects complicate causal interpretation (Hughes, 2011; McGlynn, Niens, Cairns, & Hewstone, 2004). At the same time, integrated schools educate only around seven percent of pupils, and the broader system remains structured by the Controlled-Maintained distinction.
The Northern Irish case also features a distinctive curricular instrument: the Education for Mutual Understanding programme, later folded into broader citizenship education, which sought to introduce reconciliation themes across all schools regardless of sector (Smith, 2003). Evaluations of these programmes have been mixed, with researchers noting that pedagogical content can be vitiated by the segregated context in which it is delivered (Niens & Cairns, 2005). The Northern Irish case illustrates the possibilities and limits of integration as a voluntary track within a larger separated system.
4.5 Brief Comparative Reference: Rwanda and North Macedonia
Two further cases warrant brief mention. Rwanda after the 1994 genocide adopted a unified common schooling model in which Hutu and Tutsi identities were officially erased from the curriculum in favor of a common Rwandan identity (Freedman, Weinstein, Murphy, & Longman, 2008). This represents the most assertive instance of the unified common-schooling type. Critics have argued that the suppression of ethnic categories without sustained engagement with the genocide’s history risks producing a brittle peace, in which unaddressed grievances may resurface (King, 2014).
North Macedonia, after the 2001 conflict and the Ohrid Framework Agreement, expanded mother-tongue education for the Albanian minority while maintaining a single state system. Research has shown that this arrangement, intended to balance recognition of group identity with civic integration, has nonetheless produced significant interethnic distance in schools, with limited contact between Macedonian and Albanian students (Reka, 2008). These cases extend the typology and reinforce the finding that no design escapes trade-offs.
5. Synthesis: Mechanisms Linking Educational Design to Conflict Dynamics
Comparing the cases reveals several mechanisms through which educational design shapes the maintenance or transformation of conflict. These mechanisms operate at three levels: the structural level of who interacts with whom, the curricular level of what is taught, and the institutional level of how schools are embedded in broader political arrangements.
5.1 The Structural Level: Contact and Its Absence
The most direct mechanism is the simple presence or absence of intergroup contact during the formative school years. Full separation, as in Israel and Cyprus, eliminates this contact almost entirely, leaving young people to form their understanding of the other group through mediated representations rather than personal acquaintance. Parallel systems within shared institutions, as in Bosnia, produce a paradoxical configuration in which physical proximity coexists with social distance, sometimes deepening the salience of group boundaries (Hromadžić, 2008).
Voluntary integration, as in Northern Ireland, offers contact under relatively favorable conditions to a minority of students, while unified common schooling, as in Rwanda, mandates contact but does so within a framework that may suppress legitimate identity claims. The lesson across these configurations is that contact alone is neither necessary nor sufficient to transform intergroup relations, but its structural availability sets the terms within which other mechanisms operate (Pettigrew & Tropp, 2006).
5.2 The Curricular Level: Narratives and Counter-Narratives
Curricular content interacts with structural design in complex ways. In separated systems, each group’s curriculum tends to develop without sustained engagement with the other’s perspective, producing what Bar-Tal and Salomon (2006) described as parallel collective narratives that legitimize the in-group and delegitimize the out-group. In parallel systems within shared institutions, the co-presence of competing curricula in the same building can render their incompatibility visible to students, but in the Bosnian case this has not generated systematic dialogue (Torsti, 2009).
Integrated and unified systems open the possibility of curricular engagement with multiple narratives, but realizing this possibility requires deliberate pedagogical work. The Northern Irish experience suggests that integrated structures must be paired with explicit curricular treatment of the conflict to produce sustained changes in attitudes (McGlynn et al., 2004). Where structural integration is decoupled from substantive engagement with the conflict, as has sometimes been the case in Rwanda, the result can be a surface coexistence that does not reach underlying historical wounds (Freedman et al., 2008).
5.3 The Institutional Level: Schools within Political Settlements
Educational designs are embedded in broader political arrangements that shape what schools can plausibly attempt. In contexts of pronounced power asymmetry, ostensibly neutral or integrationist designs can mask the continued dominance of one group’s culture, language, and historical narrative (Bekerman & Zembylas, 2011). The Israeli case, where even within Arab schools the curriculum is significantly shaped by Jewish-majority institutions, illustrates how separation does not necessarily entail genuine autonomy.
Conversely, in settings where political settlements have institutionalized group autonomy, as in Bosnia, structural integration faces formidable resistance from elites whose authority depends on the maintenance of group boundaries (Magill, 2010). The political economy of educational design thus matters as much as its formal architecture. Reforms that target educational structure without engaging the underlying political settlement tend to be either resisted, diluted, or absorbed into the very dynamics they were meant to challenge.
6. Conclusion
The comparative analysis offered here suggests that educational systems in divided societies cannot be understood through a simple opposition between separation and integration. Each of the four ideal-type designs identified, full structural separation, parallel systems within shared institutions, voluntary integration, and unified common schooling, embodies a distinct theory of how education should mediate intergroup relations, and each carries characteristic risks.
Three broader conclusions emerge. First, no structural design guarantees a particular outcome. Separation can preserve cultural autonomy or entrench mutual incomprehension. Integration can foster genuine encounter or produce a thin coexistence that suppresses legitimate grievance. The effects of any design depend on the curricular and pedagogical work conducted within it, and on the political conditions in which it is embedded (Davies, 2003).
Second, structural form and curricular content are not independent variables but interact systematically. A separated structure produces curricular drift between groups even where formal coordination is attempted, while an integrated structure without curricular engagement with the conflict tends toward avoidance of difficult content (McGlynn et al., 2004). Educational reform that addresses only one of these dimensions is likely to disappoint.
Third, the question of educational design cannot be separated from the question of recognition. Where one group’s identity, language, and historical experience are not adequately recognized in the broader political order, educational integration tends to be experienced as assimilation, while educational separation tends to entrench inequality (Bekerman & Zembylas, 2011). Sustainable reform requires changes in both the educational architecture and the political settlement that authorizes it.
These conclusions have implications for both research and policy. For research, they point to the need for comparative work that holds educational design in view while attending to curricular content and political context, rather than treating any of these dimensions in isolation. For policy, they suggest caution about exporting institutional models from one context to another. The Northern Irish integrated school, the Bosnian shared building, the Israeli separated stream, and the Rwandan unified classroom each emerged from particular histories and serve particular political functions. What works in one setting may not transfer, and what appears to fail may yet contain elements of value when reconfigured.
The deeper lesson is that schools cannot resolve conflicts that political institutions are unable or unwilling to address. They can, however, constitute a domain in which the relationship between groups is rehearsed each day, and the form of that rehearsal matters. Whether young people grow up imagining the other as enemy, as stranger, as neighbor, or as fellow citizen depends, in part, on how their schools are organized. The architecture of educational systems is therefore not merely an administrative question. It is one of the principal arenas in which divided societies decide what kind of future they are willing to imagine.
References
Al-Haj, M. (2002). Multiculturalism in deeply divided societies: The Israeli case. International Journal of Intercultural Relations, 26(2), 169–183.
Al-Haj, M. (2005). National ethos, multicultural education, and the new history textbooks in Israel. Curriculum Inquiry, 35(1), 47–71.
Al-Haj, M. (2012). Education, empowerment, and control: The case of the Arabs in Israel. State University of New York Press.
Allport, G. W. (1954). The nature of prejudice. Addison-Wesley.
Anderson, B. (1983). Imagined communities: Reflections on the origin and spread of nationalism. Verso.
Apple, M. W. (1993). Official knowledge: Democratic education in a conservative age. Routledge.
Arar, K., & Ibrahim, F. (2016). Education for national identity: Arab schools principals and teachers dilemmas and coping strategies. Journal of Education Policy, 31(6), 681-693.
Bar-Tal, D. (1998). The rocky road toward peace: Beliefs on conflict in Israeli textbooks. Journal of Peace Research, 35(6), 723–742.
Bar-Tal, D. (2000). Shared beliefs in a society: Social psychological analysis. Sage.
Bar-Tal, D. (2007). Sociopsychological foundations of intractable conflicts. American Behavioral Scientist, 50(11), 1430–1453.
Bar-Tal, D., & Salomon, G. (2006). Israeli-Jewish narratives of the Israeli-Palestinian conflict: Evolution, contents, functions and consequences. In R. I. Rotberg (Ed.), Israeli and Palestinian narratives of conflict: History’s double helix (pp. 19–46). Indiana University Press.
Bekerman, Z. (2009). The complexities of teaching historical conflictual narratives in integrated Palestinian-Jewish schools in Israel. International Review of Education, 55(2–3), 235–250.
Bekerman, Z. (2016). The promise of integrated multicultural and bilingual education: Inclusive Palestinian-Arab and Jewish schools in Israel. Oxford University Press.
Bekerman, Z., & Zembylas, M. (2011). Teaching contested narratives: Identity, memory and reconciliation in peace education and beyond. Cambridge University Press.
Bourdieu, P., & Passeron, J.-C. (1990). Reproduction in education, society and culture. Sage.
Bush, K. D., & Saltarelli, D. (2000). The two faces of education in ethnic conflict: Towards a peacebuilding education for children. UNICEF Innocenti Research Centre.
Cole, E. A. (2007). Teaching the violent past: History education and reconciliation. Rowman & Littlefield.
Davies, L. (2003). Education and conflict: Complexity and chaos. RoutledgeFalmer.
Dixon, J., Durrheim, K., & Tredoux, C. (2005). Beyond the optimal contact strategy: A reality check for the contact hypothesis. American Psychologist, 60(7), 697–711.
Freedman, S. W., Weinstein, H. M., Murphy, K., & Longman, T. (2008). Teaching history after identity-based conflicts: The Rwanda experience. Comparative Education Review, 52(4), 663–690.
Gallagher, T. (2004). Education in divided societies. Palgrave Macmillan.
Gellner, E. (1983). Nations and nationalism. Cornell University Press.
Hadjipavlou, M. (2007). The Cyprus conflict: Root causes and implications for peacebuilding. Journal of Peace Research, 44(3), 349–365.
Halbwachs, M. (1992). On collective memory (L. A. Coser, Ed. & Trans.). University of Chicago Press.
Hayes, B. C., McAllister, I., & Dowds, L. (2007). Integrated education, intergroup relations, and political identities in Northern Ireland. Social Problems, 54(4), 454–482.
Hromadžić, A. (2008). Discourses of integration and practices of reunification at the Mostar Gymnasium, Bosnia and Herzegovina. Comparative Education Review, 52(4), 541–563.
Hughes, J. (2011). Are separate schools divisive? A case study from Northern Ireland. British Educational Research Journal, 37(5), 829–850.
King, E. (2014). From classrooms to conflict in Rwanda. Cambridge University Press.
Magill, C. (2010). Education and fragility in Bosnia and Herzegovina. UNESCO International Institute for Educational Planning.
Maoz, I. (2011). Does contact work in protracted asymmetrical conflict? Appraising 20 years of reconciliation-aimed encounters between Israeli Jews and Palestinians. Journal of Peace Research, 48(1), 115–125.
McGlynn, C., Niens, U., Cairns, E., & Hewstone, M. (2004). Moving out of conflict: The contribution of integrated schools in Northern Ireland to identity, attitudes, forgiveness and reconciliation. Journal of Peace Education, 1(2), 147–163.
Niens, U., & Cairns, E. (2005). Conflict, contact, and education in Northern Ireland. Theory Into Practice, 44(4), 337–344.
Papadakis, Y. (2008). Narrative, memory and history education in divided Cyprus: A comparison of schoolbooks on the “history of Cyprus.” History and Memory, 20(2), 128–148.
Paulson, J. (2015). “Whether and how?” History education about recent and ongoing conflict: A review of research. Journal on Education in Emergencies, 1(1), 14–47.
Peled-Elhanan, N. (2013). Palestine in Israeli school books: Ideology and propaganda in education. I.B. Tauris.
Pettigrew, T. F., & Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90(5), 751–783.
Philippou, S. (2009). What makes Cyprus European? Curricular responses of Greek-Cypriot civic education to ‘Europe’. Journal of Curriculum Studies, 41(2), 199–223.
Reka, A. (2008). The Ohrid Agreement: The travails of inter-ethnic relations in Macedonia. Human Rights Review, 9(1), 55–69.
Salomon, G. (2004). A narrative-based view of coexistence education. Journal of Social Issues, 60(2), 273–287.
Smith, A. (2003). Citizenship education in Northern Ireland: Beyond national identity? Cambridge Journal of Education, 33(1), 15–32.
Smith, A. (2005). Education in the twenty-first century: Conflict, reconstruction and reconciliation. Compare: A Journal of Comparative and International Education, 35(4), 373–391.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S. Worchel & W. G. Austin (Eds.), Psychology of intergroup relations (pp. 7–24). Nelson-Hall.
Torsti, P. (2009). Segregated education and texts: A challenge to peace in Bosnia and Herzegovina. International Journal on World Peace, 26(2), 65–82.
Wertsch, J. V. (2002). Voices of collective remembering. Cambridge University Press.
Zembylas, M., Charalambous, C., & Charalambous, P. (2016). Peace education in a conflict-affected society: An ethnographic journey. Cambridge University Press.