Children living in conflict environments are tactical agents. Even though their agency is highly constrained, the majority of them manage to maximise the immediate circumstances of their environment. However, the dominant approach to the reintegration of former child soldiers fails to reflect children’s personal experiences and self-perception, as well as the understanding of childhood and child soldiering prevalent in the receiving communities. This thesis focuses on the previously under-researched case of Colombia. It examines the question of what kind of agency children exhibit in their recruitment and participation in the armed group, together with the implications of such agency for the reintegration process. The core of this empirical analysis is based on twenty life history interviews conducted with adult former child soldiers in Colombia at the beginning of 2014. The thesis makes an original contribution to the existing literature on child soldiers by examining the constraining factors to, and manifestations of, children’s agency in the reintegration period; most existing literature focuses on the pre-recruitment period. Recommendations are also made regarding how to incorporate the question of agency into reintegration programmes. Finally, this study provides a detailed examination of the double role of social capital: enabler and constrainer to children’s agency. Agency played a central part in the stories told by the research participants. Voluntary decisions and actions taken before and during children’s involvement in armed conflict have shown to have significant implications for their reintegration process. The study also confirms the crucial consideration of social factors in prevention and reintegration programmes. However, the thesis argues that the importance of social capital can only be understood if combined with other structural factors, examined within the context in which it functions and if its various characteristics are distinguished.

Purpose English construction, a particularly litigious industry, saw adjudication introduced in 1996 to improve cash flow and provide time-efficient, cost-effective dispute resolution. The industry perception is that adjudication no longer provides this. Mediation is a successful dispute resolution method used in many areas. It could be beneficial in construction disputes; however, there is limited evidence of significant implementation in England. The purpose of this research is to establish the current use of mediation in English construction, whether it is appropriate, and the requirements to encourage greater use. Method A mixed-method approach through case studies (20) followed by interviews (10) with key stakeholders, obtained qualitative data to develop a conceptual model. This informed the design of a cross-industry questionnaire, providing quantitative data to triangulate the findings. Findings The results demonstrated that the majority of disputes are adjudicated at a significant cost often with unpredictable outcomes. Little use was made of mediation. However, when used mediation is successful in resolving construction disputes, enabling negotiated outcomes. In addition, most users of adjudication and mediation would prefer to use mediation, where appropriate. The research also identified significant barriers, including a lack of understanding of mediation and the contractual requirement to use adjudication. The resistance to mediation was highest from subcontract organisations rather than larger contractors. Sub-contractors are generally suspicious of an offer of mediation from the main contractor. There was strong support for mediators being experts in the field of the dispute. Conclusion The identified barriers need to be removed to enable greater use of mediation. Parties to the projects (stakeholders) need to receive training in mediation, contacts need to encourage its use and mediators need to be easily accessible. Work is now ongoing, following this study, to develop this training, influence the bodies that draft standard contacts and make mediators accessible.

When a civil war ends there is a formal cessation of hostilities between the warring groups. Yet the termination of fighting does not necessarily end the violence, which can persist into post-war years. The study of post-civil war violence as a distinct phenomenon is dominated by explanations that see it as a legacy of the warring elite and an exercise in preventing them restarting war, or, as culturally embedded within interpersonal relations. This thesis draws on the group nature of civil war to offer a different perspective on this problem. It considers post-civil war violence as a product of continuing hostile relations between previous warring groups and examines the transformation of broader intergroup relations in post-war years. The rationale for the research rests on the idea that if violence is part of intergroup relations, then how might a more informed understanding of these relations facilitate effective peacebuilding interventions in countries emerging from civil war. Using qualitative methods, and a case study approach focusing on Nepal and Mozambique, I compare understandings and experiences of the transformation of intergroup relations from the perspective of political parties that claim to represent the previous warring groups, to that of people living in villages (three in each country). The study finds that violence can remain a central feature of how political parties relate to one another, even after decades of peace. To appreciate this violence requires a conceptual step away from dominant ways of thinking of violence as physical or structural. Conversely, while intergroup violence does not appear to have been a problem within the villages, the presence of violence within interparty relations is driving a pervasive and persistent fear that civil war may once again erupt. The thesis examines the impact of this violence on both the peace process and group relations to highlight the need to recognise and proactively address it.

This thesis analyses conflict resolution efforts between states, sub-state armed groups and independent third parties during the explicitly ethno-nationalist conflicts in Northern Ireland and Turkey. It considers that a predominantly armed response by states towards terrorist groups is an inadequate method for ending ethno-nationalist conflicts. It argues that exploring ethno-nationalist groups who use violence as a tactic to pursue their political goals, provide a better approach to understanding non-violent, political resolution attempts. However, the existing literature has paid relatively little attention to political attempts to bring their violent campaigns to an end. The thesis aims to close this gap through a comparative analysis of the Northern Irish conflict conducted by the Irish Republican Army (IRA), and the Kurdish conflict conducted by the Kurdistan Workers’ Party (Partiya Karkerên Kurdistan) as these conflicts have similar ethnic and territorial issues, armed campaigns, political attempts and peace processes. The Northern Irish conflict reached an agreement, but the Kurdish conflict has not. The reasons for the different outcomes are assessed by concentrating on a specific timeframe that is between the beginning of peace efforts and a peace agreement (or the failure of a peace process). The period involves the term between the emergence of the Provisional IRA and so the beginning of the ‘Troubles’ in 1969 and the Good Friday Agreement (GFA) of 1998 for the Northern Irish conflict. The conflict in Turkey is examined from 1984 to 2015 since the latest conflict resolution efforts collapsed and violence returned in July 2015. The thesis argues that ethno-nationalist conflicts are too complicated to be brought to an end solely through official negotiations. Therefore, it suggests a more comprehensive framework which begins at the pre-negotiation stage through secret contacts of the main armed protagonists and the initiatives of peace organisations which are pursued along with negotiations between the political parties and independent third parties at the negotiation stage. It investigates these two stages of conflict resolution processes through three major aspects: backchannel communications, peace and conflict resolution organisations, and official negotiations. These three aspects focus on the roles of political parties, government officials, opposition leaders, the IRA and PKK, and their political wings and intermediaries. This thesis recommends that a peace agreement between states and opposition parties, consisting of sub-state armed groups and their political wings, requires the mediation of an independent third party. It argues that a conflict resolution process may encourage conflicting sides to consider talks and to enter into a negotiation process at the pre-negotiation stage. The process then supports reaching a peace agreement at the negotiation stage.

Based on a ten-month ethnographic study that took place in an on-line collective space of Russian-speaking queer immigrants who live in Israel/Palestine, this thesis pursues the question: what are the relations between violence and belonging? Contrary to most research that addresses immigrants and queers solely as victims of violence (racial or homophobic), I argue that violence should be seen as a complex social and psychic phenomenon that is constitutive of spaces of belonging and not just a background against which belonging takes place. I further argue that violence of a particular type never stands alone and that it operates not only through particular figures but through multifaceted relations between them.

This mixed-methods study compares the processes of violent conflict escalation and de-escalation in two pairs of neighbouring, sub-national regions in Indonesia (Poso and Donggala districts) and Nigeria (Zangon Kataf and Kachia Local Government Areas). Despite similar contextual features, this thesis demonstrates that inter-group tensions have only escalated into repeated episodes of widespread violence in one of the two research sites examined in each country. This thesis argues that the onset of, or escalation in, violent communal conflict involves complex processes that shift inter- group relations back and forth along a continuum, from more peaceful interaction between groups at one extreme, towards repeated episodes of collective violence at the other extreme. In the presence of inter-group tensions, interventions and constraints at different points in the conflict trajectory may prevent tensions culminating in violence, or prevent repeated episodes of collective violence from occurring. Analysis of the evidence suggests that violence at the sub-national level is more likely to occur during periods of political-institutional change that are accompanied by economic decline. At such times, the opportunity for groups to re-negotiate their access to the state is enlarged, as there are higher stakes that encourage groups to participate in both violent and non-violent forms of contestation. Furthermore, at such times, this thesis argues that the risk of violent communal conflict increases when the heterogeneous interests and grievances of group members converge under politically salient identity frames, in opposition to other such groups. This is particularly the case if the convergence of motivations is underpinned by a local history of political or socio- economic inequalities between groups, or the unequal recognition of cultural groups by the state. Furthermore, inequalities between the elites of politically salient groups (for example, in terms of access to power and resources) drive their own interests in mobilising the wider group in collective action. However, power- and resource- sharing, as well as efforts to redress inequalities, can help to de-escalate tensions. Underpinning the shifts of inter-group tensions along the peace-violence continuum towards collective violence are those processes that focus public attention on inter- group differences rather than similarities. Such shifts are also underscored by constellations of actions and events that link past and present, and facilitate the mounting and staging of violence along salient identity group lines (such as the use of emotive group symbols, derogatory slurs, strategically targeted violence and other acts that invite violent reprisals). However, shifts towards more peaceful interaction tend to be driven by events and actions that focus public attention on group similarities and seek to redress inter-group tensions. The overarching argument of this thesis is that in the presence of inter-group tensions, sub-national outbreaks of violence are not always inevitable in plural societies. Supra-local tensions can stimulate communal violence, but repeated episodes of violence tend to occur when there are local roots, particularly those pertaining to inequalities.

Within the field of transitional justice, knowledge of many aspects of post-conflict reconciliation continues to evolve, including the use of public apologies. This thesis considers in particular the role and politics of post-conflict apologies by former rebel groups, attempting to fill a gap in the current literature on political apologies that fails to consider apologies by rebel groups. It provides initial insights into the question, 'why do rebel groups apologise in some cases, but not in others?' This is achieved by comparing three cases - two where rebel group apologies are present, Northern Ireland (Provisional Irish Republican Army) and South Africa (African National Congress), and one where there is no rebel group apology, Cyprus (EOKA) - through three variables, security, politics, and society to determine possible influences on the apology outcome and the motivations of rebel groups to offer an apology, or refuse to apologise. Within each of the independent variables, specific factors are scrutinised and compared to determine their relative influence on the rebel groups' apology outcome. These include factors such as community violence and decommissioning, international influences, public opinion and elections, civil society, symbolism, and reconciliation. Comparative analysis of these factors revealed several motivations behind rebel group apology outcomes. Specifically, I argue that the transition to politics results in rebel groups behaving in a manner similar to that of states in relation to political apologies; that civil society has a stronger influence than formalised truth or reconciliation commissions; that the 'Age of Apology' has normalised the use of political apologies and had a positive socializing effect on rebel groups to offer an apology; and that apologies are used both for the purposes of reconciliation and for political gain. The conclusions of the thesis have implications beyond the specific groups analysed. The large number of societies emerging from civil conflict, grappling to deal with post-conflict trauma often count rebel groups among the relevant stakeholders. The question of how to transition and diffuse these groups will continue to pose dilemmas for policy makers and peace workers, making academic study of rebel group apologies a priority.

This dissertation contributes to the growing literature centered on civil war mediation. Relying on a rationalist framework of conflict and actors, and employing quantitative methods, the research uncovers a number of findings relating to the features that assist and impede civil conflict peace brokers. Paper one demonstrates the importance of the relative belligerent strength. Using disaggregated dyadic data, the analysis shows that insurgents whose capacity more closely matches the state are more likely to see mediation in the first place, and ultimately end their conflict through a settlement This argument is developed in the second paper, which shows how belligerent capacity is affected by natural resources. The presence of oil is shown to increase the relative position of the incumbent, lowering the likelihood of mediation and agreement. Paper three focuses on the interaction between the characteristics of the mediator and the belligerents. It demonstrates that mediation is more likely to be accepted when the incumbent and third party share institutional similarities. Notably, non-democratic states are shown to have a significantly higher demand for mediation led by non-democratic third parties. In paper four, which is co-authored with Kristian :=== Skrede Gleditsch, we extend previous research on mediation by assessing the predictive powers of features highlighted as important determinants for mediation. Our results suggest that a two stage model of mediation and success does relatively well in out-of-sample predictions. In total the dissertation makes a number of important contributions, including: using disaggregated data to facilitate assessments of competing mechanisms; adopting an innovative modelling procedure to better capture the selection effects underpinning mediation; and proposing a new means of result validation that offers a more comprehensive assessment of statistical results. In , ,I I • ( i I-.- -.- "; ~ J.... i ---- ~:= I • this way the dissertation bridges the gap between studies of civil war mediation, and theoretical and methodological innovations within the broader civil war literature.

Education for Peace (EFP) represents one of the longest running and largest peace education programmes of its kind. What began as a small pilot study in six schools in Bosnia and Herzegovina (BiH) became in a period of approximately twelve years mainstreamed throughout the country's educational system. Despite the programme's "political success", relatively little has been written on EFP and there has been no systematic independent evaluation of its evolution. Therefore, the question of how and why EFP's political success was possible remains to be explored. To answer this question this thesis adopts a hermeneutica1-constructivist approach, thereby enabling this study to chronologically reconstruct the emergence and "mainstreaming" of the EFP programme and critically assess its adoption across the whole ofBiH. By using a hermeneutic-constructivist approach, this study makes explicit that the programme's philosophical assumptions derive from the Bahi'i faith, a fact which until now has been poorly understood and very rarely acknowledged. And while using this framework to critically assess the adoption and mainstreaming of EFP throughout BiH, this study demonstrates that EFP' s political success was helped by a number of factors, the six most prevalent being: the prevailing conditions and needs in post-war BiH; the programme's surface discourse which appears to address the objectives of the international community; the newness of peace education as a tool for peace-building and reconstruction in post-conflict societies; the unfamiliarity of the Bahi'i faith; the fact that the programme went through only one independent evaluation; and the disconnect of the programme's content and the political process of its adoption. Furthermore, this study will draw to attention the accidental and contingent nature of the adoption process of EFP, highlighting the naivety that was prevalent on all sides - the EFP people who want to help by spreading their positive messages of peace and unity, however unreflective oftheir assumptions and the governing officials, decision-makers and funders who want to help and/or be seen as being proactive. This study does not question whether EFP or those involved in funding or supporting the programme directly or indirectly are genuine in their intentions. Rather this study aims to draw attention to the unexpected outcomes and the failure to properly consider the programme's assumption that has now resulted in a religiously orientated peace education programme being mainstreamed throughout an entire country just emerging fi:om a violent ethnic-religious conflict. It allows us to ask more general questions about the future of peace education and its use as a peace-building and reconstruction tool in post-conflict settings while taking a closer and critica1100k at the political processes that allows for its nationwide implementation.

The purpose of this work is to identify how people ethically justify the use of violence or harm-doing in conflicts they experience; and by comparing the processes they use with those of normative ethics, to review whether and how well existing theory addresses the issues people actually face in terms of ethical reasoning. If someone is in a conflict, and the next move may involve harm-doing, what should they be thinking about and why? The work began from an open position as to how people might make their justifications, though with the expectation that they might do so in varying ways reflecting their own disparate experiences. An exploratory approach was therefore adopted, involving idiographic, qualitative methodology. Grounded theory was selected from a range of five such methodologies partly because of its commitment to building “substantive theory” which emerges from the data and also to provide analytic (but not population) generalisability to situations similar to those investigated. The primary data was collected from an interview study conducted amongst people in Sheffield, England. Secondary data from a profoundly different context (the Nuremburg Trials) was used as triangulation. A Model was constructed representing how people conceptualised conflict and reasoned about harm-doing in the actual conflicts they experienced. A literature review was then conducted covering the concept of conflict, four broad ethical approaches – consequentialism, deontology, contractarianism, and virtue ethics - and applied ethical writing relating to conflict–related harm-doing, including eg Just War Theory. The Model was then reviewed in terms of the literature and the literature in terms of the Model. The work contributes to the field by identifying and critiquing ways in which at the everyday, real-world level, people conceptualise conflict and reason ethically. There was a marked contrast between the Model and the conflict literature in one respect. The Model treated conflict as aggressive behaviour to the understanding of which knowledge of its purposes added little. Established theory understood it as goal-directed behaviour producing aggression because not all parties could have all they wanted. Consistently with its view of conflict, the Model saw the idea of common humanity and respect for people as the basis of ethical reasoning. As a result doing harm in self-defence and to protect the weak was all but taken for granted, provided this did not extend to large scale violence. People treated ethical reasoning as relevant to their lives and would borrow eclectically from all four broad approaches in actual situations – but rarely in any depth. This suggests the literature addresses real-world issues, but raises questions about its accessibility and practicality. The research also makes a methodological contribution. Normative ethics does not make extensive use of empirical data and part of the object was to investigate whether it could be useful. The work shows that idiographic methodology can help by identifying the meaning of real world issues for individuals, which meaning the ethical literature should address.Recommendations are made for nomothetic investigation of the prevalence of the views in the Model, and their relationship with national or religious cultures; and for particular developments of the ethical arguments against killing.