All organizations involved in war are concerned with military cohesion. Yet previous studies have only investigated cohesion in a very narrow manner, focusing almost solely on Western state militaries or on micro-level explanations. This dissertation argues for the need to broaden this perspective. It focuses on three classic sources of cohesion – coercion, compensation and constructs (such as identity and ideology) – and investigates their relevance in the Second Liberian Civil War (1999-2003). More specifically, this dissertation consists of an inquiry of how the conflict's three main military organizations – Charles Taylor’s Government of Liberia (GoL), the Liberians United for Reconciliation and Democracy (LURD) and Movement for Democracy in Liberia (MODEL) – drew on these three sources to foster cohesion. Based on thirteen months of ethnographic fieldwork with former combatants, this dissertation contains five parts: an introduction, which focuses on issues of theory and method, and four essays that investigate the three sources of cohesion in the three organizations. Essay I focuses on the LURD rebels, and provides an insider account of their strategy. It shows that even decentralized movements like the LURD can execute strategy, and contends that the LURD fought its fiercest battles not against the government, but to keep itself together. Essay II focuses on coercion, and counters the prevailing view of African rebels’ extensive use of coercion to keep themselves together. Since extreme coercion in particular remained illegitimate, its use would have decreased, rather than increased, cohesion. Essay III investigates the government militias to whom warfighting was subcontracted. In a context characterized by a weak state and fragmented social organization, compensation may have remained the only available source of cohesion. Essay IV investigates identities as sources of cohesion. It argues that while identities are a powerful cohesive source, they must be both created and maintained to remain relevant. Taken together, this dissertation argues for a more comprehensive approach to the investigation of cohesion, and one that also takes into account mezzo- and macro-level factors.

Many of the most destructive natural disasters have taken place in situations characterized by armed conflict and insecurity: the Indian Ocean tsunami in Sri Lanka and Indonesia in 2004, the floods in Pakistan in 2011, the drought in Somalia in 2011 and typhoon Haiyan in the Philippines in 2013. Surprisingly little research has systematically explored how armed conflict affects natural disaster management, and how shocks from natural disaster influence conflict dynamics. This dissertation addresses these gaps by providing a qualitative and disaggregated analysis of the micro-dynamics underpinning the relationship between armed conflict, natural disasters and cooperation. It asks: what is the relationship between natural disasters and processes of conflict and cooperation in countries affected by civil conflict? To explore this question, the dissertation offers four essays that explore different facets of this relationship, focusing on the rebel group. Examining collaboration between rebel group and humanitarian actors during disaster relief efforts in the Philippines, essay I finds that rebel group behavior after a natural disaster is shaped by the level of hostility between combatant parties and the nature of the ties with the local population. Exploring the effect of natural disasters on conflict dynamics in the case of the Philippines, essay II suggests that natural disasters hinder rebel group recruitment tactics, by increasing hardship for rebel combatants and supporters, by weakening the rebel group’s organizational structure and supply lines, and by leading to a loss of territorial control. Based on a comparative case study between Colombia and the Philippines, essay III revisits ripeness theory and argues that the level of rebel group cohesion will help to predict whether or not rebel groups stay at the negotiation table until an agreement is reached. While a typhoon affected the Philippines during the negotiations, it did not “ripen” the peace talks. Finally, article IV explores pre-disaster evacuation across conflict-affected regions in the Philippines and India, and argues that both experience of previous disaster and the level of trust in government officials influence the likelihood of people evacuating. The dissertation has important implications for both disaster management and conflict resolution, and it calls for more dialogue between both disciplines.