During the last decade there has been intense controversy over the nature of contemporary armed conflicts and their connection with religion, ethnicity, crime and natural resources. Central to the discussion is the question of why armed organisations use violence against civilians. Recent contributions underscore the self-interested behaviour of individuals and suggest that war, violence and collaboration are increasingly driven by personal ambitions rather than political goals. Combatants, warlords and politicians seem less interested in victory than in satisfying their lust for power and money, usually at the expense of the population; meanwhile, fearful but rational civilians try to exploit the opportunities that conflict throws up, engaging in individualistic alliances with armed organisations and even prompting the use of violence against their fellows. These trends, it is argued, are compounded by the convergence between crime and warfare, and the Colombian conflict is often cited as a typical example. This dissertation examines these claims by studying a recent outbreak of violence against civilians in the Middle Magdalena Valley, in Colombia, that left nearly 2,000 civilians dead and more than 110,000 people forcefully displaced. Based on data obtained from official and unofficial sources on conflict and violence, interviews with key informants and news reports, the dissertation argues that violence was used by armed organisations and, to a lesser extent, by the state, to extend and strengthen the territorial and political control they had over the region. Despite the salience of illegal economies, there is no evidence that economic motives have significantly contributed to the production of violence. Furthermore, civilian collaboration went beyond the narrow sphere of private interests as politicians, parties and social organisations took part in alliances with armed organisations and aimed to advance the interests of the social groups they represented. The dissertation thus challenges common misconceptions and influential contributions in the field of armed conflict and political violence; it also sheds light on the nature of the Colombian conflict.

This thesis analyses representations of urban violence in Colombia within four cultural products published/released in the time period 1990 to 2005 . The cultural products belong to genres commonly regarded as distinct, and divided between ‘fiction’ and ‘non-fiction’ – a novel, film, ‘testimonio’ and documentary. Methodologically, the analysis focuses on each cultural product as a whole – the text itself and its marketing paratext. In this focus on the cultural product as a whole, it also considers the role of the audience in the consumption of the cultural products and their themes. The theme the thesis specifically engages with is the representation of violent actors, and focuses in particular on their status as fourth world inhabitants. The fourth world is a theoretical category developed by Manuel Castells to describe spaces which are excluded from global networks and flows of information, resulting in ‘black holes’, such as favelas, inner city ‘ghettos’ and slums, in which inhabitants are unable to gain access to services and regular employment. The thesis looks at the development of myths surrounding these spaces and their inhabitants, and the role played by cultural products in constructing and perpetuating divisive myths. It posits a growing globally homogenised representation of the fourth world inhabitant as violent and destructive, creating a binary between fourth and first world inhabitants to which the representations in these particular Colombian cultural products are linked. Overall, the thesis argues that the representation of violent actors in Colombia, and in particular the city of Medellín in this time period, illustrates that the distinction between fiction and non-fiction has collapsed, due to the strength of myths surrounding fourth world figures.