This thesis presents a social ontology. It takes its problem, the emergence of social structure and order, and the relationship of the macro and the micro within this structure, from social theory, but attempts a resolution from the perspectives of contemporary French philosophy and complexity theory. Due to its acceptance of certain presuppositions concerning the multiplicity and connectedness of all life and nature it adopts a comparative methodology that attempts a translation of complexity science to the social world. It draws (...) both this methodology and its inspiration from the work of Michel Serres. After explaining this methodology, it presents a critique of the work of those prominent philosophers of multiplicity who have written on the social: Alain Badiou, Gilles Deleuze and Felix Guattari, and Manual DeLanda. Having argued for the need of a ‘non-unit’ of social organisation, it then unsuccessfully surveys the work of Michel Foucault and Gabriel Tarde in search of such a ‘non-unit’. It produces one by extracting elements from different theorists and then proceeds to offer a novel explanation of how these expectations first emerge from the ‘social noise’ and then go through a complex process of self-organisation to produce social structure. Apart from complexity theory, this explanation draws on the temporal ontologies of both Serres and Deleuze. In doing so, it argues that the social replication necessary for this self-organisation cannot be achieved through direct imitation. Instead, it draws on an idea from Stuart Kauffman and argues that this is achieved through autocatalysis. Finally, it argues that social structures and what is perceived to be social order are the effect of the codification, to varying degrees, of these emergent expectations. It concludes that this structure is at its most creative when on ‘the edge of chaos’, when at a point of social chaosmos. (shrink)