ABSTRACT The theoretical argument that emerges from my empirical study argues that food provisioning systems in Johannesburg, as a potential lens to further investigation of food supply systems in the developing south cannot be classified within a traditional-modern dichotomy. This dissertation proposes a new conceptual device – a food provisioning continuum – which should inform research on African food supply systems in the future. The process of locating this rich case within a broader theoretical paradigm to validate it and to provide it discursive space, however, is not objective or without friction. I argue that it is possible to choose to locate rich empirical material in different conceptual frameworks, related not only to its applicability, but also to how the research may be valued and seen to extend knowledge. The expectation of the research community and the epistemological demand of new research, for a Masters dissertation is that the scholarly work will build on and extend existing knowledge. It is assumed that thorough research will challenge the boundaries of knowledge and that the candidate, after having undergone this academic rite of passage, will graduate from being a student to being a colleague within a research community. However, the process of creating new theory and advancing existing theory is not quite an objective or frictionless process as it first appears. Research in the south is validated more highly if it is located within, or builds upon international/northern theory even by research forums in the south like the NRF. The pressure for researchers from the south to locate their research in conceptual frameworks from the north – in order to be validated – appears to be one of the rules of the game. While this is validation as part of an academic exercise may be necessary, the practise entrenches spatial or geographical hierarchies within academia and academic discourse. The epistemological process of forging new theoretical frontiers is thus a constructed, unnatural space fraught with less critical valuing systems than are expected to be present within academia, no less within the discipline of geography.