Participation in Practice: A Case Study of a Collaborative Project on Sexual Offences in South Africa

Abstract

In this article we critically reflect on ‘feminist research methods’ and ‘methodology’, from the perspective of a feminist research unit at a South African university, that explicitly aims to improve gender-based violence service provision and policy through evidence-based advocacy. Despite working within a complex and inequitable developing country context, where our feminist praxis is frequently pitted against seemingly intractable structural realities, it is a praxis that remains grounded in documenting the stories of vulnerable individuals and within a broader political project of working towards improving the systems that these individuals must navigate under challenging social and structural conditions. We primarily do this by working with non-governmental organisations (NGOs) providing gender-based violence services in research conceptualisation, design and implementation. This raises unique and complex questions for feminist participatory research, which we illustrate through a case study of collaborative, participatory research with NGOs to improve health and criminal justice outcomes for survivors of sexual violence. Issues include the possibility of good intentions/good research designs failing; the suitability of participatory research in sensitive service provision contexts; the degree(s) of engagement between researchers, service providers (collaborators/participants) and research participants; as well as our ethical duties to do no harm and to promote positive, progressive change through personal narratives and other forms of evidence. Given the demands of our context and these core issues, we not only argue that there are no ‘feminist methods’, but also caution against the notion of a universal ‘feminist methodology’. Whilst we may all be in agreement about the centrality of gender to our research and analysis, the fundamental aims and assumptions of mainstream (Western) feminist approaches do not hold true in all contexts, nor are they without variance in mode, ideal degrees of participation and importance to social context.