This thesis considers how ‘restorative justice’ has emerged as a legitimate response to crime. It presents the beginnings of a genealogical analysis of ‘restorative justice’ as it applies to criminal justice contexts. It comprises a ‘backwards-looking’ component, in which accepted historical accounts of ‘restorative justice’ are problematised, and a ‘forwards-looking’ component, in which a partial history of discourse of ‘restorative justice’ is presented. I conclude that these silenced discourses might be read as an incomplete and partial history of discourse of ‘restorative justice’. That is, ‘restorative justice’ ‘makes sense’ as an approach to criminal justice partly because of the credence of these discourses, upon which it relies, to some extent, for discursive legitimacy. These diverse and divergent discourses cast the ‘restorative justice’ project not as the unified and stable ‘movement’ as which it is usually portrayed, but as a fragmented and shifting phenomenon, comprised of a loose and heterogeneous assemblage of practices with variegated historical antecedents. Additionally, I conclude that some concerns raised by various scholars in the field – particularly in relation to the potential of ‘restorative practices’ to impact negatively on already marginalised and disadvantaged populations – are validated by this genealogy. / Doctor of Philosophy (PhD)