Abstract

Evaluations in the US (Coker et al., 2016; Johnston et al., 2018; Jouriles et al., 2018; Katz and Moore, 2013; Kettrey et al., 2019; McMahon et al., 2015; McMahon et al., 2018; Senn and Forrest, 2016) and in the UK (Fenton and Mott, 2018; Gainsbury et al., 2020; Roberts and Marsh, 2021), show that bystander interventions have the potential to reduce violence and abuse. Yet it is not clear whether: participants become active bystanders after the intervention (Katz and Moore, 2013; Kettrey et al., 2019; Levine, 2020; Storer et al., 2016); nor how and why interventions work (Levine et al., 2020; McMahon et al., 2015; Storer et al., 2016). We evaluated a bystander intervention to address what works. While we identified interactive techniques as the best modes of delivering the intervention, we found that the myriad of violence and abusive behaviours, which the intervention sought to disrupt, ensured that ‘a clear dividing line’ (Barad, 2014:169) between unacceptable and acceptable behaviours could not be established, so that bystanders know when to intervene. We argue that: becoming an active bystander is non-linear; bystander interventions need to be (re)-purposed; and concepts of harm must extend beyond a legal framework to achieve social justice.