Does torture prevention work? This is a pithy question but also a somewhat misleading one. Like “does torture work?” it presupposes a straightforward relation between an end, torture prevention, and the means to get to it. It directs us to think instrumentally about what tools we might craft and use to this end; what measures might “work” to address proximate causes. The answers we get may not be altogether wrong, but they are likely to be muddled. That is because, as Danielle Celermajer shows in The Prevention of Torture: An Ecological Approach, the question of whether or not torture prevention works does not help us to disentangle conditions that produce torture from the strategies needed to prevent it.

To do that we need a better question. Celermajer's opening one is: “what is it that causes, conditions, and sustains torture?” (8). Torture, this question recognizes, is not merely, or even significantly, a problem of the proximate causes of torturers' objectives or dispositions, or the material conditions of detention. Rather, it is caused, conditioned and sustained by what Celermajer refers to, citing Reference BronfenbrennerBronfenbrenner (1992), as its ecology.

An ecological approach to torture takes situational factors seriously. We have no shortage of evidence that these are what count when it comes to explaining systematic institutional violence. The problem, according to Celermajer, is that strategies to prevent torture have not made the most of this evidence. Everyone keeps prosecuting, monitoring, and training. Because all of these strategies seem to have some positive effects, people keep stoically plugging away. But because none are informed by ecological diagnoses of conditions that enable torture, they tend to miss the mark, or play some other role than torture prevention.

This is where Celermajer's approach comes in, and in particular, its emphasis on articulating a Theory of Change. Every torture prevention strategy has a theory, but usually it is implicit. Celermajer thinks that if it can be made explicit, and causally capacious, then it might be possible to disaggregate and examine the conditions that enable and sustain torture, and design empirically testable interventions to prevent it. That is the task she sets herself in this book.

It is a formidable task, which calls for a lot of preparatory work. After a number of chapters interleaved with vignettes that go through a typology of current prevention strategies (Chapter 1), evaluate these strategies' effectiveness (Chapter 2), discuss causality (Chapter 3), and explore the worlds that produce torture (Chapter 4), in Chapter 5 Celermajer proposes a new social imaginary for the prevention of torture, one that “can support a narrative of causality that encodes systemic factors” (214). From there she takes theory into the field, to try out a model for ecological diagnosis of torture.

Chapters 6 and 7 recount the torture prevention project that Celermajer designed and led in Nepal and Sri Lanka. With European Union funding, the project worked with local partners to establish research teams in each country, and solicited the cooperation of police and military forces. Thereafter it proposed modified training for police and soldiers that “would not be oriented to changing individuals' knowledge or values, but rather crafted in such a way as to encourage personnel to address situational factors” (285).

If conventional human rights trainings that concentrate on individual dispositions are ineffective in preventing torture (see Reference WahlWahl, 2016), then why opt for a modified type? Celermajer concedes that access was one reason. Ecologically minded torture prevention strategists have to tread a path between a theory of change that recommends certain kinds of interventions, and situational factors that constrain them. Where research points to situational factors in the police and military forces that sustain torture, the only way to get directly at these is via police and military personnel. And because human rights training “already constitutes the imaginative repertoire of cultural reform in security” (299), training is a convenient vehicle for admission into the worlds of torture.

However, admission has its costs. During the torture prevention project in Nepal and Sri Lanka, these gradually compounded. For one, the training rubric once affixed to the project could not be so easily removed. Even after the project had moved on to identify a subset of mid-level participants who would take the lead in setting up and trialing pilot projects to address factors that they themselves saw as causing, conditioning, and sustaining torture, people kept using the language of training, which made it hard for the project organizers to differentiate what they were doing from other human rights activities.

For another, the project conveners had to work with police and military personnel whom superior officers assigned to them. These were not necessarily the most committed or best placed personnel to do the kinds of work that the project had envisaged. Those who were did not get the resources or time they needed to effect significant change. Only some of the pilot projects tackled situational factors enabling and sustaining torture, though these raised interesting questions for Celermajer about how to calibrate attention to local, specific factors, and those inhering to the wider social and political ecology.

Celermajer is frank that the project did not prevent torture in Nepal and Sri Lanka. In that sense, it did not “work.” And yet, in another sense it did. It worked as an exercise to trial a diagnostic methodology. It illuminated various causal factors that contribute to torture's persistence, identified possibilities for its prevention, and pointed to constraints. In its author's concern to inform practical interventions to prevent torture with sociological theory, to learn from what the project did and did not achieve, to share its lessons with readers, to be critical rather than celebratory of its achievements, honest rather than circumspect about its shortcomings, this book is a model for us all.

Torture persists in countries around the world. Danielle Celermajer's The Prevention of Torture offers one approach toward better questions about why that is, and about what we might do about it differently. Hers is, of course, not the only one, and nor is it unique in many of its elements: even in Sri Lanka there has been at least one other attempt in recent years to devise a model for torture prevention that shares with Celermajer's a concern for situational diagnoses and interventions (Reference Cheesman, Fassin and HarcourtCheesman, 2019). But compared to others, hers is uncommonly sophisticated, far-sighted, and systematic. Above all, it comes with a compelling message: to change the ecology of the production of torture, an ecology of prevention is necessary.