The Initiative

The Violence Prevention (VIPRE) Initiative explores novel approaches to preventing (state-led) political violence. It suggests that it is possible to think about preventing violence in a similar way to that in which we think about preventing, or minimzing the damage caused by public health problems like traffic accidents or smoking. Efforts to prevent these problems focus not on ‘original causes’ of harm but on mitigating the risk of harm and damage caused by placing ‘intervening’ obstacles or ‘firewalls’ in front of these risks/harms. The project seeks to theorize, empirically explore, and test similar barriers or firewalls vis-à-vis political violence by drawing on insights from organisation studies, microsociology, and international political sociology.

The VIPRE project will: 1) theoretically develop an approach to preventing political violence exploring the role of the ‘organisation’, ‘circulation’, and micro ‘practices’ of violence (an ‘OCP’ model of prevention), 2) empirically ground this theory through a microsociological study of military training regimes focused on interrogation, detention, and counterinsurgency practices that draws on textual, visual, and ethnographic data, and 3) programatically synthesize its theoretical and empirical components to provide insights for practitioners working in the field as to how to better prevent political violences like torture, the targeting of civilians, and other related violences. The VIPRE project is situated at the center of scientific and policy-making work on security, human rights, and violence among diverse disciplines. It will contribute both to these scientific literatures and wider Swiss expertise in these areas, coordinating with pertinent organisations in Geneva, in particular.

Below, a more full and technical description of the project is outlined.


Technical Description of the VIPRE Initiative

Imagine torture. One probably thinks of a helpless victim. A dark prison cell. Guards. Suffering. And – of course – a malicious perpetrator, of one sort or another. This too has been the approach of most social scientific inquiries into violences like torture. Explicitly political violences that are categorized as human rights abuses tend to be conceptualised in terms of a relationship between perpetrator and victim whereby the perpetrator is motivated, to one degree or another, to harm their victim. This conceptualization has significant understandings for how we think about the prevention of violent human rights abuses. Consider Figure 1, which schematizes a classical understanding of preventing state-led violent abuses in terms of a human agent who begins as a ‘non-torturer’ but becomes a ‘torturer’ at a certain decision point, whereby a combination of originary (efficient) causes leads to their becoming-torturer. In order to prevent this occurrence, social science has generally theorised measures designed to prevent any reaching of a ‘decision point’ of violence. These measures – human rights, ethical or moral education, legal prohibitions, etc. – are hoped to make any such decision less likely by ‘pushing back’ against any motivating factors. In short, the majority of these preventive measures are designed to act on the subject (a policeman, an interrogator, a soldier) before they may come into contact with a potential future victim. The first and central goal of the VIPRE project is to move beyond this model.


Imagine, now, a traffic accident. One thinks of one or more vehicles, in a collision, with deaths or injuries. Crumpled crash barriers, and ambulances. Note, first, how unlike vis-à-vis torture our thinking about the context of such a traffic accident already includes preventive measures. That is; we may think of the crash barrier or the deployed airbags, or any other number of safety features used by modern vehicles. Those elements represent an entirely different type of prevention to the classical one outlined above. Consider, for example, the case of a:

4 A.M. [car] crash that was classified by police as caused by a drowsy driver. Yes, if the driver in question did not drive past his or her ‘bedtime’ (driver factor) the crash would not have occurred. However, the crash could have also been prevented by a drowsy-driver detection system (a vehicular factor), a road-departure warning system (a vehicular factor), or an effective rumble strip that alerts the driver if leaving the lane (environmental factor). (Sivak and Tsimhoni, 2008: 8)

Preventing deaths in traffic accidents involves both an ideational component (countering the view that it is socially acceptable to decide to drive excessively fast, fatigued, or intoxicated) and the introduction of material and semiotic elements (crash barriers, clearly legible road signage, etc.) that prevent harm without relying on tackling (human-ideational) causes per se. The VIPRE project refers to these types of prevention as ‘indirect’ or non-causal and seeks to build similar mechanisms preventing violent human rights abuses. The central goal of the VIPRE initiative is, thus, to both theorise and test the empirical validity of such a ‘non-causal’ or ‘indirect’ form of prevention vis-à-vis explicitly political violences carried out by states and to construct a programmatic basis for its implementation in the real world:

Main Vipre Project Goal: Theorise, test, and construct a programmatic basis for an ‘indirect’ or non-causal means of political violence prevention.

To appreciate this central goal of the VIPRE project we first must conceptualize its constitutive elements more thoroughly, as well as lay out scope conditions and definitions. Firstly, scope. Throughout when we speak of preventing violence or violent human rights abuses we are referring to violence and abuse by state forces. While the VIPRE project has the potential to extend to non-state violence, the distinct contexts of non-state violence force us to bracket them for the purposes of this stage in VIPRE’s research agenda in order to make rigorous and generalizable inferences regarding a narrower set of cases. Violent human rights abuses are defined as acts contravening codified and widely (jus cogens) held international legal norms. This partially accords with a ‘minimalist’ definition of violence (the protection of negative rights) (De Haan, 2010). Torture, more specifically, is defined as per Articles 1.1 and 16 of the 1984 Convention Against Torture (CAT). With these preliminaries aside, we move to conceptualization. The model of prevention outlined in Figure 1 suggests – in most cases – that the training of military or security forces must seek to translate ideas (respect for human rights, or the sanctity of life) into gaining purchase over a practical set of activities (interrogation, detention, combat operations, etc.). The problem, as social scientists have long known, is that training – of all kinds, but especially military or physical training – is practical and most effective when it involves learning by doing (Mauss, 1934; Wacquant, 2004). Employing the ideational privatives of human rights (do not torture) and relying on mechanisms of deterrence (“if you torture then you will be punished”) neglects the ‘logic of practicality’ (Pouliot, 2008) underlying military training and violent abuses themselves. Neglecting this logic of practicality blinds us to:

  1. The situational organisation of violence at the moment of its practice (its semiotic structure).
  2. The materiality of those violent practices, and their organisation (their material-semiotics).
  3. The subjectification of individuals towards violence through a global circulation in (visual, textual, and ‘cultural’) violent knowledges and/or materials.

In our view, studying these three elements and recognising the ways in which they might lead to violent abuses leads us to an alternative conceptualisation of prevention to that in Figure 1. We term this model the Organisation, Circulation, and Practice (OCP) model of prevention. Each of the three blind spots listed above are connected to this OCP model of prevention. Point 1 refers to the ‘organisation’ of every situational moment of practice. It refers to activities that are situationally structured and negotiated both at the moment they are enacted and through the human subjects who carry out these acts embedded in institutional or local environments that prescript, to some degree, their actions. Of concern here are the institutional environments in which practical tasks are carried out (e.g. bureaucracies), the wider situational dynamics of a place or space (e.g. fighting in a distinct culture), and the group dynamics of individuals carrying out tasks in tandem, the bodily comportment of individuals and groups carrying out tasks in concert, and so on. All these aspects ‘build meaning’ when carrying out particular tasks and may make violence more or less likely. Point Two refers to the organisation of practical actions in terms of their material as well as social aspect. Here, the basic claim is that people are “more prone to behave evilly when surrounded by evil materials.” (Clegg et al., 2013: 336; Austin, Forthcoming 2017) Bruno Latour (1999: 176) provides the example of gun control and its basic policy principle that more guns create more violence. As he puts it, the gun “enables, of course, but also instructs, directs, even pulls the trigger- and who, with a knife in her hand, has not wanted at some point to stab someone or something? Each artefact has its script, its potential to take hold of a passer-by and force them to play roles in its story.” (Ibid) Similar material processes, the OCP approach to prevention suggests, might be causing people to carry out violent abuses more commonly than if they were not present. Point 3, finally, refers to the ‘circulation’ of the knowledges and materials that make the organisation of the practice of violence possible. Human beings are sometimes ‘subjectified’ by violent knowledges that circulate very freely across borders and these knowledges make the emergence of violent human rights abuses more rather than less likely. (C.f.: Latour, 1986) For example, textual training manuals and visual sources of knowledge (films) have provided a knowledge of ‘how-to’ torture employed by violence workers across borders (Austin, 2016; Ten Brink and Oppenheimer, 2012). In addition material objects, again, sometimes contain scripts for action that encourage violence and which – like texts or videos – circulate freely across borders. For example, the circulation of electric stun guns like the Taser across borders has led to a convergence in the use of electrical torture across the world (Austin, 2016; Dymond, 2014). By placing these three elements – rather than motivation, interest, or similar processes – at the heart of its analysis of violent human rights abuses, the VIPRE project thus conceptualizes these abuses as dynamic non-deterministic processes constituted by multiple (sequentially-linked) moments of practical action that encourage or discourage an eventual ‘slippage’ into aberrant violence due to their incorporation of circulating objects, knowledges, and human persons in particular ways. It sees violence not as necessarily decided but as sometimes unknowingly transitioned towards.

Indeed, the OCP model of prevention represents a breakthrough because like the preventive measures embedded in road traffics safety schemes, we argue that it can also work to bypass the question of original (efficient) causes. That is, it does not rely on claiming human motivations must be present for violent human rights abuses to emerge. This is especially important because many individuals who have tortured others describe their actions as having lacked agency, as not having involved any identifiable ‘cause’ or choice. Violent abuses are sometimes fundamentally unexpected even for perpetrators themselves. We refer to this as the non-purposeful emergence of violence. Consider an example. In 2008, Kevin Bell was a platoon leader in the US Army, patrolling in Afghanistan when he and his team came under ambush. One man was killed. A few days later, Bell thought he had captured the man responsible for that death. He describes his thoughts with these words:

I scrambled for the right reason to make a decision. Torture. Don’t torture. Where there should have been an answer there was only darkness. It would be wrong to say that I made a choice, but I finally broke the silence to ask an unrelated question. Soon the water boiled. The tea came. We took an awkward photograph for the record, walked back to our vehicles and left… the outcome of the situation was determined by luck. (Bell, 2011: 43)

Examples like this are very common.2 In advanced military forces, where troops are trained in the laws of war, receive ethics classes, and are – in a majority of cases – never ordered to torture, the risk and actual occurrence of torture emerges nonetheless. Not due to a ‘decision’ but due to a combination of situational, material, and environmental factors, as well as the fact that they have been subjectified towards the possibility of torture through the circulation of knowledge of its practice across borders. Both social scientists and lay observers tend to fail to pay enough attention to this non-purposeful form of the emergence of violent human rights abuses – where there is no decision – due to an ‘intentionality bias’ in studying social phenomena like state-led political violence.3 Nonetheless, focusing on this non-purposeful form of violence actually situates the VIPRE project at the state of the art of both social science and neuroscience. As Erik Ringmar (2016: 1) has described it, the idea that “actions are caused by consciousness” (decided) is – at least partly – “folk wisdom.” But this wisdom is disproven by neurological research that shows we carry out many actions before we are conscious of being about to do them (Haggard, 2010). This non- deliberative understanding of action does not mean that our consciousness does not make many decisions, and cannot halt our actions, but it does imply that we frequently carry out certain behaviours unconsciously. Critically, those unconscious behaviours emerge, as Ringmar (2016: 9) continues as “a result of the patterns of behaviour which… [are] internalized by our neurological systems. That is, they are a result of habits.” The difficulty faced by military training regimes that attempt to prohibit torture is – quite simply – that merely educating people in human rights norms does not per se create such habits and, instead, there are appears to be a certain ‘norm of torture’ containing scripts for action that are circulated across borders, organized in particular ways, and which result in the emergence of particular practices of violent human rights abuses at particular points in time and space.

The VIPRE project suggests we take this non-purposeful or non-deliberative understanding of violence as a point of departure for developing an alternative and complementary type of prevention. In doing so, the OCP approach to prevention is founded on understanding the micro-practical structure of violent acts but – with Ringmar – in connecting those micro level practices to both ‘meso’ level organisational practices (e.g. bureaucratic structures) and ‘macro’ level forms of knowledge and material object circulation across borders. It asks us to consider how our micro-practical behaviours are connected to scripts and habits, as well as objects and structures, that extend far beyond ourselves in terms of the time, space, and purpose of their creation. We refer to this as a material-semiotic approach to analysing social processes. As Law (2009: 141) puts it, material-semiotic approaches “treat everything in the social and natural worlds as a continuously generated effect of the webs of relations within which they are located.” Specifically vis-à-vis torture, Austin (2016) has argued that its emergence can be seen in these terms. Rather than seeing the figure of the torturer as deliberatively ‘deciding’ to torture someone else, Austin suggests the existence of a norm of torture which exists outside the realm of human decision and/or deliberation, writing that:

Spaces of violence are structured through a set of material-semiotic relations that are equally as prescriptive as the juridical contours of human rights or related ‘liberal’ regimes… [violence is sometimes] subjectlessly emergent… whatever the means by which political exceptionalism emerges, the practices of violence drawn on afterward may not be hierarchically imposed (‘decided’). Rather, they may flow from the flat yet thick set of relations that exist between states of all type, persons of all proclivities, archaic and novel information technologies, visual and textual inscriptions, and non-human objects. (Austin, 2016: 15-16)

The VIPRE project takes this focus on the ‘subjectless’ emergence of violence forward through an interdisciplinary collaboration between Management and Organisation Studies, International Relations and Sociology (MIRS). We do so (see Section 2.2, Interdisciplinarity, below) in order to extend the insights of the material-semiotic approach as articulated by Austin within International Relations (IR) into a form in which it can be practically applied at the ‘human’ level and so leveraged towards prevention rather than any more ‘macro’ or system level explanation for the continued emergence of violent human rights abuses. Indeed, the VIPRE project is essentially seeking to answer the question Austin (2016: 16) poses at his conclusion:

What kind of political program against torture and [other] violence[s] is possible that can tackle this everyday flat circulation of torture objects, knowledges and texts?

To answer this question we must complement the OCP approach to prevention with the insights of both Management Studies and Sociology. In particular, the new microsociology of violence articulated by Randall Collins and other figures shares our focus on understanding the micro-practical contours of violent interactions at the most granular level (see Section 2.4, State of the Art, below). Of particular interest is Collins’ “micro-situational evidence” demonstrating that “violence is hard. No matter how motivated someone may be, if the situation does not unfold so that… fear is overcome, violence will not proceed (Collins, 2007: 20).” Violence will not proceed, even if a motivation is present, Collins suggests, because it always depends on something additional: emotion, affect, materiality, etc. Introducing this insight into the OCP approach is key. It suggests that observing micro- practical sequences of action, bodily movement, and facial movements might allow us to see where violence ‘slips’ into action before conscious deliberation. More than that, it would suggest that the presence of certain material objects, environmental conditions (situations), and/or human beings who have previously been subjectified by violent knowledges makes violence ‘easier’ than it would been in its absence and, following this logic, we may be able to alter the relations between violence workers and their situations, material environment, and background or ‘peripheral’ knowledges in order to make violence harder. Indeed, this approach can be seen, to conclude, as seeking to identify certain entry points into what Tim Ingold (2000:195) has described as the taskscape of violent practices. This concept is drawn in analogy to that of landscape and affirms that:

One of the outstanding features of human technical practices lies in their embeddedness in the current of sociality… just as the landscape is an array of related features… the taskscape is an array of related activities. And as with the landscape it is qualitative and heterogeneous (Ibid).

Ingold makes this claim to counter the idea that “tasks are suspended in a vacuum” and argues that we must not separate “the domains of technical and social activity” (Ibid). We must connect human motivation with a whole landscape of other supporting elements. In doing so, the VIPRE project’s model of prevention proposes we may be able to locate elements in this taskscape that can be altered so as to stop the emergence of violent human rights abuses. And this novel understanding of preventing state led violence is now schematized in Figure 2, depicted below.



Figure 2: The VIPRE Model of Prevention, click to enlarge (Austin, 2017).


Figure 2 depicts the VIPRE model of the trajectories by which an individual ‘becomes’ a torturer. In this model the taskscape of any military or intelligence practitioner likely to carry out torture is depicted in the right hand two quadrants of the schematic. These two quadrants effectively double the social spaces that must be of both scientific and public policy concern. But in doing so, notice how they also double the effective length of the ‘flow of time’ between a person transitioning from becoming a ‘non-torturer’ to a ‘torturer’. This doubling of time provides a set of as yet unknown potential preventive measures against violent human rights abuses (the question marks in the top right hand quadrant) whose present day absence from training regimes, human rights discourse, and wider policy- making discourses is – the VIPRE project contends – one of the principal reasons for the continued (re)emergence of torture and other violent human rights abuses across borders. These question marks, we suggest, are the points at which a kind of ‘road traffic safety scheme’ for political violence must be built; a set of material and semiotic preventive measures. It is thus that rather than marking the distinction between prior or pre-existing causes of violence (ideology, pathology, etc.) and the moment of becoming a torturer as a ‘decision point’, Figure 2 describes it as a transition point. This transition point marks the start of a practical sequences of activities that in the great majority of cases are unlikely to begin with the simple decision ‘I will torture.’ Instead, the taskscape of the emergence of violent human rights abuses may begin with a relatively innocuous task – manning a checkpoint, for example – or the use of a legitimate form of violence – interrogating a prisoner, for example. The presence of particular situational dynamics, peripheral (subjectifying) knowledges, material objects, and many other aspects may then render the taskscape more or less likely to enable the emergence of violent human rights abuses. VIPRE takes as its challenge reducing that risk.