Some Remarks on the Label, Field, and Heuristics of Perpetrator Research

Perpetrator Research has been around for some time. It has many merits. This is in no way meant patronisingly but with genuine appreciation. This text does not intend to praise, criticize or historicize the diverse approaches that labelled themselves or were labelled as research on those individuals who played diverse, often extremely violent parts in the context of collective violence. The following pages are a reaction to the institutionalization of perpetrator research – especially as manifest in the launch of the Journal of Perpetrator Research ( JPR). As part of this development, the design of conceptual frames that theorize collective violence seems to be a step in the wrong direction. Obviously, there is not only a long tradition of research on perpetrators but also one that puts the action, not the actor in the focus of studies on collective violence. Classical examples of such approaches are micro-sociological and social theoretical studies representing a variety of topics and epistemologies.1 However, the last few years have seen such developments also in the field of genocide studies. Let me illustrate the point. In recent years, I have attended a series of panels and conferences whose titles indicated a shift in conceptualizing collective violence. In 2015 Erin Jessee and Timothy Williams hosted a panel entitled ‘New Horizons in Perpetrator Research’ at the IAGS conference in Armenia. The INOGS conference in 2016 saw a panel with a different twist: ‘The Margins of Perpetration: Questioning and Testing the Concept of Perpetrator.’ And finally, later that same year, the Marburg Centre for Conflict Studies hosted an event called On Collective Violence. Actions, Roles, Perceptions. Hence, the concept of perpetrators first seemed to open new horizons, then was questioned and eventually abandoned. It was given up in favour of social theoretical concepts, in this case actions and roles. I have discussed the benefits and problems of such a shift and the use of these concepts in some detail elsewhere and will soon come to some of them.2 When the Journal of Perpetrator Research published its first issue at the end of 2017, it seemed to oppose this development. And, interestingly enough, in their editorial the editors referred to yet another conference in 2015 as the inspiration or starting


Some Remarks on the Label, Field, and Heuristics of Perpetrator Research
Christian Gudehus P erpetrator Research has been around for some time. It has many merits. This is in no way meant patronisingly but with genuine appreciation. This text does not intend to praise, criticize or historicize the diverse approaches that labelled themselves or were labelled as research on those individuals who played diverse, often extremely violent parts in the context of collective violence. The following pages are a reaction to the institutionalization of perpetrator researchespecially as manifest in the launch of the Journal of Perpetrator Research (JPR). As part of this development, the design of conceptual frames that theorize collective violence seems to be a step in the wrong direction.
Obviously, there is not only a long tradition of research on perpetrators but also one that puts the action, not the actor in the focus of studies on collective violence. Classical examples of such approaches are micro-sociological and social theoretical studies representing a variety of topics and epistemologies. 1 However, the last few years have seen such developments also in the field of genocide studies. Let me illustrate the point. In recent years, I have attended a series of panels and conferences whose titles indicated a shift in conceptualizing collective violence. In 2015 Erin Jessee and Timothy Williams hosted a panel entitled 'New Horizons in Perpetrator Research' at the IAGS conference in Armenia. The INOGS conference in 2016 saw a panel with a different twist: 'The Margins of Perpetration: Questioning and Testing the Concept of Perpetrator.' And finally, later that same year, the Marburg Centre for Conflict Studies hosted an event called On Collective Violence. Actions, Roles, Perceptions. Hence, the concept of perpetrators first seemed to open new horizons, then was questioned and eventually abandoned. It was given up in favour of social theoretical concepts, in this case actions and roles. I have discussed the benefits and problems of such a shift and the use of these concepts in some detail elsewhere and will soon come to some of them. 2 When the Journal of Perpetrator Research published its first issue at the end of 2017, it seemed to oppose this development. And, interestingly enough, in their editorial the editors referred to yet another conference in 2015 as the inspiration or starting point of their endeavour: Encountering Perpetrators of Mass Killings, Political Violence, and Genocide, held at the University of Winchester. The editorial consists of a brief historical account of perpetrator research and it gives an outlook on the field, naming topics and concepts that should be addressed in future issues. Therefore, engaging with this editorial seems a fair starting point for discussing the state and usefulness of perpetrator research as an academic field, as a label and as a heuristic concept.
The authors define perpetrator, victim, and bystander as subject positions. 3 But how is such a position defined? Or to put it differently: how does a researcher know when an individual should be labelled a perpetrator? I suggest mainly by evaluating her actions during a perpetration process. So, the train conductor who knowingly transports a group of individuals that have been defined (as a group), denounced, and persecuted and who are sent to a place where she knows they will be treated badly (this is not a euphemism but an abstraction) may be defined as a perpetrator. Consequently, this person becomes relevant for perpetrator research because of her actions and knowledge about the impact these actions have; this is the context in which the person acts. There may be other approaches defining perpetrators (that may, indeed, also differentiate them from bystanders or onlookers), but it is difficult to imagine doing it without referring to actions and, at some point, to the actors' subjective interpretations of their own actions as well as of those of others occurring in the context of collective violence. As the authors themselves write, not only does the term perpetrator contain an element of legal and/or moral judgment, it furthermore may exclude acts of political violence (this is the wording used by the authors) that are not covered by a definition expressing a very specific moral and legal standpoint. 4 In the text this problem is pragmatically solved by the concept of 'democratic perpetrators' in the context of Empire-building and Empire demise, namely 'settler's violence' and 'mass violence'. 5 What we have now are (1) different concepts of violence ('mass', 'political', 'settler's') and (2) a definition that is still based on 'acts' as its main criterion. Both issues are rather easy to address from a perspective that conceptualizes the subject of interest by focussing on the action and not the 'subject position' or, as Aliza Luft puts it, on social categories. 6 Acts, actions, behaviour, doing can be much more readily defined (not completely without problems, though Of course, a second, equally important element is necessary for a satisfactory or rather appropriate definition of the field: the context. This is not something 'to be also looked at' or that 'has to be taken into account'. It should by no means be separated from the action. And I suggest the context to be collective violence rather than political violence. The reason is simple: the term is more inclusive and this benefits analyses. The adjective collective indicates the frame in which individual and collective action take place and should therefore be analysed. Collective violence consists of events, actions, and relations that are based on group-attribution: humans harm and are harmed because they belong to a group or are conceptualized as such. These external and self-ascriptions are, of course, not the only causes for the execution and experience of violence; they need not even be relevant for all actors involved. But they make it possible to conceptualize such actions as justified or as at least justifiable. An attack takes place and it is the right thing to do because the targets belong to a group that may be or even has to be fought. This is, of course, but one definition of collective violence. Others, for example Steven Barkan and Lynne Snowden's, add the intention of social change as necessary criterion. 8 Yet, that would exclude cases of violence that are collective but not political. The slave hunting that led to the destruction of entire communities was mainly economically motivated -the destruction was a side effect. 9 It is, therefore, the combination of context and action that defines the field of interest. There is more. The editorial's authors are not interested exclusively in 'how perpetrators are made and unmade', but also in 'acts of perpetration' (p. 11). These acts describe the 'collective commission of mass violence' (ibid.). I argue that this is still much too narrow an approach to describe, analyse, explain and, therefore, understand collective violence. For one thing, obviously, resistance and helping behaviour are part of the social reality of persecution; understanding them is crucial for the understanding of collective violence as a whole. What is more, investigating the formation of actions that help those who are persecuted shines a light on different motives and motivations but also on relations, circumstances, frames, and the like. A combination of sociological case studies, 10 historic network analyses, 11 statistics, 12 analyses of cultural frames 13 and political science approaches 14 help to discover the complex interactions of all these aspects and levels. 15 When this is done, it becomes clear that helping actions were motivated very differently. Individuals decided by themselves to help, were asked, persuaded, pressured, paid or even did not know that their actions were of a helping nature. Some of those who helped others to escape the persecution sexually abused those in need. Again, we may be better off conceptualising such processes, relations, events etc. as actions rather than labelling the individuals involved as helpers or rescuers.
Furthermore, an analysis of actions shows that the mechanisms determining the choice of a specific action (Handlungswahl) do not have a moral bias. Even though the moral nature of an action and the consequences it has for others differ greatly from one case to the next (at worst a life is at stake), this does not necessarily affect the dynamics of the genesis of actions which concern more or less visible motivations for action. Some are convinced that they are doing the right thing, some are paid for their actions, others just do what they are asked to do. And considered from a theoretical perspective the processes that constitute actions, i.e. the processes that lead to individuals' actions in the first place are of course very insightful -regardless of the nature of the action. The striking similarities in the genesis of actions of very different moral value can even lead to the question of whether a specific psychology of violence is needed or, indeed, possible. Comparative research on very different kinds of actions and supposed non-actions -and, again, not on the social categorizations of acting individuals -may help to answer this question.
But there is still more. Collective violence only exists due to dynamics that, for example, lead to visible forms of ethnification -put more abstractly, to action relevant categorizations. 16 Yet, these processes of norm changes need definitely not intend a later persecution. The dynamics leading to such norm variations or changes moreover do not necessarily have perpetrators as their agents. In fact, non-action plays a crucial part when existing norms subside and other norms come up. 17 I suppose that this is not considered news. The point here is that the perpetrator research approach is conceptually relatively ill-equipped to deal with these aspects. Violence and Resistance (2016). 16 Timur Kuran, 'Ethnic Norms andTheir Transformation Through Reputational Cascades', The Journal of Legal Studies, 27.2 (1998), 623-59. 17 Heinrich Popitz, ' Social Norms', Genocide Studies andPrevention: An International Journal, 11.2 (2017), 3-12. There are also some basic conceptual challenges that come with the label. For example, if a perpetrator is someone who takes part in the persecution of members of a defined group and if a helper is someone who helps to escape (if only for a moment) or to weaken the persecution, how should a person that does both be defined? This is not a problem when reconstructing action, chains of actions, and action-contexts. And to do so does not mean to ignore any determinant of individual action, neither personality traits, situational settings, biases nor social figurations, to name but a few. The point of action-centred approaches is that they focus on what has happened, on what individuals do, and not on the question to what category someone belongs.
One concluding remark on 'the dynamic process of perpetration' that features prominently in the editorial. If terms like perpetrator or victim are used, it is actually necessary to stress the dynamic nature of social processes. But what is it exactly that makes them dynamic? I would say, it is how any individual sharing a specific social world relates to other individuals, institutions, norms, social categories etc. in that shared space. Each individual action is a statement that either stabilizes or destabilizes a status quo. Norm breaches, just to pick an example, occur all the time for all kinds of reasons. How others react to them -by ignoring, applauding, punishing them promotes or prevents changes. This is what a researcher at some point may identify as an element of a very specific dynamic that eventually leads to an event of collective violence. The social world cannot be adequately analysed unless it is considered to be moving, fluid and changing. Central categories in the discursive realm of genocide studies or perpetrator research like ethnicity, identity or race are in no way fixed entities but produced performatively in the course of permanent actions. Collective actions determine, first, who or which group is considered a problem at a given point in time and, second and crucially, they also determine how the problem is solved. 18 So again, the central category is action. 19 To be clear about the intention of this piece of writing: Perpetrator research is a smashing enough label and the work done under it is at times outstanding -as the papers in the journal's first issue demonstrate. And I am sure that the editors will allow for submissions that do not necessarily use the social categories or subject positions made famous by Raul Hilberg. 20 My point is rather of a heuristic nature. I have simply argued that making actions the starting point of research on collective violence helps (not morally but theoretically) to achieve a more comprehensive and appropriate understanding. I stand to be corrected.
18 This point is a crucial element of action theories. Explicitly, for example, with Hans Joas: 'It is collective action itself that defines the problems it relates to; it generates motives and identities, shapes new social relations and communities, gives rise to profound changes in identity (conversion and regeneration), produces affectively cathected symbols and leaves behind symbolic attachments capable of structuring biographies.' Hans Joas, War and Modernity, trans. Rodney Livingstone (Cambridge: Polity, 2003), p. 192. 19 Cf. Luft, p. 152. 20 As editor-in-chief of Genocide Studies and Prevention I am well aware of the shortcomings that the term and concept of genocide have and its constraining semantic force for research on diverse aspects of collective violence.