Deradicalisation according to the Radicalisation Awareness Network : a practice approach to what the EXIT working group propagates as best practice

(1)

Radicalisation Awareness Network

A practice approach to what the EXIT working group

propagates as best practice

Master Thesis

Political Science – International Relations European Security Politics

Universiteit van Amsterdam 31-08-2018

Student: Jeroen Beekink – 10370730 j-beekink@live.nl

Supervisor: Dr. R. Bellanova Second Reader: Dr. F. F. Vermeulen

(2)

(3)

Abstract

This thesis analyses the way in which the Radicalisation Awareness Network, as a part of the European Union, decides on and promotes best practices regarding deradicalisation. Across the eighteen practices that are listed by them there is real diversity of focus and goal of intervention, being both preventative and restorative at both early and late stages of radicalisation. Almost all of them focus on individual disengagement as opposed to collective disengagement. Furthermore, there seems to be a bias towards right-wing extremism and Islamism, as well on practices from North-western Europe. This analysis is done taking a practice approach to the subject.

(4)

1. Introduction

In my thesis I analysed what kind of deradicalisation programmes the Radicalisation Awareness Network (RAN), and by extension the European Union (EU), is propagating through its Collection of Approaches and Practices. More narrow, the RAN EXIT working group is taken as the actor under scrutiny. These best practices, and more specifically, their collection and publication is treated as a practice, in the Bourdieusian sense, by the RAN. This practice is the main focus and simultaneously the starting point of the research. Since the new millennium Western Europe has seen about a dozen notable terrorist attacks, starting with the attacks in Madrid in 2004 and the latest being a hostage situation in southern France at the end of March 2018. Several more, smaller attacks or would-be attacks have also happened in the Netherlands and elsewhere. These attacks in Western Europe and indeed the rest of the world can be collectively described as the fourth wave of terrorism, mainly inspired by religion (Rapoport, 2002). During the third wave in the 1970’s and 1980’s terrorists supposedly had other means and motives, leading Brian Jenkins to famously quote “terrorists want a lot of people watching, not a lot of people dead” (Jenkins, 1975). During this third wave the European Communities (now European Union (EU)) were not yet willing or able to formulate a common response to this phenomenon. Security was considered exclusively a national matter. This changed somewhat with the fourth wave. After the attacks on 9/11 the EU came with its first common policy document, the Conclusions of the Extraordinary European Council (in 2001) and the EU Action Plan on Combating Terrorism (in 2002) (Martins & Ziegler, 2017). After the Madrid bombings of 2004 the EU came with a Declaration on Combating Terrorism (Council of the European Union, 2004) and in 2005 it presented the first version of its Counter-Terrorism Strategy (Council of the European Union, 2005). This last document contains the pillar approach, Protect, Prevent, Pursue and Respond. Up until 2018 the EU revised its Counter-Terrorism Strategy twice, established three Communications and a Declaration. Lastly and perhaps most important is its European Agenda on Security from 2015. Despite these documents apparently showing increased cooperation between the EU member states, to most member states security is still a national matter (Coolsaet, 2010) (Hegemann, 2012). In the 2015 European Agenda on Security the

(8)

foreign fighters (European Commission, 2015, p. 12). One of these responses is the setting up and increasing role of the Radicalisation Awareness Network (RAN) and the RAN Centre of Excellence (RAN CoE), along with “developing effective disengagement/de-radicalisation programmes”.

I understand the RAN and the RAN CoE to be organisations working for the EU in facilitating EU member states by reviewing and collecting best practices and publishing them for others to read. They also organise meetings for frontline workers and specialists to discuss the recent developments and possible interesting programmes. The relevant practices are these meetings, the people that make up the RAN and those that attend the meetings are treated as a community of practice. Both these concepts have their roots in the work of Pierre Bourdieu and are further analysed and developed by Adler, Pouliot and Adler-Nissen, among others. During these meetings different topics are discussed and different people are invited for each separate meeting. Access to these meetings has proven to be hard and so the thesis is based on textual sources, specifically the above mentioned Collection of Approaches and Practices (Radicalisation Awareness Network, 2017). Within the Collection of Approaches and Practices document only the exit strategies in the corresponding chapter are analysed. Note that practice in my understanding has a different meaning than the best practices mentioned in said document. Practice as an analytical concept is explored and defined in chapter 2. Whenever the term ‘best practice(s)’ is used it will be about the best practices that are in the RAN document Collection of Approaches and Practices. Whenever the term ‘practice’ is used is considered to be practice in the Bourdieusian sense of the word. I focus on exit strategies because it is one of the most difficult issues of radicalisation and is even less researched than radicalisation. As the RAN itself notes, “[p]rocesses of deradicalisation and disengagement are […] complex psychological and social processes and “[a]ttempting to measure and isolate the effects of these interventions (especially in terms of success) is extremely difficult” (ibid, p. 3). Given the fact that deradicalisation is even less well understood as radicalisation and it is harder to define best practices for, I argue it is important to know what kind of programmes (collectively called best practices) the RAN, and by extension the EU, is propagating. This is because, according to the new sociology of knowledge approach, knowledge production is neither detached from social reality nor a precise reflection of it

(9)

(Adler-Nissen & Kropp, 2015). Furthermore, it is important, especially as a social scientist, to be able to understand and explain how this came to be and what the underlying

To answer this I have developed the following research question and sub-questions. My main question is: What does the Radicalisation Awareness Network consider best practices? The sub questions needed to answer this question are based on the framework of the practice approach. They, along with the main question, will be discussed in chapter 2.3.

With this thesis I hope to accomplish the following things. Research on the RAN and its working groups is scarce, it is mostly talked about second hand, as a practice or institution of the EU, not what it actually does in and on the world. I hope to add to this research. Secondly, practice approaches are fairly new to International Relations. Only in the last decade have there been serious attempts to take a practice approach to the world of IR. Consequentially, when it comes to methodology and actual research, literature is scarce and divided on how to do this. By taking the practice approach and seeing how it can be useful in understanding the world of deradicalisation and knowledge production on deradicalisation, I also hope to add to this gap.

The thesis is outlined as follows. I will first lay out the groundwork for a practice approach and will shape my research questions according to it. secondly, I will discuss EU counter-terrorism and counter-radicalisation policies in recent years and the debate in the academic community on how to understand this, to show how my work engages with the wider community. After that I provide a state of the art on radicalisation and deradicalisation in order to define the concepts that are needed to place the knowledge production by the RAN EXIT working group within a wider framework. Fourthly, I will show my research design and how I translated the previously mentioned concepts into indicators. In the chapter on analysis (6) I discuss my findings and in the final chapter I will discuss these findings and relate them back to the research question and place them back into a wider social context.

(10)

2. The practice approach

Practice theory is used as framework for the analysis. According to Adler and Pouliot it is the aim of the social researcher to understand and explain how politics works, in practice (Adler & Pouliot, 2011, p. 3). It has its basis in culture theory, poststructuralism and constructivism, but goes beyond them to “conceive the social as bundles of ideas and matter that are linguistically, materially, and intersubjectively mediated in the form of practices” (ibid, p. 14). Using practice as a way to view the social has a lot of advantages because it takes us beyond some dichotomies that seem to plague social science research. It goes beyond the structure versus agency problem because practice lets us understand both. Structure is understood in the way the habitus and field shape the dispositions and positions the relevant actors have. At the same time, they help explain the agency these actors have. Much in the same way it also moves us past the dichotomy of the logics of appropriateness and consequence and again lets us focus on dispositions and positions. Both material interest and ideas matter in shaping the position and disposition of an actor. It takes us out of the text and lets us use the text to shape our understanding of the world by making us think about the knowledge that has to be present in order for the text to be made as it is. Fourth, a practice approach goes beyond the problem of level of analysis, in a practice approach this is always the field. How to determine what constitutes the field will be clarified later. First a definition of practice will be given and explained. Than the aspects of habitus, field, doxa and symbolic power will be clarified. Thirdly, the community of practice and sociology of knowledge approaches will be explained. Lastly, some criticisms of the practice approach will be addressed and taken into consideration.

2.1 Definition and aspects

Adler and Pouliot define practice as “socially meaningful patterns of action, which, in being performed more or less competently, simultaneously embody, act out, and possibly reify background knowledge and discourse in and on the material world” (Adler & Pouliot, International practices, 2011, p. 4). Several elements of this definition need further explanation, namely socially meaningful patterns of action, competently and background knowledge and discourse. There is a difference between action, behaviour and practice, although they might seem the same (ibid, p. 5). Behaviour can be seen as a “deed performed in and on the world”, the concept of action adds a certain meaning to the behaviour. When this action is done more or less structured, patterned, it

(11)

becomes practice. This does not mean a practice approach is only concerned about structure, agency is always an integral part of it. Practice is both attached and detached of its actual carrying out. A social action can be part of a practice, but a practice does not have to constitute one action, or always the same action. Take for example the Cuban missile crisis, the planned placement of rockets on Cuba by the Soviet Union can be seen as a socially significant action, and was part of the practice of coercive diplomacy, which could have also been a practice without this socially significant action. The second aspect is that of it being more or less performed competently. This is more about social recognition than about any practice being carried out correctly or incorrectly (Ibid, p. 7). An action as part of a practice can be carried out completely incorrect but still be deemed competent by those in the rest of the field of practice. Thirdly, and especially important for this thesis is the understanding of background knowledge. In the acting out of the practice knowledge and discourse can be discerned. For an action to be carried out, certain knowledge has to be present. This is not merely distilled from what is being said, but also by way of exercise to peer into the mind and past practices of those actors that carry out the practice. In its carrying out the present knowledge is at once being made concrete and also possibly changed. In the present case, knowledge on deradicalisation is presented, it can have impact on real life situations because it can be accessed and talked about, and it also changes knowledge on deradicalisation because it selects some elements and discards others. To determine what this background knowledge is the concepts of habitus, field, doxa and symbolic power are used.

All these concepts originate from the highly regarded French sociologist Bourdieu. They are based on a relational ontology, meaning we are what we are because there is a social relation between us and the world around us. This is worked out in the concepts of habitus and field.

Habitus is defined as a “system of lasting, transposable dispositions which, integrating past experiences, functions at every moment as a matrix of perceptions, appreciations, and actions” (Pouliot & Mérand, 2013, p. 29). An actor’s habitus is built on past experiences which shaped the dispositions of the actor and shapes the perception and appreciation of reality and helps shape the actions the actor views as legitimate, what others would call a focal point. In other words, it is that which the actor is most inclined to do because of what it did and had to do in the past. Knowing an actor’s habitus lets us understand and explain better how the practice at hand came to be. Of course

(12)

their history. Some experiences will have been more significant than others. Every new situation is also different and every new experience can be slightly different from the past one. This also explains why habitus is not the sole explanatory variable of practice. The past experiences structure future action and shape agency but the field is also an important variable. Secondly, dispositions are also based on a certain know-how. In other words, the tacit knowledge the actor has to possess in order act out in the field. This is both knowledge about the object of struggle as well as knowledge of rules of the game.

The field is defined as “a social space structured along three principal dimensions: power relations, objects of struggle, and the rules taken for granted within the field” (Pouliot & Mérand, 2013, p. 30). Every field has different actors and these actors are related to each other in a hierarchical way. One actor is more powerful than the other. Power is understood as symbolic power. Because we have a relational ontology, power is also seen as a relational feature. As such “[s]ymbolic power takes place through a process of misrecognition and naturalization that is inscribed in habitus” (Pouliot & Mérand, 2013, p. 39). This does not mean a certain actor possesses no possible sources of power and still have symbolic power. It because of the powerful position in the field and its recognition by other actors that symbolic power is gained. Each field has an object of struggle. This object of struggle is what defines the field and separates it from other fields. The field ends where the effects of the field no longer exist (Ibid, p. 34). Pouliot and Mérand are not very clear what is meant by this. Changes in the field of coercive diplomacy, to use this example again, can have far reaching effects on things completely unrelated to the field. Contrastingly the economy of the United States could have taken a blow thanks to the Cuba crisis, and can, in certain ways, be seen as important for coercive diplomacy. It is up to the researcher in question to take into account the other two aspects of what defines a field and make an educated estimate of where the boundaries of the field are. Lastly, the rules that are taken for granted in the field, what Bourdieu calls the doxa. The doxa are those, written or unwritten, rules that structure the field. Being shaped by the habitus of all those in the field it can be seen as the natural way of things within the field. This harks back to the aspect of competent performance and agency within the field. An actor can act incorrectly according to the doxa but this can become a competent act after and thereby change the doxa of the field.

(13)

2.2 Community of practice and Sociology of knowledge

A community of practice is defined as “likeminded groups of practitioners who are informally as well as contextually bound by a shared interest in learning and applying a common practice” (Adler, 2008, p. 196). More specifically Adler defines it as “a configuration of a domain of knowledge that constitutes like-mindedness, a community of people that ‘creates the social fabric of learning’, and a shared practice that embodies ‘the knowledge the community develops, shares, and maintains’” (ibid, p. 199). In other words, a group of people, involved in a shared practice which embodies and shows their specific domain of knowledge. Going back to one aspect of the field, a specific object of struggle, knowledge is the object of struggle here. These communities of practice embody and act out a specific iteration of knowledge, as such they are considered to be a single actor.

Lastly, to better understand the aspect of knowledge, the new sociology of knowledge approach by Adler-Nissen and Kropp is used. In this view, knowledge and more importantly knowledge production, is neither detached from social reality nor a precise reflection of it (Adler-Nissen & Kropp, 2015, p. 159). In other words, knowledge production is not carried out completely detached of the social world. It is shaped by and interacts with the world it tries to understand and explain. On the other hand, knowledge is not a mirror-reflection of society either. To clarify they give four principles. First, the principle of symmetry, arguing we should approach all scientific knowledge claims equal, from those having almost law-like status to those deemed completely false. Secondly, to reject the internal/external divide. This means knowledge production cannot be seen detached of society. Moreover, it also puts importance on non-academic sources and sources from other disciplines. Consequentially, it lets us ask questions like how non-academic sources and ideas shape our understanding of the social world. These principles shape the view of what data to use and how to approach the data. More precisely, it argues we can and should treat all knowledge and knowledge production as equally valuable. With this in mind the choice for the RAN EXIT working group is a sound one. As will be shown the knowledge produced comes from both scientific sources as well as frontline practitioners. There is the question of whether the people that make up the working group approached the best practices they reviewed according to the same principles. While this is a valid question to ask it is beyond the scope of this thesis because this

(14)

simply not viable. Thirdly, the principle of situatedness. Taking a sociology of knowledge approach lets us place knowledge production within a certain setting, a situation or environment. The emphasis is on the scholarly discipline, or in other words the academic school. Keeping in mind that also non-academic sources are deemed relevant for this approach it is however broader than that. What is important is the notion that knowledge is believed to be produced in small communities. This is in line with the community of practice approach discussed earlier. Lastly, the principle of contextualism. This principle tells us to place the production of knowledge in a specific context, historically, geographically and socially.

2.3 Applying theory to practice. Research questions

These general principles of the practice approach need to be linked to the actual case of the RAN EXIT working group. Consequently, these form the sub question needed to answer the main question: What does the Radicalisation Awareness Network consider best practices? Some can be answered now, while others need further research and analysis first. This last group of question form the basis of the rest of the thesis.

1. What is the practice?

a. What are the repeated actions that make it a practice? b. To what extent are they competently performed? c. What is the actor in question?

2. What is the habitus?

a. What were important past experiences? b. What dispositions can we identify? 3. What is the field?

a. Where does the RAN EXIT working group stand in the field’s hierarchical power dynamics?

b. What is the object of struggle? c. What is the doxa of the field? d. Where are the limits of the field?

(15)

4. What background knowledge can be discerned from the data? 5. What is the situation in which the knowledge is produced? 6. What is the context in which the knowledge is produced?

The practice is the production and sharing of knowledge via best practices by the RAN EXIT working group. The repeated actions are the meetings they organise in which knowledge is shared and ultimately chosen to be published in their Collection of Approaches and Practices. To what extent this is done competently is more difficult. The question of it being competent is one of social recognition by other players in the field. Since this is thesis is exclusively about the working group as a single actor this question is for further research and cannot be answered here. As already mentioned the actor in question is the RAN EXIT working group. Following a community of practice approach allows me to see it as a single actor. They are bound by the context of them all being associated with the RAN EXIT working group, either as an employee working for the EU or and attendee of one of the meetings. Unfortunately, this does not allow me to look into any internal dynamics. Conversely, this is also not possible because this would require participant observation and several interviews. In other words, should this thesis prove fruitful this area of enquiry can be valuable as further research.

The habitus and the past experience which form the disposition cannot be identified so easily. These questions need further analysis before they can be answered.

Some characteristics of the field can be answered at this moment. The object of struggle is that of knowledge on deradicalisation. One could argue there is no struggle and that all tributes to knowledge and overall knowledge production is valuable in and of itself. However, since deradicalisation is such a difficult thing to grasp and even more difficult to produce knowledge about, publishing (producing) best practices constitutes a choice of what is and what is not (yet) considered knowledge about the issue. A decision is made and naturally some issues are left out. By making this choice the RAN EXIT working group is engaging in a struggle over what is and what is not knowledge about the issue. This is also what Bourdieu would see when looking at international knowledge production, “a socially situated activity that takes place in a context – a field – which is related to other fields in a dynamic struggle for power” (Berling, 2013, p. 60). The

(16)

struggle as being within the field and not between fields, in line with the rest of the practice approach explained above. An answer to the doxa of the field is also not answerable. When it comes to this last issue, I can only provide general rules for conducting scientific research.

Lastly, the background knowledge needed and embedded in the practice of knowledge production. Knowledge on deradicalisation is of course needed, but because it is such an essential part of this research this single statement is not enough. This question still needs to be answered in the theoretical framework. The question of situation is again about the internal characteristics and this is not answerable in this thesis, at least above surface level. The context is partly explained in the previous chapter on EU counter-terrorism policy but will be further analysed still.

Before going to the research question formed from the questions that could not yet be answered, a note on reflexivity. Bourdieu held reflexivity in high regard. To be reflexive as a researcher is to be aware of one’s own habitus and scholarly background. It is not clear whether he meant it should also count for the practitioners in question. Because the practice in question is knowledge production surely the scientists and other members that form the RAN EXIT working group should be reflexive about their work. To question cannot be answered at this moment. Part of the question is related to the situation the knowledge is produced in. Heeding reflexivity also forces me as a researcher to take extra care of minding my own habitus and scholarly history.

This leaves a number of questions still unanswered which are within the scope of this thesis: 1. What is the habitus?

2. What are notable past experiences? 3. What is the doxa?

4. What background knowledge is needed? 5. In what context is the knowledge produced?

These questions place the knowledge production by the RAN EXIT working group within a historical context, gives an account on the habitus, and places it (partly) within a specific field. By answering these questions not only is the main question answered, what the RAN considers best practices, but an explanation is also given how this follows out of its habitus and field.

Chapter 3 presents current EU counter-terrorism policy, which accounts for the question what the habitus and the past experiences are. In the following theoretical framework (chapter 4) an overview of deradicalisation research is given which will answer what background knowledge,

(17)

tacit knowledge, is needed. After that the research design for the rest of the thesis is given, which will show how the best practices are analysed. The analysis on the best practices from the Collection of Approaches and Practices document will show what kind of knowledge is selected by the RAN EXIT working group. I acknowledge that the question on background knowledge (chapter 4) is a large part of the thesis but is also one of the most important questions. To answer what best practices the RAN EXIT working group selected we should first have an idea of what there is to select from.

(18)

3. EU counter-terrorism policy

In this chapter I will outline the EU’s counter-terrorism policy and how to understand it. The last and so far most important document it is based on is the 2015 European Agenda on Security (European Commission, 2015). In analysing the EU’s counter-terrorism policy, most scholars agree that it has been event driven (Coolsaet, 2010) (Monar, 2015) (Bigo, et al., 2015) (Hegemann, 2012). Most notably the 9/11, Madrid 2004 and London 2005 attacks are named as catalysts that brought on more policies and better cooperation. Progress has especially been made in the ‘protect’ and ‘pursue’ pillars of the agenda, a few examples are the Dublin II Regulation, the Passenger Name Record Agreement and the European Arrest Warrant system. Monar, among others, also portrays that Europe’s response was to criminalise terrorist offences, as opposed to the United States’ War on Terror approach (Monar, 2015). Less progress has been made in the ‘respond’ and ‘prevent’ pillars, the last of these two was perceived as increasingly important after the Madrid and London attacks. These attacks showed that terrorism and radicalisation were not just an external problem but terrorist could also be ‘home-grown’. As of 2018 we can conclude that there has been significant progress in EU counter-terrorism policy. Bakker argues there is significant overlap with what science tells us to focus on, and what the EU and its member states are actually focussing on (Bakker, 2015). However, overlap between deradicalisation knowledge and practice is lacking and there is a bias towards Islamist extremism (idem). Den Boer and Wiegand also argue that there is significant convergence in counter-terrorism policy (den Boer & Wiegand, 2015). Finally, Monar also acknowledges this, adding that the EU has built a strong international character in counter-terrorism, but is limited internally (Monar, 2015). This limit is mostly due to national sovereignty over both security and social matters (Monar, 2015) (Coolsaet, 2010) (Hegemann, 2012). The present situation is one in which policy integration on traditional security issues in counterterrorism, as well as the international cooperation dimension, are fairly well developed but both it and radicalisation, under the ‘prevent’ pillar, are limited to national sovereignty. The focus herein has shifted to knowledge dissemination and ‘low politics’, because it would be more prudent and less politically sensitive (Martins & Ziegler, 2017). One of the primary outcomes of this turn to ‘low politics’ is the Radicalisation Awareness Network. However, with it there are again certain problems. Bossong defined three challenges for it, how to agree on what is ‘good’ or ‘best practice’,

(19)

how to mitigate political pressure and if it can obtain enough resources to do a proper job (Bossong, 2014).

Most research that mentions the RAN do so describing it as one of the EU’s counter-radicalisation initiatives. One article directly engages with the RAN, it focusses on the link between the War on Terror and the educational system. In a way similar to my research, it focusses on the RAN Collection of Approaches and Practices but on a different subject, namely ‘Educating young people’. They use critical discourse analysis to explore the link between the discourse of preventing radicalisation and education practices within the educational system in the various approaches and practices (Mattsson, Hammarén, & Odenbring, 2016). This article only analysed the introductory chapter of the Approaches and Practices document and only the few components that explain the practices that are concerned. These texts in the document are just summaries of what is done by the individual practice and to my opinion cannot fully represent what the practice entails. To analyse this inhibits the research and this is my main objection with it.

The habitus of the EU seems to be act after some major terrorist event has happened. The major events can be seen as catalysts for further cooperation and integration. However, it is also the habitus that when it comes to security issues, sovereignty is still very important, and cooperation has to take other “low” forms.

(20)

4. Theoretical framework

This theoretical framework serves to provide the state of the art on radicalisation and deradicalisation research. This state of the art is needed to give an overview of what recent scientific knowledge on deradicalisation tells us. Out of this overview, several indicators are taken to analyse the best practices the RAN EXIT working group has selected. Based on these indicators we will know what choice the working group has made out of the knowledge that is available. In order to properly understand deradicalisation, a small overview of radicalisation knowledge is first given. After that the overview of knowledge on deradicalisation is given.

4.1 Analysis of counter-terrorism and counter-radicalisation practices

Counter- and deradicalisation projects are very different for each country and indeed each city (Feddes, 2015, p. 14). As such, I looked at several meta-level researches to provide a general overview of what projects have been run in recent years or are still running. I examined three articles that together cover a wide range of extremist groups and studies thereof between the 1970’s and now.

Butt and Tuck analysed several counter- and deradicalisation practices from Sweden, Germany, Denmark and the Netherlands (2014). The article is from 2014 and the practices ran before that so it is some time ago but still the insights are valuable. They identify, among other things that both the national and the local level need to be concerned when it comes to these issues. A meaningful relationship between practitioners and participants is also important, which of course requires time to invest in. Lastly, they emphasize the importance of programmes tailored to each extremist individually. The research has a slight bias towards deradicalising right-wing extremists, with five of the programmes focussing exclusively on that group. Also, some of the projects ran around the end of 20th_{century through to 2010 and a lot more programmes have started since with} the increased focus on radicalisation in recent years, which could make the findings already outdated. Still the findings are based on a variety of practices and are also applicable today.

(21)

Feddes and Gallucci reviewed 55 manuscripts (assessments and/or evaluations of programmes) about counter- and deradicalisation from 1997 through to 2014 (2015). Most of the manuscripts were not peer-reviewed and most were anecdotal. Besides this the instruments and methods were, more often than not, unspecified. The way they analysed these manuscripts is very interesting because of their coding of target group, intervention, methods and evaluation. For the focus of intervention, they coded subjects as non-radical, potentially radicalised and radicalised. I will use the same coding scheme but differentiate between violent extremist and non-violent radical. The difference of which I will show below. The type of intervention is not specified and indeed I analysed this for each practice separately. The goal of the intervention was coded either preventative, suppressive or restorative and either long-term or short-term. As will be explained later on, suppressive interventions can work in different ways as well as restorative ones. Long-term interventions are deemed better than short-Long-term. Because of this I used this coding scheme as well. Lastly, they coded the practices for anecdotal, empirical or theoretical data, which I did not use. I am interested in how deradicalisation is understood and what measures are taken against it for each practice, in what way this is measured can be useful but is beyond my research.

Demant et al analysed several extremist groups in the Netherlands during the 1970’s through to the 1990’s (Demant, Slootman, Buijs, & Tillie, 2008). This is not really recent and only focused on one country but still relevant to discuss because the authors use the knowledge from that Third Wave period to apply to current issues. They found that for each of the four extremist groups they analysed factors are present on the supply, demand and context side of deradicalisation. For each group these factors work in different ways and some had a bigger impact than others. For my research I am interested in what factors are identified. I discuss them later on in this chapter.

4.2.1 Radicalisation

In order to understand deradicalisation it is first useful to have an understanding of radicalisation. A second reason is that a lot of the best practices concerned focus on both deradicalisation and preventing radicalisation. In order for me to also take this into account, and overview of knowledge on radicalisation is useful. Although some might argue the two are very different from each other it is not true that the two are completely unrelated. It is also useful because certain triggers for

(22)

disengagement work better at an earlier stage in the process and it is generally useful to have an understanding of radicalisation to understand deradicalisation.

4.2.2 Defining radicalisation

First to get an understanding of the concept of radicalisation. As a concept it is defined in varies ways by both academics and governments. To give a few examples. Dalgaard-Nielsen defines radicalisation as “a growing readiness to pursue and support far-reaching changes in society that conflict with, or pose a direct threat to, the existing order” (2010, p. 798). Here we see it is understood as a process with the goal of pursuing changes to an existing order. Demant et al. define it as “process of delegitimization, as a process wherein trust in the system goes down and one withdraws within one’s own group because the person does not feel part of society anymore” (Demant, Slootman, Buijs, & Tillie, 2008, p. 4). Again we see the aspect of it being a process and being in relation to a society. An equally important part is the aspect of the in-group. Rabasa et al. define it as “the process of adopting an extremist belief system, including the willingness to use, support, or facilitate violence, as a method to effect societal change” (Rabasa, Pettyjohn, Ghez, & Boucek, 2010, p. 1). This definition puts more focus on the willingness and eventual use of violence. The EU for example, uses a more vague definition, namely that it is “understood as a complex phenomenon embracing radical ideology that could lead to the commitment of terrorist acts” (European Commission. Migration and Home Affairs, 2018). They also see it as a process and bring in the, undefined, concept of terrorism, which also has a dimension of violence to it. The problems with defining terrorism are not relevant at this moment. Given the problems with defining radicalism and the abundance of definitions Schmid though it wise to reconceptualise it as follows (2013, p. 18):

an individual or collective (group) process whereby, usually in a situation of political polarisation, normal practices of dialogue, compromise and tolerance between political actors and groups with diverging interests are abandoned by one or both sides in a conflict dyad in favour of a growing commitment to engage in confrontational tactics of conflict-waging. These can include either (i) the use of (non-violent) pressure and coercion, (ii) various forms of political violence other than terrorism or (iii) acts of violent extremism in the form of

(23)

terrorism and war crimes. The process is, on the side of rebel factions, generally accompanied by an ideological socialization away from mainstream or status quo-oriented positions towards more radical or extremist positions involving a dichotomous world view and the acceptance of an alternative focal point of political mobilization outside the dominant political order as the existing system is no longer recognized as appropriate or legitimate.

Although it is long, it encompasses all the aspects of radicalisation that are important. The fact that it is a process, the abandonment of normal politics, the aspect of ideology, the reference to the mainstream, the importance of a group and the black and white worldview. Moreover, I coded for these factors in my analysis. All of these I will now discuss more deeply.

4.2.3 The process

Most academics and policymakers agree that radicalisation is understood as a process, as the definitions above also show (Schmid, 2013). While some include the aspect of violence, others only go as far as far-reaching changes to society, not specified by which means. This brings to the fore the issue of non-violent radicalism, violent radicalism and extremism. While some do not use the word ‘extremist’ or use it interchangeably with ‘radical’, both Schmid and Demant et al. make a clear distinction between the two. Demant et al define extremism as using violent action to act on the intention to thoroughly change the system (2008, p. 5). Schmid adds that for this to happen “individual liberties are to be curtailed in the name of collective goals, including the mass murder of those who […] disagree with that program. Restrictions on individual freedom in the interests of the collectivity and the willingness to kill massively are central to this definition” (2013, p. 9). Besides the obvious addition of dismissal of civil liberties and high level of violence this also adds to the question of deradicalisation. As Schmid puts it “radical militants can be brought back into the mainstream, extremist militants, less so” (Ibid, p. 10). Using more or less violence can be scaled using the conflict dyad (Schmid, 2013, p. 14). Both the extremist group and the state can move up and down this scale. At the first step in the dyad is persuasion politics, in which politics is done through constitutional means. The second step is pressure politics, where the state uses oppression and the non-state actor uses extra-parliamentary action such as protests and strikes. The third step

(24)

war crimes are most commonly seen in this stage. For the remainder of this chapter I will use the term extremists to refer to people involved in violent radicalism. The difference just described between radicals and extremists also explains why I code either non-violent radicals and violent extremists during the analysis.

These issues revolve around the question of where a person is in the process of radicalisation. The logic here is that, the farther a person is radicalised, the more willing he or she is to use violence. Kruglanski et al. conceptualize this as a pyramid (2014, p. 416). At the bottom of the pyramid is every person who is sympathizes with the movement. Going further up the pyramid the group gets smaller and the degree of radicalisation increases. What’s more important is how and why a person moves up this pyramid. Dalgaard-Nielsen identifies three levels to understand this, micro-, meso- and macro-level. Among individual (micro) level mechanisms are personal victimization, political grievance and ways of getting involved in a radical group (McCauley & Moskalenko, 2008). As a group, the radicals can also become more extreme, either by wanting to outdo each other or as competition for the same base of support a group can become more. At the macro level, a group (the masses) can also radicalise. Think of nationalist reactions to 9/11 or even outright hate due to a long-lasting conflict. In essence this is just a different way of labelling someone non-radical up to violent extremist. What this adds is an understanding of how large the group is and at what level we can analyse the way a person or group radicalises. I did not code this in a separate way but referred to these factors when the need arose during the analysis.

4.2.4 Group, ideology and relation to mainstream

One of the aspects of the definition by that of the group. As McCauley and Moskalenko put it “radicalisation and terrorism are made possible by bringing people into small groups” (McCauley & Moskalenko, 2008, p. 417). This notion of importance of groups is widespread (Schmid, 2013, p. 20). Why the group aspect is important for my research is further explained in chapter 4.3.2 and 4.3.3 factors for and barriers to disengagement. Secondly, the notion of a worldview or ideology is important. On the one hand it does not matter what kind of ideology is used, since it is used as approval for violent action (idem). On the other hand it does because, as Demant et al. show, groups with different ideologies walk different paths and may be susceptible to different factors for disengagement (Demant, Slootman, Buijs, & Tillie, 2008). I identify four ideologies and code for

(25)

five, left-wing, right-wing, nationalist/separatist, religious (Islamist) and religious (other). I make the distinction between Islamist ideology in groups and other religious ideologies because of potential bias in EU policy and also the practices I analyse. By not making this distinction in the code it will be lost during analysis.

Thirdly, what comes back a number of times and is often taken for granted is the relationship of what is deemed radical with what is deemed mainstream. Because what is deemed radical is always relational to what is deemed normal it is important to (broadly) conceptualise what can be considered normal. Schmid and Steiner and Önnerfors all argue that for Western societies, Western values like democracy, rule of law, pluralism, secularism, and freedom of thought and expression, among other things, are what can be considered mainstream. Although the relation to mainstream society is important it is hard to measure this in this thesis. what mainstream society is or should be is an enormous question on its own. Furthermore, it is hard to code it for such a small amount of text in each best practice description. Also, it is already measured indirectly with the type of intervention, goal of the intervention and barriers to disengagement, which will all be explained further on.

4.3.1 Deradicalisation

The first question we need to answer is to what extent deradicalisation is the same as radicalisation. If we take a scale from normal to extremist, becoming less radical is indeed the opposite of becoming more radical. Indeed, Kruglanski et al. paint this exact picture. They argue that deradicalisation is about a restoration of alternative goals (job, friends, family, etc) and values and a decrease of the ideological goal, or at least an abandonment of violence in order to achieve this goal (2014, p. 84). Demant et al. share this opinion, conceptualizing deradicalisation as “the process of becoming less radical […] both in behaviour and worldviews” (2008, p. 5). Both see being radicalized as being on a scale and thus agree that deradicalising is moving the other way on the scale. Others, like Horgan and Schmid, have a different view. Horgan argues that “the factors that determine how someone becomes involved are rarely related to the kinds of factors that determine how […] that person will engage in specific terrorist operations [and] answering questions about what keeps people involved with a terrorist movement may have surprisingly little, if any, bearing

(26)

argues that “the path to de-radicalisation is not necessarily the reverse of the path to radicalisation” (2013, p. 40). This focus on what factors lead to deradicalisation is closer to reality. On a scale of normal to extremist, moving back to normal constitutes deradicalisation but focussing on what makes it that people start to disengage and deradicalise is more interesting. This does not mean that knowledge about radicalisation is worthless when dealing with deradicalisation. Knowing how someone or some group came to be can be very useful in understanding and dealing with their issues. Also, since the practices I will analyse might include dealings with radicalisation as well as deradicalisation it is useful to have knowledge about both.

Just like radicalisation, there is little consensus on how to define deradicalisation. Two good examples are the definitions of Feddes and Rabasa et al. Feddes argues that deradicalisation “primarily involves the cessation of violent actions. With regard to beliefs, this involves an increase in confidence in the systems, a desire to once more be a part of society, and the rejection of non-democratic beliefs” (2015, p. 2). Rabasa et al. define it as “the process of abandoning an extremist worldview and concluding that it is not acceptable to use violence to effect social change” (2010, p. 1). Feddes’ main point is the cessation of violent action, where Rabasa et al. place more emphasis on changing of belief. This difference amounts to what Horgan calls disengagement and deradicalisation (2009). Disengagement amounts to stopping with violence but still having extremist worldviews. Deradicalisation is when a person or group actually takes a different worldview and renounces its extremist thought. In this definitions disengagement is a prerequisite for deradicalisation. In Horgans view however, stopping violence due to not being able to (e.g. being kicked from the group or being captured by security services) also amounts to disengagement. Most agree with this definition but I find it difficult to sustain. When using this definition, the only thing stopping the radical from using violence is the in-group or not being able to. The will to use violence and seeing violence as one of the ways to further the political goals can still be present. Making the cognitive transition to wanting to further the person’s political goals through means other than violence would then constitute true disengagement. Since most radicals are part of group, this would constitute leaving the violent group. The difference with deradicalisation is that for deradicalisation to occur a person would also have to abandon his or her extremist worldviews. This raises the question what policy governments should pursue, disengagement or deradicalisation. As Horgan argues there is no reason to assume that

(27)

disengagement leads to deradicalisation (2009). Disengagement would be a more attainable goal. Conversely, deradicalisation would mean a more durable change and has less chance to backfire, but it is very difficult to significantly alter someone’s worldviews (Rabasa, Pettyjohn, Ghez, & Boucek, 2010, p. 8). Through various ways disengagement could constitute deradicalisation however, due to rationalization or routinization for example (ibid, p. 10). It is better to view disengagement and deradicalisation as different phases in a deradicalisation process, with actual disengagement being a prerequisite for deradicalisation. I still code for either disengagement or deradicalisation as the stated goal of the practice. Given the debate on what is more preferable as a policy goal it is important to know which of the two is pursued more often.

4.3.2 Factors for disengagement

To that extent it is better to first look at factors that are known to lead to disengagement. These can be divided between push factors and pull factors as well as exit barriers. Different authors identify different factors that can lead to disengagement and deradicalisation, with significant overlap between them.

Horgan differentiates between psychological and physical disengagement. Factors leading to psychological disengagement are, among others, negative sentiments towards membership, a change in priorities and sense of disillusionment with the areas that are pursued by the group (Horgan, 2009). Physical disengagement can involve apprehension by the security services, movement into another role (forced or otherwise), being removed by the movement or, again, a change in priorities (idem). Note that, as I said earlier, several authors do not consider changing roles (forcefully or otherwise) actual disengagement.

Dalgaard-Nielsen divides the factors into three clusters. The first cluster revolves around losing faith in the militant ideology (Dalgaard-Nielsen, 2017, p. 277). There can be a realization that the world is not quite as black and white as they perceive it to be. Another factor can be the confrontation with the realities of armed conflict which trigger doubt in a person. A third possibility is encountering a person of the out-group who acts counter to the extremist’s worldviews. What is problematic is that all of these factors can just as easily contribute to the validation of the person’s extremist worldview (ibid, p. 279). It can also be one specific event or several events over a period

(28)

and leadership failure. Peers or leaders might care more about money or power than about their cause or their comrades, or they can be outright incompetent (ibid, p. 282). People can join the group who are seen as not fit for the cause. The third cluster consists of factors of changing personal priorities and costs of staying involved in extremism. Among these factors are the ageing of the extremist which triggers thinking about what to do with the rest of his or her life. Another factor is the relation to a loved one, a wife or child, and the consequences continued membership of the extremist group has on them. Dalgaard-Nielsen also lists costs for continued membership but I will discuss these below in the section about barriers against disengagement.

With all of these factors there is the question of why some seem more susceptible than others. Since this is hard to test, Dalgaard-Nielsen names a few possibilities. Some people might be more intellectually open than others. The composition and indeed the ideology of the group can be important as well. Society as a whole is important because of its acceptance (or lack thereof) of the former extremist back into its fold.

Demant et al. use slightly different categories but there is considerable overlap. Note that their factors revolve around collective disengagement and so some factors might only explain disengagement for a group as a whole. Most factors, however, can work for both collectives and individuals. The factors they use are clustered in ‘crisis of supply’, ‘crisis of demand’ and ‘crisis of context’ (Demant, Slootman, Buijs, & Tillie, 2008). On the supply side they name three factors (ibid, p. 14). The ideology of the group might not be that appealing to them anymore. Second, the organisational capacity of the group might start to fall short, either in organising social functions for group members or basic needs such as money or weapons. Thirdly, the leadership might fail due to various reasons like incompetence, greed or failure to handle a certain situation. On the demand side the authors identify two factors (ibid, p. 15). Members can grow out of the extremist group, demanding things the group cannot provide. The second factor is failure to attract new members to the group. Lastly, the matter of context (ibid, p. 19). The conflict the group derives its identity and worldview from can change or even disappear entirely. The government enact repressive measures, try to win over the less radical members of the group or try to reform, thus giving in to the extremist group. The support for the group can fail or the rival movements can attract more attention and recruits, for example Al Qaeda and Islamic State.

(29)

4.3.3 Barriers to disengagement

When wanting to leave an extremist group, certain barriers may be encountered. By nature, extremist groups are hard to enter and exit and knowledge of what barriers are encountered when wanting to leave one is difficult to obtain. Instead knowledge from cults and gangs is applied to this question, arguing that they share certain characteristics with extremist groups that make them alike (Demant, Slootman, Buijs, & Tillie, 2008, p. 19). Cults are a very tight group. Members of cults only interact with other cult members, contact with family or other friends is frequently cut off entirely. This creates social dependence on the cult. It also creates psychological dependence because the identity of the group member is completely based on the cult. Cults also require a lot of investment, in money, time and energy (ibid, p. 20). The fear of being independent of this or ‘losing’ the investment made into the cult are big barriers against exiting. For gangs operate differently. Gangs revolve around a culture of violence and protection (ibid, p. 21). The gang provides protection from both other gangs and the government, as well as a social circle for the member to be a part of. Conversely, leaving the gang puts the member in a situation where other gangs, or the previous gang can inflict violence upon the former member. Thirdly, there is a stigma on gang members, as well as cult members, barring them from entering back into society as a normal member. It follows logically that, the longer a person is involved in extremism, the bigger its commitment and investments will be, the bigger the barriers to exiting will be.

4.3.4 Phases of disengagement

Rabasa et al., Feddes and Demant et al. all provide approximately the same overview of phases of a disengagement and deradicalisation process. (Demant, Slootman, Buijs, & Tillie, 2008) (Feddes, 2015) (Rabasa, Pettyjohn, Ghez, & Boucek, 2010). The first phase is one where doubt comes to the extremist. This can be triggered by any one or a combination of the factors listed above. Also, this can be a sudden trigger or factors that have been working over a period of time. This is followed by some kind of decision making. One way for this is weighing pros and cons but there are of course varying ways humans make decisions. Assuming that the extremist than chooses to disengage, and does not radicalise further, he or she goes on to start disengaging from the extremist group. This means he or she starts developing contacts outside of the extremist group and wider

(30)

overtly or covertly. What follows is what Demant et al. call cognitive transition, where the former, extremist, identity is slowly dismantled in favour of a new one. The final phase is cognitive reorganization, where the former extremist is, to a certain extent, reintegrated back into ‘normal’ society. There is no clear-cut separation of one phase from another, instead it should be seen as a process starting from doubt up to reintegration into society. However, these phases help to identify at what point a person is in the process of disengaging. Note also that these phases mainly concern disengagement and not deradicalisation. Looking back at the difference between the two, a person can be reintegrated back into society without having left its radical views behind. It just means that the means to obtain his or her political goals are sought using ‘regular’ (read: democratic) means.

(31)

5. Research design

In this chapter I will show how I designed my research, how I selected my data and what the basis is for the codes that are used. The literature on the practice approach is not very precise when it comes to methodology. When discussing access to a certain practice, Pouliot acknowledges that it can be quite difficult at times. Participant observation is seen as the best and most direct way of access (Pouliot, Methodology, 2013). However, for apparent reasons this cannot always be done. Textual analysis is seen as a valid alternative to both observation and interviews, but it has its limits. When using textual sources, Pouliot suggests taking sources that can give insight into the practice that is researched (ibid, p. 49). Discourse can be used in the Foucauldian sense to analyse practice. Again, when analysing practice, a big part is getting to know the tacit knowledge that an actor needs to have in order to have a certain disposition and to produce a certain type of knowledge. Chapter 4 provided the overview of knowledge on radicalisation and deradicalisation. It is important to analyse what aspects of this general knowledge are chosen by the RAN. In other words, what their choice was in their knowledge production. The analysis of the Collection of Approaches and Practices document (the best practices) will provide for this.

5.1 Data sources

For my analysis I used 18 practices listed in the Radicalisation Awareness Network (RAN) list of Approaches and Practices, Exit Strategies (Radicalisation Awareness Network a, 2017). I use these data sources because they are one of the primary outputs of the RAN and by extent the EU. Of the best practices I analyse the content that is given in the Approaches and Practices document, there is a summary of each practice, along with country and organisation details (idem). Almost all practices link to a website for further information. Due to size constraints of the thesis and potential bias because some websites are not in English and others have very little content I did not analyse these sources. The added value of this is that all the practices are on equal footing, being analysed solely on the basis of what is provided in the Approaches and Practices document. The table below gives an overview of the practices and the home-countries.

(32)

# Name and organization: Country: 0. RAN Collection of Approaches and Practices. Exit Strategies

1. Work with people who commit violence or have an experience of violence. Association for Nonviolent Communication.

Slovenia

2. Success Together. Account Trust. United Kingdom 3. Fair Skills – youth cultural peer training. Cultures Interactive. Germany

4. Back on Track. Direktoratet for Kriminalforsorgon. Denmark 5. Disengagement and Critical Aftercare. Politiets

Efterretningstjeneste

Denmark

6. Radicalised individuals follow-up strategy. Entr’Autres. France 7. EXIT s.c.s. onlus. EXIT s.c.s. onlus. Italy

8. Exit Sweden. Exit Sweden. Sweden

9. Swedish method of working with formers in Exit work. Sweden 10. Aggredi programme. HelsinkiMissio. Finland 11. Exit work located within the social space. Jump. Germany 12. A guide to police empowerment conversations. National Police

Directorate.

Norway

13. Restoring Relation Project (RRP): Addressing Hate Crime through Restorative Justice. Race on the Agenda.

United Kingdom

14. Danish National Corps of Mentors and Parent Coaches. Danish Centre for Prevention of Extremism

Denmark

15. The Unity Initiative. The Unity Initiative United Kingdom. USA. Europe. Asia. 16. EXIT-Germany. Society Democratic Culture. Germany

17. Advice Centre Hesse– Religious Tolerance instead of Extremism. Violence Prevention Network

Germany

18. Taking Responsibility– Breaking away from Hate and Violence – Education of Responsibility (Verantwortungspädagogik ®). Violence Prevention Network

(33)

5.2 Content analysis

In order to research the practice of knowledge production by the RAN EXIT working group, the proposed method was content analysis. I belief Bryman understands this as qualitative content analysis (Bryman, 2012). One of the aspects of qualitative content analysis is that of inter-textuality, putting texts in context with each other (ibid, p. 555). For the analysis I made use of qualitative content analysis, specifically ethnographic content analysis. According to Bryman this approach to content analysis lets the researcher go back and forth between content, codification and analysis, providing a method that let me change the codification and add to it based on what I had read at several times during the analysis. For example, this lead me to include the concept of the train the trainer because I came across it on several occasions. Ethnographic content analysis should not be confused with ethnography. It is exclusively based on texts, where ethnography also looks at peoples and cultures (idem). Part of my analysis will also be quantitative content analysis. I will be counting themes that are based the previous chapters on radicalisation and deradicalisation. Since there is only one coder, myself, there are no problems with inter-coder reliability. However, since my bias can play a big role in the eventual analysis there is a potential problem with intra-coder reliability. I hope to take largely take this away by having thoroughly explained my concepts in the previous chapters and how I coded for them in the sub-chapter on coding below.

5.3 Coding

The table below shows all the assessments I will make and the indicators (codes) I use for them. Whenever there is no mention of an assessment a ‘X’ is coded. With this coding scheme I include the overall assessments that can be made for a best practice, according to the general knowledge overview given in chapter 4. In other words, based on what can be selected and produced as knowledge, what is actually selected and produced by the RAN EXIT working group.

For intervention focus I code for non-radical, potentially radicalised, non-violent radical or violent extremist. This is based on the code of Feddes and Gallucci (2015). I code for non-radical when there is no reference of violence, being at risk of radicalising, being part of an at-risk community and also when there is reference of a group standing opposite to the radical or extremist.

(34)

potential factors for radicalisation I coded for potentially radicalised, see chapter 4.2 for further explanation on radicalisation. I make the distinction between non-violent radicals and violent extremists because I belief there is an important difference the two. A person can be forcefully pushed out of violence by, for example, being incarcerated. However, in this situation the label violent extremist would still apply. When there is mention of radical thought but no intention of using violence (yet) I coded for non-violent radical.

In the assessment of the type of intervention I coded for material and immaterial. This gives an indication of the factors for disengagement that are met with as well as potential barriers. Material interventions are everything from help with schooling, a job, housing, mode of transport, etcetera. Immaterial intervention was coded whenever there was mention of training, dialogue or helping with acquiring skills.

The goal of the intervention is coded as preventative, suppressive or restorative. Only a restorative goal will include disengagement or deradicalisation. I also code for preventative and suppressive to know if these aspects of counter-terrorism are also present. The timeline of the goal is either short-term or long-term. Long-term is coded whenever an intervention mentions to be for 2 years or longer or ‘as long as it takes’. Everything else is coded as short-term.

I code for subject of the intervention because there is a real difference between personal and collective factors for disengagement. Individual is coded whenever there is mention of an ‘individual’ or ‘person’. Collective is coded whenever there is a clear mention of disengagement or deradicalisation of a group, in other words, more than one person at the same time using the same means.

I assess the ideology to analyse whether there is a bias towards any one or several ideologies of extremists. I code for left-wing when there is mention of ‘left-wing’, ‘socialist’, ‘anarchist’, ‘animal rights activist’ and so on. I code for right-wing when there is mention of ‘right-wing’, ‘neo-Nazi’, ‘white power’ and so on. I code for nationalist/separatist when there is mention of ‘nationalist’, ‘separatist’, ‘insurgent’, ‘civil war’, etcetera. I code for religious (Islamist) when there is mention of ‘Islamist’ or ‘jihadist’ or when there is mention of the war in Syria or Iraq. If this is the case, I also code for nationalist/separatist because based on that mention alone it is not clear whether there is a religious or nationalist/separatist ideology. Those persons that go to Iraq or Syria could also join a separatist group like the Kurds.

(35)

For the assessment of the aim I code for either disengagement or deradicalisation. Disengagement is coded when there is mention of stopping violence but no change of belief. Deradicalisation is coded when there is only mention of changing beliefs. Both are coded when both are mentioned.

For the assessment of factors of disengagement that are met by the practice I code for losing faith, group/leadership failure or changing personal beliefs. With being met by I mean the practice engages in intervention that are focussed on making the subject of the intervention either lose their beliefs, force group or leadership failure or lure them into changing their personal beliefs.

In the assessment of barriers against disengagement I code for doubt, decision making, out-group contact, withdrawal and cognitive transition/reorganisation, according to the part about phases in chapter 4.3.4. I merge the last two phases because they are quite similar in essence. The same as with the assessment of factors, what I measure here is the way in which a practice engages in ways to take away barriers for the subjects of the intervention.

Assessment: Indicators:

Intervention focus:

Non-radical Potentially Radicalised Non-violent Radical Violent Extremist Type of intervention: Material Immaterial Goal of intervention (goal):

Preventative Suppressive Restorative

Goal of intervention (timeline): Short-term Long-term Subject of intervention: Individual Collective

Ideology: Left-wing Right-wing Nationalist/Separatist Religious (Islamist) Religious (Other)

(36)

Aim of intervention:

Disengagement Deradicalisation

Factors for disengagement:

Losing faith Group/ Leadership Failure Changing Personal Priorities

Barriers: Psychological Social Mundane Threat of violence Society

Phase of

Disengagement:

Doubt Decision making Out-group Contact Withdrawal

Cognitive Transition/Reorganisation

5.4 Reflexivity

Lastly a note on reflexivity. It is important for the actor under scrutiny (the RAN) to be reflexive but also for myself. I acknowledge to come from a Western academic background. I have been schooled at the UvA, a place I consider very liberal and critical. The context of this thesis has been that of the European Security Politics project group, which had a critical outlook on politics from the start. This has also shaped the readings on the conceptual background. I have also done a course on the UvA, which shaped my current understanding of radicalisation and deradicalisation and provided some of the sources used in this thesis. One way in which this has shaped the thesis is the understanding of radicalisation and related subjects as a social problem rather than one of, say police matters. It has also shaped my understanding of the social world in general, to be one where discourse matters, and nothing is really true. In other words, the social world is shaped intersubjectively. I also acknowledge that within my understanding of knowledge and the production of knowledge as object of struggle, I believe several strands of knowledge cannot live side by side. Two realities, two knowings, cannot exist side by side, for if one is true, the other can’t possibly be.