Are Muslims in the Netherlands constructed as a ‘suspect community’? An analysis of Dutch political discourse on terrorism in 2004-2015

Ever since the terrorist attacks of 9/11, counter-terrorism legislation has been argued to be globally focused on a so called ‘suspect community’: the ‘Muslim community’. The media, politicians and scholars speak about a new wave of terrorism where ‘Islamic’ is a common key denominator. Critical research, so far predominantly focused on the United Kingdom, has pointed at unintended consequences arising from political discourse in which a ‘suspect community’ is constructed, for society as a whole and the ‘suspect community’ in particular. Building on this research, this study analyses if and how Muslims are also constructed as a ‘suspect community’ in Dutch political discourse on terrorism in the period 2004-2015. The analysis shows that political discourse in the Netherlands has shifted significantly in this period. Whereas until 2011, terrorism was framed as a problem that originates in society and that is to be solved for society as a whole, it is currently seen as a problem that originates in Islam and which needs to be addressed by the ‘Muslim community’. All members of that ‘Muslim community’ are now considered as potentially ‘suspect’ when they do not openly and explicitly adhere to Western values and take action to distance themselves from the ‘Jihadist enemy’. Further societal implications of this discourse, in which the ‘Muslim community’ is constructed as a ‘suspect community,’ are also discussed. Ever since 9/11, terrorism has arguably emerged as the most important security issue for Western states, engendering Ban impressive array of new anti-terrorism laws, agencies, doctrines, strategies, programs, initiatives and measures^ ([18]: 11). State authorities have increasingly adopted strategies which represent both ‘hard counter-terrorism’, aimed at the prevention of terrorist attacks, and ‘soft counter-terrorism’, aimed at Crime, Law and Social Change https://doi.org/10.1007/s10611-018-9802-y * M. J. (Masja) van Meeteren m.j.van.meeteren@law.leidenuniv.nl L. N. (Linda) van Oostendorp lindavanoostendorp@gmail.com 1 Department of Criminology, Leiden Law School, Leiden University, Steenschuur 25, 2311ES Leiden, The Netherlands preventing radicalization and violent extremism [26]. These counter-terrorism measures often target the threat coming from within European societies in the form of ‘homegrown terrorism’ and can potentially affect the public at large [26]. The increase in counter-terrorism legislation has been accompanied by scholarly attention warning about potential unintended consequences [27]. These new laws and policies have allegedly been underpinned by a political discourse aimed at legitimizing these practices. B[T]his political discourse has designated Muslims as the new ‘enemy within’ – justifying the introduction of counterterrorist legislation and facilitating the construction of Muslims as a ‘suspect community’^ ([23]: 646). This Bsuspect community^ [17], despite being a contested term [26], relates to the perception of the Muslim community as dangerous and thus as the legitimate focus of security measures [5]. Whereas the concept of a ‘suspect community’ was originally developed in the context of British anti-terror laws in Northern Ireland, the concept has recently been adopted by scholars to scrutinize the effects of modern-day counter-terrorism laws and policies [1, 22, 23]. According to Pantazis and Pemberton [23] the discursive construction of Muslims as a ‘suspect community’ is taking place ‘within the United Kingdom and beyond’. However, as some studies have indeed confirmed for this to be the case for the United Kingdom (e.g. [20]), it remains to be seen to what extent and how the results indeed hold true outside of the UK. In this light, the present study critically analyses political discourse on terrorism in the Netherlands, and seeks to answer the research questions: are Muslims in the Netherlands constructed as a ‘suspect community’ in Dutch political discourse on terrorism, and if so, how? The Netherlands provides a good reference point for comparison with the United Kingdom for this purpose because the two countries are situated in a similar geographical and socio-economic context. Furthermore, as in the UK, the terrorism threat in the Netherlands has been regarded as substantial for quite some time now. On the other hand, as Den Boer [11] notes: BThe Netherlands is one of the few countries in Western Europe that did not experience massive terrorist attacks and where counter-terrorism actions did not feature prominently on the political agenda.^ In addition, the Netherlands does not have a long tradition of domestic terrorism comparable to the Northern Ireland conflict [11]. Furthermore, the Netherlands has a historic reputation as being a ‘beacon of secular tolerance’ ([8, 9]: 27). According to Van der Valk [29], Dutch political discourse is relatively free of rhetoric and straight to the point. Debates are aimed at explaining goals and intentions and uncovering unwanted consequences of potential decisions, rather than creating polarization. Debates focus on achieving consensus, a typically Dutch feature, instead of creating potential differences in opinion between parties. All in all, if a ‘suspect community’ is constructed in Dutch political discourse on terrorism, it is likely to be constructed in the discourse of other Western European democracies as well. The unintended consequences of the discursive construction of a ‘suspect community’ Results from studies on the potential effects of political discourse that constructs Muslims as a ‘suspect community’ emphasize the urgency of our research aim. Similar van Meeteren M.J.(.M.)., van Oostendorp L.N.(.L.).

to studies focussing on researching the discourse itself, studies focusing on the impact of such political discourse Bare overwhelmingly focused on the UK^( [26]: 725). These studies point towards three types of unintended consequences.
The first type concerns a negative effect on police-community relations. The 'hard' approaches taken by counter-terrorism policy, which lean largely on law enforcement action, could be damaging due to a sense of grievance amongst Muslims, creating barriers to good police-community relations [23]. Disclosure of information from those communities most suited to contribute to discussions about religious strands used in terrorist propaganda, risks being hindered by suspicions surrounding these groups. 'Soft' approaches aimed at community engagement and preventing the recruiting by terrorist organizations might not succeed due to these suspicions and resulting resentment. This is especially important in light of community policing, which is viewed as an integral element in preventing radicalization and early intervention in the path towards extremism.
The second type of unintended consequence relates to this process of increased suspicion and resentment among Muslims towards authorities, and has to do with the risk that this situation will be used as a recruitment tool by terrorist organizations [17,23]. There are many factors driving the process which could eventually result in a radicalized individual with extremist violent intentions. However, the perception of injustices against Muslims -both by foreign policy and 'hard' approaches to domestic terrorism -seems to be an important factor [12,23]. It stands to reason that these grievances do not fully explain radicalization, but research has shown that the discursive creation of a 'suspect community' may actually fuel radicalization in some.
As a third and final point, political discourse which has a suspicious tone towards Muslim communities survives quite well outside of politics, as it intertwines with the general public's opinion. Poynting and Mason [25] label this phenomenon as 'permission to hate', in which the public feels legitimized to show suspicion and hate towards the 'suspect community' as they perceive the state authorities to have the same stance. Consequently, fear of this suspicious social group heightens, and therefore facilitates the deterioration of community relations in society, which seems to have occurred at quite an alarming rate in the UK [23].
To sum up, research on the situation in the UK shows a process which starts with a political discourse on terrorism that traces discursive constructions of a 'suspect community. This may legitimize 'hard' approaches over 'soft' approaches and therefore damage relations between law enforcement and communities. Further consequences are potentially increasing the risk of actually contributing to radicalization instead of combating it, and also fostering an increased polarization of society. All in all, this highlights the importance of studying political discourse on terrorism through the lens of discursive constructions of Muslims as a 'suspect community', which can have far-reaching effects.
The discursive construction of a 'suspect community' The term 'suspect community' has been coined by Hillyard [17] in connection to the Irish community in Britain, a community rendered suspect due to the anti-terrorist legislation and discourse of that time. Subsequently, the term has also been applied in research on Muslims in Britain [5]. Pantazis and Pemberton [23], drawing upon Hillyard, define a 'suspect community as follows: B… a sub-group of the population that is singled out for state attention as being 'problematic'. Specifically in terms of policing, individuals may be targeted, not necessarily as a result of suspected wrong doing, but simply because of their presumed membership to that sub-group. Race, ethnicity, religion, class, gender, language, accent, dress, political ideology or any combination of these factors may serve to delineate the sub-group.^(2009:649) The use of the term 'suspect community' in connection to the global war on terror has not gone without critique. Greer [14], for example, asserts that the term 'suspect community' is inappropriately applied to Muslims because they do not form a 'community' due to internal differences in religion, ethnicity and nationality. Indeed, the label 'community' has a tradition of being applied to 'communities' that are identified as such by researchers, but that are not perceived as communities by those who comprise them [31]. Ragazzi [26] likewise suggests it is probably better to speak of a 'suspect category' than a 'suspect community'. However, as Breen-Smyth [5] rightfully notes, Bmembers of the 'suspect community' have no essential bond with each other, since membership is defined within the imagination of non-members.^In other words, the 'suspect community' is socially constructedor 'imagined' in Benedict Andersons's termsin discourse by non-members of the community. Our understanding of a 'suspect community' is therefore a sociological one, as it is socially constructed, not by its members, but by its non-members.
Political discourse should be understood as the way a certain situation is talked and written about in the political field. Through 'language in action', group identities are constructed and processes of in-and/or exclusion take place [3]. Connecting language to the broader political context has proven to be valuable for researching processes of social and cultural change [24]. The construction of a 'suspect community' initiates in the political field. This expands beyond the political arena and becomes reproduced by both the general public and the media [16]. Therefore, this study examines the first step in this process and analyses political discourse. If signs of the discursive construction of a 'suspect community' can be found in the political arena, future analyses could focus on studying press coverage and other forms of public discourse.

Data, methods and analytical approach
The present study critically analyses political discourse on terrorism in the Netherlands to see if and how Muslims are constructed as a 'suspect community'. To select the relevant discourse on terrorism, policy documents ('kamerstukken') were first retrieved from a government website (rijksoverheid.nl), using the search term 'terrorism' and the time limits of January 1st, 2004 to July 31st, 2015. The search took place between July 31st, 2015 and the 1st of September 2015. The search started in 2004 because that was the year in which Dutch filmmaker Theo van Gogh was murdered by a radicalized Dutch-Moroccan man. Van Gogh had been critical of Islamic cultures and his violent death inspired debates about terrorism and the position of Muslim communities in Dutch society. Documents were scrutinized making use of the 'control plus find function' in Windows. Documents were initially selected when the word 'terrorism' was found more than once and when the document actually contained content. For example, agendas with topics for debate were not included. This search process yielded 202 documents, with documents ranging from one page to 174 pages.
To this corpus of policy documents, Chamber debates and discussions in the House of Representatives ('handelingen') were added. 1 In the House of Representatives, politicians speak their mind without the language of compromise that could make its way into policy documents, and as such constitute a valuable addition to the corpus of policy documents. The debates are transcribed by the government and posted online. The search for relevant Chamber debates was done using the keyword 'terrorism' at the government website www.
zoek.officielebekendmakingen.nl. This search yielded 606 documents. As this was a large number we narrowed the search by adding the keyword 'Jihad*,'yielding 114 publications.
The second step was to identify the discourse in which terrorism is actually discussed in relation to Muslims or Islam. As the word 'terrorism' functions as a primary term in political debate these days, similar to words as 'freedom', 'democracy' and 'justice' [18], terrorism is discussed in a multitude of different contexts; some of which were unnecessary for the purpose of this paper. The 202 documents and 114 Chamber debates and discussions were therefore scanned to determine if terrorism was mentioned in relation to Islam or Muslims in the Netherlands. When in doubt, they were read in their entirety. This way, it was possible to exclude many documents and debates from the total corpus.
The first reason for exclusion pertains to documents in which terrorism is not the focal point of the document or discussion, but rather, other aspects related to terrorism are. For example, terrorism is mentioned as part of situations that have Bto be prevented^, such as where government parties note what will be most important on the policy agenda for the next few years, they mention Bthe prevention of terrorism^, or Bthe threat of terrorism^. Furthermore, when terrorism is a cursory remark in documents, this is often accompanied by the words B(increasingly) international^, B(increasingly) global^, and Ba (increasing) threat^. Finally, documents were excluded under this criterion when terrorism is brought to the discussion to underline the urgency of preventing the problem that is discussed but is not the focus of the document or discussion.
The second reason for exclusion pertains to documents in which the mentioning of terrorism is often accompanied by Bcrime^, Borganized crime^, Bpiracy^or some other criminal behavior (see Baker-Beall [2] for similar observations on discourse in the EU). This is comparable to the first exclusion criteria in which terrorism is discussed, albeit in documents meeting this exclusion criterion 'the crime' of terrorism is often the focal point, amongst other crimes, most commonly Borganized crime^. Emphasizing the crime aspects of terrorism permits the government to combat it (and the radicalizing possibly leading up to it) with its full criminal justice apparatus. In such discussions, it is often mentioned how international police cooperation should be improved to combat terrorism.
The third category of documents that were excluded from the corpus concerned foreign military operations. The Netherlands takes part in quite a few operations on foreign soil, predominantly in countries in the Middle East or Africa, and some policy documents mention terrorism in relation to those operations. For example, some documents mention the mission in Mali, where Bterrorist attacks threaten our military^, and also Bpirates in Somalia^, which are also considered to be terrorists by political parties. Although these documents sometimes discuss Islam or Muslims, they are not relevant for answering the present research question, which is focused on Muslims in the Netherlands.
The fourth and final reason for exclusion pertains to documents in which specific terrorism laws are discussed and which are fairly doctrinal in nature. Terrorism, for example, appears in documents on fraud laws, insofar as they concern the Bfinancing of terrorism^. With regard to specific laws, it is also important to note the proposals for laws that deal with privacy invasions. Those (proposals of) laws often associate terrorism with other crimes to justify farreaching invasions. Infringements of privacy are justified, or considered justified, when someone is Bunder suspicion^of Bcommitting a terrorist act, helping a terrorist act come to fruition, or financing a terrorist organization^. Such legitimization pointing at the exceptional position terrorism and terrorist suspects have in society is present in documents both from earlier and more recent years.
Please see Table 1 for an overview of the four reasons for exclusion along with the criteria used. In the end, a total corpus was selected consisting of 93 policy documents (1718 pages) and 15 chamber debates and discussions (589 pages).
The analysis was performed by one of the authors which is why we cannot report intercoder reliability. The process of analysis started with open coding informed by preliminary sensitizing concepts [4]. These sensitizing concepts were drawn from the literature on suspect communities discussed in the first sections of this paper. Examples of the sensitizing concepts which initially guided the research were 'suspect', 'suspect community', 'Jihadism', 'societal groups' and 'community relations'. These functioned as a starting point for the discourse analysis by establishing a simple codebook [7]. In addition, in this phase of open coding additional codes were added to the codebook, such as 'polarization' and 'responsibility,' as the analysis unfolded.
The codebook that was used in the first phase of open coding has been further developed during the second phase of the analysis, which entailed more focused coding, providing specific content for concepts such as 'polarization', and 'responsibility' [7]. Writing memos, in combination with focused coding, helped to form abstract ideas, which allowed for the building and elucidation of categories, and which also clarified variations within and between categories of political discourse on terrorism [7,31]. Through the process of constant comparison ( [13]2006), we eventually built four core categories: problem, origin problem, polarization and responsibility problem. Furthermore, we identified two distinct periods in the time under scrutiny, both characterized by their own type of discourse. In each of these periods, each of the four categories is associated with a distinct discursive content. In other words, through analysis of political discourse on terrorism in the Netherlands along the lines of these four core categories, it becomes evident that a discursive change has taken place.
From 'communicative coexistence' to 'the Jihadist enemy' and the discursive construction of a 'suspect community' We have coined 'communicative coexistence' to refer to the discourse that was dominant from 2004 to 2011 and labelled the discourse after 2014 'the Jihadist enemy'. To clarify different discourse strands with their characteristics, a table (Table 2) has been added at the end of this section. It provides an overview of the two discourses and their most notable characteristics (the four core categories). As the shift from one discourse to the other is subtle, there is ambiguity in pinpointing the exact moment this shift happened, and the intermediate time is a transitional period which is difficult to interpret. Within these two distinguishable time periods, we pinpoint if and where discursive traces of constructions of Muslims as a 'suspect community' are noticeable. This is done by discussing policy programs, and evaluation, midterm reviews of those programs, and Chamber Debates and discussions.
Before we delve into the two time periods, a few general observations can be made with regard to discursive notions of terrorism. Within the time periods under scrutiny, terrorism is noted often as BJihadist^or Bextremist^terrorism. The context of the latter category makes it clear that this is only terrorism from Islamic groups, or groups thought to be Islamic. However, in earlier policy documents (before 2011) Bextremism^was most regularly followed by discussions of other types of terrorism. In later documents, this pattern has ended.
The word Bterrorism^appears in both time periods consistently in combination with Bradicalization^and Bextremism^. Where such references are made, the community at large, meaning the Bgeographical^and the Breligious community,^is implicated in helping the government in Bbeing vigilant upon first signs of radicalization^and Bdefendable against the lure of radicalized groups^. However, there is almost no specific reference to Muslim or Islam, which is offset by speaking of BJihadist (foreign) fighters and/or terrorists^, unquestionably concepts that are derived from (religious) beliefs within Islam. A specific reference to Muslim or Islam is only Terrorism is only mentioned as judicial pretext, noticeable when questions are asked about recent policy and/or news articles. The words BIslamterrorists^, BIslamic terrorism^, and BMuslimterrorists^are used sporadically in such events over both time periods. Policy documents dating before 2011, already focusing on the category of BJihadist terrorism and terrorist groups^, show nonetheless that a few Dutch parties (CDA, D66, PvdA, Groenlinks most notably) were aware of the risks that could arise from manifestations of extreme right ideology with regard to polarization and marginalization of Muslims. It was warned that extreme right comments against Muslims and Islam could Bfurther polarize^society. The warning Bto be careful^and Bto avoid further societal polarization^has been repeated throughout the documents, often in response to inflammatory statements.

Communicative coexistence (2004-2011)
In November 2004, a Dutch film and television director, Theo van Gogh, was shot and stabbed to death by a 26-year old radicalized Dutch-Moroccan citizen. Van Gogh had been critical of Islamic cultures and his murder caused public outrage. It also led to a debate in the House of Representatives in which the governmental parties showed awareness of the potential polarizing effects their statements could have and emphasized Bfighting together^against radicalization, and taking responsibility as a society. This is illustrated by quotes such as B(…) no 'us' versus 'them'^(Dittrich, D66), Bultimately this is an extensive responsibility for everyone.^(Halsema, Groenlinks) and Beveryone in this country has a responsibility.^(Marijnissen, SP).
In 2007, the 'Actionplan Polarization and Radicalization' (Actieplan Polarisatie en Radicalisering) (2007-2011) was introduced. In this period, phrases seem carefully worded, with notable attention to definitions. Both 'polarization' and 'radicalization' are meticulously defined, as evident in the document detailing the course of action for local governments [35]. This document states that the threat of radicalization Bis not primarily the ideas or ideology, but the expectation that the ideology will increasingly influence actions. This concerns actions with negative consequences for individuals and In addition, there is a focus on prevention and the first stages of radicalization, plus taking responsibility on a local level to foster integration processes and work against the forces of polarization. The plan outlines a careful 'fix it in society' approach to radicalization with attention to root causes in society. The focus of initiatives is on prevention: preventing the emergence of a foundation for radicalization, mostly through stopping discrimination and increasing knowledge about other cultures.
The aim is to avoid undue focus on terrorism and there is consistent attention for unintended consequences which would result from government interventions labelling a person as 'radical'. According to the action plan Bthe person can perceive the intervention as a confirmation of the negative stance of the government with respect to people who adhere to the ideas in question. This confirmation can strengthen their beliefs that a differentradicalpath is a necessary choice.^This shows that in this part of the time period (2007-2011), policy documents display a careful discursive framing of issues and high awareness of the possible consequences this governmental framing and interference could have.
The language used in the action plan is reflected in the debates that followed. The central idea is of a shared responsibility, and attention to possible unintended consequences resulting from society divided by an Bus^versus Bthem^mentality, illustrated by the quote: Bpolarization without nuance is only able to evoke misery.^(van Toorenburg, CDA). Polarization, as an unintended and undesirable consequence of counter-terrorism actions as well as the hazard of counterproductive consequences, is scrutinized: B(…) Precisely in the humiliation of Muslims, both internationally and in our own country, some find a motive to turn away from our society.^(Halsema, Groenlinks). Polarization of Dutch society is considered undesirable and hazardous, as the following quote from 2007 illustrates: BSetting aside one group of Dutch people on the basis of a single criterion, namely religion, is contrary to our values. (…) That road leads to polarization, to thinking in terms of us and them, to a sharpening of prejudice and ensconcing in one's own belief of what is right (…).^(Balkenende, CDA prime minister) It is perceived that action against terrorism is a responsibility of all members of society. In a 2004 debate, a member of Parliament (Sterk, CDA) suggested that 'moderate Muslims' speak out against terrorism more openly, as a consequence, she is fiercely attacked, as the quote below illustrates: BWith this sort of calls, mrs Sterk puts a stamp on all moderate Muslims which indicates that what goes on at the margins of society is their responsibility. (…) you suggest that they have a special responsibility.^(Dijsselbloem, PvdA) In a debate in response to the film 'Fitna' produced by Geert Wilders (PVV), the debate is again characterized by fear for labelling moderate Muslims as terrorists.
BOld images of violence and terror or maliciously associated with the so-called threat of a couple of hundred thousand Muslims in the Netherlands. As if they are all Jihadists. The comparison is invalid and particularly bad. (…) We should be alert, strengthen integration (…) work on a society that is based on respect^(Van Geel, CDA) As mentioned in the quote above, many politicians call for solutions aiming to take away the breeding ground for radicalization, and to avoid polarization in society. Society as a whole is responsible for the problem of terrorism. The previous quotes illustrate the sentiments prevailing during the period of 2004-2011. When the threat of terrorism motivated by Islamic belief is discussed, neutral terms are the norm, as can be seen in the analysis above. Approximately around 2009, one can see a slight change as the attention to Jihadism intensifies, but the overall tone remains respectful. In fact, it seems as though words are consciously chosen to ensure that Muslims are not constructed as a 'suspect community,' as politicians and policy makers seem aware of the negative consequences this might have.

Transitional period (2011-2014)
The time period between 2011 and 2014 is difficult to interpret and has been dubbed as transitional. What does stand out are references to the (ideological) forms that terrorism can take. In the earlier policy documents and chamber debates (before 2011), terrorism is referred to in a multitude of possible ways, for example terrorism related to animal activism. In later documents, this pattern has mostly ceased. Interestingly, when the focus is incidentally on some other form of terrorist action not linked to Islam, offenders are not classified as terrorists, but get labelled Bdangerous soloists^and Blone wolves^. Different 'types' of terrorism get labelled as Bextreme right^or Banimal rights activist^. After 2011, the notion of terrorism itself gradually becomes understood to mean Jihadist terrorism.

'The Jihadist enemy' (2014-present)
We have labelled the political discourse that has been dominant since 2014 'Jihadist enemy'. In August 2014, a new action plan was introduced that has been the norm for terrorist prevention since: the 'Actionprogram Integrated Approach to Jihadism' (Actieprogramma Integrale Aanpak Jihadisme). It has a distinct focus on Jihadism as the greatest threat that the Western World faces. As the title already suggests, this program is specifically aimed at Bpreventing terrorism and radicalization^related to what is often called BJihadist extremism.^It should be clear that this connotation of Jihadism is specific to Islam, which means that program measures are aimed at Bterrorism that claims an Islamic motive^.
Under 'Polarization and Radicalization' there were chapters for all ideological radicalization processes (Jihadist, extreme-right, animal rights), yet an analysis of 'Integrated approach to Jihadism' shows this attention to other ideologies has dwindled significantly. It appears the only focus for terrorism prevention is on the Bprevention of Jihadist terrorism^, and the Bradicalization^towards such a terrorist act Bstarting in Muslim communities^. Where other potential terrorist groups are mentioned, they are tucked away in footnotes or small subsections. Other ideologies from which terrorist actions can originate also do not appear to be discussed in the Chamber debates.
The period of 'Integrated Approach to Jihadism' and the policy documents published in this time, seem to have a much heavier focus on punishment of radicalized individuals, and create a distinction between them (the threat) and us (the innocent society). The threats are the Jihadist, the radicalized individuals and the surrounding population that does not contest them. That Jihadist population must be stopped at all costs, and whoever is not against them, risks suspicion of radicalism themselves. Here, there is much more of a suspect oriented approach, in which responsibility for this problem is placed with the Muslim communities. This starkly contrasts with the cautious 'fix it in society' approach that characterizes the previous period which was dominated by a discourse that we labelled 'communicative coexistence. ' After the attacks in France in 2015, the debates that followed clearly showed that the responsibility for terrorist attacks and radicalized individuals was placed both on the 'Muslim community' and Islam.
BThere is no doubt that there are moderate Muslims, but there is no moderate Islam. Of course not all Muslims are terrorists, but it is the case that about all of the terrorist today are Muslims. (…) You have to tackle the problem by its root, and that is the Islam.^(Wilders, PVV) The responsibility shift is facilitated both done by parties who are notorious for such statements such as the quote above, but also by 'moderate' parties, demonstrating the widespread nature of such ideas. The following quotes highlight the regularity with which such statements are uttered in political debate: B(…) the Jihad is a part of the Islam.^(van Klaveren, groep Bontes/van Klaveren), BIt is naive to claim that it has nothing to do with the Islam at all. (…) the Islam is the discussion point here.^(van Haersma Buma, CDA) and B(…) there is a problem within the Islam. This will have to be solved within the Islam itself.^(Zijlstra, VVD). These quotes originated from a debate prompted by a specific terrorist incident, namely the attacks in Paris in January 2015. Here, the attention is justifiably focused on terrorist acts claimed to be Islamic in nature.
Even under these conditions, there is still an awareness of the need to prevent polarization between Muslims and the rest of society by parties in the opposition: BMr. Wilders seems to suggest, that basically everyone is a suspect.^(Pechtold, D66), B(…) but then he is only helping terrorism along.^(Thieme, PvdD). They also issued warning of an increase in polarization, Bdividing people in such a way that the polarization in society becomes worse, that hate will be answered with hate.^(Roemer, SP).
Nevertheless, the dominant sentiment allocates responsibility to the 'Muslim community' for having radicalized individuals. Dijkhoff (VVD), for example, emphasizes the need to Bkeep them moderate, prevent radicalization,^where it is implicit that Bthem^refers to Muslims, and the suggestion seems to be that all moderate Muslims can potentially radicalize and thus warrant suspicion. Also prominent in this period, is the repeated calls on 'moderate' Muslims to publicly distance themselves from the Jihadist interpretation of Islam. Or, as mentioned in the debate after the attacks in Paris: BThe voice of the peaceful, on democracy and freedom focused Islam should prevail. (…) It is necessary that the Muslim community during this time of uncertainty lets the voice of democracy and freedom dominate.^(van Haersma Buma, CDA). In a Chamber debate about the approach to Jihadism, vice Prime Minister Ascher said: BThe challenge is to have all Muslims who make something of their life in the Netherlands, who work, who take their children to school (…) that they themselves also combat this phenomenon [terrorism].D espite the warnings from some oppositional parties of possible unintended consequences, the responsibility for terrorist attacks is placed on the Muslim community, resulting in a 'new form of polarization'. This new Bus^versus Bthem^discourse promotes an Bus^group with shared Western, democratic values, and a Bthem^that consists of the Jihadist enemy. The following quote is the most explicit example of this 'new polarization': BPlease note: there is indeed a case of an us and a them, but not of the stereotypical or polarizing us and them following ethnical or religious divisions (…). No, us, those are the people who cherish the freedom, the tolerance, the progression, the democracy (…). They, those are them that despise and oppose the cherished European values, sometime even with violence (…). Anyone who wants to foster unity, must do more than just speak unifying words. That person must also be willing to define the feeling of 'we' more meticulous and sharper. (...) Because only if we succeed in consolidating the unifying powers and create a community of all those people of good inclinations that cherish these values, our values, we make sure that extremist forces do not have a chance to make any headway.^(Samson, PvdA) As this quote from a major left-wing party at the time illustrates, this discourse in which responsibility for terrorism lies with Muslim communities is not exclusively the domain of right wing politicians. The present political discourse calls for Muslims to Bprovet hat they belong to the Bus^group, by explicitly distancing themselves from Jihadists: B(...) Why are there not more Muslims who take a stand with me? (…) we as a group (…) speak out that the radical Islam must be contested^(van Haersma Buma, CDA). Such questions by politicians directed at a specific group, highlights the discursive shift of responsibility for radicalized individuals from the whole society, solely to 'Muslim communities'. In the current discourse of 'us' versus 'them,' Muslim Communities are 'suspect' until they publicly reveal themselves as 'moderate' and as supportive of values such as 'freedom, tolerance, progression and democracy.' Overall, a discursive shift in terrorism discussions in Dutch political discourse can be observed, which is summarized in Tables 1. The focus of the action plan 'Polarization and Radicalization' on radicalization prevention in society, is substantively different compared to the action plan 'Integrated Approach Jihadism', where society is no longer responsible for radicalization. Responsibility is placed instead with the Muslim community, where their 'members' are pressured to stand up for freedom and against terrorism. There is no responsibility assigned to actions of society in general such as discrimination or polarization fostered by majority groups or political discourse.
The discursive shift that has taken place in Dutch political discourse is subtle yet important. No longer can the whole of society be considered responsible for the prevention of radicalization. Instead, Muslim communities are responsible for the radicalized elements in their midst, and all Muslims are possibly 'suspect' until they publicly denounce Jihadism. In other words, current dominant discourse on terrorism constructs the 'Muslim community' as a 'suspect community'. Until a member of this suspect community publicly distances him-or herself from Jihadism, and thereby asserts membership of the 'us' community instead of the 'them' community, he or she essentially remains a suspect based on this group membership.

Conclusion
The overarching aim of this research was to analyze if and how a 'suspect community' is constructed in political discourse in the Netherlands from 2004 to 2015. To achieve this objective, the contexts in which terrorism is mentioned in the Dutch political discourse were explored first. This exercise revealed that 'terrorism' today is indeed used as a primary term in political discourse, similar to words such as 'freedom', 'democracy' and 'justice' [18], which is illustrated by the versatility of contexts in which the word has been used.
Furthermore, this paper has demonstrated how an important discursive shift has taken place in Dutch political discourse on terrorism in relation to Islam and Muslims. Whereas discourse between 2004 and 2011 was labelled as 'communicative coexistence,' discourse in the period 2014-2015 was characterized as 'Jihadist enemy.' The discursive shift during this time is subtle, yet visible in both policy documents and Chamber debates. These depict a similar discursive shift in which responsibility for the problem of terrorism is passed on from society as a whole to the 'Muslim community.' These must stand up for freedom and against terrorism and until they do, they are 'suspect'. Muslims are required to explicitly state that 'their Islam' is different from the 'radical Islam', which implies that for as long as a Muslim does not state so, s/he remains a legitimate suspect. Therefore, the analysis presented here indicates that a discursive construction of Muslims as a 'suspect community' is indeed taking place in political discourse in the Netherlands.
In doing so, this study demonstrates that such processes are indeed taking place 'beyond' the United Kingdom, as Pantazis and Pemberton [23] suggest. This research therefore supports the assumption that the notion of the 'suspect community' is also created in other Western Democracies. In addition, the results contribute to the literature in showing how 'suspect communities' are created discursively, which is likely occur differently depending on the national context.
As outlined in the introduction, the construction of Muslims as a 'suspect community' can have devastating effects. Future research should investigate if and how such consequences, which were identified in the UK context, also occur in the Netherlands. Such research is beyond the scope of what we have investigated here. However, we may contribute a modest evaluation of the likelihood that these unintended consequences occur for Muslims in the Netherlands, based on observation of the Dutch context.
Previous research indicates that the likelihood of unintended consequences increases when political discourse on terrorism and counter-terrorism measures focuses on one particular societal group, which applies in this context. First, unintended consequences, such as deteriorating relations between police and communities, could presumably occur in the Netherlands. On a local level, this already seems to be the case [33]. Furthermore, Ragazzi [26] points out that, although the 'suspect community' thesis has proven its merits Bin highlighting the discriminatory aspects of counter-terrorism^, it misses out on an important aspect: the pro-active involvement of Muslims in their own policing. With recent deliberations on radicalization amongst imams in the Netherlands, and discussions on including more ethnic minority members in the police force, this proactive involvement seems an important topic of analysis for research [15].
Secondly, political discourse eventually penetrates the media, through television or through the on-and offline news. The present research has shown how Muslims are constructed as a 'suspect community', potentially affording a 'permission to hate'. What does seem clear is that there may indeed be a development of so-called 'collective hate'. This is evident most notably in the rhetoric of the leader of the Freedom Party (PVV), Geert Wilders, which distinguishes between the 'us' and 'them' group, thereby potentially promoting negative relations between the 'general' public and the Muslim minority [32]. The PVV was the second most popular party in the last national elections in 2017.
Thirdly, the creation of a suspect community in political discourse could facilitate (further) radicalization. The processes of radicalization are personal and different factors may be decisive in different cases, even though a general impact of discourse has been identified [12]. Research has shown that the process of radicalization can be intensified by perceived injustice. In fact, one of the most important determinants for moving onto a path towards radicalization and possible use of violence for getting one's ideas across, appears to be feelings of perceived injustice. No empirical data has been collected in the course of this study to determine if such injustice is perceived by the various Muslim communities in the Netherlands. The appeal for so-called 'moderate Muslims' to explicitly state that they do not support terrorism, could increase the feelings of alienation and isolation, and force people who identify as Muslims to evaluate their position in the Dutch society. De-radicalization is certainly possible and can be influenced by the narrative or discourse provided by the government [10].
Defining a target group ('suspect community') is a strategy to manage the threat of terrorism within a risk-oriented society, which the Netherlands has become [30]. Political discourse may identify a specific target group as the 'enemy within' [23]. Such a persistent focus on the Islamic background of terrorists provides the general public with the idea that Muslims are to be feared. This establishes them as a group 'other' than the majority [34].
Research into the concept of 'suspectification' as well as the potential unintended consequences is beneficial for society in general. The general public is already involved in the process, whether they know it or not. First, through political discourse public opinion is influenced by the government's portrayal of Muslims in the context of counter-terrorism [10,28]. Consequently, the general public plays an important role in the construction of suspicion [5,17] and should therefore be made aware of the likely unintended consequences. When perpetuating a crude narrative, these interconnections between policy, politics, and majority public opinion, could foster polarization in society and a culture of fear. Second, the attention to radicalization is profound within Dutch society [6,12,19]. This is evident in multiple attempts at involving the general public in a signaling function, to stop radicalization in its tracks as early as possible. Placement of the public in this position gives them a large influence on construction of suspicion.
The results of this study illustrate the creation of a 'suspect community' and the form this process takes in the Netherlands. Through comparison of the Dutch political discourse with research findings from the UK, it seems likely that unintended consequences will occur in the Netherlands. The unpalatable conclusion has to be that political discourse at this moment could generate outcomes, completely opposite to those intended by counter-terrorism policies.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.