Unifying or Divisive Threats? Anxiety about Political Terrorism and Extremism among the Swedish Public and Parliamentarians, 1986–2020 Patrik Öhberg, Daniel Brodén, Mats Fridlund, Victor Wåhlstrand Skärström and Magnus P. Ängsal University of Gothenburg, Universitetsplatsen 1, Gothenburg, 405 30, Sweden Abstract This paper contributes to the history of political terrorism through a data-driven study of the worry about the threat of terrorism and political extremism among the Swedish public and in the Parliament, 1986–2020. The aim is to explore the intersection of public opinion as expressed in national survey data and data on political action in the form of motions by Members of the Parliament (MPs), focusing on trends over time and the significance of political sympathies. Our study points to the impact of the attacks in the USA in 2001 and the wave of Islamist terror attacks from 2014 and onwards on both the public’s anxiety and the MP’s activity. It also shows the significance of political sympathies in this context with both citizens and MPs on the right being more worried about terrorism than those on the left, whereas the pattern is repeated in reverse when it comes to worry about political extremism. Through the investigation we highlight the benefits of combining parliamentary data and survey data as well as the importance of the parliamentary context, in exploring the relationship between public opinion and MPs activity on terrorism. An underlying argument is that the analysis of parliamentary data should be grounded in the context of the institutional and historical framework of the political system. Keywords 1 Terrorism; political extremism; parliamentary data; survey data; mixed-methods 1. Introduction Terrorism has had a decisive impact on Swedish society since the early 1970s, and the attacks in the US on 11 September 2001 accelerated its prominence as a critical concern. Today, terrorism is viewed by many Swedes as one of the gravest threats to society and the concern for political terror tends to take center stage in national politics [1]. In a study of fear in Swedish politics during the 20th century, historian Sverker Oredsson notes that fear of threats and disorder justify state power and various forms of state apparatus, including laws, judiciary, the military and police [2], and much international research has been devoted to the role of politics in the creation of public anxiety and fear of terrorism [see e.g. 3, 4, 5]. However, at the same time that fear of threats may serve as a unifying social factor, different people worry about different things and research has shown that the fear of terrorism is dependent on factors such as gender, ethnicity, class and ideology [6, 5, 7, cf. 8]. The study of the politics of terrorism in Sweden has largely focused on the legislation process of the counter-terrorism law [9, 10], rather than its broader contexts and dynamics. The present study, The 6th Digital Humanities in the Nordic and Baltic Countries Conference (DHNB 2022), March 15-18, 2022, Uppsala, Sweden EMAIL: patrik.ohberg@gu.se (Patrik Öhberg); daniel.broden@gu.se (Daniel Brodén); mats.fridlund@gu.se (Mats Fridlund); victor.whalstrand.skarstrom@gu.se (Victor Wåhlstrand Skärström); magnus.pettersson.angsal@gu.se (Magnus P. Ängsal) ORCID: 0000-0002-7433-8552 (Patrik Öhberg); 0000-0002-5914-1516 (Daniel Brodén); 0000-0002-5759-0027 (Mats Fridlund); 0000-0001- 6569-120X (Victor Wåhlstrand Skärström); 0000-0001-5996-5067 (Magnus P. Ängsal) © 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). CEUR Wor Pr ks hop oceedi ngs ht I tp: // ceur - SSN1613- ws .or 0073 g CEUR Workshop Proceedings (CEUR-WS.org) 145 however, turns the attention to the relationship between the public’s anxiety for terrorism and the parliamentary activity on the topic, drawing on the affordances of data-driven analysis [11; c.f. 12]]. The aim is to explore the intersection of public opinion as expressed in national survey data and political action in the form of motions by Members of the Parliament (MPs), focusing on trends over time and the significance of political sympathies. Historically as well as in present time, Swedish national politics and public life have been dominated by the left-right conflict dimension, which arguably is also the case with the terrorism discourse in Western countries [13]. To explore possible patterns of convergence and divergence, we address three complementary research questions: First, we ask, how has the concern for terrorism been reflected in public opinion and parliamentary action at different points in time? Second, we ask about the significance of political sympathies in this context and whether terrorism is a politically divisive or unifying issue. Third, we ask about the connection between concern about ‘terrorism’ and ‘political extremism’ among citizens and MPs. Since terrorism is often understood as a manifestation of political extremism, one might assume that those who are concerned about one should also be concerned about the other. However, our previous findings indicate that the two concepts may have different connotations in public opinion depending on political sympathies [1]. In the present study, we expand that discussion by also analyzing MPs’ worry about political extremism. As a starting point, the chapter discusses the two types of data used in the study: national survey data from the Swedish SOM Institute and data from the Swedish Parliament (Riksdagen) on motions by Swedish MPs. We comment on general aspects of these two types of data and argue, from a methodological standpoint, for the importance of a contextual approach to the analysis. An underlying argument is that the analysis of parliamentary data should be grounded in the context of the institutional and historical framework of the political system. We proceed by giving an overview of the Swedish public’s anxiety about terrorism and political extremism over time, discussing peaks and putting them into a contemporary historical context. Then, we turn to Parliamentary data on the MPs’ activity when it comes to putting forward motions concerning the topic of terrorism. Here, we highlight governmental position and party affiliation as key factors in the analysis. After this, we extend the analyses by including political extremism in relation to public opinion and parliamentary action, also drawing on MP survey data. In order to further understand the dual concern for terrorism and extremism, we conduct two case studies on MPs motion writings in 2010 and 2018, using distant and close reading to pursue the specifics behind the connection between the two and how public fear is used as a reference point in political rhetoric. We conclude by discussing the benefits of a context-sensitive approach to the analysis of Parliamentary data and how the study may be followed up with a refined language technology-driven analysis of the sentiments of the motions. 2. Combining survey data and parliamentary data As this study explores the intersection of public opinion and parliamentary action, we need to make some initial comments on the different types of data that we use and how to approach them in a way that embraces their quantitative qualities but still locates the analysis in their contextual dynamics. Researchers have recently started to explore Swedish parliamentary data using methods from statistics and machine learning, but the influence of political domain knowledge has been limited [c.f. 14]. By grounding the investigation of the parliamentary data in the historical and institutional framework of the political system, we may expand the perspective on the political processes and contexts behind the data. Thus, our interdisciplinary mixed-methods approach, combining political science, terrorism studies and digital history, allows for a multidimensional interpretation of the data in its contexts. By also conducting a brief close reading of parliamentary motions we provide an understanding of some specifics behind the connection between the concepts of terrorism and political extremism in political rhetoric. A point of departure for this study is that although public opinion data and parliamentary data are hardly equivalent materials, they are, nevertheless, to some extent comparable. We treat motions by MPs as indicators of political activity against the background of the principle of policy responsiveness 146 [15], and, thus, as a measurement of the extent to which the actions of the MPs reflect the public anxiety about terrorism and political extremism. 2.1. Questions in the SOM survey The SOM Institute is a Swedish university-based research organization and a national infrastructure for survey data, that since 1986 has conducted annual nationwide surveys on attitudes and habits related to society, opinion and media. The SOM surveys are based on representative samples of the Swedish population, randomly addressing selected persons living in Sweden between the ages 16–86. In the surveys there are recurrent questions asked each year, which makes it possible to track trends in the public opinion related to certain events. One recurring set of questions in the questionnaire asks, ‘If you look at the present situation, how worrying do you find the following for the future?’, (‘Om du ser till läget idag, hur oroande upplever du själv följande inför framtiden?’), with ‘terrorism’ and ‘political extremism’ being two of the several alternatives that is ranked on a four-grade scale from ‘not worried’ to ‘very worried’. Notably, neither questionnaire provides any definitions of ‘terrorism’ or ‘political extremism’, which are ambiguous and contested concepts. Terrorism is frequently used in conflicting ways by various actors, often to label the actions of their opponents as illegal or illegitimate. The familiar cliché that ‘one man's terrorist is another man’s freedom fighter’ is often quoted on good grounds [16]. Though different, the concept of political extremism essentially follows a similar pejorative logic, at times through the distinction between extremists and reformists [17, 18, 19]. Thus, while SOM provides data of the Swedish public’s anxiety about terrorism and political extremism at different points in time, this data does not, however, say anything about what forms of political violence these concerns are associated with. Consequently, our study becomes dependent on our ability to look at the issues at hand from a larger perspective. The same circumstances apply to another set of data. Additionally, we draw on a unique survey, Riksdagsundersökningarna, the Riksdag MP Survey. The first Riksdag MP Survey was conducted in 1969 and since then Swedish MPs have at 10 occasions been asked about their attitudes and behaviors related to representative democracy. Thus, the survey also provides us with data on MPs’ anxiety about political extremism (terrorism has not been a topic of interest in the survey, though). 2.2. Parliamentary action and motions The Riksdagen’s Open Data (http://data.riksdagen.se) maintains a range of Swedish Parliamentary data, including motions. In the Swedish Riksdag there are 349 legislatures from 8 political parties. The party discipline is, like in other party dominated systems, strong. MPs seldom deviate from the party line [20]. Even though there is a strong tendency to align with the party when MPs vote in the Parliament, there is still room for the MPs to express individuality [21]. The rules and the norms concerning motions are rather liberal. Each autumn, the MPs are free to submit motions to change legislation on, more or less, any subject. Most of the MPs do so and approximately between 3500–4000 motions are submitted every autumn. Given the free nature of motions, it offers ample opportunities for the MPs to demonstrate their responsiveness towards citizens’ concern. Notably, the field period for the survey and the motions both take place in the autumn each year (motions can also be written in response to government bills). In order to capture the intensity of the parliamentary activity regarding the topics of terrorism and political extremism, rather than counting the number of motions, we count MPs who signed a motion (if 3 members sign 1 motion, we count 3 ’hits’). Moreover, we focus on motions using the words ’terrorism’ and ‘political extremism’ and the latter’s compound forms, which we identify through semi- automatic and manual searches in the parliamentary motions (available at https://www.riksdagen.se/sv/ dokument-lagar/). Since both these terms are open for interpretation and political extremism in particular has many derivatives, we include ‘political extremism’ (‘politisk extremism’), ‘right-wing extremism’ (‘högerextremism’), ‘left-wing extremism’ (‘vänsterextremism’) and ‘violent extremism’ (‘våldsbejakande extremism’) in our search queries to cover its central meanings. 147 3. Public anxiety – highs and lows While the types of political violence that today is usually labelled ‘terrorism’ has a long history, it was during the early 1970s that terrorism became a domestic and publicly recognized phenomenon in Sweden, following a string of violent incidents, including the killing of the Yugoslavian ambassador (1971), the Bulltofta skyjacking (1972), the West German embassy siege (1975) and the ensuing plot to kidnap former Minister Anna-Greta Leijon [see 10]. Since the national SOM surveys was initiated in 1986, it is difficult to say anything about the public’s prior anxiety about terrorism, but it is notable that the anxiety for terrorism has never been higher than the first year of the SOM surveys (65 percent). During the 1970s and 1980s, terrorism was a significant phenomenon in Europe and the fact that so many Swedes were worried about terrorism in 1986 (Fig. 1), may be due to the fact that several terrorist attacks sponsored by Libya and Syria were carried out in Europe that year. However, it is perhaps most likely that the degree of anxiety was tied to the assassination of Prime Minister Olof Palme in 1986. The first national alarm on the night of the murder concerned the militant Croatian separatist organization Ustaša that was closely associated with the prior killing of the Yugoslavian ambassador and the Bulltofta skyjacking. Initially the police also held suspicions against the Red Army Faction (Rote Armee Fraktion, RAF), which was responsible for the occupation of the West German embassy. The police then dedicated extensive resources and public prestige in investigating whether the Kurdish organization PKK (Partiya Karkerên Kurdistan, ‘Kurdistan Workers' Party’) carried out the murder in Sweden [22]. Several Kurds residing in Sweden, linked to the PKK by the security service, had been designated as terrorists under the Swedish counter- terrorism law. The fact that the next major rise in anxiety for terrorism is after the attacks in the United States on September 11 2001 is perhaps to be expected. The hijacking and deliberate crashing of four passenger planes, aroused anxiety and fear throughout the Western world. In Sweden, 60 percent of Swedes felt worried about terrorism in 2001. Nevertheless, it is worth considering the relatively low degree of fear of terrorism in the period between the Palme murder and the 9/11 attacks. For example, the so-called Laser Man spread fear among immigrants through a series of shootings in 1991–1992 and in 1999 two neo-Nazis shot and killed a member of the Syndicalist organization. The latter year, there were also several neo-Nazi bombings, including attacks against two journalists, as well as the Malexander murders, where bank robbers who planned a terrorism campaign inspired by the RAF and IRA killed two police officers [23]. It should be stressed that the question about the anxiety about terrorism was not included in the SOM surveys in neither 1992 nor 1999. However, since anxieties were otherwise comparatively low during the 1990s, it seems at least reasonable to assume that deadly and systematic right-wing extremist violence was not primarily associated with terrorism by the public [cf. 24]. Figure 1: The Swedish public’s worry about ‘terrorism’ according to SOM-survey data 1986–2020 (Percentage of replies that include ‘very’ (mycket) worrying). 148 In the mid-2010s, there was a dramatic rise in the fear of terrorism. The increase started in 2014 (with 16 percentage points to 46 percent) and after the truck attack on Drottninggatan 2017 it reached roughly the same level (60 percent) as the years of the Palme murder and 9/11. While it could be expected that the Drottninggatan attack (which killed five people, including one child) would be reflected in the data, it should be noted that a previous attack at Drottninggatan in 2010 did not leave a significant impact on the public anxiety for terrorism. Then just before Christmas, a bomber blew himself up, failing to carry out a major attack in the shopping district (two people were slightly injured). After that incident the fear of terrorism had even dropped (from 37 to 31 percent). Figure 2: The Swedish public’s worry about ‘political extremism’ according to SOM-survey data 2011– 2020 (Percentage of replies that include ‘very’ (mycket) worrying). As noted above, the SOM data about the Swedish public’s concern about political extremism is more limited, since this particular question has only been posed since 2011. Consequently, it is impossible to compare the anxiety about terrorism and the anxiety about political extremism during, for example, the 1990s which saw a rise in neo-Nazi violence. When comparing these results as presented in Figure 2 with the results in Figure 1 above, there is, however, a tendency that the public is more concerned about terrorism than political extremism. The trend lines do follow each other with the exception of 2016 and 2017. As previously mentioned, the public become more anxious of terrorism after the 2017 truck attack in Stockholm. However, this attack did not have a major impact on the publics’ concern for political extremism. This might further indicate that citizens make a distinction between political extremism and terrorism. 4. MP’s activity – the weight of institutional context and party affiliation We now turn to the parliamentary activity on the topic of terrorism. The results in Figure 3, which shows the number of MPs that have signed a motion containing the word ‘terrorism’ during the period in focus (1986–2020), to some extent align with the ones displayed in Figure 1. One should be careful when comparing the two figures, as they measure different trends and have different scales. Observe that the figure shows the number of motion movers, and not motions, thus highlighting MP engagement and party collaboration. Nevertheless, this conveys the concern about terrorism among the MPs. Similar to in Figure 1, there is in Figure 3 a sudden rise in the proportion of motions around 2001 and 2015 that deal with terrorism, illustrating the impact of the 9/11 attacks and the wave of Islamist terror attacks from 2014 and onwards, including the 2017 Drottninggatan truck attack. But the curves in both figures are far from identical, an obvious difference being that there is no equivalent to the high public anxiety 149 about terrorism in 1986 when it comes to parliamentary activity. While the difference is difficult to explain based on our data, we might assume that while the Palme killing had an impact on public anxiety about terrorism, as we hypothesize, it seems less likely that it would have an effect on the motions in the Parliament (the fact that the perpetrator was not identified by the police would force MPs to speculate about the unsolved murder). Overall, our results point toward the importance of the parliamentary context itself. The dramatic increase in motions that use the word terrorism from 2001 and onwards are, of course, related to 9/11 and the actions by the militant Salafist jihadist network al-Qaida. A brief review indicates that the 2003 invasion of Iraq also fed into the discussion. But the increase in the intensity of motions at this point was also dependent on legislative factors and, thus, at least partly institutionally-driven. In 2002, the member states of the European Union (EU) agreed on a framework decision on combating terrorism (2002/475/JHA), which was based on a debated definition of terrorism. The framework decision can, on some level, be understood as a reaction to public opinion, but not necessarily in Sweden. The EU’s controversial definition of terrorism and the presumed effects of the framework decision on Swedish legislation generated a fair amount of debate in the Swedish Parliament. So did the Act on Criminal Responsibility for Terrorist Offences (2003:148), which was put forward by the government in 2003 to implement the EU’s framework decision and subsequently adopted by the Parliament. Figure 3: Swedish politicians’ concern about terrorism. Parliamentary motions 1986-2020 mentioning ‘terrorism’ according to the party belongings of MPs signing the motions (motion movers). Straight line is the total yearly number of motions on all topics. What is perhaps most striking in Figure 3 is that the use of the word ‘terrorism’ is particularly prevalently used by MPs in motions submitted during two phases, 2001–2005 and 2015–2020, respectively. While this lines up with the terror-related events previously discussed, it distinctly points to the significance of governmental position as a factor when it comes to motion writing concerning the topic of terrorism. In the period 2000–2006, Sweden was governed by the Social Democratic Party (‘Socialdemokraterna’ S), and 2014–2020 by the Social Democratic Party in coalition with the Green Party (‘Miljöpartiet’ MP). The drop in motions following the first phase (2001–2005) largely coincides with the change of government, from Social Democrats to a center-right coalition between the Moderate Party (‘Moderaterna’ M), the Centre Party, the Liberals (‘Liberalerna’ L) and the Christian Democrats (‘Kristdemokraterna’ KD), the so-called ‘Alliance’ in the period 2006–2010. The second increase in the number of motions 2015–2020, coincides with the Social Democrats regaining government in coalition with the Green Party in 2014. From this we may likely conclude that MPs are distinctly more active in writing motions about terrorism when their party is not in government. Perhaps the most significant result is that the right parties have been driving the writing of terrorism- related motions. During the first phase (2001–2005), it was MPs from the Moderate Party, the Liberals and the Christian Party and during the second phase (2014–2020), it was MPs from the Moderate Party, 150 the Liberals and the right-wing populist party the Sweden Democrats (‘Sverigedemokraterna’ SD), which entered the Parliament in 2010. In fact, members of the Sweden Democrats are behind an unproportionally large number of motions that concern the issue of terrorism. During the middle period, 2006–2014, when there was a center-right coalition government and the Social Democrats and the other parties more to the left were in opposition, there was a clear drop in the number of motions being brought forward that connected to the topic of terrorism. Few were written by the center-right parties and while the parties to the left increased their activity, they did not reach similar levels. In Figure 4 we look at the parliamentary activity on the topic of political extremism in the form of MPs signing motions (motion movers) using any of the four common collocates and compound forms of political extremisms: politisk extremism, våldsbejakande extremism, högerextremism and vänsterextremism. This shows that the terms were not frequently used until the mid-2010s. Previous research has argued that the concept of political extremism, and in particular as it was manifested in its sibling concept of violent extremism (‘våldsbejakande extremism’), gained momentum in the political discourse after it started to appear in reports and communications by the government and the security service in the early 2010s. Human rights scholar Dan-Erik Andersson suggests that the concept violent extremism to some extent started to replace or merge with the concept of terrorism in the policy discussion [25]. We will return to this question below in a case study on the use of various forms of extremisms in motions, but Figure 4 appears to support these prior findings, showing a dramatic increase in the use of political extremisms in motions from 2015 and onwards. Figure 4: Swedish politicians’ concern about extremism. Parliamentary motions 1986-2020 mentioning various forms of political extremism according to party belongings of MPs signing the motions. Straight line is the total yearly number of motions on all topics. 5. Terrorism and political extremism – left vs right issues? Overall, the Swedish public is about as worried about terrorism as it is about political extremism. In 2020, 49 percent of the population felt anxiety about terrorism, and 47 percent felt the same about political extremism. But if one takes into account political domicile, an underlying pattern can be distinguished. Table 1 confirms that people with political sympathies to the right have greater concern for the threat of terrorism than people with political sympathies to the left. The concern is highest among people who sympathize with the Sweden Democrats, where two thirds (66 percent) are worried about terrorism, followed by Christian Democrats (59 percent) and Moderates (51 percent), closely followed by Social Democrats (49 percent). However, only one in four among Green Party sympathizers (26 percent) feels great anxiety, and about one in three among sympathizers to the Left Party (‘Vänsterpartiet’ V) and the Liberals (31 percent and 37 percent, respectively). 151 Table 1 Citizens concern about terrorism and political extremism 2020 according to party sympathies (%) SD CD M S C L LP GP Terrorism 66 59 51 49 42 37 31 26 Extremism 31 42 45 47 47 61 62 56 Difference 35 17 6 2 -5 -24 -31 -30 Looking at anxiety about political extremism as displayed in Table 1, we can conclude that the pattern is largely repeated, although in the reverse. The anxiety about extremism is the most widespread among sympathizers to the Left Party, the Liberals and the Green Party (62, 61 and 56 percent, respectively), while the Sweden Democrats sympathizers are half as worried (32 percent). If we measure the ‘difference’ in the concern for terrorism and political extremism, we may state that people on the right are more worried about terrorism, while those on the left are more concerned about political extremism. Thus, while terrorism and political extremism may seem like ‘branches on the same tree’, most Swedes seem to make a difference between the two depending on their political views, the exception being sympathizers with the Social Democrats and the Centre Party, who are about equally worried about both (whether it is because they see them as synonymous phenomena or not is, however, hard to ascertain). The results in Table 1 could be interpreted as pointing to the possibility that people on the left are more worried about the threat of right-wing extremist violence than terrorism. However, one must keep in mind that concerns about political extremism do not have to concern acts of terror and violence per se (it may also include, for example, the impact of extremism on the social climate). While it is difficult for us to evaluate this question further based on SOM data, we may, however, make a comparison with data from the Swedish Riksdag MP Survey, showing the degree of anxiety concerning political extremism among MPs. A comparison between, on the one hand, MP’s anxiety in 2010 and 2018, and the anxiety among the public in 2011 (‘political extremism’ was not included as an alternative in the set of questions in the SOM survey in 2010) and 2018, provides us with some information about overlaps or gaps. Figure 5: Citizens and MPs concern about political extremism in 2010/11 and 2018 as related to party sympathies (%). 152 In Figure 5, two things stand out. Firstly, MPs are in general more concerned about political extremism than the public (c.f. Table 1). This is especially true for MPs on the left and from the Centre Party. MPs were also twice as concerned about extremism in 2018 than in 2010. Secondly, our data indicate that anxiety for political extremism seems to have become politicized. In 2011, citizens’ placement on the left-right scale and their fear of extremism were not significant, but in 2018 the coefficient for left-right dimension on fear for extremism is significant (about the same level as terrorism). A similar pattern emerges among the MPs. In 2010, MPs’ left-right placement was a poor indicator of their fear of extremism, while it has some predictive power in 2018. A clear illustration of the surge in this polarization on political extremism is provided by the different paths of MPs from the Green Party and the Sweden Democrats. In 2018, MPs from the Green Party were very alarmed about political extremism (92 percent), while MPs from the Sweden Democrats were not alarmed at all (10 percent). Going back to 2010, almost one third of the Sweden Democrats perceived political extremism as a threat, while barely 50 percent of Greens felt anxious. Thus, the differences in anxiety between MPs from these two parties have increased from 15 percent to 82 percent. . 6. Case study – concerns for terrorism and extremism among MPs The following brief case study of the use of political extremism in motions during two calendar years illustrates how closely the concept is associated with terrorism by MPs. The two years chosen 2010 and 2018 are chosen as they are two years for which we have the best survey data regarding MP’s anxiety. Here, we perform simple word-queries to identify motions that use political extremism. The selected limited corpus used comprises Swedish parliamentary motions for 2010 and 2018 available at www.riksdagen.se/sv/dokument-lagar/, and the search used the pre- and post-truncation search query *extremism*, to identify all occurrences of collocation and compound forms containing extremism. The search run generated in total 164 tokens of extremism words in 53 motions for the two years. A discussion of this result, comparing motions from 2010 with those submitted 2018, is followed by a close reading of the use of the abstractum ’extremism’ in the relevant motions. This allows us to grasp some of the specifics behind the connections between extremism and terrorism as well as how the perceived public concern for terrorism may be used as a reference point in political rhetoric. 6.1. Motions in 2010 In 2010, only 3 motions were put forward that concerned the topic of political extremism; one by MPs from the Social Democrats, one by a MP of the Christian Democrats and one by MPs from the Sweden Democrats. Of the 10 MPs that signed these three motions, the majority (7) were from the Social Democrats. With regard to the significance of governmental position, the Social Democrats and Sweden Democrats were in opposition at the time while the Christian Democrats were part of the coalition government. In these motions political extremism is exclusively discussed in the context of foreign policy and international relations and as something existing outside but not yet inside Sweden. Furthermore, the word extremism only occurs in three ideologically neutral forms (extremism, politisk extremism and utländsk politisk extremism) and it is used in a general or vague sense to designate violent militancy and (non-violent) sentiments among the public and politicians. At the same time the concept is connected directly or indirectly to the topic of terrorism in two of the motions. In the one submitted by the Social Democrats, extremism appears first as politisk extremism but later also in the form of the personal noun extremist and then being closely linked to terrorism by means of the head ‘terror deed’ (terrordåd) in the noun phrase ‘extremists’ terror deeds’ (‘extremisters terrordåd’) in the sentence: ‘Extremists’ terrorist attacks and the disappearances of suspected terrorists in prisons are also examples of abuses, as is the lack of guarantees of fair trials’ (‘Extremisters terrordåd och misstänkta terroristers försvinnanden i fängelser är också exempel på övergrepp, liksom bristen på garantier för rättvisa rättegångar’, 2010/11:U306, our translation). The second terrorism-related motion, by the Sweden Democrats, makes an indirect connection between extremism and terrorism through a description of 153 ‘militant islamism’ as a manifestation of foreign political extremism. If a connection is to be assumed here it is to be inferred by implication (that terrorism is an manifestation of so-called militant Islamism): ‘Border surveillance should also in the long run be strengthened so that the Armed Forces, the Coast Guard, the Security Police and other relevant authorities can effectively and in cooperation obstruct and prevent militant Islamism and other forms of foreign political extremism from operating in our country’ (‘Gränsbevakningen bör även på sikt förstärkas så att Försvarsmakten, Kustbevakningen, Säkerhetspolisen och andra berörda myndigheter på ett effektivt sätt och i samverkan kan försvåra och förhindra militant islamism och andra former av utländsk politisk extremism från att verka i vårt land’, 2010/11:Fi231, our translation). 6.2. Motions in 2018 In 2018, as many as 50 motions were put forward that use the term extremism or any of its compound forms. The motions were signed by in total 132 MPs (many of them participating in several motions) and with the motions distributed across the political parties as follows: Moderates (14), Sweden Democrats (12), Liberals (11), Christian Democrats (6), Centre Party (5), and Social Democrats (2). This means that these motions were submitted almost exclusively by conservative/right-wing and liberal parties in opposition, the exceptions being two individual motions by MPs from the Social Democrats, the largest government party (2018/129:1154 and 2018/129:2081). The 50 motions were seconded to all the 15 different parliamentary committees with the top five committees being Justice (22), Constitution (15), Culture (9), Foreign Policy (8) and Education (8). This indicates that political extremism had by 2018 become an issue relevant to all areas of politics, although most often judicial matters. Notably, the 2018 motions indicate a shift in the localities associated with extremism. While the motions from 2010 are exclusively concerned with extremism outside of Sweden (violence and political extremism in foreign countries), the forms of extremism that are referenced in 2018 are primarily associated with domestic issues. ). Turning to the linguistic features of the use of the concept of ‘extremism’, the 50 motions entail 161 token of words containing extremism as constituent or as compounds. In the motions, extremism is predominantly depicted as an abstract phenomenon. In the motions, extremism is predominantly used to denote an abstract phenomenon or a vaguely contoured ideological position or sentiment. In 2018 the concept of extremism is used more widely and more specifically compared to 2010. The 16 different forms of extremism referred to could be grouped according to ideological affinity. The most common group consists of ideologically neutral collocations våldsbejakande extremism (71), extremism (59), politisk extremism (5) and våldsam politisk extremism (1); followed by various forms of religious extremism (religiös extremism (6), våldsbejakande religiös extremism (4), islamistisk extremism (3), våldsfrämjande islamistisk extremism (1), salafistisk extremism (1)) followed by the more traditional left- and right-wing extremists vänsterextremism (3), vänsterextrem våldsbejakande extremism (1), högerextremism (2), högerextrem våldsbejakande extremism (1), and as the smallest and ideologically ambiguous group of animal rights extremisms djurrättsextremism (1) and djurrättsrelaterad våldsam extremism (1). The fact that the single most common extremism collocation is våldsbejakande extremism shows how this concept has gained room in parliamentary discourse. This seems to indicate a convergence between the general and open concept of ‘extremism’, on the one hand, and the technical-juridical and distinct notion of ‘violent extremism’, on the other. Furthermore, there is a strong connection between the use of the concepts of extremism and terrorism, since 32 of the motions in focus also use terrorism-related terms. A significant part of them (12) makes an explicit connection between extremism and the religion of Islam, as can be seen in Figure 6. The connection between extremism and left-wing (4) or right-wing (3) ideologies is weaker, but nevertheless in several cases Islamic, left-wing and right-wing extremism are mentioned together such as in the following quote: ‘The Swedish Security Police estimates that the total number [of violent extremists] is 3000, of which 2000 Islamists and 1000 from the right- and left-wing extremist environments.’ (‘Säpo bedömer att det totala antalet uppgår till 3000, varav 2000 islamister och 1000 från de höger- och vänsterextrema miljöerna.’ 2018/19:2454, our translation). 154 Figure 6: Extremism in its local discursive motion contexts. Example of how sentences with political extremism (green) are discursively connected in their respective context paragraph (yellow) and adjacent paragraphs to terror-words (purple) and to alleged existing Swedish notions of fear and threat (blue). (Motion 2017/18:4162) Finally, several motions mention anxiety about terrorism or extremism in Swedish society as part of the argument for political action, either directly or indirectly through the use of fear-inducing rhetoric. The direct use is when MPs claim that there is a fear (skräck) or worry (oro) in society or among the public. and the indirect use is when MPs refer to the existence of different threats (hot, hotbild, terrorhot) or designate specific entities as fearful (fasansfulla) or insecure (otrygga). Both strategies are illustrated by the following quote: 155 Terror attacks against European countries have become a recurring fearful moment in everyday life. The safety that a state should provide its citizens is rightfully put into question when our brothers and sisters, in Sweden and other countries, lose their lives due primarily to Islamic extremism. It is not a question of just an external threat that can be fought by the military or an insecurity that can be resolved by closed borders. The extremism is growing in our country, in insecure areas where the majority society and the rule of law have given away to parallel structures. (‘Terrorattentat mot de europeiska länderna har blivit ett återkommande, fasansfullt moment i vardagslivet. Den trygghet som en stat ska erbjuda sina medborgare ifrågasätts med all rätt när våra bröder och systrar, i Sverige och i andra länder, blir bragda om livet till följd av främst islamistisk extremism. Det är inte frågan om ett enbart yttre hot som kan bekämpas med militär eller en otrygghet som kan stoppas med stängda gränser. Extremismen växer inom vårt land, i otrygga områden där majoritetssamhället och rättsstaten har fått ge vika för parallella strukturer.’ 2017/18:4114, our translation and emphasis). 7. Conclusions In this paper, we have explored the intersection of public opinion as expressed in national survey data and political action in the form of data on motions by MPs concerning the issue of terrorism but also political extremism. Our study points to the attacks in the USA in 2001 and the wave of Islamist terror attacks from 2014 and onwards impacting the public’s anxiety and the MP’s activity. Furthermore, the investigation shows the significance of political sympathies in this context with both citizens and MPs on the right being more anxious about terrorism than those on the left. At the same time, the results indicate that the concern for political extremism among MPs has been politicized over the last decade. Throughout the paper we have argued for the importance of a contextual approach to the analysis of parliamentary data. We have found that the increase in the intensities of motions has been partly dependent on legislative initiatives and, thus, to some extent institutionally-driven. Our results also confirm governmental position being a key factor when it comes to the writing of motions about terrorism. While MPs from the center-right parties have been driving the writing of terrorism-related motions, they are decidedly less active when their party is not in government. Overall, we have showcased the analytical benefits of combining parliamentary data and survey data to trace significant patterns of convergence and divergence, including the discrepancies between the anxiety about political extremism among MPs on the left and their activity in the Parliament. This further points to the importance of a comparative approach for understanding the relationship between public opinion and MPs activities on terrorism. It is, however, difficult for us to elaborate on the contexts behind these results without further investigation. To develop the analysis, we intend to widen our perspective to include government propositions and parliamentary debate in the analysis. This will provide a more comprehensive scope on parliamentary data regarding the topic of terrorism in Sweden. Our perspective may also be deepened by analyzing the content and sentiment of motions and other parliamentary records, using language technology. The possibilities of exploring expressions of anxiety need to be pursued through a more in-depth analysis and with a range of language technology approaches, including sentiment analysis and the use of word embedding vectors, to gain a more developed understanding of the terrorism discourse in the Parliament. Another point of interest for future research is to further pursue the conceptual development of political extremism, including the collocation violent extremism, in relation to the concept of terrorism. There are signs pointing to that extremism might have to some extent replaced terrorism in Swedish political-institutional discourse. While our results show an increase in the use of extremism in motions in the second half of the 2010s, (see Figure 4 and section 6), it remains to be explored to what extent concepts of extremism have blurred the perceived boundaries between terrorism, political violence and other social phenomena deviating from mainstream political norms. 156 8. Acknowledgements The SweTerror project (https://sweterror.se) is funded by the research program Digitisation and Accessibility to Cultural Heritage Collections (DIGARV), which promotes data-driven research in the humanities and social sciences by the 2016 Swedish Government Research Bill and is jointly organized by the Swedish Research Council (VR), the Swedish Foundation for Humanities and Social Sciences (RJ) and the Royal Swedish Academy of Letters, History and Antiquities (https://www.digarv.se/en/). The national survey data from the SOM Institute used in the study can be accessed through Swedish National Data Service (SND): https://snd.gu.se/sv/catalogue/collection/national-som. 9. References [1] Brodén, Daniel, Mats Fridlund & Patrik Öhberg (2021). Vem räds terrorismen? Kluven oro och säkerhetsivrare i Sverige. In Ulrika Andersson, Anders Carlander, Marie Grusell & Patrik Öhberg (eds) Ingen anledning till oro (?). Göteborg: SOM-institutet, Göteborgs universitet. [2] Oredsson, Sverker (2001). Svensk fruktan: Offentlig fruktan i Sverige under 1900-talets andra hälft. Lund: Nordic Academic Press. [3] Glassner, Barry (1999). Culture of Fear: Why Americans Fear the Wrong Things. New York: Basic Books. [4] Altheide, David (2006). Terrorism and the Politics of Fear. Lanham: Rowman & Littlefield Publishing. [5] Best, Joel (2018). American Nightmares: Social Problems in an Anxious World, Oakland: University of California Press. [6] Haner, Murat, Melissa M. Sloan, Francis T. Cullen, Teresa C. Kulig & Cherly Lero Jonson (2019). Public concern about terrorism: fear, worry, and support for anti-Muslim policies. Socius, 5: 1–16. [7] Hibbing, John R. (2020). The Securitarian Personality: What Really Motivates Trump’s Base and Why It Matters for the Post-Trump Era. New York: Oxford University Press. [8] Bjereld, Ulf (1998). Kön och politiskt våld: Attityder under svensk efterkrigstid. Stockholm: Gidlunds. [9] Ribbing, Antonia (2000). Sveriges terroristbestämmelser – brottsprevention och demokratiska rättsstatsideal. In Janne Flyghed (ed) Brottsbekämpning – mellan effektivitet och integritet: Kriminologiska perspektiv på polismetoder och personlig integritet. Lund: Studentlitteratur. [10] Hansén, Dan (2007). Crisis and Perspectives on Policy Change: Swedish Counter-terrorism Policymaking. Diss. Stockholm: Försvarshögskolan. [11] Edlund, Jens, Daniel Brodén, Mats Fridlund, Cecilia Lindhé, Leif-Jöran Olsson, Magnus P. Ängsal & Patrik Öhberg (2022). A multimodal digital humanities study of terrorism in Swedish politics: an interdisciplinary mixed methods project on the configuration of terrorism in parliamentary debates, legislation, and policy networks 1968–2018. In Kohei Arai (ed), Intelligent Systems and Applications: IntelliSys 2021. Lecture Notes in Networks and Systems (Vol. 295). Cham: Springer. [12] Fridlund, Mats, Daniel Brodén, Leif-Jöran Olsson & Lars Borin (2020). Trawling the Gulf of Bothnia of news: a big data analysis of the emergence of terrorism in Swedish and Finnish newspapers, 1780–1926. In Costanza Navarretta & Maria Eskevich (eds) Proceedings of CLARIN Annual Conference 2020. Virtual edition: CLARIN. [13] Economou, Athina & Christos Kollias (2015). Terrorism and political self-placement in European Union countries." Peace Economics, Peace Science and Public Policy 21(2): 217–238. [14] Rouces, Jacobo, Lars Borin, & Nina Tahmasebi (2019). Political stance analysis using Swedish parliamentary data. In: Workshop Proceedings (Vol. 2364). Digital Humanities in the Nordic Countries 4th Conference. Aachen: CEUR Workshop Proceedings. [15] Pitkin, Hanna F. (1967). The concept of representation. University of California Press. [16] Stampnitzky, Lisa (2013). Disciplining Terror: How Experts Invented ‘Terrorism'. Cambridge University Press. 157 [17] Futrell, Robert, Pete Simi, & Anna E. Tan (2018). Political extremism and social movements. In The Wiley Blackwell companion to social movements. Hoboken: Wiley. [18] Jackson, Sam (2019). Non-normative political extremism: reclaiming a concept’s analytical utility. Terrorism and Political Violence 31(2): 244–259. [19] Onursal, Recep & Daniel Kirkpatrick (2021). Is extremism the ‘new’terrorism? The convergence of ‘extremism’ and ‘terrorism’ in British parliamentary discourse. Terrorism and Political Violence 33(5): 1094-1116. [20] Willumsen, David M. & Patrik Öhberg (2017). Toe the line, break the whip: explaining floor dissent in parliamentary democracies. West European Politics 40(4): 688–716. [21] Baumann, Markus, Marc Debus & Jochen Müller (2015). Personal characteristics of MPs and legislative behavior in moral policymaking. Legislative Studies Quarterly, 40(2): 179–210. [22] SOU 1999:88. Granskningskommissionens betänkande: Brottsutredningen efter mordet på Olof Palme. [23] Andersson, Christoph & Andreas Förster (2017). Nazisten som teg: Drömmen om att skapa högerextrema terrorceller i Sverige och Tyskland. Stockholm: Norstedts. [24] Lööw, Heléne (2017). I gränslandet – symbiosen mellan det organiserade och det oorganiserade. I Mattias Gardell, Heléne Lööw & Michael Dahlberg-Grundberg (eds) Den ensamme terroristen: Om lone wolfs, näthat och brinnande flyktingförläggningar. Stockholm: Ordfront. [25] Andersson, Dan-Erik (2018). Från terrorism till våldsbejakande extremism: Att institutionalisera ett nytt begrepp i svensk politik. In Malin Arvidsson, Lena Halldenius & Lina Sturfelt (eds), Mänskliga rättigheter i samhället. Malmö: Bokbox förlag. 158