“We Are a Small Country That Has Done Enormously Lot”: The ‘Refugee Crisis’ and the Hybrid Discourse of Politicizing Immigration in Sweden

ABSTRACT This article looks at mainstream political discourses about immigration in Sweden during the recent “refugee crisis”. It argues that different patterns of politicization of immigration have traditionally dominated in Sweden and focuses on Swedish mainstream politics wherein, as is shown, explicit focus on politicization via (previous as well as current) immigration-related policies still persists. However, as the analysis of Sweden's Social Democratic Party's Twitter discourse shows, a hybrid new discourse of politicization is now emerging. It allows political actors to legitimize immigration policy with often populist-like politicization and the use of new modes of online political communication.

and, hence, of political parties and groups that form governments and have sufficient opportunities to drive national/regional policy making and implementation. As such, this strategy traditionally boiled down to addressing immigration-related issues, not via open, political (and a public-sphere-based) debates but in fact through interparty and interinstitutional bargaining and the ensuing policy-making process.
Traditionally, the aforementioned politicization strategy significantly differed from its counterpart, politicization through the public sphere. Within the latter, the focus has been on a wider politicization logic whereby not only the political realm but also the wider public sphere (including the media, by now both traditional and online) would be addressed. These would be "used" to politicize certain topics, with both top-down and bottom-up political communication mobilized in order to make various politically driven "claims" (Koopmans & Statham, 2010). In case of immigration, it has by now become a matter of tradition in Europe and elsewhere that especially radical and right-wing populist parties strongly politicize immigration and carry out an agenda-setting role in immigration-related politics. They thereby spread the often strong, even racist, anti-immigration rhetoric and other patterns of mediated politicization of immigration (see e.g., Buonfino, 2004). These are often taken up by other political parties and they effectively significantly change the tone of wider public debates, making immigration onto one of the central political issues debated within and beyond politics. Hence, while the politicization through public-sphere articulation has traditionally been championed by right-wing populist politics, it has also had a spillover effect onto the wider field of politics and has also increasingly become embraced in recent years by the political mainstream (see e.g., paper on Poland, this special issue).
It is recognized here that in many national contexts the difference between the two aforementioned "types" of politicization might not be very clear and that they have often been part of one, path-dependent politicization process. The latter would in many cases pertain to subsequent or sometimes even simultaneous articulation of immigration-related views (in the wider public sphere), through its penetration into political agendas up to eventual inscription into the policy frameworks. Hence, the politicization through public-sphere articulation would often precede the politicization through policy making and legitimation, with ideas about, or in most cases against, immigration being first articulated in the public sphere, including politics and the media, and only after entering the political and policy-making domain.
However, this paper aims to showcase Sweden as a significantly different case. As is argued here, policy-driven politicization of immigration undertaken by the country's political mainstream has in fact for a very long time preceded strong articulation of immigration and refugee/asylum-seeker-related views in the public sphere in Sweden (see Rydgren & van der Meiden, 2016). This paper hence builds on the premise that, unlike in many other countries, the policy-related politicization in Sweden preceded the one based on public sphere articulation and, hence, was at the foundation of, to a large extent, reverse overall politicization logic than has usually been encountered elsewhere. As is shown below, in its current form, the policy-driven politicization of immigration draws extensively on the affordances of political communication in the context of mediatization (Esser & Str€ omb€ ack, 2014;Mazzoleni & Schulz, 1999) and online and digital politics (Chadwick, 2013;Dahlgren, 2005;Vaccari, 2013). It is hence a foundation of the, by now, hybrid politicization pattern reliant on (especially online) mediated politicized messages to legitimize policy-based responses. Its hybridity resides in the fact that the discourse of this pattern of politicization draws increasingly on combining those "typical" for several other politicizing-discursive dynamics. It is, in particular, realized though the linkage of both governance-driven policy actions/discourses and related discourse of "policy communication" (Krzy_ zanowski, 2013a), which, including the use of traditional and online media, now allows introducing policy-like measures and simultaneously creating legitimacy and prelegitimacy (Krzy_ zanowski, 2014) for the proposed policy solutions.
The analysis below is set in a wider context of ongoing and indeed often radical-especially in comparison to previous years-political change in Sweden. This has, to large extent, accelerated in the aftermath of the 2014 Swedish national parliamentary election, which brought Social Democrats back to power after almost a decade of center-right governments. It also very clearly marked a huge electoral gain for the Swedish by-now-established parliamentary right-wing populist movements, who in recent years have extensively propagated the anti-immigration discourse in Sweden. Finally, this period also coincided with a dramatic increase in the number of asylum seekers, particularly from the war-torn Syria, and culminated with the tightening of the policy regime on refugee entry and residence that was introduced by the Swedish government in November 2015.
While this paper certainly aims to avoid making simplifying, causal arguments, it is contextually vital to foreground the increasing role of (radical) right-wing and populist politics and discourses in the overall politicization of immigration in Sweden (see below) and especially in the rise of anti-immigration public attitudes (though these, to be sure, still remain relatively moderate compared to other EU countries; see Demker & van der Meiden, 2015). It is therefore also recognized here that the, by now, long-lasting presence of right-wing populist politics in the Swedish political arena has led to the fact that many views have eventually been "normalized" as standard immigration-related arguments (Green-Pedersen & Krogstrup, 2008;Odmalm, 2011). This has been significantly fuelled by the growing presence of the radical right-wing views embodied by the-by now, parliamentary-party politics of the Sweden Democrats (Sverigedemokraterna; see below; see Oja & Mral, 2013;Rydgren, 2005Rydgren, , 2006Rydgren, , 2008Rydgren & van der Meiden, 2016;Sannerstedt, 2015). However, the process has been even more radically reinforced by the widespread presence of the "uncivil society" (Ruzza, 2009), 3 which in the Swedish context has recently been particularly eager to utilize widely read, rightwing web platforms to disseminate and even further radicalize the anti-immigration messages including via often outright racist and discriminatory discourse (Ekman, 2014;Krzy_ zanowski, 2019, Krzy_ zanowski & Ledin, 2017). However, this paper contends that arguments that immigration was not, until recently, at all politicized in Sweden, including due to the general lack of parliamentary right-wing populist politics (until the early 1990s) and of wider political and public sphere debates (Dahlstr€ om & Esaiasson, 2013;Ivarsflaten, 2008;Odmalm, 2011;Rydgren & van der Meiden, 2016), certainly need further elaboration. They need to be supplemented by an in-depth look into selected strands of Swedish politics to recognize the salience of area-specific discursive traditions of politicization now increasingly "borrowing" some discursive elements from one another (indeed, as a token of discursive and political "hybridity" highlighted above). The above is especially vital as far as mainstream politics is concerned. It is within the latter, as is shown below, that the policy-oriented-politicization has been incepted, indeed long before the public-sphere-oriented politicization has developed in Sweden especially under pressure from Swedish right-wing populist politics. By the same token, the analyses of politicization of immigration in Sweden proposed to date require empirical analyses that would show that a very significant duality and lack of otherwise typical temporal subsequence (see above).
In order to capture a significant part of the aforementioned, complex dynamics of politicization of immigration in Sweden on the example of the recent "refugee crisis," this paper analyses discourses that can be deemed prototypical for the Swedish contemporary political mainstream. It looks specifically at the discourse of Sweden's Social Democratic Party (Sveriges Socialdemokratiska Arbetareparti, SAP) whose political ideas and narratives are analyzed as a token of immigrationrelated views typical of a mainstream party, which, while historically opting for policy-driven patterns of politicization, still constructs its discourses strictly in relation to policy yet by now in close combination with the wider discourse of policy legitimation or prelegitimation (Krzy_ zanowski, 2014), especially as augmented and communicated by the use of social/online media.
The paper hence looks at SAP views as that of Sweden's government party and analyses its ideas as a token of those that underpin the current policy landscape and wider governance-related discourse in Sweden. But, to be sure, the focus on the SAP also has a wider historical meaning as it is the Socialdemokraterna who have shaped the majority of Swedish postwar politics. The SAP has also long been assumed to be the standard bearer of core ideas related to Sweden's world-famous welfare system based on the now gradually dismantled Swedish "strong state" (Andersson, 2007;Lindvall & Rothstein, 2006) or of the internationally renowned Swedish social-democratic model also known as "people's home" or Folkhemmet (Andersson, 2009;€ Ostberg & Andersson, 2013). On the example of the SAP in the context of the "refugee Crisis," the paper hence looks in detail at Swedish mainstream-political, self-mediated discourses on the microblogging platform Twitter. As is shown below, the social/online media discourses are very efficiently used by the SAP to introduce and familiarize its immigration-related and asylum-related policy changes in the context of the "refugee crisis." The paper hence analyses discourses from the period of November and December, 2015, i.e. at the time when it was claimed that the Swedish state system could no longer "cope" with the large volume of incoming refugees. The paper shows that the mainstream SAP effectively constructed "the crisis" in its discourse chiefly to legitimize-and often indeed prelegitimize-the turn in immigration/refugee policy as a strategy of coping with the alleged crisis situation.

On politicization of immigration in Sweden
Sweden is a multicultural country that is often an international example of openness to diversity and equality. Since the beginning of the postwar period, it has been an obvious case of a country of immigration rather than emigration (Abiri, 2000) with, at present, over 20% of Sweden's inhabitants being of foreign origin and Sweden being one of the most diverse European societies (Schierup & A lund, 2011). Sweden is also known globally as a country driven by the principles of international engagement that, despite its relatively small size (as far as population), over most of the postwar period has often provided some very extensive humanitarian help to, and received thousands of immigrants from, many countries from across the globe that were torn by conflicts and political unrest ( € Ostberg & Andersson, 2013). Such has also been the case in the course of the most recent "refugee crisis," when as many as 80, 000 asylum seekers were registered in Sweden alone throughout the months of October and November 2015 with their overall number reaching over 170,000 by the end of 2015 (UNHCR/Global Trends, 2016). This was the biggest number of asylum seekers ever received in Sweden, surpassing even the previously historically largest group of Balkan region migrants of the early 1990s (see below).
During the initial stages of the postwar period, the so-called guest-worker schemes, similar to those elsewhere in western Europe, supported the Swedish economic boom of the 1950s and 1960s ( € Ostberg & Andersson, 2013) and made Swedish immigration policy generally very liberal (Hinnfors, Spehar, & Knapp, 2012). At the same time, asylum application numbers-especially those issued pertinent to the Geneva Convention rules-remained relatively low in Sweden. However, starting from the early 1970s the rules of economic immigration became stricter and were largely limited to the family members of those immigrants admitted to Sweden before the late 1960s. At the same time, the number of asylum applications started to increase (for details, see overview in, e.g., Schuster, 2000) as part of a trend that was augmented by Sweden's increasingly verbal engagement in international affairs-and with Swedish leaders often opposing authoritarian and colonial politics in many countries. Sweden was offering shelter to political and other refugees from, in particular, the late 1960s onwards ( € Ostberg & Andersson, 2013). This trend continued well into the 1980s and especially the early 1990s (Balkan refugees) and, most recently, since 2014.
Only from about the 1970, has the Swedish immigration policy moved into the focus of national political actors who increasingly looked for various ways to regulate both temporary and long-term immigration. Interestingly, they in most cases manage to do so as part of government politics and administration and while quite effectively keeping the topic outside of public-wide debates. From the early 1980s onwards, asylum policy followed suit (Abiri, 2000) with its ever-stronger focus on temporary rather than long-term forms of protection or on the refugee integration and naturalization. While this trend was to some extent stopped in the early 1990s-when Sweden proved very generous to thousands of asylum seekers from the war-torn Balkan region (about 100,000 in 1991 and 1992 alone; see Schuster, 2000, and below)-from the mid-1990s onwards Swedish policy became ever more conservative (although remaining relatively more liberal in this respect than other developed western European countries). This to some extent happened as part of the process of adjusting Swedish legal frameworks to the EU regulations, upon Sweden's EU accession in 1995 (Schierup & A lund, 2011;Schuster, 2000). It is rather difficult to overlook that a very prominent role in the process of policy-based politicization of immigration in Sweden has historically been played by the Swedish Social Democrats (SAP). SAP immigration policies concerned both immigrant entry (Hinnfors, Spehar, & Knapp, 2012), asylum (Abiri, 2000;Eastmond & Ascher, 2011), or immigrants' social and welfare rights in Sweden (Eger, 2010;Sainsbury, 2006;Schierup & A lund, 2011;Schuster, 2000). Despite being a social-democratic party with an "ideological foundation on public sector support, solidarity, inclusiveness and socialism with international roots' (Hinnfors, Spehar, & Knapp 2012, p. 586), the SAP often legitimized its immigration policy with its primary role of a defender of interests of the Swedish national community-as part of its key idea of Swedish Folkhemmet (Andersson, 2009;Pred, 2000). SAP also often emphasized its strong reliance on trade unions and trade union associations whose main interest has often been in protecting the already existent collective agreements built chiefly on the demands of the national work force.
The initial postwar SAP immigration policy seems to be very strongly underpinned by the logic of (political) economy. While in 1954 the SAP was behind a relatively liberal Aliens Act, passed at a time when Sweden was in great need of a foreign workforce, the economic logic also prevailed when the SAP actually put a stop to the guest-worker schemes via a number of policy decisions taken in 1972. Similarly, in 1979, the SAP stood behind provisions that eventually excluded many immigrants (especially noncitizens) from pension and other welfare provisions in Sweden. This moment also constituted a significant shift in SAP immigration politics traditionally in favor of limited entry for migrants yet supporting a rather generous welfare provision to support integration of the foreigners including refugees who were already residing in the country (Hinnfors, Spehar, & Knapp, 2012;Schierup & A lund, 2011).
While political economy was the driving force in Swedish immigration policy through the 1980s, from the 1990s onward one can observe-similarly to elsewhere in the social democratic systems in Europe (see Krzy_ zanowski & Wodak, 2009)an increase of strictly political considerations regarding immigration. At this time, the SAP was the driving force behind the passing of the widely discussed 1997 Aliens Act. Interestingly, in this and other cases, the SAP legitimized its legislative initiatives by real-world occurrences such as, e.g., the Swedish economic downturn of the early 1990s, the huge increase of both asylum seekers and refugees of the same period, or indeed the Swedish 1995 EU accession. Most of those occurrences, however, to large extent took place still during centrist and right-wing governments that preceded SAP's return to power, while during SAP's renewed government tenure the number of immigrants and especially asylum applications in Sweden started to fall rather sharply as opposed to other EU countries (see Hinnfors, Spehar, & Knapp 2012, p. 13).
The period of the mid-to late 1990s continues the above trend with immigration politics in Sweden underpinned by, in particular, arguments over the necessity of adjusting Swedish immigration and asylum policy to the EU legal framework (Schierup & A lund, 2011). However, it is rather obvious that the ongoing and in fact, at that time, increasing wide politicization of immigration in Swedish politics and the public sphere was related to the emergence and electoral success of Swedish right-wing populist parties such as the New Democracy (Ny Demokrati) of the early the 1990s, which made immigration and asylum into widely and publically debated issues. Those developments posed a requirement for the other political parties to "respond" or at least "take a stand" and present as well as legitimize new, often harsher, policy solutions on immigration. In the end effect, the political mainstream in fact came with several responses and policies (Abiri, 2000). These were, including in case of the SAP, often spiced with "objectifying" claims that the immigration and asylum situation in Sweden since the 1980s was anyway "beyond control" and thus had to be regulated (NB: Some of those arguments were, as is shown below, recontextualized in the 2015 discourse on the refugee crisis).
In the 2000s, largely recognized as a period of the further politicization of especially labor immigration by the Swedish political mainstream (Widfeldt, 2015), the SAP continued on both its traditional policy-oriented path of politicization and by simultaneously making its claims ever-more visible in the wider public sphere, including the interparty competition and media discourse. Still before its eventual departure from the government in 2006, the party continued its policy even if in an often ambiguous manner-that is, while restricting the Swedish labor market to third-country nationals and at the same time liberalizing other policies of, for example, labor market access toward EU citizens (Hinnfors, Spehar, & Knapp, 2012). Upon its eventual return to power in 2014-the moment that coincided with the further rise of especially asylum applications and the heyday of the current "refugee crisis"-the SAP continued the immigration-related policy of previous years. It mainly focused on the issues related to entry and temporary residence laws regarding especially asylum seekers.
The most politically significant of the immigration policy solutions proposed by the current SAP-led government to date-and indeed those that constitute the main reason for the discourses analyzed below-were those introduced on November 24, 2015. In the regulations, asylum-seeker entry and temporary residence were restricted by way of introducing several temporary (in most cases, initially 3-year) exemptions from the Aliens Act. The exemptions have in fact revived previous SAP ideas about the further differentiation between "quota" (or Convention) refugees and the newly arriving ones who, according to the current and the previous SAP proposals (see above), would in most cases be granted only a "temporary protection" status. The proposals also stipulated that the temporary refugee status could be transformed into a permanent one providing refugees receive employment (a requirement that often proved very difficult to fullfil). It also increased the age level for spouses in family reunification to 21 years of age. Finally, the new regulations also restricted refugee-children laws to those who at the moment of granting status would not exceed 18 years of age and included provisions on medical age-checks for asylum seekers not holding the relevant documents to prove their age. The regulations also imposed temporary ID checks on all modes of public transport into Sweden.

Design of study and methods
The analysis below is performed on a material collected during a 6-week period between November 1 and December 15, 2015. This is the period when, in the aftermath of the peak of the "refugee crisis" in Europe (throughout October 2015), the number of refugees arriving in Sweden rose very sharply. As a result, midway through the period of investigation, the change of refugee-and asylum-related policy is announced (on November 24, 2015), resulting in the furthering of the debate about how Sweden (and Swedish politics) should shape its immigration and asylum policy. The period is also marked by the terrorist attacks in Paris (Nov 13, 2015), quite obviously reflected in the government and mainstream political discourses and their focus on security and related matters. The Paris attacks were also well visible in the right-wing populist discourse of the same time, often conflating anti-refugee and Islamophobic expressions.
The analysis has a dual focus. On the one hand, it focuses on the SAP Twitter discourse and examines, at first by means of topic-oriented analysis, the overall development of the key themes and frames of the SAP's mediated political discourse in the studied period of 6 weeks. On the other hand, a closer look is also taken on SAP social media discourse of one selected day, Nov 24, 2015, when the party announced the tightening of asylum-seeker regulations. Within that second step of analysis, the focus is on how policy and related political action is discursively legitimized by the SAP as a political actor behind the policy changes. The sample discourse is hence approached as to large extent prototypical for the politicization discourses within political mainstream in Sweden.
In line with the discourse-historical tradition in critical discourse studies (DHA; see, inter alia, Krzy_ zanowski, 2010;Wodak, 2001), this study follows a perception of discourse as a social practice. The focus is on how social phenomena-in our case those related to immigration and specifically the "refugee crisis"-are constructed and represented linguistically, in this case in political discourse. This allows seeing discourse as a carrier of different forms of legitimation of discursively constructed forms of political identity and agency.
In fact, this study builds on the previous work on politics, radical rightwing populism, and anti-immigration discourses and mobilization in Europe (Krzy_ zanowski, 2012(Krzy_ zanowski, , 2013aKrzy_ zanowski & Wodak, 2009) as well as on media and migration (Krzy_ zanowski, 2014). It analytically follows the DHA of right-wing populism and anti-immigration rhetoric (especially, Krzy_ zanowski & Wodak, 2009;Wodak, 2015), which focuses extensively on the deconstruction of various discursive strategies deployed in political language and discourse to expose their role in real-world political, including politicization, strategies and actions. In the current paper, the focus is, in particular, on strategies of legitimation and how the latter are deployed in policy-driven discourses to both present, or recontextualize (Bernstein, 1990;Krzy_ zanowski, 2016) policy, and to legitimize its social and political validity.
The analysis will follow a multilevel pattern of analysis originally introduced by Wodak (2001) and later elaborated by Krzy_ zanowski (2010). It distinguishes between the entry-level (thematic, topic oriented, or content oriented) and in-depth (strategy oriented including argumentation oriented) analysis. Within the former, the focus is mainly on discourse topics that allow one to map the key tendencies in the contents of analyzed mediated texts and to sketch the tendencies in overall framing of the discussed issues. The in-depth analysis, on the other hand, focuses on the key patterns of argumentation. It will rely on the notion of discursive strategy as introduced by Reisigl and Wodak (2001) and, thereby, on the category of topoi (or argumentation schemes; Krzy_ zanowski, 2010), especially in the forms known from public and political anti-immigration discourse and as part of wider sets of discursive strategies. 4 As this study is mainly interested in how political (and especially policy) action is represented and legitimized, its focus on political-discursive strategies is supplemented by a number of categories that help in identifying different, often overlapping, types of legitimation. Here, the study follows Van Leeuwen's (2007) analysis of strategies of legitimation that have previously been effectively combined with DHA-driven examination of public discourse (see Krzy_ zanowski, 2014;Van Leeuwen & Wodak, 1999). Specifically, the analysis focuses on the so-called authorization strategies (Van Leeuwen, 2007) wherein various forms of authority are used to lend legitimacy to social and political actions. Here, the difference will be followed between what Van Leeuwen saw as "impersonal authority" (of laws, regulations, acts, or facts), "personal authority" (of individual and collective social actors), and authority of tradition (highlighting the salience of path dependency in social action). On the other hand, the analysis will also showcase strategies of moral legitimation (Van Leeuwen 2007: 98ff), which highlight the salience of either evaluative-and indeed often value-laden-language and discourse strategies (pertaining to, e.g., positive or negative assessment of past, present, or future actions) or those that construct analogies (e.g., between current and past actions) allowing for their positive/negative valuation and thereby legitimation.
The SAP social media discourse: @socialdemokrat Key themes and overall dynamics of the @socialdemokrat discourse The analyzed Twitter discourse of @socialdemokrat was produced in the relatively high level of account activity-with 201 tweets (Ts) and retweets (RTs) in total for the 45-day period of analysis. Within that period, the account was relatively interactive with a number of RTs (141 total/70%) clearly outnumbering that of Ts (65 total/30%). As for the RTs, they originated in a variety of sources. The main ones were individual Twitter accounts of key SAP politicians (especially of the widely known Swedish foreign minister @margotwallstrom), of their aides (e.g., @nataliesial, press secretary of minister for employment), and of various ministries and government agencies (@FinansdepSv, @Migrationsverk). Thus, although coming not from the main account in question, the RTs were still focally very close to the Ts as both the former and the latter focused in most cases in a positive manner on government actions and policies.
Other RTs also originated in EU-related Twitter accounts of-for example, EU commissioners (@avramopoulos) and also of Swedish EU representations (@Swe-deninEU) or foreign embassies in Sweden (e.g., @FranceinSweden). RTs from media organizations were also frequent (@Expressen, @SvD, @dagensnyheter) and they played a very specific function-they particularly intensified in periods before official government and related statements were issued and tweeted (so, e.g., in the early phase of the Paris attacks of November 13, 2015), hence, at the time when the media voice was used instead of that of the politicians.
The Twitter discourse of the SAP reveals that the @socialdemokrat was mainly interested in immigration-and refugee-related actions of the Swedish government and thereby of the key Social Democratic officials. Table 1 illustrates the key "target" semantic fields of the @socialdemokrat Twitter discourse as represented by hashtags (theme markers) used in both Ts and RTs (though it must be mentioned that only about 30% to 40% of tweets and retweets were effectively tagged). As the hashtags depict, the discourse was visibly strategically organized into two areasdomestic and international-thus showing both local and transnational salience of Swedish government politics with regard to immigration and asylum seekers/ refugees. As far as "domestic" or Sweden-related tags and topics were concerned, these were clearly led by the (#svpol), a generic tag used in debates related to Swedish politics (though widely known to be a tag used especially by the political right). In addition, general tags related to Swedish immigration policy were used (#MigPol) on a par with tags referring to government policies and actions specifically concerning the refugee crisis-for example, the government's refugee settlement program (#etableringspaket) or a program aiming at fast-tracked skills and education recognition of refugees (#snabbspar). Further, both policy areas (#utbpol, education policy) and corresponding specific areas of socioeconomic reality (#skolan, #jobb) were tagged so as to broaden the scope of wider efficiency of the governmental immigration and refugee-related activities. These were often supported by more generic, one may say, problem-specific tags such as #flykt (refuge) or by means of tagged references to social initiatives such as those directed against refugee-oriented hate actions and hate speech (#mothatbrott, against hate attacks; #NoHateSE). The latter, though not originating within government politics, were presented as if being intact with relevant government activities.
The last vital area of the domestic themes were the references to the Swedish media that were not only mentioned by means of account names (@aftonbladet), as one would expect, but tagged and made into topics of tweets. It seems the tabloid #Aftonbladet was mentioned most commonly in such a way-apparently as the main agora for immigration-related discourse of the Swedish media. Similarly, the Swedish public TV broadcaster SVT was mentioned, yet mainly by means of building references to its specific TV shows (#svtnyheter, #svtagenda), and mostly at times of the Swedish government officials' appearances therein.
While the domestic framing above seems relatively broad and mainly issue centered, the internationally tagged Ts and RTs were mostly event related. Such was the case with all tags thematically marked as (#Paris, #Parisattacks, and #terroristattack), which referred to the Paris terrorist attacks of November 13, 2015-that is, the main international event in the period of investigation. Other than that, EU-related events were tagged, mainly by means of abbreviated references to various meetings within the EU council, though with a quite obvious preference for those in which Swedish government ministers took place. The timeline of @socialdemokrat tweets during the period of investigation (see Figure 1) indicates that the discourse peaked on several occasions, two of which clearly stand out. The first peak occurred the day immediately following the Paris terrorist attacks (November 14, 2015) and marked a period during which, at first, Swedish media and international reactions (mainly via RTs) and then government response (via Ts) were communicated.
The second event, which produced the most significant peak of @socialdemokrat tweets, is dated November 24, 2015 and relates directly to the press conference in which Social Democrat Swedish prime minister and his deputy (from the traditionally immigration-friendly Green party) announced a significant change in immigration and asylum seeking and in refugee regulations in Sweden. The event was widely televised and broadcast online and is probably remembered not only for the fact that new, stricter immigration and refugee policy was announced but also for the fact that, "deeply distressed" with the need to introduce such policy, the deputy prime minister A sa Romson broke into tears midway through the conference. 5 Contrary to the Paris attacks' coverage, the November 24 government press conference was mainly communicated by means of Ts by @socialdemokrat, with some interesting interplay of RTs (see below). Overall, the press conference constituted a political (including policy) communication attempt that used Twitter as a channel for explaining and to a large extent legitimizing the political actions (to be) undertaken. Refugee crisis and the hybrid discourse of policy legitimation: The case of @socialdemokrat Looking more closely at the in-depth qualitative features of Twitter representation of the aforementioned press conference of 24/11/2015, it becomes evident that, probably just like the event of the announcement of the policy change itself, its reporting on Twitter by @socialdemokrat was a very well planned and executed as a token of mediatized political communication. Hence, the @socialdemokrat Twitter discourse of 24/11 evolved gradually, mainly as a result of using solely Ts and obviously avoiding RTs (which could not be sequentially ordered). Such "staging" of the press conference on Twitter-by now a recurrent feature of press conferences that are not only performed "live" but also online (Ekstr€ om, 2016)-could, in fact, be clearly divided into three stages (see Figure 2): Part A (Setting the Scene) in which the conference as such is announced Part B (The Announcement) in which the statement on the policy changes by Swedish Prime Minister Stefan L€ ofven takes place Part C (The Follow-Up) in which reactions to the government statements and follow-up interpretations are provided As far as Part A was concerned, this was rather straightforward and consisted of two own Ts announcing the press conference. The only interesting aspect of this part was that the Tweets were actually posted immediately before the press conference was about to start, with the second of them even starting "Just nu" (right now), indicating that the event was already in progress. Also, both of the tweets of the first part included a link to the main Swedish government website (regeringen.se). Press Infromation, 6 yet without specifying whether the link was referring to the live stream of the video conference or to the actual policy documents.
Part B, consisting solely of a series of @socialdemokrat Ts, was supposed to serve as an announcement of the new policy. However, throughout this part, comprising almost solely edited statements made by Swedish SAP Prime-Minister L€ ofven during the actual press conference (with almost all of them heavily tagged with the generic #svpol), no details of the new policy were actually announced. Instead, the majority of tweets focused on describing Sweden's past actions toward refugees and asylum seekers as a way of legitimizing the stricter immigration regulations. This way of discursively legitimizing future/imagined actions with traditional legitimation toolsalso known as pre-legitimation strategy (Krzy_ zanowski, 2014) overtly resembled the aforementioned Press Information, published at www. regeringen.se, which also focused mainly on mentioning previous rather than current refugee-related actions.
In Part B, the tweets, quite surprisingly, moved away from looking at refugees or asylum seekers and their plight and instead focused on creating the image of Sweden as not only refugee friendly but also globally responsible. This was evident in one of the opening tweets, which argued, "L€ ofven: Sverige tar sin del av det globala ansvaret f€ or m€ anniskor pa flykt. #svpol" (L€ ofven: Sweden took its part of the global response to people in refuge) and followed the impersonal authority (of previous actions) legitimation strategy. A tweet that immediately followed continued the previous argumentation but turned to "Swedish people" as its subject. There, it was claimed that, "L€ ofven: Det svenska folket har visat och visar en stor solidaritet i denna tid. #svpol" (The Swedish people have shown a huge degree of solidarity at this time), thus presenting government-represented by the prime minister-as speaking for all Swedes. At the same time, "legitimation through evaluation" also took place here, with the evaluative "huge degree of solidarity" deployed and augmented by the personal-authority-based legitimation through the reference to "the Swedish people" in the nominal position. Such references to the Swedish collective "us" were further constructed in one of the central tweets wherein L€ ofven is quoted as saying, "Vi € ar ett litet land som har gjort en enorm insats, och vi har visat stor solidaritet i en sva r tid" (We are a small country that has done enormously lot, and we have shown great solidarity in a difficult time). Here, the first part continues the legitimation through personal authority logic (by referring to, yet again, collective Swedish "we"), whereas, overall, the tweet deploys a number of strongly evaluative statements ("enormously lot," "great solidarity," "difficult time" that support the overall legitimatory claims. However, it is in the following tweet that one is able to eventually see where the statements were aiming. In the tweet, "Ingen kan tvivla pa Sveriges vilja att hj€ alpa flyktingar. Men just nu klarar vi inte mer. Fler ma ste s€ oka sig till andra l€ ander. #svpol" (Nobody can doubt in Sweden's will to help refugees. But right now we cannot handle more. More must search their place in other countries), it became obvious that the previous statements were just a foundation for the change of not only policy but also government rhetoric. The latter draws here on an analogybased strategy (of legitimizing current actions by referring to previous ones) as the main tool. Yet, from here onwards, only general statements are made and, in fact, no details of the new policy are explained except for a few (and still mostly legitimatory) statements that the "level" of asylum seeker intake in Sweden should match that in the other EU countries. The tweet "Lagen ska tillf€ alligt anpassas till miniminiva n i EU i syfte att fler ska v€ alja att s€ oka asyl i andra l€ ander. #svpol" (The level is to be temporarily adjusted to minimum level of other EU countries with the aim of more-MK: refugees-seeking asylum in other countries) comes next while legitimizing through impersonal (fact-based). It scarcely describes action toward refugees yet reveals its overall aim-that is, to radically (yet, as is claimed only "temporarily") limit refugee intake and redirect asylum seekers to other countries. However, quickly after this statement is made, most tweets return to the legitimation pattern indicated above, whereby, the focus is yet again on creating an analogy between current actions and what Sweden has done in the past. One of the Ts argues, "Vi har tagit emot 80 000 asyls€ okande under de senaste 2 ma naderna och sa ma nga barn att det motsvarar minst 100 nya skolklasser i veckan" (We have taken in 80,000 asylum-seekers in last two months and large number of children that corresponds to at least 100 new school classes per week), providing more facts (numbers) to the overall prelegitimation yet still failing to provide details of the new policy. Part C of the press conference, the Follow-Up, consists mainly of RTs from other government officials and staff interspersed with further Ts from @socialdemokrat. Interestingly, it is only in this final part that we find out the actual links to the newly announced immigration and asylum policy. Those links are not provided by the focal account but via RT from Sweden's Ministry of Justice account (@Justitiedep). In fact, the links are repeated once again in a very peculiar tweet from the Swedish foreign minister (@margotwallstrom), who not only provides the link to the policy but also legitimizes its introduction with a short preceding statement, "En akut flyktingkris kr€ aver politiskt ansvarstagande" (An acute/grave refugee crisis merits taking a political responsibility). This tweet-drawing on the evaluative legitimation strategy that constructs the crisis as "acute/grave"-is particularly vital as it originates from the account of SAP's Margot Wallstr€ om, indeed, an internationally recognized "icon" of Swedish international and humanitarian policy. Wallstr€ om's voice, hence, lends further legitimacy to the government policy and its change.
Part C also includes further legitimising voices, mainly by means of RTs of positive statements about the policy announcement. For example, an RT from Social Democratic politician Matilda Ernkrans (@ernkrans) is retweeted and states, in a declarative/speculative tone otherwise characteristic of Part B above, "Regeringen visar ledarskap i ytterst sva r situation. Sta r upp f€ or asylr€ att och s€ atter press pa gemensamt ansvarstagande inom EU. #svpol" (The government displays leadership in an utterly difficult situation. Standing up for asylum rights and putting pressure for response across the EU). The first part of this tweet follows the evaluationbased legitimation by describing the current situation as "utterly difficult" and thus discursively constructing the call for political action. On the other hand, the second part of the tweet seems equally legitimatory in nature as it claims that being for asylum/refugee rights might also mean stricter immigration and asylum laws (and hence using here impersonal legitimation through facts). Further legitimation is also provided in part C by RTs from the account of the Swedish Social Security minister Annika Strandh€ all (@strandhall), which to a large extent recontextualizes L€ ofven's earlier statements. For example, the prime minister's earlier statements about 80,000 refugees already accepted in Sweden (see above) are given a new textual context, as in "Man ska se regeringens a tg€ arder mot bakgrund av att Sverige nu tagit emot 80 000 flyktingar pa tva ma nader. Inte ett normall€ age" (One should see government's measures against a background of the fact that Sweden took in 80,000 refugees in two months. That is not a normal situation). Especially the latter part of the Tweet seems very strategic as it not only intensifies/fortifies the statement with an evaluative legitimizing claim but also implies that restoring the "normal situation" must entail a harsher immigration and asylum policy.

Conclusions
As the analysis above indicates, the recent refugee crisis has been fully entangled in Swedish politics in discourses that belong to the country's long-standing traditions of politicization of immigration and, in particular, to the mainstream-political tradition of politicization though policy. Indeed, the centrality of policy and policydriven solutions is still very obvious in the analyzed material, as would be to large extent expected of the mainstream Swedish political discourse in general as well as the SAP discourse in particular. As has been shown, both of these discourses have traditionally been centered on governance-driven solutions and have long been constructing preference toward migration-oriented policy rather than addressing immigration in the process of wider claim making and debated in the Swedish public sphere. What remains, however, very interesting about the analyzed material is the fact that, while the SAP discourse as such is still constructed around policy, especially at the thematic level, the majority of that discourse's argumentation is, in itself, mainly legitimatory or pre-legitimatory in nature. It is hence very explicitly focused on legitimizing policies sometimes to such an extent that the actual details-or contents and provisions-of the policy are almost completely omitted. This, as has been shown, is largely strategic in nature as it allows communicating policy almost irrespective of its actual contents and its de facto impact on those concerned-in the current case, migrants in general and asylum seekers/refugees in particular. Indeed, as the analysis shows, while produced and set strongly in the context of the "refugee crisis", the examined discourse is almost not at all primarily focused on refugees or their plight but on Swedish politics and its apparent efficiency in dealing with the refugee crisis. This shows the overarching role of "policy communication" (Krzy_ zanowski, 2013a) in the contemporary politics of immigration that is increasingly focused on those who undertake actions (especially, mainstream politicians) rather than those who are directly affected (migrants, refugees). As such, this is also a typical feature of key public discourse on politics/policy, which often strategically and purposefully omits or silences the "benefactors" of political and in particular policy solutions (Krzy_ zanowski, 2016).
Hence, as the analysis indicates, the refugee crisis has been constructed in the Swedish political mainstream-in this case the governing Swedish SAPwithin a hybrid mode of politicization. The latter, while relying on policy yet mainly as a certain largely undefined and vague idea that emphasizes political actors rather than action, extensively draws on the affordances of mediated political communication and, hence, to a large extent, on aspects of the traditionally distinct form of politicization in and through the public sphere. Such hybridity of the deployed politicization strategy is certainly very efficient: It allows the studied mainstream politics to address the shape of not only policy-based solutions but also wider debates that are thus informed by actions undertaken by, in this case, the SAP. Hence, in doing so, the political mainstream clearly moves beyond its traditional scope of action on immigration and also aims toward the wider public-sphere politicization of the issue traditionally reserved for nonmainstream and right-wing populist politics.
Finally, although seemingly focused on (policy) actions and (policy-based) response, the discourse of the SAP is very strongly ideological in nature. It focuses, almost irrespective of the refugee-related policy, on the discursive construction of the Swedish national "in-group" and serves to highlight both current and past refugee-related actions as positive and thus as sustaining, especially the SAP's political identity. It is, as such, also a hybrid case of the use of social/online media in the context of political communication (Krzy_ zanowski & Tucker, 2017). The social media are, effectively, entangled here in a top-down political communication process and strategy that only uses channels of social/online media to create the image of openness and interactivity (see also Calhoun 2016). They also allow the portrayal of organized political action (cf. the process of "staging" of the press conference) and thus yield legitimacy to the presented political actions.
Notes 1. Despite its by now widespread presence in public and academic discourse, the notion of a refugee crisis is approached critically in this article. It is viewed as an ideologically charged notion developed in the media and political discourse and also as a recontextualization of earlier (negativized) descriptions of large-scale developments related to immigration and asylum seeking (e.g., in the context of wars in former Yugoslavia in the early 1990s). 2. This paper draws on definitions of asylum seekers/refugees as widely accepted in the discourse of transnational institutions, such as the UN (http://www.unhcr.org/asylumseekers.html) and the EU (see http://ec.europa.eu/dgs/home-affairs/what-we-do/networks/ european_migration_network/docs/emn-glossary-en-version.pdf), and of Amnesty International (2016). I am grateful to JIRS Editor-in-Chief Anna Triandafyllidou for her suggestions regarding the issue of definitions. 3. Although he acknowledges a wide array of types of contemporary uncivil society movements, Ruzza (2009, p. 88) sees uncivil society primarily as "groups which have a self-professed antidemocratic and exclusionary political identity" and, in a majority of cases, associated with the political, including radical, right. Whereas these types can be prototypically defined as a form of civil society (especially due to its closeness to the "social" base), the uncivil society movements are significantly different from the civil society inasmuch as the former (a) act against rather than for the benefit of the common good and democratic principles and (b) are, even if unofficially, often closely linked to political parties and groups rather being voluntary bottom-up organizations and effectively a "voice" of the civil society. 4. For a more in-depth explanation of these key analytical categories see Krzy_ zanowski (2019) or Krzy_ zanowski (2010, Chapter 2). 5. See https://www.theguardian.com/world/2015/nov/24/sweden-asylum-seekers-refugeespolicy-reversal (last accessed July 21, 2016). 6. See http://www.regeringen.se/artiklar/2015/11/regeringen-foreslar-atgarder-for-att-skapaandrum-for-svenskt-flyktingmottagande (last accessed June 19, 2016).

Funding
The work presented in this paper has been partially funded by the Swedish Research Council (Vetenskapsra det) grants 21-2013-1498 ("The Journalism-Politics-PR Interplay on Twitter: Hybridized, Cross-Professional Relations on the Web") and 2016-05464 ("Interactive Racism in Swedish Online Media, Press and Politics: Discourses on Immigration and Refugees at Times of Crisis").