Logics of Othering: Sweden as Other in the time of COVID-19

‘Othering’ – the view or treatment of another person or group as intrinsically different from and alien to oneself – is a central concept in the International Relations literature on identity construction. It is often portrayed as a fairly singular and predominantly negative form of self/Other differentiation. During the first months of the COVID-19 pandemic, Sweden at first glance emerged as exactly such a negative Other. This article problematises such a view of Othering. Departing from a narrative analysis of news reporting on Sweden’s management of COVID-19 in the United States, Germany and the Nordic states, the article proposes an ideal type model with four forms of Othering – emotional, strategic, analytic and nuanced – not recognised in previous research. These types differ in their treatment of the Other as more or less significant and in involving a more or less self-reflexive construction of the self. Although narratives in all these settings drew on previously established narratives on Sweden, they followed different logics. This has implications for our understanding of Sweden as an Other in the time of COVID-19, as well as of self/Other relations in International Relations more broadly.


Introduction
When COVID-19 began to spread in the winter of 2020, Sweden quickly gained considerable international attention in media narratives around the world (Swedish Institute, 2020). Narrative meaning-making is expected to be particularly active and visible during crises (Roselle et al., 2014: 74) -or 'circumstances of radical disjuncture of an unpredictable kind' (Giddens, 1984: 61), when it seems impossible to keep erstwhile autobiographical narratives coherent and consistent. The COVID-19 crisis can thus be expected to have intensified narrative identity reconstruction around the world (Hagström and Gustafsson, 2021). Moreover, internarrativity -or the way in which new narratives draw on and are shaped by narratives already internalised by target audiences (Spencer, 2016) -makes Sweden a fairly predictable target in this context. The country's adoption of a strategy that differed from those of most other states provided observers with yet more impetus to use it in their sense-making processes in relation to a largely unknown danger. Sweden has historically received attention for being particularly liberal, rational, pragmatic, secular and so on, and been described alternately as a 'model nation' and a deterrent example. More recently, it has also become a special target for illiberal and antagonistic narratives, as far-right and foreign authoritarian -mainly Russian -media outlets have sought to undermine democratic processes and institutions around the world (Bennett and Livingston, 2018;Walker, 2018). Such media tend to represent Sweden as an Other that symbolises all that is bad about liberalism, feminism and immigration (Ramsay and Robertshaw, 2019;Wagnsson and Barzanje, 2021). Sweden stands out even among the Nordic countries, as it has often been portrayed as ungovernable and in a state of chaos -a 'lost case not to be imitated' (Deverell et al., 2021: 30). At the height of the COVID-19 crisis, Russian state-sponsored media platforms continued to depict Sweden in a consistently negative way (Hellman, 2021), using the same kind of stereotypical and malign narrative strategies that they have been applying for some time (Wagnsson and Barzanje, 2021).
However, Sweden received attention and criticism for its way of handling COVID-19 not just from Russia. International media reporting in the first month of the pandemic widely sought to explain what were perceived as poor policy choices with reference to Swedish culture (Irwing, 2020). To what extent or how did Sweden emerge as a negative Other also among its friends and allies as the COVID-19 crisis struck around the world? This question has theoretical implications as the existing scholarly literature on identity construction portrays Othering as a fairly singular and predominantly negative form of self/Other differentiation. According to this logic, the positive identity of the self emerges through the construction of a negative Other.
Based on narrative analysis that compares American, German and Nordic media reporting on Sweden in the initial 8 months of the COVID-19 crisis, we argue that Othering is more varied, and propose a novel ideal type model. We define narrative as 'a temporally, spatially and causally connected sequence of events, selected and evaluated as meaningful for a particular audience' (Colley, 2017: 4). The three cases are similar, to the extent that these states are generally considered friends and allies of Sweden, and that all three contexts at first glance featured highly negative narratives about Sweden and COVID-19. However, we find that the US narratives about Sweden and the pandemic drew on previously established narratives in a largely strategic fashion. German narratives, by contrast, featured more analytic discussions, while Scandinavian narratives were emotional, but sometimes also nuanced. While Sweden was positioned as an Other in all three contexts, more was at stake in the Nordic states. The article contributes by demonstrating that not all Othering revolves around an Other that is understood as significant, that negative and positive forms of Othering can exist in parallel and that more and less self-reflexive narratives can become intertwined.
We begin by presenting our ideal type model and explaining how it emerged abductively (Peirce, 1934), that is, as we moved back and forth between existing identity theory and empirical analysis focused on how Sweden was narrated during the first few months of the COVID-19 crisis. We then discuss how Sweden has historically been positioned as an Other -both a model nation and a deterrent example -including in current debates on immigration and 'exaggerated' liberalism in the Nordic countries. This provides the background for an analysis of 'internarrativity', addressing whether or how narratives about Sweden and COVID-19 drew on previously established narratives about Sweden as an Other. After explaining the methodology, we present our analysis, which shows how the different logics of Othering appeared in the empirical material and drew on previously established narratives in different ways.

Logics of Othering
The identity literature in and beyond International Relations (IR) takes identity to be co-constituted with difference (Campbell, 1998;Connolly, 1991;Epstein, 2011). Whereas identity/difference marks an ontological relationship, political theorist William Connolly (1991) has noted that difference is prone to be 'converted' into Otherness (p. 64). In that sense, Othering signifies the 'behavioural' relationship, whereby the self is actually constructed through representations of and in interplay with Others (Rumelili, 2004: 29).
The existing IR literature has focused on a range of different self/Other relations. Thomas Diez (2005), for example, has proposed an elaborate ideal type model with four different Others, portrayed as an 'existential threat ', 'inferior', 'violating universal principles' and 'different' (pp. 628-629). The first subcategory -which has also been referred to as 'radical Other' (Hansen, 2006: 37-41), 'oppositional Other' (Vucetic, 2017: 3), 'antagonistic Other' (Herschinger, 2012: 74) or 'adversarial' Other (Abizadeh, 2005: 45) -is clearly privileged in the existing IR scholarship. This type of Other is securitised, or narratively constructed as a source of danger, an evil or a threat to identify against. In this mode of Othering, certain ways of being or acting in the world are thus denounced or excluded, while others are glorified or embraced. The self is ascribed positively loaded terms, such as 'normal', 'realistic', 'rational', 'responsible' and 'safe', which are differentiated from their negatively valued opposites (e.g. Campbell, 1998;Diez, 2005;Neumann, 1999). There is a paradox, however, because while the Other is positioned as threatening to the self, it is also necessary for imagining and reproducing it (Campbell, 1998;Hansen, 2006). This kind of Othering can have problematic consequences, and trigger or sustain international conflict.
In Diez's second subcategory, the Other is not necessarily a threat to handle, but positioned as substandard and in need of development. In such a comparative and often hierarchical relationship, the self is narratively constructed as exceptional, or 'exceptionalised' (Hagström, 2015: 124; see also Rumelili, 2004: 35). It is portrayed as temporally 'ahead' and ascribed positively loaded terms, such as 'civilised', 'modern', 'advanced' and 'developed', which are again differentiated from their negatively valued opposites (Doty, 1996). There is arguably overlap with the third subcategory, in which the self is 'not simply seen as superior, but of universal validity' (Diez, 2005: 628).
While the existing literature is mostly focused on negative Othering, as in Diez's first three ideal types, the fourth subcategory 'differs from the previous three in that it does not place an obvious value-judgement on the other' (Diez, 2005: 628). Alternative previous research also allows for positive or exemplary Others, which become the object of admiration and emulation (Rumelili, 2004: 36;Abizadeh, 2005: 45;Hansen, 2006: 38-41), or integration (Guillaume, 2010). There is also mention of more ambivalent Others, such as the 'stranger' (Berenskötter and Nymalm, 2021), with whom the relationship is arguably more ambiguous.
We embarked on this study of narratives about Sweden in the time of COVID-19 bearing these theoretical starting points in mind. As we began to analyse our empirical material, these distinctions struck us as relevant and helpful to the extent that we found both negative and positive Othering of Sweden. However, two additional important distinctions were not accounted for. First, the existing literature mostly revolves around the roles that 'primary' or 'significant' Others play in the construction of the self (Idel, 2019;Morozov and Rumelili, 2012: 34;Neumann, 1999: 5) but this article contributes by demonstrating that not all Othering revolves around such significant Others. Although Sweden was the object of much Othering during the pandemic, it did not always emerge as a significant Other. The article also demonstrates that negative and positive Othering of Sweden often existed in parallel, symbolising the identity contestation or politics of identity that tends to occur within and across states (Berenskoetter, 2010). Second, this article makes a distinction between narratives that construct the self in a more or less self-reflexive way. A large part of the constructivist identity literature assumes that identity is negotiated self-reflexively, or through the continuous revision of autobiographical narratives 'in light of new information or knowledge' (Giddens, 1991: 20). Meanwhile, narratives constructed around a negative Other often seem too exaggerated and polarising to fit the description of self-reflexivity (Hagström, 2021: 334-335). The article thus contributes by showing that more and less self-reflexive modes of Othering exist and can be intertwined.
The resulting logics of Othering -emotional, strategic, analytic and nuanced -can be placed in a 2 × 2 matrix (see Table 1). These logics of Othering are also ideal types, and it is possible to conceive of Othering processes that have traits of different logics. We outline the particularities of each ideal type below.
What we call emotional Othering can be clarified with the help of Lacanian identity theory. Psychoanalyst Jacques Lacan conceptualised the Other as more than just a counterpoint that defines the self through its negativity, as outlined above. The Other plays a vital role in the fantasies that are conjured up in order to explain and deal with the sense of lack that epitomises the self. In these fantasies, the Other is imagined as stealing the self's enjoyment, thereby rendering the self incomplete and wanting (Solomon, 2015;Stavrakakis, 1999). The enjoyment that the self desires typically consists of abstract and vague master signifiers, such as 'security', 'democracy' or 'community'. The self can only become complete if the Other's theft of those qualities is stopped (Solomon, 2015)  or, conversely, if the Other starts to provide the self with that which is seen as missing. This Othering logic is called emotional since it is driven by the emotional desire for full identity. According to Lacanian identity theory, this is Othering par excellence, but the reason is arguably that the predominant focus has been on significant Others. Strategic and analytic forms of Othering, by contrast, do not revolve around a significant Other, or an Other that is narrated as stealing the self's enjoyment. Instead, in strategic Othering, the Other is pinpointed in a more pragmatic fashion in order to make a specific point in a political debate. While much may still be at stake for the self, the specific Other in strategic Othering can be conceptualised as merely an empty vessel on which highly exaggerated and polarised negative and positive qualities are projected. This resonates somewhat with research on strategic narratives. According to this approach, actors strategically create, mobilise, diffuse and contest narratives, for instance, to get others to do what they would not otherwise have done, or to impose their own political views on others (e.g. Miskimmon et al., 2017). This is arguably how Sweden has been used as an Other in the Russian malign information influencing discussed above.
Analytic Othering, next, is neither strongly emotional and driven by desire, nor clearly strategising. Instead, it is a process characterised by a more openly probing and selfreflexive attitude. This resonates with existing research that emphasises how narrative meaning-making can be used to make sense of changing 'realities' and uncertainties (Heinrichs et al., 2020). Here, narratives are constructed to examine whether or how the self should reject or associate itself with the Other and its characteristics and policies. While much may again be at stake for the self, the Other in analytic Othering is treated as an ostensibly neutral laboratory.
The final Othering process explored here is what we call nuanced Othering. The Other is again significant, but narratives actively seek common ground with the self instead of amplifying small differences. Nuanced Othering thus traverses the distinctions between self and Other with a greater sense of complexity and ambiguity, so that they are sometimes blurred, problematised or challenged. This is what some existing scholarship in the psychoanalytical, postcolonial and poststructuralist vein conceptualises as a more 'healthy' mode of self-identification (Mälksoo, 2015;Untalan, 2020). It has been seen to occur, for example, in border zones and on 'the margins' (Chen and Shimizu, 2019).
Regardless of logic, we assume that the narratives through which Othering occurs will almost inevitably reference and reproduce widespread and previously established narratives, including institutionalised local or national myths, in order to gain resonance (Colley, 2017;Schmitt, 2018;Hagström and Gustafsson, 2021). However, the four logics arguably relate differently to such established narratives. Emotional Othering is expected to be most clearly bound to their content and form. The Other is well known for stealing the self's enjoyment and recent occurrences are seen as confirming such tendencies. Strategic Othering, by contrast, is more consistent with the notion that there is a pool of narrative 'material' that actors can draw on as they juxtapose self and Other more or less at will. Analytic Othering, next, weighs facts in an ostensibly neutral fashion, but cannot fully escape the content and form of established narratives. The same goes for nuanced Othering, which is an identification process that actively seeks to renegotiate the content and form of previously established narratives, in order to seek common ground. Yet, it thereby also serves to prove their lingering importance.
In sum, previous conceptualisations of Othering generally centre on whether or how the Other is depicted in a negative, and sometimes more positive or ambivalent, fashion. Our proposed model adds nuance and clarity to such conceptualisations by demonstrating four underlying logics through which processes of Othering can be structured. The article illustrates the relevance of the ideal type model below by presenting the narrative analysis through which it was developed, but the analysis also displays how such logics intertwine with negative and positive Othering. Hence, our model is clearly infused with and can be combined with previous conceptualisations. It is again important to stress the abductive nature of this project. The 2 × 2 ideal type matrix did not precede the analysis of how Sweden was narrated as an Other in international media reporting in the first 8 months of the COVID-19 crisis but was developed alongside it. The next section outlines established narratives of Sweden as Other, thereby clarifying why it was fairly predictable that Sweden became an Other in the time of COVID-19.

Established narratives about Sweden as Other
There are fairly strongly entrenched narratives about Sweden that specific storytelling about the country's way of handling COVID-19 would be expected to reference and reproduce. In short, Sweden has been narrated as particularly 'progressive' since the inter-war period (Czarny, 2018: 204;Glover, 2009: 247;Kythor, 2020: 213;Musiał, 2002: 131). In the 1930s, the country became internationally known as a 'middle way' (Kythor, 2020: 213) between capitalism and communism (Andersson and Hilson, 2009: 221;Musiał, 2002: 178). The notion arguably also referred to the country's 'neutrality', or an extended period of military non-alignment (Andersson and Hilson, 2009: 221). In both cases, the middle way was epitomised by a willingness to adjust to a complex reality, and Swedes were described as the 'ultimate pragmatists' (Marklund, 2009: 268).
During the inter-war period, the international community became interested in Swedish social engineering (Andersson and Hilson, 2009: 220). The country's allegedly unique approach to economic policy, labour politics and social welfare gave rise to the notion that Sweden was a 'model' (Czarny, 2018: 201;Kythor, 2020: 214). Apart from 'progressive pragmatism', the model also involved an idealised penchant for 'rationality', 'secularism' and 'science'. According to Musiał, 'the recognition of scientific rationality as a means of solving social and political problems thus became an integrated element of the progressive image ' (2002: 95). This denotes an Other that international actors could potentially learn from and emulate.
During the early 1950s, however, negative accounts began to appear in the United States, which described Sweden as a nation of 'manipulative social engineering and patronizing social control' (Marklund, 2009: 276). Anglo-US sentiment towards Swedish neutrality after the end of the Second World War was also less than enthusiastic (Kastrup, 1985). The Swedish model was moreover criticised as creating a terrible environment for personal progress and for curbing private initiative (Marklund, 2009: 278). Thus, while the political left typically narrated the Swedish model as a kind of utopia (Andersson and Hilson, 2009: 220;Kythor, 2020: 214), the political right criticised Sweden as a '"new totalitarian" state, whose citizens lived in a state of willing servitude . . . [or a] land of drab conformity and melancholy, unrestrained social engineering and control, or, more recently, of self-satisfied smugness and stultifying political correctness' (Andersson and Hilson, 2009: 220).
Since the 1950s, Sweden has been characterised as a country that values a work/life balance, societal equality and women's rights (Kythor, 2020: 214). To some, these were again admirable traits to emulate, but to others they constituted a negative template. Sweden thus became narrated as 'free' and 'socially liberal' in a positive sense, or a place where prudence had been abandoned in a negative one (Kythor, 2020: 216-217). In recent years, Sweden has come into spotlight particularly due to its acceptance of migration. It has also emerged as a negative Other as multiculturalism and religion have become narrated as a threat to Nordic progressiveness (Agius, 2017). The best known example of the use of Sweden as a symbol for exaggerated liberalism was the reference by then US President Donald J. Trump to disorder in Sweden back in 2017: 'You look at what's happening last night in Sweden . . . ' (BBC, 2017).
While it has sometimes been difficult to distinguish narratives about Sweden from narratives about 'Scandinavia' or 'the Nordic states' (Andersson and Hilson, 2009: 220), Emil Edenborg's focus group study demonstrates that many Norwegian and Danish respondents align -at least to some extent -with the Russian media outlet Sputnik's reporting on Sweden as a state in decline, or a warning example of a place where feminism and multiculturalism have 'gone too far' (Edenborg, 2021: 2). Similarly, a recent report shows that Nordic citizens' views of Sweden have deteriorated more than those of citizens in other European countries. The most commonly reported reasons are Sweden's policies on crime and migration, and the handling of the pandemic (Swedish Institute, 2021). The empirical analysis below examines how narratives about Sweden and COVID-19 drew on these previously established narratives, and according to which logics of Othering.

Method
The empirical analysis is based on articles about Sweden and COVID-19 published between March and October 2020 in news outlets in five countries: the United States, Germany, Denmark, Finland and Norway. The latter three were later conflated as 'the Nordic states'. In order to identify media narratives that were reasonably representative of the multitude of narratives in each country/context, we focused on large news outlets that were believed to be complementary in terms of their political profiles. In the case of the United States, we collected and analysed 23 articles from the Los Angeles Times (liberal), 45 from the New York Post (conservative), 44 from the New York Times (liberal) and 41 from the Wall Street Journal (conservative). In the case of Germany, we gathered and examined 19 articles from BILD (right wing), 15 from Der Spiegel (left wing), 20 from Süddeutsche Zeitung (centre left) and 13 from WELT (right wing). Finally, we focused on 31 articles from the Norwegian Aftenposten (conservative), 30 from the Finnish Hufvudstadsbladet (liberal/conservative) and 45 from the Danish Politiken (social liberal). While the media outlets in the United States and Germany arguably provided us with a representative selection of narratives in each state, the conflation of Denmark, Norway and Finland into 'the Nordic states' is methodologically more problematic. Narratives identified in a single newspaper certainly do not represent the complete spectrum of media narratives in each Nordic country. Thus, we cannot aspire to establish a comprehensive 'Danish', 'Norwegian', or 'Finnish' narrative. Rather, differences among narratives in these national settings were noted and reported but given this limitation, the analysis primarily served to identify shared tendencies in Nordic media narratives on Sweden.
We searched the respective news outlets' websites and large-scale search engines, such as PressReader and Factiva, using the key words 'Sweden' or 'Swedish' and 'corona' or 'COVID-19' in each language. We included all articles with a distinct focus on Sweden and COVID-19 but excluded articles that merely touched on the topic in passing.
We employed narrative methods to analyse the material in search for Othering logics that revolved around Sweden and COVID-19. Narrative analysis can be conducted in several ways (e.g. Bal, 1997), but it is common to focus on narrative components. Here, we focused on: (a) how Sweden was characterised as an Other, including the adjectives and adverbs used; (b) how the self was characterised, including the adjectives and adverbs used; (c) how causality was ascribed between different traits and events; (d) how narratives connected or differentiated the past, the present and the future, or temporality; and (e) how, based on the other elements, a solution was found or lessons were drawn. Thus, in the first step, we performed a narrative analysis based on these questions to distil a limited number of narratives. Coders also noted important quotes, wrote brief summaries and included additional relevant observations.
Rather than trying to single out 'complete narratives' in each article, we took an approach of aggregation. Hence, the aim was to establish the most prominent narratives about Sweden and COVID-19 that could be derived from the entire textual corpus in each of the three contexts, and to investigate how they drew on established narratives about Sweden. Due to limitations of space, the examples in the empirical analysis and the references are, with a few exceptions, just indicators of broader tendencies.
We then looked more closely at the narratives and subplots identified in each media system, and found that each of them adhered to different ideal type logics of Othering outlined in the above model: (a) reproducing well-established self/Other distinctions emotionally; (b) tinkering with established identity constructions more strategically to make a point in domestic or international political debates; (c) reproducing self/Other distinctions more or less inadvertently in purportedly neutral analysis; or (d) blurring or problematising the logic of identity construction advanced in the established narratives, thereby also verifying its continued relevance.

The US media: strategic Othering
The US narratives about Sweden and COVID-19 drew on institutionalised narratives about Sweden as both a model nation and a deterrent example, and were quite exaggerated and polarised with little evidence of moderation or self-reflexivity. However, the distribution of supporters and critics was rather unexpected. The US Democrats and political liberals, who have traditionally supported Sweden's progressive politics, now depicted the country as a deterrent example in the context of COVID-19. Meanwhile, the traditional Republican and conservative critics of Sweden as a socialist dystopia now celebrated it as a model nation. These findings suggest that Sweden did not carry much inherent meaning in the US identification processes, but was instead used quite instrumentally as a device in political battles.
It is notable how strategically narratives in the liberal US media exploited the content of established narratives about Sweden -often featuring the country's admired characteristics -to point out how Sweden should be regarded as a deterrent example in the context of COVID-19. Part of this narrative revolved around Sweden being technologically and scientifically advanced, but arguably too scientific in 'conducting an unorthodox open-air experiment' on its own population (e.g. Goodman, 2020). The narrative also projected Sweden as too 'laissez-faire' (e.g. Goodman, 2020;Leonhardt, 2020), much 'laxer' in its approach (e.g. Lyons, 2020), and defying 'conventional wisdom' (e.g. Erdbrink and Anderson, 2020) as well as 'all science and logic' (e.g. Rainey and Feldman, 2020). Sweden was described as 'gambling' with the lives of its citizens (e.g. New York Times Editorial Board, 2020), with 'disastrous effects' (e.g. Weiland et al., 2020). These traits, moreover, were derived from Sweden's 'arrogance' and 'exceptionalism', which were pinpointed as defining features of the self-perceived 'regional hegemon' (e.g. Erdbrink, 2020). The narrative rejected the idea of Sweden as an example to follow, as the country's strategy was neither saving lives nor helping to protect the economy (e.g. Cowles, 2020). The narrative contained a temporal component, which described how Sweden had transformed from a respected, scientifically oriented nation that other countries liked to follow to something completely different: 'Swedes appear bewildered by the sudden scrutiny they are under. [. . .] The response has been cloaked in a self-image of a country that believes it behaves calmly while others are impetuous' (Milne, 2020). Indeed, the narrative advanced the idea that Sweden had 'become a bit of a regional pariah' (e.g. AP, 2020).
In the more conservative role model narrative, by contrast, Sweden was described as thoughtful and rational as it was 'concerned all along with the accuracy of early assumptions' (Bogan, 2020). A causal connection was made in the narrative that lockdowns are 'authoritarian' and do not save lives. On the contrary, they merely postpone unavoidable second and third waves, and restrain individual freedom (e.g. Sternberg, 2020). The role model narrative thus portrayed Sweden as an exemplary Other, characterising it as 'free' and 'democratic', and as appealing to individual responsibility (e.g. Clowes and Rossi, 2020). The lesson was that with its 'rationality', Sweden had managed to combat the pandemic 'and save its economy simultaneously' (Clowes and Rossi, 2020). The US conservatives emphasised the premature and unfair nature of the judgements of the (implicitly liberal) news media, which was 'maligning' Sweden (e.g. Ziegler, 2020).
In conclusion, US narratives about Sweden and COVID-19 drew on previously established narratives but were not bound to a particular form. Instead, they referenced and reproduced established narratives to fit with and reinforce the storytelling in which they wished to engage. Where sympathies dramatically switch, and former 'enemies' become 'friends', and vice versa, along neatly defined ideological lines, there is proof of strategic Othering. While Sweden figured as both deterrent and exemplary Other in US narratives, they all had this strategic Othering logic in common. Strategic Othering was arguably facilitated by the fact that the narrators were situated at vast geographical and cultural distance from the narrated, less entangled by historical bonds or shared identity and less bound by the content and form of previous narratives about Sweden.

The German media: analytic Othering
German news outlets did not seem particularly preoccupied with their own country and its identity when discussing Sweden and its strategy for combating COVID-19. There were exceptions, such as when Germany was juxtaposed with Sweden and depicted as 'prudent' rather than 'reckless' and sloppy (Rößner, 2020). However, Sweden predominantly served as a point of reference, from which to decide Germany's strategy in an analytical fashion, and ostensibly separate from issues related to identity.
Nonetheless, Sweden still emerged as an Other in German narratives. The most prominent narrative, embraced in most news articles and strengthened over time, was that Sweden had become an elapsed model nation, and the antithesis of the rest of Europe's (and the world's) superior COVID-19 strategies. Contrary to the established depiction of Sweden as a progressive, humane and safe nation, this narrative stressed extreme mortality rates (e.g. BILD, 2020a) and ethically disastrous behaviour, as some groups in society were 'sacrificed' (e.g. Hummel, 2020). In this story, Sweden had decided to pursue an outlier path like so many times before. With a pandemic raging, however, things did not turn out well. In terms of temporality, it was noted that a usually reasonable Sweden was now experiencing the same death toll as Italy and the United States, and how it was unable to accept that it was no longer a model nation -in fact, the narrative refers to Swedes as living in 'a parallel universe' (e.g. Trenkamp, 2020).
German media narratives drew on established narratives about Sweden, and the country's specific characteristics as 'vain' and 'proud', as allegedly incarnated by Sweden's former state epidemiologist Johan Giesecke (Strittmatter, 2020b). It was also characterised as 'incomprehensibly negligent' (e.g. Von Bayern, 2020) and 'headstrong'/'idiosyncratic' (e.g. Gasser, 2020). These traits were narrated as having generated both the country's peculiar approach to the pandemic and its refusal to reverse its strategy, while having led to one of the highest per capita national death tolls in the world (e.g. Strittmatter, 2020b). At the same time, German narratives also reflected on how Sweden's fall from grace was being experienced in the country, and how unusually 'painful' this must be for Swedes. They concluded that Sweden was paying a high price for the international criticism, since it targeted key components of the country's self-image: science, the welfare state and democratic values. Sweden used to be at the forefront of progressive thinking and science, but now found itself regarded as a 'pariah state' (e.g. Strittmatter, 2020a).
At first glance, the attributes discussed above might well appear to land Germany in the strategic Othering or emotional Othering parts of the matrix. However, a strong indication that German newspapers adhered to analytic Othering is that the narrators seemed to adjust their storytelling about Sweden with the passage of time and in the light of new information. During the first few months, Sweden was narrated mainly in a critical way, but from 15 July onwards, when Germans were allowed to travel to Sweden, German narratives became more open to the possibility that Sweden might have made the right choice after all when implementing its 'special strategy' (sonderweg). Supposedly in contrast to Germany, Sweden was characterised as 'free' in the sense of avoiding lockdowns and entrusting individuals to take responsibility (e.g. BILD, 2020b). German newspapers also narrated the Swedish COVID-19 strategy as rooted in a culture of high trust and mutual confidence, as well as shared responsibility between state and population (e.g. Raasch, 2020). At the same time, however, critical voices were being raised about how too much responsibility was being placed on individual citizens (e.g. Dambeck and Pieper, 2020).
In terms of emplotment, this positive narrative clarified that Sweden's trustful and free characteristics produced a better situation for the Swedish economy compared to the rest of the world (e.g. Augstein, 2020). Sweden was storied as 'brave' and 'sensible', but also as 'unfairly accused' and 'alienated' internationally because of its 'outlier' approach, and as proving that there was an alternative to lockdowns (e.g. Bockenheimer, 2020). German narratives became more uniformly critical again during the second wave, however, particularly from the end of September 2020.
In sum, German newspapers' Othering of Sweden was primarily that of cool-headed observation, careful analysis and gradual adjustment. Assessments appeared to be less influenced by the narrators' political affiliations than in the US media. German media narratives were not openly preoccupied with the question of German identity, but still drew on and referenced the content and form of established narratives about Sweden. The willingness to adjust narratives self-reflexively over time, depending on 'latest trends' from Sweden, is proof of analytical Othering.

The Nordic media: emotional and nuanced Othering
While Sweden did not emerge as a significant Other in the American and German media, Nordic media narratives positioned Sweden as more significant. While the multitude of narratives pointing in different directions makes the media debate difficult to summarise, the narrators were clearly emotionally preoccupied with Sweden and its strategy. There was a tendency, mainly in the Finnish and Norwegian media outlets, to condemn Sweden for its allegedly wrongheaded strategy. Prominent narratives revolved around Sweden's transformation from role model to nightmare, and the neighbouring countries' sense of disappointment. Some Nordic media narratives also expressed concern in a more nuanced way, however, attempting to blur or problematise self/Other distinctions.
To begin with, the Nordic media often depicted Sweden as 'irresponsible', completely turning its back on 'science' and 'reason', playing a 'hazardous game' and 'conducting an experiment' (e.g. Simin, 2020), and letting down its elders (e.g. Nilsson, 2020). It was portrayed as too arrogant and superior in its self-image to acknowledge any of this (e.g. Küchen, 2020). The Swedish authorities were described as stubbornly sticking to their strategy, even though the rest of the world -particularly the Nordic countries -had proved it wrong (e.g. Kløvedal Reich, 2020). Sweden was also characterised as 'ill-prepared' and 'repentant' that it had dismantled its supply of emergency protective equipment (e.g. Andelin and Gestrin-Hagner, 2020). This characterisation resonated with the narrative of Sweden as a rather naive 'neutral state', which had no sense of real danger, having avoided fighting wars for more than 200 years.
In the Nordic newspapers, the comparisons with Sweden often served to boost the national self. Articles in Aftenposten, for example, were fuelled by the idea that Norway was managing the pandemic better than Sweden in every conceivable way (e.g. Andreassen, 2020). Denmark, in turn, was characterised as 'wise', 'responsible' and taking a cautious path in times of great uncertainty (e.g. Fuglsang, 2020). Finland was portrayed as 'well-prepared' for a crisis and 'sensible' in its approach -often through juxtaposition with Sweden (e.g. Andelin and Gestrin-Hagner, 2020).
Like the German narratives, temporality came out in the sense that the Finnish newspaper narrated Sweden as having lost its previous international standing and selfappointed position as a humanitarian great power (e.g. Lillsunde, 2020). The failed role model narrative was moreover connected to the narrative on the Nordic 'family', in which Sweden has traditionally upheld the role of 'big brother', yielding expectations of responsibility and rule-following (e.g. Hårklau, 2020). This logic was disturbed when Sweden was seen as excessively liberal or behaving irresponsibly. As Heltberg (2020) noted: 'Once upon a time, there was a Sweden that we [Denmark] admired and begrudged from the perspective of a younger brother [lillebroderligt]. Bigger, better, wiser than us'.
There was also a narrative in the Danish news outlet that Denmark and Sweden had swapped roles. Sweden had had a reputation for strict rule-adherence whereas Denmark prioritised individual freedom, but this had now been turned upside down: 'Sweden of prohibitions' (förbuds-Sverige) was a recurring Danish phrase that no longer seemed valid (e.g. Klarskov, 2020). This is one example of how the Nordic Othering of Sweden 'struck back' at and required a reconsideration of the self and its identity. Danish identity construction as a free and somewhat unruly little brother was unsettled by Sweden's unusual behaviour, and the Swedish COVID-19 strategy was in a sense construed as 'stealing' the Danish 'enjoyment' of being the most 'freedom-loving' of the Nordic 'siblings', as well as the Nordic countries' 'enjoyment' as 'little brothers' to Sweden. In some articles, Sweden's 'failure' to handle the pandemic was portrayed almost as a personal failure that was jeopardising the close-knit relationships in the region (e.g. Munch, 2020).
Despite the criticism and disappointment, the Nordic newspapers were also filled with expressions of sympathy and understanding. The Danish and Finnish media narratives in particular suggested that despite -or because of -Swedish crisis mismanagement and the resulting problems, there was a need to show compassion (e.g. Al Fakir, 2020;Berg, 2020). There was no similar tendency in Norwegian newspaper's narratives, however, which generally had a harsher tone (e.g. Andreassen, 2020). Nuanced Othering was also visible in the defence of Sweden in view of the World Health Organisation's criticism of the country (e.g. Valbjørn Stavnsbjerg, 2020). The vilification of Sweden had allegedly caused border closures and was now threatening the Nordic community and brand (e.g. Dokk Holm, 2020;Munch, 2020). As Sweden was 'misunderstood', it became the subject of 'unfair judgements' as well as 'cold-blooded' and 'cynical' media reporting, putting further strain on Nordic relations (e.g. Didrichsen, 2020). Narratives in both the Danish and the Finnish media outlet urged that the Nordic community must be sustained after the crisis, and not allow the reputation of Norden as a unique zone of peace and welfare to fade. Thus, according to this narrative, denigrating Sweden served no purpose (e.g. Al Fakir, 2020). Other articles recommended that Sweden should follow its neighbours' lead, for instance, by copying Finland (e.g. Andelin and Gestrin-Hagner, 2020), or emulating and learning from Norway (e.g. Andreassen, 2020). Even the narratives that adopted a more nuanced Othering logic, whereby self/Other distinctions were blurred, were thus underpinned by emotional Othering and the fear that misreporting and misunderstandings were 'stealing the enjoyment' of having a Nordic community. In examples of both emotional and nuanced Othering, Sweden emerged as a significant Other, as demonstrated through a strong emotional involvement with the country and a preoccupation with questions about individual or shared identity, rather than narrow political manoeuvring or ostensibly neutral analytical discussions.
In sum, narratives in the Nordic newspapers referenced and reproduced established narratives about Sweden and adopted Othering logics that were both emotional and nuanced. This can arguably be explained by the fact that the countries have historically close relations; they are more significant to each other's identity construction but also share a collective sense of identity (Hansen, 2006: 39). People in the Nordic countries have become used to crossing borders without carrying passports, and many people travel between countries daily to work, and to visit relatives or holiday cottages. A recurring theme was thus how unnatural the border closures felt (e.g. Carsten Nielsen and Paarup Petersen, 2020).

Conclusion
In IR, narratives are commonly seen as constructing identity through practices of Othering, which tend to be understood as a rather singular and predominantly negative form of self/Other differentiation. This article demonstrates the need for a more diversified view of Othering. Far-right and Russian media narratives about Sweden ahead of and during the COVID-19 crisis were lopsided and unbalanced, systematically reporting on a limited number of key themes (Hellman, 2021;Wagnsson and Barzanje, 2021). Narratives about Sweden and COVID-19 in the United States, Germany and the Nordic states appeared similarly polarised at first sight. After further scrutiny, however, they began to look more varied and appeared to adhere to different Othering logics. We classified US narratives as strategic and serving a clear purpose in political debates, whereas German narratives were fuelled more by an analytic spirit. Narratives in the Nordic newspapers were both emotional and nuanced at the same time.
All the narratives on Sweden and COVID-19 drew on and referenced previously established narratives. Nonetheless, our findings show that narratives can be adjusted to fit current purposes. The primary example is the United States, where conservatives and liberals changed places in their views on Sweden. German narrators were more prone to take an observational stance, wishing to learn from the Swedish example rather than use it strategically. The Nordic newspapers also largely drew on pre-existing narratives of Sweden. They promoted narratives of harsh condemnation, on the one hand, and narratives characterised by compassion and understanding, on the other. Despite treating Sweden as a significant Other, many narrators in the Nordic newspapers thus refrained from negative storying that could have boosted the identity of their own countries, instead nuancing the discussion and stressing unity, thereby attempting to boost community and shared identity.
In theoretical terms, the article demonstrates that not all Othering revolves around an Other understood as significant, that negative and positive forms of Othering can exist in parallel and that more and less self-reflexive narratives can be intertwined. We hope that future research on Othering will continue to develop our ideal type model and further clarify the different Othering functions of narratives. For instance, while emotional Othering can boost self-identity, it serves to cultivate distance and may instigate conflict with Others. Analytic Othering, moreover, can yield insights for policymaking and inform difficult strategic decisions that require expert knowledge. Strategic Othering can inflict harm on the narrated, or fuel ideological debates either nationally or internationally. Nuanced Othering, finally, can help to bring about reconciliation and construct shared identity.
Future research could also investigate the consequences of different Othering logics. For example, does it matter if Sweden is narrated in a more strategic, analytic, emotional or nuanced way? Did the partly harsh narratives about Sweden during the pandemic have any persistent negative consequences for the country? Did they change the established narratives about Sweden and, if so, how? Finally, did the different Othering logics featuring Sweden affect autobiographical narratives in Sweden in any way?