Keywords

This chapteranalyzes the discursive construction of nationality and ethnicity in the context of the Overwatch World Cup 2019 and especially among the discussions of the world cup’s spectators on the live-streaming platform Twitch. Drawing on the positioning theory and the concept of banal nationalism, our study demonstrates how esports fans are active negotiators and co-creators of the esports discourse. The analysis illustrates what kind of role nationality and ethnicity take in this environment, in other words what they come to mean for those participating in the discourses defining them.

Nationalism and Esports

In this study we follow Billig’s (1995) conceptualization of nationalism, which forefronts the banal forms of nationalism alongside so-called hot or heated nationalism. Billig argued that the way nationalism is traditionally understood in research literature is by references to the extraordinary—for example to extreme right-wing movements or the creation processes of new nation states. As a counterpoint to these approaches, Billig introduced the concept of banal nationalism, which refers to the many ways in which nationalism is made into being and upheld through repetitive everyday phenomena. This includes the way nationalism is normalized, and embedded into our everyday life through media, symbols, and texts. In short, it is socially constructed in everyday social interaction. While banal nationalism can appear mundane on the surface, it can become heated or hot nationalism fast (Billig 1995). We furthermore align with Anderson’s (2006 [1983]) conceptualization of nations as imagined communities, socially constructed sources of a sense of ‘us’, and something that keeps on being adapted and transformed by its stakeholders through communication.

Banal nationalism is particularly visible in sports and sports coverage, where national symbols such as flags are constantly displayed in order to enhance and normalize the sense of national belonging (Billig 1995). The relationship between nationalism and sports is a well-studied phenomenon. Several studies have shown how sports has been used as a tool in nation-building and evoking the nationalist sentiment (Bowes and Bairner 2019; Crawford 2004; Hong 2013, King 2006).

In the context of esports and competitive gaming, nationalism and its displays have been less studied, nor has competitive gaming always been understood through the lens of sports. The concept of esports itself was only introduced in around 1999 (Wagner 2006) and debates concerning whether esports is truly a sport continue to this day (cf. Hallmann and Giel 2018; Taylor 2012; Witkowski 2012). Some of the competitive gaming communities also resist the term ‘esports’ as it is seen to commercialize and ‘tame’ the activity of competitive gaming (Ferrari 2013). However, some notable esport events like the World Cyber Games, launched in 2000, have framed themselves similarly to traditional sport competitions, and thus aimed to evoke a sense of national belonging from the viewers as well as the players (Taylor 2012; Szablewicz 2016). In more recent years, the first Overwatch World Cup, held in 2016, utilized a similar formula. Turtiainen et al. (2018) note that nationalism was a particularly strong element in the construction of the tournament broadcast. They argue that the event was sportified in many of its aspects, meaning that it drew from the representation of traditional sports (cf. Heere 2018; Jonasson and Thiborg 2010; Thiborg 2011). This was visible for instance in the display of national symbols and the way the event was narrated by the casters (Turtiainen et al. 2018). Similarly, Szablewicz (2016) highlights how in China esports operates as a state mandated spectacle whose primary function is to display nationalism and ideology.

Nationalism in esports has also been studied from the viewpoint of individual actors, such as fans and spectators. For example, Välisalo and Ruotsalainen (2019) point out the importance of nationality of favorite Overwatch esport players for the fans. Zhu (2018, 130) draws attention to the way masculinity, nationality, and ethnicity intersect in esports, paying particular attention to how ‘Asian’ players are constructed as both feminine and physically inferior to ‘Western’ players in discussion boards. Given these previous studies, it is clear that nationalist sentiment can be present in esports, but more studies are needed on the topic in order to understand the ways esports operates to reproduce the mechanisms of banal nationalism. In this study, we address this question by examining the way nationalism and its intersections with ethnicity are discursively constructed by esports viewers in the context of a large, international tournament.

Positioning Theory

Concepts such as nationality or ethnicity should not be understood as something abstract located within individuals’ minds, but rather as being progressively negotiated by actors who engage in discursive practices. Here, we focus our analysis on so-called positioning (Davies and Harré 1990), which may be understood as a conceptual tool for facilitating linguistically oriented analysis of social episodes. At its heart, positioning is a process wherein people negotiate ‘rights’ and ‘duties’. These are defined as “shorthand terms for clusters of moral (normative) presuppositions which people believe or are told or slip into and to which they are momentarily bound in what they say and do” (Harré et al. 2009, 9). Both nationality and ethnicity are examples of such ‘clusters’. We may position ourselves as well as others, and any so-called first-order positioning may be contested (second order positioning). Overall, positioning can be seen as being constructed as layers upon layers, where ongoing positioning may be related to positioning that occurred before, in a wholly other discursive practice (Harré and Van Langenhove 2010).

Positioning theory utilizes the concept of jointly produced storylines or unfolding narratives as something through which we make sense of who we and the others surrounding us are. It is through these stories that we learn to separate ourselves and others into social categories, and further allocate meanings to those categories. Social encounters may develop along multiple, interlinking storylines, and actions people take may carry multiple meanings and tie into multiple storylines simultaneously (Harré et al. 2009). For example, in the context of this chapter, the grand storylines could be those of (e)sports, media events, nationality, and ethnicity, all intersecting with one another.

Positioning theory has also been used to understand the issue of cultural stereotyping (Van Langenhove and Harré 1994). Instead of seeing cultural stereotypes as relatively stable cognitive models, a positioning viewpoint shifts the focus on stereotypes as social constructs. By making stereotypes public within a conversation, actors use them as rhetorical devices in the process of positioning. In the realm of discourse online, for example, Devlin (2016) illustrated how participants on Russian online political message boards used national or ethnic insults drawing on stereotypes as a way to construct the cultural other and to demarcate boundaries between ‘us’ and ‘them’.

In this study, we use the tenets of positioning theory as a kind of analytical lens that helps us understand the construction of banal nationalism in the context of esports. The premise of our study is that concepts such as nationality and ethnicity are ultimately mental constructs that need to be not only imagined, but also ‘modelled’ into being (Anderson, 2006 [1983]). Often, this is done via communication acts that appear mundane or ordinary (Billig 1995). These premises fit well with the idea of jointly produced storylines inherent to positioning theory. As a discursive approach, positioning theory instructs us to focus our attention to those instances, where the self, others, and relationships with other people are made visible (constructed) in actual social interaction.

Our study contributes to the emerging work on how nationality and ethnicity are made visible and relevant (discursively constructed) in online contexts, here specifically within the realm of esports. More specifically, our analysis focuses on the Overwatch World Cup 2019 and the live discussions of the world cup’s spectators on Twitch. We chose this particular event for our analysis since earlier research has demonstrated both the national elements in the broadcast (Turtiainen et al. 2018) as well as the importance of nationality for the fans of Overwatch esports (Välisalo and Ruotsalainen 2019). Our study seeks to answer the following research questions:

  • RQ1: How are nationality and ethnicity constructed in the production of the 2019 Overwatch World Cup?

  • RQ2: How do Twitch audience members discursively co-construct storyline(s) related to nationality and ethnicity?

Over the next paragraphs, we briefly introduce Twitch as a context of social interaction. Then, we present our method and describe our analytical approach, before moving on to the findings.

Twitch as a Platform

Twitch (www.twitch.tv) has been a forerunner in shaping the so-called live-streaming culture. As Johnson and Woodcock (2018) point out, Twitch has had a marked impact on the video game industry on the whole, enabling new marketing strategies as well as new forms of digital economy. In early 2021, Twitch had reached an average of nearly three million concurrent viewers, with close to ten million monthly streamers (https://twitchtracker.com/statistics). While most of the content of Twitch comes from individual players or streamers, the platform has also been a major actor in helping live-streaming esports scale up to the kind of media entertainment it is today (Taylor, 2018).

An integral part of the way Twitch spectators interact among themselves as well as with streamers is its chat function. The Twitch chat combines text with graphical icons called emotes (also: emoticons or emojis). Emotes have a textual label, but the way they normally are presented in the chat are as small images (Image 6.1). Some emotes are only visible in other than text format with a browser extension such as Better Twitch TV (BTTV).

Image 6.1
3 emojis with the facial expressions of men are labeled pixelbob, pogchamp, and poooound.

Examples of Twitch emotes in their graphical and textual format. (Source: https://www.twitchemotes.com)

Full size image

Many of the emotes used on Twitch, such as generic smileys, are similar to the ones used across other digital communication platforms. However, some of them are either specific to Twitch, or have originated there, and can be considered to be internetmemes. Internet memes are here understood as units or building blocks of culture (Shifman 2013), which are collectively negotiated, remixed, and recirculated in social media discourse (Milner 2016).

Gillespie (2010) argues that platforms such as Google and YouTube are the “the primary keepers of cultural discussions as it moves to the internet” (p. 348) which means they also face questions of responsibility on how these discussions evolve. Gillespie further asserts that by branding their service as ‘platform’, a term that suggests technical neutrality and progressive openness, these companies seek to diminish this responsibility. Following Gillespie’s framework of platform politics, Ask et al. (2019) note that initially Twitch was committed to this neutrality in their approach to the user-created content, but once they were acquired by Amazon.com in 2014, more strict guidelines started to appear. Amazon has shown signs of taking responsibility for what kind of discussion happens in twitch. For example, the popular ‘pogchamp’ emote was changed once the person it previously displayed became associated with opinions furthering violence (Kelly 2021).

Nevertheless, most of the responsibility of what kind of discussions take place in individual Twitch channels is still left in the hands of channel moderators. Since it is well-known that Twitch chat may include harassment such as sexism (Nakandala et al. 2017), and racism (Gray 2016), tournament broadcasters of the Overwatch World Cup 2019 included chat moderation. Twitch moderation typically includes both human moderators and automated algorithms. Algorithms for example time-out or ban users using words that have been black-listed and delete messages containing these words (Cai and Wohn 2019). Moderators and the moderating guidelines hold considerable power on what can and cannot be said in the chat and this also influences who gets to participate in the discussion and who is excluded (Grimmelmann 2015).

Overwatch World Cup 2019 as a Setting

Overwatch World Cup 2019 was the fourth of its kind. A total of 28 national teams participated in the tournament, which was organized between October 31 and November 2 as part of Blizzcon in Los Angeles, USA. Blizzcon is an annual convention of the game publisher Blizzard Entertainment that centers around the publisher’s games. While all World Cup matches were played in the Blizzcon or its vicinity, the preliminary qualifiers and the group stages were not played in front of a live audience. Throughout the tournament all the matches were broadcast live on Twitch. During the first days there were five overlapping streams (cf. channels) due to a number of games being played simultaneously. Due to the way spectatorship is split among multiple channels, it is impossible to accurately tell how many viewers the 2019 World Cup had. Still, one indication of the popularity of the tournament is that the semi-final that we analyze later on in this chapter gathered approximately 146,000 viewers on the official stream.Footnote 1

Data and Analysis

Studies drawing on positioning theory usually follow a social constructionist epistemology, and can be grouped under the broad umbrella of discourse analytical research interested in ‘locally’ negotiated meanings (Harré et al. 2009). Our study, which is interested in the discursive co-construction of nationality and ethnicity in the context of esports, adopts a similar analytical stance.

The data of our study consists of the public Overwatch World Cup 2019 Twitch broadcasts and the live chats connected to them. Both authors of this chapter as well as one research assistant followed the tournament live, and recorded matches and their chats as they were being played. In the end, we had 11 recordings totaling 500 minutes of video footage. We also gathered the chat logs of each recording in text format with the help of an automated tool Chatty (https://chatty.github.io/), which is freely available online.

As a first step of the analysis we engaged in a round of so-called data reduction (Guest et al. 2012). After deliberation we chose to focus on two matches. First, we chose the USA versus South Korea semifinals match for the reason that it had the highest number of viewers in the entire tournament, and featured a team from the country where the tournament was organized (USA). Another reason for including this match was that both USA and South Korea have a long history or rivalry in the context of the Overwatch World Cup. Second, we chose the Finland versus The Netherlands group stage match. Since most of the tournament games came from the group stage, this match represented a ‘standard’ match. It was played between two teams that did not rank high in the overall predictions, had fewer viewers, and therefore represented a less ‘heated’ match-up in the tournament. As an added benefit, the authors could understand both countries’ languages and cultural context well enough to catch up on the more subtle meanings possibly hidden in the chat.

The USA versus South Korea match recording was 127 minutes long. The log file of the chat for this match was 206,312 words long. The Finland versus The Netherlands match recording was 48 minutes long. The log file was 30,794 words long.

The analysis was conducted by both authors. We began the analysis by watching and re-watching the matches, as well as closely inspecting the log files. During this initial critical reading, we engaged in continuous discussions on the types of positioning we could identify in the data. We took a multimodal approach (Kress 2010) to our analysis. In the case of Twitch, there are several modes, such as writing, images, speech, and moving image, that together act as the building blocks of the social event. For the purpose of our analysis, the way different semiotic modalities interact or are combined is of importance. For example, we concentrated on how text and images join together to create new meanings. This kind of interaction is called inter-modality, or intersemiosis (O’Halloran 2011).

As Gee (2010, 117) argues, no discourse analysis is ever based on all the features present in a text, but rather aims to focus on the aspects of data that are relevant in the context and for the given task. Of course, the choice of what to include as relevant is ultimately a matter of theoretically informed qualitative interpretation. Due to the vast scope of the data, it was not feasible to ‘code’ every single line of chat or emote posted by the tournament viewers. Instead, based on our initial reading, we focused on those passages that included discursive positioning of the self or the other, and where this positioning was linked to nationality or ethnicity. After a further round of critical reading and discussions, we settled on four themes that pervaded the data and through which we can explore the discursive construction of nationality and ethnicity in the context of esport viewership. Over the next section we will explore these themes further. First, we will show how the visual and symbolic environment of the tournament sets the stage for the discursive construction of nationality. Second, we discuss the use of ethnically loaded memes as focal points, around which positioning occurs. Third, we explore how ‘the other’ and, therefore, the self are explicitly constructed. And fourth, we show how national languages may be used as a tool for in-group positioning.

Findings

Visual and Symbolic Positioning of Nationality

In line with an earlier study of the Overwatch World Cup 2016 (Turtiainen et al. 2018), our analysis shows that nationality was constructed through a variety of means during the tournament. These include the use of national flags in the visualization of the matches, using standardized jerseys with flags, as well as narrative means, such as the way the teams or players are presented in video highlights. The broadcast, especially after moving from online-only matches to those played in a physical setting, used similar tropes as traditional sports in setting the stage. An example of this comes from the beginning of the match between the USA and South Korea that was played in front of a live audience. After the Korean players were already seated, team USA entered the stage with the audience waving US flags and chanting “USA, USA”. Simultaneously, the chat reacted to the scene, as if echoing the live audience’s chanting. On top of loud, pompous music, the caster was shouting: “Put your hands together for … the Unites States of America!”. Finally, the US team was seen entering the arena with their captain carrying a large flag over his shoulders (Image 6.2).

Image 6.2
A screenshot of U S A's team members who enter the stage. Their captain carries a large flag over his shoulders. To the right of the screen is the list of comments.

Team USA entering the arena

Full size image

During tournament gameplay, nationality was continuously being enforced by presenting national flags on the screen both beside the team names as well as underneath the battle tag (nickname) of individual players (Image 6.3). In addition, in-game characters of playing teams were dressed in special ‘skins’ (cf. player uniform), allowing the whole team to be uniformly colored. This kind of extra layer of positioning by the tournament broadcasters is made even more relevant when one imagines the event without them—for example, most players’ battle tags are not easily retractable to their nationality, and the in-game characters are also otherwise identical on both sides. Without such cues, a random viewer entering the match would certainly not be able to make out who is playing whom.

Image 6.3
A screenshot of a game screen lists the Netherlands team on the top left and Finland team on the top right.

Visual cues constructing nationality during the gameplay broadcast

Full size image

Throughout the tournament, spectators also made use of the possibility of including national flags into the Twitch chat. Combined with specifically highlighting nationality in messages such as “go USA from FRANCE”, this allowed spectators to reflexively position themselves while simultaneously positioning others. Put together, this type of banal nationalism (Billig 1995) contributes to the jointly produced storyline of nations battling each other.

While the broadcast clearly utilized national tropes and symbolic means similar to other established sports, most of the audience involvement was not as straightforward to decipher. Instead of a uniform, non-contested storyline, the audience could be seen as fluctuating between affirmative and ironic or counter discourses concerning nationality and related concepts such as ethnicity. Over the next section, we will explore a key example of such a discourse—the construction of ‘Kkona’.

The Ambivalence of Ethnically Loaded Memes

In addition to nationality, also ethnicity was a relevant theme in the data. An interesting example of multimodal discursive construction of ethnicity comes from the way the meme known as ‘Kkona’ was used by the spectators (see examples below for the graphical representation of Kkona). ‘Kkona’ was a recurrent message/emote in the chat during the USA versus South Korea match, featuring over 4000 times in the chat log. The origin of the emote is a picture of a US American long-time streamer whose on-screen nickname is Kkona. The emote itself was removed from Twitch in 2018, but its use in textual format continued, and at the time of the tournament, the emote was included in the browser extension Better Twitch TV. Kkona is generally understood to refer to a stereotypical American “hillbilly” or “redneck” (Urbandictionary.com 2020; Levvvel.com 2020). Throughout the chat, Kkona was used both for reflective first-person positioning of the self, and interactive positioning of the other. The positioning of the self is visible in chat lines such as:

  • “Let’s go brothers, In God we trust ”

  • “ MAKING TRUMP PROUD”

  • “ 7 THATS MY BOYS ”

The use of ‘brothers’ and ‘my boys’ suggests that the speakers align themselves in the same in-group as the US players, and use Kkona as a tool of self-positioning. In the first quote the use of the sentence “in God we trust”, the official motto of the USA, further strengthens this. The second quote demonstrates the use of Kkona in relation to Donald Trump, who was the president of the USA at the time of the tournament. In the chat, Kkona was also often used with proverbs and punch lines widely connected to Trump (such as “Make America great again”).

The third quote above has the number ‘7’ following the text Kkona, which adds a possible racialized dimension to the use of the meme. Number seven, which can be seen as a graphical representation of an arm doing a military salute, can also be read as referring to a banned emote, ‘Trihard7’ (Image 6.4). This emote became notorious after it was spammed in the Twitch chat during a Hearthstone competition every time an African-American player Terrence “TerrenceM” Miller was shown in the camera (see, e.g., Fletcher 2020). Utilizing the number seven with the emoji Kkona can be read as a form of defiance for the ban of Trihard7. Unlike Trihard7 in the case of TerrenceM, Kkona in our data was not spammed for any particular player but rather in connection with anything related to the USA. In effect, Kkona constructs and refers to a particular kind of (white) American identity, thus being both ethnically and nationally loaded.

Image 6.4
A laughing face emoji of a man with a safety helmet, along with the number 7 mentioned on it.

Trihard7 emoji

Full size image

Throughout the chat, Kkona was also used in a sense that can be interpreted as ironic. For example, when the live audience was shown in the stream, and some members of the audience were seen holding flags of South Korea, the chat erupted with comments such as:

  • “why are all these Americans waving Korean flags”

  • “ where’s your sense of patriotism?”

  • “those Korean fans look suspiciously American ”

There appears to be a sense of ridiculing the idea of national belonging here and perhaps even the way the event itself is constructed to evoke national belonging. The third quote includes the ‘Kappa’ emote. The Kappaemote generally stands for indication of irony or sarcasm (Dictionary.com 2020). Its inclusion here suggests that while there clearly is nation-building at work among spectators, there are also levels of irony or sarcasm related to this.

The ambivalent meaning of Kkona was further visible in the way it was utilized as a tool for building counter discourse to that offered by the official stream. An example of this comes from the US versus South Korea match, where between games the Twitch stream displayed advertisements. Among the advertisements of technology, game, and phone companies commonly seen in similar broadcasts, there was also a recruitment video by the United States Air Force. The video states that today “planes are piloted from the other side of the world” while showing video footage of a soldier sitting in front of two screens in an army base, almost as if playing a video game. Overall, the video uses glorifying language in describing members of the US air force, including statements such as: “We are not just pilots and engineers, we are pioneers”, and “Join us and be the future”. While audience reactions to most ads in the chat data was rare, or simply signaled boredom, the comments to the recruitment video showed a different level of attention, including comments such as:

  • “SOME OIL NEEDS TO LIBERATED ”

  • “MAKE LOVE NOT WAR! THIS IS PROPAGANDA FOR USA”

  • “Killing children ”

  • “Join us and you too can bomb innocent civilians ”

  • “HELP US COMMIT WAR CRIMES ”

Here, Kkona continues to be associated with the USA, but takes on a much more condemning tone. While during the gameplay Kkona was often added to chants of “USA, USA” when team USA was performing well in the game, here it becomes connotated with practices of “liberating oil”, “bombing innocent civilians”, and “committing war crimes”. While it is clear that it is a particular kind of US Americans who are being positioned here, it is not clear who is doing the positioning—the use of Kkona in this context could be seen as either interactive or reflexive positioning. When looking at the way chat participants use Kkona throughout the match, multiple ways of using the emote are revealed: In some cases those who in other instances use Kkona in a positive way turn it around in this particular situation; in some cases those who in other instances cheer for USA (without Kkona) use Kkona here negatively, perhaps to signify a kind of ‘American’ they do not want to identify with; and some of the cases remain highly ambivalent, allowing multiple interpretations of the use of Kkona.

Positioning the Other

The context of the World Cup, where teams representing nationalities were in competition with each other, provided a clear framework for conceptualizing ‘the other’. This basic setup was also reflected and reinforced in the chat. For example, in the US versus South Korea match, messages positioning Koreans included such as:

  • Haksal looks cute

  • haksal is my waifu

  • KOREA OP OP OP

  • lol carpe is op

In the above excerpts, both ‘haksal’ and ‘carpe’ refer to the player tags of players in the Korean team. This kind of positioning of the Koreans is akin to the way Zhu’s (2018) description of how Asians are often perceived within ‘Western’ esportscommunities. According to Zu, Asians are othered through both feminization and roboticization (2018). This was visible in the chat through physical description of Asian players (“tiny Asians” “cute”) and framing them as ‘OP’ (Overpowered), a term used to describe powerful videogame characters, strategies, and so on. Another theme that was visible in this type of positioning was the so-called fetishization of Korean players and South Korea as the promised dreamland of esports (Taylor 2012). In the chat the skill of the Korean players was often the focal point of how they were positioned.

In addition to positioning Koreans as players, the chat included messages positioning their fans. Again, this kind of positioning often evoked certain existing stereotypes and concepts, such as referring to kpop (Korean pop-music) fandom or the concept of ‘weabu’ or ‘weeb’, which is often used to describe a Western person obsessed with Japanese culture or Asian popular culture at large:

  • “KOREAN WEEBS ”

  • “kpop fans”

  • “tiny asian fanboys in chat”

At the same time, the chat also positioned the American fans as a distinct group. This kind of positioning was, again, done mostly with the help of negatively loaded descriptions, such as the stereotype connecting US Americans with junk food and obesity:

  • “KR is so ba … wait let me take a breath KR is s … i need to take another one give me another burger ”

In the Finland versus The Netherlands match, there was an interesting case of positioning Finns as the ‘exotic’ other by both the casters and the chat participants. The casters began the segment by speaking about Finns and wondering what Finnish people do in their spare time: “I believe they fish. They fish and they complain about the cold”. They then moved on to discuss a particular Finnish player, ‘Taimou’, spending time in his home country, roaming the countryside, and going “Full on Kkona, the Finnish version”. The chat then picked up on the discussion, including messages such as:

  • “they ski and sit in saunas ”

  • “we drink and complain:)”

  • “The Finnish Summer is the best day of the year!”

  • “turpa kiinni ” [translation: shut up ]

  • “finnish are not vikings”

  • “Finland is its own thing, apparently they’re more closely related to Asian cultures”

This kind of discourse can be seen as including both interactive and reflexive positioning, which together outline the storyline of ‘Finnishness’. It resonates with the widely circulated myth of Finns being distinct from other (West) European national groups, as well as the myth of their taciturn and reserved communication style (for a critical review of the myth, see Olbertz-Siitonen and Siitonen 2015).

In-Group Positioning by Using National Languages in an International Chat

While the officialstreams were broadcasted in English, and the vast majority of the messages posted in the chat used this language, there were also occasional messages in other languages present. This was especially apparent in the Finland versus The Netherlands match, where both teams came from countries where English is not an official language. In this match, the chat included a number of messages such as “Min hart kan dit niet aan” (“My heart cannot take this”), and “hyvä pojat hyvä meno” (“good job boys, good going”). It is perhaps noteworthy to add here that moderators allowed the use of national languages in the chat. This is not to be taken for granted on channels that are aimed at an international audience, and where, for example, English is used as a lingua franca.

Using a playing team’s national language can be seen as a prime example of first-order positioning in this context. Even though the posters may be aware of the likelihood of the majority of the international audience not understanding what is being said, using the national language of one of the playing teams effectively positions them as a member of the same in-group as the players.

Some of this kind of first-order performative positioning (Harré and van Langenhove 2010) where spectators used languages of the playing teams utilized memes. Spectators evoking memes in playing teams’ national languages could be seen as engaging in a kind of double in-group positioning. Not only does one need to understand the language in question to participate in the construction of the meme, but one also has to be versed in a specific subculture. For example, in parts of the match where the opposing team was dominating, some apparently Finnish spectators posted “FIRST WE GIVE THEM SIIMA, THEN WE _PULL_ MATTO ALTA”. This combination of English and Finnish could be roughly translated as “First we cut them some slack, then we pull the rug from under them”. While the idiom “to pull the rug from under (someone)” can be used in both Finnish and English, this particular version of the meme deliberately combines both languages. It originates from the context of another popular game, Counter Strike: Global Offensive, and specifically the Finnish team ENCE.

Positioning theory argues that whenever someone positions themselves, they inevitably imply a positioning of ‘the other’ as well. At times, this parallel nature of positioning was even clearer than normally, a kind of challenge or gauntlet thrown down at the perceived adversary. For example, at the point in the match between Finland and The Netherlands where The Netherlands was leading 2–0 on a best of five match, the theme of colonization started appearing in the chat:

  • “Kolonisatie 66% compleet ”

  • “ Colonised ”

  • “G E K O L O N I S E E R D ”

Messages such as these utilize both the national language as well as the history of the nation in question to intentionally position participants into in- and out-groups. These messages are also connected to a meme playing with the colonial history of The Netherlands. We can see a similarity to how Kkona is used, where in-group members formulate an ambivalent message concerning their own group or position.

Overall, while not abundant in the data set, the use of national languages in an international (English language) chat was a consistent discursive tool that participants used to engage in intentional and deliberate positioning. Here, a parallel may be drawn to other contexts such as international business, where there is often a shared (third) language, but people may still opt to use their mother tongue in certain situations as a symbolic tool with which identification and community may be strategically constructed (e.g., Lauring 2008).

Conclusions

This study contributes to our understanding of how esports producers and viewers discursively construct nationality and ethnicity in the context of esports. From the visual presentation to the way the audience reacts, the whole production may be seen as following familiar tropes related to constructing nationality in the context of sports (Billig 1995; Bowes and Bairner 2019). However, the analysis also shows that this construction was not uni-dimensional, nor uncontested. Often, the contents of the chat could only be interpreted as being ironic and even producing a satiric counter discourse to that offered by the tournament organizers, similar to what has been witnessed in the field of contemporary politics and activism (Day 2011). For example, the pompousness often related to emphasizing nationality was ridiculed in many ways by the spectators. Also the case of ‘Kkona’, and the way the colonial history of The Netherlands was used by chat participants, illustrate how the same messages may be used both for reflexive and interactive positioning, and to simultaneously affirm and undermine existing storylines related to nationality and ethnicity. In several instances the audience could be interpreted as occupying a position of first and foremost belonging to player culture(s), and only secondarily belonging to a national or ethnic group. We argue that this kind of behavior is connected to a broader tradition of ‘trolling’ prevalent in-game cultures (Cook et al. 2018). Here, provocative messages can be seen as a kind of ‘bait’ or challenge to other participants. The ability to understand and use memes signifies the boundary of the in-group, with those who become offended or do not understand their use being left on the outside (Manivannan 2013). This also functions as a way of constructing a normative order amongst spectators which is furthermore strengthened by practices such as spamming the chat with memes, thus possibly hiding the individual messages which do not follow the established use of memes.

While the setting of the World Cup emphasizes nationality as a key category for grouping participants, our analysis shows how ethnicity may also become relevant. From the way the South Korean players and fans were characterized to the way Finnish players and Finns in general were discussed by the casters and the chat, our analysis shows that when positioning nationality, the question of ethnicity is never far behind (Bairner 2015; Rowe et al. 1998).

Finally, our analysis highlights how national languages were used as a tool for first-order self-positioning. In an international environment where English is used as lingua franca, the use of other languages carries a message in and of itself. As Brubaker (2013) argues, language may serve as a kind of central category in establishing societal inclusion and exclusion, inherently intertwined with ethnicity and nationhood. This finding works as a reminder of the importance of language for nation-building (Anderson 2006 [1983]).

Our analysis of language in use contributes to the literature on how online contexts such as Twitch allow people to locally negotiate new social practices, or what one could call digital literacy in its broad, action-oriented meaning (see, e.g., Barton and Hamilton 1998). Understanding, let alone participating in, the discourse surrounding the World Cup requires a certain level of digital and game literacy (cf. Milner 2016 on literacy required to participate in memetic discourse).

Our study also serves as a reminder of the kind of challenges that interpreting memes represent. Since memes are constantly evolving and being negotiated also from situation to situation, the interpretation relies heavily on contextual cues and knowledge. We also witnessed how memes that were originally constructed to operate in a visual mode may continue to exist in another modality (text), once they have become established enough.

As always, our study has its limitations. When large competitions such as the Overwatch World Cup are streamed to the public, there are always several unofficial streams that accompany the main one, often with multiple languages. It would have been interesting to follow the Korean, Finnish, and Dutch streams of the event in addition to the mainstream where the official language was English. This could have opened the door to other kinds of analyses, such as comparing the positioning done in the ‘national’ streams against the main channel. However, due to practical reasons such as language proficiency, and the way smaller streams are easily lost in the multitude of Twitch programming, we ended up focusing on the official streams only. Future studies could explore the way smaller, linguistically or nationally oriented streams may interact, or whether there are apparent differences in their discourse.