Impetus

Sponsored trolls and fandom community conflicts
As the marketing and community operating strategies developing together with the global Internet, the influence of the Internet, especially from the few large corporations, has become omnipresent (Singer & Brooking 2019, Cao 2020). Some states, corporations and interest groups are identified by researchers and social media users to apply their impact on the public voice through a malicious and unjust marketing and operating method called “corporate, political, and special-interest sponsored trolling” (Zannettou, Caulfield, De Cristofaro, Sirivianos, et al 2019). The term “(state) sponsored trolling” is used by English-speaking scholars to describe trolling performed by government-sponsored or directed trolls as a way of political propaganda domestically or targeting a foreign country (Zannettou, Caulfield, De Cristofaro, Sirivianos, et al 2019). However, the most prominent and alarming manifestation of sponsored trolling in China is not political, but about teenage fandom culture and the popular entertainment industry (Cui 2020).

In China, the fandom culture is quickly growing to be a considerable economic and social power (Jiang 2018). Gathering around celebrities, idols or talent groups (which mostly target the young teenage girls as the audience), fans form active, close and hierarchical communities (Gu 2020). As branding strategies for the celebrity, fan groups, supported by the celebrity’s stakeholders, bring positive influence through celebrity charity and fan activism (Jiang 2018), but they also abuse the power of their group to perform data fraud, trolling and cyberbullying (Wang 2020). As to be articulated in the following section, such trolling is not only performed by professional trolls under the direct instruction of stake-holding commercial entities, but also by individuals who are misled or recruited by the employed trolls.

Such an organised trolling campaign is influential on social media platforms globally. It is believed by researchers of global cyberspace governance that these trolling wars are not only planned by corporations or political entities, but also a rising risky factor of online communication, and even has unpredictable political outcomes (Shen 2020). In the US, online wars on political topics are believed to be planned by foreign entities to manipulate the presidential election (Singer & Brooking 2019). In China, from the Cherry keyboard incident (Chen 2019) to the Xiao Zhan incident (Wang 2020), directed trolling war has been a concerning feature of both Chinese social media and global social media platforms related to China. More importantly, from an individual’s perspective, these incidents are aggressive and hateful; they bring devastating harm to their victims’ mental health and personal safety (Gorman 2019), and they interfere the value formation and the sense of cyber citizenship of young users (Wang 2020).

Cases on sponsored trolling war were studied during the course of the project, to understand the detailed mechanism and collect materials for project development. Cases under close inspection are: “Russia-Iran sponsored trolls” (Singer & Brooking 2019, Zannettou, Caulfield, De Cristofaro, Sirivianos, et al 2019), “Cherry keyboard incident” (Chen 2019), Xiao Zhan incident (Gu 2020, Wang 2020), “accelerationist” community, and “Chinese-Thai online flame war”. As the instance from the Chinese fandom community, the case Xiao Zhan incident has been the focus of the case study. Also, three semi-structured interviews with people who have experience with trolls in fan communities were conducted in October 2020, in order to understand the first-hand individual perspective.

A typical Chinese fan community of an idol is led by four strings of power: (a) fan leaders sponsored or employed by the idol’s agency and other related companies (such as a film production company marketing the film the idol stars), (b) voluntary fans (regular fans as well as productive and/or influential fans) who are genuinely attracted by the talent or the community, (c) trolls sponsored or employed by the idol’s competing capital, (d) voluntary trolls sometimes from the competing community (such as fans of another idol) (Wu 2020, Chen 2020, Wang 2020, Shitaoda 2020). Usually, employed fan leaders and trolls conceal the fact that they have a connection with the idol’s officials, and sometimes so do the voluntary trolls (Cui 2020). The communities exist on social media platforms such as Weibo (Twitter’s Chinese equivalent), and QQ chat groups, under the name of “fan club” (Cao 2020, Cui 2020, Li 2020).

Within the community, multiple functional subgroups are initiated and managed by employed and/or genuine fans, in charge of image and video production, voting, organising campaigns, public image management (anti-hater), and “comment control” (Cao 2020, Chen 2020, Li 2020). In these groups, most of the members are regular fans, who are organised by the fan leaders to devote their labour for free (Wu 2020, Chen 2020, Gu 2020). The majority of trolls are not paid by their director, but to some extent believe what they do is correct or enjoyable (Wu 2020, Gu 2020). In the voting, public image management, and comment controlling branches, it is common for members to make use of the censorship rules of the Chinese government and weighting system of the platform, to perform data fraud (e.g., applying for many accounts and voting repetitively), to uplift their influence and to muzzle unpleasing voices (Wang 2020). Some researchers and influential social media users believe that the platforms acquiesce in or encourage such operation (Cui 2020, Shitaoda 2020).

Theoretical explanation: a social psychological focus
There are researchers from sociology, social psychology, communication studies, and global cyberspace governance trying to uncover the mechanism of sponsored trolling. Theories that are used to study online fandom wars includes decategorisation, in-group favouritism and out-group prejudice, information cocoon room, group polarisation, the spiral of silence (Cao 2020, Cui 2020, Li 2020). For example, according to Singer and Brooking (2019), a foreign power’s influence on social media in the United States can be summarised as “4Ds”: dismiss the critic, distort the facts, distract from the main issue, and dismay the audience.

According to accumulated psychological studies on social categorisation, which is commonly used to explain prejudice, identifying others as group members is a spontaneous reaction. Once an individual has identified other's social group, their ability to recognise and memorising information about the other will be altered by the stereotypes or norms they have.

More specifically, the "Xiao Zhan incident" can be an example of how social categorisation framework views people's idea and identity shifting during a conflict:

"Xiao Zhan incident" is an online conflict about whether the Chinese celebrity Xiao Zhan is guilty of an attack against "Danmei" fanfiction community. There are two major groups against each other in the conflict: the pro-Xiao group, which consist of mainly Xiao Zhan's fans, and, anti-Xiao group, whose members are mostly niche fanfiction creators. Interestingly there is little space for a neutral standpoint, but this can be explained by the theory later.

Before formally join the conflict by posting public messages, a social media user may belong to any one of the following positions: Xiao's supporter, Xiao's sympathiser, neutral (pro, against, or careless of both sides), fanfiction creator's sympathiser, and fanfiction creator's supporter. Assume that there is a post that says "I don't think Xiao's mistake is that serious, although it's not something be proud of for him." For anti-Xiao members, the post shows a clear pro-Xiao position (because it does not blame Xiao hard) and is immediately connected with the stereotypes of Xiao's supporters: "the person must be a reasonless, fanatical and spoiled middle school girl".

A more advanced tactic, however, is to make use of the categorisation process. It was very common during the Xiao Zhan incident that each side sends impostures group member to their opposing group. For instance, a pro-Xiao member might imposture themself as an anti-Xiao member, and posting bizarrely extreme anti-Xiao speeches and expose themself to the general public, to create a negative public image of the anti-Xiao group on purpose.