On the fly: How Japanese social media “watchers” improvise to counter problematic information

Abstract

Recent studies have explored how the purveyors of problematic information can mobilize online crowds by tapping into positive feelings like amusement, belonging, and optimism. Unfortunately, it is not clear how such participation and emotions can also help communities reduce the spread of problematic information. We address this gap by examining how a group of Japanese social media “watchers” monitored and countered YamatoQ, a pro-QAnon Japanese conspiracy group. We analyze interviews with 12 watchers to make empirical contributions about the how, where, and why of their activities. Participants describe a playful approach to confronting problematic information, whereby they leverage a sense of improvisation, humor, and community to meet the challenges of their work. These results extend our knowledge of volunteer efforts to address problematic information, and expand the range of perspectives represented in the literature regarding the participatory and emotional aspects of problematic information.

Introduction

On 7 April 2022, 12 people forcibly entered a Tokyo health clinic offering coronavirus vaccinations for children, and disrupted activities (Kyodo, 2022; The Asahi Shimbun, 2022). Subsequent investigations linked these individuals to YamatoQ (神真都Q), a Japanese QAnon-affiliated group that promoted narratives supporting Russian President Vladimir Putin, former US President Donald Trump, and anti-vaccination views (The Yomiuri Shimbun, 2022). The group has been active on several online platforms, and has multiple regional branches that provide an on-the-ground presence: the group has held anti-vaccination demonstrations in all 47 prefectures in Japan, with an estimated 6000 participants (The Yomiuri Shimbun, 2022). Such developments speak to how various forms of problematic information are flourishing in present social conditions. We use the term problematic information as an umbrella concept to encompass disinformation, which is intentionally disseminated to deceive, misinformation that is unintentionally inaccurate, and other forms of toxic content such as conspiracy theories (Jack, 2017). This framing also highlights the ambiguous boundaries between these categories. Where appropriate, we use the term conspiracy theory as a subtype of problematic information, as YamatoQ is commonly understood as a conspiracy movement.

Cases like YamatoQ are significant because they highlight how purveyors of problematic information mobilize audiences on multiple platforms to amplify their messages by tapping into positive emotions, such as humor (Arif et al., 2018), feeling smart (Marwick and Partin, 2022), and the desire to belong (Young, 2021). Given how such “bottom-up” approaches can weaponize online participation and these positive feelings, there is a pressing need to understand and counter such efforts.

In this article, we examine how a group of Japanese citizens grappled with some of these complex dynamics as volunteers intervening against YamatoQ. We ask : How do members of this group—who self-identify as “watchers”—collaborate with each other and understand their own work?

The term watcher (ウォッチャー), in the context of Japanese Internet culture, generally refers to those who closely follow claims of certain organizations or individuals, often focusing on problematic speech and content (Wakisaka, 2014). Our research question allows us to (1) better understand the role community-driven interventions can play in fostering a healthier information environment and (2) expand the kinds of communities represented in the literature. Such efforts have emerged across a range of media contexts and cultures: from the Bellingcat citizen journalist group’s open-source investigations (Higgins, 2021) to StopFake’s peer-to-peer counter-propaganda in Ukraine (Haigh et al., 2018). This digital volunteerism—or sometimes, vigilantism (Starbird et al., 2014)—cannot obviate the need for more structural interventions, but such work has been impactful in a variety of other contexts such as natural disaster (Starbird and Palen, 2011) and human rights investigation (Dubberley et al., 2020). To support such efforts and make them more impactful, we need to understand what these actors do and how they understand their activities. We investigate our questions by conducting semi-structured interviews with 12 watchers who collaboratively monitored and addressed problematic information propagated by YamatoQ. We also collected data from participants’ tweets and the group’s Discord to concretize what participants told us during the interviews.

Our article sheds light on the goals and underlying motivations of the watcher community. Their perspective—which is embedded in Japanese online subcultures, rather than in formal institutions—helps us advance our understanding on the processes of self-organization, how they build and maintain community ties, and the potential of emotional appeals as part of interventions to problematic information. Our findings advance empirical and theoretical understanding of how informal volunteerism emerging out of grassroots community-settings can complement the existing solutions against the spread of problematic information.

Literature review

Informal volunteer work against problematic information

As people become more cognizant of the pervasive nature of problematic information, they are motivated to take action against it. Some of these concerned individuals engage in formal volunteer work with established organizations such as Snopes, or the Digital Verification Corps at Amnesty International, where they might receive training, clearly defined responsibilities, and acknowledgment through awards or certificates (Dubberley et al., 2020). Others contribute more anonymously through decentralized efforts, where informal volunteer work can offer a lower barrier to entry and potentially greater personalization and autonomy (Whittaker et al., 2015). Studying such emergent efforts can provide insights into the diverse and creative ways people respond to problematic information.

A growing body of research in communication and social computing has delved into emerging efforts to combat problematic information beyond the confines of formal institutions. This body of research has examined a spectrum of such initiatives, including the collaborative efforts of online citizen “sleuthers” following the 2021 US Capitol riot (Reese and Chen, 2022), and the mobilization of the Korean music group BTS’ global fandom to curtail the spread of problematic content in their community (Lee et al., 2022). These inquiries have highlighted the advantages of such participatory efforts, showcasing their ability to utilize social connections, common interests, and volunteerism to enable communities to rapidly identify unverified information, source evidence, and even employ “carnivalesque responses” to debunk falsehoods (Haigh et al., 2018).

Simultaneously, this research highlights several challenges of informal volunteerism. It can be exploited by malicious actors spreading low-quality information (Cox and Maiberg, 2023), diverse participant backgrounds may compromise fact-checking (Kligler-Vilenchik, 2021), and it risks turning into digital vigilantism, where wrongdoers are targeted without due process (Starbird et al., 2014). These challenges, alongside the potential advantages of combating problematic information “from below,” highlight the importance of acquiring a nuanced understanding of digital volunteers’ experiences, perspectives, and practices in addressing problematic information. Unfortunately, there is limited research on how digital volunteers addressing problematic information understand their own work outside formal organizations, largely because they often remain anonymous and are difficult to reach (Reese and Chen, 2022). Our research addresses a critical gap by examining how a group in Japan tackles problematic information, offering insights into the role of informal participatory efforts in combating problematic information. Japan warrants attention as problematic information spreads not only through global platforms like Twitter (X) but also through regional platforms like the LINE messaging app, which has been largely overlooked in academic studies. Conspiracy beliefs have long been prevalent in Japan, driven by its insular media landscape (Alt, 2020).

Decoding YamatoQ: examining challenges and potential remedies

While scholarly exploration of grassroot efforts against problematic information remains relatively limited, more research has delved into how these phenomena thrive on the broad participation facilitated by social media platforms (Marwick, 2021). Within this perspective, online users assume a central role in generating, disseminating, and amplifying problematic information (Starbird et al., 2019). The case of YamatoQ is particularly noteworthy in this context, as it intersects with two interconnected challenges emphasized in the literature: emotional asymmetry and volatile virality (Ong and Cabañes, 2018). Although our interview protocol did not specifically address these challenges, they appeared to play a prominent role in participants’ experiences, prompting us to briefly describe them here and revisit them in the “Discussion” section to offer further insights and reflections.

Several scholars have identified a phenomenon we term emotional asymmetry—namely, the unequal emotional appeal of problematic information compared to evidence-based information. Traditional methods for promoting the latter—like fact-checking or media literacy programs—typically prioritize objective analysis, which may not address people’s emotional needs for validation, empowerment, and a sense of belonging (Wilner et al., 2023). In contrast, movements like YamatoQ deftly tap into a diverse array of emotions, including fear, self-superiority, hope, and a sense of belonging (Marwick and Partin, 2022; Young, 2021). Purveyors of problematic information also often rely on humor and insider jokes, creating “subcultural language communities” (Peeters et al., 2021: 2). This asymmetry can allow conspiracy groups to stimulate participation more effectively than efforts aimed at countering them. For movements like QAnon, which have extended their influence beyond the United States, a crucial element of this participation involves volunteerism (Starbird et al., 2019). In this context, individuals freely offer their time, skills, or resources to recruit others to their belief systems and attack skeptics, thereby fueling the movement’s expansion and influence.

These participatory dynamics can create “volatile virality,” a concept that describes the dynamic nature of problematic information within networked environments. Ong and Cabañes (2018) argue that problematic information can swiftly evolve, mutate, or migrate across various platforms owing to the actions of grassroots volunteers who not only amplify the original narratives but also remix and disseminate them in unpredictable ways, aiming to maximize engagement. In fact, groups like YamatoQ, which emerged as a mutation of QAnon by incorporating local interpretation of the movement, is a primary example of volatile virality. This volatility complicates efforts to combat problematic information, as traditional interventions may struggle to keep pace with the rapid and unpredictable nature of its spread, especially as it cross-pollinates between platforms (Wilson and Starbird, 2020).

These two challenges highlight the importance of leveraging diverse community resources to build resilience against harmful entities like YamatoQ. Addressing emotional asymmetry and volatile virality requires a multifaceted approach that recognizes the interconnectedness of online spaces, adapts to evolving trends, and highlights the crucial role of community bonds in resisting problematic information. Collaboration among institutions like social media platforms and fact-checking agencies is vital, but grassroots, volunteer-based initiatives also play a role. These initiatives not only recognize community vulnerabilities but also harness their strengths, effectively disrupting virality by utilizing local networks and resources to combat false narratives (Lee et al., 2022). In addition, they can address the emotional aspects of problematic information more authentically than formal institutions, which may face challenges, such as perceived authority and cultural insensitivity (Kiousis, 2001). To implement these strategies effectively, direct engagement between researchers and volunteers is necessary. Our study aims to facilitate this by providing an in-depth understanding of one such volunteer initiative.

Methods

Our findings are based on themes drawn from semi-structured interviews with “watchers,” who actively investigated YamatoQ. Participants were identified through purposive sampling, searching social media and analyzing follower-following relationships. This was followed by a snowball sampling approach, where initial participants referred the lead author to other members of the watcher community that places a strong emphasis on privacy and anonymity to safeguard themselves from potential reprisals. Our recruitment approach helped us establish trust and gain access to this highly protective group.

In total, we conducted 12 interviews. While it is difficult to precisely estimate the size of the watcher community due to its decentralized nature, participants indicated that the core group consisted of approximately 25 active members. With our interviews representing nearly half of this core group, we believe that our data are sufficient to support our claims regarding the motivations of the watchers and the meanings they give to their actions (Blee and Taylor, 2002). To help the reader better understand our sample, we provide participants’ demographic information in the appendix (Supplemental material). Nine interviews were conducted via video chats, while three were conducted asynchronously in writing, as those participants preferred not to use video chat. The synchronous interviews lasted between 60 and 90 minutes. All interviews were conducted in Japanese, then transcribed and translated into English.

We conducted a thematic analysis to examine the transcripts. Our analysis was both grounded, meaning bottom-up and inductive (Charmaz, 2006; Magoon, 1977), acknowledging that we were constructing one of many possible interpretations of the data. The research team conducted open coding on several transcripts, wrote memos, and met regularly to discuss observations. This helped us identify some initial patterns in the data and a codebook of 15 codes that we used to systematically analyze all transcripts. Finally, we took the coded excerpts and conducted affinity diagramming using the web-based tool Miro, refining our themes by examining relationships across codes. We then ordered these themes to write our findings.

Trustworthiness

We took steps to confer more rigor of our work. For example, we have carefully reflected on how our positionality as a cross-national, interdisciplinary team shaped how we collected and interpreted our data. Our prior research in problematic information, crisis informatics, and political communication shaped our understanding of online crowds and digital spaces, driving our interest in human-centered solutions to problematic information, particularly in non-Western contexts where approaches to conspiracy theories and argumentation may differ.

During data collection, we intentionally had the first author, a Japanese researcher, conduct all interviews. We acknowledge that this decision introduced trade-offs. For instance, her familiarity with the social context may have led us to overlook certain clarifying follow-up questions, such as assuming that we fully understood criticisms of the media raised during interviews. However, her position also allowed us to build trust with participants and avoid exoticism or orientalism. Our methodological approach further strengthened our analysis by incorporating multiple perspectives. For instance, for research results to be included, they had to be reviewed and approved by the entire team to minimize the risk of narrow interpretations.

In addition, we obtained permission to join the watchers’ private Discord server. We also collected a total of 3461 tweets, which included either of the two keywords, “YamatoQ” and “watcher,” from all of the participants’ social media accounts. In line with Shenton’s (2004) recommendations, these materials served as supplementary data to familiarize ourselves with the community and provide concrete examples to illustrate the themes emerging from our interviews. In keeping with the ethical considerations brought up by Reagle (2022), we did not directly quote messages from these data sets to respect participants’ strong wishes to remain anonymous.

Limitations

We recognize several limitations in this study. One key limitation is that we did not systematically analyze the digital traces of the group’s online activities. We made this decision due to technical constraints that made it difficult to construct a reliable data set (e.g. API [application programming interface] limitations with platforms where infiltration and monitoring occurred), and the challenge of analyzing large volumes of Japanese social media content across a cross-lingual team. As a result, our understanding of the group’s work is primarily based on participants’ perspectives rather than direct observation of their online interactions. In addition, three participants chose to communicate asynchronously in writing. While this approach offers advantages, such as including participants who may be hard to reach otherwise (Amri et al., 2021), sporadic responses sometimes disrupted the flow of conversation and limited our ability to ask timely follow-up questions. Finally, during the interviews, we learned that some participants had expanded their focus to other conspiracy theories, but our article does not address the broader scope of their activities.

Findings

We have organized our findings into three sections. First, we present how participants discussed their goals and values, providing insight into the motivations of the watcher community. Second, we report how participants described their efforts to respond to YamatoQ, highlighting the multi-channel nature of their work across various settings. Finally, we examine how participants addressed the emotional dynamics of their activities, emphasizing the value of a playful attitude—approaching problematic information with curiosity and creativity. Together, these sections shed light on how a community-driven effort tackles problematic information across diverse environments and how its members give meaning to their actions.

The watcher community’s goals

Our findings suggest that participants in the watcher community were not necessarily adversarial toward YamatoQ. Rather than engaging directly in argumentation, they emphasized a more empathetic approach to address YamatoQ. P1 rejected the notion of framing the watchers as an “Anti-YamatoQ” movement, explaining that positioning their work as a counter to YamatoQ would only strengthen the group’s beliefs: “I don’t really want to create a ‘YamatoQ vs. Anti-YamatoQ’ type of structure. I feel that online battles like that would only make [followers of YamatoQ] more stubborn.” P10 echoed this view, expressing empathy toward YamatoQ’s followers by stating, “It’s the conspiracy theories that are bad, not the people who are exposed to them.”

Alongside this non-confrontational stance, another key theme emerged: participants aimed to cultivate a more informed society. They focused on promoting individual media literacy while also building broader societal resilience to problematic information. As P11 explained,

We must make efforts to prevent ourselves from falling into conspiracy theories by improving our individual literacy and by creating an atmosphere in our society that does not tolerate the rise of conspiracy theories. I don’t believe that our society as a whole has been able to do that in response to YamatoQ, and I feel that further improvement is needed.

Here, P11 envisions an informed society where media literacy is strengthened, and the public is less tolerant of conspiracy theories. However, participants noted that a major challenge in achieving this is the lack of media coverage and, consequently, limited public concern about YamatoQ. Two participants explicitly described their activities as “grassroots journalism” to fill the gaps left by mainstream media, which they argued often overlooks YamatoQ—either to avoid amplifying its claims or due to a lack of seriousness toward the group. This, they believed, hinders public awareness. P12 emphasized that watchers helped bridge this information gap, creating space for discussions on how society should address issues like YamatoQ:

We are explaining things that are not reported by the media . . . [watchers] provide information and opinions to the general public about how we should deal with YamatoQ and how we should make sense of it.

In line with prior work (Phillips, 2018), participants described tensions within the journalistic approach, describing the delicate balance between debunking false claims and inadvertently giving YamatoQ a platform. P11 acknowledged concerns that informing people about YamatoQ might “unnecessarily increase the risk” of exposing them to the group’s ideologies. Similarly, P1 shared that he had been criticized by his followers that his activities were only exposing more people to harmful claims by YamatoQ, but he maintained “a strong belief that we should name them” because he believed that members of the public would “benefit more” if they are better informed about the group and its problematic claims.

Participants also grappled with the challenges of content moderation and anonymity, both of which raised ethical concerns within their activities. Two participants shared that they actively reported YamatoQ’s social media messages, while others worried that such actions might backfire, amplifying resentment and reinforcing conspiracy beliefs. P12, for instance, expressed concern that content moderation could hinder efforts to keep records of YamatoQ: “What happens then is that the media won’t be able to find evidence about their claims later.” Anonymity further complicated these tensions. Most participants opted to tweet anonymously to protect their privacy, diverging from traditional journalistic norms of source transparency (IFCN, n.d.). P11 recognized that this could undermine the trustworthiness of the information shared, while P7 contended that anonymity allowed them to reach users who distrust official sources, thus expanding their influence.

In keeping with this empathetic and journalistic mind-set, three participants drew intriguing parallels between the watchers and YamatoQ conspiracy theorists, noting that both groups share a strong drive to investigate and expose issues to the public by delving into complex narratives, reflecting a common human tendency to decipher the world around them—albeit with markedly different outcomes. Recognizing this similarity, P6 even expressed hope that the watcher community could offer a healthier alternative space for those drawn to conspiracy theories:

After all, watchers and conspiracy theorists are interested in the same thing. Why do some choose to be part of conspiracy movements? One of the reasons is because the activities of watchers are rarely well-known. I think it’s a shame that people who could have been one of us fell for conspiracy theories . . . You can enjoy conspiracy theories like eating junk food, but you can also enjoy it by simply doing more research to fulfill curiosity. If I can let people know that, things will get better. That’s why I do what I do.

Participants acknowledged that the empathetic stance toward YamatoQ is not universally shared within their community. For instance, P2 expressed strong feelings of anger and animosity toward YamatoQ and its ideologies, stating that he despises the group. P12 noted that a few watchers took a more confrontational approach in digital spaces, actively attacking and shaming YamatoQ followers: “There are people who curse YamatoQ members . . . [they] observe YamatoQ because they want to attack and swear at them.” However, P12 explained that he does not actively police watchers with differing values and opinions, choosing instead to “keep distance” from them and avoid interaction. “Watchers who are close to me are pretty critical of YamatoQ . . . but they won’t do anything aggressive . . . I think the way they approach what they do reflects journalistic ethics,” P12 said, though he noted that not all watchers explicitly view their activities as journalistic.

In summary, the watchers’ goals extend beyond simply dismantling YamatoQ or curbing the spread of problematic information. Rather, they seek to engage with its social and emotional dimensions, aiming to understand and empathize with its followers rather than alienate them. However, the watcher community is not monolithic. The tensions described above have led to subtle divisions within the community, which P6 described as “a price we pay for being an informal community.” Despite these differences, participants embraced this diversity, allowing room for ethical tensions to be negotiated in ways that enable them to pursue their broader strategic goals across various contexts. We now turn to how they translate this ethos into action across different spaces.

Multi-platform approach to problematic information

This section describes how the watchers worked toward their goals through multiple channels, highlighting the diversity and spontaneity in their activities and the informal nature of their community. As P11 articulated, “There’s a value in observing different platforms” because “it’s impossible to understand the big picture of what’s happening by observing only one platform.” In keeping with this perspective, we found that the watchers had four main sites of activities: physical spaces, private messaging platforms, Wikipedia, and Twitter.

Physical sites of action: information gathering and community building

While efforts to counter problematic information typically concentrate on online platforms, our participants suggested that this might be insufficient. They emphasized the significance of in-person observations and interactions in their efforts, asserting that such engagements contributed to a more holistic and grounded understanding of YamatoQ. For example, one participant shared that they tried raising people’s awareness about YamatoQ’s problematic activities by attending and documenting the trial of a YamatoQ member who was being prosecuted for breaking into vaccination clinics. Likewise, two participants recounted visiting YamatoQ’s headquarters to investigate its organizational structure and identify its members.

The physical space most commonly mentioned by participants was the anti-vaccination rallies held by YamatoQ. Participants would regularly check YamatoQ’s social media accounts to learn about upcoming rallies, openly attend and document them, and subsequently share what they observed with fellow watchers and their audiences on social media. For example, P5 mentioned, “Since January 23, [2022], I’ve gone to observe all of their rallies.” Similarly, P1 said documenting physical rallies is “the main thing I do.”

At these rallies, according to participants, watchers would meticulously document various details, including the size of crowds, what they chanted, their claims, special guest speakers, and demographics of the protestors (e.g. their estimated age groups). While four participants said they typically observed from the sidelines and focused on documenting events, another three said they would “try chatting with” YamatoQ members (P5) about topics, such as their stance on vaccines. One participant reported going a step further by staging a “counter-rally” (P2) to disrupt YamatoQ. Despite the differences in approach among participants, rallies served as key space for watchers to gather information, underscoring their strategy of combating YamatoQ wherever it operates.

Furthermore, participants shared that in-person engagements were also a way for them to meet other watchers in person and nurture a sense of community. As P6 described,

When I go to see a rally . . . that includes hanging out with other watchers and exchanging information with them . . . When we meet with each other, you feel a sense of affinity. You get close with each other very quickly. We go for dinner together and exchange information, and we become friends. We also exchange information that we can’t post on Twitter.

Taken together, these findings illuminate two key points. First, while interventions to problematic information are often discussed in the context of online spaces, participants also focused on physical spaces where people may be exposed to problematic claims. Second, physical sites were an important avenue for participants to foster a community of like-minded people, build friendships with each other, and facilitate information exchanges.

Private messaging platforms: infiltration, information sharing, and content creation

Participants’ activities extended to private messaging platforms, which are often hard to monitor due to their closed nature and encryption protocols (Gursky et al., 2022). They shared that they used two such messaging platforms, LINE and Discord, to self-organize and monitor YamatoQ.

Participants highlighted LINE as YamatoQ’s primary platform for recruitment and internal communication. Aligning with participants’ observations, YamatoQ had 72 active chat rooms with a total of over 10,000 members on LINE’s OpenChat (Fujiwara, 2022), a feature that allows people to chat anonymously. P1 mentioned that some watchers would “infiltrate” these LINE chat rooms to monitor, expose, and disrupt the conspiracy group’s activities. Infiltration, according to participants, involved covertly identifying and entering YamatoQ’s chat rooms without disclosing their status as watchers. P1 explained that when they saw problematic information, such as “racist claims,” they would screenshot it as evidence, and expose it on other social media platforms such as Twitter.

Participants emphasized the challenge of monitoring YamatoQ’s numerous chat rooms operated by regional chapters. For instance, P5 mentioned being part of 10 such chat rooms but could only manage the two largest ones due to time constraints. Monitoring these chat rooms alone was hard also because YamatoQ “might get cautious and kick me out [of these chat groups] . . . if it’s a small group, they might become skeptical of me,” added P5. To address this issue, participants distributed the activity of monitoring these chat groups among “a network of watchers”: “It was absolutely impossible to observe all of them by myself, so we loosely share this task” (P6). P2 unpacked this: “Other folks observe these chat rooms for me. . . [and] I would curate [their] information as soon as I obtain it.” P2’s quote illustrates the self-organizing nature of their activities. Instead of waiting for directives, watchers found their own roles and activities. For example, P1 focused less on infiltration, noting that “other folks are already exposing OpenChat conversations,” and instead aimed to contribute in ways others could not.

Participants also shared that Discord served as a central hub for various internal activities among watchers. Watchers would share screenshots of social media posts and chat rooms operated by YamatoQ members and share links for the information others are looking for, according to participants. A primary example of this account by participants is the seven Discord channels dedicated to sharing URLs for messages posted by YamatoQ members and news articles relevant to the group. P2 shared that watchers on Discord would also discuss how to effectively disseminate their content to a broader audience.

Participants also said that watchers on Discord participated in collaborative content creation. To concretize this account, we confirmed that watchers collaborated on constructing a publicly-accessible wiki “dictionary” of YamatoQ’s terminology, conspiracy beliefs, and profiles of its leaders; watchers crowdsourced knowledge and edited content together to ensure accuracy. They also discussed appropriate sources to cite and incorporated debunking information to the wiki.

Aligning with Chadwick et al. (2023), these findings highlight that problematic information often circulates not only on public platforms like Twitter but also on more private ones, such as LINE. The watchers’ endeavors to combat problematic information in these less-public arenas were driven by their spontaneous, ad hoc collaborations that involved task sharing, content creation, and information exchange.

“Edits wars” on Wikipedia

Three participants highlighted the significance of Wikipedia in their efforts. They described how community members spontaneously united when YamatoQ’s Japanese Wikipedia page surfaced. P3 said:

We noticed that YamatoQ’s Wikipedia emerged out of nowhere. We investigate [claims on the Wikipedia page] and replace content that has been posted or modified by YamatoQ members who seek to align the page with their own claims . . . some volunteers do this stuff.

This illustrates how the watchers collaboratively engaged in Wikipedia “edit wars” (Yasseri et al., 2012) with YamatoQ. P1 explained that winning edit wars required overpowering the opponent by force of numbers, and watchers used “as many people as possible” to edit the page simultaneously. Illustrating this account, coordination occurred on Discord, where they strategized on accurately portraying YamatoQ and ensured the information they were adding was citable and verified. For example, one watcher on Discord asserted that “it’s useless to write about [the slang used by YamatoQ] on Wikipedia. It will be deleted if there’s a problem with citations.” These efforts by participants on Wikipedia are noteworthy in light of research by Borak (2022) and the Wikimedia Foundation (2021) which highlight Wikipedia’s overlooked role in information warfare, especially in non-English editions with fewer contributors.

Twitter: direct engagement with audiences

Participants described actively engaging with other social media users by sending direct messages and replying to tweets to ensure they were factually informed. Twitter, a major platform in Japan (Newman et al., 2023), was a key tool for reaching both spreaders of problematic information and potential recipients. As P7 explained, “Fact-checking is pointless if it doesn’t reach both those spreading it and those at risk of believing it.” This perspective guided their approach to correcting false information on Twitter:

I would try to persuade the person who is spreading false information by posting an individual reply, or even if I couldn’t convince them, I would explain in the reply section [why the piece of information is incorrect] when it’s going viral so that someone else who saw the content wouldn’t be deceived. (P7)

Participants also reported using Twitter to alert audiences of YamatoQ activities in a timely manner. P6 said they would first “go on Twitter to break news” after observing rallies, because they can “deliver information more quickly” there than other platforms, such as blogs. A primary example of this activity we saw in the Twitter data set is participants sharing schedules of anti-vaccine rallies in different regions to notify nearby residents in advance and raise awareness within local communities about the danger of the conspiracy movement’s nationwide operations.

Taken together, these findings highlight how participants tried to achieve their strategic aims in a variety of settings, spontaneously responding to the spread of problematic information on multiple platforms. Interestingly, they associated their activities with positive feelings and terms such as excitement and friendship—we now dive deeper into this dynamic.

“It’s a hobby”: the playful side of misinformation resistance

Participants perceived their activities as enjoyable and fulfilling, often indicating a blurred line between “work” and “leisure.” They found value in keeping matters light-hearted for two reasons: facilitating broader audience engagement and sustaining their efforts.

Participants highlighted that playfulness helps them reach wider audiences. They noted using humor as a tool to entice people and convey their messages more effectively, rather than solely relying on presenting facts. P7 positioned this as entertainment with an educational aspect, where “producing an entertaining show” can have audiences engage with factual information:

I think the entertainment factor [in my tweets] is a reason why I can deliver information to the demographic of people who have no interest in factual medical information, unlike accounts operated by official institutions and doctors.

Participants also framed investigating YamatoQ as a form of entertainment or a “hobby,” deriving both intrigue and fulfillment from unraveling conspiracy theories and exploring the motivations behind them. P10, for example, noted that monitoring YamatoQ was fascinating because it offered a glimpse into “a very real aspect of humans . . . a manifestation of people’s anxieties.” Similarly, P12 explained that many watchers are drawn to this work because they enjoy encountering “extraordinary, weird stuff,” describing the watchers as “subculture geeks” who see conspiracy theories as a form of subculture. Aligned with their goal of providing healthier alternatives to conspiracy theorizing, several participants were also optimistic that the enjoyment derived from this investigative work could attract more individuals to their network. P12, whose tweets often included humor, polls, and quizzes related to YamatoQ, noted the appeal of this playful approach in engaging with others:

When I’m trying to be funny, it’s like, I’m trying to say that the watcher culture is fun, so everyone should join us. Or, even if you don’t wanna join us, please check it out because it’s fun. . .the main intention is to encourage the audience to be interested in it.

Yet, participants seemed to be keenly aware of the importance of not becoming overly immersed in conspiracy theories and the potential impact their activities can have on themselves. Speaking about the danger of this potential pitfall, P11 said “the mummy-collectors should be careful not to become a mummy,” referring to a Japanese idiom that describes a situation when someone trying to solve a problem becomes entangled in it or persuaded by the other side. Participants suggested that adopting a hobbyist mind-set made their efforts more sustainable, allowing them to maintain perspective and emotional balance. According to P11, “Having a sense of vocation and vision is important, indeed. But I think it [being a watcher] works better if everyone does it as a hobby.” They added that hobbies serve as a way to “maintain an appropriate distance” from harmful information encountered during their activities. To concretize this theme, we drew on data from the watchers’ Discord, which featured channels dedicated to talking about hobbies like gaming, anime, and cars—activities unrelated to YamatoQ but likely reflective of the kinds of hobbies P11 was referring to.

This approach also seemed to extend into how participants organized their activities in flexible and informal ways. For example, P5 shared that he used to be part of a more organized effort to counter Japanese hate groups that positioned him to be “more like an activist.” But such activism became “a little tough at times” as he had to commit to fixed roles, deal with internal troubles, and navigate through rigid member-to-member relationships:

If we establish a solid organization, I would have to deal with troublesome personal relationships [within the group]. I might not like the way [the organization] does things. Organizations feel uptight . . . I wanna continue doing this among a loose, not-so-organized group of acquaintances.

In summary, participants’ strategy to create entertaining content can be seen as a primary method to engage wider audiences, amplify their messages, and encourage others to join their network. Furthermore, they adopted the “hobbyist” and playful approach as a means to care for their mental health and make their activities more sustainable by avoiding a rigid organizational structure.

Discussion

In this study, we have described how an informal, voluntary group of Japanese social media users, who identify themselves as “watchers,” collaborated to investigate and counter problematic information propagated by a pro-QAnon group called YamatoQ. We now turn to discussing the roles groups like watchers can play in addressing problematic information. Considering this potential is important because, as we noted earlier, we are witnessing more grassroots efforts as people grow more cognizant of the impacts of problematic information. To discuss this further, we highlight two outstanding features of the watcher activities: improvisational and emotional dynamics offered by their informal volunteerism. We also reflect on tensions that may arise when situating such volunteer efforts within the broader societal interventions.

One of the key findings is that watchers drew on improvisational, rather than highly organized, approaches to countering problematic information. We have detailed how task sharing, like monitoring small chat groups, unfolded organically as watchers prioritized tasks based on availability. They also spontaneously responded to YamatoQ as it appeared on multiple spaces simultaneously, from Wikipedia and social media to private messaging and physical rallies. This flexible approach deserves attention in light of volatile virality of problematic information, which can unfold across multiple platforms in unpredictable ways and even spill over to the offline world (Lukito, 2020). This volatility complicates current interventions, like content moderation by social media platforms, which often lack coordination among themselves. The improvisational approach described by participants represents a potentially fruitful way to grapple with the volatility by forgoing fixed rules and roles.

Another important finding lies in empathy the watchers bring to addressing problematic information. Participants took a humanizing approach toward YamatoQ followers rather than confronting them with hostility, expressing hope that their community would be a “healthier” alternative to conspiracy movements. These dynamics were also reflected in how they countered YamatoQ using light-hearted approaches. For example, they collectively built a wiki page about YamatoQ using a Discord server, which they also used to foster friendships and have fun through discussing various subcultural topics. The framing of their activities as a “hobby” indicates that these activities became a regular practice they enjoyed. These accounts suggest that informal volunteer efforts can use communal and playful approaches to mirror and counteract the positive feelings leveraged by the purveyors of problematic information we reviewed earlier, such as belonging (Marwick, 2021) and intellectual fulfillment (Young, 2021). Community-driven interventions potentially contribute to more holistic solutions by going beyond critiquing people’s beliefs.

One way to highlight the significance of these findings is to connect them with the literature on informal volunteerism in the disaster context. Previous scholarship has highlighted noteworthy benefits of how informal volunteers support formal disaster response efforts by (1) addressing unexpected events in a timely manner, (2) meeting local needs and reaching people who might be overlooked by formal agencies, and (3) providing real-time, on-the-ground knowledge of the issue (Whittaker et al., 2015). Our findings suggest that informal volunteerism might also be able to offer similar benefits in the context of problematic information.

The previous literature has also highlighted challenges of incorporating informal volunteerism into more official efforts, such as concerns over safety and privacy protection (Park and Johnston, 2017), a lack of clear structure and accountability (Whittaker et al., 2015), and the need to cultivate trust with volunteers (Ruohonen and Backholm, 2023). Our participants taught us that there might be similar challenges in the context of problematic information. For example, they shared that different sub-groups within the watcher community sometimes worked at cross purposes; while some watchers would try to approach YamatoQ members with empathy, others would compromise that effort by vilifying them and exposing personal information. Such activities within the broader watcher community can be understood as vigilantism rather than information work. Furthermore, while the flexibility within the watchers’ informal organizational structure can allow them to exercise their own discretion to a great degree, the lack of structure can dilute the effectiveness of their activities, which impacts formal organizations’ trust in them.

Ultimately, these challenges represent opportunities for future work. Scholars can explore how we might make the work of such communities more visible, identify ways to connect them with formal institutions effectively, and investigate specific circumstances in which informal volunteerism might complement the existing solutions against problematic information. Community-based information work alone cannot replace the necessity for broader structural interventions, but we can take inspiration from tools and practices developed in the context of emergency events, such as Ushahidi (Okolloh, 2009), that facilitate information sharing and collaborations among different stakeholders. Given some informal volunteers remain active over extended periods (Atsumi and Goltz, 2014), the sustainability of such communities in the context of problematic information warrants closer examination. Such explorations will deepen our understanding of human-centered approaches to problematic information and how diverse communities can leverage their unique resources to respond effectively. Embracing these possibilities can help us address not only problematic information but also empower our online communities that provide spaces for social engagement.

Supplemental Material

sj-docx-1-nms-10.1177_14614448241302311 – Supplemental material for On the fly: How Japanese social media “watchers” improvise to counter problematic information

Supplemental material, sj-docx-1-nms-10.1177_14614448241302311 for On the fly: How Japanese social media “watchers” improvise to counter problematic information by Kayo Mimizuka, André K Rodarte and Ahmer Arif in New Media & Society