Just Here for “Danmu”: Participation and Interruption in Chinese Online Comments

(1)

Just Here for “Danmu”:

Participation and Interruption in Chinese Online Comments

Vincent Hoising So

Research Master’s Thesis

Department of Media Studies Graduate School of Humanities

University of Amsterdam 26 June 2017

Author Note

Submitted to the Department of Media Studies at the University of Amsterdam, Faculty of Humanities, in partial fulfilment of the requirements for the degree of

Master of Arts (M.A.)

Supervisors: Dr. Thomas Poell & Dr. Jeroen de Kloet Second Reader: Dr. Patricia Pisters

Correspondence concerning this Master’s thesis should be addressed to V.H.So@uva.nl

(2)

Abstract

“Danmu” is a commenting system in which viewers can post their comments directly on top of the video. These Danmu comments are projected directly on the screen. They float horizontally through the video and overlay the original content. As the video plays, viewers read the comments asserted by other viewers while watching the video simultaneously. Through an integrated theoretical and methodological approach inspired by actor-network theory, platform studies and visual culture, this study analyzes the Danmu commenting system on the Chinese video streaming site Youku. It is revealed that the technological platform features of Danmu challenge existing boundaries of traditional comment section. As a result, temporal and spatial

reconfigurations of Danmu comments lead to new forms of audience participation. By tracing the actor-network in Danmu participation, it unveils the role of the nation state within the interruptive participation. This study puts forward the idea of an

interactive, interruptive, and multi-layered “interrupted narrative.” It suggests that there is constantly shifting power dynamics among content, audience, media technology and nation state.

Keywords: Danmu, participatory culture, online commenting, screen culture, interrupted narrative

(3)

Acknowledgments

First and foremost, I would like to thank my supervisors Dr. Thomas Poell and Dr. Jeroen de Kloet for their insights and guidance. This project would not have existed without the time, effort and patience they have given me in the past year. I would also like to thank Dr. Patricia Pisters for graciously agreeing to be the second reader of this thesis.

Furthermore, my sincere gratitude goes to my peers in the Research Master's program, especially Charlie Vielvoye and Dennis Leeftink, for their continuous encouragement and support in the past two years. I would like to take this opportunity to show my appreciation towards the Media Studies Research Master's program at the University of Amsterdam, and all the faculty members who have provided assistance and feedback in the process. It is the freedom and inspiration from the program that makes this project possible.

Last but not the least, I wish to thank my family for the support and inspiration they have given me. It has been a mentally and emotionally tough year for me. Without their tireless and unconditional support, understanding and love, I could not have written this thesis.

(4)

Chapter 1: Introduction

“Danmu” (Chinese:弹幕; Japanese:弾幕/だんまく), sometimes referred to as “Danmaku,” is a commenting system in which viewers can post their comments directly on top of the video. These inserted texts are not subtitles or annotations added by video uploaders, but actual comments posted by internet users. Unlike the typical western video sites, where comments are shown below the video in a designated “comment section,” these Danmu comments are projected directly onto the screen (see Figure 1). The comments float across the video from right to left, overlaying the original video content. As the video plays, viewers read the comments asserted by other viewers while watching the video at the same time.

Figure 1. The Danmu Interface

Literally meaning the “bullet curtain” in both Chinese and Japanese, the term “Danmu” or “Danmaku” is said to be a translated military term of “barrage.” Now it is used to describe comments resembling “rapid artillery fire,” flying over the screen like a “bullet curtain” in video games, signifying the ferocity of this visual wall of

(6)

comments that allows users to engage and participate in real time. While it is referred to as “Danmu” based on its Chinese translation, due to its Japanese origins, some existing literature still refers to the unique commenting system as “Danmaku.” Since this thesis is focusing on the subject within the context of China, this paper will refer to the commenting system as “Danmu” instead of “Danmaku.”

This concept of Danmu was first introduced in the Japanese ACG (Animation, Comic, and Game) community in the year 2006 in one of the first Danmu video sites Niconico. After the unique form of commenting had gained significant attention from the Internet community, Danmu sites were introduced to China around 2008 which included popular sites such as Acfun, Bilibili, and Miomio, whereas Danmu was still considered as a minority form of commenting only popular within the ACG

community back in Japan1. It was not until around 2012-2014 when the idea of Danmu was picked up by major video sites such as Youku, Sohu, Tudo, and Tencent QQ, and thus gained household recognition in China. Today in China, Danmu has become a mainstream (and mostly default) feature in video and movie sites. Due to the unique intellectual property culture and copyright law of China, the content from these Chinese video sites is mostly free. Some of the major video sites now hold exclusive streaming rights to certain movies and TV shows, and users can pay a subscription fee to access the latest movies. With reference to the western mediasphere, these video sites are a combination of YouTube and Netflix. Like YouTube, users can view, upload, search, and share on these video sites. However, at the same time, most of the content are from other mainstream entertainment media such as movies and television programs, similar to more singular distribution patterns like Netflix. The specific feature of Danmu is not only found in both Danmu sites and

(7)

mainstream video sites, as some movie theaters in China have also been introducing this feature to the real-life cinema.

When compared with the rigid and immutable comment section that

characterizes most media sites around the world, the fluid and imbricated flow of the Danmu system appears to break and redefine the boundaries of time, visual

representation, and interactivity of how the traditional online commenting works. Danmu is unique in how it presents multi-source information on the same screen. Through the interlacing and meshing of different elements across time and space, the phenomenon of Danmu leads us to rethink our common perception of online

commenting, and challenges the existing media habit of video-viewing and

commenting. It prompts us to examine this new form of audience participation that is taking shape in China.

Danmu appears to correspond with larger shifts in audience participation, which we can observe in contemporary visual culture. Particularly interesting in this regard is the so-called “second screen phenomenon”: when people are watching TV, they also use a second screen device such as mobile phones, tablets, and laptops to express and socialize on other media platforms (Giglietto & Selva, 2014; Ma & Cao, 2017). For example, more and more people are live-tweeting when they are watching a specific television series or event, which is indicative of an attempt to create a shared viewing experience through social networking sites (Schirra, Sun, & Bentley, 2014; Wohn & Na, 2011).

Both contemporary media platforms and content creators are trying to respond to the second screen phenomenon by superimposing content and user commentary within the same platform or even within the same screen. The rising popularity of the genre of “reaction videos” on YouTube is a clear indication of the trend. Platforms

(8)

like live-streaming video service Twitch are simultaneously placing gameplay, webcam of the broadcaster, and audience chatroom within the same screen. In contemporary screen culture, we can observe more and more of these fragmented or interposing narratives and conversations. For example, in the Facebook Live videos introduced in 2016, emoticons overlay and interrupt the original video in a similar fashion like Danmu. There are also direct to home (DTH) TV set boxes that allow viewers to externally insert Twitter feeds on their TV, watching TV along with real-time tweets (Mathur, 2014).

People enjoyed extra commentary when they are watching their programs, and in response, content producers and distributors are incorporating multiple elements and stories onto one single screen. Everything adds up to a new media experience: Internet users are consuming audio-visual content while knowing what others are talking about it. These observations point out that people are constantly using their social media while simultaneously consuming audio-visual materials. Here is the logical questions: Is there another form of audience participation that derives from this new form of viewing habit? And more specifically, how does Danmu contribute to this habit?

Another trend that needs to be addressed is that producers, audiences, and

platforms are struggling to gain control of the comment section. The comment section is a major default feature of most social media, serving as an indispensable aspect of new media and allowing us to interact with each other in the digital sphere. However, since the mid-2010s, big media companies and news sites, such as CNN and WIRED, are either gradually disabling their comment sections or heavily moderating them (Finley, 2015). News organizations struggle to maintain and regulate their own comment spaces, as they are flooded by spambots, trolls, and hate speech. Since

(9)

social media has changed the means of information distribution, and users now prefer to engage in direct debates on social media, articles, and stories are shared through social media platforms such as Twitter, Facebook, and Instagram. Therefore, big media corporations such as Google and Facebook now have control over the majority of online comment sections. In the past decade, these media giants constantly

restructure their comment sections. The top comment algorithm of each platform has a different way of sorting out the comment spaces, especially regarding how they order and rank comments within the comment section. The platforms are dedicated not only to furnishing convenience to the users, but also to effectively regulating the comments by avoiding spam bots and other forms of abuse by users.

The above observation suggests a change in power relations between content, consumer, and media technology. Back in the days of the pre-internet era, when print media (e.g. books, newspapers) and broadcast media (e.g. radio, TV) dominated the mediasphere, “content was king.” The content producers held the most power and determined what was being produced and delivered in one-way communication channels. After the invention of the World Wide Web in the late 20th century, the audience (or customers) became empowered, especially after the Web 2.0 era. When internet platforms allow interaction between producers and consumers, audience participation entails a shift in media power towards audiences, as the instantaneous and decentralized form of communication allows audiences to determine what is valuable in mass media (see Benkler, 2006; Bruns, 2006; Jenkins, 2006). Nowadays, some scholars argue that “platform is king” (Hoang, 2016), as social media platforms not only control the distribution of content but also determine the presentation, and sometimes the surveillance and censorship, of audience voice in the comment sections as well.

(10)

It is evident that there is a fundamental transition of power from content producers to audiences, and platforms are taking advantage from the user’s activity through data collection and exploitation of immaterial labor, and gain power in the process (see also Couldry, 2015; van Dijck & Poell, 2013). Content producers, audiences, and platforms all play a crucial part in our increasingly variegated media environment. But how does the power dynamic operate between them today? As Danmu shows a different form of user engagement from its core design, does the unique form of audience participation in Danmu influence the change of media power between different stakeholders? This research aims to shed light on the subject

through the study of Danmu.

Through the following steps, this thesis explores audience participation and power relations in contemporary comment culture from the perspective of Danmu: The next chapter provides the conceptual and methodological framework for the analysis, starting with a review of Danmu/Danmaku literature in both English and Chinese. After identifying the common themes and missing angles from the existing literature, the chapter will continue discussing relevant research traditions concerning online commenting and participatory culture, which will aid us in understanding the phenomenon of Danmu conceptually. In order to fill in the gap within the existing literature, this chapter will explain how an integrated approach inspired by actor-network theory, platform studies, and visual culture is useful in deconstructing the technical and social aspects of Danmu. Last but not least, the selection of cases for study, analytical approach of the platform and visual-discourse analysis will also be detailed in the chapter.

Through a combination of platform analysis and visual-discourse analysis of various examples, Chapter 3 will explore how the “synchronicity” of time and space

(11)

brings in alternative forms of audience participation in Danmu. Danmu breaks and redefines the existing spatial and temporal boundaries of traditional comment sections that in turn leads to a form of collective yet interruptive co-viewing experience. The co-viewing experience will help us understand audience participation and power relations in Danmu.

Following the arguments in Chapter 3, Chapter 4 will examine how the

interruptive nature of Danmu participation brings about a new power relation between content, audience, media technology, and the nation state. It will be argued that the synchronization of time and space within the Danmu commenting system introduces a new form of narrative, namely the “interrupted narrative.” The “interrupted narrative” is interactive, interruptive, and multi-layered. This new form of narrative does not only affect how commenters react differently in the sense of participatory culture, but also how the audience consumes and interacts with audio-visual materials and user-generated materials differently in the new media age. Through the specific example of the Chou Tzuyu flag controversy, this chapter will show how the nation state of China also plays a role in the nationalistic expression observed in multiple instances in the study. By going through the interruptive co-viewing experience in the Chinese context, this thesis will demonstrate how the Chinese Danmu system initiates a constant change of power dynamics among content, platform, users, and even the nation state.

The research objective of this thesis is to explore how the unique

presentation/interruption of Danmu comments invites new forms of audience participation and facilitates a new power dynamic between producer, content, audience, and media technology. The final chapter will conclude this thesis by highlighting the theoretical contribution of this study with regard to media

(12)

(13)

Chapter 2: A New Reflection on Danmu

Despite being a relatively new phenomenon, there have already been a number of studies done about Danmu. We need conceptual reflections on the study of Danmu in order to understand new forms of audience participation and power relations in this new media phenomenon. By going through existing Danmu literature and relevant research traditions, this chapter explains why an integrated approach is desirable to study the different aspects of Danmu. The research method and choice of the cases for study are also described in this chapter.

2.1 Collective Viewing, Fragmented Narrative and “Synchronicity” of Danmu

Three common themes about Danmu are identified in the existing Chinese and English literature.2 The first theme is that Danmu constitutes a form of collective viewing experience derived from youth culture and Otaku culture. We can understand the popularity of Danmu in terms of participatory culture and fan culture (Zhang, 2014). Many authors frame Danmu as a form of Chinese youth culture, saying that young people try to redefine and empower themselves by adopting new media techniques not understood by older generations. Chinese youth are said to use

“images, data, multimedia videos and new vocabulary as ‘weapons’ that their parents do not understand”3_{(Chen, Cao, & Wang, 2013, p. 20). Zhang (2014) advocates that}

Danmu serves as a means of resistance towards corporate consumerism and

2_{While this thesis recognizes that Japanese literature on the topic of Danmu/Danmaku exists, it will}

not be discussed here due to language barrier and time limitation.

3_{Translated from: “他们以父辈不了解的图像、数据、多媒体视频、新词语作为‘武器’” (Chan,}

(14)

traditional values, as criticism and conflict emerge from the fragmented,

discontinuing, diverse, and decentralized Danmu conversation. This results in an anti-orthodox discourse in the process.

The popularity of Danmu in China is often explained by the Japanese ACG culture and Otaku culture4 (Chen et al., 2013; Shi & Zhang, 2015; Yuan, Xing, & Jiang, 2016). ACG (Anime, Comic and Games) as an adolescent subculture is becoming one of the most significant Internet cultures in China, largely due to the increasing exposure of Japanese pop culture which brings the concept of “Cool Japan” and Otaku culture to the Chinese digital natives (Lin, Ito, & Hirokawa, 2014; Zhao & Tang, 2016). According to Peng, Zhao and Teo (2016, n. p.), Danmu is an “important medium for Chinese young people who are fond of ACG subculture to communicate and disseminate their thoughts, attitudes, and emotions.”

Particularly striking is the idea of collectivism in ACG culture. Ian Condry proposes in The Soul of Anime (2013) that there is a form of collective social energy which he calls “the soul of anime” (hence the title of his book). He argues that ACG culture’s “success arises from social dynamics that lead people to put their energy into today’s media worlds” (p. 2). This collective social energy motivates its audience to “value media and use it as a means to help organize their social worlds” (Condry, 2013, p. 29).

With its ACG roots, Danmu shares this collective social energy with its user base (Chen et al., 2013; Shi & Zhang, 2015). The collective viewing experience is the key here as Danmu jumps from anime sites to mainstream video sites, thereby extending this viewing experience to other forms of content. Peng et al. (2016) emphasize the importance of perceived coolness and subcultural identity of ACG culture in

(15)

innovation adoption. They suggest that subcultural features are integrated into the design of Danmu. While Peng et al.’s argument sounds rather speculative, it raises interesting questions concerning the relationship between popular culture and media technology. Does certain popular culture influence the design of media technology? Do platform features reflect the audience’s needs? These questions call for reflection on the study of Danmu, as the relationship between media technology and its socio-cultural setting is often neglected.

Many studies have argued that youth culture and Otaku culture lead to a form of shared-watching or co-watching experience. Some studies attempt to explain this idea by invoking motivation and gratification behind this co-watching experience. Zhao and Tang (2016) look at the psychological and behavioral aspects of user behavior by exploring the motivational affordances of Danmu websites from a gamification design perspective. Based on the framework of affective response model, Liu, Suh and Wagner (2016) attempt to identify users’ satisfaction and intention with reference to their cognitive state. Some studies rely on uses and gratifications theory to explain users’ needs for company and belonging (Chen, Gao, & Rau, 2015; Yuan et al., 2016). All these studies support that there are emotional and social needs of Chinese

audience within the collective viewing experience in Danmu. However, these studies have not addressed how the roles of platform and content influence audience

participation, as most of these studies simply focus on the audience aspect and assume the audience hold the most power. This thesis hopes to investigate how content and media technology play a part in the Danmu experience. For example, does the design of Danmu incorporate the collective social energy and subcultural identity of ACG culture into its presentation?

(16)

redistribution of attention in the Danmu experience. Ma and Cao (2017, p. 771) refer to Danmu interactions as “fragmented, agrammatical, and interactionally disjointed.” Jie (2014) and Zhang (2015) observe a further fragmentation of conversation that users prefer to consume and engage with the text comments rather than the original visual content. An interesting paper discusses the postmodern aesthetic of Chinese online comment cultures of Danmu from a visual culture and aesthetic perspective (Xu, 2016). In discussing the Chinese internet phenomena of “Gailou” and Danmu, Xu argues that these forms of alternative comments “embody the aesthetic of the postmodern in that net users are increasingly disinterested in the metanarrative

(content) and instead become obsessed with multiple small narratives (comments)” (p. 439). Xu further explains that the fragmentation of content is nothing new in the field of Chinese traditional art:

“The Chinese perspective occupies both the formal and aesthetic elements of artistic production and consumption. Contrary to Western painting that is

generally presented on a fixed canvas, Chinese scroll painting provided a kinetic medium where the viewer’s perspective shifts as the scroll is unraveled. The spatial and temporal medium here is no longer confined to a single point of view, but rather the viewer is constantly interacting with multiple vantage points within the print. The departure from the Western singular perspective allows the artist to interpret works based on multiple nonlinear perspectives that operate within different temporalities” (Xu, 2016, p. 446).

Xu (2016, p. 439) argues that the fragmented presentation of Danmu “challenges Western-centric views of modernization project while opening upon alternate

(17)

narratives across different global cultures.” Xu’s argument does not only provide a historical explanation in Danmu’s spatial and temporal reconfiguration in relation to the western comment section, but also provides insight in the celebration of

fragmented narrative in terms of visual culture. This is useful in exploring how Danmu affects audience consumption and interaction with these fragmented

comments. It concerns the complex narrative construction and perception within the Danmu experience, as well as the power relationship among content, audience, and medium (as new media technology).

The third common theme is the idea of “synchronicity across time and space” (跨时空的共时性) in Danmu. This is a term several Chinese scholars have used to describe a key characteristic of Danmu, namely the synchronization of time and space in the same screen (Chen et al., 2013; Li & Wei, 2014). The idea of 共时性 is more than the translation of “synchronicity.” As the Chinese character of 共 also bears the meaning of “common,” “altogether” and “shared,” and the character of 时 implies not only “time” but also “opportunity.” Different articles refer “synchronicity” to different aspects of the Danmu community they have observed. The idea of

“synchronicity” proposes that Danmu brings together time and space of the comments on the same screen. It also brings people together across time and space (Li & Wei, 2014). This mirrors the idea of collective viewing experience discussed earlier. Other than that, Chen et al. (2013, p. 19) also state that the synchronicity of Danmu forms a “virtual tribal viewing atmosphere.”5_{It brings different forms of content and}

comments onto the screen, including shared humor, shared emotions, and shared “spiritual-homing.”6

5_{Translated from: “形成一种虚拟的部落式观影氛围” (Chen et al. 2013, p. 19)} 6_{Translated from: “精神归宿”}

(18)

In this thesis, synchronicity of Danmu refers to the merging, or what is called “synchronization,” of time and space in a commenting format. By bringing different elements together in the same space at the same time, synchronicity of Danmu makes it unique and different from the traditional bulletin-board comment system of the West. The syncing of time and space allows “multiple parallel layers of interactions that enable the reconfiguration of the temporal/spatial order contesting the unitary narrative” (Xu, 2016, p. 447). Combining with the idea of bringing the comments, the audience, and the social into the same time and space, the notion of “synchronicity” allows us to explore spatial/temporal reorientation in Danmu, and also how new forms of participation emerge from it. More importantly, when the content, audience, and platform features are all brought together, what is the contested relationship in Danmu? Who holds more power in this synchronized media experience?

Even though the existing literature has provided many useful ideas and

perspectives in approaching Danmu, there are some limitations in both the Chinese and English studies. Most Chinese articles identify the characteristics of Danmu but are unable to provide in-depth analysis with empirical support. There is not much inter-referencing within the Chinese literature, and some similar points are repeatedly mentioned. Many of these papers are written in an authoritative manner without much evidence. They do not necessarily focus purely on the negative aspects of Danmu, but tend to adopt the government’s perspective. While some papers recognize that Danmu is an empowering form of expression for the Chinese youth, the general conclusion is often about how one can regulate and control Danmu video content and comments. By implying that Danmu can and should be regulated, they fall into the trap of assuming internet users as passive audience and neglect the freedom it brings to the Chinese users.

(19)

The influence of the nation state can be observed in their discourse and

arguments. Many articles tie their findings back to the national interest, and present a paternalistic and governmental sentiment in trying to educate the public to fit the Chinese socialist values. It shows that the Chinese academic content about Danmu is governed by a sense of national political correctness. These articles show power relations not only among audience, content, and media technology, but also the nation state. It reveals a significant role of the government in the context of China.

As for the Danmu studies written in English, there are two major issues

surrounding how they approach Danmu. First, most studies simply focus on Danmu-exclusive sites such as Bilibili as case study. What makes Danmu interesting to study is the moment when it becomes the default mode of commenting on Chinese video sites, where millions of people watch movies, TV shows, news, podcast, and other forms of visual entertainment. Some studies show that the default setting on Facebook does affect users’ behavior (Acquisti, Brandimarte, & Loewenstein, 2015; Lewis, Kaufman, & Christakis, 2008; Liu, Gummadi, Krishnamurthy, & Mislove, 2011). “Default” implies convenience and normalization. Thus people treat the default as the norm and is the recommended way for everyone to follow (Johnson & Goldstein, 2003, p. 1338). Under the same logic, default settings of these major video sites affect the dominant viewing habit and commenting culture in China. This is an oversight undermining the impact of Danmu on the Chinese internet, and a missed opportunity to investigate the power relations between media technology and audience.

Second, not many of the studies have actually conducted close reading or in-depth case studies on Danmu videos, and thus they fail to discuss the influence and interconnection among content, audience, and media technology. Some studies are problematic in how they define the values in their data and establish correlations

(20)

between variables. For example, Peng et al. (2016) propose a research model without any empirical work to back up their claims. He, Ge, Wu, Chen and Tan (2016) put forward another model utilizing the dynamic herding effect to predict the popularity of videos. But it seems to lack social significance and has no implication as the study merely describes the herding effect they observed. As there is just only limited number of peer-reviewed articles out there, most of the conference papers remain experimental and only make claims without solid research methods or procedures. These studies usually fail to explain the relationships between ambiguous values (in relation to media technology), and do not analyze the Danmu comments (audience participation) in relation to the audio-visual materials (content) presented on screen. As a result, they do not address the complex relationships among Danmu comments, the video, commenters, and the viewers. This study tries to fill some of the gaps by conducting an in-depth close-reading on Danmu video and zeroing in on the changing power relations among different elements.

2.2 Participatory Culture and Online Commenting

Danmu is an online commenting system that depends on heavy audience participation. A study on the nature of online comment would require us to look into the idea of Web 2.0 storytelling and transmedia narrative. Commenting has been a key aspect of online platforms since Web 2.0 started in around the mid-2000s. The

comment section is a crucial interactive element of the platforms, with user-generated content shaping the online discussion. Users can comment on every post or video and contribute to the online discourse with their “collective intelligence” (O’Reilly, 2009). These interactions “fold into the experience of the overall story from the perspective

(21)

of subsequent readers” (Alexander & Levine, 2008, p. 47). Commenting has become part of the Internet narrative experience. Pre-Web 2.0 user-generated content can only happen off-screen and off-Internet, but today it is difficult for viewers to only see a post without taking other viewers’ contribution and reaction into consideration.

Regarding reading media text from a storytelling perspective, Jenkins discussed transmedia storytelling in his book Convergence Culture (2006). He sees audience participation as a positive empowering factor, as he further explains in his blog Confessions of an Aca-Fan:

Transmedia storytelling represents a process where integral elements of a fiction get dispersed systematically across multiple delivery channels for the purpose of creating a unified and coordinated entertainment experience. Ideally, each

medium makes its own unique contribution to the unfolding of the story. (Jenkins, 2007, n.p.).

Reading the comment section is now part of the entertainment experience.

Comments can be equally important as the original content, as they have the power to alter the original content “beyond the immediate control of a creator” (Alexander & Levine, 2008, p. 48). The comment section can provide a more sophisticated narrative experience when it draws the audience to form a collective experience. The

interpretation of other viewers will reshape and re-contextualize the content. The interactive aspect of the social media platforms, along with the unique contribution of the audience, can shape multiple complex reading or comprehension of a single matter. The audience feels like their contribution matters, and they are able to spin the original content to their liking. This is summarized effectively in a YouTube blog post

(22)

title about the new commenting system: “Turning comments into conversations that matter to you” (Janakiram, 2013). The above literature reinforces the importance of participation in Danmu’s co-viewing experience. The overall narrative is shaped by both content producers, consumers (audience) and distributors (platform).

The idea of studying audience as an active user instead of a passive consumer can be traced back to as early as the 1950s (Reinhard, 2009). Marshall McLuhan foretold that we would enter an age where “everyone is a publisher”; simultaneously, the idea of active audience theory is further recognized throughout the years (Simpson, 2014). At the latter half of the 20th century, more and more theorists recognize the ability of the audience to become an “active consumer” of media, as audiences are actively involved in determining how they engage with the mass media. Theories such as the encoding/decoding model (Hall, 1973) and the uses and gratifications theory

(McQuail, Blumler, & Brown, 1972) are brought up in the study of television audience. James Webster (1998) proposes the idea of “audience-as-agent,” stating that:

Rather than seeing people as acted upon by media, people are conceived of as free agents choosing what media they will consume, bringing their own interpretive skills to the texts they encounter, making their own meanings, and generally using media to suit themselves. (Webster, 1998, p. 194)

Other active audience theorists such as Ien Ang (1991) also establish that

audience should be considered as active social subject and not as passive object to be controlled. However, other scholars such as David Morley (1993) and John Corner (1991) warn about the overt optimism in active audience theory by questioning its

(23)

methodology of ethnographical research and the conception of media power.

In the field of popular culture of the 21st century, “web-based media have made multidirectional, audience-generated communication a reality, giving citizens the opportunity to join the party as producers rather than merely consumers” (Gross, 2009, p. 67). Most notably, Jenkins (2006) suggests that new forms of audience are active, resistant, and socially connected in the age of convergence. As more and more are claiming “audiences are dead, long live the user!” (Jermyn & Holmes, 2006; Rosen, 2006), the idea of “prosumption” started to emerge around the mid-2000s. “Prosumers” are active participants who are both consumer of content and contributor of the overall media narrative in an environment that enables user-led content

production and collaborative engagement (Bruns, 2006). YouTube, and arguably the Chinese Danmu sites, consist of the characteristics of “both a ‘top-down’ platform for the distribution of popular culture and a ‘bottom-up’ platform for vernacular

creativity” (Burgess & Green, 2013, p. 6).

Some theorists have critiqued earlier the rather one-sided celebrating view of the active prosumer. As time progresses, more scholars have written about the

exploitation of consumer and user-generated content in immaterial labor (Andrejevic, 2012) and user agency (van Dijck, 2009). Media companies and social platforms have found ways to extract and take advantage of users’ activities. The audience is no longer the sole stakeholder that is empowered by popular culture.

Corporate-controlled media platforms also have the ability to resist, exploit, and construct their own discourse in terms of taking power from the audience. The above debate shows that there are always pushbacks or waves of backlash within the academic

community, and the fundamental question of “who is being empowered” is not always set. One goal of this thesis is to map out the changing power relations in the case of

(24)

Danmu.

The seemingly empowered audience is not only the concern for participatory culture, but also within the study of online comments. In the existing literature

concerning the online comments, most of the studies are from the field of Journalism. This literature is relevant because they draw connection between audience

participation and power relations between content and audience. Starting from the mid-2000s, comments on the same page with news content began to capture Western academia’s attention with regard to online journalism (Santana, 2011). Many studies are about the content and topic of comments that respond to news stories, and the respective perceptions and judgments of readers that are related to agenda-setting and media bias (Houston, Hansen, & Nisbett, 2011; Lee & Jang, 2010; Thorson, Vraga, & Ekdale, 2010; Walther, DeAndrea, Kim, & Anthony, 2010). Some literature also discuss the interactivity of reader comments (Weber, 2014), the anonymity of the commenters (Reader, 2012), and the social identification between viewers (Walther et al., 2010). Even though some take on a pessimistic position that professional

journalism is fading away, there is a rather optimistic view that audience triumphs over content and media technology.

The general belief is that the audience holds the most power as they can appropriate the original content through interactive features online. Scholars are mostly interested in user-participation and formation of public opinion in the sense of participatory journalism. This is often connected to the notion of the public sphere (Habermas, 1962/1999), pointing at the relationship among online participation, citizenship, and an emergent online public sphere (Papacharissi, 2002; Simpson, 2014; Toepfl & Piwoni, 2015). Joke Hermes (2005) takes this discussion to a “less formal everyday practices of identity construction, representation, and ideology” (p.

(25)

4). She defines the notion of “cultural citizenship” as “the process of bonding and community building, and reflection on that bonding, that is implied in partaking of the text-related practices of reading, consuming, celebrating, and criticizing offered in the realm of (popular) culture” (Hermes, 2005, p. 10). In their book about YouTube, Burgess and Green (2013) take the idea of everyday cultural citizenship and describe the YouTube community as a “cultural public sphere.” However, Hermes (2006) also reminds us that new media technology simply provides for new and important citizenship practices, but may not necessarily produce new citizenship. This is important as we will discuss the collective participation observed in Danmu comments in later chapters.

2.3 An Integrated Approach: Actor-network Theory, Platform Studies and Visual Culture

In order to investigate how Danmu invites new forms of audience participation and facilitate new power relations among producer, content, audience, and media technology, and to respond to some of the questions and limitations raised by the literature, this thesis would adopt an integrated approach that is inspired by the actor-network theory (ANT), platform studies, and visual culture.

In his book Reassembling the Social (2005), Latour rejects the idea of considering the social as a single being that determines and influences human interaction. Instead, the social is constituted through the connections and association between people and objects. He writes, “There is no society, no social realm, and no social ties, but there exist translations between mediators that may generate traceable associations” (Latour, 2005, p. 108). Therefore the study of ANT is the study of what

(26)

he calls the “sociology of associations.” For Latour, external social explanations cannot explain human interaction. There are only networks and assemblages. He defines network as “nothing more than an indicator of the quality of a text about the topics at hand… Network is a concept, not a thing out there. It is a tool to help describe something, not what is being described” (Latour, 2005, pp. 129-131). An encyclopedic definition of the term “assemblage” states “a group of objects of different or similar types found in close association with one another” (Kipfer, 2000, p. 38), which shares a similar idea.

Within these networks and assemblages, there are actions. Latour (2005, p. 44) states that “action is not done under the full control of consciousness; action should rather be felt as a node, a knot, and a conglomerate of many surprising sets of

agencies that have to be slowly disentangled.” It is important to know that an action is not a deliberate act with a conscious intention made by an actor. Instead, action is informed and determined by the role of actor and its associations between other actors within the network. In this case, actors can be seen as participants or stakeholders “in action” within an assemblage. Latour (2005, p. 52) argues that every person and object has agency which “are always presented in an account as doing something… making some difference to a state of affairs.” In ANT, it is assumed that human and non-human actors (media technology) share an equal amount of value and agency. Utilizing ANT on Danmu can identify the components within a network and

determining contingencies between actors, thus revealing the power relations among them. Apart from audience, content and platform also hold power and can impact on the overall media experience.

Apart from ANT, this thesis also draws elements from science, technology and society (STS) studies to focus on the technological aspect of Danmu platform which

(27)

has a unique interface and website features. According to Bogost and Montfort (2007, p. 178), “ platform studies are the humanistic parallel of computing systems and computer architecture, connecting the fundamentals of new media work to the cultures in which they were produced and the cultures in which coding, forms, interfaces, and eventual use are layered upon them.” In this case, platform studies are the investigation of computing systems that “reveals the interactions between these systems and creativity, design, expression, and culture” (p. 177). It also connects the technical details of these systems to culture. In the conception of platform studies, there are multiple levels of approaching new media which include

reception/operation, interface, form/function, code, and ultimately the abstraction level of “platform.” It is said that:

If code studies are new media’s analogue to software engineering and computer programming, platform studies are the humanistic parallel of computing systems and computer architecture, connecting the fundamentals of new media work to the cultures in which they were produced and the cultures in which coding, forms, interfaces, and eventual use are layered upon them. (Bogost & Montfort, 2007, p. 178)

Bogost and Montfort (2007) stress that platform studies recognize all the other levels above (e.g., interface and code) and situate them in culture, society, economy, and history in the humanistic exploration of computing. They also add that they “seek to describe how platforms have come about as well as how they influence further cultural production” (Bogost & Montfort, 2007, p. 178). In a later article, they clarify that platform studies includes both hardware and software of all computational

(28)

platforms, and is opposed to “hard” determinism (Bogost & Montfort, 2009). The methodological approach of platform studies is particularly useful in studying the technological features of Danmu, and at the same time not omitting the social-human connection between the platform and other actors. Therefore,

conducting a platform analysis in this study is desirable in terms of exploring the connection between platform features of Danmu, original video content, and audience participation (Danmu comments). By analyzing the medium feature of the

platform,this approach can dig into media specificity and remediation (Bolter & Grusin, 1999) of Danmu and provide a more comprehensive understanding of how co-viewing is made possible through the operationalization of Danmu. It also helps reveal the power relations between media technology and other actors.

Last but not the least, it is crucial to draw ideas from visual culture and literacy studies in analysis Danmu videos as cultural artifacts. The notion of multiliteracy is useful for discussing the image-text relationship in Danmu, as it helps us understand the temporal/spatial reconfiguration of Danmu and the relationship between content and audience.

As early as the beginning of the 20th century, Walter Benjamin’s works already started to address the increasing interdependency within the image-text relationship. When images coalesce and collide with words, critical meaning and reflection emerge (see Walter Benjamin’s Archive: Images, Texts, Signs, 2015). Fairclough (2000) argues that meaning is drawn from how different sign systems interact with each other in his idea of “multisemiotic nature.” Therefore, we take both visual culture and literacy theory to look into the notion of multiliteracy which includes both the text and the visual in image-text processing. Cope and Kalantzis (1999, p. 24) further define it as “the multimodal relations between different meaning-making processes that are

(29)

now so critical in media texts and the texts of electronic multimedia.”

After the rise of post-structuralism, scholars recognize there is a dynamic

relationship between visual culture and literacy studies. Words anchor the meaning of the pictures, and vice versa. “Meaning lies within what we make of the space between discrete communicative systems” (Duncum, 2004, p. 257). Marantz (1978) states that written narratives and pictures obey their own conventions but “together they should provide a much fuller experience than either alone” (as cited in Duncum, 2004, p. 256). One does not only have to read both text and imagery together, but also need to take social practice and social context into consideration. Image or text is no longer a singular object to be analyzed, especially in the age of new media. Everything on the screen adds up to a bigger narrative. Sturken and Cartwright (2000) argue that visual culture means not only recognizing the integration of the visual representation, but more importantly how cultural forms inform human life beyond their perception. “The study of visual imagery is concerned with more than images; it is concerned with the whole context of images, their production and the lived experience of those who view and interpret” (Duncum, 2004, p. 254). Multiliteracy should be understood as a form of social practice, a dynamic reading into the social representation in modern

consumer culture (Slater, 1999).

In Danmu, the viewing of videos incorporates the reception of moving image, text, and audio. It also involves the Internet culture, opinions, and comments of other users within the overall narrative. This idea has been demonstrated by previous Danmu studies, as Xu (2016) argues that consumption and interaction of users play a huge part in the new form of visual culture of Danmu. If multiliteracy brings image and text together in the visual, Danmu further extends this idea by bringing user’s social comments into visual via “synchronicity.” There is a need to develop a new

(30)

approach to study the intricate media participation in Danmu when textual comment directly overlays and informs the images. Therefore, this thesis will approach the case studies with a combination of visual analysis and discourse analysis in relation to Danmu’s media representation. The image-text relationship, intertextuality, and societal power relations of commenting will be addressed through visual-discourse analysis.

2.4 Methodology

This study focuses on the mainstream Chinese video sites Youku. A combination of platform analysis and visual-discourse analysis will be conducted to highlight different actors and their respective associations and power relationships based on actor-network theory.

As mentioned above, many studies only look at Danmu video sites such as Acfun and Bilibili without noticing that Danmu has become a default feature in mainstream video sites. This study would instead examine mainstream Chinese video sites in order to observe how the Danmu phenomenon has impacted the general Chinese media sphere as well as the different actors within Chinese internet culture and popular culture.

The video streaming site Youku 优酷 (Tudou is owned by the same company) is chosen as case study here. According to iresearch.com.cn which monitors the index of Chinese internet activity, Youku has an estimated monthly coverage of 254 million users. It was also the most visited China video site in February 2017, seconded by Iqiyi. Other major video sites such as Sohu, Iqiyi, and Letv are also potential research objects. However, many of these Chinese video sharing sites have imposed

(31)

restrictions on overseas IP address, thus limiting the current study which is conducted outside of mainland China. Youku is the one of the few Chinese video sites that can be fully accessed due to the Google Chrome extension Unblock Youku. It allows users to gain stable access to the site content.

As this study discusses the power relations among different actors within

Danmu, it is important that a wide, significant, and representative corpus is examined. Even though this study is able to gain full access to the Youku site, there are a number of issues narrowing the selection of corpus for study. Live broadcast videos of Youku has disabled its Danmu option, which means instant comments cannot be overlaid on the video. Danmu is also not available in certain categories and genres of video on Youku, such as news, education, and documentary, due to the sensitive nature of these videos in the Chinese context. Therefore, this study focuses on TV shows and movies offered on Youku, and they also have the highest level of audience engagement.

In total, six movies and four TV shows are selected for study in this thesis. The six movies include The Mermaid (2016), Ip Man 3 (2015), Big Fish & Begonia (2016), Lucy (2014), Snowpiercer (2014), and Kung Fu Panda 3 (2016). The four TV series are In the Name of People (2017), Running Man (2016), Surprise (2013, Season 1), and Gintama (2017, Season 8). More details are shown in Table 1.

(32)

Table 1: Names and Number of Views of the Chosen Cases for Study

English Title Chinese Title Description Number of views In the Name of People

(2017)

人民的名义 (2017)

Chinese TV drama series 8.4 billion views (Average around 150 million views per episode) Running Man (2016) 奔跑男女 (2016) South Korean variety TV show 1.3 billion accumulated

views

(Average around 25 million views per episode) Surprise (2013, Season 1) 万万没想到

(2013, 第一季)

Chinese web series 924 million accumulated views

(Average around 6 million views per episode) Gintama (2017, Season 8) 银魂 (2017, 烙阳

决战篇)

Japanese anime TV series 401 million accumulated views

(Average around 5 million views per episode) The Mermaid (2016) 美人鱼 (2016) Chinese fantasy

romantic-comedy film

337 million views

Ip Man 3 (2015) 叶问 3 (2015) Chinese–Hong Kong biographical martial arts film

254 million views

Big Fish & Begonia (2016) 大鱼海棠 (2016) Chinese animated fantasy film 104 million views Lucy (2014) 超体 (2014) English-language French action

film

86 million views

Snowpiercer (2014) 雪国列车 (2014) English-language South Korean-Czech science fiction thriller film 77 million views Kung Fu Panda 3 (2016) 功夫熊猫 3 (2016) Chinese-American 3D computer-animated action-comedy film 152 million views

Based on both popularity (view count) and diversity of the selected series, these 10 shows/films are selected to ensure there are abundant relevant examples for analysis. These shows are picked from their respective sub-genres, and the selected shows are all among the top 10 most viewed videos within their respective genres.

(33)

Three local Chinese films and three English-speaking foreign films are chosen to observe the difference in how users comment on local and foreign content. Due to the sub-cultural anime roots of the Danmu culture, two animated films are chosen to see if discussion in an animated film is different from that in a live-action film. The different national origins of the samples turn out to be contest subject within the Danmu discussion. It serves as an entry point to trace the action within multiple assemblages.

Among the TV shows, each of them is selected from the sub-genre of local drama TV series, foreign variety TV series, anime TV series, and web series respectively. This is to ensure that the cases being studied cover as much genre as possible (with a guaranteed level of popularity). Due to time limitation, this study is unable to view the entirety (full season) for these four running shows within the given research period. Therefore, it is decided that three hours per show are analyzed, and content from each show is selected from the beginning, the middle, and the end of their respective series/seasons. This also facilitates investigation of the difference in Danmu comments between different periods of the serialized shows.

Earlier studies usually focus on quantitative meta-data analysis of the Danmu video or conducting interviews and surveys that primary study audience responses. They omit the relationships between content, audience, and platform, and tend to treat them as separate subjects to be studied individually. Alternatively, this study presents a close reading on Danmu videos that takes audience, content, and platform all into consideration. It requires an in-depth content analysis to map out the connection among the different actors, and trace the action and association among them in specific contexts.

(34)

overlaying of Danmu comments. The aim is to purely focus on the content and visuals with the absence of Danmu comments. In the second viewing of the videos, Danmu comments are enabled and the researcher manually records the instances when specific Danmu comments show significant impact on the screen. These instances include Danmu comments causing specific reactions from other commenters, or if they have interacted with the content or platform interface in an unexpected manner. A visual-discourse analysis is done to see how content and audience interact and accommodate with each other in the unique image-text relationship of Danmu.

In the third and last viewing, the researcher records and copies the relevant Danmu comments. This is done by utilizing the comment copying function, and taking external screenshots which record the video content, audience interaction, and platform interface at the same time. In this step, platform analysis is conducted to explore the platform’s technical features, the Danmu comment, and the moving visuals of the original content. This final step also checks if there are any missing elements or unrecognized major actors in specific contextual moments.

Later on, Chapter 3 mainly focuses on discussing the results of the characteristics and features of the platform. Examples are mainly drawn from the materials Ip Man 3, Kung Fu Panda 3, and Gintama to discuss the spatial aspect of the Danmu platform, while examples from In the Name of People, The Mermaid, and Surprise are used to illustrate the temporal aspect of Danmu participation. Chapter 4 specifically centers on the example from the TV series Running Man along with minor examples from other shows to detail the contesting relationships among content, audience, platform, and a new key actor (nation state).

(35)

Chapter 3: Interruptive participation through “Synchronicity”

This chapter presents the findings of the platform features of Danmu and discusses the implication of innovative online participation. Through platform analysis, it is found that Danmu system introduces certain novel features that reconfigure the traditional online commenting space. These platform changes invite new forms of audience participation uncustomary to the traditional comment section of the Western mediasphere.

With regard to the concept of “synchronicity” discussed in the previous chapter, Danmu does not only bring together space and time, but also content, audience, and platform, creating a new symbiotic relationship between these different actors. This chapter will trace and discuss the different modes of participation in terms of the spatiality and temporality of the Danmu platform. The spatial and temporal changes in Danmu’s platform features and online commenting culture lead to a unique form of collective viewing. They foster a kind of interruptive participation and show a new connection among audience, content, and media technology.

3.1 Spatiality of Danmu

The result from the platform analysis shows two key spatial differences between traditional comment section and Danmu. Firstly, the traditional comment section is separated from the video content; secondly, within the traditional comment section, their line-on-line bulletin-board structure implies a strict spatial hierarchy. The presentation of Danmu undermines both of such spatial boundaries as seen in the traditional comment section, and allows synchronous and fluid reading of comments

(36)

during viewing. In traditional comment structure, the content is positioned at the front and center, being the most significant actor within the network. In Danmu, however, audience and platform share an equal amount of value and agency, which disrupts the traditional power relations in online commenting culture.

The traditional structure of the commenting space originated from the bulletin-board system (BBS) in the 1980s. Electronic bulletin bulletin-boards allow users to “gather, asynchronously, to read, consider, and possibly post replies” (James, Wotring, & Forrest, 1995, p. 31). In public message boards, lines of user’s text messages are placed on top of each other in chronological order. This provides the basic

architecture of the current comment thread. Blogs and news sites, starting from 1998, have gradually introduced commenting features in their platform layouts. From that point onward until today, the comment section refers only to a separate part of the website. With the advent of Web 2.0 the comment section has become a standard for most websites’ functionality. The evolution of the Web design determines that the commenting space is placed underneath the object being referred to. Accordingly, following the logic of the top-down arrangement of comment lines, the comment section submits to the spatial hierarchy of the Web page (Erard, 2013).

The first key spatial difference between traditional comment section and Danmu is that the former is usually separated within the web page and arranged in a relatively inferior position. Joseph Reagle’s book about online comments already implies this idea in its title Reading the Comments: Likers, Haters, and Manipulators at the Bottom of the Web (2015). Most studies about online comments reference the fact that comment sections are separated and placed below the original content. By placing the comment section in a relatively inconspicuous space, it is implied the comment section is less important than the original content. The comment section is often

(37)

treated as an extra, marginalized feature of social media platforms. User-generated comments are always supplements to professionally generated original content.

Scholars view comments simply as “paratext” of the screen. Literary theorist Gérard Genette (1997) defines paratext as semi-textual fragments that surround and position the work (the main text). The content of the paratext alone is not

self-contained and it is created to change or reinforce our interpretation of another specific media text. In this case, comment and “the very buzz or media talk that surrounds any given program” are considered as paratexts (Gray, 2003, p. 72). Due to the marginal positioning of the comment section at the bottom, many treat comments as paratext, something not as important as the original content.

The basic idea of Danmu challenges the above conception by reconfiguring the traditional web page layout. The redesigned space in Danmu counters the hierarchy within the comment section as well as the Web page itself. In the traditional video site layout such as YouTube, where the comment sections are located at the bottom or the side of the Web page, the original content is placed at the top. This normative spatial hierarchy of most social media platforms signals that the original content is always the most important element on the screen. However, by merging both the original content and the comment section into the same space, comments are treated as equally important to the original content in Danmu. Xu argues with reference to traditional Chinese art that “painting and prints are ‘nonhierarchical’ in that subjects in both the foreground and background remain in scale to one another” (Xu, 2016, p. 447). With the given positioning, screen time, and visibility, Danmu comments have the same level of power and significance communicated across the screen as the original

(38)

(comments)”7_{in multiple videos, it suggests user comments are more than just a}

“supplement” to the original content, and thus changes the power relation between content and audience. The comments can have as much, if not more, social

significance than the original content.

The second key difference is that Danmu has abolished the line-on-line system, and it provides users a fluid, heterogeneous, and customizable comment experience. Audience as commenters are allowed to set the position and movement of their comments, and as viewers the viewing preferences of these comments in the designated space. Danmu transforms the structure of the Web page as well as the structure of the comment section itself. This is unlike the traditional comment section which implies serious ranking and hierarchy within the bulletin board structure. As many comment systems (such as Facebook and YouTube) have adopted “top comments,” comments are no longer arranged chronologically but in default presented based on user-rating and up-votes. These two key differences show how Danmu has the potential to bring about spatial redistribution and resignification of commenting spaces, allowing the audience to interact with the content and other viewers in a new image-text relationship. Combining the result from platform analysis and visual-discourse analysis, it is shown that the spatial freedom and the level of customization invite different forms of interaction and participation unattainable in the traditional comment section. New creative opportunities and collective

engagement bring alternative forms of participation as shown in the following examples.

7_{Translated from: “我是来看弹幕的”}

(39)

Figure 2. “I am so green”

Figure 3. “I am Ip Man Green”

There is a scene in the film Kung Fu Panda 3 when all the characters are turned into green creatures, the Danmu blows up with green comments (see Figure 2). These comments scream, “I am so green,” “I am green too,” “Not as green as I am,” and “you all turned green hahaha.” Another example in Ip Man 3, when one random comment suddenly writes “I am Ip Man Green” (The Chinese word “Ip” also means “leaf”), subsequent comments follow with different colours such as “I am Ip Man Yellow,” “I am Ip Man Purple” etc.8_{(see middle section of Figure 3).}

(40)

These two examples show the Danmu comments interact with the visual content and other comments through customization within the display area, bringing the content together with the audience in the same space at the same time. Synchronicity brings all the actors together, and also the audiences together. These repetitive comments that copy previous ones can be regarded as a form of collective viewing. The audience engages and participates in the content and comments in a way that corresponds to other user’s participation.

Figure 4. Users Playing Laser Tags

Another example will further illustrate the idea of repetitive participation. In Figure 4, it shows a number of Danmu users playing laser tags in a martial art movie for no apparent reason. As an individual comments “red laser---,” it is quickly followed by similar comments of “yellow laser ---,” “white laser ---,“ and “shock wave)})).” These creative yet “meaningless” interactions among the audience is full of immature humor and repetition, and it is considered part of the fun viewing experience by Danmu viewers. These repetitive participation can be deemed as “meaningless” because they have no connection with the content at all. Without referencing to the visual content, the users create their own context with their Danmu participation. It brings in a new relationship between audience and content. As shown

(41)

in the above example, by playing laser-tags, the audience has subverted the original program for a brief moment, challenging the dominant screen presence of the content in that particular moment. The spatial freedom of the Danmu system allows the

audience to interact with each other with the help of the platform feature, undermining the content at the same time. It is also worth noting that this form of repetitive

participation is highly fragmented, which will be further discussed later in this chapter.

Figure 5. Anime Avatar in Danmu Comments

With regard to the spatial aspect of Danmu, other forms of customization also invite alternative modes of participation. In certain popular anime shows on Youku, their Danmu systems allow viewers to comment with their personal favorite character avatar. Shown in Figure 5 we see that commenters are expressing themselves as characters in the show Gintama. These avatar comments often repeat the lines in the show, laughing at certain characters as if the commenters are part of the show, and sometimes even talking to each other as if they were the characters. For example, an avatar comment representing the character Sougo would say “Kagura, I love you!”,

(42)

then a Kagura avatar comment would reply “I love you too Sougo!” At the same time, other Danmu comments (with or without avatar) would comment they like the idea that the two characters becoming a couple, while other comments reject the idea and ask them to stop “making this a thing.” In this example, we can see the fans are at the same time practicing role-playing, writing their own fan fiction, and discussing in the same space while watching the show. In Danmu, this complex, simultaneous narrative engagement in several fan activities within the same space is only possible through the notion of synchronicity.

Through this example, we can see specific platform features have increased the interactivity of commenting culture and encouraged fan participation in ACG culture. The exclusive avatar feature of Youku attracts and draws audiences to view and interact only on this platform. In this case, a new relationship between audience, content, and platform is established. The platform has altered and transformed the connection between content and audience; and without platform as an actor, the anime fandom assemblage and this new form of audience participation will never exist in the first place.

3.2 Temporality of Danmu

Apart from the spatial configuration, temporal boundaries of the online commenting have also been challenged in Danmu, which leads to more alternative forms of audience participation. First and foremost, traditional comments in the comment section are asynchronous in nature. In traditional commenting, one has to pause the video and then comment; more often, one reads a comment and has to go back to the video and see what it refers to. Danmu, however, completely reshapes this temporal aspect of commenting. Now one can see the specific comment for that

(43)

specific time frame, as the designated comment appears alongside the video at the same time. Music streaming platform Sound Cloud’s commenting system shares a similar idea that commenters are allowed to comment directly in specific time codes. The idea behind Sound Cloud’s design is to provide specific technical feedbacks and comments to the musician. Danmu pushes the idea even further. It extends the timely feedback feature to the daily conversation and also opens up other forms of

participation in video viewing.

One such example is observed in the Chinese drama series In the Name of People. It is a political drama about an anti-corruption task force combating high-level government corruption. As the show features rare and sensitive topics, without proper context and basic knowledge of how the Chinese government works, average viewers are often confused about subtle details of the show. However, thanks to the instant interactivity of Danmu, the audience is able to follow the show with better understanding. In one of the early episodes of the series, the show involves the idea of “Shuang-gui” (双规) in its main plot. Without getting into the specifics of the political terminology, the Chinese character 规 “Gui” means “rule” or “designated,” and “Shuang-gui” can be literally translated as “double-designated.” “Shuang-gui” is an internal disciplinary process within the Chinese Communist Party, meaning the suspect must non-publicly “be present at a designated time and designated location” and provide an explanation to party investigators.

Just Here for “Danmu”: Participation and Interruption in Chinese Online Comments