Polyphonic/Pseudo-synchronic: Animated Writing in the Comment Feed of Nicovideo

AbstractNicovideo is a popular video-sharing site in Japan that incorporates several aspects of social media into its design. Key among these is the projection of user-made comments into the video display by having text scroll across the screen like an animated subtitle-track. The movement of comments across the screen and the 'pseudo-synchronicity' created by the way they are projected produces a feeling of 'live' viewing via a sense of virtual time shared between users. In this article I argue that the feeling of movement and time on the site directs users toward a certain kind of vision that, when considered alongside the modes of counter-transparent communication taken up by its user base through things like orthographic 'mistypes', is part of a shift between denotational and pictorial forms of text production that troubles the distinction between reading and other modes of vision. The article conceptualizes what kind of vision Nicovideo's interface suggests and its relationship to a distinct kind of polyphonic, anonymous communication that intersects with ideas of animation and performance. It is particularly through the intensity of textual representation that I will pursue these questions. AcknowledgementsI thank the two anonymous reviewers who offered comments on an earlier draft of this article, as well as the editorial staff of Japanese Studies for their patience and guidance through the revisions process. Further gratitude goes to participants at the New Media Workshop and the Art and Politics of East Asia Workshop at the University of Chicago, as well as the audience of the 'Cinematic Diasporas' conference held at the University of Chicago on 14 April 2012, and the 'Critical Frameworks of Transmission' graduate student conference held at UCLA on 26 October 2012, for their questions and feedback. Special thanks also to Phil Kaffen and Shunsuke Nozawa for reading earlier drafts and offering detailed feedback.Notes1 The Japanese title of the video is Shin Goketsuji Ichizoku: Bonno Kaiho – Rettsugō! Onmyoji. The video of the dancing characters was taken from the Playstation 2 game Power Instinct series (Noise Factory, 2006). It was originally uploaded on 6 March 2007, and as of 18 March 2013 has had over 13 million views and over 4 million comments entered. The video capture being used for this paper was taken on 15 February 2013: http://www.nicovideo.jp/watch/sm92 Bakhtin, Problems of Dostoevsky's Poetics, 14. My reading of Bakhtin's notion of the polyphonic is indebted to Naoki Sakai's use of the term, particularly his gloss that the polyphonic allows the reader to become a 'participant'; Sakai, Voices of the Past, 26.3 Silvio, 'Animation'.4 My use of the phrase 'counter-transparency' borrows from Shunsuke Nozawa's concept of 'counter-spectacularity'. He develops the term to describe the appearance of masks and other forms of camouflage on Nicovideo that 'avoid full representation' of the seeming grossness of the human body. Here it is intended to capture the pronounced manner in which the mediated properties of written communication are put on display by users of this site via deliberate mistypes, to give one example. Nozawa, 'The Gross Face and Virtual Fame'.5 In highlighting the 'counter-transparent', I do not mean to assume other forms of language or communication to be transparent. Rather, I wish to make a precise claim about the particular way in which the mediated properties of written communication are rendered visible in these cases. In other words, while some forms of language may feel transparent due to conventionalized use, this form of image-oriented writing and use of deliberate mistypes draws out the possibilities of not being able to understand or of misunderstanding that I think should be recognized.6 The conventionalized use of misspellings such as 'pwn' instead of 'own' in online gaming or 'teh' instead of 'the' in ironic message board posting are two such examples of netspeak in English. However, while the mistypes in Japanese that I am looking at here depend on a deliberate mis-selection of a character on the basis of its phonetic or orthographic resemblance to another via the input method editor (IME), many common forms of netspeak in English are derived from the proximity of certain keys to one another on most keyboards and typing pads.7 For this paper I will focus on the properties of the comment feed rather than the types of videos that are common on the site. A more thorough engagement with the content of some of these videos is warranted, but is beyond the scope of this article.8 Videos are also categorized by user-created tags. However, as Hamano Satoshi demonstrates, the tagging system on Nicovideo is used not only for generic classification, but also for generating new ways of making connections between distinct texts that suspend singular authorship over a work to what he calls 'Nth order derivative' collaboration. This mode of generating meta-content for the videos is dependent more on an environment of users working together (and against one another) than on author-centric modes of creation. See Hamano, 'Niconico dōga no seiseiryoku', or the English translation, 'The Generativity of Niconico Douga'.9 This style of commenting was also originally made available for videos uploaded to YouTube which could then be watched 'through' Nicovideo's interface, but this was shut down by YouTube. Kanose, 'Garapagosu na Nihon no nettokai ni', 71.10 Hamano, Aakitekucha no seitaikei, 197.11 Ibid., 210, 235–236. Miyairi and Satō also compare Nicovideo's comment feed to an experience of 'liveness'. However, they phrase this as part of what they see as a shift away from direct, physical experiences of 'liveness', in which performers are on stage before a live audience, to ones in which 'seeing' and 'hearing' are less important than 'feeling', such as the redistribution of musical performance to social media. See Miyairi and Satō, Raibushiin yo doko e iku, 110–119.12 In addition, the content of a video's comment stream can change over time because the comment display system can only show a thousand comments during a single viewing, so as more and more users comment upon a video, older comments will be made invisible. These can still be accessed by changing the date on the comment feed, allowing the viewer to access older versions of the comment feed for a particular video. However, one of the most distinctive traits of the culture of the comment feed is how users attempt to recreate or repeat previous versions of a particular video's comments that have been rotated out in favor of newer comments. The video mentioned at the beginning of this paper is one such case, in which variations on the same comments are repeated over and over as the comment feed's maximum capacity continues to roll over. This has resonance with Hamano's comparison of the site to a concert that can be 're-experienced', although with the additional quality of requiring a kind of 're-performance' or 're-animation' by the site's user base.13 See Nozawa, 'The Gross Face and Virtual Fame' for more on pseudo-anonymous communication and performance in Japan, and Nicovideo in particular. A more in-depth comparison between the comment feed of Nicovideo and crowds in physical spaces is unfortunately beyond the scope of this article, but there are clear points of resonance in the anonymous mode of membership in both bodies as well as in how individuals within each body experience sensations of movement or motion, and the potential for escalation or intensification. For more on crowds, mimesis, and affect theory, see Mazzarella, 'The Myth of the Multitude'.14 Akada, 'Web Anonymity Nurtures Closed In-groups'. Natsuno Takeshi, an assistant executive at the company that manages Nicovideo, claims up to 63% of that age group; see 'Nico nico dōga no koa ni aru mono', 88.15 The performance center is called Nicofarre. It occupies the building that once held the Avex-affiliated discotheque Velfarre. For more on the opening of Nicofarre, see 'Raibu hausu Nicofarre ohirome'.16 Inoue, Vicarious Language, 130.17 In some ways this can be compared with Soundcloud, a music-sharing website that incorporates the comments of listeners into the moving 'Waveform' time-track of a music file. However, while Nicovideo's comment feed has text that moves across the screen and appears and disappears as the video plays on, the comments on Soundcloud are static and become highlighted as the time-track cursor (or the user's mouse cursor) moves over them. This creates less a sense of pseudo-simultaneity than a kind of archive that the listener 'waits' to see appear while seeing the cursor move over each comment marker. Similarly, with the exception of the replies to individual comments (which appear as miniature 'threads' in the time-track), comments on Soundcloud appear one at a time, which also creates a very different feeling of a social community that is virtually 'watching together'.18 David Atkinson and Helen Kelly-Holmes' discussion of code-switching between English and Irish language speech in radio comedy from Ireland provides a useful comparison for thinking about fragmented understanding and the kinds of performance of community that it can invite. The radio performers and audience for these programs are not native speakers of Irish, so their command of the language is often limited and on the verge of not understanding, but the use of the Irish vernacular also helps produce a kind of linguistic in-group or even performance of border patrolling against exclusively English-speaking communities. The counter-transparent modes of communication on Nicovideo are different in that they do not involve spoken language or switching between two languages. However, the possibility of fragmented understanding as a mode or performance of community seems to be a strong point of intersection between the two groups. Similarly, the kinds of typographic play that use phonetic or orthographic resemblance in writing have some resonance with verbal code-switching, although those similarities are beyond the scope of this paper. Atkinson and Kelly-Holmes, 'Codeswitching, Identity, and Ownership in Irish Radio Comedy'.19 2channel or '2chan' is a popular Japanese-language message board that was started in 1999. Perhaps most widely introduced to the public at large via the numerous adaptations of the story of Train Man (Densha Otoko) in film, animation, and drama, the site also had a public profile associated with internet nationalism, practices of trolling and flaming, and a strict form of anonymity. For more on 2chan, see Kitada, Warau Nihon no nashonarizumu.20 Laura Miller has written about a similar kind of 'bad' writing in regard to young women using non-standard writing practices (such as borrowing characters from non-Japanese alphabets) in texting and purikura ('print club') photo-captioning. See Miller, 'Those Naughty Teenage Girls'.21 Murakami and Itō give an overview of some of the wordplays in typography used in relation to music-related internet videos in two essays: 'Dōga saabisu ni okeru shichōsha komento no bunseki', and 'Dōga tōkō saito de fuyo sareta dōga tagu no kaisōka'.22 For similar practices on 2channel, see Nishimura, 'Establishing a Community of Practice'.23 These forms of writing, of course, have many similarities with texting, in which mistypes are also used as an 'in-group' communication technique. One commonality that I would like to highlight here is the use of various forms of typographic play as a way of simulating some of the qualities of expressiveness in speech, such as tone, volume, or character. Tanaka Yukari has described the use of virtual dialects in phone messaging as a way of 'encoding intimacy' via textual simulation of spoken language, and we can find other strategies of encoding community in the textual play seen on Nicovideo. Tanaka, Hōgen cosupure no jidai, 9.24 Matsuda, 'Netto shakai to shūdango'. Sites that summarize popular threads from 2Channel (matome) will sometimes use different colors and size of fonts to emphasize well-received or funny posts.25 Patterson, 'POEMFIELDs and the Materiality of the Computational Screen', 245, 258. There is, of course, a long history of mechanical text production, engaging in multiple forms of content, that extends beyond the examples Patterson is looking at. However, what I am trying to draw out of her reading is not a historical connection between POEMFIELDs and Nicovideo, but rather a conceptual approach for thinking about the activity of reading and its relationship to problems of media determinism.26 Rodowick, Reading the Figural, 62–63.27 Because comments are shown without attribution it is difficult to say anything about the identities of individual users and what kinds of comments they might make or what videos they might watch. There have, however, been some surveys for age demographics among users of Nicovideo. Data from 2012 show that 75.9% of users who post comments to the feed of popular videos (those that rank on the site's internal voting system) are between the ages of 10 and 20, which generally seems consistent with the orientation of the site toward games, anime, and idol music. See, for example, 'Niconico dōga no comment no nenrei bunseki shitara', and 'Niconico dōga 3-oku comment nenrei bunseki'.28 Hosoma, 'Uta o sodateta kanaria no tame ni', 35.29 Video available at: http://www.nicovideo.jp/watch/sm15745067. This capture is based on the video feed from 16 October 2011. This video shows the meta-interactive potential of the comment feed quite clearly: the performance of the dancer is augmented by that of the comment artist, who is essentially producing a new variation on the original dance/video by supplementing the dancer's choreography with the images he or she is inscribing over the video.30 Video address: http://www.nicovideo.jp/watch/sm3520730. Capture taken on 1 September 2012.31 In that sense comment artists have a privileged position within the discourse of the site in that, although comments are not attributed to individuals, the role they fulfill is still one that commands a kind of prestige. For example, while users do sometimes respond to other comments, comment artists are typically the only ones to enjoy lots of personal attention or praise, sometimes even more than the content of the video and its performers. This presents an additional complication in the way individual representation is partially suspended in anonymous environments. However, these comments are still made without attribution, and the shift toward this mode of intense textual representation or performance is precisely what I am trying to articulate here.32 Gunning draws primarily on Christian Metz's essay 'On the Impression of Reality in the Cinema' and Henri Bergson's writing on movement and affect to develop this point, also making clear the difference between the perception of 'realism' in film and the materiality of reality. Gunning, 'Moving Away from the Index', 42. Gunning's language of 'participation' via vision also has resonance with Sakai's gloss on Bakhtin's notion of polyphonic narration as a way of allowing for participation by the reader.33 This is a gloss on Silvio's summary of these concepts. Most work on animation as a kind of movement of images adopts a more concrete definition, but I find Silvio's abstraction of animation to a kind of environmental vision and sense of relationality to be very insightful. Silvio, 'Animation'.34 Ibid., 427.35 Marks, 'Calligraphic Animation', 309.36 Silvio, 'Animation', 433.37 Some of the other 'A-s' in A-culture include acceleration, adolescent, anti-, and anarchy. Auerbach, 'Anonymity as Culture'.38 Manning, 'Barista Rants about Stupid Customers at Starbucks'.