Reinventing Audiovisual Contents as Seen in Asian Amateur Video-editing Cultures: Driving Your Characters MAD


Student thesis: Doctoral Thesis

View graph of relations


  • Qian CHEN

Related Research Unit(s)


Awarding Institution
Award date11 Dec 2015


This research investigates the current phenomenon of meticulously crafted and fragmented video compilations created by mass collaborative video-sharing communities, especially but not only in Japan and China. These videos, which originated as a manifestation of a Japanese medial subculture, formally resemble but are genealogically disconnected from either the historic avant-garde video assemblages dating from the 1960s (e.g., Bruce Conner) or the college-oriented MAD creative conventions in Japan since 1980s. Such acts of video compilation, which a few fan communities proclaimed as “kichiku”, most often appropriate samples from video footage from a wide range of other sources (from anime to television programs to western gay porn), for purposes of social exchange and interpretive dialogue concerning visual characteristics of the compilations themselves as well as the surrounding contexts of interface and culture. With emergence of these videos aligned to characteristic performances and architectures of the subculture’s receptive habitats i.e. the NicoNico Douga-style danmu web video-sharing communities, this research postulates that the kichiku compilations entail the digital-age zeitgeist of modernity and openness of works. But contrary to the more typically self-conscious détournement of many video collages, kichiku compilations become legible and gain their remarkable popularity (e.g. this example was once the most viewed video on Nico Nico Douga, the most popular video-sharing community in Japan) by functioning in the reception process not primarily as autonomous content but as nodes of hybrid networks of textual, audiovisual and database-centered articulation and metadata. The intense formal hybridity and almost infinitely recursive authorship of these works may be unprecedented. On the aspect of cultural meaning making, this research suggest that the kichiku compilations are more prone to be interpreted as metadata for para-narratives than as video mashups despite their structural resemblance to the latter. By preserving the independence of the characters in the original and then “deterritorialize” them with stereotypical labels of attributes (属性), these compilations serve to apply Azuma’s semi-transparent “database of characters” model of otaku narratives to wider spectra of proto-narrative elements such as the more classically narrated and filmed originals for kichiku. In the conclusive phase, this research advances to project the case of kichiku reception onto the macroscopic contemporary media landscape of “web offset” reproductive tensions where conflicts between the database and the character are stressed, by extending Benjamin’s theoretical model of artistic mechanical reproductions. Based on the recognition that receptive criteria of kichiku alleviate the web offset tensions with their specificities for both categories of the database and the character, this research confirmed the prospects of some circumstances of kichiku reception as precursors to a generic solution of the contemporary web offset tensions.

    Research areas

  • kichiku video compilations, video remixes, MAD, Kyara and Characters, semi-transparent realism, NicoNico Douga (NND)