Swimming against the data stream: plot, polyphony and heteroglossia in data-driven writing
Conference presentation proposal for ELO 2014 “Hold the Light”
Unprecedented access to real-time social data is changing the way we tell stories about ourselves. Social data is being utilised within a wide variety of electronic literature and media art from Mark Hansen and Ben Rubin’s Listening Room to the recent explosion in Twitter bots which remix digital text. These practices have been designated under the rubric “environmentally interactive” digital writing (Wardrip-Fruin 2010, p. 41). Such writing often takes the form of a data stream (Manovich 2012), representing content as a chronological flow of units of information, with the newest information being most salient.
In this presentation, we propose to examine an opposing tendency which narrativizes data rather than representing it as a stream. Such work mounts a challenge to the claim by Manovich (2000) that narrative is being displaced as a symbolic form in new media objects. Examples range from the 2013 generative novel-writing competition NaNoGenMo, which featured a number of long-form works produced via data remixing, to older projects such as the Impermanence Agent by Noah Wardrip-Fruin et al. We will share excerpts from our own work which has narrativized social data, including Enquire Within Upon Everybody, a public art project presented at the International Symposium on Electronic Art 2013; and the art installation Everything Is Going To Be OK :).
The split between data stream and narrative reinscribes the famous formalist distinction between raw story elements or fabula, and plot or syuzhet (Shklovsky 1965). Works representing data as a stream typically present a time-ordered fabula; they are closely aligned with the conceptual poetry tradition, which emphasises verbatim transcription. By contrast, narrativized works freely reorder data to suit the purposes of plot construction and often interpolate fictive text. The two types of work also sharply diverge in how they are consumed. The former tend to invite audiences to browse, skim or adopt a distant reading posture (Moretti 2013); some purely conceptual works are not even intended to be read at all (Goldsmith 2011). The latter often invite readers to engage closely and thoroughly.
We will also draw attention to two other formal qualities of data-driven storytelling: polyphony, the deployment of multiple characters’ perspectives alongside that of the author (Bakhtin 1984), and heteroglossia, the use of multiple patterns of speech and discourse to express authorial intentions (Bakhtin 1981). Originally identified in the context of the 19th century novel, these emerge in far more radical forms in data-driven digital works, which often mediate the voices of living human subjects. We will show how polyphony and heteroglossia are enabling a unique interplay between authorial voice and the voices of real-life characters.
We contend that long-form, data-driven narrative holds exciting promise and merits further exploration. It allows readers to engage with data not in a distant or discontinuous way, but to experience it as an immersive story. It also enables authors to offer novel insights into the interrelationships between data; to explicitly critique dominant discourses and not merely recontextualise them; and to construct metanarratives which help us make sense of the stories around us.
(Source: Author's introduction)
