My notes on Brooker, P. (2019). My unexpectedly militant bots: A case for Programming-as-Social-Science. The Sociological Review. https://doi.org/10.1177/0038026119840988
In this thought provoking paper, Phil Brooker takes issue with the scaremongering surroundings bots which positions them as epistemically dangerous due to their quantity and capacity to evade deception. Instead he propose sociologists engage with them as both topic and resource, able to be used critically as a way of intervening in problematic online areas. This entails an upskilling by sociologists, without which any such engagement is likely to be at an abstract distance. He argues that “programming is a vital tool for responding to emerging issues and is potentially transformative across the social sciences in terms of how we understand and intervene in the world” (3).
Sociology lags behind fields like the digital humanities and software studies for reasons which would be interesting to explore but he offers this intervention as part of a broader project to help close that gap. He argues that we should think of “computer programming as a multipurpose toolkit for understanding and intervening in the (digital) social world in lots of different ways” (7). He begins with an exploration of the social character of design, infamously manifested in bots which reproduce the discriminatory features of the environment they take as stimulus. Drawing on Lupton’a work on design sociology, he draws attention to the affinity of design and sociology, with the former increasingly attending to the social context and the role of users within it. He argues that Di Salvo’s conception of adversarial design has particular relevance, with its focus on the capacity of design to disrupt preconceptions and knock users out of habitual ways of thinking.
His focus on social media bots responds to “a shift from depicting bots as insidious infections against which platform users must inoculate themselves, to a more nuanced understanding that reflects the different functions bots hold for those that build and use them” (5). The problem of bots is often framed in terms of computational propaganda such that they are seen to be fundamentally deceptive and manipulative, coming from without and requiring decisive intervention” (5), in spite of the useful and fun purposes bots have long served on sites like Reddit and Wikipedia. He argues that using design approaches instead “opens up a space to think about understanding bot–human interaction as playful, creative, positive and/or useful both culturally and sociologically” (6).
I won’t try to summarise the superb narratives of his two bots and their development. But he positions the first, which randomly generates Facebook updates from a core list of components, in terms of the politics of obfuscation which can be seen in projects like ISP Data Pollution, RuinMyHistory and Noiszy and their significance after the Cambridge Analytica revelations. He explores the latter in terms of the digital picket line, redesigning his Zenbot to proactively inform people it was on strike on the appropriate days, using the relevant hashtag and accompanied by random positive invocations, rather than dispensing zen wisdom on command via twitter. He uses these to signify the potential direction which what he terms *programming as social science* could take for sociology, particularly in terms of its capacity to it help open the notorious black boxes of algorithms and software. On pg 16 he notes other forms PaSS might take:
engaging with new forms of data, designing new methods to investigate social phenomena, developing interactive/alternative/engaging data visualisations, building applications to nurture innovative forms of interaction with research participants, and so on.
But a few steps are needed first. We need to acknowledged the lived experience of interacting with bots, as opposed to engaging in abstract generalisations. We need to grapple with the ethical challenges posed by these potential interventions, currently undertaken with a sociological blank canvas. We need to take bot design seriously as a method of generating social knowledge, currently being taken up almost exclusively by non sociologists. I consider myself enthusiastically signed up to Phil’s project here!
Mark Carrigan 