Submit a Manuscript to the Journal
Distinktion: Journal of Social Theory
For a Special Issue on
Social Theory in an Age of Machine Learning
15 April 2022
Social Theory in an Age of Machine Learning
Machine learning (ML) systems, designed to extract patterns in data and to make predictions on that basis, are permeating a growing number of spheres in society. For example, predictive ML algorithms are increasingly being deployed in fields such as credit scoring (Fourcade & Healy, 2017; Kiviat, 2019; Rona-Tas, 2020), insurance (Cevolini & Esposito, 2020), criminal justice (Brayne, 2017; Brayne & Christin, 2020), self-driving cars (Bissell et al., 2020; Stilgoe, 2018), social media (Fourcade & Johns, 2020), warfare (Scharre, 2018), and automated trading(Hansen, 2020; Hansen & Borch, 2021), as they are at the core of modern advertisement-centered capitalism (Zuboff, 2019).
A large and growing body of literature has examined the effects these types of systems have on society. For example, the proliferation of ML systems has raised alarms about their potential biases (Zou & Schiebinger, 2018), the social inequalities they may breed (Eubanks, 2018; Noble, 2018), as well as the ways in which they transform everything from subjectivity and everyday life (Elliott, 2019; Wajcman, 2019)to labor markets (Gray & Suri, 2019; Rosenblat, 2018; Shestakofsky, 2017). Similarly, scholars have discussed the opacity of ML systems as well as the broader epistemological, ethical, political, and theoretical implications the deployment of such systems have, including what their use might mean for human-anchored conceptions of accountability, expertise, liability, and so on (Amoore, 2020; Brighenti & Pavoni, 2021; Burrell, 2016; Coeckelbergh, 2020; Collins, 2018; Dubber et al., 2020; Fazi, 2020; Pasquale, 2020).
Alongside these kinds of studies, and partly fueled by them, there is a growing appreciation that the rise of ML might have a profound impact on social theory. On the one hand, the use of ML as a new methodological tool holds the promise of detecting patterns in data which might prompt a rethinking of established notions used to describe the social world. Although this promise may yet remain unfulfilled, some scholars are confident that ML’s potential to extract non-linear patterns in data endows it with theory-generating powers superior to those of previous tools (Edelmann et al., 2020; Evans & Aceves, 2016). On the other hand, the very functioning of ML systems engenders a reconceptualization of human-centered social theory (Esposito, 2017). After all, in some domains, the actionable predictions of ML systems not merely inform human decision-making but replace it entirely. This sets them apart from previous, human-defined algorithmic systems and opens an array of questions concerning the accountability, control, ethics, liability, and politics of ML systems—but also concerning the ways their use may reshape human-machine configurations (Suchman, 2007)as well as the potential ability of inter-machining dynamics to produce social relationships of some form.
Against this backdrop, the aim of this special issue is to bring together papers that address the social theory implications of ML. We are particularly interested in two types of submissions (or combinations thereof):
- Papers that discuss how key social theory notions and approaches may need rethinking considering the ways ML systems operate. How, in other words, do the inner workings and/or actionable predictions of ML systems prompt a rethinking of established notions in social theory? Addressing this may be accomplished in both pure theory papers and papers that study ML systems empirically—for example, by focusing on their modus operandi and domain-specific applications.
- Papers that deploy ML techniques to generate novel social theory insights. How, in other words, may the data-driven inference of ML methodologies in sociology, political science, and beyond present novel insights into sociality and the social world and lead to new social theory considerations and/or open up new ways of conceiving the meaning(s) of theory (Abend, 2008)? Addressing this may be accomplished in either methods-driven or empirically oriented papers.
Note that, although we welcome both empirical and theoretical submissions, all papers must tease out their theory implications.
We welcome papers that focus on particular ML architectures (for example, deep learning) as well as papers that discuss the broader workings and implications of ML-driven systems. This special issue welcomes submissions from the disciplines such as geography, law, political science, and sociology as well as from scholars working on related issues in the humanities.
- The deadline for submitting full papers is 15 April 2022.
- Papers should be no more than 12,000 words, inclusive of tables, references, figure captions, and notes.
- This special issue is edited by Christian Borch, Copenhagen Business School. Please send any inquiries about the special issue to Christian Borch at [email protected].
View the latest tweets from Routledge_socio