For a heterodox computational social science

Abstract

The proliferation of digital data has been the impetus for the emergence of a new discipline for the study of social life: ‘computational social science’. Much research in this field is founded on the premise that society is a complex system with emergent structures that can be modeled or reconstructed through digital data. This paper suggests that computational social science serves practical and legitimizing functions for digital capitalism in much the same way that neoclassical economics does for neoliberalism. In recognition of this homology, this paper develops a critique of the complexity perspective of computational social science and argues for a heterodox computational social science founded on the meta-theory of critical realism that is critical, methodological pluralist, interpretative and explanative. This implies diverting computational social science’ computational methods and digital data so as to not be aimed at identifying invariant laws of social life, or optimizing state and corporate practices, but to instead be used as part of broader research strategies to identify contingent patterns, develop conjunctural explanations, and propose qualitatively different ways of organizing social life.

How sharing is the “sharing economy”? Evidence from 97 Airbnb markets

“…This study takes a heterodox approach to studying Airbnb, using digital data for the critical examination of the societal implications of the platform [ 24 ]. Accessing the data of platforms such as Airbnb is challenging as the company does not share their data, implying that studies are forced to rely on web-scraped data.…”

Don’t 'Research Fast and Break Things': On the Ethics of Computational Social Science

“…This new access to the computational tools availed by potent compute and high-dimensional algorithmic machinery have led to the misconception in some corners of CSS that tools themselves can, by and large, "solve all problems". Rather than confronting the contextual complexities that lie behind the social processes and historical conditions that generate observational data (Shaw, 2015;Törnberg & Uitermark, 2021), and that concomitantly create manifold possibilities for non-random missingness and meaningful noise, the computational solutionist reverts to a toolbox of heuristic algorithms and technical tricks to "clean up" the data, so that computational analysis can forge ahead frictionlessly (Agniel et al, 2018;Leonelli, 2021). At heart, this contextual sightlessness among some CSS researchers originates in scientistic attitudes that tend to naturalise and reify digital trace data (Törnberg & Uitermark, 2021), treating them as primitive and organically given units of measurement that facilitate the analytical capture of "social physics" (Pentland, 2015), "the 'physics of culture,'" (Manovich, 2016), or the "physics of society" (Caldarelli et al, 2018).…”

“…These threats to the integrity of CSS research activity manifests in a cluster of potentially unseemly alignments and conflicts of interest between its own community of practice and those platforms, corporations, and public bodies who control access to the data resources and compute infrastructures upon which CSS researchers depend (Theocharis & Jungherr, 2021). First, there is the potentially unseemly alignment between the extractive motives of digital platforms, which monetise, monger, and link their vast troves of personal data and marshal inferences derived from these to classify, mould, and behaviourally nudge targeted data subjects, and the professional motivations CSS researchers who desire to gain access to as much of this kind of social big data as possible (Törnberg & Uitermark, 2021). A similar alignment can be seen between the motivations of CSS researchers to accumulate data and the security and control motivations of political bodies, which collect large amounts of personal data from the provision and administration of essential social goods and services often in the service of such motivations (Fourcade & Gordon, 2020).…”

Editorial: The Computational Analysis of Cultural Conflicts

“…As Global Warming is paralleled by rising temperatures in our political climate, we must ask how to manage the technology that determines our ability to engage in constructive collective political action. Heeding the calls to become global stewards for a sustainable environment requires first becoming stewards of the technology that shape our collective behavior – and for that, we must understand the complex relationship between media technology and cultural conflicts (Törnberg and Uitermark, 2021 ).…”