Abstract
At the end of a special issue dealing with methods for datafication and the datafication of methods in media and communication research, it seems appropriate to reflect on some of the prominent themes on the current media and communication research agenda studying the implications of datafication. One candidate for this kind of reflection is the broad notion of digital disconnection, which has received quite a bit of scholarly attention lately (e.g. Kaun and Treré, 2018; Portwood-Stacer, 2012; Syvertsen, 2020). Sometimes the practice and fashion of disconnection is motivated by individual users’ sense of losing control to personal devices that lure them into constant availability and incessant checking of notifications and apps to pass time, instead of doing more valuable and meaningful things, whatever they may be. But the idea of disconnection also speaks to more fundamental critiques of datafication raised in scholarship, about the implications of the ongoing datafication of human existence for human agency, equality, autonomy and meaning.
The two timely books reviewed in this review essay specifically intervene in debates over disconnection, both pushing us beyond the perhaps most familiar point of view of individual media users. Karppi’s analysis of disconnection evolves around the basic idea that disconnection is a problem to tech companies, and especially social media. He carefully unveils how disconnection is dealt with from the perspective of a core player in the global data economy: Facebook. Couldry and Mejias throw a wider net, examining what they call data colonialism as an emerging social order conditioned on digital infrastructures in the broad sense. They present a sociologically and historically informed analysis of the structural, political-economic and ideological forces driving data colonialism and its implications. Both books offer critical accounts of the contemporary state of data extraction and commodification of human life. They urge us to combat it, as it fundamentally threatens human flourishing and integrity. And they offer complementary analytical perspectives to guide our ongoing efforts to make sense of datafication and assess what Couldry and Mejias call ‘the costs of connection’.
In his book, Disconnect. Facebook’s Affective Bonds, Karppi takes the reader through seven chapters each detailing specific aspects of what he labels an ideology of connectivity and its alleged antidote, disconnection. It focuses on how disconnection works, for whom it works, rather than what it is, and seeks to challenge dominant presumptions about the benefits of disconnecting. Karppi poses disconnection as a critical problem for social media platforms, ‘a problem of control, governance, and design’ (p. 7) – of keeping users engaged and contained in the platform’s system of value-creation. As such, Karppi joins forces with important voices in software studies in taking a non-human approach to the question of disconnection from social media, one that focuses on the engineering of specific affective bonds to lock-in users, to gain profit and enhance data power to condition future possibilities for us all. Karppi’s analysis is centred on Facebook, the dominant social media platform of the past decade (and still so, at our present moment). He draws on research materials collected over a 7-year period, starting around Facebook’s initial public offering in 2012. His material spans Facebook-related texts, such as marketing material, software applications, artworks, financial and legal documents, developer documents, journalistic texts and so on, all of which are used to offer compelling analyses of the problem of disconnection and the quest for commodifying user behaviour to an ever-greater extent. However, the main attraction in Karppi’s discussion of disconnection is a rich theoretical repertoire, spanning the classic works of Kittler, Tarde, Foucault, Deleuze and contemporary scholarship in the areas of critical data studies, software studies and STS. It is not an easy read, but a carefully crafted philosophical encounter with the affective bonds generated and orchestrated by social media platforms.
Over the course of seven chapters – log in, engage, participate, deactivate, die, disconnect, log out – Karppi takes the reader through detailed and deep analyses of how Facebook’s risk management reflects in the design of user experiences through the platform. If Facebook capitalizes on user engagement and participation, captures and turns it into marketable data, then user churn presents an existential crisis for Facebook. Karppi argues that ‘[. . .] to survive, Facebook needs to expand and intensify’ (p. 39) user engagement with reference to recent examples of Facebook developments such as the ‘Buy button’ and ‘Payments in Messenger’, and to Facebook’s significant efforts to expand connectivity infrastructure in emerging markets. The first three chapters offer illuminating analyses of Facebook’s efforts to keep users active on the platform. The latter four chapters turn the lens to its efforts to keep users from disconnecting in various ways to retain control. Through numerous compelling examples, Karppi alerts us to the constraints posed by Facebook on disconnection, showing how difficult (if not impossible) it is to deactivate or delete an account. The database of each individual user remains. Even death is commodified through memorial pages. Indeed, Disconnection shows that questions of agency are tricky and complex in datafied infrastructures, and complicates the simple binary of connection/disconnection. In concluding his discussion of various forms of disconnection (detoxing, migrating to alternative platforms and so on), Karppi notes that disconnection in fact breeds Facebook’s ubiquity. The threat of disconnection forces the platform to pursue new avenues for tracking and datafying users. Being everywhere-from websites to GPS navigators to video game consoles and so on- is the strategy employed by Facebook to mitigate the risks of disconnection from the platform. The result: users are engaged with Facebook even when they are not using the core social media platform, and this leads by extension to an expanding colonization of our networked environment.
In The Costs of Connection. How Data is Colonizing Human Life and Appropriating it for Capitalism, Couldry and Mejias specifically develop a theory of data colonialism to describe the pervasive tracking and monetization of social life that has come to characterize contemporary datafied societies across the globe. Complementing Karppi’s almost exclusive focus on Facebook, Couldry and Mejias set out to draw the larger picture of the capitalist and colonialist underpinnings of datafication as practised today, with the aim of raising our collective awareness and capacity for changing the path. In that sense, the book shares the core ambition of one of the most significant and publicly acclaimed analyses of datafication in the past decade, namely Shoshana Zuboff’s (2019) The Age of Surveillance Capitalism. Yet, arguably, by advancing the lens of data colonialism and drawing heavily on Marxist social theory, Couldry and Mejias have a more radical critique of capitalism in mind, one that historically ties it to colonialist efforts an appropriating, exploiting and controlling resources, redistributing benefits and spreading specific ideologies. They assert, ‘it is becoming increasingly clear that capitalism’s current growth cannot be captured simply in terms of ever-more ambitious business integration or the ever-expanding exploitation of workers’ (p. 4). What is instead at stake, argue Couldry and Mejias, is a shift in the raw material that capitalism is appropriating and controlling: it is human life itself. Mundane activities of personal life, work, education, banking, healthcare and so forth are all datafied. The data, and their value and terms of use, are controlled by parties external and often unknown to us – social media platforms, governments, data brokers, IoT businesses, cloud companies and so on, collectively dubbed ‘the social quantification sector’.
The book is organized in three parts. The first part, comprising chapters 1–3 (and an interlude on colonialist theory and the decolonial turn for readers new to this field), develops the theoretical framework of data colonialism – an emerging social order that naturalizes limitless data extraction and surveillance and cultivates ideologies that celebrate connectivity, personalization and ‘dataism’, lending a term from a prominent European media scholar, Jose Van Dijck (2014). Chapters 1 and 2 further develop a set of concepts for animating the analysis of data colonialism’s operations and its new order, which, the authors argue, is premised on developments in technological infrastructures over the past decades. Importantly, data colonialism is not just a western project. Liberal-democratic and authoritarian states compete for leadership in the game of data colonization. Nodding to this fact, Couldry and Mejias draw their many compelling examples from around the globe. The aim of part 1 is to trace the continuities between historical colonialism and today’s data colonialism and thereby better tease out the kinds and scopes of inequality, injustice and knowledge/power-relations it embeds. Like Karppi’s Disconnect, the major strength of the argument lies in a rich theoretically driven narrative that weaves together multiple strands of classic social theory – from Marx and Foucault to decolonial theory – and connects them with contemporary analyses of data justice and the legal-commercial complex regarding personal data.
Part 2, comprising chapters 4 and 5, examines the social consequences of data colonialism, specifically ‘the hollowing out of the social’ (p. 115) and the changing grounds of knowledge production in the age of datafication. The social quantification sector is characterized with reference to its privileged private capacity for the production of social knowledge, its opaqueness, its fetish of prediction rather than explanation and the pervasiveness and intimacy of its knowing that stems from the quest to collect and convert everything as data for later use. The implications are critical for human autonomy and self-integrity, justice and so forth. They are exemplified through an acerbic analysis of the quantified self-movement and its ‘know thyself’-mantra. Data colonialism, the authors go on, is also consequential for social and humanistic sciences. These are at risk of being sidelined by ‘social physics’, and other computationally based calibrations of science values and epistemologies.
Part 3 recapitalizes the book’s overall argument and sketches Couldry and Mejias’ vision for decolonizing data, posing data colonialism not as an individual problem to be fixed with disconnection, but as a collective challenge. The authors call for resistance, regulation and rejection of the ideologies and data uses that do not directly benefit the data subjects delivering input to the systems. And they turn to decolonial perspectives, among others, to ignite a participatory and civic research agenda where research moves beyond the confines of commercial constraint and corporate control, and where experimentation and collaborative engagement in decolonizing data is a collaborative endeavour. The catch-22 is, of course, that ‘there can be no benefits from the use of platforms that don’t at the same time reproduce, at root, the very power we want and need to oppose’ (p. 193).
The critical agenda sketched by significant contributions like the ones reviewed here will certainly stimulate further theorizing of and active resistance to datafication, premised on the fact that individual disconnection is futile. But it also suggests that an empirical commitment is needed as a way of moving forward, as implied by Couldry and Mejias’ call for participatory and locally grounded projects. Raising awareness to harness collective responses for change implies continued efforts to empirically map the data flows and ownership structures, the forging of affective bonds through design to lock-in users, inequalities built into automated decision-making systems and regulatory effects with respect to data extraction, mining and monetization. Such empirical charting of datafication, its methods and epistemological implications, may gain inspiration from the questions and research trajectories discussed in this special issue, to put visions of decolonization into practice. It is urgently needed to concretize what, when and how to exercise resistance, regulation and rejection to mould alternative and better human futures.