Abstract
John Cheney-Lippold identifies as a man in his thirties, but Google thinks he is a sixty-five-year-old woman. He writes, “online you are not who you think you are” (p. 6). In We Are Data: Algorithms and The Making of Our Digital Selves, Cheney-Lippold uses a cultural studies approach to investigate our “algorithmic identities,” the digital shadows that flex and flux in reaction to “near-real-time interpretations of data” (p. 26). This book is fresh, innovative, and well-researched. It offers a theoretically informed and highly sophisticated account of the ways algorithms help constitute our world, making this a key addition to the burgeoning scholarship on algorithms. Indeed, We Are Data stands in contrast to the many books that offer celebratory, vacuous, and ahistorical accounts of big data and algorithms, that fail to attend to how algorithms operate in regard to gender and race, identity, and power. Cheney-Lippold is part of a generation of scholars—including Tarleton Gillespie (Gillespie, Boczkowski, and Foot 2014), Kate Crawford (2013), Frank Pasquale (2015), Mark Andrejevic (2007), Wendy Chun (2011), and Alexander Galloway (2004)—who build on the work of pioneers—including Donna Haraway (1988), Lisa Gitelman (2013), Katherine Hayles (2002), Friedrich Kittler (1992), Tiziana Terranova (2004), Lev Manovich (2001), and Lisa Nakamura (2007)—working to bridge this gap in the digital studies literature.
The central argument of We Are Data is that algorithms—equations that process data drawn from many various aspects of daily life—increasingly shape the world we live in, what we know about ourselves and each other. This book has two main strengths. First, it theorizes algorithms as formative agents in the construction of identity, in particular gender, race, class, and citizenship. Cheney-Lippold shows us how algorithms create datafied subjectivities “layered” into our other subjectivities, how algorithms scaffold onto—as opposed to contest or subvert—our lived experiences. Second, this book develops the concept of the “measurable type,” which he contrasts with Erving Goffman’s (1959) “ideal type.” Measurable types—such as “woman” above—consist of aggregated and transcoded data organized into “digital containers of categorical meaning” (p. 19). The measurable type is a useful concept that describes the ways in which algorithms are in perpetual reflective flux, meaning algorithmic categories collapse offline and online categories in a way that “disallows any clean conceptual separation between life as data and life as life” (p. 11). In short, measurable types are “not truths but betas” (p. 90) in that they describe categories based on the data available instead of static norms or taxonomies. Measurable types are restricted in that they only measure that which is accessible and quantifiable. Their measurement never reaches a finite end, or “100 percent certainty” (p. 90).
Individual chapters in this book address the roles algorithms play in identity categorization, the control of knowledge, the formation of subjectivity, and the conceptualization of privacy. Each chapter provides rich cultural examples that Cheney-Lippold then analyzes and theorizes. For example, his fourth chapter begins with the story of a British disabled man who called emergency services to report severe pain. After a series of questions generated by an algorithmic triage system, the operator advised the man not to call an ambulance. The man died. Cheney-Lippold uses this example to investigate the ways that our bodies become visible and invisible to computerized systems such as the labor and cost-saving automated diagnosis systems implemented as part of the U.K. government’s post-2008 austerity measures. This ill man failed to fit into a measurable type read by the algorithmic system as in need of emergency aid, a failure that resulted in his death. As this example illustrates, layered algorithmic identities and measurable types matter because they are increasingly the default ways through which we become legible and visible. Indeed, one of the central arguments in We Are Data is the scale to which they matter: if we are to understand our health care, our politics, our markets, ourselves, we must also understand algorithms and data.
Cheney-Lippold seems alarmed yet also ambivalent about what he studies. For example, the ambiguity of our online identities—perhaps straddling both “man” and “woman”—holds the potential to “signal a great diversity of who we are and can be, even as different technologies of power desperately try to wrangle us into preexisting boxes of identity” (p. 32). At the same time, as the conclusion articulates, the modes of classification Cheney-Lippold tracks create an “emergent power relation” characterized by tremendous “asymmetry” because the “world is no longer expressed in terms we can understand” (p. 252). Data collection occurs, of course, in ways that are “almost never transparent” through the uncompensated labor of users (p. 257). And this lack of transparency serves particular interests. Data sets are in themselves meaningless. They must be made meaningful. Therefore, we are “made subject not to our data but to interpretations of that data” (p. 258). By identifying data sets as neither neutral nor natural but “manipulated” and “massaged,” Cheney-Lippold directly critiques the social scientific notion that we can “let data speak for itself” (p. xiii). He shows how power flows through those who claim they let data “speak.” Using Michel Foucault’s (2003) biopolitics and Gilles Deleuze’s (1992) societies of control, he persuasively articulates how measurable types are central to describing the contemporary dynamics of power as they flow through algorithms, humans, and institutions, as they are reflected and reformulated. Cheney-Lippold concludes his book raising the twin specters of capitalism and government. He notes capitalism’s history of “colonizing life with its own logic and motives” (p. 253) and the U.S. government’s efforts to leverage data in service of its “global geospatial strategies” (p. 262).
I conclude with a note about this book’s delicate balance between readability and research. Cheney-Lippold’s writing is approachable and at times even funny, but not simplistic. On the contrary, he interweaves humor with the work of heavy hitters such as Judith Butler (1993), Hannah Arendt (1998), and Walter Benjamin (1969). We Are Data is descriptive, theoretical, and politically engaged, yet also lively and clear. Those writing in this field know how difficult that is to achieve. Cheney-Lippold makes it seem effortless.