Symposium Introduction

It is my pleasure to introduce this symposium on Colin Koopman’s important book. As the effects of the corporate use of data and the age of “surveillance capitalism” (Shoshana Zuboff) becomes more and more explored within our world, Koopman’s book is a particularly welcome one to the extent that it locates this notion of data in genealogical terms, showing how its precursors stretch back beyond where its origins are oftentimes located.

This is a useful procedure to the extent that it allows us to situate and anchor these questions in broader discussions that emerge in modernity. Where it is common to conceive the various problems raised by big data and by us becoming understood chiefly in terms of data, How We Became Our Data shows how these discussions in fact connect to earlier movements in modernity around conceptions of subjectivity and agency, around the emergence of racial thinking, and around discourses of power and sovereignty. Of course, none of these multifaceted discourses have been absent, but it is incredibly useful to have them unified in the genealogical procedure that Koopman undertakes in this book. Framing things in this way also allows Koopman to develop a notion of infopower, which is meant to be located amidst other (Foucault-inspired) discourses of power.

In what follows you will read responses by Dan Smith, Jen Forestal, Corey McCall, and Verena Erlenbusch-Anderson.


Infopower, Formatting, and Inscription

The counterpoint to Colin Koopman’s brilliant How We Became Our Data is perhaps Michel Foucault’s own 1977 introduction to a never-produced book called The Lives of Miniscule Humans. Foucault had considered putting together an anthology chronicling the lives of ordinary people who had disappeared from history completely except for brief entries entered into the bureaucratic records of a prison, asylum, or hospital—entries that Foucault would stumble upon while working in the archives of the French national library: “Jean Antoine Touzard, placed in the castle of Bicêtre, 21 April 1701: ‘Seditious apostate friar, capable of the greatest crimes, sodomite, atheist if that were possible; this individual is a veritable monster of abomination whom it would be better to stifle than to leave at large.’”1 These lines are the only data that remains of the life of the unfortunate Jean Touzard, although no doubt such fleeting archival records were already a step above the fate of most humans—the degree zero in which we die without leaving behind the slightest trace of our existence.2

Koopman’s book reminds us how far removed we are from such a world. In Foucault’s imagined book, data was produced primarily by institutions charged with caring for people considered to be “abnormal”—the diseased, the criminal, the mad. Today, in an age of information, we live in a regime of “infopower” where the data on each individual is massive (“big data”) and the scourge of abnormality has instead fallen on “undocumented” people, the sans papiers. The aim of Koopman’s book is both to analyze this new apparatus of infopower and to trace its genealogy.

Koopman’s argument is that infopower is a distinctive modality of political power that is exercised through the technique of formatting, an operation that serves to “fasten” subjects so tightly to their data that “we have become our data” (ix). For Koopman, this claim means that data can now determine what it is possible for an individual to be or do. Depending on our data, a financial transaction will be approved or blocked, a college admission will be accepted or rejected, entrance to a building will be granted or denied, a job application will be successful or unsuccessful, a border will be crossed or not. Data can both “tie us down and speed us up”; it can canalize future possibilities (63) as much as it can accelerate our existence (as in the constant stream of notifications we receive “as an elicitation for ever more engagement” [13]). Koopman is one of the few philosophers who has taken the notion of formatting and turned it into a rigorous philosophical concept.3

How We Became Our Data presents a history of our present, an analysis of the roots of the regime of infopower that has supplanted—or rather coexists with—the regimes of biopower and disciplinary power that Foucault himself analyzed. The information age, Koopman shows, began to appear long before the development of the World Wide Web, or the computer, or even the development of information theory in the late 1940s (Weiner, Shannon, Turing), which are the touchstones for traditional histories. Rather, Koopman shows that the regime of infopower coalesced in the beginning of the twentieth century, when “information began to precede the person” (6) as if it were a new infrastructure, a new “historical universal” (10) into which subjects were inserted. Humans were becoming “informational persons.”

Part 1 of the book—which is a stunning tour de force—proposes a Foucauldian genealogy of three new forms of data-based identity that constitute the informational person: documentary, psychological, and racial identity. Like Foucault’s work, Koopman’s analyses are deeply grounded in archival detail, and one might say they give a new meaning to “empiricism” in philosophy. The first chapter traces the formatting of the birth certificate, the first document that fastens individuals to specific data points, such as their race, sex, name and occupation of the parents, and even the “legitimacy” of their birth. Bookended by one’s death certificate, the birth certificate became an individual’s entry point into an ever-expanding network of informatics: social security numbers, bank statements, email accounts, drivers’ licenses, passports, university transcripts, market transactions, email accounts, genetic reporting, and on and on (44, 155).

More surprisingly, the second chapter (66–107) shows how the concept of “personality” coalesced in the early twentieth century out of the endeavor to “objectively” measure the psychological traits of humans. One result of this endeavor was an astonishing book by the primary proponent of “psychometrics,” Gordon Allport, that compiled an exhaustive list of 17,953 English-language trait names—the human psyche formatted into big data.4 “Intelligent” was one of these trait-names, and the early 1900s was the era in which intelligence testing came into vogue as a method of measuring the intellect. Koopman’s radical claim is that, strictly speaking, individuals did not have “personalities” until this attempt to format human psychology came into being, much as Arnold Davidson argued that there had been no perverts until the concept of perversion, and its corresponding mode of being, was formatted in the latter part of the nineteenth century.5

The remarkable third chapter, on the seemingly mundane topic of real estate, shows how racism in the 1930s and 1940s, though officially disavowed, wound up being embedded in the algorithms of redlining, which resulted in severe racial segregation in housing, banking, and education, an Eichmann-like convergence of systemic racism with bureaucratic banality. The chapter is the apotheosis Koopman’s empirical analyses, since it shows that the entire structure of infopower in which the information person exists is inevitably racist. I suspect it will become a standard text in the theory of race.

In each of these chapters, Koopman shows himself to be a worthy successor to Foucault. He does not merely repeat what Foucault said, but does what Foucault did: patient, detailed, and philosophically informed historical research to which these brief summaries can hardly do justice. “The discursive deeds of grand theory,” Koopman writes, “gain their force through the practical elaboration of tiny techniques” (71). Koopman has weaved an extraordinary narrative from the work of the largely unknown technicians who produced the regime of infopower by devoting themselves to the “minutiae of formatting” (161).

In part 2 of his book, Koopman then turns to the necessary question, How should we respond politically to the regime of infopower? to which the fifth chapter (“Redesign”) gives a nuanced and forceful response. Infopower has opened up new possibilities, to be sure, but it has also facilitated “injustices, inequalities, and unfreedoms” (154). One is reminded of Melvin Kranzberg’s dictum that “technology is neither good nor bad; nor is it neutral.”6 Koopman persuasively shows the limitations of theories of deliberative democracy based on communicative proceduralism (Habermas, Rawls), since such theories ignore the informatics and formatting processes upon which they rest, and without which deliberation and communication would be impossible (184–87). What we need, Koopman argues, is a more profound politics of formats, a politics that not only addresses the formation of information, but the ways in which formatting occurs and the means by which information is stored, processed, compared, repurposed, distributed, and so on (182). In reflecting on Foucault’s notion of resistance, Gilles Deleuze noted in 1990 that the strikes and sabotage of factory work had given way to hacking and viruses,7 and one wonders what new forms of resistance would emerge in a politics of formats: Alternative types of formats? A jamming of current formats? A struggle against the very activity of formatting? The lawsuits against the current tech giants (Amazon, Apple, Google, Facebook) are only the surface manifestation of a formatting politics that is spread deeply into the warp and woof of the social fabric, and we can only hope that Koopman will develop a politics of formats in more detail in a future work.

In the fourth chapter (“Diagnostics”), finally, Koopman attempts to situate his analysis in a broader philosophical context. Koopman’s empirical analyses are largely restricted to the early twentieth century, roughly the period from 1913 to 1937, just as Foucault’s early works were confined to the classical period. He takes pains to show how the regime of infopower must be distinguished from the regimes of biopower, disciplinary power, and sovereign power that Foucault himself analyzed, even if they overlap and intermingle. His discussion raises several issues that, in my mind, point to fertile directions for future research.

First, Koopman seems to have opened the door to a broader understanding of the history (and even prehistory) of infopower. He approvingly cites Lisa Gitelman’s declaration that “new inscriptions signal new subjectivities” (6), and her concept of inscription is perhaps one manner of approaching that history.8 Nietzsche, for instance, had argued that the inscriptions that initially documented our identity were made directly on the body: a circumcised penis, a scarified forehead, or a tattooed body marked one as a Jew, a Nuer, or a Maori (mnemotechnics).9 Are these marks, which are inscribed directly on the surface of the body, any less a form of infopower than marks that are inscribed on a piece of paper? With the invention of writing, inscriptions were externalized and able to be “captured” by states and their bureaucracies. Moreover, teaching the techniques of inscribing marks and interpreting them (literacy) became one of the fundamental purposes of educational institutions, to the point where being “illiterate” implied that one is both uncivilized and unintelligent. Money, or capital, is itself a form of inscription, pieces of data added to or subtracted from the accounts of banks and firms. The practices analyzed by Koopman obviously presuppose these “prior” forms of infopower, and recent books have analyzed how inscription has been further transformed by the advent of digitalization, such as Cathy O’Neil’s Weapons of Math Destruction (on the role of automated algorithms)10 and Seth Stephens-Davidowitz’s Everybody Lies (on the human psychology revealed in Google data).11 Koopman himself points to the work in media studies (Kittler) and elsewhere that has started to assess the continuities and discontinuities of this prehistory of infopower, with all its singularities (writing, literacy, printing, computers, etc.).

Second, although Koopman adopts Foucault’s method of genealogy, for instance, one might argue that what Bergson once called the “retrograde movement” of knowledge is also at play.12 It is often said that, in the modern world, there have been three ages of machines—mechanical, energetic, and information machines—and each of these machines has been used as a model for comprehending nature, or objects in nature. In the seventeenth century, the idea of mechanism arose in part from the model of the watch: the world is like a machine with internal mechanisms that explain its functioning. The same happened in the nineteenth century, when the invention of the steam engine led to the development of the science of thermodynamics, and workers came to be seen as human motors with an energetic capacity or “labor power” that could be quantified and optimized (Taylorism, Fordism).13

Today, the computer seems to have become a model for almost everything, from genetics (the genetic “program” or “code”)14 to the mind (our brain is the hardware, and the mind is the software, running different programs in different modules). These are not mere metaphors or analogies. Nature is an organization of matter, and technical artifacts (machines, motors, computers) are ways in which we have learned to organize matter. Since we have a “maker’s knowledge” of our artifacts—a knowledge from the inside, as it were—we use that knowledge to comprehend the organizations found in nature, including human organizations.

Yet once technical models such as mechanism, energetics, and information emerged, it was inevitable that they would be used, in a retrograde movement, to understand the past. In his essay “Kafka and His Precursors,” Jorge Luis Borges catalogues a number of earlier writers whose works exhibit Kafkaesque elements but otherwise have little else in common. But such authors can begin to look Kafkaesque only once we have read Kafka himself. “The fact is that each writer creates his precursors,” Borges concludes, “his work modifies our conception of the past, as it will modify the future.”15 The same would seem to be true of information: the advent of informatics has modified our conception of the past. Koopman rightly suggests that the informational “precursors” he analyses were on the cusp of achieving consolidation or stabilization (176–77). But might it be equally the case that their “conditions of appearance” (26) lies in the fact that we retrospectively recognize them as informational precisely because of the later consolidation and stabilization of informatics?

Third, this leads to a last point about historicity and temporality. Koopman rightly rejects an analysis of the relation between infopower and biopower (or any other formation) in terms of simple succession or “temporal eras” (171). Succession is itself a form of temporality that derives from sovereign power, since it measures time in terms of the succession of sovereigns (“in the fifteenth year of the reign of Tiberius Caesar”). For Koopman, it is not as if infopower succeeded biopower, disciplinary power, or even sovereign power; rather, infopower inserted itself into these formations. He adopts a suggestion of Verena Erlenbusch-Anderson that replaces succession with “interpenetration and superimposition” (172). But this seems to presume a conception of temporality in which time is not successive but rather coexistent. In Capitalism and Schizophrenia, Deleuze and Guattari proposed the provocative thesis that archaic states such as Babylon and Egypt could not and did not from simpler “primitive” societies; rather, such social formations existed side-by-side in a single field of coexistence, which alone can account for their intermingling and interpenetration.16 Whether or not their solution is adequate, Koopman’s work has renewed the question of how to think of history in terms other than succession.

No doubt these are metaphysical issues of the type of Koopman rightly eschews (10, 236n84). But they point to the fact that, in its focus on the concrete practices of infopower, the implications of Koopman’s outstanding work, like all great works, goes far beyond its stated intentions.

  1. Michel Foucault, “The Lives of Infamous Men” (“La Vie des hommes infimes,” 1977), in Power, vol. 3 of Michel Foucault: The Essential Works, ed. James D. Faubion, trans. Robert Hurley et al. (London: Penguin, 2000), 157–75: 158. I’ve obviously altered the translated title: infime seems better translated as “miniscule,” “insignificant,” “minute,” or even “infinitesimal.”

  2. Similarly, in the Iliad, Achilles was told by his mother that he had a choice between two fates: the renown bestowed by an early but glorious death or the obscurity of a long but ultimately forgotten life (Homer, Iliad, book 9, lines 410–15). Achilles chose the former, and thereby achieved what the Greeks called κλέος (kleos, “glory, fame, renown”), which literally meant “what others hear about you,” whether in oral poetry or, later, written texts. Achilles’ renown stemmed less from the greatness of his exploits than the fact that they could be recorded as bits of data, morsels of information, that could be passed on to future generations.

  3. The only other philosopher I know of who has attempted to develop a philosophical concept of formatting is Michel Serres. See, for one example, his Branches: A Philosophy of Time, Event, and Advent (2004), trans. Randolph Burks (London: Bloombury, 2020), chapter 1, “Format-Father,” 3–33.

  4. See Gordon Allport and H. S. Odbert, “Trait-Names: A Psycho-lexical Study,” Psychological Monographs 47.1 (1936) i–71. Koopman’s reference so intrigued me that I found a copy of the book to peruse the trait-names. As a typical example, the entries for the letter S begin with the follow traits: “Sabbatarian, sacerdotal, sacrificatory, sad, Sadducean, sagacious, sage, sailorly, saintish, saintlike, saintly, Samaritan, Samson, sanctified, sanctimonious, sang-froid, sanguinary, sanguine, sannyasin, Sapphic, sarcastic, sardonic, satiable, satiric, satisfiable . . .”

  5. See Arnold I. Davidson, “Closing Up the Corpses,” in The Emergence of Sexuality: Historical Epistemology and the Formation of Concepts (Cambridge: Cambridge University Press, 2001), 1–29: 22.

  6. Melvin Kranzberg, “Technology and History: ‘Kranzberg’s Laws,’” Technology and Culture 27.3 (1986) 544–60: 545, doi:10.2307/3105385.

  7. Gilles Deleuze, “Postscript on Control Societies” [1990], in Negotiations, 1972–1990, trans. Martin Joughin (New York: Columbia University Press, 1995), 177–82.

  8. Gilles Deleuze and Félix Guattari similarly argue that “inscription” is one of the fundamental activities of any social formation. See Anti-Oedipus: Capitalism and Schizophrenia (1972), trans. Robert Hurley et al. (Minneapolis: University of Minnesota Press, 1983), 142: every social formation is “a socius of inscription where the essential thing is to mark and be marked.”

  9. Friedrich Nietzsche, Genealogy of Morals, trans. Walter Kaufmann (New York: Vintage, 1972), essay 2, §3, 60–62.

  10. Cathy O’Neil, Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy (New York: Crown: 2016).

  11. Seth Stephens-Davidowitz, Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us about Who We Really Are (New York: HarperCollins, 2017).

  12. Henri Bergson, “The Retrograde Movement of the True Growth of Truth,” in The Creative Mind, trans. Mabelle L. Andison (New York: Philosophical Library, 1946).

  13. See Anson Rabinbach, The Human Motor: Energy, Fatigue, and the Origins of Modernity (New York: Basic Books, 1990).

  14. See Alexandre E. Peluffo, “The ‘Genetic Program’: Behind the Genesis of an Influential Metaphor,” in Genetics 200.3 (2015) 685–96,

  15. Jorge Luis Borges, “Kafka and His Precursors,” in Selected Non-Fictions, ed. Eliot Weinberger (New York: Viking, 1999), 363–65.

  16. On the priority of coexistence over succession, see Gilles Deleuze and Félix Guattari, A Thousand Plateaus, trans. Brian Massumi (Minneapolis: University of Minnesota Press, 1987), 473: “the coexistence and inseparability of that which the system conjugates.”

  • Colin Koopman


    Reply to Daniel W. Smith

    I am grateful to all four respondents who have assembled these engaging responses to How We Became Our Data. A number of crucial themes criss-cross between their pieces as they pose challenging questions central to both the book’s account of the power of data and the background methodology operative in that account. I would like to begin with a few thoughts in response to Dan Smith’s very welcome provocations as I believe this will help establish a useful frame for the three exchanges to follow with Jennifer Forestal, Corey McCall, and Verena Erlenbusch-Anderson. Smith’s response also offers a welcome introduction in its concise summary of the book’s main claims and arguments. I have nothing to disagree with in his summary and find it a helpful description of what I was attempting in the book.

    In the course of the summary portion of his response, Smith pays me two enormous compliments, though I am not sure he intended them as quite as significant as I take them. The first is his observation that the book “does not merely repeat what Foucault said, but does what Foucault did: patient, detailed, and philosophically informed historical research.” Foucault’s genealogy is indeed my major philosophical orientation in the book, but Smith is right that I use Michel Foucault the way a novelist uses George Eliot or an essayist uses James Baldwin. The serious writer uses the writers that precede them as a model for what, with great effort and even greater luck, they themselves just might be able to pull off in some other form. But there would be no point in attempting to state again (at least at the length of a book, leaving to the side the scholarly task of the exegetical article) what Foucault has already said about discipline and biopower, just like it would never occur to any serious novelist to simply reproduce the iconic characters in Middlemarch or any serious essayist to merely rehearse Baldwin’s penetrating insights in “Nothing Personal.”

    This brings me to Smith’s second compliment. He claims that my book’s analyses “give a new meaning to ‘empiricism’ in philosophy.” This is significant because genealogical philosophy is not often labelled empiricist. And yet one reason I go to it is because of its empiricist ambitions. Smith is right that an empirical genealogy would require a new conception of empiricism, that is one no longer beholden to a simplified view of knowledge as an automated accumulation of sense data. Perhaps it is the empiricism worked out, to mention in passing a philosopher Smith and I both admire, Barry Allen, and his new book Empiricisms (Oxford University Press, 2021).

    To give a sense of why I think an empirical-genealogical orientation matters when it comes to any pressing present issue, consider a further comment of Smith’s later in his response. He notes as a key theme of the book its “conception of temporality in which time is not successive but coexistent.”

    Smith’s point about temporal coexistence is framed by him through the Deleuzian idea that “social formations” are never totalities but rather always exist “side-by-side in a single field of existence.” The cartographic correlative to Smith’s chronometric observation is that space, like time, is coextensive. Said differently, when philosophy analyses its fields, it should not assume as its methodological postulate that there is a totalizing system that exhaustively saturates any given space. The political philosopher should not assume that a given space is completed saturated by a single political system. The epistemologist should not presume that a given field of rationality is exhaustively characterized by a single epistemic formation. Every space is already a conjunctive association of a multiplicity of coexisting formations. I take it that this, in part, is what Deleuze and Guattari were focused on with their signature concept of agencements (often translated as “assemblages”), Latour with his notion of associations, and Foucault with his concept of dispositifs.

    The glue that holds this methodological approach together is not the internal rationality in a unified social totality that waits for a philosopher to intuit it. The approach holds because of the “empiricist” commitment discerned by Smith in which there is an unceasing ambition to inquire ever further into the heterogeneity that is the political. Multiplicitous time and space can be countenanced only by the energetic empiricist. The rationalist by contrast insists on a unity and totality that can be smoothed into a beautiful idea.

    Let me now situate Smith’s insight (or at least my adoption of it) in the landscape of contemporary political philosophy. There is a crucial methodological question facing the political philosopher through which we can refract Smith’s point. Is political philosophy about the totality of the system or is it about regional operations of power?

    The former approach is represented by much of classical political theory, and is more recently clearly exhibited in John Rawls’s theory of justice, which takes as its object of scrutiny what he calls “the basic structure of society.” The idea, and it is not at all unintuitive, is that society itself is held together as a single system by a political core that forms its “basic structure” in such a way that this structure represents the total universe of a society’s politics. This approach can also be fairly attributed to Jürgen Habermas, for whom the totality is focused more temporally rather than spatially, and is construed in terms of the political trajectory of “modernity” construed as a systemic whole. It is crucial to such views that the total system can be rationally apprehended. Rawls was well aware of this in positioning his own view as “ideal theory.”

    But not everyone accepts this approach. It is already clear from the above that formations of political power are construed regionally by Foucault, Deleuze, and Latour (and one certainly sees something similar at work in political agonists such as Laclau, Mouffe, and Connolly). In more analytic anglophone political philosophy, a serious challenge to Rawls’s systemic approach has been on the agenda at least since Michael Walzer’s pluralism in his Spheres of Justice (Basic Books, 1983). And we have recently witnessed what is probably the most forceful challenge to the presumption of unified totality as it operates across German critical theory. I refer to the work of Rahel Jaeggi, specifically her Critique of Forms of Life (Belknap/Harvard, 2018 [2014]), in which “the evaluation of forms of life opens up a broad and inclusive field of practical questions that cannot be subsumed under the narrower domain of questions of relevance for morality or justice” (5). (It is an interesting further question whether rationalism furtively redoubles in some of these projects, for instance in Jaeggi’s argument that segmented practices are always subject to an “organizing principle” [62] in virtue of which they, in good Hegelian fashion, can be determinately subsumed under their “concept” [118ff.].)

    In the context of a critical inquiry into present and contemporary informatics, the need for a non-totalizing mode of inquiry is particularly pressing. This is because the object of inquiry itself, information, would have us accept that it is a universal. Information today travels so well that we are bound to want to see it anywhere, literally anywhere, that we can go. In How We Became Our Data, I adopted an approach that does not seek to refute this common conception of information as a universal, but nor does my approach intend to affirm it. Rather I aim to interrogate its genealogy—I aim to ask how it came to be true.

    The universality of information romps wildly across the terrain of the contemporary. It roams nearly everywhere. But it does not roam and romp in the same way in each of its domains. It does not operate identically across high-tech genomics research laboratories and lower-profile civil service recordkeeping operations. Thus do we need an empiricism that can track its many movements and motions. Information may be a universal of our contemporary. But precisely for that reason we cannot sustain the fantasy that it is an idealization. Forms of political inquiry rooted only in idealizations, for instance those that interrogate only the conceptual dimensions of a certain idea of information, leave to the side so much of what is of interest when practices manage to mobilize nearly everywhere.

    When something like information manages to universalize itself, what is being distributed across so many regions of social practice is not just a tiny idea, but rather a multitudinous melange. This melange includes concepts, to be sure, but also much more, such as technical, infrastructural, somatic, and aesthetic conditions for practice. These elements can be mobilized across multiple segments or regions of social practice. Thus can, for example, the melange of infopolitics both chronologically and spatially overlap with disciplinary anatomopolitics in a way that does not rely on a classical conception of a singular sequential time.

    The more rigorously-empirical approach of the genealogist is appropriate for, indeed necessary to a serious regard for, the melange of politics. This is not to say that idealism is useless. It is rather to say that its utility is restricted, and perhaps most especially when it is trained on that which (like information) is unrestricted. But political philosophy ought to be useful for making sense of politics, which is almost always constituted of a heterogenous and rowdy mangle that only a brisk empiricism can keep up with.


Infopower & Democracy

Commentary on How We Became Our Data

At a time when it seems that everyone is concerned with how our data is used—lamenting the new (and problematic) dynamics introduced to this end by digital technologies—it may seem quaint to turn to the pencil-and-paper past for insight into algorithmic decision-making. Yet, as Colin Koopman argues in How We Became Our Data, it is precisely these often-obscured historical moments that reveal to us how data has shaped our very understanding of who we are. In providing us with the new vocabulary of infopower, and tracing its genesis through certain moments in the history of data collection and processing, Koopman gives us a compelling account of how—and why—we can and should rethink the work that data does to us, rather than focusing solely on what we do with data.

I very much appreciated this book. Koopman’s focus on the early decades of the twentieth century is a welcome corrective to the all-too-common insistence that we understand digital technologies (and their effects) as wholly novel. And Koopman’s genealogical analysis of infopower—in particular, its constituent parts, formatting and fastening—is exceedingly useful for developing a fuller understanding how, exactly, data works on us today. As a result of this investigation, as Koopman argues in the book’s concluding chapters, we are better equipped to deal more explicitly with the politics of information—to “confront,” as Koopman puts it, “information itself as a political problem” (22).

Koopman’s engagement with this political problem largely takes the form of his critique of deliberative democracy for overlooking infopower. Deliberative theory, Koopman argues—especially work in the Habermasian tradition—“ignores information as a site of politics” (187). While these scholars have focused primarily on processes of communication—which presuppose already-formatted information—Koopman instead suggests that a “normative political theory of information” might also investigate the “processes of information design that any and all communication presupposes” (189).

Certainly, Koopman is right that democratic theorists should pay more attention to the critical questions his genealogy introduces: namely, “How is this information [communicated in deliberations] formed? How is it formatted? What burdens are embedded in those formats?” (187). Indeed, similar questions (especially the last) have been lobbied at deliberative democratic theory for years, by thinkers such as Iris Marion Young, Lynn Sanders, and Arthur Lupia and Anne Norton, who all call attention to the restrictiveness of “rational deliberation.”1

But deliberative democracy, while useful, is not the only way to think of democratic politics. Indeed, there is another longstanding—and complementary—tradition that treats democracy not only as a communicative practice, but also as a “way of life” that is built out of social habits, attitudes, and mores.2

Thinking of democracy in this broader sense, I want to suggest here, might help flesh out the political significance of infopower, which Koopman rightly directs us towards. Indeed, I think it is worth considering that there is something distinctively democratic about infopower. And examining deeper relationship between infopower and democracy has implications for not only why we find infopower so compelling, but also how we might resist its negative potentialities.

I. Understanding Infopower

One of the important contributions of this book is Koopman’s novel concept of infopower. Infopower, he argues, “deploys techniques of formatting to do its work of producing and refining informational persons who are subject to the operations of fastening” (12). It is through infopower, in other words, that we become “informational persons”—we become beings who are understood by, and through, our data.

Importantly, Koopman characterizes infopower as a new—and distinct—mode of power. Though he acknowledges that infopower is “layered on the biopolitical, disciplinary, and sovereign powers characteristic of a more familiar moment in modernity” (189), Koopman nevertheless wants to keep these terms conceptually separated. Consider his comparison of sovereign power with infopower. While “a sovereign,” Koopman argues, “garishly expunges, or extinguishes, that which is deemed impermissible . . . a format leaves everything as it was such that a subject that is not formatted according to its terms is not committed to nonexistence but only consigned to either try again or go its own way” (168). Whereas traditional sovereign power operates through “force and brutality” (167) to prohibit certain actions, infopower does not. If someone, or something, fails to fit the format, they do not disappear; they simply become unreadable.

Yet this distinction is perhaps a bit too neat. Consider, for example, the University of Virginia’s trouble in August 2020. The university required students to test negative for the novel coronavirus before returning to campus; yet the website used to order such a test failed to list certain states, like Rhode Island and New Jersey—meaning students from those states were unable to comply, as their information did not fit the format.3 Likewise, until 2016 the Common Application required students to select either “male” or “female” before advancing to the rest of the form; students whose gender identity differed from those two options were excluded from consideration.

These are not examples of a Leviathan handing down mandates; rather they are, or at least appear to be, mere accidents of formatting. But they nevertheless do seem, in important ways, to be acts of “prohibition and permission”—perhaps more subtle and insidious in their restrictiveness, but nonetheless producing real political effects of exclusion and erasure.

And Koopman himself seems open to this interpretation. While he makes clear that infopower is “neither a variant of nor reducible to biopower or discipline” (14), he nevertheless leaves a bit less daylight when it comes to sovereign power. “If the power of information is at all expressive of sovereign power,” he notes, “then it can only be expressive of a decidedly altered shape of sovereignty. It would be a power that dispenses entirely with techniques of violence and logics of prohibition and permission” (168, emphasis mine).

But sovereign power does look decidedly different in a democratic society. In reading Koopman’s description of infopower, I was struck by its parallels with Alexis de Tocqueville’s description of democratic tyranny. In democracies, says Tocqueville, “the master no longer says: ‘Think like me or you die.’ He does say: ‘You are free not to think as I do; you can keep your life and property and all; but from this day you are a stranger among us. . . . You will remain among men, but you will lose your rights to count as one. . . . Go in peace. I have given you a life, but it is a life worse than death.’”4

What Tocqueville reminds us here is that sovereign power in a democracy is of a “decidedly altered shape.” Rather than the “clumsy weapons of chains and hangmen”5 the democratic sovereign exercises power over thought. The power of forbidding here is not explicit, nor is it particularly violent; instead, we all end up censoring—or, in Koopman’s terminology, formatting—ourselves just to fit in and be recognized by society. And if we fail to do so, we are simply “consigned to either try again or go [our] own way.” Might infopower, then, be the newest manifestation of this democratic power—another way to solicit obedience through social pressure rather sovereign fiat?

II. Enacting Infopower

Approaching infopower from this Tocquevillian perspective—rather than Koopman’s Foucauldian one—can, I think, help to uncover additional dimensions to the specifically democratic politics of infopower.

For Koopman, information’s power is found “in its promise of, and success at, effective universalization” (11). Information seems universal because it is everywhere always. The political project that Koopman takes up in the book, then, is to trouble this claim and uncover how historical and contingent forms and formats were mobilized in a way that “made information into such a powerful universal” (10, emphasis mine). By interrogating the genesis of infopower, Koopman argues, we are better equipped to resist it.

This genealogical analysis clearly details how these individual instances helped to consolidate infopower as an organizing force in modern life; it is only in turning to Tocqueville’s writings about democratic subjectivity, however, that we can find an explanation for why we all so willingly went along.

Democracy, Tocqueville reminds us, engenders a love of equality: “The particular and predominating fact peculiar to [democratic] ages,” he says, “is equality of conditions, and the chief passion which stirs men at such times is the love of this same equality.”6 Democratic subjects, in other words, cannot tolerate difference—indeed, their love of equality is such that “they will put up with poverty, servitude, and barbarism, but they will not endure aristocracy.”7

And as a specifically democratic mode of power, infopower is perhaps the most advanced means for equalizing democratic subjects. Consider Koopman’s discussion of social media profiles. These, he argues, “pin us down to prefab formats, categories, and conceptions that we then readily tie ourselves to” such that the “social media profile that is produced on the basis of all this processed information looks more or less the same for us all” (13). The homogeneity enforced by the profile’s format ensures that no one user will, or can, rise above the rest. Formats may “tie us down” (12), in other words, but in so doing they also bind our neighbors—and it is in that equalizing effect that we find infopower’s enduring appeal. While we might prefer equality in freedom, we’ll happily take the restrictions it imposes, provided they apply the same to us all.

What Tocqueville reminds us, then, is that infopower’s universalizability is not simply the result of luck or chance. While the emergence of infopower is, without question, a historical specificity, it is also one motivated in large part by the psychology of democratic subjects. It is, in other words, a specifically democratic problem. Motivated by our love of equality, we keep mobilizing infopower because it makes, as Jaron Lanier puts it, “the pack mentality as efficient as possible.”8

III. Resisting Infopower

What does all of this mean for our prospects of resistance? Koopman’s Foucauldian analysis of infopower leads him to a rather Foucauldian answer: resistance consists in “occupation, contestation, and transformation . . . a politics of critique” (193). And this critique, Koopman suggests in the final section of the book’s final chapter, should be largely aimed at experts and elites working in technocratic spaces like a “design lab . . . a code studio . . . a tech incubator . . . an engineering firm . . . a seminar table around which regularly gather a small team of collaborators . . . somebody’s mom’s garage” (194). These are the “formatters, designers, and developers [who] build the information systems that form the basements beneath our lives” (194); they are, as a result, rightly the focus of a political project aimed at resisting infopower’s potentially restrictive effects.

This approach is no doubt valuable. We should be critical of those who “format our futures” (194), asking incisive questions “about the politics of today’s designs as they are becoming tomorrow’s formats” (195). And Koopman certainly gives us a good list of questions with which to start this work.

But approaching infopower as a specifically democratic mode of power raises additional considerations when thinking of resistance. As a tool of the democratic sovereign, infopower relies on the cooperation of many to be successful. And it is infopower’s capacity to equalize that makes us all the more happy—even eager—to cooperate and oblige. Infopower, in other words, is a power wielded by the masses over ourselves, as evident in Koopman’s examples of the (volunteer women) who deployed birth registration at scale, the people who took self-assessed personality tests, and the real-estate-agent-to-be who turned to accessible manuals to learn racialized appraisal algorithms.

Taken together, these considerations force us to direct our critique beyond the information architects who create formats. Rather, we must also look to those who deploy formats daily, in often-mundane ways, and ask ourselves how we might act collectively to counteract or resist their effects. Approaching infopower as a democratic problem, in other words, directs us to also consider how we are all cooperative and complicit—even when we are not the ones doing the developing.

As members of university departments, for example, we regularly deploy formats to collect, track, and evaluate students, peers, and potential hires, though we likely do not create those formats ourselves. How might we ensure that the equality imposed by these formats is productive rather than destructive? How do we resist the tyrannical consequences of equality while still making collective decision-making possible? And, perhaps most importantly, what are the collective dimensions of this project of resistance? It is, after all, only through collective efforts that we can hope to successfully ‘harness’ infopower in the name of democratic freedom, rather than fall prey to the democratic despotism Tocqueville so feared.

How We Became Our Data gestures toward these questions but it does not provide us with answers. Nor should it; as a democratic problem, infopower requires democratic solutions. But by excavating the roots of infopower, Koopman provides us with a starting point for this collective work: a vocabulary with which to begin addressing the politics of information—together.

  1. Iris Marion Young, “Activist Challenges to Deliberative Democracy,” Political Theory 29.5 (2001) 670–90; Lynn M. Sanders, “Against Deliberation,” Political Theory 25.3 (1997) 347–76; Arthur Lupia and Anne Norton, “Inequality Is Always in the Room: Language & Power in Deliberative Democracy,” Daedalus 3 (2017) 64–76.

  2. John Dewey, “Creative Democracy—The Task Before Us,” in John Dewey: The Later Works, vol. 14, 1939–1941, ed. Jo Ann Boydston (Carbondale: Southern Illinois University Press, 2008), 224–30.

  3. Max Marcilla, “University of Virginia Answers Questions about Student Testing, New Problems Emerge,” NBC29, August 3, 2020,

  4. Alexis de Tocqueville, Democracy in America, ed. J. P. Mayer, trans. George Lawrence, Harper Per (New York: HarperPerennial Modern Classics, 1969), 256.

  5. Tocqueville, Democracy in America, 255.

  6. Tocqueville, Democracy in America, 504.

  7. Tocqueville, Democracy in America, 506.

  8. Jaron Lanier, You Are Not a Gadget (New York: Vintage, 2011).

  • Colin Koopman


    Reply to Jennifer Forestal

    Jennifer Forestal’s response to How We Became Our Data picks up from my argument, offered in the book’s fifth and final chapter, against deliberative democratic theory as sufficient for confronting the politics of data, or what I call in the book “infopower.”

    Like me, Forestal situates her approach to democracy outside of the current (but fraying) consensus around deliberativism. She cites as a preferred framework a Deweyan conception of democracy as a “way of life” that involves a heterogeneous association of habits, bodies, attitudes, values, and words (but not only words). I too endorse this kind of Deweyan analytic as a way to theoretically interrogate the functioning of democracy. To state it in the terms of last week’s exchange, I would regard Dewey’s pragmatist analytics as one more example of the kind of empiricism in political philosophy I associated with Foucault, Deleuze, and Latour. (I would note, however, that while I agree with Dewey’s methodological political theory, I depart from him with respect to key aspects of his specifically normative political theory of democratic communication as a cure for all ills. Forestal does not say if she concurs with this part of Dewey’s theory or not, but I would suspect that she too has her concerns.)

    With this in view, I want to attempt a direct response to an observation Forestal offers concerning the understanding, or conceptualization, of infopower forwarded in the book. In part 1 of her response, Forestal perceptively addresses my attempt to distinguish infopower from other modalities of power familiar in contemporary political theory. While she agrees that I successfully distinguish the political fastening (or tying down and speeding up) operated by infopower from the normalization operated by disciplinary power and the regulation operated by biopower, she is less confident that I successfully distinguish it from the prohibitions established by classical sovereign power. Forestal is right that the infopolitical format that mandates what we must do to participate (“select one of two genders”) is on its face not entirely distinct from the representative of the sovereign who stands at the gate prohibits entry to those are forbidden (“enter only if you are a male head of household”). But I would suggest that the distinction between these two modalities can be made by answering the question of “How?”

    When asked “How is infopower exercised?” my response is that it is reliant on varied techniques of formatting. Infopower fastens us to our data by specifying the shapes that data must take such that it can exist. Consider the difference between the prefab formats of Facebook-era social media in contrast to the formats of the earlier web technology of the homepage. My point is not that the mere presence of formats is itself an institution of power. For there are formats internal to both social media profiles and webpages. My point is that the politics is found in the shapes that the formats take. Everything about a Facebook profile asks for us to specify ourselves in terms of familiar categories of race, gender, and sexuality. Nothing about a webpage, as such, demands that we even so much as consider those. My point is not that Facebook is, therefore, “bad.” Rather, my point is only that we are formatted, and often in highly specific ways, by data apparatus. That can certainly lead to bad results (like the unwitting perpetuation of certain kinds of inequalities).

    By contrast, when we ask “How is sovereign power exercised?” I tend to follow Weber’s classical conceptualization of the sovereign as the monopolizer of legitimate force. Sovereign power is exercised through sword, lash, noose, or (today) paramilitarized police personnel. It is the power that at bottom relies on physical violence. Forestal suggests otherwise. “Sovereign power does look decidedly different in a democratic society,” she urges, following Tocqueville. But I would suggest that Tocqueville should be read not as theorizing sovereign power so much as anticipating Foucauldian discipline.

    Tocqueville’s work considers a form of power that has more to do with what his contemporary John Stuart Mill castigated as “the tyranny of conformity” and “the tyranny of the prevailing opinion” (Tocqueville himself employed the phrase “tyranny of the majority”). Yet that is not sovereign power. It is not consolidated in the sovereign. It is not executed by law, nor any other model of political sanction ultimately backed by a legitimated exercise of violence. Mill responds to his own worry about “tyranny” that people simply “should be eccentric” (On Liberty, ch. 3). Unless Mill is hopelessly doe-eyed (which he was not), his “tyranny” of opinion cannot be sovereign tyranny. For anyone who has faced true sovereign tyranny knows that the eccentric response is already forbidden. Mill’s “tyranny” rings metaphorically but not literally (imagine it in the ears of, say, Locke). I would suggest much the same goes for Tocqueville’s “tyranny” too (though I would be curious if Forestal would agree; for my part, I have always failed to understand how Tocqueville resolves the contradiction he poses to himself in chapter 15 of Democracy in America concerning the sovereignty of the people in democracy).

    Tocqueville’s exquisite descriptions of the political power of majority opinion emergent in American democracy are not, to my eye, descriptions of classical sovereign power. One can, of course, call that democratic power “sovereign” if one likes (and Tocqueville does), but the point is not semantic so much as analytical. We need separate concepts for separable operations of power if we are to have any chance of empirically studying the interaction among regionally and temporally coexisting modalities of power. To even be in a position to inquire into how sovereign power intersects with disciplinary power, we needed Foucault to come along and clearly delineate the two. To be able to inquire into how the garish sovereign power of state-sanctioned violence can, and sometimes does, intersect with the seemingly innocuous power of the humble technical formatting of infopower, we need to be able to analytically separate the two. Only then can we begin to empirically discern the ways that different operations of power can be made to productively leverage and exploit each other.

    In part 2 of her response, Forestal builds on her analysis by posing a Tocquevillean question concerning equality. Here I would want to hear more about why Forestal believes that the “homogeneity” of a social media profile’s format “ensures that no one user will, or can, rise above the rest.” My thought in writing the book is that the consistency established by universalizable formats can easily work in precisely the opposite way: namely, to create a terrain of obligatory information whose universality can be leveraged for the reproduction and exacerbation of inequality. This is what I had been thinking of as I constructed chapter 3 of the book as I confronted the forms and manuals that comprised the technical apparatus for residential real estate redlining. It was precisely because all houses must be coded for their race that only some of those houses (or more precisely, their neighborhoods as defined by their racial composition) could be made to “rise above the rest” and exacerbate socioeconomic inequalities.

    I should also note that in the case of redlining, all of this happened with a relatively minimal (though certainly not nonexistent) mobilization of sovereign violence. The infopolitical perpetration of racially-unequal redlining was carried out not by battalions armed with billy clubs, but by a bureaucracy of clipboard-clad professionals who diligently followed their procedure manuals and denied government-subsidized mortgages to black families whilst they approved those very same handouts for middle-class white families. I fail to see how the formats of infopower have anything resembling an “equalizing effect” in such a case. But I do wonder if Forestal might still have an insight into how it is that formats give an appearance of innocence that makes them seem so innocuous even as they are being leveraged to reproduce actual inequalities.

    Finally, in part 3 of her response, Forestal addresses the least developed part of How We Became Our Data, namely that where I only very briefly take up the crucial question how we can resist infopower. Here I would only signal my concurrence with Forestal in her call to “consider how we are all cooperative and complicit” in the operations of infopolitical injustices “even when we are not the ones doing the developing.” I think Forestal is right that this is precisely where our greatest challenge lies when it comes to the politics of data.

    One of my hopes in refocusing data politics around techniques of formatting is to make those politics more tractable than is the case in the current literature that is overly focused on algorithms. A format is often conceptually and technically available for redesign or reuse in a way that an algorithm is not. Even those of us who are not developers can articulate how to remodel a format (say a gender drop-down selector on an account registration page). That, I think, gives us some ground to travel along in conducting resistance. But I admit that this is only to establish the possibility of transformation in a space that is increasingly closed in.


From “Infamous Men” to Influencers

or, Some Notes on Compulsion and Desire in the Data Archives

In an essay first published in 1977, Michel Foucault justifies his fascination with the archival projects he had undertaken in collaboration with the historian Arlette Farge in the archives at the Bastille. This fascination with the archives is evident throughout Foucault’s oeuvre, and it would be the motivation for texts such as Herculine Barbin, Pierre Riviere, and Disorderly Families.1 Foucault claims that his intention in such texts is to gather together “singular lives, transformed into strange poems through who knows what strange twists of fate—that is what I decided to gather into a kind of herbarium” (EW3:157). One of the main sources of Foucault’s fascination with these lives is that they exist on the margins of the archive.

Christianity demands the avowal of every sin. “The Christian West invented the astonishing constraint, which it imposed on everyone, to tell everything in order to efface everything, to express even the most minor faults in an unbroken, relentless, exhaustive murmur which nothing must elude, but which must not outlive itself even for a moment” (EW3:166). Foucault detects a shift in the seventeenth century in the technology of confession. Although the administrative police report would serve a similar function as confession had, namely to reveal the “quotidian” and make it legible to sovereign power through the form of a lettre de cachet, a petition made by a subject imploring the sovereign to act on the petitioner’s behalf against an individual who has wronged them, often against a family member. “The intervention of a limitless political power in everyday relations thus became not only acceptable and familiar but also deeply condoned—not without becoming, from that very fact, the theme of a generalized fear” (168).2 Ordinary people were compelled to express themselves in these police reports precisely because they wanted sovereign power to intercede on their behalf. This is a moment when ordinary subjects enter discourse and become present to the archive precisely because the presence of sovereign power pervades society in a way that it hadn’t before (recall for example Foucault’s contrast between the spectacular effects of sovereign power and the mundane effects of disciplinary power in Discipline and Punish), but it is not the only such moment. In the first volume of The History of Sexuality, Foucault also traces the origins of such an incitement to speak, this time to refute what he calls “the repressive hypothesis.” On the contrary, Foucault shows that beginning in the seventeenth century there had been an “incitement to discourse”: Sex becomes implicated in the exercise of power in ways that it had not been before.3

I begin here, with this moment when ordinary people became legible to the ancien régime and sovereign power made itself felt in the everyday life of the people because, just as Foucault notes various similarities between the Christian confession and the eighteenth-century lettres de cachet around a self-imposed desire to speak and thereby make oneself legible to power, we can detect a similar sort of demand to make oneself legible in the various formats of data that bears an uncanny resemblance to these older police reports and lettres de cachet. We find ourselves part of yet another “birth, consequently, of an immense possibility for discourse” that takes the form of infopower (EW3:169).

Koopman carefully analyzes infopower in terms of what he calls power’s “layering effects.” That is, infopower is superimposed upon other regimes of power relations, rather than a distinct epoch of power. I believe Foucault is making a similar point with respect to sovereign power and lettres de cachet: here we see sovereign power saturating everyday life, coexisting alongside the grand spectacle of sovereign power (and Koopman correctly points out that Foucault’s ambiguity in his analysis of power relations that gave rise to the confusion that sees power simply in historical and epochal terms).

But there is another reason for beginning here: I wish to make explicit an important dimension of infopower that is only implicit in Koopman’s genealogical reconstruction of infopower in the United States during the first third of the twentieth century. How We Became Our Data does an excellent job of telling how we became subjects of data or infopolitical subjects precisely because it shows how individuals were subject to novel techniques of formatting that only subsequently became technologies of infopower: the birth certificate, the personality test, the mortgage. But there is an ambiguity in the concept “subject” that Foucault analyzes more clearly than he does the spatiotemporal dimensions of power relations: not only are individuals passive subjects composed by concrete practices and conducted through technologies of power relations; individuals also articulate themselves. We are subjects of power in both active and passive senses, and Koopman’s genealogy of infopower shows how these various formatting techniques constitute us as data subjects, but this genealogy of the data subject must be supplemented by an account of how individuals very much want to constitute themselves as data subjects. We are data subjects in both senses.

Let me be clear: showing how we articulate ourselves as data subjects precisely because we want to be subjects of data is not Koopman’s project, and I am not claiming that it should have been.4 A genealogy of the data subject ought to begin where Koopman does, with an account of the formats and technologies that render us subjects of information, and the book does an excellent job of this. Nevertheless, I believe that any account of what we should do about this articulation of the self as a data subject, including possible strategies for resistance, must acknowledge that many of us very much want to be subjects of infopower, and many of us work hard at it every day, measuring our daily accomplishments in terms of “likes” and “retweets.” We want our subjection to data, and technology companies are all too happy to oblige. In sum, we have made a transition, perhaps without even realizing it, from infamy to influencer.

Toward the end of “Lives of Infamous Men,” Foucault makes the following observation about power: “How light power would be, and easy to dismantle no doubt, if all it did was observe, spy, detect, prohibit, and punish; but it incites, provokes, produces. It is not simply an eye and ear: it makes people act and speak.” It is this dimension of incitement that is missing from Koopman’s book. Foucault’s fascination with the lettre de cachet is a literary one. He relates how “for a long time” everyday life was largely invisible unless it somehow related to something “legendary” or “fabulous” (EW3:173). Beginning in the seventeenth century, Foucault claims, everyday events could be recounted on their own terms. “The impossible or the ridiculous ceased to be the condition under which the ordinary could be recounted” (EW3:173). Consequently, people began to relate that which is “infamous,” which Foucault explains simply means the contrary of that which is glorious and heroic; in other words, the mundane and ordinary. We see here, Foucault claims, an ethical demand emerging to “tell the most common of secrets” (EW3:173). “Hence literature belongs to the great system of constraint by which the West obliged the quotidian to enter into discourse” (EW3:174).

Foucault’s “infamous men” are the ancestors of today’s influencers precisely because Facebook, TikTok, and Twitter are among the platforms we use to reveal our quotidian secrets today; put otherwise, they are among the ways that we oblige our everyday selves to enter into infopolitical discourse.

The reason it is important to make individuals’ pervasive desire to become infopolitical subjects explicit is because it complicates the question of how we might contest infopower. Koopman notes that political theories of communication merely assume information as a starting point, and hence do not attend to the politics of information itself (186–87). Koopman cites identity theft as an example. Identity theft is a pervasive problem that can only be fully understood once we have an adequate account of the politics of information. “Addressing the problem of personal information exposure cannot only be a matter of providing for valid discursive exchange, or the free processing of information” (188). Koopman concludes by observing that because information has become a pervasive feature of our lives, at times explicitly so, but more often part of the implicit background of our lives (after all, how often do we think about our Social Security numbers and what this number represents?), it makes no sense to be oppose information altogether. After all, even Ted Kaczynski has a Social Security number. Resistance must be specific and localized. “Resistance can be conducted within the operation of infopower: a resistance to this kind of fastening, a resistance to that kind of canalizing and accelerating, and a resistance that mounts these actions by repurposing and releveraging information for alternative designs. This would be a resistance of occupation, contestation, and transformation” (193).

While Koopman’s list of forms that resistance should take is not intended to be exhaustive, I believe it is also vital to ask why we desire to be subjects of information so that we can develop practices of the self that address this desire. Naturally, there are apps for this; for example, apps that limit your screen time. Or we can consider this in terms of the widespread conspiracies that have become a feature of our present. It is not just that social media facilitates the spread of misinformation. More significantly, social media makes it easier for us to exist in online echo chambers that become one’s virtual reality, and on some level this is precisely what we want. People do not become participants in echo chambers or believers in conspiracies because they want to be misled. Humans need to trust others, and they unfortunately have placed their trust in bad epistemic authorities, and therefore this trust is misplaced. Thi Nguyen helpfully distinguishes between echo chambers and epistemic bubbles.5 Echo chambers are virtual communities in which contrary views have been actively discredited because they are distrusted, whereas epistemic bubbles merely exclude opposing views. Nguyen writes, “In epistemic bubbles, other voices are not heard; in echo chambers, other voices are actively undermined. The way to break an echo chamber is not to wave ‘the facts’ in the faces of its members. It is to attack the echo chamber at its root and repair that broken trust.” Echo chambers are more problematic because epistemic bubbles can be resolved with the introduction of different viewpoints that had previously been omitted. People in echo chambers actively distrust contrary views and want to be part of a likeminded community of believers. Echo chambers function much like cults, and the difficult way out of an echo chamber lies in finding outsiders to trust.

What fascinated Foucault about the literature of infamy that developed in the seventeenth century was that it allowed access to the real, or at least “the last and most tenuous degrees of the real” (EW3:173). Koopman correctly points out that we cannot fully escape the infopolitical world we now inhabit, but perhaps it is still possible to catch glimpses of a reality outside of it in, perhaps at least in its “tenuous degrees.”

  1. Herculine Barbin, Being the Recently Discovered Memoirs of a Nineteenth Century French Hermaphrodite, intr. Michel Foucault, trans. Richard McDougall (New York: Pantheon, 1980); “I, Pierre Riviere, Having Slaughtered My Mother, My Sister, and My Brother”: A Case of Parricide in the Nineteenth Century, ed. Michel Foucault (Lincoln: University of Nebraska Press, 1982); Arlette Farge and Michel Foucault, Disorderly Families: Infamous Letters from the Bastille Archives, ed. Nancy Luxon, trans. Thomas Scott Railton (Minneapolis: University of Minnesota Press, 2017).

  2. At least for the poor of Paris, this petition was made to the police lieutenant general; only the wealthy were able to petition the sovereign directly. See Disorderly Families, esp. the introduction.

  3. Michel Foucault, The History of Sexuality, vol. 1, An Introduction, trans. Robert Hurley (New York: Vintage, 1978 [1976]); see esp. pt. 2 (“The Repressive Hypothesis”), esp. ch. 2 (“The Incitement to Discourse”).

  4. Indeed, initial efforts to provide a genealogical account of the subject of data as a desiring subject have been provided by Bernard Harcourt in his account of what he calls our “expository society” in Exposed: Desire and Disobedience in the Digital Age (Cambridge: Harvard University Press, 2015), though Harcourt’s account focuses on the surveillance dimension of this data-driven subject. More recently, Sun-Ha Hong has written on the epistemic dimension of data subjectivity, though more work needs to be done. See Technologies of Speculation: The Limits of Knowledge in a Data-Driven Society (New York: NYU Press, 2020).

  5. Thi Nguyen, “Escape the Echo Chamber,” Aeon, April 9, 2018,

  • Colin Koopman


    Reply to Corey McCall

    Last week, Jennifer Forestal posed some provocative challenges to my conceptualization of the modality of power operative where information begins to get its grip on us and thereby dispose us as what I call “informational persons.” This week, Corey McCall poses a similar set of challenges to my conceptualization of the subject who is gripped by what I have called “infopower.”

    McCall’s discussion takes off from Foucault’s two-sided conception of the subject, or more properly, from his two-sided conception of processes of subjectivation. “We are subjects of power in both active and passive senses,” he rightly asserts (following Foucault). We find ourselves inscribed into regimes of power. For example, we get loaded into a dozen or so databases within a month of being born. But we also do the work of taking up these inscriptions. As we grow older, we diligently curate our documentary identities in anticipation of getting our driver’s license or passport or mortgage. Even leaving aside all this oh-so-serious documentation, it’s easy to see how people take up their digital selves as their, well, true selves. Think of how the average preteen today displays an almost obsessive diligence in curating their online presentation on their favored platform. And it’s not just the preteens and tweens—you know how much you obsess over your Twitter feed, your profile, your departmental webpage, or that still-ubiquitous legacy technology of academic social media, your CV.

    All of this is to suggest that I agree with McCall that any account of how “various formatting techniques constitute us as data subjects” needs its other side and “must be supplemented by an account of how individuals very much want to constitute themselves as data subjects.” I am reminded of Forestal’s suggestions in the final part of her reply last week that “we are all cooperative and complicit” in the deployment of infopower. This, I agreed with Forestal, is precisely what makes the challenges of infopolitical inequality seem so formidable. This, I agree with McCall, is necessary for any adequate understanding of how what I call “the informational person” is not a victim of repressive power, nor even a poor dupe who ideologically spins out the terms of their own oppression, but is an active associated player in the production of regional assemblages of power whose effects often include the comforts of being tied down to that data that we know and well.

    What I am not all that clear on, then, is why McCall holds that “it is this dimension of incitement that is missing” from How We Became Our Data. I certainly wanted it to be there. The first sentence of the book, I had hoped, signaled it in its initializing metaphor: “We are swaddled in data.” Our data swaddling is two-dimensional in Foucault’s sense. It is wrapped around us in moments where we are as passive as can be (just like a real swaddling is wrapped around the howling infant who demands they know not what). And it is also something we take comfort in, come to define ourselves through, and eventually demand (just like a real infant comes to depend on their swaddling and howlingly demands that they be tightly wrapped up in that way they remember). Throughout the book, I was eager to point out how eager people were to be enumerated by the Social Security Board or to be algorithmically scrutinized by a pencil-and-paper informational machine promising to reveal their personality.

    Here is perhaps one clue in our different perceptions of the book’s success in conveying the incitement to data. McCall at one point emphasizes the idea of “the subject of data as a desiring subject” by referencing Bernard Harcourt’s excellent book Exposed: Desire and Disobedience in the Digital Age (Harvard University Press, 2015). Perhaps what McCall misses most in my book is its articulation of our data to our desires. This invitation to connect the subject of data to the desire for data is an intriguing one (it echoes a similar suggested offered by the brilliant Foucault scholar Daniele Lorenzini in his review of my book in a recent issue of Contemporary Political Theory). And it is also I suspect where the book may have fallen short for some readers just as it resonates the right chord for others. Allow me to attempt to explain why I chose the approach I did.

    I am cautious about articulating the way in which people take up their data through a psychological category such as the concept of desire. Perhaps I inappropriately overly-identify the notion of desire with a psychological propositional attitude (such that a desire is always a “desire that . . .” such and such take place). Or perhaps I am just too rigid in thinking of desire as even a psychological category.

    I am more inclined to emphasize the active dimension of processes of subjectivation by construing it less in terms of what people want or desire, and more in terms of what people actually do. My account of the subjective dimensions of infopower in these terms is methodologically in step with Foucault’s accounting of disciplinary subjectivation. The target of disciplinary power is trained and taught first by having something small imposed upon them and then after awhile learning to take up that imposition as part of themselves. Does it make sense to think of discipline as something that the docile subject eventually wants or desires? In some cases, yes. But in all cases? Better to say that the docile subject just is the one who is well disciplined. They are the meticulously compunctious little prisoner, soldier, line-worker, student, social-media-profile-preener, or so forth. Do they desire to be that way? Do they want to be that way? We can ask the same questions of the formats we take up into ourselves as informational persons. Does the tween desire and want to see themselves through their own Tik Tok feed? Do you desire your perfected profile? Do you want your CV to be so thoroughly organized? For my part, I do not know how to answer these questions, but more importantly I am not clear on why it should be so important to answer them in the context of our datafied subjectivity.

    We can, and sometimes we should, pose the question of the emergence of processes of subjectivation without entangling that question in an account of desire. I freely accept that for some kinds of subject, the psychology of desire is part of any viable answer to the question. But it is not therefore always relevant.

    My point, then, is this: whether desire belongs in a genealogy is an empirical question, but not a theoretical one. As it happens, I take it that Foucault’s takedown of desire toward the end of the first volume of The History of Sexuality established precisely this (though there are of course other ways to make the point without relying on Foucault). That said, perhaps McCall’s point (and I suspect this is the case for Lorenzini), is that a genealogy of desire is an important part of the genealogy of data. If so, I freely concede that this is a ripe suggestion, yet one that still remains to be empirically harvested.

    Until it is harvested, I would rather agree with the illustration of my account offered by the sociologist David Beers in his review of my book in a recent issue of History of the Human Sciences. Beers notes how, in the period I cover in the book, “people were adapting to the presence of data . . . they were also established in their data . . . [and there was] the integration of data in individual life, narrowing in upon the self and the processes and technologies of personhood” (Beers, “A History of the Data Present,” 2020). What I like about this formulation is that it concisely articulates a relation between subjects and data that suggests activity on the part of people but without appealing to a psychological category such as desire. What matters, for me and for Beers, is what we do to take up data into ourselves. Desiring our enmeshment in data is only one of the possible things we could do with it. I am not yet convinced that it is somehow the most important. But I leave it open as an intriguing possibility that warrants future empirical scrutiny for those who are interested in.

    I agree with McCall, and I think Forestal agrees with us both, that how we conceptualize our own activity in taking up our data into ourselves as we become informational persons is crucial to any viable account of, in McCall’s words, “how we might contest infopower.” I am curious to hear more about how an articulation of infopower’s formatting along the vector of our desires could help us approach the crucial question of resistance. Perhaps it is that it helps explain how we might implement resistance to the inequalities and injustices of data that are perpetrated by technologies we are so stubbornly attached to? I do still wonder, though, if it is always so clear that the attachments of subjectivation are always those of desire. We have become our data, I argue. Do we also desire our data? Or is there some other conceptualization of our active relationship to data that would be more illuminating here?


Inquiring after Foucault

Scholarship inspired by the work of Michel Foucault falls roughly into three main categories. First, there are those who approach Foucault as a key figure in the history of philosophy whose writings, biography, and context invite interpretive and reconstructive work intended to understand Foucault and his oeuvre in its own right. The second category comprises scholars who mine Foucault’s work for concepts, such as discipline or biopolitics, which are then applied to a range of phenomena about which Foucault himself had nothing to say. Finally, a fairly small number of scholars take from Foucault not his concepts but his methods in order to engage in the kind of empirico-historical mode of inquiry he modeled, elaborating new concepts along the way.

Colin Koopman’s How We Became Our Data is firmly situated in the third of these camps. As a thoroughly historical—that is, empirical—analysis of how we became the informational persons we are today, Koopman’s study examines the disparate and contingent historical events that gave rise to information as a site for the exercise of power and the formation of new subjects. Pushing against received wisdom that the politics of information is a very recent phenomenon, Koopman shows that it, in fact, predates both our current era of “big data” and the birth of information theory in the mid-twentieth century. Indeed, the “information age,” as well as information theory, find their conditions of possibility in the formation and consolidation of a regime of information in the first three decades of the twentieth century. Koopman’s detailed attention to the technical minutiae of data collection, processing, and output in the context of US birth registration, social security numbers, personality psychology, and redlining enables him to identify and conceptualize a distinct mode of power that is immanent in and circulates through these practices: infopower.

Infopower names a distinctive and contextually specific way of exercising power through an array of technical procedures that format data and fasten us to our data, thereby constituting us as particular kinds of persons who are both effects and targets of infopower. This mode of power is underpinned by a political and epistemic rationality that Koopman calls the “data episteme” and for which information itself is the foundation and goal (160). Consider, for example, that obtaining the social security number that uniquely identifies us requires us to produce other identifying information about us, such as a birth certificate or passport, which in turn rely on our ability to supply certain identification documents. The rationality of the data episteme thus establishes a self-referential regime of information that constitutes a closed circuit through which infopower can flow continuously and without interruption.

Koopman anticipates a potential objection to this diagnosis: namely, that he has not actually isolated a distinct mode of power. The objection, in other words, is that what Koopman calls infopower can be reduced to more familiar notions of power such as the Foucauldian concepts of sovereignty, discipline, and biopolitics or competing conceptualizations of control power (Deleuze), expository power (Harcourt), #datapolitik (Panagia), or datapower (Chamayou), to name just a few. Accordingly, Koopman provides a careful justification of the irreducibility of infopower. This justification, it seems to me, takes two different forms. With regard to Foucault’s concepts of sovereignty, discipline, and biopolitics, Koopman pursues a strategy of analytically distinguishing their subjects, operations, techniques, and rationalities from those of infopower. He argues, for instance, that even though Foucault’s biopolitics of the population relies on statistics and mechanisms of normalizing racism, it is a “politics of life” whose “subjects are living beings construed as populations” and whose “operation is that of regulation, typified by such techniques as public health policy, demographic management, and medical intervention” (164). Or take discipline, which “operates a power of normalization by coaxing bodies . . . to conform to the norm” through “techniques [that] include panoptic surveillance, regular examination, and meticulous training, or dressage” (165–66). Infopower, by contrast, is neither concerned with the life of populations nor with the bodies of individuals but with accounting for subjects in terms of their information, which typically captures immaterial traits. Similarly, infopower does not prohibit and exclude, as sovereign power does, but simply formats data in particular ways. Koopman maintains, however, that to claim that infopower is irreducible to sovereignty and biopower is not to say that they are incompatible. Rather, following Foucault’s emphasis on the coexistence and superimposition of different modes of power, he suggests that infopower is “deposited on, or layered on, the sedimented earlier strata of power” (168), from which it can nevertheless be analytically distinguished.

In response to the potential objection that his concept of infopower adds little to competing concepts such as Deleuze’s notion of control power or Harcourt’s concept of expository power, however, Koopman takes a different—namely, genealogical—approach. In particular, he contrasts his own approach, and the concept of infopower that emerges from it, with a certain avant-gardism that he finds, to different degrees, in Deleuze and Harcourt. Koopman points out that Deleuze’s “tendency to disregard the past” (169) and Harcourt’s lack of attention to the historical conditions of possibility of the digital age of exposure give rise to conceptualizations that may well be useful but must themselves be “properly embedded in a wider history” (171). Instead of drawing analytic distinctions between infopower, control power, and expository power, Koopman thus emphasizes infopower’s genealogical provenance. Infopower is an empirical concept produced by historical inquiry that makes available for critique the very elements that are presupposed, and therefore remain unexamined, in Deleuze’s and Harcourt’s conceptualizations.

But if it is its genealogy that differentiates infopower from control power and expository power, it seems to me that its genealogy is also what distinguishes infopower from the more familiar modes of power theorized by Foucault. As Koopman argues elsewhere, Foucault’s concepts, including biopolitics and discipline, are products of highly contextual empirical inquiry whose applicability to other contexts must be established genealogically rather than assumed axiomatically. To presuppose that Foucault’s concepts apply beyond their contexts of articulation is to “ontologize” or “transcendentalize” what are “empirical concepts yielded by historical inquiry” (574). This is why, I take it, the work undertaken in How We Became Our Data “involves pressing beyond Foucault’s conceptualizations” and instead appropriates Foucault’s method to elaborate new concepts that enable us to understand phenomena about which Foucault’s work is silent (22). I wonder, then, whether Koopman’s analytic, rather than genealogical, distinction between infopower and sovereignty, discipline, and biopolitics risks implicitly ontologizing these concepts by accepting that they indeed apply to a geographical and historical context that Foucault himself did not study. That we live in disciplinary and biopolitical times has become a truism—the kind of obviousness which Foucault urged us to critically investigate. It is precisely this refusal to give in to the allure of the obvious that is at the heart of Koopman’s work and, to my mind, one of its greatest strengths. Instead of relying on familiar concepts, which have become too self-evident to retain their critical edge, his genealogy provides us with an empirical conceptualization of the power that has made us, who inhabit a very specific historical and geographical context, who we are. It is on those who would insist that his account tells us little that Foucault hasn’t already told us to demonstrate, rather than simply assert, that this is indeed the case.

  • Colin Koopman


    Reply to Verena Erlenbusch-Anderson

    Verena Erlenbusch-Anderson perceptively, and rightly, situates my efforts in How We Became Our Data through a tripartite categorization of possible uses of Foucault in scholarly inquiry. Foucault is a subject of interpretation for some, a source of conceptual illumination for others, and a methodological model of how to conduct inquiry for still others (this is a schema that Erlenbusch-Anderson elaborates in much more detail in her article “Philosophical Practice Following Foucault” in a recent issue of Foucault Studies). She is right that How We Became Our Data takes philosophical genealogy as its methodological model without relying on concepts such as discipline or biopower to help make sense of our contemporary data politics. (In this I would hope to humbly locate the book in a lineage of other genealogical philosophers who have learned from Foucault, including Arnold Davidson, Ladelle McWhorter, Ian Hacking, Paul Rabinow, Kevin Olson, and of course Verena Erlenbusch-Anderson.)

    Genealogy requires us to be empirical (on which, see my reply to Smith above) such that we begin inquiry with a question that can only be answered after we have conducted our examinations. Only in this way can philosophy—for instance, genealogy—be an instrument for learning (rather than, say, a sophisticated way of rehearsing what we already know). Genealogical empiricism requires that we leave open the question of whether or not a concept such as discipline or biopower (or sovereign power, or desire, thinking of some of the previous replies in this exchange) applies to a given political formation. We must always be prepared to ask the question, and to submit candidate answers to rigorous and patient inquiry. This is what Foucault did in Discipline and Punish, The Birth of the Clinic, and other of his works. This is what I mean in saying that I take his genealogy as my model. I learn from what he did. But I do not stare at it in endless fascination (as do some of Foucault’s readers who are so taken with such of his concepts as discipline and biopower that they believe that these concepts simply must make sense of everything in the contemporary moment).

    How We Became Our Data attempts a genealogy of our present data-saturated moment in order to excavate from it patterns of practice that, the book then argues, are best understood through my concepts of “infopower” and “the informational person” (and also “the data episteme”). These concepts, I argue (in the fine print of the overly compunctious academic), are distinct both from Foucault’s concepts (such as “biopower”) and other candidate concepts proposed by contemporary scholars that are meant to help explain our data-laden present.

    Erlenbusch-Anderson detects a dissymmetry between how I distinguish my “infopower” from its Foucauldian predecessors (biopolitics, disciplinary anatomopolitics, sovereignty) and how I distinguish it from contemporary conceptualizations of the politics of data by other scholars in recent years. In distinguishing “infopower” from Foucault I largely rely on what I would just as well call conceptual analysis. But in distinguishing “infopower” from the landscape of contemporary scholarship on data politics I rely on the empirical details of my genealogy to make the distinction. I am not sure if Erlenbusch-Anderson means to point out this dissymmetry as a shortcoming or not. I agree with her point as it concerns the manner of exposition I offer in the book. However, I take it that any rigorous attempt to distinguish a concept involved in a genealogy must take place on two simultaneous levels: there must be both conceptual analysis and empirical genealogical specification. Perhaps it is not clearly exposited in the book, but my attempt to isolate “infopower” from any other candidate concept that purports to make sense of our informational present relies on both the analytical and the genealogical moment. I think that conceptual analysis is crucially necessary for distinguishing “infopower” from, say, Bernard Harcourt’s concept of “expository power” (in the book Exposed, cited last week in my reply to McCall). And I believe that genealogical analysis must be part and parcel of determining whether or not “infopower” (or some other predecessor concept of power such as those offered by Foucault) best makes sense of a given political formation.

    Erlenbusch-Anderson expresses the worry that a purely analytical distinction between infopower and biopower (discipline, etc.) risks ontologizing these concepts. I take it that the worry is one concerning whether one has done enough to submit these concepts to the tests of genealogical analysis to see if they help make sense of a given field of inquiry. Erlenbusch-Anderson’s worry on this point is posed not so much to How We Became Our Data as it is to a common reaction on the part of some Foucauldians to a book like mine insofar as it attempts to theorize power beyond the models excavated by Foucault. The reaction itself is not unwarranted, but asserting its contents as true, rather than submitting them to inquiry, is. In this, I agree with Erlenbusch-Anderson.

    Throughout working on the book I was asking myself if infopower was in fact the best way to conceptualize the documentary evidence I was awash in or if such canonized concepts as discipline would suffice. I was posing these questions to myself because, to employ the terms I offered above, I hoped to learn something that I did not already know. What I learned, and so I argue in the book, is that those canonized concepts help to make sense of certain aspects of those fields that I studied (for an exhibit of this, see my coauthored article on birth certificates: “Standard Forms Power,” in Constellations, from 2018), but utterly fail to make sense of them in many of their most important functions, effects, and techniques.

    This dispels at least part of the kind of worry that Erlenbusch-Anderson expresses and that I myself share. Indeed in my own research and writing, this is not just a worry but an obsession. That said, I have only addressed the worry as it applies to my book. I am not sure if Erlenbusch-Anderson also intends to express the concern as one that applies, as it were, generally. In other words, is the concern that ontologization would follow from any attempt to apply Foucault’s concepts to “a geographical and historical context that Foucault himself did not study”?

    I am not sure that, when it comes to other efforts in political philosophy, I have this worry quite as obsessively as I do when it concerns my own work. As I see it, there are three basic options here. First, one can ontologize Foucault’s concepts of power and simply assume they apply just about anywhere without thinking that one has to do empirical inquiry to show this (certain dimensions of Agamben’s work are a good example of this and that serves to demonstrate how utterly disconnected is Agamben from the style of philosophy practiced by Foucault ). Second, one can genealogically study a field to determine if Foucault’s concepts of power apply there or not (this is what I hoped to do in my book, and it is also surely what Erlenbusch-Anderson successfully does in her Genealogies of Terrorism). Third, one can provisionally and reasonably infer that a specific concept of Foucault’s applies well to a specific field of inquiry that he himself did not study and which one oneself has also not studied. There are plenty of cases where we can reasonably, though certainly only provisionally, believe that disciplinary power is operative before we actually dig into the details to study the question. But Erlenbusch-Anderson would be right to insist that any such case can only be regarded as provisional. Until we conduct the work of inquiry, we cannot truly say that we have learned. Perhaps the difference between us, and it seems rather miniscule when stated in this way, is over how much we would be willing to credit the provisional hypothesis in the meanwhile. I think that is probably a question to be answered only contextually and pragmatically.

    I want to close by reiterating my gratitude to all four commentators in this series for their illuminating insights and perceptive challenges. I especially want to acknowledge, as publicly as I can, Erlenbusch-Anderson’s insights and instigations. For all that she says in the piece above, one thing she does not say is that she raised at least a dozen other sharp and subtle criticisms when she read earlier drafts of the manuscript, and in so doing helped make it a much better book. Whether I therefrom turned the unfinished manuscript into a good book is, I wholly accept, for its readers to decide.