What Is Bio-Medial-Politics?

What Is Bio-Medial-Politics?

by Natasha Lushetich (Dundee, UK)

TO understand present-day biopolitics and how we can resist it, we need to talk about “bio-medial-politics.” The effects of datafication on freedom, affect, (perception of) time, and the political horizon—understood as the common horizon of possibility—are well-known. But datafication also implies more mysterious medial operations that cannot be related to human intentions or programming. Random and hidden machinic and algorithmic processes—and the medial efficacy they exercise—may, at this stage, be unfathomable in their effects. Still, we must theorise them, not only to understand how they change techno-human relations, but also to explore possible modes of resistance.

Biopolitics: Old and New

The first things that usually come to mind when we say “biopolitics” are Michel Foucault’s critical genealogy of power and Giorgio Agamben’s reinterpretation of sovereign power. At the end of the 18th century, after decades of anatomo-politics—or “body discipline”—that emerged as a side effect of the Industrial Revolution, biopolitics asserted itself through the calculation and management of birth and death rates as well as the classification of races, classes, sexualities, and physical/mental diseases. It turned, as Foucault put it, “’the population’ into a political problem.”1 As a technology of power, it was very different from sovereign power because of its propensity to “qualify, measure, appraise, and hierarchize rather than display itself in its murderous splendor,” as the sovereign rule over life and death had done.2

For Agamben, by contrast, sovereign power is always already bio-political. The emergence of biopolitics does not mark a break in/with the history of western politics; it merely extends the existing biopolitical imperative of the state. By “placing biological life at the center of its calculations,” Agamben writes, the modern state “does nothing other than bring to light the secret tie uniting power and bare life.”3 The concept of bare life is tied to the timeless political trope of homo sacer, a human being who, according to Roman law, may be killed but not sacrificed because s/he is reduced to bare life, bereft of political existence, and therefore expendable. It is also closely associated with the state of exception, for which Agamben’s prime example is the Nazi concentration camp, a territory placed outside the juridical order. To Agamben, the concentration camp is “the hidden paradigm of the political space of modernity.”4 In other words, it is not a surpassed historical fact or a separate space but rather a matrix that is re-enacted and re-created in contemporary modes of neoliberal coercion.

The health crisis of the last two years is, for Agamben, a crisis of biosecurity, like the one staged by the Nazi state in 1933, not only because citizens have been reduced to their bare biological existence,5 but because science has joined the “religion” of capitalism: it pursues its own goals at the expense of human lives, attempting to control the population on the basis of conflicting and often nebulous interpretations of data, and biodata in particular.6

In a book that is, in many ways, a response to Agamben’s 2021 Where Are We Now? The Epidemic as Politics, Benjamin Bratton contests the idea that what we have witnessed since March 2020 is an extension of neoliberal coercion and the culmination of biopolitics. He suggests, instead, that Agamben’s “negative” view of biopolitics is useless in our current predicament because it is soixante-huit-ard in the excessive importance it grants to the sovereign subject.7 What we need, according to Bratton, is a positive approach to biopolitics. Instead of denial and escapism, which is how he diagnoses the current condition, we should open ourselves up to the “ethics of being an object,”8 as the key problem today is not the pandemic, but “narcissistic bourgeois liberal individualism itself.”9 Unlike Agamben, who insists on the need to re-establish the agency of the political subject in one way or another, Bratton’s solution is to give up (individual) agency altogether.

Even if we accept Bratton’s “positive” biopolitics—regardless of how much it may remind us of such products as coffee without caffeine, beer without alcohol, and the post-George Bush administration without politics—the problem with both the “negative” and “positive” view is that they grossly underestimate the agentic role of mediation in the digital world. They also implicitly assume that it is for humans to decide which kind of biopolitics to choose. Unfortunately, nothing could be further from the truth. Not only because of what Jacques Rancière tells us about aesthetic politics, which differs from the Habermasian universe where sovereign individuals make informed decisions about the commons in that it is always already pre-cognitive, that is, unconsciously filters what it perceives as “intelligible speech” from “mere noise,”10 but also because of the pre-emptive operation of (many) algorithms—“code as law,” as Lawrence Lessig puts it11—with various actors, from governments to corporations, embedding their values in programmes and protocols, effectively constraining and in some cases even dictating human actions.

The Medial in Bio-Medial-Politics

I would suggest that the only way to understand what is at stake in our current crisis-ridden world is to take full cognisance of the “medial” in “bio-medial-politics.” The point here is not nomenclature—whether we should call something x, y, or z. It is to enable understanding so as to enable resistance.

A medium, in the broad sense of the word, stands for relationality and the channelling or modification of difference.12 In addition to establishing new relationships between here and there, now and then, yours and mine, all media remediate older media. They reconfigure temporal, spatial, inter-subjective, and inter-subjective-objective relations. Some of these reconfigurations can be traced to intentionality and/or a programmed course of action. Others, however, operate in ways that are as opaque as they are impactful.

But let us begin with datafication. Data architectures have been determining systems of classification in government, the corporate world, and, more broadly, the social world for almost two decades now. Datafication has also had profound effects on human affect, time, perception of time, and the political horizon. Practices such as biometrics, which “snatches” fingerprints, eye scans, and photographs from thousands of people on a daily basis, are widely used in security and identity management from border control to civic, medical, and corporate records. The combination of body datafication and what Maurizio Lazzarato has called “asignifying semiotics”—a semiotics that does not communicate representations to the conscious mind but issues commands to less-than-conscious regions, the way traffic lights issue “stop” and “go” commands through red and green lights13—has created a (not entirely unfounded) fear of ubiquitous control, even though discussions of such control have hardly gone beyond the usual suspects: governments and corporations.

Rapid circulation of data has further created responsive body-minds affectively attuned to the demands of data architectures. Consider the well-known “vibrating phone” syndrome: the mobile phone user’s hallucination of vibration when the mobile phone is not vibrating at all. As two separate groups of researchers showed as early as 2010, one reason for the high percentage of such hallucinations today (70-86%) is stress. The other is the proximity of the phone to the nervous system, with the anatomo-political effect of placing people on permanent call.14 Today, gadgets are not separate from us as the telescope was in the 18th century. Even when they are not implants, like, say, biomorphic chips, they are integrated into our nervous system.

Time, too, is “out of joint,” in a permanent state of crisis. On the one hand, most things and objects have become disposable. Designed to last a year or two, they are no longer repositories of memory but recede from experience with speed and indifference. On the other hand, digital recorders have, for decades now, been immemorialising even the smallest events, repeatedly hauling back the past and in this way erasing the thresholds between the familiar, the semi-forgotten, the deeply intimate, and the so-distant-as-to-be-unthinkable.15 Predictive algorithms have, for their part, changed l’avenir into le futur. Le futur, as Derrida noted, is “programmed to come”16 in the sense that, once born, every living being is “programmed” to die. In contast, l’avenirà venir (to come)—surpasses all forms of programmability, “natural” as well as “artificial.” L’avenir, Derrida writes, is “nothing of which it might be possible to say: ‘this is.’” It is “unforeseeable, free of ulterior expectations.”17

Related to this is the question of memory. Cultural memory is usually defined as that part of culture that cannot be transmitted by genes.18 In oral cultures, this memory is stored in finite human beings and does not exceed the capacity of the information carrier. In the digital era, however, it is stored on external mnemonic carriers with almost unlimited storage capacity and organised, to a significant degree, by gadgets and apps that follow obscure algorithmic procedures. The result is the emergence of a technological unconscious. Further, while Norbert Wiener maintains that the purpose of cybernetic organisation was to save us from entropy,19 the exponential growth of the informational mass as well as the instability of information itself—the average life of a URL is 40 days—has exposed us to drastic reversals in information that render formerly beneficial actions utterly useless, or even criminal.20

Finally, today’s technical media entail an irrevocable change in the political horizon. Antoinette Rouvroy and Thomas Berns speak of an “algorithmic governmentality,” in which political subjects are data points, knowledge production eschews absolute truth, power exercise is devoid of authority, and subjectivation is personalisation without the subject.21 The result is #datapolitik—a set of operations that regulate circulation through space and time by applying technology to practices of tracking and capture.22 Fully automated in its operations,23 #datapolitik monitors and manages difference through recursive repetition, with the ultimate aim of rendering any moving target infinitely predictable.

Subjects and Radical Mediation

Far from mere gadgets, technical media must be seen as perceptual milieus that shape the disposition of worlds. They demand that we reconsider what Richard Grusin has called “radical mediation.” Understood as the “immediacy of middleness in which we are already living and moving,”24 radical mediation is the very process out of which subjects and objects emerge.

What sort of subjects, then, are emerging from this radical mediation? As Cathy O’Neil has persuasively argued in her (by now classic) Weapons of Math Destruction, the data economy relies heavily on algorithms that encode prejudices and biases since their interpretative methods reinforce existing forms of discrimination. Such algorithms produce destructive feedback loops that remain invisible to the people they impact.25 For example, they routinely calculate recidivism risks based on data extraction and correlation methods used by the police, resulting in a further aggravation of century-long inequities such as the incarceration of black populations in the U.S.26

Similarly to such violence-perpetuating algorithms, biometric practices do not only identify bodies by the processing of biometric data, but transform these bodies in and through the classifications used, such as “asylum seeker” or “citizen.” Healthcare algorithms, likewise, decide not only who receives care and how that care is administered. They also determine the identity of the subject in need of medical care by correlating biometric health designations such as “breast cancer survivor” or “bulimic” with the individual’s gender, class, and economic status.

Resistance to such practices cannot be direct, oppositional, or even based on brilliant, carefully articulated arguments. Rather, it needs to “work with” the existing frameworks and conditions. Already in the 1990s, artivist (artistic-activist) groups such as The Yes Men engaged in culture jamming and (h)activism, creating, for instance, fake websites for the World Trade Organisation (WTO). By exposing site visitors to an overwhelming amount of details and provoking circumstantiality (the incapacity to discern the relevant from the irrelevant), their aim was to “detourn”—disrupt and redirect—the course of political summits and global legal disputes.27

Knowbotic Research, in turn, has drawn attention to tactics of invisibilisation. While oppressive practices such as racism and sexism work to render invisible the very process of rendering invisible, there is also the reverse trend, since the beginning of this millennium, of technology making it increasingly difficult to remain invisible even if one wishes to do so. In MacGhillie Just A Void (2008-2010), the group deployed a camouflage suit originally invented in the 19th century for hunting, and subsequently used for military purposes the world over, to stage invisibility in plain sight.28

In the last ten years, the Dirty New Media movement has used computing to disorder proprietary software production through databending. Helen Nissenbaum & Finn Brunton’s browser extension TrackMeNot sends randomly generated queries to search engines, obscuring user information from corporate eyes by overwhelming search engines with data. Programmers like Peder Norby use algorithms to disrupt the orientational codes of Google Maps. micha cárdenas’ Autonet (2016-preent) uses DIY protocols to create ever-changing networks of discriminated people—whether migrant, poor, or transgender—thereby sabotaging the linkage of medical, employment, and financial records—in other words, medical surveillance. Such practices rearticulate the relationship between design and chance, rule, iteration, and variability, and reconfigure the relationship between existential territories and datafied virtual domains.

Medial Efficacy and Algorithmic Indeterminacy

But these forms of resistance are by no means enough in an age where bio-medial-politics extends to less perceptible, more diffuse operations. Today, machinic and algorithmic indeterminacy have a medial efficacy all of their own. Rather than automated mathematical procedures, algorithms are emergent diagrams based on the repeatable organisation of space, time, objects, and actions. As long as it is repeatable, any such organization can appear as stable—in human terms, “just the way things are”—regardless of whether it is produced for the first, fifth, or thousandth time. If truth is taken to mean the stability of basic assumptions, then any such configuration can be “true.”29

To better understand this indeterminate process, we might turn to the medial aspect of inscription. Any kind of inscription, Cornelia Vismann argues, is inseparable from auto-praxis [Eigenpraxis]. Whether the agent of inscription is human or other-than-human,30 it is never independent of its conditions of coming into being—space, time, and environmental forces. As a result, the agent-thing iteratively steers emergent processes in new, and, for humans, unpredictable and imperceptible directions.31 All media, understood both as mediatic relations and as objects, programmes, and protocols, engage in such auto-praxis.32

Just like medial agents and environments, neural networks used in machine learning are essentially constitutive processes. Neurons “fire through a function of the strength of the connection.”33 They create media based on the mechanisms configured during training on input data. In (humanly) supervised training, the model of emergence is consecutively monitored and modified. In non-supervised learning (without human input), however, an auto-productive developmental logic emerges and stabilises through repetition.34

Most algorithmic and machinic exchanges are black-boxed and interactive alike. They are characterized by a process that McKenzie Wark calls vectoralisation—the linking and frequent (re) directing of informational flows in information ecologies.35 In addition to their opaque operation and perilous dependence on after-the-fact insights, they operate on a temporal scale that surpasses that of humans. This means that machinic exchanges generate new temporalities and new forms of indeterminacy (as random change) beyond human parameters.36

The horizon of a bio-medial-politics that is unrelated to human intention or programming is doubly worrying. On the one hand, it engenders errors; on the other, random algorithmic and machinic behaviour. This indeterminate process is often described as opaque, unknowable, and even dangerous: when code passes into algorithms and algorithms begin to create new algorithms, who knows where it all might lead? At this point, there is no obvious counteraction, or alternative, in sight. But this is not to say that we are powerless, even if we are not computer or data experts. What we need is a different method. Instead of relying on historical genealogies that can show or prove the development of a tendency, we need to study chance, contingency, plasticity, and accidentology.

It is true that the future has been usurped—at least to an extent. Many authors, such as Shoshana Zuboff, have argued this point eloquently and persuasively. But can we really oppose, in this day and age, “human futures” (seen as open) to “data futures” (seen as pre-empted or cancelled)?37 Such a separation seems as naïve as Bratton’s conflation of “positive” and “biopolitics.” What we need to do instead is understand the nature of non-programmed machinic and algorithmic behaviour, or algorithmic self-creation, and formulate tactics of resistance based on a posthuman (non-human-centred) and distributed agency which includes non-human (machinic) others and can navigate complexity.

Consider the wekinator, a software with which anyone (any non-machine-learning-scientist) can use machine learning to create non-existent musical instruments, real-time gesture-responsive choreographic applications, computer vision, composing, listening, or design systems. The wekinator does not have a specific set of functions of its own. Instead, it connects to dozens and dozens of sensors and creative coding tools and creates new functions by recombining their existing functions in hitherto non-existent, often surprising ways. Is it too far-fetched to think that, like the recursive function in computing, wekinators and similar software could present a critical account of their operation as well as that of the algorithms, protocols, and programmes they interact with, and in this way serve to dismantle our current, grim revival of Skinnerian inculcation techniques?

B.F. Skinner’s work is rooted in the idea that all behaviour is a reaction to stimuli rather than the action of some sort of essence or identity, and that it can be controlled through conditioning and reinforcement. There are ways to break the chain of reinforcement (or repeated inculcation), such as stimulus avoidance, self-administered satiation, or aversive stimulation.38 But none of these ways are, strictly speaking, applicable in our networked world, where it is not the sovereign, the state, or the corporation that exerts a biopolitical working, but our everyday (digital) environment itself, indefinable or opaque at best in its spatiality, temporality, substance, operations, and interactions. This situation can only be “detourned” with the aid of oblique digital tactics, like those practiced by The Yes Men, Nissenbaum, cárdenas, and all others who take up the challenge of confronting the “medial” in “bio-medial-politics.”


1. Michel Foucault, The History of Sexuality, Volume 1: An Introduction, translated by Robert Hurley, London: Penguin, 1990, 245.
2. Foucault, The History of Sexuality, 144.
3. Giorgio Agamben, Homo Sacer: Sovereign Power and Bare Life, translated by Daniel Heller-Roazen, Stanford: Stanford University Press, 1998, 6.
4. Agamben, Homo Sacer, 123.
5. Giorgio Agamben, Where Are We Now? The Epidemic as Politics, translated by Valeria Dani, Lanham: Rowman and Littlefield, 2021, 67.
6. Agamben, Where Are We Now?, 45.
7. Benjamin Bratton, The Revenge of the Real: Politics for a Post-Pandemic World, London: Verso, 2021, 103-108.
8. Bratton, The Revenge of the Real, 108.
9. Bratton, The Revenge of the Real, 148.
10. Jacques Rancière, Dissensus: On Politics and Aesthetics, translated by Steven Corcoran, London: Continuum, 2010, 36.
11. Lawrence Lessig, Code and Other Laws of Cyberspace, New York: Basic Books, 1999.
12. Natasha Lushetich, “Why Ask the Question?” in Lushetich (ed.), Big Data: A New Medium?, London and New York: Routledge, 2020, 6.
13. Maurizio Lazarrato, Signs and Machines: Capitalism and the Production of Subjectivity, translated by Joshua David Jordan, Cambridge: MIT Press, 2014, 40.
14. Michael B. Rothberg et al., “Phantom vibration syndrome among medical staff: A cross sectional survey,” BMJ 341: c6914 (2010); Michelle Drouin et al., “Phantom Vibrations among Undergraduates: Prevalence and Associated Psychological Characteristics,” Computers in Human Behavior 28 (2012): 1490-1496.
15. Byung-Chul Han, Le Parfum du Temps, translated by Julie Stroz, Paris: Circe, 2016.
16. Jacques Derrida, Derrida, documentary film, New York: Zeitgeist Films, 2002.
17. Jacques Derrida, Mémoires: pour Paul de Man, Paris: Galilee, 1988.
18. Jan Assmann, “Communicative and Cultural Memory,” in Astrid Erll et al. (eds.), Cultural Memory Studies, An International and Interdisciplinary Handbook, Berlin and New York: De Gruyter, 2008, 109-118.
19. Norbert Wiener, The Human Use of Human Beings: Cybernetics and Society, New York: Da Capo Press, 1950.
20. Molly Rothenberg, The Excessive Subject: A New Theory of Social Change, Cambridge: Polity, 2009.
21. Antoinette Rouvroy and Thomas Berns, “Algorithmic Governmentality and Prospects of Emancipation,” translated by Liz Carey Libbrecht, Réseaux 177.1 (January 2013): 163-196.
22. Davide Panagia and Çağlar Köseoğlu, “#datapolitik: An Interview with Davide Panagia,” Contrivers’ Review, 2017.
23. Panagia and Köseoğlu, “#datapolitik.”
24. Richard Grusin, “Radical Mediation,” Critical Inquiry 42.1 (2015): 124-148, here 128.
25. Cathy O’Neil, Weapons of Math Destruction, New York: Broadway Books, 2016, 3.
26. O’Neil, Weapons of Math Destruction; see also Faisal Kamiran and Toon Calders, “Data preprocessing techniques for classification without discrimination,” Knowledge and Information Systems 33.1 (October 2012): 1-33.
27. For more information on The Yes Men, see Natasha Lushetich, “On Ludic Servitude,” in Guillaume Collett et al. (eds), The Double Binds of Neoliberalism: Theory and Culture After 1968, Lanham: Rowman and Littlefield, 2022, 103-122.
28. For more information on this project, see Andreas Broeckmann and Knowbotic Research (eds.), Opaque Presence: Manual of Latent Invisibilities, Berlin: Diaphanes, 2010.
29. For more information on medial efficacy, see Natasha Lushetich, “Algorithms and Medial Efficacy,” The Philosophical Salon, March 2022.
30. Cornelia Vismann, “Cultural Techniques of Sovereignty,” translated by Ilinca Iurascu, Theory, Culture & Society 30.6 (August 2013): 83-93.
31. Vismann, “Cultural Techniques of Sovereignty,” 84.
32. Vismann, “Cultural Techniques of Sovereignty,” 84.
33. Lonce Wyse, “Appreciating Machine-Generated Artwork through Deep Learning Mechanisms,” in Lushetich (ed.), Big Data: A New Medium?, London and New York: Routledge, 2020, 94-112, here 96.
34. Paul M. Bodily and Dan Ventura, “Explainability: An Aesthetic for Aesthetics in Computational Creative Systems,” Proceedings of the Ninth International Conference on Computational Creativity, Salamanca, 2018, 153-160.
35. McKenzie Wark, Capital is Dead, London: Verso, 2019.
36. Yuk Hui, Recursivity and Contingency, Lanham: Rowman and Littlefield, 2019.
37. Shoshana Zuboff, The Age of Surveillance Capitalism, London: Profile Books, 2019.
38. Burrhus Frederic Skinner, About Behaviorism, New York: Knopf, 1974.

Responses