Lacan's Third Series #4
Jouissance, Doom-Scrolling, and Why Rational Choice Theory Cannot Explain Platform Engagement
You know you’re not enjoying this. You’ve known it for twenty minutes. Still, your thumb moves, scrolling through a feed that isn’t what you wanted or needed, and it isn’t making the evening feel any better as you sit on your couch, screen six inches from your face. Tomorrow, you’ll call this a waste of time. Tomorrow, you’ll do it again.
Rational choice theory predicts that this behavior should not exist. An agent who maximizes utility stops engaging when marginal cost exceeds marginal benefit. The cost is high (time, attention, sleep, mood), and the benefit, according to the user’s contemporaneous report, is low or negative. The user should stop. The user does not stop. Something else is operating, something that the dominant frameworks for understanding human behavior cannot name because their basic unit of analysis, the utility-maximizing agent, excludes it by construction.
Behavioral economics takes the analysis a step further. It highlights concepts such as variable reinforcement schedules, present bias, sunk cost fallacy, and dark patterns. However, it still relies on the idea of the rational agent, now seen as exhibiting systematic biases that cause deviations from personal interests. In this framework, scrolling past the point of enjoyment is viewed as an error—a lapse in self-regulation or a flaw in largely functional cognitive processes. Here, the user is described primarily as a pleasure-seeker who has been misled or impaired.
At this point, a different analytic approach is needed. Lacan proposes a different approach. Instead of cost-benefit calculation, he uses the distinction between desire and drive, which relies on circular logic. Here, the goal is not reaching the object but completing a loop. Platform engagement follows this drive: infinite scroll does not promise satisfaction at the bottom, but keeps the circuit going. A notification does not deliver what you want; it keeps you checking. Satisfaction comes from the act of looking, not from what you find.
Two Theories That Cannot See What Is in Front of Them
A randomized experiment paid 2,743 Facebook users to deactivate their accounts for four weeks (Allcott et al., 2020). Deactivation increased subjective well-being by 0.09 standard deviations, reduced political polarization, and freed up an average of 60 minutes per day, most of which users redirected to socializing, watching television, and other offline activities. The most damaging finding for rational choice theory was not the improvement in well-being but what happened afterward: users who had been paid to deactivate voluntarily reduced their Facebook use by 9% after the experiment ended. They valued the platform less after not using it. Their prior “revealed preferences,” the choices that supposedly demonstrated they preferred scrolling to alternatives, were inflated by the very engagement pattern those preferences were supposed to explain.
Another study found 31% of social media use is due to self-control problems, not real preference. This came from willingness-to-pay experiments that separated what users wanted from what they did (Allcott, Gentzkow & Song, 2022). When offered tools to limit their own usage, 66% of users adopted them without prompting or payment. This shows most heavy users know they use social media more than they truly want (Hoong, 2019). A Carnegie Mellon study found users regretted at least some of their social media use in 60% of sessions and regretted all of it in nearly 40%. Algorithmic recommendations produced the highest regret (Cho et al., 2021).
Revealed preference theory cannot accommodate these findings without collapsing into tautology. If choosing to scroll reveals a preference for scrolling, the theory is unfalsifiable. If the same user simultaneously scrolls and pays to prevent scrolling, their revealed preferences contradict each other. The theory requires a unified agent whose choices reflect stable preferences. The platform user is not that agent.
When behavioral economics is applied to platform design, it becomes easier to understand why these systems are so engaging. Variable ratio reinforcement—the slot-machine logic of unpredictable rewards—creates strong, lasting habits and shows up in pull-to-refresh, notification badges, and algorithmic feeds. Present bias and hyperbolic discounting show why scrolling feels better than stopping. Loss aversion and FOMO mean not checking feels worse than checking. Choice architecture shows how autoplay and infinite scroll make it easy to start and hard to stop.
However, these mechanisms share a structural assumption: the user is a pleasure-seeking agent whose cognitive biases cause her to miscalculate. Present bias means she overweights immediate pleasure relative to future pleasure. Variable reinforcement means she overestimates the probability of reward. FOMO means she overestimates the cost of missing out. In every case, the deviation is from a norm of rational pleasure-maximization, and the engagement, however compulsive, is still oriented toward pleasure. The framework cannot explain a user who scrolls while experiencing no pleasure at all, who recognizes in the moment that no pleasure is forthcoming, and who continues anyway. That user is not miscalculating. She is doing something that the theory’s categories cannot describe.
Kent Berridge’s incentive salience research provides the most suggestive bridge from the behavioral sciences. Thirty years of neuroscience have demonstrated that “wanting” (dopamine-mediated incentive salience) and “liking” (opioid-mediated hedonic impact) are neurobiologically dissociable (Robinson & Berridge, 2025). Rats depleted of 99% of brain dopamine still showed normal hedonic reactions to sugar; they still “liked” it, but lost all motivation to seek or consume it. Conversely, amphetamine injections into the nucleus accumbens increased “wanting” without enhancing “liking.” The dissociation scales to addiction: a progressive increase in drug wanting occurs without any parallel increase in drug liking, sometimes even despite liking the drug less.
Applied to platforms, the notification badge, the phone’s availability, and the pull-to-refresh gesture function as conditioned cues that trigger incentive salience. Users can want to check their feeds intensely while liking the experience minimally or not at all. Berridge’s framework identifies the neural substrate but does not address why the system is structured to exploit the dissociation, what kind of satisfaction the user actually gets, or how engagement relates to the broader structure of subjectivity. For that, a different theoretical register is required.
The Drive Does Not Seek Its Object
Lacan’s drive theory, developed in Seminars XI (1964), XIV (1966-67), and XVI-XVII (1968-70), reshapes Freud’s idea of drives. Lacan gives the four drive elements (pressure, aim, object, source) a symbolic and circular form. The drive’s pressure is constant—it does not change with time or use (Seminar XI, pp. 164-165). This never-ending pressure sets drive apart from biological needs, which fluctuate and rise or fall. The drive is always on, just like the platform feed. Both are constant systems that do not weaken with use. The drive is Lacan’s decisive innovation. Freud distinguished between the drive’s aim (Ziel) and its object (Objekt), but Lacan radicalizes the distinction: the drive can be satisfied without attaining its object because it is a partial drive and its aim is the return into circuit (Seminar XI, p. 179). The drive goes out from the erogenous zone, circles the objet a (the object-cause of desire, itself a void around which the circuit organizes), and returns. The satisfaction is not in reaching the object but in completing the trajectory. When you entrust someone with a mission, the aim is not what he brings back but the itinerary he must take (Seminar XI, p. 179).
The distinction between desire and drive is topological, not quantitative. Desire is metonymic: it slides along the signifying chain, each object encountered is never quite “it,” and the subject moves to the next. Desire is sustained by the promise that the next object will satisfy, a promise that is structurally incapable of fulfillment because the object of desire is constitutively lost. This is the logic of browsing with intent: searching for a specific piece of content, not finding it, trying a different search term. The user who browses with intent is operating in a state of desire. She wants something. The want sustains the search.
Drive is circular: it finds satisfaction not by reaching an object but by repeatedly circling its absence. The aim is always the circuit itself. As Zupančič notes, desire remains unsatisfied; drive is satisfied by the loop. For the drive, satisfaction comes not from getting what you want, but from the loop repeating. This is the scroll: the thumb moves, content appears, it’s not satisfying, so the thumb moves again. The aim is the circuit.
All drives are partial: they represent not the reproductive function of sexuality but only the dimension of enjoyment (Seminar XI, p. 177). The four partial drives organize around four objects: the oral drive (breast), anal drive (faeces), scopic drive (gaze), and invocatory drive (voice). Platform engagement activates principally the scopic drive, the compulsive looking, scrolling through images, the visual consumption of feeds, and the invocatory drive, the compulsion to post, to voice, to be heard, to receive the notification sound that signals someone has responded. The three grammatical voices of the drive map onto the structure of social media use with uncomfortable precision: to see (scrolling), to see oneself (curating a profile), and se faire voir (to make oneself be seen; posting). The circuit completes when the post generates a response, or does not, and the cycle begins again.
Lacan’s claim that every drive is virtually a death drive is not biological mysticism. The death drive names the drive’s indifference to the pleasure principle, its excessive character, its resistance to regulation. Freud identified the compulsion to repeat (Wiederholungszwang) in “Beyond the Pleasure Principle” (1920): patients repeated traumatic, unpleasurable experiences against the dictates of the pleasure principle, driven by a force more primitive than the pursuit of pleasure. Lacan reads this not as biology but as the insistence of the signifying chain: repetition belongs to the signifier, to what he calls the automatism of repetition. The compulsion to scroll, to check, to refresh is not a deviation from the pleasure principle. It is the drive circuit operating beyond it. The user does not scroll because it feels good. The user scrolls because the circuit demands completion, which in turn triggers another circuit.
Jouissance Names What Rational Choice Cannot
Jouissance is not pleasure. Pleasure operates according to the homeostatic principle: it reduces tension, maintains constancy, and keeps excitation manageable. Jouissance overruns it. Lacan describes jouissance as beginning with a tickle and ending in a blaze of petrol (Seminar XVII, p. 72). The pleasure principle functions as a barrier to jouissance: pleasure is what stops us at a respectful distance from jouissance (“Psychoanalysis and Medicine,” 1967). When the user scrolls past the point of pleasure, past the point where the content is interesting, past the point where the experience feels good, she has crossed from pleasure into jouissance. The scroll that continues despite dissatisfaction is not irrational pleasure-seeking gone wrong. It is jouissance operating as it always does: as painful pleasure, satisfaction in dissatisfaction, an enjoyment the subject does not want but cannot stop pursuing.
The phenomenology of doom-scrolling confirms the structure. The term entered common usage during the COVID-19 pandemic in 2020, and Merriam-Webster added it to the dictionary in 2023, defining it as spending excessive time online navigating unpleasant news, even when it is saddening or demoralizing. A 2024 Morning Consult survey found 31% of American adults doomscroll regularly; among Gen Z, 51%. The research confirms what users already know: daily social media exposure during the pandemic was associated with increased depression and PTSD symptoms, with the association strongest for those with preexisting vulnerability (Price et al., 2022). Users scroll through content that makes them feel worse. They know it makes them feel worse. They continue.
The standard explanations (uncertainty reduction, negativity bias, habit formation, and low self-control) describe proximate mechanisms but do not address the structure of the phenomenon. Reframed through drive theory: doomscrolling is a repetitive circuit in which the user scrolls, encounters content that circles around but never reaches the objet a, the algorithmically curated “almost-but-never-quite” relevant, important, or satisfying piece, and returns to scrolling. The aim is not the content. The aim is the scrolling itself, the return into the circuit.
Natasha Dow Schüll’s fifteen-year ethnography of machine gambling provides the empirical bridge between clinical drive theory and platform phenomenology. Her concept of the “machine zone” describes a trancelike state where daily worries, social demands, and even bodily awareness fade away (Schüll, 2012). Gamblers in the zone play not to win but to keep playing, for as long as possible, even at the cost of physical and economic exhaustion. An industry insider told Schüll that what the industry initially failed to understand was that people do not really want to be entertained; the best customers want to be totally absorbed, to get into a rhythm. The industry’s key metric shifted from “jackpot excitement” to “time on device,” maximizing duration regardless of outcome. The parallel to platform engagement metrics is direct. “Time on device” is the platform industry’s foundational metric, and it measures not satisfaction but circuit duration.
Schüll documented how machine design deliberately obscures the relationship between player action and machine outcome, creating an opacity structurally homologous to algorithmic opacity. The appearance of slot-machine symbols bears no representational connection to the technology that generates them. The machine’s interior logic is inaccessible to the player. The zone, Schüll argues, is not a failure of rational control but an emergent property of the interaction between a human subject and an opaque, responsive system designed to sustain engagement. Chris Hayes called the book essential reading for anyone interested in our experience of the 24/7 casino that is now life online. The infinite scroll is the digital equivalent of the zone-producing machine: optimized not for satisfaction but for sustained circuiting.
The Algorithm Sustains the Circuit
The previous essays in this series established that the algorithm occupies the structural position of the big Other, that users relate to it through the transferential mechanism of the subject supposed to know, and that the opacity of the algorithm’s “desire” generates the Che vuoi? that drives engagement. The drive circuit depends on this opacity. If the Other’s desire were transparent, the circuit would dissolve. The drive requires an object that is never fully attained, an Other whose desire is never fully legible, a question that is never definitively answered.
A 2025 preregistered study empirically confirmed the mechanism. Researchers compared Twitter/X’s engagement-based ranking algorithm to a reverse-chronological feed and found that the algorithm amplified emotionally charged, partisan, out-group-hostile content by 0.24 standard deviations in partisanship and 0.24 standard deviations in out-group animosity relative to chronological ordering (Milli et al., 2025). The finding that matters for the drive argument: users did not prefer the algorithm-selected political tweets. The algorithm underperformed their stated preferences while outperforming a chronological feed on engagement metrics. Engagement and satisfaction diverged. The algorithm optimized for the drive circuit (sustained engagement, continued scrolling) rather than for desire (finding what you actually want).
William Brady and colleagues demonstrated the reinforcement mechanism: platforms provide social feedback for expressions of moral outrage, increasing the likelihood of future expressions through operant learning (Brady et al., 2021). Even if platform designers do not intend to amplify outrage, design choices aimed at maximizing engagement indirectly affect moral behavior because outrage-provoking content draws high engagement. The PRIME framework identifies the content categories that algorithms amplify because humans are biologically biased to attend to them: Prestigious, In-group, Moral, and Emotional content. Environments become oversaturated with PRIME information, and users learn to produce more of it, creating a feedback loop between the algorithmic Other and the user’s drive circuit.
Jodi Dean’s analysis of communicative capitalism identifies the structural logic at its most abstract: in networked communication, the message has been replaced by the contribution (Dean, 2010). A contribution need not be understood; it need only be repeated, reproduced, or forwarded. Circulation is the content, the condition for the acceptance or rejection of a contribution. Who sent it is irrelevant. Who receives it is irrelevant. Whether it is responded to is irrelevant. The only thing that matters is that it enters the circuit. This is the logic of drive applied to the social field: the compulsive loop of posting, checking for responses, posting again enacts Lacan’s formulation that the aim of the drive is the itinerary, not what is brought back. Dean captures the loop’s temporal structure: new media forms become short loops that lock the subject into repeated attempts at enjoyment, where enjoyment is no longer the lost object of desire but the object of loss itself.
Lacan’s concept of surplus jouissance (plus-de-jouir), developed in Seminars XVI and XVII, makes the economic dimension explicit. The concept builds directly on Marx’s surplus value (Mehrwert): just as surplus value is extracted from labor beyond what is required for the reproduction of labor-power, surplus jouissance is extracted from the subject’s engagement in the signifying chain beyond what would sustain functioning. The ambiguity of plus is structural: it means both “surplus” (more jouissance) and “no more” (the cessation of jouissance). Each scroll that delivers disappointing content simultaneously shuts down jouissance and produces surplus jouissance. The drive circuit extracts its yield through the near-miss, the almost-satisfying, the content engaging enough to sustain the loop, but never satisfying enough to conclude it.
Applied to platform capitalism: algorithms optimize for surplus jouissance extraction. They identify what triggers the drive circuit, amplify those triggers through recommendations, create feedback loops that intensify compulsive engagement, and capture the surplus data generated. The platform’s metric of success, time on device, is functionally equivalent to the extraction of surplus jouissance from the user’s drive circuit. Shoshana Zuboff’s surveillance capitalism describes the economic logic of this extraction: platforms harvest “behavioral surplus,” data beyond what is needed for service improvement, and feed it into prediction products traded in behavioral futures markets (Zuboff, 2019). A Lacanian supplement does not contradict Zuboff but adds a dimension: the user is not only a victim of data extraction but a subject captured in a drive circuit that provides jouissance. Both surpluses, data and jouissance, are extracted from the same circuit. The dual extraction is what makes platforms so resistant to reform.
Addiction Is the Wrong Name
The dominant discourse frames compulsive platform engagement as addiction. Tristan Harris calls smartphones “slot machines in your pocket.” Adam Alter titled his book Irresistible. The documentary The Social Dilemma (2020, watched by 38 million Netflix households) presented platform design as a deliberate scheme to create dependency. The narrative generates a specific political grammar: tech companies are pushers, users are victims, the solution is treatment (digital detox, screen time limits, willpower training) or prohibition (age restrictions, usage caps).
The addiction frame has a structural problem that no amount of empirical refinement can fix. It assumes that rational, autonomous self-regulation is the default human condition and that compulsive engagement represents a pathological deviation from this norm. The subject, on this account, was functioning properly before the platform intervened. The platform introduced a distorting influence. Remove the influence, and the subject returns to baseline rationality. This is why the “social media addiction” discourse generates individual solutions: the dysfunction is located in the user’s brain, willpower, or psychology. An industry of digital detox clinics, self-help books, and screen-time apps has emerged to treat a condition that the DSM-5 does not recognize and the scholarly literature increasingly doubts exists as a coherent diagnostic entity (Kardefelt-Winther, 2014; Billieux et al., 2015; Panova & Carbonell, 2018).
The scholarly critiques are correct that the addiction frame overpathologizes. But they typically stop at recommending “problematic use” as a replacement term, which merely softens the pathological framing without providing an alternative structural account.
Lacan’s drive theory inverts the assumption. Drive is not a pathological deviation from rational agency. It is the normal mode of the subject’s engagement with the symbolic order. All subjects are driven. The question is not whether one is captured in drive circuits but which drive circuits one is captured in. There is no pre-drive subject who could, if only freed from distortion, engage platforms “rationally.” The rational agent of economic theory is a theoretical fiction; the driven subject of psychoanalysis is the structural reality. As Todd McGowan argues in Capitalism and Desire (2016), capitalism’s power lies not in how it represses or exploits but in how it provides jouissance: an incomplete satisfaction that compels us after the new, the better, the more. Our jouissance lies in the paths we take around the object, in how we circle the object without ever getting it. Enjoyment is the path of loss.
The reframing has political consequences. The addiction frame locates the dysfunction in the user and generates therapeutic interventions that leave the platform architecture untouched. A drive-structural analysis locates the dysfunction in the system: the business model that requires sustained engagement regardless of user satisfaction, the algorithmic opacity that sustains the Che vuoi? structure, the design features that instantiate the drive circuit’s loop. Addiction invites therapy. Drive analysis invites regulation, redesign, and structural critique.
Frances Haugen’s 2021 disclosures revealed what this looks like in practice. Facebook’s internal research found that 64% of all extremist group joins were due to recommendation tools. A 2018 algorithm change designed to boost “Meaningful Social Interactions” rewarded outrage and incentivized sensationalism because the most “meaningful” interactions, comments, shares, and reactions were disproportionately triggered by inflammatory content. The company understood the dynamic and chose engagement over safety because the business model demanded it. The algorithmic Other’s “desire” to maximize engagement was structurally misaligned with user welfare, and this misalignment was the condition of sustained drive, not its accidental byproduct.
Aza Raskin, who invented the infinite scroll in 2006, estimated that the feature wastes 200,000 human lifetimes each day. He described it as one of the first products designed not to help a user but to keep them online for as long as possible. The design does not exploit a pre-existing pathology. It creates the conditions for a drive circuit that captures subjects who were not “addicted” before they encountered the system and who are not “addicted” after they leave it, but who, while engaged, operate according to a logic that no framework premised on rational pleasure-seeking can describe.
The Circuit and the Third
The series has argued that Lacan’s analytic triad maps onto the platform coordination triad, that the algorithm functions as a material interpretant in Peircean terms, that users relate to the algorithm through the transferential mechanism of the subject supposed to know, and that they construct fantasies about its operations that function as enabling fictions. The drive argument completes the picture by identifying what the fantasmatic relation to the algorithmic Other produces: not pleasure, not utility, not satisfaction in any sense that rational choice theory or behavioral economics can recognize, but jouissance, the painful, excessive, compulsive satisfaction that the drive circuit generates through its own repetition.
The algorithmic third is the structural condition for this circuit. Without an opaque Other whose desire remains unclear, the question Che vuoi? could not be posed. Without that question, the fantasy that organizes the user’s engagement could not be constructed. Without the fantasy, the drive circuit would lack the object (the objet petit a, the cause of desire, the void around which the circuit organizes) it needs to circle. And without the circling, the compulsive engagement that platforms depend on, that users describe as toxic and addictive and time-wasting and soul-destroying, and that they return to tomorrow, would dissolve.
The implication for algorithmacy, the competency the ALC framework describes, is that platform literacy cannot be modeled on rational self-regulation. Teaching users to “make better choices” about screen time assumes the rational agent that drive theory has shown to be a fiction. Algorithmacy is not the capacity to resist the drive circuit through willpower. It is the capacity to recognize the circuit as a circuit, to identify the structural conditions that sustain it, and to develop a different relation to one’s own jouissance. This is closer to what Lacan described as the end of analysis than to what behavioral economists describe as debiasing. The analysand does not stop being a driven subject. She changes her relation to the drive. She recognizes the circuit without being captured by it. She scrolls, or does not, with something other than compulsion.
The best dissertation is the one that is completed. The best analysis of platform engagement is one that names what the user is actually doing, in the user’s own terms, without translating it into a language that flatters either the user’s self-image or the economist’s model. The user is not irrational. The user is not addicted. The user is driven, and the system is built to drive her. The question is not how to fix the user. The question is: what kind of system are we building, and what kinds of subjects does it produce?
References
Allcott, H., Braghieri, L., Eichmeyer, S., & Gentzkow, M. (2020). The welfare effects of social media. American Economic Review, 110(3), 629-676.
Allcott, H., Gentzkow, M., & Song, L. (2022). Digital addiction. American Economic Review, 112(7), 2424-2463.
Billieux, J., Schimmenti, A., Khazaal, Y., Maurage, P., & Heeren, A. (2015). Are we overpathologizing everyday life? A tenable blueprint for behavioral addiction research. Journal of Behavioral Addictions, 4(3), 119-123.
Brady, W. J., McLaughlin, K., Doan, T. N., & Crockett, M. J. (2021). How social learning amplifies moral outrage expression in online social networks. Science Advances, 7(33), eabe5641.
Cho, H., Ratan, R., Jeong, H., Palilingan, V., & Santos, M. (2021). In-the-moment regrets of social media use. Proceedings of the ACM on Computer-Supported Cooperative Work.
Dean, J. (2010). Blog Theory: Feedback and Capture in the Circuits of Drive. Polity Press.
Freud, S. (1920). Beyond the Pleasure Principle (J. Strachey, Trans.). SE XVIII.
Hoong, R. (2019). The behavioural economics of social media: A study of self-commitment devices. Stanford University.
Kardefelt-Winther, D. (2014). A conceptual and methodological critique of internet addiction research: Towards a model of compensatory internet use. Computers in Human Behavior, 31, 351-354.
Lacan, J. (1998). The Seminar of Jacques Lacan, Book XI: The Four Fundamental Concepts of Psychoanalysis (A. Sheridan, Trans.; J.-A. Miller, Ed.). W. W. Norton.
Lacan, J. (2007). The Seminar of Jacques Lacan, Book XVII: The Other Side of Psychoanalysis (R. Grigg, Trans.; J.-A. Miller, Ed.). W. W. Norton.
McGowan, T. (2016). Capitalism and Desire: The Psychic Cost of Free Markets. Columbia University Press.
Milli, S., Carroll, M., Wang, Y., Pandey, S., Zhao, S., & Dragan, A. D. (2025). Engagement, user satisfaction, and the amplification of divisive content on social media. PNAS Nexus, 4(3), pgaf062.
Panova, T., & Carbonell, X. (2018). Is smartphone addiction really an addiction? Journal of Behavioral Addictions, 7(2), 252-259.
Price, M., Legrand, A. C., Brier, Z. M. F., Adams, Z., & Lucas, G. (2022). Doomscrolling during COVID-19: The negative association between daily social and traditional media consumption and mental health symptoms. Psychological Trauma, 14(8), 1338-1346.
Robinson, T. E., & Berridge, K. C. (2025). The incentive-sensitization theory of addiction 30 years on. Annual Review of Psychology, 76, 29-58.
Schüll, N. D. (2012). Addiction by Design: Machine Gambling in Las Vegas. Princeton University Press.
Zuboff, S. (2019). The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. PublicAffairs.
Žižek, S. (1989). The Sublime Object of Ideology. Verso.