Abstract
Sleep is a state of vital recursive unavailability. In ordinary good sleep, living coordination withdraws from waking L3 and higher engagement and settles into predominantly L1 and L2 multimediated coordination, the sleeper ceasing to answer ordinary demands while embodiment continues its own metabolic work regardless. Safe sleep therefore requires a mesocosm in which a living being can temporarily relinquish vigilance without being harmed for doing so. This article argues that sleep disruption becomes a form of violence, rather than the lesser but still serious category of sleep harm, when interference with that necessary withdrawal is coercively imposed and either deliberately used, knowingly tolerated, or institutionally treated as an acceptable cost. Its violence does not consist merely in making someone tired. It lies in forcing a living being to remain available exactly when metabolic restoration requires unavailability, by a power that has stopped needing to answer for the demand. The concentration camp, read through Wolfgang Sofsky's sociological account, supplies the paradigm case, since virtually every condition necessary for sleep, time, bodily security, temperature, spatial territory, predictability, protection from intrusion, and confidence that a formally announced closure would actually hold, was systematically destabilized there at once, and the camp's own internal hierarchy shows that the asymmetry between who can see and who can be seen was itself delegated downward rather than held by the SS alone. The article then argues that Michel Foucault's contrast between productive discipline and destructive power, and Sofsky's own still sharper contrast between disciplinary institutions and camp terror, capture a real difference but risk making it categorical. Sleep exposes an unstable boundary running underneath that difference rather than a single continuous scale, since the same technique can serve either purpose depending on its relation to the organism's possibility of restoration, and no institution's form settles the question in advance. The article closes with an account of sleep disruption's three mechanisms and their cumulative consequence, a comparative survey running from interpersonal abuse through war and total institutions, and a positive counter-image: protection as the willingness to remain available so that someone else may safely cease to be.
I. When Being Kept Awake Becomes Violence
Someone wakes another person. A light stays on. A door bangs. A child cries. A guard calls reveille. A supervisor extends a shift. A bomb makes sleeping unsafe. None of these acts is intrinsically an act of violence. The conceptual problem this article sets out to solve is not how to classify waking someone as violent in general, but how to say, precisely, what distinguishes the ordinary interruption of a night's sleep from the cases in which interrupting it becomes one of the ways power reaches into a life.
The concentration camp supplies the clearest case in the historical record. Wolfgang Sofsky's sociological account of the camps describes a day beginning around four or half past four in the morning, prisoners awakened by clanging bells, whistles, or sirens, with those who did not rise immediately beaten from their bunks. Within thirty or forty-five minutes, an exhausted, still half-conscious body had to dress, clean the block, make its bed to a precision that left no margin for the time actually available, use the latrine, and assemble for roll call. The day did not simply begin early. It began through a violent conversion from sleep into total, immediate, compulsory availability, with nothing resembling the drowsy, self-paced waking an ordinary morning allows.
The other end of the day offered no more reliable a boundary. Evening roll call had no fixed limit, and every additional minute it lasted was, on Sofsky's own account, a minute subtracted directly from the prisoners' time for eating, relaxation, and sleep, a deliberate use of time itself as a form of punishment. In the most extreme cases, punitive roll calls extended through winter nights until prisoners collapsed or died where they stood. And even once lights went out at nine in the evening, the night carried no guarantee of closure, since arbitrary nighttime incursions by drunken SS officers remained possible at any hour, for reasons examined in full below. Sofsky opens his entire historical narrative with a scene of exactly this kind: the earlier, comparatively tolerable regime at Dachau broken apart when sleeping prisoners are woken by the thud of marching feet and the clang of weapons as an SS unit takes control of the camp. Terror, on the very first page of this history, announces itself as an interruption of sleep.
This licenses the question the rest of the article tries to answer with some precision. What kind of violence consists in making it impossible for another person safely to cease responding? Sleep disruption becomes violence when power penetrates the interval in which a living being must become recursively unavailable in order to undergo metabolic restoration, and does so coercively rather than merely negligently, a distinction the article develops fully once its typology is in place.
One clarification has to be made explicit from the first page. The concentration camp appears in what follows as a paradigm, the case in which the structures this article is interested in become maximally visible, not as a scale against which lesser instances of disrupted sleep are to be measured or, worse, morally equated. Sofsky himself insists that structural comparison across cases need not diminish historical difference or moral responsibility, and this article's method depends on the same insistence. Nothing that follows treats a bad night's sleep as comparable to a night in Sofsky's camps. What it does claim is that the same underlying structure, power reaching into the interval where vigilance would otherwise be allowed to lapse, is visible at very different scales and with very different moral weight, and that the camp is where that structure appears with the least possible ambiguity.
II. Sleep as Recursive Unavailability
The camp can only be read this way once the theoretical apparatus that makes the reading available has been set out, and this section supplies it.
Sleep is not inactivity. It is a state of continuous, highly organized physiological and metabolic activity, not a kind of blank interval or mere absence. Embodiment continues throughout sleep, in breathing, thermoregulation, and the ordinary work of bodily maintenance that never actually stops, whatever further physiological detail sleep science might add to that list. This article brackets that further detail deliberately, and brackets too the whole question of dreams, parasomnias, sleep talking, and the partial reactivations of symbolic or interrecursive activity such episodes can involve, since its concern is the ordinary organization of sleeping as a mesocosmic state rather than a theory of what happens inside it. Being-with is attenuated but remains present and important, a partner's or a child's proximity still mattering to a sleeping body without anything resembling waking interaction. Dwelling persists throughout, temperature, darkness, sound, weather, and the underlying circadian rhythm, all continuing to condition the sleeper whether or not any of it registers. Multimaterial arrangements, beds, walls, blankets, doors, continue operating precisely because they require no conscious supervision to keep doing their work. What largely disappears during ordinary sleep is not any of this. It is waking, L3 and higher, recursive engagement, the capacity to articulate, question, and revise. Sleep is accordingly a condition in which a living being remains fully alive, indeed intensely occupied, while becoming radically less answerable to whatever surrounds it.
Good sleep, on the strongest available formulation, tends towards predominantly L1 coordination with selective L2 permeability. The obvious mistake is to imagine that better sleep simply means less responsiveness across the board. A sleeper who failed to respond to every minor creak or momentary discomfort would indeed be sleeping more deeply in one narrow sense, but a sleeper who failed to respond to smoke, to a crying infant, to pain, to an intruder, or to genuine physical danger would not be sleeping well in any sense worth having. Successful sleep depends on an unusually sophisticated threshold, one capable of sorting, continuously and without conscious supervision, between differences that can safely be ignored and differences relevant enough to justify waking.
This threshold is exactly what a certain kind of violence learns to attack, and identifying that attack precisely gives this article its working definition of sleep disruption. Violence need not remove sleep altogether to count. It can instead make a sleeper's surroundings unpredictable enough that potentially relevant differences repeatedly cross the threshold from L1 sedimentation into L2 felt disturbance, whether a given sound, an approaching presence, an expected summons, a nearby body's movement, or the risk attached to failing to wake instantly. The sleeper is not silently interrogating these possibilities; the discrimination occurs below articulation, in a heightened and chronic permeability to exactly the differences an untroubled sleeper would leave safely sedimented. Call this forced recursive availability. Sleep disruption of this kind is not simply the removal of hours of sleep. It is the coercive maintenance of recursive availability across a period in which the organism requires withdrawal, a definition that immediately accommodates terror, abuse, war, and bullying without reducing any of them to a stopwatch measurement of hours lost, and without requiring that sleep be prevented outright before anything counts as disruption at all.
A sleeper trusts, for the duration, that the walls will not move, that the temperature will not become dangerous, that whoever else is present will not exploit the interval, and that whatever does require a response will actually be loud, close, or urgent enough to cross the threshold in time. None of this trust is usually experienced as trust at all. It is simply what allows sleep to happen without being consciously monitored, which is exactly why its removal is so easy to overlook as a form of harm in its own right. Nobody has to strike a sleeping body to have already reached it. Making the surrounding world untrustworthy in this specific, narrow sense is enough.
A further piece of the apparatus, already developed elsewhere in this corpus, forms the bridge into the historical material. An ordinary bedroom works, when it works, by arranging for almost nothing relevant to happen inside it: the walls remain walls, the bed remains still, the door stays closed, the temperature stays within bounds, and whoever else shares the space largely refrains from making demands. A successful sleeping mesocosm is, in this precise sense, an anti-event technology, engineered around the deliberate prevention of anything requiring a response. Terror reverses this arrangement rather than merely damaging it. The terrorized night is one in which something important may happen at any moment precisely because one has become unavailable, so that unavailability itself, the very condition sleep requires, becomes the condition terror is organized to exploit. That reversal is the bridge into the concentration camp, the historical case in which an entire institution was built, whether by explicit design or by the cumulative logic of absolute power, around exactly this inversion.
III. The Concentration Camp as the Destruction of the Sleeping Mesocosm
Sofsky's account does not primarily describe a specialized, deliberate policy of sleep deprivation, in the sense of a single targeted technique with sleep as its named object. Something more thorough is happening. The camp destroys the entire ecology within which restorative sleep could ever become secure, reaching it from every direction at once rather than through one isolated mechanism.
Reveille supplies the first and most immediate case. Waking at four or half past four is followed, on Sofsky's account, immediately by threat, shouting, physical punishment, and an extreme compression of time: thirty or forty-five minutes to dress, clean the block, make a bed to an exacting standard, use the latrine, and assemble. There is no leisurely waking of the kind an ordinary morning permits, no interval in which a body might gather itself before being asked to function. The significant feature is not simply that sleep was cut short, though it was, but that an institution controlled the entire transition from recursive unavailability into total availability, compressing what is, for an ordinary sleeper, a gradual and self-paced process into an instant, coerced conversion. The sleeper becomes, within minutes, countable, mobile, inspectable, punishable, and labour-ready all at once. Sleep terminates institutionally, on the guards' schedule and at the guards' signal, rather than metabolically, on any schedule belonging to the organism itself.
The evening roll call supplies the sharpest single scene in the whole record. Unlike other inspections, it carried no fixed time limit, and Sofsky is explicit about what this meant in practice: every additional minute it lasted was a minute subtracted directly from the prisoners' time for eating, relaxation, and sleep, its prolongation functioning as a temporal sanction, a form of punishment that used time itself as the instrument. A further, still more revealing detail lies buried inside this practice. When a prisoner was missing from the count, an exhaustive search would follow, and very often, Sofsky notes, there had been no real escape at all. Some prisoner, exhausted past the point of endurance, had simply curled up in a hidden corner, a sewer pipe, a pigpen, beneath the floor of a barracks, and fallen asleep exactly where they lay. If found, that prisoner was beaten, dragged to the assembly ground, and killed, treated by the institution as a fugitive rather than as a body that had done nothing more than what an exhausted body eventually and involuntarily does. This is the scene this article treats as its central image, more analytically concentrated than reveille itself, because a single event compresses metabolic necessity and institutional interpretation into direct collision. The organism finally enforces sleep because metabolic exhaustion has become total and can no longer be overridden. The institution reads that involuntary surrender as insubordination. The body sleeps. Power interprets the sleep as disobedience.
The scene is also an exact case of asymmetrical interlegibility, the condition in which one party can model another's state, intentions, and vulnerability with far greater clarity than the reverse. A guard searching for a missing prisoner reads exhaustion on a discovered body with complete confidence: the slack posture, the closed eyes, the failure to respond to a shout, all legible at a glance as sleep rather than escape. The prisoner curled into that sleep has no comparable access to the guard's side of the encounter. They cannot know whether a search is already underway, whether discovery will be read as escape or exhaustion, or what an exhausted body found in a hidden corner will be taken to mean by whoever finds it first. The camp did not simply see more of the prisoner's body than the prisoner could see of the camp's intentions. It converted that asymmetry into the very mechanism by which sleep became indistinguishable from disobedience: the guard's legibility into the prisoner's metabolic collapse was total, and the prisoner's legibility into how that collapse would be classified was zero.
This scene supplies the cleanest way of stating what the camp actually withheld. Any regime that organizes sleep at all grants, at minimum, time set aside for it and a formal permission for that time to be used. The camp granted both nominally, a schedule specifying when prisoners might sleep, and lights-out marking the moment such sleep became formally permitted. What the camp withheld was a third and decisive element: security that either grant would actually hold once given. Call these three, together, time to sleep, permission to sleep, and security while asleep. A disciplinary institution can control the first two without thereby committing anything this article would call violence, provided the third is genuinely protected. Terror is what remains once the third has been destroyed while the first two are still nominally in place. The camp did not simply take sleep away. It rendered the permission to become unavailable revocable at every moment of the night, a more exact and more damning description than any account of hours lost could supply.
Night itself offered no more reliable a boundary than the day did, and supplies the clearest illustration of security's absence from the triad just introduced. Lights went out at nine in the evening, a fixed and formally announced closure, permission in exactly the sense just defined, and yet Sofsky describes this closure as carrying no actual guarantee. Drunken SS officers might force prisoners outside barefoot into mud or snow for no announced reason at all, and Sofsky is explicit that the function of such incursions was to keep anxiety ceaseless and render terror omnipresent. The schedule announces that the day has finished. The lived world underneath the schedule continues to insist that nothing has actually finished, that the interval nominally set aside for withdrawal remains fully penetrable by power at any moment of the institution's choosing. Sleep, under these conditions, never actually closes, whatever the posted timetable says. This compresses everything the triad above was reaching for into one formulation: lights-out is symbolic closure without mesocosmic closure. An L4 schedule can declare an interval closed. Only the L1 and L2 mesocosm underneath that declaration can actually make the closure hold, and in the camp it never did.
The barracks supply the spatial and material half of the same destruction. Sofsky describes minute, exacting regulations governing how prisoners could sleep, including restrictions on what could be worn even in bitter cold or under fear of theft, and a requirement that beds be made with geometrical precision under time pressure that made the standard nearly impossible to meet, generating further, entirely predictable occasions for punishment. More fundamentally, severe overcrowding destroyed whatever personal territory sleep would ordinarily require. Some blocks operated a hot-bed system, prisoners sleeping in relays on bedding that never cooled between occupants, and Sofsky records that as many as forty prisoners could be packed onto a bunk designed for a small fraction of that number. Where crowding of this severity applied, prisoners had no meaningful choice of who lay beside them, fought over what little room, warmth, or blanket existed, and slept with a stranger's body pressed against them through the night, proximity governed entirely by touch, heat, smell, and the ordinary indignities of bodies that could not move. In one of the book's most striking single observations, describing exactly this destruction of personal space, Sofsky writes that the only way out was sleep, prisoners packed so completely into a mass of other bodies, and so powerless to protest, withdraw, or move, that unconsciousness became the sole remaining exit from an unbearable present. That sentence sits remarkably close to this article's own account of sleep as withdrawal from intolerable interrecursive exposure, though Sofsky does not go on to ask what kind of power it is that can reach into, and ultimately compromise, even that last resort.
Asymmetrical interlegibility in the camp was not a single relation running from the SS down to an undifferentiated prisoner population. Sofsky's fuller account of camp self-administration shows it delegated, in hierarchical stages, into the prisoner population itself. The SS appointed senior camp prisoners at the apex of this internal structure, who selected block chiefs for each barracks, who in turn appointed room chiefs and barrack-room assistants whose task, on Sofsky's own account, was explicitly to help keep the other prisoners under surveillance. Each rung gained the capacity to model, monitor, and report on the rung below with a clarity the rung below could not return upward. A room assistant could see who was too weak to rise at reveille, who was hoarding food, who might inform, and who had collapsed into sleep out of turn. An ordinary prisoner could not see whether a block chief's cruelty on a given day was personal, ordered from above, or a performance staged for a watching SS officer, since dual subordination meant every functionary answered simultaneously to the prisoners nominally under their care and to the SS who could remove or punish them at will.
This delegation bought its own legibility with the functionaries' own bodies. Sofsky records that block chiefs and their assistants occupied a separate, reserved corner of the barracks, kept their own beds and lockers, and in the most privileged blocks held single beds where ordinary prisoners lay two to a bunk or worse. The camp's internal elite literally slept in security purchased by the surveillance they performed on those sleeping around them. The same structure repeated itself in the work Kommandos. Kapos kept work sites under continuous surveillance, were authorized to report infractions, and could use violence without restraint, and Sofsky's account of their incentives closes the loop this article has already traced between labour and sleep: Kapos had a direct interest in the output of their own Kommando, so that a worker too exhausted to keep pace threatened the Kapo's own position, converting one prisoner's fatigue into another prisoner's incentive to drive them harder still, with violence serving, on Sofsky's own analysis, the double function of extracting labour and manufacturing the disorder that justified the Kapo's continued usefulness to the SS.
None of this makes the camp's terror, at bottom, prisoner against prisoner rather than SS against prisoner. The delegation was itself an SS technology, designed, as Sofsky notes, to relieve its own personnel of the burden of directly monitoring every barracks and every work detail. What it demonstrates is that asymmetrical interlegibility does not require a single powerful observer facing a single powerless observed. It can be built as a cascade, in which each rung buys a margin of security, privilege, or exemption from immediate violence by rendering the rung below newly visible, exhausted, and exploitable, all the way down to the prisoner with no rung beneath them, whose sleep, exhaustion, and collapse remain permanently and one-directionally legible to everyone positioned above them in a structure that gave them no comparable way of looking back.
This matters because ordinary sleep requires something closer to bodily surrender to one's surroundings than most waking activity does. In an ordinary bed, the surface beneath a sleeper becomes trusted, nonrecursive ground, present and reliable without requiring any further attention. In the barracks Sofsky describes, the sleeping surface itself became contested social territory, fought over nightly, offering no comparable trust at all. The camp invaded sleep not merely by shortening the hours available for it but through every mediation sleep actually depends on at once: embodiment, forced into immobility and packed proximity; being-with, reduced to an unchosen, often hostile crowding rather than any protective closeness; multimateriality, in bedding, space, and warmth that could never be reliably secured; and dwelling, in a temperature and physical environment nobody controlled. What was destroyed was not simply time. It was the possibility of ever sedimenting the immediate surroundings into a reliable L1 background stable enough for sleep to proceed undisturbed.
Work supplies the final piece of this section's evidence, and closes the loop rather than opening a new one. Sofsky describes camp labour as severed from ordinary productivity, prisoners forced to conserve whatever energy they could wherever supervision briefly relaxed, because exhaustion itself had become a survival threat. Workdays could run to eleven or twelve hours, breaks were minimal, and work could be interrupted at any point by arbitrary violence unconnected to output. His larger argument is that camp labour could invert the normal relation between working and living: work no longer sustained life but consumed it, so that a prisoner who became incapable of work faced disposal, while continuing to work destroyed the very capacity that continued survival required. Read through sleep, this inversion becomes a closed loop rather than a single cruelty. Work consumes metabolic capacity beyond what any night can restore. Sleep, for all the reasons already given, is inadequate and insecure. Recovery accordingly fails. Capacity deteriorates further. The work that remains becomes still more exhausting relative to what the body can supply. Diminished capacity invites further violence. And restoration becomes less possible with every cycle rather than more. This is cumulative metabolic destruction in the fullest sense, a system that does not merely deprive a body of one resource but progressively removes its capacity to recover any resource at all.
Sofsky's own analysis of what he calls prisoner time supplies the theoretical closure this section needs. Unpredictable danger, on his account, demanded constant vigilance and presence of mind, collapsing consciousness towards the immediate present because no reliable future could be anticipated at all. Read through the apparatus this article has built, that observation becomes considerably sharper than Sofsky's own vocabulary allows it to be. The camp does not simply deprive prisoners of sleep as one deprivation among several. It makes the surrender of vigilance irrational, at exactly the moment metabolism makes that surrender necessary. To sleep is a metabolic requirement no institution can indefinitely override. To remain vigilant is, inside the camp, a situational requirement no institution allows a prisoner to set aside. Absolute power forces these two necessities into direct and permanent contradiction, which supplies perhaps the strongest definition of terror this article can offer from the perspective of sleep alone: terror is a mesocosm in which survival demands the very vigilance that metabolism cannot indefinitely sustain.
IV. Foucault, Sofsky, and the Unstable Boundary Between Discipline and Destruction
This section moves from the camp's empirical anatomy to the wider question of what kind of power sleep disruption reveals.
Foucault's central innovation is not the claim that disciplinary power is benign. It is the claim that modern power very often works by producing capacities rather than merely suppressing them: training bodies, organizing time, normalizing conduct, and increasing usefulness, so that the disciplined body emerges simultaneously constrained and made more productive than it was before. Sleep, on this account, is normally subordinated to exactly this productive logic. School, barracks, factory, and hospital schedules organize when bodies sleep because tomorrow's body must again be capable of functioning, and a regime that destroyed that capacity altogether would be defeating its own purpose.
Sofsky's distinction between disciplinary power and camp terror is considerably sharper than Foucault's own framework. Disciplinary schedules, on Sofsky's account, exist to create predictability, habituation, and cumulative capacity over time; camp power uses a schedule only as a kind of armature, repeatedly breaking it through suddenness, delay, harassment, and arbitrary intrusion, so that its temporality is never developmental or pedagogical in the way a disciplinary timetable's is. Disciplinary power, similarly, seeks a docile, functioning individual; camp terror seeks a universe of uncertainty, and can drain a body through labour that serves no productive purpose at all as readily as through labour that does. Sofsky's own epilogue states the contrast in its most absolute form: terror does not generate productive capacities. It turns production itself into ravagement, and aims, in the end, at the extinction of action and vitality rather than their cultivation.
The intervention this article makes is not a rejection of that distinction. Collapsing the concentration camp into ordinary discipline would erase precisely what Sofsky's whole analysis is built to demonstrate, and nothing here proposes doing so. The mistake in the other direction is to imagine that the boundary Sofsky draws, real as it is, is also categorical, a clean line separating two entirely different species of institution rather than an unstable relation that the same techniques can occupy differently at different moments. Consider the actual techniques involved on either side of the supposed line: reveille, barracks, bed inspection, timed routines, marching, shift systems, watch duties, restricted sleeping hours, collective sleeping arrangements, punitive exercise. Every one of these belongs, in more or less transformed form, both to institutions explicitly designed to produce disciplined capacity and, in the camp, to an institution explicitly designed to destroy it. Sofsky himself repeatedly notices the resemblance rather than denying it, observing that block discipline recalls military barracks and that camp violence sometimes resembles military hazing carried past any ordinary limit rather than a wholly unrelated phenomenon. No technique is inherently productive or destructive simply by virtue of its institutional form. Reveille is reveille whether it produces a soldier or destroys a prisoner. What differs is not the technique but what happens, downstream of it, to metabolic restoration.
This licenses a concept that does the work the Foucault-Sofsky distinction needs without abandoning that distinction's real insight. Productive discipline normally requires what might be called a metabolic floor. The worker, soldier, student, or patient a disciplinary regime is organizing must be preserved sufficiently to continue working, fighting, learning, or recovering, which means sleep, however compressed, badly distributed, or inconveniently timed a disciplinary regime makes it, is never actually external to that regime's own purposes. It is one of its own hidden reproduction requirements, quietly built into even the most demanding schedule, because a regime that destroyed recovery altogether would eventually destroy the very capacity it exists to extract.
Destructive power is power that no longer needs to respect that floor, provided the concept is treated as an ideal type rather than a binary switch a given institution simply has set one way or the other. Camp power, on Sofsky's own account, still wanted labour output from many of its prisoners some of the time, which means even the most destructive regime in the historical record did not operate with perfect indifference to metabolic capacity throughout. Productive institutions, conversely, can and do knowingly exhaust workers they regard as replaceable. Military organizations sometimes destroy bodies quite deliberately in pursuit of a strategic, productive end. A hospital can institutionally deprive its own staff of sleep for years while continuing to demand their highest level of performance from them, without any single actor involved intending destruction as such. The productive and destructive poles describe a variable and unstable relation between extraction and restoration that any real institution's techniques can occupy differently at different moments, rather than two mutually exclusive species of power sorted cleanly into disciplinary and totalitarian boxes, and rather than points on a single scale that would make camp terror merely a more intensified degree of ordinary discipline. No technique is inherently productive or destructive simply by virtue of its institutional form, a stronger and more exact claim than any image of a continuous scale could support. Sleep is what lets an observer see, with unusual precision, exactly how a given institution's techniques are actually related to restoration at a given moment, since it is restoration, more directly than almost anything else an institution touches, that reveals whether extraction has been allowed to run past the point any future use of the body could still justify.
None of this weakens Sofsky's historical argument. Sofsky is not claiming that camp power always, everywhere, and without exception abandoned every productive purpose. He is claiming, correctly, that camp power was prepared to abandon that purpose whenever terror required it, in a way no disciplinary institution, by definition, could remain a disciplinary institution and still permit. The metabolic floor concept sharpens this rather than disputing it. What distinguishes the camp is not that every one of its techniques was uniquely destructive in form, since reveille, roll call, and barracks discipline all have close disciplinary cousins outside it. What distinguishes the camp is that none of its techniques were reliably tethered to any floor at all, so that the same instrument that might, elsewhere, preserve a body for tomorrow could, here, be allowed to destroy it today without the institution suffering any corresponding loss of purpose.
V. War: The Organized Redistribution of Sleep and Vigilance
Broadening outward from the camp while keeping war as the closest comparison available, a further pattern becomes visible: no institution, however powerful, can abolish the metabolic requirement itself, only redistribute who must meet it and when.
Military technology can overcome distance, darkness, communication delay, and difficult terrain. It cannot overcome the need for metabolic restoration, which means every army, however advanced, faces a constant and unavoidable distribution problem: who, exactly, may become unavailable now, given that someone, somewhere, must remain available at every moment the wider organization continues to operate.
Sentry duty is the clearest institutional answer to that problem. Some soldiers remain recursively available specifically so that others may safely become unavailable, and the extreme historical severity of the punishment for sleeping on watch becomes considerably more intelligible once this structure is made explicit. The sleeping sentry has not violated an arbitrary rule invented for its own sake. They have broken an interrecursive delegation on which other people's actual sleep depended, converting a shared arrangement of mutual protection into a moment in which nobody at all was watching. Call this delegated vigilance: the organized transfer of recursive availability from those who are permitted to sleep to those who are required, for a time, to remain awake on their behalf.
Military training deliberately destabilizes its own recruits' sleep for a distinct and more ambiguous reason, one that sits directly on the boundary the previous section drew. Training regimes that impose sleep deprivation do so partly in order to produce a capacity to continue functioning under future disruption, meaning discipline here deliberately approaches the destructive threshold in the service of an eventual, productive goal, and it is exactly this proximity that makes the Foucauldian distinction blur rather than dissolve. A technique that damages an organism in the short term can be, and routinely is, justified institutionally as building a greater future capacity. The honest question this raises is not whether such training is disciplinary or destructive in the abstract, but at what point, for any given recruit, training through deprivation stops cultivating a capacity and starts merely consuming one that will not return.
Pharmacological intervention supplies a final, more recent variation on the same underlying problem. Stimulant use to sustain wakefulness across sustained campaigns has a long documented history in modern warfare, and the conceptual point it illustrates is precise: a stimulant can postpone the felt demand for sleep without in any way abolishing the underlying metabolic stake that demand exists to signal. Technology of this kind displaces the burden of sleep in time rather than eliminating it, buying an army additional hours of availability against a debt that a chemical alone can defer but never actually discharge.
Taken together, these military cases sharpen rather than dilute the argument developed through the camp. War, unlike the camp, generally still needs its soldiers to function again tomorrow, which is exactly why war organizes sleep through delegation and training rather than simply destroying the conditions for it outright the way camp terror was willing to do. But the same vocabulary applies to both, because both are, at bottom, institutions managing an identical, non-negotiable metabolic fact through very different relations to the metabolic floor introduced above. An army that crosses too far past that floor, exhausting its own soldiers past the point recovery training or sentry rotation can repair, stops being a fighting force capable of continuing the war it was organized to win, which is precisely the constraint camp terror, aimed at something other than continued useful function, did not have to observe. War therefore occupies a position between the two poles this article has been describing throughout, close enough to ordinary discipline that its sleep technologies remain recognizably protective in intent, yet close enough to destructive extraction, in its most severe forms, that the line between training a body and consuming one can become difficult to draw even for the institution doing the training.
VI. Other Forms of Sleep Violence
The camp and the army are extreme and organized cases. The same structure recurs, in smaller and less organized forms, across a much wider range of ordinary and semi-ordinary settings, and surveying several of them briefly shows that the theory developed above is not an artefact of the single historical case that generated it.
Interpersonal and sexual violence supplies the smallest-scale analogue to camp terror, and in some respects the most direct one. Such violence can exploit exactly the vulnerability of a sleeping body and the reduced availability of any protector during the hours when most households and communities are least alert. Where nocturnal threat recurs, a further and distinct violence typically follows the first: subsequent sleep itself becomes unsafe for the survivor, so that the original attack continues to operate long after it has physically ended, now working through the victim's own diminished ability to withdraw rather than through any further direct action by the original perpetrator. This connects directly to the delegated vigilance introduced above, inverted into its most damaging form. Where a functioning household or relationship depends on some member's trustworthy wakefulness or reliable absence during the night, an abuser occupies exactly the position sentry duty is meant to fill, and betrays it in the most direct way available, converting the one relation meant to make another person's unavailability safe into the mechanism that makes it dangerous instead.
Workplace bullying supplies a related but structurally distinct case, because here nobody need ever enter the victim's bedroom for the same mechanism to operate. Relational harm accumulated during the working day continues, metabolically, into the night that follows it, disturbed sleep functioning as one of the more externally legible traces by which such harm sometimes becomes visible to colleagues, investigators, or a court. This licenses a category worth naming in its own right: indirect sleep violence, in which a perpetrator need not physically wake a target at all. Threat and anticipatory rumination persist instead, so that the victim remains interrecursively captured by an unresolved daytime relationship while lying, physically alone, in an otherwise undisturbed bed.
War directed against civilian populations, rather than fought between organized militaries, makes an entire population's sleeping mesocosm unstable at once. Night ceases to be reliably uneventful. Shelter may or may not actually protect. Every unexplained sound requires interpretation before it can safely be dismissed. Family members remain alert on one another's behalf precisely because no institution is doing so for them. The mere possibility of attack, whether or not an attack ever actually follows, can be sufficient on its own to prevent the kind of recursive withdrawal sleep requires, which is exactly why counting hours slept is such an inadequate measure of what this kind of violence actually does. A population that technically obtains eight hours in a shelter each night, every one of them spent listening rather than resting, has not thereby escaped the harm this article is describing. The destruction here lies specifically in the inability to trust the night at all, independently of how many hours of it are technically available.
Prisons and solitary confinement raise a related question that deserves cautious treatment rather than confident empirical claims about any particular system, since the point here is structural rather than a survey of penal practice. Sleeping arrangements have shifted historically from collective wards and shared barracks towards private cells, an arrangement now widely considered more humane and more dignified for most purposes, yet the same arrangement can, in a different institutional register, be converted directly into a punishment through prolonged isolation, in which the near-total absence of any other social presence makes the reduced-vigilance state sleep requires considerably harder to achieve rather than easier. The broader diagnostic question a prison raises, whatever its specific regime, is who controls lights, noise, night checks, bedtimes, waking, privacy, and company, since the answer to that question reveals how deeply institutional authority reaches into embodied life more reliably than a prison's own formal rules do.
Hospitals and psychiatric institutions belong in this survey for a different and, in some ways, more useful reason: they make comparison ethically productive rather than merely illustrative. Nighttime checks, shared accommodation, and scheduled interruptions can all disrupt sleep while being genuinely intended, and often successfully functioning, to preserve health or save a life, which means not every institutional interruption of sleep counts as violence under the definition this article develops below. The relevant questions, in a hospital as much as in a prison or a camp, become questions of necessity, proportionality, predictability, the possibility of eventual recovery, the patient's or prisoner's own capacity to consent to or contest the interruption, and whether a given interruption is actually protecting the person it disturbs or merely serving the institution's own convenience at that person's expense. Framing the comparison this way gives the article criteria capable of doing real diagnostic work, rather than a piece of rhetoric that would make every night nurse's round sound like a small act of terror.
What unites this otherwise disparate set of cases, domestic violence, workplace harm, civilian bombardment, imprisonment, and medical care, is not their moral equivalence, which plainly does not hold, but their shared structure. In every one of them, some agent or condition, whether a person, an institution, or an indiscriminate danger, reaches into the interval a living being needs in order to stop responding, and the question that actually distinguishes violence from its absence is never simply whether that interval was shortened. It is whether the interval remained safely closed to further demand for as long as the organism required it to be, and, where it did not remain closed, whether anyone or anything had a legitimate and proportionate reason for reopening it.
VII. Total Institutions: Sleep Control Is Not Yet Sleep Violence
Monastic life provides a particularly clear case in which highly detailed regulation of sleeping and waking need not in itself constitute sleep violence. Catholic monastic orders have historically regulated where, when, and with whom their members sleep and wake in considerable detail, governing the hour of rising, the hour of retiring, interruptions for night office, and the physical arrangement of dormitories. Yet it would be absurd to conclude from the thoroughness of this regulation alone that total regulation of sleep simply amounts to violence.
The relation among purpose, participation, predictability, withdrawal, and restoration differs entirely between the two cases, even where the formal degree of regulation looks comparable on paper. This licenses a distinction among four related but genuinely different phenomena, easy to collapse into one another if the differences are not held apart deliberately. Sleep coordination is the ordinary organization of sleep for the sake of a household's or a community's shared living, present in any functioning family or institution and carrying no negative connotation whatsoever. Sleep discipline is the more formal, institutionally prescribed timing of sleep and waking, present in monasteries, schools, and militaries alike, and compatible with genuine restoration so long as the schedule it imposes respects the metabolic floor discussed earlier. Sleep extraction is the reduction of restorative opportunity specifically in order to increase some other output, present wherever an institution knowingly trades a participant's recovery for its own productive advantage. And sleep violence, the concept this article has been building throughout, is the coercive undermining or penetration of necessary restorative withdrawal, present wherever an institution or agent treats a participant's vulnerability during that withdrawal as available for exploitation rather than as something to be protected.
A monastery organized around a predictable and openly stated rule, an accepted purpose that requires its members to remain capable of prayer, labour, and community life the following day, and some meaningful possibility of raising a grievance or eventually leaving, sits comfortably within sleep coordination and sleep discipline, and nowhere near sleep violence, however early its bell or however communal its dormitory. Predictability, purpose, an accepted rule, protected restoration, and some possibility of exit or contestation, rather than any single variable such as consent on its own, are what keep this case on the discipline side of the boundary this article has been drawing throughout. Total institutions become analytically interesting precisely because they maximize an institution's capacity to decide when its members may become unavailable, which means the same total reach that makes a monastery's discipline so thorough is what would also make an abusive version of the same institution's control so dangerous, were its predictability, purpose, and openness to exit ever to be withdrawn.
This yields a compact diagnostic worth calling the sleep test for institutions, a short set of questions capable of revealing more about an institution's actual distribution of power than its own formal rules typically do on their own. Who controls when a person may stop responding. Can institutional demands still reach someone who has, by the institution's own rule, been granted permission to withdraw. Is the coming night actually predictable, or merely scheduled. Is adequate restoration genuinely protected, or only nominally allowed for. Can sleep be interrupted without justification, and if so, by whom. Who holds the authority to wake whom, and on what basis. Who watches while others sleep, and under what obligation. And what happens, specifically, to a body that can no longer remain awake. A monastery, a barracks, a hospital, and a concentration camp can each be run through this same short list, and the answers, rather than the buildings' outward resemblance to one another, are what actually sort them. The camp, run through this list, answers every question in the register of exploitation. The monastery, run through the same list, answers every question in the register of accepted, purposeful, predictable constraint. What the list supplies is a way of asking the right question of an institution rather than settling for its outward appearance, which is exactly what a formal comparison of bedtimes and dormitory sizes alone could never do.
VIII. A Typology of Sleep Disruption as Violence
The cases assembled across the preceding sections can now be pulled together explicitly. Defining violence simply as the fact of lost sleep will not do. Three distinct mechanisms, rather than a list of independent categories, describe how sleep is actually attacked, and a fourth term names not a further mechanism alongside them but their cumulative consequence when any combination of the three continues unresolved.
Temporal disruption names control exercised over bedtime, reveille, duration, the timing and frequency of interruptions, and enforced waiting, of exactly the kind the evening roll call demonstrates in its purest form, an institution literally subtracting minutes owed to rest and converting that subtraction into a deliberate instrument of punishment.
Spatial-material disruption names the conditions that make sleep physiologically or materially insecure regardless of how much time is nominally set aside for it: cold, overcrowding, absent or inadequate bedding, noise, unwanted light, and surroundings that cannot be trusted not to shift. Sofsky's account of the barracks, the hot-bed system, and forty prisoners to a single bunk supplies the clearest instance of this mechanism operating at its most severe.
Interrecursive disruption names the specific case of making withdrawal from another person impossible: threat, bullying, abuse, arbitrary surveillance, and unpredictable demand, all of them converting what should be low-bandwidth, trust-based being-with into the heightened L2 permeability described in Section II, this article's most distinctive signature of sleep under attack.
These three mechanisms are not parallel to a fourth. They are what generates a fourth, cumulative metabolic destruction, whenever expenditure is allowed to outpace restoration for long enough: fatigue, impaired functioning, the progressive inability to recover, and eventual exhaustion of a kind no single night could reverse even under ideal conditions. The labour-sleep loop traced in Section III is exactly this convergence in its most complete form, temporal disruption through the roll call, spatial-material disruption through the barracks, and interrecursive disruption through the constant threat of nighttime incursion and hierarchical surveillance, none of them acting in isolation, all three together producing the closed loop of metabolic destruction that section described. Most of the smaller-scale cases considered above combine at least two of the three mechanisms rather than one alone. Domestic abuse combines interrecursive disruption with, very often, its own cumulative metabolic consequences. Civilian bombardment combines spatial-material disruption with an interrecursive disruption that requires no direct human antagonist at all, only an indiscriminate threat capable of making an entire population mutually vigilant on one another's behalf.
One further distinction has to be drawn before a working definition can be stated responsibly, because without it the definition risks calling anything that damages sleep an act of violence, which would flatten exactly the moral distinctions this article has tried to keep visible throughout. Call sleep harm the serious, sometimes severe damage to restorative capacity that follows from disruption of any of the three mechanisms above, regardless of why it occurred. A negligently designed hospital rota, a landlord's failure to fix a broken window, an understaffed care system nobody has properly examined, can all produce sleep harm of real severity. None of them, on this article's account, need count as sleep violence. Reserve that stronger term for harm that is coercively imposed and either deliberately used as an instrument, knowingly tolerated once its effects are apparent, or institutionally treated as an acceptable cost of extracting some other output. The nighttime incursions already described are violence under every clause of that definition at once. A rota drawn up in good faith, never audited, and quietly ruinous to the staff working it, is harm that may need urgent correction without being violence in the same sense, unless and until whoever controls it learns of the damage and tolerates it regardless. This distinction is what keeps the concentration camp analysis from being diluted by its own generality. Violence, on this account, is not synonymous with anything that damages sleep, but is what happens specifically when damage to sleep is imposed, tolerated, or extracted by a power that has stopped needing to answer for it.
The three mechanisms and the harm-violence distinction together yield a single working definition, since every case considered above can be located somewhere within it. Sleep disruption becomes violence, rather than remaining the lesser but still serious category of sleep harm, when an asymmetric relation of power coercively interferes with another being's necessary capacity for restorative recursive unavailability, through temporal, spatial-material, or interrecursive disruption or some combination of the three, and does so as something deliberately used, knowingly tolerated, or institutionally treated as an acceptable cost.
A further qualification belongs beside the definition rather than after it. Even among cases that clear the coercion-and-toleration threshold just described, intent still matters enormously to how a given case should be judged. The nighttime incursions already described were carried out with deliberate, calculated terroristic intent, the clearest and most severe form violence under this definition can take. An institution that knowingly maintains a schedule it has been told is ruinous to recovery, but does so for administrative convenience rather than out of any wish to harm anyone, still meets the definition's coercion and toleration clauses, and still counts as sleep violence on this account, but belongs to a different moral register from calculated terror even so. The definition, built this way, keeps intent doing exactly the moral work only intent can do, distinguishing degrees of culpability among cases that are already violence, without asking intent to do the prior work of separating violence from harm in the first place. That earlier, more basic separation is what the coercion and toleration clauses are for.
IX. Protection as the Opposite of Sleep Violence: Watching Over Sleep
The substantive argument closes here on its positive counter-image rather than on the camp alone.
A recurring motif runs through nearly every example of sleep's protection this article has touched on: a parent watching over an infant, one partner remaining alert for another, a nurse checking on a sleeping patient, a sentry standing guard over an encampment of sleeping soldiers, and, across a great many religious traditions, a deity imagined as watching over sleeping humanity when no earthly protector is available. Children's stories return to the same image so often that it barely needs pointing out: safety, in the world these stories construct, means that somebody remains awake, or at least remains close, while the child does not have to.
The deepest form sleep protection can take, on the evidence this article has assembled, is a simple relational structure rather than any particular technology or institution: I remain available so that you may safely become unavailable. Violence, throughout every case this article has examined, reverses exactly this relation. I exploit your unavailability, or I force you to remain available to me, are the two directions that reversal can take, and every instance of sleep violence surveyed above falls into one or the other.
This gives the article an ethical formulation it can rest on without needing to argue for anything as contested as a universal right to sleep, a claim this article has deliberately avoided making. The basic stake identified throughout is narrower and, for that reason, considerably harder to dispute: protected recursive unavailability, the condition in which a living being can trust that becoming unavailable will not itself be used against them. Whatever legal or political language eventually develops around sleep, the underlying stake it would need to protect is already visible here, in the plain relational fact that somebody, somewhere, is willing to remain available so that somebody else does not have to be.
This is, in the end, the simplest possible description of what the concentration camp abolished, and what every lesser case of sleep violence surveyed in this article abolishes in its own, smaller register. Not sleep itself, which the organism will eventually take regardless, by force if necessary. The trust that made taking it safe.
X. Conclusion: Terror Is the Destruction of the Safe Night
Sofsky's concentration camp is, in his own analysis, an institution of organized uncertainty. A timetable exists, but does not bind the power administering it. Rules exist, but generate further occasions for arbitrary intervention rather than protection from it. Lights go out, but the night that follows never becomes trustworthy. Work consumes the body's capacity without ever guaranteeing its restoration. Spatial order manufactures disorder in the very act of imposing itself. Even involuntary sleep, produced by exhaustion too complete to resist, becomes an occasion for punishment rather than being recognized for what it is. Sofsky describes the cumulative effect of all of this as the destruction of continuity, trust, sociality, action, and, eventually, vitality itself.
No single scene in the record states this more exactly than the one this article has returned to as its central image: an exhausted prisoner curling into sleep in a hidden corner, found, and killed as a fugitive. The body sleeps. Power interprets the sleep as disobedience. That single collision, between a metabolic necessity no institution can indefinitely override and an institution that refused to recognize unavailability as anything other than insubordination, is the concentration camp's whole logic compressed into one event, and everything else this article has traced, the roll call, lights-out, the delegated hierarchy of barracks surveillance, the labour, is that same logic distributed across an entire institution rather than concentrated into a single night. Sleep brings these otherwise separate observations into a single frame that Sofsky's own vocabulary, built around space, time, and absolute power, does not quite supply on its own. The concentration camp stands as the paradigm of sleep violence developed in this article not because sleep deprivation was ever its explicitly named or centrally administered policy, but because it abolished, systematically and from every direction at once, the protected interval in which any organism could safely withdraw from power.
The Foucault-Sofsky argument can now be closed with the same care it was opened with. Foucault was right that modern discipline often produces capacities rather than merely crushing them. Sofsky was right that camp terror cannot be reduced to a more intense version of disciplinary normalization. But sleep shows why the boundary between the two can never be drawn as a single, categorical line. The same institution can discipline a body in one register and deplete it in another. The same timetable can coordinate a population's restoration or destroy it, depending on nothing about its outward form. The same barracks can protect its sleepers through collective security or expose them to permanent surveillance. The same military training can cultivate genuine resilience or cross, unnoticed even by those administering it, into metabolic destruction. The difference was never simply whether an institution organizes bodies, since every institution considered in this article does that. It is what kind of future that organization leaves those bodies still capable of having.
Productive discipline must preserve enough of the body for tomorrow. Destructive terror can suspend that requirement. Sleep is where that difference becomes metabolically visible, sooner and more precisely than almost anywhere else an institution touches a life. And because real institutions can simultaneously use, exhaust, preserve, and destroy the bodies within them, often without any single actor intending any of it as a whole, the boundary between the two is never as clean as the distinction itself suggests. Sleep disruption marks the point at which the organization of a life begins to turn into violence against that life's own capacity to restore itself.