Abstract
Sleep presents law with an unusually revealing problem of legibility. It is metabolically indispensable, socially organized, economically consequential, and sometimes central to questions of safety, care, harm, and responsibility, yet sleep itself is remarkably difficult for law to apprehend directly. Ordinary, non-dreaming sleep is organized predominantly below the levels of waking articulation and institutional engagement that law requires, while law itself exists only through symbolic stabilization at those higher levels. Sleep and law therefore occupy different recursive domains, and their encounter necessarily takes the form of conversion rather than direct address. Sleep becomes legally visible only after it has been converted into some other recognizable object: a medical disorder, a period of required rest, a safety risk, negligence, incapacity, economic disadvantage, evidentiary inference, harm, or a contribution to care. This article argues that this pattern is not accidental but an ontological consequence of how differently sleep and law are each organized. Legal systems can protect the conditions in which sleep may occur, regulate activities whose safe performance depends on adequate sleep, assign responsibility for sleep-related conduct, and compensate certain consequences of lost sleep. They cannot directly legislate a good night's sleep into existence, and no institutional refinement closes that gap, because closing it would require law to act on sleep merely as sleep rather than through a further juridically relevant predicate. Sleep thereby reveals a more general problem of metabolic burden: living beings incur metabolic costs that legal institutions cannot normally receive in their original embodied form, costs that must undergo attribution, evidentiary conversion, biocommensuration, and domain-specific translation before they become actionable. Sleep is a privileged case because the metabolic burden of lost or interrupted sleep is obvious, recurrent, and potentially severe, while remaining peculiarly resistant to direct legal representation.
I. A Vital Need That Law Barely Names
The German Grundgesetz, the constitutional document under which one of the world's most elaborately codified legal systems operates, contains no direct mention of sleep or any closely related term. This is a striking absence, not because constitutions ought necessarily to enumerate biological necessities one by one, but because sleep occupies such an enormous portion of human life and is non-negotiable at the level of metabolism in a way remarkably few other activities are. A constitution can be silent about hobbies, diet, or leisure without anyone finding the silence odd. Silence about an activity that consumes roughly a third of every life, and without which no other right could be exercised for very long, is a different kind of silence.
Law elsewhere is anything but silent. It contains elaborate doctrines concerning work, home, injury, care, privacy, driving, capacity, health, employment, culpability, and damages, an apparatus built up over centuries and covering almost every corner of a life lived awake. Yet sleep itself, the activity underneath all of these domains, largely disappears from the doctrine built to govern them.
The initial temptation, faced with this absence, is to go looking for an implicit right to sleep buried somewhere in existing doctrine, a right the law has simply failed to name. That temptation is worth resisting as the organizing question of this article, not because such a right would be uninteresting, but because chasing it risks skipping past a more basic and more answerable question. Before asking whether law protects sleep as a right, it is worth asking what has to happen to sleep before law can recognize it as anything at all.
The master argument of what follows can be stated as a single, repeated question, asked afresh of each domain law touches sleep through: what does sleep have to become here before law can act on it? For medicine, the answer is a disorder. For the law of driving, it is foreseeability and fault. For working-time regulation, it is protected non-working time. For family law, it is economic disadvantage. For the law of employment and harm, it is impairment, injury, duty, or liability. For the law of evidence, it is a legally receivable trace. Sleep itself is never the answer, in any of these domains, and that is the master argument this article develops: sleep becomes visible before law primarily through conversion.
Each of these conversions will be examined in turn. But the deeper puzzle they all share is worth stating clearly at the outset, because it is the question the rest of this article is trying to answer. Why must an activity so central to life, so obviously consequential, and so universally shared, become something else before law can act upon it at all?
The stakes of answering this well are not merely academic. Whatever a legal system fails to name, it also fails to protect directly, and whatever a legal system can only address through conversion inherits every distortion the conversion introduces along the way. A parent's exhaustion, a shift worker's fatigue, a patient's disturbed nights: each of these has to travel a considerable distance, through medicine, through evidentiary procedure, through economic calculation, before it can register as anything a court, a regulator, or a legislature is equipped to act on. Understanding that distance, and what happens to sleep while it is being travelled, is the task of this article.
II. Recursive Incommensurability: Why Sleep and Law Cannot Meet Directly
Living Value Theory supplies a direct explanation for why this pattern recurs so consistently across such different areas of law, and the explanation is stronger than a mere mismatch of capacities. Ordinary, non-dreaming sleep is organized predominantly below the levels of waking L3 and L4 recursive engagement that externally checkable symbolic activity requires. The sleeping body continues an entire multimediated mesocosmic coordination of its own: embodiment continues, thermoregulation, postural adjustment, immune activity, memory consolidation; being-with continues in attenuated form, a partner's or a child's presence still mattering without sustained interaction; dwelling continues, the sleeper still subject to night, season, and temperature whether or not any of it is noticed; and multimaterial arrangements, the mattress, the walls, the locked door, continue to scaffold the sleeper throughout.
Even multisymbolic materials can remain physically present around a sleeping body without this making any difference to the point. A passport can lie in a drawer beside the bed. A court order can determine, quite independently of anyone's awareness of it that night, who is legally permitted to occupy the house the sleeper is sleeping in. An alarm clock can display symbolic time on its face throughout. None of this amounts to active symbolization. What has dropped out during ordinary sleep is not the presence of symbols in the sleeper's surroundings but the sleeper's own recursive engagement with them as symbols, the checking, questioning, and revising that distinguishes a symbol actually in use from a symbol merely lying nearby.
Law is organized in almost the opposite direction. A statute, a claim, testimony, a judgment, a legal category, an evidentiary inference, or a monetary award exists as law only through symbolic stabilization at L3 and above, articulated, checked, revised, and eventually fixed into a portable, general form other cases can draw on. Law cannot descend into L1 and simply receive the metabolic process operating there. Sleep and law, on this account, occupy different recursive domains. Sleep is a multimediated mesocosmic coordination organized predominantly at L1 and L2. Law is a multisymbolic coordination that exists only at L3 and above. Law and sleep cannot meet at the same recursive level, and their encounter therefore necessarily takes the form of conversion.
Call this recursive incommensurability, provided the term is handled carefully. It names a direct incommensurability between two modes of organization, not an impossibility of translation between them. The whole of this article is about the translations that bridge exactly this incommensurability, and none of what follows should be read as claiming those translations cannot succeed. They succeed constantly. What they cannot do is dissolve the incommensurability itself, since a successful conversion is still a conversion, not a disappearance of the gap it crosses.
This needs one careful qualification before it is pushed any further, because an unqualified version of the claim invites an objection this article does not need to weather. Law can, obviously, establish as a factual matter that someone was asleep at a given time, and can recognize the word sleep linguistically without any difficulty at all. What law cannot ordinarily do is act on sleep merely as sleep. A finding that a defendant was asleep is not, by itself, actionable; it becomes actionable only once a further, juridically relevant predicate attaches to it, risk, harm, disorder, incapacity, disadvantage, duty, or entitlement. The incommensurability this article traces therefore sits not at the level of naming or factual recognition, which law manages easily, but at the level of receiving sleep's own living, metabolic organization as the object a legal decision directly addresses.
This yields a stronger claim than the one this article might otherwise reach for, worth stating in its sharpened form. It is not merely that very little of sleep survives the journey from L1 to L4 in its original form, but that nothing could survive that journey unchanged, because becoming legally legible is itself a transformation of recursive level and mediation, not a relabelling of something that stays the same underneath. A sleep disorder is not sleep. Eight hours of protected rest is not sleep. Fatigue treated as a safety risk is not sleep. Foreseeability is not sleep. Economic disadvantage is certainly not sleep. Two hours of lost working capacity is not sleep. Even the ordinary, entirely truthful statement I slept badly last night is already an L3 retrospective articulation of an L1 and L2 process that was not available in that articulated form while it was actually occurring. Each of these is a conversion product, and none of them is the thing converted.
This has a direct consequence for evidence that the rest of this article develops at length, and it is worth stating in its strongest form here. There can be no legal evidence of sleep that remains inside sleep's own recursive mode, because the production of evidence already constitutes an exit from that mode. A doctor's record, a rota, a payslip, a witness statement, a sleep diary, a wearable's score, an employment history, and courtroom testimony, none of these reproduce the sleep event itself. Each constructs an L3 or L4 trace standing for some selectively relevant feature of it, chosen according to whichever legal question the trace is being produced to answer.
This also explains something that might otherwise look like an inconsistency across branches of law rather than what it actually is. Different branches of law can, and routinely do, produce radically different versions of the same night's lost sleep without any one of them necessarily being wrong. Medicine asks whether there was a disorder. Employment regulation asks whether adequate rest time was provided. Negligence asks whether the person should have anticipated impairment. Family law asks whether the loss produced an economic disadvantage. The law of evidence asks what surviving traces establish whichever of these conversions is being claimed. None of these questions is, or can be, what was this sleep in the fullness of its lived metabolic coordination. That question sits outside law's ontology entirely, not because law has failed to ask it well enough, but because no branch of a symbolic institution operating at L3 and above could ask it in the first place.
This finally supplies the fuller explanation for the observation this article opened with. The remarkable thing about the Grundgesetz's silence on sleep is not merely that a constitution happens not to mention a biological necessity. A constitution is a paradigmatically L4 multisymbolic stabilization, and it can protect the conditions under which L1 and L2 sleep becomes possible by symbolizing home, privacy, bodily integrity, and working time, each of which it can reach because each is already, in its constitutional form, a symbolic object rather than a living process. What no constitution can do, however finely drafted, is incorporate the sleeping process itself into its own symbolic operation, because that operation exists only at the level sleep has, for the duration, withdrawn from. The absence is not an oversight so much as what recursive incommensurability looks like when the institution in question is a country's most fundamental law.
III. The First Great Translation: Sleep Into Medicine
What does sleep have to become here before law can act on it? In medicine, the answer is a disorder, and a 2018 review of the legal aspects of sleep medicine makes an instructive case study precisely because its ostensible topic is sleep and law in general, yet almost every legal intersection it discusses is already mediated through sleep medicine before the discussion begins. Obstructive sleep apnoea, excessive daytime sleepiness, parasomnia, polysomnography, sleep laboratories, and clinical reporting obligations frame the entire review, which presents legal interaction with sleep as something that arises through the manifestations of sleep disorders and the responsibilities of the clinicians who diagnose and treat them.
Medicalization, considered as a conversion process, follows a recognizable sequence. A disturbance of sleeping becomes a symptom, the symptom becomes a diagnosis, the diagnosis becomes a documented impairment, and the impairment becomes legally relevant to something else entirely, fitness to drive, employer liability, medical screening obligations, or a duty to report. Obstructive sleep apnoea becomes relevant to litigation not because a court has any independent interest in respiratory physiology, but because a diagnosis supplies exactly the kind of stabilized, portable, L4 category a legal proceeding can act on. Parasomnia becomes relevant to criminal responsibility for a comparable reason: medical classification of a sleep disorder can be brought to bear on questions of automatism or mitigation in a way an undiagnosed, merely asserted account of sleepwalking cannot.
This is an important pathway, and nothing in this article disputes its usefulness. But it would be a serious mistake to let sleep and law collapse into sleep medicine and law, and the reason is worth stating directly. Sleep does not have to become pathological to become legally relevant. A parent repeatedly awakened through the night by a child's illness need not have insomnia in any diagnosable sense. A junior doctor may have entirely normal sleep physiology and still have insufficient opportunity to sleep, for reasons having nothing to do with their own body. A driver may simply have stayed awake too long, without anything resembling a sleep disorder ever entering the picture. A worker may be exhausted without qualifying for any recognized diagnosis at all.
Medical recognition, in other words, is one conversion available to sleep, not the master conversion every other legal encounter with sleep must eventually pass through. Several of the most consequential legal questions this article goes on to examine, working time, economic disadvantage, evidentiary reconstruction of a shared history, never require a diagnosis at all. They require only that ordinary, undiagnosed, universally shared sleep be shown to matter, which is a considerably harder thing for law to recognize than a documented disorder, precisely because it arrives without the ready-made stabilization a diagnosis supplies.
The review's own data illustrate how much weight this single conversion is asked to carry once it is available. Litigation connected to obstructive sleep apnoea is more often generated by perioperative complications, adverse outcomes following surgery on patients with known or suspected apnoea, than by any failure to diagnose the condition in the first place, with reviewed cases showing death or serious injury in a substantial majority of outcomes and average awards running into the millions of dollars. This pattern is itself informative. Once a diagnosis exists, or ought to have been made, the legal system has a stabilized category to reason with, negligence in monitoring, inadequate informed consent, mismanaged anaesthesia, and litigation clusters around exactly the points where that category was, or should have been, in play. Where no diagnosis exists, one particularly powerful route to legal stabilization is unavailable, whatever the underlying severity of what actually occurred.
Reporting obligations complete the same pattern from the regulatory rather than the adjudicative side. Several jurisdictions require physicians to notify licensing authorities when a patient with excessive sleepiness continues to drive despite treatment recommendations, and some states impose licensing restrictions tied specifically to a sleep disorder diagnosis. Here the medical category is not merely a convenience for litigation after the fact. It is the trigger written directly into the regulatory apparatus, so that a disturbance of sleep becomes actionable by a public authority only once it has already been converted into a name a medical board or a motor vehicle department is equipped to act on.
IV. Culpability: Translating Sleepiness Into Foreseeability and Fault
What does sleep have to become here before law can act on it? In the law of driving, the answer is foreseeability and fault, and a second pathway running through the courts rather than through medicine reveals that legal relevance need not depend on a diagnosis at all. Early twentieth-century case law on sleepy driving already shows a legal system converting sleepiness into fault without ever asking whether the driver suffered from any recognized disorder. In a 1925 Connecticut case, a driver who dozed at the wheel and crashed argued that sleep occurs without warning and that he should therefore be excused from a duty he could not consciously have discharged while unconscious. The court disagreed, holding that sleep, unlike a sudden blackout, is typically preceded by recognizable warning signs, fatigue and a dulling of the senses among them, and that a driver who continued driving through those warning signs had failed a duty that attached before sleep began rather than during it. The legal transformation at work is precise and worth naming exactly: sleepiness became foreseeability, foreseeability became a failure to act on that foreseeability, and the failure to act became negligence, all without any need to establish a sleep disorder as such.
Statutory reform has since made a version of this same conversion considerably more explicit. Some jurisdictions now treat a defined period without sleep as itself sufficient evidence from which recklessness may be inferred, a fixed number of consecutive hours awake operating as a bright, evidentiary line a court can apply without having to reconstruct a driver's subjective state of mind at all. The conversion here runs directly from a measured duration to a legal inference, skipping medical diagnosis entirely and skipping, too, most of the uncertainty that ordinarily attends questions of what a defendant did or did not appreciate about their own condition.
What makes this pathway conceptually distinctive is where, in time, the legally relevant event is located. Once a driver is actually asleep, ordinary conscious responsibility is sharply attenuated, for exactly the reasons the recursive incommensurability already described: a sleeping person cannot deliberate, cannot weigh a decision, cannot exercise the kind of agency the law ordinarily requires before it will assign fault. Law therefore moves backward in time rather than trying, impossibly, to locate responsibility inside the sleep episode itself. The operative question becomes not what happened while the driver was asleep, but when, before that, the driver should have recognized their own approaching metabolic state and stopped. Sleep itself becomes legally relevant only through a moment of prior recursive discernment that a court can then evaluate on ordinary, waking terms.
This licenses a distinction worth holding onto with some precision. There is a considerable gap between the modest claim I fell asleep and the much stronger claim I was negligent in allowing myself to reach the point at which I fell asleep, and the entire apparatus of foreseeability doctrine exists to convert an uncertain, embodied, gradually developing causal chain into something a court can stabilize sufficiently to assign responsibility. The doctrine does not deny that sleep arrived involuntarily, but insists only that the conditions making sleep likely were, or should have been, recognized in time.
Sleep therefore reveals something about the temporal structure of culpability that other cases of harm do not display nearly as cleanly. Responsibility attaches not to what happened during an incapacitated state but to earlier, fully conscious conduct, even though the metabolic decline connecting the two is continuous and gradual rather than sudden. This gives sleep an unusual theoretical importance for any account of fault, because it forces the law to draw a sharp evaluative line, this was the moment blame attached, across a process that has no comparably sharp physiological boundary of its own.
This pathway also supplies a clear illustration of recursive incommensurability translated into the specific vocabulary of tort and criminal law. The court is not, and does not claim to be, examining the driver's L1 metabolic state directly. It is examining the driver's own prior L2 and L3 engagement with that state, whether the felt warning signs of fatigue crossed the threshold into something the driver noticed, and whether noticing carried an obligation to act. Foreseeability doctrine is, in this sense, a legal technology for locating the moment at which an L2 disturbance should have been escalated into an L3 decision, before the L1 process it warned about took over involuntarily. Sleep supplies an unusually clean test of this technology precisely because the disturbance it is built to catch, drowsiness, has a well-documented, recognizable phenomenology of its own, heavy eyelids, wandering attention, a drifting sense of time, which is exactly why courts have been willing to treat it as the kind of warning a reasonable person is expected to heed.
V. Working Time: Law Protects the Opportunity to Recover, Not Sleep Itself
What does sleep have to become here before law can act on it? In working-time regulation, the answer is protected non-working time, not sleep itself, and a third pathway avoids both medicine and fault entirely, working instead through the regulation of time. Maximum working-hours regulation, of the kind embodied in the European Working Time Directive and its application to junior doctors, offers a particularly clear illustration, because sleep is rarely if ever the explicit subject of the regulatory language, even though sleep and metabolic recovery are among the indispensable embodied conditions such regulation indirectly protects, alongside health, accident risk, overwork, and family life more generally. A limit on weekly working hours, framed in terms of safe staffing, patient safety, and sustainable rostering, has major consequences for how services are organized and staffed, and the language in which those consequences get discussed among the managers implementing such a limit tends to be organizational and operational rather than physiological. The translation at work here can be stated plainly: a metabolic need for recovery becomes non-working time, non-working time becomes a maximum permitted number of hours, and the maximum permitted number of hours becomes an enforceable regulatory limit that inspectors, rotas, and compliance officers can check against a roster without ever having to ask whether anyone actually slept.
This gap between the regulation's aim and its actual object is worth dwelling on, because it exposes something the earlier sections have already suggested in more general terms. Law cannot require that protected rest time becomes actual sleep. A junior doctor released from duty at the end of a shift may spend the following hours sleeping, caring for a child, commuting home, socializing, or lying awake unable to switch off after a demanding day, and the regulation has no purchase on which of these actually occurs. What it can and does control is the temporal opportunity, the block of time during which institutional demands are suspended for long enough that sleep becomes possible, without being able to reach any further into what a person actually does with that suspended time. This is intervention distance, already identified in other contexts, appearing here in explicitly legal form: the regulation intervenes at the level of institutional demand rather than at the level of the living outcome the demand is ultimately meant to protect.
Working-time law is, in this sense, best understood as an indirect sleep technology rather than a sleep guarantee. It creates a zone in which institutional demands are suspended sufficiently for sleep and recovery to become possible, without legislating sleep itself into existence. The law does not, and structurally cannot, legislate sleep. It legislates withdrawal from work, and trusts, without being able to verify, that withdrawal from work will translate into the metabolic recovery it exists to protect.
A 2009 report on NHS service reorganization, drawing on interviews with the managers responsible for implementing exactly this kind of working-time limit, shows the same regulatory apparatus pulling in two directions inside a single institution. Managers describe the pressure of operating services running, in their own words, at as much as one hundred and fifteen per cent of notional capacity, and at least one manager describes performance targets as something that would, in their own phrase, keep me awake at night. Read alongside the working-time regulation itself, this produces a striking institutional contrast. One regulatory instrument protects a defined block of time away from work because metabolic recovery matters enough to be written into binding rules. A separate and simultaneous set of institutional pressures, targets, capacity constraints, performance management, generates exactly the kind of recursive demand capable of invading sleep through anxiety and a sense of permanent performance pressure, in the very people responsible for administering the protective rule. Law and institutional practice can therefore protect metabolic withdrawal in one domain while actively undermining it in another, within the same organization, at the same time, without any contradiction visible at the level of formal compliance.
This double movement deserves to be named as a general pattern rather than treated as an isolated irony of one health system's management culture. Any institution large enough to have both a formal duty to protect rest and an informal culture of continuous performance pressure is capable of satisfying the first while quietly defeating it through the second, and a regulatory audit checking only formal compliance with working-time limits will have no way of detecting this, since the informal pressure leaves no trace in the roster at all. The metabolic burden such pressure generates does not disappear simply because the hours on paper are compliant. It is displaced instead into exactly the register the following sections go on to examine.
VI. From Sleep Loss to Metabolic Burden
The conversions examined so far, medical, culpatory, regulatory, share a feature worth making explicit before the article turns to its most demanding case. None of them treats sleep loss as the master category needing explanation. Each treats sleep loss as one instance of something broader, and it is worth naming that broader category directly. Call it metabolic burden: the embodied cost of sustaining living processes under mesocosmic conditions that intensify demand, restrict recovery, redistribute care, or require continued recursive availability from a body that would otherwise be permitted to rest.
This needs one boundary drawn clearly, since not every metabolic cost qualifies. All living requires metabolic expenditure, the ordinary energetic cost of continuing to be a living, coordinating organism at all. Expenditure becomes burden specifically when mesocosmic conditions intensify demand beyond what ordinary recovery can absorb, constrain the restoration that would otherwise offset it, or distribute the demands of living unevenly across the people sharing a household or a relationship. This boundary matters because treating all expenditure as burden would make the category too broad to do any diagnostic work.
Sleep is one major source of metabolic burden, but far from the only one. Pregnancy, feeding an infant, illness, physical caregiving, chronic pain, sustained vigilance, night work, and long-term caregiving for a dependent relative all impose other embodied costs, and several of them interact directly with sleep rather than standing apart from it. Framing the underlying category as metabolic burden rather than sleep deprivation is not merely a matter of vocabulary. Sleep deprivation tends to suggest either a clinical pathology or an extreme, sharply measurable deficit, the kind of thing a sleep laboratory could quantify in hours. Metabolic burden is considerably broader, and closer to how most of the relevant harm actually accumulates in ordinary life. A person may lose a little sleep every night for years, sleep lightly while listening for a child, wake repeatedly without ever fully rousing, recover poorly from one disturbed night before the next arrives, and perform demanding daytime care throughout, without any of this ever producing a recognized sleep disorder a physician could name. Yet it may accumulate into a substantial embodied cost all the same, precisely the kind of cost the medical conversion examined earlier is poorly equipped to capture, since nothing about it crosses the threshold into diagnosable pathology.
Metabolic burden, once identified, must itself undergo further legal translation before it can do any work in a legal proceeding, since law cannot ordinarily compensate metabolic burden as such. It requires a recognizable downstream object to attach to. Metabolic burden becomes an impairment where a diagnosis is available. It becomes a safety risk where conduct affecting others is at stake. It becomes an economic loss where a career or an earning capacity has been measurably affected. It becomes a care contribution where a household's internal division of labour is being reconstructed. It becomes a legally cognizable injury only where one of these more specific categories has successfully received it.
This yields the article's central analytic sequence, worth stating once and setting apart, because every section from here on is an elaboration of one part of it.
mesocosmic demand → metabolic burden → recursive discernment → causal attribution → biocommensuration, where required → institutional translation → evidentiary conversion → legal legibility
Mesocosmic demand generates metabolic burden. Metabolic burden becomes available to recursive discernment, the sleeper's or carer's own sense that something has been costly. Recursive discernment must then survive the much harder passage into causal attribution, an account of what specifically caused the felt cost. Where competing burdens or competing accounts must be weighed against one another, causal attribution passes through biocommensuration, a stage not every case requires. Causal attribution, biocommensurated or not, must in turn undergo institutional translation into whichever category a given branch of law is built to recognize, and that translation must survive its own evidentiary conversion before it achieves legal legibility, in a form that usually bears only a partial resemblance to the lived cost that started the whole sequence. Different branches of law enter this sequence at different points, medicine intervening early, at the level of diagnosis, family law intervening late, at the level of economic consequence, and this difference in entry point is itself one of the more consequential facts about how law and sleep actually relate.
A single example makes the sequence concrete before the article turns to its most demanding case. Consider a parent whose child suffers from a chronic condition requiring attention several times a night, over a period of years. The mesocosmic demand is the child's condition itself, not chosen by anyone and not itself a legal fact. The metabolic burden is the cumulative cost of interrupted, fragmented sleep borne by whichever adult responds most often. Recursive discernment is that adult's own, entirely reliable sense that the years have been exhausting. Causal attribution is the much harder claim that this specific exhaustion, rather than some other feature of a demanding life, produced some specific later consequence, reduced earning capacity, a foreclosed promotion, a health problem of its own. Institutional translation is whatever legal category eventually receives that claim, should the parents' relationship end and the question of who bore the cost become a matter for adjudication. Every one of these stages can fail independently of the others, and a claim can founder at any point along the sequence even when every earlier stage in it was entirely real. This is exactly the situation the most demanding case this article considers is built to examine in detail.
VII. Section 28 as a Privileged Extreme Case: Biocommensurating Metabolic Burdens
Scots family law's statutory remedy for cohabitants on separation supplies the privileged extreme case this article has been building towards, because it does not merely translate sleep loss into a legal category the way the earlier sections have described. It may have to reconstruct who bore a metabolic burden, compare burdens borne by different people, compare burdens located at different times, identify their downstream consequences, and translate those consequences into a single figure of economic disadvantage, all within one proceeding. An existing analysis of that remedy already describes the mediational conversion requirement it imposes: embodied and relational costs become legally legible only insofar as they can be converted into the statute's own multimaterial vocabulary of economic advantage and disadvantage. That analysis does not yet draw out fully what happens when the specific embodied cost in question is sleep, and it is worth developing that extension in some detail, because sleep exposes distinctions the analysis's own examples do not force into view.
Begin with a distinction that is easy to elide but does a great deal of work once separated out properly. Suppose it is entirely uncontested between two former partners that a child's chronic eczema created persistent nocturnal disturbance for the family, night after night, for several years. That single, agreed fact establishes only the source of a metabolic demand. It does not establish how that demand was distributed between the two parents, which is a separate question requiring separate evidence. What caused the family's metabolic demand and how was that demand distributed between the two people living with it are different questions, and a court, or the parties themselves, can be entirely right about the first while remaining deeply contested about the second. This gives a principle worth stating in its own right: causal attribution of metabolic demand must be distinguished from attribution of metabolic burden, since agreement about where a demand came from settles nothing about who actually carried it.
The reason this distinction matters so much for sleep specifically is that the burden in question is very rarely lived as two neatly separable ledgers, one for each parent. Care performed at night is, in its lived form, interrecursive rather than individually assignable in any simple way. One parent wakes at a sound. The other may also wake, whether or not either intends to. One gets out of bed to attend to the child. The other listens from the next room, ready to help if needed, contributing something real even while remaining physically still. One parent may take on more daytime care as compensation for a harder night. Roles shift by day, by job schedule, by illness, by whichever parent happens to have an early start the following morning. The existing analysis already explains that litigation over such a history must individuate what was actually an interrecursive history, in which neither partner's contribution is separable from the other's, into separately attributable quantities a court can assign to one party or the other. Sleep makes this individuation unusually difficult precisely because so much of what happened at night was, by its nature, coordinated rather than performed by one person acting alone.
A further complication belongs specifically to sleep and deserves to be treated as a new insight in its own right rather than a restatement of the individuation problem just described. Night-time burden is not self-evidently heavier than daytime burden, and nothing about the clock settles the comparison in advance. If one parent bears the greater share of night-time waking while the other manages daytime feeding, medication, comforting, medical appointments, and supervision, there is no reason built into the situation itself why the night-time contribution should automatically receive greater weight in an eventual accounting. It may in fact be more consequential, because interrupted sleep fragments metabolic recovery in a way daytime tasks, however demanding, do not. But that is an empirical and causal claim requiring its own evidence, not something the mere fact of occurring at three in the morning rather than three in the afternoon can establish by itself.
Call this further operation temporal weighting, and distinguish it carefully from temporal location. Temporal location simply asks when a burden occurred. Temporal weighting asks how much that temporal position should count once burdens are being compared against one another, and it is not itself an intrinsic property of the burden the way magnitude or duration are. It is a comparative judgement performed upon those properties, and separating the two makes clear that no clock reading, however precise, can substitute for it.
This licenses a compact framework worth setting out explicitly, since it clarifies what a court, or the parties themselves, are actually being asked to compare whenever a history of this kind is reconstructed. Metabolic burden has a magnitude, how intense the demand is at the moment it occurs. It has a duration, how long and how repeatedly it recurs across a relationship's whole history. It has a distribution, who actually bears it, which as already shown cannot simply be read off from who bears the underlying demand. It has a temporal location, when it occurs relative to sleep, work, and other competing demands and opportunities for recovery. And it is subject to temporal weighting, the comparative judgement, performed on top of the first four properties, about how much a given temporal location should count once differently timed burdens are set against each other. Clock time alone settles none of these, and settles the last least of all. Care given at three in the morning may be metabolically costly precisely because it fragments recovery in a way clock time does not capture on its own. Care given at two in the afternoon may be more economically consequential if it repeatedly displaces paid work, a connection that has nothing to do with metabolic cost as such and everything to do with which hours of the day happen to overlap with an employer's expectations.
This is where the concept this article needs earns its name. Biocommensuration is the process through which heterogeneous embodied burdens are rendered sufficiently comparable for an institution to attribute their distribution and consequences across persons. A Section 28 court is not required to determine, in any comprehensive sense, who suffered the greater metabolic burden overall. Its task is narrower: to determine facts only insofar as they bear on economic advantage or disadvantage. But where competing accounts of caregiving and embodied expenditure bear on that narrower question, as they very often do, the court is forced into partial acts of biocommensuration whether or not it names them as such, comparing different bodies, different times of day, and different opportunities for recovery, against one another, well enough to support the narrower economic finding the statute actually requires. This is why raw arithmetic, however carefully the underlying hours are counted, cannot settle the matter on its own. The arithmetic presupposes a prior judgement about which hours, and which kinds of burden, should count for how much, and that judgement is precisely what biocommensuration is doing, whether or not the court names it.
A final distinction closes this section and prevents its central claim from collapsing into something simpler and less accurate than intended. Metabolic biocommensuration, the question of who bore how much living burden, is not the same operation as economic commensuration, the question of what economically recognizable consequences followed from that burden. The two need not coincide, and assuming that they do is one of the most consequential errors available to anyone reasoning about a case of this kind. A significantly unequal metabolic burden could, in a particular case, produce comparatively little economic asymmetry between the parties, where both retained broadly similar earning capacity regardless of who slept worse. An equal metabolic burden, conversely, could produce sharply unequal economic consequences, where one partner's career happened to be more sensitive to interrupted sleep than the other's, for reasons having nothing to do with how the burden itself was shared. Keeping these two commensurations distinct prevents sleep loss from being treated as automatically synonymous with economic disadvantage, a conflation the statute's own multimaterial vocabulary makes tempting but that the underlying reality, properly examined, does not support.
VIII. Evidentiary Conversion: The Burden Must Leave the Right Kind of Trace
Once sleep loss, or the metabolic burden it forms part of, has been translated into a legal claim of economic disadvantage, the evidence supporting that claim must undergo a parallel translation of its own, and this section develops a point implicit in everything preceding it into an explicit general proposition. It is not enough, in litigation of the kind just described, to establish that a person slept less. The relevant chain of proof typically has to establish something considerably longer: that less sleep produced a metabolic burden, that the burden reduced some specific capacity or foreclosed some specific opportunity, that the reduced capacity or foreclosed opportunity had an identifiable effect on a career or an income, and that the effect can be expressed as an economic disadvantage in the statute's own terms.
Every substantive conversion, in other words, requires a corresponding evidentiary conversion, and the two do not automatically travel together. This is larger than sleep, and worth stating as a general principle of legal conversion as such: legal conversion changes not only what must be proved, but what can count as proof. Where a phenomenon is translated into economic disadvantage, the evidence must demonstrate an economic trajectory. Where it is translated into negligence, the evidence must establish foreseeability and breach. Where it is translated into a disorder, the evidence must establish diagnostic criteria. Sleep merely makes an already general feature of legal translation unusually easy to see. Embodied costs paid in health or in time, rather than in money, can remain legally elusive even where they are entirely real, if they cannot be connected to a further, economically expressed effect, and this is not a flaw specific to any one statute. It follows directly from what a legal category built around economic advantage and disadvantage can, by its own terms, receive.
This needs to be handled with some care, because it invites a mistaken inference that is worth heading off explicitly. The absence of a visible career disadvantage does not prove that sleep loss, or the caregiving burden generating it, never occurred. What it does mean is that an asserted conversion from metabolic burden to economic disadvantage requires further explanation wherever the observable career trajectory does not exhibit the effect the claim depends on. Proof of burden and proof of legally relevant consequence are different achievements, and a court's inability to find the second is not evidence against the first, even though it will, as a practical matter of litigation, function as though it were.
The deepest evidentiary problem in this whole area, however, is one the underlying material already identifies in a different context and that sleep exemplifies with unusual force. Long, cumulative, ordinary forms of contribution, precisely the kind sleep-related caregiving typically takes, are exactly the forms least likely to leave any documentary trace behind, for the simple reason that nobody keeps a contemporaneous record of an intimate relationship's ordinary nights while those nights are still ordinary. Nobody ordinarily records exactly when a child woke, who got up, how long any treatment took, whether the other parent also woke, and when either of them returned to sleep. The burden in question is real precisely because it is woven into the texture of ordinary intimate life rather than conducted as a documented transaction, and retrospective litigation, years later, asks an L4 institution to reconstruct a history that was never lived in documentary form and cannot be made to yield one after the fact.
Where secondary evidence exists at all, it tends to be indirect and unevenly available: medical records showing a health consequence, employment records showing a change in hours or role, correspondence in which a difficulty was mentioned in passing for some other purpose entirely, the testimony of friends or relatives who happened to notice. None of these was created to document the burden in question, which means their evidential value depends heavily on chance rather than on the actual severity of what they are being asked to prove. A parent who happened to mention their exhaustion to a doctor treating an unrelated complaint may find that offhand remark doing more evidentiary work, years later, than years of unremarked, undocumented nights ever could. This is not a flaw introduced by any particular judge or any particular statute so much as a predictable consequence of asking evidentiary procedure, built to receive documented transactions, to receive a history that was never transacted at all.
IX. Harm, Employment and the Other Routes Through Which Sleep Becomes Visible
A brief comparison closes the survey of routes, and its job is explicitly comparative rather than cumulative: to show that the conversion model predicts legal heterogeneity rather than being embarrassed by it. Different branches of law do not disagree about what sleep really is. They ask different institutional questions and therefore construct different legal objects from the same underlying metabolic process.
Disturbed sleep can become legible as evidence of harm in a register distinct from anything considered above. Workplace bullying and abuse, difficult to prove directly, can become legally visible partly where the underlying relational disturbance leaves an embodied trace a court or investigating body can point to, sleep disturbance being one evidentiary trace among others available for this purpose, nonspecific rather than distinctively credible, since a great many things beyond workplace harm can disturb sleep. The conversion here runs from an interrecursive harm through disturbed sleep as an embodied consequence and out the other side as corroborating evidence, sleep functioning not as the object of the claim but as one trace supporting a claim about something else.
Employment law supplies a further route through vicarious liability, familiar by now in outline. Employers can be held responsible for employees who fall asleep on duty and cause harm, particularly in safety-sensitive occupations, where a documented history of excessive sleepiness can shift liability onto an employer who failed to screen for it. A well-known trucking case and two national rail crashes attributed to undiagnosed sleep apnoea among operators both show the same structure, the legal object under examination being fitness, duty, and liability rather than sleep as such.
A final route runs through incapacity, and deserves mention because it exposes a real gap rather than a working conversion. Consciousness, on the best available medical evidence, is not an all-or-nothing state but a spectrum, and the law's usual insistence on a sharp line between culpable and non-culpable states sits uneasily with a phenomenon that resists exactly that kind of binary treatment, most visibly where a parasomnia is raised as a defence to violent conduct committed while the defendant was, by every available account, asleep. This is not a conversion working imperfectly so much as a conversion not yet achieved, and the resulting uncertainty falls on defendants and claimants rather than on the institutions asked to resolve it.
X. What Law Can Never Do: Guarantee Sleep
It is worth bringing the article back, at this point, to the more general theory of sleep this whole discussion has been applying rather than restating from scratch. The legal conversions examined above are useful, and several of them are more than that: without the medical conversion, sleep apnoea would go undiagnosed and untreated in exactly the perioperative settings where it kills people; without working-time regulation, junior doctors would have no protected time at all; without the economic conversion, a lifetime of unpaid nocturnal caregiving would have no legal existence whatsoever on separation. None of this article's argument should be read as diminishing any of these achievements.
But none of these conversions abolishes the basic limit that has structured the whole discussion. Law may protect rest time, limit working hours, assign liability, recognize economic disadvantage, compensate injury, regulate professional fitness, and recognize medical disorder. It cannot say you shall sleep well tonight and make the statement true by saying it, in the way a court's judgment can make a debt legally owed simply by declaring it so. This is because sleep remains a recursively mediated living process rather than a mechanically controllable output responsive to institutional command in the way a payment, a licence, or a working roster is.
A working-time rule can create time. A building regulation can create shelter. An employment duty can remove a demand. A damages award can compensate a loss already suffered. None of these interventions directly produces sleep, and none of them could be redesigned, however cleverly, to do so, because sleep is not the kind of thing an external command can bring about in the person commanded. Legal intervention with respect to sleep is therefore necessarily indirect and mesocosmic, working on the conditions surrounding sleep rather than on sleep itself, and this is worth stating as strongly as the evidence supports: law can protect the conditions of sleep, regulate the consequences of sleep loss, and translate metabolic burdens into actionable categories once they have occurred. It cannot command the living process itself into occurring.
This makes sleep an unusually clear demonstration of a distinction worth generalizing well beyond this one topic. There is a difference between legal control over conditions and legal control over living outcomes, and a great deal of confused thinking about what law can achieve in areas touching health, wellbeing, and embodied life more generally comes from collapsing that distinction, treating the successful regulation of a condition as though it were equivalent to guaranteeing the outcome the condition was meant to make possible. Sleep will not let that collapse go unnoticed, because the gap between the two is, in this one case, almost impossible to miss once it has been pointed out.
XI. Conclusion: Sleep Before the Law
Return, to close, to the constitutional absence with which this article began. The absence of sleep from foundational legal texts is no trivial curiosity, and by now it should be possible to see exactly what kind of fact it is. It exemplifies a general structural feature of the relation between sleep and law rather than a gap any particular drafter overlooked. Sleep rarely appears before law as sleep. It arrives as illness, as fatigue, as risk, as foreseeability, as working time, as care, as economic loss, as evidence, as harm.
This article has shown that the movement from sleep to any one of these legal objects involves several distinct operations, each worth naming precisely because each can fail independently of the others. Recursive conversion is required because sleep and law are organized in different, incommensurable recursive domains, and no institution can act on the first while remaining inside the second. Mediational conversion is required because embodied and relational processes must enter symbolic, and very often multimaterial, categories built for other purposes. Causal attribution is required because law must decide which consequences followed from which burden, a decision the underlying living history does not supply ready-made. Biocommensuration is required wherever burdens lived by different bodies, and at different times, must be compared against one another to reach a single figure or a single finding. And evidentiary conversion is required because the evidence offered in support of a claim must become legible in the very same terms as the claim itself, a requirement ordinary, undocumented life is poorly equipped to satisfy.
The Scottish cohabitation remedy examined at length above is the extreme but privileged case among all of these, because it may require every one of these operations to succeed at once, in a single proceeding, applied to one of the most intimate and least documented forms of coordination in ordinary family life.
One distinction developed at length in that discussion deserves restating here, because it resolves a confusion this article's whole argument could otherwise invite. Unequal metabolic burden need not produce unequal economic consequences, and equal metabolic burden can produce sharply unequal economic consequences, since the two are separate acts of commensuration answering separate questions. The legal question a remedy of this kind actually asks is never simply who suffered more. It is what legally relevant economic advantage or disadvantage resulted, and the first question may inform the second without ever being able to substitute for it.
Three of the concepts this article has needed to coin along the way are likely to travel considerably further than sleep itself. Metabolic burden names the embodied cost of sustaining living processes under uneven mesocosmic conditions, a category with obvious application to caregiving, illness, and chronic vigilance well beyond the bedroom. Biocommensuration names the partial, purpose-bound comparison of heterogeneous embodied expenditures across bodies, times, and recovery conditions, a comparison courts perform routinely without naming it. And evidentiary conversion names the more general fact that legal conversion changes not only what must be proved but what can count as proof at all. Sleep is simply where all three become unusually difficult to miss, because so much of what matters happened precisely while nobody was documenting, articulating, or institutionally classifying it.
The broader theoretical result reaches well past this one statute, and it is worth stating in its most general form before the article's final sentence. Sleep does not become legally visible simply by becoming noticeable, to the person living through it or to anyone around them. It becomes visible only when some feature of sleeping, of not sleeping, or of bearing the metabolic consequences of sleep loss, has been translated into a category the relevant legal institution is already equipped to recognize. Everything this article has traced, medicine, foreseeability, working time, economic disadvantage, evidence, harm, is a record of what that translation costs, and of what it leaves behind.
Law can recognize what sleep costs, what its loss causes, how its burdens are distributed, and what consequences follow. But each recognition requires a conversion, and no conversion is sleep itself. And none of them, however completed, can guarantee that sleep itself will occur.