Abstract

This dissertation advances a unified account of mental energy as a situated inference rather than a consumable stuff. The central claim is that the feeling of having or lacking energy is the organism’s estimate of the expected value of staying present and exerting control over the near horizon given predicted bodily costs, background reward conditions, social safety, learned threat, and time. The work binds a humane phenomenology of effort to disciplined biology and theory without treating persons as machines. It begins from the lived body as a medium of world relation in which vitality and fatigue are modes of openness or withdrawal rather than numbers on a gauge, and it retains William James’s insight that effort unlocks reserves only when the world confers permission to try and when belief makes investment plausible. These philosophical anchors set the explanandum that subsequent science must answer to rather than replace (Merleau Ponty 146 to 164; James 113 to 146).

The biological ground is predictive regulation. Allostasis describes how nervous, endocrine, immune, and behavioral systems coordinate in advance of anticipated demand, with chronic uncertainty driving costly high gain control and cumulative wear. This prospective budgeting explains why identical objective reserves can feel affordable in one context and unaffordable in another and why the language of spoons became a faithful vernacular for lived cost rather than for fuel depletion (McEwen and Stellar 2093 to 2101; McEwen 33 to 44; Sterling 5 to 15). Interoceptive models then specify how the brain predicts its own body and updates those predictions with ascending visceral evidence. In this loop, affect is not ornament but the organism’s best guess about its condition, and fatigue or vitality registers confidence or doubt that alignment between predicted and sensed can be sustained without further harm (Barrett and Simmons 419 to 429; Craig 655 to 666; Critchley, Wiens, Rotshtein, Ohman, and Dolan 189 to 195; Seth 565 to 573).

The decision to invest control requires value. The expected value of control model and average reward accounts of vigor formalize when organisms spend effort and when they conserve, recasting fatigue as a rational signal of unfavorable terms rather than as moral failure. This decision variable absorbs changes in predicted cost and expected payoff, which explains why safety, clarity of purpose, and reliable response from the world restore energy without any new fuel being added (Shenhav, Botvinick, and Cohen 217 to 240; Niv, Daw, Joel, and Dayan 507 to 520; Kurzban, Duckworth, Kable, and Myers 661 to 679). Social life is not a backdrop. Cues of exclusion and threat recruit pain systems and shift immune transcription toward inflammatory profiles that raise the bodily price of remaining available. Belonging and credible care reverse those taxes and make the next hour cheaper, which shows why energy is as much a civic and relational phenomenon as it is a private sensation (Eisenberger, Lieberman, and Williams 290 to 292; Slavich 265 to 295; Cole 31 to 37).

Learning governs whether priors about danger and cost can change. Modern exposure science emphasizes inhibitory learning rather than momentary habituation, with expectancy violations that limit retrieval of threat appraisals. Under constrained windows, reconsolidation based updating can further reshape traces, although boundary conditions and mixed human replications counsel restraint. Dissociation is acknowledged as a distinct mode in which interoceptive access is filtered rather than overwhelmingly intense, requiring different routes to revision. Time quietly controls price through sleep and circadian phase, so that identical labor may be cheap or expensive depending on when it is attempted, which explains ordinary reversals of capacity that otherwise appear mysterious (LeDoux 155 to 184; Brewin, Dalgleish, and Joseph 670 to 686; Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23; Monfils, Cowansage, Klann, and Ledoux 951 to 955; Schiller, Monfils, Raio, Johnson, Ledoux, and Phelps 49 to 53; Lanius, Brand, Vermetten, Frewen, and Spiegel 701 to 708; Borbély 195 to 204).

The manuscript holds a critical stance toward unbounded unification. The free energy principle is used only as a conceptual umbrella that links perception, action, and learning around expected surprise, while all falsifiable claims are placed within specific literatures on interoception, control allocation, social genomics, learning, and sleep that can lose to data. Popular autonomic narratives are engaged with anatomical caution, privileging the inflammatory reflex and reproducible cardiac vagal measures while avoiding contested phylogenetic stories. The ethical arc treats reports of low energy as intelligent testimony about oppressive prices rather than as evidence of inadequate will, and it frames dignity as the right to steward one’s predictive budget within communities that can either lower or raise avoidable costs. The dissertation is written to withstand adversarial critique by specifying the constructs it unifies, the boundaries of its scope, and the decisive findings that would force revision. The payoff is a theory that keeps persons and their histories in view while making precise claims about how a life becomes affordable again and why a world that answers back makes people feel like they have energy.

Introduction

The problem of mental energy is older than any laboratory. People have always needed a way to say that the next hour is too expensive to live and that sometimes an hour that looked impossible becomes suddenly possible. Everyday metaphors developed to mark those facts. The language of spoons captured the experience of counting capacity and of spending carefully through a day. The language of small injuries captured the way many are undone not by one catastrophic event but by a relentless arrival of minor demands that accumulate and sharpen into overload. These vernaculars succeeded because they were honest about survival. Yet they eventually falter, because they do not say what sort of thing energy is, how it can be known without reducing persons to instruments, why it rises in one place and collapses in another, or what would count as a mistake in a theory that claims to explain it. The present work answers to that lack by building a theory that begins from lived experience, keeps faith with dignity, and then binds that experience to disciplined accounts of regulation, inference, valuation, social life, learning, and time.

The beginning is philosophical on purpose. The lived body is not a chassis but the first horizon of meaning. Maurice Merleau Ponty’s account of embodiment makes this unavoidable. He argues that the body is the general medium for having a world. That claim reframes capacity as a style of world relation rather than as a level of fuel. A posture does not only reflect a state. It makes a world available or unavailable. Fatigue and vitality are not private numbers. They are modes of openness and withdrawal that arrive before explicit judgment. Any science that cannot answer to this first fact will talk past the person it claims to describe. This dissertation therefore keeps phenomenology as an authority on what must be explained. It insists that the first person is not an optional color but a datum that directs the rest of the argument and that constrains what counts as explanatory success (Merleau Ponty 146 to 164).

William James already anticipated the practical heart of the matter. He observed that effort sometimes unlocks reserves that the person could not have predicted. He refused to mystify this access. In his telling, reserves are not magic. They are licensed by belief and by a social world that confers permission to try. A person who expects welcome will risk more. A person who expects humiliation will conserve. This moral insight is not a sentimental add on. It previews a central technical claim of the present work, namely that willingness to invest control is a bet on expected returns that depends on value, cost, and the credibility of safety. The theory must therefore keep the social and ethical field in view if it is to do justice to the phenomenon of energy rather than reduce it to a private variable of will or chemistry. James’s voice reminds us that a person’s reserves belong as much to the community that receives or rejects them as to the individual who carries them. This is not a flourish. It is a boundary condition on any account that hopes to be both true and useful (James 113 to 146).

The ground then shifts to biology, but it does not leave the person behind. Classical homeostasis imagines regulation as negative feedback around fixed points. The picture is tidy and misleading. Organisms do not thrive by waiting for errors and correcting them after the fact. They live by anticipating what will be required and by budgeting mediators to meet those requirements in advance. Allostasis is the name for this prospective coordination across nervous, endocrine, immune, and behavioral systems. Under conditions of volatility and credible hazard, predictions are often wrong, and the controller must run at higher gain to keep the organism inside viable bounds. The cost of running hot accumulates as allostatic load, a pattern of wear that inscribes itself across blood pressure, glycemic control, inflammatory tone, sleep architecture, mood, and the narrowing of perceived possibility. The feeling of energy tracks this budgeting. When predicted demands can be met at tolerable cost, the organism experiences vitality. When predicted demands require high gain correction, fatigue is a prudent signal and withdrawal a rational policy. This explains why identical objective reserves can feel usable in one room and impossible in another. The theory will later formalize this fact without betraying its human center (McEwen and Stellar 2093 to 2101; McEwen 33 to 44; Sterling 5 to 15).

To avoid impressionism, the dissertation specifies how the brain arrives at the feelings that accompany these budgets. Interoceptive models treat affect as the organism’s best guess about its internal condition, given what it expects and what it senses. Agranular visceromotor cortices issue predictions about the state of the body that are compared with ascending visceral signals. Mismatch drives autonomic and endocrine correction along with the feelings that people name as heaviness, tightness, urgency, or ease. In this loop, fatigue and vitality are not raw readouts of physiology. They are inferences about whether predicted and sensed can be kept aligned without further harm. The language of spoons becomes theoretically sharp at this point. I have no spoons translates to the model’s expectation that alignment will be too costly to maintain during the next span of time. The commitment to this account comes with responsibilities. The dissertation will distinguish interoceptive accuracy, sensibility, and metacognitive insight so that changes in endorsement are not confused with changes in perception and so that reports are tied to objective signatures when appropriate. The aim is not to distrust the first person. The aim is to elevate it by showing how it connects to the loop that generates it (Barrett and Simmons 419 to 429; Craig 655 to 666; Critchley, Wiens, Rotshtein, Ohman, and Dolan 189 to 195; Garfinkel, Seth, Barrett, Suzuki, and Critchley 65 to 74; Mehling et al. e48230; Mehling et al. e0208034).

The organism does not only feel. It decides whether to invest control. Here the theory relies on work that treats effort as a costed choice. The expected value of control model places a computation in dorsal anterior cingulate that integrates predicted payoff, cost, and efficacy and that determines when additional control is worth deploying. Average reward accounts of vigor then explain why time feels expensive in barren environments and why response speeds improve when background reward is rich. In this frame, fatigue is not a character defect. It is a rational signal that present terms do not justify continued investment. Energy returns when expected value improves or when predicted costs fall. This decisional grammar explains why credible safety, clear purpose, and trustworthy response from others restore vitality even when a person has not slept more or eaten differently. It also exposes the ethical risk of exhortation. If expected value does not change, asking for more effort is a request for waste. The conceptual move is spare and powerful. It preserves agency and meaning while pulling the idea of energy out of the folklore of fuel and placing it in the light of value and cost where it can be examined without shame (Shenhav, Botvinick, and Cohen 217 to 240; Niv, Daw, Joel, and Dayan 507 to 520; Kurzban, Duckworth, Kable, and Myers 661 to 679).

Social life enters again with the full force of biology. Cues of exclusion and status threat recruit nociceptive networks and shift gene expression toward inflammatory programs that were adaptive under frequent bodily injury but that impose metabolic and affective taxes when they persist. The result is an increase in the price of remaining available to experience. Belonging, fairness, and credible care reverse those taxes and lower the cost of presence. The alliance between social neuroscience and social genomics allows the theory to say how a hostile meeting, a silent team, or a humiliating rule sets the organism to pay more for the next hour and why repair lowers that price. This is not metaphor. It is a convergent body of work that connects the civic to the cellular and that restores respect to the person who declines a further demand in an environment that punishes engagement (Eisenberger, Lieberman, and Williams 290 to 292; Slavich 265 to 295; Cole 31 to 37).

Lives change when priors change. The theory must therefore acknowledge the rules by which learning can release a person from costly habits of prediction. The modern exposure literature emphasizes inhibitory learning. The point is not to eliminate arousal for its own sake. The point is to create expectancies that contradict threat predictions and to stage retrieval so that new expectancies win when old ones are cued. Under defined conditions, brief reminders can render consolidated memories labile and open a window for reconsolidation based updating. The best human demonstrations are promising. Failures and boundary conditions mean that a sober account will treat reconsolidation as a possible amplifier rather than as a general recipe. Dissociation requires further care. Presentations dominated by overmodulation often reflect altered access to interoceptive signals rather than hyperarousal, and therefore demand distinct routes to revision that restore safe connection to the body as a precondition for expectancy violations to matter. The theory has room for these differences and treats them as tests of its ability to remain human while staying precise (LeDoux 155 to 184; Brewin, Dalgleish, and Joseph 670 to 686; Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23; Monfils, Cowansage, Klann, and Ledoux 951 to 955; Schiller, Monfils, Raio, Johnson, Ledoux, and Phelps 49 to 53; Lanius, Brand, Vermetten, Frewen, and Spiegel 701 to 708).

Time is a regulator that almost never receives its due. The two process model of sleep shows that homeostatic pressure and circadian phase govern the price of exertion. Identical tasks are cheap at one hour and expensive at another. The experience of being tired yet agitated in the evening and of finding tasks unexpectedly possible after slow wave rich sleep should not be treated as mood alone. It is a signature of temporal architecture that any serious theory must include. The benefit of placing time inside the theory is not to prescribe routines. It is to explain variance in energy that will otherwise be attributed to personal weakness or to mysterious fluctuations when the cause is the clock (Borbély 195 to 204; Borbély e13598).

The introduction closes by marking its own limits and by stating how the theory can lose. The free energy principle is used only as a conceptual umbrella that unifies perception, action, and learning around expected surprise. It is not asked to serve as a directly testable claim in this context because that use would invite unfalsifiability and would displace more specific and accountable literatures. Popular stories about autonomic hierarchies are engaged with anatomical caution. The inflammatory reflex and cardiac vagal indices are treated as real and useful but not as essences that explain every case. Where neurometabolic failure or neurodegeneration dominate, the theory will not overreach. It will mark its scope and import other frameworks. Most important, the work specifies findings that would force revision. If manipulations that should change expected value do not move willingness to invest control or reported vitality, then the valuation core is incomplete. If shifts in social safety and temporal alignment do not change felt cost, then the interoceptive and allostatic arms are overstated. If procedures that maximize expectancy violation do not generate inhibitory signatures under reasonable parameters, then the learning component is underspecified. Declaring these outcomes in advance is not a rhetorical device. It is an ethical one. A theory that cannot lose cannot help.

What remains is to write the account in a voice that keeps persons visible while making claims sharp enough to be tested by others. The promise of such a theory is clarity without cruelty. It allows one to say why the next hour might be unaffordable, how it can become affordable again, and why a world that answers back can give people their energy. The chapters that follow argue that this is not wishful. It is what predictive regulation, interoception, value, social safety, learning, and time already teach when they are read together with care.

Chapter 1. Phenomenology of effort, fatigue, and dignity

An account of mental energy that hopes to be both true and humane must begin with how a life feels from the inside. The person who says I cannot today is not taking a reading from a hidden gauge. The person is offering a situated judgment about the price of remaining available to a world that at this moment offers little help. The judgment is embodied before it is computed. It is a posture toward possibilities that can open or close the next hour. Phenomenology gives the vocabulary for this first fact. Maurice Merleau Ponty reads the body as our general medium for having a world, which means that vitality and weariness are modes of world relation rather than private numbers. Tonic tension, breath, and readiness are not shadows of an underlying quantity. They are the means by which meaning is organized at the scale of a day. When the world is disclosed as navigable the body leans forward and what counts as doable expands. When the world is disclosed as hostile or unresponsive the body tightens and what counts as doable contracts. To interpret these shifts as changes in a reservoir is to miss the way a person becomes available or unavailable to possibilities through the lived body itself. The claim obliges the theory that follows. Energy must be treated as style of access to a situation, not as fuel in a tank. This obligates care with measurement and with ethics because the voice that says the hour is unaffordable is not only reporting a state. The voice is also defending dignity under circumstances that have become expensive to endure (Merleau Ponty 146 to 164).

William James turns the same insight toward action and its moral terms. He notes that effort sometimes unlocks reserves and that persons discover they can continue where they believed they could not. He refuses to mythologize the discovery. Reserves are licensed by belief and by a credible invitation from the world. A person who expects reception will risk more than a person who expects humiliation. The practical consequence is immediate. Withdrawals should not be treated as weakness. They should be read as intelligent refusals to waste the self where the terms on offer make investment unlikely to be returned. This position does not romanticize refusal. It situates it within a moral field that must be named if any subsequent biology is to remain answerable to the life it hopes to explain. A theory that ignores this field will talk past the person it intends to help and will produce measures that are rigorous in form yet blind to meaning. James’s insistence that effort is relational will later align naturally with the claim that willingness to invest control is a bet on expected returns that depends on value, cost, and the credibility of safety. The opening task is to secure this human register so that later chapters can translate it into constructs without erasing the voice that gives the subject matter its significance (James 113 to 146).

The first critique appears at once. Phenomenological description can be accused of lyric precision without mechanism. The answer is not to abandon description. The answer is to let description set the explanandum and to braid it with disciplined accounts of how bodies regulate in advance and how feelings arise from the loop between prediction and sensation. That braid is what allows a humane theory to become testable without becoming inhuman. The present chapter therefore fixes several non negotiables that subsequent chapters must honor. First, capacity is a mode of world relation rather than a private quantity. Second, refusal can be rational when conditions make investment a bad bargain. Third, the terms of that bargain include meanings that are not epiphenomenal. If later biology cannot speak to these facts, the biology must be refined rather than the facts denied.

A second critique targets first person reports. Critics worry that the language of heaviness, pressure, or vacancy is unreliable or circular. The response is to refuse both extremes. We cannot treat reports as transparent truth, and we cannot treat them as noise. Reports are disciplined data about stance. The dissertation will later pair reports with objective indices of interoceptive accuracy and with behavioral measures of effort pricing. The pairing is not methodological fetish. It follows from the phenomenological claim that feeling and action are two aspects of the same orientation toward a situation. Divergence between what is said and what is done is itself a datum that requires explanation. Convergence strengthens the case that a stance has changed in the sense the person intends.

A third critique points to cultural and developmental variability. If vitality and fatigue are modes of world relation, the world in question is not the same everywhere or at every age. The theory anticipates this point by refusing universal psychologizing. It will treat social safety, reliability of response, and credible belonging as variables whose values depend on local ecologies and histories rather than as constants. The core claim is not that lives everywhere feel the same. The core claim is that across cultures and across development persons make embodied judgments about whether the immediate future is affordable, and that these judgments are sensitive to norms of reception, to promises that institutions keep or break, and to the learned statistics of harm and help. Where those conditions differ the thresholds for withdrawal and for renewed engagement will differ. A theory that cannot absorb such differences has confused a narrow habitus for a general principle.

With these critiques faced, the chapter turns to define its key terms so that they can constrain later chapters. Effort will mean a purposive investment of control in the service of a valued end under conditions where the outcome is uncertain. Fatigue will mean a felt judgment that further effort would exact a cost that cannot be justified by the returns that are likely under current circumstances. Dignity will mean the right to steward one’s availability without being compelled to spend the self where the terms are exploitative or humiliating. Each term is relational rather than intrinsic. Each term presupposes a situation and a history. This orientation lets the theory avoid two traps. The first trap is the temptation to treat energy as the sum of biochemical stores in isolation from appraisal, meaning, and time. The second trap is the temptation to treat energy as pure narrative untethered from the body’s regulators. The aim is to hold both together. The lived body sets the problem. The body’s regulators set constraints on plausible solutions. The person remains the unit of analysis, not the instrument reading nor the abstracted system.

The phenomenological picture yields three consequences that will be carried forward as demands on biology and decision theory. First, identical objective reserves can produce opposite stances depending on the world as it is disclosed. One must therefore prefer explanatory languages that admit conditionality and anticipation rather than languages that assume linear consumption of a resource. Second, the experience of sudden access to reserves cannot be reduced to an injection of fuel. It must be explained as a change in appraisal and in embodied permission supported by credible shifts in the situation or in belief. Third, the ethical status of refusal requires that any clinical or organizational application of the theory be transparent about the model of cost and value it is using, and that it invite the person to edit that model in light of aims that the person endorses. Later chapters will specify adjudication criteria precisely so that refusals remain audible as knowledge rather than being pathologized in the name of efficiency.

The chapter also anticipates a methodological worry that will follow the dissertation if it is not addressed here. If vitality and fatigue are styles of access to the world, attempts to measure them as if they were free floating states will distort what they are. The remedy is not to abandon measurement. It is to design measures that are tethered to situations and to time. First person accounts can be elicited with structured prompts that reference the demands and the returns that are expected in the near horizon rather than asking for decontextualized numbers. Behavioral measures of effort should be tied to tasks whose value and cost parameters are made explicit so that a refusal to exert control can be read as an intelligible outcome rather than as a generalized deficit. Physiological measures should be interpreted as convergent evidence about the cost of regulation only when they move with changes in prediction and appraisal rather than in isolation. These methodological consequences are not procedural details. They are direct implications of the phenomenological stance.

Finally, the chapter draws a line around what it will not claim. It will not claim that all instances of withdrawal are rational in a narrow sense. Persons misjudge, become frightened, or become captured by avoidance that reduces life. The theory does not bless every refusal as wise. It insists that refusals be interpreted before they are condemned and that the conditions that set price be examined where the person reports a debt that cannot be paid. It will not claim that feelings are infallible guides. It will claim that feelings are the organism’s testable hypotheses about its near future and that those hypotheses deserve the same respect and scrutiny that one affords to any model that must be updated by contact with the world. It will not claim that embodiment exhausts meaning. It will claim that every meaning that matters to energy is lived through a body that anticipates and that is altered by credibly different terms.

The chapter now hands the argument to biology with a clear mandate. The next chapter must show how the body anticipates rather than reacts and how chronic uncertainty inscribes itself as wear and as narrowness of perceived possibility. It must do so in a way that permits error and revision. It must remain answerable to the person’s report that an hour felt impossible on Tuesday and became possible on Thursday without any new substance added to the body. It must refrain from replacing a life with a mechanism. It must treat the mechanism as the means a life uses to continue. Only an account that can meet these demands deserves the name of theory in a field where lives are at stake and where a wrong description can become a wrong response.

Chapter 2. From homeostasis to allostasis, predictive regulation as the ground of energy

A theory that treats mental energy with seriousness must replace the picture of a thermostat with the picture of a strategist. Classical homeostasis imagines a controller that waits for errors and then pushes the system back toward fixed set points. That picture is tidy and historically important, yet it fails to account for the daily fact that organisms live forward. They do not stand watch at a dial. They cultivate expectations about what will be required and they budget across systems to meet those requirements in advance. The modern name for this stance is allostasis, regulation through change. It is an economy of anticipation rather than a ritual of restoration. In this frame subjective energy is not a hidden fluid that rises and falls. It is the felt summary of whether the organism believes it can keep itself within viable bounds while the next period of demand arrives. When the forecasted mediators that would be required to maintain order are affordable, life feels possible. When the forecast is grim, fatigue is an intelligent signal and withdrawal becomes a rational policy rather than a moral failure. This is the interpretive move that allows a humane phenomenology to clasp hands with disciplined biology without sacrificing either voice.

The biological content of allostasis is specific. Across the autonomic, endocrine, immune, and behavioral axes, the body organizes a conditional plan for the near future and then updates the plan as new evidence arrives. Catecholaminergic bursts support rapid mobilization, hypothalamic pituitary adrenal activity supports medium horizon metabolic and immune adjustments, inflammatory mediators prepare tissue for damage at the price of later wear, and behavior reorganizes around expected challenges or reprieves. The same mediator can be protective or damaging depending on dose, context, and time. Bruce McEwen’s program distilled the logic into a durable grammar. Mediators are necessary for adaptation and survival, yet when uncertainty and prediction error remain high the cost of running high gain control accrues as allostatic load that can be read in the slow drift of blood pressure, glycemic control, visceral fat, sleep architecture, and mood, and in the quiet narrowing of what a person believes is livable in a week or a year (McEwen and Stellar 2093 to 2101; McEwen 33 to 44). The old set point picture encouraged a search for a missing ingredient when energy fell. The allostatic picture encourages an audit of the model that predicts the near future and of the price that model implies.

This shift from correction to budgeting matters at the level of theory because it clarifies what kind of object energy talk is. It is not a thing. It is an expectation. Peter Sterling states the principle without romance. The controller learns the statistics of its niche and makes proactive adjustments that minimize costly deviations while preserving flexibility. In this light the sentence I am out of energy becomes a plain report about expected affordability given the present niche statistics. The same objective reserves can feel usable in one hour and unusable in another because the world is not offering the same terms and the learned model does not assign the same predicted costs. The inflammatory tone of a team that punishes initiative raises the price of presence. A stable promise from others lowers it. The circadian and sleep context gates that price further. All these variables enter the predictive budget before a person speaks, and what is spoken is a summary judgment about whether the next hour can be lived without unacceptable harm. The utility of allostasis is that it joins these disparate facts under one structure of explanation that earns prediction rather than offering metaphors that happen to feel right in the room (Sterling 5 to 15).

There are four recurring critiques of the allostasis frame that must be addressed directly to secure its place as the biological ground of this dissertation. The first critique claims that allostasis is descriptive aggregation rather than mechanism, a label placed atop many moving parts. The answer is twofold. First, the descriptive synthesis already has explanatory bite, because it places the time dimension at the heart of regulation and thereby renders intelligible patterns that homeostasis treats as aberrations rather than as the price of anticipation. Second, the present work embeds allostasis within an explicit inferential architecture in the next chapter, where interoceptive prediction and precision are specified as the computations that implement the budgeting stance that allostasis names. In that way the project retains the ecological wisdom of allostasis while satisfying the demand for algorithmic clarity.

The second critique claims that allostatic load popularizes a basket score that invites category errors. It can be tempting to index wear by a composite and then misattribute causation to any component that correlates with the composite. The response is to keep the concept of load as a qualitative pattern across time rather than as a single decontextualized number, and to interpret any biomarker as convergent evidence only when it moves with changes in prediction and with changes in the felt price of control. Cortisol can be adaptive or costly. Heart rate variability can index flexible control or reflect respiratory confounds. Inflammatory markers can protect tissue or tax mood and cognition depending on context and duration. The theory therefore refuses to elevate any one measure to essence and insists that any physiological signal be interpreted as part of a coherent temporal story in which anticipation, correction, and wear track together under specified conditions (McEwen 33 to 44).

The third critique concerns generality and scope. If allostasis is the true language of life, does it then explain everything and therefore explain nothing. The remedy is boundary setting. Allostasis is the correct frame for ordinary life under uncertainty, for trauma sequelae where threat priors keep prediction error high, and for chronic stress where volatility and lack of control are structural. Allostasis is not sufficient where neurometabolic failure, endocrine catastrophe, or degenerative processes set the primary terms. In such cases the predictive budget is constrained by deficits that must be named in their own literatures. The present theory marks those edges and refuses panacea.

The fourth critique concerns causality from social life into biology. Some readers worry that claims about exclusion, status threat, and reliability of others are moralized stories painted onto biology after the fact. The answer is that social signals have been shown to recruit nociceptive networks and to retune immune transcription in conserved patterns that reflect ancestral hazards and modern burdens. The biology carries social marks because social life has always been a matter of life and death for a social species. When the world signals that one is unwelcome or unsafe, the controller prudently shifts toward high gain modes that raise momentary survival probability at the price of later wear. When the world signals that one is received, the controller prudently relaxes those settings and lowers cost. A theory that refuses this linkage to protect a narrow account of mechanism removes from the organism the very context that decides whether the next hour is affordable. The linkage is part of the mechanism because the niche is part of the organism’s predictive budget.

The consequences of an allostatic stance for the concept of energy can now be spelled out without reduction or sentiment. First, subjective vitality is the affective face of successful budgeting under expected demand. The organism predicts that mediator deployment will be affordable and therefore permits action, and that permission is felt as readiness. Second, subjective fatigue is the affective face of a forecast that sustained high gain control would be required and would exact unacceptable future costs. The organism withholds, and that prudence is felt as heaviness and thinned possibility. Third, the same individual can move between these stances without any change in objective reserves, because what has changed is the expected price of keeping order over the next stretch of time. Fourth, because the expected price is a function of model quality as well as of environment, two persons with similar objective reserves and similar environments can report different energy because they have learned different statistics of help and harm. The theory therefore predicts meaningful person by context interactions rather than any simple mapping from a resource to a feeling.

Critics who prefer crisp levers will ask how this structure differs from a promissory story. The answer is in the precision of demands the structure makes on downstream theory and on evidence. If allostasis is correct, then any manipulation that reduces credible uncertainty should lower mediator gain and subjective price. A day made more predictable by reliable time signals should produce a cheaper next hour than a day stripped of cues. A context that confers social safety should allow lower inflammatory tone and lower autonomic effort for the same behavior compared with a context that humiliates or threatens. If these relations do not obtain after reasonable care in measurement and design, then the structure must be trimmed or replaced. The present work anchors those tests in literatures that already contain candidate signatures of change without asking the allostasis concept to conjure those signatures out of rhetoric.

The frame also constrains how to read common confusions. It is tempting, for instance, to infer that low morning cortisol is a sign of resilience because high cortisol is often associated with acute stress. Allostasis makes the picture more serious. Under chronic load a flattened diurnal rhythm can reflect a system that has lost dynamic range and that pays other prices in metabolic and immune domains. It is tempting to infer that high heart rate variability is always desirable. Under certain demands a healthy controller should be able to reduce variability to marshal action. The person who cannot do so at need is not flexible. These examples are not procedural footnotes. They show how a predictive budget reinterprets signals that would be misread under a reactive set point model.

Allostasis finally gives the bridge from the first chapter’s human register to the next chapter’s inferential content. If regulation is prospective, then the controller must carry a model of what the body will need and a confidence about the precision of that model. Feelings are the organism’s report about the quality of that model and about the likely cost of enforcing it. A person can then correctly say that an hour is unaffordable when the model expects costly mismatch even if the objective reserves would be sufficient under different statistics. The controller may be wrong in either direction. It may run hot because it learned that the world would not help and that learning is now out of date. It may run too cool because it learned that help would arrive when it no longer does. The remedy in both cases is not to shame the person for reporting fatigue or to congratulate the person for reporting vitality. The remedy is to engage the model that is being used to price the next hour and to expose it to conditions that earn revision. That is an inferential task as much as a biological one, and it belongs to the next chapter.

Before handing off, the chapter names two boundaries that protect against overreach. The first boundary is temporal. The useful unit of analysis is the trajectory, not the point measurement. An intervention or a measurement protocol that refuses time will misread both adaptation and wear. The second boundary is ethical. To say that fatigue is a prudent price signal is not to argue that persons should obey it in every case. Lives often ask for costly action in the service of goods that persons endorse even when those actions are expensive. The theory therefore does not convert a budget into a dictate. It offers an explanation of why the price feels as it does and a disciplined way to speak about what would lower that price without violating meaning.

In conclusion, allostasis supplies the philosophical and biological ground for a theory of mental energy that takes persons seriously and that earns prediction. It tells us why a life can become unaffordable under volatility even when reserves remain, why safety and reliable time can restore possibility without adding any substance, and why social worlds that answer back in humane ways change physiology in register with felt capacity. The next chapter specifies the computations by which the brain makes and revises these budgets through interoceptive prediction and precision, so that the argument can move from a strategic picture to the mechanisms that implement it while continuing to honor the human voice that opened the work (McEwen and Stellar 2093 to 2101; McEwen 33 to 44; Sterling 5 to 15).

Chapter 3. Interoceptive inference, how the body becomes a prior for feeling and action

The move from homeostasis to allostasis turns regulation into a forward looking enterprise. Interoceptive inference explains how that enterprise is carried out at the level of feeling and action. On this view the brain issues descending predictions about the state the body ought soon to occupy, compares those predictions with ascending visceral evidence, and adjusts both physiology and belief so that the organism can remain within viable bounds while demands arrive. Feeling is not a decorative after effect. Feeling is the organism’s best current guess about its internal condition and about whether alignment between predicted and sensed can be sustained without unacceptable cost. Read this way, the common report of fatigue is not a reading from a fuel gauge. It is a forecast that near future mismatch will require high gain corrections that the organism judges to be overpriced. The corresponding report of vitality is a forecast that predicted and sensed will remain in register at tolerable cost. These claims give muscle to the phenomenology of the first chapter and place it inside a mechanism that earns prediction rather than relying on persuasive images alone (Barrett and Simmons 419 to 429; Craig 655 to 666; Critchley, Wiens, Rotshtein, Ohman, and Dolan 189 to 195).

The anatomical and functional scaffold for this claim is now well described. Lamina I spinothalamic pathways and vagal afferents carry information about temperature, visceral stretch, chemosensation, and cardiovascular states to posterior insula and to brainstem nuclei. Agranular and dysgranular visceromotor cortices in anterior insula, cingulate, and orbitofrontal regions generate predictions about interoceptive state and send autonomic and endocrine commands that attempt to make the world of the body match what is expected. The anterior insula in particular is argued to integrate predicted and sensed interoception into a momentary model of the embodied self, with mismatch registered as affective urgency to revise the model or to act on the body and the environment. From this architecture follows a sober interpretation of many ordinary experiences. When the heart accelerates under a learned threat prior, predicted arousal is large and precision on threat cues is high, so the same physiology that could have been felt as readiness is instead felt as dread. When predictions are revised in a context that proves safe, the same physiology becomes legible as energy, the cost of control falls, and the person reports that the next hour is affordable again (Craig 655 to 666; Critchley, Wiens, Rotshtein, Ohman, and Dolan 189 to 195; Barrett and Simmons 419 to 429).

Precision is the hinge in this account. The organism not only predicts states but also assigns confidence to those predictions and to the sensory channels that can correct them. Under credible danger the system will raise precision on exteroceptive threat cues and lower precision on benign interoceptive variation in order to avoid costly misses. Under credible safety the system can permit more variance in both directions and still remain within bounds. Feelings therefore carry information about confidence as well as about state. In practical terms this means that what a person names as heaviness or lightness is shaped by how much trust the system places in its own ability to keep predicted and sensed aligned as conditions shift. The theoretical benefit is that interoceptive inference accounts for both the content and the conviction of feelings without positing a hidden essence of energy that moves up and down like a liquid in a tube. The ethical benefit is that confidence can be earned. Precision can be redistributed by experience and care rather than by exhortation, which honors the first chapter’s claim that dignity depends on the world answering back in ways that permit a person to continue (Seth 565 to 573; Barrett and Simmons 419 to 429).

To prevent slippage between rhetoric and evidence the dissertation adopts a measurement architecture that maps directly onto the theory. Interoceptive accuracy refers to performance on objective tasks that require detection or discrimination of internal events, such as heartbeat counting under controlled pacing or heartbeat discrimination with metacognitive scoring. Interoceptive sensibility refers to a person’s beliefs and attitudes about their own bodily awareness, often measured by questionnaires. Interoceptive metacognitive insight refers to alignment between confidence and objective accuracy. These constructs move together in some conditions and dissociate in others. After an intervention a rise in sensibility without a rise in accuracy may reflect endorsement rather than calibration. A rise in accuracy without insight may place a person at risk of over interpretation of benign signals. The theory therefore treats changes in these channels as distinct sources of evidence and interprets their joint pattern against the background claim that fatigue and vitality are inferences about future alignment under cost constraints. The point is not to multiply measures. The point is to keep concepts honest by tracking prediction, error, and confidence at the levels the theory names, and to resist the temptation to declare victory when a questionnaire moves but the organism’s guessing about itself has not become more precise in a way that lowers the price of control (Garfinkel, Seth, Barrett, Suzuki, and Critchley 65 to 74; Mehling et al. e48230; Mehling et al. e0208034).

Several recurrent critiques of interoceptive inference deserve to be engaged directly. One claims that predictive accounts are too general to be falsified and that any pattern of data can be re described as a change in prediction or precision. The response is to specify signatures that count in favor of the theory and signatures that count against it. If credible increases in social safety do not produce predictable shifts in affective interpretation of identical visceral states, the theory is weakened. If manipulations that specifically alter interoceptive prediction produce no change in first person stance or in autonomic deployment while exteroceptive predictions are held constant, the theory is weakened. If persons who improve objective interoceptive accuracy under conditions that preserve safety do not also report lower price of engagement when bodies fluctuate, the theory is weakened. The point is not to create brittle tests. It is to convert a helpful unifying frame into a set of disciplined expectations that can lose to data and can therefore learn.

A second critique is that interoceptive training is sometimes offered as a curative generality without a mechanistic target. The theory sets the target. Training that increases accuracy without context and without a credible reduction in expected cost may amplify vigilance and raise price. Training that privileges sensibility without accuracy may inflate confidence without calibration. The desired movement is a coordinated change in prediction, precision, and environment such that benign fluctuations are expected and trusted, threat signals are neither minimized nor over weighted, and the cost of maintaining alignment falls. Under chronic threat or humiliation such calibration will not hold, because the world continues to make the body expensive to inhabit. The synergy with social safety is therefore structural rather than optional. Persons regain energy when predictions about the self and the world become kinder and when the world behaves kindly enough to make those predictions true often enough to update the model. Where either condition fails, the theory predicts relapse into expensive modes even in the presence of skill (Slavich 265 to 295; Eisenberger, Lieberman, and Williams 290 to 292; Cole 31 to 37).

A third critique insists on heterogeneity, especially in trauma related presentations. Persons who dissociate often describe distance from bodily signals rather than intrusive arousal. The theory acknowledges dissociation as a distinct failure mode of interoceptive inference. Precision on interoception is down weighted or the channel is gated off from appraisal, sometimes in the service of survival in contexts where signals were unbearable. In such cases the path to a cheaper next hour is not to force contact with intense sensation before any permission to feel has been won. It is to restore safe access to gentle internal cues and to link them to meanings and actions that do not annihilate a fragile self. Only then can expectancy violations about safety propagate through the network to change predictions about what the next moment will cost. This stance protects persons from a one size model and protects the theory from erasing difference behind a graceful unifier. It also sets up the later ethical claim that refusal can be rational even inside therapy when the price of contact would be to endanger the very continuity that the therapy hopes to preserve (Lanius, Brand, Vermetten, Frewen, and Spiegel 701 to 708).

Two further limits keep the account from promiscuity. First, there are presentations in which predictive models are constrained by primary neurometabolic, endocrine, or degenerative processes. In those contexts the price of presence is set by failures or saturations that are not primarily inferential. The theory will not overreach into those domains. Second, there are narratives that attribute interoceptive phenomena to phylogenetic autonomic hierarchies that remain contested at the level of anatomy and development. The present work adopts a conservative stance. It privileges the inflammatory reflex and reproducible cardiac vagal metrics as one set of signals in a larger inferential loop. It refuses to elevate any single autonomic story to the status of essence. This restraint is not a concession to fashion. It is a requirement of coherence in a theory that must remain revisable by evidence and that must not import what cannot be defended into an already ambitious synthesis.

The conceptual gains of interoceptive inference for a theory of energy are now clear and practical. First, it explains why identical physiology can feel like readiness in one context and exhaustion in another, because feeling rides on prediction and on confidence. Second, it clarifies why skills and schedules change price only when they change expectations and precision, which prevents the field from confusing routines with mechanisms. Third, it places dignity back at the center, because a person’s report about heaviness becomes a testable hypothesis about their near future and about the terms the world is offering, not a confession of lack. Finally, it aligns directly with the value theoretic account that follows. The decision to invest control depends on the expected value of control, and expected value depends on predicted cost. Predicted cost in turn depends on interoceptive models of what it will take to keep the organism within bounds. When those models are pessimistic the value of control falls, and the person withholds. When those models become kinder through safety, time kept well, and learning that truly updates priors, the value of control rises, and the person spends. In this way the third chapter hands the argument to the fourth without any seam between feeling and choice. The organism that guesses well about itself can afford to continue, and the organism that is forced to guess in danger will pay too much to remain. The theoretical and ethical task is to help the first condition become common without pretending that a life is a machine. Interoceptive inference gives us a way to do that by speaking in a single voice about body, meaning, and action while keeping every important claim answerable to data and to the person whose life is being described (Barrett and Simmons 419 to 429; Seth 565 to 573; Garfinkel, Seth, Barrett, Suzuki, and Critchley 65 to 74; Mehling et al. e48230; Mehling et al. e0208034; Slavich 265 to 295; Eisenberger, Lieberman, and Williams 290 to 292; Cole 31 to 37; Lanius, Brand, Vermetten, Frewen, and Spiegel 701 to 708).

Chapter 4. Control as a scarce good, value theoretic foundations of effort

The organism that predicts its bodily future must still decide whether to spend itself. Interoceptive inference explains how feelings arise from predictions about alignment between sensed and expected internal states. Those feelings are not the end of the story. They are inputs to a decision about whether to invest control in the service of a goal under uncertainty. A theory of mental energy that remains faithful to persons must therefore say what it means to spend control and why the same individual will sometimes choose to continue and sometimes choose to conserve. The claim defended in this chapter is spare. The felt economy of energy is a decision variable. It summarizes whether deploying additional control is expected to pay, net of its costs, given beliefs about the world, beliefs about the body, and the time that remains. The analytic spine comes from the expected value of control account of anterior cingulate function and from average reward theories of vigor that treat time as a scarce resource. Together they turn energy from a mystified reservoir into a judgment about value that can change with meaning and context without reducing meaning to a price tag. They also honor the moral point that withdrawal can be prudent when the terms on offer would waste the self, while keeping open the possibility that new terms can make the next hour worthy again (Shenhav, Botvinick, and Cohen 217 to 240; Niv, Daw, Joel, and Dayan 507 to 520).

The expected value of control framework proposes that a supervisory system estimates the benefit of adding control to a process, subtracts the expected cost of doing so, and weighs the probability that control will succeed. It then allocates control when the expected net benefit is positive. On this view there is no mystical force of will. There is a calculation, sometimes rapid and tacit, about whether extra precision and persistence will change outcomes in a way that matters for the agent. The dorsal anterior cingulate appears to encode components of this calculation, integrating payoff, cost, and efficacy signals, and broadcasting a bias for action when the calculus favors investment. The details of neural implementation are debated and will not be overstated here, yet the behavioral logic is clear and tractable. When contexts become stingy, when success feels unlikely, or when the price of control is predicted to be high because the body will exact a toll, the decision variable falls and the agent slows or disengages. When contexts become generous, when success feels likely, or when the price of control is predicted to be low because the body will cost less to mobilize, the decision variable rises and the agent accelerates. The phenomenology of vitality and fatigue follows those movements, which binds the felt stance of Chapter 1 and the predictive regulation of Chapter 3 to a decision rule that can explain why identical reserves can support opposite choices across hours and places (Shenhav, Botvinick, and Cohen 217 to 240).

Average reward theories of vigor add a quiet but decisive layer. They treat time as an opportunity cost. When the background rate of reward is rich, time spent moving slowly leaves more on the table, so agents increase vigor because the opportunity cost of delay is high. When the background rate of reward is poor, time spent moving quickly is not compensated by additional gains, so agents slow because the opportunity cost of delay is low. The experience of energy maps onto this ecology. In barren environments even small efforts feel too expensive to repeat because they displace other scarce options without promise of return. In responsive environments even large efforts feel affordable because each moment spent engaged tends to be answered. Persons therefore report energy not because more substance has been added to them, but because the exchange rate for time has been altered in their favor. This reframing protects dignity in a way that simple exhortations cannot. To ask for vigor in an environment that pays poorly is to ask for waste. To increase vigor by increasing meaningful return honors the person’s intelligence about the terms under which life is livable (Niv, Daw, Joel, and Dayan 507 to 520).

Opportunity cost accounts of subjective effort then place these ideas within a single felt metric. The unpleasantness of effort is not an intrinsic toxin that control secretes. It is the signal that other valuable actions are being forgone while one persists here. The more attractive the alternatives, the higher the felt cost of staying. The less attractive the alternatives, the lower that cost. Much that has been labeled laziness can be reinterpreted, without romance, as a calculation that present terms do not justify further investment. Much that has been labeled grit can be reinterpreted, without cynicism, as a calculation that present terms do justify continued investment because the world is answering in kind. The practical upshot is not that effort is cheap talk. It is that effort is a wise bet when value is credible and a poor bet when value is counterfeit. The theory insists that ethical and institutional questions are therefore first order determinants of energy and not decoration for a private struggle between will and weakness. A workplace that humiliates makes time barren and increases the opportunity cost of staying. A community that receives persons as persons makes time rich and decreases that cost. The felt economy of energy tracks those facts with precision even when language lags behind them (Kurzban, Duckworth, Kable, and Myers 661 to 679).

The integration with interoceptive inference is direct. Predicted bodily costs enter the control calculus as a price term. When the organism expects that keeping predicted and sensed aligned will require high gain corrections for the next hour, the cost of control rises and the expected value of control falls. When the organism expects that alignment can be maintained at low gain through safety, sleep, or calibration, the cost of control falls and the expected value of control rises. Feelings about the body are therefore not noise but variables that alter the decision to spend. This interpenetration rescues both sides from simplification. It saves decision theory from a bloodless formalism by making cost depend on a living body that guesses about itself. It saves interoceptive science from romanticism by making feelings answerable to whether action pays. The result is a single account in which a person declines to continue because their body predicts that doing so will exact an unjustifiable price in the present ecology, and accepts to continue when that prediction changes because the world or the body has credibly changed in ways that lower cost or increase return.

Several critiques of this value based account are predictable and deserve engagement at the level of theory rather than in the appendix. One critique accuses the framework of reductionism. It worries that rendering energy as a decision variable will strip it of meaning and treat human action as an economic game in which the only currency is payoff. The reply is twofold. First, value is not money here. Value is whatever matters to the agent. It includes moral goods, relational goods, craft goods, and the meanings that a person endorses as worth living for. Second, formalizing the choice to spend control does not trivialize it. It clarifies that when a person continues in the face of pain for the sake of a loved one or for the sake of a promise, the expected value of control is high because the value term is large, not because the cost term is small. In this frame courage is not a miscalculation. It is a lucid assignment of weight to an end that justifies cost. The account therefore protects dignity better than mechanistic stories that reduce persistence to trait or than sentimental stories that ignore price.

A second critique targets biological localization. It cautions that anterior cingulate cannot be identified with a homunculus that decides, and that signals attributed to expected value of control may reflect multiple overlapping processes. This caution is sound. The theory does not require a cortical sovereign. It requires only that some system integrate information about payoff, cost, and efficacy in a way that biases allocation of control. The dorsal cingulate is a strong candidate among others and the evidence that such integration occurs is robust, yet the claims that matter for this dissertation are behavioral and phenomenological. They can stand even if the precise mapping from computation to structure is later revised. That humility is a feature rather than a weakness. It keeps the account revisable by evidence instead of asking current imaging conventions to bear weight that history shows they cannot yet carry with certainty (Shenhav, Botvinick, and Cohen 217 to 240).

A third critique warns against normative creep. It asks whether expected value of control theories are being used to justify social arrangements by declaring that those who withdraw have simply made rational choices in barren ecologies. The concern is real. The correct response is to restore causality at the level of institutions. If a school, clinic, or firm creates barren ecologies by refusing reliable response, the theory does not excuse withdrawal as an unchangeable fact. It indicts the environment and hands designers concrete variables to alter. Make returns credible. Reduce avoidable costs by reducing humiliation and by providing time that aligns with bodies. Increase efficacy by giving persons tools they can learn and by structuring tasks so that control can succeed and be recognized. None of these are prescriptions for procedure in a theory chapter. They are constraints on any application that claims to respect the theory. The dignity of the agent remains central. The decision variable should not be gamed to squeeze more labor from an unchanged world. It should be used to make a world in which the agent’s willingness to invest is matched by a world that deserves the investment.

A fourth critique asks whether the account gives affect its due. If effort is always a bet on value, does it leave space for moods and for irrational discouragement. The reply is to fold affect into the price of control. Moods are not noise. They are historical priors about how the next hour is likely to go. They alter expected payoff, expected cost, and expected efficacy. A grey morning lowers the prior that the world will answer back, thereby decreasing expected value before any new evidence has arrived. That change is not a moral failure. It is a prediction that can be revised by credible disconfirmations such as reception by others, successful small acts, or the visible stabilization of a chaotic day. The account therefore situates affect without pathologizing it and without giving it sovereignty. Mood becomes a starting point for the decision rather than an unchallengeable verdict.

The theory yields several high level implications that can be stated without dropping into procedure. First, if the background reward rate rises through credible success and recognition, persons will report more energy because time becomes expensive to waste and the decision variable moves in favor of engagement. Second, if the predicted bodily cost of exertion falls through sleep, safety, or calibration, persons will report more energy because control becomes cheaper to deploy. Third, if efficacy rises because a task has been fitted to a person’s current skills and because tools genuinely extend their reach, persons will report more energy because attempts are more likely to succeed. Fourth, if alternatives that were once attractive become less so because they fail to pay, the felt cost of staying with a difficult but meaningful task can fall even when the task is unchanged. None of these imply a technocratic fix. Each implies that energy is intelligible in ordinary language once value and cost are placed at the center.

Two boundaries protect the account from overreach. There exist presentations in which neurometabolic failure or primary endocrine catastrophe sets cost independent of appraisal. In such cases expected value of control will be low regardless of context because the price term reflects defects that are not under inferential control. The theory defers to those literatures rather than stretching to cover what they already explain better. There also exist social arrangements that punish persons who try in ways that ruin their willingness to continue. In such cases energy is not restored by telling a better story about value. It is restored by changing the world that makes value incredible. The account will not allow a cognitive gloss to serve where structural repair is required.

The chapter closes by restating the central link that holds the dissertation together. Phenomenology taught that vitality and fatigue are modes of world relation rather than numbers. Allostasis taught that regulation is anticipatory and that the price of keeping order can rise in chronic uncertainty. Interoceptive inference taught that feelings are predictions about whether alignment between predicted and sensed can be maintained without harm. The present chapter adds the rule by which those feelings become choices. Agents invest control when expected value exceeds cost. When value collapses or cost rises, agents conserve. The felt economy of energy is this choice seen from within. When a person says that the next hour is unaffordable, a humane and exact theory hears that expected value has fallen below a threshold set by their history and their world. When a person says that energy has returned, the same theory hears that predicted costs have fallen or that returns have become credible again. The next chapter will carry this decision variable into the social field. It will ask how belonging and humiliation change the biology that sets price, how institutions scaffold or tax the expected value of staying present, and how a civic world that answers back becomes the simplest instrument by which energy returns without magic and without coercion (Shenhav, Botvinick, and Cohen 217 to 240; Niv, Daw, Joel, and Dayan 507 to 520; Kurzban, Duckworth, Kable, and Myers 661 to 679).

Chapter 5. Social safety as biological permission to be a person

The decision to spend oneself is not taken in isolation from the regard of others. For a social organism the credibility of reception changes the cost of remaining available, and that change is written into tissue as well as into expectation. Social Safety Theory gathers diverse findings under a single claim. Cues of rejection, exclusion, and status jeopardy recruit systems for the registration of pain and the control of vigilance while shifting immune transcription toward a proinflammatory profile that prioritizes short term wound readiness over long term repair, a shift that raises the price of keeping order when such cues persist across days and years (Slavich 265 to 295; Cole 31 to 37). The phenomenology is familiar. In environments that humiliate or fail to answer, bodies tire of being presented. In environments that receive and reciprocate, bodies lean forward. The theory treats these ordinary experiences as evidence about a predictive budget rather than as ornaments to a private psychology. Belonging and its absence alter priors about hazard, precision assignments to signals, and the mediator settings that make the next hour cheap or expensive to live.

Two biological lines fix this claim. The first is neuroimaging work showing that structured exclusion recruits regions implicated in the construction of aversive bodily states and in the assignment of salience, including the dorsal anterior cingulate and anterior insula. The finding does not say that social and nociceptive pain are identical. It says that the organism uses overlapping resources to register the cost of social violation and bodily insult, which is why humiliation is not a metaphorical injury but a state with bodily consequences that shape the next decision to appear or withdraw. The organism that expects such cues assigns higher precision to threat and expends more effort to anticipate and counter it. The price of presence rises and is felt as fatigue that rationally conserves the self until terms improve (Eisenberger, Lieberman, and Williams 290 to 292). The second line is social genomics. Under credible social threat the conserved transcriptional response to adversity increases expression of proinflammatory genes while dampening antiviral programs, a pattern that alters mood, pain sensitivity, and metabolic handling and that makes vigilance and withdrawal more attractive choices on the next page of life. When affiliation becomes credible again, these set points can move in the opposite direction, lowering the bodily cost of remaining available and restoring the expected value of control that Chapter 4 identified as the proximate determinant of engagement (Cole 31 to 37).

The integration with interoceptive inference is direct. Social safety changes how the organism weights exteroceptive cues of evaluation and interoceptive cues of arousal. Under threat the system assigns high precision to ambiguous social signals and low tolerance for benign visceral variation, which causes the same heart or gut to be felt as evidence of impending failure rather than as ordinary fluctuation. Under safety the system can down weight ambiguous threats and increase tolerance for visceral noise without losing contact with real danger. Feelings of energy and heaviness track these reweightings. They are inferences about whether predicted and sensed can be held in register at acceptable cost while one remains present in a field of others who may help or harm. This connection prevents the theory from treating social life as a late influence on a finished calculation. Social life is an input to prediction and to the price term that prediction implies.

Critiques of this social account arise in several forms and must be met before the chapter proceeds. One critique alleges reverse causation. It proposes that low energy and dysphoria cause withdrawal, and that withdrawal is mistaken for social threat rather than being its cause. The reply is that the literatures cited above include manipulations and natural experiments in which social cues are altered while internal resources are held as constant as is feasible, and the downstream effects are measured on pain networks, immune transcription, cortisol dynamics, and next day affect. The direction of effect is bidirectional, but the claim that social context drives biology and felt cost is not an inference from correlation alone. It is a measured influence that remains after reasonable covariates are considered, including prior mood and sleep (Eisenberger, Lieberman, and Williams 290 to 292; Slavich 265 to 295). A second critique targets specificity. It notes that patterns like the conserved transcriptional response to adversity may be invoked too widely. The answer is restraint. CTRA is treated here as one mechanism by which social statistics enter physiology, not as a total explanation. Where inflammatory changes are absent or move contrary to prediction, the theory must accommodate other routes by which safety changes cost, including shifts in autonomic tone, endocrine set points, and the precision assignments that drive interoceptive interpretation (Cole 31 to 37).

A third critique concerns reproducibility and cultural scope. Many demonstrations derive from samples with particular demographics and from laboratory tasks that stage exclusion in stylized ways. The dissertation does not claim a universal threshold for what counts as exclusion, nor does it claim that every culture marks status and belonging with the same signals. It claims that across cultures social worlds teach persons what to expect from others, and those expectations change mediator settings and precision assignments in ways that alter the price of remaining available. Where signals differ, priors will differ, and the mapping from cue to price will differ. The theory therefore treats social safety as a general determinant whose local implementation must be learned rather than as a fixed parameter. This stance is compatible with the strongest evidence and protects the account from the overreach that would convert a disciplined pattern into a thin universal.

A fourth critique warns against overreading network overlap between social and physical pain. The best response is conceptual clarity. Overlap is not identity. The dissertation does not claim that the brain confuses a cold shoulder with a cut. It claims that a social species reuses machinery for salience assignment and aversive valuation across harms that threaten continuity, bonding, or status, and that the reuse makes adaptive sense because the same bodily preparations can serve both. The practical inference survives the critique. If an institution produces humiliation and hypervigilance, the next hour will cost more to live even in the absence of physical threat, and persons will rationally conserve. If an institution removes humiliation and provides credible reception, the next hour will cost less to live even if demands remain high, and persons will rationally spend.

Daily life data bind these claims to ordinary time. Diary and sampling studies show that minor social stressors and hassles alter the next day’s affect and endocrine output even when major events are absent, and that the accumulation of small costs predicts withdrawal and narrowed repertoire. These patterns are not the whole story of energy, yet they show why environments that steadily tax social prediction produce fatigue out of proportion to measurable caloric expenditure. They also show why relief is often experienced as surplus energy when nothing tangible has been added. The world has become cheaper to inhabit for a time because it has stopped asking for so much anticipatory work, and the organism is correct to spend more of itself in such a day (Almeida 64 to 68; Stawski, Almeida, Lachman, and Tun 2654 to 2665).

Autonomic mechanisms provide an additional bridge without inviting contested narratives. The inflammatory reflex places vagal pathways within a circuit that modulates cytokine production, which links reception and safety to changes in inflammatory tone through plausible routes. Cardiac vagal indices, read cautiously and in context, can serve as coarse signatures of flexible control that tend to improve when social uncertainty falls. The chapter does not elevate these measures to essence, nor does it import claims that remain disputed at the level of anatomy and development. It uses them as convergent physiological channels through which social statistics can lower or raise the marginal cost of staying present, in alignment with the caution already stated in earlier chapters (Tracey 853 to 859).

The conceptual consequences can now be made precise. First, social safety shifts the background reward landscape that Chapter 4 tied to vigor. When a world answers back, the average return to action rises, time becomes expensive to waste, and response speeds and willingness to engage increase without any additional fuel. Second, social safety reduces predicted bodily costs by lowering high gain control of autonomic and immune systems that would otherwise be deployed against expected threat, which raises the expected value of control by shrinking the price term. Third, social safety changes precision assignments in a way that allows benign variation in interoception to be interpreted as tolerable rather than as evidence of impending failure, which further lowers price and widens the range of actions that feel possible. These three changes are not optional context. They are constitutive determinants of the felt economy that the dissertation seeks to explain.

Two boundaries keep the argument honest. There exist cases where the biology of cost is set by primary disease processes that are not downstream of social statistics. In such cases social safety can do less than the present account describes, and the theory must defer to specialized literatures on neurometabolic failure, degenerative disease, or endocrine catastrophe. There also exist contexts in which social inclusion is offered in forms that demand self abandonment or denial of testimony. Such inclusion does not lower price in the sense respected here. It shifts cost to a different register and is often experienced as exhaustion by other means. The theory therefore treats dignity as a constraint on what counts as safety. Permission to be a person requires reception without erasure.

The chapter anticipates misuse. One might attempt to convert these claims into instruments for extracting more labor by simulating safety without changing the conditions that create hazard. The theory forbids such use on its own terms. If safety cues are not credible, precision will shift toward threat, mediators will run hot, and the price of presence will remain high. A rhetoric of inclusion without a practice of reception will not restore energy. The decision variable in Chapter 4 will continue to fall, and persons will continue to conserve. The account therefore aligns descriptive truth with ethical requirement. Worlds that are just are also cheaper to inhabit. Worlds that are unjust are expensive even for those who survive them.

The handoff to the next chapter is natural. If safety and belonging alter the priors and prices that govern the next hour, then learning is the route by which those priors change and remain changed once conditions genuinely shift. The next chapter asks how predictions that once protected can be revised without erasing history, why inhibitory learning is the mechanism of record for durable change, how reconsolidation based updates may amplify revision under strict boundaries, and how dissociation requires different sequencing so that access to interoceptive life becomes tolerable before expectancy violations can matter. The aim is continuous with the aim of this chapter. To keep persons visible while stating exactly how a life becomes affordable again and how the world’s answer becomes part of the body that must carry that life forward (Slavich 265 to 295; Cole 31 to 37; Eisenberger, Lieberman, and Williams 290 to 292; Almeida 64 to 68; Stawski, Almeida, Lachman, and Tun 2654 to 2665; Tracey 853 to 859).

Chapter 6. Learning, prediction, and the durable revision of costly priors

A predictive organism that budgets its future must also learn which worlds are likely and which can be relinquished. The question for a theory of mental energy is not whether memory stores events. The question is how learned expectations about danger and effort change the felt price of remaining available and how those expectations can be revised without erasing a life. The answer given here keeps three claims in view. Threat learning is real and it is not the same thing as conscious fear. Durable change in threat governed behavior is best understood as inhibitory learning that limits the retrieval and expression of earlier associations rather than as deletion. Under restricted conditions, brief reminders can render consolidated traces labile and allow reconsolidation based updating, yet boundary conditions and mixed demonstrations in humans require caution. The chapter binds these claims to the dissertation’s prior argument by showing how each learning process alters priors, precision, and expected value of control and thereby changes what an hour will cost (LeDoux 155 to 184; Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23; Monfils, Cowansage, Klann, and Ledoux 951 to 955; Schiller, Monfils, Raio, Johnson, Ledoux, and Phelps 49 to 53).

The starting point is survival learning in circuits that do not require language. Joseph LeDoux’s program separated the detection of threat and the rapid preparation for defense from the conscious experience that persons call fear. Signals that predict harm can come to trigger autonomic, endocrine, and behavioral mobilization through pathways that reach the amygdala and its extended partners at speeds and in formats that do not wait for narrative explanation, while higher cortical systems construct feelings, interpretations, and plans around those preparations on a slower clock. This separation allows the theory to dignify first person reports without confusing them with the processes that set bodily cost. It also prevents a common error in clinical and cultural discourse in which a failure to feel safe is treated as the sole locus of the problem when the dominant work is being done in systems that have learned to expect harm and that mobilize accordingly regardless of what a person says to themselves in a quiet room. When those systems predict hazard the body runs hot, precision weights tilt toward threat, and the expected value of control falls because the price term inflates even before a decision is consciously taken to lean in or to withdraw. Any adequate account must therefore locate change at the level of retrieval and expression of learned associations and not only at the level of articulated belief (LeDoux 155 to 184).

Modern exposure science gives the most coherent explanation of how such change is earned. Michelle Craske and colleagues reframed exposure from an exercise in within session habituation to a program that maximizes inhibitory learning. The person enters a situation that their model predicts will be dangerous or intolerable. The situation unfolds in a way that violates those expectancies while the person remains engaged long enough to register the mismatch. New learning accrues that competes with and limits the retrieval of the older association. The old trace is not erased. It is fenced. The practical signatures of inhibitory learning follow directly from this mechanism. Between session return of anxiety can occur because the original association still exists and is available to retrieval cues that differ from the learning context. Generalization is earned by varying context and cues so that the new association wins when the old one is cued across places and times. Retrieval is supported by reminders that call the new learning to mind at the point of need. The theory therefore predicts why a person can experience genuine relief in a session yet find themselves haunted on a bus a week later and why change becomes durable when expectancy violations are large, contexts are varied, and reminders are correctly used at retrieval rather than treated as talismans. The conceptual payoff for the present dissertation is immediate. When inhibitory learning succeeds the predicted bodily cost of staying present falls in those situations because the model no longer assigns high precision to catastrophe, and the expected value of control rises because the price term is smaller and the probability of success is larger. When inhibitory learning fails or is not attempted, nothing in the value calculation changes, and exhortation to try harder will predictably fail because the organism is right about the cost given what it now expects (Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23).

The endurance of change matters as much as its mechanism. A long literature on relapse phenomena shows that extinguished responses can return with time, with context change, or with reminders of the original unconditioned stimulus. This observation used to be treated as evidence that extinction is weak. In the inhibitory learning frame it becomes an expected feature of a system that protects itself against premature erasure. The organism retains access to earlier associations so that it can re deploy them if the world turns hostile. What looks like relapse can therefore be read as the system doing what it evolved to do under ambiguity. The task for a humane theory is not to accuse the person of backsliding but to ask what retrieval cues and what forms of variability will make the newer, cheaper model available at the right moments so that price stays low in the wild and not only in the room. That question returns the discussion to expected value. If the new learning does not win when the world is noisy, the expected value of control will remain low, and the person will conserve energy not because their will is defective but because the cost of pretending that the world has changed when retrieval insists that it has not is indeed too high to pay without help.

Reconsolidation based updating promises a more radical route by which older traces can be modified rather than fenced, yet the promise comes with strict conditions. Monfils and colleagues showed in rodents that a brief reminder could render a fear memory labile and that extinction training applied within a limited time window after that reminder could produce persistent attenuation of fear responses, while Schiller and colleagues demonstrated conceptually similar effects in humans using an analogous retrieval extinction sequence. The excitement is intelligible. If older priors about the danger of the body or the world could be updated directly when briefly opened to revision, the cost of staying present might fall in a way that does not require continual competitive retrieval. Yet the boundary conditions are tight and human replications are mixed. Age and strength of the memory, timing and duration of the reminder, the presence of prediction error during retrieval, and individual differences in defensive style all appear to matter for whether a trace becomes labile and whether the subsequent learning writes into it. The prudent stance is to treat reconsolidation as a real phenomenon that can sometimes be recruited but not as a general method. That stance preserves the theoretical integrity of inhibitory learning as the central explanation for durable change and protects persons from interventions that promise erasure where no such editing is likely to occur. It also keeps the dissertation honest about falsifiable claims. If robust, well controlled studies using appropriate boundary conditions do not show labile window effects in humans beyond what is predicted by strong inhibitory learning, then reconsolidation will return to a special case in a book about prediction and control rather than holding center stage as a mechanism of ordinary change (Monfils, Cowansage, Klann, and Ledoux 951 to 955; Schiller, Monfils, Raio, Johnson, Ledoux, and Phelps 49 to 53).

Trauma related heterogeneity, particularly dissociative presentations, requires separate treatment to prevent theory from flattening lives into a single story of arousal. Ruth Lanius and colleagues have argued that the dissociative subtype of posttraumatic stress involves overmodulation of limbic activity and altered network dynamics, including disturbed interactions among salience, default mode, and executive networks. Persons with such presentations often report distance from interoceptive life rather than bombardment by it. In an interoceptive inference frame this looks like down weighted precision on bodily channels and gating between sensation and appraisal. In that state the attempt to stage expectancy violations against feared sensations or contexts will fail to propagate because the signals that carry the violation are not granted access to the model at the right level. The humane and exact implication is that learning must often be sequenced differently. Safe access to gentle internal cues must be restored and made tolerable before retrieval competitions about external contexts will matter, and cognitive work must be aligned with the task of allowing the body to be felt without annihilation. Only then will inhibitory learning about the world bring down the price of presence because only then will the organism have a model in which bodily life is again part of what must be predicted and governed rather than a source of noise to be ignored. A theory that ignores this sequence will blame persons for failures that are in fact correct predictions given the way their models were shaped to preserve continuity under assault (Lanius, Brand, Vermetten, Frewen, and Spiegel 701 to 708; Brewin, Dalgleish, and Joseph 670 to 686).

The learning rules that secure durable change can be re read now in the language of the dissertation without becoming procedural. Expectancy violation is not an abstract demand. It is the necessity for the generative model to encounter evidence that contradicts its high precision predictions about danger and cost. Variability of context is not a technique. It is the means by which the new model is made available across retrieval cues so that the expected value of control rises in the places where a life is actually lived and not only in the room in which the model was first revised. Retrieval cues are not charms. They are designed reminders that bias the competition in favor of the new model when old cues attempt to call the earlier association to the front. These translations matter because they keep the center of gravity on prediction, price, and value rather than on rituals that can be mechanically reproduced without altering what the organism expects.

Sleep marks the temporal edge of learning and brings earlier chapters back into the frame. Consolidation and reconsolidation processes engage across sleep in ways that can either strengthen the guardedness of older associations or stabilize the accessibility of new ones. The two process model does not in itself declare how memory traces will move, yet it secures a minimal truth for the present theory. The same learning attempt will lower future price differently depending on whether it is embedded in a temporal architecture that supports consolidation of the desired model. A person who leaves an encounter in which expectancy was violated and then traverses a night of sleep that restores precision and reduces homeostatic pressure will face tomorrow with a different expected cost than the person who leaves the encounter and then pays a debt of wakefulness that inflates price. The point is not to prescribe routines. It is to respect time as a component of how priors change and how cost is computed at the next sunrise (Borbély 195 to 204; Borbély e13598).

Several critiques must be answered before the chapter closes. A first warns that inhibitory learning is an elegant story that can be post hoc fitted to any outcome because it accommodates relapse without having to admit failure. The reply is the same insistence on signatures that the dissertation has used throughout. If expectancy violation is small and context variability is absent, then durable change should not be claimed. If violation is large and variability is present yet the new learning does not win more often at follow up than in the room, then the account must be revised to include mechanisms that have not been honored here or to accept that earlier learning is less labile than the theory predicted. A second critique alleges that reconsolidation has been promised beyond what the literature can bear. The present stance guards against that error by making reconsolidation a boundary constrained special case and by letting inhibitory learning and interoceptive revision carry the explanatory load for most durable changes in perceived price. A third critique asks whether a value theoretic lens reduces the heroism of staying present to a calculation. The reply is the same as in Chapter 4. Value is what matters to a person. When a parent decides to continue because a child’s safety is at stake, the expected value of control is high because love is a large term, not because cost is small. The language of expected value re dignifies such choices by placing their weight where it belongs rather than in a vaporous language of fuel and will.

The gains for the whole dissertation can now be stated precisely. Learning rules that earn durable change alter priors about harm and about the body and they do so in a way that reduces predicted cost and raises the expected value of control. When change is earned the person reports that energy has returned not because they have been filled but because what they expect from the next hour has become kinder and more accurate. When change is not earned the person reports that energy is gone because the world continues to ask for high gain control or because the model remains too expensive to maintain even when the world has improved. The theory invites humility. There are lives in which the world remains dangerous and in which conservation is wisdom. There are lives in which worlds have improved and models will catch up only under very specific conditions and with time. There are lives in which biology sets cost for reasons outside prediction. By marking these differences while refusing to abandon a unified account of prediction, value, and time, the chapter keeps the project exact and humane.

The handoff to the next movement is now prepared. If learning changes what the next hour will cost, then time governs when the same work is cheap or expensive to attempt and sleep governs whether what is learned is kept in a form that helps tomorrow. The next chapter honors that temporal law and returns to the dissertation’s promise to make a life affordable again without mythology, by showing how circadian phase and homeostatic pressure gate price and why the same effort is well spent at one hour and wasteful at another. The line that runs from phenomenology through predictive regulation, interoception, value, social safety, and learning meets the clock at last, and the clock answers in the same voice. Lives become possible in time kept well. Lives become too expensive to live in time kept badly. The theory earns its keep only if it can continue to say exactly why.

Chapter 7. Time kept well, sleep and circadian gating as regulators of cost

The same work is sometimes cheap and sometimes ruinous, and the price difference is time. A predictive organism does not move through a neutral hour. It moves through a temporal field in which the cost of keeping predicted and sensed aligned is lowered or raised by homeostatic sleep pressure and by circadian phase. If mental energy is the estimate of the expected value of control over the near horizon, then time alters both terms in that estimate at once. It changes predicted bodily cost by determining how much gain a controller must apply to hold regulation together, and it changes expected payoff because vigilance, working memory, and inhibitory control rise and fall with circadian phase in ways that alter the probability that control will succeed. The promise of this chapter is modest and decisive. Sleep homeostasis and the circadian pacemaker are constitutive variables in the felt economy of effort. They are not procedural trivia or lifestyle ornaments. They are the clockwork against which an organism prices the next hour (Borbély 195 to 204; Borbély e13598).

The two-process model remains the most economical frame for these claims. Process S tracks the build up of sleep pressure with time awake and its dissipation during sleep. Process C is the circadian component generated by the suprachiasmatic pacemaker and expressed across physiology and behavior as a roughly twenty four hour rhythm that gates sleep propensity, alertness, temperature, endocrine release, and performance. The point for present purposes is not to arbitrate among refinements of the model. It is to state that the energetic price of staying present depends on where a person stands on S and C. High S increases the gain a controller must apply to keep the body within bounds. Unfavorable C reduces the baseline from which cognition and vigilance can be recruited without cost. Together they create windows in which control is expensive and likely to fail, and other windows in which control is cheap and likely to pay. A theory that ignores these windows will call a person weak on one page and heroic on the next for doing the same thing under different clocks. A theory that centers these windows can speak with precision about why an hour that was unaffordable on Tuesday night became affordable on Wednesday morning without any new substance being added to the person (Borbély 195 to 204; Borbély e13598).

Forced desynchrony experiments isolate these components and answer the obvious critique that temporal effects might reflect monotony, expectancy, or demand characteristics. When participants live on imposed days that are too long or too short for the circadian pacemaker to follow, sleep pressure and circadian phase can be varied independently while neurobehavioral performance is measured. The results are consistent. Lapses of attention, reaction time slowing, and errors in tasks that require sustained control follow both the time since awakening and circadian phase, with worst performance arriving when high homeostatic pressure coincides with the circadian trough and best performance when low pressure coincides with the circadian peak (Dijk and Czeisler 112 to 129). For this dissertation the lesson is direct. The expected value of control falls in the trough because predicted costs are high and the probability of success is low for reasons that have nothing to do with will, and it rises at the peak for reasons that do not require romance. A humane theory treats these variations as the law of the organism’s day rather than as excuses for flagging effort.

Chronic partial sleep restriction provides a complementary test of mechanism. When sleep is cut back across a week in otherwise healthy people, psychomotor vigilance, working memory, and executive control degrade in a dose dependent fashion, while subjective ratings of sleepiness plateau and dissociate from objective impairment. The dissociation matters. It shows that the organism can feel less impaired than it is, which counsels against trusting self report alone when the clock says cost is high. It also shows why days of curtailed sleep quietly increase the price of later hours. The controller must apply more gain to sustain the same behavior, and the expected payoff of control falls because errors and lapses become more likely in the cognitive systems that translate intention into action (Van Dongen, Maislin, Mullington, and Dinges 117 to 125). The interoceptive account in earlier chapters can now be read with more acuity. Fatigue after curtailed sleep is not mood alone. It is the organism’s forecast that matching predicted and sensed will require continued high gain corrections that are poorly compensated by success under the current cognitive state. The price signal is appropriate even when a person is motivated to deny it.

Circadian phase adds a second law to this economy. Alertness, inhibitory control, and working memory improve and degrade with the clock independent of sleep pressure. These changes alter the likelihood that control will succeed and hence alter expected value even when homeostatic debt is small. The practical meaning is familiar to anyone who has tried to do demanding work in the early circadian morning or to make careful decisions late at night. The theoretical meaning is that time can raise or lower the price of the same plan because the organism’s cognitive tools are themselves rhythmic. To speak of energy without the clock is therefore to speak without one of the principal determinants of whether a person is correct to persist or to conserve at a given hour (Dijk and Czeisler 112 to 129).

If time changes price so deeply, the question shifts from whether sleep matters to how sleep participates in revising priors and controlling cost. Slow wave rich sleep reduces homeostatic pressure and restores the dynamic range of systems that must respond to the next day’s demands. It also participates in the consolidation and integration of new learning, including the inhibitory learning that the previous chapter identified as the mechanism of record for durable change. This is not a claim that sleep erases fear or writes safety by itself. It is the more modest and defensible claim that sleep influences which traces are accessible and in what form when the organism meets the next day’s retrieval cues, and therefore influences whether the cheaper model wins when it needs to. A person who experiences expectancy violation today and then traverses a night of restorative sleep faces tomorrow with a different expected price than a person who experiences the same violation and then pays a night of debt. Without importing contested stories about stages and their dedicated psychological functions, it is safe to say that sleep is the temporal medium in which the organism’s model of what the next day will cost is stabilized or fragilized, and that this medium must be included if price is to be explained without sentiment (Borbély e13598).

Circadian misalignment adds a structural tax that illustrates how social worlds enter the time economy. Shift work, jet lag, and social jet lag place behavior at phases that oppose internal rhythms. The result is elevated homeostatic pressure at the wrong times, altered endocrine release, and impaired performance even when total sleep time is controlled. A theory that centers expected value of control will predict that persons living under misalignment rationally conserve and report fatigue because costs are high and success is compromised by design. The point is not to prescribe scheduling policy in a theory chapter. It is to fix the causality. When institutions impose misalignment, the organism pays. When institutions provide temporal scaffolds that align with bodies, the organism pays less. Energy rises because price falls, not because a person has become more virtuous (Dijk and Czeisler 112 to 129).

The interoceptive loop meets the clock at several points. High homeostatic pressure increases prediction error because ascending signals contain noise that cannot be efficiently suppressed. Under such conditions the system often assigns high precision to threat cues and low tolerance for benign visceral variance, which yields the familiar state of being tired and overinterpreting ordinary fluctuations as signs of failure. A humane and exact account will not accuse the person of catastrophizing in a vacuum. It will say that the clock made pessimism cheap by raising prediction error and lowering confidence in correction. The remedy in concept is simple. Reduce S, align C, and the same interoceptive variation can be interpreted as tolerable again because predictions can be kept in register without high gain and because the probability of success has risen. The decision to invest control then follows without exhortation because expected value has improved.

A recurrent critique of temporal accounts is that they risk slipping into prescription and that they veer toward the managerial. The reply is to hold the line between theory and procedure. The claims advanced here are explanatory. They state why identical hours carry different prices and why organisms that refuse to continue at certain phases are often correct. They also bind the clock to the social and ethical claims of earlier chapters. Persons who live under schedules that preclude alignment are not failing a test of will. They are paying the biological tax their institutions have levied. Persons who discover that work once impossible becomes possible after sleep are not witnesses to a miracle. They are reporting the lawful fall in price that follows from lower S and a kinder C. A theory that cannot say these things cannot keep faith with lives that are organized by time whether or not those lives are allowed to notice it.

A second critique warns that sleepiness and circadian phase are often confounded with motivation and meaning, and that time will be used to ignore the person’s aims. The reply is to integrate rather than to displace meaning. The expected value of control is large only when the aim matters. The clock alters cost and probability of success; it does not tell anyone what to live for. The chapter therefore secures a division of labor. Persons remain sovereign over ends. Time tells the truth about when certain attempts are likely to be cheaper or dearer to pursue. A mature theory respects both truths at once.

There are limits to what time can do inside this theory. Where primary neurometabolic disease or degenerative processes set price, the relief that follows from alignment will be partial. Where humiliation and hazard remain high, the relief that follows from alignment will be temporary because the social term in the price equation remains unfavorable. Where dissociation down weights interoceptive channels, the relief that follows from alignment may arrive later because the organism must first relearn that bodily signals are tolerable. Naming these limits protects theory from panacea and protects persons from being told that a night of sleep will cure a life organized against them.

Three precise consequences flow from the chapter’s argument and will be carried into later synthesis. First, any account of energy fluctuations that does not include S and C is underspecified. Observed variance in vitality across days and hours must be modeled as a function of homeostatic pressure and circadian phase before being attributed to traits or to private will. Second, adjudication of the dissertation’s central claims must condition on time. If manipulations of social safety or interoceptive prediction are tested at unfavorable phases, null results cannot be cleanly interpreted. If the same manipulations succeed at favorable phases, claims of mechanism cannot be made universal. Third, the ethics of asking for effort must respect the clock. To call for maximal control at the circadian trough under high pressure is to ask for waste and to court error that will be wrongly blamed on character or commitment. To arrange demands when S is low and C is kind is to reduce avoidable price, and the organism’s report that energy has returned will be the honest correlate of a world that has stopped making the body expensive to inhabit.

Time does not compete with the earlier chapters. It completes them. Phenomenology taught that capacity is a way of being in a world rather than a number. Allostasis taught that regulation is anticipatory and that running hot exacts a price. Interoceptive inference taught that feeling is prediction about the body and its confidence. Value theory taught that control is invested when it is expected to pay. Social safety taught that belonging and humiliation set biological terms for cost. Learning taught that priors can be revised and made durable. Time now teaches that all of this is gated by S and C in every ordinary day. The final synthesis will carry these pieces together into a single statement of what makes an hour affordable and how a life that has become too expensive can be made less so without mythology and without coercion (Borbély 195 to 204; Borbély e13598; Dijk and Czeisler 112 to 129; Van Dongen, Maislin, Mullington, and Dinges 117 to 125).

Chapter 8. A unified conceptual model of control energy under uncertainty

The previous chapters have laid the pieces. The present chapter binds them into a single construct that can be stated in ordinary language without losing rigor. The claim is that perceived energy at a given moment is the organism’s estimate of the expected value of continuing to exert control over the near future, given what it predicts about bodily costs, background reward, social safety, learned threat, and time. The estimate is not mystical or private. It is a public object in the sense that each of its determinants can be named, constrained by evidence, and moved by conditions that other observers can verify. The determinants are predictive load that sets how hard regulation must work to keep life within bounds, the quality and confidence of interoceptive models that forecast the next hour inside the skin, the social safety prior that tunes vigilance and inflammatory tone, the valuation system that prices control under a background rate of return, the learned structure of danger and relief that governs which predictions are retrieved, and the temporal alignment of sleep pressure and circadian phase that gates both cost and probability of success. Energy is the felt summary of this joint problem. When the joint problem looks affordable, the organism is correct to report vitality. When it looks overpriced, the organism is correct to report fatigue. The task of theory is to specify each determinant, explain their interactions, and state what would count as a contradiction of the claim (McEwen and Stellar 2093 to 2101; Sterling 5 to 15; Barrett and Simmons 419 to 429; Critchley, Wiens, Rotshtein, Ohman, and Dolan 189 to 195; Shenhav, Botvinick, and Cohen 217 to 240; Niv, Daw, Joel, and Dayan 507 to 520; Slavich 265 to 295; Cole 31 to 37; Borbély 195 to 204; Dijk and Czeisler 112 to 129; Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23).

Predictive load names the work required to keep the organism within viable bounds while the near future arrives. In the allostasis frame, volatile niches, uncertain contingencies, and threat priors push controllers to run at higher gain. The price of running hot accumulates as allostatic load across cardiovascular, metabolic, immune, and sleep axes and as a slow narrowing of what feels livable. Felt energy tracks this budgeting problem because the body does not wait for errors. It anticipates and prices them. On days when the world is credibly predictable, mediator settings can be economical and the felt price of presence falls. On days when volatility is high, mediator settings rise to maintain order and the felt price rises with them. The same reserves can therefore feel spendable in one room and unavailable in another without contradiction because price is a function of model and niche rather than of a hidden reservoir (McEwen and Stellar 2093 to 2101; McEwen 33 to 44; Sterling 5 to 15).

Interoceptive models translate budgeting into feeling. Agranular visceromotor cortices send predictions about internal state, ascending visceral signals constrain them, and the comparison yields autonomic and endocrine action along with affect. Precision weights determine how much confidence is assigned to predictions and to error signals. In this loop, vitality is a judgment that predicted and sensed can be held in register at tolerable cost, while fatigue is a judgment that sustained alignment would exact an unacceptable price. The value of this account is not poetry. It is predictive discipline. If priors become kinder under credible safety and calibration, the same heart and gut will be felt as readiness rather than as omen and the price of control will fall for reasons that can be measured in both first person reports and autonomic deployment. If priors remain threat weighted, the same physiology will be felt as a reason to conserve. The language of spoons becomes sharp here. I have no spoons is the organism’s declaration that predicted alignment will be too costly to maintain through the next set of calls (Barrett and Simmons 419 to 429; Craig 655 to 666; Critchley, Wiens, Rotshtein, Ohman, and Dolan 189 to 195).

Valuation ties feeling to decision. The expected value of control account proposes that a supervisory system invests control when predicted benefits exceed predicted costs, taking into account the likelihood of success. Average reward theories of vigor add that when the background rate of return is poor, time becomes cheap to waste and agents rationally slow. When the background rate of return is rich, time becomes expensive to waste and agents rationally speed. The experience of energy follows these economies. In barren environments even small efforts feel extravagant because they displace other scarce options without promise of return. In responsive environments even large efforts feel affordable because the world answers back. This decisional frame rescues fatigue from moralism and restores dignity to refusal without denying agency. It also renders tractable predictions. If the expected payoff of engagement rises while bodily state is held constant, willingness to invest control and reported vitality should rise in step. If the predicted cost of exertion falls while payoff is held constant, the same outcomes should follow. If neither class of manipulation moves behavior or felt stance when measured with care, the model is wrong or incomplete (Shenhav, Botvinick, and Cohen 217 to 240; Niv, Daw, Joel, and Dayan 507 to 520).

Social safety is not a late modifier. It is a determinant. Cues of exclusion and status threat recruit regions that construct aversive bodily states and shift immune transcription toward inflammatory profiles. Precision assignments tilt toward threat, pain thresholds fall, and the cost of holding regulation together rises. Credible belonging reverses these settings and lowers price by both inferential and biological routes. The theory therefore expects that reliable reception will raise reported energy even when calories, sleep hours, and tasks are unchanged, because the near future has become cheaper to inhabit. It also expects that humiliation will tax energy out of proportion to visible workload, because the organism is paying in anticipation to survive a social world that does not promise to help (Eisenberger, Lieberman, and Williams 290 to 292; Slavich 265 to 295; Cole 31 to 37).

Learning determines whether priors about danger, body, and world remain costly. When expectancy violations are large, contexts vary, and retrieval is supported, new associations compete with old ones and limit their expression in the places where a life is actually lived. The price of presence falls because the model no longer assigns high precision to catastrophe in those settings. Under restricted conditions, brief reminders can render consolidated traces labile and allow deeper updating. Human demonstrations are promising and bounded. The present synthesis therefore keeps inhibitory learning as the mechanism of record and treats reconsolidation as an amplifier when boundary conditions are respected. Dissociative presentations require a distinct route in which safe access to interoceptive life is restored before expectancy violations can propagate through the network. In each case learning is not a technique. It is the way a predictive system earns the right to price the next hour more kindly without lying to itself (Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23; Monfils, Cowansage, Klann, and Ledoux 951 to 955; Schiller, Monfils, Raio, Johnson, Ledoux, and Phelps 49 to 53; Lanius, Brand, Vermetten, Frewen, and Spiegel 701 to 708; Brewin, Dalgleish, and Joseph 670 to 686).

Time gates all of the above. High sleep pressure raises prediction error and the gain required to keep order. Unfavorable circadian phase reduces the baseline from which cognition and vigilance can be recruited. The same effort is therefore cheap at some hours and expensive at others. Forced desynchrony and partial restriction experiments show that lapses, slowing, and errors follow both time since awakening and the clock. Subjective ratings often underestimate impairment. The model does not read these results as lifestyle counsel. It reads them as the law of cost. When S is low and C is kind, expected value rises and energy returns without any magic. When S is high and C is hostile, expected value falls and conservation becomes wisdom regardless of intention (Borbély 195 to 204; Dijk and Czeisler 112 to 129; Van Dongen, Maislin, Mullington, and Dinges 117 to 125).

The interactions among determinants are not simple addition. They are conditional. A favorable background reward rate raises expected value only if predicted bodily cost is not prohibitive. Improved interoceptive calibration lowers cost only if social safety allows precision to be redistributed away from threat. Strong inhibitory learning reduces price only if time permits consolidation and retrieval in the wild. Temporal alignment lowers cost only if learned priors do not insist that benign sensations portend catastrophe. These conditionalities explain why interventions that look obvious on paper sometimes fail. They also protect persons from being blamed for outcomes that follow from interactions they did not choose. The model therefore insists that adjudication be conditioned on these relations rather than on any single lever that is assumed to operate in isolation.

A theory that unifies across levels attracts two families of critique. The first warns against an umbrella so large that it shields itself from evidence. The reply is to push testable claims into specific literatures and to use unifying principles to organize questions rather than to settle them by decree. The free energy principle remains in the background as a conceptual umbrella that links perception, action, and learning around expected surprise. It is not asked to bear empirical weight here, and it is not used to overwrite more specific accounts that are already answerable to data in the domains at issue. Where free energy talk adds clarity about why prediction and correction are joined, it is retained. Where it would function as a bell that rings in place of a mechanism, it is set aside in favor of interoceptive inference and valuation that can lose and therefore learn (Friston 127 to 138; Barrett and Simmons 419 to 429; Seth 565 to 573).

The second critique charges that any model that renders energy as expected value risks collapsing meaning into arithmetic and risks excusing injustice as rational choice in barren ecologies. The reply is ethical and technical at once. Value here is whatever matters to an agent. It includes moral, relational, and craft goods. Rendering effort as a priced choice restores honor to persistence when persons spend themselves for ends they endorse and restores respect to refusal when the terms on offer would waste the self. The model indicts barren ecologies rather than excusing them, because it states with precision how humiliation and unreliability raise cost and lower expected value. It thereby hands designers concrete targets for repair rather than a rhetoric for extracting more labor from unchanged worlds. The measure of seriousness is whether the model helps persons change the near future without lying to them about price.

The model can now be stated compactly and then restated in a way that invites contradiction. A person’s report of high energy means that, given current and expected demands, interoceptive predictions, background reward, safety, learned priors, and time, the expected value of exerting control is above threshold. A person’s report of low energy means the opposite. If a manipulation that credibly increases expected payoff produces no change in willingness to invest control or in reported vitality while costs and time are held constant, the valuation arm is too weak. If a manipulation that credibly lowers predicted cost through safety, interoceptive calibration, or alignment produces no such change while payoff is held constant, the interoceptive and allostatic arms are too weak. If an exposure sequence that strongly violates expectancies across varied contexts and with supported retrieval does not reduce the felt price of presence in those contexts at follow up, the learning arm is too weak. If none of these classes of manipulation move behavior or felt stance under reasonable measurement, the model is wrong. This is not theatrical humility. It is an operational definition of seriousness.

Several boundary conditions must be marked to protect against overreach. There exist presentations in which primary neurometabolic failure, endocrine catastrophe, or degenerative processes set the price of presence almost regardless of prediction and valuation. In such cases the present model defers to those literatures and limits its claims to the variance it can plausibly explain. There exist stimulant induced and manic states in which value, cost, and time are distorted by neuromodulatory settings that break the usual link between expected value and felt energy. In such cases reports of energy do not carry the same information about affordability and the model must be applied with caution. There exist worlds in which structural violence keeps safety low and background reward barren across years. In such cases conservation is wisdom and any model that treats refusal as pathology has left the terrain of truth.

The payoff of a unified model is not a sleek portrait of the mind. It is a language clear enough to be shared across philosophy, clinical science, social biology, and chronobiology without erasing the person who must live the next hour. It permits an account of why identical reserves support opposite choices in different rooms, why reliable reception is felt as energy, why time kept well changes price, and why learning sometimes sticks and sometimes does not. It also returns dignity to a familiar sentence. I cannot today is a precise hypothesis about near future affordability, not an admission of failure. The chapters that follow take up the ethical and civic consequences of speaking this way. If energy is a priced permission to continue, then dignity is the right to steward one’s predictive budget, and institutions are either scaffolds that lower avoidable costs or machines that tax a life. The theory earns its place only if it can keep those sentences true in rooms where the next hour has to be lived, and only if it can revise itself when those rooms teach otherwise (McEwen and Stellar 2093 to 2101; Sterling 5 to 15; Barrett and Simmons 419 to 429; Critchley, Wiens, Rotshtein, Ohman, and Dolan 189 to 195; Shenhav, Botvinick, and Cohen 217 to 240; Niv, Daw, Joel, and Dayan 507 to 520; Eisenberger, Lieberman, and Williams 290 to 292; Slavich 265 to 295; Cole 31 to 37; Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23; Monfils, Cowansage, Klann, and Ledoux 951 to 955; Schiller, Monfils, Raio, Johnson, Ledoux, and Phelps 49 to 53; Lanius, Brand, Vermetten, Frewen, and Spiegel 701 to 708; Borbély 195 to 204; Dijk and Czeisler 112 to 129; Van Dongen, Maislin, Mullington, and Dinges 117 to 125; Friston 127 to 138).

Chapter 9. Dignity, agency, and the ethics of energetic life

A theory of mental energy is not complete when it can predict variation in cost and choice. It is complete enough for human use only when it can say what is owed to persons who must live inside those costs and choices, and what is forbidden when the language of energy is placed in the hands of institutions that can raise or lower price. The claim defended in this chapter is spare and binding. Dignity consists in the person’s right to steward a predictive budget without coercion, and agency consists in the capacity to author one’s pattern of expenditure in light of ends that matter. The corollary is practical and moral. The report of low energy is presumptively intelligent testimony about unfavorable terms, not a confession of defect, and any environment that treats such testimony as weakness rather than as knowledge is doing epistemic and physiological harm at once. Jonas supplies the first premise by reminding us that organisms are centers of concern, continuously engaged in the work of remaining intact in the face of uncertainty; there is a basic normativity in the living that cannot be reduced to a machine’s equilibrium (Jonas 1 to 22). Nussbaum supplies the second by arguing that emotions are appraisals of value and threat in a life, not froth upon reason; they are intelligent commentaries on what matters and whether the world is answering back (Nussbaum 1 to 30). Together they forbid any account that would treat the person’s refusal to continue as merely a noise term to be optimized away. The refusal may be miscalibrated; it is never meaningless.

The dissertation’s earlier chapters supply the technical substrate upon which this ethical claim must rest. If perceived energy is the organism’s estimate of the expected value of staying present given predicted cost, background reward, social safety, priors, and time, then a report of low energy says that the expected value of control has fallen below threshold under the model the person now has and the world the person now inhabits. To treat that report as failure is to change neither the model nor the world while adding humiliation to the price term that was already too high. To treat it as testimony is to ask what has been learned, what the world is offering, how interoceptive precision has been distributed, and how the clock is taxing the controller. Because those variables are public objects, they can be altered without trespassing against agency. Because they bear directly on the person’s ability to continue, neglecting them is not neutral but an act that imposes cost in the name of efficiency. In this light, dignity becomes a constraint on explanation. Any use of the theory that converts predictive budgets into levers to extract more expenditure from unchanged conditions is a misuse on the theory’s own terms, because expected value has not risen and predicted cost has not fallen; the organism is merely being asked to spend into a loss while smiling (Shenhav, Botvinick, and Cohen 217 to 240; Niv, Daw, Joel, and Dayan 507 to 520; Slavich 265 to 295; Cole 31 to 37; Borbély 195 to 204).

The first ethical risk is paternalism disguised as benevolence. Once energy is rendered as expected value of control, it is tempting for designers, clinicians, or managers to declare that they know which ends should count for the agent and to manipulate price so that expenditure aligns with those ends. The corrective is conceptually simple and morally hard. Value in this theory is not external payoff. Value is whatever matters to the person. It includes moral, relational, craft, and spiritual goods, articulated or tacit. A humane application must therefore disclose the model being used to describe price and return, must invite the person to edit that model in light of ends the person endorses, and must accept refusal when proposed ends are alien. The alternative is moral injury cloaked in a calculus. Jonas is instructive here. Striving is not neutral push and pull. It is the organism’s own orientation toward what it takes to be good for it; to replace that orientation with a surrogate is to violate the phenomenon one pretends to protect (Jonas 1 to 22).

A second risk is testimonial injustice, the systematic downgrading of certain people’s reports about their own states. Because cultural and institutional power often decides whose fatigue is believed and whose is scorned, the same sentence I cannot today will lower price for some by eliciting reception and will raise price for others by eliciting suspicion. The theory has no safe harbor from this fact. It can only forbid its own use as a technology of disbelief. Reports are treated as first order data about stance that must be paired with convergent evidence from interoception, behavior, and time, not to override testimony but to understand what it is reporting. Where reports are dismissed because of identity or rank, the theory predicts physiological cost along the routes already named, which makes testimonial injustice not only a moral failure but a biological tax that the ignored will pay with narrowed repertoire tomorrow (Barrett and Simmons 419 to 429; Garfinkel, Seth, Barrett, Suzuki, and Critchley 65 to 74; Eisenberger, Lieberman, and Williams 290 to 292). To pretend neutrality in such settings is itself a stance with a price.

A third risk is technocratic conversion, in which measures intended as proxies for cost become targets that replace the person. Because indices like heart rate variability, cortisol slope, or inflammatory transcription are legible and aggregate, institutions will be tempted to optimize them without reference to the person’s aims. The theory denies them that permission. Physiological signals are convergent evidence about the price of control only when they move with changes in prediction and valuation that the person can recognize as making the next hour more or less livable. Where signals are moved by manipulation that leaves value alien and cost subjectively high, the apparent gains are counterfeit. The body can be quieted by suppression without any restoration of dignity. The respectful demand is that proxy changes be accompanied by reports that life has become cheaper to inhabit in the domains that the person names as constitutive of their good. Anything less is a category error dressed in significance (McEwen 33 to 44; Tracey 853 to 859).

A fourth risk is distributive blindness. Background reward rates and credible safety are not randomly assigned; they are socially produced. If vigor rises where returns are rich and falls where returns are barren, then just worlds will be worlds in which returns to action are broadly available and humiliation is rare. To explain conservation in barren ecologies as a rational decision is not to excuse injustice. It is to describe the path by which injustice becomes physiology and then becomes narrowed life. The ethical upshot is that institutional design belongs inside a theory chapter, not to prescribe policy but to fix causality. A classroom that answers students’ bids with reception rather than with contempt lowers price and raises expected value, which changes presence. A clinic that treats reports of cost as knowledge rather than as resistance lowers price and raises expected value, which changes adherence. A workplace that protects time that aligns with bodies lowers price and raises expected value, which changes what people can do without harm. Each of these sentences is a restatement of the model in a civic key; none introduces a new principle beyond the insistence that justice is cheaper to inhabit than domination (Slavich 265 to 295; Dijk and Czeisler 112 to 129; Van Dongen, Maislin, Mullington, and Dinges 117 to 125).

The chapter must also answer a charge sometimes raised against humane theories, namely that they sanctify refusal and thereby paralyze action that a life requires. The reply is to separate endorsement from avoidance. In the present account, courage is the clear-eyed choice to invest control for ends that matter even when the price is high; it is not the denial of price. To tell the truth about cost is to honor courage when it is exercised and to refuse shaming when refusal is wise. James remains a guide. Hidden reserves appear not by exhortation but when the person believes expenditure will be met by a world that confers permission to try and by companions who will help bear the cost. The model holds this without romance. When expected value rises through credible reception and intelligible purpose, persons discover that energy has returned because the world has made it rational to spend (James 113 to 146).

An ethics of measurement follows naturally. Because feelings are predictions about the body and its near future, and because those predictions are susceptible to learning and to the clock, measures must be embedded in time and context to avoid doing violence to what they name. The demand is not procedure; it is respect. Ask for reports in relation to the next span of demand, not as free-floating numbers. Pair those reports with interoceptive accuracy and metacognitive insight so that confidence can be distinguished from calibration. Read autonomic and inflammatory indices only when they move with changes in prediction and valuation that the person recognizes as making life cheaper to inhabit. Refuse summaries that erase outliers when the outliers are those whom the world has made expensive to be. The theory thus requires an ethics of representation in its own practice, because an organism that is misdescribed is an organism whose predictive budget is being rewritten without consent.

Finally, the model must state its own limits in ethical terms rather than only in technical ones. There exist lives in which price is set by neurometabolic failure, degenerative disease, or endocrine catastrophe. In those lives the right response to reports of cost is compassion and specialized care, not a demand that prediction and valuation do work they cannot do. There exist lives in which safety is structurally absent. In those lives conservation is wisdom and any language that shames it is a collaborator with injustice. There exist lives in which dissociation has kept a self intact at the price of distance from bodily life. In those lives the restoration of contact is a sacred and precarious act; it must not be demanded at speed nor celebrated as a victory when the person’s sense of continuity is imperiled. By speaking these limits in the voice of the theory, the dissertation guards against hubris and protects persons from being made to pay with their bodies for the mistakes of those who speak for science.

The chapter closes by binding ethics back to prediction. To honor dignity is to lower avoidable price by making safety credible, by aligning time with bodies, by teaching models to expect what the world can now deliver, and by raising the background return to engagement in domains the person names as good. To honor agency is to refuse to substitute one’s own ends for the person’s, to disclose the model being used, and to accept revision when the person says what matters. To honor truth is to keep every claim answerable to evidence and to lived testimony together, because either alone can mislead. The final chapter will make adjudication explicit and will say how the theory must lose when the world says it is wrong, not to perform humility but to remain a servant of lives that are trying to keep faith with themselves under demand.

Chapter 10. Adjudication, limits, and the path to revision

A theory earns its way not by eloquence but by specifying what would count as its mistake. The present account claims that perceived energy is the organism’s estimate of the expected value of continuing to exert control over the near horizon given predicted bodily costs, background reward, social safety, learned threat, and time. To adjudicate that claim we must name the core constructs with enough discipline that rival accounts can meet them on common ground, state the decisive contrasts that would force revision, and mark boundary conditions that protect against overreach without turning the model into a self-sealing umbrella. The chapter proceeds in that order and returns repeatedly to two nonnegotiables the dissertation has kept from the start. First, persons remain the unit of analysis. Second, the theory must be revisable by evidence and by lived testimony together, because either alone can mislead.

Begin with construct clarity. Expected value of control is a decision variable that integrates predicted payoff, predicted cost, and efficacy for the act of adding control. The claim is not that there exists a cortical sovereign that “chooses” in the philosophical sense, nor that any single region suffices as its locus. The claim is that some neural machinery integrates these terms in a way that biases allocation of control, and that this integration is reflected in behavior that can be observed without knowing neural details. The dorsal anterior cingulate is a plausible contributor among others; it need not be an oracle for the argument to stand (Shenhav, Botvinick, and Cohen 217 to 240). Predicted cost is not an abstract penalty. It is the price implied by interoceptive models of how difficult it will be to keep predicted and sensed aligned while the near future arrives. That price is higher when allostatic demand is high, when precision is tilted toward threat, and when mediator settings must run at higher gain to maintain order; it is lower when credible safety, temporal alignment, and calibration make low-gain regulation adequate (McEwen and Stellar 2093 to 2101; Barrett and Simmons 419 to 429; Sterling 5 to 15). Background reward is not a market metaphor. It is the organism’s learned rate of return to action over the recent past. When that rate is rich, time becomes expensive to waste and response vigor rises; when it is barren, time becomes cheap to waste and response vigor falls (Niv, Daw, Joel, and Dayan 507 to 520). Social safety is not sentiment. It is a biologically mediated prior about reception and exclusion that alters inflammatory tone, vigilance, and pain processing in ways that change the cost term even when material resources are momentarily unchanged (Eisenberger, Lieberman, and Williams 290 to 292; Cole 31 to 37; Slavich 265 to 295). Learned threat names the structure of retrieval and expression that modern exposure science has reinterpreted as inhibitory learning, with reconsolidation-based updating as a boundary-conditioned special case (Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23; Monfils, Cowansage, Klann, and Ledoux 951 to 955; Schiller, Monfils, Raio, Johnson, Ledoux, and Phelps 49 to 53). Time is not lifestyle. It is the two-process gate on cost and probability of success through sleep pressure and circadian phase (Borbély 195 to 204; Dijk and Czeisler 112 to 129).

Against that scaffold we can now state decisive contrasts. The model predicts that manipulations that reliably increase expected payoff will increase willingness to invest control and reported vitality when predicted cost and time are held as constant as feasible. If, for instance, background reward is made richer in a way the agent values while interoceptive state and temporal phase remain steady, then response vigor and subjective energy should move up together; failure to observe such co-movement after careful measurement would weaken the valuation arm (Niv, Daw, Joel, and Dayan 507 to 520; Shenhav, Botvinick, and Cohen 217 to 240). The model predicts that manipulations that lower predicted cost will have the same effect when payoff is held steady. If credible social safety reduces inflammatory tone and threat precision, or if temporal alignment reduces sleep pressure during an otherwise identical task, then the felt price of remaining available should fall and willingness to invest control should rise; absence of such joint movement would weaken the interoception and allostasis arms (Cole 31 to 37; Borbély 195 to 204). The model predicts that when inhibitory learning is accomplished—large expectancy violations, varied contexts, supported retrieval—the price of presence will drop in those contexts at follow-up rather than only in the room; if expectancy is repeatedly violated under strong parameters and yet the cost of presence does not fall in the wild, then the learning arm is overgeneral (Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23). The model predicts that each determinant’s effect is conditional on the others. If background reward is raised in a body that is paying high homeostatic price at the circadian trough, the benefit to willingness to invest control will be smaller than the same manipulation at favorable phase; null effects under misalignment do not falsify valuation and must not be used to overrule time (Dijk and Czeisler 112 to 129; Van Dongen, Maislin, Mullington, and Dinges 117 to 125). These are not procedural recipes; they are constraints on interpretation. They say what movements should co-occur when a determinant is moved, and they say when null results indict a specific claim versus when a boundary condition has been violated.

Rival accounts merit engagement at equal altitude. Strong depletion models claim that a consumable resource (often imagined as glucose or a generalized willpower store) is the primary determinant of effort, and that declines in energy reflect literal depletion. The present theory honors metabolic cost inside predicted cost, yet it rejects depletion as the central grammar for ordinary variance. It predicts that identical caloric state can support opposite choices across contexts because price is a function of prediction, value, safety, and time. If, after careful control of those determinants, manipulations of caloric availability alone robustly and directly restore willingness to invest control and subjective vitality in ways that persist across contexts and times, then the resource thesis would demand incorporation; the weight of extant evidence aligns better with opportunity-cost and expected-value accounts (Kurzban, Duckworth, Kable, and Myers 661 to 679). Strongly phylogenetic autonomic hierarchies claim that vagal “states” explain most variance in social engagement and energy. The present theory treats autonomic signals as convergent channels that reflect and influence cost, not as essences. If stable, anatomy-consistent demonstrations show that specific vagal modes alone, independent of prediction and valuation, determine willingness to invest control and felt energy across contexts, then revision is due; at present, the safer reading retains the inflammatory reflex and cardiac vagal indices as partial signals and resists essentialism (Tracey 853 to 859). Grand unifications under the free energy principle sometimes promise elegant necessity in place of disciplined prediction. This dissertation keeps free energy as a conceptual umbrella—perception, action, and learning minimize expected surprise—while placing every falsifiable claim within specific literatures that can lose to data. If a free-energy instantiation makes an empirically novel prediction about cost, value, or time that outperforms the present synthesis while remaining testable, then the umbrella would do more work here; absent that, the specific accounts carry the load (Friston 127 to 138; Barrett and Simmons 419 to 429).

Measurement philosophy must match the theory to avoid adjudicating against a straw figure or for a mirage. First-person reports are disciplined data about stance; they cannot be the sole arbiters. They must be paired with behavioral indices of willingness to invest control under known payoff and cost parameters and with interoceptive measures that distinguish accuracy, sensibility, and metacognitive insight so that a movement in endorsement is not mistaken for a movement in calibration (Garfinkel, Seth, Barrett, Suzuki, and Critchley 65 to 74; Mehling et al. e48230; Mehling et al. e0208034). Physiological proxies must be read as convergent evidence only when they move with theoretically predicted changes—lower inflammatory tone and altered autonomic deployment when safety becomes credible, restored diurnal dynamics when sleep pressure falls—rather than as targets that replace the person (Cole 31 to 37; McEwen 33 to 44). Time is not a nuisance covariate. Forced desynchrony and chronic restriction literatures require that cost and efficacy be interpreted against sleep pressure and circadian phase; experiments that ignore S and C cannot adjudicate claims about valuation or interoception because they have left the price gate off the page (Dijk and Czeisler 112 to 129; Van Dongen, Maislin, Mullington, and Dinges 117 to 125). The phenomenological anchor remains in force: the explanandum is the lived judgment that the next hour is or is not affordable; metrics are servants of that voice, not replacements for it (Merleau Ponty 146 to 164; James 113 to 146).

Three families of results would force revision at scale. First, dissociations that cut the model’s joints. If increases in background reward, delivered through ends the agent values, repeatedly fail to increase willingness to invest control or reported vitality when predicted cost and time are steady, the link from value to vigor must be weakened or reframed (Niv, Daw, Joel, and Dayan 507 to 520). If credible reductions in predicted cost through safety and temporal alignment repeatedly fail to move either behavior or reports when payoff is steady, the tie from interoception and allostasis to price must be loosened (Barrett and Simmons 419 to 429; Borbély 195 to 204). If inhibitory-learning-optimal expectancy violations across varied contexts with supported retrieval repeatedly fail to reduce price at follow-up, then the learning arm is overstated and alternative mechanisms—habit architectures, model-based control failures, or neuromodulatory constraints—must be brought forward (Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23). Second, cross-level contradictions. If cardiac vagal indices and inflammatory signatures change in directions that should lower cost, but behavior and felt stance do not move correspondingly across well-specified contexts, then the reading of those signals inside predicted cost needs correction (Tracey 853 to 859; Cole 31 to 37). Third, adversarial counterexamples. If there exist stable subpopulations for whom energy tracks a consumable “resource” independent of prediction, valuation, safety, learning, and time—beyond cases of primary neurometabolic, endocrine, or degenerative disease—then the model’s scope statement must be narrowed to exclude them without euphemism.

The model’s limits are not defects to be concealed; they are coordinates that keep dignity in view. Mania and stimulant intoxication are states in which felt energy and willingness to invest control rise while cost and success probability are misestimated by neuromodulatory settings that break the usual link between expected value and stance; here the model cannot use reported energy as an index of affordability and must defer to disease frameworks that explain why the estimate itself is untrustworthy. Primary neurometabolic failure, endocrine catastrophe, and neurodegeneration can set price in ways that inference cannot cheaply revise; here the model explains variance only at the margins and must not be asked to do more. Structural hazard—poverty, discrimination, coercive institutions—keeps background reward barren and safety low across years; here conservation is wisdom and any application that pathologizes refusal colludes with injustice in the name of science (Slavich 265 to 295). Dissociative presentations are distinct failure modes in which precision on interoception is down weighted or gated from appraisal; here the sequence of change must restore safe access to bodily life before expectancy violations can propagate (Lanius, Brand, Vermetten, Frewen, and Spiegel 701 to 708; Brewin, Dalgleish, and Joseph 670 to 686). These limits are not apologias. They are the theory speaking in its own grammar about where it must stop and import other languages.

Finally, the path to revision. The theory invites adversarial tests precisely because it is a synthesis across levels that can drift into charisma if it is not constrained. Predetermine contrasts where the direction of change is fixed by the model; precommit to interpretations conditioned on time; publish nulls as amendments rather than as quiet omissions. Where evidence forces contraction, contract in print. Where rival accounts outperform the model on a domain it claims, import them without pride and redraw the map. Where lived testimony contradicts tidy inference, treat the contradiction as a signal that the model has missed a determinant or misweighted a term. The aspiration is not a closed system. It is a perched framework that can be knocked into better shape by contact with persons and with data, because the object of study is not an engine but a life that is constantly repriced by a world that answers back.

If the model survives such adversarial attention, it does so because it continues to make ordinary sentences precise without cruelty. I cannot today remains a hypothesis about expected value falling below threshold under current priors and clocks; I can again remains a hypothesis about price falling or return rising in a way the person recognizes as true. The theory’s utility is then moral as well as explanatory. It hands designers and companions a language for lowering avoidable cost without demanding cheerful expenditure into loss, and it hands persons a way to read their own fatigue as intelligence in need of a kinder world rather than as a defect to be conquered. Should some other account do these things more truthfully, the present one should yield. That is the last nonnegotiable: the work serves lives, not the other way around.

Conclusion. A humane science of energetic life

The dissertation set out to make an ordinary sentence exact without losing its human weight. I cannot today was treated as a hypothesis about the expected value of continuing to exert control over the near future given predicted bodily cost, background reward, social safety, learned threat, and time. The path to that sentence passed through a phenomenology that keeps the lived body as the first horizon of meaning, a biology of predictive regulation that budgets life in advance, an inferential account of interoception that turns feeling into an estimate with confidence, a decision theory of control that prices effort without belittling courage, a social biology that reads belonging and humiliation as first order determinants of cost, a learning science that explains durable revision without myth of deletion, and a chronobiology that restores the clock to its rightful place as a gate on price and success. The pieces were selected because each bears disciplined evidence and because together they explain why identical reserves can support opposite choices in different rooms, why reception is felt as energy, and why time kept well can make the same work cheap that yesterday was ruinous (Merleau Ponty 146 to 164; McEwen and Stellar 2093 to 2101; Barrett and Simmons 419 to 429; Shenhav, Botvinick, and Cohen 217 to 240; Eisenberger, Lieberman, and Williams 290 to 292; Craske, Treanor, Conway, Zbozinek, and Vervliet 10 to 23; Borbély 195 to 204).

The first contribution is conceptual. Energy was redefined as a situated inference rather than as a fluid. The definition is not rhetoric. It makes specific, testable contact with measures at three levels at once. First person reports are treated as disciplined data about stance toward an immediate horizon. Behavioral indices of control investment under known payoff and cost parameters register the same stance from the side of action. Physiological channels report the gain being applied to keep predicted and sensed aligned and the wear that accrues when high gain control is run across weeks. The three channels need not move together, and the theory predicts when they will part and when they will reunite. This multi level discipline protects the person from being reduced to an instrument and protects the science from being seduced by a single proxy. The same discipline answers the familiar worry about unfalsifiability. The unified model forecast co movement patterns that rival accounts do not forecast, and it stated in advance the findings that would force revision or contraction of scope. To repeat those standards here is to say that this work binds its fate to data and to lived testimony together, which is the only ethical posture when lives are the subject matter (Garfinkel, Seth, Barrett, Suzuki, and Critchley 65 to 74; McEwen 33 to 44).

The second contribution is integrative without being totalizing. The free energy principle was retained as a conceptual umbrella that links perception, action, and learning around expected surprise, yet every falsifiable claim was placed inside specific literatures that already admit loss to evidence. The gain from unification is not a triumphal slogan. The gain is a shared grammar that allows philosophy of embodiment, social genomics, decision theory, exposure science, and sleep science to speak about the same hour in the same person without erasing the differences in method that keep each field honest. The umbrella is used when it clarifies why prediction and correction belong together. It is set aside when a more specific mechanism carries the weight. This selectivity is not timidity. It is fidelity to the ambition of an account that wishes to remain useful across rooms rather than impressive in a single auditorium (Friston 127 to 138; Barrett and Simmons 419 to 429; Seth 565 to 573).

The third contribution is ethical and so is scientific at once. Dignity was defined as the right to steward a predictive budget without coercion, and agency as the capacity to author one’s pattern of expenditure in light of ends that matter. The theory explained why humiliation raises the price of presence by recruiting pain networks and inflammatory programs, why credible belonging lowers that price by changing both priors and mediators, and why exhortation in barren ecologies asks persons to spend into a loss. The ethical arc is therefore not an ornament at the end of a technical essay. It is an inference from the model’s primitives. To honor dignity is to lower avoidable cost by changing worlds and clocks in ways that the person recognizes as making life cheaper to inhabit, and to do so while keeping the person’s ends sovereign in the calculus of value. A theory that does not say these things in its own voice is not yet a humane science of energetic life, and a practice that uses these ideas to extract more work from unchanged conditions is a misuse on the theory’s own terms (Jonas 1 to 22; Nussbaum 1 to 30; Slavich 265 to 295; Cole 31 to 37).

Several predictable gaps were faced and turned into work to be done rather than into excuses for drift. Strong resource depletion accounts can be reconciled with the present view by locating metabolic limits inside the predicted cost term, yet a decisive challenge remains open. If manipulations of caloric availability alone, with prediction, valuation, safety, and time held steady, restore willingness to invest control and reported vitality in a durable and context general way, then the resource thesis would deserve a larger place than it now holds in this synthesis. Strong autonomic hierarchy claims can be respected at the level of partial signals while avoiding essence language that outruns anatomy and development. Here too an empirical door remains open. If specific vagal modes are shown to determine willingness to invest control and felt energy across contexts independent of prediction and valuation, then the present account would require revision. Strong reconsolidation promises are tempered by boundary conditions and mixed human demonstrations. If future work identifies reliable human windows and parameters that consistently write into older traces beyond what strong inhibitory learning already predicts, then the learning arm will be rewritten to give that mechanism more central space. These challenges do not threaten the project’s integrity. They expose it to correction, which is the only way the enterprise deserves to survive. The same stance governs limitations that are not gaps but edges that protect against harm. Primary neurometabolic failures, endocrine catastrophes, degenerative processes, and manic or stimulant states set price and distort valuation in ways that this model does not pretend to control. Structural hazard keeps safety low and background reward barren across years for many. In those lives conservation is wisdom and the right response is repair of the world, not a demand for expenditure that the numbers do not justify (Kurzban, Duckworth, Kable, and Myers 661 to 679; Tracey 853 to 859; Monfils, Cowansage, Klann, and Ledoux 951 to 955; Schiller, Monfils, Raio, Johnson, Ledoux, and Phelps 49 to 53).

The positive program that follows from these admissions is simple to state and hard to enact. Design for credible reception so that the social safety prior can move and the inflammatory tax can fall. Design for time kept well so that sleep pressure and circadian phase are allies rather than adversaries of effort. Teach models to expect what the world can now deliver using learning rules that maximize expectancy violation, vary contexts, and support retrieval in the wild, with dissociative presentations sequenced so that safe access to interoceptive life is restored before further tests are asked. Raise the background rate of honest return so that time becomes expensive to waste and vigor rises without romance. These lines are not prescriptions in a theory chapter. They are constraints on any application that claims fidelity to the argument. They are also a practical ethics, since each lowers price without asking a person to spend against the numbers. The discourse of energy is thereby returned to the person whose life is being priced. The sentence I can again becomes a statement about a world and a body that have become kinder to inhabit. The sentence I cannot today remains a statement about a world and a body that are charging too much, spoken by someone who has earned the right to be believed.

The final word belongs to phenomenology and to James, because the first and last test of any model of energetic life is whether it can still recognize a person. The body remains our general medium for having a world, which means that every claim has to return to a posture toward possibility that can be felt from the inside and recognized from the outside. Effort still reveals reserves when the world confers permission to try, and it is still an act of sense to conserve when the terms are wrong. The theory has offered an architecture that makes these sentences precise, that ties them to a biology and a decision rule, that keeps them answerable to adversarial evidence, and that refuses to let precision become cruelty. If later work shows that some other architecture does these things more truthfully, the present one should yield. Until then, it gives a clean account of why a life can become affordable again. Prediction becomes kinder and more accurate. The world answers with reception rather than humiliation. Time aligns so that required control is cheap and likely to succeed. Under those conditions the report that energy has returned is not a mood. It is the organism’s correct estimate that the next hour is worth the price, and that continuing to be a person in that hour is a good bet. That sentence is what the work exists to protect and to explain, which is as much philosophy as science and as much ethics as either.

Works Cited

Almeida, David M. “Resilience and Vulnerability to Daily Stressors Assessed via Diary Methods.” Current Directions in Psychological Science, vol. 14, no. 2, 2005, pp. 64 to 68, https://doi.org/10.1111/j.0963-7214.2005.00336.x.

Barrett, Lisa Feldman, and W. Kyle Simmons. “Interoceptive Predictions in the Brain.” Nature Reviews Neuroscience, vol. 16, no. 7, 2015, pp. 419 to 429, https://doi.org/10.1038/nrn3950.

Borbély, Alexander A. “A Two Process Model of Sleep Regulation.” Human Neurobiology, vol. 1, no. 3, 1982, pp. 195 to 204.

Borbély, Alexander A., et al. “The Two Process Model of Sleep Regulation Forty Years On: A Reappraisal.” Journal of Sleep Research, vol. 31, no. 4, 2022, e13598, https://doi.org/10.1111/jsr.13598.

Brewin, Chris R., et al. “A Dual Representation Theory of Posttraumatic Stress Disorder.” Psychological Review, vol. 103, no. 4, 1996, pp. 670 to 686, https://doi.org/10.1037/0033-295X.103.4.670.

Cole, Steven W. “Social Regulation of Human Gene Expression: Mechanisms and Implications for Public Health.” American Journal of Public Health, vol. 99, suppl. 1, 2009, pp. S31 to S37, https://doi.org/10.2105/AJPH.2008.159491.

Craig, A. D. “How Do You Feel? Interoception: The Sense of the Physiological Condition of the Body.” Nature Reviews Neuroscience, vol. 3, no. 8, 2002, pp. 655 to 666, https://doi.org/10.1038/nrn894.

Craske, Michelle G., et al. “Maximizing Exposure Therapy by Optimizing Inhibitory Learning.” Behaviour Research and Therapy, vol. 58, 2014, pp. 10 to 23, https://doi.org/10.1016/j.brat.2014.02.003.

Critchley, Hugo D., et al. “Neural Systems Supporting Interoceptive Awareness.” Nature Neuroscience, vol. 7, no. 2, 2004, pp. 189 to 195, https://doi.org/10.1038/nn1176.

Dijk, Derk Jan, and Charles A. Czeisler. “Contribution of the Circadian Pacemaker and the Sleep Homeostat to Sleep Propensity, Sleep Structure, Electroencephalographic Slow Waves, and Sleep Spindle Activity in Humans.” Journal of Sleep Research, vol. 4, suppl. 2, 1995, pp. 112 to 129, https://doi.org/10.1111/j.1365-2869.1995.tb00221.x.

Eisenberger, Naomi I., et al. “Does Rejection Hurt? An fMRI Study of Social Exclusion.” Science, vol. 302, no. 5643, 2003, pp. 290 to 292, https://doi.org/10.1126/science.1089134.

Friston, Karl. “The Free Energy Principle: A Unified Brain Theory.” Nature Reviews Neuroscience, vol. 11, no. 2, 2010, pp. 127 to 138, https://doi.org/10.1038/nrn2787.

Garfinkel, Sarah N., et al. “Knowing Your Own Heart: Distinguishing Interoceptive Accuracy from Interoceptive Awareness.” Biological Psychology, vol. 104, 2015, pp. 65 to 74, https://doi.org/10.1016/j.biopsycho.2014.11.004.

James, William. The Energies of Men. Moffat, Yard, 1911.

Jonas, Hans. The Phenomenon of Life: Toward a Philosophical Biology. Northwestern University Press, 2001.

Kurzban, Robert, et al. “An Opportunity Cost Model of Subjective Effort and Task Performance.” Behavioral and Brain Sciences, vol. 36, no. 6, 2013, pp. 661 to 679, https://doi.org/10.1017/S0140525X12003196.

Lanius, Ruth A., et al. “The Dissociative Subtype of Posttraumatic Stress Disorder.” Depression and Anxiety, vol. 29, no. 8, 2012, pp. 701 to 708, https://doi.org/10.1002/da.21889.

LeDoux, Joseph E. “Emotion Circuits in the Brain.” Annual Review of Neuroscience, vol. 23, 2000, pp. 155 to 184, https://doi.org/10.1146/annurev.neuro.23.1.155.

McEwen, Bruce S. “Stress, Adaptation, and Disease: Allostasis and Allostatic Load.” Annals of the New York Academy of Sciences, vol. 840, 1998, pp. 33 to 44, https://doi.org/10.1111/j.1749-6632.1998.tb09546.x.

McEwen, Bruce S., and Eliot Stellar. “Stress and the Individual: Mechanisms Leading to Disease.” Archives of Internal Medicine, vol. 153, no. 18, 1993, pp. 2093 to 2101, https://doi.org/10.1001/archinte.153.18.2093.

Mehling, Wolf E., et al. “The Multidimensional Assessment of Interoceptive Awareness MAIA.” PLOS ONE, vol. 7, no. 11, 2012, e48230, https://doi.org/10.1371/journal.pone.0048230.

Mehling, Wolf E., et al. “The Multidimensional Assessment of Interoceptive Awareness, Version 2 MAIA 2.” PLOS ONE, vol. 13, no. 12, 2018, e0208034, https://doi.org/10.1371/journal.pone.0208034.

Merleau-Ponty, Maurice. Phenomenology of Perception. Translated by Donald A. Landes, Routledge, 2012.

Monfils, Marie H., et al. “Extinction Reconsolidation Boundaries: Key to Persistent Attenuation of Fear Memories.” Science, vol. 324, no. 5929, 2009, pp. 951 to 955, https://doi.org/10.1126/science.1163064.

Niv, Yael, et al. “Tonic Dopamine: Opportunity Costs and the Control of Response Vigor.” Psychopharmacology, vol. 191, no. 3, 2007, pp. 507 to 520, https://doi.org/10.1007/s00213-006-0502-4.

Nussbaum, Martha C. Upheavals of Thought: The Intelligence of Emotions. Cambridge University Press, 2001.

Schiller, Daniella, et al. “Preventing the Return of Fear in Humans Using Reconsolidation Update Mechanisms.” Nature, vol. 463, no. 7277, 2010, pp. 49 to 53, https://doi.org/10.1038/nature08637.

Seth, Anil K. “Interoceptive Inference, Emotion, and the Embodied Self.” Trends in Cognitive Sciences, vol. 17, no. 11, 2013, pp. 565 to 573, https://doi.org/10.1016/j.tics.2013.09.007.

Shenhav, Amitai, et al. “The Expected Value of Control: An Integrative Theory of Anterior Cingulate Cortex Function.” Neuron, vol. 79, no. 2, 2013, pp. 217 to 240, https://doi.org/10.1016/j.neuron.2013.07.007.

Slavich, George M. “Social Safety Theory: A Biologically Based Evolutionary Perspective on Life Stress, Health, and Behavior.” Annual Review of Clinical Psychology, vol. 16, 2020, pp. 265 to 295, https://doi.org/10.1146/annurev-clinpsy-032816-045159.

Stawski, Robert S., et al. “Associations among Daily Stressors and Salivary Cortisol: Findings from the National Study of Daily Experiences.” Psychoneuroendocrinology, vol. 38, no. 11, 2013, pp. 2654 to 2665, https://doi.org/10.1016/j.psyneuen.2013.06.023.

Sterling, Peter. What Is Health? Allostasis and the Evolution of Human Design. MIT Press, 2020.

Tracey, Kevin J. “The Inflammatory Reflex.” Nature, vol. 420, no. 6917, 2002, pp. 853 to 859, https://doi.org/10.1038/420853a.

Van Dongen, Hans P. A., et al. “The Cumulative Cost of Additional Wakefulness: Dose-Response and Individual Differences in Neurobehavioral Functions and Sleep Physiology.” Sleep, vol. 26, no. 2, 2003, pp. 117 to 126, https://doi.org/10.1093/sleep/26.2.117.