Prologue: A Closure, A Body, A Remainder
Michelle Davison did not first encounter the State’s judgment as an argument. She encountered it as a seizure. According to the pleadings later recited by the United States Court of Appeals for the Sixth Circuit, she learned that a fraud determination existed only when the Internal Revenue Service notified her that her tax refunds were being intercepted, and the interceptions continued across multiple tax years. The operative institutional event had already occurred. A determination had been generated, penalties had attached, collection machinery had begun to move, and the documentary file could present itself as procedurally complete long before the person against whom it was acting possessed a usable account of what had happened or how to contest it (Cahoo, 2019, pp. 6 to 7). The order of experience matters. Institutional closure arrived first in the file. Exposure arrived later in the body, in time, in vigilance, in the forced labor of reconstructing an event from consequences.
That order is not incidental to the argument of this book. It is the mechanism. In the Michigan Unemployment Insurance Agency’s MiDAS period, as described in the same litigation record, a claimant could be flagged through discrepancy logic or income spreading routines, sent a multiple choice questionnaire that did not disclose the factual basis for suspicion, and then be auto adjudicated for fraud if the claimant failed to respond within ten days or answered in ways the system treated as inculpatory. The system sent questionnaires to online accounts, including dormant accounts, without taking additional steps in many cases to ensure the claimant actually knew the questionnaire existed. It then generated determinations and restitution demands, assessed severe penalties, and triggered collections pathways that included tax intercepts and wage garnishment. Appeal rights formally existed, but the record alleges that many people did not learn of the determination until after the ordinary appeal window had expired, while the agency’s help line was functionally inaccessible at scale (Cahoo, 2019, pp. 4 to 6). By the time the person enters the scene as a speaker, the institution has already spoken several times in forms, codes, and timelines. The individual is then asked to translate that speech into contestation at personal cost.
The point of beginning here is not to make the prologue carry the entire indictment. It is to prevent a recurrent analytic error at the threshold. When institutions are studied from the vantage of mandate, procedure, and output, closure appears first as a competence problem, a documentation problem, or a legality problem. Those are real problems. They are also incomplete descriptions. What they often obscure is the transfer event by which unresolved uncertainty, evidentiary insufficiency, classification error, or procedural compression leaves the institutional file in the form of declared closure and reappears in embodied life as unpaid labor, sleeplessness, reputational damage, interrupted speech, survival administration, and the long temporal afterlife of having to remain available to a system that has already declared itself done. In that sense, the question is not whether institutions close. They must close. The question is where the remainder of closure goes, who carries it, and under what authority that transfer is treated as rational.
The phrase remainder of closure names what persists after a decision has become actionable. It does not name everything unresolved in the world. It names what the closure event itself cannot absorb without loss, what it displaces, what it leaves to be borne elsewhere, and what may be either declared, distributed, contested, or hidden. In the MiDAS litigation record, one can see the elements of this grammar before the grammar is named. The State’s documentary outputs had legal and administrative force. The grounds of suspicion were often opaque to those targeted. Notice and timing were structured in ways that could make formal recourse metabolically unavailable. Penalties and collections moved ahead of usable explanation. The institution retained a vocabulary of determination while offloading the work of uncertainty resolution onto claimants who lacked synchronized access to the facts, the forum, and the clock (Cahoo, 2019, pp. 4 to 7). This is not a story about bad software replacing good governance. It is a story about an institutional grammar of closure that software intensified, accelerated, and rendered legible in a particularly unforgiving way.
It is also a story that became visible in law because persons and advocates forced it into reviewable form. The record that later emerges in appellate opinions and audits is already a second order event. It is a record produced after exposure, after nonresponse, after collections, after appeals were missed or denied, after lawyers, judges, reporters, auditors, and claimants themselves had already expended immense labor to convert lived asymmetry into something institutions recognize as admissible contradiction. The Michigan Office of the Auditor General’s 2016 performance audit on claimant services identified defects in claimant communications, including the need for the agency to provide claimants with facts and rationale when determining false or misleading information, and noted failures in processing undeliverable mail, both of which bear directly on notice, contestation, and the ability of persons to understand allegations in time to respond (Michigan Office of the Auditor General, pp. 1 to 2). Later judicial opinions, looking backward over the MiDAS period, would treat the risk of error as unusually high and the notice practices as constitutionally deficient in important respects. The Michigan Court of Appeals described a fraud determination that was “plainly deficient” for failing to identify underlying allegations, and emphasized that even multiple notices considered together did not necessarily provide reasonable notice of the grounds a claimant had to defend, especially in the automated setting (Scott v. Michigan Unemployment Insurance Agency, pp. 4 to 7). What appears in doctrine as notice insufficiency and error risk appears in the lives of affected persons as the custody question this book pursues.
The governing claim of this manuscript follows from that question and extends beyond any single administrative failure. Contemporary institutions increasingly secure practical authority by converting indeterminacy into defensible closure while externalizing unresolved remainder into embodied life. They do this under many legitimacy grammars and not always in bad faith. Courts must decide on incomplete records. Agencies must process at scale. Employers must classify. Hospitals must triage. Platforms must moderate. Insurers must adjudicate. Police, schools, benefits offices, licensing boards, and technical systems all operate under conditions in which waiting for complete certainty would itself constitute a form of harm or nonperformance. Closure is therefore a necessary institutional wager. This book does not begin by treating closure as deception. It begins by insisting that the legitimacy of closure cannot be evaluated only by the internal regularity of the procedure or the formal existence of a result. It must also be judged by custody of uncertainty, by burden distribution, by recourse conditions, and by the public reviewability of the grammar under which closure is justified, especially where closure imposes disproportionate bodily and temporal cost.
Public reviewability of grammar is part of the doctrine because institutions do not act under a single legitimacy language. A police order, a benefits denial, a regulatory emergency measure, and an algorithmic risk score may each be defended under distinct grammars of authority even when they produce similar forms of burden transfer. Some appeal to legality and rule fidelity, some to expertise and rational administration, some to democratic mandate, some to necessity, some to risk prevention, some to efficiency, some to technical reliability. These grammars do not only describe decisions after the fact. They shape what counts as sufficient evidence, what forms of explanation are owed, who is recognized as a party, what timelines are treated as reasonable, and which burdens are named as unfortunate side effects rather than adjudicative defects. If a doctrine of legitimacy is to assess closure across domains, it cannot accept institutional self description as dispositive. It must ask which grammar is actually operative and whether that grammar, as used in practice, licenses hidden uncertainty transfer under the cover of procedural completion. The problem of grammar identification enters here as a methodological demand before it appears later as doctrine.
I also need to state, at the beginning and without softening language, the methodological debt that governs this book. This manuscript is in debt to lines of exposure like Davison’s, and that debt limits what the theory may claim, abstract, or ignore. The debt is not sentimental homage to injury. It is a disciplinary relation. It means that conceptual elegance is not evidence of explanatory adequacy. It means that any argument in these pages that increases theoretical range by dulling the visibility of burden transfer is suspect. It means that the book cannot purchase coherence by converting lived asymmetry into neutral system language and then congratulating itself for analytical balance. It means that the scene at the beginning is not a threshold ornament that authorizes later abstraction. It is a test that the middle chapters must repeatedly fail, revise against, and reapproach. The theory is answerable to whether it clarifies the transfer event or conceals it.
That debt changes the structure of the book. The chapters that move most heavily through institutional doctrine and technical design are required to return functionally to the prologue’s instance and ask, in plain terms, whether their concepts sharpen or obscure the mechanism first shown. If they obscure it, the prose must revise. This is not a performance of humility. It is a writing constitution meant to resist a familiar failure of academic argument, which is to grow more sophisticated at the price of becoming less answerable to what gave rise to the inquiry. Several chapters later, when the manuscript reaches the problem of contestability climate, distributed exclusion, and perimeter production, the risk of atmospheric vocabulary without accountability becomes especially high. The return to the opening instance is therefore not a stylistic callback. It is a debt test. The same holds when the manuscript reaches documentary due process in generative environments. If the synthetic frame displaces the person into an abstraction called “the user” or “the affected party” without preserving the embodied temporality of exposure, then the analysis will have reproduced the transfer it claims to diagnose.
The book’s subtitle names the field of this debt. Embodied remainder is the lived carriage of unresolved closure. Institutional closure is the production of actionable endings under declared or concealed insufficiency. The legitimacy crisis of the synthetic age is not reducible to misinformation, model bias, or automation error, though all of those matter. The deeper crisis is that synthetic systems lower the marginal cost of producing closure outputs while leaving contradiction, verification, and attributable custody comparatively expensive. When text, image, record, summary, recommendation, and rationale can be generated at scale, institutions can appear increasingly articulate while becoming less attributable. They can produce more explanation in form while preserving less responsibility in substance. They can satisfy documentary appetite without improving evidentiary custody. They can simulate responsiveness while increasing the metabolic price of contestation. This book argues that, under these conditions, legitimacy doctrine must become more attentive to custody rather than less, because the abundance of documentary closure can itself function as a concealment mechanism.
I do not claim exhaustiveness. Even at the level of legitimacy traditions, the architecture of this book is selective and openly so. The chapters that stage conflict among Arendtian, Fullerian, Weberian, and administrative grammars do not settle the field and do not treat excluded traditions as absent because they are irrelevant. The omission of Indigenous governance frameworks, Islamic administrative traditions, and postcolonial critiques from the initial adjudicative staging is a real loss and not a neutral simplification. Part of the argument of this book is that declared insufficiency is more legitimate than concealed insufficiency, and that principle binds the book’s own composition. I state the perimeter here so the reader can evaluate the doctrine that follows with its exclusions named rather than hidden inside universal language. Later chapters will also show why grammar identification itself can reproduce historical exclusions when the available archive has already been shaped by classificatory regimes that treat some traditions as background rather than as living sources of legitimacy.
The formal architecture is therefore governed counterpoint rather than linear proof. Institutional chapters move toward provisional closure under declared insufficiency. Embodied and ontological chapters preserve remainder where closure cannot honestly absorb it. Movement I defines terms and stages legitimacy conflict without forced synthesis. Movement II follows the transfer event into embodied burden, administrative settlement, documentary due process, and contestability climate. Movement III turns to mismatch and perimeter, where the outside of custody is produced not only by intentional exclusions but also by cumulative sequences of separately defensible decisions. Movement IV offers bounded repair, institutional alternatives, and finally a constitutionally constrained doctrine of uncertainty custody that includes grammar identification, threshold failures, and grammar sensitive balancing under reviewable reasoning. The final chapter returns to the opening line of exposure and refuses the false comfort of doctrinal cadence by naming the gap between identifying incompatibility and making revision politically available.
The reader should know from the beginning that this manuscript will not end by claiming to have solved remainder. The best doctrinal forms in this book discipline closure. They do not redeem it. They can expose hidden transfer, require attributable reasoning, preserve metabolically usable recourse conditions, and force institutions to state their insufficiencies in public. They cannot abolish nonfinality. They cannot make all burdens commensurable. They cannot guarantee political revision where power depends on opacity, fragmentation, or the privatization of cost. They cannot convert every line of exposure into a record that institutions will treat as legible. A doctrine that pretends otherwise becomes another closure technology. This is why the appendices are constitutionally subordinated as instruments under critique. They are tools for bounded use under declared insufficiency. They are not a second ending.
The scene with which I began therefore carries two obligations at once. It demands that the analysis remain concrete enough to identify how a closure event migrates into lived time, and it demands that the book refuse the temptation to treat that migration as a pathology confined to one failed system or one scandalous period of automation. The MiDAS record is an instance with unusual visibility because the documentation, litigation, and audit trail became public, because the error rates were astonishingly high, and because the gap between form and fairness became difficult to deny in court (Cahoo, 2019, pp. 5 to 7; Scott, pp. 5 to 7). Yet the analytic question it exposes is much wider and much older than that system. Institutions close under uncertainty. Bodies carry what closure cannot absorb. Legitimacy, in the synthetic age, can no longer be judged by closure alone. It must be judged by what happens to the remainder, by who is made to hold it, and by whether the grammar that authorizes that transfer can itself be named, reviewed, and contested in public.
The pages that follow begin from that burden of proof.
Chapter 1
The Remainder of Closure
Every institution that governs anything consequential must decide before the world has finished disclosing itself. Courts rule on incomplete records. Agencies adjudicate under deadlines and volume pressures. Hospitals triage before causal certainty is available. Schools classify, employers discipline, insurers deny or approve, regulators issue guidance, and technical systems rank, filter, and trigger actions under conditions in which waiting for exhaustive knowledge would itself become a form of abandonment. The problem is therefore not whether institutions should close. They must. The problem is how closure is made, what is admitted as unresolved when closure is made, and where what remains unresolved goes once the decision acquires force.
The argument of this chapter begins from that practical premise and then imposes a discipline that institutional discourse often resists. We need a grammar that distinguishes several terms that are routinely collapsed in administrative, legal, technical, and public argument, because each collapse conceals a different mode of institutional responsibility. When uncertainty is treated as ambiguity, semantic conflict is mistaken for evidentiary limitation. When indeterminacy is treated as undecidability, ordinary rule application difficulty is inflated into a metaphysics of impossibility. When nonfinality is treated as administrative defect, institutions are asked to complete what no institution can complete. When remainder is treated as background noise, closure appears cleaner than it is. The legitimacy questions that animate this book cannot be stated well until these distinctions are made.
The chapter therefore performs a definitional and adjudicative task under self limitation. It defines the terrain on which later chapters will argue. It does not yet resolve the conflict among legitimacy traditions, and it does not yet settle which institutional grammars should govern which domains. Those questions arrive in the next chapter. Here the work is narrower and foundational. I establish the core terms, define closure as a necessary institutional wager rather than deception by default, distinguish closure from custody evasion, and introduce the difference between declared insufficiency and concealed insufficiency, which later chapters will treat as doctrinally load bearing.
The first term is uncertainty. I use uncertainty in a disciplined institutional sense rather than as a synonym for everything unresolved. The classic distinction in Knight remains useful because it separates calculable risk from conditions in which probabilities cannot be stably assigned or known in a way that supports confident optimization (Knight). Keynes also matters because he insists that rational judgment often operates under incomplete information and unequal evidential weight rather than under a clean probabilistic field (Keynes). Institutions rarely inhabit pure Knightian uncertainty or pure measurable risk. They operate in mixed environments where some variables are modeled, others are estimated by convention, and still others are simply unknown but action is still required. Uncertainty, in this book, names limitations in what can be known, stabilized, or assigned probabilistic confidence for the purpose of justified action at the moment of closure. It is an epistemic and evidentiary condition relative to a decision function.
That definition needs to be held away from ambiguity. In many technical and decision theoretic literatures, ambiguity can name uncertain probability distributions, especially in the wake of Ellsberg’s demonstrations of ambiguity aversion (Ellsberg). That usage is important, and I do not dismiss it. For the purposes of this manuscript, however, ambiguity names something narrower and different. It names plurality of meaning, interpretation, or semantic uptake in the signs, categories, instructions, or reasons through which institutions communicate and govern. A notice may be ambiguous even when the underlying evidence is strong. A policy term may be ambiguous even when probabilities are known. A classifier output label may be ambiguous in relation to the operative legal category even when the model confidence score is numerically high. Ambiguity therefore concerns the meaning bearing vehicles of closure rather than the total evidentiary environment in which closure is made.
This distinction is not semantic fastidiousness. It is a burden allocation distinction. If a claimant cannot understand what allegation must be answered because the notice is vague or the category is opaque, the problem is not exhausted by uncertainty in the agency’s evidence. The institution may have transferred interpretive labor through ambiguous communication while presenting the resulting difficulty as a subject side failure to respond. Ambiguity can function as a custody device because it redistributes the work of clarifying what is at stake after closure has begun to move. Later chapters will show how this becomes more acute in synthetic documentary environments where institutions can produce more text than they can attribute.
The next term is indeterminacy. I use indeterminacy to name a condition in which the available rules, standards, evidence, or institutional purposes underdetermine a single outcome without thereby making decision impossible. Hart’s discussion of open texture remains central because it demonstrates that rule governed systems necessarily contain penumbral cases in which application requires judgment beyond mechanical subsumption (Hart). Administrative law and organizational practice intensify this condition because officials must often act at the intersection of multiple mandates, incomplete records, conflicting policy aims, and time constraints. Indeterminacy is therefore not a failure of institutional order. It is a structural condition of rule application in complex domains.
Indeterminacy differs from uncertainty and ambiguity in at least this sense. Uncertainty concerns what is not adequately known. Ambiguity concerns what is not adequately clear in meaning. Indeterminacy concerns what is not uniquely yielded by the available normative and evidentiary structure, even when relevant facts and terms may be sufficiently known for action to proceed. A case can be determinate in fact but indeterminate in norm. It can be clear in language but indeterminate in weighting. It can be uncertain in evidence yet relatively determinate in procedural response because the institution has adopted default rules under insufficiency. These combinations matter because different institutional duties attach to each.
The fourth term is undecidability. I use this term with deliberate narrowness because it is easy for it to become a gesture of sophistication that exempts argument from institutional judgment. In this book, undecidability does not name a generalized condition in which no decisions can be justified. It names a harder threshold condition in which the decision cannot be fully derived from existing rules or criteria without a supplemental act of judgment that cannot itself be authorized by the same rule set in a non circular way. The language of undecidability is associated, in legal and philosophical debates, with efforts to show that decision always contains an irreducible moment of responsibility beyond algorithmic deduction (Derrida). I retain the term because institutions often deny precisely this moment and then launder discretion as neutral necessity. At the same time, I refuse to use undecidability as a license for romantic voluntarism. Most institutional decisions are not undecidable in this stronger sense. They are difficult, contested, and imperfect under conditions of uncertainty and indeterminacy. Inflating those conditions into undecidability obscures the many places where better custody practices are available.
If uncertainty, ambiguity, indeterminacy, and undecidability describe different aspects of incompletion at the point of action, nonfinality names something else. Nonfinality is not primarily an epistemic or procedural condition. It is a temporal and social condition of lived worlds that exceed institutional endpoints. Arendt’s account of action and plurality is instructive because she refuses the fantasy that human consequences terminate at the moment of official completion (Arendt, The Human Condition). Institutional acts enter webs of relation whose effects continue, mutate, and reappear beyond the file. A court judgment may be final for appellate purposes while a family conflict remains ongoing. A benefits determination may be closed in the record while the person’s material and affective exposure continues. A workplace investigation may conclude while reputational and relational consequences remain active. Nonfinality, in this book, names this persistence of lived and social consequence beyond procedural closure. It is neither a defect to be eradicated nor a reason to suspend all institutional decision. It is a constitutive condition that legitimacy doctrine must respect.
The distinction between nonfinality and remainder is the chapter’s central move. Nonfinality is broader than institutions and broader than closure. Remainder is narrower and specifically tied to closure. Remainder names what persists, is generated, or becomes newly visible after a closure event and cannot be honestly absorbed within the closure’s declared account of itself. It is what the decision leaves behind, pushes outward, displaces, or makes another site carry. Remainder can take evidentiary form, as when a decision proceeds despite acknowledged uncertainty that must later be monitored. It can take institutional form, as when an appeal burden, verification burden, or contradiction burden is transferred downstream. It can take embodied form, as when persons must bear the temporal, reputational, cognitive, and physiological costs of contesting or surviving a closure that the institution treats as complete. It can take political form, as when a closure stabilizes one conflict by privatizing another.
Remainder should not be mistaken for error alone. A closure can be accurate and still produce remainder. A triage decision may be justified and still leave distributive and embodied remainder that must be governed. A court may rule correctly and still generate enforcement burdens, interpretive uncertainties, and downstream asymmetries that are not reducible to legal error. The concept is therefore broader than mistake and narrower than consequence in general. It tracks what closure does not, cannot, or does not admit that it cannot absorb.
This is why the phrase remainder of closure matters more than the phrase unresolvedness alone. Institutions often speak as if what remains unresolved after a decision belongs to the world rather than to the closure event. Sometimes that is true. Often it is only partly true. The chapter’s grammar insists that we ask a sharper question. Which unresolved elements are simply features of an unfinished world, and which have been redistributed by the form, timing, and justificatory grammar of closure itself. Later chapters will call this question the custody question. Here it is enough to establish that remainder is not a poetic supplement to decision. It is an analytic category for tracing burden transfer.
Closure itself must now be defined in a way that resists both institutional apologetics and anti institutional romanticism. Closure is the production of an actionable institutional endpoint under conditions of insufficiency. It is actionable in the sense that it authorizes a next step, allocates a burden, triggers a consequence, stabilizes a status, or temporarily settles a contest such that ordinary institutional operations may proceed. Closure is not identical with truth, and it is not identical with finality. A closure may be provisional, reviewable, reversible, or explicitly temporary. What makes it closure is that it allows authority to move.
Calling closure a wager is therefore precise rather than rhetorical. The institution stakes authority on the adequacy of proceeding now rather than later, here rather than elsewhere, by these criteria rather than those alternatives. Weber’s account of bureaucratic and legal rationality helps illuminate why institutions seek closure in documentary and procedural form, because calculability and file based reproducibility are conditions of administrative continuity (Weber). Fuller’s work is equally important because the inner morality of law depends on forms of publicity, clarity, congruence, and prospectivity that make official action answerable to standards rather than pure command (Fuller). Neither thinker collapses closure into deception. The wager can be legitimate. It can also be abusive. The point is that closure should be analyzed as a structured commitment under insufficiency, not as the simple arrival of certainty.
Once closure is seen as a wager, the central distinction of this chapter becomes available. Closure is not the same thing as custody evasion. An institution can close responsibly under uncertainty. It can declare what is insufficient, specify what grounds action anyway, assign attributable reasons, preserve contestable pathways, and disclose the burdens its closure is imposing. In such cases, the institution still produces remainder, but it does not hide the existence or distribution of remainder behind the language of completion. Custody evasion begins when the institution secures the practical benefits of closure while obscuring the unresolvedness, burden transfer, or justificatory grammar that closure depends upon. It is a mode of taking credit for decisiveness while privatizing the costs of insufficiency.
Custody evasion can take many forms, and later chapters will examine them in detail, but the concept can be stated now without catalog. It occurs when unresolved uncertainty is concealed as certainty, when ambiguity is offloaded as user error, when indeterminacy is laundered as neutral rule output, when discretion is exercised without attributable reasoning, when recourse exists formally but not metabolically, or when burdens generated by closure are distributed in ways that no institutional site is required to recognize as a custody question. Custody evasion is therefore not synonymous with bad motive. It can be designed, tolerated, emergent, or routinized. What unites these forms is that the institution benefits from closure while disavowing responsibility for the remainder its closure produces or redistributes.
The distinction between declared insufficiency and concealed insufficiency now enters as an internal criterion for evaluating closure practices. Declared insufficiency exists when an institution makes visible the limitations under which it is acting and binds that visibility to the form of its closure. The declaration must be more than a ritual disclaimer. It must affect the action’s reasoning, scope, reversibility, recourse conditions, or burden allocation. A temporary emergency measure with explicit expiration and review conditions can count as closure under declared insufficiency. A benefits determination made on incomplete records with clearly stated evidentiary limits and a genuinely usable contradiction pathway can count as closure under declared insufficiency. A technical system output presented as advisory, with traceable provenance and attributed human responsibility for action, can count as closure under declared insufficiency.
Concealed insufficiency, by contrast, exists when institutions act under limitations they do not disclose, or disclose only in ways that do not govern the closure’s force. This includes obvious cases where evidence is weak and represented as decisive. It also includes more subtle cases where the institution knows that notice is likely unusable, that categories are unstable, that contradictory evidence is expensive to surface, that staffing conditions make formal recourse practically inaccessible, or that the operative grammar of authority has shifted from one register to another without public acknowledgment. Concealed insufficiency is not cured by boilerplate caveats. A disclaimer that does not alter attributable reasoning or recourse conditions can itself function as concealment by performing honesty while preserving closure’s full force.
This difference will matter later in the doctrine chapter because threshold failures depend on whether uncertainty transfer is hidden, attributable, and contestable. It also matters in Chapter 7, where the legitimacy of institutional closure under nonfinal conditions turns on whether institutions can admit what they cannot absorb without punishing subjects for inhabiting what remains. For now the distinction has a simpler role. It prevents the familiar slippage in which every institutional acknowledgment of complexity is praised as reflexive sophistication even when nothing in the burden distribution changes. Declared insufficiency is not an aesthetic of humility. It is a custodial practice.
The grammar developed here also clarifies a recurrent error in contemporary debates about automation and generative systems. Much public argument treats the crisis of synthetic governance as a crisis of falsehood, hallucination, or bias alone. These are serious problems. They are not the whole problem. A system can generate outputs that are locally accurate and still intensify custody evasion by lowering the cost of closure relative to the cost of contradiction, verification, and attributable reasoning. Synthetic systems multiply the volume of notices, summaries, classifications, and justifications that can be produced in form. They do not automatically improve the institution’s custody of uncertainty. In some settings they worsen it by enabling closure to appear more articulate while becoming less reviewable. The distinction between closure and custody evasion is therefore indispensable in the synthetic age because it prevents critique from narrowing to content accuracy while ignoring burden transfer and recourse conditions.
At this stage a possible objection should be faced directly. One might argue that the chapter’s distinctions are too refined for institutional practice and that governance requires coarser categories. The objection mistakes the function of the grammar. The argument is not that every frontline actor must carry a philosophical lexicon into every decision. The argument is that without these distinctions institutions routinely collapse different kinds of insufficiency and then misassign responsibility. A claimant is blamed for “nonresponse” when the problem was ambiguous notice. A reviewer invokes “uncertainty” when the issue was concealed indeterminacy in criteria weighting. An agency cites “discretion” where the operative act was an undecidability claim used to avoid attributable reasons. Precision at the level of doctrine and design is what makes simpler operational practices fairer rather than more cumbersome.
Another objection comes from the opposite direction and claims that by normalizing closure as a wager this chapter grants too much to institutions, especially in settings marked by historical domination. That objection is morally and politically important. It keeps the chapter from drifting into procedural neutrality. The answer is that refusal to name closure as necessary does not protect vulnerable persons if institutions continue to close anyway. It only drives the analysis toward denunciation without doctrine. The wager formulation is intended to make institutions more accountable, not less, because wagers can be specified, reviewed, bounded, and judged in relation to their burden transfers. A theory that begins by demanding certainty from institutions will fail to diagnose how authority actually operates. A theory that begins by excusing closure because uncertainty is universal will fail to diagnose custody evasion. The present grammar is built to avoid both failures.
The chapter’s terms can now be held together in a single institutional sequence. An institution faces uncertainty because relevant evidence is incomplete and future effects are not fully knowable. It faces ambiguity because its categories, notices, or reasons may be semantically unstable or under clarified. It faces indeterminacy because available rules and purposes do not uniquely determine one outcome. In rarer cases it faces undecidability in the stronger sense, where no closure can be fully derived without an irreducible supplementary judgment. It acts anyway and produces closure. The closure interacts with the nonfinality of lived and social worlds that exceed procedural endpoints. What closure cannot absorb, or does not admit that it cannot absorb, persists as remainder. The legitimacy question then turns on custody. Was insufficiency declared or concealed. Was the burden distribution visible or hidden. Were reasons attributable. Was recourse metabolically usable. Was the operative grammar of closure publicly reviewable.
This chapter does not yet answer those legitimacy questions across traditions. It cannot, and it should not pretend to. Arendtian, Fullerian, Weberian, and administrative grammars authorize closure differently and assign value to publicity, expertise, legality, procedure, and action in conflicting ways. Indigenous governance frameworks, Islamic administrative traditions, and postcolonial critiques will later press even harder on what counts as authority, standing, and justifiable burden. The present chapter establishes the terms under which those conflicts can be staged without conceptual confusion. It defines the terrain. It does not resolve the jurisdiction of the theories that will contest it.
A final clarification is needed before the chapter closes. Remainder is not an argument against decision. It is an argument against the misdescription of decision. Institutions that acknowledge remainder are not thereby weak. They may become more legitimate because they cease presenting closure as total absorption. Institutions that hide remainder are not thereby strong. They may become fragile because hidden burdens return as distrust, litigation, noncompliance, attrition, reputational crisis, and political delegitimation. The synthetic age intensifies this fragility because closure can now be mass produced in textual and documentary form, which increases the temptation to treat articulation as custody. It is not. Articulation can conceal as effectively as silence if the grammar of closure is not reviewable and the burden of insufficiency is not publicly governed.
The remainder of closure, then, is the name for what a decision leaves in motion while claiming enough stability to act. The task of this book is not to abolish that remainder. No institution can do that honestly. The task is to develop a way of judging legitimacy that asks where the remainder goes, who carries it, by what grammar that transfer is justified, and under what recourse conditions the transfer can be contested. The next chapter turns to the conflict among legitimacy traditions and begins to construct uncertainty custody as a cross cutting adjudicative test rather than a replacement theory.
Chapter 2
Legitimacy in Conflict
Chapter 1 established a grammar of closure, insufficiency, nonfinality, and remainder. That grammar was necessary, but it is not yet enough to judge institutions, because institutions do not justify closure under a single account of authority. They speak in different legitimacy languages, often more than one at once, and they can shift among them as pressure changes. The same decision can be defended as legally regular, administratively necessary, democratically authorized, technically rational, procedurally fair, or publicly accountable without those claims converging into a single standard. A closure that looks justified under one grammar can appear evasive under another. A doctrine that fails to stage this conflict will either collapse into abstraction or mistake one tradition’s criteria for universal adjudication.
This chapter therefore does not attempt harmony. It stages a conflict among four legitimacy grammars that recur across contemporary closure regimes and that frequently overlap in real institutions without becoming equivalent. I call them, for shorthand, Arendtian, Fullerian, Weberian, and administrative legitimacy grammars. The names do not indicate pure schools in practice. They indicate dominant justificatory orientations that shape how closure is authorized, what counts as a defect, and where burden transfer becomes visible or disappears. The task of the chapter is to show how each grammar can justify closure under insufficiency and where each becomes vulnerable to hidden uncertainty transfer and custody evasion. The chapter does not settle which grammar should govern all domains. It defines the conflict so that later doctrine can proceed without conceptual fraud.
I also need to state the chapter’s perimeter openly. The staged conflict is not exhaustive, and the omissions are not neutral. The architecture of this chapter excludes sustained treatment of Indigenous governance frameworks, Islamic administrative traditions, and major postcolonial critiques of legitimacy, standing, and defensibility. This exclusion produces a real loss in at least three ways. It risks over-centering state and bureaucratic forms as the primary scene of legitimacy, it risks narrowing the archive of lawful and accountable authority to Euro-Atlantic categories, and it risks treating contestability as if it were always already inside the institutions whose closure is under scrutiny. Later chapters, especially the doctrine chapter, will have to acknowledge that any grammar identification protocol built on available institutional records may reproduce precisely these exclusions if archival presence is mistaken for normative relevance. For now, I bracket these traditions as a declared insufficiency, not as a claim that they are secondary. Their pressure remains active at the edge of the chapter and returns later as a doctrinal constraint.
The first grammar in conflict is Arendtian. To invoke Arendt in a book on closure and legitimacy is to invite a misreading unless the terms are handled carefully. Arendt does not give a bureaucratic theory of administrative justification. She gives a political account of action, plurality, judgment, natality, and world building under conditions in which certainty and mastery are never fully available (Arendt, The Human Condition; Arendt, Between Past and Future). Her importance here lies in the way she locates legitimacy not primarily in correctness of outcome but in the maintenance of a common world where action and judgment can appear, be contested, and remain meaningful among plural actors. In this register, closure is never self-grounding. It is legitimate only insofar as it stabilizes enough world to permit further action and judgment without pretending to end politics.
An Arendtian grammar therefore authorizes closure as provisional worldly stabilization. Institutions can decide, and sometimes must decide urgently, but they do not thereby acquire metaphysical finality. Their closure is legitimate when it preserves a space in which plurality remains politically real, where those affected are not reduced to mute objects of administration, and where judgment remains answerable to a shared world rather than absorbed into necessity claims. This grammar is especially powerful against closures that masquerade as purely technical but in fact settle political questions while denying that politics is occurring. It is also powerful against the fantasy that legitimacy can be reduced to procedural throughput, because Arendt insists that politics is not exhausted by administration and that the public realm is not identical with state management.
At the same time, an Arendtian grammar is vulnerable in ways that matter for this book’s custody question. One vulnerability is the historic tendency, in Arendt’s own distinctions, for labor and the social to be treated in ways that can under-illuminate the administrative production of embodied burden (Arendt, The Human Condition). If closure is judged primarily by whether it preserves a space of appearance and political action, the mundane but crushing transfer of uncertainty into time, exhaustion, translation labor, and household administration can become politically invisible precisely because it is experienced in domains coded as social, private, or prepolitical. A second vulnerability follows from the same strength that makes the grammar attractive. Because it is attuned to action and plurality, it can be less precise about how large institutions routinize burden transfer through forms, categories, timing, and document design long before anyone reaches a scene of recognizably public judgment. In practice, this can leave administrative remainder under-theorized until it erupts into scandal or constitutional litigation. An Arendtian lens can diagnose depoliticization with force and still miss how closure’s remainder is privatized at low visibility and high frequency.
The second grammar is Fullerian. Fuller matters here because he offers not just a defense of legality but a demanding account of law as a purposive enterprise of subjecting conduct to the governance of rules, where legality has an internal morality expressed through publicity, generality, prospectivity, clarity, consistency, possibility of compliance, relative constancy, and congruence between official action and declared rule (Fuller). A Fullerian legitimacy grammar authorizes closure when institutions act through forms that preserve reciprocity between ruler and ruled rather than treating subjects as objects of ad hoc command. In this grammar, closure is legitimate not because it is correct in every instance, but because it is produced through a mode of governance that can be followed, understood, and contested by those subject to it.
This grammar is indispensable for the present project because it directly illuminates how closure can fail before one even reaches distributive outcomes. If notices are unintelligible, if standards are shifting, if officials act incongruently with declared rules, if compliance is impossible within the actual conditions imposed, then closure may be formally recorded and practically actionable while still lacking legitimacy in a Fullerian sense. Fuller also provides a language for why concealment matters. A closure built on concealed insufficiency is often also a closure that undermines reciprocity, because it demands compliance or acquiescence under conditions the institution itself knows are not adequately disclosed or governable by ordinary subjects.
Yet a Fullerian grammar has limits that become visible once the custody question is foregrounded. Fuller’s internal morality can show that a regime is failing as law while still under-specifying how burdens are distributed when legality forms are present but metabolically unusable. A notice may be clear in syntax and public in form, yet timed, routed, or framed so that contradiction is prohibitively costly for those affected. A recourse pathway may exist and be legally legible while remaining functionally inaccessible because of pace, translation burdens, sanctions, or cumulative depletion. In such settings the institution can appear to satisfy a significant portion of legality’s internal demands while externalizing the residual cost of insufficiency into bodies and households. The point is not that Fuller ignores burden. It is that a Fullerian grammar alone can mistake formally improved reciprocity for adequate custody when the decisive asymmetry lies in the usability, not only the existence, of reciprocal forms.
The third grammar is Weberian. Weber’s relevance is almost unavoidable in any serious account of modern closure because he gives the most durable account of legal rational authority, bureaucratic office, calculability, hierarchy, file based administration, and impersonal competence as the organizational conditions of large scale rule (Weber). A Weberian legitimacy grammar authorizes closure through office bound procedure and role competence. In this grammar, closure is legitimate when it is rendered through authorized channels, according to rule and jurisdiction, by officials occupying defined offices whose acts can be documented, repeated, and integrated into ongoing administration. The attraction of this grammar is not moral warmth. It is continuity. It makes complex coordination possible. It permits institutional memory. It disciplines arbitrariness in one register by routing decisions through impersonal forms.
This grammar is especially important for the present argument because it allows one to say plainly that institutions must often close in order to remain institutions. Weber is unsentimental on this point. Administration cannot operate as a continuous seminar in first principles. It requires files, categories, deadlines, and routinized decisions. A doctrine of uncertainty custody that failed to recognize this would not be a doctrine but a refusal of governance. The Weberian grammar also clarifies why documentary closure carries so much practical authority. The file is not a neutral container of facts. It is an instrument through which the institution stabilizes what counts as actionably real.
The Weberian vulnerabilities, however, are severe when viewed through the lens of remainder. The very strengths of bureaucratic rationalization can conceal burden transfer by distributing decision fragments across offices, roles, and systems such that no single actor is institutionally positioned to see the full custody event. Formal rationality can preserve means end coherence inside each unit while producing cumulative exclusions in aggregate. Role bounded action can protect predictability and yet diffuse responsibility for embodied costs generated by closure. Fragmentation can make injustice appear as throughput. What later chapters call distributed exclusion is, in one register, a Weberian risk taken to scale. No single rule need be abusive. No single official need act in bad faith. The aggregate can still produce a custody outside that no office claims as policy. Weber gives the architecture of this risk even where he does not frame it in custodial terms.
There is another Weberian vulnerability that bears directly on contemporary synthetic systems. Because legal rational authority values fileability, standardization, and administrative legibility, institutions can become highly responsive to documentary expansion while remaining weak in attributable custody. In practical terms, they can absorb more records, more outputs, more coded reasons, and more procedural traces without improving the conditions under which affected persons can understand, contest, and survive closure. Synthetic generation intensifies this vulnerability because it multiplies documentary form at low cost. A Weberian institution under digital pressure may appear more procedurally articulate while becoming less answerable in the concrete sense this book requires.
The fourth grammar I stage is administrative legitimacy in a more specific and contemporary sense than Weberian bureaucracy alone. By this I mean the legitimacy claims that emerge from the modern administrative state’s combination of expertise, procedural regularity, reason giving, balancing, delegated discretion, and state capacity. In the United States context, this grammar is visible in doctrines and practices that justify administrative closure not only by formal legality but by institutional competence, comparative error management, and procedural design. Mashaw’s work is central because he shows that administrative justice is not reducible to one value and often operates through multiple models, including bureaucratic rationality, professional treatment, and moral judgment, each with different legitimacy commitments (Mashaw, Bureaucratic Justice). Administrative legitimacy in this sense authorizes closure when institutions can plausibly claim that they are using specialized competence and procedural mechanisms to manage complexity under resource constraints in a manner that is publicly defensible, even if imperfect.
This grammar has real normative force. It captures what neither a purely Arendtian nor a purely Fullerian or Weberian account can fully absorb, namely the necessity of delegated expertise in high volume and technically complex environments where elected bodies cannot directly adjudicate every case and courts cannot substitute for ongoing administration. It also provides a vocabulary for explicit tradeoffs. The administrative state often governs through comparative judgments about error, cost, speed, and capacity. The familiar balancing structure in U.S. due process doctrine, exemplified by Mathews v. Eldridge, is an obvious instance, where private interest, risk of erroneous deprivation, probable value of additional safeguards, and governmental burden are weighed to evaluate procedural sufficiency (Mathews v. Eldridge). Whatever one thinks of the doctrine’s outcomes, the grammar is visible. Closure is justified through structured balancing under institutional constraints.
The vulnerability of this grammar, and the reason it cannot govern this book without modification, lies in what it tends to count and what it tends to undercount. Administrative balancing can become a sophisticated mechanism for normalizing burden transfer when embodied and temporal costs are either unmeasured or misclassified. It can price contradiction as if all parties experience procedural cost symmetrically. It can treat recourse pathways as available because they exist on paper while ignoring metabolically decisive differences in time, translation burden, sanction expectation, and cumulative depletion. It can also use expertise to immunize grammar shifts. An agency may defend one closure as technically necessary, another as legally mandated, a third as administratively efficient, and a fourth as emergency action, all within a single operational regime, without publicly acknowledging that the operative legitimacy grammar has moved. In such settings administrative legitimacy becomes vulnerable to grammar shopping even when each local justification sounds plausible.
If these four grammars are allowed to appear only as rivals, the chapter would fail. The real problem is not rivalry alone but overlap without declared hierarchy. Modern institutions regularly combine them. A single closure may be defended in Fullerian terms of rule fidelity, Weberian terms of office competence, administrative terms of expert balancing, and Arendtian terms of preserving the conditions of a common world by acting rather than remaining inert. The combination can be honest. It can also be opportunistic. The institution may invoke Fullerian legality when challenged on arbitrariness, Weberian role necessity when challenged on responsibility, administrative balancing when challenged on burden, and democratic or public world imperatives when challenged on urgency. The resulting justificatory field is often internally unstable but externally effective because affected persons must contest not only a decision but the very grammar in which the decision is being justified.
This is why uncertainty custody enters in this chapter as a cross cutting adjudicative test rather than a replacement theory of legitimacy. It does not purport to decide in advance whether Arendtian, Fullerian, Weberian, or administrative values should dominate in every domain. It asks a different question that each grammar must answer if it is to justify closure under insufficiency without concealing remainder. Where does the unresolved uncertainty go. How is burden distributed. What recourse conditions exist in metabolically usable form. Are the reasons attributable. Is the insufficiency declared or concealed. Is the operative justificatory grammar publicly reviewable and contestable. A closure may remain legitimate under more than one grammar while performing differently on these custodial dimensions. The cross cutting test does not dissolve conflict among traditions, but it can make hidden burden transfer visible across them.
The point is easiest to see when each grammar is subjected to the same custodial questions. An Arendtian closure that preserves public action but privatizes the labor of surviving administrative nonresponse may fail custodially even if it appears politically necessary. A Fullerian closure that improves clarity and congruence but leaves recourse metabolically inaccessible may fail custodially despite formal legal gains. A Weberian closure that secures calculable administration through routinized categorization may fail custodially if it distributes responsibility so effectively that no office bears attributable reasoning for embodied remainder. An administrative closure that passes balancing review may fail custodially if the balancing record underestimates temporal cost, misprices contradiction, or treats simulated recourse as sufficient. The test does not announce that any one of these grammars is false. It asks whether, in operation, they are being used to justify closure while hiding what closure displaces.
At the same time, a cross cutting test creates its own risks, and the chapter must state them before using the concept further. The first risk is false commensurability. If uncertainty custody is treated as a single metric, it will flatten the very conflicts this chapter is trying to preserve. Arendtian concerns about public world maintenance are not reducible to administrative error costs. Fullerian reciprocity is not interchangeable with Weberian calculability. Administrative capacity constraints are not dissolved by political critique. The cross cutting test must remain adjudicative and diagnostic rather than absorptive. The second risk is theater. Institutions can learn the language of custody and continue the same closure practices by producing declarations, checklists, and burden statements that do not alter attributable reasoning or recourse conditions. Later chapters, especially Chapter 11, therefore keep threshold failures and anti theater constraints inside the doctrine. The third risk is perimeter blindness. A custody framework built solely from records generated by dominant institutions may reproduce exclusions by treating archival presence as normative authority.
This is the point at which the chapter’s declared perimeter has to become more than a disclaimer. The exclusion of Indigenous governance frameworks, Islamic administrative traditions, and postcolonial critiques is not only a bibliographic omission. It changes what this chapter can perceive. Indigenous political thought and legal orders often place relational responsibility, land, kinship, and collective continuity in positions that disrupt the state centered assumptions embedded in all four grammars staged here. Postcolonial critiques expose how legitimacy, standing, and defensibility have historically been co-produced with conquest, classification, and racialized administration, which means that closure cannot be judged solely by internal procedural virtues when the institutional field itself is historically stratified (Fanon; Spivak; Mbembe). Islamic administrative and legal traditions, in their historical and contemporary diversity, press on the relation between governance, normativity, juristic authority, and state power in ways that unsettle the simple equation of legitimacy with modern bureaucratic form (Hallaq). By excluding sustained engagement with these traditions here, the chapter narrows the range of available legitimacy grammars and risks overvaluing the self descriptions of modern administrative states. Naming that loss is not redress. It is the minimum condition for using the staged conflict without pretending exhaustiveness.
The conflict among the four grammars also reveals a more technical problem that the book cannot postpone forever. A cross cutting test can only evaluate closure if one can identify, in a given setting, what legitimacy grammar is actually operative. That identification cannot be left to institutional declaration alone. Institutions have incentives to describe their own authority in whichever grammar minimizes present scrutiny and maximizes discretionary room. Nor can grammar identification be inferred from formal mandate alone, because institutions often act under hybrid justifications that shift with decision type, emergency posture, evidentiary condition, and audience. A benefits agency may be Weberian in ordinary intake, administrative in fraud analytics, Fullerian in formal notice design, and quasi Arendtian in public emergency justification, all within a single program. If uncertainty custody is to function as a serious adjudicative test, it needs a public and contestable method for identifying the operative legitimacy grammar in the closure setting under review.
That doctrinal problem is not solved in this chapter, and it should not be finessed. At this stage I can only mark its necessity and specify the cost of ignoring it. Without a grammar identification method, the cross cutting test collapses into either abstraction or opportunism. It becomes abstract if it applies the same custodial expectations in all settings without regard to the goods and constraints each grammar legitimately foregrounds. It becomes opportunistic if institutions are free to declare whichever grammar makes the closure easiest to defend after the fact. Chapter 11 will address this by constructing a publicly reviewable grammar identification protocol tied to mandate, decision function, closure form, standing structure, and historical modes of justification, and by subjecting that identification to contestable external review. For now, the conflict staged here is enough to show why such a protocol is not optional.
The chapter can now close with the narrower accomplishment it was designed to achieve. It has shown that closure legitimacy is not a singular criterion but a field of contesting grammars. It has shown that each of the four staged grammars authorizes closure under insufficiency in a way that captures something real and normatively significant. Arendtian grammar preserves the primacy of plurality and public world conditions against the reduction of politics to administration. Fullerian grammar preserves reciprocity and rule based answerability against arbitrary command and procedural degradation. Weberian grammar preserves the organizational conditions of continuity, fileability, and office bound action against romantic anti institutionalism. Administrative legitimacy grammar preserves the normative importance of expertise, delegated capacity, and explicit tradeoff management in complex systems that cannot be governed by direct political judgment alone. None of these can simply be discarded.
It has also shown that none is self sufficient in the face of hidden burden transfer. Arendtian grammar can under-read the administrative production of embodied remainder. Fullerian grammar can over-credit formal reciprocity while undercounting metabolic inaccessibility. Weberian grammar can normalize distributed custody evasion through routinized fragmentation. Administrative legitimacy grammar can convert burden transfer into balancing variables and permit grammar shifts without public acknowledgment. The result is not theoretical despair. It is the exact condition under which a cross cutting adjudicative test becomes necessary.
Uncertainty custody is that test, but only in a provisional and constrained sense at this stage. It is not a replacement constitution for legitimacy. It is a discipline of judgment that asks each grammar to answer for what happens to insufficiency once closure is made actionably real. It asks where remainder is located, whether burden transfer is visible, whether recourse is genuinely usable, whether reasons are attributable, and whether the grammar doing the justificatory work can itself be named and contested in public. The unresolved issue, and the one that now becomes doctrinally unavoidable, is how that operative grammar is to be identified without permitting either institutional self certification or theoretical overreach. The book can proceed only by carrying that problem forward in the open.
Chapter 3
Embodied Remainder and Asymmetric Exposure
She discovers the determination through the consequence and not through the reasoning.
The money does not arrive. A refund is intercepted. A wage has been touched. A balance that had already been assigned to rent, groceries, debt, or the ordinary choreography of staying afloat is suddenly absent, and the first intelligible fact is not an allegation but an extraction. Only afterward does the person begin to learn that an institutional sequence has already unfolded somewhere else in documentary time, that a file has moved, that a classification has hardened, that a determination has become actionable, and that the institution has already spoken in forms, clocks, and codes before the subject of the action had a metabolically usable chance to answer. In the MiDAS litigation record, as later described in federal and state court opinions, claimants alleged exactly this inversion of ordinary fairness, with fraud determinations, penalties, and collection consequences arriving through automated or semi-automated pathways that many people encountered first as aftermath rather than as reviewable accusation (Cahoo v. SAS Analytics; Scott v. Michigan Unemployment Insurance Agency). The opening line of exposure from the prologue returns here because the mechanism cannot be seen if one begins with doctrine alone. The scene is not there to intensify tone. It is there to establish sequence.
What appears, from the institutional side, as process completion often appears, from the exposed side, as temporal dislocation. The institution has one clock and the person has another, and the institution’s clock is the one that counts as governance. The agency flags a discrepancy, issues a questionnaire, runs a response window, records nonresponse or treats a response as inculpatory, issues a determination, assesses penalties, and initiates collection pathways. The person discovers a consequence, reconstructs the event from fragments, searches for reasons, waits on a line, gathers documents, attempts to identify what exactly is being alleged, and learns that the timing that governs recourse is not the timing of discovery but the timing of issuance. The difference is not administrative inconvenience. It is the first form of asymmetric exposure. One sequence is treated as action and the other as delay even when the second sequence is the first moment at which the person can actually begin to know what has happened.
This asymmetry is easy to flatten if one stays inside documentary categories. The file can present itself as coherent because it converts unresolvedness into successive forms of procedural legibility. A discrepancy becomes a questionnaire. A questionnaire becomes a response state. A response state becomes a determination. A determination becomes debt. Debt becomes collection. Each transformation can be internally intelligible to the system because each is documentable. The claimant’s sequence is different. It begins not as documentation but as bodily and temporal reorganization. Vigilance rises. Attention shifts. Ordinary planning is interrupted. Speech becomes strategic because any account may have to be given to an agency worker, an employer, a spouse, a landlord, a preparer, or a judge before the person even has a stable statement of the grounds of accusation. The system calls this a response problem. The body experiences it as an exposure problem.
Only now, after the mechanism has been shown, can the chapter state the analytic claim it is responsible for earning. One person’s institutional remainder becomes another institution’s administrative normality.
The sentence names a transfer that occurs across an institutional ecology and not only within one office. The agency’s unresolved uncertainty about identity continuity, reporting accuracy, notice receipt, evidentiary sufficiency, or fraud inference does not vanish when a determination is issued. A portion is converted into closure and rendered actionable. The remainder is redistributed. Some of it is externalized onto the claimant as vigilance, contradiction labor, time loss, reputation management, and survival administration. Some of it reappears downstream in other offices as routine workload, call volume, appeal backlog, returned correspondence, enforcement action, compliance review, audit finding, litigation risk, or settlement cost. The original uncertainty, once translated through closure, no longer presents itself at each subsequent site as uncertainty. It presents itself as operational demand. This is what the sentence means. The exposed person carries nonfinality as life, while the institution re-encounters the same unresolvedness as normal administration.
The point is structural and should not be misheard as a moral accusation against every official inside the chain. Institutions are organized to metabolize unresolvedness by redistributing it. They routinize what they cannot resolve. Weber helps explain why this occurs because bureaucratic continuity depends on file form, role differentiation, calculability, and repeatable procedures that make large-scale governance possible (Weber). What Weber does not furnish by himself is an account of what happens when routinization is achieved by exporting uncertainty into embodied life while preserving documentary coherence inside the file. The person’s burden returns to the institution only after it has changed form. It returns as missed deadline, escalated call, complaint, appeal, hearing request, arrears problem, litigation, or public scandal. By then the closure event that generated the burden may be several procedural translations away from view. Each office can be locally intelligible and still no institutional site can see the full custody transfer.
The distinction that keeps this chapter honest is the distinction between embodied nonfinality and imposed remainder. Embodied nonfinality is a general condition of human life. Persons do not live as closed records. Grief exceeds ritual. Illness exceeds diagnosis. Conflict exceeds judgment. Institutional action enters lives that were already unfinished and will remain unfinished after the file closes. A theory of legitimacy that mistakes all nonfinality for institutional injury becomes incoherent because it demands that institutions complete what no institution can complete. Imposed remainder is narrower. It names the portion of unresolvedness that institutions generate, amplify, or redistribute through closure practices in ways that produce asymmetric exposure. The distinction matters because the concept of remainder becomes useless if it absorbs every hardship following a decision.
The MiDAS context makes the distinction concrete. A claimant who has lost employment already inhabits uncertainty, financial strain, and social vulnerability that no unemployment system can abolish. That baseline nonfinality is not itself proof of custody evasion. The institution becomes answerable in a different register when its closure mechanisms add hidden burdens through opaque allegation, unusable notice, compressed timing, or recourse forms that are available in law and unavailable in practice. The legal language in the litigation and later opinions names due process, notice, and error risk, and those categories are indispensable because they place the problem within constitutional doctrine rather than mere sympathy (Cahoo; Scott). This chapter asks what those same failures look like from the side of lived exposure. It does not replace doctrine. It prevents doctrine from becoming a view from nowhere.
Phenomenology enters at this point because the problem under analysis is not only whether information was absent, but how a person’s practical world is reorganized when institutional action precedes usable orientation. Merleau-Ponty is helpful here not as a prestige citation but as a discipline against administrative abstraction. The body is not a passive receptacle into which institutional consequences are deposited. It is the living site through which the world becomes navigable, threatening, proximate, delayed, possible, or foreclosed (Merleau-Ponty). When an institution acts in a way that the person cannot yet intelligibly locate, the injury is not reducible to data deficiency. It is a disturbance of practical orientation. What had been background becomes a monitoring task. Mail, portals, unknown numbers, payroll irregularities, agency correspondence, and tax records acquire heightened salience. The future becomes narrower because planning must reserve capacity for possible institutional recurrence. Administrative closure is therefore not only a legal event with downstream consequences. It can be a reconfiguration of the field of attention.
This is one reason the chapter had to begin with consequence before explanation. If the analysis starts from doctrine, burden appears as additive. A delay. A fee. A hearing. A document request. A call queue. All of those may be present. Yet the exposed person often experiences the burden first as a changed relation to anticipation. What can be safely ignored shrinks. What must be checked expands. The institution may record only discrete contacts, while the person bears a continuous reservation of readiness. In this sense, administrative remainder is not just what a subject must do in response to closure. It is what a subject must remain prepared for after closure has already asserted authority. The temporal cost is not exhausted by logged interactions. It includes the occupation of cognitive and affective bandwidth by a system that has demonstrated its ability to act before being answerable.
Trauma theory can clarify this structure if it is used with restraint. It would be analytically undisciplined to classify every episode of bureaucratic exposure as trauma in a clinical sense. The category would lose explanatory force and would flatten distinctions among kinds of harm. At the same time, trauma theory offers a developed account of what happens when systems of threat, opacity, and nonresolution reorganize safety, vigilance, and temporal orientation. Herman’s account remains instructive because she emphasizes that overwhelming exposure alters trust and the relation to future time, and that the afterlife of exposure can persist long after the originating event is institutionally coded as complete (Herman). The relevance here is limited and precise. Administrative closure, especially when accompanied by penalties, accusation, and opaque recourse, can generate rational forms of anticipatory labor and hypervigilance that institutions later misread as disorderly behavior, noncompliance, inconsistency, or affective excess. The point is not pathologizing the claimant. The point is that the claimant’s altered orientation may be an intelligible response to asymmetric exposure.
Affective theory deepens this account because institutional documents do not only transmit content. They organize subject positions. Ahmed’s work is particularly useful in this regard because she treats emotion as relational orientation and social circulation rather than private interior residue (Ahmed). An accusation, notice, or determination does not simply tell someone what has occurred. It places that person within a field of expectation, suspicion, and demanded movement. In a fraud context, classification carries social and moral force even before its evidentiary basis becomes reviewably available to the person so classified. The claimant may have to narrate the state’s accusation in settings where the person cannot yet narrate the accusation’s grounds. Shame, fear, anger, and uncertainty are not detachable sentiments added after the fact. They are among the media through which the remainder of closure is carried and through which burden becomes socially legible or misrecognized.
Here the phrase asymmetric exposure can be sharpened further. The asymmetry is not simply that institutions possess more information. Institutions themselves often act under profound uncertainty. The asymmetry lies in the direction and cost of conversion. Institutions can convert their uncertainty into authoritative action. Individuals must convert institutional action back into intelligible uncertainty before contestation becomes possible. The institution moves from not knowing enough to act to acting anyway under a justified grammar. The subject moves from being acted upon to trying to determine what would count as an answer. These are not reciprocal epistemic positions because the temporal burden and consequences of error are not symmetrically distributed.
Predictive neuroscience is useful here because it offers a vocabulary for why administrative uncertainty can become metabolically heavy even when the formal procedural demands look modest in isolation. Predictive processing and active inference frameworks, in broad outline, treat cognition and action as ongoing processes of prediction, update, and regulation in which organisms maintain workable models of their environments while minimizing maladaptive surprise through perception, action, and learning (Friston; Clark). One need not endorse every strong metaphysical claim made in that literature to see its relevance to institutional exposure. A person navigating ordinary life depends on being able to form sufficiently reliable expectations about what events matter, what actions are required, and what cues can be deprioritized. Opaque institutional closure disrupts precisely this economy of expectation.
The burden is not exhausted by having to answer one notice or attend one hearing. It lies in the forced expansion of prediction labor under conditions where the threat source is powerful, intermittent, and imperfectly intelligible. If a person cannot model when an agency will act, what counts as adequate response, how penalties are triggered, or whether prior attempts at clarification alter anything, then the person must allocate ongoing cognitive and physiological resources to monitoring and readiness. Seth and Friston’s work on interoceptive inference is especially relevant because it links prediction and regulation to felt bodily states rather than treating cognition as detached symbolic computation (Seth and Friston). Administrative uncertainty can therefore become embodied remainder in a literal and disciplined sense. The burden is carried through altered regulation of attention, arousal, and anticipation, not only through explicit legal reasoning.
This framework also helps correct a persistent institutional self-description. Institutions often defend closure by invoking the necessity of acting under uncertainty, and at one level they are right. No court, agency, hospital, or platform waits for exhaustive certainty before acting. Predictive and inferential frameworks make this condition harder to romanticize because they remind us that uncertainty is not an administrative special case. It is a general condition of situated action. The legitimacy question then becomes custodial rather than metaphysical. Who gets to discharge uncertainty through authoritative closure, and who is required to absorb the residual error budget in lived time. McEwen’s work on allostatic load, used carefully, supports this shift because it shows how repeated adaptation demands can accumulate physiological and behavioral costs over time without requiring a simplistic one-to-one mapping from any single stressor to any single outcome (McEwen). The chapter does not seek a biomarker theory of justice. It seeks to make visible the fact that institutions can impose ongoing regulation demands on bodies while coding the generating event as complete in the file.
An objection is now necessary, because the chapter risks appearing to redescribe ordinary bureaucratic dysfunction in a vocabulary that is too theoretically dense for the phenomenon. The objection should be taken seriously. Understaffing, poor interface design, fragmented systems, weak supervision, and bad notice practices explain a great deal of administrative harm. The present argument does not deny this. It insists that those defects have a form when they become patterned and recurring. They become mechanisms of burden transfer. The theoretical vocabulary is not used to inflate the moral drama of bureaucracy. It is used to prevent legal and organizational descriptions from concealing the lived carriage of unresolvedness. If a system repeatedly secures documentary closure by shifting uncertainty resolution into households, then what looks operational from one side is embodied governance from another.
A second objection comes from fairness to institutions. Agencies administer large volumes under statutory directives, anti-fraud mandates, budget constraints, and political scrutiny. Errors will occur, and some burdens will be unavoidable. This chapter accepts that premise because the book as a whole rejects the fantasy of total closure. The issue is not whether institutions can eliminate all remainder. They cannot. The issue is whether they declare insufficiency, make reasons attributable, preserve metabolically usable recourse, and disclose the burdens their closures generate, or whether they secure the practical advantages of decisiveness while privatizing the cost of unresolvedness. Once this distinction is in view, it becomes possible to criticize custody evasion without demanding impossible institutional innocence.
The opening scene can now be read again without reducing it to anecdote. A claimant who learns of a determination through a tax intercept does not first receive a reason. She receives a reordered life schedule. Bills must be re-ranked. Explanations must be improvised. Documents must be sought for an allegation whose basis may still be unclear. Emotional composure becomes labor. The institution, by contrast, meets the same event through partitioned categories that are each operationally intelligible and cumulatively evasive. She is a claimant record, a response state, a collection case, a call event, a returned-mail issue, an appeal, a backlog item, a compliance finding, a plaintiff. None of these categories is false. The problem is that their sequence can distribute visibility so effectively that no single institutional site is required to name the transfer by which closure’s remainder has been externalized into embodied life. This is how one person’s institutional remainder becomes another institution’s administrative normality.
The chapter must also guard against a predictable institutional response once embodied burden becomes part of the analysis. Acknowledgment can become extraction. Institutions can begin demanding more narratives, more attestations, more measurements, more disclosures, and more intimate evidence in order to prove the burden they helped generate. They can build dashboards of suffering and call this accountability. The concept of embodied remainder should not be conscripted into that project. The claimant’s burden in the opening scene is not unreal until it is measured. It is institutionally obscured because closure’s grammar does not yet treat it as a governance variable. The first repair is not unlimited capture. It is attributable closure, usable notice, recourse that can be used before depletion, and explicit burden allocation. Later chapters will ask what bounded instruments can aid that work without reproducing the same transfer in audit form.
The distinction between embodied nonfinality and imposed remainder also prepares a boundary that Chapter 7 will need. Institutions cannot become liminal in the mode of embodied subjects. They cannot refuse closure altogether and remain institutions. Persons often must inhabit nonfinality because life exceeds procedural endpoint. A legitimacy doctrine that forgets this asymmetry will either romanticize institutions for “holding complexity” while they continue to externalize burden, or condemn institutions for failing to complete what cannot be completed. This chapter supports neither mistake. It shows that the problem lies in the form of closure and the custody of what closure cannot absorb.
The MiDAS record is therefore best understood here as an exposed mechanism and not as a singular scandal that can be quarantined to a failed automation episode. The automation layer intensified and revealed the pattern by compressing timing, scaling document production, and widening the gap between closure output and contradiction capacity. Yet the underlying transfer is older and wider. Benefits systems, workplace investigations, school discipline procedures, insurance adjudications, platform moderation regimes, and many other closure settings can require people to reverse-engineer the grounds of institutional action while absorbing the temporal and reputational costs of uncertainty carriage. The synthetic age will intensify this structure because it lowers the cost of generating closure forms and justificatory text while leaving attributable custody and recourse usability comparatively expensive. That argument belongs in fuller form to later chapters, but this chapter establishes the threshold condition that those arguments must satisfy. Technology analysis that loses the embodied sequence of exposure is already reproducing the transfer it claims to critique.
I end, then, by restating the methodological debt in operational terms, because this chapter is the first in the book that risks expansion through theory at the price of dulling the line of exposure that justified the inquiry. A concept in this manuscript earns its place only if it clarifies the mechanism by which institutional closure externalizes unresolved remainder into embodied life. Concepts that increase explanatory range while dulling exposure are suspect and must be revised or constrained. Phenomenology is useful here only if it keeps the reorganization of practical world in view. Trauma theory is useful only if it illuminates anticipatory labor without flattening all administrative harm into one category. Affect theory is useful only if it shows how institutional forms position subjects within burdened relations. Predictive neuroscience is useful only if it clarifies metabolic cost without reducing legitimacy to physiology. The opening instance remains the test. If this chapter has done its work, it has made that test harder to evade in the chapters that follow.
Chapter 4
Administrative Settlement and Custody Evasion
Settlement is often misdescribed as what happens after conflict. In administrative life it is more often the form through which conflict is made governable before anyone can honestly say it has been resolved. The term is therefore broader here than negotiated compromise in litigation, though that meaning remains important. I use administrative settlement to name the institutional act of converting an open, burdened, or contested situation into an actionable arrangement by allocating time, evidentiary labor, explanation duties, categories, and risk across persons and offices. A benefits determination, a compliance finding, a fraud classification, a corrective action plan, a payment schedule, a denial with appeal rights, a conditional reinstatement, a deferred review, or a closure letter can all function as settlement in this sense. Each produces enough order for institutions to proceed. Each also distributes what remains unsettled.
The point of beginning with this wider frame is to avoid a recurring distortion in legitimacy analysis. If settlement is treated only as a voluntary bargain reached by parties under relatively symmetrical conditions, then the central administrative question disappears. Most persons subject to institutional closure do not encounter settlement as negotiation among equals. They encounter it as an allocation architecture that arrives in forms, deadlines, reasons, categories, and documentary pathways that prestructure what can be contested, when contradiction is still usable, and which costs must be privately absorbed before any forum will recognize a dispute. In this setting, settlement is not the opposite of adjudication. It is one of adjudication’s most common outputs and one of administration’s most common techniques.
This chapter extends the prior chapter’s argument about embodied remainder by shifting the analytic focus from exposure to allocation. Chapter 3 showed how institutional closure can become embodied remainder through asymmetric exposure, such that a person first experiences a consequence and only later acquires a usable account of the reasoning that produced it. The present chapter asks how that asymmetry is organized and stabilized. Its answer is that administrative settlement functions as a burden allocation event. Institutions may legitimately allocate burden in settlement. They cannot govern without doing so. Custody evasion begins not at the mere fact of allocation, but when allocation is hidden inside procedures presented as neutral, exhaustive, or complete. The decisive distinction, then, is not settlement versus no settlement. It is visible allocation versus invisible allocation.
This distinction requires care because settlement has been criticized from more than one direction, and the chapter needs both lines of criticism while collapsing into neither. Fiss’s classic argument against settlement remains indispensable because it refuses the comforting assumption that settlement is always an unqualified civic good. He insists that adjudication has public functions, including articulation of norms and the disciplining of power, that private settlement can displace or suppress when the resulting agreement is treated as an adequate substitute for judgment (Fiss). Galanter’s account of repeat players and one-shotters also remains central because it shows how asymmetries in experience, resources, timing capacity, and institutional familiarity shape outcomes long before any formal legal principle is applied, making ostensibly consensual or procedural results deeply structured by unequal position (Galanter). Mnookin and Kornhauser, from another angle, illuminate how bargaining takes place in the shadow of formal legal order, which means that settlement is never simply outside institutions but patterned by rules, expectations, and procedural baselines that allocate leverage before a dispute begins (Mnookin and Kornhauser). Administrative settlement, as I use the term, extends these insights beyond litigation proper. The shadow is not only law. It is also the shadow of file architecture, notice design, queue management, categorization systems, and the practical usability of recourse.
To say settlement is a burden allocation event is to specify what institutions do when they close under insufficiency. They allocate not only outcomes but obligations of conversion. Someone must convert facts into recognized evidence. Someone must convert an institutional reason into an answerable claim. Someone must convert a deadline into an action sequence that fits inside ordinary life. Someone must convert categorical designations into narratives that can survive administrative scrutiny. Someone must carry the cost when these conversions fail, arrive late, or cannot be performed in the grammar the institution recognizes. Settlement identifies, distributes, and often conceals these conversion burdens. This is why the chapter treats settlement as an event rather than a static result. The relevant question is not only what the settlement says. It is what work it assigns, to whom, under what timing, and with what visibility.
Four mechanisms dominate the administrative settlement process analyzed here. They are timing, reason giving, categorization, and documentary form. The distinction among them is analytic rather than ontological, since in practice they operate together. I separate them because each hides burden in a different way and each can also be governed more honestly under declared insufficiency.
Timing is the first allocator because it determines whether a person encounters closure as a decision open to contestation or as a consequence already in motion. Administrative systems routinely defend timing choices as capacity necessities, and often they are. Deadlines, queues, escalation protocols, automated triggers, mailing cycles, and review windows are not incidental logistics. They are the architecture through which institutions prioritize finite attention and maintain throughput. Yet timing also decides who bears the cost of uncertainty. If a response window runs before notice becomes usable, if collections proceed before reasons become attributable, if appeals are available only after penalties attach, or if delay itself imposes cumulative depletion that alters what a person can do next, then timing has allocated burden while appearing to be a neutral feature of process design. What matters for custody is not whether every timing burden can be removed. What matters is whether timing is treated as a governance variable and disclosed as an allocation choice rather than hidden as mere administration.
The sociolegal literature on procedure and institutional behavior has long shown that time is not simply a background medium but a form of power. Lipsky’s account of street-level bureaucracy demonstrates how scarcity, queues, and routinization shape frontline governance in ways that become policy in effect, even when never publicly declared as policy (Lipsky). Administrative settlement deepens that insight because timing decisions do not only mediate access to services. They can convert uncertainty into closure by shifting the practical burden of response onto those least able to synchronize their lives with institutional clocks. What looks, from inside the office, like cycle time management can look, from the exposed side, like forced triage among work, childcare, illness, translation, and survival.
Reason giving is the second allocator. Fuller’s emphasis on clarity, publicity, and congruence helps explain why reasons are not ornamental to legitimate governance but part of the reciprocity relation between institution and subject (Fuller). Yet administrative reason giving also functions as a settlement mechanism in a more specific sense. Reasons define the object of contradiction. They determine what must be answered, what evidence is relevant, what forum is triggered, and which future pathways remain open or foreclosed. A reason that is syntactically clear but evidentially thin can still shift enormous investigative labor onto the subject. A reason that names a category without disclosing the factual predicate can force the person to contest a moving target. A reason that is formally attributable but operationally generic can perform accountability while preserving the institution’s discretion to revise its basis without admitting inconsistency. In each of these cases reason giving is not absent. It is allocative.
This is why the chapter distinguishes attributable reasons from usable reasons without treating them as mutually exclusive. Attributability asks whether a decision can be linked to a responsible institutional actor, pathway, or function whose justificatory commitments can be reviewed. Usability asks whether the reason, as delivered, permits a subject to understand what is being asserted, what would count as contradiction, and what sequence of action remains metabolically possible. Institutions can satisfy one condition without the other. A decision can be attributable and still unusable. It can also be legible to insiders and opaque to those who bear the closure’s consequences. Custody evasion often occurs in precisely this gap, where institutions can claim they provided reasons while the burden of rendering those reasons contestably actionable is externalized.
Categorization is the third allocator and often the most underestimated. Categories do not only describe. They route. Bowker and Star’s work remains indispensable because it shows how classification systems embed practical and moral order, distributing visibility, normality, exception status, and administrative tractability in ways that can become invisible precisely because the categories function so routinely (Bowker and Star). In administrative settlement, categorization determines which pathways a person enters, what presumptions attach, which documentation is recognized, what timelines apply, and what remedies become thinkable. A fraud category, an overpayment category, a noncompliance category, a clerical error category, an eligibility category, a risk category, or an emergency category does not simply label a situation. It settles part of the dispute by assigning the grammar in which subsequent burdens will be interpreted.
The allocative force of categorization appears most clearly when categories compress mixed realities into singular administrative identities. A person becomes, for purposes of the file, a claimant, debtor, nonresponder, suspected fraud actor, or high-risk case, while the institution’s own mixed and unfinished role in producing the situation is dispersed across systems, offices, and process stages. This compression is often necessary for administration to proceed. The legitimacy problem arises when category design and category deployment conceal the transfer they effect. If the category carries punitive implications, evidentiary presumptions, or recourse burdens that are not publicly governed as allocation choices, then the institution has settled more than it has admitted.
Documentary form is the fourth allocator and the one that increasingly mediates the other three. Weber’s account of files, offices, and written records remains essential because it clarifies why modern institutions attach authority to documentary form. Documents stabilize memory, coordinate actors, preserve jurisdictional boundaries, and make large-scale governance reproducible (Weber). Administrative settlement relies on these capacities. Yet documentary form also allocates burden through format, provenance, field structure, mandatory inputs, portal design, evidentiary admissibility rules, template language, and record retention practices. A form can require the subject to translate lived complexity into categories the system can process while relieving the institution of reciprocal translation obligations. A portal can count access as notice without verifying practical use. A template can create the appearance of explanation while withholding the factual specificity needed for contradiction. A record can preserve the fact of issuance while losing the conditions of receipt. Documentary form, then, does not merely record settlement. It helps produce it.
At this point the chapter’s central distinction can be made more exact. Visible allocation occurs when an institution discloses, in attributable and reviewable form, what burdens its closure imposes, why those burdens are being allocated as they are, what limitations constrain the closure, and what recourse remains materially usable under those conditions. Invisible allocation occurs when burdens are imposed through timing, reasons, categories, or documentary forms that are treated as neutral process givens or complete outputs, such that the institution secures the benefits of closure while the subject bears the practical work of uncertainty resolution without public recognition that such work has been assigned. The difference is not primarily rhetorical. A visible allocation can still be harsh, and a hidden allocation can still conform to many local procedural rules. The distinction matters because custody evasion begins at concealment, not at severity alone.
It is important here to resist two opposite simplifications. The first says that if allocation is visible then it is legitimate. That is false. Institutions can openly allocate burdens in ways that remain unjustifiable, discriminatory, or politically indefensible. Visibility is a necessary but not sufficient condition for custodial legitimacy. The second says that any burden on the subject proves hidden allocation. That is equally false. Institutions must allocate burdens, including documentation duties, response obligations, and timing constraints, if they are to govern at all. The chapter’s claim is narrower and more exacting. Custody evasion begins when the institution denies, obscures, or naturalizes burden allocations that are in fact constitutive of the closure it is defending.
This distinction also clarifies why administrative settlement can be both more and less coercive than it appears. It can be less coercive than overt command because it often leaves formal options intact, including appeal, hearing, reconsideration, supplemental submission, or later judicial review. It can be more coercive than overt command because those options may be burdened by timing, translation, cost, and informational asymmetries that are neither publicly priced nor institutionally owned. A regime can therefore be richly procedural and still custodially evasive. Indeed, procedural abundance can intensify invisibility by making burden transfer appear as the subject’s failure to navigate available pathways. This is one of the reasons the book cannot treat recourse as a binary variable and later must distinguish nominal from metabolically usable contestability.
Administrative settlement also has a political valence that the chapter must keep in view. Settlement is often praised as an index of pragmatism because it allows institutions to continue operating under conflict rather than waiting for comprehensive resolution. There is truth in that praise. Institutions that refuse settlement may simply shift burdens elsewhere through delay, neglect, or paralysis. Yet settlement can also function as a depoliticizing technology when burden allocation is hidden within technical process language. A closure can present itself as administrative necessity while effectively deciding a distributive question, a dignity question, or a standing question that the institution then denies it has decided. Arendt’s warning about the reduction of politics to administration remains relevant here, but the present chapter adds a different emphasis. The danger is not only depoliticization in the public realm. It is also the privatization of unresolvedness in embodied life, where political burdens are carried as household management, emotional labor, and time loss instead of appearing as publicly governable allocations (Arendt).
The chapter can now return, as required, to the prologue’s line of exposure and test whether this framework clarifies or obscures the mechanism first shown. The diagnostic question is simple. Does the account of settlement as burden allocation explain how a person in the MiDAS context could experience closure first as a tax intercept, wage garnishment, or penalty consequence rather than as a usable accusation, or has the analysis drifted into an abstract theory of administration that no longer sees the transfer event. It clarifies the mechanism only if the answer is concrete. Timing clarifies because response windows and collections sequencing could make institutional clocks govern before discovery became actionable (Cahoo v. SAS Analytics). Reason giving clarifies because notices and determinations could fail to identify the factual basis that claimants needed to answer, thereby assigning investigative labor to the subject while preserving closure’s force (Scott v. Michigan Unemployment Insurance Agency). Categorization clarifies because fraud designations routed claimants into punitive and evidentiary pathways that altered the burden of contradiction. Documentary form clarifies because portal-based questionnaires, notice channels, and record structures shaped what counted as notice, response, and nonresponse, regardless of whether those forms were metabolically usable in practice (Michigan Office of the Auditor General). If these mechanisms are not kept at the level of the lived sequence, the chapter fails its debt. The framework is justified only insofar as it makes the original transfer more visible and more attributable.
That return matters because settlement discourse is peculiarly prone to abstraction. Institutions and scholars alike can become enamored of procedural design and fail to ask who is carrying the gap between formal pathway and practical use. This chapter therefore treats settlement not as a neutral stage between decision and aftermath but as the very site where closure’s burdens are composed, distributed, and frequently hidden. The next chapters will extend the analysis into documentary due process in generative environments and contestability climate, where the same allocative mechanisms operate at greater scale and under more intense theater risk. For now the chapter has to secure one proposition and hold it firmly. Administrative settlement is not suspect because it allocates burdens. It is suspect when it allocates them invisibly and then calls the result neutral process.
The force of this proposition becomes clearer if one follows burden distribution across institutional layers rather than focusing only on the first decision point. A settlement that appears administratively efficient in one office may generate contradiction labor in another, reputational cost in a workplace, debt management burden in a household, and cumulative depletion that later appears as noncompliance or missed recourse. None of these downstream costs need be intentional to be real. Galanter’s repeat-player analysis is useful again here because institutions often have structural advantages in absorbing iterative process costs, while one-shot participants experience those same costs as acute and consequential (Galanter). Administrative settlement can therefore produce asymmetry even where each local step is procedurally defensible on its own terms. This is the beginning of what later chapters will call distributed exclusion, but the mechanism is already visible here in burden allocation form.
The chapter should also acknowledge that institutions can make genuine progress at this level. They can redesign notice so that reasons become usable rather than merely issuable. They can stage consequences to preserve contestation before irreversible penalties attach. They can disclose category effects and evidentiary presumptions. They can build documentary pathways that preserve provenance and receipt conditions rather than only issuance traces. They can publicly state where insufficiency remains and adjust closure scope accordingly. These are not trivial reforms. They can materially improve custodial legitimacy. What they cannot do, and what this chapter has not claimed, is eliminate remainder. Settlement can be made more accountable without becoming complete.
That final point matters because arguments like this one are often misread as demands for infinite process. The chapter has argued the opposite. Legitimacy does not require institutions to preserve all nonfinality or refuse settlement until every uncertainty is resolved. Institutions must close. They must allocate burdens. They must act under insufficiency. What legitimacy requires is that burdens generated by closure be disclosed and contestably governed rather than hidden inside procedures represented as neutral or complete. Administrative settlement becomes custodially defensible when it acknowledges its own allocative work and submits that work to reviewable standards. It becomes custody evasion when it disclaims that work while still taking the authority that only settlement can confer.
The next chapter will test this argument under a different pressure, where synthetic documentary abundance lowers the cost of producing closure outputs and increases the temptation to confuse articulation with attributable custody. But the core lesson belongs here and must remain stable across the manuscript. Settlement is not the end of conflict. It is the institutional distribution of what conflict, uncertainty, and nonfinality will cost, and to whom.
Chapter 5
Documentary Due Process After Generative Text
The prior chapter argued that administrative settlement is a burden allocation event and that custody evasion begins when allocation is hidden inside procedures presented as neutral or complete. That argument now enters a changed documentary environment. The change is not that institutions have suddenly become textual. Institutions have long governed through files, notices, forms, templates, memoranda, and written reasons. The change is that generative systems collapse the marginal cost of producing documentary outputs that look specific, reasoned, responsive, and individualized, while leaving the costs of contradiction, verification, and attributable custody comparatively high. The result is not only a technical shift in office workflow. It is a legitimacy shift in the cost structure of closure.
This chapter therefore asks a narrower question than the broad public debate over artificial intelligence often asks. It does not ask whether generative systems can think, whether they will replace professionals, or whether they should be permitted in the abstract. It asks what happens to documentary due process when institutions can generate closure text faster than affected persons can test it, and faster than institutions themselves can preserve accountable custody over how that text was produced, adopted, and used. In this setting, the danger is not exhausted by error. The danger is laundering. Generative text can launder insufficiency into apparent articulation. It can produce the documentary appearance of individualized attention, reasoned consideration, and procedural completeness while redistributing the labor of verification and contradiction outward onto the person who bears the closure.
The phrase documentary due process names a narrower and more precise concern than fairness discourse in general. Due process in the constitutional tradition has never been reducible to paperwork, and any attempt to make documentary design stand in for political judgment, legal rights, or institutional accountability would repeat the very substitution this manuscript opposes. Yet documentary conditions are not incidental to due process. Notice must be intelligible and timely enough to be meaningful, not merely issued (Mullane v. Central Hanover Bank & Trust Co.). Opportunities to be heard must be practically tethered to the stakes and risks of deprivation rather than abstractly preserved in procedure (Goldberg v. Kelly; Mathews v. Eldridge). Reasoned decision making in administrative law is not satisfied by conclusory text that masks the absence of accountable consideration (Motor Vehicle Mfrs. Ass’n v. State Farm). Documentary due process, in the sense used here, concerns the written and record-based conditions under which institutional closure becomes contestable, attributable, and reviewable in practice.
Generative systems alter those conditions by changing what is cheap and what remains expensive. It has become cheaper to draft notices, summarize records, generate explanatory paragraphs, produce correspondence variants, compose appeal responses, draft investigative timelines, generate hearing summaries, and render procedural language in multiple tones and formats. It remains expensive to verify the factual grounding of each generated sentence, preserve the provenance of the inputs and transformations that shaped it, identify who is accountable for adopting it as an institutional act, and maintain contradiction pathways that do not force subjects to re-investigate the institution’s entire documentary chain in order to contest one adverse action. The asymmetry matters because closure authority travels through the document the person receives, while custody often resides in processes and systems the person cannot see.
The resulting temptation is structural. Institutions under throughput pressure can confuse the availability of better wording with the presence of better reasons. They can treat generated articulation as evidence that a file has been adequately reviewed. They can produce individualized sounding notices at scale and infer that procedural dignity has been increased, when what has actually increased is only the persuasive or pacifying surface of the closure output. They can even improve readability in a local sense while worsening attributable custody in a systemic sense if the path from source evidence to generated reason to final action is not preserved in reviewable form. This is what I mean by documentary laundering. The unresolved insufficiency in the institution’s knowledge, evidence, or deliberation is not eliminated. It is translated into more polished documentary closure and then externalized as contradiction labor for the subject.
The mechanism should not be confused with deception in every instance. Generative systems can produce useful summaries, translations, and drafts, and institutions can use them in ways that genuinely reduce friction. The problem is that the same affordances that support assistance also support custody evasion when they are inserted into adverse or consequential closure pathways without documentary disciplines proportionate to their effect. A generated summary can help an overwhelmed caseworker orient to a file. It can also become the de facto record through which downstream officials and the affected person encounter the case, especially when source materials are voluminous, inaccessible, or procedurally segregated. A generated explanation can improve clarity. It can also create a misleading impression that the institution has considered reasons it has only templated. A generated notice can increase responsiveness. It can also multiply contradictory or shifting formulations across communications, forcing the subject to resolve institutional inconsistency at personal cost.
This is why the chapter does not frame the problem as one of “automation bias” alone, though that literature remains relevant. The central issue is not merely that officials may over-trust machine outputs. It is that generative documentary systems reconfigure settlement’s allocative mechanisms described in Chapter 4. They alter timing by accelerating issuance and response cycles without necessarily improving usable notice. They alter reason giving by multiplying plausible explanations faster than accountable evidentiary mapping can occur. They alter categorization by generating increasingly fine textual distinctions that may obscure stable category effects and presumptions. They alter documentary form by making the document itself an active site of synthesis, compression, and rhetorical closure rather than a relatively passive record container. In each case, the institution’s capacity to close expands faster than its capacity to preserve attributable custody over the closure.
The legal and technical scholarship already contains partial vocabularies for this problem, and the chapter relies on them while shifting their center of gravity. Citron’s account of technological due process remains foundational because it shows how computational systems can degrade accountability when procedural rights are formally preserved but operationally displaced by opaque processing and error-prone infrastructures (Citron). Kroll and colleagues make an equally important contribution by insisting that accountability for algorithmic systems is not only a matter of revealing code but of designing institutional mechanisms that permit verification and review of system behavior in context (Kroll et al.). Burrell’s analysis of opacity in machine learning helps clarify why institutional claims of transparency can be misdirected when they focus on technical internals while leaving the actionable social opacity intact, including the relation between input assumptions, institutional goals, and downstream effects (Burrell). What generative text adds is a documentary acceleration of these dynamics. Opacity is no longer only in the classifier or scoring model. It moves into the reasons, summaries, notices, and records that mediate due process itself.
The scale of this shift is tied to the character of generative language systems. Large language models and related generative architectures produce text by modeling statistical regularities in training data and context rather than by retrieving a stable evidentiary chain for each sentence, which is one reason they can be so productive and also why they can generate fluent but unsupported or distorted outputs (Brown et al.; Bender et al.). Foundation model scholarship has correctly emphasized the breadth of downstream adaptation and the sociotechnical risks introduced by deployment contexts that differ from development assumptions (Bommasani et al.). In administrative documentary settings, this means that the key risk is often not the model in isolation but the institutional composition in which it is used. The same generative component can function as a harmless drafting aid in one workflow and as a laundering device in another, depending on whether the generated text becomes advisory, evidentiary, dispositive, or merely clerical in relation to closure.
This composition problem is precisely where documentary due process becomes urgent. Administrative law and due process doctrine do not generally require institutions to narrate every internal cognitive or clerical step that precedes a decision. Nor should they. Institutions need drafts, internal discussion, and workflow efficiency. The demand at stake here is narrower and more exacting. When generative text is materially involved in producing the reasons, summaries, or record representations through which a consequential closure is justified, the institution must preserve a reviewable distinction between evidentiary custody and textual articulation. Without that distinction, the person subject to closure can no longer tell whether they are contesting the institution’s evidence, its inferences, its summary, its template, its generated wording, or some unstable mixture of all five.
Anti laundering disciplines begin at this distinction. I use the phrase disciplines rather than safeguards because the problem is not only technical control but institutional practice under pressure. The first discipline is documentary role declaration. Institutions must specify, in attributable and reviewable form, what function a generative system performed in a given closure pathway. Did it summarize submitted evidence, draft correspondence, suggest reasons, classify issues, generate translations, produce interview notes, or recommend an action sequence. Without role declaration, generated text can move across functions after the fact, allowing advisory prose to become dispositive rationale while preserving plausible deniability about what the system “really” did. Role ambiguity is a primary laundering channel because it lets institutions enjoy the benefits of synthetic articulation while disclaiming accountability whenever a defect is exposed.
The second discipline is provenance custody for consequential text. This does not require public release of every model parameter, proprietary prompt library, or internal tool configuration in every case. It requires something more practical and more legally salient. When generated text is used in a closure pathway, the institution must preserve a reviewable chain linking the final text to the source materials, the operator or office that initiated generation, the model or system version used, and the human or institutional actor who adopted the text as part of official reasoning. Documentation scholarship on model cards and datasheets has shown the value of structured metadata for accountability at system and dataset levels (Mitchell et al.; Gebru et al.). The present claim extends that logic into adjudicative and administrative documentary practice. Without provenance custody, generated reasons become rhetorically legible but institutionally ownerless.
The third discipline is source-to-reason congruence. Fuller’s concern with congruence between declared rule and official action can be re-read here as a documentary requirement at the level of generated explanation (Fuller). A generated notice or determination may be grammatically clear and stylistically individualized yet still fail due process if the reasons it states cannot be traced to identifiable evidence, standards, and decision criteria actually used in the closure. Congruence in this chapter does not mean exhaustive transparency of institutional cognition. It means that a person contesting closure should be able to identify what proposition is being asserted, on what evidence or record basis, under what governing category, and whether unresolved uncertainty remains declared or concealed. Generative systems are especially dangerous when they produce polished reasons that exceed the institution’s evidentiary custody, because the excess is then externalized as contradiction labor. The subject must prove not only that the decision is wrong but that the institution’s own stated rationale is a synthetic overstatement.
The fourth discipline is action-tier separation. Not every generated text in an institution warrants the same custody obligations. Internal drafting for nonconsequential communications is not identical to text that triggers denial, penalty, collection, classification, or sanction. Documentary due process after generative text requires institutions to differentiate action tiers and prohibit certain forms of adverse closure from relying on nonattributable or nonreviewable generated rationale. Put more plainly, if the institution cannot say who owns the reasons and how they connect to the record, the text should not carry adverse authority. This chapter does not yet formulate the full doctrinal threshold rule, but the direction is already clear and must be stated plainly. In environments where synthetic documentary production is operationally central, non attributable closure becomes a threshold risk rather than a secondary defect.
That sentence foreshadows Chapter 11 because it marks a shift in how defects are classified. In earlier documentary regimes, one might treat weak attribution or vague drafting as a remediable procedural deficiency to be balanced against capacity constraints. Under generative conditions, the same deficiency can become constitutive of the closure’s legitimacy because the institution’s ability to produce persuasive text so dramatically outpaces the subject’s ability to verify, contextualize, and contradict it. The old assumption that wording defects are secondary breaks down when wording is itself the medium through which closure authority is scaled.
The fifth discipline is contradiction parity at the documentary level. Chapter 4 argued that settlement allocates conversion burdens through timing, reasons, categories, and forms. Generative systems intensify this because they can produce multiple layers of explanatory text, summaries, and supplemental communications, each of which may be internally coherent and collectively contradictory. A subject should not be required to resolve institutional documentary inconsistency created by synthetic generation before gaining access to contestation. This means contradiction pathways must attach to the closure record as delivered, including the generated text actually relied upon, and must preserve version history when reasons are corrected, supplemented, or withdrawn. Otherwise institutions can continuously regenerate articulation in response to challenge while treating each new text as clarification rather than amendment, shifting the cost of tracking and reconciliation to the subject.
The sixth discipline is declared insufficiency at the point of synthetic articulation. Generative systems can produce confidence of tone without confidence of basis. The institution must therefore distinguish between what it knows, what it infers, what it cannot yet establish, and what is being provisionally treated as actionable under a declared standard. This is not a request for machine uncertainty scores to be dumped into notices as pseudo-scientific theater. It is a request that institutions not use synthetic fluency to erase epistemic limitation. Declared insufficiency already functions in the book as a constitutional distinction between honest closure and concealed closure. In a generative documentary environment, that distinction becomes operationally documentable. The closure text must not imply evidentiary completion where only procedural progression exists.
The seventh discipline is reviewable adoption, which is different from ceremonial human in the loop language. It is now common for institutions to answer concern about automated or generative systems by asserting that a human reviews outputs. This is often too vague to carry legitimacy weight. Reviewable adoption asks a harder question. What was the reviewer’s function, what materials were available, what aspects of the generated text were checked, what authority did the reviewer have to reject or revise the output, and what record exists of that adoption. A human signature can itself become laundering if it is appended to generated text under throughput conditions that make substantive review implausible. The issue is not whether a human touched the text. The issue is whether accountability and contradiction can attach to the adoption as an institutional act.
None of these disciplines should be mistaken for a comprehensive cure, and the chapter’s argument would become self-undermining if it suggested otherwise. Documentary regimes can satisfy all seven in partial form and still remain custodially weak because the deeper mismatch between closure and lived nonfinality persists. A person may receive a clearly attributed, source-linked, versioned, reviewably adopted denial and still face burdens of time, translation, income instability, caregiving, and reputational injury that the institution cannot absorb. The aim of documentary due process after generative text is therefore not to make closure complete. It is to prevent synthetic articulation from disguising incompleteness while intensifying its transfer.
At this point the chapter must perform its required return to the prologue’s line of exposure and ask whether the technology frame clarifies the original burden transfer mechanism or displaces the person into abstraction. The diagnostic can be stated in the same terms used in Chapters 4 and 6. Suppose the opening scene remains unchanged in structure. A person discovers institutional action through consequence before obtaining a usable account of the reasoning. Does the analysis of generative text sharpen what happened or simply add a fashionable vocabulary. It sharpens only if it makes the cost transfer more precise. Generative systems clarify the mechanism when we see that institutions can now produce more notices, more explanations, more “individualized” correspondence, and more internal summaries at lower cost than they can preserve attributable reasoning and usable contradiction pathways. In that environment the exposed person may receive more text and still less custody. The burden shifts from obtaining any explanation to determining which explanation is operative, grounded, and contestable. If the chapter were to celebrate documentary abundance as procedural improvement without asking that question, it would have drifted from the prologue and failed its debt. The point of the return is to keep the person, not the tool, as the site where legitimacy is tested.
This debt test also reveals why anti laundering disciplines must be designed for both designed failures and distributed exclusions, even though the full distinction is formalized later. Some documentary defects are visible and intentional enough to be named as designed failures, such as undisclosed model use in adverse notices, missing provenance records, or nonreviewable template generation in penalty decisions. Others emerge cumulatively through ordinary workflow composition. A call center uses generated summaries, a caseworker relies on them to draft a notice, a supervisor edits for tone but not evidentiary mapping, a portal displays only the final text, and the appeals unit sees a compressed record that omits earlier generated variants. No single step is obviously abusive. The aggregate can still produce nonattributable closure and recourse simulation. Documentary due process after generative text must therefore guard against both explicit misuse and distributed laundering through routine institutional composition.
The chapter also needs to resist an opposite danger, which is romantic anti-documentarism. Faced with synthetic text, one may be tempted to conclude that documentary reason giving is now inherently suspect and that institutions should return to oral, discretionary, or minimally articulated decision making. That would be a serious regression. Persons subject to closure are usually less protected when institutions can act without leaving reviewable traces. The response to synthetic documentary abundance cannot be documentary retreat. It must be documentary discipline. Weber remains relevant here precisely because the file is still a condition of contestable governance, even if the file now includes generated layers that complicate custody (Weber). The task is not to abandon written administration but to re-specify what counts as attributable documentary action under changed production conditions.
NIST’s risk management frameworks and the broader technical governance literature are useful in this regard, but only if their role is kept bounded. They can support organizational practices of documentation, measurement, and risk review, and the generative AI profile appropriately highlights the need to manage GAI-specific risks across the lifecycle (Tabassi; Autio et al.). Yet no framework of this kind can itself substitute for due process, adjudicative accountability, or custody doctrine. The danger in many institutional settings is framework laundering, where organizations adopt risk management language and documentation routines that improve compliance posture while leaving closure pathways nonattributable and recourse metabolically unusable. This chapter uses risk governance materials as technical companions, not normative sovereigns.
The doctrinal consequence, then, begins to take shape before the doctrine chapter formally arrives. If institutions wish to use generative systems in documentary pathways that bear on consequential closure, they must treat attribution, provenance, and contradiction usability as constitutive features of legitimacy rather than after-the-fact quality improvements. Where those conditions cannot be met, the honest response is not synthetic fluency plus procedural theater. It is declared insufficiency and constrained closure scope. The institution may still need to act, but it must not borrow the confidence of generated text to conceal what it cannot presently justify or govern.
This leaves the chapter with the limit it was always obligated to declare. Documentary reform can matter greatly. Anti laundering disciplines can reduce synthetic opacity, preserve attributable reasoning, and improve the practical conditions under which persons understand and contest closure. They can slow the conversion of generative fluency into administrative authority and make documentary pathways more reviewable. But they cannot abolish the structural mismatch between closure and lived nonfinality. The person who bears institutional remainder still lives beyond the file, beyond the notice, and beyond the revised protocol. Documentary due process after generative text can discipline synthetic closure. It cannot complete what closure, even honestly governed, leaves unresolved.
Chapter 6
Contestability Climate
The preceding chapters have moved from grammar to exposure to allocation. Chapter 1 distinguished closure from custody evasion and established remainder as what closure cannot honestly absorb. Chapter 3 showed how that remainder can become embodied through asymmetric exposure, where the person encounters consequence before usable reasoning. Chapter 4 argued that administrative settlement is a burden allocation event and that the decisive distinction is not allocation versus no allocation, but visible allocation versus invisible allocation. This chapter extends those claims by shifting the unit of analysis once again. The question is no longer only what procedures exist, or what burdens a particular settlement allocates, but what kind of environment institutions produce for contestation as such.
I call that environment contestability climate.
The term is chosen to resist a persistent error in legal and administrative analysis. Institutions often assess contestability by inventory. Is there notice. Is there appeal. Is there reconsideration. Is there hearing. Is there ombuds review. Is there hotline access. Is there judicial review after exhaustion. These questions are not trivial. In some settings they are the difference between domination and any possibility of answer. Yet they remain insufficient because they treat contestation as a pathway property when, in practice, contestation is also a climate property. A pathway can exist and remain unusable. A right can be preserved and remain uninhabitable. A hearing can be offered under conditions that price contradiction beyond the range a person can metabolically sustain. In such cases institutions remain formally recourse capable while functionally noncontestable.
Contestability climate names the aggregate conditions that determine whether a nominal recourse pathway becomes metabolically usable for actual persons under actual burdens. It concerns pace, timing compression, sanction expectation, translation burden, reputational pricing, evidentiary orientation, and cumulative depletion, among other factors, insofar as they shape whether contestation can occur before closure’s consequences become irreversible or too costly to resist. The emphasis on climate does not replace procedure. It re-situates procedure inside the conditions that decide whether procedure can be used.
This distinction between nominal pathways and metabolically usable pathways is the chapter’s first obligation and the one that later doctrine will rely upon most heavily. A nominal pathway is a formally available route for review, contradiction, or reconsideration recognized by institutional rules and documentary forms. It may satisfy important legal baselines and can matter enormously in systems otherwise structured by pure command. A metabolically usable pathway is narrower. It is a pathway that can be traversed without requiring a person to incur levels of temporal, cognitive, affective, reputational, linguistic, or material cost that functionally extinguish the possibility of sustained contestation. The phrase metabolically usable is not metaphor for intensity alone. It carries forward the prior chapter’s argument that institutions can externalize uncertainty into bodies and schedules, such that the feasibility of contestation depends on reserves of attention, time, money, composure, and support that institutions often treat as infinitely available but do not themselves supply.
The distinction should not be mistaken for sentimentalism. It is a governance distinction. If institutions can satisfy contestability by creating nominal pathways whose use presupposes extraordinary reserves, then contestability becomes a legitimating fiction. The institution can continue to describe itself as reviewable while the burden of making review real is privatized. The familiar due process cases already contain the seeds of this problem, even where they do not articulate it in climatic terms. Mullane insists that notice must be reasonably calculated to apprise interested parties, not simply issued into the world (Mullane v. Central Hanover Bank and Trust Co.). Goldberg recognizes that procedural sufficiency in deprivation settings cannot be judged as if timing and stakes were external to fairness (Goldberg v. Kelly). Mathews, for all its balancing framework, makes clear that procedure is evaluated in relation to risk, value, and governmental burden, which means that procedural existence alone is never the whole test (Mathews v. Eldridge). Contestability climate develops a parallel point at the level of institutional analysis. The existence of a path is not the same as the usability of a path.
Climate becomes visible most clearly when separately unremarkable procedural choices are viewed as a single environment rather than as isolated design features. A deadline that looks administratively reasonable in one office combines with a notice that is technically clear but evidentially thin, a queue that introduces uncertainty about response sequencing, a portal that collapses multiple document states into a final display, and a sanctions regime that makes error costly to the subject long before the institution’s own uncertainty has been reduced. None of these features, considered alone, must be malicious to produce a custodially weak climate. The aggregate can still communicate a practical message that is more decisive than any formal statement of rights. Contest if you can, but do so on the institution’s clock, in the institution’s language, against consequences already in motion, and at your own metabolic expense.
Lipsky’s analysis of street-level bureaucracy remains central here because it shows how resource scarcity, routinization, and coping practices become policy in effect, even without legislative declaration (Lipsky). What the climate framework adds is a way of naming how those effects accumulate at the point of contestation. A frontline worker’s shortened interaction, a standardized script, a referral to another office, a requirement to resubmit documentation, or a warning about timelines can each be locally reasonable under conditions of scarcity. Yet in aggregate they can produce a climate in which contestation is theoretically present and practically foreclosed. The problem is not reducible to bad faith at the frontline. It is that institutional systems can generate noncontestability as an emergent property while each actor remains procedurally intelligible within their role.
Galanter’s repeat-player account intensifies the point because contestability climate is not experienced symmetrically (Galanter). Institutions and repeat actors often possess infrastructural familiarity, timing literacy, and process reserves that allow them to inhabit procedural delay or multiplicity as routine administration. One-shot participants, by contrast, experience the same environment as immediate risk. A two-week delay in clarification can be ordinary queue management for the institution and rent destabilization for the person. A request for one more document can be prudent verification for the office and a lost shift, childcare scramble, or reputational explanation for the claimant. The climate concept is therefore not a generalized complaint about difficulty. It is a way of tracing how the same procedural atmosphere distributes cost differently by position.
The chapter’s use of the term climate also carries a second purpose, which is to prevent contestability from being reduced to a single procedural threshold. Institutions often seek legitimacy metrics that can be cleanly counted because counts travel well in management and compliance settings. Number of appeals filed. Number of notices sent. Average response time. Number of hearings held. These measures can be useful. They can also become theater when they stand in for the environment that determines who can use the pathway before depletion or sanction makes the exercise functionally irrational. A high appeal rate can indicate robust contestability in one context and crisis-driven desperation in another. A low appeal rate can indicate satisfaction in one context and climate suppression in another. Contestability climate does not abolish measurement. It changes what measurement must be interpreted against.
The climate itself can be analyzed through the recurring burden vectors already implicit in the prior chapters. Time pressure is the most obvious and perhaps the most often normalized. Institutions treat time as an administrative resource, which it is. Persons experience time as a finite substrate already allocated among labor, care, illness, commuting, and survival. A contestability pathway that presumes discretionary time where none exists is not simply inconvenient. It converts institutional uncertainty into private scheduling crisis. Feeley’s account of process burdens in criminal justice remains instructive beyond that domain because it shows how process itself can become punitive independent of outcome, not only through formal sanctions but through the costs of appearing, waiting, and remaining available to institutions (Feeley). Contestability climate extends that insight across administrative settings. A pathway can punish through the conditions of access even when the final decision is eventually revised.
Sanction expectation is the next vector and one that institutions systematically understate when they describe recourse as available. Contestation does not occur in a neutral psychological field. It occurs under anticipated consequences that may include financial penalty, benefit interruption, intensified scrutiny, reputational marking, employment consequences, or the simple fear that challenging an institution will worsen one’s position. Hirschman remains useful here because voice is never exercised under conditions of cost-free neutrality; the choice among exit, voice, and acquiescence is structured by expected response and institutional loyalty conditions (Hirschman). In many closure settings exit is impossible or catastrophic, which means the meaningfulness of voice depends on whether the institutional climate makes contestation survivable. If subjects rationally expect retaliation, escalation, or prolonged exposure for contesting, the nominal availability of review can coexist with climate suppression of actual challenge.
Translation burden is equally central and often hidden beneath the language of clarity. Chapter 4 showed that reason giving can be attributable and still unusable. The climate framework adds that even usable reasons can require extensive translation labor across legal, administrative, technical, and ordinary life grammars. A person may understand the words of a notice and still not know what the institution is really claiming, what evidence would answer it, what prior document matters, or how to narrate a mixed life event in the categories the system recognizes. Bowker and Star’s work on classification remains relevant because categories do not simply organize information. They structure what can appear as normal, relevant, or answerable within an institutional field (Bowker and Star). Contestability climate becomes thinner where institutions assume that subjects will perform high-order translation between lived experience and classificatory grammar without recognizing that this translation is itself a burden allocated by closure.
Reputational pricing is a further vector that often escapes formal doctrine while shaping practical recourse decisively. Contesting a closure can expose a person to labels and inferences that travel beyond the immediate proceeding. A fraud classification, a noncooperation notation, a performance concern, a compliance flag, or a risk marker can alter how later actors receive the person’s account before the underlying issue has been fairly adjudicated. Goffman’s work remains useful because stigma is not simply a social attitude but an interactional condition that reorganizes credibility and self-presentation under scrutiny (Goffman). In contestability terms, reputational pricing means that the cost of challenge is not limited to time or paperwork. The person may have to contest while already carrying a category that makes contestation itself appear suspicious. Institutions can thereby preserve nominal review while raising the social cost of using it.
At this point the climatic approach can be misunderstood in two ways, and both need to be refused. One misunderstanding treats climate as a diffuse cultural mood beyond governance, as if naming atmosphere were a retreat from institutional accountability. The chapter argues the opposite. Climate is an institutional product, even when no single office designs it in full. It is produced by recurring choices about timing, queue architecture, notice design, sanctions sequencing, classification, staffing, and documentary pathways. The second misunderstanding treats climate as too soft for doctrine because it lacks the crispness of discrete rights violations. This objection carries weight if climate is used as a substitute for threshold analysis. It does not carry weight if climate is used, as here, to diagnose how nominal rights and pathways become nonusable in aggregate. Doctrine already attends to practical effect in many settings. The climate concept extends that attention without dissolving legal specificity.
The most important contribution of the chapter, however, lies in its treatment of cumulative effects. Institutions often defend contested closures by showing that each individual decision point was reasonable within local constraints. The notice met baseline requirements. The deadline followed policy. The queue was due to staffing. The categorization reflected the available record. The portal functioned as designed. The appeal right existed. Any one of these claims may be true. The climate framework asks a different question that no single local defense can answer. What was the cumulative effect of these separately reasonable decisions on the practical possibility of contestation for the person carrying the closure’s remainder.
This cumulative effect is where distributed noncontestability emerges. Distributed noncontestability names a condition in which no single actor, rule, or procedural step fully blocks review, yet the aggregate environment makes contestation functionally unavailable to many subjects. The unavailability is distributed because it is produced across sites. It is noncontestability because the resulting suppression is not merely inconvenience but a practical foreclosure of contradiction before depletion, sanction, or timing renders challenge inert. The concept matters because without it institutions can continuously defend local reasonableness while no one is required to own the aggregate closure climate.
The cumulative mechanism also clarifies why high visibility reforms often fail to produce custodial effect. An institution may improve one pathway element, such as notice readability, hearing scheduling, or portal interface design, and truthfully report procedural progress. Yet if timing compression, sanction expectation, translation burden, and reputational pricing remain unchanged, the climate may remain functionally noncontestable. The reform performs well on a local metric and poorly on the burden distribution the book is tracking. This is not an argument against local reform. It is an argument against confusing pathway repair with climate repair. Institutions can become procedurally busier and climatically colder at the same time.
The chapter’s debt to Chapter 3 remains active here because the climate concept would become empty if it were detached from embodied remainder. The reason metabolically usable pathways matter is that contestation occurs inside finite lives. A pathway can be legally real and climatically unusable because the person must remain employed, care for children, recover from illness, navigate disability, preserve housing, manage debt, or simply continue speaking coherently under conditions of accusation and uncertainty. The prior chapter drew on predictive and stress literatures to show how institutions can impose ongoing regulation demands through opaque closure. This chapter extends that claim into recourse design. Contestability depends not only on formal opportunity but on whether institutions structure challenge in a way that presumes or destroys the reserves required to use that opportunity (McEwen).
Procedural justice scholarship, especially Tyler’s, provides an important but incomplete companion here (Tyler). The literature rightly emphasizes that perceived fairness, voice, neutrality, and respectful treatment affect compliance and legitimacy judgments. Contestability climate affirms that insight while adding a custodial constraint. Perceived fairness can improve while usable contestation remains weak if the institution becomes more courteous, more articulate, and more processually legible without altering timing, burden distribution, or sanction sequencing. In other words, procedural justice can be aesthetically upgraded while climate remains suppressive. This is another route by which anti theater discipline must remain active in the manuscript. Legitimacy signaling is not the same as custodial effect.
The climate framework also permits a clearer relation between Chapters 4, 6, and 8 than the manuscript could previously state. The cumulative noncontestability mechanism described here and the distributed outside developed in Chapter 8 are the same aggregate mechanism operating in different registers. In this chapter the mechanism appears as atmospheric and organizational. It names how the environment of contestation becomes suppressive through accumulations of timing, translation, sanction expectation, reputational pricing, and depletion. In Chapter 8 the same mechanism will appear as jurisdictional and classificatory. It will name how persons or claims are cumulatively pushed beyond meaningful custody through standing rules, evidentiary gates, timing bars, status distinctions, and related devices, even where no single exclusion is sufficient to explain the result. The continuity between these chapters matters because it shows that custody evasion is not only a discrete act. It can be an emergent structural property of procedurally defensible decisions in aggregate.
That continuity must be earned rather than declared, which brings the chapter to its required functional return to the prologue’s line of exposure. The question is whether the climate framework sharpens the reader’s understanding of the original burden transfer mechanism or whether it converts lived difficulty into atmospheric vocabulary without accountability. The answer must remain concrete. In the MiDAS-related exposure pattern described in earlier chapters, the person encounters consequence before usable orientation and then confronts a recourse field shaped by timing compression, punitive classification, documentary opacity, and uncertainty about what counts as a valid answer (Cahoo v. SAS Analytics Inc.; Scott v. Michigan Unemployment Insurance Agency). The climate concept clarifies this only if it helps explain why formal recourse options could coexist with widespread practical difficulty. It does so when it names the aggregate environment in which notice, response windows, penalties, communication channels, and reputational implications interacted to raise the cost of contradiction beyond what many claimants could sustain while maintaining ordinary life. It fails if it remains an atmospheric gloss that no one can operationalize or contest. This repeated return is not rhetorical self-reference. It is a method of disciplined attention, and in that sense it honors Weil’s insistence that thought must be tested by what affliction exposes rather than protected by abstraction (Weil).
The return also clarifies why contestability climate is not reducible to subjective feeling, even though subjective experience is one of its indicators. Climate is inferable from institutional design and outcomes as well as from reported experience. One can examine response windows, sequencing of penalties, average wait times, rate of successful appeals after default consequences, frequency of repeated document requests, translation access conditions, and distributional differences in recourse uptake. Yet every such measure remains partial because climate is relational and adaptive. Institutions can improve metrics while shifting burden elsewhere. They can lower average response times while increasing the evidentiary complexity of notices. They can expand help channels while preserving sanction-first sequencing. They can increase appeals filed while extending final resolution time beyond practical utility. Climate analysis therefore requires both measurement and interpretation, and it must remain alert to how institutions learn to perform contestability.
This is the point at which the anti-performance discipline of the manuscript must be stated again in explicit terms. Any climate measure is theater-prone. Institutions can build contestability dashboards, publish user experience reports, sponsor listening sessions, and revise templates while leaving the aggregate burden distribution substantially unchanged. They can generate a procedural image of openness that itself becomes a barrier to critique because dissenters are then cast as unreasonable in the face of visible reform efforts. The chapter’s purpose is not to produce a new compliance surface called climate. It is to give later doctrine and instruments a way to ask whether pathways can actually be used before depletion and whether cumulative decisions have produced noncontestability that no actor claims.
A further limit also needs declaration. Contestability climate analysis can diagnose pathways and conditions. It cannot, by itself, settle the legitimacy conflict staged in Chapter 2. Different institutional grammars will weigh climate defects differently and may disagree about what level of burden is unavoidable in a given domain. A war powers setting, a benefits adjudication system, a public health emergency response, and a professional licensing regime may all exhibit contestability climate problems under different constraints and with different tolerated risks. The climate framework is therefore a cross cutting diagnostic, not a universal balancing formula. It tells us where contestation is being made functionally unavailable and through what aggregate conditions. It does not by itself determine the full political or legal remedy.
The chapter can now close by securing the proposition it was designed to establish. Contestability is not adequately judged by the existence of nominal recourse pathways. It is a climate property produced by the aggregate institutional conditions under which persons must attempt contradiction, explanation, and review while living with closure’s consequences. A pathway becomes metabolically usable only when the burdens of time, translation, sanction expectation, reputational pricing, and evidentiary orientation are governed in ways that make challenge practicable before depletion or irreversibility forecloses it. Separately reasonable decisions can accumulate into distributed noncontestability that no single actor owns. This cumulative mechanism is not accidental to modern administration. It is one of the primary ways custody evasion persists under procedural defensibility.
The next movement will carry this analysis to the mismatch and perimeter chapters, where the same aggregate mechanism appears as the production of outsides, some designed and some distributed. For now the key result is narrower and harder. Institutions can be formally recourse capable and climatically noncontestable at the same time. Any legitimacy claim that ignores that possibility is already misdescribing the burden it asks others to carry.
Chapter 7
Liminality Without Capture
By the time this manuscript reaches the present chapter, a certain temptation has become almost irresistible. After tracing hidden burden transfer through closure, after showing how documentary form can launder insufficiency, and after arguing that nominal recourse can coexist with climatically suppressed contestation, one begins to want institutions that do not close so quickly, institutions that can remain with uncertainty rather than convert it into defensible force. In contemporary administrative and design discourse, this desire is often expressed through the language of liminality. We ask whether institutions can hold ambiguity, whether procedures can remain open longer, whether systems can create space for transition, emergence, complexity, or becoming. The desire is understandable. It is also dangerous if handled without precision.
This chapter begins from that temptation because it is the right temptation. It is right in what it refuses. It refuses the fantasy that every conflict can be fully adjudicated into peace, that every discrepancy can be conclusively known, that every body can be brought into administrative legibility without remainder. It refuses the recurring institutional habit of treating nonfinality as a defect in subjects rather than a condition of life. It refuses the violence by which institutions preserve their own image of coherence through premature or concealed closure. If the earlier chapters did not induce this desire for a more patient and less extractive relation to uncertainty, they would have failed.
Yet the chapter cannot stop with desire, because liminality is one of the most easily romanticized terms in the contemporary theoretical lexicon. Its anthropological origins are disciplined and specific. Van Gennep’s account of rites of passage situates liminality within structured transitions marked by separation, threshold, and incorporation, with obligations, authorities, and forms that make the threshold socially held rather than merely indefinite (van Gennep). Turner expands the account by showing how liminality can suspend ordinary status arrangements and generate forms of communitas, but he also emphasizes that liminal states are not simply free play. They are patterned, bounded, and often governed by ritual authority, danger, and sequence (Turner, The Ritual Process). Once liminality is abstracted from these constraints and imported into institutional theory as a generalized virtue of openness, the term can become a cover for very different realities, including prolonged uncertainty, delegated burden, and suspended accountability.
The prior chapters have already given us the resources to see why. Institutions secure practical authority by closing under insufficiency. Persons live through nonfinality whether institutions admit it or not. These are not symmetrical conditions. A person can remain in unfinishedness because a person is a living, embodied, relational being whose life exceeds any one procedure. An institution that attempted to inhabit liminality in the same mode would cease to function as an institution and would likely redistribute its indecision as intensified burden on those subject to it. The problem, then, is not solved by asking institutions to become liminal. The problem is to determine what a legitimate institution can owe to nonfinality without pretending to embody it.
This is the mismatch the chapter has to work through, and it must do so without evasion. The mismatch appears at several levels at once. Anthropologically, liminality concerns transformations of status and social form under bounded conditions that do not map cleanly onto routine administration. Phenomenologically, nonfinality names the fact that lived experience exceeds procedural endpoint and that institutional closure often arrives in the middle of ongoing life rather than at any natural completion. Politically, institutions are authorized to act, and delay or indecision can itself become a mode of domination when consequences continue while decision responsibility remains suspended. Juridically, contestability requires pathways and standards, which means some closure forms must exist if recourse is to be anything more than an appeal to institutional mood.
One can now see why the modern appetite for “holding space” inside institutions requires a more exact vocabulary than it often receives. Winnicott’s account of transitional phenomena and potential space has been enormously generative for thinking about intermediate zones where play, symbolization, and relation can occur without immediate foreclosure (Winnicott). In therapeutic and developmental registers, this insight can be humane and illuminating. In institutional settings, however, the translation is not straightforward. A welfare agency, a court, a licensing board, a university disciplinary office, or a benefits adjudicator cannot simply become a transitional object relation for the public. Institutions act through categories, deadlines, and enforceable effects. If they borrow the language of holding without specifying burden distribution, accountability, and closure conditions, they may produce not transitional care but suspended exposure.
It is nevertheless important not to retreat too quickly into anti-liminal suspicion. Institutions that deny all nonfinality generate their own pathologies. Scott’s account of legibility remains useful because it shows how state and administrative projects can produce violence when they force living complexity into simplified schemes and then treat the scheme as reality (Scott). Bowker and Star, in a different register, show how classification systems distribute consequence by deciding what can count, what can travel, and what becomes error or residue (Bowker and Star). These critiques matter here because they explain why institutional capture of liminality is a problem. When institutions insist on total capture, they do not abolish remainder. They relocate it into bodies, households, and communities while preserving documentary order. The earlier chapters have already shown this mechanism in closure, settlement, and contestability climate. Liminality enters now as the name for what institutions repeatedly encounter but cannot absorb without loss.
At this stage, the chapter might seem to be moving toward a familiar resolution. We might say that institutions should recognize liminality, build more flexible procedures, create protected thresholds, and honor transitional states without forcing premature closure. There is truth in each of those impulses, and later chapters will indeed propose bounded instruments and doctrines that discipline closure under insufficiency. But the chapter is under a stricter compositional obligation than synthesis, and the argument cannot be responsibly completed from concept alone. It has to remain answerable to a pressure the chapter did not script in advance.
That pressure emerged, for me, not from a philosophical text but from a line of administrative adjudication that appears at first to support the chapter’s opening desire. In Social Security disability litigation, federal courts often confront records in which agency adjudication has been procedurally deficient, factually incomplete, or legally erroneous, and the ordinary remedial instinct is to remand for further proceedings rather than order immediate benefits. This looks, at first glance, like institutional humility. The court resists premature closure. It preserves factual development. It acknowledges insufficiency. It gives the system another opportunity to decide correctly under law. If one approached the present chapter from theory alone, one could take these remands as a model of institutional nonfinality properly held.
The arrest occurs when one follows what repeated remand means in lived time.
In Benecke v. Barnhart, a case involving Social Security disability benefits and fibromyalgia, the Ninth Circuit declined another remand for further proceedings and ordered benefits, emphasizing the years of delay already imposed and warning against a pattern in which the agency effectively receives repeated chances to deny while the claimant continues to wait under conditions the court recognized as already overlong (Benecke v. Barnhart). What stopped the chapter’s earlier line of movement was not simply the outcome, but the form of the court’s reasoning. The opinion did not say that more process is always worse or that remand is categorically illegitimate. It exposed a limit in the very remedy that had looked, in the abstract, like principled openness. There comes a point at which preserving institutional nonfinality no longer protects accuracy or fairness but instead transfers the cost of institutional insufficiency into a person’s life through delay, uncertainty, and depletion.
This is not a prepared paradox. It is a category arrest. The chapter had been moving toward a defense of institutional openness to liminality, and the case shows a setting in which “more openness” can function as custody evasion if it takes the form of serial procedural extension borne by the subject. The point is not that the court solved the entire problem. Ordering benefits in one case does not abolish the mismatch between administrative adjudication and lived nonfinality. The point is that the case forces a distinction the chapter had not yet stated with enough force. Institutions may need to acknowledge what they cannot know or absorb, but they cannot offload that acknowledgment as indefinite procedural exposure for the person.
Once this arrest is taken seriously, the governing principle of the chapter can emerge in the sequence it needed. The right lesson from liminality is not that institutions should become liminal in the mode of embodied subjects. The right lesson is that institutions must govern at the threshold without claiming threshold possession. They must close where closure is required for action, rights, payment, protection, or review, but they must do so under declared insufficiency, with attributable reasoning, and with contestable pathways that do not punish persons for inhabiting the nonfinality institutions cannot absorb. The demand is neither total capture nor institutional self-mystification as “holder of complexity.” It is disciplined closure under custodial limits.
This principle can now be stated in its constitutional negatives and bounded positive, and it must be stated plainly because the rest of the manuscript will depend on it. Institutions cannot abolish nonfinality without violence because living subjects, social worlds, and historical injuries exceed administrative endpoint. Institutions cannot convert all remainder into metrics and records and still claim legitimacy because the act of conversion itself redistributes burden and can erase the very exposure it claims to govern. Institutions cannot become liminal in the mode of embodied subjects while remaining institutions because authority, allocation, and enforceable action require closure forms that embodied life does not. Legitimate institutions close under declared insufficiency, preserve contestable pathways, and refrain from punishing subjects for inhabiting nonfinality.
Each sentence in that sequence carries a pressure the prior chapters have prepared. The first sentence answers the legibility temptation diagnosed by Scott and the classificatory overreach traced by Bowker and Star. The second answers the documentary laundering and theater risks intensified by generative systems. The third answers the chapter’s own initial temptation to redeem institutions through liminality discourse. The final sentence gives the bounded positive that the doctrine chapter will later formalize. The wording matters because the positive claim remains bounded. Institutions are not asked to redeem nonfinality. They are asked not to punish it, not to hide what closure exports, and not to treat their own insufficiency as a private cost borne publicly.
The arrest produced by Benecke also clarifies why the language of “staying open” can be so politically unstable across domains. In some contexts, people demand open pathways because institutions close too fast and too opaquely. In others, people demand closure because institutions have kept them in procedural suspension while consequences continue to accumulate. These are not contradictory demands if one tracks burden distribution. Both are demands against custody evasion. One resists premature closure that conceals insufficiency. The other resists indefinite nonclosure that converts institutional insufficiency into lived waiting. The criterion that allows us to distinguish them is not an abstract preference for openness or decisiveness. It is whether the institution’s way of handling uncertainty distributes burden in attributable, reviewable, and metabolically survivable form.
Turner becomes more useful again at this point, though in a stricter way than organizational theory often permits. Liminality in Turner is dangerous, transformative, and socially consequential, but it is not an indefinitely privatized condition administered by remote files. It is held by forms, sequences, and authorities that acknowledge the threshold as threshold (Turner, The Ritual Process). Modern institutions often produce the opposite. They create threshold conditions without threshold acknowledgment. Subjects are suspended, classified, reviewed, deferred, flagged, reprocessed, or remanded, yet the institution describes each step as ordinary procedure rather than as burdened transit. The result is not liminality in the anthropological sense. It is liminality without social holding and without shared obligation, which is precisely why the term becomes analytically and politically useful only when paired with custody.
This pairing also prevents another mistake, which is to romanticize refusal of legibility as if all noncapture were emancipatory. For many subjects, noncapture means not safety but abandonment. Hartman’s work on the afterlives of domination and archival violence is instructive here because what escapes institutional capture is not thereby protected from harm; it may instead be rendered unredressable, unwitnessed, or structurally unavailable to public remedy (Hartman). The chapter therefore cannot equate “preserving remainder” with simply leaving people outside records or metrics. The task is not to celebrate invisibility. It is to prevent institutions from claiming total custody while making remainder ungovernable for those who carry it. In some settings this requires more record, more reason, more attributable process. In others it requires less extraction, less forced narration, and narrower closure claims. The bounded principle is intentionally exacting because the direction of repair depends on how burden is being transferred.
Arendt’s political vocabulary helps secure the final turn of the chapter. If legitimacy concerns the maintenance of a common world in which judgment and action remain possible, then institutions must neither collapse politics into administration nor abandon administration to procedural mystique (Arendt, Between Past and Future; Arendt, The Human Condition). The mismatch this chapter identifies is not a defect to be solved by importing liminality as a management style. It is a standing condition of political and administrative life in which closure and remainder remain co-present demands. Institutions owe actionable decisions and reviewable reasons. Persons owe no obligation to become finally legible to the institutions that govern them. Where institutions require such final legibility as the price of recognition or recourse, they punish nonfinality and thereby convert a condition of life into a defect of the governed.
The chapter can now state the governing principle for the rest of the manuscript in the form required by the architecture. Every instrument proposed after this point is a closure instrument under critique. None may claim to absorb remainder. This sentence is not a disclaimer added for modesty. It is a constitutional consequence of the arrest. Once the chapter has recognized that institutional attempts to “hold liminality” can become deferred punishment, and that institutional attempts to eliminate liminality can become violent capture, every subsequent proposal must be designed and judged within that interval. Audits, recourse assessments, safety cases, documentary protocols, doctrine itself, all remain closure instruments. They may discipline closure, expose burden transfer, and preserve contestability. They do not complete what they govern.
That constitutional consequence is easiest to ignore precisely when institutional design becomes most sophisticated. A highly articulated doctrine can begin to resemble a final moral technology. A well-designed instrument can generate the appearance that remainder is now measured, managed, and safely housed in governance. The chapter’s arrest is a warning against that seduction. If a doctrine makes closure more reviewable but leaves subjects carrying the same undeclared burdens, it has improved visibility without custodial effect. If an instrument reduces one designed failure while intensifying distributed suspension, it has rearranged remainder rather than absorbed it. If a system speaks the language of liminality, trauma awareness, or complexity while extending procedural exposure in lived time, it has converted ethical vocabulary into administrative weather.
The relation to the previous chapters can now be seen with sharper resolution. Chapter 3 named asymmetric exposure. Chapter 4 named hidden burden allocation in settlement. Chapter 5 named documentary laundering under generative conditions. Chapter 6 named contestability climate and distributed noncontestability. The present chapter does not replace those mechanisms with a new theory of thresholds. It identifies the mismatch that makes them recurrent. Institutions repeatedly encounter nonfinality that they cannot absorb, and in trying to govern anyway they are tempted either to deny remainder through closure theater or to export remainder through suspended process. Liminality without capture names the discipline of refusing both errors.
It is important, before closing, to say what this chapter has not established. It has not given a domain-general rule for when closure must occur rather than remand, extension, or further review. It has not solved the political problem of who can force institutions to declare insufficiency rather than hide it. It has not yet specified the doctrinal thresholds by which hidden uncertainty transfer, nonattributable closure, or metabolically unusable recourse become nonderogable failures. It has not answered how grammar identification will be publicly reviewed when institutions strategically shift justificatory language. Those tasks remain. The chapter has done something prior and indispensable. It has narrowed the field of acceptable error in the manuscript’s own reasoning. The book may no longer appeal to liminality as a cure for closure, and it may no longer imagine closure as a cure for liminality.
The final movement of the chapter returns, briefly and deliberately, to the line of exposure that has disciplined the middle chapters. A person discovers the consequence before the reason. The institution can answer this fact in at least three ways. It can deny the significance of the sequence and insist the file is complete. It can multiply process and documentary articulation while leaving the person to carry the cost of waiting, translation, and contradiction. Or it can close under declared insufficiency, make reasons attributable, preserve metabolically usable recourse, and refrain from treating the person’s ongoing nonfinality as procedural defect. The first response is capture through concealment. The second is capture through suspended exposure. The third is not redemption. It is legitimacy under custody.
That is enough for this chapter, and it is all it should claim. The chapters that follow may now propose doctrine and instruments, but they do so under a condition this chapter has forced into the open. The threshold cannot be possessed. It can only be governed with declared limits, reviewable burdens, and a refusal to make subjects pay for the institution’s inability to complete the world.
Chapter 8
The Outside of Custody
The prior chapter established a limit that the manuscript can no longer evade. Institutions cannot become liminal in the mode of embodied subjects, and efforts to govern uncertainty by either total capture or indefinite suspension each produce their own forms of custody evasion. That result now requires a different analytic object. If institutions close under insufficiency and if nonfinality persists beyond closure, then the legitimacy question is not only how institutions decide, explain, and permit contestation. It is also how they produce an outside.
I use the phrase outside of custody to name the persons, claims, times, harms, and evidentiary materials that fall beyond meaningful institutional responsibility while remaining materially shaped by institutional action. This is not simply an outside of law, and it is not a romantic outside of freedom. The outside analyzed here is often manufactured inside governance itself. It is produced through jurisdiction, classification, timing, standing, documentary practice, and procedural sequencing. Institutions do not need to expel a person entirely from legal order to place that person outside meaningful custody. They need only configure the conditions under which burdens remain real while review, remedy, or attributable obligation become functionally unavailable.
The chapter therefore treats perimeter production as political anatomy. The metaphor is deliberate. Perimeters are not abstractions floating over institutions. They are drawn, maintained, revised, and inhabited through organizational forms and public authority. They distribute visibility, legibility, risk, and recourse. They decide which unresolved burdens count as governance problems and which are privatized as misfortune, noncompliance, personal failure, or background condition. The outside of custody is not what institutions simply fail to reach. It is often what they reach through consequence while declining to reach through obligation.
The argument proceeds through two interlocking modes because the outside is produced in more than one way. The first mode is the designed outside. Here exclusions are intentionally produced through jurisdictional narrowing, standing restrictions, emergency classifications, evidentiary gates, contractual waivers, status distinctions, timing bars, and related mechanisms that can operate as custody evasion instruments. The second mode is the distributed outside. Here no single rule appears abusive and no single actor owns the exclusion, yet a sequence of separately legitimate classifications, timing decisions, burden allocations, and review thresholds cumulatively pushes persons or claims beyond the reach of meaningful custody. The distinction matters, but the deeper point of the chapter is that both modes can generate the same practical result. A person remains governed by effects while ungoverned by remedy.
The designed outside is easiest to identify and therefore easiest to condemn, but it is not always illegitimate. Institutions require boundaries. Jurisdiction cannot be infinite. Standing cannot dissolve every distinction between public concern and adjudicable injury. Evidentiary gates cannot be abandoned without destroying the possibility of reasoned decision. Timing limits can be necessary to preserve finality, coordinate administration, and prevent endless re-opening. Emergency classifications can be unavoidable where institutions face urgent threats and incomplete information. The chapter does not treat perimeter drawing as suspect by default. It treats perimeter drawing as an exercise of public power whose legitimacy depends on whether its burden distribution is attributable, reviewable, and contestably governed.
That condition is more exacting than familiar legality tests because the designed outside can be perfectly lawful in local form while still functioning as custody evasion in aggregate effect. Standing doctrine offers a clear example. A system may require injury, causation, and redressability as conditions for adjudication, and those requirements can serve real institutional functions, including separation of roles and prevention of generalized managerial judging (Lujan v. Defenders of Wildlife). Yet standing rules also allocate who may convert diffuse or structural harm into reviewable claim. If an institution or legal order narrows standing in ways that systematically block those who bear cumulative burdens from contesting the closure regimes that generate them, then the perimeter is doing more than sorting docket competence. It is deciding whose remainder must remain private. The legitimacy issue is not exhausted by whether the rule is doctrinally familiar. It includes whether the rule’s perimeter effects are acknowledged as burden allocation rather than naturalized as neutral architecture.
The same holds for evidentiary gates. Every adjudicative system must decide what counts as admissible, sufficient, timely, or reviewable evidence. These decisions are not optional. They are constitutive of adjudication. Yet evidentiary gates can also become custody evasion instruments when they are structured around the institution’s preferred records while discounting the forms in which affected persons can actually document lived burden, sequence, and exposure. Bowker and Star help make this visible because classification and record systems do not simply order information. They decide which traces travel and which traces disappear into noise (Bowker and Star). In custody terms, an evidentiary regime can preserve high procedural regularity while producing a perimeter around institutionally legible proof, leaving embodied remainder real but non-adjudicable. The harm is not that all experience is not admitted whole. No system can do that. The harm is that the evidentiary boundary is treated as pure necessity when it is also a distributive choice about whose forms of knowledge can enter closure and recourse.
Emergency classifications sharpen the same problem because they intensify both necessity and abuse potential. Agamben’s work on exception is valuable here not because every emergency measure confirms the strongest thesis of sovereign abandonment, but because it forces attention to the way extraordinary classifications can suspend ordinary procedural expectations while preserving the appearance of lawful order (Agamben). Foucault, from another direction, shows how security rationalities govern populations through variable thresholds, circulation management, and differential tolerances of risk rather than through a single juridical command structure (Foucault). Read together, these traditions help clarify a recurring modern pattern. Emergency and security classifications rarely remove governance. They redistribute it. Some subjects are governed through heightened intervention, others through deferred protection, and others through procedural thinning in the name of urgency. The designed outside in such settings is often not a zone of no power. It is a zone of asymmetrical power where burdens remain enforceable while custody obligations are narrowed.
Status distinctions operate similarly and often more durably. Arendt’s reflections on statelessness and the fragility of the “right to have rights” remain indispensable because they expose how formal membership categories mediate access to public world, remedy, and recognition rather than simply describing pre-political belonging (Arendt, The Origins of Totalitarianism). Mamdani and Ngai extend this insight in distinct historical registers by showing how colonial and immigration regimes institutionalize differential personhood through administrative categories that are not incidental to governance but central to it (Mamdani; Ngai). The relevance to this chapter is direct. A status distinction can place a person inside institutional reach for surveillance, extraction, or sanction while placing the same person outside robust custody for review, relief, or recourse. This is perimeter production by design, and it is often defended as ordinary administrative sorting. The chapter’s claim is not that all status distinctions are invalid. It is that legitimacy analysis must treat status as a burden-allocation device rather than as a neutral precondition of procedure.
Timing bars and jurisdictional narrowing complete the picture of the designed outside because they convert temporal and forum boundaries into substantive distributions of who can still matter. Chapter 6 already argued that time pressure can make nominal pathways metabolically unusable. The present chapter adds that timing can also produce a perimeter in stricter legal and administrative terms. A filing deadline, exhaustion requirement, issue-preservation rule, waiver doctrine, or jurisdictional transfer can be locally intelligible and still operate as a custody boundary when combined with notice deficiencies, translation burdens, or sanction-first sequencing. The outside produced by timing is especially difficult to contest because institutions can describe exclusion as a consequence of temporal order rather than as a burden allocation decision. Yet a regime that counts institutional issuance time while discounting discovery time, or that recognizes formal availability while disregarding practical sequence, has already drawn a perimeter around whose delay counts as governance and whose delay counts as fault.
Contractual waivers and procedural consent mechanisms create another designed outside that modern institutions increasingly rely upon. Here too the chapter avoids a simple denunciation. Waiver, consent, and contractual allocation can be legitimate instruments of coordination and risk management. They become custodially suspect when institutions treat formal assent as complete transfer of burdens that subjects could not meaningfully evaluate, negotiate, or metabolically refuse. In such settings the perimeter is drawn through the fiction that closure was voluntarily privatized. The issue is not whether signatures were obtained. It is whether institutions are using contractual form to convert public burden allocation into private assumption while continuing to shape outcomes through asymmetric power, information, and sequencing.
If the chapter ended here, it would have identified important perimeters and still missed the mechanism that most often defeats reform. Designed exclusions attract attention because they can be named, litigated, and symbolically targeted. They also permit institutional response in the language of correction. Revise the rule. Expand standing. Extend the deadline. Modify the evidentiary threshold. Clarify emergency criteria. Improve waiver disclosures. Each reform can matter. Yet institutions frequently remain custodially weak after such reforms because the outside is also produced in a second mode that does not depend on any single exclusion.
The distributed outside emerges when cumulative decisions, each separately defensible, push persons or claims beyond meaningful custody in aggregate. This is not simply a softer version of the designed outside. It is a different anatomy of responsibility. No single office can be shown to have intended exclusion. No single rule can be isolated as the decisive barrier. No single procedural defect fully explains the loss of recourse. Instead, a person moves across an environment of classifications, queues, handoffs, evidence demands, timing expectations, communication gaps, and reputational markings that together produce an outcome more exclusionary than any one component appears to authorize. The person is not formally outside the system. The person is moved through it in ways that yield an outside as result.
This chapter must now state clearly what the previous chapter prepared and what later recourse analysis will depend on. The cumulative noncontestability mechanism described in Chapter 6 and the distributed outside described here are the same aggregate mechanism operating in different registers. In Chapter 6 the mechanism appeared as atmospheric and organizational. It explained how timing pressure, sanction expectation, translation burden, reputational pricing, and depletion could suppress contestation even when pathways remained nominally open. In the present chapter the same mechanism appears as jurisdictional and classificatory. It explains how standing limits, documentation thresholds, issue-preservation practices, category routing, timing bars, and forum boundaries can cumulatively move persons or claims beyond meaningful custody even when each procedural step is locally defensible. The continuity is not rhetorical. It is diagnostic. Custody evasion is therefore not only a discrete act. It can be an emergent structural property of procedurally defensible decisions in aggregate.
This is the point at which much institutional self-description becomes analytically misleading. Institutions often respond to exclusion claims by demanding the identification of a single wrongful act, a single offending rule, or a single bad actor. That demand makes sense in some legal settings because liability and remedy often require specification. It becomes distorting when imported into legitimacy analysis as if aggregate exclusion were unreal until personified. Galanter’s repeat-player framework helps explain why this demand recurs. Institutions are organized to fragment process into routinized encounters and role-bounded decisions, which allows them to absorb cumulative effects as normal administration while one-shot participants experience the same accumulation as total environment (Galanter). Lipsky adds the frontline dimension by showing how coping practices under scarcity become policy in effect without formal declaration (Lipsky). The distributed outside names what appears when these institutional strengths of fragmentation and routinization are viewed from the side of burden carriage.
Eubanks’s work on automated and data-driven welfare governance provides a particularly clear contemporary illustration of this structure, not because technology is uniquely culpable, but because it reveals how multiple procedural and classificatory decisions can converge to produce exclusion without a singular dramatic denial event (Eubanks). A screening system, a documentation requirement, a scheduling rule, a database mismatch, a referral threshold, and an appeal window can each be defended in isolation. Their composition can still produce a perimeter around those least able to perform the translation and time labor the system requires. The same is true well beyond welfare administration. Labor classification regimes, school disciplinary systems, housing enforcement, immigration processing, insurance adjudication, and platform moderation can all generate distributed outsides through cumulative burden transfer rather than explicit expulsion.
The political consequence is severe because distributed outsides are difficult to contest using institutions’ preferred evidentiary forms. Designed exclusions can sometimes be shown through text. A rule says no. A category applies. A deadline passed. A forum lacks jurisdiction. Distributed exclusions often require synthetic demonstration across time, offices, and records, which means the burden of proving exclusion can itself reproduce the exclusion. A person must show not one closure but a pattern of closures, not one denial but an ecology of delays and reroutings, not one misclassification but a chain of category effects. Institutions can then redescribe the pattern as anecdotal accumulation or unfortunate complexity. The outside persists because the proof form required to name it exceeds what many subjects can gather while living under its consequences.
This is one reason the chapter insists that perimeter production is political anatomy rather than technical defect. The distributed outside is not an accidental residue that disappears once enough process engineering is applied. It expresses a deeper feature of institutional governance under complexity and scarcity. Institutions preserve actionability by partitioning responsibility, stabilizing categories, and narrowing review tasks. These are often necessary organizational moves. They also create conditions under which cumulative exclusion becomes easy to produce and hard to own. The legitimacy question is therefore not whether institutions can operate without perimeters. They cannot. The question is whether perimeter production is rendered visible, attributable, and revisable, especially where burdened subjects remain governed by consequences generated inside the perimeter.
The chapter now needs to say more directly why high-visibility reforms so often fail to produce custodial effect, since the next chapter on recourse economy will depend on this diagnosis. High-visibility reforms usually target designed blockages because designed blockages are legible, litigable, and communicable. An institution revises a notice template, extends a filing deadline, creates a new review office, broadens a category, adds an ombuds channel, or narrows the scope of an emergency classification. These reforms can produce real gains. Yet if distributed exclusions remain intact, the reform’s custodial effect may be low despite high symbolic value. The pathway exists but the climate remains suppressive. The forum opens but the evidentiary chain remains fragmented. The deadline extends but sanction-first sequencing still depletes the person before contestation matures. The classification changes but status-based distrust and translation burden continue to shape uptake. Institutions then report procedural progress while the outside is reproduced through composition.
This is not an argument against visible reform. It is an argument for perimeter-aware reform. A doctrine or instrument that can detect only designed failures will repeatedly overestimate improvement because it cannot see the aggregate mechanisms that reconstitute exclusion after local repair. This is why the appendices, later in the manuscript, are constitutionally required to state whether they are better at detecting designed failures than distributed exclusions and what that asymmetry means for interpretation. It is also why Chapter 11 will need a threshold stage and grammar-sensitive balancing that can name hidden transfer and metabolically unusable recourse without collapsing into checklist proceduralism. The outside of custody cannot be reduced to rule audit alone.
The chapter’s own argument must remain bounded, however, and this is where self-limitation becomes part of the substance rather than a ritual of modesty. No doctrine eliminates the outside. No legal order, administrative regime, or institutional design can bring every burden, every remainder, every historical injury, and every emergent exclusion fully inside accountable custody. Arendt’s lesson about political belonging, Foucault’s lesson about governmental distribution, and the sociolegal lesson about repeat-player asymmetry each point in this direction from different angles. Every governance form draws lines, allocates attention, and leaves some matters unresolved or under-attended. The problem is not the existence of an outside as such. The problem is the denial, concealment, or naturalization of perimeter production where institutional closure and institutional effect are actively creating the burdens that then appear outside remedy.
A responsible doctrine therefore does not promise perimeter abolition. It makes perimeter production more visible, contestable, and revisable. Visibility, in this sense, means that institutions must disclose the classificatory, jurisdictional, temporal, and documentary boundaries through which closure obligations are limited, together with the burden distributions those boundaries predictably generate. Contestability means that affected parties and reviewing bodies can challenge those boundaries as burden-allocation decisions rather than merely submit to them as neutral architecture. Revisability means that perimeter effects are not treated as fixed necessities when evidence accumulates that they are generating systematic custody evasion, whether by design or through distributed aggregation.
The chapter also closes a gap that could otherwise mislead the reader about political agency. To say that the outside is emergent in distributed form is not to say it is impersonal fate. Emergence does not absolve institutions of responsibility. It changes the form responsibility must take. Where no single rule or actor owns the exclusion, institutions remain responsible for building review mechanisms capable of seeing aggregate effects, for refusing reform metrics that count only local throughput, and for acknowledging when procedurally defensible decisions produce custodially indefensible perimeters in combination. If they refuse this responsibility, emergence becomes a shield for custody evasion.
The relation to the prologue’s line of exposure remains indirect here but still decisive. The person who first encounters consequence before reason is already at risk of being moved toward an outside if subsequent pathways demand forms of proof, timing, and translation that the person cannot sustain. The earlier chapters tracked that risk through settlement, documentary form, and contestability climate. The present chapter names the perimeter result when those mechanisms accumulate. The person is still acted upon, still classified, still timed, still recorded. What falls away is not governance but custody.
That is the chapter’s central claim and the condition under which the next chapter must proceed. The recourse economy can only be understood if we first understand that recourse failures are not exhausted by visible blockages. Some arise from designed exclusions. Others arise from distributed outsides built out of procedurally defensible decisions in aggregate. Any reform that sees only the first will spend legitimacy on symbolism and leave burden where it was.
Bracing
Chapter 9
The Recourse Economy
The prior chapter argued that the outside of custody is produced in two interlocking modes, one designed and one distributed, and that both can yield the same practical condition in which persons remain governed by consequence while ungoverned by meaningful remedy. That result now forces a sharper claim than the manuscript has yet made. If closure is to remain legitimate under conditions of declared insufficiency, recourse cannot be treated as an auxiliary virtue, a reputational enhancement, or a residual safeguard appended after institutions have already optimized for throughput. Recourse is a first order institutional budget.
I use the term budget advisedly and against its narrower administrative usage. The chapter is not speaking only of appropriations, though appropriations are part of the problem and often the public face of it. Recourse, in the sense required by this book, depends on a budget of time, attention, contradiction capacity, evidentiary custody, translation support, adjudicative patience, reversible sequencing, and institutional tolerance for being challenged by those over whom closure has already been exercised. An institution budgets for recourse whenever it decides how much contradiction it can be made to absorb before acting, after acting, and against its own preferred account of the case. Institutions that refuse to budget for recourse do not thereby escape recourse costs. They externalize them.
This is why the present chapter is the incompatibility chapter. Earlier chapters developed diagnostics, distinctions, and debt tests. This chapter must now name combinations that cannot coexist with meaningful recourse, however polished the surrounding procedure appears. The argument is not that every institution must maximize challenge at all times or that every closure must remain indefinitely reopenable. Institutions must act, allocate, and sometimes finalize. The claim is more exacting and more bounded. Certain recurring arrangements are incompatible with recourse adequacy because they preserve the image of review while making contradiction practically unavailable, temporally inert, or documentary ownerless. They do not merely weaken recourse. They counterfeit it.
Recourse adequacy, as the chapter uses the term, is therefore not measured by pathway count alone. It is measured by whether contradiction can occur in time, against attributable reasons, under conditions that do not demand extraordinary depletion from those carrying closure’s burdens. This carries forward the distinction between nominal and metabolically usable pathways developed in Chapter 6, while giving it a more institutional and allocative register. A recourse system is adequate when the institution has budgeted enough contradiction capacity to permit challenge before consequence hardens into irreversibility, and when the costs of mounting that challenge have not been privatized beyond practical use. It is inadequate when the institution advertises review but prices access to contradiction in currencies the exposed person cannot sustain.
The economic language matters because institutions habitually speak as if recourse were an external demand imposed on otherwise rational administration. In practice, recourse is one of the core terms on which administrative rationality is constituted. Throughput without recourse is not efficiency in any meaningful legitimacy sense. It is closure acceleration. A regime that can issue thousands of determinations, denials, classifications, or sanctions but cannot absorb contradiction except as backlog, complaint, or litigation has not solved the problem of governance. It has shifted the contradiction budget outward onto households, workers, claimants, and downstream institutions. Lipsky’s account of street level scarcity already showed how rationing by queue, script, and coping routine becomes policy in effect (Lipsky). The recourse economy argument extends that insight by insisting that contradiction capacity is itself rationed, and that its rationing must be treated as a distributive choice rather than a background fact.
Hirschman remains indispensable at this juncture because the relation between exit and voice clarifies what institutions often misdescribe as user choice (Hirschman). In many closure settings, exit is unavailable, exorbitant, or itself a form of injury. A claimant cannot simply exit unemployment adjudication without forfeiting benefits. A worker cannot always exit a disciplinary regime without losing income and reputation. A tenant cannot easily exit an enforcement process without housing consequences. A person classified as ineligible, noncompliant, or high risk may be unable to exit the category’s downstream effects. When exit is constrained in this way, voice is not a discretionary feature of institutional life. It is a condition of non-domination. A recourse economy that budgets too little for voice under conditions of constrained exit is not austere. It is structurally coercive.
The chapter’s first incompatibility can now be stated with the force it requires. Contradiction priced beyond metabolically usable range is incompatible with meaningful recourse.
Pricing here includes money but is not limited to money. Boddie and later decisions in the access to justice line already recognized that formal rights can be nullified when fees or procedural costs block entry to forums needed for fundamental interests (Boddie v. Connecticut; M.L.B. v. S.L.J.). Those cases matter because they expose a constitutional intuition broader than filing fees alone. A review pathway can exist in law and fail in use when the cost of entry or continuation exceeds what the affected person can bear. The present chapter extends that intuition across administrative and institutional settings where the decisive price is often mixed. It may include wage loss from repeated appearances, childcare disruptions, transportation, document retrieval fees, translation costs, time off work, reputational explanations to employers, emotional regulation under accusation, or the sheer labor of reconstructing an institutional chain from fragmented communications. Feeley’s account of process as punishment remains relevant because it demonstrates how process burdens can punish independently of final disposition (Feeley). In recourse terms, pricing contradiction beyond usable range does not simply deter frivolous challenge. It suppresses challenge as such while preserving the institution’s claim that challenge was available.
The metabolically usable qualifier is not rhetorical excess. It prevents the chapter from collapsing into a universal demand for low-friction procedure. Some recourse burdens are inevitable and sometimes necessary. Institutions may require deadlines, forms, evidence, and appearances. The incompatibility arises when those burdens are set or sequenced such that the practical exercise of contradiction requires reserves that the institution knows many affected persons will not have and has chosen not to provision or accommodate. At that point recourse becomes a sorting mechanism for surplus capacity rather than a governance obligation. The institution can then redescribe low uptake as acceptance, nonresponse, or claimant failure, while the actual determinant is the price structure of contradiction.
The second incompatibility follows directly from Chapters 3 and 6. Nominal review coupled with temporal depletion is incompatible with custodial legitimacy.
The phrase temporal depletion names more than delay, and the distinction is important. Delay can sometimes protect fairness by allowing evidence gathering, reconsideration, or de-escalation. Temporal depletion occurs when institutional timing and sequencing consume the period during which review would still be practically meaningful, while consequences continue to operate against the person. The institution can then point to review in form while the subject experiences review as a process whose utility expires before completion. This is the temporal analogue of nominal pathways in a suppressive contestability climate. The pathway exists. The life in which it could matter has been depleted.
The due process canon already contains scattered recognition of this problem. Goldberg is often read for pre-termination hearings and welfare entitlements, but what gives the opinion enduring force is its refusal to pretend that timing is separable from deprivation in a subsistence context (Goldberg v. Kelly). Mathews allows more balancing flexibility, yet its framework still turns on risk, value, and burden in a way that makes sequence and timing unavoidable considerations (Mathews v. Eldridge). Outside constitutional doctrine, the social security remand line discussed in Chapter 7, including Benecke, shows a related limit from another angle. A formally available remand can become a mechanism of depletion when repeated procedural extension transfers the cost of institutional insufficiency into years of waiting borne by the claimant (Benecke v. Barnhart). The recourse economy framework generalizes the point. An institution that sequences review after penalties, after irreversible losses, or after prolonged depletion has budgeted for review theater rather than recourse adequacy.
This incompatibility is one reason the chapter insists that recourse must be understood as a budget and not only as a right. Rights language is indispensable, and the manuscript does not abandon it. Yet institutions can affirm rights while starving the capacities that make those rights usable in time. They can preserve appellate channels while underfunding adjudicators, preserving sanction first workflows, fragmenting case records across offices, or requiring repeated re-entry into the same process after each provisional decision. The right remains. The recourse budget is absent. When this occurs, legitimacy claims become fragile in a particular way. The institution can still narrate itself as reviewable. Those who carry closure’s remainder experience review as a temporal consumption mechanism.
The third incompatibility carries forward Chapter 5 and must be stated in the sharper register the outline demands. Documentary proliferation without attributable custody is incompatible with contestable closure.
This incompatibility is becoming more central under generative and synthetic documentary conditions, but it is not confined to them. Institutions have long been able to multiply documents, notices, summaries, memos, and procedural artifacts in ways that increase textual volume without increasing accountability. What has changed is the speed and scale at which plausibly reasoned documentary outputs can now be produced, revised, and recirculated. As argued in Chapter 5, this alters the cost structure of legitimacy by making closure text cheaper than contradiction, verification, and provenance. The recourse economy consequence is now direct. When closure is justified through proliferating documents that cannot be reliably tied to responsible actors, stable evidentiary chains, or reviewable adoption decisions, the burden of restoring documentary coherence is imposed on the person seeking recourse. Contestation then becomes a prior contest over what the institution itself has done.
Administrative law has long resisted one version of this through the rule against post hoc rationalization. Chenery insists, at least in principle, that agency action must be judged on the grounds the agency actually invoked, not on later arguments invented to defend the result (SEC v. Chenery Corp.). State Farm similarly requires reasoned explanation tied to the agency’s action rather than conclusory assertion (Motor Vehicle Manufacturers Association v. State Farm). These doctrines matter because they preserve a rudimentary form of attributable custody in reasons. The recourse economy argument extends their logic beyond formal judicial review settings and into the broader documentary environments where closure now occurs. If institutions can proliferate explanations, templates, summaries, and generated notices without preserving attributable custody over which reasons are operative and why, then recourse degrades into interpretive labor imposed on the governed. The person is required to reverse engineer the institution’s closure grammar before any substantive contradiction can even begin.
The incompatibility claim here is not technophobic and not anti-documentary. Documentary articulation remains essential to reviewable governance. The incompatibility arises when institutions substitute documentary abundance for accountable reasoning and then count the abundance as procedural adequacy. In such settings, more text can mean less recourse because each additional artifact expands the subject’s burden of reconciliation. A closure regime that cannot provide attributable custody over consequential documentary action may still function as administration. It cannot honestly claim contestable closure.
These three incompatibilities are the chapter’s spine, but they do not yet explain why reform so often misfires. For that we must re-enter Chapter 8’s distinction between designed and distributed outside, because recourse failures arise in both modes and are often mistaken for only one. Some recourse failures are designed blockages. A statute narrows standing. A rule imposes a short filing deadline. A policy requires exhaustion through a channel known to be unusable. A waiver clause strips review rights in practice. A penalty attaches before any contradiction opportunity exists. These failures can sometimes be identified, litigated, and revised with relative clarity. They are visible enough to anchor high profile reform.
Other recourse failures are distributed exclusions. No single rule bars review. No single office denies a hearing. No single actor intends to foreclose challenge. Yet timing compression, category routing, staffing limits, documentary fragmentation, repeated handoffs, unclear reasons, reputational pricing, and sanction expectation combine so that the practical capacity to contest is exhausted before review becomes useful. This is the same aggregate mechanism Chapter 6 named as distributed noncontestability and Chapter 8 named as the distributed outside, now viewed through the budget lens of recourse. The institution has not formally abolished recourse. It has built a recourse economy in which contradiction capacity is too scarce, too late, or too ownerless to function.
The distinction matters because recourse reform that addresses only designed blockages while ignoring distributed exclusions will predictably produce high symbolic value and low custodial effect. This warning has to be stated in stronger terms than the manuscript has previously used because the pattern is now recurrent across domains. Institutions revise visible rules, improve forms, or add appellate bodies and then declare progress, while the recourse economy remains structurally hostile to those without surplus time, documentation literacy, or social insulation. The result is legitimacy inflation without custodial repair. Reform succeeds in the language of architecture and fails in the lived sequence of challenge.
A common institutional response to this diagnosis is to invoke scarcity as if it were an external law. We would provide fuller recourse, the response goes, but the caseload is too large, the staffing too thin, the budget too constrained, the risk environment too severe, the demand too volatile. Some of this is undeniably true. Scarcity is real, and no serious theory of institutions can proceed as if contradiction capacity were unlimited. The recourse economy framework does not deny scarcity. It changes the normative meaning of how scarcity is used. Scarcity does not excuse invisible rationing of recourse. It requires publicly contestable rationing. If recourse is a first order institutional budget, then institutions must declare how contradiction capacity is being allocated, what forms of challenge are being prioritized, what burdens are being shifted to subjects, and what insufficiencies remain. Otherwise scarcity becomes a standing alibi for custody evasion.
This point connects the chapter back to Chapter 2’s legitimacy conflict without attempting premature synthesis. A Weberian administrative grammar may emphasize calculability and role order. A Fullerian grammar may emphasize congruence, publicity, and reciprocity. An Arendtian grammar may emphasize the maintenance of a common world in which judgment remains possible. Administrative pragmatism may emphasize throughput and governability under conditions of volume. The recourse economy argument does not replace these grammars. It introduces an adjudicative test across them. Whatever grammar governs closure, a regime that prices contradiction beyond usable range, couples review to temporal depletion, or proliferates documentary closure without attributable custody is failing at recourse in a way that directly bears on legitimacy. The weighting and thresholds will later be grammar sensitive. The incompatibilities are prior.
The chapter must also address a more subtle objection, one that often appears among institutional reformers rather than institutional defenders. The objection says that recourse adequacy should be pursued primarily by making challenge easier at the margins through user centered design, better interfaces, plainer language, and procedural nudges. These interventions can be valuable and the manuscript does not dismiss them. The problem is that they often operate at the surface of the recourse economy while leaving the budget structure untouched. A clearer portal does not cure sanction first sequencing. A friendlier notice does not create attributable custody for proliferating reasons. A better chatbot does not solve documentary ownerlessness. A navigable appeal form does not make the pathway metabolically usable if the review arrives after the relevant life function has collapsed. User centered recourse can therefore become another high visibility reform that improves encounter quality while preserving contradiction scarcity.
None of this means institutions are doomed to recourse failure or that recourse adequacy is immeasurable. It means measures must be tied to the economy rather than the inventory. Institutions can ask how often review occurs before consequences harden, how often attributed reasons remain stable across challenge, how often appeal success follows default or depletion rather than timely contradiction, how burdens differ by status and resource position, how many documentary revisions occur without clear provenance, and how often claimants abandon pathways at points predicted by cost spikes rather than by merits resolution. Such measures remain partial and susceptible to theater, as Chapter 6 warned. Yet they are closer to the budget question than pathway counts alone.
The chapter’s relation to political economy should now be explicit. Recourse costs are often displaced not only onto individuals but onto other institutions and publics. When primary systems under-budget contradiction capacity, the remainder reappears as crisis line volume, emergency assistance demand, civil legal aid overload, judicial backlog, employer conflict, household debt, and public distrust. This recirculation is one reason recourse adequacy is not a narrow administrative luxury. It is a distributional and civic question. Galanter’s repeat player analysis helps explain why institutions can tolerate such displacement longer than one-shot participants can bear it, since the institution experiences the costs as spread and routinized while the subject experiences them as concentrated and decisive (Galanter). The recourse economy concept makes visible what that asymmetry allows institutions to hide. They are not economizing in the abstract. They are deciding where contradiction will be paid for.
The chapter can now advance the stronger reform warning in the exact form the architecture requires. Recourse reform that addresses only designed blockages while ignoring distributed exclusions will produce high symbolic value and low custodial effect. The sentence is intentionally severe because the temptation to declare success after visible reform is one of the main ways institutions preserve legitimacy under conditions of persistent burden transfer. Designed blockages should be repaired. The warning is not a counsel of purity. It is a counsel against misdescription. If distributed exclusions remain, the outside of custody will be regenerated through timing, classification, documentation, and depletion even after the visible barrier has been removed.
This is also where the chapter must avoid its own overreach. Naming incompatibilities does not by itself supply a full reform blueprint. It clarifies what arrangements cannot coexist with meaningful recourse, and it sharpens how the next chapter on safety cases for persons should evaluate institutional claims of risk governance and protective design. It does not yet tell us how recourse budgets are to be politically secured in each domain, who will force declaration of contradiction rationing, or what coalition can convert diagnostic clarity into institutional redesign. Those questions belong partly to doctrine, partly to politics, and partly to the final coda.
The chapter therefore closes by naming in advance the gap that the final movement of the book must not evade. Establishing incompatibility does not by itself make revision politically available. One can show, rigorously and publicly, that contradiction is priced beyond usable range, that review is coupled to temporal depletion, or that documentary closure lacks attributable custody, and still find that the institutions in question retain legal insulation, political support, budgetary immunity, or administrative path dependence sufficient to preserve the existing recourse economy. The diagnostic remains necessary because without it institutions can continue calling closure reviewable when it is not. But necessity is not availability. The difference must be named now so that the coda can ask the right question later, not whether incompatibility has been proven, but what forms of collective action, redesign, leverage, and public judgment can make recourse adequacy politically buildable where it is presently denied.
The result of the chapter is therefore intentionally hard and incomplete. Recourse is not a peripheral safeguard but a first order institutional budget. Institutions routinely under-budget it while preserving procedural images of review. Some failures arise through designed blockages. Others arise through distributed exclusions that no single actor owns. Certain arrangements are incompatible with meaningful recourse and should be named as such, not balanced away by administrative self-description. Yet even this clarity does not itself produce reform. It produces a truer account of what institutions are doing when they close and a more disciplined basis for judging what they owe before closure can still be called legitimate.
Chapter 10
Safety Cases for Persons
The preceding chapter established that recourse is a first order institutional budget and that certain arrangements are incompatible with meaningful recourse, even when institutions retain the documentary and procedural appearance of review. That result leaves the manuscript at a dangerous and productive point. It is dangerous because the argument has now named incompatibilities without yet supplying an instrument that institutions can use to discipline closure prospectively, before burden transfer hardens into ordinary administration. It is productive because this is precisely the point at which a bounded transfer from another domain becomes possible.
The transfer at issue is the safety case.
In high hazard industries, a safety case is not an abstract promise to be careful and not simply a collection of compliance artifacts. In its strongest form, it is a structured, reviewable argument by a duty holder that identifies hazards, states the claims being made about risk control, specifies the evidence supporting those claims, discloses assumptions and limits, and submits the adequacy of that architecture to external scrutiny before catastrophe is treated as the primary teacher. The safety case tradition took recognizable form in part through hard lessons such as Piper Alpha, where post hoc explanation of failure could not substitute for prospective demonstration of how hazards had been understood, controlled, and governed under uncertainty (Cullen). Leveson’s systems work and Rasmussen’s analysis of risk management in dynamic societies make the deeper point even clearer. Failures in complex systems are rarely exhausted by component fault. They emerge through interactions among organizational incentives, control structures, degraded feedback, and the drift of practice under pressure (Leveson; Rasmussen). A safety case, at its best, disciplines these dynamics by forcing institutions to articulate what could go wrong, what controls exist, what evidence supports confidence, and where uncertainty remains.
It is easy to see why this discipline attracts the present project. Contemporary person affecting closure regimes also operate under uncertainty, complexity, and throughput pressure. They also generate harm through interactions among classification, timing, documentation, delegation, and review structures rather than through singular dramatic acts alone. They also normalize drift, local rationality, and distributed responsibility. A burdened claimant, worker, student, tenant, patient, or resident can be exposed to severe consequence through an institutional control structure that looks procedurally intelligible at every local node and remains catastrophically weak in aggregate. The attraction of safety case reasoning is therefore real and should not be minimized. It offers a way to move from retrospective blame toward prospective burden of justification. It offers a way to force institutions to show their hazard understanding before they claim the authority of closure.
Yet the transfer is also perilous, and the chapter has to state the peril before it states the instrument. The safety case tradition emerged in domains where hazards are often modeled in relation to physical systems, operational controls, and tolerability frameworks oriented toward accident prevention. Person affecting closure regimes are not offshore platforms, aircraft systems, or chemical plants. Persons are not components. Rights bearing subjects are not risk nodes. Institutional legitimacy is not reducible to a reliability target. If safety case discipline is imported into person affecting settings without this boundary, the transfer will reproduce the very custody evasion the manuscript has been diagnosing. Institutions will begin speaking the language of safety while reclassifying political conflict as risk management and converting questions of rights, standing, burden distribution, and public accountability into technical documentation exercises.
This is the category error the chapter must avoid. A safety case for persons is not a case about the person as an object of risk. It is a case about the institution’s closure pathway as a source of foreseeable harm to persons. The object of discipline is the institutional act under insufficiency, not the subject’s supposed propensity for failure, fraud, error, escalation, noncompliance, or instability. This distinction is the chapter’s constitutional boundary, and if it is breached the entire transfer collapses into administrative paternalism or audit theater.
Once that boundary is secured, the bounded transfer can be stated with precision. What may be borrowed from the high hazard tradition is not an acceptable loss calculus for human dignity and not a technocratic fantasy of complete control. What may be borrowed is the discipline of prospective justificatory structure under declared insufficiency. The institution that seeks to impose consequential closure must show, in a reviewable form, that it has identified foreseeable hazards of that closure, specified the controls it claims will prevent or limit those hazards, preserved attributable evidence for those claims, declared the uncertainties it cannot resolve, and maintained recourse and reversal conditions proportionate to the stakes. The point is not to certify innocence in advance. The point is to shift the burden from post hoc institutional explanation to ex ante institutional justification.
That shift matters because person affecting closure regimes recurrently present their harms as unfortunate exceptions discovered only after implementation. The manuscript’s prior chapters have shown why this presentation is often false or incomplete. Hidden uncertainty transfer, nonattributable documentary closure, temporal depletion, and distributed noncontestability are not rare anomalies that appear only after deployment. They are foreseeable hazards of institutional design under pressure. A safety case for persons becomes useful precisely where institutions can no longer plausibly say they did not know that such hazards existed in kind, even if they could not predict each instance in detail. The task is to require institutions to show how they have addressed known hazard classes before closure is exercised at scale.
To say this rigorously, the chapter has to define the instrument more carefully than reform discourse typically does. A person centered safety case is a structured, attributable, and publicly reviewable argument, produced by the institution seeking to exercise consequential closure, that its closure regime can operate under declared insufficiency without crossing specified nonderogable hazards and with declared controls for foreseeable burden transfer pathways, together with monitored recourse and revision conditions. Each part of this definition is doing work.
It is structured because institutions otherwise diffuse justification across policy memos, training decks, legal boilerplate, vendor claims, and post hoc explanations, which makes contradiction burdensome and accountability evasive. It is attributable because Chapter 5 has already shown that documentary proliferation without custody destroys contestable closure. It is publicly reviewable because Chapter 2 established that legitimacy grammar cannot be left to unilateral institutional declaration. It is produced by the institution seeking closure because the burden of justification belongs first to the duty holder, not to those who bear the closure’s risks. It is under declared insufficiency because no institution can honestly certify complete knowledge of social complexity, and synthetic confidence is one of the principal laundering mechanisms this manuscript opposes. It is bounded by nonderogable hazards because some failures cannot be traded away through balancing. And it includes monitored recourse and revision conditions because a static assurance document in a dynamic administrative environment becomes obsolete quickly and often theatrically.
The chapter can now say what hazards such a safety case must treat as first order in the framework of this book. The first is hidden uncertainty transfer, where the institution converts unresolved indeterminacy into closure while externalizing the burden of uncertainty resolution into embodied life without disclosure. The second is nonattributable closure, where consequential action cannot be reliably tied to accountable reasoning, evidence custody, and reviewable adoption. The third is recourse collapse through pricing or sequencing, where contradiction exists in form and fails in use because the institution has budgeted too little time, support, or reversible sequencing. The fourth is temporal depletion, where review pathways mature only after consequence has consumed the practical utility of challenge. The fifth is documentary laundering, intensified in generative text environments, where articulation outpaces attributable evidentiary custody. The sixth is perimeter drift, where designed or distributed exclusions move persons or claims outside meaningful custody while the institution continues to govern by effect. The seventh is extractive measurement, where institutions demand escalating self-documentation or exposure from subjects in the name of safety or fairness without demonstrating that the extraction has custodial effect. These hazards are not exhaustive, but they are not optional in a person affecting closure regime that seeks legitimacy.
The significance of naming these hazards prospectively is not that it guarantees their elimination. High hazard engineering never promised that safety cases abolish accidents, and the strongest traditions in the field are explicit about residual risk, uncertainty, and organizational drift (Reason; Leveson; Rasmussen). The significance is that institutions must no longer be allowed to treat these hazards as accidental byproducts discovered only after persons have already borne them. A safety case for persons changes the temporal order of justification. Before closure is scaled, the institution must show what it knows these hazards to be, what controls it claims to have, what evidence supports those controls, and what recourse and reversal pathways remain if those controls fail.
The language of controls is itself treacherous in person affecting settings, and the chapter must discipline it. In engineering contexts, controls may include redundancy, alarms, interlocks, barriers, shutdown logic, operator training, and maintenance regimes. In person affecting closure, controls are not mechanical analogues but institutional design constraints. They include attributable reason giving, provenance-preserving documentary practices, sanction sequencing that preserves meaningful contradiction before irreversible harm, recourse budgets calibrated to metabolically usable pathways, independent review with authority to halt or narrow closure, category governance that discloses presumptions and routing effects, and change control procedures that preserve the reviewability of revisions. The transfer is functional, not literal. The institution is not asked to mechanize persons. It is asked to identify and govern the pathways by which its own closure practice can harm them.
This functional translation makes Leveson especially useful and also in need of modification. Her systems theory of accidents rejects simple component blame and focuses on control structures, feedback loops, and unsafe control actions (Leveson). The analogy to administrative closure is strong. Harm often emerges not from one malicious official or one broken rule but from the interaction of automated triage, documentation practices, category thresholds, staffing pressure, and deferred review. A safety case for persons can borrow this systems orientation to identify unsafe institutional control actions such as imposing sanction before attributable reasons are delivered, issuing generated documentary closure without provenance custody, routing claims into punitive categories without disclosing evidentiary predicates, or permitting repeated remand and review extension without temporal depletion analysis. What cannot be borrowed unmodified is the implicit tendency to treat optimization and control sufficiency as the central normative task. In person affecting governance, the institution’s control objective is itself part of the legitimacy question and may require political contestation rather than technical refinement.
That last sentence marks the first severe capture pressure on the instrument. Institutions can adopt risk language in order to displace conflict over rights, standing, and distributive burden into an auditable vocabulary of control effectiveness. Power’s analysis of audit culture remains indispensable because it shows how verification practices can become organizational rituals that produce assurance value, reputational protection, and manageability even when substantive performance remains uncertain or unchanged (Power). Porter’s account of trust in numbers deepens the warning. Quantified and standardized forms of justification often travel as neutral and impersonal, which is part of their administrative appeal, yet that very impersonality can conceal whose judgments, values, and exclusions have been embedded in the metric and whose burdens are rendered commensurable for management purposes (Porter). A person centered safety case is acutely vulnerable to this capture. An institution can produce hazard registers, assurance diagrams, scoring matrices, and review dashboards that display procedural seriousness while preserving the same custody evasion structures in practice.
The danger is not hypothetical. The manuscript’s prior chapters have already documented the ingredients of such capture. Chapter 4 showed how burden allocation can be hidden inside neutral-seeming procedure. Chapter 5 showed how documentary abundance, especially under generative conditions, can launder insufficiency into plausible articulation. Chapter 6 showed how institutions can perform contestability while preserving a suppressive climate. Chapter 8 showed how designed repairs can leave distributed outsides intact. Chapter 9 showed how recourse can be under-budgeted while review is still advertised. A safety case discipline inserted into this environment can become yet another layer of procedural rhetoric unless the instrument is constitutionally constrained by the very hazards it is supposed to govern.
This is why a person centered safety case cannot be a compliance dossier alone. It must be adversarially inhabitable. The institution’s claims about hazard control must be framed so that affected parties, independent reviewers, and downstream adjudicators can identify what is being claimed, what evidence is offered, what assumptions are doing hidden work, and where declared insufficiency remains. Jasanoff’s work on regulatory science and public reason is instructive here because expertise in governance is not discredited by contestation. It is constituted by the conditions under which claims can be scrutinized, challenged, and revised in public and institutional forums (Jasanoff). A safety case for persons that is technically rich and publicly inert will fail in the exact way the manuscript has tracked across other domains. It will increase institutional articulation while externalizing contradiction labor onto those least able to carry it.
The instrument is also prone to a subtler form of capture that the chapter must name more sharply. High hazard traditions often include tolerability reasoning, proportionality judgments, and practicability language, which can be appropriate in domains where institutions must manage residual risk under finite resources. If imported carelessly into person affecting closure, these logics can become a legitimating language for the managed sacrifice of recourse and burden asymmetry. Institutions may begin to argue that a certain rate of wrongful classification, opaque closure, or delayed review is “reasonably practicable” in light of throughput demands, or that temporal depletion is a tolerable cost of system efficiency. At that point the safety case is no longer a discipline on closure. It is a grammar for pricing persons. The manuscript cannot accept that transfer. Some hazards in person affecting closure are not ordinary tradeoff variables. They belong in the nonderogable threshold zone that the doctrine chapter will formalize.
This is the chapter’s second constitutional boundary. Safety cases for persons may structure balancing only after threshold failures are excluded. They cannot legitimate hidden uncertainty transfer, nonattributable closure, or recourse priced beyond metabolically usable range by assigning them low probabilities or high mitigation costs. The moment the instrument permits that move, it becomes a laundering device for exactly the incompatibilities Chapter 9 identified.
A third capture pressure comes from scale and vendorization. Institutions increasingly procure risk and compliance systems, decision support tools, and documentation platforms from external vendors who can supply templates, metrics, and assurance language at speed. In such settings a safety case program can be adopted as organizational posture while actual hazard understanding remains shallow and heavily outsourced. The risk is amplified in generative documentary environments, where institutions may use synthetic systems to draft the very safety cases meant to govern synthetic closure. The result can be recursive procedural theater. The institution appears to have a sophisticated assurance architecture while the underlying evidentiary custody, recourse budgeting, and burden distribution remain unchanged. Chapter 5’s anti laundering discipline therefore applies with full force here. If a safety case document itself is produced through nonattributable or unreviewable synthetic processes, the instrument is compromised at the point of articulation.
A fourth capture pressure concerns distributed exclusions. Safety case disciplines are often better at detecting designed failures than distributed ones because the method encourages explicit hazard statements, identifiable controls, and traceable evidence. This is a strength and a weakness. It is a strength because many institutional harms are indeed produced by designed sequences, explicit thresholds, and known workflow choices that should be surfaced and governed. It is a weakness because distributed outsides and distributed noncontestability frequently emerge from interactions no single office owns and no single document captures. A person centered safety case that reports only designed hazard controls may therefore improve local reliability while leaving aggregate custody evasion largely intact. The chapter must say this plainly because the appendices later will be required to disclose this asymmetry for every instrument. Safety cases for persons are likely to detect designed failures more readily than distributed exclusions unless they include explicit aggregate testing obligations tied to contestability climate, perimeter effects, and recourse economy outcomes.
The bounded schema now comes into focus more fully. A person centered safety case worthy of the name must begin by specifying the closure regime under review and the stakes of consequence, including what forms of deprivation, sanction, classification, or reputational impact the regime can impose and at what scale. It must then identify foreseeable hazard pathways in the manuscript’s sense, including burden transfer, documentary laundering, temporal depletion, and perimeter production, rather than limiting hazard identification to system downtime or clerical error. It must state the controls claimed for each hazard in attributable and reviewable form, including controls on timing, category governance, documentary provenance, recourse sequencing, and reversal authority. It must present evidence custody rather than assertion alone, which means the institution must preserve the record by which it knows what it claims to know. It must declare insufficiency, including data limits, conceptual uncertainty, expected blind spots, and domains where the regime’s closure claims are intentionally narrowed. It must define recourse and revision triggers, including what counts as a near miss, what indicators signal recourse climate suppression, and who has authority to halt, narrow, or redesign the closure pathway. It must provide an adversarial entry point for challenge from affected parties and independent reviewers. And it must preserve change history so that revisions to controls, categories, or documentary practices remain reviewable rather than disappearing into iterative updates.
This may sound exacting. It is exacting. High hazard safety case disciplines were exacting precisely because the costs of ungoverned confidence were catastrophic. The present argument is that person affecting closure can also be catastrophic, albeit in dispersed and administratively normalized forms. Wrongful deprivation, prolonged procedural exposure, reputational marking, and nonreviewable classification may not appear as a single explosion, but they can produce cumulative life damage that institutions are structurally incentivized to undercount because the harms are distributed across households and time. A safety case for persons is justified only if it refuses that undercounting and forces institutions to state, before closure, what burdens their regime can generate and how they propose to govern them.
The chapter should now reconnect to Chapter 9 with precision. The recourse economy named incompatibilities that no institution should be allowed to balance away under the banner of procedural adequacy. A safety case for persons provides a prospective vehicle for making institutions show, before deployment or continuation of a closure regime, that they are not building those incompatibilities into the architecture. It can require institutions to demonstrate how contradiction remains metabolically usable, how review is sequenced before depletion where stakes require it, how documentary closure remains attributable, and how distributed exclusion risks are being monitored even where no single rule appears defective. What it cannot do is substitute for political contestation over how much contradiction capacity institutions must fund, who counts as affected, or which closures should be narrowed as a matter of justice rather than risk management. The instrument disciplines institutional argument. It does not dissolve the politics the argument enters.
This is the point at which Foucault’s warning about governmentality should be heard without caricature. Techniques of security and risk management can expand the state and institutional capacity to govern through apparently rational distributions of uncertainty and tolerable harm, often with great effectiveness and with equally great depoliticizing force (Foucault). The answer is not to abandon technique. The answer is to constitutionalize technique. In the terms of this book, that means every safety case for persons must operate under declared insufficiency, public reviewability, attributable reasoning, contestable remainder, and the doctrine’s threshold constraints. It must be built so that its own grammar can be challenged and its own exclusions can be seen, including the historical and classificatory exclusions that Chapter 11 will require reviewing bodies to confront.
The chapter has one final limit to declare before it closes. Safety cases, even in their strongest engineering traditions, do not guarantee safety. They can improve foresight, discipline argument, surface assumptions, and reassign the burden of justification toward those who create hazards, but they remain vulnerable to drift, institutional forgetting, false confidence, and capture. The same is true here, and more so, because person affecting closure regimes are saturated with normative conflict that cannot be stabilized by technical demonstration alone. A flawless argument form can still carry unjust premises. A rich hazard register can still omit the burdens of those historically excluded from the institution’s evidentiary and classificatory horizons. A well run review process can still occur inside a political order that refuses to fund recourse. The chapter would betray its own architecture if it obscured these limits in the excitement of instrumental design.
What the chapter has established, and what the next chapter will need, is narrower and stronger. A bounded transfer of safety case discipline into person affecting closure regimes is possible and useful if the transfer is constrained against category error and capture. The transferable core is prospective, structured justification under declared insufficiency by the institution that seeks closure. The nontransferable core is any logic that prices persons, rights, or recourse incompatibilities as ordinary tolerability variables. The instrument can expose and discipline foreseeable burden transfer pathways. It can make institutions state their hazards, controls, evidence, assumptions, and recourse triggers before closure is scaled. It can create a record that later doctrine can evaluate. It cannot settle the legitimacy question by itself.
Safety cases for persons, then, are what this movement has demanded they be. They are closure disciplines under critique, not legitimacy guarantees.
Chapter 11
A Doctrine of Uncertainty Custody
This chapter arrives late because it had to. A doctrine written earlier in the manuscript would have been an evasion of method. It would have converted the book’s central claim into an evaluative scheme before the mechanisms of burden transfer had been sufficiently exposed, before the distinction between declared and concealed insufficiency had been sharpened, before contestability climate and the outside of custody had been differentiated, and before Chapter 7 forced the manuscript to relinquish the fantasy that institutional virtue lies in becoming liminal. The doctrine could not be responsibly stated until the manuscript had first disciplined what it was allowed to forget.
What follows is therefore not a replacement theory of legitimacy and not an attempted synthesis that dissolves the conflict staged in Chapter 2 among Arendtian, Fullerian, Weberian, and administrative grammars. It is a constitutionally constrained adjudicative doctrine designed to test closure under conditions of uncertainty, especially where closure imposes disproportionate bodily and temporal cost. It does not tell us what legitimacy is in final form. It tells us how legitimacy claims that depend upon closure should be examined when institutions secure practical authority by converting indeterminacy into defensible decisions while exporting remainder into embodied life. In that sense it completes the cross cutting aspiration introduced in Chapter 2 without pretending to complete the traditions it traverses.
The doctrine proceeds through three integrated moments. Their order is not procedural ornament. Each moment exists to prevent a recurrent failure that the earlier chapters have already named. The first moment prevents grammar laundering by requiring public and contestable identification of the operative legitimacy grammar in the closure setting under review. The second moment prevents tradeoff laundering by establishing a threshold screen of nonderogable failures that cannot be balanced away through administrative fluency or institutional scarcity claims. The third moment permits grammar sensitive balancing among the remaining criteria, but only after threshold failures have been excluded and only through publicly declared weighting and reviewable reasoning. This architecture is exacting because the manuscript has shown that institutions are now highly practiced at appearing reviewable while redistributing the burden of making review real.
The first moment is grammar identification.
Chapter 2 argued that legitimacy conflicts are not noise around institutional closure but constitutive of how closure is justified and contested. One institution speaks in the idiom of rule congruence and reciprocity. Another invokes mandate, expertise, and administrability under volume. Another appeals to procedural regularity and role order. Another relies on political authorization and delegated competence while quietly borrowing moral force from a different grammar when public scrutiny intensifies. These are not simply rhetorical choices after the fact. They shape what counts as a sufficient reason, what burdens are visible, what failures matter, and what recourse is treated as obligatory. A custody doctrine that ignored this would collapse into hidden universalism. It would claim neutrality while silently importing one grammar as sovereign.
For that reason the operative legitimacy grammar in a closure setting cannot be declared solely by the institution under scrutiny. Institutions are too well positioned to invoke one grammar to secure discretion and another to evade custody obligations. A regulator may invoke expertise and emergency discretion when resisting challenge to evidentiary opacity, then invoke formal notice and pathway inventory when challenged on contestability climate. A platform may invoke private ordering and contractual assent to narrow public obligations while invoking quasi public function when seeking deference. A court or agency may invoke administrability to justify summary closure and later invoke individualized fairness to resist structural critique. None of these moves is conceptually impossible in the abstract. The problem arises when grammar switching is used to convert scrutiny itself into a moving target.
The doctrine therefore requires a provisional grammar identification through a publicly reviewable record before balancing or threshold analysis begins. The record must include, at minimum, the institution’s mandate, the decision function actually exercised, the closure form employed, the standing structure of affected parties, and the historical modes of justification used in comparable cases. Each element is necessary because institutions often describe one function while performing another. A body may present itself as conducting technical verification while in practice allocating social risk and sanction. A process may be labeled advisory while producing closure effects that function as adjudication. A pathway may be described as open while standing structures and timing rules make challenge functionally selective. Grammar identification must therefore follow the decision function as exercised, not merely the institutional self-description attached to it.
The requirement that historical modes of justification be included in the record is equally important. Institutions do not appear as fresh reasoning subjects in every case. They inherit justificatory habits, archival classifications, and settled vocabularies that shape what kinds of closure are imaginable and what kinds of burden remain nameless. Weber’s insistence on the routinizing force of administration remains indispensable here because office forms, files, and competencies stabilize not only action but justification (Weber). Fuller’s insistence on publicity, congruence, and fidelity to declared forms also matters because a doctrine that cannot see how institutions have historically justified comparable closures will misread selective novelty as principled adaptation (Fuller). The point is not antiquarian completeness. It is to expose the grammar conditions under which a closure claim is made before the institution can hide inside whichever justification it finds most protective under immediate pressure.
Provisional identification alone, however, would remain vulnerable to pretext if the institution’s record were treated as self authenticating. The doctrine therefore requires independent external review in a contestable forum. The reviewing body tests for pretext, mismatch, and self serving reclassification. Pretext occurs when the declared grammar appears to function primarily as a shield against obligations the institution’s actual closure role would otherwise trigger. Mismatch occurs when the decision function, closure form, and burden effects align with one grammar’s obligations while the institution claims another grammar’s immunities. Self serving reclassification occurs when the institution narrows the description of the closure setting to avoid scrutiny of standing structure, temporal effects, or evidentiary custody. These are not rare pathologies. They are ordinary institutional survival practices in environments where legitimacy is continuously demanded and contradiction capacity is unevenly distributed.
At this point the doctrine must incorporate a pressure the manuscript could not responsibly ignore once Chapter 8 clarified perimeter production. A record based grammar identification protocol can reproduce historical exclusions when the available record is itself shaped by traditions that omit, subordinate, or misclassify Indigenous, postcolonial, or non Western legitimacy frameworks. The doctrine must therefore refuse the comfortable fiction that archival absence equals normative irrelevance. Arendt’s work on statelessness, Ngai’s on impossible subjects, and Mamdani’s on colonial governance each show in different ways how administrative and legal archives preserve power through categories that decide who appears as a proper subject of legitimacy claims in the first place (Arendt, The Origins of Totalitarianism; Ngai; Mamdani). Hartman’s archival method deepens the warning by showing how the record may preserve domination while erasing or distorting the lives that bear it (Hartman). A doctrine that treats the available record as complete will reproduce perimeter production inside its own method.
For that reason the reviewing body must allow challenge to the provisional grammar on grounds of historical classification exclusion. Where the record is plausibly shaped by traditions that suppress relevant legitimacy frameworks, the reviewing body may require supplementary evidentiary submissions, including expert, community, and historical materials, in order to test whether the protocol’s own available options are already perimeter produced. This is not an invitation to unbounded relativism and not a denial that institutions must eventually act. It is a procedural discipline of epistemic honesty. The doctrine acknowledges that grammar identification is itself a closure act performed under uncertainty and historically shaped visibility. If the doctrine cannot admit that, it will become another instrument that launders exclusion through method.
That admission leads directly to the doctrine’s recursive oversight sentence, which must be stated in full constitutional spirit because it prevents the reviewing body from occupying a false position outside custody. Any reviewing body applying this protocol is itself a closure instrument and is therefore subject to the same custody obligations it applies, including reviewable grammar identification, declared insufficiency, attributable reasoning, and contestable remainder. This sentence is not decorative recursion. It is the condition under which the doctrine avoids reproducing the very immunized adjudicative posture it critiques. A reviewing body that demands grammar honesty from others while concealing its own grammar, that demands declared limits while speaking in final voice, or that demands attributable reasoning while issuing opaque summaries would collapse the doctrine into hierarchy without custody.
Once the first moment is complete, the doctrine moves to the threshold screen.
Chapter 9 identified arrangements that are incompatible with meaningful recourse and argued that some failures cannot be traded away through administrative self-description. Chapter 10 then showed why safety case disciplines for persons must be constrained by nonderogable hazards if they are not to become a technical language for pricing persons. The threshold screen is where those arguments become adjudicative doctrine. Its purpose is to prevent the doctrine from degenerating into checklist culture or balancing theater in which everything is commensurated and institutions with enough documentary fluency can justify any burden transfer as regrettable but necessary. Threshold analysis does not eliminate judgment. It narrows what judgment is allowed to do.
Three nonderogable failures remain core at this stage. Hidden uncertainty transfer is the first. Where the institution converts unresolved indeterminacy into closure while concealing that transfer and externalizing the cost of uncertainty resolution into embodied life, the doctrine treats the closure as threshold defective. Chapter 1 established the distinction between declared insufficiency and concealed insufficiency, and the entire manuscript has depended on it. Institutions may act under insufficiency. They may not conceal insufficiency and borrow authority from the appearance of completion. This threshold is the doctrine’s primary defense against synthetic confidence, administrative overstatement, and procedural naturalization of what the institution does not know.
Non attributable closure is the second threshold failure. Chapter 5 argued that documentary due process after generative text requires a reviewable distinction between evidentiary custody and textual articulation. Chapter 9 extended that argument by identifying documentary proliferation without attributable custody as incompatible with contestable closure. The doctrine now elevates that incompatibility into threshold form. Where consequential closure cannot be reliably tied to accountable reasoning, evidence custody, and reviewable adoption, the institution may not treat the defect as secondary. A closure that no one can own in reasons is not made legitimate by volume, speed, or technical sophistication. It is made harder to contest. The doctrine treats this as constitutive because the manuscript has shown that modern institutions can now produce persuasive closure text more cheaply than they can preserve accountable documentary custody. Without a threshold rule, balancing would simply reward the institution most capable of generating explanation surfaces.
Recourse priced beyond metabolically usable range is the third core threshold failure. Chapter 6 established the difference between nominal and metabolically usable pathways. Chapter 9 named contradiction priced beyond usable range and nominal review coupled with temporal depletion as incompatibilities. The threshold screen now operationalizes that insight in the doctrine’s strongest form. A closure setting fails at threshold where the practical capacity to contradict, review, or seek reversal is priced in time, money, translation labor, sanction risk, reputational exposure, or sequential depletion beyond a range that the institution knows many affected persons cannot sustain and has not provided means to accommodate or offset. The emphasis on known and unaccommodated conditions matters. The doctrine is not claiming that every person must experience recourse equally or that institutions can abolish all burden. It is claiming that a regime cannot preserve legitimacy by advertising pathways whose use is functionally reserved for those with surplus reserves.
These threshold failures are called core because the doctrine remains open to domain specific additions. A custody doctrine that could never admit new nonderogable hazards would quickly become obsolete or strategically gamed. The doctrine therefore permits extension at the threshold stage under declared criteria and reviewable argument, especially where new closure technologies or organizational forms generate hazards analogous in structure to hidden transfer, ownerless reasoning, or unusable recourse. Yet the core trio is nonnegotiable within this manuscript’s architecture because each emerged repeatedly across the preceding chapters and because each marks a point at which balancing can otherwise become laundering.
Threshold analysis is especially vulnerable to gaming, and the doctrine must say how. Institutions can minimize thresholds by redescribing consequential closures as preliminary or informational while consequences are already in motion. They can fractionate decisions so that no single step appears to carry enough weight to trigger threshold scrutiny. They can redefine affected populations narrowly to make recourse pricing appear tolerable in aggregate. They can treat declared insufficiency as a formal disclaimer while maintaining documentary and public rhetoric of completion. They can preserve attributable signatures while hollowing substantive review through throughput targets. The doctrine cannot eliminate gaming, but it can force gaming into a reviewable record. Threshold determinations must therefore state the closure form as experienced by affected parties, not only as labeled by the institution, and must include reasoning on sequencing, consequence onset, and practical recourse conditions. Where institutions repeatedly restructure pathways to evade threshold classification, the reviewing body may treat this pattern itself as evidence of custody evasion intent or effect.
If a closure setting passes threshold, the doctrine moves to its third moment, grammar sensitive balancing among the remaining criteria.
Balancing is the doctrine’s most dangerous and most necessary element. It is necessary because institutions operate across domains with different mandates, risks, and temporal constraints, and because legitimacy grammars really do disagree about what burdens can be tolerated and what forms of closure are obligatory. A doctrine that prohibited balancing would either become domain blind or expand threshold categories until almost all adjudication collapsed into categorical prohibition. It is dangerous because balancing is where institutions most readily hide value choices behind technical form. The doctrine accepts balancing only after threshold screening and only through declared weights tied to the provisionally identified and independently reviewed legitimacy grammar.
The criteria to be balanced remain the ones named in the book’s final architecture. They include declared uncertainty, burden distribution, evidentiary custody, contestability path, reversibility, temporal cost, and embodied load. The earlier chapters have already supplied the substance of each term. Declared uncertainty concerns whether the institution has honestly specified what remains unresolved and where closure claims are intentionally narrowed. Burden distribution concerns where the costs of uncertainty, contradiction, delay, and translation are allocated and whether that allocation is visible or hidden. Evidentiary custody concerns the institution’s ability to trace consequential reasons to accountable records and adoption decisions. Contestability path concerns the existence and usability of recourse under actual climate conditions rather than pathway inventory alone. Reversibility concerns whether closure effects can be stayed, narrowed, or unwound if challenge succeeds and under what conditions. Temporal cost concerns the sequencing of decision and review in relation to the life functions and stakes at issue. Embodied load concerns the bodily, cognitive, affective, and relational burdens imposed by the closure and its recourse regime, especially where institutions export uncertainty into ongoing vigilance and self regulation.
The doctrine does not assign fixed universal weights to these criteria because doing so would covertly install one legitimacy grammar as master. In a setting governed by a strongly adjudicative grammar, evidentiary custody and reversibility may carry especially heavy weight. In a setting of emergency public administration, temporal cost and declared uncertainty may interact differently, though not in ways that erase threshold constraints. In a setting where common world maintenance and public accountability are central, burden distribution and contestability path may require stronger public justification than administrability alone would yield. The doctrine therefore requires that weighting be publicly declared and explicitly tied to the operative grammar as independently reviewed. The weighting declaration itself is contestable. A reviewing body may reject weights that are inconsistent with the identified grammar, that ignore the actual closure function, or that functionally reintroduce threshold failures through low weighting of criteria that carry their substance.
This is where Chapter 2’s conflict among legitimacy traditions returns in its most productive form. Arendt does not become a checklist factor, Fuller does not become a compliance rubric, Weber does not become a permission slip for closure acceleration, and administrative pragmatism does not become a default sovereign. Instead the doctrine requires that institutions and reviewers show which grammar they are invoking, how that grammar justifies the weighting pattern, and what burdens that grammar is willing to name rather than hide. The result is not harmony. It is forced candor. Institutions may still defend different closures under different grammars. What they may not do, if the doctrine is followed, is move silently between grammars to retain discretion and shed custody.
Balancing, like threshold analysis, is intensely gameable. The doctrine therefore pairs each criterion with foreseeable gaming pathways and requires reviewers to address them in the reasoning record. Grammar shopping is the most obvious. Institutions may frame the closure setting to trigger a grammar more favorable to their preferred weights. Threshold minimization remains active here too, with institutions presenting severe burdens as downstream contingencies outside the closure under review. Performative declaration is another familiar move. An institution may openly declare uncertainty but in a manner so generic or inflated that declaration no longer constrains closure claims or supports recourse. Recourse simulation is the balancing stage analogue of contestability theater. Pathways are added, interfaces improved, or ombuds functions created while temporal depletion, sanction sequencing, or evidentiary ownerlessness remain unchanged. The doctrine cannot prevent institutions from attempting these moves. It can require that weighting and criterion analysis be written in a way that makes the moves legible and contestable.
The anti theater discipline that has governed the manuscript from the middle chapters onward therefore remains active inside the doctrine itself. This matters because doctrines are especially prone to self neutralization through procedural performance. Once institutions learn the vocabulary of declared insufficiency, burden distribution, and contestability, they can begin speaking it fluently while preserving underlying transfer structures. Chapter 10 warned that safety case practices can become audit ritual. The same is true here. A custody doctrine can become a legitimating script if applied as form rather than as burden tracing. The doctrine’s answer is not purity and not distrust of all institutional learning. Its answer is recursive reviewability. Grammar identification, threshold screening, and balancing reasoning must each remain open to challenge on grounds of mismatch between declared method and observed burden effects. Where institutions repeatedly satisfy formal doctrine and continue to generate the same hidden transfers, the reviewing body must treat that recurrence as evidence against the institution’s custodial claim, not as proof that the doctrine is too demanding.
At this point the chapter must make one of its most difficult distinctions because self limitation can easily become harmless humility. The doctrine distinguishes epistemic limitation from political unavailability.
Epistemic limitation concerns domains in which robust application is constrained because relevant concepts or evidence cannot be sufficiently stabilized for the doctrine’s criteria to operate with integrity. Some closures occur in conditions of rapid factual uncertainty, contested causation, or genuinely indeterminate evidentiary forms where precise burden tracing or reversible sequencing cannot be specified in the manner the doctrine prefers. In such settings the doctrine may still require declared insufficiency and threshold vigilance, but its balancing outputs may be coarse, provisional, and explicitly limited. To deny this would be to force false precision into the doctrine and thereby reproduce the closure confidence the manuscript has criticized.
Political unavailability is different and must be named plainly. Some domains resist custodial doctrine not because evidence or concepts are unstable but because immunized discretion, emergency authority, opacity dependent governance, fragmented responsibility, or unreviewable classification power make custodial review structurally unavailable. Institutions in these domains may possess abundant data, stable procedures, and sophisticated documentation while preserving legal or political insulation against the very forms of grammar challenge, threshold scrutiny, and balancing contestation the doctrine requires. Chapter 8’s analysis of designed and distributed outsides is directly relevant here. A doctrine can be epistemically applicable and politically blocked. The difference matters because if the chapter describes political blockage as mere complexity or uncertainty, it launders power into method.
The doctrine must therefore say when it is being refused. Where a closure regime is organized to preclude independent grammar review, to withhold attributable reasoning, to prevent metabolically usable recourse, or to immunize category and timing choices from contestation, the doctrine’s honest output may be not a weighted assessment but a statement of structural unavailability. This statement is not a failure of doctrine. It is one of the doctrine’s most important forms of truthfulness. It tells the reader, the subject, and the broader public that the issue is not that uncertainty could not be responsibly adjudicated, but that the institutional order has arranged itself against custodial adjudication.
This distinction also clarifies the relation between doctrine and politics that the coda will later have to confront. A doctrine can expose hidden transfer, identify threshold failure, and declare political unavailability. It cannot by argument alone compel institutions to become reviewable, fund recourse, or surrender immunized discretion. Chapter 9 has already prepared that conclusion. Establishing incompatibility does not make revision politically available. The doctrine’s work is indispensable because it sharpens what is being denied and prevents institutions from naming denial as neutral administration. It remains bounded because naming is not yet redesign.
The chapter should now show, at least briefly, how the doctrine relates to the instruments under critique that follow in the appendices. An uncertainty custody audit can assist grammar sensitive balancing if it preserves declared insufficiency and burden tracing rather than collapsing into scorecard assurance. A recourse adequacy assessment can support threshold analysis where metabolically usable pathways and temporal depletion are operationalized without recourse simulation. A documentary due process protocol for generative environments can strengthen evidentiary custody and nonattribution screening if it resists synthetic fluency as a substitute for ownership. An embodied burden review can inform temporal cost and embodied load criteria if it avoids extractive measurement and remains answerable to the line of exposure rather than turning affliction into data exhaust. In each case, however, the doctrine stands above the instrument in one constitutional sense. Instruments may help generate evidence and structure review. They do not decide threshold normativity by themselves. They remain closure instruments under critique, subject to the same custody obligations and the same gaming pathways they are meant to reveal.
The doctrine can now gather its architecture in the form of an adjudicative sequence without reducing itself to a procedural checklist. First, identify the operative legitimacy grammar provisionally through a public record tied to mandate, actual decision function, closure form, standing structure, and historical justificatory practice. Then submit that identification to independent external review capable of testing pretext, mismatch, self serving reclassification, and historical classification exclusion, including through supplementary evidentiary materials where archival options are already perimeter produced. Second, apply a threshold screen that rejects closures marked by hidden uncertainty transfer, non attributable closure, or recourse priced beyond metabolically usable range, while permitting domain specific threshold additions under declared and reviewable reasoning. Third, where threshold is passed, conduct grammar sensitive balancing among declared uncertainty, burden distribution, evidentiary custody, contestability path, reversibility, temporal cost, and embodied load through publicly declared weights tied to the reviewed grammar and open to contestation. At every stage, identify foreseeable gaming pathways and test the institution’s reasoning against observed burden effects rather than documentary fluency alone. And at every stage, treat the reviewing body as a closure instrument subject to its own custody obligations.
That sequence is demanding, but demand alone is not the point. The doctrine is designed to redistribute burden in institutional argument. For too long, the burden has fallen on affected persons to prove that what they are carrying is not private misfortune but exported remainder, to prove that nominal pathways are unusable, to prove that proliferating documents are ownerless, to prove that cumulative exclusions are not anecdotal complexity. The doctrine does not eliminate that labor, and it would be dishonest to imply otherwise. It does, however, require institutions and reviewing bodies to carry more of the justificatory weight before closure can continue to pass as legitimate by default.
The chapter closes with the sharpened constitutional sentence of insufficiency that its own architecture requires. The doctrine can discipline closure, expose burden transfer, and preserve recourse conditions. It cannot eliminate remainder, resolve legitimacy conflict in final form, or make politically unavailable revision available by argument alone.
Yes. There is one more chapter in the book’s main architecture before the appendices under Instruments Under Critique.
Chapter 12, The Governed Remainder
A book such as this one can fail at the end in two opposite ways, and both failures now stand close. It can mistake diagnosis for victory and write as if naming incompatibility had already altered institutions that depend on it. Or it can retreat into moral atmosphere and treat the political problem as too contingent, too empirical, or too compromised for final address. The first failure flatters doctrine. The second flatters critique. Neither can be accepted here.
The manuscript began with a line of exposure. An institution reached documentary closure. A person continued to carry unresolved cost in time, vigilance, translation, reputation, and speech. The question that followed was not whether closure was always false, nor whether institutions could live without it, but where the remainder of closure goes, who carries it, and under what authority that transfer is treated as rational. Everything since then has been an attempt to refuse two habits at once. The first is the institutional habit of treating remainder as private residue once file order has been achieved. The second is the intellectual habit of treating remainder as a philosophical excess detached from administrative sequence, procedural form, and burdened life.
The result, if the argument has held, is not a completed theory of legitimacy. It is a changed perception of what institutional closure is and what it asks others to bear. Closure now appears less as the opposite of uncertainty than as a managed wager under insufficiency. Remainder appears less as unfortunate spillover than as a structured outcome of how institutions secure decisional authority. Contestability appears less as pathway inventory than as a climatic and economic condition. Recourse appears less as an accessory to fairness than as a first order budget for contradiction. Documentary articulation appears less as evidence of accountability than as a domain requiring custody of reasons, provenance, and adoption. The outside of custody appears less as a natural perimeter than as a political anatomy produced in designed and distributed forms. Doctrine appears less as a final harmonizer of legitimacy conflict than as a constrained adjudicative practice that can discipline closure, expose burden transfer, and preserve reviewable argument under conditions it cannot complete.
These shifts in vocabulary are not semantic refinements. They are changes in institutional obligation. If the book has done its work, then some things can no longer be said with a straight face. Institutions can no longer claim that uncertainty disappears because documentation is abundant. They can no longer call review meaningful because pathways exist in abstract form. They can no longer invoke complexity as if complexity itself dictated hidden transfer. They can no longer treat cumulative exclusion as unreal because no single rule appears abusive. They can no longer speak of safety, fairness, or due process while refusing attributable custody over consequential closure. They can no longer call their own grammar neutral when grammar selection decides which burdens count and which disappear from admissible sight.
Yet this is also the point at which the book must resist self-congratulation. A doctrine can expose, and exposure can fail to move power. A threshold incompatibility can be demonstrated with public clarity and still remain institutionally survivable for the actors who benefit from it. A reviewing body can state structural unavailability honestly and still lack leverage to alter the conditions that make review unavailable. The manuscript has reached, and cannot evade, the political availability gap.
The distinction must be stated again because the end of a book invites rhetorical compression. Doctrinal incompatibility and political revisability are not the same thing. To show that a closure regime externalizes uncertainty without declaration, prices contradiction beyond metabolically usable range, or proliferates ownerless reasons is to establish a normatively serious claim. It is not yet to make revision available. Revision is a political achievement. It depends on forms of organization, sequencing, coalition, institutional rivalry, legal strategy, crisis interpretation, budgetary struggle, and public imagination that doctrine can illuminate but not furnish by argument alone. The book would betray its own account of custody if it converted this limit into a final sentence of resignation. The point is not that doctrine is weak. The point is that doctrine is one form of work among others, and the burden of making incompatibility actionable cannot be laundered back into theory once theory has named it.
This is where the coda has to ask, not prescribe, and ask in a way that does not displace the question onto an abstract future public. What political forms, coalitional capacities, legal sequences, institutional redesigns, or crisis conditions might make revision available where the doctrine has established incompatibility? The book cannot answer this once and for all because the answer varies by domain, by regime, by available leverage, and by the distribution of burden and organization in the setting at issue. But it can refuse false alternatives. Revision will not come only through litigation, though litigation matters. It will not come only through administrative design, though design matters. It will not come only through legislative reform, though statutes matter. It will not come only through moral witness, though witness matters. The question is how these forms are composed under conditions in which exclusion is often distributed, contradiction is under-budgeted, and institutions have learned to metabolize critique as procedural improvement without custodial effect.
Unger’s pressure is indispensable here because he forces the issue of institutional imagination without permitting either romantic voluntarism or deterministic fatalism. Institutions are not natural facts, and the arrangements that make custody evasion appear inevitable are not beyond remaking simply because they are presently routinized. At the same time, remaking does not occur because critique has exposed contingency. It occurs when organized capacity is built to convert contingency into alternative structure, and when legal, administrative, and political forms are made available to sustain that conversion rather than absorb it back into the old grammar (Unger, False Necessity; Unger, What Should Legal Analysis Become?). The present manuscript can meet that pressure only by refusing to let its final question float free of positions from which the gap is lived.
For administrators and institutional designers, the political availability gap appears first as a question of benefit and refusal. Once custody evasion has been named in the forms traced by this book, the issue is no longer whether hidden transfer exists in the abstract. The issue is whether those who design and maintain closure regimes will continue to benefit from transfer while describing themselves as constrained by inevitability. Some will indeed be constrained by law, budget, hierarchy, and throughput targets. The point is not to deny those constraints. The point is to ask what they do with the margin they possess, how they describe the constraints they invoke, whether they disclose insufficiency or conceal it, whether they build recourse budgets into redesign or spend all reform capital on visible throughput gains, and whether they are willing to make perimeter production legible even when the institution would prefer to preserve the fiction of neutral architecture. The gap, from this position, is not only a structural condition. It is a practice of institutional candor and institutional disobedience in small and large forms.
For lawyers, judges, and reviewers, the gap appears as a question about adjudicative form itself. Can legal and quasi-legal forums bear burdens they have historically privatized? Can they treat timing, embodied load, and documentary ownerlessness as institutional questions rather than unfortunate externalities? Can they discipline grammar shopping without covertly installing their own grammar as beyond review? Can they say when a closure regime is politically unavailable to custodial doctrine without translating refusal into harmless complexity? The doctrine proposed in the prior chapter gives one architecture for this work, but architecture does not supply courage, jurisdiction, or appetite. Adjudicative actors operate inside traditions that have their own immunities, their own docket pressures, their own repeat-player dependencies. The coda cannot command transformation here. It can only insist that the legitimacy of adjudicative forms now depends, in part, on whether they continue to naturalize the transfer of uncertainty into lives they never had to count.
For labor advocates and organizers, the gap appears as a problem of coalition, timing, and leverage under distributed exclusion. This manuscript has repeatedly argued that custody evasion often emerges not through a single spectacular denial but through cumulative procedures, classifications, and depletions that are hard to narrate in the institution’s preferred proof form. That very fact creates an organizational challenge. Distributed burdens often produce isolated experiences before they produce a common claim. The political work, then, is partly epistemic and partly material. It involves turning dispersed remainder into shared intelligibility without flattening differences in exposure, and turning shared intelligibility into leverage before institutional adaptation converts the claim into symbolic reform. This may involve union structures, worker centers, legal aid coalitions, community organizations, mutual aid formations, policy campaigns, and domain specific alliances that can connect designed blockages to distributed outsides. The coda does not prescribe one path because domains differ, and because coalition forms themselves are historically situated and contested. What it does insist upon is that the recourse economy and the outside of custody are not problems that institutions will solve because they have been well described. They are problems that must be made costly, visible, and politically non-optional.
For those who have carried institutional remainder without language for it, the gap appears in the hardest register of all because it joins possibility to cost without guarantee. One of the claims of this book has been that persons are repeatedly required to bear closure’s unresolved burdens in forms that institutions misname as personal difficulty, noncompliance, delay, confusion, or affective excess. To name those burdens as exported remainder can be clarifying. It can also be expensive. Contestation demands time, repetition, documentation, self-exposure, coordination, and risk. Refusal can protect dignity and foreclose immediate remedy at the same time. Survival may require strategic quiet in settings where the doctrine would name threshold failure but no usable forum exists. The coda cannot turn these conditions into a heroic politics without reproducing the same transfer it has criticized. It can, however, insist that the cost of contestation is itself part of the custodial question, and that any politics of revision that depends upon endless self-translation by the burdened has already conceded too much to the order it opposes.
The distributed readership named here does not produce multiple endings. It changes what the final question is. The question is no longer whether the argument has identified a better doctrine than its rivals, though doctrinal clarity matters. Nor is it whether institutions can be persuaded to acknowledge uncertainty more honestly, though candor matters. The question is how the burden of making custodial revision available is distributed once incompatibility has been named and once the institutions under scrutiny have demonstrated their capacity to preserve legitimacy through visible reform and distributed evasion. Who can force grammar identification into public view when grammar opacity preserves discretion? Who can make recourse budgeting contestable when contradiction scarcity appears as technical necessity? Who can preserve evidentiary custody when documentary abundance itself becomes a mode of rule? Who can insist on protected nonclosure for embodied life without converting institutions into theaters of suspended decision? These are not questions the book can answer by one final synthesis. They are the questions the book is obligated to leave behind in a form that resists misrecognition.
The coda also owes a final clarification about what “the governed remainder” means, because the phrase can be misunderstood in two opposed directions. It does not mean that remainder can be fully governed, captured, or absorbed into institutional design if doctrine and instruments become sufficiently sophisticated. Chapter 7 already foreclosed that fantasy, and Chapter 10 placed every instrument under critique for exactly this reason. Nor does it mean that remainder should be left entirely ungoverned as a kind of sacred outside beyond institution, law, or review. Chapter 8 foreclosed that romance by showing that what is left outside is often not protected but abandoned, privatized, or differentially exposed. The governed remainder names a more difficult posture. It names an institutional and political commitment to govern the transfer, visibility, contestability, and burden distribution of remainder without claiming to eliminate remainder itself. It names accountable closure and protected nonclosure as co-present demands.
This co-presence is not a compromise formula. It is the condition of honest institutional life in a synthetic age. Systems now produce closure at unprecedented speed, density, and fluency. They can classify, summarize, route, predict, flag, draft, and decide with an efficiency that invites old fantasies in new technical dress. The fantasy of completion returns as document. The fantasy of neutrality returns as optimization. The fantasy of fairness returns as interface. The fantasy of accountability returns as trace logs without custodial ownership. The book’s governing claim has been that legitimacy in such conditions must be judged in part by the custody of uncertainty. The coda adds what the prior chapters have made unavoidable. Custody is not a property of statements alone. It is a political achievement of institutions, forums, coalitions, and publics that accept the cost of keeping closure reviewable while refusing to punish life for exceeding it.
One final return to the prologue is now required, and it must be austere. An institution closes. A person continues. The book cannot stop the first event by argument, and it cannot redeem the second by vocabulary. What it can do is alter the terms under which the transfer between them is seen, justified, and contested. It can make hidden transfer harder to call rational. It can make ownerless closure harder to call accountable. It can make unusable recourse harder to call available. It can make perimeter production harder to call necessity. It can make doctrinal self-limitation harder to confuse with political innocence. It can insist that where revision is unavailable, the unavailability be named as power and not misdescribed as method.
That is not cadence. It is an interval, and the interval is the only honest ending this manuscript can offer. Accountable closure and protected nonclosure remain co-present demands under conditions no doctrine can complete. The remainder is not solved. It is governed, contested, and carried in public sight or it is exported again.
Instruments Under Critique
Appendix A
An Uncertainty Custody Audit
This appendix is offered under the constitutional limits already established in the body of the manuscript. It is an instrument, not a settlement of the argument. It is designed to discipline institutional description and expose burden transfer under declared insufficiency, and it remains vulnerable to capture, theater, selective uptake, and grammar dispute. It does not absorb remainder. It does not convert custodial legitimacy into a score. It does not replace the doctrine in Chapter 11. It helps generate a reviewable record for custodial judgment, and it does so only to the extent that the institution and its reviewers accept the burden of making closure practices inspectable in forms that can be contested.
The need for an uncertainty custody audit follows directly from the pattern traced across the manuscript. Institutions routinely describe closure in terms of correctness, compliance, throughput, service quality, or technical adequacy while leaving unanswered the prior question of where uncertainty has gone, who carries its unresolved cost, and under what authority that transfer is treated as rational. The audit intervenes at that level. It is a burden tracing instrument before it is an efficiency instrument, and a legitimacy support instrument before it is a performance management tool. Its central aim is to force a closure regime to state, in attributable and reviewable form, what uncertainty remains unresolved at the point of closure, how the institution distributes the costs of that unresolvedness, what recourse and reversal conditions remain in practice, and what grammar of legitimacy is being invoked to justify the arrangement.
The phrase grammar is not ornamental here. Chapter 11 has already shown that the same institutional pathway can be defended as administratively necessary, legally regular, technically expert, or publicly accountable depending on which justificatory grammar is made operative. An uncertainty custody audit that ignores this problem becomes a laundering device almost immediately, because institutions can fill the instrument with data while silently choosing the grammar under which the data will count. For that reason, the audit begins not with metrics but with a provisional grammar declaration that is explicitly marked as contestable and incomplete until independently reviewed. The instrument is therefore not self activating. It presupposes the doctrinal requirement that grammar identification cannot be monopolized by the institution under scrutiny.
This also means that the audit has a built in applicability condition. Where grammar identification is under live dispute and no contestable forum is available to test the dispute, the audit may still be conducted, but its outputs must remain conditional. In such settings the audit can surface burden transfer, documentary ownerlessness, recourse pricing, temporal depletion, and perimeter effects, but it cannot honestly convert those findings into a single custodial conclusion without declaring that the conclusion depends on an unsettled grammar question. This is not a defect of the instrument. It is one of its principal honesty functions. The audit must not simulate certainty in the very situations where the doctrine has said uncertainty is partly about the public reviewability of the grammar itself.
The audit is best understood as a structured inquiry into a closure regime rather than a one time questionnaire about a single decision. This distinction matters because the manuscript has already shown that many of the most consequential failures are distributed rather than discrete. A single closure event may look defensible in isolation while the regime that produces many such closures is custodially unsound in aggregate. The audit therefore asks the institution to identify the closure form under review, the range of consequences it can impose, the standing structure of affected persons, the timing and sequencing architecture, the documentary chain through which closure is produced and justified, the recourse pathways actually available, and the known interfaces with downstream institutions that absorb unresolved burdens. The resulting record is not simply descriptive. It is the first stage at which the institution can no longer pretend that closure is exhausted by the moment at which a file is marked complete.
The discipline of the audit lies in how it compels the institution to distinguish declared insufficiency from concealed insufficiency. An institution may legitimately close under uncertainty. The audit does not require omniscience and would become absurd if it did. What it requires is a clear statement of what the institution does not know, cannot know at the time of closure, or has chosen not to resolve before acting, together with a statement of how that unresolvedness is expected to travel after closure. If the institution cannot produce this statement, or produces only generic disclaimers detached from the actual closure pathway, the audit has already surfaced a custodial defect even before any balancing analysis begins. Institutions often know how to write disclaimers. The audit is designed to test whether the disclaimer actually governs the closure claim, the recourse design, and the burden distribution, or whether it functions as legal insulation attached to a rhetoric of practical certainty.
The evidentiary dimension of the audit is equally central. The instrument requires not only that the institution describe its closure pathway, but that it preserve attributable custody over the evidence and reasoning by which the pathway is justified. This requirement is especially important in regimes saturated by generated text, templated notices, layered vendor systems, and policy updates that diffuse responsibility across software, compliance teams, contractors, and front line staff. The audit therefore asks a simple but institutionally disruptive question in a form that must be answered with specificity. When consequential closure occurs, who can identify the operative reason, the record that supports it, the rule or policy source that authorized it, the person or office that adopted it, and the change history by which the pathway acquired its present form. Where this chain cannot be maintained, the audit should not record a mere documentation weakness. It should record a potential non attributable closure hazard of the sort the doctrine treats as threshold relevant.
At the same time, the audit must resist the opposite failure, which is to reward institutions for textual abundance. The manuscript has already argued that documentary proliferation can decrease contestability when articulation outruns custody. An institution can satisfy shallow audit culture by producing manuals, policies, notices, scripts, quality assurance summaries, dashboards, and generated explanations in great quantity while preserving no coherent answer to the question of ownership, adoption, and reviewability. The uncertainty custody audit therefore treats textual volume as neutral at best and potentially suspicious absent attributable custody. It asks whether documentation clarifies burden transfer and recourse conditions for those affected and for independent reviewers, not whether the institution can produce many words about process.
The recourse portion of the audit is the point at which the instrument must most clearly incorporate Chapters 6 and 9. It does not ask only whether review pathways exist. It asks how contradiction is budgeted and sequenced. The institution must state what forms of challenge are available, at what stages, under what timing constraints, with what evidentiary expectations, with what risks attached to filing or delay, and with what practical supports or accommodations for translation, disability, work conflict, childcare, documentation retrieval, or other known barriers that make pathways nominally open and functionally unusable. The audit is not a universal accommodation mandate in disguise. It is a burden tracing demand. If the institution knows that contradiction is routinely priced beyond metabolically usable range for a meaningful portion of affected persons and has chosen not to redesign, offset, or narrow closure claims accordingly, the audit should make that choice visible as recourse rationing rather than neutral operational reality.
Temporal sequencing receives its own attention in the audit because institutions often hide recourse inadequacy inside the order of operations. A pathway may look fair in formal terms while consequence begins immediately and review matures only after the practical utility of challenge has been exhausted. The audit therefore requires the institution to state, in the concrete temporal grammar of the regime, when consequence begins, when reasons become attributable, when challenge can first be initiated, when review is likely to be completed under ordinary and stressed conditions, and what harms become irreversible or substantially harder to repair during that interval. This is the point at which many closure regimes reveal that they are relying on post consequence review while continuing to describe themselves as reviewable in a general sense. The audit does not resolve whether every such sequence is illegitimate, since that is a doctrinal question tied to stakes and grammar. It does force institutions to stop concealing the sequence itself.
The audit’s treatment of embodied burden is deliberately constrained, because this is one of the domains where instruments become extractive quickly. The point is not to produce more intimate metrics of distress in order to optimize burden transfer. The point is to prevent institutions from describing burdens as purely procedural when those burdens predictably require bodily vigilance, cognitive labor, repeated self translation, affective regulation under accusation, or prolonged uncertainty management that reorganizes ordinary life. The audit therefore permits institutions to document embodied load only in forms tethered to closure design decisions and recourse conditions, and it requires a statement of why each requested burden indicator is necessary for custodial governance rather than merely useful for managerial curiosity. If an institution cannot explain why a proposed measure helps redesign closure or recourse conditions, the instrument should presume non collection. This is one of the appendix’s anti extractive constraints and is included because the book has repeatedly argued that institutions often respond to hidden burden transfer by demanding more disclosure from the burdened.
The perimeter portion of the audit is where the appendix must most clearly satisfy the honesty requirement introduced in the book’s final architecture. The instrument must state whether it is better at detecting designed failures than distributed exclusions, and it must explain what that asymmetry means for interpretation. The answer, in the present form of the instrument, is that it is generally stronger at detecting designed failures. It can identify explicit timing bars, standing restrictions, waiver clauses, category gates, documentary ownership gaps, sanction first sequencing, and visible recourse pricing with reasonable clarity if the institution is compelled to provide the relevant records. It is weaker at detecting distributed exclusions that emerge only across time, offices, and interacting pathways, especially where no single unit can see the cumulative burden and where the institution lacks or resists cross system tracing. This weakness does not invalidate the instrument. It imposes an interpretation rule. A strong audit result on designed failures cannot be taken as evidence that the closure regime is custodially sound overall. It may indicate only that the instrument has not captured the distributed outside. The audit must therefore include a declared asymmetry statement and a distributed exclusion caution whenever the regime under review involves multi office handoffs, fragmented documentary systems, or known downstream burden displacement.
To strengthen detection of distributed exclusions without pretending to solve the problem, the audit includes an aggregate pressure analysis that is expressly marked as partial. The institution must identify recurrent handoff points, repeat documentation demands, category reroutings, timing accumulations, and downstream institutional referrals known to shape recourse climate and closure burden in practice. It must also identify which of these it can currently observe and which remain outside its visibility because of legal, technical, contractual, or organizational fragmentation. This part of the instrument is as much a disclosure of institutional blindness as it is a performance assessment. If an institution cannot see where its closure pathways offload burden downstream, the audit should not permit a conclusion of custodial adequacy. It should instead record a declared visibility limit that bears directly on the credibility of the institution’s closure claims.
Because every modern instrument is vulnerable to theater, the uncertainty custody audit includes anti theater constraints built into its own use conditions. It may not be represented as a certification of legitimacy. It may not be reduced to a composite score for procurement, public relations, or executive benchmarking. It may not be completed solely by compliance or communications units without attributable participation from the offices that actually design, authorize, and administer closure. It may not omit adverse examples on grounds of reputational sensitivity if those examples are necessary to show how the closure regime handles contradiction, reversal, or evidentiary instability. It may not treat low challenge rates as evidence of procedural success without analyzing contradiction pricing and temporal depletion. It may not present pathway inventory as recourse adequacy without a usability and sequencing account. These constraints are not procedural fussiness. They are lessons extracted from the manuscript’s central diagnosis that institutions are increasingly sophisticated at converting scrutiny into controlled evidence of their own seriousness.
The audit is also vulnerable to a specific capture pattern in synthetic administrative environments. Institutions may use generative systems to draft audit narratives, summarize case patterns, and articulate control claims faster than they can verify evidentiary custody or recourse conditions on the ground. The result can be an audit that is rhetorically exact and operationally hollow. To counter this, the instrument requires provenance declaration for the audit itself. If generated text, automated summarization, or vendor prepared language is used, the institution must identify where, for what function, under what review protocol, and with what attributable adoption. The purpose is not to prohibit assisted drafting. It is to prevent the audit from becoming an instance of the very non attributable closure culture it is supposed to discipline.
A further capture risk arises when institutions use the audit to displace political conflict into managerial calibration. An uncertainty custody audit can be absorbed into internal governance as a process improvement tool while the institution continues to avoid external contestation over standing, rights, budget, or threshold failures. The appendix must be explicit here. The audit can support doctrine. It cannot replace politics. A regime may perform the audit faithfully and still be politically committed to closure patterns the doctrine would condemn. In such settings the instrument’s most important output may be a documented record of custodial incompatibility and structural unavailability rather than a redesign plan. This is not a failure of the instrument but a reminder of its constitutional subordination to the manuscript’s larger claim about political availability.
The relation between the audit and Chapter 11 should now be clear. The audit does not decide the doctrine’s first moment because grammar identification remains contestable and may require supplementary historical and community evidence beyond what the institution can or will provide. The audit can, however, discipline the institutional side of the grammar record by forcing the institution to state mandate, function, closure form, standing structure, and justificatory practice in attributable form. The audit does not decide the threshold screen, but it can surface evidence relevant to hidden uncertainty transfer, non attributable closure, and recourse priced beyond metabolically usable range. The audit does not perform grammar sensitive balancing, but it can structure evidence for declared uncertainty, burden distribution, evidentiary custody, contestability path, reversibility, temporal cost, and embodied load. In short, the audit is an evidentiary and descriptive support instrument for custodial adjudication. It becomes dangerous when it is mistaken for adjudication itself.
The appendix closes where it should, with declared insufficiency. This instrument can clarify closure pathways, expose burden transfer, and make recourse rationing and documentary ownerlessness harder to hide. It can improve institutional candor where good faith exists and generate reviewable records where contestation is possible. It can also be gamed, narrowed, scored, vendorized, and ritualized. It is stronger on designed failures than distributed exclusions, and it depends on contestable grammar identification for full doctrinal use. It should therefore be used as a burden tracing discipline under critique, never as a final warrant that a closure regime is legitimate because it has been audited.
Instruments Under Critique
Appendix B
A Recourse Adequacy Assessment
This appendix is offered under the same constitutional limits that governed Appendix A and the body of the manuscript. It is an instrument under critique, not a warrant of legitimacy, and it remains subordinate to the doctrinal architecture developed in Chapter 11. Its purpose is narrower and more exacting. It is designed to make institutions state, in attributable and reviewable form, whether the contradiction capacity they claim to provide is actually usable by those who bear closure’s burdens, and whether the timing, sequencing, evidentiary structure, and cost distribution of review permit recourse to function before consequence hardens into practical irreversibility. It does not convert recourse into a customer experience metric. It does not collapse custodial legitimacy into pathway inventory. It does not answer the political question of who will force institutions to fund contradiction where they presently refuse to do so. It supports that struggle by making recourse rationing harder to hide.
The need for a recourse adequacy assessment follows from the manuscript’s central argument in its strongest institutional form. Institutions often preserve authority by displaying pathways of challenge while under-budgeting the conditions that make challenge usable in time. They can announce appeals, ombuds channels, hearings, reconsideration stages, complaint portals, and judicial review while pricing contradiction beyond metabolically usable range, sequencing review after deprivation, diffusing evidentiary custody across offices and vendors, and preserving documentary ambiguity about who decided what and on what grounds. In such settings institutions still appear reviewable. The person subjected to closure experiences review as a secondary burden, an interpretive labor, or a delayed ritual that matures after the life function at issue has already been damaged. The recourse adequacy assessment is built to test that gap between pathway form and contradiction use.
The instrument therefore begins from a premise the manuscript has already defended against multiple simplifications. Recourse is not a procedural ornament attached to otherwise complete closure. It is a first order institutional budget for contradiction. The assessment asks the institution to describe and justify that budget, not in the language of aspiration, but in the language of allocation. How much contradiction can be absorbed before closure is issued, after closure is issued, and after adverse consequences have begun. What forms of challenge are recognized as institutionally meaningful rather than merely receivable. What time is allocated to challenge before loss becomes irreversible or prohibitively costly to unwind. What supports are built for those who must carry translation, documentation, and scheduling burdens in order to be heard. What offices own review authority and what offices merely absorb grievance traffic without power to alter closure. What reasons remain attributable across challenge and revision. These questions are not exhaustive, but they force institutions to stop treating recourse as an uncosted virtue.
Like the uncertainty custody audit, this assessment cannot proceed honestly without a provisional and contestable grammar declaration. The adequacy of recourse depends in part on what kind of closure is being justified, under what mandate, and within what legitimacy grammar. A high volume administrative screening process, an adjudicative determination with rights bearing consequences, an emergency action under temporary authority, a contractual dispute resolution pathway, and a hybrid automated triage system linked to human review do not carry identical recourse obligations in timing, reversibility, or evidentiary form. The instrument cannot decide these obligations by hidden importation. It must require the institution to state the grammar it believes governs the closure setting, the actual decision function exercised, the consequence profile of closure, and the standing structure of affected persons. That declaration remains conditional until independently reviewed under the doctrine. Where grammar identification is materially contested and no contestable forum exists to test the dispute, the instrument must state that its adequacy findings are provisional and cannot be treated as final custodial conclusions.
This requirement is not procedural excess. Institutions regularly defend recourse adequacy by switching grammars midstream. They may invoke throughput and administrability to justify compressed review timelines, then invoke formal legal availability to claim fairness, then invoke consumer style voluntarism to privatize burdens that are in fact nonoptional for affected persons. A recourse assessment that does not force grammar candor becomes a managerial report with legal vocabulary. The instrument’s first honesty function is to expose how the institution is defining the closure setting before it begins describing review.
The unit of analysis for the assessment is also crucial and must be stated with care. The instrument is not a survey of individual satisfaction with appeal experiences and not a narrow audit of whether a single appeal channel exists. It is an inquiry into a recourse regime. That means it examines the architecture through which contradiction can occur across the relevant closure pathway, including pre-closure opportunities where they exist, post-closure review stages, evidentiary access, sequencing of sanctions and stays, decision ownership, timelines in ordinary and stressed conditions, and the interfaces with downstream institutions that bear the consequences of delayed or failed review. This broader unit is necessary because recourse failure is often distributed across stages rather than concentrated in a single denial point. A person may be allowed to file, acknowledged, routed, re-routed, asked for more documents, given a hearing date, issued a provisional response, and technically heard at every step while remaining unable to secure contradiction before the relevant harm is complete. A narrower instrument would record pathway presence. This one is designed to trace contradiction viability.
The core analytic discipline of the assessment is the distinction between nominal recourse and metabolically usable recourse. The institution is required to show not only that challenge pathways exist in law or policy, but that they are usable within the embodied and temporal conditions under which the burdened are expected to act. The phrase metabolically usable is retained from Chapters 6 and 9 because it prevents a recurring laundering move. Institutions frequently treat any pathway that can be described as available as if it were thereby adequate. The assessment blocks that inference. A pathway is not presumptively adequate if it requires reserves of time, money, documentation fluency, schedule flexibility, transport, childcare, language access, or affective self-regulation that the institution knows many affected persons will not have and has not provisioned for. To say this does not require a fantasy of frictionless process. It requires institutions to stop describing contradiction scarcity as natural when it is in fact a budgeted feature of closure design.
The instrument’s treatment of cost therefore extends beyond filing fees and direct payments, though those remain important and legally salient in many domains. It asks the institution to specify the mixed costs of challenge as they are experienced in the actual sequence of recourse. Time away from work, repeated appearances, document retrieval expenses, interpretation and translation burdens, lost benefits during review, reputational exposure, risk of retaliation, uncertainty maintenance, and repeated self-explanation are all recourse costs if they predictably shape who can contradict and when. An institution need not eliminate all such costs to claim adequacy. It must, however, identify them, justify their distribution, and show what accommodations, sequencing choices, or narrowing of closure claims have been adopted in light of them. Where the institution has no account of these costs, the instrument should not infer that the costs are absent. It should record recourse blindness as a deficiency in the institution’s contradiction budget record.
Temporal sequencing is treated by this instrument as constitutive rather than secondary. Institutions often discuss review and consequences as separate policy domains, and that separation is one of the main ways recourse theater is produced. A pathway can appear procedurally rich while being temporally null because sanction, deprivation, classification effects, or reputational marking attach immediately and review concludes only after the practical stakes have migrated, hardened, or become irreversible in relevant part. The recourse adequacy assessment therefore compels a time based account of closure and contradiction together. The institution must identify when closure becomes operative, when reasons become attributable, when challenge can first be initiated, when challenge can realistically be completed under ordinary and elevated load, whether interim relief or stays exist, who can grant them, and what harms continue to accrue while review is pending. If the institution cannot provide this sequence, or provides only nominal timelines detached from observed case flow, the assessment should treat the gap as a serious recourse adequacy problem and not a clerical reporting issue.
This sequencing inquiry is especially important in systems that rely on repeated provisional determinations, remands, or iterative reconsideration processes. Such regimes can preserve an appearance of patience and review while converting time itself into a depletion mechanism. The instrument therefore asks whether review stages provide substantive contradiction opportunities or merely recycle the burden of persistence onto the affected party while preserving the institution’s claim that process is ongoing. A recourse regime that repeatedly restarts the burdened without attributable correction, clear narrowing of issues, or meaningful protection against interim harm may be procedurally active and recourse inadequate at the same time. The assessment is designed to make that possibility legible.
The evidentiary and documentary dimension of recourse adequacy follows directly from Chapters 5 and 9. Contradiction cannot function if the object of contradiction is unstable, ownerless, or inaccessible. The institution must therefore show what records a person can obtain, at what stage, in what form, with what delay, and whether those records identify the operative reason, the rule source, the decision owner, and the evidentiary basis of closure in a way that permits challenge to begin. The assessment treats documentary abundance without attributable custody as a recourse hazard, not a procedural asset. In many contemporary systems, the burdened receive multiple notices, summaries, status updates, generated explanations, and partial records while remaining unable to identify which reason actually governs the closure under review. In such cases the institution has not simply made contradiction difficult. It has shifted the labor of reconstructing the institution’s action onto the person who seeks recourse. The instrument records this as documentary recourse degradation.
At the same time, the assessment must avoid rewarding institutions for withholding records under the theory that less documentation means less confusion. The relevant question is not quantity but contradiction relevance. Records are adequate for recourse when they permit the burdened and a reviewer to identify what was decided, by whom, on what grounds, with what evidentiary support, and through what policy or rule authority, in time for challenge to matter. The instrument is therefore aligned with attributable reasoning, not minimalist disclosure. This is one of the points at which the assessment supports rather than replaces doctrinal threshold analysis. Where documentary ownerlessness or instability prevents this level of contradiction relevance, the assessment may generate evidence of non attributable closure, but only the doctrine decides threshold consequence.
The instrument also has to examine institutional ownership of recourse itself. Many institutions now route grievances through offices that can receive, summarize, and de-escalate but cannot revise closure or compel reconsideration. These offices may still be valuable, and the appendix does not dismiss them. The problem arises when institutions count them as recourse in the same sense as adjudicative or decisional review. The recourse adequacy assessment therefore asks the institution to distinguish reception, mediation, explanation, escalation, recommendation, and revision authority, and to state which offices possess which powers. It also asks how often recommendations are adopted, how long escalation takes, and whether the person is informed of the distinction between being heard and obtaining review. This distinction is especially important in environments that have invested heavily in customer support style interfaces. A system can be highly responsive in tone and deeply inadequate in contradiction capacity.
Because Chapter 8 established the centrality of designed and distributed outsides, this appendix must satisfy the same honesty requirement as Appendix A and do so in the specific register of recourse. The recourse adequacy assessment is generally stronger at detecting designed blockages than distributed exclusions. It can identify explicit filing barriers, waiver clauses, truncated deadlines, missing stays, ownerless reasons, inaccessible records, absent review authority, sanction first sequencing, and visible cost burdens with relative clarity if records are available. It is weaker at detecting distributed recourse collapse produced by accumulations across agencies, vendors, courts, employers, social service systems, and other downstream actors that no single institution sees as part of its own recourse regime. This asymmetry has direct interpretive consequences. A favorable assessment result within one institution cannot be read as evidence that the person’s recourse environment is adequate overall when contradiction viability depends on cross system sequencing and cumulative burdens outside the institution’s own visibility. The instrument must therefore include a declared asymmetry statement and must identify the boundaries of the recourse environment it can and cannot observe.
To mitigate, though not solve, this weakness, the appendix includes a cumulative burden inquiry that asks the institution to state where its recourse pathway depends on external documentation, referrals, co-determinations, legal aid availability, employer tolerance, transportation systems, or parallel proceedings. The institution must also state what it knows about abandonment, delay, or adverse outcomes associated with these dependencies and what it does not know because no tracking or coordination exists. This is an inquiry into institutional dependency and visibility, not a demand for impossible total system knowledge. Its purpose is to prevent institutions from claiming recourse adequacy in a vacuum while relying on unmeasured external capacities to make contradiction viable. Where an institution’s recourse pathway presupposes resources or conditions it neither provides nor monitors, the assessment should record that presupposition as a recourse externalization risk.
The instrument is vulnerable to theater in forms that are now familiar enough to require explicit anti theater constraints. It may not be represented as proof that recourse is adequate because challenge pathways exist and are staffed. It may not be reduced to a customer satisfaction survey, a response time dashboard, or a complaint volume trend report, since each of these can improve while contradiction capacity remains unchanged or even worsens. It may not treat low appeal rates as presumptive evidence of fair closure absent analysis of contradiction pricing, sanction risk, and temporal depletion. It may not treat average timelines as adequate without distributional analysis of delay and reversibility, because averages routinely conceal recourse collapse in the very cases where closure stakes are highest. It may not treat the provision of explanatory text as recourse support absent attributable custody and contradiction relevance. And it may not merge offices with no revision authority into recourse counts in ways that inflate pathway inventory while obscuring where actual review power resides.
Synthetic administrative environments introduce a distinct capture pattern. Institutions can use automated triage, summarization tools, and generated communications to create the appearance of responsive recourse while narrowing access to human review, obscuring issue framing, or accelerating routing into closure preserving channels. A recourse adequacy assessment that ignores these mediation layers will register procedural activity while missing contradiction filtration. The instrument therefore requires institutions to disclose where automation shapes intake, routing, prioritization, summary generation, issue classification, or response drafting in the recourse pathway, and whether affected persons can contest the framing and categorization imposed by those systems before or during review. The concern is not automation as such. The concern is whether automation becomes a hidden allocator of contradiction capacity while the institution continues to describe recourse as if all pathways were equally open and humanly reviewable.
The assessment also identifies weak applicability domains because this appendix is required to state where its own methods lose traction. It is weak in settings where the institution cannot or will not preserve attributable data about timelines, reasons, outcomes, and revision authority. It is weak where recourse is intentionally informal and undocumented in ways that prevent burden tracing. It is weak where the operative closure regime spans multiple sovereign or quasi sovereign actors and no forum can compel a shared record. It is weak where retaliation risk is high and reporting itself suppresses truthful accounts of contradiction cost. It is weak where grammar identification remains deeply contested and politically insulated from review, since adequacy judgments then risk importing a false normative baseline. In these domains the instrument may still produce valuable descriptive fragments, especially about local sequencing or documentary access, but it should not be used to issue strong custodial claims without explicit declarations of insufficiency.
The relation of this appendix to Chapter 11 can now be stated precisely. The recourse adequacy assessment does not decide grammar identification, but it can discipline institutional descriptions of decision function, closure form, and standing structure. It does not itself impose the doctrine’s threshold screen, but it can generate evidence relevant to the core threshold failure of recourse priced beyond metabolically usable range and can illuminate how temporal depletion and documentary ownerlessness interact with that failure. It does not perform grammar sensitive balancing, but it can structure evidence on contestability path, reversibility, temporal cost, burden distribution, evidentiary custody, and embodied load in ways that make balancing less vulnerable to institutional fluency and more answerable to observed contradiction conditions. Like Appendix A, this instrument is best understood as a record building discipline under critique. It is dangerous when institutions use it to certify themselves against standards that only an independent and contestable doctrine can apply.
The appendix closes with declared insufficiency in the spirit required by the manuscript’s final constitution. A recourse adequacy assessment can make contradiction rationing visible, sharpen temporal sequencing analysis, expose documentary recourse degradation, and distinguish reception from review in ways that improve institutional candor and doctrinal scrutiny. It cannot by itself secure recourse budgets, eliminate distributed exclusions, or compel institutions to treat contradiction as a first order obligation when political incentives reward closure acceleration. It is stronger on designed blockages than on distributed recourse collapse, and it depends on contestable grammar identification for full legitimacy use. It should therefore be used as a bounded instrument for tracing the actual conditions of contradiction, never as a final assurance that because recourse has been assessed, recourse is adequate.
Epilogue
The Interval Kept Open
A book that spends this much time studying closure cannot end by pretending it has escaped it. The temptation is strongest precisely here, after doctrine has been stated and instruments have been drafted. The prose can begin to feel competent in the wrong way. It can begin to speak as if naming transfer, formalizing screens, and designing bounded audits had brought the remainder under sufficient control. It has not. If this manuscript has any discipline worth preserving at the end, it is the discipline of refusing to confuse better custody with completion.
The title has always contained that refusal. The Custody of Uncertainty does not promise the elimination of uncertainty, and it does not sanctify uncertainty as a virtue beyond administration, law, or collective judgment. It names a burden of handling. It names a question of who carries unresolvedness when institutions act, whose time absorbs the unresolved work, whose body becomes the site of administrative afterlife, and whose speech is forced into repeated forms in order to convert private endurance back into public intelligibility. That question remains after the chapters, after the doctrine, and after the appendices because institutions will continue to close, and lives will continue beyond the file.
What should change, if the argument has earned its place, is the moral and political status of that continuation. The person who continues after closure is not the residue of an otherwise completed process. The person is part of the process the institution has chosen not to count inside its own closure claim. This sentence carries the full weight of the book. It is why the prologue had to begin with a line of exposure rather than a definition. It is why Chapter 3 had to earn its analytic claim through instance. It is why Chapters 4 through 6 had to return to the prologue and test whether their concepts clarified or obscured what had first been shown. It is why Chapter 7 had to be written toward arrest rather than from a prepared sophistication. It is why Chapter 11 had to make grammar identification publicly reviewable and refuse the institution the right to name its own legitimacy grammar without contest. And it is why the appendices were written as instruments under critique rather than as operational salvation. The argument has tried, in each of these ways, to prevent institutions from taking the unresolved work they externalize and calling it someone else’s life problem.
There is a second discipline that belongs to the epilogue and not only to the doctrine. The manuscript has insisted that closure is often necessary. Institutions cannot inhabit nonfinality in the manner of persons without becoming derelict or violent in another register. The point has never been to romanticize suspension. Arendt’s account of action and world building remains useful here because she understood that durable institutions and durable worlds require forms of stabilization, recording, judging, and promising that cannot remain indefinitely in experiential openness (The Human Condition). Fuller also remains indispensable because institutional order without publicity, congruence, and accountable form does not protect freedom simply by avoiding a decision (The Morality of Law). The book has therefore not argued against closure as such. It has argued against the hidden transfer of closure’s remainder. The distinction matters most at the end because readers who agree with the diagnosis may still be tempted by an anti-institutional conclusion the manuscript does not support.
What it supports is harder. It supports accountable closure and protected nonclosure as co-present demands. Institutions must decide under insufficiency, and they must not borrow legitimacy from false completion. Persons will continue to carry what institutions cannot honestly absorb, and they must not be punished for continuing in conditions the institution itself has produced. These are not symmetrical burdens. Institutions carry public authority and can formalize transfer. Persons carry embodied cost and often must translate that cost into admissible form under conditions of depletion. The doctrine and appendices were built to move a portion of that justificatory burden back toward institutions and reviewers. The epilogue cannot claim that this redistribution has happened because the book exists. It can only state that any legitimacy claim now made in the vocabulary of fairness, safety, efficiency, or due process should be judged in light of whether it has accepted or refused that burden.
This is where the synthetic age enters again, and it enters not as a topical addendum but as the environment in which the old temptations have acquired new speed. Generated text, automated routing, predictive scoring, and optimization regimes do not create closure, but they alter its economics. They make it cheaper to produce the appearance of explanation. They make it easier to circulate decision artifacts detached from attributable adoption. They make policy drift harder to see in real time while making policy language easier to proliferate. They intensify the possibility that institutions will answer contradiction with more documentation rather than more custody. In such conditions the crisis of legitimacy is not only that institutions make mistakes. Institutions have always made mistakes. The crisis is that institutions can now preserve public confidence in closure form while deepening the opacity of how unresolvedness is converted into ordinary burden. That is why the book has tied legitimacy to custody of uncertainty rather than to output quality alone. Output can improve while custody worsens. Fluency can increase while recourse degrades. Visibility can be simulated while responsibility diffuses.
If the manuscript has a wager, it is that a change in grammar can alter what institutions are forced to defend. This is a modest wager and a serious one. It is modest because grammar does not produce coalitions, budgets, jurisdiction, or courage. It is serious because institutions live on admissible descriptions, and the line between what must be defended and what can be ignored is often drawn in language before it is drawn in law. When hidden transfer is named as hidden transfer rather than complexity, when ownerless closure is named as a custody defect rather than a technological side effect, when unusable recourse is named as contradiction rationing rather than pathway presence, when distributed exclusion is named as a structural property rather than anecdotal friction, the burden of justification moves. It does not move automatically, and it does not move far enough on its own, but it moves. Unger’s work has long insisted that institutional arrangements endure partly because they present themselves as the natural shape of practical reason rather than as revisable settlements of power (False Necessity). The present book attempts a narrower version of that anti fatalist labor. It tries to make certain closure arrangements harder to describe as inevitable once their custody structure is visible.
Still, the epilogue must not mistake harder to justify for politically unavailable no longer. The political availability gap remains, and it remains in differentiated form. Some readers will meet this book from inside institutions that benefit from custody evasion while sincerely experiencing themselves as constrained by throughput, law, or scarcity. Some will meet it from professions that mediate between rule and person and have built traditions of procedural virtue around distinctions this manuscript has tried to pressure. Some will meet it from collective formations that already know, in practical and often dangerous ways, that contradiction is a budgeted privilege and not a neutral public good. Some will meet it from lives in which institutional remainder has been carried for years without stable language and where any new language arrives together with a renewed demand to explain oneself. The epilogue cannot unify these positions into a single moral voice without falsifying the very asymmetries it has spent the book exposing.
What it can do is hold them in the same field of obligation. The administrator is not absolved by scarcity if scarcity is used to conceal transfer rather than declare insufficiency. The reviewer is not absolved by form if form preserves ownerless reasons and unusable challenge. The designer is not absolved by innovation if innovation cheapens closure and externalizes verification. The advocate is not absolved by diagnosis if diagnosis remains unorganized and therefore easy for institutions to metabolize as reform theater. The reader who has carried remainder is not obligated to turn endurance into pedagogy for the comfort of the institution. These are different obligations under unequal conditions. Their coexistence is precisely what the title’s word custody has been trying to hold. Custody is not a sentiment. It is an allocation of burdens, powers, and answerability under conditions where certainty is unavailable and action continues.
There is one final risk at the end of a manuscript such as this, and it is a risk of style rather than doctrine. The prose can begin to aestheticize incompletion. It can make nonclosure sound like refinement. That would be a betrayal of the line of exposure from the prologue. For many people the interval after institutional closure is not philosophically glamorous. It is expensive. It is humiliating. It is metabolically draining. It can reorganize work, care, sleep, and speech. It can make ordinary tasks feel contingent on documentary weather. It can produce a second job of self explanation. The book has argued for protected nonclosure because persons are not institutions and because forcing finality where there is no honest finality is a form of violence. But protection is not romanticization. The epilogue must say this plainly so the argument does not drift into an ethics of eloquent endurance. The point is not to dignify suffering after the fact. The point is to alter the institutional conditions under which suffering becomes the hidden substrate of closure.
That alteration will not arrive through one channel. Doctrine can discipline and expose. Instruments can trace and constrain. Litigation can force records into view and alter consequence sequencing. Administrative redesign can move contradiction capacity upstream and narrow closure claims under declared insufficiency. Labor and community organization can convert distributed burden into leverage and make recourse budgets politically costly to starve. Public language can change what counts as an acceptable institutional answer. Crises can open revision that ordinary periods suppress. None of these are sufficient alone, and no serious reader should want a final chapter that pretends otherwise. The political task after this book is compositional. It concerns how these forms of action are arranged so that institutions cannot continue to benefit from custody evasion while appearing responsive to critique.
If the manuscript has any right to ask something of its readers now, it is not agreement in the abstract. It is a practical refusal of misdescription. Refuse the sentence that treats hidden transfer as unfortunate complexity when the transfer is structured and knowable. Refuse the sentence that treats pathway presence as recourse adequacy when contradiction has been priced out of use. Refuse the sentence that treats generated explanation as accountable reason when no one can own the decision chain. Refuse the sentence that treats distributed exclusion as no one’s responsibility because no single office can see the whole. Refuse the sentence that calls political blockage epistemic humility. These refusals are not the politics itself, but they are among the conditions under which politics can become more exact.
The epilogue returns, finally, to the interval. The book cannot close the interval because the interval is where lives continue after institutions stop speaking. It can only insist that the interval be brought into public judgment as part of what closure is. Weil wrote in ways that remain difficult and sometimes severe about attention as a moral discipline, a refusal to replace reality with the consolations of premature interpretation (Gravity and Grace; Waiting for God). This manuscript has borrowed that discipline in a civic register. It has asked for institutional and scholarly attention to where the remainder goes before theory congratulates itself on order. To keep the interval open is not to refuse decision. It is to refuse the lie that decision finishes what it has displaced.
That is the epilogue’s final claim and its final limit. The custody of uncertainty is not a final settlement. It is a public obligation to govern closure honestly, to expose and contest burden transfer, to preserve contradiction where institutions would prefer speed, and to protect the forms of nonclosure without which embodied life is forced to counterfeit completion. Everything else remains the work of institutions, forums, coalitions, and persons acting under conditions no book can complete.
Absolutely. I cleaned and consolidated the list into a single deduplicated MLA 9 style Works Cited, standardized capitalization, punctuation, publisher naming, journal titles, and repeated entries. I also preserved distinct editions where your list clearly cites different editions of the same work.
Works Cited
Ahmed, Sara. The Cultural Politics of Emotion. 2nd ed., Edinburgh University Press, 2014.
Agamben, Giorgio. State of Exception. Translated by Kevin Attell, University of Chicago Press, 2005.
Arendt, Hannah. Between Past and Future. Penguin Books, 2006.
Arendt, Hannah. The Human Condition. 2nd ed., University of Chicago Press, 1998.
Arendt, Hannah. The Origins of Totalitarianism. Schocken Books, 2004.
Autio, Chloe, et al. Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile. National Institute of Standards and Technology, 2024.
Bender, Emily M., et al. “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?” Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, Association for Computing Machinery, 2021, pp. 610-623.
Benecke v. Barnhart. 379 F.3d 587. United States Court of Appeals for the Ninth Circuit. 2004.
Boddie v. Connecticut. 401 U.S. 371. Supreme Court of the United States. 1971.
Bommasani, Rishi, et al. “On the Opportunities and Risks of Foundation Models.” arXiv, 2021, arXiv:2108.07258.
Bowker, Geoffrey C., and Susan Leigh Star. Sorting Things Out: Classification and Its Consequences. MIT Press, 1999.
Brown, Tom B., et al. “Language Models Are Few-Shot Learners.” Advances in Neural Information Processing Systems, vol. 33, 2020, pp. 1877-1901.
Burrell, Jenna. “How the Machine ‘Thinks’: Understanding Opacity in Machine Learning Algorithms.” Big Data & Society, vol. 3, no. 1, 2016.
Cahoo et al. v. SAS Analytics Inc. et al. United States District Court for the Eastern District of Michigan. No. 2:17-cv-10657. Order on Final Approval of Settlement, 9 July 2020.
Cahoo v. SAS Analytics Inc. United States Court of Appeals for the Sixth Circuit, Nos. 18-1295 and 18-1296, 3 Jan. 2019.
Canguilhem, Georges. The Normal and the Pathological. Translated by Carolyn R. Fawcett and Robert S. Cohen, Zone Books, 1991.
Citron, Danielle Keats. “Technological Due Process.” Washington University Law Review, vol. 85, no. 6, 2008, pp. 1249-1313.
Clark, Andy. “Whatever Next? Predictive Brains, Situated Agents, and the Future of Cognitive Science.” Behavioral and Brain Sciences, vol. 36, no. 3, 2013, pp. 181-204.
Cullen, William Douglas. The Public Inquiry into the Piper Alpha Disaster. Her Majesty’s Stationery Office, 1990.
Derrida, Jacques. “Force of Law: The ‘Mystical Foundation of Authority.’” Deconstruction and the Possibility of Justice, edited by Drucilla Cornell et al., Routledge, 1992, pp. 3-67.
Ellsberg, Daniel. “Risk, Ambiguity, and the Savage Axioms.” Quarterly Journal of Economics, vol. 75, no. 4, 1961, pp. 643-669.
Espeland, Wendy Nelson, and Mitchell L. Stevens. “Commensuration as a Social Process.” Annual Review of Sociology, vol. 24, 1998, pp. 313-343.
Eubanks, Virginia. Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor. St. Martin’s Press, 2018.
Fanon, Frantz. The Wretched of the Earth. Translated by Richard Philcox, Grove Press, 2004.
Feeley, Malcolm M. The Process Is the Punishment: Handling Cases in a Lower Criminal Court. Russell Sage Foundation, 1979.
Fiss, Owen M. “Against Settlement.” Yale Law Journal, vol. 93, no. 6, 1984, pp. 1073-1090.
Foucault, Michel. Security, Territory, Population: Lectures at the Collège de France, 1977-1978. Translated by Graham Burchell, Palgrave Macmillan, 2007.
Friston, Karl. “The Free-Energy Principle: A Unified Brain Theory?” Nature Reviews Neuroscience, vol. 11, no. 2, 2010, pp. 127-138.
Fuller, Lon L. The Morality of Law. Revised ed., Yale University Press, 1969.
Galanter, Marc. “Why the ‘Haves’ Come Out Ahead: Speculations on the Limits of Legal Change.” Law & Society Review, vol. 9, no. 1, 1974, pp. 95-160.
Gebru, Timnit, et al. “Datasheets for Datasets.” Communications of the ACM, vol. 64, no. 12, 2021, pp. 86-92.
Goffman, Erving. Stigma: Notes on the Management of Spoiled Identity. Prentice-Hall, 1963.
Goldberg v. Kelly. 397 U.S. 254. Supreme Court of the United States. 1970.
Hallaq, Wael B. The Impossible State: Islam, Politics, and Modernity’s Moral Predicament. Columbia University Press, 2013.
Hart, H. L. A. The Concept of Law. 3rd ed., edited by Penelope A. Bulloch and Joseph Raz, Oxford University Press, 2012.
Hartman, Saidiya. Lose Your Mother: A Journey Along the Atlantic Slave Route. Farrar, Straus and Giroux, 2007.
Herman, Judith Lewis. Trauma and Recovery: The Aftermath of Violence from Domestic Abuse to Political Terror. Basic Books, 1992.
Hirschman, Albert O. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. Harvard University Press, 1970.
Jasanoff, Sheila. The Fifth Branch: Science Advisers as Policymakers. Harvard University Press, 1990.
Keynes, John Maynard. A Treatise on Probability. Macmillan, 1921.
Knight, Frank H. Risk, Uncertainty and Profit. Houghton Mifflin, 1921.
Kroll, Joshua A., et al. “Accountable Algorithms.” University of Pennsylvania Law Review, vol. 165, no. 3, 2017, pp. 633-705.
Leveson, Nancy G. Engineering a Safer World: Systems Thinking Applied to Safety. MIT Press, 2011.
Lipsky, Michael. Street-Level Bureaucracy: Dilemmas of the Individual in Public Services. 30th anniversary expanded ed., Russell Sage Foundation, 2010.
Lujan v. Defenders of Wildlife. 504 U.S. 555. Supreme Court of the United States. 1992.
M.L.B. v. S.L.J. 519 U.S. 102. Supreme Court of the United States. 1996.
Mamdani, Mahmood. Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism. Princeton University Press, 1996.
Mashaw, Jerry L. Bureaucratic Justice: Managing Social Security Disability Claims. Yale University Press, 1983.
Mathews v. Eldridge. 424 U.S. 319. Supreme Court of the United States. 1976.
Mbembe, Achille. On the Postcolony. University of California Press, 2001.
McEwen, Bruce S. “Protective and Damaging Effects of Stress Mediators.” New England Journal of Medicine, vol. 338, no. 3, 1998, pp. 171-179.
Merleau-Ponty, Maurice. Phenomenology of Perception. Translated by Donald A. Landes, Routledge, 2012.
Michigan Office of the Auditor General. Claimant Services: Unemployment Insurance Agency, Talent Investment Agency, Department of Talent and Economic Development. Performance Audit Report No. 641-0318-14, Apr. 2016.
Mitchell, Margaret, et al. “Model Cards for Model Reporting.” Proceedings of the Conference on Fairness, Accountability, and Transparency, Association for Computing Machinery, 2019, pp. 220-229.
Mnookin, Robert H., and Lewis Kornhauser. “Bargaining in the Shadow of the Law: The Case of Divorce.” Yale Law Journal, vol. 88, no. 5, 1979, pp. 950-997.
Motor Vehicle Manufacturers Association of the United States, Inc. v. State Farm Mutual Automobile Insurance Co. 463 U.S. 29. Supreme Court of the United States. 1983.
Mullane v. Central Hanover Bank and Trust Co. 339 U.S. 306. Supreme Court of the United States. 1950.
Ngai, Mae M. Impossible Subjects: Illegal Aliens and the Making of Modern America. Princeton University Press, 2004.
Pasquale, Frank. The Black Box Society: The Secret Algorithms That Control Money and Information. Harvard University Press, 2015.
Porter, Theodore M. Trust in Numbers: The Pursuit of Objectivity in Science and Public Life. Princeton University Press, 1995.
Power, Michael. The Audit Society: Rituals of Verification. Oxford University Press, 1997.
Rasmussen, Jens. “Risk Management in a Dynamic Society: A Modelling Problem.” Safety Science, vol. 27, nos. 2-3, 1997, pp. 183-213.
Reason, James. Managing the Risks of Organizational Accidents. Ashgate, 1997.
Scott, James C. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. Yale University Press, 1998.
Scott v. Michigan Unemployment Insurance Agency. Michigan Court of Appeals, No. 364292, 21 Sept. 2023.
SEC v. Chenery Corp. 332 U.S. 194. Supreme Court of the United States. 1947.
Selbst, Andrew D., et al. “Fairness and Abstraction in Sociotechnical Systems.” Proceedings of the Conference on Fairness, Accountability, and Transparency, Association for Computing Machinery, 2019, pp. 59-68.
Seth, Anil K., and Karl J. Friston. “Active Interoceptive Inference and the Emotional Brain.” Philosophical Transactions of the Royal Society B: Biological Sciences, vol. 371, no. 1708, 2016, article 20160007.
Spivak, Gayatri Chakravorty. “Can the Subaltern Speak?” Marxism and the Interpretation of Culture, edited by Cary Nelson and Lawrence Grossberg, University of Illinois Press, 1988, pp. 271-313.
Tabassi, Elham, et al. Artificial Intelligence Risk Management Framework (AI RMF 1.0). National Institute of Standards and Technology, 2023.
Turner, Victor. The Ritual Process: Structure and Anti-Structure. Aldine de Gruyter, 1969.
Tyler, Tom R. Why People Obey the Law. Princeton University Press, 1990.
Unger, Roberto Mangabeira. False Necessity: Anti-Necessitarian Social Theory in the Service of Radical Democracy. Verso, 2004.
Unger, Roberto Mangabeira. What Should Legal Analysis Become? 2nd ed., Verso, 2002.
van Gennep, Arnold. The Rites of Passage. Translated by Monika B. Vizedom and Gabrielle L. Caffee, University of Chicago Press, 1960.
Weber, Max. Economy and Society: An Outline of Interpretive Sociology. Edited by Guenther Roth and Claus Wittich, University of California Press, 1978.
Weil, Simone. Gravity and Grace. Translated by Emma Craufurd, Routledge, 2002.
Weil, Simone. Waiting for God. Translated by Emma Craufurd, Harper & Row, 1951.
Weil, Simone. Waiting for God. Translated by Emma Craufurd, Harper Perennial, 2009.
Winnicott, D. W. Playing and Reality. Routledge, 2005.
Jonathan Johnson-Swagel 
Leave a comment