A Critique of Practical Institutional Reason

Authored Conditions, Substitution, and the Validity of Institutional Judgment

Photo by Fine Photographics on Unsplash

A Note to the Reader Before the Trial Convenes

What follows is a 69-page, 34,500-word immanent critique of institutional reason. I am under no illusions about the patience of the internet, and I know exactly what it means to drop a treatise of this scale onto a platform built for quick consumption. The essay relies on Kantian cognitive architecture, formal decision theory, and the structural mapping of administrative power. It is dense, and it demands your attention.

But this was not written from the safety of a seminar room. The philosophical machinery you are about to read was excavated directly from the floor of the institution. It is the result of over fifteen years spent standing at the clinical counter, navigating infectious disease protocols, fighting against the fifteen-minute scheduling window, and watching the institutional file systematically displace the living patient.

I wrote this because the violence of modern institutional life is not an accident; it is a designed equilibrium. And a designed equilibrium can be put on trial.

If academic philosophers happen to stumble across this docket, they will likely recognize the mechanics of the Copernican inversion, but they may bristle at the operatic framing of the courtroom. They will find my formal proofs waiting for them in the appendix, but they will still ask why I didn’t simply write a standard, dispassionate journal article.

To them, my answer is this: I can speak your dialect fluently; I just chose to speak in a voice that has blood in it.

The trial convenes below.

---

The Motion to Dismiss

Every trial begins before it begins. Before evidence, before argument, before verdict, there is a prior question: does this court have the power to hear this case at all? Lawyers call it jurisdiction. Philosophers rarely name it, and they should, because every critique in the strict sense is a trial, and every trial can be ended on that one question before a single fact is heard. A tribunal without jurisdiction is not a tribunal that rules badly. It is not a tribunal at all. Its findings are wind. So the question of jurisdiction is not preliminary in the sense of minor. It is preliminary in the sense of founding: everything the court will later say depends on its right to speak, and that right must be established first, or nothing said afterward carries weight. This is why the defense reaches for it before all else. A defendant who can show the court has no power to hear the case need never answer the charge. The charge simply falls, unheard, and the defendant walks out having conceded nothing, because nothing was ever properly asked.

So let me tell you plainly what I am doing here. I claim that institutions reason. I claim that there are conditions under which their reasoning about persons remains valid, and that those conditions can be named, mapped, and enforced. I intend to put institutional reason on trial: its receptions, its judgments, its records, its repairs. Not one institution. Not one bad clinic or one cruel agency or one failing school, held up as a scandal and then forgotten when the news cycle turns. Institutional reason itself: the mode of cognition by which modern persons are known, sorted, recorded, judged, and kept, wherever it operates and under whatever letterhead. That is the scale of the proceeding. And a proceeding at that scale can be dismissed at that scale, in a single stroke, before the first witness is called, if the court’s power to convene is not secured at the outset. I am securing it at the outset. That is the whole work of this essay, and it is enough work for one essay, because if this fails, nothing after it can begin.

And I should tell you where you sit, because this proceeding has no marble, no bailiff, and no seal. You are not the gallery. You are the court. A case against institutional reason does not convene in the defendant’s building. It cannot. The defendant does not hear cases against itself in any forum that could return an adverse verdict, and an institution that judged itself under its own procedures would be judge, defendant, and clerk at once, which is precisely the arrangement critique exists to break. So the trial convenes elsewhere: here, in the open, before the one bench no institution appointed. That bench is persons, one at a time, reading. No commission seated you. No form qualified you. No board reviewed your credentials and found them adequate to the office you now hold. Your qualification is of another kind entirely, and it is the only kind that could be legitimate for this particular trial: you have stood at the counter, received the letter, waited inside the schedule, been classified by the category, and signed where indicated. You know the defendant by long acquaintance, from below, as the decided-upon know the deciding apparatus. That acquaintance is your standing. The person who has been judged by institutional reason is exactly the person competent to sit in judgment on it, and no one else is. So the motion to dismiss will be argued to you, and you will rule on it. I can hear the motion already. It is the strongest motion the defense has, and I intend to state it better than the defense would, because a jurisdiction won against a weak version of the objection is worth nothing, and I do not intend to win anything worth nothing.

The motion goes like this. Only persons reason. Institutions process. A hospital does not think; clinicians think, and the hospital is merely the building they think in, the walls around the cognition, no more a reasoner than a library is a reader. To speak of institutional reason, the motion continues, is to speak in metaphor, a figure of speech we reach for out of convenience and then mistake for a fact. And to write a critique of institutional reason is to commit a category error: an indictment served on the weather, a subpoena issued to an avalanche, a summons nailed to the door of a storm. Kant could put reason on trial because reason was defendant and judge at once, a claim-making capacity able to examine its own claims by its own standards (Kant [1781] 1998). That is what made his tribunal coherent: the thing on trial could answer, could be wrong, could reconsider, could sit in judgment on itself. An institution, says the motion, is none of this. It makes no claims; it only produces outputs, the way a mill produces flour. It has no self to examine; there is no institutional mind behind the paperwork, only paperwork. The whole project, on this account, is a long apostrophe addressed to furniture, an earnest speech delivered to a filing cabinet, and the court should dismiss for want of a defendant. There is no one in the dock. There is only a machine, and you do not try a machine, because a machine has committed no act, holds no view, and owes no answer.

That is the motion. I have stated it at full strength, and I will add that it is not stupid. It is the intuition most people carry without ever examining it, and it has the great advantage of being half true: it is indeed persons, not buildings, who think. If the motion were wholly wrong it would not be dangerous. It is dangerous because it is half right, and the half that is right is doing the work of concealing the half that is wrong.

I am going to deny that motion. But I will not deny it by assertion, and I will not deny it by sentiment. Sentiment is what the defense expects: that the case for institutional critique rests on grievance, on the accumulated hurt of everyone the apparatus has wounded, and that if the grievance can be acknowledged and soothed, the jurisdictional question can be waved past. It cannot. Grievance is not a case. A wound is not a warrant. Your acquaintance with the defendant seats you; it does not carry me. I will deny the motion by proof, and I will prove it out of the defendant’s own filing cabinet, because that is how this kind of case has always been won and the only way it can be won cleanly. The civil rights bar did not defeat segregation by reciting its convictions, however just those convictions were. It entered the state’s own statutes into evidence, the company’s own job requirements, the district’s own maps, the registrar’s own records, and it let the defendant’s documents testify against the defendant; the doctrine that a requirement neutral on its face can be unlawful in operation was won exactly this way (Griggs v. Duke Power Co., 401 U.S. 424 [1971]). The paper convicted the hand that filed it. I will do the same. The institution’s own paperwork will establish this court’s jurisdiction over the institution, and the exhibits will be entered in their turn: the denial letter, the chart, the case file, the transcript, the intake form, and above all the appeal form, that last one printed and distributed by the very office whose judgment it exists to contest. Some of this paper has your name on it. You have held these exhibits in your hands before you ever read this sentence. That is why you are competent to weigh them.

Here is the burden I accept, and I state it precisely so that you can hold me to it and find me in default if I fail to carry it. Critique, in the strict sense, is legitimate only where its object bears certain marks. This is not my stipulation; it is what the practice of critique has always required, and I will read the requirements off Kant’s own procedure rather than invent them (Kant [1781] 1998). I will name those marks. I will show that Kant’s object, human reason itself, bore each of them, which is why his tribunal was coherent. And I will show, mark by mark, from the defendant’s own exhibits, that institutional reason bears every one. If it bears them all, the category error dissolves, the motion to dismiss fails, and the court has jurisdiction. That is the first task, and it establishes permission.

Then I will do something Kant had no occasion to do, because his object made it impossible. I will enter a sixth mark that Kant’s object never bore and never could bear, and the sixth mark will convert the permission to critique into the obligation to critique. It is one thing to show that institutional reason may be put on trial. It is another to show that it must be. The five marks open the court. The sixth compels its use. And the hinge between them is a single fact about institutional reason that has no parallel in the reason Kant examined, a fact I will prove from the paperwork like everything else: the conditions of institutional reason have authors.

Along the way, two sets of prior conditions will come into evidence, and they are the payload of this proof, the freight the whole tribunal was built to deliver. The first set: the conditions under which institutional cognition is possible at all, the forms that must be in place before any person can appear to an institution as anything, settled before any person arrives at any counter. The second set: the moral conditions that stand upstream of every institutional judgment and condition its validity before the judgment is ever made. The first set proves the defendant exists, that there is a reasoning apparatus here and not merely a heap of unrelated acts. The second set proves the trial is owed, that this apparatus stands under obligations it did not invent and cannot repeal. Cognition fixes what the institution can know. Morality fixes what its knowing owes. Both are prior. Both are conditions. Both are settled at the moment of design, before the person walks in and out of sight of anyone standing at the counter. One apparatus, read twice: once for what it can do, once for what it must answer for.

The record follows.

What Critique Requires

Critique, in ordinary usage, has collapsed into a synonym for finding fault. To critique a film is to list what failed in it. To offer critique of a proposal is to say what you would change. The word now means little more than considered disapproval, an opinion with reasons attached, and because that is what it has come to mean, a reader encountering the phrase critique of institutional reason will expect what the word now promises: a catalog of institutional failings, a bill of complaints, a case for the prosecution built out of everything the apparatus has done wrong. That expectation is precisely wrong, and I have to break it now, at the threshold, or everything that follows will be misread as grievance dressed in philosophical clothing. So let me say what the word meant before it went soft, because I am using it in the hard sense and the whole proceeding depends on the difference.

Critique is not complaint. Complaint reports a wound and asks for redress; it takes the standards of judgment for granted and argues that they were misapplied. Critique is not exposure. Exposure reveals a hidden fact, drags into daylight what was meant to stay concealed, and rests its force on the revelation; but a thing fully exposed, every operation visible, every document public, could still stand in perfect need of critique, because critique is not about what is hidden. Critique is not commentary with a grievance, the running annotation of a hostile observer who has decided the verdict in advance and marshals detail to support it. It is none of these, and it is more demanding than all of them. Critique is a tribunal: the examination of a claim-making capacity, conducted by that capacity’s own standards, to fix the boundaries of its valid use and to name the illusions it generates from its own structure. Every clause in that sentence is load-bearing, so I will slow down on each.

A claim-making capacity: critique takes as its object not an event, not a thing, not a person, but a power that issues claims, assertions that hold themselves out as correct. Conducted by that capacity’s own standards: this is what separates critique from external attack. The tribunal does not import a foreign measure and convict the object of failing to meet it. It holds the object to the standards the object itself presupposes, judges the capacity by its own lights, and finds it wanting only where it violates what it already claims to honor. Kant makes this the constitution of the tribunal itself: reason has no dictatorial authority, and its verdict is never anything but the agreement of free citizens, each of whom must be able to voice his objection and even his veto without penalty (A738–39/B766–67). To fix the boundaries of its valid use: the aim is not to abolish the capacity but to map it, to mark where its verdicts are sound and where they exceed their warrant, to draw the line between legitimate exercise and overreach. This is the express task Kant sets himself, a tribunal that will secure reason in its rightful claims while dismissing those that have no ground, and do so not by decree but according to reason’s own eternal and unchangeable laws (Axi–xii). And to name the illusions it generates from its own structure: the tribunal is required, rather than merely permitted, precisely because the capacity produces, out of its own constitution, a characteristic distortion that ordinary correction cannot reach. Not a mistake it happens to make. A distortion it must make, given what it is.

Kant’s first Critique does exactly this, and it is worth seeing that it does, because the whole method I am about to run is his method transposed. He does not accuse reason of crimes. He does not stand outside reason and indict it in the name of something higher, because there is nothing higher than reason in whose name one could speak. He asks, from within reason and by reason’s own standards, under what conditions reason’s verdicts are good, where its writ runs out, and why it keeps ruling beyond its jurisdiction anyway. That last question is the signature of genuine critique, and Kant answers it with a doctrine of illusion that is native and not accidental: there is a transcendental illusion that does not cease even after it has been detected and its nullity laid bare, an illusion inseparable from human reason, one that continues to press upon reason after every exposure (A297/B354). Reason does not merely err at its boundaries; it is driven, by its own structure, to press past them, to issue verdicts about matters on which it can have no valid verdict, and to keep doing so even after the boundary has been shown. The tribunal exists to discipline that structural overreach, which is why it must be standing and not occasional. You do not convene a permanent court to correct an occasional mistake. You convene it because the object before it generates its characteristic illusion perpetually, from its own machinery, and requires perpetual discipline in consequence.

Read off Kant’s practice, then, the object of a critique must satisfy five marks. These are not a checklist I have imposed from outside; they are the conditions that made Kant’s own tribunal coherent, the features his object had to possess for the examination to be possible and necessary at once. I ask you to hold each one as I state it, because the entire jurisdictional argument consists in showing, one mark at a time and from the defendant’s own paper, that institutional reason possesses every one of them. If it does, the tribunal is as coherent against institutional reason as Kant’s was against reason itself. If it fails even one, the motion to dismiss succeeds and I have no case. So here are the five, stated plainly, each with the reason it matters.

First, the object must make claims: assertions that can be right or wrong, not mere events that occur. This is the threshold mark, the one the motion to dismiss denies most directly. Where nothing claims validity, there is nothing to examine, because critique examines validity claims and nothing else. A landslide makes no claim; it simply happens, and however much damage it does, there is no assertion in it to be found correct or incorrect. If institutional outputs are landslides, mere events in the causal order, the tribunal has no object and the case ends. Everything turns, at the first mark, on whether the denial letter asserts or merely occurs.

Second, the object must be one activity: unified enough that there is an it to put on trial, not a heap of unrelated acts. A tribunal needs a defendant, a single thing that acts and can be held to answer for its acts. If what looks like institutional reason is really just a scattering of separate human judgments that happen to occur under one roof, then there is no it, only a crowd, and the proper procedure is to examine each member of the crowd separately, which dissolves the case against the institution as such. The second mark asks whether institutional cognition has genuine unity or is merely an aggregate misdescribed as a single agent.

Third, the object must operate under prior conditions: a structure that precedes and shapes its every operation, which the critique can map. This is the mark that makes critique more than opinion, and it is the hinge of Kant’s whole enterprise. His revolution was to reverse the order of dependence between cognition and its object, proposing that objects must conform to our cognition rather than our cognition to objects, so that the forms of experience are contributed by the cognizer in advance of any encounter (Bxvi–xviii). Reason operates under conditions, forms that precede and shape every act of cognition, and because those conditions can be mapped, the boundaries of valid use can be drawn with precision rather than asserted with confidence. An object with no prior structure offers nothing to map; one can only react to its outputs case by case. An object with a prior structure can be understood at the level of its conditions, and its characteristic overreach can be located in those conditions rather than chased through its endless particular acts. The third mark asks whether institutional reason has a mappable prior structure of this kind.

Fourth, the object must generate illusion from its own structure: not occasional error, which ordinary correction handles, but a standing distortion produced by the very apparatus that makes the activity possible. This is the mark that makes critique necessary rather than merely possible. Kant is precise about the difference: the transcendental illusion is not the sort of logical mistake that vanishes once one attends to it, but a natural and unavoidable illusion that persists even in the one who knows it for what it is, as the sea still appears higher at the horizon to the eye that understands why it is not (A295–97/B352–54). If an object erred only occasionally and by accident, ordinary correction would suffice; you would fix the error and move on, and no standing tribunal would be warranted. What warrants a standing tribunal is a distortion the object produces structurally, unavoidably, from the same machinery that lets it function at all, so that the distortion cannot be corrected away because it is not a malfunction but a feature. Only structural illusion necessitates a standing discipline. The fourth mark asks whether institutional reason has a native illusion of this kind, one built into its conditions and therefore ineradicable by good intentions or better staff.

Fifth, the object must be capable, at least in principle, of examining itself: otherwise the examination is external legislation, and Kant is emphatic that reason accepts no dictates. Reason’s verdicts, he insists, are not imposed from above but reached as the agreement of free participants, each entitled to voice reservations and even to withhold assent, so that the authority of the tribunal is the authority of free examination and nothing else (A738–39/B766–67). This matters because a tribunal that judged its object by a standard the object could not in principle apply to itself would not be critique but conquest, one power subjugating another in the name of an alien measure. Genuine critique is self-examination: the capacity turned upon itself, judged by standards it already holds, disciplined by a tribunal it could in principle convene from within. The fifth mark asks whether institutional reason can host its own examination, whether the tribunal I am convening completes something the object already contains rather than imposing something foreign upon it.

Those are the five marks. Make claims, possess unity, operate under prior conditions, generate structural illusion, and be capable of self-examination. Kant’s object bore all five, which is why his Critique was possible and necessary at once. The jurisdictional question is whether institutional reason bears them too, and I will answer it mark by mark, from the exhibits, refusing at every step to argue from anything but the defendant’s own documents.

And note, before the defense objects that I am stretching Kant to cover institutions, that Kant stretched first. This matters, because the natural line of attack is to say that critique was designed for reason as such, a universal human faculty, and that to point it at hospitals and welfare offices is to smuggle a method out of its proper domain and press it into service where it does not belong. But the domain was never so narrow, and Kant said so himself, in the founding document, at the founding moment. In a footnote to the first preface he declared his age the age of criticism, to which everything must submit, and he named the two powers most tempted to claim exemption from it: religion, through its holiness, and legislation, through its majesty. Exemption, he wrote, only awakens just suspicion, and nothing that exempts itself can claim the honest respect that reason grants only to what has survived free and public examination (Axi note). Read the second of those two powers again. Legislation. The enacted order, the machinery of statute and rule, the institution in its most authoritative form. Kant named the institution as a would-be exemptee from critique in the very text that founds critique, and he named it precisely because it would be tempted, by its majesty, to claim the exemption that critique cannot grant to anything. The method was pointed at the institution from the first page. I am not stretching it toward institutions. I am collecting on a summons Kant issued and left unserved.

The lineage that followed took the license and ran with it, each turning critique upon an object larger and more structural than a single reasoning mind. Marx subtitled his masterwork a critique of political economy, aiming the method not at a thinker but at an entire system of production and the categories by which it understands itself (Marx [1867] 1976). Horkheimer turned critique upon the social order that produces the theorist, insisting that the one doing the examining is himself constituted by the arrangement under examination, so that critique of society is never conducted from a clean position outside it (Horkheimer [1937] 2002). Foucault excavated the reason embedded in government, in discourse, in the quiet procedures by which knowledge and administration produce the subjects they claim only to describe (Foucault [1969] 1972). The tradition, in other words, has spent two centuries pointing critique at exactly the kind of object I am pointing it at now, and the method has held.

But I do not rest on precedent, because precedent is not proof. That a line of formidable thinkers pointed critique at systems does not establish that the pointing was licit; it establishes only that it was done, and a shared error, however distinguished its authors, is still an error. The lineage shows that the ambition is not idiosyncratic. It does not show that the ambition is warranted. What warrants it is the marks, and only the marks: the demonstration that institutional reason actually bears every feature Kant’s object bore, so that the tribunal against it is coherent for the same reasons Kant’s was coherent, and not by loose analogy or borrowed prestige. So I will not lean on Marx’s authority, or Foucault’s, or Horkheimer’s, or Kant’s own. I will lean on the denial letter and the appeal form. The proof is the marks. So I run them.

The First Mark: The Defendant Speaks in Claims

Watch what an institution actually says, because its grammar will convict it.

The denial letter does not report that an event has occurred. It says: you do not qualify. Read the difference carefully, because the whole first mark lives in it. A report of an event describes something that happened in the world and leaves the describing outside the thing described; the rain fell, the file closed, the meeting ended, and the sentence recording it makes no claim except that the event took place. But you do not qualify is not the record of an event. It is a verdict. It asserts that a determination has been made and that the determination is correct: that measured against the standard, you fall on the wrong side of it, and that this is not the clerk’s opinion but the truth of your case. The chart does not record that a word was uttered. It says: noncompliant. Not the patient said he had not taken the medication, which would be a report, but noncompliant, which is a finding, a classification asserted as accurate, a claim about what the person is that holds itself out as correct and stands ready to be relied upon by everyone downstream who reads it. The case file does not note that an investigation ended. It says: unfounded. Not the inquiry concluded on this date, but unfounded, a word that reaches back over everything the person alleged and pronounces it groundless, a validity claim of the most consequential kind, asserting that what was said to have happened did not happen or does not count. The transcript does not describe. It says: failing. Not the student scored below a threshold on these dates, but failing, a status conferred and asserted as earned, a judgment the institution stands behind and will act upon.

Line them up and hear what they have in common. Approved, denied. Eligible, ineligible. Competent, incompetent. Credible, not credible. Founded, unfounded. Compliant, in violation. Every one of these is an assertion that can be right or wrong. Every one holds itself out as correct. Not one of them is the grammar of weather. Weather has no true and false in it; a storm is neither accurate nor mistaken, it simply occurs, and no sentence in the language of meteorology asserts that the storm was the correct storm. But every word on that list asserts exactly that: that the classification fits, that the determination is sound, that the finding is what the case in truth warrants. This is not the grammar of events. This is the grammar of judgment, and grammar is not decoration. The grammar carries a commitment, and the commitment is the whole of what I need at this mark: each of these utterances claims to be correct, and a thing that claims to be correct is, by that very fact, a thing that can be examined for whether it is. That is what a claim is. A claim is an assertion that submits itself, in the act of being made, to the question of its own validity, and thereby makes that question askable (Habermas [1992] 1996). To say noncompliant is to raise the question whether the person was in fact noncompliant, and to raise that question is to open the very door critique walks through. The institution does not merely permit the examination of its outputs. It invites the examination by the grammar in which it issues them, because to assert is to stake correctness, and to stake correctness is to make correctness reviewable.

Weber saw the same structure from the other side, from the side not of grammar but of power, and the convergence is worth marking because it closes off an escape route in advance. Modern institutional authority, he observed, does not rest on force alone and does not even primarily rest on force. It rests on a standing claim to legality: the belief, maintained by the apparatus and demanded of those subject to it, that its commands issue from enacted rules and that those rules have been correctly applied to the case at hand (Weber [1922] 1978). This is authority that lives or dies by a claim to rightness. The bureaucratic order does not say obey because I am strong. It says obey because this determination was reached lawfully, under the rules, by the proper office, and is therefore correct. Strip the claim to correct application away and the authority collapses into naked command, which is precisely the thing modern institutions are at pains to insist they are not. So the claim to correctness is not incidental to institutional power. It is the load-bearing wall of institutional power. The claim is the institution’s own currency, the thing it spends to purchase compliance, and it spends validity every time it speaks. Every letter, every finding, every classification is a withdrawal from an account the institution insists is full: the account of its own correctness. And an apparatus that stakes its authority on being correct has, in the same motion, staked its authority on being the kind of thing that could be found incorrect. There is no other kind of correctness. To claim it is to submit to the possibility of its refutation.

Now let me enter the exhibit that settles this mark, and I want you to appreciate that the defendant printed it, at its own expense, and hands it across the counter to anyone who asks: the appeal form.

You cannot appeal an avalanche. Sit with the sentence, because it is not a figure of speech, it is a test, and the test is decisive. You cannot file a grievance against the weather. You cannot request reconsideration from a hurricane, submit a rebuttal to an earthquake, or ask a landslide to review its determination in light of new evidence. The idea is not merely futile; it is incoherent, a category error of exactly the kind the motion to dismiss accuses me of committing. And the reason it is incoherent tells us everything. Appeal machinery is intelligible only where an output is understood, by the very party that produced it, to be the kind of thing that can be wrong. There is no appeal from an event, because an event cannot be mistaken. There is appeal only from a judgment, because only a judgment can be in error, and appeal is the procedure by which the question of that error is put. So the existence of an appeals office is not a neutral administrative convenience. It is a confession. Every appeals office, every reconsideration procedure, every grievance channel, every notice of the right to contest printed at the bottom of the letter, is a signed admission by the institution that its outputs are judgments and not events, that they make claims and not merely occur, that they can be wrong and stand ready to be tested for it. The institution has answered the first mark itself, in the affirmative, and printed the answer on a form.

In litigation there is a name for a statement a party makes that cuts against its own position, and the name is worth having here because it captures the peculiar force of this exhibit. It is called an admission against interest, and it carries special weight precisely because parties do not, as a rule, volunteer statements that damage their own case; when they do, the statement is presumed reliable for that very reason. The appeal form is the institution’s admission against interest. It would be squarely in the institution’s interest, faced with a critique of its reasoning, to deny that it reasons at all, to retreat to the position that it merely processes, merely records, merely produces outputs that are neither correct nor incorrect but simply issued. The appeal form forecloses that retreat, because the institution printed it before the critique arrived, distributed it at volume, built offices to receive it, and staffed those offices with reviewers whose entire function presupposes that the original determination was a judgment capable of being wrong. Nor will it help the defense to answer that the confession was extracted, that due process law compelled the form and the institution printed it under order rather than conviction. Extracted by what, and on what theory? Process attaches only to determinations. No legislature has ever mandated an appeal from the weather, because there is nothing in weather to appeal, and the law that compels the form is the same recognition written one level up: that these outputs are judgments, and that judgments answer. Compelled or volunteered, the admission reads the same. The defendant has testified against itself, in writing, in advance, at every counter it operates. I move the appeal form into evidence, and I note that the defendant is the one who drafted it.

The defense has one move left at this mark, and I will make it for them, at full strength, because a mark proven against the strongest version of the objection is the only kind worth having. The move is to concede the grammar and deny the substance. Yes, the defense says, institutions speak in the grammar of judgment; yes, they issue findings phrased as claims to correctness; yes, they print appeal forms. But all of this is costume. Underneath the grammar of judgment there is only power, and power is not the sort of thing that is right or wrong. Power is only strong or weak. The denial letter wears the robes of a verdict, but strip the robes away and there is no verdict underneath, only a stronger party imposing its will on a weaker one and dressing the imposition in the language of correctness to make it easier to swallow. The claim to correctness, on this account, is ideology: a useful fiction that launders force into legitimacy. And a fiction cannot be put on trial for getting the facts wrong, because it was never in the business of getting the facts right. It was in the business of ruling.

I want to grant this move its due before I show what it costs, because it is not cynicism for its own sake; it is a serious position with serious defenders, and there is truth in it. Institutional judgment is entangled with power at every point, and the grammar of correctness does often launder force. But watch precisely what the move must claim in order to defeat my case, and watch what that claim costs the defendant, because the cost is fatal. Take the dilemma and hold both horns. Either the institution’s outputs are genuine validity claims, assertions that hold themselves out as correct and can therefore be examined for whether they are, in which case the first mark is proven and my tribunal has its object. Or they are not validity claims at all, but mere exercises of power in the costume of judgment, events dressed as verdicts, force with a vocabulary. Follow the second horn all the way down, because the defense never does. If the denial letter makes no claim to correctness, if it is only the stronger party imposing its will, then it has no authority whatsoever, only strength. Its determination binds the way a fist binds, not the way a verdict binds. And in that case obedience to a denial letter is not the honoring of a legitimate judgment but mere submission to force, on precisely the level of a subject who obeys because he fears the blow. Compliance with an unfounded finding becomes superstition, the reading of institutional entrails, a ritual deference to marks on paper that assert nothing and mean nothing beyond the power of the office that produced them.

No institution that claims legitimate authority can accept the second horn, and this is the pivot of the entire mark. Not one hospital, not one agency, not one court, not one school will say of its own determinations that they carry no claim to correctness and bind by force alone. Every institution insists, and must insist, that its determinations bind because they are correct, or at the very least because they were correctly produced under valid rules by the proper authority. That insistence is the whole basis of the distinction the institution draws, and stakes its legitimacy on, between itself and a mere extortion racket. The extortionist also issues demands backed by power. What the institution claims that the extortionist does not is that its demands are right: lawful, warranted, correctly determined, owed. Remove the claim to correctness and you have not described institutional authority more honestly; you have abolished it, and reduced the institution to the racket it insists it is not. So the defense’s final move is one no institution can afford to make, because to make it is to saw off the branch its own authority sits on. The institution cannot say I make no claims and retain the authority it exercises, because that authority is constituted by the claim. It is trapped by its own legitimacy. To keep its power it must speak the grammar of judgment, and to speak the grammar of judgment is to make claims that can be examined, which is to walk, in full robes, into the courtroom it would rather avoid. The costume it wears to command is the summons that compels its appearance. Power that must dress as judgment to rule has already conceded that judgment is the standard, and a standard is a thing one can be measured against and found wanting.

The first mark is proven, and proven from the defendant’s own grammar and the defendant’s own forms. The defendant makes claims.

The Second Mark: The Judgment That Is Nobody's

Now the defense changes tactics, and this motion is subtler than the last, which is why it is more dangerous. The first mark’s objection was a frontal denial: institutions make no claims. That failed, because the grammar and the appeal form convicted the defendant out of its own filing cabinet. So the defense retreats to a stronger position, one that concedes the ground just lost and fights on new terrain. Very well, the defense says: grant that claims are being made, grant that the outputs are judgments capable of being right or wrong. It does not follow that the institution makes them. The institution is shorthand, a convenient collective noun for the persons who actually do the reasoning. When we say the hospital decided, we mean a physician decided, and a nurse recorded, and a case manager processed, and an administrator signed. The claims are real, but they are the claims of individual human beings, each of whom can be named, examined, praised, or blamed as an individual. So the proper motion is not to dismiss the case but to sever it: to break the single proceeding against the institution into its true components, a set of separate proceedings against the individual persons who made the individual judgments, and to conduct each one on its own. Examine the clinicians and the clerks one by one, the motion goes, and the shorthand dissolves. There was never an institution in the dock. There was only a crowd of persons, and a crowd is tried one member at a time.

This is the motion to sever, and I oppose it, and I oppose it on a formal ground first, before I say a word about what it costs, because the formal ground is decisive on its own and does not depend on any appeal to consequences. The motion to sever rests on a hidden premise, and the premise is false. The premise is that the judgment of a collective is nothing but the aggregate of the judgments of its members, so that if you examine each member you have, in the examining, exhausted the collective. Take that premise away and the motion collapses, because if the collective’s judgment is not the sum of its members’ judgments, then examining the members one by one leaves the collective’s judgment wholly unexamined, and severance is not analysis but evasion. So everything turns on the premise, and the premise is demonstrably false, and I can demonstrate its falsity with nothing more than three judges and two questions.

Collegial bodies that aggregate their members’ judgments can arrive at conclusions that no majority of their members individually holds. This is not a paradox I have invented for the occasion; it is a known result, first noticed in the practice of collegial courts, where it is called the doctrinal paradox, and later generalized in social choice theory under the name of the discursive dilemma (Kornhauser and Sager 1993; List and Pettit 2011). Here is the mechanism, and I ask you to follow it in full, because it is the formal heart of the second mark and it settles the question by itself. Take a panel of three judges deciding a case of liability, and suppose, as the law of contract in fact requires, that liability depends on two distinct findings, both of which must be affirmative for liability to follow: first, that a valid contract existed, and second, that the contract was breached. Now poll the panel judge by judge on the two issues.

Judge one finds that a valid contract existed and finds that it was breached. Both findings affirmative, so Judge one concludes: liable. Judge two finds that a valid contract existed but finds that it was not breached. One affirmative, one negative, so Judge two concludes: not liable. Judge three finds that no valid contract existed but finds that if one had existed the conduct would have breached it. One negative, one affirmative, so Judge three concludes: not liable. Now count. On the bottom-line verdict, the panel divides two to one for acquittal: Judge one alone votes liable, Judges two and three vote not liable. Poll the verdicts and the panel acquits. But now poll the issues instead of the verdicts. On the first issue, whether a valid contract existed, Judges one and two both say yes; only Judge three says no. Two of three find a valid contract. On the second issue, whether the contract was breached, Judges one and three both say yes; only Judge two says no. Two of three find a breach. So on every issue the panel’s majority is affirmative: a valid contract existed, and it was breached, and liability follows necessarily from those two premises. Poll the issues and the panel convicts.

Sit with what has just happened, because it is the whole point. The very same three judges, with the very same individual opinions unchanged, produce acquittal when you aggregate their verdicts and conviction when you aggregate their reasons. The panel’s judgment, reached issue by issue in the way courts often must reason and often do, is a judgment that a majority of the panel rejects at the level of the bottom line: two of the three judges would acquit, yet the panel, reasoning through its premises, convicts. Whose judgment is the conviction, then? It is not Judge one’s, though Judge one alone voted to convict, because the conviction rests on the panel finding a contract, which required Judge two, and on the panel finding a breach, which required Judge three, and Judges two and three both voted to acquit. It is not Judge two’s judgment; he found no breach. It is not Judge three’s; she found no contract. It is not the judgment of any one of them, and it is not the sum of them, because the sum, polled as verdicts, goes the other way. It is the judgment of the structure: of the rule that says aggregate the issues rather than the verdicts, applied to these three minds. There is an it, a judgment that belongs to the body and to no member of the body, and I have just shown it to you with three judges and two questions. The premise the motion to sever depends on, that the collective’s judgment is nothing but the aggregate of its members’ judgments, is false, because here is a collective judgment that is not any member’s judgment and is not their aggregate either. It emerges from the procedure by which their several judgments are combined, and it lives at the level of the procedure, where no individual examination can reach it.

This example does not prove that every collective is automatically a unified reasoner. It proves the narrower and stronger point I need: collective judgment can exist at the level of procedure rather than at the level of any individual mind. The institutional case then turns on whether the organization has standing decision structures that gather inputs, combine them under rules, authorize outputs, preserve those outputs, and act through them. Where such structures exist, the judgment belongs neither to a ghostly group mind nor to any single member, but to the organized procedure by which the institution makes claims.

And the formal result only names, in the clean setting of a three-judge panel, what every operator inside a real institution already knows in the mess of practice. The file is assembled by many hands across many shifts, and the case it constitutes was never held whole in any single mind at any single moment. One person intakes, another examines, a third records, a fourth reviews, a fifth authorizes, and the determination that emerges was composed by all of them and grasped in its entirety by none. The determination then persists while every person who touched it rotates out: the nurse changes, the physician moves on, the caseworker transfers, the administrator retires, and the determination stays, binding successors who never judged the matter and were not present when it was judged. And here is the most telling feature of all, the one that breaks the severance motion beyond repair. The determination can stand while every single individual in the room doubts it. The nurse doubts the discharge. The physician doubts it. The case manager doubts it. And the discharge proceeds anyway, because the system generates it, because the criteria are met and the boxes are checked and the pathway runs, and the output of the system is the institution’s judgment even though not one person in the building would have reached it on their own. Sever this case into its individuals and you will find, at the end of your examination, a room full of people who each doubted the outcome and an outcome that happened regardless. You will have examined everyone and explained nothing, because the judgment was never in any of them. It was in the arrangement that combined them.

This is no mystery ontology, and I want to be careful here, because the defense’s last refuge is to accuse me of mysticism, of conjuring a group mind that hovers over the institution like a ghost. I am doing nothing of the kind. There is no institutional soul, no collective consciousness, no spooky entity above and beyond the persons and the procedures. There is something far more ordinary and far more real: a decision structure. Organizations possess internal decision structures, standing procedures that specify how inputs are gathered, how they are combined, who may sign, what counts as authorized, and by what steps a set of individual contributions becomes the organization’s own act rather than anyone’s private opinion. It is these structures, and not any employee’s state of mind, that make it literal rather than loose to say the corporation decided (French 1984). When we say the hospital discharged the patient, we are not speaking figuratively and we are not summing five people’s opinions; we are naming the output of a decision structure that took five people’s contributions as inputs and produced, by its own rules, an act attributable to the institution as such. The structure is the reasoner. The persons are its organs, and the fifth mark will show that they are never only that. And reasoner here is a functional title, not a psychological one: it names the site at which validity claims are produced, maintained, and revised under rules, which is all critique has ever needed its object to be.

And the outputs of that structure are not shadows of real facts located elsewhere; they are facts in their own right. Institutional facts are real facts, brought into being by the collective acceptance of constitutive rules, rules of the canonical form X counts as Y in context C (Searle 1995). This piece of paper, in this context, counts as a denial, and its counting as a denial is not a metaphor for something more basic happening in someone’s head; it is the institutional fact itself, as real as any fact about the person, and consequential in ways no private opinion could be. This number counts as a risk score. This entry counts as a finding. This signature counts as authorization. The counting is not a report on cognition happening somewhere else. The counting is the cognition. The institution thinks by counting one thing as another under its constitutive rules, and the results of that counting are the institution’s judgments, made by the structure, attributable to the structure, and reachable only by a tribunal that takes the structure as its object.

And now let me tell you what the severance motion actually purchases, because the defense will never say it out loud. Grant severance, insist that only individuals may be examined and that the institution as such is never in the dock, and watch what becomes of every institutional harm. It decomposes. It breaks into a sequence of individual acts, and here is the trap: each individual act, taken in isolation, is very often reasonable, defensible, even correct. The nurse charted accurately what she observed. The physician followed the protocol as written. The case manager applied the criteria that governed her. The reviewer confirmed that the boxes were checked. The administrator signed what the file supported. Examine each one and each has a perfect account of themselves, because each did their part correctly, and the harm was in no one’s part. The harm was in the composition, in the way the correctly-performed parts combined into an outcome that wounded. So severance does not locate the wrong more precisely. It dissolves the wrong entirely, distributing it so finely across so many individually blameless acts that it comes to rest nowhere. This is not analytical rigor. It is acquittal by dispersion: a thousand defendants, each with a genuine and complete alibi, and a harm standing in the middle of them with no author of record. The severance motion is a machine for producing harms that no one committed.

Which is exactly why the tribunal must refuse it. The judgment that is nobody’s is not the judgment that least needs a court. It is the judgment that most needs one, because it is the only judgment no other court can reach. Individual accountability, wherever it applies, is already provided for; there are boards and licenses and suits and reviews to reach the individual who errs. What has no forum, what falls through every existing procedure precisely because it belongs to no individual, is the wrong that lives in the structure, the harm composed of blameless parts, the determination that every person doubted and the system produced anyway. If that judgment cannot be tried as the institution’s own, it cannot be tried at all, and a whole domain of the most consequential judgments made about persons in modern life passes entirely beyond the reach of accountability. The motion to sever is the mechanism of that escape, and I will not grant it.

Severance denied. The second mark is proven, and proven formally, from the structure of collective judgment itself and not from any appeal to a group mind. The defendant is one.

The Third Mark: The Printed A Priori

Here is the heart of my case, and the first set of prior conditions comes into evidence. The first two marks established that there is a defendant and that it speaks: institutional reason makes claims, and those claims belong to the structure and not merely to the persons who staff it. But making claims and being unified are not yet enough to warrant a tribunal in the strict sense. A capricious tyrant makes unified claims. What critique requires, the feature that lifts it above mere complaint and makes it possible to map an object rather than merely react to its outputs, is that the claims issue from prior conditions: a standing structure that precedes every particular judgment and shapes what any judgment can be. This is the mark that turns the proceeding from prosecution into critique. And it is the mark at which the correspondence between institutional reason and the reason Kant examined becomes more than a loose analogy: not identical in metaphysical status, because institutional conditions are historical and authored, but functionally parallel in the role they play within an order of cognition.

Kant’s revolution was an inversion, and I have to state it precisely because everything in this section is built on running it a second time. The picture Kant overturned was the natural one: that in knowing, the mind conforms itself to objects, receiving their imprint the way wax receives a seal, so that cognition is a copying and the object is prior. Kant reversed the direction of dependence. Cognition does not passively copy objects; objects, if they are to be cognized at all, must conform to the prior forms of the cognizer (Bxvi–xviii). The mind is not wax taking an imprint; it is a set of forms into which whatever is to be experienced must first be poured, and nothing can be experienced that will not fit the forms. Space, time, and the categories are not features the mind finds in the world; they are conditions the mind imposes on anything that is to count as world for it. So whatever appears, appears already shaped, already conditioned, already formatted by structures that were in place before the encounter and that decide, in advance, the terms on which anything can appear at all. The object as known is, in part, constituted by the knower’s prior forms. That is the Copernican turn, and it is the most consequential single move in modern philosophy.

Now perform the inversion a second time, institutionally, and watch it hold. The natural picture of institutional reception is the copying picture: that the institution simply registers the person as she is, takes down her situation, records her facts, and that the record conforms to the person the way wax conforms to a seal. Reverse it, exactly as Kant reversed it. The institution does not passively receive the person. The person, if she is to appear institutionally at all, if she is to register as anything the institution can recognize, act on, or record, must first conform to the prior forms of the order. She must fall within a jurisdiction, arrive within a schedule, fit a category, satisfy an evidentiary standard, enter a file. What cannot be poured into those forms does not appear; it is institutionally invisible, a reality the apparatus cannot see because it has no form in which to receive it. The person as received is therefore not the person as she is; she is the person as constituted, in part, by conditions settled before she arrived and imposed on her appearance as its price of admission. Map those conditions and you have mapped the institutional a priori: the standing forms that decide, before any particular person walks in, what a person can be for this institution.

I need to fix the term before it can mislead. By institutional a priori, I do not mean an a priori in Kant’s strict, universal, authorless sense. I mean a historically situated and framework-relative a priori: prior to any particular institutional judgment, constitutive of what can count as an institutional object within that order, and revisable because authored. These forms are not necessary for experience as such. They are necessary for institutional cognition within a given order. Their necessity is therefore not metaphysical in Kant’s original sense but operative, constitutive, and historical.

They map with unsettling functional precision, role for role, and I will walk the whole correspondence, because the precision is itself the argument. It is one thing to say loosely that institutions have their categories the way minds have theirs. It is another to show that every distinct element of Kant’s cognitive architecture has an institutional analogue occupying a comparable functional position within an authored order of cognition, so that the parallel is not a metaphor drawn across two fields but one structure appearing in two materials. That is what I mean to show.

Kant’s forms of intuition were space and time: the where and when of any possible appearance, the two frames within which anything must be given before it can be thought at all. Nothing can be experienced that is not somewhere and somewhen; space and time are the forms of receptivity itself. The institution’s forms of intuition are jurisdiction and schedule, and they occupy the identical position: the where and when within which any person must be given before the institution can process her at all. Jurisdiction is institutional space. It is the whole apparatus of catchment area, venue, standing, service area, network coverage, eligibility by residence: the determination of where a person may appear and, appearing there, as whom. A person outside the jurisdiction does not appear weakly; she does not appear at all, exactly as an object outside space is not a faint object but no possible object. Schedule is institutional time. It is filing deadlines, decision windows, appointment lengths, review cycles, statutes of limitation, retention periods: the determination of when a reality is permitted to count, for how long it will be entertained, and how long it will be held against the person after it has ceased to be true. A claim filed after the deadline is not a weak claim; it is no claim, unreceivable, invisible in institutional time the way an event outside time is no event. Space and time, jurisdiction and schedule: the forms of institutional intuition, the where and when of institutional appearance.

Kant’s categories were the pure concepts under which any object must fall to be thought at all: unity, plurality, cause, substance, and the rest, the concepts without which the given remains an unthinkable blur. To be an object of thought at all is to be brought under the categories. The institution’s categories are its classification scheme, and they do precisely this work: they are the concepts under which any person must fall to be institutionally thinkable. Diagnosis codes, eligibility criteria, offense classes, benefit tiers, grade bands, risk levels: these are the pure concepts of the institutional understanding, the finite set of types under which the person must be subsumed before the apparatus can think her. As what does this reality register? As this code, this class, this tier, this band. A reality that fits no category is not thought poorly; it is not thought at all, because the apparatus has no concept under which to bring it, exactly as, for Kant, an intuition to which no category applies is not a dim thought but no thought whatever. The classification scheme is the table of institutional categories, and to be received is to be subsumed under one.

Kant’s schematism was the hidden mediation between the pure category and the concrete case: the procedure by which an abstract concept like cause is connected to the sensible particulars that fall under it, the rule that tells the understanding how to recognize an instance of the concept in the manifold of intuition. Kant confessed it was obscure, calling it an art concealed in the depths of the human soul whose real workings nature is unlikely ever to let us read (A141/B180–81). The institution’s schematism is its evidentiary rule, and here the parallel yields a striking dividend, which I will collect in a moment. The evidentiary rule is precisely the procedure that connects the pure institutional category to the concrete person: it specifies what counts as showing that the category applies. Which documents prove residence and thereby establish jurisdiction. Which findings prove disability and thereby trigger the eligibility category. Which conduct proves risk and thereby warrants the classification. The category disability is abstract; the evidentiary rule is what tells the apparatus how to recognize an instance of it in the particular person standing at the counter, and the answer takes the form of a documentation requirement: these papers, this proof, this showing, and no other. The schematism of the institution is the rule of evidence, and note the dividend, because it will matter: in Kant the schematism is a hidden art, buried, unreadable, a procedure of the soul that resists all inspection. In the institution the schematism is printed. The documentation requirement is published. The hidden art of the human mind is, in the institution, a form you can hold in your hand, which means the most obscure link in Kant’s chain is, in the institutional case, the most visible.

And Kant’s supreme condition, the keystone on which the whole architecture rests, was the transcendental unity of apperception: the principle that the I think must be able to accompany all my representations, because a representation that could not be taken up into one self-consciousness would be nothing to me, not a cognition at all, but a mental content belonging to no mind (B131–32). The unity of the knower is the condition of there being knowledge rather than a scatter of unownable impressions. Every representation must be gatherable into one consciousness, or it is not a representation for anyone. The institution does not possess apperception in Kant’s strict sense. It has no self-conscious “I” to which representations are mine. The analogy therefore cannot be identity. What the institution possesses is a functional analogue of apperceptive unity: the file is the authorized unity into which dispersed representations must be gatherable if they are to count as one case for the institution. An entry that cannot be taken up into the file is not thereby unreal, but it fails to become an institutional representation available for judgment. In this limited and strictly institutional sense, the file performs the work that apperceptive unity performs for Kantian cognition: it gathers a manifold into the unity under which judgment can proceed. The file is not an institutional “I” in the full Kantian sense; it is the institutional analogue of the “I think,” the unity without which its representations cannot become one case.

But here the correspondence, having held at the level of function rather than identity, produces one dark inversion, and I ask the court to hold it, because it is the single most important structural fact in this entire proof and I will return to it at the sixth mark and again in the moral conditions. In Kant, the unity of apperception serves the knower. The I think is mine; it gathers my representations into my consciousness for my cognition; the unity exists for the sake of the self whose unity it is. In the institution, the unity is inverted in its ownership. The file gathers the person’s representations, but not for the person and not into the person’s self-consciousness. It gathers them into the institution’s, and it does something the Kantian unity never does: it outlives the person’s presence, persists when she has gone, travels where she cannot follow, and then returns to confront her as a second self, a synthesized institutional person assembled from entries she may never have seen, asserting things about her she may not recognize, and demanding that she answer for it. In Kant the unity serves the one it unifies. In the institution the unity is extracted from the one it represents and turned to face her as an object she must account for. The file is the institution’s functional analogue of the I think, and it produces a second institutional self the person did not author, cannot fully inspect, and cannot escape. Hold that. It is where cognition and morality will meet.

Add one final element and the faculty architecture is complete: the structure of authorization, the determination of who within the institution may judge, who may sign, who may override, who may escalate, whose determination is provisional and whose is final. This has its Kantian counterpart too, in the spontaneity of the understanding, the active power that does the actual judging as against the mere receptivity that takes in the given; but I will not press that parallel as hard as the others, because the institutional point stands on its own. Authorization is the institution’s structure of judgmental power: not everything entered into the file becomes a judgment, and what converts an entry into a determination is a signature from an office authorized to make it. The architecture is now whole. Jurisdiction and schedule, the forms of intuition. The classification scheme, the categories. The evidentiary rule, the schematism. The file, the functional analogue of apperceptive unity. Authorization, the power of judgment. Six elements, each occupying the functional position of its Kantian counterpart, together constituting the institutional a priori.

These conditions are prior in the strict operative sense, and I want to be exact about what that means, because the whole force of calling them a priori depends on it. They are not prior merely in the trivial sense of coming earlier in time. They are prior in the constitutive sense: they are settled before any particular person arrives, and they decide in advance the terms on which she can appear. Before this person walks in, it has already been determined where she may appear, which is jurisdiction; when her reality will be permitted to count and for how long, which is schedule; as what she can register, which is the categories; by what showing she can establish that a category applies, which is the evidentiary rule; into what unity her facts will be gathered, which is the file; and under whose signature a determination about her becomes binding, which is authorization. Every one of these was fixed before she existed as a case, and every one of them shapes what she can be as a case. She does not encounter the institution as a free reality that the institution then neutrally records. She encounters a set of forms already in place, and she becomes institutionally real only in and through them, only to the extent that she can be poured into them. Kant put the two-sidedness of this with a formula I can translate exactly: thoughts without content are empty, intuitions without concepts are blind, and the two stems of knowledge are barren in isolation, yielding cognition only in their union (A50–51/B74–75). The institutional translation holds at the level that matters for this argument: not identity of metaphysical status, but correspondence of cognitive function. Forms without persons are empty: a classification scheme with no one to classify is an idle table, a jurisdiction with no one within it an empty map. Persons without forms are blind, or rather invisible: a person who fits no category, falls in no jurisdiction, meets no evidentiary standard, enters no file, cannot be seen by the institution at all, however vividly real she is in herself. Only in the union of person and form does an institutional cognition occur, exactly as, for Kant, only in the union of intuition and concept does knowledge occur. The person supplies the content. The forms supply the concepts. Neither alone yields an institutional fact.

And here I want to note a methodological gift the institutional case hands the critic, because it inverts the single hardest labor in Kant’s book and it explains why this proof can be so much more direct than his. Kant had to excavate his a priori. The forms of human cognition are not lying on the surface where anyone can read them; they are buried in the operation of cognition itself, presupposed by every experience but never given as an object of any experience, and the whole eight hundred pages of the first Critique are a sustained work of inference toward structures that no one can inspect directly, that show themselves only in what they make possible, and that must be reconstructed by argument because they can never be observed. Kant is reasoning backward from the fact of experience to the conditions that must hold for experience to be possible, and the conditions remain forever inferred, never seen. Institutions spare the critic this entire labor, because institutions arrive pre-excavated. Their a priori is not buried; it is externalized by construction, printed and published and distributed as a condition of the institution’s own operation. The forms of institutional cognition are on intake forms. They are codified in policy manuals, tabulated in code sets, promulgated as eligibility regulations, published as documentation requirements. The intake form is quite literally a table of categories with a printing date at the bottom. What Kant had to infer, the institution has already printed. In litigation there is a word for the compelled production of the other side’s internal documents, and the word is discovery, and I am pleased to report that in this case discovery is complete before the trial begins, because the defendant published its own a priori as a matter of course. Institutional reason is reason with its conditions showing. Kant had to reconstruct his object from its effects. I can read mine off the paperwork, because the paperwork is the a priori, laid out, dated, and signed.

One objection at this mark deserves a full answer, because it is the best philosophical shot the defense has left and the section is not secure until it is met. Kant’s conditions, the objection runs, were necessary and universal: one set of forms, the same for every possible cognizer, holding always and everywhere, not up for revision by anyone. That is what made them transcendental. Institutional conditions are nothing like this. They are contingent and local: this state’s eligibility rules and not that state’s, this hospital’s intake fields and not another’s, this year’s code set and not last year’s, all of it revisable by the next legislature, the next administrator, the next software vendor. Where everything is contingent and local and revisable, the objection concludes, there is no transcendental structure at all, only sociology: the empirical study of how this particular institution happens to be arranged, with none of the necessity that would make talk of an a priori anything more than a flattering metaphor. Call it a category scheme if you like, but do not dignify a revisable administrative form with the name Kant reserved for the necessary conditions of experience as such.

The objection is serious, and the answer is not to deny the contingency but to show that contingency at the level of content is fully compatible with necessity at the level of form, and that this exact combination is already established doctrine under a precise name. The name is the relativized a priori, and philosophy of science arrived at it by meeting this very structure in the history of physics. There are principles within a scientific framework that function as genuinely a priori relative to that framework: they are not empirical findings within the theory but the constitutive conditions that must be in place for the theory’s empirical questions to have meaning at all, the framework-defining principles that make measurement and test possible rather than being themselves measured or tested. And yet these constitutive principles change when one framework gives way to another, so that what is a priori is relativized to a framework rather than fixed for all time: constitutive within its framework, revisable across frameworks (Friedman 2001). Contingency across frameworks and constitutive necessity within them are not in tension; they are the two halves of a single well-understood structure. Foucault named the same stratum from the side of discourse rather than physics, and gave it almost the name I need: the historical a priori, the set of positive conditions that determine, in a given period, what can be said, what can count as a serious statement, what can be known, conditions that are fully real and fully constitutive for their age and yet themselves transformable, belonging to history rather than standing outside it (Foucault [1969] 1972). I take the category from Foucault and depart from him on one point, deliberately: where the archaeology brackets the author and treats the conditions as anonymous strata, the sixth mark will reinstate the signature at the level of the operative form, where authorship is not a theoretical posit but a documented act with a drafting session and a budget line. Between them, the philosophy of science and the archaeology of knowledge have already established the exact category the institutional a priori occupies: necessary in form, contingent in content; constitutive within a framework, revisable across frameworks; transcendental in function, historical in existence.

Apply that category and the objection dissolves cleanly. It is contingent that this state sets the residency requirement at twelve months rather than six; it is not contingent that there is some jurisdictional form determining where a person may appear, because an institution with no jurisdiction at all is not a more open institution but not an institution, incapable of receiving anyone because incapable of determining whom it receives. It is contingent which categories this scheme contains; it is necessary that there be some scheme of categories, because an apparatus that subsumed persons under no concepts could not think them at all. It is contingent what this evidentiary rule demands; it is necessary that there be some evidentiary rule, some determination of what counts as showing that a category applies, or the categories could never be applied to anyone. The occupants of the positions are legislated, revisable, local, contingent, the proper study of sociology. The positions themselves are necessary for institutional cognition as such, and it is the positions that constitute the institutional a priori. The form is transcendental in function. The content is enacted.

Mark that last word, because it is the hinge on which the second half of this entire proof will turn, and I am placing it here deliberately so that you will remember where it entered. The content is enacted. What is enacted has enactors. Space and time and the categories were enacted by no one; they have no author, and that authorlessness is why Kant’s critique could only ever be cartography, a mapping of conditions that no one made and no one can be called to answer for. The institutional a priori is not authorless. Every position in the architecture I have just laid out was filled by someone: someone drew the jurisdiction, someone set the schedule, someone drafted the categories, someone specified the evidence, someone designed the file, someone chartered the authorization. Enacted conditions have enactors, and enactors can be summoned. I will collect on that at the sixth mark, where it converts everything proven so far from a permission into an obligation. For now I only plant it: the word enacted is the seam where cognition opens onto morality.

One further objection must be met before this mark closes, and it is the sharpest a Kantian can raise, so I will raise it myself rather than wait. In Kant, the forms are the sole mode of givenness. Nothing reaches the mind except through them, which is why the thing in itself remains forever out of reach and why no appearance can ever be laid beside the reality and checked against it. If the institutional parallel held at that seam as well, the person behind the file would be the institution's noumenon, unreachable in principle, and every contest would be finished before it began, because the appeal arrives on a form, enters the file, and becomes one more representation among the representations it disputes. The parallel does not hold at that seam, and the break is not a flaw in the proof; it is a finding of the proof. Institutional forms are not the person's sole mode of givenness. The person can stand at the counter beside her own record. She can speak over the file in the room where the file is read. The encounter exists, however brutally the schedule compresses it, and the gradients I will establish at the fourth mark are gradients precisely because a person is present for the record to outcompete; nothing outcompetes a noumenon. The institutional a priori, unlike Kant's, is porous: it formats the person's appearance without exhausting her availability. Hold this alongside the dark inversion of the file, because the two will meet at the end of the proof. The porosity is what contest will draw on when I argue that an appeal is epistemology and not mercy, since a procedure that summons the reality into the room beside its representation is possible only where the reality can still enter the room. Kant's cognizer could never confront the appearance with the thing. The institution can be made to. The forms decide how she appears. They do not decide that she is, and they cannot make her unavailable; they can only make her easy to ignore, which is a different crime with a different remedy.

The third mark is proven. The defendant operates under prior conditions; the conditions form a structured institutional a priori parallel to Kant’s at the level of cognitive function; and unlike Kant’s a priori, this one is in the record, printed, dated, and produced in discovery by the defendant itself. The court has before it not merely a reasoner that makes claims, but a reasoner whose forms of reasoning are laid open for mapping. Critique in the strict sense is now not only possible but equipped.

The Fourth Mark: The Native Illusion

Critique in the strict sense is necessary, and not merely permitted, only where the object breeds illusion from its own structure. This is the mark that separates a standing tribunal from an occasional correction, and I have to be exact about the distinction, because the whole necessity of the enterprise rests on it. Everything makes mistakes. A cognizer that erred only occasionally, by inattention or bad luck or missing information, would need no tribunal and no critique; it would need ordinary correction, the local repair of the local error, after which the matter would be closed. You do not convene a permanent court to handle mistakes that ordinary vigilance can catch. What warrants a standing discipline, a permanent tribunal built into the very operation of the faculty, is a different and far more serious thing: an illusion the object produces structurally, unavoidably, from the same machinery that lets it function at all, so that the illusion cannot be corrected away because it is not a malfunction but a property of correct functioning. Only a native, structural illusion necessitates a standing discipline, because only a native illusion returns after every local correction, generated afresh by the apparatus each time it operates.

Kant is explicit that reason’s characteristic illusion is exactly of this kind, and the point is so central to his enterprise that he draws a sharp line to mark it. There is a difference, he insists, between the mistakes that logic corrects and the illusion that critique must discipline. Ordinary logical error is remediable by attention; once you see the fallacy, it is gone. Transcendental illusion is not like this at all. It is natural and unavoidable, seated in the very nature of reason, and it does not disappear even after it has been exposed and its nullity demonstrated. It arises from principles whose application, entirely legitimate within experience, reason is irresistibly driven to extend beyond experience, where it has no validity; and this drive is not a habit reason can be broken of but a disposition belonging to reason as such (A293–98/B349–55). Kant gives the image that has become the standard illustration, and it is precisely the right one because it captures the persistence: the illusion survives knowledge of the illusion. Even the astronomer, who knows with full certainty that the moon at the horizon is no larger than the moon overhead, and who can explain exactly why the appearance arises, cannot make the horizon moon look smaller; it goes on appearing larger, because the source of the appearance is structural and lies below the level at which knowledge operates (A297/B354). Knowing the truth does not dissolve the appearance, because the appearance is not produced by a belief that knowledge could correct; it is produced by the constitution of the faculty itself. And that is exactly why the remedy cannot be a one-time correction. Because the pull is structural, the discipline must be structural too: a standing tribunal, permanently in session, because the illusion is permanently regenerated. You cannot correct a structural illusion once and be done. You can only institute a standing discipline against it.

Institutional reason has its native illusion, and it meets Kant’s specification at the level relevant to this argument: it is generated by the apparatus itself, it is unavoidable given how institutional cognition works, and it survives its own detection. I will name it plainly and then prove that it is structural rather than incidental, because the naming is easy and the proof is what matters. The native illusion of institutional reason is substitution: the standing tendency of the representation to displace the person, so that the file becomes the object of response in the person’s place, and the partial recorded world comes to be treated as the whole of the reality it was only ever meant to represent. The institution builds a representation of the person in order to act on the person, and then, by a drift built into the very conditions of its cognition, it begins to act on the representation instead, responding to the record as though the record were the reality, until the proxy is no longer consulted as a stand-in for the person but treated as the person entire. And when substitution runs to completion across an order, the effect is not a single error but a collapse of the order’s contact with the reality it was constituted to serve: response comes to rest entirely on the proxy, and the person as she is drops out of the institution’s field altogether, present in the world but absent from the cognition that decides her.

The claim I have to prove is not that this happens, which anyone who has dealt with institutions knows, but that it is transcendental in function within institutional reason: that it is generated by the same apparatus that makes institutional cognition possible at all, so that it cannot be eliminated by better intentions or better staff, because it is a property of the conditions themselves and not of the people operating under them. I prove it by following the gradients along which substitution occurs, and the decisive fact is that every one of these gradients is not a defect in the apparatus but a direct consequence of the very forms established at the third mark. The file, which the third mark identified as the institution’s functional analogue of apperceptive unity, its indispensable condition for gathering a person into one case, turns out to be the mechanism by which the person is displaced. The condition of institutional cognition is the engine of institutional illusion. That is what makes the illusion structurally critical rather than merely empirical: it flows from the institutional conditions of possibility themselves. There are four gradients, and every one is built in.

The first gradient is availability. The institution can act only on what it holds, and what it holds is the representation, never the person. This is not a shortcoming; it is a structural necessity following directly from the file. The person herself is present only intermittently, for the length of an encounter, and then she leaves; the file is present always, permanently available, open on the desk and in the system whenever the institution turns its attention to her case. So at every moment when the institution acts and the person is not physically present, which is almost every moment, the only thing available to be acted upon is the representation. Availability weights all response toward the file for the simple reason that the file is what is there. The person is the intermittent object; the record is the continuous one; and cognition necessarily fastens on what is continuously available to it.

The second gradient is durability. Encounters are events, and events end; records are objects, and objects persist. The conversation is over the moment it concludes, retrievable thereafter only as whatever was written down of it, but the note endures indefinitely, carried forward through the schedule established at the third mark, surviving in the file long after the encounter that produced it has vanished. So time itself, institutional time, the schedule that is one of the forms of institutional intuition, weights response toward the representation, because the representation is what time preserves and the encounter is what time dissolves. What persists accumulates authority simply by persisting, and what persists is the record.

The third gradient is auditability. The operator does not act in a vacuum; the operator answers to review, and this accountability, which sounds like a safeguard, is in fact a third engine of substitution. Review reads documents. When the operator’s conduct is examined, whether in an audit, a supervisory review, a legal proceeding, or a quality check, what is examined is the record: what was charted, what was documented, what appears in the file. What was not charted did not happen, so far as review can determine, and what was charted is what the operator will be held to. So accountability itself flows to the record rather than to the person, and since attention follows accountability as surely as water runs downhill, the operator’s attention is drawn to the chart, because the chart is what the operator will answer for. The very mechanism meant to hold the operator responsible binds the operator’s attention to the proxy, because the proxy is the thing under review.

The fourth gradient is mobility, and it is the most totalizing of the four. The person cannot travel through the institution; only the representation can. The person exists at one place at a time, but her file moves: across desks, between departments, from one office to the next, from one institution to another entirely, carried along the pathways the apparatus is built to move records through. And every downstream site that receives the file meets the proxy without ever meeting the person, because the person is not what was sent. The receiving clinician, the reviewing officer, the transferring caseworker, the next institution in the chain: each of them encounters the person solely in the form of the representation that arrived, and for every one of them the person simply is the file, because the file is the only thing that came. So for most of the institution, most of the time, at every site the person herself never physically reaches, the proxy is not experienced as a stand-in for an absent reality. It is experienced as the entirety of the reality, because it is the only version of the person present at all. Mobility means that the representation is, for the vast majority of the institution’s operations, not a representation of the person but the whole of the person as far as that part of the institution can ever know.

Availability, durability, auditability, mobility. Set them side by side and see what they establish together. On every one of the four gradients, the file outcompetes the person it represents: the file is more available, more durable, more auditable, and more mobile than the person, on every axis along which institutional attention is allocated. Attention flows to what is available, and the file is more available. It flows to what persists, and the file persists longer. It flows to what one answers for, and one answers for the file. It flows to what is present, and downstream the file is the only thing present. The migration of institutional response from the person to the proxy is therefore not an aberration, not a failure of an otherwise sound system, not the fault of careless or callous operators. It is the equilibrium state of institutional cognition: the resting condition that the system settles into whenever nothing actively holds attention on the person, because every structural gradient runs from the person toward the file. Substitution is not what goes wrong with institutional attention. Substitution is where institutional attention comes to rest when nothing holds it elsewhere, and holding it elsewhere requires continuous effort against a continuous structural pull, which means the pull reasserts itself the instant the effort relaxes. And the pull does not distribute its injuries evenly: persons already thickly filed, heavily policed, medically disbelieved, administratively burdened, racially marked, disabled, poor, unhoused, or otherwise overexposed to institutional judgment meet substitution earlier, more often, and with fewer exits.

And like Kant’s moon, the illusion survives its own detection, which is the final proof that it is structural and not a matter of belief. The seasoned clinician who knows perfectly well that the chart is only a partial record still feels the chart’s verdict arrive in the mind before the patient’s own account does, because the chart was read first and read when the patient was not there. The veteran caseworker who knows perfectly well that a closed file does not mean a resolved need still experiences the closed file as a closed matter, because closure is what the record registers and the record is what the institution runs on. Knowing better does not dissolve the pull, exactly as the astronomer’s knowledge does not shrink the moon, and for exactly the same reason: the pull is not a belief that knowledge could correct. It is a gradient built into the conditions of the work, operating below the level at which knowledge operates, regenerated every time the operator turns to the file because the file is structurally where the operator must turn. Kant marked, at the very passage already in evidence, the distinction this remedy will live in: the illusion cannot be prevented, but the deception can. The astronomer still sees the risen moon swollen and is not fooled by it, because a standing discipline sits between the seeing and the assenting (A297/B354). That is the exact shape of what follows. The pull toward the proxy will not be dissolved by anything, and the surrender to the pull can still be refused, provided the refusal is built rather than felt. This is the formal reason, visible now with full clarity, that the conditions of institutional reason had to be conditions, structures, load-bearing members of the apparatus, rather than exhortations addressed to the people who staff it. An exhortation is a correction aimed at belief, and this illusion is not seated in belief; it is seated in structure, and it defeats every exhortation by regenerating itself structurally the moment the exhortation’s effect fades. You do not counsel vigilance against gravity, because vigilance is intermittent and gravity is constant, and the constant force wins every time attention lapses. Against a structural pull you do not exhort. You build: you install standing structures that hold attention on the person by force of design, because only a structural remedy can hold against a structural illusion. The tribunal this proof convenes is one such structure, and the conditions it will enforce are others.

One more thing before I leave this mark, and I ask the court to hold it precisely, because it is the second of the two seams along which cognition opens onto morality, and I am placing it here so that it is already in the record when I collect on it. In law, a harm that flows predictably from the design of a thing is a foreseeable harm, and foreseeability is the seed of duty: the law imposes obligations with respect to harms that a reasonable designer could anticipate, precisely because they could be anticipated, and it does not excuse the designer who looks away from what the design predictably does. I have just proven, from the four gradients, that substitution is the predictable output of the institutional apparatus itself: not an accident that befalls it but the equilibrium its own structure produces, derivable in advance from the forms established at the third mark. That is foreseeability in the legally recognizable sense. The harm is not a surprise; it is a consequence readable off the design before the design is ever operated. Foreseeability is now in the record. At the fourth mark it establishes only this: that the characteristic harm of institutional reason is structurally predictable. At the sixth mark, joined to the fact that the structure has authors, it will mature into something heavier, because a foreseeable harm flowing from an authored design is the classic form of a duty owed and, where the design is maintained after notice, of a duty breached. For now I only lodge it. The harm is foreseeable, and foreseeability is where responsibility begins.

The fourth mark is proven. The defendant does not merely err; it manufactures its characteristic error by design, generating substitution from the same forms that make its cognition possible, regenerating the illusion after every correction, and producing it as the predictable equilibrium of its own structure. The illusion is native, it is structural, and it survives detection, which is precisely the profile that makes a standing tribunal necessary rather than optional.

The Fifth Mark: The Tribunal Within

Kant’s last requirement is the most demanding of the five, and it is the one on which the legitimacy of the entire proceeding finally rests, so I will state it with care. Critique must be self-examination: reason judging reason, the faculty turned upon itself and disciplined by its own standards. If it is anything less than this, if it is one authority standing outside another and pronouncing on it in the name of some external measure, then it is not critique at all but mere legislation, one power imposing its will on another and calling the imposition judgment. Kant is unambiguous about this, and about why it matters. Reason, he says, has no dictatorial authority; there is no tribunal above reason that could hand down verdicts to it from outside, because there is nothing above reason in whose name such verdicts could be issued. Reason’s verdict is therefore never a decree received from elsewhere. It is nothing other than the agreement of free citizens, each of whom must be permitted to express, without hindrance, his objections and even his veto (A738–39/B766–67). The authority of the critical tribunal is the authority of free and public examination conducted by the faculty upon itself, and no other authority is available or would be legitimate. So the defense raises its final objection, and it raises it with real force, because if it holds, everything proven so far is insufficient. Grant, the defense says, that institutional reason makes claims, possesses unity, operates under prior conditions, and generates a native illusion. It still does not follow that institutional reason can sit as its own tribunal. Perhaps it is simply an object, a structure that reasons but cannot examine its own reasoning, in which case your critique is not self-examination at all but exactly the external legislation Kant forbids: you, standing outside the institution, issuing decrees to it in the name of standards it does not hold and cannot apply to itself. If institutional reason cannot judge itself, then your tribunal, whatever else it is, is not critique.

The objection is serious, and I will not answer it with argument in the first instance, because it can be answered with evidence, and the evidence is decisive. Look at what institutions already contain. The answer to whether institutional reason can examine itself is not a matter for speculation; it is stipulated in the institutions’ own organizational charts, printed in their own bylaws, funded in their own budgets. Appeals. Grievance procedures. Internal audits. Ombuds offices. Ethics consultations. Peer review. Quality assurance committees. Morbidity and mortality conferences. Dissent channels. Notice-and-comment procedures. Reconsideration processes. Institutional review boards. Each of these is, at minimum, an organ by which institutional reason can turn upon its own outputs and ask whether they were correct. Each is a mechanism internal to the institution whose function, at least in principle, is to examine institutional determinations for error. They are proto-critical organs: partial, often weak, frequently underfunded, sometimes purely ornamental, but still evidence that institutional reason already contains immanent procedures for reviewing its own claims by standards it recognizes as its own. The appeals office asks whether the determination was right by the institution’s own criteria. The audit asks whether the operation conformed to the institution’s own rules. The ethics consult asks whether the action met the institution’s own professed commitments. In every case the standard applied is the institution’s own, and the examiner is the institution itself, turned reflexively upon its own acts.

This does not require pretending that these organs usually perform critique fully or faithfully. Many are defensive, narrow, procedural, captured by liability concerns, or designed to preserve the institution from embarrassment rather than expose it to truth. But even in their weakest form they concede the point that matters here: institutional reason already knows that its own determinations require procedures of review. The presence of a compromised tribunal is not evidence that no tribunal exists. It is evidence that the tribunal exists in truncated form and must be completed.

This is the fact that converts the whole proceeding from moralizing into immanent critique, and I want to be exact about why, because the distinction is the difference between a legitimate tribunal and an illegitimate one. If I came to the institution from outside, bearing standards it did not hold, and convicted it of failing to meet them, I would be moralizing: imposing my measure on an object that never accepted it, which is precisely the external legislation Kant forbids. But that is not what I am doing, and the institution’s own organs prove it. I hold the defendant to standards its own appeal machinery already presupposes. The institution has already conceded, by building an appeals office, that its determinations can be wrong and ought to be examined for error. It has already conceded, by convening a mortality conference, that its outcomes must be reviewed against its own commitments. It has already conceded, by printing a grievance procedure, that the persons it acts upon are entitled to contest its actions and have the contest heard. Every one of these concessions is a standard the institution set for itself, and my critique does nothing more than hold the institution to the standards it has already, in its own architecture, confessed that it owes. The critique of institutional reason does not import foreign standards from outside and press them upon an unwilling object. It completes and radicalizes an examination the institution’s own structure has already confessed to needing, by building the very organs that conduct it. The critic is not an outside authority shouting at the institution. The critique is the institution’s own logic of appeal, followed to its conclusion and denied the funding-starvation and jurisdictional timidity that keep it ornamental. The standard is the defendant’s. I have only refused to let the defendant apply it in bad faith.

And here institutional reason turns out to possess a resource that Kant’s object conspicuously lacked, one that makes the self-examination not merely possible in principle but staffed in fact. Kant’s reason had to examine itself with nothing but itself; there was no plurality of persons through whom the reflexive examination could be conducted, only the single faculty turning upon its own operations by argument. Institutional reason is different in a way that decisively favors the tribunal: it runs through persons, and persons can host the critique from inside the apparatus, at the very point where its cognition is executed. Every act of institutional cognition is carried out by a human being who retains the capacity to examine that act even as she performs it, and this means the critical tribunal has, within the institution, a standing staff of potential judges. The clinician who charts against the chart, who records what the record would otherwise miss, is institutional reason examining itself through a person. The reviewer who reopens the file the system had closed, who refuses to let closure stand as resolution, is the tribunal operating from within. The committee member who registers the objection, who casts the veto Kant reserved for the free citizen and refuses assent to a determination she judges wrong, is the Kantian free citizen appearing inside the institution and exercising exactly the right Kant said the critical tribunal requires. The tribunal is not merely possible within institutional reason. It is staffed, by every operator who retains the capacity to judge the institution’s judgments while standing inside its machinery, which is to say by everyone the institution employs to reason on its behalf.

Which sharpens, into something with an edge, the verdict on those orders where the organs are missing. If the presence of self-critical organs is what makes institutional reason a legitimate object of critique on its own terms, then their absence is not a neutral fact and certainly not a mark of rigor. Consider an order that provides no appeal, no reconsideration, no grievance channel, no correction pathway, no standing for the person to contest what has been determined about her: an order that issues determinations and permits no examination of them. Such an order is not thereby a stricter or more disciplined order, though it may present itself that way, as tough-minded, as decisive, as unwilling to second-guess itself. It is a dogmatic order, in the exact Kantian sense of the word: a claimant of validity that exempts its own claims from the examination that validity requires, that asserts correctness while refusing the one procedure by which correctness could be tested. And Kant told us in advance what such exemption earns. I entered the phrase into the record at the outset, and I will not spend it a second time; the court has it already, and it was just when Kant wrote it and is just now. An order that exempts its determinations from examination awakens exactly the suspicion Kant said exemption awakens, and forfeits exactly the honest respect he said only free examination can earn. Pre-critical institutional reason, the institutional reason that issues judgments and forbids their contest, is not severe. It is invalid: it claims a validity it will not submit to testing, which is to claim a validity it has not earned and cannot support. And it knows this, at some structural level it cannot quite admit, which is precisely why even the harshest institutions print the word appeal somewhere, build some organ of reconsideration however starved, gesture at some channel of contest however narrow. The gesture is a confession. The institution knows that a validity claim exempt from examination is naked, and it prints the appeal form to cover itself, which means that even in covering itself it concedes the standard I am holding it to.

And I owe the court a reconciliation before this mark closes, because I opened this proceeding by declaring that the trial could not convene in the defendant's building, and I have spent this mark praising tribunals the defendant built inside it. The two claims divide the labor correctly, and the division must be stated or it will be mistaken for a contradiction. What Kant required to be internal is the standard, and here it is internal: every criterion this court applies is one the defendant's own organs already confess. What may be internal is the machinery, and it is, in the truncated organs entered above. What cannot be internal is the seat, and the reason is a disanalogy with Kant's object that must be named. Kant's reason was not a party. It had no budget to protect, no liability to fear, no interest an adverse verdict could injure, which is why reason could examine reason without the arrangement collapsing into judge, defendant, and clerk. An institution is a party. It has everything to fear from its own verdict, and the fifth mark's own exhibits, the starved ombuds office, the grievance channel built for ornament, show what an interested party does with a court it funds. So the seat moves outside, not to import a foreign standard but to protect the domestic one: Kant demanded free and public examination, and an examination conducted on the defendant's budget is neither. The law this court applies was found inside the building. The bench that applies it cannot be paid by the defendant. That is the whole reconciliation: the standard stays the object's own; only the purse and the docket leave the premises.

The fifth mark is proven, and with it the last of Kant’s five requirements is met. Institutional reason can be tried by its own lights, because it possesses, in its own appeals and audits and ethics consults and dissent channels, the organs of exactly the self-examination critique requires; and those lights are already burning, low and underfunded and hemmed in, in offices the institution built and then starved, staffed by persons who retain the capacity to judge the institution from within it. The tribunal I am convening is not foreign to the defendant. It is the defendant’s own proto-critical architecture, completed, radicalized, and denied the bad faith that keeps it ornamental. The examination is immanent rather than foreign. The standard is the defendant’s own. Critique of institutional reason is therefore not foreign legislation imposed from outside; it is the completion of what the institution’s own organs were already, however feebly, attempting to perform.

The Sixth Mark: The Authored A Priori

Five marks establish legitimacy. Institutional reason makes claims, possesses unity, operates under a mappable prior structure, generates a native and detection-proof illusion, and can host its own examination through the very organs and persons it already contains. The motion to dismiss is dead. The category error is dissolved. There is a defendant, the defendant reasons, and the defendant can be tried by its own lights. Critique of institutional reason is not a category mistake but a coherent tribunal, as coherent against institutional reason as Kant’s was against reason itself. That much is proven, and it is a great deal. But it establishes only permission. It shows that institutional reason may be critiqued. It does not yet show that it must be, and the distance between may and must is the distance between an academic exercise and an obligation. I now cross that distance, and I cross it with the one mark Kant’s object never bore and never could bear. I ask for your full attention here, because this is the hinge of the entire proceeding, the point at which the case stops being philosophy about institutions and becomes a case against them.

Begin with the feature of Kant’s a priori that his whole method depended on and that has gone almost unremarked because it seemed too obvious to state. Kant’s a priori has no author. Space and time were legislated by no one. The categories were drafted in no session, adopted by no vote, signed by no hand. They are the conditions of any possible human experience, and they simply are: given with the structure of the mind, prior to every act of cognition, but produced by no act of anyone’s will. This is why they can be mapped but not amended. You can chart the forms of intuition and the table of categories with all the precision Kant brought to them, but you cannot petition to change them, cannot appeal them, cannot hold anyone responsible for them, because there is no one who made them and nothing about them that could have been otherwise. And this is precisely why the first Critique ends where it ends, in cartography and nothing more. Its final office is to draw the boundaries: here are the conditions, here are the limits of their valid use, here is the territory within which reason may legitimately operate and beyond which it falls into illusion. Live within them, Kant says, because there is nothing else to do with conditions that no one authored. There is no one to summon. There is no one who signed. The map is the whole of what critique can deliver when its object has no author, because a map is what you make of a territory that no one is answerable for.

The institutional a priori is signed.

That sentence is the pivot of this essay, and everything before it was built to make it land and everything after it collects on what it means. The forms established at the third mark, the forms that constitute what a person can be for an institution, are not given with the structure of anything. They were made. Someone drew the catchment line that determines who falls within the jurisdiction and who falls outside it. Someone set the filing deadline that determines when a reality is permitted to count and when it arrives too late to be received. Someone chose the fields on the intake form, and in choosing them chose what the institution would be able to see about a person and what would have no place to be recorded. Someone wrote the code set, fixing the finite list of categories under which any person must be subsumed to be thought at all. Someone fixed the documentation requirement, the institutional schematism, deciding which papers would count as proving residence and which findings would count as proving disability and which conduct would count as proving risk. Someone designed the file and chartered the authorization, deciding what would be gathered into the institution’s unity and whose signature would convert an entry into a binding determination. Every one of these was an act, performed by a person or a body of persons, at a particular time, for particular reasons, and every one of them could have been performed otherwise.

I can be entirely concrete about this, because at least one of the forms carries its authorship in the published record where anyone can read it, and I enter it as a specimen of the whole. Consider the compressed clinical encounter, the visit fixed at something close to fifteen minutes, which is a temporal form in the exact third-mark sense: a schedule that determines how long a person’s reality is permitted to unfold before the institution and how much of it can be received. That form has authors, and I can cite them the way I cite any monograph. The resource-based relative value scale that reshaped how clinical work is valued and timed was developed by a named research team working in the late 1980s, who set out explicitly to estimate and quantify physician work and translate it into a relative scale (Hsiao et al. 1988). It was subsequently adopted into Medicare’s payment methodology at the start of the 1990s and became the template through which the temporal conditions of clinical encounters were structured across the system. The authorship here is a chain rather than a single hand: the scale, the adoption, the fee schedule, the scheduling template built to fit them. But a chain of signatures is not anonymity; it is a docket, and the same holds where the chain now ends in software, since a vendor’s default and a procurement officer’s acceptance are signatures too. Diffusion of authorship multiplies the summons. It does not void it. I am not, at this mark, arguing about whether that scale was wise or unwise; that is a separate question and not mine here. I am establishing a single fact with total precision: the temporal form inside which millions of compressed encounters have since occurred is not a natural condition of clinical reality but an authored one, with named designers, an identifiable adoption, a funding structure, and a date. Every rushed visit since occurred inside a form that someone built. The schedule was set. It did not fall from the sky.

And the specimen is not a peculiarity of medicine, because the same signature line runs beneath the temporal forms of every order this proceeding has named. The unit by which the school transcript counts learning at all is the Carnegie unit, promulgated by the Carnegie Foundation for the Advancement of Teaching in 1906 as a standard measure of instructional time and carried into general adoption within a generation. The procedural clock of the court docket, the deadlines and windows inside which a legal reality is permitted to count, runs on the Federal Rules of Civil Procedure, drafted by a named advisory committee under the Rules Enabling Act of 1934 and adopted in 1938. A payment scale, a foundation's standard, a committee's draft: three orders, three temporal forms, three sets of names in the published record. Substitute either of them for the clinical specimen and the sixth mark runs unchanged.

Generalize the specimen and the point is complete. The conditions of institutional cognition are not found, the way Kant’s conditions are found, lying in the structure of the mind for the philosopher to excavate. They are enacted, budgeted, procured, promulgated, and maintained, the way any built thing is. Jurisdiction is drawn. Schedule is set. Categories are drafted. The evidentiary schematism is promulgated. The file is designed. Authorization is chartered. Every single position in the third mark’s architecture, every form that constitutes what a person can institutionally be, has a signature line beneath it, an author or a body of authors who put it in place and could have put something else in place instead. And here the litigation metaphor stops being a metaphor and becomes the literal description of the situation. In the third mark I said that discovery was complete before the trial began, because the institution had published its own a priori. Now the meaning of that fact matures. Those published forms are not merely a map of the institution’s cognition. They are exhibits, each one bearing the name of the party who authored it, and the mapping of the institutional a priori is therefore not cartography at all. It is the discovery of evidence. Every intake form is an exhibit with a defendant’s name on it.

The normative bridge is this: whoever authors a condition that structures another person’s institutional appearance assumes a duty to keep that condition answerable to the person whose appearance it governs. The duty does not arise because the author intended harm. It arises because the authored form exercises constitutive power over someone who did not author it. A rule that determines what can count as need, credibility, eligibility, risk, compliance, disability, competence, or deservingness does not merely organize an administrative process. It helps determine the terms on which a person can become institutionally real. To author such a rule is therefore to exercise practical authority over another person’s appearance before the order. That authority is legitimate only if the condition remains corrigible by the person it governs, accountable for foreseeable failures in reception, and open to repair when the form substitutes its representation for the reality it was built to receive.

The defense sees where this is going and raises the objection it has been saving, which is the strongest one available and must be met in full. The forms, the defense says, are neutral. They apply to everyone identically. The deadline falls on every applicant alike; the category admits or excludes by the same criteria for all; the documentation requirement demands the same papers from everyone who comes. No one was singled out. No one intended any harm. Whatever difficulty a particular person encounters is the impartial operation of a rule that plays no favorites, and you cannot convict an author of wrongdoing for writing a rule that treats all comers the same. Intent is absent, partiality is absent, and without them there is no wrong to charge, only the inevitable friction of any system that must have rules at all.

The civil rights tradition has already met this objection squarely, in a holding I am proud to enter into this record, and the answer it gave is the sixth mark stated in judicial prose. In Griggs v. Duke Power Co., the Supreme Court confronted authored conditions of exactly this kind: a high school diploma requirement and a standardized general intelligence test, imposed as conditions of employment and advancement, both perfectly neutral on their face, applying to every applicant by the same terms, and both operating to exclude Black workers from advancement at drastically higher rates. The employer’s defense was precisely the defense I have just stated: the requirements were neutral, applied to all, and adopted without discriminatory intent. The Court rejected that defense and rejected the premise beneath it. What the statute reaches, the Court held, is not only overt and intentional discrimination but practices that are fair in form but discriminatory in operation (Griggs v. Duke Power Co., 401 U.S. 424, 431 [1971]). Facial neutrality is no defense when the operation of the condition produces the exclusion. The authorship cannot hide behind the neutrality, because the neutrality is a property of the form’s wording and the harm is a property of the form’s operation, and it is the operation that the law reaches.

Understand what that holding recognized, because it is the whole of what I need and it was won in a court of law before I ever brought it to a court of philosophy. It recognized that a requirement is conduct. A form is conduct. A criterion is conduct. To design a condition is to perform an act, the act has consequences that follow from the design, and those consequences are chargeable to the authors and maintainers of the design whether or not any author intended them, because the design produced them and the authors produced the design. Intent is not the hinge of responsibility here. Authorship is. The one who writes the condition answers for what the condition does, in the same way and for the same reason that the one who builds a thing answers for how the thing operates, regardless of what was in his mind when he built it. And it is no accident that the law arrived at this recognition through the cases of the despised, through the workers a neutral requirement kept down. The people most decided-upon by authored conditions have always been the first to discover that the conditions were authored, because they are the ones who feel, in the operation of the form against them, that a form is not a fact of nature but the act of a hand that could have written otherwise. The privileged experience the institution’s forms as the neutral background of the world. The excluded experience them as decisions, because for them they are.

Let me be exact about what I take from the case and what I do not, because a lawyer will note, correctly, that the doctrine has had a contested career: the holding is statutory rather than constitutional, the constitutional cases went the other way and required a showing of intent (Washington v. Davis, 426 U.S. 229 [1976]), and later Courts have narrowed what Griggs began. None of that reaches what is entered here. I do not cite the holding for its reach; I cite it for its recognition: that a requirement is conduct, and that conduct is chargeable in its operation. A court saw that once and said it plainly, and the seeing is not unsaid by the history of how far other courts were willing to carry it. The philosophy needs the recognition. The litigation over its scope belongs to the litigators.

And authorship, I must remind the court before the defense retreats to it, is not a past event that happened once and passed into history. It is a subscription, renewed continuously, and this changes everything about the responsibility it carries. The form is reprinted every year. The code set is relicensed on a schedule. The regulation is readopted, the policy manual reissued, and the budget line that keeps the compressed encounter compressed is renewed every single cycle by someone who could decline to renew it. Every reprinting is a re-signing. Every renewal is the authoring performed again, by present hands, in present conditions of knowledge. And the present conditions of knowledge are the crux, because institutions keep meticulous records of their own failures. They maintain complaint logs. They track appeal reversal rates, which record how often their determinations were found wrong on contest. They compile disparity data, which record how their neutral forms operate unequally across populations. They keep grievance files, incident reports, audit findings. Every one of these records is notice, in the exact legal sense: documented knowledge, held by the institution, of how its own conditions operate and whom they harm. And a condition maintained after notice of its harms is categorically different from a condition whose harms were not yet known. It is no longer inherited error, the innocent perpetuation of a form whose effects no one had seen. It is knowing conduct, renewed with the harms on file, re-signed by authors who possess, in their own records, the evidence of what they are re-signing. The receipts are in the defendant’s own cabinet, filed by the defendant’s own hand.

Now I join this to what I lodged at the fourth mark, and the joinder is the engine of the whole conversion from may to must. At the fourth mark I proved foreseeability: substitution is not an accident that befalls the institution but the predictable equilibrium its own structure produces, derivable in advance from the forms themselves. I said then that foreseeability was the seed of duty and that I would collect on it here. I collect on it now. Set the two facts together. First, the characteristic harm of institutional reason is foreseeable, readable off the design before the design is operated. Second, the design is authored and re-authored continuously by identifiable hands who hold, in their own records, notice of exactly that harm. A foreseeable harm, flowing from an authored and maintained design, imposed on persons who had no hand in the authoring and no power to amend it: this is not the structure of a tragedy. Tragedy is what befalls people through conditions no one made and no one can answer for, and if the institutional a priori were authorless, tragedy is all this would be, and cartography is all critique could offer, exactly as with Kant. But the institutional a priori is not authorless, and so this is not tragedy. It is the precise structure of a cause of action: a harm, a design that produces it, authors who maintain the design, and notice that closes off the excuse of ignorance. What is authored can be otherwise, because what was made by a choice could have been made by a different choice. What can be otherwise, and wounds, and is renewed with notice of the wound, is answerable. The mapping of the institutional a priori is therefore never mere cartography, however much it may resemble Kant’s in method. It is the assembly of evidence against answerable parties, and every intake form is an exhibit with a name on it.

The legal analogy matters because it names a structure of responsibility, but the obligation I am drawing is not merely legal. It is moral before it is legal. Because authored institutional conditions structure another person’s possible appearance before the order, any condition governing access to recognition, care, credibility, liberty, standing, or relief carries a duty of answerability to the reality it governs. The reason is simple: the person did not author the form, yet the form helps decide what the person can institutionally be. A power that constitutes another’s appearance without that person’s authorship must remain open to that person’s contest, correction, and demand for repair. Otherwise authored structure becomes unanswerable rule over someone else’s reality.

From this mark, the second set of prior conditions follows, and I said at the outset that this second set was the deeper half of the proof’s payload. These are the prior conditions of institutional morality: the moral facts that stand upstream of every institutional judgment and condition its validity, exactly as the cognitive conditions established at the third mark stand upstream of every institutional judgment and condition its possibility. The first set fixes what the institution can know. This second set fixes what its knowing owes. There are four, and I will state and ground each one on the materials this proceeding has already placed in the record, importing nothing from outside, because a moral condition asserted rather than derived would be exactly the foreign legislation the fifth mark forbade.

The first prior moral condition is the priority of the person, and I derive it directly from the third and fourth marks rather than asking you to grant it. Recall what the third mark established: the institution’s entire apparatus is built to receive a reality. The forms of intuition receive the person into institutional space and time; the categories receive her under institutional concepts; the schematism receives her evidence; the file receives her into institutional unity. Every one of these is a form of reception, and reception has a logical presupposition that cannot be evaded: there must be something there to be received, something real prior to and independent of the receiving, or there is nothing for the forms to format and the entire third-mark architecture is a machinery for receiving nothing. An apparatus built to receive persons presupposes persons who are there to be received, real before the reception and independent of it, exactly as a sense organ built to receive light presupposes light that exists whether or not the organ receives it. The person is therefore prior to her institutional reception, not as a premise I import but as a condition the apparatus itself requires in order to be the receiving apparatus the third mark proved it to be. And the fourth mark sharpened this into something the person feels in her own case. Recall the dark inversion of the file: the institution’s unity of apperception gathers the person’s representations not for her but into itself, and then turns the synthesized case back to face her as a second self she must answer for. But she can only be confronted by that second self, can only experience the file as a distortion or a stranger or a version of her that leaves things out, if she is someone prior to and other than the file, someone the representation is a representation of. The very possibility of the person standing before her own record and finding it inadequate presupposes that she exceeds it, that she was real before it was assembled and remains real beside it. So the priority of the person is not borrowed from anywhere. It is entailed twice over by what this essay has already proven: reception presupposes something received, and the confrontation with the file presupposes someone the file fails to exhaust. Persons are morally real before any order receives them; the order’s reception answers to that prior reality and never confers it. And this fixes, once and for all, the direction of answerability, which is the entire moral yield of the condition. The institution answers to the person, because the person is the prior reality the institution was built to receive and by which its reception can be measured and found wanting. The person does not answer to the form. She is the reality the form exists to serve, and any answering she is required to do to the form is a courtesy the form must earn by receiving her faithfully, not a debt she owes it by nature.

The second prior moral condition is authorial answerability, and it follows immediately from the central result of this mark. Because the a priori is enacted rather than found, whoever enacts and maintains it answers for what it can and cannot receive. This condition relocates responsibility to where the decisive acts actually occur, and the relocation is the whole of its force. There is a responsibility that attaches at the moment of decision, when the operator applies the form to the person at the counter, and that responsibility is real. But it is downstream of a prior responsibility that attaches at the moment of design, when the author fixes what the form will be able to receive and what it will be structurally unable to see. The clerk who applies the field faithfully is not the author of the field’s blindness; the author of the field is. So the responsibility for what the institution can and cannot cognize runs upstream, past the operator at the counter, past the office, to the drafting committee that wrote the criteria, the legislature that set the deadline, the vendor whose software determined what the chart could hold and what it would force into a box or drop entirely. My summons runs to all of them, because they are the parties who authored the conditions under which every downstream determination was made. And this yields a rule that the condition demands and that I state plainly: there can be no anonymous constitutions. A condition that constitutes what a person can institutionally be, and that cannot be traced to an author who answers for it, is a determination imposed on persons by no one they can call, which is precisely the situation in which authored power disguises itself as natural fact and escapes all accountability. Much of institutional reality has drifted into exactly this condition, constituted by forms whose authorship has been lost, diffused, or hidden behind proprietary systems no affected person can petition. Against that drift the condition sets its rule. Every field has an author. Every author has a name. Every name can be called. A constitution that cannot be attributed cannot be legitimated, because legitimacy is owed to persons and can be demanded only from someone, and a form with no answerable author is a claim to authority that has placed itself beyond the reach of the only parties entitled to question it.

The third prior moral condition is non-sovereignty, and I state it as a theorem, derived from results already proven, rather than as a preference I am urging. Assemble the premises. The third mark proved that the institution’s cognitive apparatus is derivative: built to receive a reality it did not create and, being finite and formal, cannot exhaust, because the person is never exhausted by the forms that receive her. This mark has proved that the apparatus is authored: put in place by parties who answer for the build. Now draw the conclusion that follows necessarily from those two premises. A cognitive apparatus that is both derivative, receiving a reality it did not make, and authored, built by answerable parties, cannot issue final judgments about the original reality it was built to receive. It cannot, because a derivative representation is answerable to what it represents, and an authored instrument is answerable to the parties who authored it and to the reality they built it to serve; nothing that is answerable in both these ways can be the court of last resort about the reality it answers to. Therefore every institutional validity claim about a person is a subordinate claim, reviewable in principle by the reality it is about, which is to say by the person it judges. The institution’s determination that you are noncompliant, ineligible, unfounded, failing, is not a final truth about you handed down from a sovereign court. It is a subordinate claim made by a derivative and authored apparatus about a reality that exceeds it, and it stands permanently subject to review by that reality. The institution’s court is never the final court, because the institution’s cognition is never the sovereign origin of the reality it judges but always a derivative reception of it. This is not a limit I am imposing on institutional authority from outside. It is a limit that follows from what institutional authority is: the authority of an apparatus built by someone to receive something it did not create, which is by its nature the authority of a subordinate and not a sovereign.

The fourth prior moral condition is constitutive contestability, and it is the condition in which the fourth mark and the sixth mark meet and jointly yield a conclusion sharper than either produces alone. Hold the two results together. The fourth mark proved that institutional reason manufactures a specific, foreseeable, detection-proof error: substitution, generated structurally by the same forms that make its cognition possible, and surviving every attempt to know it away. This mark has proved that the conditions of institutional reason are authored and maintained with notice, re-signed continuously by parties who hold the records of their own failures. Now draw the joint conclusion. A final or person-defining judgment issued without any machinery of contest, inside a system known to manufacture a specific error, by authors who are on notice of that error, is not a stern judgment or a rigorous one. It is void as a claim to final validity, and I mean void in that strict sense: a validity claim that has structurally exempted itself from the only discipline capable of catching its own known and characteristic failure. Consider what the claim amounts to. The system is known to produce substitution. Contest is the one procedure by which a particular instance of substitution could be caught, the one mechanism through which a person could say the file is not me, the record has replaced the reality, and force the apparatus back into contact with the person it was built to serve, a contact the third mark's porosity holds open in principle. To issue a judgment and foreclose contest is therefore to issue a judgment while disabling the sole corrective for the very error the system is known to generate. That is not toughness. It is the deliberate blinding of the apparatus to its own characteristic failure at the exact point where the failure would show. Contestability, then, is not a courtesy appended to institutional judgment after the fact, a mercy the institution may extend to the persons it judges if it is feeling generous.

This does not mean that every institutional act requires the same form of contest, or that every minor classification must trigger a full adversarial proceeding. Contestability is proportional to the force of the judgment. The more final, person-defining, durable, mobile, or consequential the institutional determination becomes, the stronger the required machinery of contest must be. A temporary scheduling notation does not require the same procedure as a finding of noncompliance, ineligibility, incompetence, risk, neglect, or fraud. But wherever the judgment claims durable authority over the person’s standing, credibility, access, liberty, care, or identity before the order, contestability is no longer optional. It is a validity condition of institutional judgment, in the way that answerability to evidence is a condition of empirical assertion. The first mark established what asserting is: to assert is to stake correctness, and a stake that can never be collected is not a stake. A judgment that has arranged, in principle, never to meet what could show it wrong has not hardened itself; it has stopped asserting, because it has withdrawn from the only game in which its words counted as claims. An institutional judgment that cannot in principle be contested by the person it is about, once it claims final or person-defining authority, has in the same way severed itself from the only thing that could correct its known and structural error, and has thereby forfeited the final validity it claims. The appeal is not mercy. The appeal is epistemology: the procedure through which an institutional judgment maintains its contact with the reality that alone can validate it. An order that forecloses contest has not toughened its judgments. It has forfeited them, because it has cut them off from the only discipline that could make them valid claims rather than unreviewable assertions.

Now set the two tables side by side, because together they are the entire apparatus this proof was built to expose, and seeing them together is seeing the whole of what has been established. The prior conditions of institutional cognition, from the third mark: jurisdiction, schedule, categories, evidentiary schematism, documentary unity, authorization. These fix what the institution can know: the forms into which a person must be poured to appear at all. The prior conditions of institutional morality, from this mark: the priority of the person, authorial answerability, non-sovereignty, constitutive contestability. These fix what the institution’s knowing owes: the moral facts that stand upstream of every determination and condition its validity. The first table is the architecture of institutional cognition. The second is the architecture of institutional obligation. And the decisive feature they share, the feature that this entire proof has been driving toward, is that both are settled before any person ever arrives at any counter. The forms are in place before she walks in, and so are the obligations, because the obligations follow from the forms and from the persons the forms exist to receive. Which means the moral situation of institutional life is decided earlier than anyone standing inside it can see. It is not decided at the counter, in the encounter between the operator and the person, though that is where it appears to be decided and where both parties experience it as being decided. It is decided upstream, at the moment of design, in the drafting session and on the signature line, before the operator was trained and before the person applied. The operator at the counter is often applying, faithfully and in good conscience, conditions whose moral character was fixed by someone else long before either of them arrived. So the wrong, when there is a wrong, has in the usual case already happened by the time the person walks in. It did not happen at the window. It happened in a drafting session she will never hear about. It happened on a signature line she will never see. It is happening now, as this sentence is read, in rooms the person will never enter, on forms she has not yet been handed, authored by parties she will never be able to name unless the rule of this mark is enforced and the constitution is made to bear its authors’ names.

The sixth mark is proven, and with it the conversion is complete. The conditions of institutional reason have authors. The authors have notice, filed in their own records, of the foreseeable harms their conditions produce. And an authored condition is conduct, chargeable to its authors in operation whether or not they intended its effects, exactly as a facially neutral requirement that excludes in operation is chargeable to the party who imposed it. Five marks opened the court and established that institutional reason may be tried. The sixth mark compels the trial and establishes that it must be, because what is authored, foreseeable in its harm, maintained after notice, and imposed on persons who could not consent to it and cannot amend it is not a condition to be mapped and lived within. It is a wrong to be answered, and the answering is owed: owed by the institutions to the persons they were built to receive, owed by the authors to the realities their forms decide in advance, owed by critique to the actual architecture in which modern persons are known, judged, sorted, and kept. And an owed answering that no bench will hear is not a debt deferred but a debt dissolved, extinguished by the recess itself, because answerability exists only where someone can be made to answer, and a right that no forum will seat dies in the silence. A court with jurisdiction over a continuing injury does not hold that jurisdiction as a discretion. The court is not merely open. It is obligated to sit.

Closing Argument

Let me give you the case as it now stands, element by element, because I have carried a burden and I intend to show it discharged before this court and leave nothing for the reader to take on faith.

An activity is a legitimate object of critique, in the strict sense Kant gave the word, if it claims validity, possesses unity, operates under mappable prior conditions, breeds illusion from its own structure, and can host its own examination. Those are the five marks, read off Kant’s own procedure rather than invented for the occasion, and I have proven each one from the defendant’s own exhibits, refusing at every step to argue from anything but the paper the institution itself produced. Institutional reason claims validity in every letter it sends, in the grammar of approved and denied and eligible and unfounded, and it confesses the claim in every appeal form it prints, because you cannot appeal a thing that made no claim, and the appeal form is the defendant’s own admission, drafted in advance and distributed at every counter, that its outputs are judgments capable of being wrong. It possesses unity: the judgment that is nobody’s, the determination every person in the room may doubt and the system produces anyway, generated by decision structures no single member’s mind contains, and I showed you that unity with three judges and two questions, where the very same three opinions yield acquittal polled one way and conviction polled another, so that the judgment belongs to the structure and to no member of it. It operates under a printed a priori: jurisdiction and schedule, category and schematism, file and signature, each functioning within institutional cognition as a historical and authored analogue to the conditions Kant mapped in reason, reason with its conditions showing, produced in discovery before I ever asked, because the institution published its own conditions as a matter of course. It breeds substitution along four gradients that no good intention flattens: availability, durability, auditability, mobility, on every one of which the file outcompetes the person, so that the migration of response from person to proxy is not the aberration of institutional life but its equilibrium, an illusion that survives its own detection exactly as the moon at the horizon survives the astronomer’s knowledge, because the pull is not a belief but a gradient built into the conditions of the work. And it already contains the organs of its own tribunal: appeals, audits, grievances, ethics consults, dissent channels, every one a proto-critical organ by which the institution turns upon its own outputs, staffed by the very hands that execute its cognition, the clinician who charts against the chart and the committee member who casts the veto Kant reserved for the free citizen. That last mark is my proof that this trial is self-examination and not foreign rule, that I hold the defendant only to standards its own architecture already confesses it owes. Five marks. The motion to dismiss is denied. The category error dissolves. There is an it, the it makes claims, and claims are what tribunals are for.

Then the sixth mark passes the case from standing to sentence, from may to must, and it does so on the strength of the one feature Kant’s object could never possess. Kant’s conditions had no author. The point is not that institutions possess a transcendental subject, but that they possess historically authored conditions performing transcendental work within institutional cognition. Space and time and the categories were legislated by no one, could be mapped but never amended, and answered to no one, which is why the first Critique could end only in cartography: here are the conditions, live within them, for there is no one to summon. These conditions have authors. They have budgets, renewal cycles, and complaint logs. Someone drew the jurisdiction, set the schedule, drafted the categories, fixed the evidence, designed the file, chartered the authorization, and someone re-signs each of these every cycle it is renewed. And they are imposed on persons who were morally real before the first field was drafted, real because the whole apparatus is built to receive them and reception presupposes something there to be received, persons who never signed the form that decides them and hold no power to amend it. From which the chain follows and closes. An authored condition is conduct, chargeable to its authors in operation whether or not they intended its effects, exactly as a facially neutral requirement that excludes in operation is chargeable to the hand that imposed it. Conduct performed with notice of its harms, and the institution keeps the notice in its own files, in its own disparity data and its own reversal rates, is knowing conduct. And knowing conduct that forecloses the contest by which its own known error might be caught is void where it claims final or person-defining authority, because it has severed its judgments from the only discipline that could make them valid claims rather than unreviewable assertions. So the critique of institutional reason is not an academic genre and not borrowed prestige and not grievance dressed in philosophical clothing. It is owed: owed by the institutions to the persons they claim to serve and were built to receive, owed by the authors to the realities their forms decide in advance and in absentia, owed by philosophy itself to the actual architecture in which modern persons are known, judged, sorted, recorded, and kept.

Kant wrote that his age was the age of criticism, to which everything must submit, and in the same breath, in the same footnote, he named the two powers most tempted to exempt themselves, and one of them was legislation, the enacted order in its most authoritative form. He named the institution as a would-be exemptee in the founding document of critique itself, and then he left the summons unserved, because his object had no author and the tools to reach an authored object were not yet in his hand. That exemption has run two and a half centuries. I am here to tell you it ends where this proof concludes. Institutions reason. Their reason claims to be right. A claim to be right is a submission to the question of whether it is, and that submission is a summons, and the summons has now been served: on the clinic, on the school, on the court, on the welfare office, on the vendor whose software decided what the chart could hold, on the legislature that set the deadline, on every drafting committee that ever decided in advance, in a room the person would never enter, what a person could be.

So I ask this court for exactly what I have proven and not one thing beyond it: jurisdiction over institutional reason, and the duty to use it.

Bring the chart. Bring the record. Bring the file. Bring the authors of the forms. The court is open, and the docket is long.

---

Foreseeable Critiques and Replies

1. The discursive dilemma does not prove a ghostly group mind.

Critique.
The second mark may appear to rely too heavily on the discursive dilemma. A hostile reading would say that the three-judge example proves only that aggregation rules can generate outcomes discontinuous with individual member judgments. It does not prove that an institution is a unified Kantian rational subject, nor that the institution possesses a mind capable of bearing moral responsibility.

Response.
That objection mistakes the role the discursive dilemma plays in the proof. The dilemma is not asked to prove a group mind. It is asked to defeat a severance motion. The motion to sever rests on the hidden premise that a collective judgment is nothing but the aggregate of individual judgments, so that examining each individual exhausts the collective. The discursive dilemma falsifies that premise. It shows, with three judges and two questions, that collective judgment can arise at the level of procedure rather than at the level of any individual mind. That is all it is required to prove.

The positive account of institutional attribution is supplied elsewhere: by French’s internal decision structures, Searle’s constitutive rules, and List and Pettit’s theory of group agency. Those sources are not invitations to mysticism. They are the opposite. List and Pettit’s account is precisely a way to explain group agency without a ghostly group mind. To cite them is not to smuggle the ghost back in. It is to cite the exorcism.

The second mark requires claim-attributing unity, not psychological subjecthood. Kant did not place a human person in the dock. He placed reason there: a claim-making capacity whose conditions could be examined. The institutional analogue requires enough unity for claims to belong to an “it” whose conditions can be mapped. Decision structures supply that unity. Moral responsibility, in my argument, does not attach to a mystical institutional soul. It attaches at the sixth mark to the authors and maintainers of the conditions under which the structure reasons. The paper therefore runs on two levels: examination attaches to the structure’s claims; obligation attaches to the structure’s authors. The objection collapses those levels and then objects to the collapse.

For that reason, the revision clarifies the word “reasoner” at the second mark. “Reasoner” is not used psychologically. It names the functional site at which validity claims are produced, maintained, and revised under rules. That is all critique requires its object to be.

2. The institutional a priori is not a vulgarized Kantian a priori.

Critique.
The third mark may seem to confuse empirical artifacts with transcendental conditions. Intake forms, files, schedules, and documentation requirements are historical, revisable, material objects. They are not space, time, the categories, or the transcendental unity of apperception. A hostile reading would say the argument wants to have it both ways: treating institutional forms as transcendental when establishing cognition, then treating them as authored legal conduct when establishing duty.

Response.
That is not an inconsistency. It is the thesis.

The argument does not claim that institutional forms share Kant’s transcendental status. It claims that they perform a transcendental function within a historically bounded order of cognition. The paper says this at the point of first use: the parallel is “not identical in metaphysical status,” because institutional conditions are historical and authored, but functionally parallel in the role they play within an institutional order.

The distinction matters. Reichenbach separated the apodictic from the constitutive sense of the a priori. Friedman’s relativized a priori develops that insight: constitutive principles can define what counts as an object within a framework while remaining historically revisable across frameworks. That is the category institutional forms occupy. They are not necessary for experience as such. They are necessary for institutional cognition within a given order. Their necessity is operative, not metaphysical; constitutive, not eternal.

The claim that “one cannot sue a transcendental condition” is therefore correct, and the paper says so. Kant’s a priori has no author. That is why Kant’s critique could end in cartography. Institutional conditions differ precisely because they are authored. The sixth mark does not contradict the third. It collects on the difference the third mark already established.

The same holds for the file. The file is not the transcendental unity of apperception in Kant’s strict sense. It is the institution’s functional analogue of apperceptive unity: the unity into which dispersed entries must be gathered if the institution is to treat them as one case. The file is contingent in its material form. Some documentary unity is necessary for institutional cognition of a person at all. Without such unity, the institution cannot think its entries together.

Nor does the appeal to Foucault erase this difference. Foucault is cited for the category of historical conditions that are constitutive for a period while remaining transformable. The paper departs from Foucault on authorship. Archaeology may bracket the author. The sixth mark reinstates the signature at the level of the operative form: the form that was drafted, budgeted, procured, adopted, printed, coded, purchased, licensed, or renewed. Deeper anonymous strata do not immunize proximate signed conditions. The summons runs to the layer that has names.

Nothing is elevated here. The movement runs downward. The institutional a priori is discovered in the humility of the form. The bathos is the finding. The scandal is not that a form has become metaphysical. The scandal is that the conditions deciding what a person can institutionally be are often as lowly, revisable, and authored as a printed field.

3. The RBRVS example is a specimen, not the foundation of the argument.

Critique.
The sixth mark may appear to rely too heavily on the resource-based relative value scale and the compressed clinical encounter. A hostile reading would say that the RBRVS is a hyper-local artifact of gamified American healthcare administration, not proof of a general authored structure across institutional orders such as schools, courts, welfare systems, prisons, or licensing bodies.

Response.
The RBRVS example is not the inductive foundation of the sixth mark. It is a specimen. Its task is limited: to show, with unusual documentary clarity, that one temporal form of institutional cognition has identifiable authors, an adoption history, a funding structure, and a date. The general claim does not run from “RBRVS exists” to “all institutional forms are authored.” It runs from the nature of institutional forms themselves. Institutional forms are promulgated artifacts. Promulgation has promulgators.

The critic’s counterexamples prove the rule. Court procedure has authors: rules committees, statutory adoption, judicial promulgation, recorded amendments. Educational time has authors: the Carnegie unit was promulgated by a named foundation in the early twentieth century and became a governing unit of academic recognition. Welfare eligibility matrices have authors: they are codified in regulation, revised through administrative procedure, and preserved through public records such as the Federal Register. Legal classifications, school transcripts, risk instruments, intake categories, parole matrices, diagnostic thresholds, and documentation rules all bear chains of authorship. Their authorship may be diffuse, delegated, bureaucratic, or vendor-mediated, but diffusion is not anonymity. A chain of signatures is still a docket.

The RBRVS was chosen because its signature is unusually clean: named researchers, a published article, statutory adoption, and institutional uptake. It is not offered as typical in every incentive structure. It is offered as legible in its authorship. Where the rule under proof is that authored forms have authors, the best specimen is the form whose name is easiest to read.

The revision therefore adds parallel specimens from education and law, alongside a sentence on chain authorship. The sixth mark does not depend on a quirk of American healthcare. It depends on the authored character of institutional conditions wherever they are drafted, adopted, implemented, maintained, and renewed.

4. Griggs does not ground the epistemology; it answers the neutrality defense.

Critique.
The moral conditions may seem to make an illicit leap from Griggs v. Duke Power Co. to epistemology. A hostile reading would say that a statutory employment discrimination case about disparate impact cannot establish a general rule that institutional judgments without contestability forfeit their truth claims. On this reading, the argument launders a philosophical conclusion through the moral authority of civil rights law.

Response.
That objection misassigns the load. Griggs is not the source of the contestability theorem. It is used at one joint only: to answer the neutrality-and-intent defense. The defense says that a facially neutral rule, applied equally and adopted without discriminatory intent, cannot be wrongful merely because it wounds in operation. Griggs defeats that premise by recognizing that a facially neutral requirement can be conduct chargeable in its operation.

The epistemological conclusion is derived elsewhere. It follows from the internal architecture of the argument. First, the fourth mark establishes that institutional reason structurally generates a characteristic error: substitution, the displacement of the person by the institutional representation. Second, the sixth mark establishes that the conditions producing this error are authored and maintained after notice. Third, contest is the procedure by which a person can force the institution back into contact with the reality its representation may have displaced. Therefore, when a final or person-defining institutional judgment forecloses contest, it disables the only corrective for the error the system is known to generate. Such a judgment does not cease to exist. It forfeits the final validity it claims.

The conclusion is scoped. The argument is not that every unappealable institutional statement ceases to be a truth claim. The argument is that a judgment claiming final or person-defining authority over a person, while structurally refusing the procedure through which its known characteristic error could be corrected, cannot retain the final validity it claims. It remains a subordinate institutional claim. It is not sovereign truth.

The revision therefore fences Griggs carefully. The case is statutory, not constitutional. Washington v. Davis required intent in the constitutional line, and later courts narrowed the reach of disparate-impact doctrine. The paper does not borrow more than it needs. It takes from Griggs only the recognition that a requirement may be conduct chargeable in operation. The rest of the moral argument is carried by authorship, notice, substitution, and the norm of assertion.

The verification analogy has also been replaced. The stronger grounding is internal to the first mark: to assert is to stake correctness, and a stake that can never be collected is not a stake. A judgment that arranges in advance never to meet what could show it wrong has not made itself rigorous. It has withdrawn from the discipline that makes its claim to correctness answerable.

Closing

The pattern across these critiques is uniform. Each replaces a functional claim with an ontological one and then refutes the replacement: structure-as-subject for structure-as-claim-site; transcendental status for transcendental function; inductive base for specimen; derivation-from-precedent for answer-to-objection. But the paper asserts none of those replacements. Its vocabulary is functional from the first appearance of the institutional a priori, and its moral conclusion runs through authorship, notice, substitution, and answerability, not through the ontological promotion of a form.

On the charge of rhetorical grandstanding, I will say only this: the courtroom is not ornament laid over the argument. It is the argument enacted. An essay that claims jurisdiction must convene, and the conceit is the convening.

On the recommendation to reject: the review identifies, by my count, five points requiring clarification or sharpening, but no premise that must be abandoned. That is the profile of a revision, not a rejection. I have made all five revisions: I added a functional gloss on “reasoner” at the second mark; inserted a sentence at the third mark clarifying the deliberate departure from Foucault; added two further specimens and a chain-of-authorship sentence at the sixth mark; added a scoping paragraph on Griggs, with Washington v. Davis entered among the cases cited; and replaced the verification analogy with an assertion-norm grounding. These revisions address the review’s stated objections while preserving the article’s central argument.

The defendant remains in the dock.

---

Foreseeable Judicial Critiques and Replies

Because this essay speaks in the language of courts as well as philosophy, a separate set of objections should be anticipated from jurists. The following replies do not claim that the doctrine already exists in this form. They clarify what the argument would require of law if courts took institutional reason seriously as an authored structure of judgment.

A protocol before the motions are heard. No justice named below has ruled on this framework, and I put no words in any mouth. Each motion is constructed from the published record: opinions, dissents, concurrences, books, and law review articles bearing the author’s own signature, extrapolated no further than this proceeding’s method requires, which is to state every objection at the full strength its author’s commitments would give it, and then to answer it without remainder. The bench is taken as presently composed. If the personnel change, the motions survive them, because each motion is less a person than a school, and the schools are older than the seats. I argue the hostile wing hardest, and first, because a jurisdiction defended only against friendly judges has not been defended.

1. Justice Thomas: the motion from original meaning.

Critique.
The Constitution’s original public meaning contains no duty of answerability for institutional design. The judicial power is the power to decide cases under enacted law, not to enforce a philosopher’s validity conditions, and the essay’s prior moral conditions are natural law freelancing of exactly the kind that produced substantive due process: obligations derived from the structure of things rather than from any text the people ratified. The constitutional line is settled: discrimination requires discriminatory purpose, and effects alone prove nothing (Washington v. Davis, 426 U.S. 229 [1976]). Nor can the essay retreat to statute, because its crown-jewel precedent is itself illegitimate. His dissent in Inclusive Communities argued that disparate-impact doctrine sprang from agency practice rather than the text Congress enacted, and that racial imbalance does not disclose discrimination, since a disparity may have a hundred causes and a statistic is not a story (Texas Department of Housing and Community Affairs v. Inclusive Communities Project, Inc., 576 U.S. 519 [2015], Thomas, J., dissenting). The essay cites as judicial vindication a holding that was never the law Congress wrote. If the people want contestability rules, Article I is open for business. The remedy lies with legislatures, not with tribunals of readers.

Response.
Take the last sentence first, because it is not a defense against the sixth mark; it is the sixth mark. To say the remedy lies with the enactors is to stipulate everything the mark asserts: that the conditions are enacted, that enactors exist and can be found, that what was made by choice can be remade by choice, and that answerability runs to identifiable authors. This proceeding was never limited to Article III. Its bench is the public, and its summons runs also to the drafting committee and the legislature, which is precisely where the motion points. The dispute is venue, and the venue proposed is one the essay already subpoenaed. A motion that concedes the defendant’s identity, the authored character of the conditions, and the address for service of process has not defeated the case. It has entered an appearance.

Second, the demand for text and history is met by more history than most of the text he enforces. Coke voided the College of Physicians’ power to judge and profit from its own cause in 1610 (Dr. Bonham’s Case, 8 Co. Rep. 107a, 77 Eng. Rep. 638 [C.P. 1610]). Madison carried the maxim into the founding itself: “No man is allowed to be a judge in his own cause” (Madison [1787] 2003). And the leading originalist scholarship on the clause locates the original core of due process in the guarantee that deprivations run through contest before a tribunal other than the deciding power (Chapman and McConnell 2012), the same structural suspicion of institutional self-judgment that animates the conservative canon on administrative power (Hamburger 2014). Non-sovereignty and constitutive contestability are not imports from critical theory. They are the two oldest maxims in his own inheritance, audi alteram partem and nemo iudex in causa sua, wearing administrative clothes. An originalist may refuse my vocabulary. His sources make the refusal expensive.

Third, the mechanism he demanded is supplied. His Inclusive Communities dissent has one honest engine: imbalance without mechanism proves nothing. Granted, entirely, and let the record show this proof never argues from disparity. The fourth mark supplies what the dissent said was missing: a named error, a derivation of that error from the design itself, four gradients readable off the forms in advance, and notice of the error filed in the defendant’s own cabinet. That is not statistics plus suspicion. That is design, foreseeability, and notice: the ordinary architecture of responsibility in the private law he has never questioned, where the builder answers for the foreseeable operation of the build without anyone asking after his heart (MacPherson v. Buick Motor Co., 217 N.Y. 382 [1916]).

Fourth, his own reports enter the record. In Perez he wrote separately to condemn deference that lets the author of a rule also fix the rule’s meaning, concentrating powers the Constitution separates (Perez v. Mortgage Bankers Association, 575 U.S. 92 [2015], Thomas, J., concurring in judgment). That is the non-sovereignty condition in his prose: the writer of the form may not be the final judge of the form. And in TransUnion he dissented to insist that at the founding the violation of a private right was itself an injury the courts would hear, no downstream damage required (TransUnion LLC v. Ramirez, 594 U.S. 413 [2021], Thomas, J., dissenting). That is this essay’s upstream thesis, argued from original practice, against his own wing, over four signatures. I do not need Justice Thomas to adopt the framework. I need the court to notice that he has been enforcing two of its clauses for years, under other names, and that his motion asks me to abandon positions he has published.

2. Justice Alito: the motion against sociology in robes.

Critique.
The framework is disparate impact generalized into a metaphysics, and disparate impact is dangerous enough at statutory scale. His principal dissent in Inclusive Communities warned that effects liability pressures decisionmakers toward numerical management and punishes actors innocent of any wrong (Inclusive Communities, 576 U.S. 519 [2015], Alito, J., dissenting); the essay would extend that logic from employment tests to every deadline, field, and code in institutional life. It abolishes the distinctions that make responsibility administrable and just: act and omission, intent and effect, rule and application. It converts the neutral administration of general rules, the central achievement of the rule of law, into presumptive misconduct, and it does so in a vocabulary imported from the academic left; strip away the Kant and what remains is critical theory litigating under an alias. Its practical meaning is that every institution becomes a standing defendant and every disappointed applicant a plaintiff with a philosophy. Worse, the void doctrine abolishes finality itself: if unappealable judgments forfeit validity, the losing party never has to stop. And it wraps all of this in Griggs, borrowing the moral credit of the civil rights movement for a project the movement never endorsed.

Response.
The scope objection attacks a rule the essay does not contain. The framework voids one thing: finality claimed for person-defining judgments where contest was structurally foreclosed, and it voids only the finality, leaving the judgment standing as the subordinate claim non-sovereignty already showed every institutional claim to be. The proportionality paragraph is in the text: a scheduling notation is not a fraud finding, and the machinery of contest scales with the force of the judgment. Every neutral rule suspect: not asserted. Finality abolished: not asserted, and the law itself has long treated constitutionally adequate notice as a condition of binding judgment, not a decorative courtesy appended after the fact (Mullane v. Central Hanover Bank & Trust Co., 339 U.S. 306 [1950]).

The intent objection fares worse, because the second basis of responsibility the essay adds is the one his own private-law tradition installed a century ago and has never regretted. Cardozo did not ask what Buick intended; he asked what the wheel would foreseeably do (MacPherson, 217 N.Y. 382 [1916]), and the modern law of design defect asks the same of every manufacturer with notice on file (American Law Institute 1998). A steering column designed to fail is chargeable without a single bad heart in the company. A form designed such that substitution is its equilibrium, renewed after notice, is the same structure in different material. If the sixth mark is critical theory, then products liability is critical theory, and Justice Alito is invited to say so in print.

The floodgates have a levee, and he polices it. The law has separated general rules, contested through politics, from individualized determinations, contested through hearings, for over a century (Londoner v. City and County of Denver, 210 U.S. 373 [1908]; Bi-Metallic Investment Co. v. State Board of Equalization, 239 U.S. 441 [1915]), and the essay’s proportional contestability tracks that line exactly. And where does principled suspicion of self-judging institutions lead? To the holding he joined in Jarkesy: when the government seeks civil penalties for claims resembling traditional legal actions, the defendant may be entitled to an Article III court and a jury rather than an agency’s in-house tribunal (SEC v. Jarkesy, 603 U.S. 109 [2024]). The principle that emptied the Commission’s home court is the principle this essay states in general form. He may distinguish the SEC from the welfare office, but the distinction must be named. The principle cannot do the work silently. As for Griggs, the essay’s scoping paragraph already surrendered everything the motion demands: statutory not constitutional, Washington v. Davis acknowledged, the later narrowing conceded. The case is cited for one recognition, that a requirement is conduct chargeable in operation, and the motion nowhere denies the recognition. It only resents the source.

3. Justice Barrett: the motion from faithful agency.

Critique.
The most analytically serious motion, in two counts. Count one, authorization: federal judges are faithful agents of enacted law, and her scholarship argued that even long-pedigreed substantive canons sit uneasily with that office, because they permit courts to bend enacted text toward unenacted values (Barrett 2010). The essay’s validity conditions are a substantive canon of maximal strength: a judicially enforced presumption against institutional finality, derived from moral philosophy rather than from any text. Whatever its merit as philosophy, the faithful agent has no commission to enforce it. Count two, finality: the essay treats finality as epistemic arrogance, but finality doctrines are themselves authored conditions serving persons: repose for the defendant, reliance for third parties, an end to litigation for a system that must decide at scale. Res judicata, limitations periods, appellate deadlines: each is a form with authors, and by the essay’s own sixth mark those authors deserve the same charity the essay extends to persons contesting them. So the void doctrine is a dilemma: either it is ordinary due process in costume, in which case the essay adds nothing, or it is a new legal status, in which case it needs a source the essay never supplies. And her work on precedent adds an edge: she has argued that precedent binds softly where reasoning is weak (Barrett 2017), so appeals to Griggs’s reputation will move her not at all. She will grade the reasoning.

Response.
The dilemma omits its third option: the essay is an account of why the enacted guarantees are what they are, and an account is not a rival. When Mullane treated notice and the opportunity to be heard as constitutional conditions of binding judgment (Mullane, 339 U.S. 306 [1950]), when Armstrong required the hearing at a meaningful time and in a meaningful manner (Armstrong v. Manzo, 380 U.S. 545 [1965]), when Congress commanded that agency action fall where the agency failed to consider an important aspect of the problem (Motor Vehicle Manufacturers Association v. State Farm Mutual Automobile Insurance Co., 463 U.S. 29 [1983]), the law enacted, piecemeal, the validity conditions this essay derives whole. The faithful agent needs no new commission to enforce commissions already issued. What the essay supplies is the reason the commissions were worth issuing, and a faithful agent may read the reason without disloyalty. A theory of the statute is not a canon against it.

Count two is accepted in full, and it costs the essay nothing, because the essay drew the distinction first. Preclusion is not epistemic finality. A judgment may be final for repose, res judicata attaching, limitations running, and remain what the third moral condition showed every institutional claim to be: subordinate, a derivative representation answerable in principle to the reality it judges. The essay’s void is scoped to the case the law’s own finality doctrine already refuses to protect: finality claimed where contest was never available. Where contest was available and had, finality earns its repose, and the proportionality paragraph says so in terms. The two regimes do not collide. Hers governs when the contest ends. Mine governs whether it was ever allowed to begin.

And her canon-skepticism, applied evenhandedly, is the essay’s ally, because the presumption this essay attacks is itself an unenacted canon: the working judicial habit of treating the institutional record as presumptively conclusive, the file as the fact. No Congress enacted that presumption. It accreted, exactly as substitution accretes, along the gradients of availability and auditability the fourth mark mapped, and a judge who distrusts unenacted canons should distrust that one first, because it is the one applied every day, in every record review, without anyone ever having voted for it. Her reasoning-over-reputation test is welcomed rather than feared. The essay rested Griggs on one recognition and fenced everything else precisely so the reasoning could be graded on its own. Grade it. The whole document is an audit invitation.

4. Chief Justice Roberts: the motion for a limiting principle.

Critique.
The institutionalist’s motion, and the most practical. Grant everything: institutions reason, the forms are authored, substitution is real. A court must then ask the question philosophy never has to answer: where does it stop? If authored conditions maintained after notice ground answerability, then every deadline, field, catchment line, and code set in American life is a lawsuit awaiting a theory, and the judiciary becomes the standing design-review board of the administrative state, a role it has neither capacity nor warrant to fill. His court reviews decisions, not enterprises; it has held that institutional interests too unmeasurable to examine fail examination for that reason (Students for Fair Admissions, Inc. v. President and Fellows of Harvard College, 600 U.S. 181 [2023]); and the umpire’s office is to call the play, not to redesign the stadium. The essay’s proportionality paragraph is a gesture, not a rule: person-defining and durable are adjectives, and adjectives do not decide cases. Until the framework states an administrable test, it is a brief without a cause of action.

Response.
The test, stated as a test: contest is owed where a determination is individualized rather than general, the Londoner and Bi-Metallic line the Court has policed for more than a century, and the required machinery scales by the factors his court applies every day: the private interest, the risk of error, the burden of additional process (Mathews v. Eldridge, 424 U.S. 319 [1976]). The essay adds one thing to that calculus, and it is a gift rather than a burden: content for the risk-of-error prong. Mathews has always asked courts to estimate the risk of erroneous deprivation with no theory of where the errors come from. The fourth mark is the theory: substitution, a named, structural, documented error-risk, derivable from the design in advance, with the institution’s own reversal rates and disparity data as its measure. That is not less administrable than current doctrine. It is current doctrine with its blank filled in.

As for where the principle stops, ask first where it started, because the Chief Justice started it. Loper Bright holds that the institution administering a law may not be the final judge of what the law means (Loper Bright, 603 U.S. 369 [2024]): non-sovereignty, in his own majority’s prose. SFFA holds that institutional self-justification too unmeasurable to examine fails: the fifth mark’s verdict on ornamental self-review, his prose again. Jarkesy holds that, where the government seeks civil penalties for claims resembling traditional legal actions, the defendant may be entitled to an Article III court and a jury rather than an agency’s in-house tribunal: not the whole of this essay, but one recent expression of its suspicion toward institutions keeping their own docket (Jarkesy, 603 U.S. 109 [2024]). Three times since 2023 his court refused to let a claimant of validity be the unexaminable measure of its own claims. The essay does not ask him to open a door. It reads the hinge he has turned three times and gives its principle a name. And the umpire metaphor is accepted at full value, because an umpire is precisely what the essay demands: a judge of the play who is not a player. The fifth mark is nothing but the requirement that institutions accept one.

5. Justice Kavanaugh: the motion from concrete harm.

Critique.
The sharpest doctrinal weapon on the hostile side, because he has already litigated this essay’s central object, the false file, and ruled against it. In TransUnion, thousands of people carried credit files falsely flagging them as potential terrorists and criminals; for the 6,332 whose files were never sent to a third party, he held there was no Article III injury at all: “No concrete harm, no standing” (TransUnion, 594 U.S. 413 [2021]). That is the substitution thesis presented to the Supreme Court and rejected: a false representation, sitting in a database, undisseminated, is not a legal wrong. The essay locates the wrong at design, upstream even of the false entry; Article III will not travel even as far downstream as the entry itself. Add his methodological writing, clear rules over mushy standards (Kavanaugh 2016), and the settled instinct that transformative frameworks require clear congressional authorization (West Virginia v. EPA, 597 U.S. 697 [2022]), and the motion completes: the framework is unadministrable, unauthorized, and already rejected at its most sympathetic point of entry.

Response.
Read TransUnion as evidence rather than obstacle, because it drew this essay’s map and conceded its premise. For the 1,853 class members whose false files traveled, he found concrete harm, and he found it by analogy to defamation: the file, once published, wounds the way a libel wounds. Hold that analogy to the light. A libel is an assertion; only assertions defame. In selecting defamation as the closest traditional harm, the opinion concedes that the file is a claim-making artifact, a thing that says something about a person and can say it falsely. That is the first mark, granted in the act of limiting the sixth. The motion cannot run on the defamation analogy and deny that the file asserts.

What remains is a disagreement about tense, not about truth. Article III opens its ledger at publication; the essay opens the moral ledger at design; and the two are joined by the fourth mark’s mobility gradient, which his own facts illustrate: files are built to travel, and dissemination is the file’s function rather than its accident. A false entry awaiting transmission may not yet be publication for Article III purposes, but it is already a designed risk whose later operation is chargeable to the design once the record is used, transmitted, or relied upon. The law he knows handles that intertemporal structure without strain: in design defect, the wrong is authored years before the injury it awaits, and the injury, when it comes, is chargeable to the design (American Law Institute 1998). The essay does not quarrel with his timing rule. It identifies whom the clock, when it strikes, will summon. And if he wants the originalist answer to his own timing rule, it is already filed, from his right: Thomas’s dissent in the same case, four signatures deep, arguing that at the founding the violation of a private right was itself actionable, no downstream damage required. The essay’s upstream thesis is on the United States Reports, and the senior originalist put it there.

On administrability and authorization, the compile targets are itemized and pre-authorized: contestability compiles to due process (Mullane; Mathews); authorship-with-notice compiles to arbitrary-and-capricious review, the failure to consider an important aspect of the problem (State Farm, 463 U.S. 29 [1983]); non-sovereignty compiles to independent review. Nothing on that list required new congressional authorization when enacted, and nothing in the essay asks for more than enforcement. The major questions doctrine polices agencies claiming transformative new power from old text. The essay claims no power. It describes where the old text’s power was always pointed, which is a reading, and readings are what judges are commissioned to give.

6. Justice Gorsuch: the motion filed against his own exhibits.

Critique.
He will not attack the structure; he will attack the register and the remedy. The vocabulary codes left, and he will suspect that beneath it lies a license for judicial supervision of institutional design untethered from text, history, liberty, or jury: structural equity by another name, the freelancing he entered public life to oppose (Gorsuch 2019). He will demand that every operative claim be restated as a specific, historically grounded guarantee, fair notice, hearing, jury, nondelegation, or be dismissed as policy. And he will add an edge the left flank never will: if authored forms are the disease, then the essay’s remedy of more mandated structures, more standing organs, more required process, is more authorship, and the framework risks prescribing the pathogen as the cure. His own book documents the human toll of too much law (Gorsuch and Nitze 2024). The true remedy for too many forms is fewer forms.

Response.
Enter his own exhibits first. Over Ruled is a book-length catalog of persons crushed by authored conditions no one can trace or contest: the fisherman, the monks, the citizen lost in the agency maze (Gorsuch and Nitze 2024). Those are this essay’s exhibits with his byline on the spine. His Dimaya concurrence opens with four words that could caption the fourth mark, “Vague laws invite arbitrary power,” and grounds fair notice in the founders’ own grievances (Sessions v. Dimaya, 584 U.S. 148 [2018], Gorsuch, J., concurring). His Gutierrez-Brizuela concurrence condemned deference for letting one body write, interpret, and apply the rule, the judge-in-its-own-cause arrangement non-sovereignty forbids (Gutierrez-Brizuela v. Lynch, 834 F.3d 1142 [10th Cir. 2016]). And his Niz-Chavez majority held the government to the discipline of its own single form, refusing to endow the executive with maximum bureaucratic flexibility at the expense of the person the form decides (Niz-Chavez, 593 U.S. 155 [2021]). Restate the essay in his registers and nothing is lost: porosity plus contest is fair notice; non-sovereignty is his anti-deference; the anonymous constitution is the untraceable rule his book was written to indict. The motion objects to the accent, and the accent is not load-bearing.

The edge about remedy cuts, and it misses. More forms are not the remedy, and the essay never prescribes them as one; it prescribes answerable forms, a different axis entirely. A jurisdiction can shrink its code set and still foreclose contest; it can multiply its forms and keep every one signed, appealable, and corrigible. Fewer forms is a policy the sixth mark neither requires nor forbids, because the mark governs the relation between any form and the person it decides, however many forms there are. And his remedy concedes the mark exactly as Thomas’s does: to demand fewer and simpler authored conditions is to stipulate that the conditions are authored and that their authors could choose otherwise, which is the whole of the proposition on trial. Over Ruled does not rebut the sixth mark. It is the sixth mark, written for an audience that would never read Kant, and this proceeding claims it as an amicus brief filed early.

Last, the coding is refused on the record. The summons in this proceeding does not check party registration. It runs to the welfare matrix and the fishing regulation, the Medicaid denial and the tax letter, the school transcript and the licensing board. Substitution does not caucus. The gradients pull on the chart of the nursing home resident and the docket of the small businessman with identical force, because they are properties of files, not of ideologies. A framework that indicts the DMV and the SEC in the same paragraph is coded to the counter, and everyone stands at the counter eventually.

7. Justice Kagan: the motion for a rule, with a warning.

Critique.
Two-handed, and the second hand is the sharper. First: show the doctrine. Philosophy does not decide cases; tests decide cases. What is the cause of action, the standard of review, the remedy? Until the framework compiles into ordinary law, it is a mood. Second, from her Loper Bright dissent: the framework’s global suspicion of institutional cognition proves too much. Agencies know things courts do not: the science, the record, the thousand technical judgments no generalist can audit (Loper Bright, 603 U.S. 369 [2024], Kagan, J., dissenting). A doctrine treating every institutional determination as presumptively substitutive licenses judges to displace experts, and the displacement does not restore the person. It substitutes the judicial record for the agency record, one file for another, with less expertise behind it. The essay’s own logic condemns its likeliest judicial use.

Response.
The compile, itemized. Contestability compiles to procedural due process (Mullane; Mathews; and Goldberg, where the interest is survival and process must precede the deprivation, Goldberg v. Kelly, 397 U.S. 254 [1970]). Authorship-with-notice compiles to State Farm: a form maintained after documented notice of its distorting operation is a failure to consider an important aspect of the problem, reviewable as arbitrary and capricious. And non-sovereignty compiles to the structural-review line she wrote herself: Axon holds that being subjected to a proceeding whose very structure is unlawful is a “here-and-now injury,” contestable before any final order issues (Axon Enterprise, Inc. v. Federal Trade Commission, 598 U.S. 175 [2023]). That is this essay’s upstream thesis in her own majority: the structure of the deciding apparatus, and not only its output, is a reviewable wrong. Her narrow version, that review must reach the recognition infrastructure where distortion is documented and noticed, is accepted with gratitude as the doctrinal floor. The essay wrote the reasons. She may write the rule.

The warning is met head-on, because it is the deepest objection any justice can file: is judicial review itself substitution, one file replacing another? Yes, and the essay’s own conditions govern it, which is the answer. The fourth moral condition never says replace the institution’s cognition with a better cognizer’s. It says keep every cognition answerable to the person it is about. A reviewing court that reads the record without the person has substituted exactly as the agency did, and the framework indicts it identically; the remedy is never a wiser file but a summoned person, the reality in the room beside its representation. So the framework does not license generalists to displace experts. It conditions both finalities, the expert’s and the judge’s, on the same porosity. And expertise is precisely the resource substitution wastes, since the expert who never meets the patient behind the chart is expensive cognition spent on a proxy. Deference to expertise and contest by the person are not rivals. The second is how the first stays in contact with its object, and her dissent’s best argument, that agencies know the territory, is an argument for keeping the territory in the room.

8. Justice Sotomayor: the motion from the named.

Critique.
Friendly, and sharper for it. The framework speaks of the person in the abstract, and the abstraction is a choice with a history. The wounds do not fall on persons in general; they fall on the named: Black, brown, poor, disabled, unhoused, the people for whom the form was never neutral background but always a decision. A theory of institutional harm that never says race risks becoming one more facially neutral instrument, rigorous in form and evasive in operation. Her dissents supply the record: warrant databases holding millions of entries, concentrated where the policing is, converting a suspicionless stop into a lawful arrest (Utah v. Strieff, 579 U.S. 232 [2016], Sotomayor, J., dissenting), and the insistence in Schuette that constitutional silence about race is itself a statement about race (Schuette v. Coalition to Defend Affirmative Action, 572 U.S. 291 [2014], Sotomayor, J., dissenting). The motion: a critique of legibility that leaves its own demographics illegible has reproduced the defect it names.

Response.
Strieff is the fourth mark in a case caption, and the essay claims it. The warrant database is the file; the hit is substitution; the person on the sidewalk was displaced by the record before the officer’s hand moved; and her dissent named the person’s condition in five words this essay would be proud to have written: “just waiting to be cataloged.” Note who wrote the majority she was answering. Justice Thomas held that the record’s discovery attenuated the illegal stop, which is to say the proxy was permitted to launder the encounter, the file validating what the person’s own conduct never could. Her dissent is not an objection to this framework. It is the framework, reported from the bench, a decade early.

And the mechanism is the gift, aimed exactly where she aims. The standing conservative complaint against effects liability, from Washington v. Davis through the Inclusive Communities dissents, is imbalance without mechanism: the statistic is not a story, so the disparity proves nothing. The gradients are the story. They trace the causal path from authored form to unequal wound: whose files run thickest, whose records travel farthest, whose categories carry the most freight, whose contest is priced beyond reach. The framework does not abstract from race. It explains how facially raceless forms deliver raced outcomes, which is precisely the explanatory gap her opponents have exploited for fifty years, and the essay’s Griggs section already holds her historical point: the excluded discover authorship first, because for them the form was never weather. What her motion adds is the register of names, and the amendment is accepted in full. The person of this proof has always had a body, a neighborhood, and a race the form pretended not to see while sorting by it.

9. Justice Jackson: the motion against paper rights.

Critique.
The deepest friendly objection, from the one member of the bench who has worked both sides of the sentencing grid. Contest is itself a form. The appeal has a jurisdiction, a schedule, categories, an evidentiary rule, a file: a complete third-mark architecture of its own, authored by the same class of authors. So the fourth moral condition can be satisfied on paper and defeated in fact. An appeal that requires counsel the person cannot afford, literacy the form assumes, time the schedule forecloses, and stamina the process is built to exhaust is not a corrective; it is a second gauntlet that launders the first, and its existence legitimates, since the institution points to the channel as proof of answerability while pricing the channel beyond use. Her SFFA dissent states the general principle: “deeming race irrelevant in law does not make it so in life” (SFFA, 600 U.S. 181 [2023], Jackson, J., dissenting), and deeming contest available in law does not make it available in life. A framework that certifies the right without pricing its exercise has built one more ornamental organ, and the fifth mark taught the court what those are worth.

Response.
Guilty as charged, and charged first by the essay. The fifth mark’s indictment of ornamental organs, the starved ombuds office, the grievance channel built for show, is her motion stated as evidence, and the fourth moral condition was drafted with the loophole closed: contest must be capable of catching substitution in fact, and the adequacy standard is already enacted, a hearing at a meaningful time and in a meaningful manner (Armstrong v. Manzo, 380 U.S. 545 [1965]), with Goldberg’s recognition that where the interest is survival the process must arrive before the deprivation, not after it (Goldberg, 397 U.S. 254 [1970]). An appeal priced beyond use is not a cheap appeal. It is no appeal, and under the fourth condition the judgment it decorates has forfeited the same finality as if the channel had never been printed. The unwinnable appeal does not satisfy this framework. It violates it twice: once as foreclosed contest, once as false answerability.

So the motion is granted, and continued. What she names is not a flaw in the right but the price of honoring it: the political economy of contest, resources, representation, translation, time, the conditions under which a person can actually stand beside her file and be heard over it. That is the enforcement arm, and it is the next docket. This proceeding was always jurisdictional, the securing of the court, and the working drawings of the standing structures come after the verdict that they are owed. Her office and this essay’s divide the labor the way remedy and right always have: the essay establishes that the debt exists; she is pricing it; and a debt priced is a debt taken seriously, which is more than a debt denied. Her motion is entered not as an objection but as the first order of business for the court this essay convened.

Closing.

Nine motions, one pattern, and the pattern is this essay’s method returned in judicial keys. The hostile wing defends four lines: intent, finality, standing, and the judicial commission. The essay crosses none of them; it explains all four, and every explanation was drawn from the movants’ own paper. The originalist was answered out of Coke, Madison, and his own dissents. The textualist was answered out of clauses Congress enacted and courts already enforce. The proceduralist was answered out of the map his own standing opinion drew. The institutionalist was answered out of a hinge he has turned three times since 2024. The skeptic of the administrative state filed this essay’s exhibits as a book. And the friendliest justices offered no mercy worth less than the hostility: one supplied the case caption, one supplied the price list, one offered to write the rule. The hostile motions are denied out of their own authorities. The friendly motions are granted as amendments. And one fact survives all nine, because no motion touched it: the forms are authored, the authors are on notice, and the paper is already in the cabinet. The bench above may narrow the holding. It cannot narrow the record. The court that cannot be defunded remains in session.

---

Joshua Sandifer, NP-BC, is an adult-gerontology primary care nurse practitioner, infectious disease clinician, and independent scholar whose work examines personhood, recognition, and institutional life. His developing framework, Transordoism, investigates how institutions come to know, classify, judge, and act upon persons through forms, records, categories, thresholds, and procedures. He is the author of the Clinical Legibility framework and writes at the intersection of healthcare ethics, critical theory, disability, bureaucracy, and moral philosophy. His broader project, The Legibility Project, argues that institutions do not merely fail persons through individual bias or error, but through recognition infrastructures that determine in advance what can be seen, believed, documented, and repaired.

---

References

American Law Institute. 1998. Restatement (Third) of Torts: Products Liability. St. Paul, MN: American Law Institute Publishers.

Barrett, Amy Coney. 2010. “Substantive Canons and Faithful Agency.” Boston University Law Review 90: 109-82.

Barrett, Amy Coney. 2017. “Originalism and Stare Decisis.” Notre Dame Law Review 92: 1921-44.

Chapman, Nathan S., and Michael W. McConnell. 2012. “Due Process as Separation of Powers.” Yale Law Journal 121: 1672-1807.

Foucault, Michel. (1969) 1972. The Archaeology of Knowledge. Translated by A. M. Sheridan Smith. New York: Pantheon Books.

French, Peter A. 1984. Collective and Corporate Responsibility. New York: Columbia University Press.

Friedman, Michael. 2001. Dynamics of Reason. Stanford, CA: CSLI Publications.

Gorsuch, Neil M. 2019. A Republic, If You Can Keep It. New York: Crown Forum.

Gorsuch, Neil M., and Janie Nitze. 2024. Over Ruled: The Human Toll of Too Much Law. New York: Harper.

Habermas, Jürgen. (1992) 1996. Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy. Translated by William Rehg. Cambridge, MA: MIT Press.

Hamburger, Philip. 2014. Is Administrative Law Unlawful? Chicago: University of Chicago Press.

Horkheimer, Max. (1937) 2002. “Traditional and Critical Theory.” In Critical Theory: Selected Essays, translated by Matthew J. O’Connell, 188-243. New York: Continuum.

Hsiao, William C., Peter Braun, Douwe Yntema, and Edmund R. Becker. 1988. “Estimating Physicians’ Work for a Resource-Based Relative-Value Scale.” New England Journal of Medicine 319 (13): 835-41.

Kant, Immanuel. (1781) 1998. Critique of Pure Reason. Edited and translated by Paul Guyer and Allen W. Wood. Cambridge: Cambridge University Press.

Kavanaugh, Brett M. 2016. “Fixing Statutory Interpretation.” Harvard Law Review 129: 2118-63.

Kornhauser, Lewis A., and Lawrence G. Sager. 1993. “The One and the Many: Adjudication in Collegial Courts.” California Law Review 81 (1): 1-59.

List, Christian, and Philip Pettit. 2011. Group Agency: The Possibility, Design, and Status of Corporate Agents. Oxford: Oxford University Press.

Madison, James. (1787) 2003. “Federalist No. 10.” In The Federalist Papers, by Alexander Hamilton, James Madison, and John Jay, edited by Clinton Rossiter. New York: Signet Classics.

Marx, Karl. (1867) 1976. Capital: A Critique of Political Economy. Vol. 1. Translated by Ben Fowkes. Harmondsworth: Penguin Books.

Reichenbach, Hans. (1920) 1965. The Theory of Relativity and A Priori Knowledge. Translated by Maria Reichenbach. Berkeley: University of California Press.

Searle, John R. 1995. The Construction of Social Reality. New York: Free Press.

Weber, Max. (1922) 1978. Economy and Society: An Outline of Interpretive Sociology. Edited by Guenther Roth and Claus Wittich. Berkeley: University of California Press.

---

Cases cited

Armstrong v. Manzo, 380 U.S. 545 (1965).

Axon Enterprise, Inc. v. Federal Trade Commission, 598 U.S. 175 (2023).

Bi-Metallic Investment Co. v. State Board of Equalization, 239 U.S. 441 (1915).

Dr. Bonham’s Case, 8 Co. Rep. 107a, 77 Eng. Rep. 638 (C.P. 1610).

Goldberg v. Kelly, 397 U.S. 254 (1970).

Griggs v. Duke Power Co., 401 U.S. 424 (1971).

Gutierrez-Brizuela v. Lynch, 834 F.3d 1142 (10th Cir. 2016).

Londoner v. City and County of Denver, 210 U.S. 373 (1908).

Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024).

MacPherson v. Buick Motor Co., 217 N.Y. 382, 111 N.E. 1050 (1916).

Mathews v. Eldridge, 424 U.S. 319 (1976).

Motor Vehicle Manufacturers Association v. State Farm Mutual Automobile Insurance Co., 463 U.S. 29 (1983).

Mullane v. Central Hanover Bank & Trust Co., 339 U.S. 306 (1950).

Niz-Chavez v. Garland, 593 U.S. 155 (2021).

Perez v. Mortgage Bankers Association, 575 U.S. 92 (2015).

Schuette v. Coalition to Defend Affirmative Action, 572 U.S. 291 (2014).

Securities and Exchange Commission v. Jarkesy, 603 U.S. 109 (2024).

Sessions v. Dimaya, 584 U.S. 148 (2018).

Students for Fair Admissions, Inc. v. President and Fellows of Harvard College, 600 U.S. 181 (2023).

Texas Department of Housing and Community Affairs v. Inclusive Communities Project, Inc., 576 U.S. 519 (2015).

TransUnion LLC v. Ramirez, 594 U.S. 413 (2021).

Utah v. Strieff, 579 U.S. 232 (2016).

Washington v. Davis, 426 U.S. 229 (1976).

West Virginia v. Environmental Protection Agency, 597 U.S. 697 (2022).