Followable

What a rule must be in order to bind at all

This essay is the fourth of our instruments, following The Problem of Absolutes, Standing, and Addressability. The first traced the missing criterion beneath rankings of subjective magnitude over measured standards. The second identified what is violated when a good-faith challenge to a governing standard is closed rather than answered. The third found the architecture under which challenge survives — answered, carried, bounded. The present essay steps back one pace further, to a question prior to all three: before a standard can be defended, before a challenge to it can be answered or even raised, the standard must be capable of being followed. The question is what a rule must be in order to bind at all. The answer arrives by convergence: a Prussian philosopher, an American professor of jurisprudence, and a constitutional court — three separated lines of work across two centuries, none citing the others’ vocabulary — carve at the same joint. Obligation presupposes possibility. A system of rules that cannot be complied with fails as law, not merely as good law. A statute whose meaning must be guessed at gives no notice and licenses arbitrary enforcement. Independent arrival is evidence: when separated instruments find the same joint, the joint is probably real.

The essay then dissects the modern form of the failure, which is no longer the hidden text but the hidden trigger: rules that are perfectly public and epistemically impossible — conditioned on facts the addressee cannot observe, solvable only by acts the rule itself forbids, sealing a datum that sanctioned channels themselves can return as a byproduct. What such a rule cannot produce is discriminating compliance or targeted deterrence. What it produces instead is universal exposure — latent liability distributed across everyone the rule governs, held in reserve, available for enforcement from outside the addressee’s position. A rule whose boundary its addressee cannot locate and whose liability cannot be cleared through informed conformity is not a norm; it is a stored instrument.

The close returns the finding to the series’ spine. A rule is a species of address: it speaks to an agent and presupposes that the agent can answer, and compliance is the answer. A rule whose operative conditions its addressee cannot ascertain does not address him as an agent — it positions liability over him. The unfollowable rule is the standing violation in imperative form.

What cannot be followed does not command.

Suetonius records a complaint about the emperor Gaius — Caligula — that has outlived nearly everything else the man did. Having imposed new taxes, he was pressed into publishing the law; so he published it in minute letters, posted in a cramped and narrow place, so that no one could take a copy. [1] The obligation to make the rule public was discharged. The possibility of knowing the rule was not.

The complaint already contains its remedy: publicity. Carve the law in stone, post it in the forum, read it aloud. If the citizen can reach the text, this failure is cured.

The cure no longer reaches, because the failure has moved. The modern unfollowable rule is not hidden — it is everywhere: printed, distributed, acknowledged by signature, reviewed annually. The text could not be more public. What is hidden now is the trigger — the condition under which the duty attaches. The citizen holds the rule in his hand and cannot determine, from anything he is permitted to observe or to ask, whether this is a moment it governs. The tyrant posted the law where it could not be read. The modern failure posts it everywhere and makes it impossible to locate — not the text, but the line.

This essay’s claim is that followability is not a virtue of good rules but a condition of rulehood — that a rule which cannot be followed does not bind weakly or harshly or unfairly, but fails to bind, and continues to exist only as something else wearing a rule’s clothes. The argument proceeds by convergence, then by dissection, then by return to the ground this series has been surveying from its first instrument.

Ought and can

The canonical modern formulation of the condition is the Kantian dictum that ought implies can: obligation presupposes possibility. [2] A command to do what cannot be done fails as a command — not because it is cruel, but because the concept of obligation has nothing to attach to. Duty is addressed to agency; where agency cannot reach, there is nothing standing in the place the duty was sent.

Two distinctions guard the dictum from abuse. First, impossibility is not demandingness. A rule may be difficult, costly, even ruinous to follow and bind perfectly well; the sentry’s post is no less a duty for being held in winter. The dictum excuses nothing merely hard. Second, impossibility comes in two kinds — and the second is this essay’s extension of the dictum, not a distinction Kant drew; the fair-notice doctrine below will reinforce it. Physical impossibility is the textbook case: the act cannot be performed. But there is also epistemic impossibility: the act could be performed with ease, if only the agent could know when the duty attaches — and the agent cannot know. The performance is trivial; the ascertainment is impossible.

The modern failure examined here is epistemic. The rules of interest command no flight and no lifting of mountains; they impose duties whose application conditions are unobservable — duties keyed to facts about other minds, other histories, other statuses that the addressee has no lawful or possible way to determine. Such a rule passes every physical test and fails a prior test: at the moment of action, the agent cannot establish whether he is inside the rule or outside it. He can only act, and find out later — from someone else, with authority, after the fact.

A rule like that is not followed by anyone. It is survived.

Three lines, one joint

The dictum would carry less weight if it stood alone. It does not. Three separated instruments, working in different centuries with different materials and different purposes, arrive at the same finding — and in this series, independent arrival is itself evidence. It is the witness test applied to ideas: when lines that never coordinated converge on one joint, the joint is probably in the object and not in the instruments.

The first line is Kant’s, already stated: persons are the beings who act on reasons they can grasp, and a rule they cannot grasp gives them no reason — it merely stands over them.

The second line is Lon Fuller’s. In The Morality of Law (1964), [3] Fuller tells the story of a king — Rex — who tries to make law and fails eight separate ways: by deciding cases ad hoc without rules at all; by keeping his rules secret; by legislating retroactively; by writing rules no one can understand; by writing rules that contradict each other; by demanding conduct beyond the power of the governed; by changing the rules so often that no one can conform; and by administering the rules in ways that do not match their announcement. Fuller’s point is engineering: these eight are not aspirations that good legal systems meet and bad ones miss. They are the conditions under which the word law refers to anything at all. A system can fail them completely while producing texts that look exactly like statutes. Two of the eight do this essay’s work: the rule requiring the impossible, and the incongruence between the rule as announced and the rule as administered.

The third line belongs to a court. American vagueness doctrine holds that a criminal statute is void where — in the formulation of Connally v. General Construction Co. (1926) — “men of common intelligence must necessarily guess at its meaning.” [4] Grayned v. City of Rockford (1972) states the doctrine’s rationale with unusual clarity — a twin rationale that maps this essay’s territory exactly: a vague law fails to give fair notice to those who would obey it, and it licenses arbitrary and discriminatory enforcement by those who administer it. [5] The two harms are one artifact viewed from its two ends. And the doctrine’s cleanest specimen is Coates v. City of Cincinnati (1971), which struck down an ordinance criminalizing assembly conducted “in a manner annoying to persons passing by.” [6] The defect was not that annoyance is trivial. The defect was where the trigger lived: in the minds of passersby. The citizen on the sidewalk cannot read the interiors of strangers; his duty attached or did not attach according to a fact he had no way to observe. The Court’s own terms were vagueness: no standard of conduct specified at all, a standard unascertainable by those it governed. On this essay’s terms, that is a followability holding — the line could not be found by the person expected to stand behind it.

A philosopher, a professor, a court. Agency, legality, notice. One joint: the line must be findable by the person expected to stand behind it.

The anatomy of the unfollowable rule

The modern specimen has a recurring anatomy: a generator, and two mechanisms that seal it.

The unobservable trigger. Write it as algebra first: a duty D is imposed on an addressee a, conditioned on a property P of another person b:

D(a) if P(b) — where P(b) ∉ Obs(a).

Obs(a) is the set of everything the addressee can lawfully see, ask, or verify — and the trigger of his duty lies outside it. The duty lives in one person; its condition lives in another; and the rule supplies no operator connecting them. Coates is the constitutional instance; P was the annoyance of passersby. But the structure is general, and it needs no legislature. Consider a checkpoint guard given two orders: turn away everyone who lacks a permit; and never ask anyone for a permit. Each order is clear. Each is followable alone. Together they impose a duty whose trigger the guard cannot lawfully access — so he must guess who the permitless are, and the only instruments available for the guessing are appearances: bearing, dress, manner, everything a permit is not. The compliance precondition is a classification the rule gives him no means to perform — and, in the usual case, a classification of exactly the kind such rules exist to prevent. He is ordered to sort what he is forbidden to see. The rule does not merely fail to prevent the sorting-by-appearance; it requires it, as the only remaining path to compliance.

The self-sealing solver. In many rules of this anatomy, the variable has exactly one lawful announcer: the bearer of the property. Only the man himself can tell you the fact about him on which your duty turns. And the announcement channel — the question — is what the rule forbids. This is worse than a rule unsolvable in silence; it is a rule that seals its own exit. The single path from ignorance to informed, discriminating compliance runs through the paradigm violation. Ask, and you have breached. Do not ask, and you cannot know whether every next sentence breaches. The rule locks its door from the inside and holds the occupant responsible for being in the room.

The reconstructible prohibition. The third mechanism completes the trap. Suppose the rule seals a single biographical datum — call it D — while leaving every adjacent question not merely permitted but encouraged. The difficulty is that a biography is not a set of separable boxes. A biography happens somewhere: at some school, in some tongue, through some trades, keeping some festivals, counting in some first language. The sealed datum is distributed through nearly every fact about a person that ordinary acquaintance touches, and the sanctioned questions return it as a byproduct — reliably, innocently, in the first minutes of ordinary talk. And now observe the position of the compliant asker. He asked the encouraged question; the answer delivered the sealed datum into his possession; and there is no duty he can now perform, because un-knowing is not an act. The rule manufactured the knowledge through its own approved door — and treats the sentence after as breach. That is not a boundary. It is a tripwire placed behind a door the rule itself props open.

What it produces

A rule with this anatomy cannot produce discriminating compliance, because the boundary between permitted and prohibited cannot be located. The addressee can assure nonviolation only by retreating behind every possible location of the line — and that is avoidance, not obedience. A rule is followed when its addressee can determine what it permits and act on the determination; a rule that can be “complied with” only by abandoning the domain has supplied no rule of conduct. It cannot produce targeted deterrence of the specified conduct, because the addressee cannot locate the condition that distinguishes the permitted act from the prohibited one. What it produces is a third thing, and the third thing has a structure worth stating slowly.

It produces universal exposure. Since the boundary cannot be found, and since ordinary human exchange reconstitutes the sealed datum continuously, the rule distributes latent liability across everyone it governs. Some violate. Some accidentally comply. Some retreat from the domain entirely. Those who remain cannot determine, from the rule, whether they are violating or complying — and those unwilling to bear that uncertainty have no option but the retreat. Exposure ceases to be an event and becomes a condition.

And ambient exposure changes what enforcement is. The information the addressee needs before acting may be available to the enforcement system independently — through a record, a complaint, a later disclosure. The asymmetry is epistemic, not necessarily temporal: the fact that could have guided him is available elsewhere, but not to him when he must act. That does not make selective enforcement inevitable; it makes the selection unanswerable from below. The announced rule speaks to everyone; the administered rule reaches whoever is selected; and for those who remain within the governed domain, nothing the rule supplies determines whether they will later stand as compliant or in breach.

Name the object honestly, then. A rule whose boundary its addressee cannot locate, and whose liability he cannot clear through informed conformity, is not a norm. It is a stored instrument — latent liability distributed across a population in advance, not clearable through informed conformity within the governed domain, dischargeable at discretion. Whether anyone built it for that purpose is immaterial; the capacity exists once the rule does, and it needs no villain. Its value to any eventual holder lies precisely in its unfollowability — not because followable rules leave nothing to enforce; they are violated too, and enforcers select among those violations too. But a followable rule gives the governed a route to clear the liability by conforming their conduct, and the magazine empties as they take it. The unfollowable rule withholds the route. Its magazine stays loaded.

Vagueness doctrine saw both ends of this artifact and named them half a century ago: no fair notice at one end, arbitrary enforcement at the other. What the doctrine states as twin harms, this anatomy shows to be one mechanism. And the doctrine’s remaining insight completes the account: people governed by an unlocatable line do not stand near where they guess it might be. They retreat — far behind it, into the only zone that is certainly safe, which is silence. The chilling is not a side effect of the vague rule, and it is not compliance — the distinction between obedience and avoidance holds here. It is the only reliable nonviolation strategy the rule leaves available, which means the silence such a rule produces can never be claimed as obedience — least of all by the system that produced it.

The seeing and the saying

There is a deeper reason the anatomy keeps recurring, and it lies in what such rules attempt to govern.

Discriminare: to discern, to separate, to tell apart. Before the word acquired its modern weight, it named the faculty of judgment itself — and judgment has no purchase in a homogeneous field. If every wall is red, there is nothing to discern until a blue wall appears; the discernment is not an act performed upon the walls but the condition of seeing walls at all. Perception of difference is involuntary. A man discerns his own youth from his middle age without consenting to the comparison; he cannot decline to notice what noticing is. The seeing cannot be legislated, because it is not conduct. Only the saying is conduct — the treatment, the utterance, the act — and conduct can be specified, observed, and judged.

This yields the diagnostic that separates a followable rule from its counterfeit, and it can be run by anyone in one step: universalize the rule and see what survives. A duty stated on conduct, applied to everyone — ask no one about X; treat every person by standard Y — is coherent, followable, and enforceable without a single classification. It may be wise or unwise, but it is a rule. Now examine the version that fails: the same duty, made conditional on an unobservable property of the other person. If deleting the conditional clause repairs the rule — if the universal form is followable and the conditional form is not — then the clause was the defect, and the defect is a classification requirement. The conditional rule pretends to govern the seeing while enforcing on the saying, and a rule with that shape manufactures universal exposure by structure: it demands a sorting it forbids the means to perform, of a difference no one can decline to perceive.

In the class of rules examined here, the legitimate object of governance is conduct, not involuntary perception. Where conduct legitimately turns on a status — age, license, office — the status is one the addressee can see, ask, or verify by a route the rule leaves open. The moment a rule’s application requires knowing what cannot be known, it has left the domain of conduct for the domain of perception — and in that domain there are no rules, only liabilities.

A rule is a species of address

Which returns the finding to the ground this series has been surveying since its first instrument.

A rule is not merely a text with an imperative mood. It is an address: it speaks to an agent, and it presupposes that the agent can answer. Compliance is the answer. That is what distinguishes a rule addressed to persons from a specification imposed on material — the pipe schedule does not presuppose that the pipe can grasp it, because the pipe is not being asked anything. A person is. And persons, as the first line of this essay’s convergence has held since 1785, are the beings who act on reasons they can grasp — always ends, never merely means. A rule an agent cannot grasp gives that agent no reason. It does not address him. It occupies him.

So the unfollowable rule is not a defective communication. It is a change of relation, performed under the grammar of communication. Its addressees are not treated as participants whose compliance is being requested and whose reasons are being engaged; they are treated as terrain — a surface across which liability is positioned in advance, to be activated later, from elsewhere, by choices they cannot see. That is the precise structure the second of our instruments named at the level of the room: the demotion of a participant to an object. The unfollowable rule is the standing violation in imperative form: the addressee cannot say, from the rule, where the boundary lies at the moment conduct is required; the line becomes legible, if at all, only from an external position unavailable to him when he must act.

And this extends the design test stated at the close of Standing. If a speak-up system is coherent only where the person required to identify deviations may raise the possibility that the governing standard is itself the deviation, then the paradigm legitimate challenge is now identifiable: it is the followability challenge — the claim that the standard, as written, supplies no route to informed conformity within the domain it governs. That challenge is prior to every other, because its answer determines whether the standard’s commands are commands at all. A system that closes it has not merely silenced a participant. It has closed the one question on which its own authority to require anything depends.

What followability asks

The remedy, as at the close of our first instrument, is modest — almost embarrassingly so.

State duties on conduct, not on the unobservable properties of others. Make every trigger something the addressee can see, or ask, or verify by a route the rule itself leaves open. Put the line where the person expected to stand behind it can find it. And when the followability challenge is raised — this, as written, supplies no route to informed conformity — answer it as what it is: the one challenge a rule-making authority can never be entitled to leave unanswered or to extinguish by closure, because it is the audit of the authority’s own product. A rule that can survive the question needs no shelter from it. A rule that cannot has already told you what kind of instrument it has become.

None of this requires bad intent, and the argument is stronger without it. A rule with this anatomy produces universal exposure, a standing selection problem, retreat into silence, and a stored instrument — by itself, mechanically, with every administrator acting in good faith and nobody anywhere intending the magazine. The anatomy does the work. That is precisely why the anatomy, and not the intent, is the thing to inspect.

There is a register far older than jurisprudence in which the requirement is stated not as doctrine but as the law’s own boast:

“For this commandment that I am commanding you ⌊today⌋ is ⌊not too wonderful for you⌋, and it is not too far from you. It is not in the heavens ⌊so that you might say⌋, ‘Who will go up for us to the heavens and get it for us and cause us to hear it, so that we may do it?’ And it is ⌊not beyond the sea⌋, ⌊so that you might say⌋, ‘Who will cross for us to the other side of the sea and take it for us and cause us to hear it, so that we may do it?’ But the word is very near you, even in your mouth and in your heart, ⌊so that you may do it⌋.” [7]

Every clause is an ascertainability claim. Not in the heavens, not beyond the sea — the law itself disclaims the two places an unfindable line lives: above the addressee’s reach, and on the far side of what he can cross. And the closing purpose clause states this essay’s whole thesis in seven words: near, so that you may do it. An ancient self-description of law makes the same claim.

The word is near, or it is not a word.

A command that cannot be found is not waiting to be obeyed. It is waiting to be used.

Notes

1. Suetonius, Life of Gaius (Caligula) 41; paraphrased, not quoted.
2. “Ought implies can” is the standard scholarly distillation of Kant’s position, not a quoted sentence; the persons-as-ends formulation is Immanuel Kant, Groundwork of the Metaphysics of Morals (1785).
3. Lon L. Fuller, The Morality of Law (New Haven: Yale University Press, 1964), ch. 2; the eight failures follow Fuller’s enumeration and order.
4. Connally v. General Construction Co., 269 U.S. 385, 391 (1926).
5. Grayned v. City of Rockford, 408 U.S. 104, 108–109 (1972).
6. Coates v. City of Cincinnati, 402 U.S. 611 (1971); the ordinance language as quoted by the Court, Cincinnati Code of Ordinances § 901-L6.
7. W. Hall Harris III et al., eds., The Lexham English Bible (Bellingham, WA: Lexham Press, 2012), Dt 30:11–14.

Keywords: followability; ought implies can; internal morality of law; vagueness; fair notice; epistemic impossibility; unobservable trigger; selective enforcement; discretionary enforcement; rule of law; address; standing; compliance architecture

Our Gods Research Program

The instruments

• O’Connor, T. The Problem of Absolutes. Zenodo, 2026. https://doi.org/10.5281/zenodo.21866538
• O’Connor, T. Standing. Zenodo, 2026. https://doi.org/10.5281/zenodo.21883564
• O’Connor, T. Addressability — The seat is conferred, never seized. Zenodo, 2026. https://doi.org/10.5281/zenodo.21926668
• O’Connor, T. Followable. Zenodo, 2026. https://doi.org/10.5281/zenodo.21971832
• O’Connor, T. The Extraction Invariant. Zenodo, 2026. https://doi.org/10.5281/zenodo.22806212

Program spine

• O’Connor, T. (2025). An Epi-Phenomenological Series in Thrice – A Quantum Junkies’ Riddle – Our Garden. Zenodo. https://doi.org/10.5281/zenodo.17926796
• O’Connor, T. (2026). The Bare, Naked Lie: The Architecture of Recursion. Zenodo. https://doi.org/10.5281/zenodo.18287548
• O’Connor, T. (2026). A Trick Question, The Cascade, and a Cup Worthy of Exile — A Script’ Junkies’ Riddle — Our Destiny (M5). Zenodo. https://doi.org/10.5281/zenodo.18452359
• O’Connor, T. (2026). The Veil: A Structural Autopsy of Epiphany — A Revelation Junkies’ Riddle — Our Apokalypsis. Zenodo. https://doi.org/10.5281/zenodo.19040340
• O’Connor, T. (2026). Quadratic Failure Scaling in Pairwise-Coupled Systems. Zenodo. https://doi.org/10.5281/zenodo.19600761

Copyright

Copyright 2026 Tony O’Connor—All Rights Reserved

License

Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0). Share and cite freely with attribution. No commercial use. No derivatives. Full licence text: creativecommons.org/licenses/by-nc-nd/4.0

Citation

O’Connor, T. (2026). The Problem of Absolutes. Our Gods Research Program. Zenodo. https://doi.org/10.5281/zenodo.21866538

Contact

Permissions beyond the scope of this license are handled through the Our Gods Research Program.