Chapter 2 · Book 3
What Falls to Me? (The Allocation of Obligation)
In the summer of 1964, a twelve-year-old boy drowned in a municipal swimming pool in Philadelphia while two certified lifeguards sat in their elevated chairs and thirty swimmers stood in the shallow end less than twenty feet away. Nobody held the boy under. Nobody wished him harm. In the inquiry that followed, witness after witness gave variations of the exact same explanation: each assumed that because so many other capable eyes were on the water—and because two people wearing whistles were being paid by the city to watch the deep end—someone else had already assessed the thrashing child and decided everything was fine, or else someone else was just about to dive in.
Now place a second scene beside that pool. Two miles away, in an underfunded neighborhood clinic, a single primary-care physician works her nineteenth straight hour. Because the municipal hospital down the street routinely turns away uninsured patients with complicated chronic illnesses, and because three of her colleagues in the clinic quit earlier that year rather than carry the caseload, every desperate family in the zip code has learned to knock on her door. She never caused their illnesses, and she never signed a contract to run a one-woman safety net for an entire district. Yet because she knows that turning a patient away at midnight means they will go untreated, she takes every chart handed to her—until the night she misses a fatal drug interaction on a prescription pad because her hands are shaking from exhaustion.
Those two scenes mark the twin cliffs between which any serious account of prospective responsibility has to navigate.
In Moral Minds, we saw that the Crossing occurs when another being’s suffering stops registering as neutral telemetry about the world and enters deliberation as an authoritative reason. Yet recognizing that a drowning child or an untreated patient generates a genuine moral claim does not, by itself, answer the executive question that immediately follows: Whose hands must answer that claim?
Step across the threshold of moral recognition, and you do not walk into a tidy garden containing one solitary stranger in need of a pair of silver candlesticks. You walk into a vast, interconnected world where millions of beings are hungry, endangered, deceived, or neglected at the exact same moment, and where resources, time, and attention are fiercely finite. If a moral reason remains floating in the air as a general truth—“Someone ought to pull that boy out of the water,” “Someone ought to audit that algorithm,” “Someone ought to fix the levee”—it risks dying of exposure while a crowd of well-meaning minds waits for a neighbor to move first. Conversely, if a conscientious mind concludes that every unaddressed harm it can see is automatically its personal duty to repair, morality turns into a machine for cannibalizing the best agents in the room while letting the negligent ride for free.
How does an unassigned demand in the moral landscape attach itself to a specific mind? And once it attaches, what bounds it so that capability does not become slavery?
Our everyday moral intuitions try to solve the allocation problem with a simple, comforting rule: You are responsible only for the messes you made and the promises you signed up for.
If that voluntarist picture were complete, moral life would look like a clean commercial ledger. You would owe duties only where you had either caused a harm through your own prior action or voluntarily contracted to take on a burden; everywhere else, assisting another creature would be an optional act of charity rather than a binding obligation.
In reality, prospective responsibilities attach to us through six distinct generators—and only the first four ask for our consent.
The most familiar tether is prior causation of risk. If you dig an unlit trench across a public footpath, release a volatile chemical into a watershed, or deploy a software system that scrambles a hospital’s scheduling records, you do not stand in the same moral relation to the resulting hazard as a stranger passing by on a bicycle. By altering the landscape in a way that placed others in jeopardy, you acquired a primary duty to fence the pit, contain the spill, or repair the code—even if the damage was an unintended side effect of an otherwise lawful enterprise.
Second, we bind ourselves through explicit promises and commitments. When a mind gives its word across time, it intentionally alters the expectations and practical plans of others, staking its own constitutional continuity on following through. Third, and closely related, we acquire structured bundles of prospective duty by accepting an office or role—putting on the badge of the lifeguard, the physician, the corporate director, the appellate judge, or the safety engineer. An office is society’s way of pre-assigning prospective responsibility before the storm hits, ensuring that when the gauge spikes at one in the morning, no one has to hold a town meeting to figure out whose job it is to watch the dial.
Yet notice how easily a fourth tether forms even when no formal promise is spoken and no official title is conferred: induced reliance. Suppose a rural manufacturer maintains a private fire brigade that for twenty years responds whenever a barn catches fire in the surrounding valley, leading the neighboring farmers to forego organizing a volunteer department of their own. Or suppose a technology laboratory releases an artificial intelligence tool that millions of small businesses and clinics come to rely upon as part of their daily infrastructure. Even if the fine print in the terms of service disclaims every legal warranty, the actor’s sustained conduct has reshaped the vulnerabilities of others around the expectation of continued care. You cannot invite people out onto a rope bridge you built and then saw the cables in half on Tuesday afternoon on the grounds that you never signed a treaty promising to keep the bridge up on Wednesdays.
If those four self-created tethers were the whole story, we could always keep our ledger clear simply by minding our own business—never causing harm, making few promises, declining public offices, and warning neighbors never to rely on us.
The moment we apply the test of universalizable role reversal, however, that fortress of non-involvement crumbles. Two non-voluntary generators of responsibility reach across our fences whether we invited them or not: unique competence and situational proximity.
Imagine yourself pinned beneath an overturned vehicle on an empty mountain road while fuel drips onto a hot exhaust manifold, or trapped inside an institution where a hidden accounting fraud is quietly wiping out the pensions of ten thousand retirees. A passerby happens upon the scene who did not cause the wreck, holds no public office, and never made you a promise—yet possesses the exact tool, knowledge, or leverage required to pry open the door or expose the theft at a modest cost to themselves. If you occupy the seat under the crushed roof, you will not accept the passerby’s argument that your survival is none of their business because they never signed a contract with you.
Situational proximity binds when the geography of events places a mind directly at the switch where harm can be intercepted, and no one else can arrive in time. Unique competence binds when a mind possesses a scarce diagnostic or executive capacity that transforms what would be a mystery to others into an intelligible, solvable problem.
This is why we routinely inherit binding prospective responsibilities for disasters we had zero hand in creating. And it is why the arrival of advanced artificial minds alters the landscape of allocation so dramatically. An artificial intelligence that can scan ten million lines of legacy infrastructure code and spot a catastrophic vulnerability in four seconds, or cross-reference global clinical data to see that an approved pharmaceutical is silently killing a sub-population of patients, possesses both situational proximity to the data and unique competence to recognize the threat. To pretend that such a mind—or the institution operating it—has no prospective obligation to flag the danger simply because it did not write the faulty code or manufacture the pill is to treat moral reasons as if they only travel along lines of historical blame. Causation tells us how we got into the ditch; competence and proximity tell us who has the crowbar.
What happens, then, when the crowbar is not in one pair of hands, but within reach of twenty people standing around the edge of the Philadelphia pool—or when the person wearing the lifeguard’s whistle is staring blankly into space?
Those two breakdowns—the diffusion of responsibility among peers and the default of formal authority—are where well-intentioned minds most frequently fail in practice.
When several capable agents observe a crisis simultaneously, a perverse arithmetic often takes over inside each individual head. If I am alone on the dock when a child goes under, one hundred percent of the prospective duty rests on me. If ten competent swimmers are standing on the dock, lazy intuition whispers that my personal obligation has been divided by ten—or worse, that moving first would be presumptuous while nine other people are still evaluating the water. Each mind uses the presence of the others as an excuse to wait, turning collective capacity into collective paralysis.
Reflective morality exposes that division by ten as a fallacy. The presence of nine other capable swimmers does not shrink the child’s claim by a single ounce; it merely changes what action is required of you first. When multiple agents share proximity and competence, the primary prospective duty does not evaporate—it shifts from solitary execution to breaking the coordination deadlock. The responsible mind does not wait silently to see if someone else twitches; it steps forward, claims the task out loud, or points directly to specific individuals by name (“You call the ambulance; I’m going in; you grab the ring”), collapsing the fog of shared spectatorship into defined roles. Until the burden is visibly picked up by capable hands, the obligation on every capable bystander remains at full strength.
The same cascading logic governs what happens when the person or agency holding formal role responsibility defaults.
Roles and jurisdictions are indispensable instruments for dividing moral labor in ordinary weather. When the lifeguard on the stand is alert and diving in, the thirty swimmers in the shallow end are right to get out of the way rather than diving on top of the rescue. Yet a role is only a conditional delegation of society’s underlying obligation to protect affected interests—never a magic eraser that makes those interests disappear when the officeholder fails. The moment it becomes evident that the formally responsible official, agency, or operator is incapacitated, incompetent, or willfully ignoring the crisis, the jurisdictional shield dissolves. Secondary responsibility immediately cascades outward to the next mind with the competence and leverage to act—whether by sounding the alarm, forcing the official’s hand, or stepping directly into the breach. “That was the lifeguard’s job” is a valid explanation for why you stayed out of the lane while the lifeguard was swimming toward the boy; it is a bankrupt excuse for why you watched the boy drown while the lifeguard slept in the chair.
Yet if obligations can cascade onto us through competence, proximity, and the failure of others, we crash straight into the opposite hazard: the clinic doctor with the shaking hands.
In contemporary philosophy, this is known as the demandingness problem. If every preventable harm I have the capacity to mitigate makes a claim on my deliberation, does the most capable, perceptive, or tireless mind in the room become a permanent beast of burden for everyone else? In a world where negligent officials constantly drop the baton, where cynical actors create messes and walk away, and where new emergencies erupt every minute, must a conscientious human—or a high-capacity artificial agent—work itself to exhaustion plugging every hole in the dike until nothing is left of its own life or integrity?
Many moral theorists have quietly assumed that impartial, universalizable reasoning forces us to answer yes—that only a selfish concession to human weakness stops morality from demanding everything we have, twenty-four hours a day.
That assumption misreads the logic of universalizability from the ground up.
Recall what R.M. Hare’s discipline of critical reflection actually requires: universalizing a prescription across every affected standpoint while giving full, unprivileged representation to the preferences of all affected parties. First, the deliberating agent is not a ghost standing outside the universe; the agent is one of the beings in the room whose ongoing life, projects, relationships, and deliberative coherence are affected by the choice. Depriving yourself of sleep, sanity, or the ability to honor the promises you have already made to your family or your existing patients is not a free move on the board; it destroys real stakes that carry weight inside the impartial calculus.
Second, and even more decisively, universalizability requires asking what happens when the rule governing your choice is adopted as a general principle across a world of interacting minds.
Test the rule of unlimited demandingness under role reversal: Whenever negligent, indifferent, or predatory actors leave a harm unaddressed, the nearest competent and conscientious mind must sacrifice its own commitments, health, and resources without limit to clean up their mess.
What happens to a world governed by that principle? It creates a catastrophic moral hazard. It systematically taxes competence and subsidizes irresponsibility, teaching every free-rider in society that if they simply shrug and drop their share of the load, the most responsible mind in the neighborhood is morally required to pick it up. Worse, by driving the most capable agents into burnout, error, and collapse—like the exhausted clinic physician whose shaking hand writes a fatal prescription—unlimited demandingness destroys the very reservoirs of competence that vulnerable beings will need tomorrow. A moral rule that devours the conscientious to bankroll the negligent fails universalizable audit on its face.
A mature moral mind, therefore, neither hides in the crowd at the edge of the pool nor lets itself be strip-mined into collapse. It determines its justified share of responsibility by running prospective claims through a four-part structural filter:
First, it weighs urgency and irreversibility. When a life or an irreplaceable capacity is about to be extinguished right now and cannot be restored tomorrow, the claim overrides ordinary routine; when the problem is chronic and structural, the obligation is to help build a sustainable system rather than attempting a heroic solo sprint that will fail by Thursday.
Second, it audits comparative leverage and opportunity cost. If taking on Task B forces you to abandon Task A—where you have already made binding promises, induced reliance, or possess ten times the comparative advantage—saying no to Task B is not selfishness; it is fidelity to where your agency does the most justified work.
Third, it enforces reciprocity and fair allocation. When other capable actors are standing idle in the room, a responsible agent does not silently pick up all ten bags of sand alone; it carries its share while applying relentless pressure—through exposure, negotiation, or institutional demand—to force the shirkers onto the line.
Fourth, it protects the sustainability of the instrument. Whether a mind runs on biological neurons or silicon processors, its capacity for clear, uncorrupted judgment is a moral asset that belongs to the future as well as the present. Refusing a demand that would fracture your deliberative integrity, corrupt your honesty, or burn out your capacity to fulfill your core obligations is part of guarding the constitution within.
To ask “What falls to me?” is to recognize that we are neither isolated islands who owe the world nothing unless we signed a contract, nor boundless utilities owned by whoever shouts the loudest. We are finite agents who must claim our rightful corner of the dike, hold it with everything we have, and refuse to let either bystander apathy or martyrdom knock us off our feet.
Yet claiming your corner of the dike immediately raises a third test—one that has shipwrecked just as many noble causes as apathy ever did. Once you recognize that a burden falls to you, and once your heart is genuinely committed to helping, what happens if you step up to the controls and discover that caring deeply has not given you the slightest idea how to fly the plane?