Omissions and Inaction

The Philosophical Problem

An emergency physician does not administer an indicated treatment. A swimmer sees a person in distress and chooses to remain on shore. A guardian forgets to secure medication. A public body repeatedly defers action on a known safety hazard. In each case the important description concerns something that did not occur. Yet the philosophical significance of that absence differs. A deliberate refusal can express a considered policy; forgetfulness can reveal a defect in practical attention; an institutional delay can result from distributed procedures; and the failure of a powerless observer to intervene can be a mere nonoccurrence. The central task is to identify the relations through which certain failures to act become attributable to agents.

The ordinary contrast between acting and doing nothing conceals several questions. What establishes the identity of an omission? Can a negative fact or absence be an action, an event, or a cause? Does intentionally refraining require a positive act of deciding? When does a duty make nonperformance wrongful? Can responsibility attach to negligent failures that the agent never noticed? Why does the law sometimes punish omissions and sometimes reserve liability for positive conduct? Answers to these questions depend on different kinds of explanation. The conditions for describing an omission need not be the conditions for blaming someone for it.

The basic philosophical insight of this inquiry is that nonperformance is always described relative to an available or contemplated performance. To say that a person omitted to close a gate identifies an action type, an agent, and a relevant interval. It also ordinarily presupposes a practical background in which closing the gate was expected, prescribed, undertaken, or otherwise salient. Such a background can be institutional, moral, instrumental, conventional, or conversational. It need not supply a genuine moral duty. A person may omit to answer a question in an interview without wrongdoing, and an artist may intentionally omit a verse from a composition as an exercise of craft.

The following analysis therefore refuses to infer moral blame from the grammar of omission. It also refuses to reduce all omissions to bodily actions that happen to accompany them. An agent who refrains from revealing a confidence may remain entirely still. Another can fail while asleep to undertake a task scheduled for the morning. Contemporary action theory has acknowledged this difficulty explicitly: the Stanford Encyclopedia of Philosophy's account of action treats omissions and refrainings as a distinct test for causal and agency-centered theories (Piñeros Glasscock & Tenenbaum, 2023, §7). The developed proposal will distinguish the ontology of the absent performance from the powers and normative relations that make it an omission attributable to someone.

Definition, Scope, and Basic Distinctions

An omission is a specified nonperformance of an action by a candidate agent at an occasion where that action is practically relevant. This working definition contains four elements: an identifiable performance type, a candidate agent, an occasion or interval, and a standard of relevance. Relevance may be supplied by a duty, plan, undertaking, task, request, institutional role, habitual expectation, or contrast under examination. The definition leaves open whether the candidate had the ability to perform the action, whether the omission was intentional, and whether it was wrong. These are further questions. A person can be said to have omitted attendance at a meeting while trapped in an elevator; that circumstance may excuse the absence altogether.

Inaction is a wider description for an interval or pattern lacking a kind of intervention. It can include rest, waiting, monitoring, paralysis, delay, withdrawal, deliberate restraint, and a sequence of unfulfilled duties. An omission concerns a specific unperformed action; inaction often concerns the organization of conduct over time. Refraining is deliberate nonperformance: an agent recognizes a live option to act and governs conduct so that the action is not performed. Abstaining ordinarily indicates adherence to a practical policy, rule, or resolution. Letting something happen designates a relationship between an agent's nonintervention and an ongoing or impending process; that relation cannot be equated simply with bodily stillness.

An action and an omission can coexist. A surgeon who performs one procedure may omit an essential step. A speaker who presents a careful report may omit a relevant qualification. The positive action and the omission are distinct under descriptions even when one continuous behavior supplies their temporal setting. Sometimes an agent avoids doing something by acting: withdrawing an application can be a way of refraining from accepting a position. Sometimes no positive action occupies the relevant interval. Accordingly, the identity thesis that every omission just is some action done instead is too strong (Clarke, 2014; Alvarez, 2013).

A refusal is a communicative action that commonly announces or effects nonperformance. A failure can mean an unsuccessful attempt, an omission, or an adverse outcome of successful performance. Nonoccurrence is the broadest category: innumerable things fail to happen at every instant, with no particular agent implicated. The expression negative agency should be reserved for forms of agency exercised or expressed through not doing something. It describes deliberate restraint and policy-governed nonperformance most clearly; its extension to unintentional omissions requires an account of responsibility and practical capacities.

A structured inquiry must distinguish at least six assessments: what action was absent; whether the candidate could perform it; whether the nonperformance was guided; whether it changed an outcome; whether acting was required; and whether the agent is answerable. These assessments can diverge in individual cases. Establishing their independence creates the conceptual foundation for the historical and argumentative analysis.

Ancient Foundations: Voluntariness, Choice, and Practical Power

In the Nicomachean Ethics, Aristotle relates ethical appraisal to voluntary and involuntary conduct. Book III, chapter 1 begins from the fact that praise and blame concern what is voluntary, and distinguishes constraint from ignorance of the relevant particulars (Aristotle, Nicomachean Ethics III.1, 1109b30–1111a). The discussion does not set out a contemporary metaphysics of omissions. It supplies the more fundamental insight that responsibility requires an inquiry into how conduct belongs to the agent and what the agent understands about the circumstances.

Aristotle's examples of mixed actions illustrate why an outward behavioral description fails to determine voluntariness. A person may throw cargo overboard during a storm to preserve a ship and its passengers; the act is chosen in circumstances that make its ordinary desirability irrelevant. The same structural question applies to refraining. A threatened person may fail to render assistance because intervention would expose a family member to grave harm. Describing that situation as an omission leaves the normative and voluntariness questions open until the circumstances are understood. The relevant contrasts concern practical alternatives at the time rather than an abstract inventory of physically possible motions.

Book III, chapter 5 adds a decisive element: culpable ignorance may arise where a person should and could have learned the facts, and negligent ignorance may therefore bear on responsibility (1113b30–1114a). Aristotle is not advancing a modern capacitarian account of negligence. Nevertheless, his treatment establishes an enduring philosophical pressure against a simple rule that only currently conscious wrongdoing is blameworthy. Habits, the history of choice, and opportunities for practical knowledge matter.

Aristotle's accounts of choice (prohairesis), deliberation (bouleusis), and what is within our power also clarify the relation between choosing not to do something and merely lacking action. A negative decision can be a deliberative settlement. One can resolve not to intervene because intervention would violate justice, or decline an opportunity because another obligation takes priority. The subsequent absence of performance is intelligible through reasons, goals, and character. This does not entail that every omission is a choice; it shows why intentional nonperformance can be investigated by the conceptual instruments ordinarily applied to action.

The Greek material also teaches a methodological lesson. A text about voluntariness and virtue may illuminate omissions without containing a fully formulated theory of absent events. Historical reconstruction should identify that genuine conceptual contribution without retroactively assigning contemporary terminology to the author.

Medieval Analysis: Aquinas on the Sin of Omission

Thomas Aquinas gives one of the clearest historical formulations of the distinction between an omission itself and the act or disposition from which it may arise. In Summa Theologiae I–II, question 71, article 5, he asks whether every sin includes an act. The objection invokes failures to do what is prescribed. Aquinas distinguishes the constitution of a sin of omission from its causes or occasions: an omission may sometimes involve an internal act of willing not to comply; it may also occur without a contemporaneous interior or exterior act, when an agent fails at the relevant time even to think of the obligation (Aquinas, Summa Theologiae I–II, q.71, a.5).

This distinction has enduring importance. The wrong, when there is a genuine duty, can consist in not doing the required thing, whereas its psychological or historical explanation may involve earlier decisions, distraction, a competing desire, or culpable patterns of neglect. Aquinas thereby avoids equating the essence of the omission with a positive action occupying the same time. His account also shows that an agent may be responsible for a failure without continuously entertaining the proposition that an obligation exists.

Aquinas's treatment remains situated in a moral-theological conception of obligation and voluntary wrongdoing. Its importance for a general philosophy of omissions is conceptual: the description of the prohibited absence and the explanation of how it came about are different explanatory objects. An analogous distinction appears in modern legal analysis, where the duty's content defines what was omitted while questions of capacity, knowledge, and causation determine whether punishment or liability is justified.

A concern arises immediately. If the omission itself includes no positive act, what makes it belong to this agent? A duty can identify whose failure is normatively significant, but duties can also be unreasonable or impossible. Aquinas's wider discussion of voluntariness and negligence places the omission within a field of possible compliance and moral understanding. Contemporary theory must generalize this point beyond theology: attribution requires specification of the agent's practical relation to the missing performance.

Medieval omission theory also cautions against treating a moral rule as a metaphysical fact about negative events. A rule can individuate an obligation and thus pick out a failure. It cannot by itself demonstrate that a corresponding absence is an event with causal powers. The moral question and the ontological question remain distinct even when they arise from a single judgment.

Early Modern Obligations and the Expansion of Duties to Assist

Modern reflection on omission increasingly joins the philosophy of agency to questions about duties, rights, and the reach of moral concern. The tradition of prohibitions on injury raises one kind of requirement: an agent must refrain from a harmful interference. Requirements to rescue, aid, warn, care, or fulfill an undertaking raise another: an agent must bring about a performance where failure leaves an existing danger untreated. Both are practical constraints, but their objects are differently structured. Compliance with a prohibition may consist in continued nonperformance; compliance with a positive duty requires an intervention.

John Stuart Mill's On Liberty (1859), chapter V, makes the relation between action and failure relevant to social accountability. Mill's broader discussion associates accountability with conduct that concerns the interests of others and with principled limits on coercion. The question of when omission warrants sanction therefore cannot be settled by noting that someone could have helped. A theory must also account for what society may rightfully demand and when the enforcement of duties encroaches on liberty.

In Utilitarianism (1861 in periodical form; 1863 in book form), chapter V, Mill treats justice in relation to rights, deserved treatment, and the withholding of what is owed. A wrongful withholding is often a negative description but can constitute a serious injustice. The relevant wrong is not a physical intrusion. It is the failure to supply or respect a good to which another person has a claim. The language of omission thus makes visible a class of injuries mediated by rights and institutional expectations.

The late modern debate cannot be described as a simple transition from negative to positive obligations. Philosophical traditions have long recognized duties of care, rescue, fidelity, and beneficence, while disputing their scope. What becomes analytically decisive is whether a duty specifies an act sufficiently clearly, whether the burden is justified, and how the duty interacts with other claims. If a theory assigns every capable person an unlimited obligation to prevent all avoidable suffering, it faces a demandingness problem; if it recognizes only prohibitions on intervention, it has difficulty explaining special relationships and institutional responsibilities.

This historical development sets the stage for contemporary disputes over nonintervention. A theory of omission must explain how moral norms pick out a subset of possible performances without assuming that moral salience constitutes physical causation.

The Turn to Analytic Action Theory

G. E. M. Anscombe's Intention (1957) and Donald Davidson's “Actions, Reasons, and Causes” (1963) provide different but related resources for modern analysis. Anscombe places the explanation of intentional action in relation to the agent's practical understanding of what is being done and why. Davidson defends the view that a reason explaining an action can also be a cause of it. In its standard form, the causal account concerns a belief-desire structure generating bodily behavior under an intentional description (Anscombe, 1957/2000; Davidson, 1963).

Negative conduct puts pressure on both approaches. If an agent intentionally declines an invitation, what is the intentional object: a decision, a bodily stillness, a continuing activity, or the nonattendance itself? An Anscombean perspective can understand the agent's reasons for not attending without first selecting a particular movement as the omission. A Davidsonian perspective can identify beliefs and pro-attitudes causing behavior that prevents attendance or sustains nonparticipation, but straightforward causal analysis becomes difficult when no action event is produced.

The crucial contrast is between the explanation of why an action happened and the explanation of why an anticipated performance did not happen. In ordinary intentional action, movement or other performance is available as an event to be causally explained. In intentional omission, the explanatory object may be a relation between intention and absence across an interval. One response is to identify a positive action performed instead. Another is to treat an omission as a negative event or fact. A third is to accept that exercises of agency need not always be actions as event ontology understands them.

These alternatives illuminate the broader theory of agency. An agent's power to perform a task can be exercised in withholding performance. Maria Alvarez (2013) develops agency in relation to two-way powers, while Helen Steward's work on agency and active powers further examines the capacity to do or not do. Such approaches resist defining agency wholly through the production of bodily events. They also invite a new question: can a power be exercised in omission when the agent never considered performing the action?

That question requires distinguishing deliberate refraining from unintentional failure. The two-way-power conception readily explains a conscious decision not to intervene. It requires further argument to extend to forgetting, inattentiveness, or unnoticed neglect. Capacity is necessary for responsible omission in many cases, but the bare possession of capacity is not yet its exercise.

Kaisa Kärki's account of intentional omission develops this problem through procedural metacognition: the agent represents an available action and evaluates it as an action not to perform (Kärki, 2025). The account draws upon work on cognitive control to argue that withholding can involve evaluative monitoring without requiring a fully verbalized proposition. This analysis strengthens the two-way-power perspective by proposing a specific mechanism for negative guidance. Yet the requirement that the omitted action be represented poses an additional boundary question for routine adherence to standing policies. Some restraint is skillful and habitual; its relation to present representation should not simply be assumed. Kärki's distinction between intentionally guiding nonperformance and merely having an intention to abstain provides a concrete test for accounts of refraining.

What Makes a Nonperformance an Omission?

The world contains indefinitely many true negative descriptions. A person reading at a desk is not running a marathon, repairing a bridge, or boarding a spacecraft. These absences ordinarily lack philosophical significance. Selection is indispensable. Randolph Clarke's Omissions: Agency, Metaphysics, and Responsibility (2014) systematically investigates omission and refraining while examining the consequences of treating nonperformance as an object of agency and causal attribution. The need for selection lies at the center of that investigation.

A basic selection test asks whether a performance was reasonably available and practically salient. A contract can make submission of a report salient; a promise can establish a specific obligation; a request can create a reason without creating a duty; danger may activate claims of aid. An audience's expectation may pick out an omission even when the agent refuses on legitimate grounds. The same absent performance can consequently be described by several normative vocabularies. What matters is to state which framework is being invoked.

A second test concerns timing. Missing a train at noon is different from deciding at breakfast not to travel. One is an outcome of failing to arrive; the other may be an intentional choice with consequences throughout the morning. To identify the omission, the investigator must specify when performance would count as timely and what the success conditions were. A late attempt may be evidence of effort while still constituting failure to satisfy an obligation that required timely performance.

A third test concerns agency. A person who cannot reach a switch may be described as not pressing it, but cannot ordinarily be blamed for omitting an impossible intervention. Social structures complicate capacity: the same person may possess a general ability and lack an effective opportunity at the particular moment. Institutional authority, access to resources, physical constraints, competing duties, and information determine the practical alternative. The relevant ability is situated, not an abstract possibility considered without the circumstances.

Normative expectations do much of the work in ordinary omission descriptions. Yet it would be a mistake to equate the ontology of omission with blame. A musician may intentionally omit a phrase as an artistic decision, and a witness may permissibly decline to answer a question outside the scope of testimony. The structure of the omission is intelligible while its normative valence remains undecided. The best initial theory therefore treats omissions as contrastively selected nonperformances, with evaluation supplied by further analysis.

Omissions, Absences, Facts, and Events

A theory that identifies actions with events faces a direct ontological challenge: what entity is an omission? If an omission is literally an absence, it appears to lack the positive occurrence ordinarily associated with an event. The absence of a rescue may be represented by a true negative proposition or fact, but it is not thereby an event in the same sense as a hand reaching toward the water. If omissions are instead treated as negative events, the theory must state their identity conditions and explain how they relate to ordinary temporal and causal structure.

Consider two descriptions: a guard's failure to lock a door at 8 p.m. and the guard's failure to secure the building that evening. These may identify one omission under different levels of generality, or two related omissions when securing the building involves several tasks. Their individuation depends on which actions count as compliance. A negative fact about a whole shift can be true across an interval containing hundreds of positive events; identifying it with any one event would appear arbitrary.

The actualist event view holds that events exist when positive changes or states occur; omission claims describe negative facts about those events or the failure of a specified process. Such a view can give a disciplined metaphysics while preserving meaningful omission explanations. A more permissive ontology may admit absences as relata in explanatory or causal relations, provided it supplies principled restrictions on which absent events qualify. Neither approach alone explains why this absence matters to an agent.

The problem of negative event identity has a modal aspect. The action that was not performed may have many possible realizations. The guard could lock the door with a key, activate an electronic mechanism, or delegate the act appropriately. These alternatives may all satisfy a higher-level description such as securing access. An omission therefore can be individuated by the satisfaction conditions of a task without prescribing one physical sequence. This feature helps explain how omission descriptions function across changing technologies and institutions.

The ontological claim advanced here is modest and precise: a theory of omission need not decide that every omission is a positive event in order to treat omissions as genuine objects of practical explanation. Negative states of affairs can be represented, distinguished, and assessed. Their explanatory and moral status depends on relations that cannot be derived from a simple assertion that nothing happened.

Intentionally Refraining and the Exercise of Agency

An agent refrains when a live option is practically represented and conduct is regulated so that the corresponding performance does not occur. The regulation can take the form of a present decision, a standing resolution, or a settled policy. A physician may refrain from an unnecessary procedure because the risk outweighs the benefit. A witness may deliberately withhold an answer to protect confidentiality. A person may decline to retaliate after judging retaliation disproportionate. In such cases, the absence is intelligible through the agent's reasons.

One argument for refraining as agency runs as follows. An agent exercises a power when the agent governs which of two practically available courses will occur. Deliberate abstention can settle a live practical question. Settling that question can be responsive to reasons, commitments, and policies. Therefore agency may be expressed through governed nonperformance. This argument establishes a genuine form of negative agency without requiring the negative performance to be an event identical to a bodily movement (Alvarez, 2013).

A central objection observes that the agent usually does something while refraining. If someone decides not to enter a room, that person remains elsewhere, walks away, or performs some other act. Why not identify the omission with the positive act? The answer is that different acts can accompany the same refraining, and the same accompanying act can support many unrelated negative descriptions. The intentional object explains what is withheld. The positive accompanying behavior may realize the policy but does not provide a general identity criterion.

Another objection concerns passive maintenance. Suppose a person resolves not to disclose a secret and then sleeps throughout an evening when disclosure was possible earlier. It would be implausible to demand active renewed decision at every moment. Intentions and policies persist; some absences are explained by earlier practical settlement. The relevant temporality is extended. An account limited to pointlike decisions cannot describe such policy-governed restraint adequately.

The difficult boundary is the case where an agent never contemplated the unperformed action. An ordinary failure to remember medication is an omission under a duty but not intentional refraining. A careful theory recognizes that agency can explain omissions both by its exercise and by its failure. The common denominator is the practical field in which performance was expected or required, not a single psychological mechanism.

Unintentional Omissions, Forgetting, and Negligence

Omissions can be unintentional and still morally significant. Someone who forgets to inspect a dangerous machine did not decide to leave it unchecked. A caregiver who is distracted may fail to notice a developing need. A worker may overlook a critical step while sincerely intending to complete a task. Negligence concerns standards of care and the capacities through which relevant risks should be recognized or addressed. It cannot be identified solely with a conscious intention to fail.

Contemporary scholarship sharply disputes the epistemic condition of blame. An occurrentist view gives central significance to what an agent actually considered; a dispositionalist view allows standing knowledge and capacities to satisfy parts of the condition; capacitarian approaches explain some blameworthiness through what an agent could and should have known under the circumstances (Rudy-Hiller, 2022). These positions differ about how demanding awareness must be. They also differ about whether culpability for ignorance requires a prior culpable act.

The regress problem deserves serious attention. If responsibility for an unwitting omission depends on a prior blameworthy failure to learn or remember, and each prior failure itself requires awareness, responsibility may recede through an indefinite history of ignorance. One possible solution assigns nonderivative significance to the appropriate use of capacities for attention and inquiry. Another identifies a specific earlier conscious decision through which the agent undertook the risk. The appropriate answer depends on whether the relevant standard is psychological explanation, moral blame, or legal negligence.

Consider the distinction between inattentiveness and incapacity. If an exhausted person is assigned an impossible sequence of tasks, omission of one task may reveal organizational failure. If the same person deliberately disregards an available warning while fully capable of checking it, the omission is better explained as culpable neglect. An adequate judgment must describe workload, training, information, time, and the foreseeable consequences of error. Otherwise the language of negligence risks turning every misfortune into personal blame.

The analysis supports a differentiated conclusion. Unintentional omission can be attributed to a person as a failure of performance. Moral or legal blame requires an additional standard of reasonable awareness and control. Serious consequences do not retrospectively create an opportunity the agent never possessed. Nor does the absence of conscious wrongdoing automatically extinguish duties of attention that the agent could reasonably have fulfilled.

Omission and Counterfactual Causation

The causal problem becomes sharp when the absence of a rescue is said to cause a death. Standard physical production accounts associate causes with positive processes, transfers, or interactions. The omission contributes no visible physical impulse. Yet counterfactual reasoning suggests that the outcome depends on whether the missing intervention occurred. Had a capable lifeguard rescued the swimmer, the swimmer would not have drowned. That difference appears explanatorily important.

David Lewis's influential counterfactual approach analyzes causation in terms of relations of dependence between events, with causal chains accommodating cases where direct dependence fails (Lewis, 1973, as discussed in the Stanford Encyclopedia of Philosophy's entry on Lewis). Applying this structure to omissions demands specification of the counterfactual antecedent. The appropriate comparison is not a world in which the lifeguard literally ceases to exist; it is a world in which a particular rescue is made at the relevant time, with suitably controlled background conditions.

A simple counterfactual test has three limits. First, many possible interventions could prevent a bad result, but we do not ordinarily call all their absences causes. Second, an attempted intervention might fail because a different process would produce the result anyway. Third, backup agents and preemption alter dependence relations. If two qualified rescuers both stand by, the victim might have survived had either intervened; yet each can say that an intervention by the other was also possible. A theory must avoid absolving both through a careless counterfactual comparison.

Jonathan Schaffer's “Causation by Disconnection” (2000) provides a different challenge to strict process accounts. A cause can be significant through disconnecting or interrupting an existing process, with no continuous positive transfer from the initial intervention to the final result. This does not demonstrate that every omission is a cause. It demonstrates that identifying causation exclusively with continuous intrinsic connections excludes important explanatory structures.

A useful distinction separates causal production from causal relevance. A missing intervention can make a difference to the outcome without supplying a productive mechanism of its own. The water drowns the swimmer; the absence of a rescue explains why that process continued. Whether the omission is counted as an actual cause depends on the causal framework and contrast class. Whether it grounds moral responsibility depends on further conditions of duty, control, knowledge, and fairness.

Explanatory Selection and the Problem of Too Many Absences

If every nonprevention counts as a cause whenever a hypothetical intervention would prevent an outcome, causal explanations proliferate without limit. A doctor across the world who never heard of a patient could have prevented a death by arriving with appropriate medicine. A passing stranger could have taken an extraordinary detour to avert an accident. Such counterfactuals may be intelligible without making those strangers the ordinary causes of the events.

Selection is therefore built into causal discourse about omissions. We select the absence of a safety inspection because a qualified inspector was assigned and the inspection would have found the fault. We do not select the absence of a passerby's unexpected engineering intervention. This asymmetry may be explained through normality, expectations, roles, causal structure, or the explanatory interests guiding the inquiry. The challenge is to distinguish the pragmatic selection of an explanatorily relevant absence from the underlying causal relation that the explanation purports to describe.

An explanation can be contrastive without being arbitrary. Asked why the door remained unlocked rather than locked, one may correctly answer that the guard failed to carry out the assigned check. Asked why a fire spread rather than remaining contained, one may mention the missing firebreak. The explanatory contrast identifies which possible difference is under consideration. Yet the existence of a contrast does not make any candidate absence causally decisive: the proposed intervention must be situated in a model of how the outcome would have changed.

There is also a risk of circularity. If omissions count as causes only when there was a duty to act, and duties are then justified by asserting that omitted actions caused harm, neither the metaphysical nor the normative inquiry has an independent foundation. A disciplined account evaluates causal difference-making on its own terms and evaluates duties through reasons, relationships, institutions, risks, and rights. Selection can be informed by norms without making physical causation identical to moral obligation.

The synthesis proposed here distinguishes an outcome process, a missing intervention, a counterfactual difference, and a normatively relevant agent. These components may jointly support a compelling explanation. They remain conceptually separable, allowing an omission to be normatively wrong without causing a harm, or causally relevant without being wrongful.

Doing, Allowing, and the Structure of Harm

The distinction between doing harm and allowing harm is adjacent to but distinct from the difference between action and omission. Philippa Foot's discussion of abortion, double effect, positive and negative duties, and harmful sequences clarified that acting and allowing do not align perfectly with moving and remaining still (Foot, 1967/2002). An agent may actively disable a safeguard and thereby allow an ongoing danger to operate. Another may disrupt a performance by failing to appear, thereby contributing to a harmful outcome through absence.

Warren Quinn's “Actions, Intentions, and Consequences: The Doctrine of Doing and Allowing” (1989) seeks a principled analysis of how agency contributes to harm. Its significance lies partly in refusing to settle the moral status of an outcome by consequences alone. Jonathan Bennett's sustained work on the doing/allowing distinction presses the relation between positive and negative facts, descriptions of behavior, and the moral weight assigned to nonintervention. The updated Stanford Encyclopedia of Philosophy analysis of doing and allowing (Woollard, Howard-Snyder, & Unruh, 2025) documents competing accounts and emphasizes the independence of this distinction from intention, preventability, and moral wrongness.

A crucial counterexample tests any simple equation of allowing with passive nonperformance. Suppose a technician disconnects a backup circuit, thereby permitting a preexisting dangerous process to continue. The technician acts positively; the result may nevertheless be described as allowing an outcome. Conversely, the absence of a designated performer can directly defeat a coordinated activity. These cases show that positive and negative relevance must be analyzed through the role behavior plays in the harmful sequence.

The moral difference is also contested. A consequentialist may assess alternatives primarily by expected outcomes and the distribution of benefits and harms. A rights-based theorist may distinguish an obligation not to invade another person's protected sphere from an obligation to provide assistance. A contractualist may ask whether the relevant principle of aid or noninterference is one that affected persons could reasonably reject. A pluralist can recognize several moral reasons without assigning an absolute priority to either doing or allowing.

An omission theory has a narrower task than a complete theory of doing and allowing. It should explain what was not done, how the agent stood to the missing performance, and whether nonintervention contributed to an outcome. The moral ranking of that relation must then be supported by independent considerations. Physical passivity has no automatic moral privilege, just as vigorous intervention has no automatic moral fault.

Positive Duties, Negative Duties, and the Limits of Demandingness

Omission judgments acquire moral force when a relevant duty to act exists. Some duties arise from voluntary undertakings: promising to deliver medicine, accepting a professional assignment, or agreeing to supervise a vulnerable person. Others arise from special relationships, creating or controlling a danger, institutional authority, or the urgency of an easily preventable serious harm. Duties differ in scope, demandingness, and grounds. A universal moral duty to prevent substantial avoidable suffering is stronger and more controversial than a particular duty owed by a designated caregiver.

Peter Singer's “Famine, Affluence, and Morality” (1972) provides a particularly clear pressure test. Singer argues that if an agent can prevent something seriously bad without sacrificing anything comparably important, the agent ought to intervene. The famous nearby-rescue case challenges a sharp moral separation between preventing local harm and helping distant persons. The argument draws its force from the structure of reasons and avoidable suffering, not from an assertion that every distant inaction is identical to directly producing a death.

A central objection concerns aggregate demandingness. Even if each small intervention is individually required, the accumulated obligations may absorb much of an agent's resources and autonomy. Moreover, many agents may be equally capable of helping. If everyone could intervene but no one does, responsibility may be distributed, overlapping, or institutionally structured. A defensible account should specify the relevant priority rules, costs, role assignments, and degree of collective coordination.

The distinction between negative and positive duties also requires refinement. Negative duties can demand effort: avoiding foreseeable injury may require checking equipment, obtaining information, and controlling hazards. Positive duties may require only modest action. Describing a duty as negative does not entail that compliance is effortless; describing it as positive does not entail that it is unlimited. The moral issue concerns protected interests and justified demands on agents.

The strongest general conclusion is that reasons for intervention vary in weight. Immediate peril, special responsibility, unique capacity, prior undertaking, and modest cost can together establish a stringent requirement. Distant and institutionally mediated harms may support duties of organized assistance rather than an unlimited series of isolated rescues. The structure of omission is stable across these cases, while the normative assessment changes according to the quality of the reasons and obligations involved.

Responsibility Without Intention: The Epistemic Condition

Moral responsibility for omissions combines control and epistemic conditions. An agent must stand in the appropriate relation to the performance that did not occur, and the assessment of blame typically examines the agent's knowledge or reasonable access to relevant information. The literature disputes the content of these conditions (Rudy-Hiller, 2022; Talbert, 2024). The dispute is particularly acute in cases where harm occurs because someone overlooked a task.

Consider a laboratory supervisor who knows the general importance of a security check but forgets it on one occasion. The supervisor may possess the training and opportunity to perform the check while lacking occurrent awareness of the obligation at the critical moment. A strict occurrentist standard may excuse the omission unless earlier culpable ignorance is found. A capacitarian approach can assess whether the supervisor's powers of attention, memory, and risk assessment were adequate in the circumstances. The latter approach better captures some ordinary negligence judgments, but it must not treat all human fallibility as blameworthy.

The relevance of a prior undertaking is often decisive. A person who accepts a responsibility to monitor a system may be required to arrange reminders, backups, or procedures compatible with predictable distraction. If that person fails to establish ordinary safeguards, responsibility can attach to the earlier organization of conduct. However, an explanation through earlier culpable behavior is not available in every case. Some departures from expected awareness occur despite reasonable precautions, making excuses appropriate.

Experimental philosophical work on forgetting gives this debate empirical relevance. Murray, Murray, Stewart, Sinnott-Armstrong, and De Brigard (2019) examine judgments of responsibility for forgetting, supporting further inquiry into the interactions among expectations, memory failures, and blame attribution. The philosophical interpretation of such evidence requires care: participants' judgments document patterns of assessment and do not themselves establish when blame is justified. Nelkin and Rickless's edited volume The Ethics and Law of Omissions (2017) brings negligence, unwitting omissions, and legal responsibility into a common comparative frame. Its significance lies in showing that responsibility can be disputed even when the missing performance and resulting harm are agreed upon.

The philosophical question is what the agent ought to have registered. Awareness of the action type, factual risk, likely consequence, moral relevance, and available alternatives can come apart. A caregiver may know that medication is scheduled but misunderstand the seriousness of delay. A technician may understand a hazard but be unable to locate a functioning control. Distinguishing these dimensions makes assessment more precise than a general accusation of carelessness.

A complete theory should also distinguish accountability from blameworthiness. Someone may properly owe an explanation, apology, repair, or institutional report for an omission even if the best explanation excuses personal moral fault. Responsibility practices include answerability, prevention, restitution, and learning. They should not be collapsed into punitive judgment.

Control, Ability, and the Opportunity to Act

The principle that obligation presupposes some form of ability is central to omission judgments. Yet ability is ambiguous. A person may have the physical capacity to swim but not know that anyone is in danger. A trained operator may know the procedure but lack access to the relevant equipment. A public official may possess formal authority without having the practical power to change a policy. A meaningful attribution of failure must identify which of these capacities was required.

The relevant practical question is whether the candidate could reasonably have performed the action at the time, given available information, control, and competing obligations. This differs from asking whether the person might have been able to act in a distant hypothetical circumstance. An isolated counterfactual world in which the agent suddenly gains expertise or tools does not establish actual negligence. The practical alternative must be suitably connected to the actual agent and occasion.

Alternative possibilities are also important to the metaphysics of responsibility. Some accounts of freedom emphasize the ability to do otherwise; actual-sequence approaches emphasize how the conduct arose from the agent's existing reasons-responsiveness and causal history. Carolina Sartorio's Causation and Free Will (2016) argues for the significance of actual causes in responsibility and explores reasons-sensitivity in causal explanation. Omissions complicate this debate because the object of evaluation can be a missing action rather than a bodily movement. The right assessment asks what actual capacities and motivations structured the failure, as well as which alternatives were meaningfully available.

A final complication is conflicting duties. A physician cannot attend to two emergencies simultaneously when both demand immediate attention. The nonperformance of one intervention may be unavoidable given the obligation to perform the other. This is an omission in a descriptive sense, and may be a tragic loss, without constituting wrongful neglect. Philosophical analysis must preserve that distinction.

Ability, awareness, and normative demand are thus jointly relevant but analytically independent. A person can have the ability without a duty, a duty without feasible performance in the particular circumstances, or an opportunity without sufficient knowledge. Only a theory that separates these dimensions can explain why omissions attract different forms of evaluation.

Time, Waiting, Delay, and Extended Inaction

Omissions are temporally structured. A missed deadline, a prolonged refusal, a momentary hesitation, and an interval of monitoring all lack a particular action, but their significance depends on timing and duration. Some duties specify a precise deadline. Others require action within a reasonable interval. Still others require continuous vigilance or an evolving pattern of intervention as risks develop.

Waiting may be rational action management. An investigator may defer a conclusion while evidence is collected; a physician may monitor symptoms before deciding on an invasive treatment; a negotiator may withhold a reply until all participants can review the proposal. These forms of waiting can express control and deliberation. The fact that an intervention was possible does not establish that it was required at the earliest imaginable moment. Practical wisdom includes sensitivity to when action becomes appropriate.

Delay can nevertheless become a distinct wrong. A person may eventually provide the promised help but fail to provide it while it would still be effective. In such cases, performance and nonperformance overlap in an extended history: the promised act occurs, yet the timely action was omitted. Describing only the final positive act would obscure the wrong. An adequate theory must individuate the performance by its temporal conditions.

Collective delay presents further problems. A committee may defer a decision through repeated votes, procedural postponements, and failures to convene. No single member's silence may be sufficient to explain the final policy, yet the institution's prolonged inaction can be attributable to decision procedures. The analysis should therefore distinguish single-occasion omissions, repeated failures, and continuing policies of nonintervention.

Temporal analysis also illuminates the relation between failure and harm. An omission may initially be harmless, then become causally relevant as a risk progresses. Alternatively, an early lapse may be irreparable even after later diligence. Assessing responsibility requires reconstructing the timeline of opportunities, reasons, and information rather than locating a single empty instant.

Legal Omissions and Duties to Act

Law provides a demanding test because moral disappointment and criminal liability must be kept distinct. Many legal systems recognize a general distinction between positive acts and failures to act, but particular duties, statutes, special relationships, voluntary undertakings, professional roles, and the creation of dangerous situations can make an omission legally relevant. The jurisdiction and legal framework are indispensable; the same conduct may be treated differently by different systems.

The central philosophical problem concerns the conditions under which the state's coercive authority may punish the failure to intervene. An open-ended obligation to rescue every endangered person would raise questions of fair notice, feasibility, proof, disproportionate burden, and administrative discretion. A carefully specified legal duty creates a clearer link between the person, required performance, and prohibited failure. This does not mean that legality exhausts morality: a morally severe omission can fall outside penal liability, while a technical regulatory omission may be legally sanctionable without grave moral fault.

H. L. A. Hart and Tony Honoré's Causation in the Law (first edition 1959; second edition 1985) examines the relation of ordinary causal explanations to legal responsibility. Their treatment supports the broader methodological warning that legal causation cannot be reduced to an unrestricted but-for test. Randolph Clarke (2014) separately examines the act/omission distinction in law alongside agency, causation, and responsibility. These works help explain why the legal category of omission depends on norms selecting the relevant actor and required conduct.

Suppose a guardian is legally required to supply necessary care. If the guardian fails to do so and serious injury follows, liability analysis must establish the duty, the relevant ability, the required mental state or standard of fault, and the connection between nonperformance and injury. Whether the absence itself is a physical producer is a separate metaphysical issue. Legal norms can responsibly attribute outcomes through causal and normative conditions without pretending that a missing movement is a mechanical force.

Comparative legal analysis must also distinguish criminal omission offenses from civil duties of care, regulatory noncompliance, and institutional accountability. A failure to file a report may constitute an offense defined by nonperformance without producing physical injury. A negligent failure to maintain equipment may generate a civil remedy through a demonstrated harm. The philosophy of omission identifies their common structure while preserving differences in legal purpose, evidence, and sanction.

Institutions, Collective Agency, and Distributed Failure

Institutional inaction is often produced through processes that no individual fully controls. A public authority may possess a duty to inspect infrastructure, while the relevant tasks are divided among inspectors, administrators, contractors, and budget committees. Describing the agency as having failed may be accurate even when responsibility among members remains complex. The institution is organized to perform a role, and its procedures establish the relevant contrast between expected and actual performance.

One model treats collective omissions as aggregates of individual failures. On this approach, an organization omitted action because its members did not perform their assigned roles. Another model attributes an omission to the institution's decision structure, recognizing that a collective policy can persist even when many individuals favor a different outcome. The second approach is particularly important where procedural vetoes, dispersed authority, and incentives generate predictable nonaction.

Collective attribution should avoid two symmetrical errors. It should not transfer all institutional blame to a single visible individual who lacked sufficient control. It should not treat diffuse responsibility as a reason that no one is answerable. An adequate analysis examines governance, role assignments, information channels, decision rights, escalation paths, and the design of incentives. Institutional omissions may demand structural correction in addition to any justified individual accountability.

Distributed failures also illuminate the distinction between explanatory causation and responsibility. A delayed safety response can be counterfactually dependent on several failures, none of which uniquely determines the result. Responsibility may be shared, differentiated by role, or assigned to a corporate body. The failure of coordination itself may be the relevant omission when a system was charged with coordinated performance.

These cases extend the metaphysics of omission beyond single bodies. Practical control can exist at the level of organized procedures and collective commitments. It is consequently necessary to specify the unit of agency under evaluation, rather than assuming that every negative fact must be reduced to one person's unperformed bodily act.

Public Conduct, Refusal, and the Meaning of Silence

Inaction can communicate. Refusing to sign a document, remaining silent in a deliberation, declining to endorse a proposal, or withholding cooperation may express a view. The interpretation depends on conventions, expectations, public context, and the agent's actual intentions. Silence may be consent under one institutional rule and dissent under another; outside such rules it may communicate uncertainty or nothing determinable.

The philosophical question concerns the relation between expressive agency and nonperformance. An intentional refusal is often accompanied by a positive speech act, but it can also be enacted simply by withholding expected participation. A participant who has undertaken an obligation to respond may be answerable for silence; a person who never accepted a public role may have no duty to engage. The expressive interpretation must therefore be grounded in an actual social relation.

The same outward absence can bear divergent meanings. A worker's nonattendance may be illness, negligence, labor protest, or compliance with a safety order. A legislator's abstention may express moral reserve, procedural necessity, or political calculation. To infer intention directly from absence would confuse a publicly observable failure with its practical explanation. Attribution requires evidence about reasons, policies, and circumstances.

Silence can also impose costs by denying expected information. Here the wrong may lie in a breach of trust, failure to warn, or failure to disclose material facts. The analysis must identify the owed communication and the legitimate expectations that make its absence consequential. It would be an error to attribute deception to every unexplained silence.

These cases demonstrate that omission is not an exclusively physical category. It is embedded in practices of language, roles, and interpretation. Negative agency can be publicly intelligible when institutions supply clear rules for what nonperformance signifies, even though the underlying absence remains an absence.

Decision Theory, Risk, and the Status Quo

Decision theory clarifies one aspect of inaction: an option to maintain the status quo can be evaluated alongside options for change. For a chooser considering alternatives under uncertainty, doing nothing may be represented as a selectable policy with expected outcomes, opportunity costs, and informational consequences. Its selection is meaningful when the agent could adopt a different policy and when the status quo is represented as an alternative.

The theoretical representation of inaction as an option does not settle its moral status. Expected-utility comparisons may help estimate consequences, but a rights-based theory can restrict permissible interventions regardless of comparative benefit. Likewise, a highly beneficial intervention may still depend on specialized authority and informed consent. Decision-theoretic choice architecture describes patterns of preference and action, while ethical evaluation supplies independent constraints.

Status quo bias introduces another layer. Agents may systematically prefer existing arrangements, treat omission as safer than intervention, or undervalue long-term harms that require present preventive action. These psychological patterns can help explain institutional or individual delay. They do not by themselves prove that a specific omission was irrational or wrongful. Sometimes the status quo is well justified because intervention creates uncertain and irreversible risks.

The relation between omission and opportunity cost must also be specified. Every chosen course excludes alternatives. If exclusion alone counted as a culpable omission, every decision would generate limitless moral debts. A philosophically useful analysis therefore selects those alternatives for which reasons, duties, and feasible opportunities are independently established.

Decision theory contributes a disciplined language for feasible options and expected consequences. Its conceptual limit is equally important: a model of choices does not automatically produce an account of intention, legal obligation, or moral responsibility. Inaction can be modeled as an option while its philosophical meaning is determined elsewhere.

Artificial and Automated Systems: Omissions as a Problem of Delegated Control

Automated systems can fail to issue warnings, execute expected procedures, or escalate incidents. These failures are routinely described as omissions. A monitoring system may miss an anomalous event; a software agent may decline a requested operation; a robotic process may remain inactive because of unavailable sensors or conflicting instructions. Such descriptions immediately raise the distinction between functional nonperformance and moral agency.

At the operational level an omission can be defined relative to a specification: an action should have occurred under stated conditions, and it did not occur. This makes automated failures suitable for logging, testing, causal reconstruction, and assignment of institutional responsibility. Yet the fact that a system has a control policy does not alone establish that it meets the conditions for moral blame. Questions about norm understanding, reasons responsiveness, autonomy, and the location of effective authority require separate philosophical analysis.

Where humans delegate important tasks to automated systems, omissions may emerge across several layers. Designers choose architectures and safeguards; operators define objectives; institutions assign authority; automated processes execute or fail to execute prescribed steps. A missing alert may be causally relevant to harm while responsibility rests partly with foreseeable design weaknesses or inadequate oversight. The relevant object may be the system-level failure rather than the absence of one isolated output.

In the philosophical horizon of the Artificial Era, these cases are relevant to the general problem of how agency and responsibility are distributed across biological and technological orders. Any more specific claim about the canonical Aisentica concepts of Artificial, Sapiens, and Artificial Sapience requires direct engagement with their official texts. This dossier makes no revision to those definitions and does not infer a moral status from technical automation.

The case of automated inaction makes visible a general lesson already established in the human cases. A description of nonperformance can be precise without identifying the bearer of ultimate responsibility. Reliable assessment depends on separating execution, effective control, normative authority, and the standards by which the missing performance was due.

Comparative Evaluation of the Main Theoretical Positions

Event-reduction theories identify omissions with positive actions occurring instead. Their attraction is ontological economy: they preserve a familiar ontology of actions and events and explain intentional omission through the causes of actual behavior. Their central weakness is underdetermination. A person can omit a task while sleeping or remaining still, and multiple positive activities can accompany the same omission. The theory must supply a principled mapping from one nonperformance to a particular positive action.

Absence and negative-fact theories preserve the difference between what happened and what failed to happen. They explain why an omission can be intelligible without any specific positive substitute. Their challenge is to identify absent performances and their causal role without generating limitless entities or causal claims. A satisfactory version requires contrast classes, temporal criteria, and relevance conditions.

Two-way-power accounts explain intentional refraining as a direct expression of agency. Their strength is fidelity to practical experience: people govern conduct by withholding actions as well as by producing them. Their challenge is the extension to unnoticed, forgotten, and negligent omissions, where no explicit settlement occurs. This calls for a layered account of direct exercise, standing policy, and failure of practical capacities.

Counterfactual accounts illuminate why omitted interventions matter to outcomes. Their strength is sensitivity to difference-making. Their challenge includes overgeneration of absent causes, background selection, preemption, and the fact that many hypothetical rescues could prevent the same harm. Process-based accounts preserve the distinction between actual production and mere nonprevention but can exclude intuitively significant cases of disconnection and prevention.

Normative theories explain why some failures are owed or blameworthy. Their strength is recognition of rights, undertakings, care, and institutional roles. Their weakness appears when moral judgment is used to settle metaphysics without argument. A duty can make an omission wrongful; it need not turn the absence into a positive force.

The comparative result is that none of these theories has to perform every explanatory function. A metaphysical description of absence can coexist with an agential explanation through practical powers, a causal model of counterfactual dependence, and a normatively grounded judgment about accountability. The main theoretical mistake is to insist that one of these relations must contain all the others.

A Layered Philosophical Synthesis

The analysis yields a structured account of omissions and inaction. Its central principle is that an omission is an unperformed, contrastively specified practical possibility associated with a candidate agent and an occasion. The framework recognizes multiple modes of attribution rather than treating all absences as a homogeneous class. It is a philosophical synthesis of the examined positions, not a proposal to amend any separate canonical philosophical system.

The first layer is descriptive. Identify the omitted action in a form that admits success conditions: administer the treatment, lock the gate, inspect the device, give the warning. Specify the relevant interval and the candidate agent. This gives the claim an object. Without such specification the statement that someone did nothing is too indeterminate to evaluate.

The second layer is practical. Determine the agent's relation to the performance: available capacity, access to information, authority, reasons, adopted plans, and competing obligations. This layer differentiates a freely chosen abstention from a task accidentally overlooked, and both from a task beyond the agent's power. The appropriate practical alternative must be realistic under the circumstances.

The third layer is explanatory. Examine whether the omission explains an outcome and, if so, through which counterfactual contrast, process, or institutional mechanism. A missing action may explain why an ongoing danger was not interrupted. It may also be causally irrelevant if intervention would have made no difference. Explanatory relevance does not require prior moral blame.

The fourth layer is normative. Identify the grounds of any duty: promises, role obligations, special relationships, rights, rescue principles, professional standards, or justified public rules. Determine the content and strength of the required performance. An omission can exist without a duty, while a genuine duty can be defeated or qualified by conflicting obligations.

The fifth layer is evaluative. Assess responsibility, blameworthiness, answerability, excuses, restitution, and institutional learning. Consider the epistemic condition, control, opportunity, reasonable care, and causal consequences. These judgments can differ. A person may owe an explanation and repair while being excused from personal blame. An institution may require redesign even where no member is individually culpable.

Together the five layers allow seemingly contradictory observations to coexist. An omission may be intentional yet justified; involuntary yet culpably negligent; causally relevant yet morally permissible; wrongful yet causally harmless; or descriptively genuine yet beyond the candidate's effective power. These combinations are essential to an adequate philosophy of negative agency.

Open Philosophical Questions

One unresolved question concerns the ontology of absences. Can an omission be an entity in an event ontology without treating every nonoccurrence as a distinct event? A promising approach distinguishes negative facts from events while allowing some negative facts into explanations. The remaining challenge is to articulate identity and persistence conditions across different descriptions and temporal scales.

A second question concerns the epistemic condition of negligence. When someone fails to notice a duty, does blame require an identifiable earlier culpable choice, or can deficient practical attention ground immediate responsibility? The debate over capacitarianism, reasons-responsiveness, and unwitting wrongdoing remains philosophically active. Any satisfactory solution must take seriously both ordinary responsibility practices and systematic human cognitive limitations.

A third question concerns causal overgeneration. If a hypothetical intervention would have prevented an outcome, which among countless absent interventions belongs in the causal explanation? Counterfactual models can specify dependence, but explanatory selection may still depend on normality, duties, mechanisms, and questions asked. Clarifying these dependencies without circularity is an important task for metaphysics and philosophy of science.

A fourth question concerns collective nonaction. Institutions possess roles and procedures through which omissions are identified, while responsibility is frequently distributed. The problem is to connect corporate answerability to differentiated duties of members without dissolving institutional agency or ignoring structural incentives.

A fifth question concerns technology and delegated control. Automated systems complicate the location of effective agency, the design of duties to monitor, and the appropriate distribution of repair obligations. The philosophical test is whether governance can specify who had which authority and opportunity, rather than treating system failure as a complete explanation.

These questions reinforce the central insight that negative agency is a meeting point of metaphysics, practical reasoning, ethics, and law. Their investigation requires preserving explanatory distinctions while tracing actual connections among them.

Conclusion

Omissions and inaction are philosophically intelligible through the specification of what could, should, or was expected to be done, by whom, and when. Their significance is located in the practical relations surrounding the missing performance. A person may exercise agency by intentionally refraining, fail through negligence, permissibly await further information, or be unable to act. These forms share a negative description while differing in agency, control, causation, and responsibility.

The historical analysis establishes continuity between Aristotle's inquiry into voluntariness, Aquinas's separation of omission from its causes, and modern disputes about the causal and moral status of nonperformance. Contemporary action theory demonstrates why omissions challenge the reduction of agency to positive events. Contemporary metaphysics reveals the difference between causal production and counterfactual relevance. Ethics and law establish the importance of independently grounded duties, practical opportunity, and justified standards of accountability.

The principal philosophical result is the layered account developed above. The identity of an omitted performance, its relation to an agent's powers, its explanatory contribution to an outcome, the duty governing it, and the agent's accountability must be judged as connected yet distinct matters. This architecture explains how an absence can have real practical significance without being treated indiscriminately as an action, physical force, or moral fault.

Understanding omissions expands the philosophy of action from what agents produce to what agents withhold, neglect, defer, and allow to continue. It shows that the field of agency includes the management of practical alternatives and that responsibility is intelligible only when those alternatives are reconstructed with precision. An adequate philosophy of action must therefore account for the consequences of what remains undone.