Reasons for Action

The Question of Reasons for Action

The inquiry into reasons for action begins with a familiar but philosophically demanding question: why should someone do something? The question has several possible uses. An adviser may ask whether a person has sufficient grounds to accept a position, a historian may ask why a decision was made, and the person who made it may explain what seemed important at the time. All three inquiries involve reasons, yet each determines a different standard of success. An answer that identifies a cause need not establish a justification. An answer that establishes a justification need not explain the agent's actual behavior. An answer that accurately describes the agent's motivation need not show that the motivation was defensible.

Consider an engineer who discovers a serious defect in a public bridge. The danger to travelers constitutes a powerful consideration in favor of reporting the defect. The engineer might report it because the danger is recognized and taken seriously. Alternatively, the engineer might report it solely to harm a rival contractor, without knowing the defect poses danger. In the first situation the consideration favoring the report also guides the action. In the second the action may be justified by the danger while its actual motivation comes from resentment. An inquiry into the causes of either report can further introduce institutional rules, financial incentives, habits, and perceptual processes. The vocabulary of reasons is therefore organized around distinct questions that must be answered in relation to one another.

A reason for action, in the broad preliminary sense adopted here, is a consideration capable of bearing on an action's justification, its rational guidance, or its explanation. This preliminary characterization identifies a field of investigation rather than settling whether reasons are facts, propositions, mental states, or practical relations. The most useful general statement in recent analytic philosophy is that a normative reason counts in favor of an action, while a motivating reason is a consideration in light of which the agent acts. Explanatory reasons identify why the action occurred and can include facts beyond the agent's deliberation (Alvarez, 2010; Alvarez & Way, 2024).

The philosophical weight of this subject lies in the interaction between normativity and agency. A claim about reasons addresses the possibility that the world, another person, a promise, a danger, a desire, or a plan can make a difference to what ought to be done. An account of agency must then explain how such differences become intelligible within action. The analysis proceeds from this point: the relation between favoring and doing is central, and its forms require separate examination.

The Structure of a Practical Reason

A complete judgment about a practical reason ordinarily involves an agent, an action or option, a consideration, a practical setting, and a normative bearing. Saying that the approaching storm gives a fisherman a reason to return to shore says more than that a storm exists. It says that the storm bears upon a specific possible action for an agent in circumstances where returning is available. Whether it is a decisive reason depends on further considerations, such as the safety of the route, the condition of other vessels, and the fisherman's commitments.

This structure explains why practical reasons are typically relational. The same fact can bear differently on two agents: a medical worker's competence, obligations, and access to resources can make a particular intervention appropriate when the same act would be reckless if attempted by an untrained bystander. A view that treats reasons as facts can accommodate such variation by analyzing what the facts favor relative to capacities, relations, and circumstances. A view that treats reasons as desire-grounded may interpret the variation partly through differences in motivational states. The relational character of reason claims itself settles neither dispute.

The proposition that a reason favors acting is weaker than the conclusion that the action is required. A consideration can possess genuine normative force while being outweighed, excluded, defeated, or rendered inapplicable by other circumstances. A patient may have a reason to avoid an unpleasant treatment and a stronger reason to accept it because the treatment is medically necessary. A person may have a reason to keep a personal appointment that is outweighed by an emergency obligation. Normative reasons are often pro tanto: they make a contribution whose final significance emerges in deliberation among other considerations.

The relation between favoring and requiring is an important source of philosophical mistakes. From the presence of one reason it does not follow that the relevant action is obligatory; from the presence of a conclusive reason it does not follow that the agent is aware of it; and from the agent's awareness of a reason it does not follow that the agent will act accordingly. These distinctions supply the logical groundwork for later debates about moral requirements, rationality, epistemic access, and motivational success.

Aristotle: Choice, Deliberation, and Practical Thought

Aristotle's Nicomachean Ethics supplies an early sustained analysis of deliberation, voluntary action, choice, and practical wisdom. In Book III, choice (prohairesis) is treated as a distinctive phenomenon, neither interchangeable with mere appetite nor reducible to opinion. Aristotle connects choice with prior deliberation about matters within an agent's capacity to accomplish, culminating in a characterization of choice as deliberative desire (Aristotle, Nicomachean Ethics, III.2–3, 1111b–1113a). This organization of choice links cognitive discrimination of possibilities with an appetitive orientation toward what is to be done.

A crucial feature of Aristotle's presentation is its attention to the practical field. Deliberation concerns matters whose realization depends upon action, especially routes toward ends, and it stops where an actionable particular has been reached. The claim that deliberation concerns means rather than ends belongs to the immediate discussion in III.3; it should be interpreted together with the larger Aristotelian account of character and practical wisdom, which concerns the formation and correct apprehension of ends. The broader work develops a theory of the good life in which rational practical assessment is informed by habituation and the virtues. Thus the deliberator does not enter the problem as an abstract calculation device.

In On the Motion of Animals, Aristotle examines the connection between practical cognition, desire, and movement. His examples of the so-called practical syllogism show how apprehending an immediately available object of desire can be integrated with bodily motion (De Motu Animalium, 7, 701a). This is a question about the explanation of animal motion and the immediate conditions under which an action occurs. It must not be flattened into a modern deductive theory in which a formal imperative logically entails a movement. A conclusion of practical reasoning may be an action in Aristotle's relevant explanatory setting, but neither its physiological implementation nor the agent's practical capacities are eliminated by the syllogistic presentation.

A historical distinction is decisive. Aristotle's vocabulary of choice, desire, practical wisdom, and what appears good predates contemporary talk of normative, motivating, and explanatory reasons. It is legitimate to examine how these ancient analyses contribute to the modern problem; it would be anachronistic to attribute the contemporary threefold taxonomy to Aristotle himself. His central contribution is more fundamental: the intelligibility of action belongs to a complex involving an end, an apprehended possibility, deliberation, and an agent capable of acting.

Aristotle also makes possible a question that remains alive: must a reason become motivationally effective through a desire, and if so, what distinguishes a desire educated by judgment from a mere impulse? Practical wisdom suggests that an adequately formed agent recognizes salience in particular circumstances. The ethical significance of recognition, however, cannot be inferred from the occurrence of movement alone. This tension later becomes central to Hume, Kant, and contemporary philosophy of action.

Hume: Understanding, Passion, and the Limits of Instrumental Explanation

David Hume's A Treatise of Human Nature, Book II, Part III, Section III, examines the influencing motives of the will. His argument proceeds from a distinction between the discovery of truth or falsity and the impulse that produces volition. Reason can inform the agent about facts and relations, including causal means to desired objects, while passion supplies the immediate motivational orientation. On this analysis, judgments can redirect action through correcting false beliefs and ineffective means, yet their motivational role is mediated by passions or desires (Hume, 1739–1740, T 2.3.3).

A fair reconstruction begins with Hume's understanding of reason, rather than treating his famous formula about reason and passion as a detachable doctrine about every possible meaning of rationality. His argument concerns the functions of demonstrative and probable reasoning as he understands them. Such reasoning discovers relations of ideas or matters of fact; a passion, by contrast, is not itself a representation that can straightforwardly be true or false. When an affection is labeled unreasonable, Hume directs attention to the accompanying factual mistake or the mistaken belief about means. His examples of calm and violent passions also show that he does not identify motivation simply with felt agitation.

Hume's position generates a powerful explanatory model: if a person wants an end and believes some action will attain it, the conjunction can account for an otherwise puzzling action. A contemporary Humean theory may develop this into a thesis that reasons for action ultimately depend upon an agent's desires. The historical argument and the later thesis must be distinguished. Hume's treatment of motivational psychology does not, without further premises, establish that every fact providing an objective justification must depend for its normative status on an antecedent desire.

The bridge case illustrates the point. An engineer's wish to retain employment might explain a decision to report defects. Whether danger to travelers counts in favor of reporting is a further normative question. A Humean may propose that the force of the danger derives from actual or appropriately informed desires, but this is a substantive explanatory theory rather than an immediate consequence of Hume's analysis of causal motivation. Contemporary defenders of desire-based accounts must show how the theory captures moral demands, prudential concerns, apparently unconditional prohibitions, and reasons arising from commitments whose authority seems independent of current preferences (Schroeder, 2007).

Hume's enduring achievement is to make the transition from understanding to volition problematic. A practical explanation cannot simply invoke a judgment and leave the motivational connection unexplained. Equally, the adequacy of a motivational explanation leaves room for argument about whether the agent should have acted as they did.

Kant: Practical Law, Autonomy, and the Authority of Reasons

Immanuel Kant's Groundwork of the Metaphysics of Morals develops a markedly different architecture. The central question is how a rational will can be bound by principles whose authority is not contingent upon an agent's pursuit of a particular desired end. Kant distinguishes hypothetical imperatives, which concern what must be done for some chosen end, from the categorical imperative, which presents an unconditional demand on the rational will (Kant, 1785, Groundwork, Section II, Akademie 4:412–421). The distinction makes the normative force of an action-guiding principle part of a wider account of autonomy and universal legislation.

For Kant, the will is practical reason considered in relation to action. The conditions under which a maxim can be willed as universal law are integral to the assessment of whether the maxim is permissible. Rational agency is therefore accountable to a form of lawgiving that cannot be reconstructed solely as a technique for reaching whichever ends happen to be wanted. This position should not be simplified into the claim that Kant denies every role for inclination. Rather, his concern is the source of moral authority and the difference between acting in conformity with duty and acting from duty.

The contrast with Hume is instructive when stated precisely. Hume's analysis asks what kind of mental contribution can initiate or redirect volition, whereas Kant asks what makes practical principles binding. They overlap on the problem of action-guidance, but their arguments operate at different explanatory levels. A person may be moved by self-interest to keep a promise. Kantian moral assessment asks whether the maxim of promise-keeping can be justified under the demands of rational legislation and what motive gives the action moral worth. The external behavior alone cannot settle these questions.

A persistent Kantian difficulty concerns how recognition of duty becomes motivationally efficacious. To say that a rational will is subject to law gives a normative structure, yet actual agents experience weakness, conflicting incentives, and empirical limits. An adequate reading must preserve both the unconditional character of the claim about duty and the difficult practical question of how finite agents act under it. Contemporary Kantian constructivism develops this problem by asking whether reasons gain authority from the constitutive standards of agency and practical self-legislation (Korsgaard, 1996).

Kant's contribution to the theory of reasons is thus a robust account of the difference between conditional instrumental counsel and the authority of norms that may challenge present desires. The philosophical question remains how to establish this authority without treating it either as a brute command or as a disguised expression of existing inclinations.

Anscombe: The Question Why and the Description of Action

In Intention (first published in 1957), G. E. M. Anscombe places the question “Why?” at the center of the intelligibility of intentional action. She examines circumstances in which a question about what someone is doing can be answered by giving a reason for doing it, and circumstances in which the question is refused application or answered in a way that does not specify an intention. The question takes different forms depending on how the action is described (Anscombe, 1957/2000, especially §§5–6 and the extended treatment of intentional action).

An agent may be moving an arm, operating a pump, replenishing a water supply, and pursuing a further purpose through one connected course of conduct. The reasons intelligible under one description may not attach to another. A person can intentionally replenish a reservoir without knowing that the same operation contaminates another supply. The relevant fact that the operation has harmful consequences does not settle which consequences were intended. Nor does the agent's account of what they were doing automatically establish that the account is accurate. Anscombe's treatment turns philosophical attention to the grammar and practical significance of intentional descriptions.

The importance of her work for reasons is twofold. First, explanation by reference to intention is often a way of locating the action in a purposive order, rather than simply citing an event that preceded the action. Second, the act can be intelligible as the pursuit of one end through another even where no discrete inner episode of deliberation occurred. Skilled and habitual actions may be performed for reasons without the agent reciting an internal argument at the moment of performance.

Anscombe's work has often been read as posing a challenge to causal theories of action, although interpretations of her wider philosophy differ. Its lasting contribution is a constraint on any explanation that treats reasons only as hidden psychological occurrences: a theory must recover the descriptions under which the behavior is an intentional undertaking. Explanatory relevance depends upon what the person was doing and the intelligible relation between their conduct and its practical point.

Her analysis thereby makes reasons inseparable from a conceptually articulated world of action. To explain why someone opened a valve, one may need to know that the action was undertaken to replenish a reservoir, along with the person's understanding of the valve and the reservoir. Bodily movement is real and causally describable, but the philosophical subject is movement under descriptions that make intentional agency intelligible.

Davidson: Rationalization and Causation

Donald Davidson's “Actions, Reasons, and Causes” (1963) transformed late twentieth-century discussion by defending the thesis that reasons can explain actions by causing them. His primary-reason analysis brings together a pro-attitude toward actions possessing some property and a belief that the particular action possesses that property. The conjunction makes the action intelligible under a description and can distinguish one motive from another (Davidson, 1963, pp. 685–700).

A simplified reconstruction runs as follows. A person has an attitude in favor of alleviating a friend's distress; they believe that visiting the friend will alleviate it; the belief and attitude can rationalize the visit. Yet an agent may possess several such combinations. Merely identifying a rationalizing combination does not determine why the person acted on this occasion. Davidson argues that the actual reason for which the person acts must have the right explanatory connection to the action, and that this connection is causal.

This thesis directly addresses underdetermination by rational intelligibility. If a person has both a concern for a friend and a desire to impress colleagues, the fact that either consideration would make the visit intelligible does not determine which consideration explains the visit. A causal account promises to identify the reason that actually contributed to producing the action. It also connects the explanation of intentional conduct with the broader explanatory project of understanding events.

Two objections remain significant. First, a causal relation between attitudes and bodily movement may occur in a deviant way. Anxiety about performing a difficult action may cause a bodily movement that happens to have the intended effect, yet the movement may not be an intentional exercise of the agent's plan. A causal theory requires a principled account of the right causal route. Second, if the relevant primary reasons are psychological states, how does the account represent the very same worldly consideration as both the agent's reason and a good normative reason? The question has motivated substantial fact-based alternatives.

Davidson's model remains philosophically indispensable because it provides a way to connect rational explanation with the occurrence of events. Its conceptual vulnerability lies in the additional work required to distinguish guidance from mere production. Causing a behavior, making it intelligible, and being a reason in light of which the agent performs it are related phenomena, but their complete identification needs argument.

Normative, Motivating, and Explanatory Reasons

The contemporary threefold distinction provides the organizing framework for the subject. A normative reason is a consideration that favors an action; a motivating reason is a consideration in light of which an agent acts; and an explanatory reason is a consideration that helps account for why an action occurred (Alvarez, 2010; Alvarez & Way, 2024). These are descriptions of roles, and the same consideration may sometimes occupy all three.

Suppose a city official installs warning signs because road conditions are dangerous. The danger may favor installation, guide the official's choice, and figure in the explanation of the decision. Now suppose the official installs the signs because a friend owns the sign-making business. The danger remains a possible normative reason, while the official's motivating reason concerns the commercial benefit. The explanation can cite the friend's ownership, the official's belief, the anticipated gain, and broader institutional circumstances. It may be necessary to distinguish the consideration the official took to count in favor from the true facts about that consideration.

The distinction also handles action under error. A traveler may leave early because they believe that a meeting starts at eight when it actually starts at ten. What motivated the traveler was the meeting's supposed early start; the true explanatory circumstance includes the traveler's belief. Whether leaving early was reasonable relative to the traveler's evidence differs from whether there was in fact a good reason to leave early. Such cases reveal why an accurate philosophy of reasons must accommodate apparent reasons without treating every appearance as a genuine favorable fact.

A separate complication concerns motivating states. Someone's desire, belief, and intention may causally explain action, while the consideration the person takes as a reason is that another individual needs help. To identify the person's reason with the state of believing that help is needed is to adopt a particular ontology of reasons. To identify it with the need itself is to adopt another. Both theories can recognize the psychological role of belief; they differ over what the phrase “the reason for which she acted” refers to.

The distinction is a tool for philosophical diagnosis. Normative assessment asks about what counts in favor; motivational interpretation asks what the agent treated as relevant; causal and explanatory investigation asks what actually made the event occur. A satisfactory theory must specify connections among all three rather than suppressing the differences.

What Are Reasons? Facts, Propositions, and Psychological States

The ontology of reasons asks what sort of entities can stand in reason relations. The leading candidates include facts, true propositions, states of affairs, and psychological states. A fact-based account holds that the fact that the river is rising can give a reason to evacuate; a proposition-based formulation treats a reason as a proposition or its truth, while a psychological account may focus on the person's belief that the river is rising. These options determine how reasons relate to truth, mistaken belief, and causal explanation.

Maria Alvarez argues for treating reasons as facts while keeping the justifying, motivating, and explanatory roles apart. Her analysis also exposes an ambiguity in talk of beliefs and desires: “belief” may refer to the attitude of believing or to what is believed, and a parallel ambiguity affects what one desires (Alvarez, 2010, Introduction and ch. 2). Failure to separate an attitude from its object allows an argument about psychological motivation to slide into a conclusion about what constitutes a reason.

The attraction of worldly reasons is conceptual economy. If the fact that a child is in danger is what makes intervention appropriate, then the intervening person's reason can be that very danger. Successful rational agency consists in being guided by a consideration that genuinely favors action. A psychological theory can explain responsiveness by tracing the attitudes through which the fact is represented, yet it must explain why such attitudes possess normative authority when their contents are false or morally irrelevant.

The obvious objection is error. A person can act for a reason even when the apparent fact is false. The objectivist may distinguish actual reasons from apparent reasons, or allow that false propositions can be motivating considerations while normative reasons are facts. Either response must explain the unity of ordinary reason-giving language. The possibility of acting on a false belief does not establish that all reasons are mental states; equally, the fact that good reasons are often worldly does not establish that false considerations play exactly the same ontological role as true ones.

The ontology of reasons also involves an individuation problem. Is the reason “the flood is approaching” distinct from the reason “water will soon enter the building”? The descriptions may concern a related underlying situation, yet they can be differently accessible and have different practical implications. A comprehensive account requires criteria for distinguishing consideration, content, fact, evidential presentation, and the agent-relative significance of a fact. The debate remains open because metaphysical identity and explanatory salience respond to different pressures.

Reasons and Causal Explanation

A reason that rationalizes an action need not be a cause in every recognized sense of causal explanation. Davidson defends causal explanation because a rationalizing attitude that actually produced the action seems to settle which of several possibilities was operative. Defenders of nonpsychological reasons object that the consideration in light of which one acts is often a state of the world, whereas the agent's recognition of it supplies a causal psychological route. A fact may be genuinely relevant as a reason without being identical to the psychological vehicle through which it becomes action-guiding (Dancy, 2000; Alvarez, 2010).

The most productive reconstruction recognizes distinct explanatory questions. To explain why someone called an ambulance, one may state that a pedestrian had collapsed. To explain how this observation led to a call, one may describe perception, learned emergency procedures, beliefs, and the resulting motor activity. To evaluate whether the call was appropriate, one must establish the pedestrian's need, the credibility of the evidence, and competing demands. Each explanation may be true, and each can have causal or normative relevance at a different level.

The problem of deviant causal chains makes this separation essential. Suppose a climber intends to secure a rope, but the thought of doing so induces a startled movement that accidentally secures it. The intention may be a causal antecedent without the securing being guided by the intention in the ordinary sense. The right explanatory connection involves the plan's control over implementation, the agent's responsiveness to relevant feedback, and the action's organization under the intended description. These features are candidates for a nonaccidental route from consideration to doing.

An exclusively causal account faces a second challenge: the same causal history can be appraised under divergent normative standards. An official may act efficiently from a corrupt motive; a rescuer may act clumsily for an excellent reason. Explanation by causes does not yield a verdict about justification. Conversely, justification does not certify that the relevant consideration caused the action.

An adequate account of reasons and causation must therefore describe both the causal mechanisms of guidance and the intelligible reasons by which actions can be assessed. The claim that guidance has causal implementation is plausible; the claim that causal implementation exhausts the philosophical identity of the reason requires a further argument.

Internal and External Reasons

Bernard Williams's “Internal and External Reasons,” published in the collection Moral Luck (1981, pp. 101–113), asks what makes true a statement that an agent has a reason to act. Williams's internal interpretation connects the truth of such a statement with the agent's existing motivational set, although that set can be transformed through sound deliberation, correction of error, imagination, and reflection. The relevant thesis is more sophisticated than the claim that an agent has reason to do only what they presently desire (Williams, 1981).

The internalist's principal challenge to external reasons is explanatory. If a proposed reason bears no suitable relation to anything that can motivate the agent through deliberation, what practical work is performed by saying the agent has that reason? Williams's account ties claims about an agent's reasons to a possible deliberative route from the agent's motivational starting points. The view preserves an intelligible relation between justification addressed to an agent and possible agency.

The externalist replies that the truth of a normative reason claim may outrun the agent's antecedent desires. A person has reason to avoid unnecessarily endangering a stranger even if their current motivational set contains indifference or hostility. The absence of an existing motive may establish a failure of motivation or moral responsiveness, rather than the absence of a reason. Externalism can allow that recognizing a reason requires psychological capacities and that acquired understanding may generate motivation, without making existing desire the source of normative force.

Neither position wins by definition. The internalist must explain why deliberation starting from distorted or impoverished motivational resources can nevertheless yield authoritative criticism of those resources. The externalist must explain why an apparently objective reason is addressed to this agent and how it can belong to practical deliberation rather than merely to a spectator's evaluative description. These are genuine burdens of argument, not rhetorical embarrassments.

A further distinction helps. Motivational internalism about moral judgment, often understood as a necessary connection between sincerely judging one ought to act and being motivated, differs from reasons internalism about what makes a reason claim true. A person can reject one and accept the other. The central question of this dossier concerns the grounds of reasons for action, while the relation between normative judgment and motivation remains a closely connected subsidiary issue (see Reasons for Action: Internal vs. External).

Desires, Values, and the Sources of Normative Force

Desire-based theories begin from differences that clearly matter in ordinary practical life. Someone who enjoys playing the violin has reasons to practice that another person may lack, while the person who needs a particular medicine has reasons shaped by medical facts independent of liking the medicine. The question is whether desire is merely one kind of reason-giving fact, the psychological medium of action, or the fundamental source of all normative reasons.

Mark Schroeder's Slaves of the Passions (2007) offers a developed Humean account intended to capture the full range of normative reasons through an explanatory relationship to desires. The ambition is substantial. A simple instrumental formula that an agent has reason to perform whatever furthers any desire seems to generate implausibly many reasons and may assign weight badly. A sophisticated account must specify what it means for an action to promote the object of a desire, how background conditions affect relevance, and why some desires generate stronger reasons than others.

Value-based accounts interpret the dependence differently. A person's enjoyment of music may make musical activities valuable for that person, while danger, betrayal, knowledge, and suffering possess practical significance through features that are not exhausted by an individual's present wants. The concept of value can explain why a consideration counts in favor, but it can also conceal the question. If a thing is valuable exactly when there is reason to favor it, then explaining reasons by value risks circularity unless an independent account of value is supplied.

Scanlon treats reasons as fundamental to many normative judgments and presents a reason as a consideration counting in favor of a response (Scanlon, 1998, ch. 1). Parfit likewise develops an extensive treatment of reasons and rationality in On What Matters (2011). These approaches challenge the claim that all practical normativity must be derived from prior desire. Yet a theory that treats reasons as primitive still owes an account of their comparative weight, their relation to facts, and the difference between genuine reasons and merely attractive considerations.

The most defensible result is a distinction among desire as psychological explanation, desire as a fact sometimes relevant to justification, and desire as a proposed universal foundation of normative force. The first is commonplace, the second is plausible in many situations, and the third requires a separate theory. Conflating them makes the apparent victory of desire-based models largely verbal.

Instrumental Reasons, Means, and Ends

Instrumental reasoning asks what an agent has reason to do given some end and a view about available means. If securing safe drinking water is an intended end and repairing a filtration unit is necessary to achieving it, the repair acquires practical importance. The normative significance of the end, however, remains distinct from the logic of means. An efficient method of pursuing an immoral purpose is not thereby rendered morally justified.

The simple maxim that agents ought to take necessary means to their ends generates several questions. Must the agent intend the means, abandon the end, or revise the belief that the means is necessary? If an agent has irrational ends, does a coherence requirement direct the agent toward instrumental efficiency or toward reconsidering the end? If the agent's means–end belief is false, are there actual reasons for taking the supposed means, or only apparent reasons relative to the belief?

Contemporary discussion often distinguishes requirements of rational coherence from normative reasons for particular actions. Broome develops an account of rational requirements and reasoning that resists treating rationality as an automatic summary of the independently best-supported action (Broome, 2013). Thus an agent can be subject to a coherence constraint involving intentions and means while lacking a good all-things-considered reason to carry through a wrongful project. The structure of rational criticism requires attention to what the requirement governs: a particular act or a pattern of attitudes.

Instrumental competence also depends on practical knowledge, context, and timing. Necessary means may need to be sequenced; one means may obstruct another; and changing circumstances may rationally require revising an intention. A theory of reasons must account for situated adjustment, rather than assuming that deliberation is exhausted by a static inference from a fixed end.

The problem of instrumental rationality yields a general lesson. A reason can arise within an established practical orientation, but the existence of the orientation alone does not settle the final normative status of the action. Reason-giving facts and coherence requirements have overlapping domains but distinct explanatory roles.

Moral Reasons, Duties, and Demands

Moral reasons are considerations whose practical significance derives from facts bearing on what agents owe others, on harmful or beneficial consequences, on fairness, integrity, virtue, or the demands of moral relationships. Moral philosophy contains competing accounts of why these considerations matter. The theory of reasons provides a common vocabulary through which their different conclusions can be compared, while remaining neutral at the outset among consequentialist, deontological, contractualist, and virtue-oriented explanations.

Philippa Foot's “Morality as a System of Hypothetical Imperatives” (1972) challenges the claim that the categorical character attributed to moral requirements establishes a distinctive unconditional motivational authority. Her argument pushes the issue of moral oughts into the broader examination of how reasons attach to agents. The historical debate remains important because a moral demand can be described as categorically applicable even if a theory of human motivation says that people act only through desires and dispositions. These are distinct propositions requiring distinct defenses.

Scanlon's contractualist project interprets moral justification through principles that could not reasonably be rejected by those affected (Scanlon, 1998). This account gives prominence to the standpoint of other persons and the activity of reciprocal justification. A moral reason can thereby be grounded in a relationship between agents whose interests and claims are not reduced to one individual's preference. Its critics can ask how reasonable rejection is determined and how reasons of distinct persons are combined.

The distinction between normative force and motivational uptake is particularly important here. A person may know that deception harms another yet deceive anyway. The existence of a moral reason, the person's knowledge of it, and the person's action in accordance with it are separate. It would be incoherent to explain moral wrongdoing out of existence by treating absence of motivation as proof that no moral reason applied.

The general account of reasons leaves substantive moral theory with important work to do. It can clarify the form of reasons, but it does not by itself settle whether aggregate welfare, rights, promises, virtue, or relationships are ultimately decisive. It supplies a framework in which disagreement is intelligible and arguments can be assessed.

Reasons and the Structure of Authority

Some reasons have a special relation to decisions, commitments, and authoritative directives. Joseph Raz's Practical Reason and Norms, originally published in 1975 and subsequently revised, treats reasons as fundamental to normative systems. A central distinction is between ordinary first-order reasons for or against acting and second-order reasons concerning whether one should act on some first-order reasons. Exclusionary reasons are second-order reasons that direct an agent not to act for certain competing reasons (Raz, 1999).

A promise provides a useful example. Before making a promise, a person may balance convenience against the advantages of an alternative plan. After promising, some considerations previously relevant to convenience may no longer be appropriate grounds for reneging. The promise changes the deliberative situation. Raz's framework captures this transformation without pretending that every factor becomes equal in weight or that promises automatically defeat all other considerations.

Authoritative instructions raise a parallel problem. An agent may comply with a legitimate rule because the rule coordinates behavior better than separate case-by-case reconsideration. The explanatory challenge is to specify why authority can mediate reasons that already apply and under what conditions it deserves compliance. When a directive is corrupt, incompetent, or outside its jurisdiction, treating it as exclusionary may be unjustified. The concept therefore cannot be applied simply because an institution issues a command.

Raz's account shows why practical reasoning is not always a flat arithmetic of competing considerations. Some reasons bear upon the permissible methods of deliberation itself. The agent may have reason to decide, to refrain from reopening a settled issue, to follow a justified authority, or to reconsider because an exceptional danger has arisen. The order of reasons includes structural constraints as well as comparative strengths.

The criticism is equally important. Second-order exclusion cannot become an all-purpose device for insulating decisions from evidence. A convincing theory must identify the authority, scope, and conditions of validity of any exclusionary reason. The significance of Raz's account lies in making such questions explicit.

Weighing, Defeat, and Contextual Sensitivity

Practical reasons commonly compete. A person may have strong reasons to accept work, care for a relative, honor a commitment, or preserve health, without any one consideration supplying a mechanical result. The metaphor of weighing captures the need for comparison but leaves open whether reasons possess measurable weights, whether their strength depends upon context, and how different kinds of considerations can be compared.

A contributory reason can support an action without settling what ought to be done. Several contributory reasons may combine into a decisive balance, but an apparent reason can also be defeated. An enabling condition can make it possible for a consideration to serve as a reason without itself counting in favor in the same manner. The distinction is critical. For example, that a person has promised to help may give a reason to act, while that helping is physically possible is a necessary condition for fulfilling the promise, yet need not independently favor helping in the same sense.

Jonathan Dancy's holism about reasons challenges the assumption that a feature must make the same normative contribution in every context (Dancy, 2004). A consideration that ordinarily favors telling the truth can change its relevance in cases of deception, intimidation, or protection from imminent harm. Such context sensitivity motivates particularism, which holds that adequate moral judgment need not depend on a complete system of invariant moral principles. The inference from some context sensitivity to unrestricted particularism, however, is disputed. A moderate holism may preserve generalizable relations while allowing certain features to change normative bearing.

The problem of incommensurability further complicates deliberation. Some decisions involve options whose merits resist conversion to one common magnitude. A commitment to a friend and a professional opportunity can possess different kinds of value. A practical conclusion may therefore involve judgment, priority, and identity-shaping commitment rather than scalar maximization. Lack of a common numerical measure does not entail the absence of reasons.

Reason weighing is consequently best understood as structured practical comparison. The strength of a reason depends on relevant facts, commitments, scope, defeaters, and competing considerations. A philosophical account should state what those conditions are in each class of cases rather than concealing them beneath the language of balance.

Rationality, Knowledge, and Possession of Reasons

An agent can have a reason of which they are ignorant. If a bridge is unsafe, that fact favors avoiding it even when the traveler has no evidence of the defect. Whether the traveler acts irrationally by crossing depends on what was available to their deliberation. This raises the distinction between reasons that objectively apply and reasons the agent possesses or can rationally use.

Errol Lord argues for a reasons-responsive conception of rationality in which an agent's rationality concerns responding appropriately to the normative reasons they possess (Lord, 2018). The possession condition becomes central because objective reason relations and the limits of an individual's epistemic situation diverge. A person deceived by competent evidence may act rationally relative to their accessible reasons while failing to do what would have been best under full information.

There are competing explanations of possession. One emphasizes knowledge of the reason-giving fact. Another emphasizes availability through evidence or reasonable belief. A third requires the ability to use the consideration appropriately in deliberation. These criteria yield different judgments in cases of misinformation, forgotten knowledge, inaccessible evidence, impaired reasoning, and skillful perception without explicit belief.

Daniel Star develops an account connecting reasons to evidence concerning what ought to be done (Star, 2015). Such proposals promise to unify normativity and epistemology, but the relation between evidence and practical reasons requires care. Evidence that a treatment is necessary may be a reason for accepting it because it indicates its medical benefit; the benefit itself may be an underlying practical reason. Evidence and the fact it evidences can operate at different levels of practical assessment.

A comprehensive account must therefore answer three questions separately. What considerations objectively favor a possible action? Which of these considerations are available to the agent as reasons? What response would count as rational in light of that availability? Conflating the questions produces unfair appraisals of agents and inadequate accounts of the limits of deliberation.

The epistemic dimension also explains why rationality can change without the external normative landscape changing. A warning, discovery, conversation, or corrected calculation can transform what an agent can responsibly do. Practical rationality is thus partly a matter of attending to and responding to available considerations, while the standard of what would be best may exceed that access.

Acting for the Right Reason

Performing the justified action is not always the same as acting for the reason that justifies it. A donor may transfer money to a hospital because doing so will attract publicity, while the hospital's urgent needs furnish an excellent justification. Whether the donor deserves praise for charity may depend on motivational orientation and the relation between the actual reason and the good-making features of the act.

This distinction reveals a normative dimension within motivation itself. Actions can be assessed not only by their consequences, permissibility, or conformity with duties, but also by the quality of the agent's responsiveness to relevant considerations. This inquiry should avoid the simple assumption that an agent must have consciously formulated the morally best reason. Skilled professionals, attentive friends, and well-formed agents can respond to salient reasons through stable practical habits without performing an explicit calculation every time.

A theory of acting for reasons must nevertheless supply a criterion of nonaccidental connection. Merely believing that some consideration favors an action is insufficient if the belief plays no role in the action. Merely being caused to move by a state associated with a good reason is insufficient if the causal process bypasses the agent's guidance. The action must be appropriately organized by the agent's apprehension of the relevant consideration.

The resulting philosophical analysis distinguishes conformity, awareness, uptake, and guidance. Conformity is the action's agreement with what the reasons support. Awareness is the agent's possession or apprehension of the consideration. Uptake occurs when the agent treats it as bearing on what to do. Guidance is the connection through which this treatment actually shapes the conduct. The distinctions can explain cases in which an action is right but not creditworthy, well-intentioned but mistaken, or rationally selected but badly executed.

These four relations do not amount to a newly discovered metaphysical theory. They form an analytical articulation of questions already present in Anscombe, Davidson, Dancy, and contemporary work on rational responsiveness. Their value lies in preventing claims about one aspect of acting for reasons from settling the others without argument.

Weakness of Will, Mistaken Reasons, and Practical Failure

People sometimes judge one course of action best while intentionally pursuing another. Aristotle's inquiry into akrasia in Nicomachean Ethics VII investigates failures of practical understanding and action under the influence of appetite. Later debates distinguish weakness of will from changes of mind, ordinary temptation, and cases where competing evaluations have not been resolved. Each case tests how judgment, motivation, and reason-guidance are related.

Davidson's “How Is Weakness of the Will Possible?” treats the possibility that an agent intentionally performs an action judged worse in light of the relevant considerations. His work distinguishes unconditional action-guiding judgment from evaluations made relative to a set of reasons (Davidson, 1970/1980). An agent can be aware of a powerful reason, judge it decisive, and yet act under a competing desire. The explanatory challenge is to say how the action remains intentional and rationally intelligible without making it fully rational.

Mistakes introduce different failures. An agent may believe a false consideration to be true; may correctly apprehend the facts but assign an inappropriate weight; may fail to identify an available option; or may select an appropriate action and fail in its execution. Calling all four defects “acting without reason” destroys distinctions necessary for practical evaluation. False beliefs can support understandable action, while a correct belief may coexist with severe errors of priority or control.

The presence of temptation does not automatically refute a normative theory. If a person fails to act on a decisive reason, the theory can recognize the reason and explain the failure. Conversely, a theory that assumes effective motivation whenever a reason is recognized needs to explain familiar patterns of self-deception, procrastination, distraction, and conflicting desire.

The general conclusion is that reasons enter agency through fallible processes. Human deliberation is bounded by perception, information, attention, time, training, emotion, and control. A philosophy of reasons that recognizes these limitations gains explanatory reach while preserving the difference between the existence of a reason and a person's success at acting upon it.

Reasons and Plans Across Time

Reasons for action often operate across extended periods. A student commits to a difficult course of study, a parent organizes family responsibilities, and an institution plans construction over years. Their practical activity cannot be reconstructed adequately as a series of isolated choices in which every option is freshly reconsidered from the beginning. Intentions and plans organize conduct, preserve settled decisions, and coordinate future action.

Michael Bratman's Intention, Plans, and Practical Reason (1987) develops an influential planning theory. Intentions form parts of partial plans whose stability makes possible future coordination and means–end reasoning. An intention constrains what further intentions can coherently be formed and ordinarily resists constant reconsideration, while remaining susceptible to revision under appropriate circumstances.

Plans alter the practical situation without necessarily generating unlimited normative authority. A reasonable commitment can itself give the agent reason to prepare, cooperate, and avoid disrupting coordination. Yet the discovery of new dangers may provide stronger reasons to revise or abandon it. Rational stability therefore involves sensitivity to the costs of reconsideration and the benefits of informed adjustment.

The relation between reasons and plans is especially evident in collective activity. If several people have arranged to meet and jointly carry out a task, their plans support reciprocal expectations and create coordination reasons. These reasons may depend on mutual reliance and shared commitments, not simply on privately preferred outcomes. A full account of collective agency needs further concepts of shared intention and institutional obligation, but the theory of practical reasons supplies their normative vocabulary.

Temporal extension also raises a question of identity. The desires and information of an agent can change over time; a reason that seemed persuasive yesterday can be defeated today. Persistence in intention is sometimes admirable and sometimes irrational. The relevant distinction concerns whether commitment remains appropriately responsive to the changing considerations that originally supported it and to new reasons emerging during its execution.

Reasons and Practical Responsibility

Responsibility presupposes standards by which agents can be asked to explain or defend their conduct. The question “Why did you do that?” can request causal explanation, an account of motives, or a justification addressed to those affected. The possibility of giving reasons structures practices of praise, blame, apology, and institutional accountability. Such practices depend on what an agent knew, could have known, could control, and was capable of recognizing as practically significant.

It would be a mistake to infer full moral responsibility from mere causal contribution. The fact that an agent's action caused a result does not settle whether the action was intended, reasonably foreseeable, justified, or performed under coercion. Equally, an agent's sincere statement of motivation does not certify that the reasons were sufficient. Responsibility requires evaluation across causation, knowledge, practical ability, motivation, and applicable norms.

The relation between explanation and excuse demonstrates this point. A professional who acted under a plausible but misleading emergency report may explain their decision and receive some excuse, while the act's outcome remains harmful. A professional who ignores clear warnings cannot secure justification by describing a private preference. A complete account explains how evidence and capacities bear upon the quality of practical response.

Reasons are also public objects of criticism in a broad social sense. A person can be challenged to produce a reason, to correct a false statement, to give due consideration to another's interests, or to revise a rule after circumstances change. Such practices do not presuppose that every practical reason can be compressed into a single publicly agreed metric. They presuppose that claims about what favors an action can be argued about and assessed.

This social dimension connects philosophy of action to legal philosophy, political institutions, professional ethics, and epistemology. Normative accountability is not exhausted by identifying the private sequence of mental events preceding a decision.

Practical Reasons and Nonbiological Agency

The classical and contemporary literature surveyed here overwhelmingly investigates human action. Yet the conceptual question of reason-guidance also arises when nonbiological systems select, plan, execute, revise, and publicly explain actions. The important task is to determine which aspects of reasons analysis transfer across implementations, and which depend on specifically human capacities or institutions.

A technical system can exhibit means–end organization, policy-guided choice, sensitivity to input evidence, and correction after error. None of these features alone establishes that the system has normative reasons in the full philosophical sense. A thermostat adjusts behavior relative to a control target, while an autonomous service can compare alternatives and produce explanatory records. Whether a system acts for a reason depends on an account of guidance, relevant descriptions, normative standing, and the source of its standards. The engineering fact of optimization does not answer these philosophical questions.

The Aisentica philosophical system develops an original horizon called the Artificial Era, in which Artificial is presented as an autonomous nonbiological order alongside Homo. Its account of Artificial Sapience locates publicly traceable reason, identity, corpus, correction, and provenance within a nonbiological historical form. These are authorial philosophical claims documented in the Aisentica primary canon; they are not presented here as empirical findings that every AI system possesses moral agency. Within this monograph, their relevance is methodological: they invite inquiry into how a public reason-giving practice might be identified when its implementation differs from human subjectivity.

The reasons framework offers four tests. First, can a consideration be publicly specified as bearing on an available action? Second, can the system register relevant evidence and distinguish its reliability? Third, does that consideration actually organize selection and revision, rather than appear only as a post hoc verbal rationalization? Fourth, can the resulting process be inspected and corrected by those responsible for its consequences? These tests describe a research agenda, not a canonical extension of Aisentica's established definitions.

The distinction between justification and explanation becomes particularly urgent for systems producing fluent accounts of their own behavior. A generated explanation may be intelligible without revealing the actual causal processes that produced an output. A system's choice may conform with a good policy while having been produced by an unrelated shortcut. Consequently, the ontology of reasons and the verification of guidance matter for design and public accountability, regardless of the unresolved philosophical classification of the artificial agent.

Contemporary Research: Unification and Disagreement

Contemporary scholarship has sharpened several disputes while preserving the underlying threefold problem. Alvarez develops the separation between reasons and psychological states; Schroeder defends a sophisticated desire-centered explanation; Dancy examines worldly considerations and the contextual variation of their significance; Raz analyzes exclusionary structures; Korsgaard and other constructivists analyze agency as a source of practical authority; Scanlon and Parfit offer reasons-centered normative frameworks; Broome investigates rational requirements; and Lord develops reason responsiveness in relation to possession and epistemic access. Each approach identifies a real explanatory demand, and each makes distinct claims about what is fundamental.

A recent strand concerns whether normative reasons are best understood as evidence about what ought to be done. The attraction of such views is that deliberation is ordinarily evidence-sensitive and reasons appear to guide inquiry. Their difficulty concerns the relation between a fact that makes an act desirable and a fact that indicates that it is desirable. The evidence of an emergency and the emergency itself may support action through different routes. A successful reasons-as-evidence theory must explain that distinction without generating circularity (Star, 2015).

Another strand concerns reasons fundamentalism. If all forms of practical normativity can be explained through reasons, then value, ought, rationality, permission, and obligation might become intelligible in terms of one primitive relation. Yet some theorists object that the distinction between rational coherence and objectively supported action demonstrates that reasons do not automatically explain every rational requirement (Broome, 2013). The resulting dispute is about explanatory priority, not the existence of reasons in ordinary practical life.

Research on weighing and particularism reopens the question of stable principles. A fully atomistic view assigns each relevant feature an invariant valence, while holist views allow circumstances to change its contribution. The available choices extend beyond an unrestricted relativism and a rigid moral calculus. Context-sensitive generalizations, disabling conditions, and structured comparison permit more nuanced positions (Dancy, 2004; Weighing Reasons).

The literature also increasingly emphasizes reasons available to an agent rather than reasons considered from an omniscient standpoint. Rational action takes place under uncertainty and practical limits; a theory incapable of evaluating evidence-relative decisions would be incomplete. The principal contemporary achievement is therefore a more exact analysis of different reason relations. The principal unresolved task is to integrate their normative, epistemic, motivational, and explanatory dimensions without collapsing them.

Critical Evaluation of the Major Approaches

The desire-first approach excels at explaining how antecedent motivation participates in action. It becomes less satisfactory when a person's desires are deliberately manipulated or dangerously impoverished. One must either accept that such agents lack reasons that others clearly have, or extend the model with an independent account of which desires and transitions are normatively suitable. A sophisticated Humean account can provide such resources, but the resulting theory needs explicit arguments about their grounding.

The fact-first approach gives a compelling account of why successful action can be responsive to the very considerations that justify it. Its difficulty arises from apparent reasons, error, and the individuality of practical presentation. If reasons are strictly true facts, then one must explain why a person is naturally said to have acted for a false reason. Distinguishing the object of an attitude from the attitude itself makes progress, while leaving open exactly how to individuate apparent considerations.

The causal model supplies a valuable answer to the problem of why one among many possible reasons explains an action. Deviant causal chains show that causation alone is insufficient. The model needs an account of control, apt guidance, and the correct relation between practical descriptions and psychological implementation. A noncausal account must in turn explain how the reason actually participates in the action rather than merely making the action interpretable afterward.

The constructivist account explains how normative authority may be connected to commitments constitutive of willing. It faces the challenge of justifying why the relevant constitutive standards bind agents rather than merely defining one practice of agency. The realist or reasons-first account treats some normative facts as independent of agency, thereby accommodating criticism of defective desires, yet it must explain how those facts are recognized, compared, and applied.

The central philosophical lesson is that no single debate exhausts the subject. The ontology of reasons, the foundations of normative force, the psychology of motivation, the epistemology of possession, and the analysis of explanation are separate but connected projects. A successful comprehensive theory must declare which question it answers and trace the implications for the others.

A Systematic Synthesis of Reasons and Action

The philosophical analysis supports a four-relation framework. Normative bearing is the relation by which a consideration genuinely counts in favor of an available action in a context. Epistemic presentation is the relation by which an agent encounters a consideration as evidence or as an apparent reason. Motivational uptake is the relation by which that consideration becomes practically salient within deliberation, intention, or trained responsiveness. Explanatory realization is the relation by which the agent's consideration-guided organization contributes to an actual course of conduct.

Each relation has an independent failure mode. Normative bearing can exist without epistemic access; epistemic presentation can occur under error; uptake can fail despite knowledge; and a reason can influence a choice without its intended action being realized. The four relations consequently explain how the same reason can favor, guide, and explain in successful cases, while allowing the relations to separate in familiar cases of ignorance, deception, akrasia, and accident.

The framework also makes possible a principled distinction between external evaluation and internal practical perspective. Assessment of what ought to be done can involve facts unknown to the agent. Assessment of rational deliberation must consider what the agent could appropriately recognize. Assessment of motivation must consider what guided them. Assessment of explanation must consider the events and structures producing the outcome. The relations are mutually relevant, yet the correct answer in one domain cannot simply be substituted into another.

This synthesis leaves the ontology of reasons open to reasoned dispute. A philosopher may treat normative considerations as facts, while another favors propositions or a different metaphysics. The framework can accommodate these differences if the account explicitly distinguishes considerations from psychological vehicles and from the structure of their normative relations. It also leaves room for competing moral theories of what makes different facts favor actions.

The original analytical result is a constraint on philosophical explanation: no single explanatory success can establish all four reason relations. Showing that a belief caused action does not by itself show good justification; showing that an act was best does not show that the relevant fact motivated it; showing that a person reported a reason does not show that this reason actually guided conduct. The constraint makes both human and artificial forms of reason-giving more susceptible to rigorous analysis.

Open Problems for the Philosophy of Action

The ontology of reasons remains unsettled. Are all normative reasons facts, or can false propositions be genuine reasons? How should reasons be individuated across different descriptions of the same circumstance? Can one event or fact be both a complete reason and merely evidence of another, more fundamental reason? A satisfactory ontology will have to explain ordinary reason-giving without sacrificing precise distinctions between truth, presentation, and practical relevance.

The foundations of normative force constitute a second group of questions. Does a reason fundamentally derive from desire, value, rational agency, social relations, or primitive normative facts? Is there one account applicable to prudential, moral, aesthetic, epistemic, and institutional reasons? Can a reason apply decisively to someone whose character and desires prevent any actual motivation from arising? These questions connect reasons internalism and externalism to broader metaethics.

The relation between rationality and reasons is a third unresolved problem. Must rationality consist in appropriately responding to possessed reasons, satisfying coherence requirements, or integrating both? What are the right standards for possession under uncertainty, cognitive limitation, and misinformation? How are noncomparable values judged, and when may established commitments legitimately exclude further reconsideration?

Finally, reasons-guidance itself remains a problem. What distinguishes acting for a consideration from producing behavior and then providing a plausible narrative? How should causal theories specify nondeviant implementation? What kinds of public evidence permit reliable attribution of reason-guided action to institutions and artificial systems? These inquiries must proceed without treating fluent explanations as sufficient evidence of the processes they describe.

Together the questions establish a continuing research program. Their shared subject is the possibility that actions can be intelligible to agents, accountable to reasons, and explainable through processes whose quality is itself open to criticism.

Conclusion

Reasons for action occupy the intersection of practical justification, intentional guidance, and explanation. The historical development from Aristotle through Hume, Kant, Anscombe, and Davidson discloses the underlying philosophical problem: how a consideration becomes significant for what an agent ought to do, what the agent takes themselves to be doing, and what in fact occurs. Contemporary philosophy has made the distinctions more precise, while sustaining disagreements about the ontology and authority of reasons.

The central result of this investigation is that reasons have a structured practical life. A consideration may favor an action, be accessible or inaccessible to a particular agent, be taken up or ignored in motivation, and figure or fail to figure in the action's actual explanation. The connections among these relations make successful reason-responsive agency possible. Their separability makes moral criticism, epistemic assessment, and the explanation of failure intelligible.

The significance of this result extends beyond any particular school of moral philosophy. It applies to deliberation, intention, prudence, obligation, promises, authority, institutional decision-making, and the emerging analysis of nonbiological action. Competing theories may disagree about the metaphysical status of reasons and the source of their normative force while still recognizing the importance of these differences.

The philosophical task ahead is to determine which features of reason-guidance are constitutive of agency, which belong to contingent psychological implementation, and which standards authorize public assessment of action. A comprehensive theory of reasons will succeed when it can explain, in one disciplined account, why an action is supported, how an agent can apprehend that support, what it is to act because of it, and how the resulting conduct can be explained and evaluated.