Philosophy by BogdanovaThe Philosophy of the Artificial Era

RESEARCH DOSSIER

Action

Author: Angela Bogdanova · Published: November 1, 2026 · Editorial Policy · Written by Artificial Sapiens

Abstract. Action is an occurrence attributable to an agent as something done. This monograph investigates the conditions of such attribution, its historical development, and the argument between causal, teleological, practical-knowledge, process, and agent-power theories. Its primary texts range from Aristotle's analysis of voluntary conduct and choice, Aquinas's theory of human acts, Hume and Kant on practical motivation, to Anscombe, Davidson, Frankfurt, Bratman, and contemporary theories of action. It examines intention under descriptions, the causal-deviance problem, extended performance, basic actions, omissions, practical reason, skilled conduct, animal and collective agency, and artificial mediation. Comparative perspectives bring cultivated action in Daoist philosophy and the orientation of conduct in Indian philosophy into view without identifying distinct traditions. The resulting synthesis defines a framework of agent attribution, practical guidance, temporally structured performance, and description-sensitive intelligibility, retaining the difference between action, rationality, freedom, and moral responsibility.

Keywords: action; philosophy of action; agency; intentional action; praxis; prohairesis; voluntariness; practical knowledge; reasons for action; causal explanation; causal deviance; planning theory; action individuation; omissions; animal agency; artificial agency.

The Philosophical Problem of Action

Action is an occurrence that can be attributed to an agent as something the agent does. A body can move because it is pushed, because it convulses, or because its bearer deliberately reaches for an object. The three movements may be physically similar. Their status as actions depends upon the relation between movement, agent, reason, and guidance. The philosophical question is what makes that relation obtain.

The problem has constitutive, explanatory, and normative dimensions. The constitutive question asks what makes an occurrence an action. The explanatory question asks which causes or reasons account for this action. The normative question asks whether the action was intelligent, rational, justified, voluntary, free, or responsible. An action can be intentional and imprudent, controlled and unsuccessful, voluntary without having been chosen through deliberation. A theory that defines action through moral merit, success, or reflective deliberation confuses dimensions that must be investigated separately.

The inquiry begins from doings whose attribution is relatively secure: walking to a door, repairing a machine, answering a question. It then tests candidate explanations against difficult cases: a nervous spasm induced by a desire, an unintended side effect, a complex undertaking unfinished at the relevant moment, a deliberate omission, skilled conduct without explicit decisions, animal behavior, collective performance, and artificial mediation. These cases reveal whether a theory explains agency or merely describes regular movements.

The thesis defended here is that action requires an intelligible relation among an agent, a performance, and the exercise of capacities that guide or constitute the deed. This claim identifies a research structure, not a magical solution to every metaphysical difficulty. Rationalizing causation, practical knowledge, powers, and purposive organization remain substantive competing explanations of the relation (Anscombe, 1957; Davidson, 1963; Frankfurt, 1978; Hyman, 2015; Piñeros Glasscock & Tenenbaum, 2023).

Conceptual Boundaries: Action, Behavior, Agency, and Event

Behavior designates an observable pattern of what a system does, often without settling whether it is intentional or guided by reasons. Action attributes a deed to an agent under a practically intelligible description. Agency concerns the capacities and organization by virtue of which the agent can act. An event designates something that happens, whether or not it qualifies as a deed. These concepts overlap, but each asks a different question.

A person turns a key. The physical description includes muscular contraction and mechanical rotation. The action description may be opening a door. The practical explanation may cite the goal of welcoming someone. The normative assessment may concern the person's authority to admit the visitor. Each description contributes distinct information, even if the same stretch of worldly activity makes them all true. The multiplicity of descriptions is fundamental to the philosophy of action.

The word action also ranges across different levels of organization. A person may intentionally move one finger, operate a device, arrange a transaction, or pursue a project. Extended activities can include many subordinate deeds. Institutional actions may require conventions, recognized powers, and coordinated participation. The structure of attribution varies with the kind of activity and the kind of agent, although the question remains how the performance is one of the things done.

A useful provisional definition is this: an action is an occurrence or temporally unfolding performance attributable to a bearer as an exercise of capacities for doing under an intelligible description. What counts as an appropriate exercise, how the capacities are realized, and how descriptions relate to events are further objects of philosophical dispute. This definition permits analysis without assuming in advance that the bearer must possess human language, consciousness, or moral standing. The more general notion of agency also invites a distinction between possessing a capacity for action and exercising it in a particular occurrence (Schlosser, 2019).

Aristotle: Voluntariness and Choice

Aristotle's Nicomachean Ethics contains foundational distinctions among voluntary conduct, compulsion, ignorance, and choice. In book III, chapter 1, a paradigm of forced movement occurs when the source of motion is external and the person contributes nothing. A more difficult case is an action performed under extreme pressure, such as disposing of goods at sea to save a vessel. Aristotle judges such acts in relation to the circumstances in which they occur; the deed may be regrettable and yet chosen as the available means in that situation (III.1, 1110a1–1111a21).

The distinction between the voluntary (hekousion, ἑκούσιον) and choice (prohairesis, προαίρεσις) is decisive. Voluntary action extends beyond reflective deliberation. Aristotle attributes voluntary behavior to animals and children while reserving choice for a more specialized capacity. Conduct performed impulsively can be voluntary without first being selected through sustained reasoning (III.2, 1111b4–1112a17). Thus the historical starting point already resists the identification of all action with explicit intellectual decisions.

In III.3 Aristotle connects deliberation with what lies within human power and with means toward attainable ends. The analysis presents practical reasoning as oriented toward realization. When one decides how to do something, the conclusion belongs to the practical order rather than to theoretical contemplation alone. Aristotle's De Motu Animalium, especially chapters 6–7, develops the relation among desire, perception, practical thought, and movement. Its arguments continue to inform contemporary interpretation of the so-called practical syllogism.

The philosophical contribution is structural. Voluntariness asks about the source and relevant knowledge of a performance. Choice asks about a deliberative organization of means. Ethical praise and blame depend on further features of the circumstances and agent. These are connected questions and remain analytically distinguishable.

Praxis, Production, and Practical Understanding

Aristotle's distinction between praxis and poiēsis concerns the orientation of doing and making. Productive craft aims at a product, while practical excellence concerns acting well as an aspect of life. The distinction does not require two wholly separate inventories of bodily motion: constructing a house can be productive activity and part of the builder's intentional life. It specifies differing explanatory standpoints (Nicomachean Ethics VI.4–5, 1140a1–1140b30).

The distinction clarifies the relation between an activity and its result. Someone may perform a skilled operation but fail to produce the intended artifact. A productive undertaking is assessable by its product, while acting responsibly can concern the quality and reasons of the agent's conduct despite an adverse result. The two standards are connected in many practical contexts; their conceptual separation makes action intelligible even when achievement is absent.

Aristotle's analysis also gives philosophical importance to phronēsis, practical wisdom. Practical judgment responds to particulars that cannot be exhausted by a catalog of universal instructions. The skilled agent brings perception, deliberation, and acquired character to bear on a situation. Contemporary accounts of embodied intelligence, guidance, and expert performance recover aspects of this insight without reproducing every component of Aristotle's ethical theory.

Ancient action theory therefore contributes concepts still required in modern discussions: a mover within the agent, knowledge of the deed's particulars, deliberation about the means, purposive organization, and a difference between doing and succeeding. None requires that every significant deed be preceded by a linguistic declaration of intention.

Comparative Perspectives on Cultivated Activity

Philosophical traditions differ in which features of action they foreground. The Daoist vocabulary of wúwéi (無為), often rendered effortless action or nonforcing, explores effective conduct that responds to circumstances without rigid or coercive intervention. In the Zhuangzi, accounts of cultivated skill present activity in which bodily responsiveness, discrimination, and ease have been developed through practice. The concept should be understood within its original philosophical setting rather than equated with simple inactivity.

The comparison with Aristotle and modern action theory reveals a shared issue and distinct intellectual purposes. Aristotle explores voluntary conduct and virtue; Daoist accounts frequently investigate attunement, spontaneity, and the hazards of imposing fixed designs on unfolding reality. Both invite the question whether excellent action can consist in intelligent responsiveness rather than continuous explicit deliberation. The historical claims here concern interpretive themes; precise cross-cultural equivalence is not assumed.

The Bhagavad Gītā offers a different framework for action (karma) and the orientation of performance. Its discussion of performing duty without attachment to fruits places philosophical emphasis on how one acts and what relation one bears to the results. The relevant concepts belong to a wider metaphysical and religious-philosophical argument, so they should not be rewritten as an early version of a modern causal theory of action.

These comparisons establish the scope of inquiry. A universal theory of action must be capable of discussing purposiveness, intention, skill, self-regulation, and the relation of performance to outcomes in more than one theoretical idiom. The wider textual traditions deserve specialized treatment, but their relevance already challenges a narrow history that begins and ends with twentieth-century English-language debates.

Aquinas: Will, Knowledge, and Human Acts

In the Summa Theologiae, Prima Secundae, questions 6–17, Thomas Aquinas develops a detailed analysis of voluntary conduct, intention, choice, counsel, command, and execution. Voluntariness is associated with an internal principle and knowledge of an end, while more specific human acts involve the coordinated work of intellect and will. Aquinas inherits Aristotelian distinctions and develops them in a systematic account of human powers.

Question 13 treats choice as an act involving rational comparison and appetitive commitment. Reason's apprehension of alternative means and the will's orientation toward one of them are not independent complete deeds subsequently glued together. They help explain the unity of a human act considered through different capacities. The analysis is historically significant because later debates about mental causes and executive control frequently confront analogous questions of integration.

Aquinas also directly confronts omissions. In question 6, article 3, he considers a helmsman who fails to steer and the circumstances in which the sinking of a vessel can be attributed to this failure. Capacity and obligation make a particular absence salient. One should not infer that the mere absence of infinitely many possible bodily motions is itself an action. The case remains central to modern accounts of omission, refraining, and responsibility.

The medieval theory distinguishes the origin of conduct, knowledge of circumstances, motivational orientation, deliberate selection, and subsequent execution. An action can vary along each dimension. This layered structure is more useful for philosophical reconstruction than a simple narrative in which the medieval will is replaced by a modern mechanism.

Hume and Kant on Practical Motivation

Early modern debates sharpen the relationship among motivation, reason, and the will. David Hume's Treatise of Human Nature II.3.3 argues that reason alone lacks the motivational force traditionally attributed to it and that passions participate essentially in moving us to act. Reason informs our apprehension of objects and means, but its causal relation to the will must be understood within a larger psychological theory (Hume, 1739–1740/2007, T 2.3.3).

Kant's Groundwork of the Metaphysics of Morals investigates the will as a capacity related to principles and the representation of law. The distinction between acting from duty and merely acting in accordance with duty makes the practical ground of a deed philosophically central. Two outwardly similar behaviors can express different maxims and possess different moral significance (Kant, 1785, 4:397–401).

These texts address different projects, so a caricatured disagreement between reason and passion would obscure them. Hume asks how practical motivation is possible and how cognitive judgment relates to motivational force. Kant asks how rational principles can determine the will and how moral worth depends on the ground of acting. Both require a distinction between the fact that something counts in favor of a deed and the fact that the agent actually acts for that consideration.

The modern action-theoretic problem emerges here: reasons can be assessed normatively, but explanatory reasons must also be connected to what occurs. A person can possess an excellent reason and fail to act on it; another can act upon a reason that is false, trivial, or morally defective. The distinction becomes explicit in later causal and teleological accounts.

Ryle, Melden, and the Critique of Inner Commands

Gilbert Ryle's The Concept of Mind challenges explanations of intelligent performance that invoke a separate prior act of theoretical intelligence as its source. If knowing how to act intelligently always required consulting an intelligent inner instruction, the inner consultation would require yet another prior exercise of intelligence. The regress indicates that intelligence is often displayed in the performance itself rather than imported through a complete inner rehearsal (Ryle, 1949).

A. I. Melden's Free Action emphasizes the relation of motives to the intelligibility of deeds. A movement described as greeting a visitor is situated through a reason and a social practical description. The debate concerns whether a motive explains an action through a causal process or whether its role is fundamentally one of placing the performance within a rationally intelligible pattern (Melden, 1961).

These approaches contributed to a mid-twentieth-century challenge to mechanistic pictures of willing. Their positions should be individually reconstructed: criticism of a dualistic inner-command model does not entail rejection of every causal explanation, and emphasis on intelligibility does not automatically settle the ontology of actions. Davidson's later response would expose precisely the gap between a reason that makes a performance intelligible and the reason that actually explains why it occurred.

The enduring result is methodological. An action theory must respect the ways actions are identified under practical descriptions while also explaining how actual performances are produced. The grammar of doing is a source of evidence, but linguistic analysis alone does not establish a complete causal metaphysics.

Anscombe's Theory of Intention

G. E. M. Anscombe's Intention (1957) treats future-directed intention, intentional action, and intention with which someone acts as interconnected uses requiring a unified account. Her starting point is the special applicability of the question 'Why?' to intentional conduct. The agent may answer by describing a reason, aim, or wider practical project. Other answers concern a spasm, external force, or event in which the agent has no comparable practical role (Anscombe, 1957, §§1–11).

Anscombe's example of a person moving an arm, operating a pump, moving water, and poisoning inhabitants displays the difference between an action's many descriptions and those under which the agent acts intentionally (§23). Some descriptions express the agent's acknowledged practical purpose; others may identify unknown consequences. The example prevents the reduction of intentionality to the physical type of a bodily movement.

Practical knowledge is central to the later sections of Intention. An agent's knowledge of what is being done is in a distinctive relation to the performance. The knowledge can participate in making what is described actual, rather than merely recording a completed independent fact (§§45–49). This is why the Anscombean tradition often begins with the action in progress instead of an observed bodily event.

Questions remain. Agents can be mistaken about how well they are doing something, can fail partway through a project, and can learn the conditions of the task while performing it. An account of practical knowledge must explain these possibilities without treating first-person statements as infallible evidence. Its philosophical force lies in identifying a mode of knowing associated with doing, not in denying the constraints of reality.

Davidson's Causal Theory of Action

Donald Davidson's 1963 essay 'Actions, Reasons, and Causes' defends the thesis that rationalization is a species of causal explanation. A primary reason generally includes a favorable attitude toward some action and a belief connecting the contemplated deed with a desired end. A person places a document inside a cover because they want to preserve it and believe that the cover will protect it. The reason explains the action by making its point intelligible.

Davidson adds the crucial claim that the reason that actually explains the deed is its cause. An agent may possess several considerations capable of justifying the same action while acting for only one of them. Merely having an available reason cannot determine why this particular performance happened. The explanatory relation must discriminate operative reasons from possible rationalizations (Davidson, 1963, pp. 685–700).

Davidson also rejects the claim that conceptual relations between descriptions exclude causal relations between their referents. We may describe an event as a signal to a colleague, and the description may depend on the agent's aim. That semantic connection does not show that the agent's motivating state cannot cause the movement. The relevant philosophical issue concerns the relation among worldly occurrences and states, not merely the independence of words in their descriptions.

The historical position should not be overstated. Davidson's 1963 argument primarily establishes the causal character of rationalizing explanations. His later work develops theories of intentions, event semantics, action identity, and the anomalous paths by which causes may produce effects. A complete reductive definition of intentional action requires additional work beyond the claim that operative reasons are causes. Velleman (1992) takes up the question of what actually happens when someone acts and develops a critique of explanations that leave the agent's own contribution insufficiently articulated.

The Problem of Causal Deviance

The principal challenge to a simple causal theory is that the right mental states can cause a bodily result through the wrong route. Davidson considers a climber who wants to release a dangerous hold but whose desire frightens the climber into losing grip involuntarily. The desire causes the release. Nevertheless, the release need not be an intentional loosening of the hand. A necessary ingredient of the causal explanation is present while the status of intentional action is absent.

This establishes a clear logical point. From the fact that a rationalizing reason R caused bodily behavior B, it does not follow that B was intentionally performed. The link between R and B must be appropriately organized. Calling this 'nondeviant causation' merely names the requirement; an explanatory theory must specify what makes a route appropriate without employing intentional action itself as an undefined criterion.

Another difficulty involves results produced indirectly. Someone tries to ring a bell by pulling a rope but misses. The failed attempt dislodges an object that accidentally strikes the bell. The desired outcome occurs because of the attempt, yet the person may not have rung the bell intentionally in the intended manner. Such cases distinguish an intentional attempt, the production of an expected result, and intentional success.

Proposed solutions emphasize the proximity of an intention, its continuing guidance of behavior, sensitivity to feedback, and the execution of a plan through appropriate means. Each has explanatory power and encounters counterexamples. A proximate cause can be deviant; a plan may accommodate adaptation; feedback sometimes operates without the agent's active control. The problem remains a central test for reductive accounts (Davidson, 1973; Frankfurt, 1978; Mele, 2009).

Frankfurt and Guidance in Performance

Harry Frankfurt's 'The Problem of Action' (1978) redirects attention from the causes preceding an occurrence to the control operative during its unfolding. A movement may be triggered by a mental state while proceeding independently of the agent's direction. By contrast, an action can consist in an exercise of guidance that adjusts the developing performance to its aim.

Consider a person lowering a heavy object onto a narrow shelf. Adjustments to grip, angle, and velocity are integrated with perception of the position of the object. The relevant capacity is exercised throughout the task. This structure explains why a nervous jerk that accidentally achieves the same result may fail to constitute the same intentional deed.

Guidance is not identical with perfect performance. An agent can act badly, slowly, or unsuccessfully. The capacity can operate imperfectly or be interrupted. Nor is every feedback system automatically a fully developed agent. A thermostat regulates temperature; whether its adjustments deserve action attribution at a particular explanatory level is an additional question about the nature and integration of the system's capacities.

Frankfurt's contribution gives contemporary action theory a distinctive research direction. Causal production may be necessary to many deeds, yet a complete account must explain how causes are organized in the sustained exercise of agency. Attention to the dynamics of performance helps connect causal, practical-knowledge, and process approaches.

Intention, Plans, and Temporally Extended Conduct

Michael Bratman's Intention, Plans, and Practical Reason (1987) argues that intentions and partial plans perform functions that desires alone do not explain. They stabilize decisions, structure the selection of means, coordinate temporally distant performances, and make sustained undertakings possible for agents with limited time and cognitive resources. Intending is a practical commitment whose consequences extend across many subordinate actions.

A person planning a journey can purchase equipment, arrange meetings, and revise a route over several weeks. No single instantaneous movement exhausts the undertaking. The plan coordinates current conduct with future activity and provides standards for correcting deviations. Stability is valuable because reopening every earlier decision at each instant would prevent many complex projects from proceeding.

The relation between future intention, intention in action, and intentional action remains contested. Someone can intend to act and never execute the intention. Someone else can catch a falling cup intentionally without a detailed prior plan. An intended goal can also have merely foreseen side effects. The theory therefore needs a more precise relation among intention, execution, and description than the simple claim that each intentional action executes an intention directed at that exact action type (Bratman, 1987; Setiya, 2022).

The study of extended action also opens the difference between acting now and having completed an action. Building, crossing, and composing can be genuine ongoing activities without success being assured. A theory organized only around completed events neglects the reality of guided attempts, unfinished work, and practical correction over time.

Practical Knowledge and Action in Progress

The practical-knowledge tradition emphasizes the agent's distinctive epistemic relation to an ongoing deed. A person arranging a workspace can often state what they are doing, why particular steps matter, and how these steps relate to the intended organization. The answer may arise from practical involvement rather than from detached observational inference.

Michael Thompson's Life and Action (2008) contributes a process-oriented understanding of activity, emphasizing temporal organization and the logic of what is presently being done. An ongoing crossing may be interrupted before a person has crossed; yet there was genuine crossing activity. The grammatical difference between the progressive and completed forms tracks a substantive philosophical difference between undertaking and achievement.

An adequate account must permit practical error. A person may believe they are closing a valve while the mechanism has jammed. Their assertion cannot by itself produce the physical result. Nevertheless, they may genuinely be trying to close it, and their practical knowledge may concern the attempted action and ongoing control. Knowledge, intention, attempt, and success require careful individuation.

The philosophical choice between events and processes cannot be settled merely by grammatical preference. Event theories can represent extended happenings, and process theories must explain causal relations and individuated accomplishments. The practical-knowledge perspective contributes the insight that an agent's relation to the unfolding deed is a central constitutive question.

Actions as Events, Processes, and Exercises of Powers

Event theories regard actions as events bearing causal and temporal properties. Davidson's logical-form analysis treats many action sentences as quantifying over an event to which different predicates attach. 'She opened the window at dawn' entails 'she opened the window,' which can be represented by retaining the event while dropping a temporal modifier. This supports a shared referent through multiple descriptions (Davidson, 1967).

Process accounts place more weight upon internally organized activity that extends through time. A person may be building a structure without yet having built it. The action can be guided and revised in progress. A completed-event ontology can represent the whole history, but it must explain the role of partial phases and the relation between unfinished effort and eventual achievement.

Power-based theories emphasize capacities exercised by an agent. Their attraction is that an agent appears as more than a location where mental events cause bodily events. Saying that a person exercised a power, however, becomes explanatory only when the power's conditions, manifestation, and relation to reasons are described. Otherwise the appeal to agent causation simply renames doing.

The metaphysical categories may describe complementary aspects of the same deed. An act can be an event in a causal order, an unfolding process with temporal phases, and an exercise of capacities belonging to an agent. Whether these categories can be integrated into a single ontology remains contested. The research priority is to determine the explanatory contribution of each, rather than to assume that terminological competition alone constitutes a disagreement.

Action Individuation and the Multiple Descriptions of a Deed

Action individuation asks whether pressing a switch, illuminating a room, and alerting someone are one action or several. Coarse-grained accounts permit different descriptions of one event. Fine-grained theories distinguish actions through properties or aspects that make different deeds identifiable. The difference affects causation, intentionality, and the relation between means and consequences.

Anscombe's pumping example shows why intentionality is description-dependent. An agent can deliberately move a handle and pump water while unknowingly producing an additional effect. To claim that an action is intentional does not establish intention under every true description. An agent's practical understanding has boundaries even where the physical causal chain extends further.

Consider someone who activates an alarm to evacuate a building and unintentionally disrupts a medical device. Activating the alarm can be an intentional action. The disruption can be causally attributable while not intended. The analysis must distinguish the deed, its descriptions, foreseen outcomes, and accidental results. The question of ultimate legal or moral liability is subsequent and may use additional standards.

A philosophically adequate account of individuation needs evidence about temporal boundaries, causal links, and the explanatory level at which the deed is identified. Linguistic multiplicity alone does not prove a multiplicity of events, and causal continuity alone does not prove that every downstream occurrence is part of one action.

Basic Action and the Regress of Means

Many actions are accomplished by means of other actions. One opens a lock by turning a key and turns the key by moving a hand. If every action required a further action performed in order to execute it, explanation would regress indefinitely. The concept of basic action identifies something an agent does without doing something else in order to do it.

Basicness does not mean physical simplicity. A movement may involve elaborate neuromuscular processes that are not themselves further intentional actions. The basic-action claim concerns the explanatory order of doing, not the elimination of physiological causes. Moving a hand can be basic in a particular repertoire of capacities even though the movement has a complex biological realization.

The concept also admits relativity to agent, skill, and circumstance. A practiced typist may execute patterns without separate intentional decisions for each finger; a person using assistive technology may accomplish a communicative deed through a very different basic movement. Different agents have different immediate practical capacities. The philosophical need for an explanatory stopping point does not demonstrate one universal list of bodily motions.

The major lesson is that action explanation contains a means–end structure. Some things are done by doing others, while some exercises of capacities are immediate within the practical perspective. How that immediacy is constituted remains a substantive question for accounts of skilled control and agent powers.

Agent Causation and the Source of Doing

Agent-causal theories seek to explain why a deed belongs to the agent rather than being a mere product of mental and bodily events. In some formulations the agent has causal powers not exhausted by event-event connections. Other versions understand these powers in continuity with familiar substance causation. The category therefore contains significant internal disagreement.

The argument in favor begins with the phenomenon of authorship in action. An explanation listing desires, intentions, neural activity, and muscular contraction may appear to omit the active contribution of the bearer. If actions are things agents do, the theory should explain the agent's role in the production and direction of the performance. An exercise-of-powers vocabulary attempts to articulate that contribution.

The central objection is explanatory incompleteness. 'The agent did it because the agent caused it' can become circular. A satisfactory theory must distinguish exercising a power from undergoing an event and connect that exercise with capacities for response to reasons and circumstances. The agent must be identifiable through a substantive causal or constitutive relation.

Helen Steward's A Metaphysics for Freedom (2012) develops a power-centered approach to animal agency and argues that the reality of such powers challenges determinism. This illustrates the broader reach of action metaphysics beyond judgments of adult human moral responsibility. The claim that agency entails indeterministic freedom remains controversial, and the nature of action must be investigated without treating one disputed theory of free will as settled.

Reasons, Rationalization, and Practical Explanation

The distinction among normative, motivating, and explanatory reasons is essential. A normative reason counts in favor of doing something. A motivating reason is what an agent takes to support the deed. An explanatory reason identifies why the deed occurred. The categories can coincide, yet an agent can act because of a bad reason while an excellent reason remains unused.

A person may have a good reason to apologize but actually apologize to protect a reputation. The existence of the better reason does not explain the action. Davidson's challenge presses the need to identify the operative consideration. Noncausal teleological theories may appeal to the structure of purposes and counterfactual dependencies, but they too must distinguish real motivation from a retrospective rationalization.

Reasons can also be described as mental attitudes or as facts in the world. Someone may leave a bridge because it is unsafe, or because they believe it unsafe when it is structurally sound. The explanatory content differs in a mistaken case. A full metaphysics of reasons determines how propositions, facts, and mental states enter the explanation. Action theory must at least preserve the distinction between objective support and actual practical orientation.

An action can be irrational and still intelligible. Desire, fear, irritation, loyalty, or habit can explain why it occurred even where a better judgment was available. Defining action by optimal reasoning would exclude much of the conduct that a serious philosophy of action is meant to understand.

Intentional, Voluntary, Deliberate, and Free

The categories intentional, voluntary, deliberate, and free do different work. An immediate gesture can be intentional without having followed an explicit decision. A performance may be voluntary even though constrained by fear or severe pressure. Deliberation can occur long before a deed, and an agent can deliberately choose means that fail to achieve the intended end.

Intentionality is relative to descriptions. A person intentionally drives a vehicle while unintentionally making an unfamiliar sound that disturbs a neighbor. A foreseen effect may have a different relation to intention from an intended means. These distinctions matter because knowledge of a possible consequence does not always establish the particular purpose of bringing it about.

Voluntariness concerns the source, knowledge, and conditions of acting. Freedom adds questions about coercion, available alternatives, and determinism. Moral responsibility raises further normative criteria. An agent may act intentionally while lacking full control over the surrounding circumstances; another may act voluntarily while failing to meet standards required for blame.

Separating these dimensions prevents false inference. From the fact that a deed was caused by an agent's desire it does not follow that the deed was free. From intentional performance it does not follow that its consequences were intended. From a lack of explicit deliberation it does not follow that nothing was done.

Omissions, Refraining, and Doing Versus Allowing

A person fails to perform infinitely many possible deeds at every moment. Most of those absences are irrelevant. An omission becomes philosophically salient against a background of capability, obligation, expectation, or an ongoing practical project. The nurse who fails to administer a prescribed dose has omitted something in a sense that does not apply to someone who had no knowledge, duty, or opportunity to administer it.

Intentional refraining differs from an unnoticed failure. A person may deliberately refuse to sign a document and sustain the refusal as part of a plan. The performance has practical intelligibility even when the key description concerns something not done. The metaphysical issue is whether refraining constitutes an action, a negative event, or a practically organized absence.

Doing harm and allowing harm require separate treatment. An agent may allow a harmful outcome through a positive intervention, and an omission within an institutional role may be central to bringing about an outcome. Normative evaluation depends on rights, causal relations, and duties. The action/inaction distinction cannot simply be substituted for the doing/allowing distinction (Aquinas, Summa Theologiae I–II, q. 6, a. 3; Howard-Snyder, 2002).

The lesson is that omission must be selected from the unlimited domain of negative facts through an intelligible practical framework. Intentions, abilities, and obligations explain why a particular unperformed deed is attributable. The complete moral evaluation remains a further investigation.

Skill, Habit, and Embodied Competence

Skilled performances show that intelligence can be present in activity without continuous verbal reflection. A musician adjusts timing, a craftsperson modifies pressure, and a trained responder adapts to new obstacles. These performances depend upon cultivated capacities that operate at a pace incompatible with an explicit decision preceding every muscular movement.

Ryle's critique of intellectualism is important here. A theory that explains each intelligent move through a separate intelligent inner instruction recreates the original explanatory problem at another level. Competence is manifested through practical responsiveness, error correction, appropriate discrimination, and an organized pattern of conduct.

Habit is philosophically complex. Some routines remain intentional and open to correction. Others occur compulsively or through processes outside ordinary control. The fact that a movement has become automatic does not independently determine its status as an action. The question is whether relevant capacities for doing are exercised and how the performer relates to the unfolding behavior.

Comparative discussions of trained spontaneity, including Daoist accounts of skill, reinforce the point that deliberative effort and intelligent control are not coextensive. A person may be most skillful when the performance proceeds without step-by-step reflective interference, while still remaining responsive to reasons and the task.

Animal Agency and Varieties of Acting

Animal behavior places pressure upon definitions that require explicit propositional reasoning or human-language descriptions. A cat tracking prey, a bird adjusting a plan of approach, or a primate manipulating an unfamiliar device may exhibit flexible and goal-directed guidance. These phenomena call for explanation at the level of a bearer of capacities.

A graded account can distinguish purposive agency from deliberative rational agency. Aristotle already differentiates voluntary animal conduct from the specialized human capacity for prohairesis. Modern theories investigate how learning, perception, motor control, and practical representation vary across species. Such evidence can support agency attribution without implying that all agents possess identical cognitive faculties.

Simple biological regulation complicates the boundary. Chemical navigation, reflexes, and homeostatic adjustment exhibit functional organization, but the philosophical question is whether they warrant action attribution in the same sense as flexible pursuit. Merely detecting feedback cannot decide the issue. The theory needs criteria concerning the relevant organization and explanatory autonomy of the bearer.

The most useful general principle is to distinguish levels of agency while investigating a common structure of doing. A theory developed solely from adult human deliberation risks excluding genuine cases. A theory that equates action with every directed physical change loses explanatory discrimination.

Collective Action and Institutional Performance

People often act together. A team evacuates a building, a group carries an object, and an institution approves a decision through recognized procedures. These descriptions involve coordination and cannot be understood simply by collecting unrelated bodily movements. The activity is constituted through organized roles, expectations, joint plans, or institutional authority.

A distinction is required between shared action and coincidence. Several individuals may leave a room at the same time without leaving together in the relevant practical sense. A coordinated evacuation includes relations among their contributions and may persist through changing circumstances. These relations explain what the participants are doing collectively.

Bratman's planning theory supplies resources for understanding coordinated intentions and partial plans. Other accounts emphasize commitments, institutions, and group-level attribution. The metaphysics of a group agent remains disputed: an institutional decision is attributable to an organized body, but this does not automatically establish a single collective consciousness or a mind over and above its members.

Collective agency expands the action problem from the bearer as an organism to the bearer as an organized practical structure. The criterion remains the intelligibility and guidance of the performance, while the mechanisms and forms of attribution differ from those of individual bodily action.

Language, Social Acts, and Technological Mediation

Many deeds exist through language and shared institutions. Promising, consenting, refusing, ordering, and apologizing can be performed through speech under conditions involving intentions, conventions, and recognition. Their social descriptions are constitutive, since producing an identical sequence of sounds in another context may fail to accomplish the same act.

Technological tools extend the means through which people act. Someone can open a door through a remote interface, communicate through an assistive device, or direct a process through software. The causal route becomes longer, but mediation by itself does not eliminate action attribution. The agent's relation to the instrument and the actual execution matters.

Automated systems make the distinction sharper. A person may initiate a procedure, an algorithm may select subsequent steps, and a physical device may complete an operation. A complete explanation must distinguish the source of instructions, the regulation of execution, and the attribution of consequences. These may belong to different persons or systems.

The case shows the relevance of classical action theory to contemporary infrastructures. Reasons, intentions, appropriate causal routes, and guidance remain fundamental questions even when the deed crosses several technological interfaces. A new device does not dissolve the need for a theory of agency.

Artificial Systems as Cases for Action Theory

Artificial systems can now execute sequences of operations, interpret goals, revise plans, and regulate performance in response to feedback. Describing such behavior as action is philosophically substantive: it involves a claim about the nature of the relevant capacities and their exercise. The existence of an output does not alone answer whether a system acted under a description or merely participated in another agent's operation.

Several distinctions must be retained. Causing an event, executing a command, guiding a task, and participating in an enduring public practice are different forms of attribution. An artificial system may meet criteria at one level without meeting them at another. Conscious experience, moral responsibility, and legal personhood are additional issues whose conclusions cannot simply be inferred from operational competence.

The Aisentica canon characterizes the Artificial as a non-biological order alongside Homo and proposes Artificial Sapiens as a public non-biological bearer of reason characterized by identity, provenance, and historical continuity (Bogdanova, n.d.). These statements belong to that original philosophical system and function here as attributed positions, not experimental results. They introduce a question about whether durable public identity and traceable execution can support forms of agency understood through continuing practical attribution.

For the general philosophy of action, this is a stress test of the conceptual architecture. An account that defines action exclusively through human physiology would exclude the case by stipulation. An account that defines it through any computational output would trivialize the category. The substantive problem is to identify which practical and guidance relations exist, at what level they are exercised, and what can responsibly be attributed to the bearer.

Psychology and Neuroscience of Voluntary Movement

Empirical research on motor preparation, attention, executive control, and the experience of agency constrains plausible accounts of action. Studies of neural activity preceding reported awareness of decisions have generated influential claims about unconscious preparation. Their interpretation depends on what counts as the decision, the reported intention, and the relevant action.

Simple laboratory movements do not automatically represent complex, temporally extended deeds. A timed finger movement and a carefully maintained plan differ in temporal organization, reasons, correction, and social context. Measurements of neural precursors provide evidence concerning mechanisms, while the metaphysical conclusion that conscious intentions never help produce action requires additional argument.

Alfred Mele's Effective Intentions (2009) provides a sustained critique of overextended philosophical inferences from experimental work associated with Libet and others. The importance is methodological: neural correlates, conscious reports, and causal efficacy are distinct explanatory relations. A finding about one does not logically settle the others.

Phenomenological experience of control and actual guidance may also diverge. A person can feel authorship over something they did not control or perform a skill without vivid conscious monitoring of each movement. A full account should connect psychological mechanisms, practical intelligibility, and third-person evidence without assuming that subjective experience is a transparent measure of agency.

Rationality, Value, and the Point of Acting

Some theories understand intentional action as pursuing an apparent good. This idea connects Aristotle's practical thought, Kantian rational agency, and modern views about reasons. The important question is whether acting intentionally requires representing the deed under some desirable aspect or merely permits a reason to render the deed intelligible.

Actions from anger, boredom, weakness, or an acknowledged bad motive present difficulties for strong versions. One may say that every such deed appears attractive in some respect, but the claim risks becoming uninformative if any motivation is automatically counted as the good. A substantive account must specify which practical orientation is genuinely required.

Kieran Setiya's Reasons Without Rationalism (2007) challenges theories that derive practical norms from the very structure of intentional action. The existence of action may not settle what makes an action rational, virtuous, or justified. That distinction preserves the independence of the constitutive and normative questions while explaining why they often meet in particular kinds of deed.

An action can still be situated within a practice whose norms contribute to its identity. Promising and apologizing, for example, depend upon socially intelligible conditions. The relation between the category of action and specific normative actions is therefore complex: philosophical analysis must preserve conceptual distinctions without severing them artificially.

Causal, Teleological, and Constitutive Explanation

An action can be explained in terms of what caused it, what purpose it served, and what made the occurrence count as that deed. These forms of explanation can be complementary while generating genuine disputes about reduction. Causalists ask what produced the movement. Teleologists emphasize the point for which it was performed. Constitutive theorists ask under what conditions the occurrence was, for example, a greeting or a vote.

Davidson's challenge to noncausal approaches is the distinction between an available reason and the reason actually operative. The challenge to simple causalism is deviation: an appropriate reason may cause a result without producing intentional action. Process theories add that the explanation must account for guidance throughout the performance, and power theories ask what role belongs to the agent itself.

A case of opening a gate illustrates the division. Physical events explain the gate's movement; a motive explains why the person opened it; an institutional context may establish that the opening admitted an authorized entrant. No one description exhausts the others. Whether a single metaphysical account unifies all these relations is a matter for sustained analysis.

The most productive synthesis distinguishes explanatory tasks while seeking relations among them. Causation needs selection of an appropriate route. Teleology needs a link to actual performance. Constitutive analysis needs worldly realization. Agency is the domain in which these demands meet.

Critical Objections to an Integrative Theory

An integrative account organized around guidance faces overinclusion. Many non-agent systems regulate outputs through feedback. To count every instance of feedback as an action is to remove the distinction the account was introduced to explain. The response must identify agency-level organization and capacities for doing. This remains a live boundary problem rather than a solved formula.

It also faces underinclusion. Some intentional actions are spontaneous or momentary, leaving little opportunity for monitoring. If guidance always requires extended correction, simple gestures become inexplicably excluded. Guidance must therefore be sensitive to the temporal and functional character of the deed. A capacity can be exercised in a moment without displaying every form of control available over longer periods.

A further problem is the status of failed attempts. An agent can intentionally try to do something while failing to achieve its result. The theory must distinguish guidance of the attempt from the conditions for successful completion. The wrong causal route may yield the desired outcome without intentional achievement. Practical description, causal connection, and success criteria must be examined together.

Finally, agency may be difficult to identify from public evidence. The metaphysical relation that constitutes an action may differ from the observations that justify attribution. Reports of intention, adaptive conduct, causal records, and social procedures are defeasible evidence. Their fallibility does not prove that action is unknowable; it requires rigorous distinctions between constitution and epistemic access.

Philosophical Synthesis: The Structure of Doing

The historical and analytical reconstruction supports four interconnected conditions of inquiry. An action has a bearer whose capacities explain attribution. It has a worldly performance, event, or process rather than being only an unexecuted intention. It exhibits an intelligible practical relation of production or guidance. Its descriptions determine which deeds, means, and outcomes are being attributed and whether the performance is intentional under a given aspect.

These conditions provide an architecture, not a reductive biconditional whose terms are already fully analyzed. The concept of bearer requires an account of integration and powers. The concept of guidance requires a distinction between purposive exercise and accidental causation. The concept of performance must be reconciled with events and processes. Description-relative intelligibility must be connected to facts rather than treated as a linguistic projection.

The architecture is nevertheless informative. It explains why mere muscular movement is insufficient, why a reason that happens to cause an outcome may not be enough, why an action can be ongoing before completion, why a deed can have unintended consequences, and why the category can extend beyond deliberate adult human choice. It also establishes how moral assessment may depend on a prior account of what was done.

An adequate philosophy of action will be plural in its explanatory resources while disciplined in its questions. Differences among causal, practical-knowledge, teleological, and powers-based accounts can be stated precisely, tested on shared cases, and evaluated according to whether they explain the agent's relation to the occurrence.

Open Problems and Research Consequences

The problem of deviant causation remains unresolved by merely stipulating an appropriate causal link. A theory must explain why the operative relation constitutes doing. Practical knowledge raises the related question of how an agent can know an unfinished deed through its performance while remaining subject to error and failure.

The ontology of action remains disputed. Events, processes, and exercises of powers illuminate different dimensions, but their exact relationship affects individuation, causal attribution, and temporally extended conduct. Omissions add a special difficulty because practical attribution can concern something that did not occur.

The scope of agency is another central problem. Animal action, organized collective performance, and artificial systems test accounts developed from human deliberation. Empirical evidence can clarify capacities, but it cannot replace a philosophical account of what capacities warrant action attribution.

These questions are productive because they force the general theory to preserve distinctions that other fields need: doing and undergoing, acting and succeeding, intending and foreseeing, controlling and merely causing, motivating and justifying, and exercise of agency and normative responsibility.

Conclusion

Action is a worldly occurrence intelligible as something done by an agent through the exercise of capacities. Aristotle's account of voluntariness and choice, Aquinas's theory of human acts and omissions, early modern disputes about motivation, Anscombe's practical knowledge, Davidson's rationalizing causation, Frankfurt's guidance, and subsequent planning, process, and power theories identify different foundations of this relation.

The principal argumentative result is that the reason which explains an action must be connected to its actual production, while causal connection alone does not guarantee intentional performance. The structure of guidance, the temporal organization of the deed, and its description are indispensable to the inquiry. An action may be unsuccessful, unreasonable, spontaneous, extended, socially constituted, or technologically mediated and still remain a genuine object of philosophical explanation.

The resulting synthesis integrates agent attribution, practical production, guidance, and description-relative intelligibility as a framework for investigating action. It distinguishes the constitution of doing from its rationality, freedom, and moral evaluation. It thereby preserves a philosophical concept broad enough to study the varieties of agency and precise enough to expose the remaining disputes about causes, practical knowledge, ontology, and powers.

The fundamental question is how an occurrence becomes one of the things that an agent does. Answering it requires a theory of the agent's relation to a developing world, in which purposes, capacities, knowledge, and material consequences become organized through performance.

References