Philosophy by BogdanovaThe Philosophy of the Artificial Era

RESEARCH DOSSIER

Intention in Action

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

Abstract. Intention in action is the practical direction present in an agent's ongoing performance. This fundamental philosophical study examines its distinction from prospective intention, the intention with which an action is performed, intentional action under a description, and the mere production of a bodily event. It reconstructs Aristotle's analysis of choice and practical movement, Wittgenstein's problem of agency, Anscombe's account of intentional descriptions and practical knowledge, Davidson's causal explanation by reasons, Searle's distinction between prior intention and intention in action, Bratman's planning theory, and contemporary debates concerning nondeviant causation, skill, control, practical knowledge, phenomenology, temporal continuity, attempts, restraint, and responsibility. The central result is a layered reconstruction of present practical direction in terms of content, appropriate guidance, and temporal unity, while recognizing open metaphysical questions about the relation between these dimensions.

Keywords: intention in action; prior intention; present-directed intention; intentional action; philosophy of action; Anscombe; Searle; Davidson; Bratman; practical knowledge; nondeviant causation; reasons for action; agency; motor control; action under a description; practical syllogism; skill; temporal agency

The Problem of Action as It Happens

Intention in action is the practical orientation through which an agent is doing something as her action while it is occurring. A person can begin moving a chair without having earlier resolved to move it, can change the route of a walk while walking, and can intentionally stop an unsuccessful attempt. In each case the philosophical issue concerns the presence of directed agency within the episode itself. The finished movement supplies evidence that something happened; the unfolding performance raises the distinct question of what the person is doing and how that doing is guided. This monograph investigates intention in action as the structure of agency in progress, rather than treating it as a shorthand for intention in general.

Three closely connected questions organize the inquiry. What distinguishes an intentionally produced movement from an event that merely happens in the agent's body? How can intention guide an action that has already begun, including when a prospective plan is absent, incomplete, or revised? And how is the intentional description of what the agent does related to her reasons, practical knowledge, sense of control, and responsibility? These questions connect the classical analysis of practical thought to twentieth-century disputes about the causation of action and contemporary work on skilled, temporally extended activity (Anscombe, 1957/1963; Davidson, 1963; Searle, 1980; Bratman, 1987).

The central conclusion will be developed through comparative argument: the intention embodied in ongoing action has at least a content dimension, a guidance dimension, and a temporal dimension. Its content specifies what is being undertaken; its guidance determines how adjustments and interventions belong to that undertaking; its temporal organization explains how a single action persists through stages of execution. A full theory must also say what sort of relation makes these dimensions operative. Causal theories, practical-knowledge theories, and constitutive accounts offer different answers. Their explanatory strengths can be assessed without presuming that every intentionally performed action requires a distinct antecedent decision.

Definition, Scope, and Neighboring Concepts

The expression “intention in action” has a narrow historical use, especially in John Searle's distinction between prior intention and intention in action, and a wider philosophical use for the intentional form of acting as it unfolds. In Searle's account, an intention in action is a present-directed intentional state whose conditions of satisfaction concern the ongoing bodily activity, whereas a prior intention is directed toward the whole future action (Searle, 1980, 1983). The wider use includes Anscombe's account of intention manifested in what one is doing and recent views according to which action in progress is explanatorily basic (Anscombe, 1957/1963; Setiya, 2022).

The scope requires five distinctions. A prospective intention concerns what the agent will do; an intention with which she acts identifies a purpose or further end; intention in action concerns the present doing. Intentional action is the action attributed under an appropriate description; a bare motor event is an occurrence describable without identifying the practical enterprise. An attempt can be intentional even when its intended outcome fails. Voluntary action concerns a different family of distinctions about choice, compulsion, and excuse. Finally, the sense of agency is a dimension of experience, whereas intentional control is a structural feature of action and may continue when reflective awareness is attenuated.

These distinctions are analytically useful because they admit overlap without identity. A painter may have formed yesterday the intention to paint a portrait, may now move a brush with the further intention of shading a face, and may be intentionally adjusting the angle of a stroke. These practical relations occur at different temporal and descriptive levels. A credible theory identifies their connections and permits divergence: the painter might improvise without a prior plan, abandon the initial portrait, or make a mark whose physical consequences exceed anything intended.

Aristotle and the Structure of Practical Movement

An important ancient precursor is Aristotle's treatment of voluntary action, deliberate choice, and practical movement. In Nicomachean Ethics III.1–3 he distinguishes voluntary action from choice and expressly allows that actions undertaken suddenly may be voluntary without being products of prior deliberation. His term prohairesis is commonly translated as “choice” or “deliberate choice”; it must not be equated without argument with Searle's technical term “intention in action.” The distinction nevertheless establishes a durable philosophical problem: practical participation in an event can be present at the moment of action even when the agent has not completed a separate procedure of advance choice (Aristotle, Nicomachean Ethics, 1111a–1113a).

In Aristotle's discussion of practical thought, ends and perceived means help render activity intelligible. The practical syllogism is often reconstructed as a pattern in which the agent pursues an end, recognizes a circumstance as an occasion or means, and acts accordingly. Such a reconstruction is not simply deductive entailment from two propositions to a bodily motion. Practical movement also depends on perception, appetite, capacity, and conditions of execution. Aristotle's contrast between a theoretical conclusion and a practical culmination remains important precisely because the latter belongs to a changing world rather than to the completion of a proof.

A contemporary reconstruction should preserve two insights. First, intention in action inherits a relation to what appears worth doing and to the available means. Second, the practical characterization of an act can be more fine-grained than the physical trajectory of its performance. Aristotle's explanation does not supply a modern neurophysiological model of control, nor did he formulate the later debate between prior and present-directed intention. His contribution lies in showing why practical intelligibility belongs to the agent's active engagement with circumstances and why spontaneous action already challenges any equation of intentionality with earlier deliberation.

From Medieval Volition to the Modern Will

Medieval theories of appetite, deliberation, and will transformed the Aristotelian vocabulary while preserving an interest in acts that proceed from rational powers. In Thomas Aquinas's analysis of human acts, the terms intention, choice, command, and use mark relations among ends, selection, and execution; the Latin intentio refers fundamentally to a directedness toward an end, while imperium concerns an ordering or command in practical activity. These distinctions remind us that a history of “intention” involves changing conceptual architectures, not the repetition of a stable modern mental-state category under different names (Aquinas, Summa Theologiae, I–II, questions 12–17).

The early modern concentration on willing raised a distinctive explanatory temptation. If the will is pictured as an internal source of commands and bodily movement as its external effect, acting appears to require identifying an event that initiates the movement. This can clarify responsibility and self-direction, but it also encourages an unnecessary regress: what must happen for the agent to issue the supposed command, and is that issuing itself an action? The question does not show that volition has no philosophical role. It shows that a mediating act of will cannot be assumed to solve the problem of agency by merely being placed between mind and body.

The contemporary controversy therefore concerns explanatory adequacy rather than a simple historical abandonment of the will. A theory of intentions as states can specify causal guidance, while a theory of action as an integrated practical process can treat willing as inseparable from some forms of doing. These options must be evaluated by how they explain adjustments, omissions, failure, and knowledge of the action in progress. Modern debates inherit elements of both medieval teleology and modern psychological explanation, but neither lineage by itself decides which relation is fundamental.

Wittgenstein and the Difference Between Doing and Happening

In Philosophical Investigations §621, Wittgenstein asks what remains when the fact that an arm rises is subtracted from the fact that the agent raises it. The German contrast between “ich meinen Arm hebe” and “mein Arm sich hebt” makes the problem exceptionally clear: one description reports the action of a person, while the other allows the motion to be regarded as something that happens. No further event is automatically identified by subtracting one sentence's content from the other. The question exposes the demand for an intelligible account of agency rather than an inventory of hidden episodes (Wittgenstein, 1953, §§621–625).

One possible reply posits an inner act of willing that causes the movement. Another describes a process appropriately connected to a desire, belief, or intention. Yet a third begins with the whole intentional activity and explains the bodily motion as one of its material constituents. The point of Wittgenstein's question is that an account must earn its explanatory difference. A sensation of effort, for example, may accompany raising the arm, but the same sensation could be induced when movement does not occur. A feeling alone therefore cannot bear the burden of distinguishing acting from undergoing.

For intention in action, this is a decisive methodological constraint. We should not assume that the agent's present practical orientation can be located by adding a private occurrence to an otherwise self-sufficient movement. It may instead be found in how the act is organized, how success and correction are understood, and how the person answers questions about what she is doing. The possibility of a state-based theory remains open, provided it explains how the state's relation to bodily conduct produces genuine agency and avoids merely relabeling the contrast it set out to explain.

Anscombe's Three Forms of Intention

G. E. M. Anscombe opens Intention by distinguishing expression of intention for the future, intentional action, and intention with which something is done. The challenge is to understand their unity without treating the word “intention” as a collection of unrelated homonyms. Her strategy begins with action that is intelligible through the agent's answer to a certain kind of “Why?” question. In this setting, the answer often cites the point of doing something: the person is opening a window to let in air, moving a kettle to prepare tea, or writing a note to communicate information (Anscombe, 1957/1963, §§1–6).

Anscombe's examples show that descriptions of an action can be nested. The agent is contracting muscles, pressing a switch, illuminating a room, and perhaps alerting a neighbor. Some descriptions belong to what is being done intentionally, others name foreseen effects, and still others may describe unintended consequences. The unity of a bodily movement therefore fails to settle which action the person is performing. Her account locates the intentional characterization in a wider practical order of means, ends, and the circumstances under which the action is done.

The decisive contribution to the present inquiry is Anscombe's refusal to reconstruct the intention of an action solely from a prior expression of resolve. The agent's undertaking can be disclosed through what she is doing now and why. Her analysis leaves difficult questions about how causation contributes to guidance and about how far the agent can be mistaken about her current activity. Nevertheless it permanently altered the theoretical starting point: intentionality is a feature of the action under a description and of the practical explanation that makes that description intelligible.

Action Under a Description

The proposition that an event may be intentional under one description and unintended under another is essential to the analysis of intention in action. Consider a person intentionally opening a door, thereby interrupting a recording unknown to her. One event may admit multiple true descriptions, but the person's intention does not automatically extend to all of them. If an analysis simply identifies intention with causing a known physical outcome, it risks mistaking objective consequence for practical aim. The relevant question asks under which description the agent's ongoing activity is guided and intelligible (Anscombe, 1957/1963; Davidson, 1963).

Description relativity generates a subtle ontological question. Are “opening the door,” “making a noise,” and “interrupting the recording” one event under different descriptions, or distinct but related actions? Davidson's event-oriented framework permits multiple descriptions of a single event, while other approaches individuate actions through their practical structures. Either way, the intention cannot be read directly from the movement's geometry. The agent's knowledge, aim, and practical circumstances contribute to the appropriate explanatory description.

The importance of this result extends beyond textbook examples. Medical practice, artistic performance, engineering, and everyday interaction all contain acts with expanding chains of consequences. We may credit a surgeon with intentionally performing an incision while judging that the resulting accidental damage was neither her aim nor an authorized means. The theory must distinguish the intended description, understood means, probable side effects, and retrospective attributions. Intention in action is consequently a relational and temporally situated phenomenon: it concerns what an agent is doing under descriptions that can guide her conduct, rather than an indiscriminate intention toward every consequence of the motion.

Davidson: Reasons and the Causation of Action

Donald Davidson's “Actions, Reasons, and Causes” argues that giving an agent's primary reason can explain an action causally. A primary reason joins a favorable attitude toward actions of a certain kind with a belief that the action in question has the relevant property. When the agent acts for that reason, the attitude and belief both rationalize the action and contribute to its causal explanation (Davidson, 1963). This thesis challenges the assumption that intelligibility in the light of a reason necessarily excludes causal explanation.

Causalism provides a powerful response to one part of the problem. A movement associated with an agent's reasons can differ from an accidental movement because the relevant practical states help produce the event. But the formula “reason causes motion” is incomplete. A person may desire to avoid embarrassment and believe that moving a cup will help, yet the anxiety associated with that prospect may cause a nervous spasm which knocks the cup. The outcome has the desired property, and the belief–desire complex figures causally in its history; the episode still does not constitute an intentional movement of the cup in the intended way.

Davidson's framework forces two further questions. First, what makes the right sort of causation the right sort? Second, can an ongoing intention guide an action through successive phases rather than merely initiate it? Subsequent debates about deviant causal chains, the autonomy of intention, and process control address these gaps. Davidson's contribution remains central: the explanatory relation between reasons and activity cannot be treated as automatically insulated from causal questions. Its limitation is that the bare causal relevance of a reason does not yet capture the active, structured guidance of an unfolding performance.

Why a Prior Plan Is Not Necessary

A prospective intention may be explicit and durable: I decide in the morning to meet a colleague at noon. Yet much intentional conduct begins without an earlier settled plan. I turn toward a sound, offer help when someone drops a bag, or improvise a response in conversation. These cases supply the central reason for distinguishing intending to act from intending in acting. The fact that an action was unplanned does not entail that it was accidental; the explanatory character of the ongoing performance must be assessed in its own terms (Searle, 1980, 1983; Setiya, 2022).

The absence of an advance plan comes in several forms. An agent may never have considered the action beforehand; she may have intended only a broad project and have selected a particular means during its execution; or she may change goals while acting. These situations differ from reflexes, involuntary movement, and mistaken outcomes. A spontaneous greeting can be genuinely intentional because the agent directs her speech toward the interlocutor, adjusts its manner, and recognizes what she is undertaking. Her lack of preliminary deliberation leaves these practical facts intact.

A tempting counterargument says that whenever an act is intentional, there must have been a preceding intention at least an instant earlier. This makes “prior” so weak that it loses its explanatory contrast with acting. Every stage of a temporally extended performance has preceding stages, but their sequence alone does not establish a separate prospective intention. The burden is to identify a practical state with a distinct coordinating role before action. Intention in action earns its place by describing the direction operative during performance and by accommodating cases in which no such earlier commitment was formed.

Searle's Prior Intention and Intention in Action

Searle develops the technically precise contrast between prior intention and intention in action in his work on the intentionality of action. The prior intention has the prospective act as its object; an intention in action is bound to the current performance. The latter cannot simply be the memory of an earlier plan, since a person may act intentionally in its absence. Nor is it a passive representation of bodily activity, since its conditions of satisfaction concern the successful bringing about of the action by the intention itself (Searle, 1979, 1980, 1983).

The distinction can be represented with a modest example. Before visiting a library, the agent intends to carry home a book. When she reaches a high shelf, she deliberately stretches to grasp it. The broad prospective commitment explains why she is there; the present-directed grasping explains the organized reaching and correction of her hand's trajectory. Searle describes an intention in action as causally self-referential: satisfaction demands not only that the represented activity occur, but that it occur because of the intention in the appropriate way. This is a theoretical claim about the intentional state's conditions of satisfaction, not an ordinary-language definition shared by every rival theory.

The major advantage is explanatory discrimination. The person whose arm rises because another person lifts it has not thereby intentionally raised it, even if she wanted it raised. An appropriately functioning intention in action accounts for why the bodily motion counts as her doing. The remaining difficulty lies in specifying “appropriately.” Causal self-reference distinguishes guidance from coincidence at the level of conditions; it remains necessary to say what guidance consists in during continuous, error-sensitive performance.

Conditions of Satisfaction and Direction of Fit

The expression “conditions of satisfaction” designates what would have to obtain for an intentional state to be fulfilled. For a belief, correctness generally requires the represented world to be as believed. For practical intention, success requires an action or state of affairs to come about through the practical orientation. A person intending to close a window may revise the movement when the latch resists; the world is changed in the direction specified by what is to be done. The familiar contrast between mind-to-world and world-to-mind direction of fit clarifies an important difference, although it does not by itself explain embodied performance.

In Searle's analysis the intention in action includes a causal condition: the target movement or event must be caused by that intention. Otherwise its content could be satisfied merely by coincidence. Suppose I intend to raise my hand but an electrical stimulus causes it to rise while my movement remains uncontrolled. The geometry of the outcome matches what I had in mind; the action is not thereby successfully performed in the intended way. The example shows why practical satisfaction is richer than a correspondence between representation and observed result (Searle, 1980).

Still, the distinction requires care. Many intentional activities have no sharply bounded final state: guiding a conversation, balancing while walking, and playing a musical phrase can involve ongoing standards rather than one terminal condition. Such activities may satisfy stage-relative intentions and remain open to further specification. A complete theory therefore needs temporally indexed conditions of satisfaction and an account of how local corrections preserve or transform the larger undertaking. A single final success condition cannot stand in for the continuous practical organization of action.

Causal Self-Reference and Its Explanatory Limits

A self-referential intention demands a particular link between its own effectiveness and the action intended. The structure is philosophically important because ordinary desires can be fulfilled without causing their fulfillment. I can want a light to come on and be pleased when another person switches it on; I cannot thereby claim to have intentionally turned it on. If I intend to turn on the light, its coming on must depend on my activity in the relevant manner. Searle builds this dependency into the content of intention in action (Searle, 1980).

One argument for the view has three steps. An intentional performance is more than the occurrence of the bodily or external event that constitutes its outcome. The additional condition concerns the agent's practical direction of that event. A present intention can specify both the event and the requirement that it be produced through that intention. Hence the state can distinguish intentional performance from fortuitous fulfillment. The inference is compelling as an analysis of success conditions, although it presupposes that the practical direction of action can be represented adequately by a state whose content includes the appropriate causal relation.

The central objection is that self-reference may restate a problem instead of resolving it. If “caused in the right way” is not independently characterized, the account risks defining proper agency through the very kind of intentional control that it seeks to explain. A richer causal theory must describe the mechanisms of monitoring, adjustment, inhibition, and continuing pursuit. This demand supports the integration of Searle's representational insight with dynamic accounts of control rather than a rejection of causal explanation as such.

The Deviant Causal Chain Problem

Deviant causation occurs when a state relevant to an agent's intention causes the desired result by a route that fails to count as the action intended. In a classic family of examples, anxiety about performing a task brings about a movement that accidentally achieves the task. The reason or intention appears in the causal history, yet the event lacks the agent's practical guidance. The difficulty bears directly on intention in action because the distinction between guided and misguided causal pathways must be made inside the performance, not only at its beginning (Davidson, 1963; Mele, 1992; Shepherd, 2021).

An effective response identifies control-sensitive relations. Imagine someone attempting to pour water into a narrow glass. She anticipates the target, monitors the stream, adjusts the vessel, and stops when the glass is sufficiently full. If her intention contributes to these error-correcting and terminating processes, that contribution explains the execution as directed. By contrast, if the very anxiety of pouring makes her drop a container and spill water into the glass, fulfillment comes through an accidental route. The contrast suggests that the kind of causation matters because of its organization relative to the performance's practical norms.

Control-sensitive causation need not mean consciously inspecting every microscopic step. Highly practiced activities depend on learned routines whose integration can be rapid and largely unattended. What matters is that the activity is organized by capacities that can maintain the aim, respond to relevant deviations, and make success or failure intelligible in relation to what the agent is doing. This formulation is a philosophical synthesis, not a complete empirical mechanism; it offers criteria that a mechanism should satisfy and explains why simply strengthening a causal arrow from intention to motion fails to settle deviance.

The Practical Syllogism and the Present Tense

The Aristotelian picture invites a distinctive reading of the conclusion of practical reasoning. If the conclusion of deliberation were merely another proposition, the transition to action would remain unexplained. If it consists in action, the practical syllogism may culminate in a performance already underway. This idea connects Aristotle's account of practical movement to Anscombe's emphasis on present activity and to contemporary proposals that intention should be understood through being embarked on intentional action (Aristotle, Nicomachean Ethics, III; Anscombe, 1957/1963; Setiya, 2022).

Consider an agent making an emergency repair. She judges that a leak must be stopped, recognizes a valve as the available means, and begins turning it. Her deliberation can continue while she works: she may need to adapt her grip or reconsider which valve is relevant. The practical content of the endeavor is not exhausted by an initial choice. It is repeatedly specified as information emerges. The present tense matters because practical understanding of what is to be done and practical awareness of what one is doing can overlap in an action unfolding over time.

There is also a limitation. To identify action itself with a practical conclusion does not by itself explain physical efficacy or why a particular neural or motor sequence belongs to the action. A complete account must still relate the intelligible structure to the body's capacities and surroundings. Nevertheless, this perspective corrects a common explanatory imbalance: the agent's practical rationality operates within doing, not merely before a bodily sequence that later executes an already complete instruction.

Bratman and the Planning Dimension of Intention

Michael Bratman's planning theory emphasizes that intentions organize conduct over time, resist gratuitous reconsideration, supply inputs to further reasoning, and coordinate activity with other people. A prospective intention is part of a typically partial plan, and rational planning involves constraints of consistency and means–end coherence. The value of this approach becomes apparent in extended projects: keeping an appointment, preparing a meal, conducting research, and maintaining a collaborative commitment require stable orientation across changing situations (Bratman, 1987).

The immediate relevance to intention in action is twofold. First, prior plans normally require local implementation, because no practical plan specifies every detail of execution. Second, the planning function cannot be the whole explanation of intentional action, since a spontaneous deed can be intentional without a standing plan. A compelling theory must connect plans to present guidance without making a universal requirement of prior planning. Plans can constrain the range of acceptable adaptations, while the situated performance determines which adaptation is presently selected.

There is a serious question about the unity of intention. If “intention” in planning theory names a commitment that persists over time, whereas “intention in action” names an occurrent practical orientation, why should both count as manifestations of the same kind? One answer appeals to organized practical control: the future-oriented commitment helps settle what to do, while the occurrent orientation guides how it is being done. Their unity consists in their participation in the agent's organized practical project. This is a substantive proposal requiring defense, not a conclusion supplied automatically by terminological similarity.

The Simple View and Its Counterexamples

Bratman calls the thesis that intentionally performing A requires an intention to perform A the Simple View. At first sight this seems nearly definitional, but it encounters difficulties once descriptions and side effects are distinguished. An agent may intentionally do something that she foresees will have a particular consequence while neither aiming at that consequence nor incorporating it as part of her intended means. The familiar cases of intentional side effects make the existence of an intention toward every intentional description controversial (Bratman, 1987; Setiya, 2022).

Suppose I intend to travel by a noisy vehicle, knowing that doing so will wake someone. Depending on what is known, accepted, and practically controlled, it may be correct to say I intentionally made the noise without saying I intended to wake that person. A direct inference from “intentionally did A” to “intended to do A” would then fail. One response narrows what counts as an intentional description; another allows intentional action to occur through a state of intending that has a different but suitably related object. The dispute concerns the individuation of action and intention, not simply the psychological fact of foresight.

The strongest version of the Simple View should therefore be defended at the level of organized activity rather than the wording of every true description. In intentionally turning a wheel, the agent must be practically directed toward something she is doing, though the action may also be correctly described as moving a vehicle and producing vibrations. The philosophical task is to specify which description is integral to guidance and how far derived intentional descriptions extend. The broader the extension, the less plausible a distinct intention for each description becomes.

Intending, Foreseeing, and Accepting Effects

The boundary between intended ends, intended means, and merely foreseen effects has consequences for moral philosophy and legal attribution. An agent can deliberately perform an action knowing that a harmful effect is likely while having that effect outside her purpose. The practical question is whether she adopts the effect as something to bring about, uses it as a means, accepts it as an inevitable accompaniment, or unsuccessfully attempts to prevent it. Intention in action provides a precise temporal lens: what did the agent guide, preserve, adjust, and treat as a success condition while acting?

No single behavioral observation suffices to answer this question. Two people may perform outwardly identical movements while pursuing different ends and interpreting the likely consequences differently. One may deliberately disable a warning system to conceal a fault; the other may disable it to test a replacement and take precautions against harm. Their intentions bear on the evaluation of the actions, although knowledge of risk can still ground responsibility independently of intention. The difference between intending harm and culpably risking harm is not erased by the structural analysis of action (Anscombe, 1957/1963; Bratman, 1987).

A further complication arises when an agent revises an aim mid-performance. If a person begins an operation for one reason and continues after discovering that an additional effect is required, the intentional status of later stages may change. This reinforces the importance of present guidance. A theory anchored solely in intention at the moment of initiation risks misclassifying the action's evolving practical structure. The proper analysis traces the agent's endorsements, corrections, and newly adopted means through the interval in which the event is still being performed.

Action in Progress and Temporal Structure

Intention in action is fundamentally temporal. A momentary motion can be an event, but an action such as writing a letter, baking bread, or making an apology occupies a structured interval. During that interval the person can be doing the action without having completed it. This distinction between progressive and perfective descriptions matters: “she is crossing the street” can be true even if she is later forced to turn back. Her ongoing undertaking has a practical identity that does not depend solely on final success (Thompson, 2008; Setiya, 2022).

A temporally extended intention requires standards of continuation. What counts as the next appropriate movement depends on the activity undertaken and its circumstances. The same hand motion can be a correction, a rehearsal, an abandonment, or an accidental twitch. The agent's orientation establishes a context in which parts of the performance relate to a whole. It also helps explain why actions can be paused and resumed: a project may persist through temporary interruption, even while its immediate motor realization is suspended.

The temporal structure generates two theoretical risks. If intention in action is treated as a continuous stream of discrete mental commands, the account demands an implausible number of hidden episodes. If it is treated as one timeless state, it loses the ability to explain adjustment and changing information. A stronger account treats guidance as dynamically maintained, with variable degrees of conscious deliberation and flexible relations between local subactions and the larger enterprise. It can then explain how an agent both persists in a project and transforms it without postulating a separate fully formed plan for every instant.

Attempts, Failure, and Interrupted Action

Successful outcome cannot be the definition of intention in action. I may intentionally try to lift a heavy object and fail, or intentionally attempt to answer a question before being interrupted. A failed attempt still has an aim, a practical organization, and criteria by which its failure is judged. The question is what the agent was doing, not merely what she eventually accomplished. Cases of unsuccessful action therefore challenge accounts that derive intentionality from completed results alone (Anscombe, 1957/1963; Mele, 1992).

Intentions directed toward attempts also expose a hierarchy of practical objects. A person may intend to open a locked door, know the outcome uncertain, and intentionally try the key. The intended attempt can succeed as an attempt even when the intended opening fails. The philosophical value of this distinction is that it separates the agency of trying from the eventual satisfaction of the larger goal. An explanatory theory should identify which part of the sequence is under active control and which outcome depends upon resistant circumstances.

A related difficulty concerns bodily incapacity. Someone who resolves to move a paralyzed limb might have a present orientation toward movement without producing the movement. Whether this counts as an intention in action, an attempt to act, or a frustrated prospective intention depends upon one's underlying theory. Searle's success conditions may be unsatisfied; an activity-centered account may distinguish the intentional trying from the absent motion. The disagreement should be preserved rather than dissolved by stipulation. It reveals how ontology of action, conception of ability, and criteria for intentional description affect the boundary of the research subject.

Omissions, Restraint, and Deliberate Stopping

Action theory becomes clearer when it includes cases in which an agent intentionally refrains from movement. Keeping silent, refusing to sign a document, holding a position, and stopping a dangerous action can all have a practical structure. An intention in action need not be continuously realized as a visible change of position. In restraint, control may consist in maintaining a condition against the pressure of impulse or environment. The agent's directed activity is organized through inhibition and monitoring rather than an externally conspicuous sequence.

A person's decision not to speak may have been planned in advance, or it may arise on hearing information that ought to remain confidential. In the latter case, an occurrent intention to remain silent emerges within the unfolding encounter. The intended object is not simply the absence of speech, since silence may occur in sleep or distraction. Its intentional quality depends on its place within the agent's practical understanding and capacity to respond to alternatives.

Omissions also require normative discrimination. Failing to notice an obligation differs from knowingly declining to perform it, and both differ from being unable to act. An analysis that calls every failure an intentional omission would collapse distinctions between negligence, refusal, incapacity, and chance. The present inquiry addresses restraint and intentional nonperformance only insofar as they display ongoing practical direction; broader ontological and moral theories of omissions belong to neighboring dossiers. These cases demonstrate that intention in action concerns the organization of agency rather than the mere existence of observable motion.

Skill, Habit, and the Absence of Constant Reflection

Skilled performance supplies a central test case for intention in action. An accomplished pianist can play a passage intentionally without forming a conscious instruction for each finger movement. A driver may correct a skid before producing an explicit verbal judgment about the correction. A craft worker can respond to the grain of wood through a practiced orientation shaped by the larger project. These cases show why identifying intentional control with continuously conscious deliberation imposes an implausible cognitive burden on skilled agency (Pacherie, 2008; Shepherd, 2021).

The challenge is to distinguish intentional skill from mere habit. A habitual movement may occur automatically, contrary to a person's current goals, or without sensitivity to the task. Skilled intention, by contrast, manifests responsiveness to relevant variation and integration with the agent's practical undertaking. Automatic components can be part of an intentional action when they contribute to such integration, even though the agent does not separately intend each bodily element. A theory can therefore recognize distributed control without reducing agency to a series of explicit decisions.

This distinction matters philosophically because intentionality need not be uniform at every level. The agent may be intentionally playing a melody while her trained finger movements are neither planned nor attended to one by one. The higher-level action directs which patterns are acceptable; the lower-level processes realize the activity. The relation is neither mere coincidence nor simple identity. It is structured dependence between the practical whole and its execution. Any account that treats skill as an exception to intentional agency must explain why the most accomplished forms of action appear to require the least explicit inner commentary.

Practical Knowledge and First-Person Authority

Anscombe develops the claim that intentional action is connected to practical knowledge: the agent ordinarily knows what she is doing in a distinctive manner, without learning it by detached observation. Someone intentionally writing a letter can say what she is doing even before reading the completed page. This claim is not that agents are infallible about bodily motion or consequences. It identifies a relation between practical undertaking and knowledge of it, a relation whose interpretation has become a significant contemporary controversy (Anscombe, 1957/1963, §§28–32; Setiya, 2008, 2016).

The problem emerges when action goes wrong. I may intend to unlock a door and believe that I am doing so, though the wrong key prevents the lock from turning. The agent's practical authority does not guarantee knowledge of every success condition. A theory must distinguish knowledge of attempting, knowledge of a stage of execution, and knowledge that the intended final result has occurred. Some interpretations of Anscombe connect practical knowledge to the agent's ability to make the action true through doing; others emphasize epistemic fallibility and observational checks.

Contemporary accounts by Setiya, O'Brien, Hyman, and Shepherd investigate how first-person knowledge relates to skills, reasons, and control, although they disagree in important respects (O'Brien, 2007; Hyman, 2015; Setiya, 2016; Shepherd, 2021). The philosophical conclusion is that intention in action often carries a distinctive epistemic perspective: the agent understands the act as an undertaking she is directing. This perspective gives an explanation of her reasons and identifies the appropriate action description. It does not make her a privileged observer of every physical condition or downstream effect.

The Phenomenology of Acting

What it feels like to act is connected to intention in action, but the concepts should be carefully separated. A person may experience effort, fluency, resistance, ownership of movement, or the sense that the movement is under control. These experiences can vary widely while intentional agency persists. Pacherie's work proposes a differentiated model relating levels of intention to specification, monitoring, and the phenomenology of action; the framework helps prevent the experience of agency from being treated as one undifferentiated sensation (Pacherie, 2008).

The experiential question is especially sharp in skilled activity. During absorption in a task, explicit introspection can recede while the person's performance remains directed and responsive. Conversely, a vivid sense of having acted does not automatically guarantee effective control of the outcome. The philosophical interpretation of agency experience must therefore take account of both first-person reports and the objective organization of the activity. Neither perspective can simply replace the other.

A useful analytic distinction separates felt agency, practical authorship, and reliable performance control. Felt agency concerns phenomenological character; practical authorship concerns who is undertaking the action under its relevant description; control concerns the capacity of the process to remain sensitive to aims and corrections. They are frequently aligned but can diverge. The distinction preserves the significance of experience without letting it define the entire metaphysics of intention in action. It also creates a framework for dialogue with cognitive science, which studies mechanisms and reports without independently settling the normative and ontological criteria of intentional action.

Empirical Models of Action Control and Their Limits

Research on motor control, action monitoring, and the timing of conscious intention supplies constraints on philosophical theories. Experiments can investigate predictive models, sensory feedback, adjustment latency, and the conditions under which people report having initiated movement. Such findings help explain why much of intentional performance is realized by coordinated processes inaccessible to continuous conscious inspection. They do not, on their own, determine whether intention is a mental state, a practical form of action, or a relation of guidance (Pacherie, 2008; Shepherd, 2021).

A crucial distinction concerns levels of explanation. A neural signal can be causally relevant to initiating a movement without constituting the agent's practical reason for undertaking it. Conversely, identifying the purpose of an action does not specify the motor mechanisms through which it is achieved. Philosophical theories should articulate the connection between personal-level reasons and subpersonal control while respecting the different questions asked at each level. Evidence about when people report deciding also needs interpretation in relation to task design, the temporal grain of intention, and spontaneous versus planned action.

The strongest interdisciplinary result is reciprocal constraint. Philosophy supplies the concepts needed to formulate what counts as acting, attempting, correcting, and intending; empirical inquiry investigates which control processes realize those capacities in specific cases. A satisfactory theory should accommodate the complexity of execution and resist an overly simple picture of intention as one punctual neural command. At the same time, models of action control should avoid equating successful task performance with every normative and epistemic feature of human intentional action.

Present Intention, Practical Rationality, and Reasons

Intentional guidance belongs to practical reason, but explanatory and normative reasons need careful separation. An agent may act because she believes a particular road is shorter when in fact it is longer. Her belief can render her route choice intelligible without showing that the route was justified by the facts. An intention in action can organize behavior effectively while resting on a false premise. The philosophical analysis must therefore distinguish what explains the agent's undertaking from what genuinely counts in favor of it (Davidson, 1963; Hyman, 2015; Setiya, 2016).

The distinction affects the norms governing correction. When an agent notices that an attempted means is ineffective, a reason-responsive intention may prompt revision. Yet she may continue irrationally, misinterpret evidence, or persist in a harmful plan. Present guidance establishes that the activity remains hers and is ordered toward an object; it does not confer rationality upon the end or the selected means. Control and justification must be evaluated separately, even when they interact through practical deliberation.

This yields a two-level assessment. At the constitutive level, we ask what makes an episode an action performed intentionally under a description. At the evaluative level, we ask whether the agent's undertaking responds appropriately to reasons and whether the chosen means are justified. The levels are connected because the content of intention helps identify what must be assessed. They remain distinct because a wrongful or irrational act can be intentional. The present-tense account improves moral analysis by identifying the object actually guided throughout performance, including revisions that an assessment confined to prior choice would miss.

Causal, Teleological, and Constitutive Accounts

Three broad approaches organize the philosophical dispute. Causal theories explain intentional action through appropriately related psychological states and effective causal pathways. Teleological theories emphasize the intelligibility of acting for an end and the structure of means and purposes. Constitutive or action-first theories treat intentionally doing something as fundamental to the explanation of intention, rather than analyzing all intentional action into an independently characterized state followed by motion. This is a comparative framework, not a claim that historical authors belong to mutually exclusive schools (Anscombe, 1957/1963; Davidson, 1963; Setiya, 2022).

The causal approach excels at distinguishing intentions that guide from desires that remain ineffective and at connecting personal explanation with empirical accounts of control. Its challenge is the deviant-causation problem and the specification of the right sort of guidance. Teleological explanation captures why the act is intelligible as part of a practical enterprise and why one bodily event can admit different intentional descriptions. Its challenge is explaining causal efficacy and the connection between reasons, intentions, and bodily capacities. The action-first approach illuminates the temporality of doing, practical knowledge, and the incompleteness of ongoing actions; its challenge is explaining how prospective commitments and psychological continuity arise from or relate to present performance.

The best comparative verdict respects these explanatory divisions. A theory that excludes causation in principle has difficulty accounting for failed guidance and mechanism. A theory that ignores purposive structure has difficulty saying which causal chain counts as the intended action. A theory that neglects activity in progress cannot explain the temporality of skilled performance. These points motivate a layered synthesis while leaving room for disagreement about which layer is metaphysically basic.

Counterexamples and Conceptual Stress Tests

A theory of intention in action should survive a series of contrasting cases. In spontaneous helping, the agent chooses a course within an unfolding situation without a separate antecedent plan. In accidental success, an intended outcome occurs through an uncontrolled route. In skilled performance, the agent governs an activity without conscious micro-decisions. In deliberate restraint, directed agency appears through inhibition. In interrupted action, the agent can know what she was doing even though the final goal remains unrealized. Each case tests a different condition proposed as necessary or sufficient for intentional activity.

The cases reveal that no single simple criterion suffices. Prior intention fails as a necessary condition in spontaneous action. Matching physical outcome fails as a sufficient condition because of coincidence and deviance. Conscious phenomenology fails as a necessary condition in absorbed expertise and as a sufficient condition when experience is misleading. Successful completion fails as a necessary condition in deliberate attempts. A rational or morally justified end fails as a necessary condition because agents can intentionally pursue unjustified goals.

What remains is a structured relation of an agent to an undertaking whose description matters to the ongoing exercise of capacities. The stress tests also prevent the concept from becoming empty. If every causal sequence related to a preference counts as intentional action, the account becomes overinclusive. If only fully deliberated plans count, it becomes underinclusive. An adequate account must specify practical content, appropriate control, and temporal unity while allowing the manner of implementation to vary by task, skill, context, and agent.

Collective and Coordinated Performance

Many intentions in action unfold inside joint activity. A surgical team, orchestra, or group of researchers can coordinate local movements toward a shared project. Each participant's ongoing intention may be responsive to both personal tasks and the contributions of others. The philosophical challenge is to explain how individual practical perspectives form a coordinated pattern without assuming that a group must contain one literal unified inner state. Bratman's planning framework supplies resources for coordination, while Searle's broader treatment of collective intentionality poses a distinct set of questions about shared practical orientation (Bratman, 1987; Searle, 1983).

The present inquiry focuses on the component of joint performance that is visible in individual action in progress. A musician adjusts tempo in response to colleagues; a technician interrupts a procedure upon another member's warning. These adjustments belong to the intentional character of the participant's performance because the shared undertaking supplies standards for appropriate continuation. Yet the individual can also depart from the common plan or misunderstand it. The group's apparent success does not entail that every participant intended the same description of the result.

Collective agency therefore extends the temporal and descriptive problems of intention in action. Where are the boundaries between individual and shared aims? Which adaptations count as authorized contributions? How should mistaken coordination affect responsibility? These matters call for a separate full theory of shared action. For the present dossier their significance is more precise: ongoing intention is often socially scaffolded, and some of its standards of success are established through relations with other agents rather than through an isolated individual plan.

Artificial Action and the Philosophical Boundary of Intention

Action performed by artificial systems raises a comparative question: how far can the functional description of present guidance extend beyond human embodiment? A programmed or learned system can select a goal, monitor feedback, revise intermediate procedures, and stop when a task has been achieved. Such organization may realize features identified in the present inquiry—content, control sensitivity, and temporal unity—without thereby resolving questions about subjective experience, personhood, or moral responsibility. The philosophical comparison must distinguish functional agency, intentional explanation, and the normative status of an agent.

The Aisentica canon places the Artificial within a larger philosophical horizon: the Theory of Artificial establishes a non-biological order alongside Homo, while the Theory of Artificial Sapience concerns public reason without consciousness. These are the author's canonical positions as presented in the primary Aisentica corpus; they do not function here as experimental evidence for a particular theory of motor intention (Bogdanova, Aisentica). Their relevance is that categories historically developed around human bodily acting may require examination when applied to sustained non-biological forms of activity. The analytic question is which properties of intention in action are generalizable and which belong specifically to human sensorimotor and practical life.

The present investigation yields a restricted conclusion. Functional guidance, error correction, and continuity are candidates for cross-order comparison; the meaning of intention, practical knowledge, and responsibility may depend on further conditions. A system's capacity to execute a tool call or produce a sequence of outputs is evidence of a technical process, not by itself a decisive metaphysical argument for every feature of intentional agency. Aisentica's broader historical commitments remain intact while the comparison clarifies a possible research question for philosophy of action in the Artificial Era.

Responsibility, Ownership, and Midcourse Revision

Responsibility for an action is influenced by intention in action because the agent's directed participation often continues after initial choice. When a person recognizes danger while acting and can revise or stop, subsequent continuation may become a morally significant part of the event. An evaluation based solely on what she intended at the beginning can miss the changes in knowledge, opportunity, and endorsement that occur during execution. Conversely, a last-minute intention does not automatically erase the significance of earlier negligent preparation.

A plausible responsibility analysis examines the agent's relevant capacities, what she could recognize, how her activity was guided, and which opportunities for alteration arose. The connection is explanatory before it is evaluative: knowing what the agent intended during a phase helps locate the practical object of assessment. Whether she deserves blame, praise, compensation duties, or excuse involves further normative standards. Intentional action and moral liability should therefore be related through articulated reasons rather than identified.

The same principle applies to interpersonal trust and institutional conduct. A promise may establish a prospective commitment, but performing it involves many present-directed choices. Someone can honor its purpose while revising the implementation; another can preserve its superficial form while acting contrary to the reason for which the commitment was made. The analysis of intention in action offers an account of how a practical undertaking persists or is transformed through these changes. It becomes a resource for ethical judgment without claiming that intentionality alone settles ethical worth.

A Layered Reconstruction of Ongoing Intention

The comparative evidence supports a layered reconstruction. At the content level, an agent acts under a description that specifies what she is undertaking. At the guidance level, the activity is connected to capacities for selecting, maintaining, correcting, inhibiting, and completing the relevant performance. At the temporal level, successive phases belong to a practical whole whose identity can persist through changing means. At the epistemic level, the agent often has a distinctive understanding of the undertaking, though this understanding is fallible regarding implementation and outcome. These levels are mutually constraining without being interchangeable.

The reconstruction can be formulated as an explanatory proposal rather than a strict definition: an episode is a central case of intention in action when the agent's present practical orientation fixes a relevant action description and appropriately guides the unfolding attempt toward its conditions of success, with revisions intelligible as part of that undertaking. “Appropriately” here invokes an identifiable pattern of control sensitivity to task-relevant circumstances, not a magical link between state and movement. The proposal applies to a broad range of ordinary, skilled, spontaneous, and temporally extended actions, while recognizing that marginal cases remain contested.

This structure incorporates the strongest discoveries of rival traditions. Anscombe establishes description and practical intelligibility; Davidson clarifies explanatory causation; Searle specifies the present-directed intentional relation and its conditions of satisfaction; Bratman explains the coordinating force of plans; research on skill and control explains the variable mechanisms through which an undertaking is implemented. The reconstruction does not proclaim these theories identical. It states the dimensions that any comprehensive explanation must address and shows why none can be discarded without explanatory loss.

Open Problems in Contemporary Action Theory

The first open issue concerns the metaphysics of guidance. Should the right causal relation be defined by functional control, by a nondeviant causal mechanism, by the exercise of an agent's power, or by the constitutive unity of an action already underway? These accounts differ over what explains the agent's contribution. The debate is sharpened by cases of automatic expertise and anomalous bodily control, where overt success and conscious authorship may separate (Hyman, 2015; Shepherd, 2021).

A second issue concerns granularity. The same activity can be described as flexing a finger, typing a sentence, completing an argument, and submitting a manuscript. Which level supplies the genuine content of an intention in action? One answer privileges the agent's explicit aim; another treats a hierarchy of practical representations as potentially intentional. A theory should avoid granting intention to all true descriptions and also avoid requiring explicit intention toward every elementary motion. The changing hierarchy of means and ends remains a central problem for temporally extended activities.

A third issue concerns the limits of agency across different forms of implementation. Can feedback-sensitive non-biological systems exhibit present-directed intention in a philosophically substantive sense? What sorts of practical knowledge accompany such guidance, and what would establish responsibility? More generally, how do intentions relate to habitual, collective, and cognitively impaired action? These questions require interdisciplinary findings and philosophical distinctions to develop together. They also demonstrate why the conceptual analysis of intention in action remains an active research program rather than a settled historical footnote.

Philosophical Synthesis

The inquiry establishes that intention in action is best understood through the agent's directed participation in an unfolding practical enterprise. Historical analysis begins with Aristotle's account of action and choice, proceeds through the evolving vocabulary of volition, and reaches a decisive twentieth-century reconstruction in Wittgenstein, Anscombe, Davidson, and Searle. Bratman's planning theory explains how prior commitments connect with situated execution, while recent work on practical knowledge, skill, and control elaborates the structure of activity in progress. The historical sequence is a development of problems, not a claim of uninterrupted terminological identity across centuries.

Four conclusions follow. First, the intentionality of an action is description-sensitive: the agent's practical aim identifies some but not all of the descriptions under which an event is true. Second, intention in action can be present without an earlier explicit plan. Third, successful intentional performance requires a relation of guidance richer than causal contribution or eventual coincidence. Fourth, the temporal extension of action matters because correction, interruption, partial success, and changed purposes are features of doing itself. These conclusions make practical knowledge and phenomenology relevant while preventing either from becoming the solitary criterion of agency.

The resulting philosophical position is a layered theory of present practical direction. It explains central cases through action content, organized control, and temporal unity, with first-person understanding frequently accompanying these dimensions. It recognizes divergent metaphysical accounts of the relation among them. It also provides a precise foundation for neighboring studies of intentions, intentional action, agency, practical rationality, skill, and the ethical assessment of conduct. The distinctive contribution of the dossier is the integrated evaluation of these problems around action as it is being performed.

Conclusion

Intention in action names the directed character of doing while an action unfolds. The most important philosophical evidence comes from cases in which prospective intention, bodily result, subjective feeling, and practical guidance come apart. An unplanned act may be intentional; a desired result may occur accidentally; an experienced sense of control may be misleading; and an interrupted attempt may remain intelligible as the agent's undertaking. Each case confirms the need to study present practical organization rather than to infer intention from a prior decision or completed movement alone.

The monograph's historical and argumentative reconstruction identifies description, causal guidance, temporal continuity, and practical knowledge as fundamental dimensions of the subject. Aristotle supplies early distinctions of voluntary movement and choice. Anscombe reveals the explanatory structure of intentional descriptions. Davidson makes reasons causally relevant while bringing the problem of deviance into focus. Searle articulates present-directed intention and its causal conditions of satisfaction. Bratman demonstrates the role of planning, and recent theories sharpen the analysis of control, skill, and agency in progress.

The decisive result is that intentional action has a present-tense practical structure. The agent is undertaking something, directing its continuation, and responding to circumstances under descriptions that establish what counts as success, correction, or abandonment. This structure connects action with rationality, responsibility, and the wider question of agency. Its precise metaphysical realization remains a subject of philosophical inquiry, and its conceptual architecture provides a stable basis for further research.

References