How the Mind Experiences Causing Action

Opening scenario

You press a key and an object moves across a screen. The intention was yours, your finger moved as expected, and the object responded immediately. Nothing about the event seems mysterious. You caused it.

Now introduce a delay. Make the object travel in a different direction. Let it sometimes move without a key press and sometimes remain still after one. Add an unseen system that can influence the result.

You still know that you pressed the key. But did you cause what happened on the screen?

The experience of causing an action usually feels direct. We do not ordinarily pause to infer that we raised an arm, opened a door or moved an object. Yet the mind does not have a detector that simply reads causation from reality. It combines information about what was intended, what movement was expected, what the body did, what happened next and what else might have produced the result.

When these signals align, authorship feels obvious. When they conflict, the experience can weaken, divide or change.

Authorship begins before movement

The foundations of authorship can be laid before the body starts moving.

A person has a goal, such as moving an object to the right. Different responses may be possible. The mind evaluates them, selects an action and prepares the required movement. In this way, the action becomes connected to a current direction: this is what I am about to do.

That intention need not take the form of a conscious sentence. Skilled and familiar actions often unfold without the person silently announcing each movement in advance. Goals, learned expectations and action preparation can organise behaviour before a separate intention becomes noticeable.

The ease of selection can also influence the developing experience of authorship. When an action is selected fluently, without much ambiguity or conflict, this can provide an early indication that behaviour is proceeding under control. When competing responses remain active, confidence in the action may be weaker.

These early signals provide only part of the answer. Fluent selection cannot guarantee that the body will perform the intended movement, and a clear intention cannot establish that a later event was caused by it. The outcome has not happened yet.

Intention places an action within a self-directed plan. It does not certify every consequence that follows.

This connects authorship with the processes described in Decisions Compare Values and Executive Functions Coordinate Behaviour. Valuation helps establish a direction. Executive coordination helps maintain the goal, select a response and manage alternatives. These processes contribute to authorship, but they are not identical to the experience itself.

Prediction accompanies movement

Preparing an action makes several predictions possible.

Before your finger reaches a key, the motor system can anticipate where it will go and what the movement should feel like. It may also predict immediate consequences: pressure against the fingertip, the sound of the key or a change on the screen.

Once movement begins, these expectations can be compared with incoming information. If the finger moves in the anticipated direction and the expected contact occurs, the correspondence supports authorship. If the hand moves elsewhere or visible feedback differs sharply from the performed movement, authorship may weaken.

This is the basic idea behind comparator accounts of agency: predicted consequences are related to sensed consequences, and their correspondence helps identify events as self-generated.

But prediction matching cannot explain the entire experience. People can retain authorship despite some discrepancies, particularly when their goal is clear and they remain engaged in controlling the movement. They can also experience control over events they did not cause when those events happen to fit their expectations.

Prediction is one contributor to authorship, not an infallible internal witness.

The body and world answer back

Once movement begins, prediction is joined by feedback.

Proprioception supplies information about the position and movement of the body. It helps you sense that your finger is extending, your arm is rising or your hand is turning even when you are not watching.

Touch reports contact, pressure and resistance. Vision shows whether the body and surrounding objects are moving as expected. Hearing can confirm a key press, a closing door or another immediate effect.

Interoception contributes a different kind of bodily information. It concerns the body's internal physiological condition, including effort, arousal and visceral sensation. These signals can shape the broader experience of acting, but they do not serve the same function as proprioception. Proprioception more directly informs the mind about where the body is and how it is moving.

Feedback also allows action to be adjusted. If a hand drifts from its target, the movement can be corrected. If an object is heavier than expected, force can change. If the environment shifts, the action can be redirected.

An action therefore need not unfold exactly as first predicted to remain experienced as self-caused. Monitoring and correction can help preserve the broader experience of controlling it, even when the original prediction was imperfect.

This bodily contribution connects with Interoception, while the coordination of goals, feedback and adjustment connects with Executive Functions Coordinate Behaviour. Together they reflect the account developed in The Many Processes Called the Self: the acting self emerges through coordination among several processes rather than from one isolated inner controller.

Causing a movement is not the same as causing its result

Suppose you press the key exactly as intended. You experienced the preparation, movement and contact. Your authorship of the key press is not seriously in doubt.

The object on the screen then moves.

It is tempting to treat these events as one question: did I do that? But at least three questions are involved:

• Did I produce this bodily movement?
• Did that movement contribute to the external outcome?
• How much of the outcome should be attributed to me?

The answers can diverge.

You may have caused the key press while software independently selected what happened next. The system may have delayed, modified or overridden your command. Another person may have sent a competing instruction. The object might have moved at that moment regardless of what you did.

In other situations, your action may be one genuine contributor without being the sole cause. You turn a handle while someone else pushes a door. Several people enter commands that jointly alter a shared system. You start a process whose result also depends on machinery, rules and changing conditions.

Authorship over movement can therefore remain strong while authorship over outcome is uncertain or partial.

This distinction is easy to miss because ordinary environments often make bodily actions and external effects correspond reliably. A switch turns on a light. A hand moves a cup. A spoken request produces a response. The chain feels continuous.

When the chain becomes unreliable, the mind must estimate where personal contribution ends and other causes begin.

Outcomes reshape authorship

The result supplies new evidence about what caused it.

One source is correspondence. Did the outcome resemble what was intended or expected? If you selected movement to the right and the object moved right, this supports the connection. If it moved left, disappeared or changed in an unrelated way, authorship may weaken.

Timing also matters. An outcome that appears immediately after an action is generally easier to connect to it than one that appears after an unusual delay. But temporal proximity is only a clue. Unrelated events can occur close together, while genuine consequences may take time to appear.

A stronger source of evidence is contingency: did the action change the likelihood of the outcome?

If the object almost always moves after your key press and rarely moves without it, there is strong evidence that the action controls the result. If it moves just as often when you do nothing, the fact that it once moved after your press provides much weaker evidence.

"It happened after I acted" does not necessarily mean "It happened because I acted."

The mind can learn these relationships across repeated encounters. At first, an unfamiliar system may produce uncertainty. With experience, the person learns which actions reliably lead to which effects, how long those effects usually take and when another process is likely to intervene.

Context affects the interpretation. The same delay may be normal for one system and suspicious in another. A result that ordinarily suggests personal control may become ambiguous when several agents can produce it. A familiar mechanical failure may explain a mismatch that would otherwise suggest lost control.

Outcome information can therefore modify an experience that began before movement. The mind is not merely checking a completed judgment. It is updating the relationship between intention, action and consequence.

Alignment makes authorship feel fluent

Authorship is often strongest and least noticeable when several signals converge.

The action fits the current goal. Selection is relatively fluent. The movement follows the intention. The body feels and looks as expected. Adjustment remains possible. The outcome occurs at an expected time, resembles the intended effect and reliably depends on the action. No stronger alternative cause is apparent.

Under these conditions, the mind has little conflict to resolve. The action simply feels like something the person did.

That fluency does not provide direct access to objective causation. A coincidental event may occur at exactly the expected time. A hidden system may produce the selected outcome without using the person's action. An interface may appear responsive while its underlying result is predetermined.

Because the event fits the person's intention and expectation, it may still feel self-caused.

Conflict reveals the construction

Different conflicts disturb different parts of an action.

The intended movement may differ from the executed one. The movement may occur correctly while visual feedback shows something else. The body may perform as intended while the external result fails. The expected outcome may arrive only after a delay, or it may occur frequently without the action. Another agent may intervene between movement and consequence.

Later information can also overturn the first interpretation. A person may initially believe their action caused an outcome and then learn that the system was disconnected. Or they may assume the action failed and later discover that its result was delayed.

These conflicts do not always eliminate authorship.

A visual distortion may be discounted when intention, proprioception and contextual knowledge agree. A delayed outcome may still feel self-caused when the delay is expected. An imperfect movement may remain authored when the person detects and corrects it.

At other times, the experience weakens. The action can feel less controlled, the outcome can seem externally generated, or the person may remain unsure how much they contributed.

Authorship is therefore graded rather than simply present or absent. An event can feel fully, partly, ambiguously or not at all self-caused.

Authorship can be overestimated or underestimated

Because authorship is estimated, errors can occur in both directions.

People can overestimate control when predicted events follow their actions. If an outcome matches an intention, occurs at the expected time and has no obvious competing cause, the available evidence may favour self-attribution even when actual control is weak.

People can also underestimate authorship. A genuine action may feel insufficiently self-caused when bodily feedback is unusual, consequences are delayed, another agent is present or the result differs from what was intended.

In shared situations, attribution may be divided. The presence of other possible agents can reduce an individual's experienced control, even when that individual made a real contribution.

The desirability of a result may influence interpretation as well. Positive outcomes are often more readily self-attributed in explicit judgments, while people may distance themselves from negative consequences. But this pattern is not universal. It varies with expectation, context and how agency is assessed.

Misattribution cannot be reduced to self-serving bias. Prediction, sensory reliability, learned contingency, affect and social context can all change how authorship is assigned.

Authorship is not a verdict of responsibility

The experience that an action was mine is relevant to responsibility, but it does not settle it.

Subjective authorship asks whether something felt like my action or consequence. Objective causal contribution asks whether my action actually helped produce the event. Control capacity concerns what I could predict, prevent, redirect or stop.

Moral and legal responsibility introduce further questions about intention, knowledge, capacity, coercion, negligence, alternatives and applicable standards.

These categories can diverge. A person may feel fully responsible for an event they did not cause. Another may feel detached from an outcome to which they objectively contributed. Strong emotion may alter experienced control without changing the physical relationship between action and consequence.

A person can also contribute causally without controlling every part of the process. But causal contribution is only one input into an assessment of responsibility. It does not by itself determine blame, liability or obligation.

The experience of authorship is evidence about a person's mental life. It is not an independent ruling on what they caused or what responsibility they bear.

An estimate that usually works

Return to the key press.

You decide to move the object. Action preparation anticipates the movement. Proprioception and touch report the finger pressing the key. Vision reports the object moving. The result occurs at the expected time and reliably depends on the press.

The signals converge, so authorship feels immediate.

Introduce delay, distortion or another possible cause, and the experience becomes less certain. The mind must determine which signals remain reliable, whether the action controlled the outcome and how much of the event belongs to the person.

This does not make authorship imaginary. It makes it constructed in the practical sense: the mind combines incomplete information to estimate what is happening.

The experience of causing an action is neither a direct perception of causal reality nor an arbitrary story added afterwards. It is a continuously updated estimate assembled from what the mind intended, predicted, sensed, observed and later understood.