Causal Deviance

The Problem of Appropriate Causation

The central problem of causal deviance is whether an agent's reasons can cause an event without making that event the intentional action those reasons would ordinarily explain. The answer is affirmative, and that result has shaped the philosophy of action since the rise of causal theories of action in the twentieth century. The phenomenon is familiar even when its technical name is not. A person plans to lower a glass, becomes agitated by the very intention to lower it, and drops it in a spasm. The intention has made a causal difference; the glass falls as envisaged; yet the person's dropping it is not the intentional execution of the plan. The causal description is true, while a particular description of intentional agency is false.

The philosophical difficulty is sharper than the observation that accidental events occur. The movement arises from the agent's relevant psychological life. It is not caused wholly independently of the intention. The reason that would justify, rationalize, and perhaps normally motivate the act is present. It even initiates the events leading to the expected physical result. What is missing is the relation that would make the resulting behavior an exercise of the agent's practical capacity. Donald Davidson's work made this difficulty central to the causal theory of action, and his later discussion acknowledged how difficult it is to define the necessary manner of causation (Davidson, 1963; 1973/2001, pp. 79–80).

The research question is therefore precise: what constitutes the appropriate manner in which reasons, intentions, or practical commitments produce an action? That question has at least three dimensions. The metaphysical dimension asks what differentiates an intentional action from an event that a psychological state merely causes. The explanatory dimension asks when a true causal explanation is an explanation of acting for a reason. The epistemic and normative dimension asks whether exercising control, manifesting practical knowledge, or responding to a plan is necessary for successful attribution. These dimensions are analytically separable, but a theory that handles only one frequently leaves the deviance problem intact.

This dossier treats causal deviance as its own philosophical subject. General theories of causation, the existence of free will, the definition of intention, and the complete metaphysics of agency are adjacent questions. They enter the argument when they bear directly on why a causally successful sequence can fail as an intentional performance. The guiding thesis developed here is that a causal chain belongs to an intentional action when it realizes a practical mode of governance attributable to the agent, relative to the description under which the event is intentional. The phrase "practical mode of governance" requires further analysis; it cannot function as a disguised synonym for nondeviance.

The Conceptual Grammar of Deviance

The term deviance is relational. A causal sequence is deviant relative to an attempted explanation of an event as the agent's intentional doing. Physical causation by itself is indifferent to this appraisal. A nervous reaction and an intentional movement both have causes; what differs is how their causal histories relate to the agent's practical orientation. The predicate "deviant" designates a failure of the proposed action explanation, not a violation of causal law.

Suppose an agent has a reason R to perform an action of type A, and event E is a bodily movement or downstream result. Three claims must be distinguished: R rationalizes A; R causes E; and E is the agent's intentional performance of A. Neither the first nor the second claim alone entails the third. Their conjunction still permits counterexamples. This is the basic structure of the objection to a simple causal sufficiency condition. In schematic form, the invalid inference is: R rationalizes A; R causes an occurrence that fits the description A; therefore the agent intentionally A-ed. The counterexample identifies a causal route under which the premises remain true and the conclusion fails.

Four distinctions keep this structure intelligible. A reason can be normatively relevant without causing anything; it can cause a movement without guiding it; a guided attempt can fail to realize its goal; and an intended goal can occur through an accidental chain begun by an otherwise intentional attempt. Accordingly, a reason's explanatory relevance, its causal efficacy, the agent's control, and the result of performance cannot be collapsed into a single property.

The relevant unit of assessment is often an event under a description. An arm movement may be intentional as reaching for a switch, while its unexpected consequence—alerting a stranger—is not intentional. In deviant cases, the problem can arise before any intentional action is established, or after an intentional basic action has already occurred. This difference eventually grounds the distinction between basic or primary deviance and consequential or secondary deviance (Mele & Moser, 1994; Schlosser, 2007).

The ordinary word accident must also be handled carefully. An accidental outcome may be the unintended side effect of an intentional action; an accidental bodily movement may fail to be an action under any intentional description; a planned outcome may arise accidentally after a genuine attempt. These cases are causally and philosophically different. Any theory that treats all of them as one undifferentiated failure of intention loses the explanatory structure that made the deviance problem significant.

Aristotelian Foundations: Origin, Choice, and Performance

Aristotle did not formulate the modern counterexample of deviant causal chains. His inquiry nevertheless provides a fundamental distinction on which the later issue depends: the distinction between a movement originating in an agent and a movement that the agent performs voluntarily. In Nicomachean Ethics III.1, Aristotle distinguishes compulsion, ignorance, and voluntary action, identifying a source of movement within the agent as central to voluntary performance (1110a1–1111a22). He also insists that evaluation concerns what was done in the concrete circumstances. The example of jettisoning cargo in a storm shows that the same type of movement can be evaluated differently when its occasion and chosen end change.

Aristotle's account cannot simply be translated into modern event-causal theory. His terms hekousion (voluntary) and prohairesis (choice or decision) belong within an account of ethical character, practical reasoning, and the agent as the origin of motion. The philosophical antecedent of deviance lies in the demand that actions be attributable to the agent in the right practical manner, rather than in the mere observation that the agent's internal state preceded a movement. That antecedent does not settle whether an event-causal account can meet the demand.

In Nicomachean Ethics III.2–3, choice is related to deliberation about matters within our power. The agent deliberates about means, and the practical conclusion concerns what to do. A movement that arises because deliberation provokes a disabling tremor would be internally caused in one sense yet fail to realize the deliberative connection relevant to a chosen act. This is a reconstruction using Aristotelian distinctions, not a claim that Aristotle himself supplied Davidson's later example.

The older framework also identifies an enduring tension. The origin of a movement and its practical explanation may both lie within a person, yet action requires an organized relation among desire, judgment, decision, capacity, and execution. The emphasis on practical organization becomes important in modern control accounts. It gives causal theorists a constructive task: specify how events can instantiate a practical order without simply postulating an additional mysterious agent outside the causal sequence.

Anscombe: Intentional Descriptions and Practical Knowledge

G. E. M. Anscombe's Intention (1957) brought the question of action under a description into the foreground of twentieth-century philosophy. An intentional action is characteristically one to which a certain application of the question "Why?" is appropriate: the agent's answer can disclose the end and means in the practical structure of what is being done. The account links intentionality with a form of practical knowledge and treats intentional descriptions as essential to the explanation of conduct. Anscombe's investigation is a major source for the claim that bodily occurrence alone cannot settle the intentional status of an action.

The famous case of a person operating a pump while unknowingly poisoning inhabitants of a house illustrates the plurality of descriptions. The agent may intentionally move a handle and replenish a water supply while not intentionally poisoning anyone. The consequences can be physically connected to one performance without sharing its intentional description. The point anticipates a central component of consequential deviance: causal production of a result does not immediately authorize an intentional description of that result.

Anscombe's treatment of practical knowledge has a second relevance. When agents act intentionally, they can in characteristic cases say what they are doing in a way grounded in the practical activity rather than inferred from observation. This is an epistemic relation to ongoing action. It is not adequately represented by the proposition that an earlier intention happened to be followed by a bodily event. The knowledge in question is bound to the agent's practical relation to the performance. Later theories dispute how broad and demanding this relation should be, especially in skilled, automatic, or unsuccessful activity.

The strongest contribution of Anscombe to the present problem is methodological. She asks what makes an event intelligible as a doing from within a practical description. A causal theorist may accept that question and offer a causal explanation of its answer. But the fact that the practical description is indispensable means that the philosophical problem cannot be answered by identifying just any causal link between prior mental content and a matching physical outcome. Contemporary proposals that unite ability with practical knowledge recover this insight without treating Anscombe and causalism as mutually exclusive intellectual programs (Kearl, 2023).

Davidson's Causal Theory of Reasons

Donald Davidson's “Actions, Reasons, and Causes” (1963) transformed the modern debate by arguing that explanations of actions through reasons are a kind of ordinary causal explanation. An agent's primary reason combines a favorable attitude toward actions of a kind with a belief that the action under the relevant description has the features that make it desirable. A primary reason rationalizes an action by making its attraction intelligible to the agent. Davidson maintains that the reason explaining why the agent actually acted must also have played a causal role.

This claim addresses a genuine explanatory problem. A person may possess many reasons that would justify performing an action. Merely showing that one reason fits the action does not identify the reason for which the person acted. Someone may want both exercise and an opportunity to meet a friend, and either desire could make a walk intelligible. What distinguishes the operative reason from an unexercised consideration? Davidson's answer is its causal role in bringing about the action (Davidson, 1963, especially pp. 685–693). The explanatory discrimination that causation provides is a substantial gain over theories that count any suitably justifying attitude as an actual explanation.

Davidson's theory distinguishes causal explanation from a strict psychophysical law. To say that a person's belief and desire caused a particular action does not entail that a universal law formulated in those intentional terms is available. This point matters because deviance is a problem about the relation between token causes and their products: a true singular causal statement can hold in both ordinary and problematic sequences. An inability to state strict laws of action does not remove the need to discriminate their different practical structures.

It is tempting to formulate causalism as a simple biconditional: a movement is intentional under a description precisely when it is caused by a reason that rationalizes it under that description. Yet deviant cases defeat this unrestricted formulation. Davidson's own later discussion of freedom and action explicitly acknowledges that relevant beliefs and desires can produce a movement in a manner incompatible with the intended exercise of action. The doctrine that reasons can be causes survives; the stronger proposal that any appropriate-content reason-cause is sufficient does not.

The distinction between a theory of reason explanation and a complete theory of intentional performance is decisive. Davidson gives powerful grounds for the first claim. The deviance problem concerns the additional conditions required for the second. The history of post-Davidsonian action theory can be read as a sequence of attempts to preserve the explanatory achievement while repairing the missing condition of appropriate causal production.

The Climber: Basic or Primary Deviance

Davidson's climber case is the canonical test of the theory. A climber holds another person on a rope. The climber wants to be relieved of the other's weight and recognizes that loosening the grip would accomplish this. The thought of doing so unnerves the climber, whose grip loosens as a result. The beliefs and desires that would rationalize deliberately releasing the rope have causally contributed to the releasing movement, yet the climber has not intentionally let go (Davidson, 1973/2001, pp. 79–80).

The force of this example rests on a subtle preservation of causal relevance. If the climber had released the rope solely because a rock struck the hand, the theory could deny that a relevant reason was causally operative. Davidson instead has the very reason produce the nervous state and the movement. The proposed reason-explanation is therefore neither an arbitrary rationalization nor a false description of causation. The failure arises in the internal organization of the causal route. The agent's apprehension defeats the sort of control that the reason would ordinarily guide.

Consider a sequence R → N → M. R is a reason state that rationalizes relaxing the hand; N is involuntary nervousness induced by R; M is loosening the grip. The chain is physically possible and causally continuous. Yet treating M as an intentional release under description A demands more than identifying R as its ancestor. N is not merely an intervening event. It changes the practical status of the entire relation by replacing execution of a decision with an incapacitating reaction. That is why a proposed solution must distinguish causal intermediaries that implement control from those that disrupt it.

This is often called basic deviance or primary deviance, because the problematic movement fails to be the agent's intentional basic action. The expression "basic action" is itself theory-dependent, but the point is stable: the movement lacks the intentional character needed to launch the intended performance. One cannot solve the case merely by tracing the intended outcome back to an earlier intentional act, since the explanatory chain contains no appropriately intentional execution of the relevant movement.

The climber example is a philosophical test of sufficient conditions. It does not show that intention and causal explanation are irrelevant to action. On the contrary, the example works because it preserves those relevant connections while exposing what they leave unaccounted for. A philosophical response should therefore identify the precise missing feature rather than dismiss the case as an implausible story. Its structural possibility is enough to test a universal analysis.

The Hunter: Consequential or Secondary Deviance

Davidson also describes a person who aims to kill someone, fires a shot, misses, and nevertheless brings about the intended death when the shot startles animals that trample the target (Davidson, 1973/2001, pp. 78–79). Here an intentional basic action exists: aiming and firing are performed deliberately. The intended ultimate result occurs. Yet the death happens by a circuitous and unforeseeable process that was not the agent's chosen way of killing.

The issue is therefore different from that in the climber case. What is in question is whether the outcome can properly be described as an intentional killing. It may be true that the agent intentionally fired and caused the death; the inference to intentionally killing by that path is contentious. This is consequential or secondary deviance, since the action under its basic description is intact while a further outcome seems improperly counted as an intentional achievement.

The hunter case forces a distinction between accomplishing an intention and acting intentionally toward an outcome. The agent can have the end in mind and perform a purposeful preliminary action. If the end occurs through a wholly unintended mechanism, the historical fact of intending it need not make its occurrence an intentional achievement. The result is causally connected to the effort, but the practical relation of means to end was not realized as the agent conceived or could manage it.

There are borderline cases. Someone throws a stone toward a bell and the stone ricochets before striking it. Whether ringing the bell counts as intentional may depend on the action plan, the degree of foreseeable deviation, and the way the agent's relevant competence was exercised. The distinction is therefore not simply direct versus indirect causation. Many intentional activities operate through long, indirect chains, including cooking, farming, medicine, and complex technical operations. What matters is how the chain relates to a practically structured endeavor.

The contrast between climber and hunter clarifies the demands on a solution. A theory must explain appropriate production of the agent's initial action and appropriate extension from that action to further effects. One criterion may govern both levels in an abstract sense, but its application differs: movement initiation and regulation in the first case, outcome generation through a selected or tolerably variable means in the second.

Event Individuation and Action Under a Description

Davidson's ontology of action commonly permits one event to fall under many descriptions. Turning a key, unlocking a door, and alerting a neighbor may describe related aspects or consequences of a single occurrence, depending on the individuation theory. The important point for deviance is that intentionality belongs to an event as described. A person can act intentionally in one respect and bring about other events unintentionally. Without that distinction, consequential deviance becomes impossible to diagnose accurately.

The philosophy of action has proposed different accounts of how many actions occur in complex sequences. Alvin Goldman (1970) distinguishes basic actions from actions generated by them, providing conceptual resources for saying that the agent intentionally performs one action while bringing about another in a different sense. Davidson's event-oriented approach and Goldman-style generation accounts disagree in ontology and individuation, but they can both recognize an explanatory difference between an intentionally initiated event and a later result.

This difference matters for responsibility. An unintended consequence can still be something an agent does, and it may be foreseeable or attributable to negligent conduct. To deny that an end was intentionally achieved through a deviant sequence does not automatically deny that the agent caused it, that it was the agent's doing under another description, or that the agent may be responsible for it. Intentionality, causal responsibility, and moral responsibility require distinct analyses.

A plausible theory must therefore keep its classification indexed to an explanatory description. Otherwise a report that the person "intentionally fired" is illegitimately transformed into a report that the person "intentionally caused every downstream consequence." The issue is a general limit on causal inheritance: intentional status cannot simply be transferred along every causal connection from an intentional act to everything the act brings about.

The same limit governs technical mediation. Intentionally pressing a command button may initiate a lengthy engineered process, but whether its final output counts as a planned achievement depends on what the agent intended, could competently influence, and was entitled to expect. Physical distance and causal length are imperfect proxies for that relation.

Why Temporal Proximity Cannot Solve the Problem

One family of responses appeals to the proximity of the intention or motivating reason to the action. Alfred Mele's work explores the distinctive executive role of proximal intentions and the organization of intentional conduct (Mele, 1992). Intuitively, the closer an intention is to execution, the less opportunity there is for an uncontrolled and irrelevant causal route to intervene. This intuition identifies something important: intentions formed long before an event are insufficient by themselves to determine which later movements express them.

The difficulty is that causal deviance can be extremely fast. A thought can immediately produce a tremor, reflex, or panic-induced movement. No long temporal interval is needed for the defect. Conversely, a long sequence of steps may constitute a competent intentional action: a pianist performs a piece, a person prepares a meal, or a programmer executes a planned multistage task. The length of a sequence is compatible with strong action guidance when the intermediate processes realize the plan.

The relevant notion of proximal must therefore be causal-functional rather than merely chronological. A proximal executive intention may structure current preparation, initiation, and adjustment of movement. But if the condition is described only as "an intention close enough to action that the causal chain is right," the proposed explanation is circular. It must explain why a proximal state has its executive significance, which intermediaries implement it, and how the agent's capacity is manifested.

Proximity remains a useful diagnostic, especially when distinguishing a remote desire from an immediately active intention. It is not a complete criterion. The problem demonstrates why temporal or spatial properties of an event sequence cannot substitute for the action-relative relation of governance.

Frankfurt and the Exercise of Guidance

Harry Frankfurt's “The Problem of Action” (1978) is a turning point because it shifts attention from action's antecedents to the organization of behavior as it occurs. Frankfurt argues that an action should be understood through the agent's guidance of bodily movements rather than through a bare causal story that begins with a desire or intention. This proposal directly responds to the threat posed by deviant causal chains: even when the relevant mental antecedents exist, the agent may fail to guide the movement.

Guidance is not identical with constant conscious monitoring. Skilled agents routinely act through learned patterns, and deliberate reflection can sometimes interfere with performance. The philosophical significance of Frankfurt's approach lies in his claim that action's distinctiveness is exercised in the conduct itself. Agency is shown in a practical relationship to unfolding behavior, including the capacity for adjustments appropriate to its purpose.

The account immediately faces an explanatory challenge. What makes a movement guided? To say that movement is guided when it is an action merely restates the target concept. An adequate theory must identify the mechanisms or capacities through which the agent can sustain or modify behavior in relation to the practical end. Feedback, adaptive correction, attention, and integration with an active plan are candidates, but none can be declared necessary in all forms of action without argument.

Frankfurt's emphasis on guidance provides a powerful diagnosis of the climber: the mental state produces the movement, yet the agent does not guide it as a release. Nervousness has causally displaced the practical organization of loosening the hand. The diagnosis is explanatorily valuable even if it does not yet offer independently sufficient conditions for guidance. It directs research to the ongoing relation between agent and movement.

The further issue is whether guidance must be exercised continuously. Some actions involve ballistic movements that cannot be corrected once launched, and some are initiated by a controlled decision and then run automatically. Guidance must therefore be understood over the appropriate temporal structure of a task: sometimes direct adjustment, sometimes competent selection and release of an organized routine, and sometimes monitoring of a sequence whose components cannot be continuously interrupted.

Searle's Intention in Action and Causal Self-Reference

John Searle's “The Intentionality of Intention and Action” (1979/1980) distinguishes prior intentions from intentions in action and develops the notion of causal self-reference. A prior intention concerns something to be done; an intention in action is involved in the action as it occurs. The content of an intention is not simply that a bodily movement occur. It includes, in Searle's analysis, a requirement that the intention itself cause the relevant condition of satisfaction in the manner represented. His account was designed in part to address deviant causal chains.

This proposal makes causal production internal to the content of intention. The simple statement "I intend that my hand rise" can be true in a deviant case if the intention makes me tremble and the hand rises. A self-referential intention includes a condition that the hand's rising be brought about through the exercise of that very intention in action. The cause is part of what would satisfy the intention, not merely an external event that happens to precede satisfaction.

The strategy improves the description of what would count as fulfillment. It explains why a merely coincidentally matching outcome should not satisfy the practical intention. Yet a familiar objection remains: what identifies the correct way in which an intention in action causes a bodily movement? If causal self-reference requires that the intention cause the result appropriately, the original difficulty reappears as a condition inside the representation. If appropriateness is given independent criteria, those criteria do the explanatory work.

Searle's account is therefore best treated as clarifying the intentional content and the satisfaction condition of action, rather than as a universally accepted completed solution. It shows that the causal relation can be constitutive of what is intended; further analysis is required to distinguish realization through an executive capacity from accidental production of the intended bodily condition.

Bratman: Plans, Stability, and Extended Agency

Michael Bratman's planning theory places intentions in the architecture of temporally extended agency. In Intention, Plans, and Practical Reason (1987), intentions are practical commitments that coordinate deliberation and conduct across time. They settle some questions, constrain subsequent reasoning, organize means, and support the execution of complex activities. A theory of causal deviance gains from this framework an account of the sustained practical relations that separate a directed sequence from a merely causally connected series of events.

Plans need not specify every muscular contraction or microphysical step. They provide organizing constraints for performance. The appropriate level of specification depends on the action. An intention to walk to a library permits ordinary variations in gait; an intention to perform a delicate surgical maneuver may require close monitoring and exact spatial control. Whether a variation is deviant depends partly on the grain of the plan and the competence legitimately engaged by the task.

This dependence helps address consequential deviance. A builder intentionally constructs a wall with the help of tools, materials, other people, and many indirect causal processes. The result is an intentional achievement because the chain follows a practical organization the builder can intelligibly adopt, regulate, and assess. By contrast, an intended result accidentally produced by an unforeseeable process may be causally traced to the builder's activity without being generated through the plan as an intentional achievement.

Planning theory alone, however, cannot settle every deviance case. The climber may possess a perfectly coherent plan while nervousness prevents its execution. A plan can be in place without governing a particular movement. A satisfactory plan-based explanation must specify how the plan is operative in producing and sustaining conduct, and how the agent's capacities permit departures and corrections. The difference between possessing a plan and executing it is one form of the general difference between causal antecedence and practical governance.

The most promising role for Bratman's framework is consequently structural. It explains how an intention can remain connected to an activity across time and variable circumstances. It supplies the continuity condition that a momentary model of action lacks, but it must be combined with a theory of competence and execution to resolve basic deviance.

Mele and Moser: A More Discriminating Theory of Intentional Action

Alfred Mele and Paul Moser's “Intentional Action” (1994) is an important effort to distinguish intentional performance from states and events that merely fit an agent's practical psychology. Their intervention belongs to an expanding literature in which the hard question is no longer whether mental causes matter, but how those causes participate in the production of intentional conduct. The terminology of primary and secondary deviance has become associated with their analysis.

Mele's Springs of Action (1992) assigns a distinctive executive role to intentions in the causation of behavior. The account recognizes that the motivational force of beliefs and desires differs from the practical resolution of what is to be done. A proximal intention is a candidate for initiating and coordinating action in a way a general preference is not. This distinction is useful in explaining why merely wanting an outcome cannot settle whether one intentionally produced it.

Any executive account must still explain the case in which a proximal intention triggers an involuntary reaction. Calling a psychological state "executive" because it is intended to guide movement does not show that it actually guides a movement in the case at hand. The theoretical burden is to identify a route that manifests the executive function, not merely the state that usually has it.

Mele and Moser's work also directs attention to intentionally doing something under one description and having it cause another event under a broader description. Since deviance arises at different sites in the action sequence, a complete theory may need differentiated conditions for intentional initiation, intentional continuation, and intended consequences. The philosophical reward is a more exact account of action as a structured event rather than a single belief-desire-to-movement transition.

The remaining question is whether executive intentions can be described independently enough to avoid the circularity that repeatedly threatens theories of nondeviant causation. The issue remains active in later control, teleofunctional, and attention-based research.

Bishop, Brand, and the Causal Organization of Agency

Myles Brand's Intending and Acting (1984) and John Bishop's Natural Agency (1989) belong to attempts to give agency a naturalistic causal organization adequate to practical explanation. They direct attention toward the functions of mental states in the initiation and regulation of behavior, rather than restricting causal explanation to a single antecedent state. Both are relevant to the question of what sorts of causal relations constitute an action.

Bishop's project aims to explain natural agency within a causal framework. A central attraction of such an approach is that the distinction between voluntary and involuntary movements can be related to capacities and organized causal mechanisms without assuming that agents stand outside physical causation. The claim that a mental antecedent caused a movement remains compatible with a further distinction between successful execution of an agent's capacity and its accidental disturbance.

Brand's theory gives intending a role in acting that reaches beyond a simple belief-desire pair. If intentions are to be practical states with a characteristic function, an explanation can look for the conditions under which that function is realized. The philosophical task is to distinguish a causal process that implements the intention from a pathological, unrelated, or interfering effect of the same state.

A frequent worry about naturalistic theories is that they shift the problem into unexplained talk of "normal" functional pathways. Biological systems can yield unusual but controlled actions, and familiar pathways can fail to guide. An adequate account cannot equate nondeviance with statistical normality. It must specify the organizational relation that confers practical relevance: what is controlled, by whom, through which capacities, and under what action description.

Naturalistic organization is therefore a promising research direction and a demand for independent empirical detail. It should not be presented as a ready-made metaphysical proof of a single complete solution.

Schlosser and Content-Sensitive Causal Explanation

Markus Schlosser's “Basic Deviance Reconsidered” (2007) argues that the apparent gulf between basic and consequential deviance may be smaller than often supposed. His account examines the possibility that the mental states and events central to action are causally effective in virtue of their contents and that a satisfactory explanation can appeal to the organization by which action-relevant content shapes performance. The later exchange in “Bending It Like Beckham” (2010) confronts objections and further explores control.

The important insight is that not every bodily sequence produced by an intention is produced as the means represented by that intention. A causal intermediary can transmit, transform, or interrupt the way in which the intentional content makes a difference. The same agent may intend to turn a wrist, and a disturbance may lead to that wrist's moving; the intentional content can still fail to be exercised as the operative pattern of control.

A content-sensitive criterion becomes informative when it connects represented conditions with independently describable functional transitions. For example, the agent's representation of a target can guide corrections of trajectory, while the motor system adjusts movement to maintain the target relation. If the target representation only produces panic, which in turn produces random movement, the content has entered the etiology without guiding the movement in the relevant way.

Yet a content account encounters its own challenge: some highly skilled actions involve sparse conscious representation, while simple movements can be performed under representations that do not specify their causal realization. The explanatory work cannot be done solely by matching content to outcome. It must characterize the capacity through which the content—or an associated practical orientation—structures execution.

Schlosser's contribution is significant because it invites a unified treatment of deviance along the whole causal chain. Instead of positing a wholly distinct kind of anomaly at each level, the theorist can ask whether practical organization is preserved when intention becomes movement and when movement produces its further consequences. This yields a fruitful methodological ambition even where proposed necessary and sufficient conditions remain disputed.

Velleman: The Disappearing Agent

J. David Velleman's “What Happens When Someone Acts?” (1992) challenges the familiar causal sequence from desire and belief to intention and bodily movement. Such a sequence may explain why an event occurred without presenting the agent in the proper explanatory role. The challenge is often described as the problem of the disappearing agent: a story of causal mental antecedents can appear to leave the person as an intermediary or site at which events occur rather than as one who acts.

Causal deviance is a particularly acute manifestation of that worry. The climber's psychology is causally active, but the agent does not perform a chosen release. Simply assigning a causal role to mental states cannot by itself distinguish an agent acting from an agent being overtaken by an effect of those states. The distinction requires explanation of the agent's participation in the causal sequence.

The disappearing-agent problem should not be confused with a proof that event causation is incapable of explaining action. A sophisticated event-causal theory may identify guidance, capacities, self-regulation, and practical commitments as patterns instantiated by events and states. The real challenge is whether its explanation gives the agent a distinct role that is not merely asserted after the causal chain has been described.

Velleman's criticism therefore sets a criterion for adequacy: the theory must make an agent's control intelligible, not merely exhibit events whose causal ancestry includes a desire. Whether this can be achieved by a suitably detailed causal functional account remains a substantive question rather than a settled division between naturalism and agency.

Agent Causation and Anti-Reductionist Responses

One family of responses maintains that action is fundamentally an exercise of an agent's causal power, and that the relevant connection is not exhausted by relations among distinct preceding mental events. Agent-causal accounts propose that the agent as substance or enduring individual originates action in a way that cannot be analyzed wholly into ordinary event causation. Such positions seek to preserve authorship where a chain of states threatens to replace the doer with happenings.

The appeal is clear in a deviant case. It seems reasonable to distinguish the climber intentionally letting go from a grip loosening under the influence of the climber's fear. An agent-causal theorist can say that only the former involves the relevant origination by the agent. But the challenge is to specify the agent-causal relation sufficiently to discriminate genuine exercises from involuntary effects, especially when both occur within the same person's psychological and bodily system.

If agent causation is introduced as an unexplained primitive that holds exactly in intentional actions, the counterexamples are excluded by stipulation. Such an account may be part of a broader metaphysics, but it supplies limited analysis of the mechanism of deviance. If it is developed through powers, capacities, and a theory of their exercise, it can compete more substantively with sophisticated event-causal accounts.

Anti-reductionist accounts inspired by Anscombe and by process-oriented traditions raise a different objection. They argue that intentional activity should be characterized as a practical unity from the beginning, and that decomposing it into an intention, a bodily event, and an external connection misrepresents its nature. This view can treat the deviant sequence as failing to instantiate the practical process itself. Its burden is to explain how that process is individuated and how empirical descriptions of causal organization relate to it.

The two families of responses have distinct ontological commitments. Their common contribution is a refusal to assume that every event caused by an appropriate mental state counts as the agent's doing. The theoretical dispute concerns the best explanation of this refusal.

Attention and Agentive Control

Wayne Wu's “Experts and Deviants: The Story of Agentive Control” (2016) develops a distinctive answer by placing attention at the center of action guidance. Wu argues that theories focusing on intentions and antecedent causes frequently omit the functional role of attention in selecting, organizing, and sustaining the behavior that constitutes an action. The relation between intention and attention can help explain both ordinary controlled performance and failures of agency, including deviance and distraction.

This proposal gives a more determinate interpretation of guidance. An agent selects an action-relevant target, perceptual feature, or task, and attentional organization facilitates the execution of a corresponding capacity. In skilled performance, attention can be absorbed in the task without taking the form of explicit verbal deliberation about each movement. In a deviant sequence, the intention may cause behavior while failing to organize the relevant attention-to-action relation.

Attention must nevertheless be distinguished from conscious awareness. An agent may be consciously anxious about acting while failing to control the movement; a practiced agent may guide intricate conduct while attending primarily to the environment rather than the muscles. The functional allocation of processing to a task may matter more than introspective intensity. This helps explain why "the agent was thinking about doing it" is not a reliable criterion of intentional execution.

The hypothesis also faces counterexamples. Some reflexlike movements are integrated into a larger intentional activity; some automated routines are reliably initiated and monitored without continuous focal attention. A complete theory must specify which attentional functions are constitutive of control in a given context and which are merely frequent aids. Wu's work substantially clarifies the candidate mechanisms, even as the universal scope of an attention condition remains an open question.

Practical Knowledge, Ability, and Recent Research

The relation between practical knowledge and causal control has become a major site of contemporary synthesis. Timothy Kearl's “What We Know When We Act” (2023) argues that a capacity-constituting kind of knowledge can help explain how an agent's reasons nondeviantly cause and sustain purposive behavior. The point is not that intentional action always requires the ability to give an explicit explanation of each movement. Rather, competent activity can manifest a kind of knowledge that belongs to the exercise of an ability.

The proposal responds to a genuine difficulty faced by both causalism and a demanding epistemic account. Skilled athletes, drivers, and craftspeople may act intentionally while lacking articulate knowledge of the microdetails of performance. Conversely, an agent can state a correct intention while failing to manifest any relevant competence in the resulting movement. Ability-constituting knowledge is meant to identify the practical relation that unites purposive orientation with competent execution (Kearl, 2023).

One attraction of this view is that knowledge becomes part of the performance explanation instead of an independent mental accompaniment. The agent's ability gives shape to what it is to exercise an intention successfully. If an agent's reason merely triggers a panic response, the resulting movement does not manifest the relevant ability, even though the reason remains a cause. In normal action the same reason may enter a process through which the capacity is exercised.

There is a challenge concerning fallibility. Agents can act intentionally but clumsily, and genuine abilities can be exercised unsuccessfully. If every failure of competence were classed as causal deviance, the account would become too restrictive. The correct distinction is between failing while exercising a capacity and producing an outcome through a process that bypasses the capacity altogether. The former may be poor or unsuccessful intentional action; the latter is the paradigm case of deviant production.

The contemporary literature thus shifts the problem from an abstract ban on unusual causal chains toward a positive account of how reason-responsive capacities are exercised. This move deserves scrutiny, because the notion of exercise cannot be accepted as a final primitive unless its relation to actual causal organization is explained.

Success, Failure, and the Scope of the Intended Result

A proper theory must separate the intentional performance of an attempt from the successful achievement of its goal. A person intentionally throws a ball toward a target and misses. The throw can be fully intentional despite the failure. Conversely, a ball may be dislodged by a person's nervousness and strike the target precisely as intended, without there being an intentional throw. Success and intentional action are independently variable.

The independence can be expressed as four possible cases. An appropriately guided action may succeed; an appropriately guided action may fail; a deviant causal sequence may produce the intended physical result; and a deviant sequence may fail even to produce that result. A theory that defines intentionality by success misclassifies at least two of these cases. A theory that defines it merely by a prior intention plus eventual correspondence misclassifies the deviant-success case.

The distinction also bears on attempts. An unsuccessful attempt may be an action under a description of trying, aiming, or initiating a means. If the agent intentionally aims and releases an arrow that is blown off course, the effort remains intentional. If a spasm releases the arrow after a thought about shooting, the release may fail to be intentional. The classification turns on the practical mode of production, not the later arrow trajectory alone.

Consequential deviance complicates this structure. The hunter intentionally shoots, but the intended death occurs through an unexpected animal stampede. One may distinguish the intention to kill, the intentional attempt, the accidental manner of causing death, and the full action as described. A careful account should not assume that an unintended route necessarily erases causal or legal responsibility. It establishes a narrower point about whether the end was intentionally achieved in the way required by the description.

An adequate theory should explain how outcome descriptions inherit intentional status from an intentional process. A promising rule is that an outcome may be described as an intentional achievement when the outcome's generation falls within the practical means and variation tolerated by the agent's guided undertaking. This rule is a substantive proposal, requiring analysis of what "falls within" means for complex, risky, and technologically mediated activity.

Counterfactual Responsiveness and Its Limits

Control is commonly connected with responsiveness to alternatives. An agent who guides a movement may adjust it if the target shifts, stop if circumstances change, or alter a plan when a means becomes unavailable. Counterfactual tests ask what would have happened under relevant variations. If the person would still have performed an essentially identical movement despite profound changes in the guiding task, that fact may undermine a claim that the task actually governed performance.

But a simple counterfactual condition is neither necessary nor sufficient in every case. A skilled throw can be intentionally made even when the ball is already in flight and no correction is physically possible. A person may act competently despite having no opportunity to revise the action. On the other side, a reflex might vary with environmental stimuli while remaining outside the agent's intentional control. Mere sensitivity to change is a feature of many natural systems; it does not independently establish agency.

Counterfactual responsiveness must be related to the agent's actual capacity and relevant task. We ask whether the agent exercised a procedure or skill capable of different behavior under task-relevant variations, not whether any imaginable condition would have altered the event. For extended action, relevant alternatives can include recalibration, inhibition, or the selection of further steps. For a launched action, they may concern earlier selection, timing, and initiation.

There is also an epistemic problem. Counterfactual claims about another person's internal organization may be difficult to verify. A theory that makes them central needs an account of evidence: behavior over repeated trials, the structure of skill acquisition, motor correction, intentional reports, and circumstances of interruption. Philosophical analysis supplies conditions of relevance; empirical inquiry can help test whether a particular case satisfies them.

Counterfactual criteria therefore refine the idea of governance but cannot replace an account of agency with a general law of system responsiveness.

Motor Control and Empirical Explanation

Research on motor control offers potential resources for explaining how intentions become organized movements. Predictive models, sensory feedback, motor preparation, error correction, and learned policy execution describe ways organisms coordinate bodily processes across time. Philosophers may use this work to make guidance more concrete than the phrase "caused in the right way." Theories of forward models and motor prediction, for example, help explain how expected sensory consequences can inform adjustment of ongoing movement.

The philosophical inference must be made carefully. A neural correlate of movement preparation is not itself a definition of intentional action. Involuntary processes can involve sophisticated feedback, while intentional acts can be executed through relatively automatic learned routines. An empirical model can identify a causal mechanism contributing to control, but whether a given mechanism constitutes intentional agency depends on how it is integrated with the agent's practical orientation and the description of the activity.

The climber example invites a mechanistic question: which circuits or processes connect deliberation to motor execution, and which connect threat or panic to a grip release? The presence of different causal routes may make the difference between guided release and involuntary loosening empirically intelligible. Nevertheless, a hypothetical neurophysiological account of a case cannot settle the normative and descriptive conditions of intentionality without philosophical interpretation.

It follows that deviance is neither a purely armchair puzzle with no scientific relevance nor an empirical classification that laboratories can discover independently of a concept of action. Philosophical work determines what contrast is under investigation; empirical work helps explain the mechanisms by which the contrast is realized.

The distinction is particularly important when studying complex movement. Control can be distributed across levels of processing, with explicit planning at one level and rapid unconscious adjustment at another. A theory that identifies control with conscious supervision of all motor intermediaries would exclude much ordinary skilled agency and would therefore fail as a general analysis.

Habit, Skill, and Automatic Execution

Skilled action provides some of the strongest tests of any proposed solution. An experienced driver changes gears, a pianist anticipates a musical transition, and a craftsperson adjusts a tool without rehearsing propositions about each muscular movement. Such conduct can be intentional despite the involvement of automated processes and a relatively low level of focal reflection. The practical organization of the activity is manifested in responsiveness to goals and constraints, not necessarily in continuous conscious issuance of commands.

A successful deviance theory must therefore allow nonconscious processes to implement an agent's intention. The mere presence of an intervening process does not make causation deviant; otherwise almost every bodily action would be disqualified. What matters is whether the intermediary belongs to the competent execution of the relevant undertaking. A learned routine may be precisely the normal medium through which the agent controls the action.

The contrast is between an automatic process that constitutes the exercise of a skill and an involuntary process that frustrates it. In one case a movement occurs because a trained pattern responds to the task; in the other a panic reaction happens to yield a movement with the intended physical configuration. Both processes may be causally complex, yet their practical roles differ. The difference must be established by the explanatory organization, including the agent's history of competence, current orientation, and available adjustments.

This perspective also prevents a naive equation of intentional action with maximal voluntary effort. An expert who attempts to monitor each microstep may perform worse. Action can be more completely guided when complex subroutines are delegated to acquired skill. The philosophical challenge is to explain how higher-level intention and lower-level routine compose a single intentional performance.

Cases of distraction introduce gradations. A driver may continue on a familiar route while attention is elsewhere. Whether a particular segment is intentionally driven depends on the kind of intention operating, the knowledge and capacities manifested, and the proper description of the activity. There need not be a sharp boundary between optimal focused control and every form of intentional action. The deviance problem concerns cases in which the relevant practical connection fails, not every imperfection in concentration.

Reasons, Mental Causation, and Explanation

The debate also intersects with the metaphysics of mental causation. Davidson's original argument relies on the claim that reasons can be causes. Some later philosophers distinguish normative reasons, understood as facts counting in favor of action, from motivating psychological states, understood as beliefs, desires, or intentions. An agent may act because a fact is a reason, while the causal mechanism involves states in which that fact is represented. This distinction prevents a simple shift from explanatory rationality to a claim that abstract normative relations literally exert forces.

Causal deviance can be formulated under either vocabulary. If an agent's psychological state represents a justifying reason and causes an accidental movement, the problem remains. If reasons are facts rather than mental events, the theory must explain the connection between those facts, the person's responsiveness, and the bodily occurrence. The central issue remains the manifestation of practical orientation in the production of the performance.

An explanation that an agent raised a hand because a colleague asked for help may be acceptable in ordinary circumstances without tracing every neural event. The legitimacy of such explanations presupposes a background of normal guidance. The deviance example forces us to examine the background precisely because the reason-explanation seems formally satisfied while the expected practical relation is absent.

The strongest philosophical position is therefore neither to replace reasons with neurophysiology nor to insulate rational explanation from causal inquiry. It is to ask what explanatory relation connects the content of a reason, the agent's ability, and the event's causal history. The shape of that relation determines whether a reason explains an action as something done rather than merely causing what happened.

This distinction also helps interpret evidence. The fact that an intention precedes an event, predicts it, or is correlated with neural preparation cannot alone show that the intentional content exercised its alleged control. The epistemic task is to establish relevant causal organization, and the metaphysical task is to explain what makes that organization agentive.

Moral and Legal Responsibility After Deviance

Causal deviance has implications for moral psychology and responsibility. An agent may have a culpable intention while the intended harm occurs by an unexpected chain. The moral evaluation of the intention, the assessment of intentional action, the attribution of causation, and the allocation of responsibility need not coincide. A person can be blameworthy for an attempt even when the hoped-for harm does not occur; harmful consequences may be attributable in some circumstances without constituting intentional achievement of the precise result.

Mark Alicke and David Rose (2012) examine how judgments of culpable control vary across kinds of causal deviation, connecting action theory with empirical moral psychology. Their work demonstrates why philosophers should be attentive to distinctions among disrupted attempts, alternative causes, foreseeable side effects, and remote consequences. These patterns do not erase the technical problem of deviance; they reveal that ordinary judgments of blame bring additional evaluative dimensions into an already complex causal story.

A philosophical account should resist drawing conclusions about legal liability directly from a stipulative description of intentional action. Legal systems have their own requirements concerning fault, causal contribution, foreseeability, and intervening events. A deviant chain can affect some of these requirements while leaving others intact. The present research establishes conceptual distinctions applicable to legal analysis without prescribing one jurisdiction's doctrine.

There is also a moral asymmetry between intending a bad outcome and accidentally producing it, which must be examined independently of the causal route. A vicious intention may reveal something about the agent even when no voluntary movement follows. Conversely, an accidental outcome can produce grave harm without making the harmful effect the agent's intended achievement. Deviance illustrates why a moral theory needs more than a map of material consequences.

The explanatory moral is modest but important: responsibility attribution requires an explicit account of what the agent controlled and what was attributable to the exercise or failure of their capacities. A causal description is necessary for many such judgments, but it cannot silently perform all their normative work.

Long Causal Chains, Institutions, and Technology

Contemporary action often extends through tools, software, institutions, and distributed systems. A person may intentionally order a vehicle to stop, authorize a payment, launch a computation, or coordinate a collective project. In such cases the intended effect may occur after numerous mediating processes involving entities and systems beyond the agent's direct bodily control. An account that demands immediate intervention at every stage would make ordinary technological and social agency unintelligible.

The question is how an initiating intention remains practically related to downstream events. The answer depends on authorized delegation, known operational roles, competent selection of means, monitoring appropriate to the task, and the agent's capacity to revise or withdraw instructions where the system permits it. Such capacities may be exercised through interfaces and institutions rather than direct muscular correction. The relevant form of governance is often distributed and procedurally structured.

Consider an operator who intends a machine to place a component. If the machine executes an authorized program that tolerates ordinary variability, the placement can plausibly be attributed as an intentional technologically mediated achievement. If the operator's command produces an unrelated malfunction that accidentally places the component in the specified position, the result is causally traceable to the command but may be deviant relative to the intended operation. The difference is the relation between selected procedure and achieved result.

This framework can also inform philosophical analysis of artificial systems. Evaluating whether a system exhibits its own agentive control is a distinct question from asking whether it is causally instrumental in a human agent's plan. Philosophical inquiry must specify what the system's capacities are, how its goals or instructions are represented, which processes guide its behavior, and how to attribute outcomes. A causal sequence connecting a human request to a machine output does not by itself determine the ontology of either agent.

The horizon of Artificial Era may make these distinctions increasingly important, but their content is established through philosophical analysis of action and causation rather than an assumption about every artificial artifact's status. This dossier does not revise definitions in the Aisentica canon. Its direct subject is the general structure of causal deviance, with technologically mediated cases used to test the generality of the proposed account.

Toward a Noncircular Account of Practical Governance

A satisfactory synthesis begins by identifying four jointly relevant relations. First, practical orientation: the agent has a reason, intention, commitment, or purposive capacity making a description of the conduct intelligible. Second, causal efficacy: that orientation figures in the actual production of the event, not merely in a story that would have justified it. Third, practical governance: the causal production proceeds through the exercise of capacities that select, sustain, or organize behavior relative to the task. Fourth, achievement: when an intentional outcome is ascribed, the route by which it is produced lies within the means or permitted variation of the governed undertaking.

Each relation has its own evidential and philosophical role. The first distinguishes intelligibility from blind occurrence. The second distinguishes operative motivation from idle rationalization. The third distinguishes acting from being affected by one's own motives. The fourth distinguishes the intentional pursuit of an end from the fortuitous occurrence of that end. These relations are not presented as a newly established metaphysical law; they are a diagnostic framework distilled from the historical and contemporary debates.

The third condition demands specification. I propose that practical governance is the actual realization of an agent's task-relevant capacity through processes that maintain the organized dependence of performance on the practical orientation. This formulation directs inquiry to abilities, selection, correction, inhibition, and task-relative causal organization. It also permits different implementations: immediate motor guidance, learned skill, plan-governed extended conduct, and responsible delegation. The reference to "task-relevant" means that the distinction is indexed to what the agent was undertaking, not simply to the mechanical regularity of an isolated process.

The criterion must avoid circularity. "Processes that maintain organized dependence" should be investigated through specific features: whether action-relevant information determines selection among available movements; whether a capacity's usual modes of correction can be exercised where appropriate; whether intermediate processes implement rather than disable the practical orientation; and whether the result is generated through the means the undertaking authorizes. These are families of independently assessable features, not a claim that any one measurable marker universally defines action.

The framework diagnoses the climber as a failure of practical governance at the transition to movement: an intention-related state causes a nervous response that bypasses the capacity for controlled release. It diagnoses the hunter as a failure of the achievement relation: an intentional shot causally produces death by a route outside the relevant means of aiming and shooting. It recognizes an unsuccessful throw as intentional action without successful achievement. It recognizes skilled automatic performance when acquired subroutines implement the task. Finally, it recognizes that intentional technological action can be mediated by long causal sequences without becoming deviant solely because of their length.

The synthesis clarifies the structure of the problem while leaving open the ultimate ontology of agency. A defender of event causation can seek a naturalistic account of these relations; a defender of process-oriented practical knowledge can explain how they arise from the unity of intentional activity; an agent-causal theorist can embed them in an account of the agent's exercise of power. The framework permits comparative evaluation by forcing each theory to identify the feature that makes the difference in the hardest cases.

Critical Objections to the Governance Synthesis

The first objection is that the framework merely renames the problem. If "practical governance" means whatever occurs in intentional action, a deviant case will be excluded by definition, and nothing will have been explained. The response is to treat governance as a theoretical relation requiring specification in terms of capacities, control variables, actual causal dependencies, and the practical description of the task. The proposal has explanatory content insofar as it identifies independent distinctions and tests; its adequacy depends on the success of more detailed theories implementing them.

A second objection concerns lucky competence. An agent may attempt a difficult action, lose control briefly, and nevertheless succeed after an unexpected correction. Some such successes count as intentional achievements; others seem accidental. The proposed answer is task-relative and graded. Losing precise control does not always erase all intentional status, and a competent performance can include recovery from unpredictable disturbances. The issue is whether the eventual result is still generated through an exercise or restoration of the relevant capacity. Borderline cases should be investigated rather than forced into a binary classification at the outset.

A third objection concerns minimal agency. Infants, animals, and highly practiced performers may act intentionally without articulate goals, deliberation, or self-representation. A theory with overly intellectualized requirements would exclude genuine practical capacities. The response is to distinguish represented reasons and explicit planning from purposive, environmentally situated abilities. In some actions the agent's practical orientation is manifested in embodied selection and adjustment, rather than in verbalizable belief-desire propositions. What the theory needs is a principled account of the relevant intentional description, not a uniform demand for reflective deliberation.

A fourth objection is that any causal chain could be redescribed at an appropriate level to look controlled. The answer is evidential. Not every description is equally supported. The agent's task, history of skill, actual opportunities for adjustment, target-directed behavior, and mechanism of interruption constrain the relevant account. The explanatory description must preserve what the agent undertook and what the event actually realized.

A fifth objection concerns determinism. If every stage of action has physical causes, how can the agent govern it? The notion of governance proposed here is compatible with deterministic and indeterministic metaphysical background assumptions, because it concerns the organization of causal processes into practical capacities. Whether this organization supplies the sort of freedom required for ultimate moral responsibility is a further and independent question.

Open Philosophical Questions

The debate still has genuine unresolved problems. One is whether nondeviant causation can be given necessary and sufficient conditions applicable to every kind of action. The diversity of bodily movement, habitual skill, deliberative action, social coordination, and technological delegation may resist a single compact analysis. This does not excuse vagueness: it may instead require a family of related action kinds with explicit unifying constraints.

A second problem is how to individuate the relevant causal process. Is an action an event with an earlier intention among its causes, an ongoing process of practical activity, a manifestation of power, or a structure that admits different metaphysical descriptions? An account may classify the climber similarly under several ontologies while interpreting the source of the classification differently. The deviance problem alone does not settle these metaphysical disputes.

A third problem concerns the epistemology of control. What evidence establishes that the relevant competence was actually exercised rather than merely available? Behavioral reliability, neurophysiology, responsiveness to perturbation, and the agent's practical knowledge provide partial evidence, but their relation is not yet captured by an uncontested theory.

A fourth problem is how far an intended result may depart from a plan while remaining an intentional achievement. Unexpected opportunities, error correction, chance events, indirect tools, and complex institutions complicate the boundary between authorized means and accidental consequences. This is an especially pressing problem where actions are temporally extended or distributed across agents and machines.

A fifth problem is the status of practical knowledge in nondeviant action. Is knowledge constitutive, evidential, or a typical manifestation of competent agency? Kearl's ability-based analysis offers an important reconciliation, while stronger Anscombian and causalist theories continue to disagree about the order of philosophical explanation.

These questions define a research program. Progress requires careful attention to the intentional description at issue, the actual causal route, the agent's active capacities, and the evidential basis for judging that those capacities governed what happened.

Philosophical Synthesis

Causal deviance reveals an essential fact about action: the physical occurrence of an intended event and the exercise of an intention are distinguishable achievements. Davidson established the importance of causal efficacy in explaining acting for a reason. The deviance cases show that causation by a rationalizing state is too permissive as a complete criterion of intentional performance. Anscombe makes intentional description and practical intelligibility central; Frankfurt makes ongoing guidance central; Searle clarifies the role of causal production within intention's satisfaction conditions; Bratman supplies temporal organization; Mele and Moser refine the distinction between intentional performance and merely resulting events; Schlosser investigates the contribution of action-relevant content; Wu identifies attention as a mechanism of control; and Kearl connects nondeviant guidance with the manifestation of ability-constituting knowledge.

These approaches identify different parts of the same explanatory burden. An adequate account must identify an agent's practical orientation, establish its actual causal influence, explain how the agent's relevant capacities were exercised through the production of the event, and, when a further result is attributed as intentionally accomplished, show how that result belonged to the governed practical undertaking. The distinction between basic and consequential deviance is useful because it identifies different points at which this account can fail. Their common theme is the disruption of a practical relation by a causal sequence that remains physically continuous.

The principal philosophical result of this inquiry is a four-relation diagnostic of intentional performance: orientation, efficacy, governance, and achievement. Each relation can be present or absent under different descriptions of a complex event. Orientation without efficacy explains an idle reason. Efficacy without governance explains the classic primary deviance cases. Governance without success explains competent but unsuccessful action. Success without a governed path to the result explains consequential deviance. This framework preserves the independence of causal truth, rational intelligibility, agency, and success.

The framework does not declare the metaphysical debate settled. It establishes the task that successful theories must discharge. The right causal route is one in which the practical capacities of an agent are actually manifested in the production and organization of the performance. Whether this is ultimately explicated through event causation, process metaphysics, agent causation, practical knowledge, or a combination remains open to further philosophical and scientific investigation.

Conclusion

The problem of causal deviance begins with a counterexample and ends by revealing the architecture required for a philosophy of intentional action. It is possible for a reason to rationalize a proposed action and cause a corresponding event without the agent performing that action intentionally. The relevant failure can occur at the production of bodily movement, at the transition from an intentional act to an intended consequence, or in an extended sequence whose organization ceases to be governed by the undertaking.

Historical reconstruction identifies the enduring importance of voluntary origin and practical description. Davidson's causal insight remains indispensable for distinguishing the reasons that actually explain behavior from reasons an agent merely possesses. The subsequent literature demonstrates that an intention's causal efficacy must be explained in terms of its role within an organized exercise of agency. Control, planning, attention, skilled ability, and practical knowledge supply distinct resources for meeting that demand.

The decisive conclusion is that intentional agency is a structured relation of practical production. A reason's occurrence, an intention's presence, a movement's physical realization, and an end's eventual achievement each make different contributions. Causal deviance occurs when one of those contributions is mistaken for the organized exercise of the agent's capacity that an intentional description requires. A complete philosophical theory must show how that capacity is realized through causal processes, which variations remain compatible with its exercise, and which outcomes genuinely belong to what the agent intentionally did.