# Free Will Through Authored Foreclosure ## A Diachronic Theory of Accrued Self-Authorship **Author:** Dustin Patzer ## Abstract This paper proposes one possible location for a logical hinge against the traditional determinism/indeterminism dilemma. It does not offer an exhaustive theory of human freedom, a census of free acts, or a foundation from which all other freedom is derived. Freedom is operationally defined against determinism: a freedom-bearing event must occur where the future genuinely could have been otherwise. This requires **ontic multiplicity**, not merely represented alternatives or uncertainty in an embedded observer. Foreclosure requires a further condition: **exclusive actualization**. One continuation becomes actual in the relevant history and alternatives cease to remain available as continuations of that history. A branching ontology on which every supported continuation remains actual does not satisfy foreclosure merely by containing plural branches. Indeterminism or plurality alone, however, supplies only openness. It does not explain authorship. The proposed witness-location is a developmentally acquired, diachronically sustained process in which a recursively continuous agent holds an operative feature of itself as an alterable object, represents at least one possible successor variance distinct from a projected default continuation, adopts that variance as an operative target, and repeatedly directs thought, action, attention, environment, and experiential curriculum toward that target under changing feedback. The target need not remain verbally or imagistically identical at every moment. It may be reconstructed, refined, or partially revised while preserving an identifiable direction of transformation. The agent appears within this process as the present representing locus, the represented object of alteration, and the anticipated and later realized bearer of the intervention's effects. In this proposed expression, recursion is the primary diachronic mechanism of accrual. The result of one qualifying self-object trajectory can alter the agent who performs the next trajectory: later target-generation, maintenance, correction, environmental access, and intervention-capacity may therefore operate from a constitution partly shaped by earlier self-directed activity. Recursion is not required for the first nonzero contribution, and it does not guarantee increase. It explains how an earlier contribution can enter the causal constitution from which a later contribution arises. Humans do not possess direct write-access to neural parameters. They possess limited **curriculum access** to the conditions under which later learning occurs. The distinctive Stage 2 claim is therefore not merely that a higher-level representation alters lower-level processing. It is that a represented possible variance of the continuing agent can function as a recurrent control target: the agent maintains or reinstates it across perturbation, interprets new prediction errors and consequences relative to it, changes means while preserving or revising direction, and thereby repeatedly contributes to the constitution of the later agent. Where more than one continuation is genuinely realizable and one becomes exclusively actual, occurrence irreversibly extinguishes alternatives. This is **foreclosure**. Where a foreclosure is materially routed through the specified self-object trajectory, it is a candidate **authored foreclosure**. The routing criterion is interventionist: varying the content of the represented self-transformation should systematically vary relevant properties of the later transition band relative to suitable matched controls. The relevant open transition must lie on the causal pathway from the self-object project to the later-agent change being credited, and the intervention must materially alter that transition's own band. Mere collocation is insufficient. The trajectory must also be prospectively and causally identifiable rather than reconstructed after success. The constructive core is **band-authorship under liveness**. Deterministic self-modification supplies routing without live alternatives. Bare chance supplies live alternatives without agent-shaped routing. The proposed witness binds the agent's content-sensitive shaping to the same open band from which one continuation becomes exclusively actual. The conjunction is therefore not compatibilist self-modification plus unrelated noise. It is agent-shaped foreclosure of genuinely live continuations. Constitutive luck is fully conceded. The first qualifying agent is overwhelmingly built by causes it did not author, and exact outcomes may retain substantial present or outcome luck. The paper denies only that lucky provenance or stochastic resolution automatically reduces a sustained, content-sensitive, self-directed causal contribution to zero. Across recursive loops, the later agent may bear a larger contribution from prior qualifying interventions because their effects can enter the constitution that initiates and maintains further interventions. Earlier freedom-bearing trajectories may become sedimented into stable dispositions and character. No monotonic increase is claimed: self-authorship can plateau, reverse, fragment, or diminish. The paper's logical form is a conditional existential proof schema. Any adequately established ontically open and exclusively actualizing event would defeat exhaustive determinism. A single adequately established trajectory in which the same materially relevant transition class is open, exclusively actualizing, and self-object-routed would additionally defeat the universal inference from indeterminism to zero authorship. Adequate establishment need not consist in direct inspection of a token's unrealized alternatives. It may be type-level and statistical. The paper does not claim that such a witness has already been empirically established. It identifies one coherent candidate location, states the physical, developmental, causal, and actualization premises required, and specifies what one successful instance would prove. It expressly brackets downstream questions of culpability, blame, punishment, legal responsibility, and basic-desert moral responsibility. # 1. The dilemma and the limited target The traditional dilemma presents two branches: 1. **Determination:** If an event is fully determined, it could not have occurred otherwise. 2. **Indetermination:** If an event is not determined, its occurrence appears attributable to luck rather than the agent. The theory accepts the pressure from both horns. It does not defend freedom on the deterministic branch. A deterministic system may represent alternatives, deliberate, revise itself, respond to reasons, and recursively modify its own future inputs. Those are real differences in mechanism. But if every state, representation, revision, and output could never have failed to occur exactly as it did, the process contains no genuinely open future. It is a complex trajectory through the only physically realizable history. The theory also rejects the claim that indeterminacy alone creates freedom. A decay event or random perturbation may extinguish live possibilities without authoring anything. The target is therefore narrower: > Can one identify a physically open event whose foreclosure is materially attributable to a historically continuous agent through a metacognitive process directed at changing that agent's later self? The paper proposes one location where that conjunction may occur. It does not claim that this is the only, maximal, or best possible location. It is selected because it appears to present both requirements in unusually visible form: - genuine multiplicity, if Stage 1 obtains; and - strong attribution through intentional self-object-directed intervention. The proposed architecture has three nested units: 1. **Event level:** an authored foreclosure is a materially routed foreclosure of live alternatives. 2. **Trajectory level:** a sustained self-variance project maintains or reconstructs a represented target across perturbation and adaptively reorganizes means toward it. 3. **Recursive agent level:** the effects of an earlier qualifying trajectory partly constitute the agent who may initiate, sustain, revise, or abandon a later one. The first two levels identify one candidate freedom witness. The third explains how isolated contributions may become accrued self-authorship over time. It must not be read backward as a requirement that the first qualifying agent already have authored itself. --- # 2. Stage 1: ontic openness ## 2.1 Operational definition of freedom For this paper, freedom is used in an incompatibilist sense: > A freedom-bearing event must occur where more than one continuation was physically realizable from the same complete relevant prior state and where one continuation became actual in the relevant history while the alternatives ceased to remain available as continuations of that history. In schematic notation only: \[ S_t,C_t \rightarrow P(S_{t+1}), \qquad |\operatorname{support}(P)|>1, \qquad \operatorname{EA}(S_t,C_t,S_{t+1}) \] Here, \(S_t\) represents the complete relevant prior system-state, \(C_t\) the relevant context, and the non-unit support represents multiple physically realizable continuations. \(\operatorname{EA}\) denotes exclusive actualization: one supported continuation becomes actual in the relevant history while the alternatives cease to remain available as continuations of that history. The notation is illustrative, not a completed formal model. The multiplicity must be **ontic**. It cannot consist only in: - an observer's ignorance; - the agent's inability to predict itself; - represented conditional alternatives; - deterministic chaos; - pseudo-random sampling; - incomplete measurement. If exhaustive determinism is true, the support of the physically realizable next-state distribution is one, whatever broader distribution an embedded agent or observer represents. ## 2.2 Foreclosure, liveness, and exclusive actualization A **foreclosure** occurs when one continuation becomes actual and irreversibly extinguishes one or more other continuations that were genuinely available before the event. The liveness requirement is not an optional embellishment. It is analytic to foreclosure. In ordinary and legal usage, one forecloses an option, interest, right, or continuation that existed as available before being extinguished. If nothing was genuinely available to be lost, nothing was foreclosed in the relevant sense. Foreclosure also requires **exclusive actualization**. Plural support is insufficient if every supported continuation remains actual. On an Everett-style universal branching interpretation, unitary evolution is normally treated as deterministic at the universal level and all decohered branches remain within the universal state. Such a structure may contain branch-relative uncertainty and divergent successor histories, but it does not contain foreclosure in the present stipulated sense unless a further argument establishes that some live continuations genuinely cease to be actual. The theory therefore requires an ontology on which one continuation becomes actual in the relevant history while alternatives are extinguished as possible continuations of that history. Collapse-type interpretations, objective-collapse theories, or non-quantum indeterministic models could in principle satisfy that condition; the paper endorses none of them here. A deterministic process may: - resolve an observer's uncertainty; - calculate several conditional trajectories; - suppress represented candidates; - prune a search tree; - produce one output rather than another under a conditional model. But if only one complete physical continuation was ever realizable, it does not foreclose an ontically live future. Likewise, if every supported continuation remains actual, there is plurality without foreclosure. This analytic point establishes: \[ \text{foreclosure} \Rightarrow \text{live alternatives} \land \text{exclusive actualization}. \] It does **not**, by itself, establish that all defensible concepts of authorship or freedom must involve foreclosure. That further step is the paper's explicit incompatibilist commitment: the relevant freedom is authorship exercised over futures that genuinely could have been otherwise and of which one becomes exclusively actual. Action, policy formation, self-constitution, and causal production can all be freedom-bearing on this account, but only where they participate in the foreclosure or restructuring of live possibilities. Along a uniquely necessitated trajectory, they may constitute complex self-production without constituting freedom in the stipulated sense. ## 2.3 Modal austerity does not rescue determinism Determinism can be stated without using the word *possible*: a complete prior state together with the laws entails every later true proposition. In a familiar formulation, the conjunction of a proposition describing the complete state of the world at one time and the laws of nature entails every proposition describing the world at another time. That entailment formulation does not remove the modal issue. It supplies the basis for the modal gloss that no alternative continuation was physically realizable from the same complete state under the same laws. The determinist therefore cannot preserve the claim that the future **could never have been otherwise** while treating every appeal to physical possibility as meaningless. If possibility and counterfactual structure are rejected entirely, then claims of inexorability, necessity, and could-not-have-been-otherwise are rejected with them. The dispute is therefore not between modality and no modality. It is between: - a transition structure with one physically realizable continuation; and - a transition structure with more than one physically realizable continuation, together with an account of actualization. This point does not establish openness. It prevents the deterministic conclusion from being insulated by simultaneously using and disavowing the modal consequence of entailment. ## 2.4 The first unresolved problem is physical The first unresolved question is prior to neuroscience: > Does genuine indeterminism exist anywhere in reality, and does any such process produce exclusive actualization rather than merely plural description or universal branching? If no indeterminism exists anywhere, Stage 1 fails. If plural continuations exist but none is extinguished, foreclosure as stipulated also fails. Only after those physical conditions obtain do the human questions arise: 1. Does an open transition enter an integrated agent–environment control loop? 2. Does it survive or amplify rather than average into irrelevance? 3. Does it affect a cognitively significant transition? 4. Is that transition on the causal pathway from a self-object intervention to the later-agent change being credited? 5. Is the open transition historically and contextually structured into a bounded band rather than arbitrary multiplicity? 6. Does the self-object process materially alter that very band? The relevant transition need not be wholly intraneural. A deliberately recruited environmental transition may qualify where it is integrated into the agent's control loop, materially shaped by the represented self-change project, and lies on the pathway to later self-constitution. Incidental environmental or bodily openness does not qualify. The theory therefore requires: > **Historically and contextually informed, cognitively integrated, materially relevant, banded ontic openness with exclusive actualization.** This is a physical and empirical posit. It is not established by phenomenology, unpredictability, novelty of thought, or current neuroscience. ## 2.5 Event time and phase time: a corollary If time in the strongest event-generating sense consists in occurrences that irreversibly alter or foreclose future possibility space, then exhaustive determinism contains no ontic foreclosure. Its total history is fixed as one trajectory. Embedded observers may experience uncertainty resolving over time, but from the total deterministic description nothing becomes impossible that was previously possible. The sequence is therefore closer to **phase traversal** through a fixed history than to event time generated by genuine modal closure. This corollary is not required for the authorship argument, but it explains why the difference between epistemic resolution and ontic foreclosure matters beyond vocabulary. --- # 3. Why openness is not authorship Ontic multiplicity is necessary but insufficient. A radioactive decay, under an irreducibly stochastic interpretation, may extinguish live possibilities without authoring anything. A cognitively relevant stochastic transition would likewise not become free merely because its outcome was not necessitated. The luck objection is therefore correct against bare indeterminism: > If the agent has no authorship-relevant relation to the structure, direction, or uptake of an open event, the occurrence is merely suffered by the agent. The theory treats ontic multiplicity as the raw material on which freedom may operate, not as freedom itself. Its second stage asks whether an agent can become materially involved in structuring, routing, and recursively incorporating a genuine foreclosure without becoming an uncaused chooser outside the causal order. The primary proposed answer is not generic higher-level control: it is a self-object loop in which the continuing agent intentionally intervenes upon a represented possible later constitution of itself. --- # 4. Stage 2: the recursively diachronic agent ## 4.1 Developmental emergence Human beings do not begin as self-authors. Early development produces a bounded locus of perception, affect, learning, and action before the organism can hold its own operative regime as an object of contemplation. Autobiographical thought can occur while the self remains only the implicit perspective from which thought proceeds. The developmental distinction is: - **Self as locus:** the organism is the bounded first-person centre from which experience and thought occur. - **Self represented within cognition:** the organism models itself as a participant in autobiographical or future events. - **Self as object of intervention:** the organism represents an operative feature of itself as alterable and intentionally acts toward changing the later self. The third capacity is the proposed Stage 2 hinge. ## 4.2 The maturational wetware floor The loop has a biological floor. A suite of capacities must sufficiently develop before an operative feature of the self can be held as an alterable object and used to guide intentional self-revision. Relevant capacities plausibly include: - bodily and agentive self-models; - autobiographical continuity; - semantic self-knowledge; - counterfactual construction; - metacognitive monitoring; - future-self continuity; - attentional and inhibitory control; - the ability to select environments and practices for their expected effects on the later self. These capacities do not come online at one clean instant. Development is graded, asynchronous, and domain-specific. The theory distinguishes two gates: ### Temporal-maturational gate The relevant capacities do not yet sufficiently exist or integrate. This is the ordinary infant and young-child case. The gate is expected to open gradually through maturation and expected developmental input. ### Developmental-access gate The component capacities exist in principle but remain weakly integrated, unavailable under challenge, domain-limited, or never recruited into stable self-object-directed intervention. This may form a persistent low-access basin without implying literal inescapability. The distinction matters because the infant lacks authorship for want of sufficient operative hardware, not merely for want of prior authored history. ## 4.3 The technical meaning of self-object-directed intervention Not every action that affects the self qualifies. The proposed witness-process has the following structure: 1. The agent represents an operative feature of itself. 2. The feature is represented as alterable. 3. The agent constructs or entertains at least one possible future, outcome, policy, disposition, or successor-self configuration distinct from the projected default continuation. 4. That possible variance is adopted as an aspirational or operative target. 5. The adoption may itself be heavily conditioned by current constitution, bodily state, context, and luck. 6. The agent holds, repeatedly reconstructs, or deliberately returns to that target across time. 7. Incoming sensory evidence, prediction errors, affective shifts, failures, temptations, and environmental constraints perturb the process. 8. The agent evaluates those perturbations relative to the represented target and alters tactics, attention, conduct, environment, relationships, or curriculum accordingly. 9. The later agent inherits some consequences of the process and may renew, revise, abandon, or recursively continue it. The target might concern: - a habitual reaction; - an attentional or sampling policy; - an interpretive regime; - an evaluative tendency; - a behavioural disposition; - a recurring relational response; - an addictive or compulsive trajectory; - a future pattern of inference or action. The central form is: > I am a system that tends to generate response X under conditions C; I can represent a possible variance of myself that instead makes Y more available; I will intentionally direct thought and arrange experience or conduct toward altering the later agent in that direction. The representational loop is diachronically closed. The present agent is the representing locus; that same continuing agent is represented as the object of alteration; and the anticipated successor is a later state of the locus that will inherit the intervention's effects. This recursive identity relation, together with target maintenance and adaptive correction through time, gives the example its distinctive Stage 2 force. Representational plurality is not required. One represented variance against a projected default is enough. The theory locates physical multiplicity in Stage 1, not in the number of futures consciously entertained. This is narrower than first-order self-regulation, autobiographical thought, or ordinary goal pursuit. It is also not restricted to linguistic or narrowly semantic cognition. The relevant representations may be verbal, visual, spatial, affective, motoric, relational, or multimodal. The requirement is content-sensitive agent-level representation, not one privileged format. ## 4.4 Curriculum access, not parameter access Humans cannot directly inspect or edit synaptic weights, precision parameters, learned representations, or low-level transition rules. They can intentionally alter the conditions under which later learning occurs through: - attention; - rehearsal; - repeated enactment; - environmental selection; - exposure or avoidance; - information seeking; - symbolic externalization; - commitment devices; - social scaffolding; - deliberate practice; - treatment, medication, or institutional support; - selection of recurring contexts. The causal route is therefore indirect: \[ \text{self-model} \rightarrow \text{represented possible variance} \rightarrow \text{sustained target} \rightarrow \text{curriculum intervention} \rightarrow \text{possible later change} \] Again, the notation is schematic. No single biological write mechanism is assumed. Reinforcement learning, inhibitory learning, habit formation, memory updating, reconsolidation under appropriate boundary conditions, and broader plasticity may contribute differently across cases. ## 4.5 Sustained directed thought and target maintenance The proposed witness is not exhausted by the moment at which a possible successor is selected. Much of its anti-luck force lies in what follows. A possible successor-self can become a recurrent actual cause when the agent maintains or repeatedly reinstates it as an operative target: \[ R_{t_1} \rightarrow I_{t_1} \rightarrow R_{t_2} \rightarrow I_{t_2} \rightarrow R_{t_3} \rightarrow I_{t_3} \] where the representation may be refined while remaining recognizably directed toward the same transformation. Relevant signatures include: - **target persistence:** later cognition and conduct remain systematically related to the represented variance; - **error correction:** failure changes the means rather than automatically dissolving the project; - **implementation flexibility:** different circumstances produce different actions serving the same or revised direction; - **pattern-sensitivity:** changing the represented successor changes the longitudinal pattern of maintenance, correction, and intervention; - **recursive uptake:** earlier interventions alter the agent performing later passes; - **return after interruption:** the project can be reconstructed after fatigue, distraction, relapse, or competing demands. This signature has an established action-theory backbone. Bratman's planning theory treats intentions, plans, and policies as relatively stable structures that organize deliberation, coordinate means across time, and constrain reconsideration. Holton's account of resolutions emphasizes the work of maintaining intentions under temptation, while Ainslie's analysis of intertemporal conflict and commitment devices shows how agents can restructure later choice environments. The present paper applies those diachronic-agency resources to a narrower self-object content: the maintained plan concerns alteration of the continuing agent's own later operative profile. A Fristonian implementation gloss is available but not required: a higher-level target belief may be maintained with enough precision to continue constraining lower-level inference and action, while prediction error revises local means, timing, and intermediate models. The theory does not equate target maintenance with rigid immunity to error. The relevant pattern is persistence with adaptive correction. Luck may affect every constituent transition. The claim is not that one instant is insulated from luck. It is that luck alone is not an adequate explanation of a temporally extended trajectory that repeatedly covaries with a represented target across changing perturbations. ## 4.6 Recursion as the mechanism of accrual The theory's subject is not an isolated act but a **recursively diachronic agent**. Recursion is used here in a specific causal and diachronic sense. It is not mere repetition, temporal persistence, self-reference, or generic feedback. A process is recursive in the relevant sense where the output of one self-object trajectory alters the operative agent-state that generates, selects, sustains, and corrects a later trajectory. Schematically: \[ A_n \rightarrow M_n(A_n) \rightarrow R_n \rightarrow I_n \rightarrow A_{n+1}, \] followed by: \[ A_{n+1} \rightarrow M_{n+1}(A_{n+1}) \rightarrow R_{n+1} \rightarrow I_{n+1} \rightarrow A_{n+2}. \] Here \(A_n\) denotes the operative agent-state at pass \(n\), \(M_n\) a self-model, \(R_n\) a represented possible variance, and \(I_n\) a directed intervention. The notation is schematic rather than a completed update rule. The recursion lies in the carry-forward relation. Effects of \(I_n\) may alter the later agent's: - probability of noticing a self-relevant problem; - available successor-self representations; - target-maintenance and error-correction capacities; - attentional and sampling policies; - intervention repertoire; - environmental and social access; - tolerance for perturbation, failure, or temptation; and - propensity to initiate or resume another self-object project. The later object and locus of intervention is therefore not necessarily the originally inherited agent. It may be an agent whose dispositions, habits, environments, conceptual resources, and intervention-capacities already bear effects of earlier self-directed trajectories. Recursion does not explain the first qualifying contribution from nothing. The first candidate authored foreclosure may issue from an agent whose constitution is overwhelmingly unchosen. Recursion explains what becomes possible after such a contribution: the result may enter the constitution from which a later contribution is generated. Constitutive luck and prior self-directed contribution can therefore be transmitted together. This creates the possibility of recursively accumulated self-authorship over time. A later agent may operate from a profile bearing a larger contribution from prior qualifying trajectories, and its next trajectory may add, revise, counteract, or erase some of that contribution. The claim is comparative, path-dependent, and non-monotonic. --- # 5. Authored foreclosure and accrued self-authorship ## 5.1 Event-level authored foreclosure A foreclosure is a candidate **authored foreclosure** where: 1. more than one continuation was genuinely realizable and one became exclusively actual; 2. the relevant open transition lay within an integrated agent–environment control loop and on the causal pathway from the self-object intervention to the later-agent change being credited; 3. the transition formed part of a prospectively individuated, diachronically sustained self-object trajectory; 4. the agent intentionally directed thought or action toward a represented possible variance of its later self; 5. the represented target made a material, content-sensitive causal difference to the support, weighting, accessibility, persistence, strategy selection, or uptake of the very open transition band in which the foreclosure occurred; and 6. the event is attributable to the historically continuous agent without requiring that the agent be the sufficient contrastive cause of the exact micro-resolution. The fifth condition combines **band-sensitivity**, **relevance-binding**, and **materiality**. Band-sensitivity asks whether changing the represented transformation changes the successor-state distribution of the relevant transition class. Relevance-binding requires that openness and routing not merely coexist somewhere within the same body, interval, or extended project. Materiality requires that the altered transition lie on the pathway to the later-agent difference actually credited to the project and make a non-trivial difference to that pathway. An incidental quantum event in a side process, even if minutely affected by the agent's state, does not create an authored foreclosure. The open transition may be neural or environmental. Environmental openness qualifies only where the agent intentionally recruits or shapes the relevant setup, the transition is integrated into the self-change project, and its result is incorporated into the later trajectory. The condition is functional and causal, not spatial. Authorship attaches directly to the qualifying contribution and trajectory. It does not radiate from a reflective organism to every stochastic event occurring within it. Autonomic, sub-personal, accidental, merely collocated, causally immaterial, and unrelated foreclosures remain unauthored. ## 5.2 The trajectory-level witness In one proposed expression, selfhood, possible freedom, and relevant authorship arise within intentionally directed, sustained thought applied to a represented variance of the self across time. The proposed authorship relation does not arise merely because a higher-level representation alters lower-level processing. Automatic priors, externally installed goals, and stochastic control systems may also display top-down organization. The stronger attribution arises from the trajectory's recursive diachronic structure. The current agent represents an operative feature of its own constitution, represents possible successor constitutions of that same continuing agent, selects or adopts one variance as a target, maintains or reconstructs that target through perturbation, repeatedly changes means relative to incoming feedback, and later inherits the consequences. The agent therefore appears within the process in at least three related positions: - the present representing locus; - the represented object of alteration; and - the anticipated and later realized bearer of the effects. The trajectory, rather than an isolated stochastic draw, is the primary unit of Stage 2 analysis. Not every constituent transition must independently reinstantiate the full witness. The trajectory may include broadly open transitions, heavily banded open transitions, deterministic sub-processes, interruptions, and later reconstruction. At least one transition must remain ontically open and exclusively actualizing, and the sustained self-object project must materially affect that transition's own band. Openness merely present somewhere else in the trajectory or organism does not satisfy the conjunction. The trajectory supplies diachronic unity, but recursion supplies accrual. A single qualifying trajectory may establish one nonzero contribution. Only where its effects alter the later agent-state from which another self-object trajectory is generated does the contribution become recursively carried forward into subsequent authorship. The trajectory must also be individuated prospectively rather than gerrymandered retrospectively. Its identity conditions should include: - a target, direction, or constraint identifiable before the credited outcome; - causal continuity between earlier and later target-instantiations through memory, rehearsal, standing policy, external scaffolding, or another traceable route; - contemporaneous use of the target in error correction and strategy revision; - outcome-independent criteria for deciding whether later revisions remain part of the same project; and - pattern-sensitivity, such that materially changing the represented transformation would alter the longitudinal intervention pattern. The target need not be continuously conscious, verbal, or unchanged. It may be periodically reinstantiated, externally scaffolded, or revised while retaining causal continuity. What is excluded is a post hoc narrative that defines the project as whatever sequence happened to produce the result later credited. ## 5.3 Interventionist routing: content as a causal difference-maker The paper adopts an interventionist account of the routing relation (Woodward 2003). In Woodward's terms, a causal relation is supported where an appropriately specified intervention on one variable would produce a systematic change in another under a suitably invariant relationship. Pearl's \(do\)-operator (Pearl 2009) provides a convenient notation for the idealized contrast. The central anti-luck decomposition distinguishes: - \(O_I\): **occurrence** — that intervention \(I\) occurred; - \(R_I\): **representation** — the content and organization of the represented self-transformation; and - \(E_I\): **efficacy** — the intervention-attributable effect of that representation on the relevant successor band and later trajectory. The occurrence of \(I\) may be determined, ontically open, or luck-shadowed. That fact does not settle what representational organization \(I\) possessed or what causal difference that organization made. Let: - \(B_R\) denote relevant properties of the successor band following an intervention organized around represented target \(R\); and - \(B_0\) denote the comparison band in a nearest relevant trajectory lacking the qualifying self-object intervention, or in an empirically matched control condition. Then, schematically: \[ E_I := \Delta(B_R,B_0), \] where \(\Delta\) does not imply literal vector subtraction. It denotes intervention-attributable differences in support, weighting, accessibility, persistence, generalization, strategy selection, or later uptake. Let \(K\) denote a transition class characterized at a level capable of cross-register investigation. Decompose the relevant prior state as \(S_t=(Z_t,R_t)\), where \(R_t\) is the represented self-transformation and \(Z_t\) is the matched prior-state background excluding that representational variable. **Band-sensitivity** requires that intervention on the represented transformation change the successor-state distribution associated with that class while the matched background is held fixed: \[ P(S_{t+1}\in\cdot\mid Z_t=z,K,do(R_t=r_1)) \neq P(S_{t+1}\in\cdot\mid Z_t=z,K,do(R_t=r_2)). \] For a freedom-bearing witness, the actual full prior state \(S_t^{(a)}=(Z_t=z,R_t=r_a)\) must also be ontically open and exclusively actualizing: \[ \left|\operatorname{support}P(S_{t+1}\in\cdot\mid S_t=S_t^{(a)},K)\right|>1, \qquad \operatorname{EA}(S_t^{(a)},K,S_{t+1}). \] The interventionist comparison and the openness condition therefore quantify over different but related state descriptions. Band-sensitivity varies \(R_t\) across matched \(Z_t\); ontic openness is assessed relative to the resulting complete actual prior state. This is also the structure approximated by the matched-controls empirical program: hold the non-target background as stable as possible while varying represented content, then separately investigate whether the relevant realized transition class had objectively plural support and exclusive actualization. These expressions are schematic. Their purpose is to exclude conjunction by collocation and to keep the relata properly typed. A self-object project plus an unrelated open event elsewhere in the organism is not enough. The routed representation must materially alter the successor-state band of the same class to which the physical openness claim applies, and that transition must lie on the pathway to the later-agent effect credited to the intervention. Pattern-sensitivity and band-sensitivity perform different jobs. Pattern-sensitivity helps individuate a prospectively identifiable sustained project by asking whether its planning, maintenance, and correction track the represented target. Band-sensitivity is the routing test: it asks whether changing the represented target changes the successor-state distribution of the relevant transition class. The witness does not pass the routing criterion merely because pattern-sensitivity was used to identify the trajectory. Changing what successor-self transformation is represented and maintained should systematically change the direction of the resulting intervention and later band. The effect should not be reproduced merely by matched arousal, generic effort, exposure, therapist attention, or non-self-directed planning. Actual experiments cannot surgically set an internal representation while holding every other variable fixed. The notation states the causal ideal. Empirical work must approximate it with active controls, component ablations, within-person target variation, and manipulation checks separating content from effort, expectancy, difficulty, arousal, and exposure. This routing criterion is deliberately **modal-neutral**. A deterministic system may also exhibit content-sensitive self-constitution. Interventionist routing supplies the sourcehood or attribution factor only. Ontic liveness and exclusive actualization supply the independent leeway condition. ## 5.4 The first instance and iterated rollback The first minimally authored trajectory cannot require that the acting self-model, target-maintenance capacity, or intervention disposition already be substantially produced by prior authorship. That would make the category circularly unavailable. The first qualifying agent is overwhelmingly unchosen. Its developmental history, concepts, values, bodily state, attentional capacities, and access to self-object representation are constituted by causes it did not author. The theory concedes this fully. The occurrence–representation–efficacy decomposition identifies what iterated rollback does and does not show: 1. **Occurrence \(O_I\).** The token intervention may be determined, ontically open, or luck-shadowed. 2. **Representation \(R_I\).** The intervention may possess a particular target-relative organization concerning a possible variance of the continuing agent. 3. **Efficacy \(E_I\).** That content may systematically organize attention, conduct, environmental selection, error correction, and later uptake relative to a no-intervention or differently routed comparator. An iterated rollback objection resets the world to a point before the intervention and observes that different replays may produce intervention \(I\), intervention \(J\), or no intervention. This shows that \(O_I\) may itself require no prior authored provenance and may retain present luck. It does not show that, when \(I\) occurs, \(R_I\) makes no systematic causal difference or that \(E_I=0\). Noise can trigger, interrupt, or influence a project, and a lucky fluctuation can produce isolated target-congruent behaviour. Noise alone, however, does not explain a stable counterfactual mapping in which changing the represented successor changes the direction of a sustained, error-corrective trajectory and the materially relevant open band across matched conditions. The first contribution is therefore minimal and conditioned. It is attributed not because the agent authored the capacities from which it acted, but because the continuing agent's represented possible variance repeatedly organized what that agent did to the conditions from which its successor emerged. ## 5.5 Constitutive luck, luck transmission, and conditioned contribution Constitutive luck is not merely an infection of the agent. It is constitutive of the agent. The original band, evaluative standards, self-representational capacity, and available aspirations are overwhelmingly unchosen. A luck-transmission principle argues that an exercise of an unchosen capacity inherits the luck of that capacity's acquisition unless the agent possessed freedom-level control at an earlier point. Applied recursively, the demand threatens regress. The paper accepts transmission of **limitation** but rejects automatic transmission of **annihilation**: > Unchosen provenance limits the extent, independence, and downstream normative significance of authorship; it does not by itself reduce a present, content-sensitive, diachronically sustained agent contribution to zero. A capacity need not have self-authored provenance in order for its exercise to be causally attributable to the system exercising it. Anti-luck epistemology supplies more than a remote analogy. Levy's account of chancy luck draws upon the modal-fragility tradition developed by Pritchard. Within the broader epistemic-luck literature, however, Pritchard (2005) distinguishes knowledge-undermining veritic luck from other forms of luck, while Greco (2009) treats knowledge as success through ability and Sosa (2007) treats apt belief as success because of competence. These distinct approaches are not one unified theory, but they share the refusal to require that a knower author the provenance of the faculties through which epistemic success occurs. The comparison does not prove that freedom-attribution must follow epistemic attribution. It establishes an attributional asymmetry that Levy's freedom-specific argument must defend: why does unchosen provenance condition but not annihilate epistemic credit, while it is alleged to annihilate even partial authorship-credit? The modal analysis of luck itself does not supply that disanalogy. The causal contribution is an empirical claim. Its classification as **partial authorship** rests on the paper's defended non-ultimate conception: authorship is a content-sensitive causal contribution made by the continuing agent to what later occurs or what that agent later becomes. A critic can deny the contribution, accept the contribution while reserving the word *authorship*, or propose a stronger attribution condition. That classificatory concession creates a targeted **definition-squeeze**. If the stronger condition is simply luck-free provenance or complete self-authorship of the constitution from which the exercise issues, it encounters Strawson's Basic Argument: the agent would need already to possess the constitution required to author that constitution. Such a standard does not identify a higher realizable grade of authorship; it fixes the extension of authorship at zero by construction. Other stronger standards—such as primitive agent-causal settlement or control over the exact token—must be assessed on their own terms, but cannot be treated as conceptually mandatory merely because the contribution remains conditioned. The resulting dispute bottoms out cleanly: does provenance-luck transmit through a sustained content-sensitive exercise so as to erase the classification *partial authorship*, or does it limit a real attribution relation without annihilating it? Nothing in the existence of constitutive luck alone answers that question. ## 5.6 Recursive accrual at the agent level The continuing agent, rather than the isolated event, is the bearer of **accrued self-authorship**. This term names a potentially increasing contribution from prior qualifying trajectories to the agent's present operative profile, not a substance, binary status, or guaranteed upward path. The central constructive claim is the **recursive accrual thesis**: > Where a qualifying self-object trajectory materially alters the later operative profile of the continuing agent, and that altered agent subsequently performs another qualifying trajectory, the earlier contribution becomes part of the causal constitution from which the later contribution arises. Repeated instances may therefore produce accrued self-authorship without requiring ultimate self-origination or monotonic increase. The causal composition of a later agent can be represented schematically as: \[ A_{n+1} = F(U_n, X_n, \Omega_n, C_n), \] where \(U_n\) denotes unchosen inherited and developmental conditions, \(X_n\) effects attributable to prior self-object-directed trajectories, \(\Omega_n\) stochastic and uncontrolled influences, and \(C_n\) wider contextual causes. This is not an additive metric or a claim that those contributions can presently be numerically separated. It displays the conceptual point that unchosen provenance is not the only history carried forward. Where a trajectory affects later habits, attentional tendencies, policy dispositions, environments, conceptual resources, relationships, treatment access, or learning conditions, the later agent is partly constituted by prior self-object-directed activity. Later ordinary cognition need not itself involve explicit self-object contemplation to bear derivative authorship. An inference or action may emerge from priors, precision policies, sampling tendencies, or accessible representations partly shaped by earlier qualifying trajectories. Earlier interventions may also alter the probability that later interventions occur. A person may become more likely to notice a problematic disposition, form a successor-self target, seek support, maintain a project, and return after failure. Occurrence-luck may therefore be heaviest at the first instance and become less dominant across successful recursive accrual, without ever disappearing. Recursion thereby unifies four features otherwise easy to treat separately: 1. **Accrual:** prior self-directed contribution enters the later operative profile. 2. **Sedimentation:** repeated contributions become embodied in dispositions, policies, environments, and character. 3. **Intervention reliability:** earlier passes may increase the probability and stability of later passes. 4. **Curriculum expansion:** one pass may alter the tools, people, contexts, and learning resources available to the next. None of these consequences is guaranteed. Recursion can transmit failure, rigidity, maladaptive targets, fragmentation, or reduced access as readily as improvement. ## 5.7 Potential increase without monotonicity Self-authorship may increase, but no monotonic law is claimed. A qualifying trajectory may: - fail entirely; - have only local effects; - produce unintended changes; - conflict with another self-directed project; - be displaced by illness, coercion, trauma, environment, or ordinary development; - be reversed; - reduce later metacognitive access; - successfully alter one policy while damaging another; - narrow one band while opening or closing others. Accordingly, self-authorship can increase, remain stable, plateau, fragment, reverse, or diminish. No claim is made that the contribution of origin approaches zero or that enough time guarantees increasing ownership. The paper does not yet possess a defensible scalar measure of the fraction of a present band attributable to authored rather than unchosen causes. For now, authorship contribution is primarily ordinal and comparative. ## 5.8 Liveness, stabilization, and inherited expression Probability concentration is not determinism. If \(P(A)=0.99\) and \(P(B)=0.01\), and both continuations remain genuinely physically realizable, the transition is highly banded but still ontically open. Successful self-formation may make target-congruent outcomes increasingly robust. This can reduce their modal fragility and the degree of chancy outcome luck without eliminating constitutive luck. It can also make some later behaviours effectively automatic or, in principle, strictly determined by an acquired disposition. Three distinctions follow: 1. **Fresh event-level freedom:** a transition is a new authored foreclosure only where local ontic liveness, exclusive actualization, and agent-level routing are all present. 2. **Trajectory-level witness:** the extended self-variance project may include highly constrained or deterministic sub-transitions organized by the same sustained target. 3. **Inherited expression:** later determined conduct may express a disposition partly formed through earlier freedom-bearing trajectories without constituting a fresh authored foreclosure. Two currencies must be kept separate. - **Modal liveness** concerns whether more than one continuation is physically realizable at a transition. - **Reachable-state richness** concerns the range of substantively different states, activities, relationships, projects, and environments available across a longer horizon. A deterministic agent can gain reachable-state richness. Narrowing a maladaptive local band may therefore reopen relationships, projects, health, work, mobility, and future self-intervention without proving that ontic liveness has increased anywhere. Recovery from addiction illustrates this independent benefit: stabilizing sobriety may narrow the local use/no-use band while greatly enlarging the person's reachable life-space. The freedom-specific redistribution claim is conditional. If ontic openness occurs within that enlarged reachable space, successful self-formation may relocate or increase the occasions on which openness can become agent-relevant. It is not a structural theorem that option-richness contains more ontic liveness. The theory therefore predicts neither that the most self-formed agent is simply the least free nor that every expansion of life possibilities is an expansion of modal openness. It predicts that successful formation changes where fresh freedom may occur, how earlier authorship is expressed, and which wider states become reachable. ## 5.9 Actualized states, open transitions, and sedimented authorship At any time \(t\), the actualized state \(S_t\) is fixed in the limited but important sense that it is now a fact. It is the realized product of prior causal history, including prior stochastic resolutions and authored or unauthored foreclosures. Alternatives previously open to that event are no longer co-actual with it. This must not be confused with deterministic necessitation. A state can be determinate as actual without having been the only physically realizable outcome before it occurred. Actuality answers what is; determinism answers whether anything else could have become actual from the same complete relevant prior state. Ontic openness therefore attaches primarily to the transition relation: \[ S_t,C_t \rightarrow P(S_{t+1}). \] Foreclosure occurs when one live continuation becomes exclusively actual and alternatives cease to remain available. Once that happens, the foreclosure is historical. Its effects may nevertheless be sedimented in the new state: the later agent can carry habits, policies, memories, environments, and learning conditions partly produced by the earlier trajectory. This supports a positive **sedimentation thesis**: > In one proposed expression, freedom-bearing trajectories characteristically convert some open, agent-routed transitions into relatively stable dispositions from which later conduct proceeds. Accrued self-authorship is the sedimented contribution of those trajectories to the continuing agent. The metaphor does not imply conservation of a substance called freedom. It identifies a historical relation. Earlier authored foreclosures can become embodied in character, policy, environment, and intervention-capacity. Later robust or determined conduct may express that sedimented authorship without constituting a fresh foreclosure. Freedom is therefore not adequately located in a timeless snapshot. At the event level, it concerns an agent's relation to an open transition and its foreclosure. At the trajectory level, it concerns sustained self-object-directed organization across perturbation. At the agent level, accrued self-authorship is a history-dependent property of a recursively diachronic agent presently constituted partly by prior qualifying trajectories. # 6. The path-dependent and banded structure ## 6.1 Path-dependent in the primary sense Human cognition is modelled as a path-dependent generative process in which each realized state becomes part of the conditioning context for what follows. Schematically: \[ P(S_{t+1}\mid S_{\leq t},C_t) \] where \(S_{\leq t}\) represents the relevant realized history and \(C_t\) present context. The primary term is **path-dependent**. “Autoregressive” is retained only as a limited analogy: what becomes actual at time \(t\) changes the structured band of what is reachable at time \(t+1\), without necessarily determining a unique successor. The paper does not assert token-by-token human cognition or import a complete language-model architecture. A thought can alter: - which memory is retrieved next; - which contradiction becomes salient; - which perspective is simulated; - which affective response gains precision; - which action becomes available; - which future thought-band is entered. ## 6.2 Historically and contextually informed bands The relevant multiplicity is not unrestricted. The current band is constrained by: - biological endowment; - developmental history; - memory; - learned representations; - current bodily state; - environmental context; - attention; - active goals; - evaluative commitments; - prior self-object-directed interventions. The theory therefore does not posit a chooser standing outside history. It posits a historically constituted distribution with more than one live continuation. ## 6.3 Counterfactual simulation as reparameterization Representing a non-actual possibility is itself an actual cognitive event. It can change the later state of the deliberating system. Counterfactual simulation may: - make a previously inaccessible trajectory reachable; - increase or decrease the probability of an existing trajectory; - narrow or widen the active band; - alter affective weighting; - reveal a new affordance; - suppress an alternative; - produce fixation or rumination; - change later action and learning. The general claim is therefore not that imagination always expands the reachable manifold. It **reparameterizes the reachable-state distribution**. ## 6.4 Bounded release from invariant top-policy enactment Stochasticity does not itself create authorship. It may, however, prevent the currently highest-ranked policy from being invariably enacted. A deterministic argmax process always realizes the top-ranked continuation, however slight the margin. A bounded stochastic process permits lower-ranked but still supported continuations to occur in proportion to their support. The warranted claim is modest: > Bounded stochasticity can permit exploration of supported alternatives that deterministic top-policy enactment would never realize. This may supply learning opportunities and alter future bands. It is not intrinsically more agentic at the instant of sampling, and excessive stochasticity can dissolve rather than enhance ownership. --- # 7. The luck objection and the location of freedom ## 7.1 Freedom is not located in a privileged instant The standard luck objection freezes the process at one resolution: > At the instant A rather than B occurred, what made A yours rather than luck? The theory rejects that as the exclusive location of freedom. It does not require an agent to determine an indeterministic draw without determining it. That demand recreates the original dilemma inside a privileged moment. In one proposed expression, the freedom-bearing structure is a temporally extended self-variance trajectory. It includes: 1. the historically formed agent-band; 2. the representation of present and possible successor constitutions; 3. adoption and maintenance of an operative target; 4. repeated target-relative interpretation of incoming feedback; 5. flexible intervention through thought, conduct, environment, and curriculum; 6. at least one ontically open transition whose own band is materially affected by that process; and 7. later uptake into the continuing agent. The exact micro-resolution may remain lucky. The paper's claim is not that every constituent event loses all luck, but that luck at one or many resolutions need not exhaust the authorship status of the organized trajectory. ## 7.2 Rollback and iterated rollback In van Inwagen's canonical rollback argument (2000), the universe is reset to the same complete prior state and replayed. If outcomes converge upon a distribution, the objector concludes that a particular token result was a matter of chance. Where rollback begins after a qualifying intervention has already shaped the band, it repeatedly samples from the routed band. It shows that the intervention did not uniquely settle the exact token outcome. It does not compare the routed band with a relevant no-intervention or differently routed comparator and therefore does not show that the intervention made no causal difference. Where rollback begins before the intervention, different interventions or no intervention may occur across replays. This establishes that the intervention's occurrence may itself be open and luck-shadowed. It still does not establish that the intervention's content lacks systematic causal efficacy when instantiated. The same interventionist criterion answers both versions: vary the represented transformation, and ask whether the sustained trajectory and successor-band properties vary correspondingly across suitable controls. Rollback variance is consistent with authored openness. Treating variance itself as attribution-excluding presupposes that authorship requires unique settlement of the token outcome. ## 7.3 The exact contrastive-settlement demand The theory does not deny probabilistic contrastive explanation. Agent-facts may make one continuation much more likely than another. It rejects the stronger demand that freedom requires a complete prior agent-fact sufficient to settle why this exact token was A rather than B. Under genuine ontic openness, such complete settlement would eliminate the liveness whose authorship is under examination. The hidden standard is: \[ \text{To count as authored, an undetermined outcome must have been determined by the agent.} \] That standard analytically excludes every event-causal incompatibilist account. The paper instead distinguishes the work done by the two conditions: - ontic liveness makes more than one continuation real; - agent-level routing attributes a content-sensitive causal contribution to the continuing agent. Neither component must perform the other's job. ## 7.4 Residual luck, transmission, and the classification boundary Luck remains at inherited constitution, developmental opportunity, target generation, intervention occurrence, exact resolution, success, and unforeseen consequence. The paper denies only: \[ \text{residual or transmitted luck} \Rightarrow \text{zero agent contribution}. \] The self-object trajectory may be overwhelmingly luck-conditioned while still systematically preserving a represented variance, correcting course relative to it, and altering the later agent. Constitutive luck limits the contribution and blocks any simple inference to ultimate sourcehood or desert. It does not independently demonstrate that the contribution is absent. The remaining boundary is partly classificatory. The paper calls that real, conditioned, non-ultimate causal contribution **partial authorship**. A theorist who defines authorship as luck-free provenance or exact contrastive settlement will reject the label. The disagreement should be stated rather than hidden. The targeted definition-squeeze is now available. A critic who rejects the label solely because provenance is not luck-free must explain why conditioned agent-level causal authorship is conceptually impossible or adopt an ultimate-sourcehood condition. The latter risks specifying a non-instantiable type. This is a defensive inversion of Strawson's Basic Argument (1994). Strawson argues that ultimate responsibility would require an agent to be responsible for the constitution from which its action issues, while any attempt at intentional self-formation proceeds from an earlier constitution for which the same question recurs. The present paper accepts the impossibility of causa-sui provenance but rejects the inference to zero partial authorship. The squeeze is not exhaustive. A critic may instead require primitive agent causation, exact token settlement, or another stronger relation. Those proposals require separate defence. The paper's claim is narrower: luck-free total provenance cannot be treated as the default meaning of authorship when it is structurally unavailable to temporally constituted beings. Authorship is itself a causal relation; authorship wholly outside causal constitution would be causally idle, while primitive agent-causal sourcehood relocates the unexplained origin into an irreducible exercise of causal power. Nothing in this inversion establishes basic-desert responsibility. It establishes only that absence of causa-sui provenance cannot, without further argument, entail absence of every lesser causal authorship relation. ## 7.5 No homuncular or primitive settler The theory does not posit a further substance-self above the process that chooses which stochastic continuation wins. The agent is, for present purposes, the historically continuous, bodily realized, representationally integrated trajectory that bears the target, performs corrections, marshals its causal ecology, and inherits the result. The agent need not be an additional causal item inserted beside its physical realization. Primitive agent causation, as defended by Chisholm and in modern form by figures such as O'Connor, is therefore unnecessary for the present theory. Saying that an enduring substance directly causes an act through a basic causal power does not explain where that power came from or why it is exercised in this direction at this time. If prior conditions sufficiently settle the exercise, the view returns to determination. If nothing settles it, the primitive exercise functions as an uncaused causal terminus. The proposed witness instead locates partial sourcehood within the organization of ordinary causal processes across time: \[ \text{represented variance} \rightarrow \text{sustained target} \rightarrow \text{feedback-sensitive correction} \rightarrow \text{altered causal ecology} \rightarrow \text{later agent}. \] --- ## 7.6 The mere-addition objection: compatibilism plus noise The strongest additive objection says that Stage 2 routing is already available to the deterministic twin, while the exact open outcome remains luck-influenced. The open witness therefore appears to be nothing more than compatibilist self-modification plus ambient chance. That description treats the two factors as merely collocated. The theory's relevance-binding condition denies the decomposition. - In the deterministic twin, the self-object process may causally shape the later agent and the mechanism's conditional response profile. But from the complete actual state, the physically realizable successor structure is degenerate: only one continuation has support. There is therefore no nondegenerate live band for the routing to shape and no alternative continuation to foreclose. - In bare chance, live alternatives may be exclusively actualized, but no sustained self-object process materially shapes the band from which actualization occurs. - In the proposed witness, the represented self-change project materially alters the very band whose live continuations are then exclusively actualized. The constructive core is therefore: \[ \boxed{\text{band-authorship under ontic liveness and exclusive actualization}.} \] The conjunction adds a relation absent from either component alone: the agent is a content-sensitive partial source of an objectively nondegenerate successor structure from which a genuinely non-necessitated continuation becomes actual and alternatives are foreclosed. Where the governing physical theory supplies a numerical chance measure, this may be described as an agent-shaped objective chance distribution; the theory does not require every admissible indeterministic ontology to provide such a measure. The deterministic twin has source-like routing without leeway. Bare chance has leeway without source-like routing. The witness locally binds leeway and sourcehood. A critic may still deny that this bound relation merits the word *authorship*. But once the open band, the content-sensitive shaping, the material pathway, and exclusive actualization are conceded, the objection has reduced to the classification boundary and the stronger standard must be defended. It is no longer an argument that the conjunction contains no additional causal or modal fact. # 8. Nearest comparators and theoretical position ## 8.1 Kane's self-forming actions: the nearest libertarian relative Robert Kane's theory of self-forming actions (Kane 1996, 2009) is the nearest established libertarian relative of the present proposal. Kane does not require every later free action to be indeterministic. He locates ultimate responsibility in earlier undetermined self-forming actions through which agents help form the wills, motives, and characters from which later conduct proceeds. His paradigmatic cases involve torn decisions in which an agent has strong reasons on both sides and is voluntarily striving toward incompatible outcomes. Whichever effort succeeds helps form the later will. The common ground is substantial: - indeterminism must be embedded in an organized purposive process rather than added as an isolated randomizer; - freedom is not located solely in micro-control of the exact neural resolution; - earlier qualifying events can alter the later will or agent-band; - present agency is historically structured; - self-formation can confer derivative significance upon later conduct. Kane also argues that an agent may retain macro-level purposive control despite indeterministic neural interference. That point is close to the present paper's insistence that authorship need not consist in controlling individual neurons or settling the exact micro-outcome. The differences are nevertheless decisive. | Kane | Accrued Self-Authorship | |---|---| | Ultimate responsibility is central. | The target is nonzero partial authorship. | | Paradigmatic SFAs are torn choices between competing motives. | The primary witness is a sustained self-object trajectory directed toward a represented successor self. | | Both competing outcomes are voluntarily and rationally willed. | No requirement that every live branch be endorsed or simultaneously pursued. | | The choice outcome helps form the will. | The agent repeatedly reshapes the successor band through target maintenance, feedback correction, and curriculum access. | | Kane seeks responsibility for later actions through tracing to SFAs. | Later conduct may express accrued authorship, while culpability and desert remain bracketed. | | The project is a general libertarian account. | The project seeks one existential witness-location. | The present theory does not merely weaken Kane's threshold. It changes the temporal and representational shape of the primary witness. Kane's central structure is approximately: \[ \text{divided present will} \rightarrow \text{undetermined choice} \rightarrow \text{formed later will}. \] The present proposal is: \[ \text{current agent} \rightarrow \text{represented possible variance} \rightarrow \text{sustained target} \rightarrow \text{adaptive curriculum intervention} \rightarrow \text{altered later agent}. \] Its anti-luck strategy also differs. Kane aims to show that either outcome can be the agent's achievement because the agent is voluntarily striving toward both. The present theory permits the exact token outcome to retain outcome luck. Its authorship claim attaches to the agent's sustained, content-sensitive contribution to the trajectory and later band. Kane should therefore be treated as the nearest libertarian relative, not as either an unacknowledged duplicate or an opponent to be dismissed. ## 8.2 Ismael: self-governance and self-constitution without the present Stage 1 Jenann Ismael supplies a powerful naturalistic account (Ismael 2011, 2016) of self-governance. Her self-representational loop models a system that integrates information about itself and its environment, locates itself within a representation, forms a plan, issues control to its machinery, observes consequences, and updates the representation. The control centre need not be a literal inner commander; it can be distributed across the organization of the system. This is close to the present paper's Stage 2 picture of agent-level representational causation. Higher-level representations can marshal lower-level physical processes without requiring a ghost outside the substrate. Ismael also treats causal structure as a resource for intervention and self-governance rather than merely a chain that constrains the agent. This closely neighbours curriculum access: the agent exploits causal pathways in its environment and body to alter what later occurs. The difference lies in the modal condition. Ismael's open future is grounded primarily in the situated perspective and intervention structure of an embedded deliberator: agents possess records of the past rather than the future, intervene asymmetrically toward what comes next, and participate in producing that future. Her account can preserve meaningful practical openness and self-governance without requiring that the same complete physical state possess multiple objectively realizable continuations. The present paper deliberately demands more for the stipulated form of freedom. A deterministic Ismaelian self-governor would satisfy much of Stage 2 and could genuinely self-modify, but would fall at Stage 1. Thus: - Kane is the nearest comparator for indeterminism plus self-formation; - Ismael is the nearest comparator for naturalized self-representational governance; - the present proposal conjoins ontic liveness with a narrower sustained self-object trajectory and seeks only one partial-authorship witness. ## 8.3 Levy: hard luck and the remaining point of collision Neil Levy's *Hard Luck* (2011) develops a modal account of chancy luck together with a pervasive account of constitutive luck, then argues that luck undermines free will and moral responsibility. Levy does not deny the reality of intentional machinery, self-reflection, or historical self-modification. His challenge is that the capacities, values, and standards through which an agent acts are themselves constitutively lucky, while indeterministic resolution adds present luck. The present paper agrees that the agent is constituted by luck and that exact outcomes may be chancy. Its narrower disagreement is whether those facts force every agent-level contribution to zero. The interventionist criterion changes the location of the dispute. If a represented successor-self target systematically organizes a sustained trajectory and changes the later agent, the causal contribution is no longer hypothetical. Levy may still deny that a contribution with lucky provenance qualifies as freedom-level authorship. The disagreement then concerns the attribution standard rather than the existence of the contribution. Levy's modal account also creates two internal pressures. First, as successful recursive self-formation makes target-congruent outcomes robust across a larger proportion of nearby relevant worlds, the modal-fragility component of chancy luck may decrease. This does not eliminate constitutive luck, but it shows that recursive band-shaping can change the kind and degree of luck present in later conduct. Second, Levy develops his account of chancy luck from modal machinery associated with Pritchard's epistemic-luck analysis. Yet prominent epistemic credit theories permit attribution to competent exercises of wholly unchosen capacities. These are distinct theories rather than one package, so no contradiction follows. The burden is instead one of attributional disanalogy: Levy's freedom-specific argument must explain why provenance-luck annihilates authorship-credit while the epistemic traditions from which the modal machinery develops permit competent exercises of unchosen faculties to retain credit. Because this paper brackets basic-desert responsibility, success here would not refute Levy's entire normative project. It would challenge the narrower claim that luck-conditioned indeterminism universally excludes all partial authorship. ## 8.4 Pereboom: the disappearing-agent objection Derk Pereboom's disappearing-agent objection targets event-causal libertarianism at precisely the point where antecedent agent-involving events leave open which decision occurs. If the beliefs, desires, efforts, and intentions do not settle the result, the objection asks what role remains for the agent in the difference between outcomes. Indeterminism appears to leave a gap rather than add control. The present theory accepts that the agent need not settle the exact token. Its answer is that the agent's contribution is not missing merely because token settlement remains open. The represented target, sustained project, and curriculum intervention settle or reshape properties of the band: which continuations are supported, how they are weighted, what becomes accessible, and how the result is incorporated. The agent's contribution is therefore to the open transition structure, not to a mysterious final shove selecting one member. Pereboom can reply that this remains insufficient for freedom-level control. The disagreement is then exact: does control require settlement of the token, or can content-sensitive authorship of a live band constitute nonzero partial sourcehood? The present paper chooses the second and binds it to exclusive actualization. ## 8.5 Mele: modest libertarianism and probability shaping Alfred Mele's *Free Will and Luck* (2006) is a close event-causal neighbour. Mele develops soft or modest libertarian responses to present luck, including accounts in which agent-involving deliberative processes influence probabilities without determining the exact result. He also uses developmental considerations to explain how agents can acquire freedom-relevant capacities from initially non-agentic or non-responsible beginnings. The overlap with band-routing is substantial: antecedent agent-involving events can shape the probability landscape, and freedom need not require control over the micro-resolution. The present theory differs by locating its primary witness in a sustained self-object trajectory rather than an ordinary decision episode, by requiring relevance-bound alteration of the same open band, and by making recursive uptake into the later agent central. Mele therefore supplies important prior art for probability-shaping and the present-luck reply, while the current proposal adds the diachronic successor-self architecture and accrued sedimentation. ## 8.6 Planning agency, resolutions, and commitment Stage 2 extends rather than replaces established theories of diachronic agency. Bratman's planning theory explains how intentions, plans, and policies stabilize practical organization, coordinate means across time, and structure later deliberation. Holton emphasizes resolutions and the maintenance of intention under temptation. Ainslie analyzes intertemporal conflict and the use of commitment to alter later incentive structures. The present theory adds a distinctive object and modal conjunction. The plan or resolution is directed at an operative feature of the continuing agent; the agent uses curriculum access and environmental pre-arrangement to alter the later system that will encounter future temptations and evidence. Planning theory explains target persistence and adaptive means. Authored foreclosure asks whether that planning process materially shapes a genuinely open and exclusively actualizing transition on the pathway to later self-constitution. ## 8.7 Balaguer, Franklin, and the event-causal libertarian lineage Mark Balaguer argues that the libertarian question can reduce to an open scientific question about whether certain neural events are appropriately undetermined. Christopher Franklin develops a minimal event-causal libertarianism that seeks the metaphysical simplicity of compatibilist causal machinery plus indeterminism in the right role and location. Helen Steward and John Bishop offer further naturalistic causal accounts of agency, although with materially different metaphysical commitments. These approaches are close to the paper's Stage 1 discipline and its rejection of a primitive agent-causal shove. The present proposal differs by specifying a particular diachronic self-object witness, requiring exclusive actualization, and insisting that the open transition be materially shaped by the same project whose later effects are credited. A broader two-stage lineage runs from William James through Daniel Dennett and later models that separate chance generation of alternatives from reason-guided selection. The current proposal is not simply a two-stage generator-selector model. It does not place randomness only in the production of candidate ideas followed by deterministic selection. It locates possible freedom in a path-dependent trajectory that repeatedly reshapes the live band itself, maintains a represented successor target through feedback, and sediments the resulting history into the later agent. ## 8.8 Leeway, sourcehood, Frankfurt, and adjacent positions The paper's two factors align with an established distinction: - **leeway:** genuine alternatives or plural physical support; - **sourcehood:** the relation in virtue of which the action or transition is attributable to the agent. Fischer's semicompatibilist program separates responsibility-relevant sourcehood or guidance control from alternative possibilities. Frankfurt-style cases likewise motivate the claim that attribution need not itself depend upon alternatives. The present paper accepts that lesson for Stage 2: deterministic routing can be real agent-level self-governance. It nevertheless stipulates that **freedom**, as opposed to attribution alone, requires ontic leeway and exclusive actualization. Its distinctive move is to conjoin incompatibilist leeway with a trajectory-level sourcehood account at the same transition pathway. Other adjacent literatures include: - **possible selves and future-self continuity:** represented future identities as organizers of present interpretation and action; - **niche construction and ecological control:** alteration of environments and recurring inputs as a means of changing later behavioural and cognitive dispositions; - **implementation intentions and metacognitive intervention:** mechanisms by which represented targets become durable cue-sensitive policies; - **Vargas's agency-cultivation model:** a normative cousin of curriculum access and recursive self-formation, relevant if responsibility is later unbracketed. The claimed contribution should therefore be stated narrowly: **band-authorship under liveness** joins exclusive ontic foreclosure to a sustained self-object intervention trajectory, formalizes attribution through content-sensitive causal routing, permits non-monotonic recursive accrual and sedimentation, and seeks one existential witness while bracketing desert. # 9. The twin-systems objection Consider two realization-isomorphic systems: - the same actual state sequence; - the same represented alternatives; - the same self-models; - the same phenomenology; - the same behaviour; - the same functional organization. In one system, the complete prior state uniquely necessitates the outcome. In the other, multiple continuations are physically realizable within a historically and contextually constrained band. The theory accepts that the systems may be indistinguishable at the realized and functional levels. The open twin in this comparison must include exclusive actualization. An Everettian universal state is not simply the open member of the pair: on the usual reading it evolves deterministically and retains all branches rather than foreclosing alternatives. They are not identical in foreclosure history. - The deterministic system contains represented branching and causal self-modification, but no live continuation is extinguished because no other continuation was physically realizable. - The open system, if Stage 1 obtains, contains actual foreclosure of live continuations. Thus the phrase **identical histories** is equivocal. The systems can have identical realized histories. They cannot have identical foreclosure histories if one contains genuine modal closure and the other does not. At the level relevant to this theory, a stipulation of identical foreclosure histories would be incoherent; at the level of realized sequence, twinhood remains coherent. Under the stipulated realization-isomorphism, neither introspection nor agent-level behaviour could discriminate the open twin from the deterministic twin. Evidence for the difference would have to arise, if at all, from physical theory and distributional investigation of the relevant transition class. This does not make ontic openness epiphenomenal. Openness is not an additional event or force expected to produce an extra behavioural effect after the realized sequence has been fixed. It is a modal property of the transition relation itself: a fact about which continuations were physically realizable before one occurred. The twins may therefore be identical in realized sequence while differing in whether that sequence foreclosed genuine alternatives. The theory therefore does not add freedom as an unexplained inner glow to functionally identical systems. It identifies freedom partly with a relation between the process and genuinely available continuations. The twin case isolates that incompatibilist commitment rather than dissolving it. A deterministic seeded generator and an otherwise architecturally comparable generator driven by a genuinely indeterministic entropy source illustrate the distinction in principle. The example does not prove that any physical entropy source is genuinely indeterministic. It shows what would differ if one were. --- # 10. The conditional existential hinge ## 10.1 One proposed witness-location The theory proposes one location at which an authored foreclosure may be most readily identifiable: > A diachronically sustained self-object trajectory in which the agent represents a possible variance of itself, adopts and maintains or reconstructs that variance as an operative target, and repeatedly directs thought, conduct, environment, and curriculum toward altering the later agent under changing feedback. The theory does not claim: - that this is the only form of freedom; - that every free act must take this form; - that this is the maximal or most defensible possible location; - that freedom elsewhere is licensed by this case; - that the prevalence of such events has been established. Other locations or formulations may exist. ## 10.2 The asymmetric proof schema The two horns are not defeated symmetrically. Any adequately established ontically open and exclusively actualizing event would establish: 1. **Exhaustive determinism is false.** Foreclosure analytically requires genuinely live alternatives. Authorship is idle for this result; an ontically open decay would suffice. A single adequately established ontically open and exclusively actualizing trajectory containing the Stage 2 contribution would additionally establish: 2. **Indeterminism does not universally entail zero agent authorship.** The trajectory's exact resolutions may retain luck, but the continuing agent's represented self-variance made a sustained, content-sensitive causal contribution to the open process and its later uptake. The determinism/indeterminism dilemma is often used as a universal freedom-excluder: - determined events are not free; and - undetermined events are attribution-excluding luck. One witness containing ontic liveness, exclusive actualization, and agent-level self-object routing of the same materially relevant band would falsify that universal exclusion under the paper's defended non-ultimate conception of authorship. The conclusion is not that the witness contains no luck. It is that luck does not exhaust its authorship status. A critic may reserve the word *authorship* for a stronger relation, but must then defend that relation rather than treating zero attribution as analytically supplied by indeterminism. ## 10.3 Conditional, not completed proof The proof schema is valid only conditionally: > If one ontically live, exclusively actualizing, self-object-routed authored foreclosure exists, exhaustive determinism and the universal zero-authorship reading of indeterminism both fail. The paper does not claim that the witness has already been produced. The empirical burden remains to show, in one case: - ontic multiplicity; - cognitive integration of that multiplicity; - a sustained self-object target rather than merely transient self-description; - pattern-sensitive trajectory individuation and band-sensitive routing under a defensible interventionist comparison; - material contribution to at least one open foreclosure; - recursive uptake into the continuing agent. ### 10.3.1 Cross-register localization and type-level establishment Relevance-binding sharpens the metaphysical criterion but creates an evidentiary problem. Evidence for ontic openness is physics-level and distributional; evidence for self-object routing is psychological, behavioural, and neurocognitive. Neither register by itself certifies that the same token transition instantiated both properties. Adequate establishment therefore need not target direct token-level inspection of an unrealized alternative. It may proceed through a common transition class \(K\) that is independently identifiable in both registers: 1. **Physical classification:** physical theory and evidence support that members of \(K\), or a nonzero proportion of them, possess plural ontic support together with exclusive actualization. 2. **Neurocognitive localization:** the sustained self-object process is shown to operate through members of that same \(K\), not merely through a neighbouring or collocated class. 3. **Interventionist routing:** varying the represented target changes the successor-state distribution associated with \(K\). 4. **Trajectory relevance:** those transitions materially affect the later self-relevant trajectory and its uptake. 5. **Existential inference:** convergent evidence supports, with stated confidence and physical assumptions, that at least one token instantiated freedom-eligible physical structure—plural ontic support with exclusive actualization—together with routed contribution. Mere verbal overlap between a physically eligible class \(K\) and a routed class \(K'\) is insufficient. Let \(F=O\land EA\), where \(O\) denotes objectively plural support and \(EA\) exclusive actualization. The bridge requires either a common mechanistic identification or a statistical lower bound on the intersection between freedom-eligible transitions \(F\) and routed transitions \(R\) within a shared population. Schematically, if defensible lower bounds \(p_F\) and \(p_R\) estimate the relevant proportions, then \(p_F+p_R>1\) guarantees a positive intersection by the Fréchet bound. Under physical interpretations that retain all supported continuations rather than exclusively actualizing one, \(p_F=0\) for purposes of this witness, so this cross-register route cannot establish it. The real evidentiary problem is obtaining warranted bounds and ensuring that “routed” means alteration of the freedom-eligible band itself, not merely causal activity nearby. This type-level and statistical inference is conditional and model-mediated, as are many physical existence claims. The phrase **adequately established** therefore means convergent evidence that a physically characterized open transition class is instantiated within, and materially routed by, the identified self-object process. It does not mean direct observation of a token's unrealized alternatives. The theory's achievement is to specify the conjunction, the cross-register bridge required to investigate it, and a proposed location—not to announce completed verification. ## 10.4 No census, exhaustive catalogue, or licence Because the logical form is existential, the theory needs no census of freedom's distribution to obtain its hinge result. The paper is not attempting to identify every possible form, location, or degree of freedom or authorship. It proposes one candidate witness whose successful establishment would show that the excluded category is non-empty. Questions about: - how common authored foreclosures are; - which populations possess the relevant capacities; - whether other forms of freedom exist; - how much authorship any person accrues; - what legal or moral consequences follow; are important but logically subsequent. A successful witness would prove that the excluded category is non-empty. It would not establish that every human act, every adult, or every domain contains freedom. --- ## 10.5 Culpability and basic-desert responsibility are bracketed This paper concerns the existence and possible diachronic accumulation of freedom and authorship as against the determinism/indeterminism dilemma. It does not presently argue from authored foreclosure to culpability, blameworthiness, punishment, legal liability, or basic-desert moral responsibility. Those downstream questions require additional premises concerning, among other things: - normative competence; - knowledge and foreseeability; - reasons-responsiveness; - coercion, incapacity, and excuse; - the degree and domain of accrued authorship; - the relation between partial authorship and desert; and - institutional purposes of legal or moral responsibility. A successful witness would establish that at least one freedom- or authorship-bearing event exists in the stipulated sense. It would not, without further argument, establish that any particular agent deserves blame or punishment for any particular act. # 11. Functional developmental ordering The theory distinguishes capacities rather than fixed human kinds. ## Level 0: bare ontic resolution A genuinely open event occurs without an agent-level band or self-model. Nothing is authored. ## Level 1: historically conditioned foreclosure A developing organism's inherited and learned structure constrains open outcomes, but no operative feature of the self is held as an alterable object. The organism may own the process only in a thin biological sense, not in the relevant authorship sense. ## Level 2: self-object representation The agent can represent some of its own dispositions, policies, or regimes as objects of contemplation. ## Level 3: minimally authored foreclosure The agent represents a non-actual variance, adopts and sustains it as a target, and intentionally intervenes through available thought and curriculum channels. The first qualifying trajectory may arise from an overwhelmingly unchosen self. ## Level 4: recursively accrued self-authorship Later interventions operate from a constitution partly affected by earlier qualifying interventions. This level is the accrual mechanism, not a prerequisite for Level 3: the first qualifying trajectory can arise from an overwhelmingly unchosen agent. Because the continuing agent can repeat the self-object loop from an altered constitution, successive qualifying interventions may cumulatively enlarge the self-authored contribution to its later operative profile. This is a possibility, not a developmental law: accumulation may be local, interrupted, reversed, fragmented, or outweighed by unchosen causes. These are functional distinctions. Development may be gradual, asynchronous, domain-specific, interrupted, reversible, or incomplete. --- # 12. Empirical grounding and seam discipline ## 12.1 What Stage 2 can reasonably draw upon The Stage 2 mechanism is a conservative synthesis of established human capacities for: - episodic and counterfactual construction; - self-referential cognition; - metacognitive monitoring; - future-self representation; - affective evaluation; - implementation intentions; - deliberate practice; - environmental self-selection and commitment; - learning-induced changes in habit and policy. These capacities show that a represented non-actual self can become an actual causal input into present attention, behaviour, and later learning. Possible-selves research treats represented future identities as motivational structures (Markus and Nurius 1986). Constructive episodic-simulation research supports the capacity to recombine past information into imagined future episodes (Schacter and Addis 2007). Implementation-intention research shows that cue-linked if–then plans can improve goal enactment (Gollwitzer 1999). Future-self manipulations using age-progressed representations have shifted intertemporal choice toward later rewards (Hershfield et al. 2011). Commitment-device research shows that agents sometimes voluntarily restructure later environments to constrain future action (Bryan, Karlan, and Nelson 2010). Intensive longitudinal personality research further suggests that volitional trait-change goals can predict later change and that completed trait-relevant behavioural challenges predict change more reliably than merely accepting change goals (Hudson and Fraley 2015; Hudson et al. 2019). These findings do not establish the full sustained trajectory or freedom witness, but they supply empirical shape for content-plus-implementation routing and curriculum access. The paper's distinctive empirical burden is therefore not to establish from nothing that represented plans, future selves, and environmental pre-arrangements can affect later behaviour. It is to establish a sustained self-object trajectory, its material causal pathway into later constitution, and its conjunction with an independently supported open and exclusively actualizing transition class. None of these Stage 2 literatures establishes ontic openness. ## 12.2 Candidate neurocognitive resources, not a discovered authorship circuit Hippocampal, default-network, medial-prefrontal, valuation, and control systems are implicated in internally generated cognition, episodic construction, self-referential processing, evaluation, and source monitoring. The warranted claim is that these literatures identify candidate resources for parts of the proposed loop. The unwarranted stronger claims would be that: - memory and imagination are literally one mechanism running in opposite directions; - the default-mode network is a unitary constructive-simulation organ; - counterfactual simulation is the brain's sole or exact baseline operation; - four canonical modules implement the loop; - identity-directed self-revision has been mapped as an integrated pathway. The composition of these resources into an identity-directed, curriculum-writing process remains inferred from parts rather than directly measured. ## 12.3 Functional rather than phenomenal format Vivid visual imagery is not required for all counterfactual or self-object cognition. Language, imagery, spatial models, bodily simulation, unsymbolized structure, and other representational formats may support different aspects of the process. Aphantasia and hyperphantasia therefore caution against identifying simulation with one phenomenal display format. They do not establish that all underlying functions are identical across formats. ## 12.4 Learning pathways Humans do not directly rewrite their own neural parameters. Self-object-directed projects must operate through learning. Possible pathways include: - reinforcement and habit learning; - inhibitory learning; - repeated attentional allocation; - memory updating; - reconsolidation under appropriate conditions; - environmental and social niche construction. No one pathway is assumed to be universal, and reconsolidation should not be treated as an automatic write channel whenever a person contemplates change. Retrieval can render some consolidated memories labile and susceptible to reconsolidation-dependent updating (Nader, Schafe, and LeDoux 2000), but prediction error, memory strength, age, and other boundary conditions constrain when reconsolidation occurs (Fernández, Boccia, and Pedreira 2016). ## 12.5 Libet-class and decoding studies Libet-class studies do not prove that a complete decision is unconsciously made before awareness (Libet et al. 1983). They show that neural processes relevant to spontaneous motor action can develop before reportable awareness of intention. Later decoding studies have sometimes predicted arbitrary left/right responses modestly above chance before reported awareness (Soon et al. 2008). Such prediction may reflect an evolving bias, action tendency, or partially formed state rather than irreversible commitment. Stochastic-accumulator models show that readiness-potential-type results are compatible with evolving fluctuations and threshold crossing rather than an already completed unconscious choice (Schurger, Sitt, and Dehaene 2012). None of these findings establishes ontic neural stochasticity. Nor do arbitrary motor tasks directly test the metacognitive self-object process proposed here. ## 12.6 The physical precondition remains open Radioactive decay and other quantum phenomena provide ubiquitous behaviour modelled probabilistically in standard quantum theory. Whether that probability is ontic depends upon unresolved interpretive and physical questions. Foreclosure additionally requires exclusive actualization; universal branching by itself is insufficient. The physical precondition is not hostage to one quantum interpretation. A contested alternative program argues that classical determinism depends upon treating infinitely precise real-valued initial conditions as physically instantiated. Gisin (2021) and Del Santo and Gisin (2019) develop finite-information approaches under which classical physics may admit fundamental indeterminacy. These proposals are not established physical orthodoxy and have attracted technical criticism. They are relevant because they show that the determinism question is not exhausted by choosing among familiar quantum interpretations. Neural variability is empirically pervasive across molecular, cellular, network, and behavioural scales (Faisal, Selen, and Wolpert 2008). But measured neural noise may arise from thermal effects, unresolved deterministic dynamics, measurement limitations, or genuinely indeterministic sources. Variability does not itself identify ontic openness. The theory therefore uses physical indeterminism as a candidate precondition, not as an established datum. The decisive questions are: 1. Is any process genuinely indeterministic? 2. Does it involve exclusive actualization rather than retained universal branching? 3. Can such indeterminacy enter an integrated agent–environment transition pathway? 4. Can it be amplified to cognitively significant trajectories? 5. Can higher-level history and context structure the relevant probability band? 6. Can the self-object process materially alter that same band? # 13. Predictions and research program The paper generates a staged research program rather than one decisive experiment. ## 13.1 Physical precondition Investigate whether any genuine physical indeterminism exists, whether it can be distinguished from deterministic chaos or hidden-state complexity, and whether its ontology includes exclusive actualization. The program should compare collapse-type quantum models, deterministic and branching interpretations, and contested finite-information approaches to classical indeterminism without treating any one as established. ## 13.2 Neural amplification Investigate whether microscopic indeterminacy, if present, is amplified to cognitively significant neural or agent-integrated trajectories rather than averaged into irrelevance. Tegmark's canonical decoherence-timescale analysis argues that cognitively relevant neural degrees of freedom should be treated as effectively classical because estimated quantum decoherence times are far shorter than neural dynamical timescales (Tegmark 2000). Any proposal relying upon sustained cognitively relevant quantum coherence must answer that argument directly. Rapid decoherence does not itself prove exhaustive determinism or establish that no quantum-origin indeterminacy can affect later classical neural dynamics; it constrains the mechanism through which such influence could occur. Neural-noise studies should therefore be combined with mechanistic source identification and amplification analysis rather than treating trial-to-trial variability as proof of ontic openness. ## 13.3 Banded cognitive dynamics Test whether neural and behavioural transitions are better modelled as historically and contextually structured probability bands than as noisy readouts from unique underlying trajectories. This alone would not establish ontic openness, but it would specify the architecture through which openness could matter. ## 13.4 Self-object routing Compare matched trajectories that vary the content or presence of an explicit successor-self target while controlling generic reflection, effort, exposure, expectancy, arousal, and external support as closely as possible. The theory predicts target-specific differences in: - persistence across perturbation; - strategy revision after error; - environmental and curriculum selection; - generalization beyond trained examples; - later policy-band properties; - recursive uptake into subsequent self-intervention. Longitudinal studies should prospectively individuate the trajectory. They should record target content before the credited outcome, track later reinstantiations and revisions, identify contemporaneous target-relative error correction, and distinguish causally continuous projects from retrospective narrative reconstruction. Retrospective self-report may supplement but should not define trajectory identity. The interventionist ideal is to estimate band-sensitivity: whether changing \(R\), the represented transformation, changes the successor-state distribution of the physically and neurocognitively localized transition class \(K\). Pattern-sensitivity should be assessed separately to establish prospective trajectory identity. Empirical designs approximate rather than literally instantiate the \(do\)-operator. Because interventions on represented content are inevitably fat-handed, designs should use active controls holding generic reflection constant, component ablations contrasting self-directed with equally specific non-self-directed targets, manipulation checks separating target content from engagement and expectancy, and within-person target variation where ethically and practically possible. The trajectory is a time-varying exposure. Earlier states affect later target engagement and later outcomes, while prior engagement can itself alter later confounders. Ordinary regression may therefore be biased. Marginal structural models, inverse-probability weighting, the g-formula, or related g-methods provide candidate tools for estimating sustained intervention effects under treatment–confounder feedback (Hernán and Robins 2020). Addiction recovery, sustained volitional personality change, and durable modification of attentional or relational dispositions are candidate testbeds. These cases must be studied without treating cognition as sovereign over medication, social structure, treatment access, or other causal supports. ## 13.5 Temporal and developmental gates Differentiate: - inability due to immature component capacities; and - low deployment despite mature capacities. The two gates should predict different responses to maturation, scaffolding, training, and environmental change. ## 13.6 Recursive accrual Use longitudinal designs to test whether successful self-object trajectories alter later attention, policy, environmental selection, target maintenance, recovery after failure, and the probability of initiating further self-revision projects. The distinctive recursive prediction is not merely serial recurrence. It is mediated carry-forward: an earlier trajectory changes an intermediate agent-state, and that altered state helps explain the content, reliability, or efficacy of a later trajectory. Designs should therefore distinguish \(I_n ightarrow A_{n+1} ightarrow I_{n+1}\) from stable trait differences, common environmental causes, and retrospective narrative continuity. The existential target creates a single-case/statistical bridge problem. Population-level estimates can identify routing regularities, but one witness is a trajectory. Candidate designs include N-of-1 interrupted time-series, dense within-person measurement, repeated target engagement and withdrawal where ethical, and mechanistic evidence linking intervention phases to later-agent changes. No single design will establish ontic openness; the aim is to combine within-person routing evidence with the cross-register physical classification of the same transition class. The expected pattern is not monotonic improvement. The prediction is only that prior qualifying trajectories can become causal inputs into later trajectories and ordinary cognition, and that the propensity to perform another self-authorship pass can itself be shaped. Studies should distinguish modal liveness from reachable-state richness. They should also develop cross-register localization strategies capable of identifying the same transition class in physical and neurocognitive descriptions, permitting type-level statistical inference to a nonempty intersection of openness and routing. A narrowed maladaptive band may coexist with an expanded set of later life trajectories even under determinism. Whether that enlarged space contains more agent-relevant ontically open transitions is a separate, Stage-1-dependent question. ## 13.7 Authorship attribution Develop comparative rather than initially scalar measures of authorship contribution. Candidate contrasts include: - intervention-driven versus drift-driven change; - self-object-routed versus externally imposed curriculum; - target-specific causal effects versus generic effort or exposure; - sustained error-corrective trajectories versus isolated target-congruent events; - persistent policy change versus transient compliance; - later generalization versus context-bound performance; - fresh authored foreclosure versus inherited expression of prior authorship. --- # 14. Failure conditions and graduated floors The theory has ordered failure modes. ## Floor 1: full authored-foreclosure theory The full theory obtains if: - ontic openness exists, one continuation becomes exclusively actual, and the relevant openness is cognitively or agent-environment integrated and banded; - a sustained self-object trajectory is prospectively individuated through target continuity, contemporaneous correction, and pattern-sensitivity; - the relevant transition lies on the causal pathway from that trajectory to the later-agent change being credited; - intervention on the represented target produces a material band-sensitive difference in a transition class independently supported as ontically open; - relevance-binding is satisfied, so the routed contribution alters the open band itself rather than merely co-occurring with unrelated openness; - cross-register evidence adequately localizes openness and routing to the same transition class and supports a nonempty token intersection; and - the resulting foreclosure is recursively incorporated into the continuing agent. ## Floor 2: cognitively open but unauthored inference If cognitively relevant ontic openness with exclusive actualization exists but the self-object loop is merely a readout or makes no material contribution, the fixed-film picture is false but authored freedom remains unproved. This is indeterministic cognition without certified self-authorship. ## Floor 3: deterministic self-modification If self-object-directed change occurs but all physical transitions are uniquely necessitated, the Stage 2 mechanism remains a real form of recursive causal self-modification. It is not freedom under the operational definition. ## Floor 4: exhaustive determinism If reality is exhaustively deterministic, no ontic foreclosure occurs anywhere. The theory reports zero freedom in the stipulated sense rather than relabelling fixed causal recursion as freedom. --- # 15. Open problems 1. **Physical indeterminism:** Can any indeterminism be established rather than merely represented in a successful probabilistic theory? 2. **Exclusive actualization:** Which physical ontologies permit one live continuation to become actual while alternatives are genuinely extinguished? 3. **Cognitive integration:** If indeterminism exists, does it materially affect cognition or an integrated agent–environment control loop? 4. **Banded structure:** Can history and context structure ontically open transitions at an agent-relevant scale? 5. **Trajectory identity:** What prospective, outcome-independent criteria preserve enough identity in a successor-self target across revision, interruption, and changing representational format? 6. **Relevance-binding and materiality:** What evidence shows that the self-object intervention altered the very open transition band at issue and that the transition lay on the pathway to the later-agent change credited? 7. **Cross-register localization:** What shared transition class can support both the physical openness claim and the neurocognitive routing claim, and what statistical evidence would warrant an existential inference to their token-level intersection? 8. **Routing criterion:** Which interventionist comparison best distinguishes target-specific band-sensitivity from generic effort, expectancy, arousal, or external scaffolding? 9. **Comparator band:** How should the no-intervention or differently routed \(B_0\) be specified philosophically and operationally? 10. **First-instance boundary:** What evidence is sufficient to identify the first minimally authored trajectory? 11. **Recursive carry-forward:** What evidence shows that effects of an earlier qualifying trajectory entered the operative constitution generating a later trajectory rather than merely preceding it? 12. **Luck transmission:** Does constitutive luck merely limit later attribution, or does it transmit so as to eliminate freedom-level authorship? What additional premise would justify annihilating transmission? 13. **Mere addition:** Does relevance-bound band-authorship under liveness add a genuine sourcehood relation, or can it still be reduced to deterministic self-modification plus unrelated chance? 14. **Classification:** When does a real conditioned agent-level contribution warrant the term partial authorship rather than causal ownership alone? 15. **Liveness and reachability:** How should local band narrowing, global possibility-space expansion, and inherited expression of earlier authorship be related? 16. **Agent individuation:** How should the agent trajectory be integrated with bodily continuity, actual history, autobiographical identity, and social identity? 17. **Developmental floor:** What component capacities are necessary for sustained self-object-directed intervention, and how do they develop? 18. **Learning mechanism:** Which pathways mediate successful self-directed change in different domains? 19. **Causal estimation:** Which N-of-1, longitudinal, and g-method designs can handle target engagement as a time-varying exposure with treatment–confounder feedback? 20. **Measurement:** Can authorship contribution be operationalized without pretending that causal contributions are clean additive fractions? 21. **Other witness-locations:** Are there other forms of authored foreclosure that do not involve explicit identity- or regime-directed self-modification? # 16. Conclusion This paper proposes one possible witness-location for freedom and authorship. It is not an exhaustive theory, taxonomy, or catalogue of every free act or form of agency. Its structure has two necessary stages and three nested units of analysis: 1. **Stage 1 / event level:** Reality must contain genuine ontic multiplicity with exclusive actualization, and at least one materially relevant open transition must be routed through the agent's self-object process. 2. **Stage 2 / trajectory level:** A developmentally produced agent must represent a possible variance of itself, adopt and sustain or reconstruct that variance as an operative target, and repeatedly direct thought, action, environment, and curriculum toward changing the later agent under incoming feedback. 3. **Recursive agent level:** Effects of an earlier qualifying trajectory may enter the constitution of the agent who generates, maintains, and revises a later trajectory. The first two levels identify one candidate freedom witness. The third is the primary diachronic engine of accrued self-authorship. The primary Stage 2 witness is therefore a trajectory, not an isolated choice; accrual begins where the result of one trajectory partly alters the system that performs the next pass. The routing relation is interventionist. Pattern-sensitivity identifies a prospectively continuous target-directed trajectory; band-sensitivity asks whether changing the represented transformation alters the successor-state distribution of the relevant transition class. Where both are supported across appropriate controls, the representation is an agent-level causal difference-maker. That fact is compatible with determinism and therefore supplies only the attribution half of the account. Where at least one ontically open and exclusively actualizing transition has its own band materially affected by that routed process, occurrence forecloses real alternatives and the locally bound conjunction supplies the proposed freedom witness: \[ \boxed{\text{ontic liveness} \land \text{exclusive actualization} \land \text{self-object-routed agent contribution} \rightarrow \text{candidate freedom witness}.} \] Constitutive luck is fully conceded. The first agent is built by causes it did not author; the intervention's occurrence and exact results may remain lucky; and no trajectory escapes history, embodiment, or causal dependence. The paper rejects only the inference that lucky provenance or residual outcome luck automatically reduces a sustained, target-sensitive, self-directed contribution to zero. Across recursive loops, the later agent may bear a larger contribution from prior qualifying trajectories and may become more likely or able to initiate further ones. This is the paper's central accrual mechanism: authored foreclosure supplies a nonzero event-level contribution; sustained trajectory supplies diachronic unity; recursion allows the contribution to enter the constitution from which later contributions arise. That accumulation is non-monotonic. Successful self-formation may stabilize some local transitions while expanding a broader reachable future space. Reachable-state richness is not itself ontic liveness; only if genuine openness occurs within the enlarged space does that expansion create additional locations where authored foreclosure may occur. Later determined conduct may express earlier authorship without constituting a fresh authored foreclosure. The theory's constructive core is band-authorship under liveness: the continuing agent materially shapes the same live successor structure—whether represented by a numerical chance measure or only by objectively nondegenerate support—from which exclusive actualization forecloses alternatives. This is not deterministic self-modification plus ambient chance, because routing and openness are relevance-bound on the causal pathway to the later-agent change. Recursion then supplies the diachronic carry-forward: earlier qualifying trajectories may sediment into character and intervention-capacity and thereby alter the agent from which later trajectories proceed. The theory does not claim ultimate sourcehood, freedom from history, exhaustive control, direct access to neural parameters, or a primitive agent-causal power. It attempts to locate partial sourcehood within the temporally extended organization of ordinary causal processes. Nor does it infer culpability, blameworthiness, punishment, legal liability, or basic-desert moral responsibility. Its logical ambition is narrow: - any adequately established ontically open event would defeat exhaustive determinism; and - one adequately established ontically open and exclusively actualizing trajectory containing the specified agent-level contribution would show that indeterminism does not universally entail zero authorship. The paper has not yet produced that witness. It has identified one proposed expression, specified its physical and causal requirements, distinguished it from nearby theories, and stated how evidence from physical and agent-level registers would have to be joined through a common transition class before the existential inference could be warranted. --- # 17. Principal philosophical and empirical references ## 17.1 Philosophical references - Ainslie, George. *Breakdown of Will*. Cambridge University Press, 2001. - Balaguer, Mark. *Free Will as an Open Scientific Problem*. MIT Press, 2010. - Bishop, John. *Natural Agency: An Essay on the Causal Theory of Action*. Cambridge University Press, 1989. - Bratman, Michael E. *Intention, Plans, and Practical Reason*. Harvard University Press, 1987. - Bratman, Michael E. *Structures of Agency: Essays*. Oxford University Press, 2007. - Chisholm, Roderick M. “Human Freedom and the Self.” *The Lindley Lecture*. University of Kansas, 1964. - Dennett, Daniel C. “On Giving Libertarians What They Say They Want.” In *Brainstorms: Philosophical Essays on Mind and Psychology*. MIT Press, 1978. - Dennett, Daniel C. *Elbow Room: The Varieties of Free Will Worth Wanting*. MIT Press, 1984. - Everett, Hugh III. “‘Relative State’ Formulation of Quantum Mechanics.” *Reviews of Modern Physics* 29 (1957): 454–462. - Fischer, John Martin. *The Metaphysics of Free Will: An Essay on Control*. Blackwell, 1994. - Franklin, Christopher Evan. *A Minimal Libertarianism: Free Will and the Promise of Reduction*. Oxford University Press, 2018. - Frankfurt, Harry G. “Alternate Possibilities and Moral Responsibility.” *Journal of Philosophy* 66 (1969): 829–839. - Frankfurt, Harry G. “Freedom of the Will and the Concept of a Person.” *Journal of Philosophy* 68 (1971): 5–20. - Greco, John. “Knowledge and Success from Ability.” *Philosophical Studies* 142 (2009): 17–26. - Holton, Richard. *Willing, Wanting, Waiting*. Oxford University Press, 2009. - Ismael, Jenann. “Self-Organization and Self-Governance.” *Philosophy of the Social Sciences* 41 (2011): 327–351. - Ismael, Jenann. *How Physics Makes Us Free*. Oxford University Press, 2016. - James, William. “The Dilemma of Determinism.” *Unitarian Review* 22 (1884): 193–224; reprinted in *The Will to Believe and Other Essays in Popular Philosophy* (1897). - Kane, Robert. *The Significance of Free Will*. Oxford University Press, 1996. - Kane, Robert. “Libertarianism.” *Philosophical Studies* 144 (2009): 35–44. - Levy, Neil. *Hard Luck: How Luck Undermines Free Will and Moral Responsibility*. Oxford University Press, 2011. - Mele, Alfred R. *Free Will and Luck*. Oxford University Press, 2006. - O'Connor, Timothy. *Persons and Causes: The Metaphysics of Free Will*. Oxford University Press, 2000. - Pearl, Judea. *Causality: Models, Reasoning, and Inference*. 2nd ed. Cambridge University Press, 2009. - Pereboom, Derk. *Free Will, Agency, and Meaning in Life*. Oxford University Press, 2014. - Pritchard, Duncan. *Epistemic Luck*. Oxford University Press, 2005. - Sosa, Ernest. *A Virtue Epistemology: Apt Belief and Reflective Knowledge, Volume I*. Oxford University Press, 2007. - Steward, Helen. *A Metaphysics for Freedom*. Oxford University Press, 2012. - Strawson, Galen. “The Impossibility of Moral Responsibility.” *Philosophical Studies* 75 (1994): 5–24. - van Inwagen, Peter. *An Essay on Free Will*. Oxford University Press, 1983. - van Inwagen, Peter. “Free Will Remains a Mystery: The Eighth Philosophical Perspectives Lecture.” *Philosophical Perspectives* 14 (2000): 1–19. - Vargas, Manuel. *Building Better Beings: A Theory of Moral Responsibility*. Oxford University Press, 2013. - Woodward, James. *Making Things Happen: A Theory of Causal Explanation*. Oxford University Press, 2003. ## 17.2 Empirical and scientific references - Bryan, Gharad, Dean Karlan, and Scott Nelson. “Commitment Devices.” *Annual Review of Economics* 2 (2010): 671–698. https://doi.org/10.1146/annurev.economics.102308.124324. - Del Santo, Flavio, and Nicolas Gisin. “Physics without Determinism: Alternative Interpretations of Classical Physics.” *Physical Review A* 100, no. 6 (2019): 062107. https://doi.org/10.1103/PhysRevA.100.062107. - Faisal, A. Aldo, Luc P. J. Selen, and Daniel M. Wolpert. “Noise in the Nervous System.” *Nature Reviews Neuroscience* 9, no. 4 (2008): 292–303. https://doi.org/10.1038/nrn2258. - Fernández, Rodrigo S., Mariano M. Boccia, and María E. Pedreira. “The Fate of Memory: Reconsolidation and the Case of Prediction Error.” *Neuroscience & Biobehavioral Reviews* 68 (2016): 423–441. https://doi.org/10.1016/j.neubiorev.2016.06.004. - Gisin, Nicolas. “Indeterminism in Physics, Classical Chaos and Bohmian Mechanics: Are Real Numbers Really Real?” *Erkenntnis* 86 (2021): 1469–1481. https://doi.org/10.1007/s10670-019-00165-8. - Gollwitzer, Peter M. “Implementation Intentions: Strong Effects of Simple Plans.” *American Psychologist* 54, no. 7 (1999): 493–503. https://doi.org/10.1037/0003-066X.54.7.493. - Hernán, Miguel A., and James M. Robins. *Causal Inference: What If*. Chapman & Hall/CRC, 2020. - Hershfield, Hal E., Daniel G. Goldstein, William F. Sharpe, Jesse Fox, Leo Yeykelis, Laura L. Carstensen, and Jeremy N. Bailenson. “Increasing Saving Behavior Through Age-Progressed Renderings of the Future Self.” *Journal of Marketing Research* 48 (2011): S23–S37. - Hudson, Nathan W., and R. Chris Fraley. “Volitional Personality Trait Change: Can People Choose to Change Their Personality Traits?” *Journal of Personality and Social Psychology* 109, no. 3 (2015): 490–507. https://doi.org/10.1037/pspp0000021. - Hudson, Nathan W., Daniel A. Briley, William J. Chopik, and Jaime Derringer. “You Have to Follow Through: Attaining Behavioral Change Goals Predicts Volitional Personality Change.” *Journal of Personality and Social Psychology* 117, no. 4 (2019): 839–857. https://doi.org/10.1037/pspp0000221. - Libet, Benjamin, Curtis A. Gleason, Elwood W. Wright, and Dennis K. Pearl. “Time of Conscious Intention to Act in Relation to Onset of Cerebral Activity (Readiness-Potential): The Unconscious Initiation of a Freely Voluntary Act.” *Brain* 106, no. 3 (1983): 623–642. https://doi.org/10.1093/brain/106.3.623. - Markus, Hazel, and Paula Nurius. “Possible Selves.” *American Psychologist* 41, no. 9 (1986): 954–969. https://doi.org/10.1037/0003-066X.41.9.954. - Nader, Karim, Glenn E. Schafe, and Joseph E. LeDoux. “Fear Memories Require Protein Synthesis in the Amygdala for Reconsolidation after Retrieval.” *Nature* 406 (2000): 722–726. https://doi.org/10.1038/35021052. - Odling-Smee, F. John, Kevin N. Laland, and Marcus W. Feldman. *Niche Construction: The Neglected Process in Evolution*. Princeton University Press, 2003. - Schacter, Daniel L., and Donna Rose Addis. “The Cognitive Neuroscience of Constructive Memory: Remembering the Past and Imagining the Future.” *Philosophical Transactions of the Royal Society B* 362, no. 1481 (2007): 773–786. https://doi.org/10.1098/rstb.2007.2087. - Schurger, Aaron, Jacobo D. Sitt, and Stanislas Dehaene. “An Accumulator Model for Spontaneous Neural Activity Prior to Self-Initiated Movement.” *Proceedings of the National Academy of Sciences* 109, no. 42 (2012): E2904–E2913. https://doi.org/10.1073/pnas.1210467109. - Soon, Chun Siong, Marcel Brass, Hans-Jochen Heinze, and John-Dylan Haynes. “Unconscious Determinants of Free Decisions in the Human Brain.” *Nature Neuroscience* 11 (2008): 543–545. https://doi.org/10.1038/nn.2112. - Tegmark, Max. “Importance of Quantum Decoherence in Brain Processes.” *Physical Review E* 61, no. 4 (2000): 4194–4206. https://doi.org/10.1103/PhysRevE.61.4194. ## Compact terminology - **Ontic openness:** More than one physically realizable continuation from the same complete relevant prior state. - **Exclusive actualization:** One live continuation becomes actual in the relevant history while alternatives cease to remain available as continuations of that history. - **Foreclosure:** Irreversible extinguishing of one or more genuinely live continuations through occurrence. - **Banded openness:** Ontic multiplicity tightly constrained by history, architecture, and present context. - **Actualized state:** The presently realized fact, including sedimented effects of prior foreclosures; its actuality does not imply that it was uniquely necessitated. - **Self as object:** Representation of an operative feature of the continuing agent as alterable content within cognition. - **Represented variance:** A possible future outcome, disposition, regime, or successor-self configuration held as a candidate alteration of the continuing agent. - **Sustained self-variance trajectory:** A target-directed process in which the agent maintains or reconstructs a represented variance, interprets feedback relative to it, changes means, and later inherits the effects. - **Curriculum access:** Indirect influence over later learning through attention, practice, environment, exposure, treatment, social structure, and action rather than direct parameter editing. - **Pattern-sensitivity:** The prospective-individuation condition under which changing the represented target would change the longitudinal pattern of maintenance, correction, planning, and intervention. - **Band-sensitivity:** The Woodwardian interventionist routing condition, expressible using Pearl's intervention notation, under which changing the represented transformation changes the successor-state distribution of the relevant transition class across suitable controls. - **Relevance-binding:** The requirement that the self-object intervention alter the band of the ontically open transition itself rather than merely coexist with unrelated openness. - **Materiality:** The requirement that the bound transition lie on the causal pathway to the later-agent change being credited and make a non-trivial difference to that pathway. - **Prospective trajectory individuation:** Identification of a sustained project through pre-outcome target specification, causal continuity, contemporaneous target-relative correction, outcome-independent identity conditions, and pattern-sensitivity. - **Cross-register localization:** Identification of one transition class in both the physical openness register and the neurocognitive routing register, sufficient for a model-mediated statistical inference that at least one token instantiated both properties. - **Reachable-state richness:** The range of substantively different states, projects, relationships, and environments accessible across a longer horizon; distinct from ontic liveness. - **Band-authorship under liveness:** Material, content-sensitive shaping of the same ontically open and exclusively actualizing transition band whose foreclosure is credited to the agent. - **Authored foreclosure:** A genuine foreclosure materially routed through a sustained self-object-directed trajectory under relevance-binding, materiality, and exclusive actualization. - **Occurrence–representation–efficacy decomposition:** Separation of whether an intervention occurred, what self-transformative content organized it, and what causal difference that content made. - **Conditioned partial authorship:** A real agent-level causal contribution that remains historically constituted, non-ultimate, and compatible with origin and outcome luck. - **Accrued self-authorship:** The potentially increasing, but non-monotonic, contribution of prior qualifying trajectories to the continuing agent's operative profile and future intervention-capacity. - **Recursive accrual:** The carry-forward relation through which effects of one qualifying trajectory enter the operative constitution that generates, selects, sustains, or corrects a later trajectory. - **Sedimented freedom:** The historical incorporation of earlier freedom-bearing trajectories into relatively stable later dispositions, environments, policies, and intervention-capacities. - **Inherited expression:** Later conduct issuing from a disposition partly formed through prior authored trajectories without necessarily constituting a fresh authored foreclosure. - **Recursively diachronic agent:** A continuing agent for whom effects of earlier self-directed trajectories can alter the operative constitution that generates and regulates later trajectories. - **Hinge proof:** The conditional existential result that any ontically open event defeats exhaustive determinism, and one ontically open, exclusively actualizing, self-object-routed trajectory additionally defeats the universal inference from indeterminism to zero authorship.