Beyond the Illusion: Philip Pettit's Reimagining of Agent Control in a Deterministic World

Introduction: The Brain, Free Will, and Pettit's Intervention

The concept of free will, the deeply ingrained human intuition that individuals consciously choose their actions, has long served as a fundamental pillar for understanding human agency and, crucially, moral responsibility. This ancient philosophical puzzle, however, faces profound challenges from contemporary scientific discoveries (Startup, 2021). The enduring debate questions whether human actions are genuinely chosen or are, in fact, causally determined by antecedent events and the laws of nature (Startup, 2021). Most philosophical frameworks tightly link free will with moral responsibility, viewing it as a metaphysical prerequisite for accountability (Startup, 2021).

Recent neuroscientific findings have intensified this centuries-old discussion, leading some to propose that free will might be an illusion (Libet, 1985; Soon et al., 2008). Discoveries, particularly those related to brain activity preceding conscious awareness, suggest that the brain may initiate actions before a person consciously forms an intention to act (Libet, 1985; Libet et al., 1983; Shibasaki & Hallett, 2006). Benjamin Libet's seminal experiments in the 1980s, for instance, revealed that the Bereitschaftspotential (BP), an electrical signal in the brain, precedes a conscious decision to move by approximately half a second, implying an unconscious pre-determination of action (Libet, 1985; Libet et al., 1983; Shibasaki & Hallett, 2006). Subsequent fMRI studies have further demonstrated the predictability of choices seconds before conscious awareness, deepening the challenge to traditional notions of conscious control (Soon et al., 2008).

It is within this complex and empirically charged landscape that Philip Pettit, a distinguished philosopher, offers a critical intervention with his article 'Neuroscience and Agent Control'. Pettit contends that these scientific revelations do not necessarily dismantle the concept of agent control or free will. Instead, he argues that they necessitate a fundamental re-evaluation of how these concepts are understood (Pettit, 2007). His work aims to bridge the gap between psychological aspects of free agency and free will and broader political theories of freedom, proposing a unified theoretical framework (Pettit, 2001). He begins from the foundational premise that freedom is inextricably linked to responsibility (Pettit, 2001). The emergence of neuroscientific data has shifted the discourse from purely abstract metaphysical arguments to empirically informed challenges. This development compels philosophers to either reconcile their conceptual frameworks with scientific data or dismiss the empirical findings' relevance. Pettit's approach exemplifies the former, as he seeks to revise the understanding of free will to accommodate scientific observations, rather than rejecting either scientific evidence or the concept of free will itself. This highlights a significant trend in contemporary philosophy: the growing interdisciplinary dialogue and the imperative for philosophical concepts to be empirically grounded.

Challenging the "Act-of-Will" Picture: Neuroscience's Insights

Historically, a prevalent assumption in philosophical thought posited that actions are subject to agent-control if and only if they originate from conscious "acts of will." This traditional view typically characterizes willing as a pure, direct mental event, distinct from the actions it prompts (thus avoiding an infinite regress of prior wills), and capable of being initiated by the agent on demand (Pettit, 2007). This aligns with the pervasive intuitive feeling that an individual is "in charge" and that conscious intention directly causes actions (Pettit, 2007). Proponents of this view often suggest that, in principle, it should be possible to precisely track the moment a person's will forms within the brain and to observe how the brain switches as a conscious will either emerges or does not (Pettit, 2007).

However, neuroscientific experiments have presented a compelling counter-narrative to this traditional "act-of-will" picture. Libet's experiments, as noted, demonstrated that brain activity (the readiness potential) precedes the conscious awareness of an intention to act by approximately half a second (Libet, 1985; Libet et al., 1983; Shibasaki & Hallett, 2006). This suggests that the brain "commits to a decision" unconsciously before that decision registers in conscious awareness (Libet, 1985; Libet et al., 1983; Shibasaki & Hallett, 2006). Further, more advanced studies utilizing fMRI technology have shown that researchers can predict simple choices, such as whether a subject will press a left or right button, up to 10 seconds before the subject becomes consciously aware of their choice. This finding further undermines the notion of conscious initiation as the primary driver of action (Soon et al., 2008).

Some philosophers and neuroscientists, such as Sam Harris, interpret these findings as evidence that our intuitive belief that conscious intention initiates action is an illusion. They propose that conscious awareness might function as a "readout" or a retrospective interpretation of brain activity already in motion, rather than its causal origin (Harris, 2012). Even the concept of a "veto" power—the ability to consciously inhibit an unconsciously initiated action, as proposed by Libet—has been challenged by research suggesting that even this inhibitory process might itself be unconsciously determined (Isham et al., 2017; Libet, 1985; Libet et al., 1983; Shibasaki & Hallett, 2006). Experiments involving mirror boxes and magnetic stimulation also contribute to the argument that the conscious experience of willing might be an illusion if it is solely tied to the traditional act-of-will model (Nielson, 1963; Pettit, 2007).

Pettit acknowledges these neuroscientific findings but critically argues that they do not, in themselves, prove that agent-control is an illusion. Instead, he maintains that these findings demonstrate the inadequacy of the traditional conception of agent-control, particularly its reliance on the "act-of-will" picture (Pettit, 2007). He contends that if willing is indeed always necessary for an agent to control an action, and if the act of will must be subject to conscious experience and activation, then the traditional model of control is indeed illusory (Pettit, 2007). 

However, this conclusion does not extend to agent-control itself. Pettit's work is recognized for developing theoretical views of agent control that are compatible with scientific results, specifically countering claims that empirical science renders control or will illusory (Pettit et al., 2016). The neuroscientific data, particularly concerning the readiness potential, presents a direct causal challenge to the "act-of-will" model: if brain activity precedes conscious intent, then conscious intent cannot be the sole originator of the action. This empirical observation necessitates a re-evaluation of what "control" and "will" truly signify. Pettit's response is not to deny the empirical data, but to assert that the traditional philosophical concept of will is insufficient. This exemplifies a characteristic move in neurophilosophy: when empirical findings conflict with a philosophical concept, the concept itself must be rigorously examined and potentially revised. This approach underscores that philosophical theories must be robust enough to integrate scientific findings, rather than existing in isolation from empirical reality.

Pettit's Reimagined Agent Control: Responsibility and Reasons-Responsiveness

Pettit critically examines and ultimately dismisses simpler conceptions of control, such as "rational control" and "volitional control." Rational control describes actions that are appropriately related to an agent's beliefs and desires, regardless of what those beliefs and desires might be. Volitional control, on the other hand, refers to actions that align with an agent's second-order desires—what they wish to desire. Pettit illustrates the limitations of these concepts with the example of an addict: if an addict's strongest desire is for a drug fix, but they do not want to desire the drug, taking the drug demonstrates rational control but not volitional control (Pettit, 2001).

Pettit argues that both rational and volitional control are too simplistic because they fail to account for the impact of external factors on an agent's freedom. For instance, in a scenario of hostile coercion, such as "Your money or your life!", an individual's rational and volitional control may remain technically unaffected, yet their freedom is undeniably diminished (Pettit, 2001). This critique sets the stage for Pettit's more comprehensive account of freedom, which he posits is fundamentally and a priori connected to responsibility and democracy (Pettit, 2001). The concept of second-order desires, notably discussed by Harry Frankfurt, is also recognized in neurophilosophy as crucial for a robust understanding of free will (Frankfurt, 1969; Frankfurt, 1971).

Pettit proposes an alternative conception of agent control that harmonizes with neuroscientific findings. He views volition not as an instantaneous, singular act but as a continuous process that is characteristic of a particular type of agent (Pettit et al., 2016). This perspective implies that a conscious state alone does not guarantee voluntary action; rather, voluntary action arises from a continuity of processes operating under conscious attentional control (Pettit et al., 2016). Pettit argues that an agent can still be "reason-responsive" even if they are not deliberatively aware of the precise reasons for their actions at the moment of acting (Pettit et al., 2016). He advocates for a "sophisticated view of moral responsibility" that transcends the simplistic notion that conscious awareness of reasons prior to action is a prerequisite for accountability (Pettit et al., 2016).

A key element of Pettit's framework is the concept of "virtual control." Under this notion, an agent is considered reasons-responsive if their cognitive system monitors automatic decisions and intervenes only when a deviation or error occurs (Pettit et al., 2016). This approach effectively accommodates the observation that individuals do not always possess explicit knowledge of the reasons for their actions, acknowledging the pervasive role of automaticity and influences that operate outside deliberative awareness, as highlighted by behavioral, cognitive, and neurosciences (Pettit et al., 2016).

Central to Pettit's broader theory of freedom is the ideal of "non-domination." This principle objects to any exposure to a power of arbitrary interference, irrespective of whether that power is actively exercised (Pettit, 2001). Such potential for arbitrary interference diminishes an agent's freedom by undermining their "discursive control" over their interactions (Pettit, 2001). For Pettit, genuine freedom entails the capacity to participate in open discourse as equals, a capacity that is compromised when responsibility for actions is limited (Pettit, 2001). This aspect of Pettit's work extends individual agency into a broader social and political context, underscoring that freedom is not solely an internal psychological state but is also shaped by external social structures that either enable or constrain responsible agency (Pettit, 2001).

By redefining agent control in terms of reasons-responsiveness and virtual control, Pettit successfully preserves a meaningful sense of moral responsibility, even in light of the significant role played by unconscious processes. Responsibility, in this framework, is linked to an agent's inherent capacity to respond to reasons, even if the deliberation of those reasons is not always consciously present at the exact moment of action. The widely held philosophical view that free will is intimately connected to moral responsibility finds a robust defense in Pettit's approach, which begins from the intuition that there is no coherent sense in which someone could act freely yet not be held responsible for their actions, asserting this connection as a priori (Pettit, 2001).

Pettit's shift from an "act-of-will" model to one based on "reasons-responsiveness" and "virtual control" represents a move towards a more functionalist understanding of agency. Rather than searching for a specific, consciously initiated "moment" of will, his focus is on the overall capacity of the system (the agent) to produce outcomes that are appropriately aligned with its reasons, even if the underlying mechanisms are unconscious or automatic. This perspective aligns with how cognitive science often explains complex behaviors without requiring constant conscious oversight. If a system can reliably generate actions consistent with its guiding principles and can self-correct when deviations occur, then it demonstrates control. This allows for a robust concept of agency that is compatible with a brain that largely operates through unconscious processes.

To further clarify the distinction, a comparison between the traditional and Pettit's reimagined conceptions of agent control is presented below:

Feature	Traditional "Act-of-Will" Picture	Pettit's Agent Control
Basis of Control	Conscious, instantaneous "act of will" (Pettit, 2007)	Reasons-responsiveness; virtual control (monitoring and intervention) (Pettit et al., 2016)
Role of Consciousness	Originator of action (Libet, 1985; Soon et al., 2008)	Monitoring, attentional control, part of a process; not necessarily the origin (Jo et al., 2014; Kondrat, 2025; Pettit et al., 2016)
Compatibility with Neuroscience	Challenged by readiness potential and unconscious pre-determination (Libet, 1985; Soon et al., 2008)	Compatible; redefines control to accommodate empirical findings (Pettit, 2007; Pettit et al., 2016)
Link to Moral Responsibility	Requires uncaused origination or alternative possibilities [1]	Based on the capacity to respond to reasons, even if processes are determined [2]

Navigating Determinism: Pettit's Compatibilist Stance

Strong metaphysical determinism posits that the entire history of the universe is fixed, allowing for only one possible historical trajectory. This view extends beyond the mere determination of the future by the past; it asserts that the universe's state at any given time, combined with the operative laws of nature, completely dictates all subsequent states (Chen, 2020; Penrose, 1989). Consequently, all events, including human actions, are necessitated and could not have occurred otherwise (Chen, 2020). Roger Penrose (1989) famously articulated this, stating that strong determinism means "not just a matter of the future being determined by the past; the entire history of the universe is fixed, according to some precise mathematical scheme, for all time" (Chen, 2020; Penrose, 1989). This implies the elimination of all alternative nomological possibilities (Chen, 2020).

Traditionally, strong determinism presents a formidable challenge to the concept of free will. Incompatibilists argue that if all actions are predetermined, agents lack the ability to "do otherwise," a condition often considered essential for free will and, by extension, moral responsibility.[1] Hard determinists, accepting the truth of determinism, conclude that free will and moral responsibility are illusory (Startup, 2021). The intuitive understanding that one "could have acted otherwise" is a common way to formalize free action, and most individuals agree that moral responsibility cannot be assigned for actions that could not have been avoided (Startup, 2021). The "Consequence Argument" in incompatibilism articulates this challenge: if past facts and natural laws entail all future facts, and individuals have no power over the past or natural laws, then they have no power over future facts.[1, 3]

Pettit's redefined concept of agent control, rooted in reasons-responsiveness and non-domination, provides a foundation for a compatibilist stance. Compatibilism is the philosophical thesis that free will is compatible with determinism.[1] By moving away from the "act-of-will" picture—which often implies a need for an uncaused cause or libertarian freedom, inherently incompatible with determinism—Pettit's framework preserves a meaningful sense of control. He focuses on an agent's capacity to respond to and be guided by reasons. Even if these underlying processes are ultimately determined, a robust form of control is maintained.

This perspective aligns with the work of philosophers like Harry Frankfurt, who argued that alternative possibilities are not always a necessary condition for moral responsibility. Frankfurt-style cases illustrate scenarios where an agent can be held responsible for an action even if they could not have done otherwise (Frankfurt, 1969). Pettit's emphasis on reasons-responsiveness resonates with this, as it shifts the focus to the internal structure of the will and its responsiveness to reasons, rather than external alternatives. If free will is not about an uncaused, instantaneous "act of will," but rather about a system's capacity to be "reasons-responsive" and to exercise "virtual control" (monitoring and intervening when necessary), then determinism does not necessarily negate this capacity. A determined system can still be sufficiently complex to process reasons, weigh options (even if the weighing process itself is determined), and exhibit self-correction. The "freedom" in this context resides in the nature of the control mechanism—its responsiveness to reasons and its non-dominated status—rather than its ultimate causal origin being outside the deterministic chain. This perspective reframes the concept of "could have done otherwise" from a metaphysical sense to a conditional one, such as "could have done otherwise if I had willed differently" or "could have done otherwise if my reasons had been different," which is entirely compatible with a deterministic universe where those reasons and wills are themselves part of the determined causal flow.

By retaining a robust understanding of agent control, Pettit's framework also safeguards the basis for moral responsibility. If agents are reasons-responsive, they can be held accountable for their actions because their actions reflect their values, beliefs, and capacities for deliberation, even if these elements are ultimately part of a deterministic chain.[1] The challenge presented by neuroscience (micro-level brain activity preceding conscious awareness) and strong metaphysical determinism (a fixed universe) is that they appear to undermine agency at the macro-level of human experience. Pettit's resolution involves re-conceptualizing agency such that it emerges from these determined micro-processes. His concepts of "reasons-responsiveness" and "virtual control" function as higher-level descriptions of a system's behavior that are fully compatible with its underlying deterministic nature. This suggests that free will, or agent control, can be understood as an emergent property of complex determined systems, rather than a disruption of determinism. This represents a pivotal philosophical move that allows for both scientific understanding and the preservation of crucial humanistic concepts.

Interdisciplinary Synthesis: A Holistic View of Agency

The contemporary debate surrounding free will and agent control is inherently interdisciplinary, drawing crucial insights from various fields. Neuroscience provides the empirical data that challenges traditional notions of conscious, instantaneous will, particularly through findings like the readiness potential (Libet, 1985; Libet et al., 1983; Shibasaki & Hallett, 2006; Soon et al., 2008). It underscores the significant and often underappreciated role of unconscious processes in the initiation of action (Levy, 2014).

Neurophilosophy, positioned at the intersection of neuroscience and philosophy, plays a vital role in re-evaluating and redefining fundamental concepts such as free will, agency, and consciousness in light of empirical data.[4] Pettit's work stands as a prime example of this interdisciplinary endeavor, as he revises the concept of agent control to accommodate neuroscientific insights (Pettit, 2007; Pettit et al., 2016). This field actively questions whether free will requires redefinition given experimental data, often proposing models like compatibilism to reconcile determinism with a meaningful form of agency.[4] Neurophilosophy also grapples with the "explanatory gap" of consciousness and its intricate connection to free will and moral responsibility.[4]

Moral philosophy, in turn, confronts the profound implications of determinism and neuroscientific findings for the concept of moral responsibility.[2] Pettit's arguments enable the preservation of moral responsibility by linking it to reasons-responsiveness, rather than requiring metaphysical indeterminacy or a problematic "act-of-will" (Pettit, 2001; Pettit et al., 2016). The central question in moral philosophy remains whether moral responsibility can be reconciled with a deterministic universe, and whether alternative possibilities are truly necessary for it.[2] Frankfurt's influential work, which suggests that alternative possibilities are not always required for moral responsibility, is particularly relevant here, reinforcing the idea that responsibility does not necessarily depend on the ability to have "done otherwise" in a literal, counterfactual sense (Frankfurt, 1969).

Cognitive science further contributes to this holistic understanding by exploring the intricate mechanisms of decision-making, intentional action, and agency (Horst, 2015). This field reveals the complexity inherent in the formation of intentions and the initiation of actions, often involving skilled deliberative activity and attentional control, rather than a single, discrete conscious act (Horst, 2015). Cognitive neuroscience and psychology provide substantial evidence that actions are frequently shaped by unconscious information, and while some argue that conscious awareness of all relevant facts is necessary for responsibility, others, like Levy, contend it is indeed crucial (Levy, 2014). The "threat from above" to the will, which acknowledges that actions adapt to social dynamics of which individuals are often unaware, further highlights the distributed and complex nature of cognition and will (Pettit et al., 2016).

This interdisciplinary synthesis demonstrates a co-evolutionary process. Neuroscientific findings establish empirical constraints and present conceptual challenges. Neurophilosophy and cognitive science then engage in the crucial work of refining and reformulating philosophical concepts, such as free will and agent control, to ensure consistency with these empirical findings. Subsequently, moral philosophy re-evaluates the implications for human practices, including the assignment of responsibility. This is not a unidirectional process where science dictates philosophical conclusions, but rather a dynamic interaction where each field mutually informs and pushes the boundaries of the other. The outcome is a more nuanced, empirically grounded, and robust understanding of human agency. The debate surrounding free will and determinism serves as a compelling illustration of this necessary and productive co-evolution between empirical science (neuroscience, cognitive science) and conceptual analysis (neurophilosophy, moral philosophy), ultimately leading to more comprehensive and realistic models of human agency.

Conclusion: Beyond Illusion – A Robust Account of Freedom

Philip Pettit's 'Neuroscience and Agent Control' offers a sophisticated and compelling response to the neuroscientific challenges to the traditional understanding of free will. He successfully elevates the debate beyond a simplistic "free will is an illusion" dichotomy by arguing that the core problem lies not with the existence of agent control itself, but with an outdated and empirically unsupported "act-of-will" conception of it (Pettit, 2007).

By reimagining agent control as "reasons-responsiveness" and "virtual control," Pettit provides a theoretical framework that effectively accommodates the complex interplay of conscious and unconscious processes that characterize human action (Pettit et al., 2016). This allows for a robust understanding of agency that is compatible with strong metaphysical determinism, thereby preserving the essential foundation for moral responsibility.

This perspective suggests that while our intuitive sense of conscious, instantaneous control might be challenged by empirical findings, our fundamental capacity for meaningful agency, thoughtful deliberation, and accountability remains intact. This understanding of human freedom is more nuanced and empirically informed, encouraging an integrated view of human nature that embraces both our biological and social realities.

I don't personally find Pettit's intervention to be convincing. Often what is happening in debates between philosophers about difficult metaphysical questions like David Chalmer's hard question of consciousness and Philip Goff and Galen Strawson's conceptions of consciousness in their work about panpsychism is a reworking of concepts and definitions. Panpsychists are frequently not talking about what most people think of when they think of consciousness, for example. There are numerous easy ways of seeing this. For example, there is a distinction between phenomenal consciousness and conscious awareness in philosophical literature. 

Pettit sets out to change concepts and definitions: to reimagine agent control. Does this save free will from strong metaphysical determinism? It's not what philosophers like to call "a knock-down argument", but it's one of the best efforts available.

References

Chen, E. K. (2020). A strongly deterministic theory of physics is one that permits exactly one possible history of the universe. Synthese, 197(12), 5275-5296.

Frankfurt, H. (1969). Alternative possibilities and moral responsibility. Journal of Philosophy, 66(23), 829-839.

Frankfurt, H. (1971). Freedom of the will and the concept of a person. Journal of Philosophy, 68(1), 5-20.

Harris, S. (2012). Free Will. Free Press.

Horst, D. (2015). Deciding as intentional action: Control over decisions. Philosophical Issues, 25(1), 164-184.

Isham, E. A., Wulf, K. A., Mejia, C., & Krisst, L. C. (2017). Deliberation period during easy and difficult decisions: re-examining Libet's “veto” window in a more ecologically valid framework. Neuroscience of Consciousness, 2017(1), nix002.

Jo, H. G., Wittmann, M., Borghardt, T. L., Hinterberger, T., & Schmidt, S. (2014). First-person approaches in neuroscience of consciousness: Brain dynamics correlate with the intention to act. Consciousness and Cognition, 26, 105-116.

Kondrat, V. (2025). Neuronal World: Illusionistic Explanation of the Empirical Reality. Journal of NeuroPhilosophy, 4(1).

Levy, N. (2014). Consciousness and moral responsibility. Oxford University Press.

Libet, B. (1985). Unconscious cerebral initiative and the role of conscious will in voluntary action. Behavioral and Brain Sciences, 8(4), 529-539.

Libet, B., Gleason, C. A., Wright, E. W., & Pearl, D. K. (1983). 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(3), 623-642.

Mele, A. R. (2010). Libet on Free Will: Readiness Potentials, Decisions, and Awareness. In W. Sinnott-Armstrong & L. Nadel (Eds.), Conscious Will and Responsibility: A Tribute to Benjamin Libet (pp. 23-33). Oxford University Press.

Mele, A. R. (2011). Free Will and Science. In R. Kane (Ed.), Oxford Handbook on Free Will, 2nd Edition. Oxford University Press.

Nielson, T. I. (1963). The effect of manipulation of the visual feedback from a drawing hand on the perceived location of the hand. Perceptual and Motor Skills, 17(1), 105-110.

Penrose, R. (1989). The emperor's new mind: Concerning computers, minds, and the laws of physics. Oxford University Press.

Pettit, P. (2001). A theory of freedom: From the psychology to the politics of agency. Polity Press.

Pettit, P. (2007). Neuroscience and Agent Control. In N. Sebanz & W. Prinz (Eds.), Distributed Cognition and the Will: Individual Intentionality and Joint Intentionality (pp. 1-18). Princeton University.

Pettit, P., Holroyd, J., & Nachev, P. (2016). Neuroscience and Agent-Control. In N. Sebanz & W. Prinz (Eds.), Distributed Cognition and the Will: Individual Intentionality and Joint Intentionality (pp. 1-18).

Shibasaki, H., & Hallett, M. (2006). What is the readiness potential? Clinical Neurophysiology, 117(11), 2341-2356.

Soon, C. S., Brass, M., Heinze, H. J., & Haynes, J. D. (2008). Unconscious determinants of free decisions in the human brain. Nature Neuroscience, 11(5), 543-545.

Startup, R. (2021). Free Will and Determinism: Resolving the Tension. Open Journal of Philosophy, 11(4), 1139-1145.