Stop Press: Problem of Free Will Dissolved!#
Determinism, Pseudo-Determinism, and the Freedom Worth Wanting
Johannes Siedersleben, assisted by Claude Sonnet 4.5
October 2025
Do we genuinely choose our actions, or are our decisions merely the inevitable output of prior causes? Either our choices are determined by preceding events (in which case we seem not to be truly free), or they arise from some form of randomness or supernatural intervention (in which case they aren’t really our choices at all). Here is my claim: This dilemma rests on a false premise—that determinism and genuine agency are incompatible. By examining what we actually mean by “self” and “choice,” we can dissolve rather than solve the traditional free will problem.
Getting Started#
There are, in principle, three possible accounts of human decision-making:
External intervention: Some non-physical entity (soul, divine agent, immaterial will) influences our decisions from outside the natural order.
Fundamental randomness: A random generator, whether quantum or otherwise, introduces genuine indeterminacy into the decision process.
Causal determination: Our choices are fully determined by our genetic inheritance and our life experiences.
I reject the first option as unscientific and unsupported by evidence. I reject the second as incoherent: random choices are not free choices—they are merely dice rolls that happen to occur in our brains. I embrace the third option, but with an important qualification that I call pseudo-determinism.
The Repository Constitutes the Self#
Consider what we are. We possess a genetic endowment—perhaps a few megabytes of information encoded in DNA. We also possess the accumulated residue of our experiences: thirty years of life amounts to roughly one billion seconds, each of which we have lived through, processed, and mostly forgotten. This represents many terabytes of information. These two components—genes and experience—constitute what I am. There is nothing else to invoke, no homunculus sitting behind the eyes, no ghost in the machine.
When I make a decision, it depends on some bytes in this personal repository. The repository is me. Therefore, when my decisions flow from it, they flow from myself. This is self-determination in the most literal sense. The apparent threat posed by determinism—that I am merely a puppet of external forces—evaporates once we recognize that these “external forces” constituted me in the first place and continue to operate as me.
Pseudo-Determinism: The Boiling Water Analogy#
Consider a pot of boiling water. Most people would describe the bubbles as appearing randomly—unpredictable in their size, location, and timing. But if we could replicate the initial conditions down to the quantum level and run the process again, the exact same bubbles would form in exactly the same sequence. The bubbles are not genuinely random; they are pseudo-random. Their behaviour is fully determined by initial conditions too complex for us to track.
Human choices exhibit the same character. They feel free and unpredictable—even to the person making them—because the repository from which they emerge is fantastically complex. We cannot track all the variables, cannot predict the output until it arrives. Yet given identical initial conditions (identical genes, identical experiences, identical brain states), the same decision would emerge every time.
The butterfly effect illustrates this principle: a butterfly’s wing-flap in Tokyo might contribute to a thunderstorm in Cornwall, not because the connection is mysterious or magical, but because atmospheric systems exhibit sensitive dependence on initial conditions. The thunderstorm is fully determined by prior physical states, yet its occurrence is practically unpredictable. Similarly, my decision to have coffee rather than tea this morning was determined by countless prior causes stretching back through my personal history, yet no one—including myself—could have reliably predicted it in advance.
The term “pseudo-determinism” acknowledges both aspects: the ontological determinism (it is fully caused) and the epistemic unpredictability (we cannot, even in principle, track all the causes). This distinction matters enormously for how we understand ourselves. We are not helpless victims of external forces; we are complex systems whose outputs genuinely emerge from our own internal structure.
Freedom as Self-Determination#
This view aligns with a distinguished philosophical tradition. [Spinoza, 1677] argued that freedom consists in acting from one’s own nature, not in acting without cause. To be free is to be self-determined, not undetermined. A falling stone is not free because its motion is imposed externally. A living organism exhibits a kind of freedom when it moves according to its own internal purposes. A human being exhibits the highest form of freedom when decisions flow from rational deliberation, from the integration of values and experiences, from the rich repository that constitutes the self.
[Dennett, 1984] calls this “the freedom worth wanting”—the ability to respond to reasons, to have one’s actions flow from one’s own deliberative structure rather than from simple stimulus-response or external coercion. A thermostat is not free, even though it “responds” to temperature. But a human being is free, because the internal complexity is so vast that choices emerge from a genuinely autonomous process.
What drops away on this view is libertarian free will [Kane, 1996, O'Connor and Franklin, 2024, van Inwagen, 1983]—the metaphysically mysterious ability to have done otherwise in absolutely identical circumstances. But this was never a coherent concept to begin with. If I could have done otherwise in identical circumstances, the variation would have to come from either randomness (which I reject as incompatible with genuine agency) or from some causally efficacious but non-physical soul (which I reject as unscientific). The “freedom” promised by libertarian free will is either incoherent or undesirable.
Implications for Justice#
If someone commits a crime, and their choice was determined by their repository—perhaps shaped by an abusive childhood, genetic predispositions, or social deprivation—what follows for punishment and responsibility?
On my view, society retains legitimate interests in responding to criminal behaviour:
Deterrence: Creating incentives that shape future behaviour
Rehabilitation: Modifying the repository that generates harmful choices
Incapacitation: Temporarily removing dangerous individuals from circulation
What drops away is retributive justice—the idea that punishment is justified because the wrongdoer “deserves to suffer” in some metaphysical sense. Punishing someone for having made a deterministically caused bad decision is like punishing a computer for executing buggy code. We can quarantine the computer, fix the bug, or deter other systems from running similar code—but vindictive punishment serves no rational purpose.
This is not a soft-hearted position but a clear-eyed consequentialist one. The medieval logic of “an eye for an eye” should be abandoned in favour of approaches focused on reducing harm and improving outcomes. Notably, societies that have moved in this direction—particularly Scandinavian countries with rehabilitation-focused prison systems [Lappi-Seppälä, 2012, Pratt, 2008]—demonstrate dramatically better outcomes than those clinging to retributive models.
Challenges#
Several objections might be raised against this position.
The Simulation Argument#
Suppose we could perfectly simulate a human being—replicate all their genes, all their accumulated experiences, every detail down to the quantum level. We run the simulation: the person is offered coffee or tea, and chooses coffee. We reset the simulation with absolutely identical initial conditions and run it again. On my view, the simulated person must choose coffee every single time. But doesn’t this reveal that “pseudo-determinism” is just determinism in disguise? What work is the “pseudo” prefix actually doing?
Response: The simulation argument does not trouble me. Yes, identical conditions would produce identical choices—that is precisely what determinism entails, and I accept it. The “pseudo” prefix does real work not by introducing ontological indeterminacy but by capturing the practical and experiential difference between simple determinism (a billiard ball’s trajectory) and complex determinism (a human choice). The latter feels free because even the agent cannot predict the outcome until it emerges from deliberation.
The Irrelevance of Subjective Experience#
If our choices are fully determined by prior causes, what status does the subjective experience of deliberation have? When I feel that I am weighing options, considering reasons, and choosing between alternatives, is this experience merely an epiphenomenon—a kind of theatre that plays no causal role? If the decision is already determined before I begin to deliberate, doesn’t the feeling of choice become a kind of illusion?
Response: The subjective experience of deliberation is not epiphenomenal. Deliberation is part of the causal process by which the repository generates a choice. I am looking for the significant bytes in my repository. When I weigh reasons, consider consequences, and integrate values, these mental operations are not idle theatre—they are the mechanism by which my complex repository produces decisions. The feeling of choice corresponds to something real: the genuine unpredictability (even to myself) of which pattern in my repository will dominate.
Residual Responsibility#
Even granting that retributive punishment is unjustified, doesn’t this view eliminate the possibility of genuine moral responsibility? If a person’s actions flow inevitably from their repository, and they did not choose their genes or their early experiences, in what sense are they responsible for what they do? And if responsibility evaporates, don’t we lose something essential to human dignity and moral community?
Response: Responsibility remains coherent on this view, though reconceived. We hold people responsible not because they possess libertarian free will but because treating them as responsible agents—praising, blaming, reasoning with them—changes their repository in ways that improve future behaviour. [Dennett, 2003] captures it well: we are responsible because we are the kind of beings whose behaviour responds to being held responsible. This is a pragmatic, forward-looking conception of responsibility compatible with determinism.
The Self-Modification Problem#
If I come to believe that my future choices are determined by my current repository, and if I can deliberately modify that repository (through education, therapy, habit formation), doesn’t this create a strange loop? My current self deterministically causes modifications that will deterministically shape future choices—but the decision to modify seems to presuppose a kind of agency that transcends determinism. How does the determinist account for the person who says, “I don’t like who I am, so I’m going to change”?
Response: The self-modification problem is not paradoxical but illuminating. When I decide to change myself, that decision itself flows from my current repository—perhaps from a value of self-improvement, or from dissatisfaction with current patterns, or from social pressures I’ve internalized. The “agent” doing the modifying is not some free-floating homunculus but the repository itself, exercising its capacity for self-reflection and self-modification. This is recursion, not contradiction.
Prediction and Control#
If in principle my choices are determined by my repository, then in principle they are predictable by anyone with sufficient knowledge of that repository. This seems to open the door to manipulation. A sufficiently advanced AI or neuroscientist who understood my repository could predict and potentially control my choices. Doesn’t this reduce human autonomy to something illusory—vulnerable to prediction and manipulation in ways that undermine genuine freedom?
Response: The possibility of prediction and manipulation does not eliminate freedom in the sense that matters. Even if an omniscient observer could predict my choices, those choices still flow from my repository rather than being imposed externally. The relevant question is not “Could someone predict this?” but “Did this choice emerge from my own deliberative structure, or was it coerced?” Manipulation threatens freedom not because it makes behaviour predictable but because it bypasses or overrides the agent’s own reasoning process.
Conclusion#
The freedom worth wanting is not metaphysical immunity from causation. It is the capacity to act from one’s own nature, to have one’s choices flow from the rich, complex repository that constitutes the self. This is the only freedom that ever existed, and it is sufficient for morality, dignity, and human flourishing. The traditional free will problem, properly understood, dissolves once we recognize that we are regions of the causal order rather than spirits imprisoned within it.