Free Will and Determinism: Could We Have Chosen Otherwise?
Do our brains decide before we do, does randomness rescue freedom, and what does “could have chosen otherwise” actually mean? A guide to determinism, compatibilism, Libet-style experiments, and practical self-government without premature conclusions.
When we ask whether we have free will, it can seem that only two options exist. Either a person is a completely free originator of action, or a biological machine merely following a chain of prior causes. The philosophical and scientific debate is far more branched than that.
We first need to separate three questions. Is nature deterministic? Is human behavior predictable? And what counts as enough control for a decision to be called free? These questions are connected, but they are not identical.
It also matters what we mean by saying “I could have chosen otherwise.” This can mean that with exactly the same past state of the world and the same laws of nature, a different outcome could have occurred. Or it can mean a more practical capacity: I can understand reasons, inhibit an impulse, change my action when the reasons change, and recognize the action as my own.
For that reason, the free-will problem cannot be settled by one experiment, one physical theory, or one philosophical slogan. A more disciplined approach separates the metaphysical question from measurable capacities for agency and asks what each kind of evidence actually licenses us to conclude.
What determinism actually claims
Causal determinism says, roughly, that the state of the world at a time together with the laws of nature fixes what follows. If the past and the laws were exactly the same, the future would be the same as well. This is a strong metaphysical and physical thesis, not merely the claim that events have causes.
Determinism is not the same as fatalism. Fatalism suggests that some outcome will occur regardless of what we do. Under determinism, our actions, reasons, conversations, plans, and inhibitions are themselves parts of the causal chain. If a warning changes a person’s decision, the warning was not irrelevant; it was one of the causes of the different outcome.
Nor is it established that our world is wholly deterministic. Classical physics contains many deterministic models, but determinism is not completely straightforward even within classical mechanics. Quantum mechanics makes standard probabilistic predictions, while different interpretations and formulations treat fundamental indeterminacy differently. Contemporary physics therefore does not simply hand us either “the universe is proven deterministic” or “the universe is proven indeterministic.”
Predictability is not the same as determination
Even if a system were deterministic, it would not follow that we could predict it in practice. Chaotic systems can be extremely sensitive to initial conditions: small measurement errors grow over time, limiting long-range prediction even when the underlying equations are not random.
The reverse is also true. A system can contain stochastic processes while remaining highly predictable at the level of large populations. Weather probabilities, disease risks, or population-level behavioral patterns can carry useful predictive information without determining an individual act with certainty.
Successful prediction of behavior is therefore not automatically proof of determinism. It shows that some information is related to a later outcome. How much control, alternative capacity, and authorship the person has is a further question.
Randomness by itself is not freedom
If determinism is troubling because it appears to close off alternative futures, an obvious escape route is tempting: add randomness. But a random event does not become my free decision merely because it was not predetermined.
If an important choice were settled by a random quantum event, the result would be less determined, but not necessarily more mine. That is why contemporary incompatibilist theories that connect freedom with indeterminism usually require something more: agent control, an appropriate role for reasons, authorship, or a particular placement of indeterminacy within the decision process.
This is an important boundary for popular appeals to quantum physics. Quantum indeterminacy, where an interpretation treats it as fundamental, does not by itself explain personal agency. Indeterminism may leave room for more than one outcome; it does not by itself explain who or what uses that room in an intelligible way.
Three main families of answers
The first family is incompatibilist: if determinism is true, freedom in the relevant sense is not possible. Within this family, libertarians hold that at least some human actions are not determined in a way that destroys freedom; skeptics or hard incompatibilists argue that we do not possess the required kind of freedom whether the world is strictly deterministic or includes chance.
The second family is compatibilism. Compatibilists do not require a free agent to break the laws of nature or create a cause from nothing. They focus on whether action issues from the person’s reasons, values, character, and capacity to respond to reasons without coercion, manipulation, or a relevant failure of control.
A third stance is not one theory but restraint: leave the metaphysical problem open while studying concrete components of agency. Inhibition, planning, weighing reasons, learning from consequences, and revising decisions can be investigated even without a final answer to whether a different world could have emerged from an exactly identical past.
What does “could have chosen otherwise” mean?
The sentence has at least two important meanings. An incompatibilist can understand it categorically: given the very same past and the very same laws of nature, a genuinely different choice had to remain possible. If determinism permits only one future, that form of alternative possibility is excluded.
Many compatibilists reformulate the question. What matters need not be whether I could miraculously choose differently while the entire universe remained unchanged, but whether my decision system can recognize and respond to reasons. If I learned that a medicine was poisonous, I would reject it; if I discovered that key information was false, I would revise the plan. Such reasons-responsiveness can be part of control.
The debate about moral responsibility is therefore more complicated still. Frankfurt-style cases influentially challenged the principle that responsibility always requires the literal ability to do otherwise. Part of contemporary debate focuses instead on the source of action and the quality of control, not only on the number of metaphysically open alternatives.
What Libet actually measured
In the early 1980s Benjamin Libet and colleagues compared electrical brain activity before a spontaneous wrist movement with a participant’s later report of when the urge or intention to move first became conscious. The so-called readiness potential, averaged across trials, began several hundred milliseconds before the reported moment of conscious intention.
The result became famous because it can be told as a simple story: the brain started the decision before the person knew about it. But the experiment did not directly observe a metaphysical decision and then measure its cause. It compared a brain signal, averaged over repeated trials, with a retrospective timing report about subjective experience during a very simple, arbitrary movement.
Libet himself did not conclude that all conscious control was necessarily excluded; he discussed the possibility of a late conscious veto. Yet the proposed veto is not an experimentally established solution to the free-will problem. The more important lesson is that the temporal priority of one neural signal does not by itself justify a leap to a conclusion about the whole of human agency.
Why the readiness potential is not a simple “decision signature”
Later work has shown that the interpretation of the readiness potential is more open than it first appeared. Schurger and colleagues showed that an averaged readiness-like pattern can arise when spontaneous fluctuations in neural activity stochastically reach a movement threshold and trials are then aligned to the time of movement.
This model does not prove free will. It does show that the early onset of the averaged signal need not be the moment when the brain has already completed a decision. Modern reviews also emphasize that the subjective timing of intention is difficult to determine reliably and that Libet’s clock may measure something more indirect than the moment at which a decision is actually formed.
The readiness potential therefore remains an important phenomenon in the study of self-initiated action, but its philosophical interpretation is not simple. Neural processing before a conscious report is expected; the unresolved question is what that processing represents and when a decision is sufficiently formed to count as chosen.
What do predictions several seconds earlier mean?
In a well-known 2008 experiment, patterns of activity in prefrontal and parietal cortex predicted which button a participant would press several seconds before the participant reported a conscious decision. The result matters, but the decoding was not perfect: prediction was above chance rather than a certain readout of an already completed future choice.
Neural activity before a reported decision may contain early biases, preparation, an emerging preference, or part of an evidence-accumulation process. If a signal raises the probability of an outcome above chance, that is not the same as proving that the decision was already irrevocably fixed.
More importantly, such tasks are usually arbitrary: press left or right for no important reason and with no meaningful consequence. Maoz and colleagues compared arbitrary decisions with deliberate decisions carrying real consequences and found a clear readiness potential for arbitrary choices while it was strongly diminished or absent for deliberate choices. This does not prove freedom, but it warns against treating a laboratory button press as a complete model of human decision-making.
From brains to responsibility
Every decision has a neural basis. Finding brain activity before an action does not reveal a rival to the person who decided; it reveals part of the process through which the person decides. For neuroscience to eliminate responsibility, an additional argument would be needed showing why the realization of decision-making in the brain entails the absence of relevant control.
This does not mean that control is always equal. Threats, physical coercion, powerful manipulation, severe impairment of judgment, or inability to understand consequences can reduce authorship and responsibility. That is why law, ethics, and ordinary judgment have long distinguished intentional action from accident, compulsion, and diminished control.
The practical question is therefore not only “does free will exist in an absolute sense?” but also: what capacities for control did this person have in this situation, what information was available, which reasons could the person respond to, and how much was the action shaped by direct coercion or manipulation?
Inner monarchy as a practice, not a metaphysical proof
In THY-REALITY’s language, inner monarchy means self-government: the capacity not to hand every action over to the first impulse, fear, group, or command. That idea does not require the claim that we are uncaused originators outside nature. It requires a more modest and testable claim: people differ in capacities for self-control, reflection, learning, and responsiveness to reasons, and those capacities can be strengthened or weakened.
If compatibilism is correct, such capacities may be an important part of freedom. If a libertarian theory is correct, they would still be required for meaningful use of a deeper metaphysical freedom. If skepticism about free will is correct, they remain important for reducing harm, forming habits, and making behavior more reflective. The practical value of self-government therefore does not depend on first settling a centuries-old metaphysical dispute.
The most defensible conclusion is twofold. Science shows that decisions are processes containing both unconscious and conscious stages and that brains prepare actions before we can report them. It does not show that one experiment has proved the nonexistence of free will. Philosophy, meanwhile, shows that the answer also depends on what we require freedom to be: an open alternative future, authorship of action, responsiveness to reasons, or enough control for responsibility.
Sources and further reading
- THY-REALITY — Naravni zakon in vprašanje svobode / Natural Law and the Question of Freedom (LOCKED): practical freedom, constraints, control, knowledge and responsibility without claiming a settled metaphysics of free will.
- THY-REALITY — Svoboda brez odgovornosti ni svoboda / Freedom Without Responsibility Is Not Freedom (LOCKED): control and epistemic conditions of responsibility; inner monarchy as self-government.
- THY-REALITY — Notranji kompas in jasna izbira / Inner Compass and Clear Choice (LOCKED): reasons, reflection, values and disciplined choice without treating intuition as infallible.
- THY-REALITY — Ali avtoriteta odpravi osebno odgovornost? / Does Authority Remove Personal Responsibility? (LOCKED): coercion, control, authorship and distributed responsibility.
- THY-REALITY — Notranja monarhija, zunanja anarhija: red brez vladarja ni kaos / Inner Monarchy, Outer Anarchy: Order Without a Ruler Is Not Chaos (LOCKED): self-government as control over one’s own conduct, not metaphysical omnipotence.
- THY-REALITY — Um, možgani in zavest: tri stvari, ki jih pogosto zamenjujemo / Mind, Brain and Consciousness: Three Things We Often Confuse (LOCKED): brain, mind and consciousness kept conceptually distinct.
- Hoefer, C. — Causal Determinism. Stanford Encyclopedia of Philosophy, substantive revision 2023. Determinism concerns whether laws plus a state of the world necessitate later states; its truth in physics remains nontrivial.
- O’Connor, T.; Franklin, C. — Free Will. Stanford Encyclopedia of Philosophy. Overview of alternative-possibilities, sourcehood, libertarian, compatibilist and skeptical approaches.
- McKenna, M.; Coates, D. J. — Compatibilism. Stanford Encyclopedia of Philosophy, substantive revision 2024. Compatibilism holds that freedom and moral responsibility may be compatible with determinism.
- Clarke, R.; Capes, J. — Incompatibilist (Nondeterministic) Theories of Free Will. Stanford Encyclopedia of Philosophy. Indeterminism alone is not sufficient; theories add accounts of control, causation or agency.
- Libet, B.; Gleason, C. A.; Wright, E. W.; Pearl, D. K. — Time of Conscious Intention to Act in Relation to Onset of Cerebral Activity (Readiness-Potential). Brain 106(3), 623–642 (1983).
- Schurger, A.; Sitt, J. D.; Dehaene, S. — An accumulator model for spontaneous neural activity prior to self-initiated movement. PNAS 109(42), E2904–E2913 (2012).
- Soon, C. S.; Brass, M.; Heinze, H.-J.; Haynes, J.-D. — Unconscious determinants of free decisions in the human brain. Nature Neuroscience 11, 543–545 (2008). Choice-related information was decodable seconds before reported awareness, but not with deterministic accuracy.
- Maoz, U.; Yaffe, G.; Koch, C.; Mudrik, L. — Neural precursors of decisions that matter—an ERP study of deliberate and arbitrary choice. eLife 8:e39787 (2019). Readiness potentials were robust for arbitrary choices and strongly diminished/absent for deliberate choices.
- Brass, M.; Furstenberg, A.; Mele, A. R. — Why neuroscience does not disprove free will. Neuroscience & Biobehavioral Reviews 102, 251–263 (2019). Review of inferential limits of Libet-style paradigms.
- Schurger, A.; Hu, P. B.; Pak, J.; Roskies, A. L. — What Is the Readiness Potential? Trends in Cognitive Sciences 25(7), 558–570 (2021). Reassessment of the RP and its relevance to classic free-will arguments.
- Dominik, T.; Mele, A.; Schurger, A.; Maoz, U. — Libet’s legacy: A primer to the neuroscience of volition. Neuroscience & Biobehavioral Reviews 157, 105503 (2024). Systematic review of the Libet literature and major methodological and conceptual criticisms.