Free will is the longest-running unresolved problem in Western thought: whether what you just did was up to you, or whether it merely felt like it was.
free-will
The capacity — if it exists — of an agent to genuinely choose between alternatives in a way that is not wholly determined by prior causes.
The problem it answers
When you make a choice, is the choosing genuinely up to you, or is it the inevitable output of prior causes — your genes, your upbringing, the firing of neurons — over which you had no control? The question matters because almost everything we take for granted morally and legally depends on the answer. Praise, blame, punishment, gratitude, regret, responsibility — each of these makes sense only if the person we are praising or blaming could have done otherwise.
This is the problem of free will, and it is genuinely unsolved. After 2,500 years of argument, philosophers have not reached consensus on whether free will exists, what it would consist in if it did, or whether it matters whether it does. The disagreement is not for lack of intelligence applied; it is because the problem sits at the intersection of three pictures — the manifest image of agency, the scientific image of causal closure, and the moral image of responsibility — that resist easy reconciliation.
The core claim
There is no single core claim about free will; the field is structured by the disagreement. The central positions are:
- Libertarianism (the metaphysical kind, not the political): free will exists and is incompatible with determinism. When you make a free choice, you genuinely could have done otherwise, all prior conditions being identical.
- Hard determinism: free will does not exist. Every event, including every choice, is the inevitable consequence of prior conditions plus the laws of nature. The feeling of free choice is illusion.
- Compatibilism: free will exists and is compatible with determinism. What we mean by free is not uncaused, but acting from your own reasons and desires, free from coercion. A choice can be both determined and free in this sense.
- Hard incompatibilism: free will requires conditions that neither determinism nor indeterminism can provide. Therefore free will does not exist, regardless of whether the universe is deterministic.
Most contemporary philosophers are compatibilists, though hard incompatibilism has serious recent defenders. Libertarianism is the historical default and the position of folk psychology.
History in one paragraph
Aristotle opens the problem in the Nicomachean Ethics by distinguishing voluntary from involuntary action and asking which deserves praise or blame. The Stoics inherit a deterministic cosmos (everything follows from logos) and develop sophisticated compatibilist accounts: Chrysippus argues that responsibility requires only that your character be the proximate cause of action, not that you could have done otherwise. Augustine introduces the theological frame: how can we be responsible for sin if God knows in advance what we will do? His answer — that foreknowledge is not causation — sets the terms for a millennium of Christian debate. Aquinas synthesizes Aristotelian and Augustinian accounts. The Protestant Reformation reopens the wound: Luther's Bondage of the Will (1525) flatly denies free will; Erasmus defends it. The Enlightenment produces mechanistic determinism in Hobbes and Spinoza, who is the most uncompromising determinist in the canon. Kant rescues freedom by locating it outside the causal order altogether: as phenomena we are determined; as noumena we are free. Hume gives the canonical compatibilist statement: liberty is the absence of constraint, not the absence of causation. The 20th century brought new tools. Frankfurt cases — thought experiments by Harry Frankfurt — challenge the requirement that free will demand the ability to do otherwise. Galen Strawson's basic argument aims to show that no matter how free will is analyzed, ultimate responsibility is impossible. Empirical results from Libet's neuroscience experiments (1983) seemed to show that brain activity precedes conscious decisions, reigniting the debate. The argument continues.
The compatibilist move
The compatibilist refuses the framing that free will and determinism are opposed. What we mean by free will, the compatibilist argues, is not uncaused choice; it is choice that flows from the agent's own character and reasoning, without external coercion. On this analysis, a choice can be perfectly determined by your beliefs, desires, and history — and still be yours, and still be free.
The move is powerful. It preserves the practices that matter — holding people responsible, distinguishing voluntary from coerced action, allowing for genuine deliberation — without requiring metaphysically extravagant assumptions. Hume put it cleanly: by liberty, then, we can only mean a power of acting or not acting, according to the determinations of the will. The slave is unfree because their action does not flow from their own will; the deliberating person, however determined by their history, is free because their action does.
The critics' response: compatibilism redefines the question rather than answering it. The folk concept of free will does involve the ability to do otherwise, the felt sense of open future, the conviction that the choice is up to me in a way no analysis of caused action can capture. Compatibilism saves the word but loses the thing.
The libertarian alternative
Metaphysical libertarians insist that free will requires that, at the moment of choice, you could genuinely have done otherwise — not in some attenuated sense, but really. This requires that the choice not be fully determined by prior conditions plus the laws of nature.
The immediate problem is that the alternative to determinism is not freedom but randomness. If your choice is not determined by prior reasons, then it is just a flip — and a random outcome is no more yours than a determined one. The libertarian needs a third option: a kind of causation that is neither deterministic nor random, where the agent themselves is the genuine source.
The most sophisticated attempts — Robert Kane's self-forming actions, the agent-causal theories of Roderick Chisholm and Timothy O'Connor — try to make sense of agent causation as a primitive: a kind of causing that originates with the agent and is not reducible to event-causation among prior states. Whether this is intelligible is contested. Whether it fits with the natural sciences is even more contested.
The hard incompatibilist position
Galen Strawson's basic argument runs roughly: to be truly responsible for an action, you have to be responsible for the character that produced it. To be responsible for your character, you would have had to choose it. But to choose your character, you would already have had a prior character, which you would also have had to choose, and so on infinitely. Therefore ultimate responsibility is impossible.
The hard incompatibilist accepts the conclusion: we are not ultimately responsible for what we do. Derk Pereboom and Strawson argue that we should accept this and revise the moral practices that depend on the old picture. Punishment as desert (you deserve this for what you did) becomes incoherent; quarantine and rehabilitation models remain available. Praise, blame, and pride lose their deepest grip but can be retained in a thinner sense.
This is the position that takes the philosophical pressure most seriously and gives back the largest changes to practice. Many find it unlivable.
Common confusions
- Free will is not the same as freedom from coercion. A prisoner in a cell may have free will in the metaphysical sense but is not politically free. A person at gunpoint may be coerced but still possess free will in the deep sense. The two questions are different.
- The experience of free will is not evidence of free will. Determinists can grant that we experience deliberation and choice; the question is what that experience tracks.
- Quantum indeterminism does not solve the problem. Even if the universe is fundamentally indeterministic at the quantum level, randomness at the synaptic level is not freedom — it is noise. The libertarian needs more than chance.
What it isn't
Free will is not political freedom. Political freedom concerns the absence of external constraints — laws, social pressure, force. Free will concerns the internal structure of agency itself, independent of external conditions. The two are easily conflated and the conflation is rarely useful.
Free will is also not self-control. A person with strong self-control may have less free will than a more impulsive person, by some libertarian lights, because their actions are more reliably predictable from their character. Self-control is a virtue; free will is a metaphysical condition. The relation between them is contested.
Finally, free will is not the same as moral responsibility. They are usually treated together, but it is conceptually possible to have one without the other. Hard incompatibilists deny free will but try to preserve some form of moral practice. Strict determinists may deny both. The combinations are open.
Live debates
- Do the Libet experiments matter? Libet's results — measurable brain activity ~300ms before reported conscious decision — were taken as empirical evidence against free will. The interpretation is now widely contested. The experiments may show only that conscious awareness of decisions lags the decisions themselves — which is interesting but does not settle the metaphysics.
- Can free will be naturalized? Compatibilists generally think so; libertarians and hard incompatibilists generally think not. Whether agent causation can be made consistent with physics is the live question.
- What should we do about moral practice? If free will turns out to be limited or absent, what should change about how we praise, blame, and punish? The applied debate runs from criminal justice reform to the ethics of psychiatry.
Why this still matters
Free will is the philosophical question with the largest gap between how it feels and what argument can establish. The felt sense of authorship is overwhelming. The arguments for it are surprisingly thin. The arguments against are surprisingly strong. Sitting with that gap is good intellectual discipline.
For Philosophy: the free will question is the entry point to a remarkable web of related problems — personal identity, moral responsibility, the relation between mind and brain, the structure of agency. Almost any direction you push from here leads somewhere interesting.
For Deconversion: a great deal of inherited religious frameworks (especially Christian ones) rest on a libertarian picture of free will — you genuinely could have done otherwise, you are therefore responsible, divine judgment is therefore just. If you leave the framework, what happens to the picture? Most people don't notice that their everyday moral psychology still runs on assumptions that the inherited framework was load-bearing for. Working through the free will question carefully is one of the most useful exercises in conscious worldview reconstruction.
Further reading
- Virtue — what depends on free will being real
- Augustine — the foundational Christian treatment
- Kant — freedom located outside the causal order
- Spinoza — the most uncompromising determinist
- Bondage of the Will — Luther's denial
- Belief Systems — the deconversion connection
This is a Pillar concept. Satellite concepts should link here as their parent.