The Door Is Not Locked: Free Will as a Capacity to Respond to Reasons

Free will can seem to disappear the moment a choice is described scientifically. If every event follows from earlier events together with the laws of nature, then my decision to read a book, apologise to a friend, or tell a lie was fixed long before I was born. It is tempting to think that freedom must be a gap in this chain: a moment at which I somehow escape the past and the laws of physics. Since no such escape appears possible, we conclude that human beings are merely complicated objects being pushed through the world.

I think this conclusion is too quick. I defend a broadly compatibilist view: we do have free will, provided that the ability to do otherwise is understood as a real capacity to recognise and respond to reasons. Free will is not the power to keep the entire past and laws of nature identical while producing a different future. Free will is the capacity to deliberate, to understand what matters, and to act differently when one’s judgement changes. The important question is less whether our actions have causes than whether our reasoning is among the causes that guide them.

The argument from causal determinism, given in Text 1, is powerful because it begins with two claims that seem undeniable. I cannot alter the past, and I cannot alter the laws of nature. If the past and those laws together determine my present action, then it appears that I cannot control that action either. Even if I feel as though I am choosing, that feeling may seem to be only the final stage of a process that was already fixed.

However, the argument moves too easily from “I did not create the causes of my action” to “my action is not under my control.” These are not the same claim. I did not create the laws of aerodynamics, but a pilot can still control an aircraft. The pilot’s control does not consist in standing outside physics and giving the plane an uncaused push. It consists in responding to changing conditions: reading instruments, adjusting direction, and correcting mistakes. The laws of physics make such control possible; they do not make it meaningless.

A thermostat also changes its behaviour when its environment changes. That is not yet agency. What matters in human deliberation is not merely that our behaviour varies with causes, but that we can ask whether a reason is good, whether our evidence is reliable, and whether a desire deserves to govern us. A person can discover that they were frightened, biased, or mistaken, and revise their judgement. This does not lift us outside causation. It makes us a special kind of cause: one capable of examining and correcting its own reasons.

A person is not a ghost placed outside the causal order, watching a body make decisions. A person is the organised system of memory, attention, character, imagination, and judgement through which some causes operate. When I pause before speaking because I realise that my words will hurt someone, the fact that my realisation has causes does not make it less mine. Indeed, if it had no connection to my understanding, values, or reasons, it would seem less like a free decision, not more.

This suggests that the difference between free and unfree actions is not whether they are caused, but how they are caused. Freedom is not simply present or absent. A blackmailed person may reason lucidly: handing over the money may be the sensible thing to do. Yet the blackmailer has made the alternative prohibitively costly, and so has compromised the conditions under which the person can choose. This is different from a decision to sign a business contract after considering its risks, where another person has not set the terms of deliberation through a threat. Causation alone does not remove freedom; coercion can diminish it because it narrows a person’s practical control.

This account must still answer the question in the prompt: could a person really have done otherwise? The answer depends on what an ability is. A swimmer has the ability to swim even while standing on dry land. A glass is fragile even while it remains unbroken on a shelf. Abilities are not shown only by what happens in one exact moment; they are shown by how something would respond in relevantly different circumstances.

The same is true of human choice. To say that someone could have acted otherwise is to say that their decision-making could have responded differently if they had understood the situation differently, encountered a stronger reason, or judged the balance of reasons in another way. Someone who refuses to drive after learning that they are dangerously tired shows this kind of capacity. Their decision is not random. It changes because they understand a reason that matters.

If “could have done otherwise” means that, with every fact about the past and every law of nature held fixed, I could still produce a different action, then causal determinism rules that out. I do not deny it. My claim is that this is not the most useful test of freedom. It asks whether one completed history can have two endings, rather than whether a person has the capacity to notice reasons, reconsider, and change course when the situation gives them reason to do so. A person who cannot leave their house because of a paralysing phobia lacks this ability, even if they sincerely want to leave. A person who stays inside after hearing that a dangerous storm is approaching may choose freely, because their decision changes in response to evidence they can understand.

Even if causal determinism is overcome, Text 2 presents a different threat of logical fatalism. Text 2 is strongest when it is not treated as an argument about causation. God’s belief need not force my action in order to create a problem. The real question is whether the proposition “Yesterday God believed that I would choose tea today” is fixed in the same way as the proposition “It rained yesterday.” I do not think it is. The first statement is grammatically about yesterday, but part of what makes it true concerns today. In a complete history in which I choose coffee, nobody reaches back and changes God’s past mind; rather, it was always true of that history that God believed I would choose coffee. This does not mean that I can alter God’s actual past belief. It means that God’s belief and my action naturally belong together in each possible history. This reply is controversial, but it shows that foreknowledge alone does not settle the issue. The fatalist needs the further claim that every statement about the past is equally beyond our counterfactual power. If free will is a capacity to respond to reasons, then an infallible observer does not remove that capacity. Even though God’s belief occurs beforehand, it tracks my choice just as passively as a perfect recording tracks the decisions in a chess game.

The most serious remaining objection concerns the origins of our character. I did not choose my genes, my childhood, or many of the influences that shaped what I value. If these formed the person who now weighs reasons, why should I receive credit or blame for what follows?

There is a harder version of this objection. Imagine that a neuroscientist quietly arranged my character so that I would always reach a particular conclusion. I might still give reasons for my action and feel that I had chosen it. We would hesitate to call that freedom. The determinist can then ask whether ordinary upbringing is merely slower, less visible manipulation.

Causal influence alone cannot be the difference, since education, friendship, and reflection also shape us. The better distinction is whether an influence leaves a person able to question and revise the motives it has produced, or installs a result while protecting it from revision. A belief learned from one’s family can later be examined, rejected, or endorsed for new reasons; an implanted command is designed to bypass that process. The boundary will not always be sharp, but it is real enough to explain why we distinguish persuasion from indoctrination, treatment from control, and an action that is mine from one merely occurring through me.

I have not shown that anyone is the ultimate author of themselves. Under determinism, no one is. But ultimate self-creation is an impossible standard, not a sensible condition of agency. To choose a character, a person would first need a character with which to make that choice, and so on without end. What matters is whether a person can understand reasons, test their own motives, and guide their conduct through capacities that are genuinely theirs. A caused action need not be an alien action. In that ordinary but demanding sense, the door is not locked.