The Problem

Suppose every thought you have, every decision you make, every movement of your body is the necessary result of prior brain states — which in turn result from prior physical states, in a chain stretching back to the Big Bang. If that is true, in what sense are you free to do otherwise than you do? And if you are not free, how can you be held morally responsible for your actions?

This, in its simplest form, is the problem of free will. It is one of the oldest questions in philosophy and remains unresolved.

Historical Background

The Stoics and Compatible Determinism

The Stoic philosophers, especially Chrysippus of Soli (c. 280–c. 207 BC), confronted an early version of this problem. For the Stoics, the universe is governed by a divine reason (logos) that determines all things. Yet Chrysippus argued that responsibility and determinism are compatible.

His cylinder argument (preserved by Cicero in De Fato): when someone pushes a cylinder, the external push initiates the motion, but the shape of the cylinder — its inherent nature — determines how it rolls. Similarly, external circumstances trigger our impressions, but how we respond — whether we assent or withhold assent — is determined by our own character. Virtue and vice are genuinely ours, even in a determined universe.

Epicurus and the Swerve

Epicurus (341–270 BC) took the opposite path: he introduced a random deviation (clinamen) in the motion of atoms — a physical indeterminacy that would break the chain of necessary causation and create room for freedom. This solution was criticized in antiquity: random atomic swerves are not the same as free choice. Randomness does not ground moral responsibility.

Augustine, Aquinas, and Medieval Free Will

Augustine (354–430) and Thomas Aquinas (1225–1274) approached the problem within the theological framework of divine providence and grace. If God foreknows all human actions, how can humans be free? Augustine distinguished divine foreknowledge (God knowing what we will do) from divine compulsion (God forcing us to do it). Aquinas argued that free will resides in the intellect and will deliberating together: this deliberation is freedom, even if causally determined by the agent’s rational nature.

The Three Fundamental Positions

1. Hard Determinism (Incompatibilist)

Determinism holds that every event — including every human action and decision — is the necessary consequence of prior states of the world plus the laws of nature. If this is true, and if free will requires the genuine possibility of having done otherwise (Principle of Alternate Possibilities), then free will does not exist.

Baron d’Holbach (1723–1789), in The System of Nature (1770), is the most radical Enlightenment advocate: humans are machines moved by necessary laws; the illusion of freedom results from ignorance of the causes that determine us.

Practical consequence: If there is no free will, retributive moral responsibility — punishing people because they deserve punishment — is unjustified. One might still defend punishment for consequentialist reasons (deterrence, protection), but not retribution.

2. Compatibilism

Compatibilism holds that freedom and determinism are compatible. Being free does not require that the action could have been different in the absolute metaphysical sense, but only that the action flows from the agent’s own reasons, desires, and deliberations — without external coercion or internal pathological compulsion.

David Hume (1711–1776), in the Treatise of Human Nature (1739), proposed that liberty means acting according to one’s own motivations, without external impediment. Causal necessity (that motives cause actions) does not threaten freedom — it presupposes it: if there were no necessary connection between character and action, responsibility would dissolve.

Harry Frankfurt (1929–2023), in “Freedom of the Will and the Concept of a Person” (Journal of Philosophy, 1971), offered the most influential contemporary compatibilist account. Frankfurt distinguishes:

  • First-order desires: wanting to do X
  • Second-order desires: wanting to have (or not have) the desire for X

A free person is one whose first-order desires align with second-order desires — one acts from the will one wants to have. An addict who acts compulsively but wants not to want the drug is not free: there is dissonance between the levels. This criterion does not require physical alternate possibilities.

3. Libertarian Free Will (Incompatibilist)

This position holds that genuine free will (requiring real alternate possibilities) exists and is incompatible with determinism — so determinism is (at least partially) false.

Kant (1724–1804) offers an elegant solution via transcendental dualism: as a phenomenon (empirical being, object of scientific knowledge), the human being is mechanically determined by natural causality. As a thing-in-itself (noumenal being, moral agent), the human being is free. This freedom is not observable or theoretically demonstrable, but is postulated by practical reason — it is a condition for the possibility of morality.

Agent causation (Roderick Chisholm, 1964) is another libertarian proposal: beyond event-causation, there is causation initiated by the agent as agent — a causal initiation irreducible to the physical chain of events.

The Neuroscience Challenge

Benjamin Libet (1916–2007) performed experiments in 1983 in which subjects flexed their wrists whenever they chose, while brain activity was monitored. Results suggested that the readiness potential (neural preparation) preceded conscious awareness of the intention to act by approximately 300–550 milliseconds.

Strong interpretation (anti-free will): the decision occurs in the brain before it becomes conscious — consciousness merely registers, it does not initiate.

Important critiques (Daniel Dennett, Adina Roskies, and others): The experiment does not show what it seems to show. First, what subjects report may not correspond exactly to the moment of intention. Second, even if consciousness is informed after the neural onset, the brain processes are processes of the agent — the boundary between “I” and “my brain” is artificial. Third, the experiment measured simple movements, not complex deliberations involving reasons and values.

Moral Responsibility: Strawson’s Reactive Attitudes

P.F. Strawson (1919–2006), in “Freedom and Resentment” (Proceedings of the British Academy, 1962), proposed a pragmatic approach that deeply influenced the debate. Strawson observes that our practice of moral responsibility is internal to a form of life — it is rooted in reactive attitudes (resentment, gratitude, indignation, love) that constitute human interpersonal relationships. To ask whether these attitudes are “metaphysically justified” given determinism is to step outside this form of life toward an “objective” standpoint that is unavailable in practice.

Conclusion

The problem of free will remains without consensus resolution. What philosophical debate has produced is rigorous clarification of the available positions:

  • Hard determinism: freedom is illusory; retributive responsibility unjustified
  • Compatibilism: freedom and determinism are reconcilable; responsibility rests on the agent’s authorship of actions
  • Libertarian free will: genuine freedom exists and requires indeterminism

The question is not merely theoretical: it bears on fundamental legal, moral, and therapeutic practices.

← Articles