The Epicurean paradox explained: the riddle of evil Epicurus probably never wrote

Explained

The Epicurean paradox runs like this: if a god wants to remove evil but cannot, he is weak; if he can but will not, he is malevolent; if he neither can nor will, he is no god at all; and if he both can and will, where does evil come from?

It is one of the most quoted arguments in the philosophy of religion, and it usually travels with Epicurus’s name attached. The attribution is shaky. No surviving text of Epicurus contains the argument, the earliest version we have comes from a Christian writer who quoted it in order to refute it, and the English wording most people know was written by David Hume in the eighteenth century. This guide sets out what the argument says, where it really comes from, what it does and does not prove, and how it grew into the modern problem of evil.

Quick answer

The Epicurean paradox is a dilemma about evil and a god who is all-powerful and all-good: either he cannot stop evil, will not, or both, so why does evil exist? It is credited to Epicurus, but our earliest source is Lactantius (On the Anger of God, chapter 13, early 4th century AD), and many scholars think it came from ancient sceptics instead. The famous English version is Hume’s, from Dialogues Concerning Natural Religion (1779).

The Epicurean paradox explained: the riddle of evil Epicurus probably never wrote

What the Epicurean paradox says, step by step

Four options, and why none of them looks comfortable

The argument works by listing every combination of two things a god might have: the will to remove evil and the power to remove it. It then shows that each combination clashes with what believers in a perfect god usually say about him. It belongs to the same family of ancient questions about fate, fortune and the gods that runs through Greek and Roman thought, which our guide to existentialism for readers picks up much later, when the question turns from “why does God allow this?” to “what do I do in a world that gives no answer?”

  1. Willing but not able: the god is weak, so not all-powerful.
  2. Able but not willing: the god is malevolent, or at least indifferent, so not all-good.
  3. Neither willing nor able: the god is both weak and malevolent, so why call him a god?
  4. Both willing and able: this is the only option that fits a perfect god, but then where does evil come from?

The force of the argument sits in the fourth line. The first three describe gods that many people would not worship; the fourth describes the God of classical theism, and the argument asks why such a being’s world contains so much pain. Everything that follows in the history of the debate is, in one way or another, an attempt to answer that fourth question or to deny that the dilemma is set up fairly.

Did Epicurus write it? The Lactantius trail

The earliest witness is a Christian apologist, writing six centuries later

Epicurus lived from about 341 to 270 BC and wrote a great deal, but only a small part survives: three letters and a set of maxims preserved by the biographer Diogenes Laertius, plus fragments. None of them contains this argument. The oldest text that does is De Ira Dei (On the Anger of God), written in the early fourth century AD by Lactantius, a Christian rhetorician and adviser at the court of Constantine. In chapter 13 he sets out the four options, credits them to Epicurus, and then answers them. In other words, our only witness is a hostile one who quoted the argument to knock it down.

That alone does not make the attribution false, since Lactantius may have known Epicurean writings now lost. But there are reasons for doubt. A closely related argument against divine providence appears in the third book of Sextus Empiricus’s Outlines of Pyrrhonism, a handbook of Greek scepticism from around the second century AD, and Sextus does not credit it to Epicurus. Several scholars have suggested that the four-way form comes from the sceptical Academy, the school associated with Carneades, which made a habit of arguing against the Stoic idea of a caring providence. On this reading, later writers pinned it on Epicurus because he was the best-known ancient critic of meddling gods.

There is also a problem of fit. The argument sounds like an argument for atheism, and Epicurus was not an atheist. He believed the gods exist; he simply denied that they run the world or care what we do. A thinker with that view had no need for a dilemma that ends with “then there is no god”. The safest summary is that the argument is Epicurean in spirit, because it attacks the idea of gods who govern human affairs, but that its exact authorship is unknown.

Hume's Philo and the wording everyone quotes

The 1779 Dialogues gave the riddle its modern English voice

The lines most people meet online, often printed over a photo of a Greek bust, are David Hume’s. In Part X of Dialogues Concerning Natural Religion, published in 1779, three years after his death, the sceptical character Philo says: “Epicurus’s old questions are yet unanswered. Is he willing to prevent evil, but not able? then is he impotent. Is he able, but not willing? then is he malevolent. Is he both able and willing? whence then is evil?”

Two details matter here. First, these are the words of a character in a dialogue, not a direct statement of Hume’s own view, and the book is built so that no speaker simply wins. Second, Hume himself credits the questions to Epicurus, which shows that the attribution was already conventional by his time. The step from Hume to the viral “Epicurean paradox” flowchart, with arrows branching from “Is there evil?” to “Then why not stop it?”, is modern: the flowchart is an internet graphic of unknown origin, not an ancient diagram.

Why the Epicurean paradox is a dilemma, not a disproof

Three common misreadings

It is not a proof that no god exists. The argument targets one specific idea of God: a being who is all-powerful, all-knowing and perfectly good at the same time. It says nothing against a limited god, a god who does not care about humans (the gods of Epicurus), or the impersonal, rational divine order of the Stoics described in our guide to what the Stoics actually taught. Process theologians, for instance, accept that God is not all-powerful in the classical sense, and the dilemma has little grip on them.

It is not a paradox in the strict logical sense. A logical paradox, like the liar sentence, produces a contradiction from premises that all seem true. The Epicurean argument is a dilemma (some call it a trilemma, folding the third option into the others): it lays out the options and argues that each is unacceptable to the believer. A believer can escape by rejecting a premise, for example by arguing that a good god might have a reason to permit some evil. Whether such reasons are convincing is the real debate.

It is not Epicurus’s main teaching. Epicurus is remembered for an ethics of calm pleasure, freedom from fear and friendship. His surviving writings spend far more time on why we should not fear death or the gods than on whether a perfect god could allow evil. Reading him through this one argument is a bit like reading Nietzsche only through “God is dead” or only through amor fati: the slogan travels, the philosophy behind it stays at home.

From an ancient riddle to the modern problem of evil

Logical and evidential versions, and the word theodicy

Philosophers today call the family of arguments the problem of evil, and they split it in two. The logical problem claims that God and evil cannot both exist: the set of statements “God is all-powerful”, “God is wholly good” and “evil exists” is inconsistent. Its classic modern statement is J. L. Mackie’s paper “Evil and Omnipotence”, published in the journal Mind in 1955, which argued that the standard religious replies either fail or quietly give up one of the three claims.

The best-known reply came from Alvin Plantinga, whose free-will defence appeared in its most cited form in God, Freedom, and Evil (1974). Plantinga did not try to explain why God actually permits evil. He argued only that it is possible that a world with free creatures who sometimes choose wrongly is better than a world without freedom, and that even an all-powerful God cannot make free creatures who are guaranteed always to choose well. If that is even possible, there is no strict contradiction. Many philosophers, including some atheists, came to accept that this answers the logical version, though not everyone agrees.

The debate then moved to the evidential problem. In “The Problem of Evil and Some Varieties of Atheism” (1979), William Rowe granted that God and evil might be logically compatible but argued that the amount of apparently pointless suffering makes God’s existence unlikely. His best-known example is a fawn trapped in a forest fire, burned and dying slowly, with no one ever knowing. What good could that serve? Replies to Rowe include “sceptical theism”, the view that we should not expect to see God’s reasons, since a finite mind may simply be unable to grasp them.

Older still is the tradition of theodicy, a word Gottfried Wilhelm Leibniz coined for his 1710 book Essais de Théodicée, from Greek words for “god” and “justice”. Leibniz argued that God chose the best of all possible worlds, evil included. Voltaire mocked the idea in Candide (1759) after the Lisbon earthquake of 1755. Before Leibniz, Augustine had argued that evil is not a thing at all but a lack or corruption of good, entering the world through the misuse of free will. A second strand, traced to the second-century bishop Irenaeus and developed by John Hick in Evil and the God of Love (1966), sees the world as a place of “soul-making”, where hardship is the setting in which moral and spiritual growth becomes possible.

Five replies to the Epicurean paradox, and their critics

Each answer, who made it, and the strongest objection
ResponseKey figureWhat it claimsMain objection
Evil as privationAugustine (4th-5th century)Evil is a lack of good, not a created thing; it enters through the misuse of free will.Pain and cruelty feel like real presences, not mere absences, and natural disasters involve no choice.
Soul-makingIrenaeus; John Hick (1966)A world with real hardship is needed for moral and spiritual growth.Much suffering crushes people rather than growing them, and animals suffered long before humans existed.
Best possible worldLeibniz (1710)God chose the best of all possible worlds; its evils are part of the best overall design.It seems easy to imagine a slightly better world; Voltaire’s Candide made the claim look cruel.
Free-will defenceAlvin Plantinga (1974)It is possible that God cannot create free creatures guaranteed never to choose evil, so God and evil are consistent.It answers the logical version only, and says little about earthquakes and disease.
Sceptical theismStephen Wykstra and others (1980s)We should not expect to see God’s reasons, so apparently pointless evil is weak evidence against God.If God’s reasons are beyond us, we may also lose grounds for saying anything about God’s goodness.
Main responses to the problem of evil, simplified. Each has defenders and critics among both religious and non-religious philosophers.

The table simplifies a large literature, and each line has been refined many times. It also leaves out positions that refuse the terms of the debate: philosophers who conclude that the evidence favours atheism, believers who hold that faith does not depend on solving the puzzle, and traditions in which God is not understood as an all-powerful individual at all. For a general reader, the useful lesson is that the paradox’s four lines turned into two distinct questions: whether God and evil are consistent at all, and whether the evil we see makes God improbable.

What Epicurus himself believed about the gods

Happy, deathless and uninterested in us

Epicurus’s own view is stated in the first of his Principal Doctrines: a blessed and immortal being has no troubles of its own and causes none to anyone else. The gods, on his account, live in perfect tranquillity, and that is precisely why they do not run the world. Governing nature, punishing the wicked and answering prayers would be work and worry, and worry is incompatible with perfect happiness. The cosmos runs by itself, through atoms moving in the void.

This makes Epicurus an opponent of providence rather than of gods. His target was the fear that the gods watch, judge and punish, a fear he thought ruined human lives. That puts him at odds with the Stoics, his great rivals, who saw the universe as ordered by a rational, provident divine principle. The Stoic sayings collected in our page of Stoic quotes with sources show the opposite attitude: accept what happens, because it belongs to a rational whole. In a sense, the Epicurean paradox is aimed at the Stoic god, which may be why later writers found it natural to give it to Epicurus.

Where Yalom meets Epicurus: death, not evil

A different part of the philosophy

Irvin Yalom does draw on Epicurus, but not on the problem of evil. In Staring at the Sun (2008), his book on the fear of death, he presents Epicurus as one of the first thinkers to treat that fear as something philosophy can help with. He discusses the argument that where death is, we are not, so death cannot be bad for the one who dies, and the symmetry argument, best known from the Roman Epicurean poet Lucretius: we feel no distress about the endless time before we were born, so why feel it about the endless time after we die?

Yalom’s interest is practical rather than theological. He is less concerned with whether the arguments are logically airtight than with how they can loosen the grip of what he calls death terror. That is a useful contrast with the paradox. The paradox concerns gods and evil, and its link to Epicurus is doubtful; the arguments about death are securely Epicurus’s own and, for a modern reader, arguably his more lasting gift.

Reading the Epicurean paradox in the original sources

Short texts, all available in public-domain translations
  • Lactantius, On the Anger of God, chapter 13: the earliest surviving version, with the author’s reply. Short, and readable in older English translations.
  • Sextus Empiricus, Outlines of Pyrrhonism, Book 3: the sceptical argument against providence that many scholars see as a cousin of the paradox.
  • Epicurus, Letter to Menoeceus and Principal Doctrines: what Epicurus actually said about the gods and death, in a few pages.
  • Hume, Dialogues Concerning Natural Religion, Parts X and XI: the classic modern discussion of evil, in dialogue form.
  • Mackie, “Evil and Omnipotence” (1955), and Plantinga, God, Freedom, and Evil (1974): the two texts that framed the twentieth-century debate.

For the argument in fiction, two novels handle it better than most essays. In Dostoevsky’s The Brothers Karamazov, Ivan refuses to accept any divine harmony bought with the suffering of children; our guide on where to start with Dostoevsky explains when to read it. In Camus’s The Plague, the doctor Rieux and the priest Paneloux watch a child die and draw opposite conclusions; see our guide to Camus’s books in order for where it sits in his work.

Frequently Asked Questions
What is the Epicurean paradox in simple terms?

It asks why evil exists if a god is both able and willing to prevent it. If he is willing but unable, he is weak; if able but unwilling, he is not good; if neither, he is not a god; if both, evil should not exist.

Did Epicurus really say it?

Probably not in this form. No surviving work of Epicurus contains it. The earliest source is the Christian writer Lactantius in On the Anger of God, chapter 13, and many scholars suspect a sceptical origin.

Who wrote the famous English version?

David Hume, through the character Philo in Part X of Dialogues Concerning Natural Religion, published in 1779. Philo introduces the lines as Epicurus’s old questions that are yet unanswered.

Does the Epicurean paradox prove that God does not exist?

No. It targets a god who is all-powerful, all-knowing and wholly good at once. Believers can reply by giving reasons a good god might allow evil, and the argument does not touch a limited or uninvolved god.

How is it different from the problem of evil?

It is an early form of the problem of evil. Modern philosophers split that problem into a logical version, made famous by J. L. Mackie in 1955, and an evidential version, developed by William Rowe in 1979.

Is the Epicurean paradox a real paradox?

Not in the strict sense. It is a dilemma or trilemma: it lists options and argues that each is unacceptable. A strict paradox derives a contradiction from premises that all seem true.

Conclusion

The Epicurean paradox is older, and less Epicurean, than its name

The Epicurean paradox is a sharp four-way dilemma about evil and a perfect god, but its link to Epicurus rests on a hostile Christian source and a convention that Hume repeated. It is not a disproof of every god, and it is not a strict paradox. What it did was frame a question that philosophers still argue about: whether God and evil are consistent, and whether the evil we see makes God unlikely. Read Lactantius and Hume for the argument itself, Mackie, Plantinga and Rowe for the modern debate, and Epicurus’s own letters for a philosophy more concerned with calm than with refuting the gods.