QuestionsDoes God exist?The best of all possible worlds
Does God exist?TheismThe problem of evil

The best of all possible worlds

Gottfried Wilhelm Leibniz 1646 · 1716We live in the best of all possible worlds.
See profile
What it claims

Our world is the best possible, and it contains evil: evil is not proof that God could have done better.

Why

A perfectly wise and good God sees all possible worlds and cannot but choose the best one.

In short

"The best of all possible worlds" is an argument by Gottfried Wilhelm Leibniz (Theodicy, 7-10, 21; Summary, objection 1) on the question "Does God exist?" (answer: theism). Its claim: Our world is the best possible, and it contains evil: evil is not proof that God could have done better. According to AskPhilosophy's logic check (Z3 solver), the argument is valid only if one grants a premise the author leaves unstated (P3). AskPhilosophy presents 5 objections, 5 of them with a documented reply, and the original text behind each premise, checked word for word.

Theodicy, 7-10, 21; Summary, objection 1
Valid with an implicit premise

The most famous answer to the problem of evil. If God is perfect, he chose the best possible world. So the evil we see does not prove that God could have done better: even the best of worlds contains some.

P1God exists and is perfect in power, wisdom and goodness (Leibniz believes he has established this, notably with the argument from contingency).
P2A perfectly wise and good being, who sees all possible worlds, cannot fail to choose the best.
P3 · implicit premiseAmong all possible worlds there is one that is better than all the others: no perfect tie, and no endless series of ever better worlds.
P4There is evil in the world God made.
Therefore our world is the best possible, and it contains evil: evil is not proof that God could have done better.

click a premise to reject it

If you accept every premise, logic forces you to accept the conclusion. So the debate is about the premises.
Checked by a logic solver: the premises cannot all be true while the conclusion is false.
Without the implicit premise: Without P3, the computer finds a situation where God is perfect but no world is better than all the others (for instance an endless ladder). Nothing then guarantees that our world is the best.
Valid with P3. Leibniz partly defends it (without a best world, he says, God would have created none), but mostly assumes it. Above all, the argument starts from God to judge the world, not from the world to prove God: it shows that evil does not contradict God's existence, provided God has already been proved some other way. Leibniz accepts this: we must judge "ab effectu", from the choice God made.
Analogy. A general prefers a great victory with a slight wound to a situation with neither wound nor victory. The wound is a real evil, but the whole is better with it. This is Leibniz's example: we do not judge a battle, or a world, by a single one of its details.

Objections

5

Each objection is a conversation: click to see who replies and where the debate stands.

Objection anticipated by Leibniz (Bayle) attacks the conclusion

A world without any evil was possible, and God could even have created nothing. Whoever makes a work containing evil, when he could have avoided it, does not choose the best.

Leibniz replies
Leibniz
The best choice is not always the one that avoids evil: an evil may be bound up with a greater good. Like the general who prefers a victory with a wound to no victory, a world with evil can be better, overall, than a world without evil.
Where the debate stands (our summary)The core disagreement is what a perfect choice requires: avoiding all evil, or aiming at the best whole. Leibniz judges the world as a whole, not evil by evil. Whether one can claim that an evil is bound up with a greater good without seeing it is an open point, taken up in the threads on Hume and Voltaire.
Objection anticipated by Leibniz attacks premise 3

There is no best possible world, just as there is no largest number: one can always imagine a more perfect creature, hence a better world.

Leibniz replies
Leibniz
What holds for a particular creature, which can always be surpassed, does not hold for the whole universe, which extends through all eternity and forms an infinite.
Robert M. Adams 1972
Adams offers another way out, which abandons Leibniz's: a perfectly good God is not obliged to create the best possible world. He can create a good world and love his creatures without wronging them or lacking goodness, whether or not a better world exists. The problem of an ever surpassable world then disappears, but so does the principle of the best on which Leibniz's argument depends (P2).
Where the debate stands (our summary)Leibniz answers that the whole universe, being infinite, can be the best even if each creature can be surpassed. Adams offers another way out: a good God is not obliged to create the best. The core disagreement: does perfect goodness require the optimum, or only a good world? An open point.
Hume (Philo, in the Dialogues) 1779 · attacks premise 1

I am willing to grant that evil is compatible with a perfect God. But starting from the world as it is, a mixture of good and evil, one can never conclude to a perfectly good God. It is not the consistency that is rejected, but the inference.

a defender replies
No reply possible: objection raised in 1779, after the author's death (1716).
William Paley defender of the argument · 1802
Paley argues that a benevolent Creator can be inferred from the world as it is. In the vast majority of natural contrivances, he says, the visible purpose is the well-being of living things, and pain is a side effect rather than the purpose of a contrivance. The mixture of good and evil therefore does not, in his view, tell against the Creator's goodness (Natural Theology, chapter 26).
Charles Darwin 1860
Darwin writes to Asa Gray (22 May 1860) that he cannot see, as others do, design and beneficence on all sides: there seems to him too much misery in the world. He cannot persuade himself that a beneficent and omnipotent God would have created the ichneumon wasps on purpose to feed within the bodies of living caterpillars.
Where the debate stands (our summary)Hume grants that evil is compatible with a perfect God: he refuses that this perfection can be inferred from the world. Paley thinks the inference possible; Darwin does not see in nature the benevolence Paley reads there. The core disagreement: what does a world mixed of good and evil show about the character of its author? Leibniz, for his part, establishes God by other arguments before speaking of evil.
Voltaire (Lisbon earthquake, Candide) 1759 · attacks the conclusion

The Lisbon earthquake (1755) killed tens of thousands. Telling the victims that all this serves the best of worlds explains nothing and makes us callous. In Candide, the philosopher Pangloss keeps repeating that "all is for the best" amid disasters.

a defender replies
summary: their text is not yet in our corpus
No reply possible: objection raised in 1759, after the author's death (1716).
Jean-Jacques Rousseau defender of the argument · 1756
Rousseau answers Voltaire that one should say "the whole is good", not "all is good" for each. He adds that most of our physical ills are our own work: nature had not gathered in Lisbon twenty thousand houses of six or seven storeys.
see the argument
Where the debate stands (our summary)Voltaire refuses to say "all is well" without being able to show it; Rousseau distinguishes the whole from the parts and recalls the human share of evils. They are not quite speaking of the same thing: the victims' suffering on one side, the good of the whole on the other. Rousseau's supporters rely on the human share of evils, his critics object that it does not cover natural evils (see Hume's argument from evil).
D
Dostoevsky (Ivan Karamazov) 1880 · attacks premise 2

Even if an innocent child's suffering served the harmony of the whole, a good being should not accept that price. A choice that makes some pay for the best of the whole is not the most good choice (The Brothers Karamazov).

Leibniz replies
summary: their text is not yet in our corpus
No reply possible: objection raised in 1880, after the author's death (1716).
Marilyn McCord Adams 1989
Adams, a believer, does not defend Leibniz here: she accepts Ivan's demand, and holds that a good of the whole is not enough to justify one person's suffering. She offers another answer: God must be good to each person, by ensuring that even the most horrendous evils are overcome within that person's own life, through an intimate relationship with him, so that this life is on the whole a great good for her (Horrendous Evils and the Goodness of God, 1989).
Where the debate stands (our summary)Ivan refuses a good of the whole paid for by an innocent. Even among believers, this objection has led philosophers to leave the best-whole answer behind: Adams asks that God be good to each person, not only to the whole. This answer rests on a life beyond death, which its opponents refuse to grant without proof. An open point: no widely accepted answer.