PhilosophersGottfried Wilhelm Leibniz

Gottfried Wilhelm Leibniz

We live in the best of all possible worlds.

Rationalism1646 – 1716 GermanyRationalistBelieverMathematicianMetaphysical optimist

Overview Arguments · 6 Works Quotes Lineage

Monadology, 32-38; Principles of Nature and Grace, 7-8

Sufficient reason: why something rather than nothing?

What it claims

There exists a necessary being, the ultimate reason of all things: God.

Why

Everything has a reason, and the whole series of things that might not have been cannot hold its own reason: it lies outside the series.

Valid with an implicit premise

If nothing happens without a reason, then the whole world must have a reason for existing. Leibniz argues that it can never be found within the world itself, even going back to infinity. Unlike Aquinas's third way, which starts from things that come to be and pass away, he starts from a general principle, sufficient reason, and applies it to the whole world: even an eternal world, he says, would need a reason. This is the argument "from contingency", often considered the strongest version of the cosmological proof.

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 world where the principle holds for each thing and each state is explained by the previous one, but the whole series has no reason: it is simply there. No necessary being is then required.

Valid with P3. Leibniz asserts it (he seeks "the reason for the existence of the world, which is the whole assemblage of contingent things") without defending it separately: he moves from "each thing has a reason" to "the collection of all things has a reason". Without it, nothing forces us to look for a reason for the whole, and the conclusion no longer follows. The debate mainly concerns P1 (is the principle true without exception?), P3 (is the whole one more thing?) and P5 (must the necessary being be God?).

Analogy. A geometry book was copied from another, itself copied from another, and so on forever. Each copy is explained by the previous one. But why are there such books at all, and why written this way? Going back up the chain, even to infinity, does not answer the question. This is Leibniz's own example.

Objections

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

Objection anticipated by Leibniz attacks premise 4

If the world is eternal, each state is explained by the previous state, without end. So everything is explained, and there is no need to go outside the world.

Leibniz replies

Leibniz

Explaining each copy by the previous one never says why there are books rather than nothing, nor why they are written this way. Even an eternal world does not contain the complete reason for its existence.

Where the debate stands (our summary)Leibniz answers that explaining each copy by the previous one does not say why there are books rather than nothing. Those who reject this demand (Hume, Russell, below) deny that the whole must have a reason beyond its parts: the core disagreement is the reach of the principle of sufficient reason.

Spinoza 1677 · attacks premise 2

If everything has a sufficient reason, then nothing could have been otherwise: everything follows necessarily from God's nature. The principle of reason destroys the contingency the argument needs. (Today this problem is called "modal collapse".)

Leibniz replies

Leibniz

Two kinds of necessity must be distinguished. This world is chosen because it is the best: that is a moral necessity, the necessity of the wise. But its contrary implies no contradiction: it is not necessary in the metaphysical sense, so it remains contingent.

Peter van Inwagen 1983

Summary: van Inwagen argues that the principle of sufficient reason, applied to the conjunction of all contingent truths, leads to an impasse: the proposition that would explain that conjunction could be neither contingent (it would then be part of the conjunction and would have to explain itself) nor necessary (whatever it entails would then be necessary, the conjunction included). He concludes that the principle must be rejected.

See the rest of the debate (1)

Alexander Pruss 2006

Summary: Pruss answers that a "sufficient" explanation need not logically entail what it explains, and that the free action of an agent can explain without necessitating. The principle therefore does not imply that everything is necessary.

Where the debate stands (our summary)Leibniz distinguishes moral and metaphysical necessity. Van Inwagen (1983) judges that the principle, taken without exception, still leaves a dilemma; Pruss (2006) replies by appeal to the free action of an agent, which assumes a certain conception of freedom. The core disagreement concerns the meaning of "sufficient" and the nature of freedom.

Hume (Cleanthes, in the Dialogues) 1779 · attacks premise 3

Once the cause of each part has been explained, everything has been explained. Uniting the parts into a "whole" is an arbitrary act of the mind, like uniting provinces into a kingdom: that whole needs no extra cause.

Leibniz replies

No reply possible: objection raised in 1779, after the author's death (1716).

William Rowe 1975

Summary: Rowe, himself a non-believer, nevertheless judges that the objection fails. He argues that when each member of a collection is explained by another member of the same collection, it does not follow that the collection itself is explained. Explaining each dependent being does not explain why there are dependent beings rather than nothing. The Hume-Edwards principle (explaining the parts is explaining the whole) therefore does not tell against this argument. Rowe nevertheless concludes that the argument is not a proof, since we do not know that the principle of sufficient reason is true.

Where the debate stands (our summary)Hume and Russell judge that once the parts are explained, the whole is; Rowe answers that this is not so when the parts explain one another. The core disagreement is the principle of sufficient reason: self-evident for Pruss, a mere postulate for Russell (see Russell's argument).

Kant 1781 · attacks premise 5

Even granting that a necessary being exists, nothing in experience proves that it is God, the perfect being. To take that step, the cosmological proof must assume that supreme perfection implies necessary existence: it is the ontological proof in disguise, and that one fails. (Leibniz himself accepted the ontological proof, provided one first shows that God is possible: Monadology, 45.)

no reply

No reply possible: objection raised in 1781, after the author's death (1716).

Where the debate stands (our summary)Leibniz himself accepted the ontological proof, provided one first shows that God is possible (Monadology, 45): Kant's objection therefore reaches him from a different angle than Aquinas. The reply by a merely "factual" necessity (Reichenbach, 1972), discussed in Aquinas's argument, fits poorly here: Leibniz's necessary being carries the reason for its existence within itself. An open point.

Bertrand Russell 1948 · attacks premise 1

Nothing guarantees that everything has a reason. The universe may be a brute fact: it is just there, and that is all. Demanding an explanation of the whole is a postulate, not something evident (radio debate with Father Copleston, BBC).

a defender replies

summary: their text is not yet in our corpus

No reply possible: objection raised in 1948, after the author's death (1716).

Alexander Pruss defender of the argument · 2006

Summary: Pruss judges the principle of sufficient reason self-evident, and adds a practical argument: declaring a fact "brute" should only be a last resort.

Where the debate stands (our summary)Russell treats the principle as a postulate, Pruss as self-evident. Neither side has a proof accepted by the other: an open point of core disagreement.