Arguments for Theism

The arguments for the existence of God — the enterprise of natural theology, the attempt to establish or support theism by reason from premises not drawn from revelation — form the historical core of the philosophy of religion. They divide into a small number of recurring families, each isolating a different feature of the world (or of the concept of God itself) and arguing that it points to a divine source: the sheer existence of contingent things, their evident order, the bare concept of a perfect being, the authority of morality, the testimony of religious experience. No single argument commands assent, but their cumulative weight is, for many theists, the real case.

This page is the map. It classifies the arguments, states each in capsule form with a pointer to its own page, and explains the two strategic choices that shape how the case is built: whether each argument aims to demonstrate or merely to raise the probability of theism, and whether the arguments are meant to stand alone or to combine into a cumulative case. The companion page sets out the arguments against.

A taxonomy of the arguments

The classical distinction, due to Kant, is between arguments that proceed independently of experience and those that start from it:

• A priori arguments reason from concepts alone, without appeal to any observed feature of the world. The sole major instance is the ontological argument: from the very concept of God as the greatest conceivable (or a maximally perfect, or a necessary) being, that such a being must exist.
• A posteriori arguments start from some observed fact and argue that God is its best or required explanation. The major families:
  • The cosmological argument — from the existence of contingent things, or the fact that anything exists, or that the universe began, to a first cause or necessary being.
  • The teleological / design argument — from the order, complexity, or fine-tuning of the world to an intelligent designer.
  • The moral argument — from the existence of objective moral values or obligations to a divine ground or lawgiver.
  • The argument from religious experience — from the widespread, apparently perceptual experience of God to its veridicality.
  • Further arguments — from consciousness, reason, beauty, desire, miracles, the applicability of mathematics, and common consent.

A cross-cutting distinction concerns the strength claimed:

• Demonstrative arguments purport to be valid deductions from certain premises — proofs, in the strict sense (Anselm's and Aquinas's self-understanding).
• Probabilistic / inductive arguments claim only to make theism more probable than not, or more probable than its denial — they are inferences to the best explanation, weighed like scientific hypotheses (Swinburne).

The modern consensus, even among theists, is that the demonstrative ambition mostly fails — each classical "proof" has a contested premise — and that the live question is the probabilistic one: does the total evidence favour theism?

The arguments in capsule

The ontological argument (a priori)

From the concept of God as that than which nothing greater can be conceived (Anselm), or as a maximally perfect being, to God's existence: a being that exists in reality is greater than one existing only in the understanding, so the greatest conceivable being cannot exist only in the understanding. Plantinga's modal version argues that if a maximally great being is possible, then (in the modal logic S5) it exists necessarily and hence actually. The perennial objection, from Kant, is that existence is not a predicate — not a great-making property that could be packed into a concept. Full treatment: the ontological argument.

The cosmological argument (a posteriori)

From the existence of the contingent world to a being that explains it. Three sub-families: the contingency argument (Leibniz: by the Principle of Sufficient Reason, the totality of contingent things needs an explanation outside itself — a necessary being); the causal/first-cause argument (Aquinas's first three Ways: a regress of caused causes requires an uncaused first cause); and the Kalām argument (al-Ghazālī, revived by Craig: whatever begins to exist has a cause; the universe began to exist; so it has a cause). Objections cluster on causation (Hume, Kant), the fallacy of composition, and "why not stop at the universe itself?". Full treatment: cosmological arguments.

The teleological / design argument (a posteriori)

From observed order to an intelligent orderer. Paley's watch-on-the-heath analogy infers a designer from biological contrivance; Hume had already pressed the decisive objections (weak analogy, the possibility of order without design, the leap to a single, perfect designer). After Darwin removed the need for a biological designer, the argument's centre of gravity shifted to cosmic fine-tuning: the physical constants lie in the extraordinarily narrow range permitting life, which (it is argued) is better explained by design than by chance — to which the standard reply is the multiverse plus the anthropic principle. Full treatment: teleological arguments.

The moral argument (a posteriori)

From the existence of objective moral values and duties to a divine foundation for them. Kant argued that morality requires us to postulate God (and immortality) to make sense of the highest good; modern versions (Adams, Craig) argue that objective moral obligation is best explained as the command or nature of God. The pivotal objection is the Euthyphro dilemma: is the good good because God wills it (then morality is arbitrary) or does God will it because it is good (then morality is independent of God)? Full treatment: the moral argument.

The argument from religious experience (a posteriori)

From the vast body of human experience that seems to be of God — mystical union, the numinous, a sense of presence — to the reality of its object. Swinburne's Principle of Credulity holds that, absent special reason for doubt, experiences should be taken at face value; Alston models the experience of God on ordinary perception. Objections: the conflicting deliverances of different traditions, the availability of naturalistic explanations, and the non-checkability of the alleged perceptions. Full treatment: the argument from religious experience.

Further arguments

Smaller or more specialised: from consciousness and from reason (Plantinga's evolutionary argument against naturalism), from beauty, from desire (Lewis), from miracles, from the unreasonable effectiveness of mathematics, and from common consent. Individually modest, several are offered as strands in the cumulative case. Full treatment: other arguments.

The cumulative-case strategy

No classical argument is a knock-down proof, and critics treat this as fatal. But theists since Newman and, rigorously, Richard Swinburne (The Existence of God, 1979) reply that this misreads the logic. The arguments are evidence, not proofs, and evidence accumulates. The natural framework is Bayesian confirmation theory: each consideration $e_i$ confirms theism $h$ to the extent that it is more probable given God than given no God,

$$P(h\mid e_1\wedge \cdots \wedge e_n)>P(h)\text{whenever each}\frac{P(e_i\mid h)}{P(e_i\mid \neg h)}>1.$$

Several individually weak arguments can jointly raise the posterior probability of theism above $\frac{1}{2}$, just as several weak strands make a strong rope, or several pieces of circumstantial evidence convict. Swinburne argues that the existence of a contingent universe, its order and fine-tuning, consciousness, moral awareness, providence, and religious experience are each more to be expected if there is a God than if there is not, and that their conjunction (with the simplicity of theism as a hypothesis weighing in its favour as a prior) makes God's existence more probable than not.

Two cautions about the strategy:

• Strands can be redundant or even subtractive. Evidence only accumulates if the pieces are independent and each genuinely raises the likelihood ratio; a critic will argue that some "strands" are explained away together (e.g. by evolution and cosmology) or that the problem of evil is a powerful piece of evidence on the other side that the cumulative case must net against.
• The prior matters. The verdict depends heavily on the prior probability assigned to theism and on the simplicity of theism as a hypothesis — both contested. The atheist runs the same Bayesian machinery to the opposite conclusion, treating evil and divine hiddenness as decisive disconfirmation.

Where this leaves us

Natural theology offers not one argument but a portfolio: an a priori route (the ontological argument) and several a posteriori routes (the cosmological, teleological, moral, and experiential arguments), each isolating a different explanandum. The honest modern assessment is that none is a demonstration, and the real debate has shifted to the probabilistic question — whether, on the cumulative balance of all the evidence, theism is the better explanation of the world we find. That balance cannot be struck without putting the other side of the ledger on the scale: the arguments against theism, above all the problem of evil. And even a successful argument would settle only the metaphysical question; whether belief requires such arguments at all is the separate issue of faith and reason.