Who Bears the Burden of Proof for God's Existence?

What Is the Question?

The question of who bears the burden of proof is procedural, but it shapes the whole debate about God. If the theist bears the burden, the theist must produce arguments, and the atheist can simply note that the arguments fall short. If the atheist bears the burden, the atheist must produce arguments against theism, and the theist can rest in the default. The procedural question determines who has to argue first, and that, in turn, can determine the outcome, since the side that does not have to argue wins by default.

The IEP article on atheism treats this as the central procedural dispute in the contemporary debate. Antony Flew's 1950 paper "Theology and Falsification" set the modern terms, and Flew's 1972 essay "The Presumption of Atheism" made the procedural case explicit: atheism, properly understood as negative atheism, is the default stance, and the theist bears the burden of producing sufficient evidence to displace it. The theist has replied, with varying degrees of success, that theistic belief is itself a default stance, grounded in the sensus divinitatis or in the manifest design of the world, and that the atheist's "default" is itself a substantive position that needs defense.

Historical Background

The procedural question is modern. The medieval scholastics did not ask it, because they took the existence of God to be demonstrable, and the dispute was about the quality of the demonstrations, not the assignment of the burden. The question became pressing once the demonstrations were disputed. Hume, in the Dialogues Concerning Natural Religion through the voice of Philo, argued that the theistic arguments, especially the argument from design, fall short of demonstration, and that the proportionate response is suspension of belief. Hume did not frame the point in terms of a "burden," but his stance implies one: the theist must produce the arguments, and if the arguments are weak, the theist loses.

Kant's critical philosophy sharpened the procedural question. In the Critique of Pure Reason, Kant argued that speculative reason cannot settle the question of God's existence, and that the ideas of reason function as regulative ideals rather than objects of knowledge. The Kantian stance places the burden on theoretical reason, which cannot discharge it, and relocates the question to practical reason, where moral faith can be rational without theoretical proof.

The modern debate was set by Clifford's "ethics of belief" and by Flew. Clifford's dictum, "it is wrong always, everywhere, and for anyone, to believe upon insufficient evidence," treats the burden of proof as universal: whoever believes bears the burden of providing sufficient evidence for the belief. Flew applied this to theistic belief specifically and argued, through the parable of the gardener, that theistic belief often becomes unfalsifiable through endless qualifications, and that an unfalsifiable belief has been eroded without contradiction, not defended. The burden, Flew concluded, lies with the theist.

Flew's Parable of the Gardener

Flew's parable is the classic statement. Two explorers come upon a clearing in a jungle. One says a gardener tends it; the other says no gardener tends it. They wait, but no gardener appears. The believer says the gardener is invisible. They set traps and bloodhounds, but no gardener is caught. The believer says the gardener is intangible, silent, and undetectable. The skeptic asks what the difference is between an invisible, intangible, undetectable gardener and no gardener at all. The believer's answer, Flew says, has eroded the original claim until it says nothing.

The parable is an argument about falsifiability and the burden of proof. A claim that cannot be falsified by any possible evidence has been insulated from criticism, but at the price of saying nothing about the world. The theist who responds to every challenge with a qualification that makes God undetectable has not defended God's existence; they have eroded the claim until it is contentless. The burden, Flew concluded, lies with whoever makes the claim, and the theist makes the claim, so the theist bears the burden.

The parable has been criticized. The theist can reply that the gardener analogy is unfair, because God is not posited as a cause among causes, to be detected by traps and bloodhounds, but as the ground of being itself, the source of the existence and order of the world. On this view, God is not detectable in the way a gardener is, but that is not a defect of the claim but a feature of the kind of being God is. The reply pushes the debate back into the metaphysics of God, and into the question of what kind of claim theistic belief is.

The Presumption of Atheism

Flew's 1972 paper, "The Presumption of Atheism," generalized the point. Flew argued that atheism, properly understood as the absence of belief, is the default epistemic position, and that the theist bears the burden of producing sufficient evidence to displace it. The argument parallels the presumption of innocence in law: the accused is presumed innocent until proven guilty, and the believer is presumed not to believe until given sufficient evidence. The "presumption of atheism" is not a claim that God does not exist, but a procedural claim about who must argue first.

The presumption has been criticized from both directions. Some atheists, including the New Atheists, think it is too weak, and that the positive case against theism, drawn from the problem of evil and divine hiddenness, should be made. Some theists, including Alvin Plantinga, think it begs the question: the presumption treats atheism as the default, but that is what is in dispute, and the theist can argue that theistic belief is itself a default stance, produced by the sensus divinitatis and warranted without inferential evidence.

The Theist's Reply

The theist's reply has two strands. The first is the Reformed epistemology line, developed by Plantinga: theistic belief can be properly basic, rational without inferential evidence, if it is produced by a reliable mechanism. The sensus divinitatis is such a mechanism, on the theist's view, and so theistic belief is not a substantive addition that needs evidence but a default output of a properly functioning cognitive system. The burden, on this view, is not on the theist but on whoever would challenge the default. See the companion entry on reformed epistemology and on the rationality of faith.

The second strand is the natural theology line, developed by Aquinas and continued by Richard Swinburne. The theist does not deny the burden but accepts it, and produces the arguments: the cosmological, the teleological, the moral, the ontological, the argument from religious experience. The cumulative case, even if no single argument is conclusive, may be sufficient to discharge the burden and to make theistic belief rational. Swinburne's Bayesian treatment is the most systematic modern version, and it assigns a probability to theism that, he argues, exceeds 0.5 on the cumulative evidence.

The Agnostic's Position

The agnostic takes a distinctive view. The agnostic says the evidence, on both sides, is balanced, and that the honest stance is to suspend judgment. On the procedural question, the agnostic declines to assign the burden to either side, treating the question as genuinely open. The agnostic's position is hardest to defend, because it requires showing that the balance of evidence is genuinely level, but it has the merit of treating the question as open rather than closed in either direction.

The Problem of Divine Hiddenness

A more recent procedural argument comes from the literature on divine hiddenness. J. L. Schellenberg has argued that if a loving God exists, God would ensure that any person capable of relationship with God would be in a position to enter that relationship, and that the existence of reasonable nonbelief, of people who have sought God and not found God, counts against theism. The hiddenness argument is a positive argument against theism, but it also bears on the procedural question: if theism predicts that God would be more plainly visible, and God is not, then the theist has an explanatory burden that the atheist does not.

Contemporary Relevance

The procedural question matters because it shapes the lived debate. The New Atheism, with its popular platform, treated the presumption of atheism as obvious and theistic belief as a prejudice in need of correction. The Reformed epistemology tradition, with its academic platform, treated theistic belief as a default and the atheist's challenge as a substantive burden. The two sides have not met, because they have not agreed on the procedural question that frames the debate.

The deeper lesson is that the procedural question cannot be settled in the abstract. The answer depends on substantive claims about what kind of belief theistic belief is, what kind of being God is, and what rationality itself requires. The debate about the burden of proof is a window into the deeper debates about epistemology and metaphysics, and those debates are not settled. The honest answer, in the contemporary literature, is that the procedural question is itself a substantive philosophical question, and that the work of philosophy is to keep it open rather than to close it by procedural fiat.

Sources

• Internet Encyclopedia of Philosophy, "Atheism" (burden of proof section) — https://iep.utm.edu/atheism/
• Stanford Encyclopedia of Philosophy, "Philosophy of Religion" — https://plato.stanford.edu/entries/philosophy-religion/
• Stanford Encyclopedia of Philosophy, "Religious Epistemology" — https://plato.stanford.edu/entries/religious-epistemology/
• Stanford Encyclopedia of Philosophy, "David Hume" — https://plato.stanford.edu/entries/hume/

Related Topics

• What Is Atheism? — the default stance Flew defends.
• What Is Agnosticism? — the suspension of judgment.
• Is Faith Rational? — the question the procedural debate bears on.
• What Is Reformed Epistemology? — the theist's procedural reply.
• Does God Exist? — the central question.
• Philosophy of God — the wider domain.