Faith and Reason

Even if some argument for God succeeded, a further and in some ways deeper question would remain: what is the proper relation between faith — the characteristic attitude of religious commitment — and reason — the evidence-weighing, argument-following faculty by which we govern belief generally? Must religious belief be based on evidence and argument to be rational, or can faith be legitimate, even admirable, precisely where it outruns the evidence? The dispute is not about whether God exists but about the epistemic standing of believing that he does, and it cuts across the theism/atheism divide: there are evidentialist theists and fideist ones, just as there are evidentialist atheists.

This page sets out what "faith" might mean, the four classic ways of relating it to reason, the evidentialism vs. fideism axis that organizes the modern debate, and the two famous pragmatic attempts — Pascal's Wager and James's "will to believe" — to license belief on non-evidential grounds. It is the gateway to religious epistemology, where Reformed epistemology, religious experience, and miracles develop the themes technically.

What is faith?

"Faith" is not one thing, and several of the field's confusions trace to conflating its senses:

• Faith as belief-that (propositional) — assent to propositions, e.g. that God exists. The question is then whether such assent is rationally held.
• Faith as trust (fiducial, Latin fides) — personal trust in God, on the model of trusting a person, not just believing a proposition. Trust can be reasonable even where proof is unavailable, as trust in a friend is.
• Faith as faithfulness — a practical commitment, a way of living, an orientation of the will, rather than a cognitive state at all.
• Faith as a virtue or gift — in much theology, a disposition not wholly under voluntary control, "infused" rather than reasoned into.

A recurring fault line is doxastic voluntarism: can one choose to believe? If belief is involuntary (we cannot simply decide that the wall is green), then injunctions to "have faith" must target trust, attention, or action rather than belief directly — a point that bears on both Pascal's Wager and James.

Four models of the relationship

The historical positions can be sorted into four stances on how faith and reason relate:

• Conflict / incompatibility. Faith and reason are rivals; where they meet, one must yield. This is the shared premise of two opposite camps — the rationalist critic who concludes that reason discredits faith, and the fideist who concludes that faith must override reason. Tertullian's credo quia absurdum ("I believe because it is absurd") and the verificationist dismissal of religious language are both, in this sense, conflict positions.
• Incompatibilist fideism. Faith is higher than reason and not answerable to it; the attempt to ground faith in argument is a category mistake or even a spiritual failing. Kierkegaard: faith is a passionate "leap," embracing the "objective uncertainty" of the paradox that the eternal entered time; to demand proof is to evade the risk that constitutes faith. Wittgensteinian philosophers of religion (D.Z. Phillips) make a gentler version: religious belief belongs to its own "form of life" with its own grammar, not assessable by external evidential standards.
• Synthesis / harmony. Faith and reason are distinct but cooperative, two complementary routes to one truth. Aquinas's is the paradigm: reason can establish the preambles of faith (that God exists, by the Five Ways) and defend faith against objection, while the mysteries of faith (Trinity, Incarnation) exceed but do not contradict reason. "Grace perfects nature; it does not destroy it." Natural theology is the joint where the two meet.
• Independence / incommensurability. Faith and reason address different domains and cannot conflict because they never compete — the religious counterpart of Gould's NOMA. Faith governs meaning, value, and the existential; reason governs empirical fact. Whether this protects religion or trivialises it (by making it immune to all evidence, and so empty of factual content) is the standing objection.

Evidentialism vs. fideism

The modern debate distils the four models into a single axis: must religious belief be proportioned to evidence?

The evidentialist demand

Locke held that faith is assent on the testimony of God, but that reason must first certify that a revelation is genuinely from God — so faith never licenses believing beyond what reason approves. W.K. Clifford ("The Ethics of Belief," 1877) pressed this into a stark ethical principle:

It is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence.

On this view, faith unsupported by evidence is not merely intellectually defective but morally culpable — a violation of one's duties to the community of inquirers. Applied to religion, the evidentialist makes the arguments for theism decisive: if they fail, belief is unwarranted.

Fideist and externalist replies

Two very different responses resist the demand:

• Fideism rejects the demand itself: religious faith is not the kind of thing that requires or admits evidential support, and Clifford's principle, applied to the existential commitments by which we actually live (in persons, in projects), is unliveable. This is Kierkegaard's and, in a key, James's line.
• Reformed epistemology rejects the demand as self-undermining and too strong: Clifford's own principle is not itself supported by sufficient evidence, and consistently applied it would forbid the perfectly rational beliefs of perception, memory, and other minds, which we hold without argument. If those are rational while properly basic, belief in God may be too — grounded in experience and the sensus divinitatis rather than inferred. This is the most influential contemporary answer and is developed in Reformed epistemology.

The evidentialist's rejoinder is the "great pumpkin" objection: if belief in God can be properly basic without evidence, what stops any belief — in the Great Pumpkin, in astrology — from claiming the same status? Drawing a principled line between properly basic religious belief and basic nonsense is the central task the Reformed epistemologist inherits.

Pragmatic and prudential arguments

A distinct tradition argues that even without sufficient evidence, belief can be rational to adopt on practical grounds — that the rationality of belief is sometimes a matter of prudence or the will, not of evidence.

Pascal's Wager

Pascal grants that reason cannot settle whether God exists ("reason can decide nothing here") and reframes the question as a decision under uncertainty. Treating belief as a wager with the outcomes set out in a decision matrix:

Since any non-zero probability of God multiplied by an infinite payoff dominates any finite alternative, the expected value of belief is infinite and the rational choice is to believe (and, since belief is not directly voluntary, to cultivate it — "take the holy water," follow the practices, and belief will come). The Wager is a landmark in decision theory as much as theology, but faces three standard objections:

• The many-gods objection. The matrix assumes a binary (the Christian God or none); a different deity who punishes belief, or rewards a different faith, scrambles the expected-value calculation, and there are indefinitely many such hypotheses.
• The doxastic-voluntarism objection. One cannot believe at will, so the Wager can at most recommend acting so as to induce belief — and a God who sees the calculating motive might not be impressed.
• The mixed-strategy and infinite-utility objections. With infinite utilities the arithmetic misbehaves: any non-zero probability of belief yields infinite expectation, so even a wavering or token believer "wins," which trivialises the result.

James — the will to believe

William James ("The Will to Believe," 1896) answers Clifford directly. When a choice between hypotheses is living (both are real options for you), forced (not choosing is choosing — suspension has the same practical upshot as denial), and momentous (a unique, weighty, irreversible stake), and when the evidence cannot decide, then we have the right to let "our passional nature" settle the matter. Clifford's rule — suspend belief whenever the evidence is insufficient — is itself a passional choice (the choice to fear error more than to lose truth), and in forced, momentous cases suspension can cost us the truth and the good by withholding the trust that alone makes certain truths accessible (some goods, like friendship, exist only if first trusted). James does not license belief against evidence, only beyond it in the special class of forced, momentous, evidentially undecidable options — of which the religious is the paradigm.

Where this leaves us

The faith–reason question is orthogonal to the existence question: one can be a theist who thinks belief needs the arguments (the evidentialist theist), a theist who thinks it does not (the fideist or Reformed theist), or an atheist on either ground. The modern debate has largely moved past the bare "conflict" picture to a contest between the evidentialist demand (Clifford) and the claim that religious belief can be rational without inferential evidence — whether by a pragmatic license (Pascal, James) or by being properly basic. The latter route hands the question to religious epistemology, where the standing of belief grounded in religious experience and the testimony of miracles is worked out in detail.