Faith and Reason

Suppose a believer is shown, to their own satisfaction, that every argument for God's existence fails.

There are two things they might say next. The first is that nothing has changed, because their belief was never resting on those arguments and was never the kind of thing that rests on arguments. The second is that something serious has been lost and has to be answered. Both replies are given by careful people who have thought about it, and the difference between them is a difference about what faith is supposed to be doing.

What the word covers

The definition is deliberately disjunctive because the English word covers several attitudes that come apart, and most quarrels about whether faith is rational are quarrels about which one is in view.

On the propositional model, faith is believing certain claims. Someone has faith when they hold that God exists, that God is good, and whatever else their tradition asserts. This is the model that makes faith straightforwardly comparable with other beliefs, and therefore straightforwardly assessable by whatever standards apply to belief.

On the trust model, faith is a relation to a person rather than to a set of claims. It is what you have in a friend, and it involves reliance and commitment rather than mere agreement that certain things are so. Believing that the friend exists is presupposed by trusting them, but it is not what the trust consists in.

On the venture model, faith is a commitment made where the evidence does not settle the matter, and the venturing is essential rather than regrettable. Kierkegaard's formulation in the Concluding Unscientific Postscript of 1846 is the standard reference point: faith is an objective uncertainty held onto with a passion that is not proportioned to the evidence, and the absence of objective settlement is part of what makes it faith rather than knowledge.

On a fourth family of models, faith does not require belief at all. A person can be committed to living as though something is true, and can order their life around it, without believing it in the sense that involves thinking it more likely than not. Whether that deserves the name is disputed, and the motivation for it is that many people who describe themselves as faithful report a good deal less confidence than the propositional model attributes to them.

Aquinas's account is worth having in outline because it sits between several of these. He treated faith as midway between knowledge and opinion: it has the firmness of knowledge, since the believer does not hold the articles tentatively, without having the seeing that knowledge involves. What the believer assents to are propositions, and the reason for assenting is that they have been revealed by God. So faith on this model is a species of belief on testimony, with the special feature that the testifier cannot be mistaken or deceptive.

That model has a difficulty which Descartes pressed and which has not gone away. The propositions are accepted on the authority of the revealer. That there is such a revealer, and that these are the revelations, is not itself something the believer sees. If it is accepted by faith, the account is circular. If it is established by argument, then faith rests on natural theology after all, and the last two sections of this article become unavoidable.

Belief and evidence

Hume gave the moderate statement of the principle: the wise person proportions belief to the evidence. W. K. Clifford gave the severe one, in "The Ethics of Belief" in 1877, holding that believing anything on insufficient evidence is wrong always, everywhere, and for anyone. Clifford's formulation is a moral claim rather than merely an epistemic one, and it is stated with no exceptions.

His argument for it is better than it is usually given credit for. He begins with a shipowner who convinces himself, without checking, that his emigrant ship is seaworthy, and whose sincerity does not make him less responsible when it sinks. The general lesson Clifford draws is that beliefs are not private, because they issue in action and are passed to others, so that a habit of believing without checking damages more than the believer.

Three difficulties attach to the strong version.

It looks self-undermining. The principle itself is a belief, and the evidence supporting it is not obviously sufficient by its own standard, since it is a normative claim rather than an empirical finding.

It threatens to condemn nearly everything anyone believes. Reliance on memory, on the existence of other minds, on the general reliability of perception, and on the future resembling the past are all held without independent evidential support, and attempts to supply it are notoriously circular. Either the principle is false, or evidence has to be read widely enough to include whatever supports those, in which case the interesting question becomes whether religious belief also qualifies.

And stated as a moral duty it presupposes that belief is under voluntary control, which is disputed. A person who cannot help finding a claim compelling has not obviously done anything they can be blamed for.

None of this shows that religious belief meets the standard. It shows that the standard has to be stated carefully, and that stating it carefully is what turns this into a live argument rather than a slogan. The connection to the question of who is obliged to argue is direct: evidentialism plus a failure of the theistic arguments gives a conclusion about what a reasonable person should believe, and any position that denies evidentialism is denying that inference.

Faith not answerable to reason

Three things should be said about this term before anything else, because it is used carelessly more often than not.

Almost nobody accepts the label for themselves. It functions overwhelmingly as an accusation, and the people it is applied to generally deny that they hold what it attributes to them.

The standard attributions are contested by scholars of the figures involved. Tertullian is the stock example of the fideist who embraced belief because it is absurd, and current scholarship reads him as attacking the presumption of philosophers rather than the use of reason. Kierkegaard is the second stock example, and the more careful reading is that faith for him goes beyond what philosophical reason can deliver without being contrary to reason. The set of theses grouped as Wittgensteinian fideism, on which religion is a self-contained practice immune to outside criticism, is not well supported by what Wittgenstein wrote, and the philosopher most associated with the label denied holding it.

And the position that survives the scholarship is more modest than the caricature. Its claim is not that reason should be abandoned. It is that a particular account of reason, namely the evidentialist one, is too narrow, and that it can be reasonable to decline to submit certain commitments to it.

Set the caricature aside and the genuine varieties are these. One holds that reason has limits, that religious questions lie past them, and that faith operates in the space that remains. A second holds that religious language and practice have standards of their own, in the way that other complex human practices do, and that judging them by standards imported from natural science misdescribes what is being judged. A third, more radical, holds that faith should not seek support even where support is available, because a faith resting on argument is resting on something that could be taken away.

The objections are serious.

The first is that the exemption generalises. If a commitment can be held without reasons because it is held with sufficient depth, or because it belongs to a practice with internal standards, then the same is available to every incompatible religion and to positions nobody would defend. A licence with no restrictions certifies nothing.

The second is the charge of self-defeat. Fideism is argued for. The reasons given for thinking that reason has limits, or that evidentialism is too narrow, are reasons, and offering them is an appeal to the tribunal being dismissed. The reply available here is that the argument is internal, showing that evidentialism fails by its own standards, which is not the same as claiming exemption from reason.

Two sources, two projects

The division is a division of sources rather than of subject matter. Both are about God, and the difference is what each is permitted to use as a premise.

On the classical arrangement, associated with Aquinas, they divide the ground between them. Some truths about God can be established by reason: that God exists, that God is one, that God is not a body. Others exceed what reason can reach and are known only because they have been revealed, and the standard examples in the Christian tradition are the Trinity, the Incarnation, and the world's having had a beginning in time. Reason cannot demonstrate the second class and is not idle with respect to it, since it can show that the alleged demonstrations of their impossibility fail.

That arrangement gives natural theology a specific job: not to produce the whole of religion, but to establish enough that accepting a revelation is not arbitrary. Everything in the intermediate tier of this subject is natural theology in this sense, and the arrangement explains why. Arguments from the existence of the universe, from its order, from the concept of God, and from morality all proceed from premises available to anyone.

The objection to natural theology that carries most weight is not that it fails. It is theological, and it holds that the project is improper.

The charge is that natural theology puts God in the position of an object to be investigated by human beings on their own initiative and by their own standards, which inverts the relation it is supposed to be describing. On this view the God arrived at by argument is a philosophers' construct rather than the God of the tradition, and a faith resting on such arguments rests on human competence rather than on what has been given. Karl Barth pressed this in the twentieth century with unusual force, replying to a fellow theologian's modest case for a natural point of contact with a pamphlet titled simply Nein!, and the position has a long history in the Reformed tradition before him.

The reply available to natural theology is that the objection assumes what it needs to show. If reason is among the things God gave, using it on the question is not an intrusion, and the alternative leaves the believer unable to say anything at all to someone outside the tradition. Which side is right on this is one of the deepest divisions in the subject, and it is a division within religion rather than between religion and its critics.

Next

This article has left the central question open on purpose, since answering it belongs to the tiers that assess the arguments. What can be settled now is what those arguments are trying to do. The last article in this tier asks what form a case for or against God's existence can take, why hardly anyone now claims to have a proof, and what it means to say that the evidence makes a hypothesis likelier without establishing it.