What Is the Argument from Reason?

Direct Answer

The argument from reason is a family of arguments aiming to show that materialism — the view that everything is physical and fully explicable by physical laws — cannot account for the phenomenon of reasoning itself, and that if materialism were true, no belief could be genuinely rational, including the belief in materialism. The argument has been given its classic modern form by C. S. Lewis in Miracles (1947) and defended by philosophers such as Victor Reppert and Alvin Plantinga.

The core idea is this: when you reason from premises to a conclusion, the sequence of your thoughts is supposed to be governed by logical relations. The conclusion follows because the premises entail it; you believe the conclusion because you see that it follows. Reason is the grasp of these logical relations — entailment, consistency, probability — and it is these relations that justify beliefs.

Materialism, the argument continues, explains our beliefs as the causal products of physical processes: neural events cause other neural events, in accordance with the laws of physics. But physical causation is not logical inference. A physical state is caused by its physical antecedents, not by the truth of a proposition or the validity of an argument. If all your beliefs are fully caused by non-rational physical processes, then the fact that a belief is true, or that it follows from the evidence, is irrelevant to its occurrence: you would believe whatever the physics makes you believe. But a belief that is produced by non-rational causes, without regard to truth, is not a rational belief — and if no belief is rational, then materialism itself cannot be rationally believed. The argument from reason concludes that materialism is self-undermining, and that reason must have a source that is not purely physical.

The argument targets not only crude behaviorism but any view — including many forms of functionalism and physicalism — that treats belief formation as a purely causal process with no place for normative, logical justification.

Historical Context

The argument from reason has deep roots. Plato's account of knowledge, Aristotle's logic, and the rationalist tradition (Descartes, Leibniz, Spinoza) all insisted that reason is a distinctive faculty that grasps necessary, non-empirical truths. The problem of how a physical world could give rise to such a faculty was raised sharply by Kant, who gave reason a transcendental structure. In the twentieth century, the argument took on its modern anti-materialist form: if the mind is a physical system, can it really grasp logical relations that are not physical causes?

C. S. Lewis popularized the argument in Chapter 3 of Miracles (1947). Lewis's version aimed at naturalism: naturalism explains all our beliefs as the products of non-rational natural causes; but this explanation applies equally to the beliefs of the naturalist, so the naturalist has no rational basis for believing naturalism. The argument was revived and given rigorous treatment by Victor Reppert (C. S. Lewis's Dangerous Idea, 2003) and Alvin Plantinga, whose "evolutionary argument against naturalism" extends the same structure: if our cognitive faculties were produced by unguided processes with no regard for truth, the probability that they are reliable is low, and naturalism is self-undermining. The mathematician and physicist Roger Penrose has pressed a related argument in The Emperor's New Mind (1989): if mathematical truth is grasped by Gödelian insight, and insight is not a computational process, then the mind cannot be a purely computational or physical system.

Key Arguments & Debates

The argument has several forms. The logical form: reason involves grasping logical relations, and logical relations are not physical causal relations; therefore the capacity to grasp them cannot be a physical capacity. The evolutionary form (Plantinga): if our beliefs are produced by natural selection for survival, with no regard for truth, then the reliability of our cognitive faculties is low; but we rely on those faculties to assess naturalism; therefore naturalism is self-undermining. The free-will form: reasoning requires that we can follow the argument because it is valid, not merely because we are caused to believe — a capacity that physical determinism seems to exclude.

The responses are extensive. The first line of defense is to deny the premise that physical processes are non-rational: a materialist can argue that physical processes implement rationality — that the brain's causal organization realizes the logical structure of reasoning, so that causes and reasons coincide. The second response is to deny that rationality requires a special non-physical cause: a belief is rational if it is formed by reliable processes that track truth, whether those processes are physical or not (this is the reliabilist and naturalized-epistemology response, associated with W. V. O. Quine and Alvin Goldman). The third response is to accept the argument's conclusion about strong rationality but deny that this refutes materialism: perhaps rationality, like everything else, is a natural phenomenon that physical science can explain, and the intuition that it cannot is a residue of dualism. The fourth response (Dennett's, in Freedom Evolves and elsewhere) is that reason and computation can be understood naturalistically: our brains are the result of a long process in which "good tricks" for truth-tracking were selected, and the capacity to reason is a physical capacity, full stop.

The debate is thus another form of the mind-body problem: it asks whether the normative dimension of thought — the difference between following a reason and being caused — can be naturalized.

Contemporary Relevance

The argument from reason remains relevant in debates about AI and rationality. If reasoning is a matter of grasping logical relations, can a large language model — which produces text by predicting tokens, not by grasping entailment — genuinely reason? The argument's structure suggests that a system whose outputs are purely causally determined by training statistics may produce correct-looking answers without engaging in reasoning, and that genuine rationality may require more. The argument also bears on the epistemology of AI: if we cannot trust our own reasoning under materialism, we can hardly trust a machine's. In the philosophy of mind, the argument from reason is a standing challenge to reductive physicalism, a contribution to the free will debate, and a reminder that any complete account of mind must explain the normativity of thought — why beliefs can be justified or unjustified, and what makes reasoning reasoning.

Related Concepts

• Free will: the related capacity that causal accounts seem to threaten.
• The Libet experiment: the empirical debate about conscious causation.
• Intentionality: the aboutness that makes beliefs truth-evaluable.
• The explanatory gap: the gap between physical explanation and normative thought.
• Critical thinking and logic: the practice that presupposes the reality of reason.

Further Learning

• Internet Encyclopedia of Philosophy: C. S. Lewis and the Argument from Reason
• C. S. Lewis, Miracles (1947), Chapter 3.
• Alvin Plantinga, Warrant and Proper Function (1993), chapters 11–12.
• Victor Reppert, C. S. Lewis's Dangerous Idea (2003).