Argument from Reason

Rational inference requires that one belief be held because another is seen to be true — a relation of logical ground and consequent. Physical causation is a relation of prior state and subsequent state governed by law, indifferent to logical validity. If mental events are exhaustively describable in the second vocabulary, the argument runs, the first has no work to do, and we do not actually reason at all; we merely undergo neural transitions that happen to correlate with valid patterns.

Lewis's formulation in the revised third chapter of Miracles is the well-known statement, written after Elizabeth Anscombe's 1948 critique forced him to sharpen it. That exchange is worth knowing: Anscombe distinguished 'irrational' causes from 'non-rational' ones and argued Lewis had conflated them, and Lewis's revision addressed the point rather than ignoring it.

Victor Reppert's C. S. Lewis's Dangerous Idea breaks the argument into several sub-arguments, which is the most useful way to engage it. The argument from intentionality asks how physical states come to be about anything. The argument from truth asks how physical states can be true or false. The argument from mental causation asks how the logical content of a thought, as opposed to its physical properties, can cause anything. The argument from the psychological relevance of logical laws asks how abstract, necessary laws bear on concrete brain events. The argument from the unity of consciousness asks what binds a multi-premise inference into a single act of reasoning.

Separating these matters because they have different strengths. Intentionality is a recognized hard problem in philosophy of mind independent of any theological interest — Fodor called it the deepest problem in the field. The mental causation argument connects to the general worry about causal exclusion, pressed by Jaegwon Kim, who is a physicalist and treats it as a serious internal difficulty for physicalism.

Plantinga's EAAN is distinct and probabilistic. Natural selection favors adaptive behavior, not true belief. Behavior is caused by neurophysiological properties; on materialism the semantic content of belief is either identical to or supervenient on those properties, and content per se does no adaptive work. So the probability that our cognitive faculties are reliable, given naturalism and evolution, is low or inscrutable.

If that is right, the naturalist who accepts it has a defeater for all their beliefs including naturalism itself, which Plantinga presents as a self-defeat rather than a demonstration that naturalism is false.

The standard reply is that true beliefs are adaptive because acting on false beliefs about predators and food generally kills you. Plantinga's counter is that selection sees behavior, and various content-plus-desire pairings can generate identical behavior — the tiger-avoider who believes the tiger is a large friendly cat best greeted by fleeing behaves adaptively while believing falsely. Critics such as Elliott Sober and Michael Ruse argue that this multiplies improbable pairings and that systematic false-belief systems would be selected against; the exchange is technical and unresolved.

Non-reductive physicalism holds that mental properties supervene on physical ones without being reducible, so reasoning is real at its own level of description, just as software is real without being an extra substance. The pressure point is Kim's causal exclusion argument: if the physical is causally closed, the supervening mental property appears to do no causal work of its own.

Functionalism holds that mental states are defined by their causal roles, so a state's being a belief that P consists in its playing the right role, and logical relations can be implemented in physical systems just as they are in computers. The theist's reply is that computers do not reason; they are interpreted by beings who do, and derived intentionality presupposes original intentionality somewhere.

Eliminativism, associated with Paul and Patricia Churchland, bites the bullet and denies that propositional attitudes exist at all. Critics on both sides regard this as self-refuting, since to assert eliminativism is to express a belief.

The argument does not establish Christianity, or even theism directly. It establishes at most that rational inference sits awkwardly within strict naturalism, and that a worldview positing a rational source of mind accommodates it more comfortably.

It should also be used carefully. The argument does not claim that naturalists cannot reason well — they plainly do, and saying otherwise is both false and insulting. The claim is about whether their metaphysics can account for the reasoning they in fact perform. Keeping that distinction visible is the difference between a serious argument and an insult dressed as one.