Chapter 10 develops the second half of Plantinga’s central thesis: There is superficial concord but deep conflict between science and naturalism. Plantinga begins by noting that naturalists routinely claim science as their ally, crediting Charles Darwin in particular with having displaced theism. Quoting paleontologist Stephen Jay Gould and biologist George Gaylord Simpson, he observes that the naturalist case against theism depends not on quantum mechanics or chemistry but on the claim that evolution is unguided. He argues that this claim confuses the scientific theory of evolution with a philosophical add-on. Plantinga asserts that nothing in evolutionary science itself rules out the possibility that God guided the evolutionary process, perhaps by acting at the quantum level to bring about specific mutations or preserve certain populations. The unguided-evolution claim supports naturalism but is not itself science, and conflating the two, he argues, is a mistake.

Plantinga then characterizes naturalism as a “quasi-religion” because it addresses the ultimate questions identified by philosopher Immanuel Kant: Is there a God? Do humans have freedom? Is there life after death? Naturalism answers all three negatively.

The chapter’s main argument focuses on the reliability of human cognitive faculties, including memory, perception, a priori intuition, and induction. From a theistic standpoint, drawing on Thomas Aquinas, these faculties are reliable because God created humans in his image, including the capacity for knowledge. Plantinga argues that a naturalist cannot rationally hold the same confidence. He introduces the concept of “Darwin’s Doubt,” a concern that Charles Darwin expressed in an 1881 letter, and that philosophers Friedrich Nietzsche, Thomas Nagel, Barry Stroud, and Patricia Churchland have echoed: if human cognition is the product of unguided natural selection, there is no particular reason to trust it as a truth-producer. This led to Churchland’s theory that the purpose of the nervous system is to help an organism feed, flee, fight, and reproduce. Natural selection, therefore, rewards adaptive behavior, not accurate belief. 

Plantinga argues that if, as Churchland asserts, evolution only cares about survival-producing behavior, there is no strong reason to expect our beliefs to be mostly true. On page 50, he expresses this in the form of the following equation, where P=Probability, N=Naturalism is true, E=Evolution is true, and R=Our cognitive faculties are reliable: 

(1) P(R/N&E) is low. 

In other words, the probability that our cognitive faculties are reliable if both naturalism and evolution are true is low. 

To argue for this premise, Plantinga incorporates materialism (the view that human beings are purely physical) into naturalism, since almost all naturalists hold it. A belief, under materialism, is a neural structure with two distinct kinds of properties: Neurophysiological (NP) properties governing its causal effects on behavior, and content properties—the propositions the belief is “about.” Materialists disagree on how these are related: Reductive materialists (such as Francis Crick) hold that content properties are simply complex NP properties; nonreductive materialists hold that content is distinct but arises from, or is determined by, NP properties.

Under either view, Plantinga argues, natural selection acts on NP properties because they produce adaptive behavior. The content those NP properties happen to generate is, from evolution’s perspective, incidental—selection is indifferent to truth. He addresses the objection that animals need accurate beliefs to survive (as a frog must track a fly’s position), responding that survival requires reliable neural indicators—structures correlated with environmental features—but indicators need not involve belief at all. Bacteria navigate using internal magnets without forming beliefs about magnetic north. Indication and belief content are separate matters, and nothing in the evolutionary process guarantees that adaptive NP properties produce true content. The probability that any given belief’s content is true, Plantinga concludes, is roughly 0.5—no better than chance.

With 1,000 independent beliefs each at 0.5 probability of being true, the likelihood that three-quarters or more are true (a minimal standard for reliability) is astronomically small—less than 10⁻⁵⁸. He defends his use of the Principle of Indifference (that equal probability should be assigned to equally supported alternatives) by noting it is routinely employed in science, including statistical mechanics.

This establishes the argument’s remaining premises: 

(2) Anyone who accepts N&E and recognizes that P(R/N&E) is low thereby acquires a defeater—a belief that makes it irrational to continue holding a prior belief—for R. 

(3) A defeater for R is a defeater for every belief produced by one’s cognitive faculties, including N&E itself. 

(4) If accepting N&E gives one a defeater for N&E, then N&E is self-defeating and cannot be rationally accepted.

Conclusion: N&E cannot be rationally accepted.

Plantinga argues that this defeater cannot be neutralized by, for example, a scientific test of one’s cognitive faculties, because any such test would circularly presuppose the very reliability under question. He also addresses the “conditionalization problem”—whether the naturalist might hold some additional belief X that, combined with N&E, restores high probability for R—and argues that no admissible candidate exists; R itself cannot function as its own defeater-deflector. Plantinga closes by offering a variant of the argument, focused specifically on the faculty producing metaphysical beliefs (such as naturalism itself), since such beliefs have no obvious connection to reproductive fitness. 

The chapter concludes by restating the book’s overarching thesis: There is “superficial conflict but deep concord between science and theistic belief,” and “superficial concord [but] deep conflict between science and naturalism” (349). The genuine conflict, Plantinga asserts, lies not between science and theistic religion but between science and the “quasi-religion” of naturalism.