Alvin Plantinga, a prominent philosopher of religion and epistemology, has put forward an argument that challenges the combination of metaphysical naturalism and evolutionary theory-an argument that has generated considerable debate in contemporary philosophy. This is the "Evolutionary Argument Against Naturalism" (EAAN). In its strongest form, the target is not evolution by itself, nor science as such, but the conjunction of naturalism and unguided evolution: if that conjunction gives us a serious reason to doubt the reliability of our own cognitive faculties, then it also gives us a reason to doubt the beliefs produced by those faculties, including naturalism itself.
Plantinga's Argument
The core of Plantinga's argument is usually stated in terms of a probability claim: if naturalism (N) and evolution (E) are both true, then the probability that our cognitive faculties are reliable (R) is either low or inscrutable. Plantinga's point is not merely that evolution values survival rather than truth. It is that natural selection directly selects behavior, not beliefs as true beliefs; and, depending on one's theory of mind, many different combinations of belief and desire could produce the same adaptive behavior. A person might run from a tiger because he truly believes it is dangerous, or because he falsely believes the tiger wants to play but also believes running is the best way to play. Plantinga then argues that someone who accepts N&E and also accepts that P(R|N&E) is low or unknowable acquires a defeater for trusting their cognitive faculties in general. That defeater would include, not merely beliefs about "metaphysics," but also the belief in naturalism itself.
This line of reasoning echoes a concern Charles Darwin expressed in an 1881 letter to William Graham: "With me the horrid doubt always arises whether the convictions of man's mind, which has been developed from the mind of the lower animals, are of any value or at all trustworthy." That quotation is often used in this debate, but it should be handled carefully: Darwin was expressing an epistemic anxiety, not formulating Plantinga's later argument or conceding that evolutionary accounts of mind are self-defeating.
Strongest Replies
There are several significant responses to Plantinga's argument in the academic literature. Serious critics-including philosophers such as Elliott Sober and Branden Fitelson-do not usually claim that evolution makes human cognition infallible or perfectly truth-oriented. Their more careful claim is that Plantinga has not shown that reliability is improbable or inscrutable given naturalism and evolution. Two replies are especially important.
Truth-Tracking is Generally Adaptive
One primary response is that truth-tracking is not an accidental luxury but often part of the best explanation of adaptive behavior. Evolution does not select "truth" in the abstract, but organisms that reliably detect predators, food, spatial relations, other agents' intentions, and cause-and-effect patterns will generally do better than organisms whose internal maps are mostly false. On this view, natural selection gives us reason to expect many domain-specific cognitive systems to be reliable in ordinary environments, even if they are biased, fallible, and uneven. The reply is strongest when it focuses on perception, memory, basic induction, social cognition, and practical reasoning rather than claiming that evolution guarantees reliability in every intellectual domain.
This reply also resists Plantinga's "many belief-desire combinations" move. Critics argue that bizarre false-belief scenarios may be logically possible, but not all logical possibilities are equally probable under a mature account of cognition. If beliefs are integrated with perception, learning, prediction, and action, then widespread falsehood would usually be costly. The naturalist therefore need not prove that all evolved beliefs are true; the more modest burden is to explain why a creature like us would have enough reliable faculties to learn, correct errors, and build cumulative inquiry. That modest claim is plausible, though it does not by itself establish naturalism.
The Argument Proves Too Much
Another important criticism is that the argument risks generating a skepticism broader than Plantinga needs. Plantinga himself is aware of this; indeed, the breadth of the defeater is part of the argument. He wants to say that accepting N&E undercuts trust in the faculties that produce all of one's beliefs, including scientific and philosophical beliefs. The critic's question is whether this move is too sweeping to be dialectically stable. Plantinga uses evolutionary theory only conditionally, for the sake of argument, but if the reasoning undermines the very cognitive practices by which we assess probabilities, evidence, and inference, then it becomes hard to say exactly what remains available to establish the key probability premise.
The stronger version of this objection is not, "skepticism is uncomfortable, so the argument is false." Rather, it asks whether Plantinga can non-question-beggingly justify the crucial claim that P(R|N&E) is low or inscrutable while relying on the same cognitive faculties whose reliability is being challenged. If the argument only shows that human cognition is fallible, the naturalist can accept that. If it shows radical unreliability, it threatens the reasoning used to reach that conclusion as well. The pressure point is therefore not mere discomfort with skepticism, but whether the argument can be made without cutting away its own evidential support.
Measured Conclusion
Plantinga's argument raises important questions about the reliability of human cognition under a naturalistic evolutionary framework. The strongest responses suggest that he has not established the decisive probability claim: that reliable cognition is unlikely or inscrutable given naturalism and evolution. What the debate more clearly supports is a humbler conclusion: our faculties are fallible, shaped by practical pressures, and especially vulnerable outside familiar domains, but this does not by itself make them globally untrustworthy.
The adaptive value of many kinds of truth-tracking provides a plausible naturalistic explanation for why perception, memory, learning, and ordinary reasoning often work well enough to support successful action and inquiry. At the same time, naturalists should concede that evolution does not guarantee reliability in every area. Abstract metaphysics, ultimate explanations, moral theory, and long-range theoretical reasoning may require additional epistemic safeguards: logic, peer criticism, formal methods, empirical testing where possible, and awareness of bias.
Ultimately, Plantinga's argument does not conclusively demonstrate that naturalism and evolution cannot account for reliable cognition, and it certainly does not by itself establish theism. Its value is more diagnostic: it presses naturalists to explain how evolved minds can be trustworthy enough for philosophy, science, and self-critique. A careful naturalist can answer by appealing to the adaptive role of reliable cognition, the corrective power of communal inquiry, and the distinction between fallibility and wholesale defeat. The debate remains open, but the strongest lesson is methodological humility rather than a quick victory for either side.