The Evolutionary Argument Against Naturalism

The Evolutionary Argument Against Naturalism (EAAN) claims that if both naturalism and evolution are true, we have a powerful reason to doubt the reliability of our cognitive faculties - which means we have a reason to doubt naturalism itself. Alvin Plantinga first presented the argument in his 1993 book Warrant and Proper Function and refined it over the following three decades, making it one of the most discussed arguments in recent philosophy of religion. With a soundness score of 20/100, the EAAN is an ingenious self-defeat argument that highlights a genuine epistemological puzzle, but it rests on questionable assumptions about how evolution shapes cognition and ultimately fails to establish its central claim.

The Core Argument

Plantinga’s argument has a precise formal structure:

1. Let N = naturalism (no supernatural beings exist) and E = evolution (our cognitive faculties arose through natural selection and random genetic variation).
2. The probability that our cognitive faculties are reliable (R), given N&E, is low or inscrutable: P(R|N&E) is low or inscrutable.
3. If P(R|N&E) is low or inscrutable, then anyone who accepts both N and E has a defeater for R.
4. A defeater for R is a defeater for any belief produced by our cognitive faculties - including the belief in N&E itself.
5. Therefore, N&E is self-defeating: accepting it gives you a reason to reject it.
6. Theism does not face this problem, because on theism God designed our cognitive faculties to be reliable.

The argument does not claim evolution is false or that naturalists are irrational. Rather, it claims that the combination of naturalism and evolution is internally incoherent - that a consistent naturalist who accepts evolution should lose confidence in all of her beliefs, including her naturalism.

Why Plantinga Thinks Evolution Cannot Guarantee Truth

Behavior vs. Belief

The central engine of Plantinga’s argument is the claim that natural selection acts on behavior, not on the truth of beliefs. A gazelle that runs from a lion survives. But Plantinga contends that many different belief systems could produce the same fleeing behavior. Perhaps the gazelle believes the lion is dangerous and runs away. Or perhaps it believes the lion is a friendly creature and is running toward it by a roundabout route. Or perhaps it wants to play tag. As long as the behavior is adaptive, selection is indifferent to the underlying beliefs.

If beliefs are “invisible” to natural selection in this way, then there is no evolutionary pressure specifically favoring true beliefs. The probability that our cognitive faculties reliably produce true beliefs, given only naturalism and unguided evolution, would be no better than chance.

The Probability Claim

Plantinga argues that P(R|N&E) - the probability that our cognitive faculties are reliable given naturalism and evolution - is either low or inscrutable. He considers several ways beliefs might relate to behavior under a naturalistic framework:

• Epiphenomenalism: Beliefs are byproducts of brain states with no causal influence on behavior. If so, selection cannot shape beliefs at all, and R is extremely unlikely.
• Semantic epiphenomenalism: Beliefs have causal power through their neurophysiological properties, but their content (what they are about) is causally irrelevant. Selection shapes the neural hardware, but truth or falsity plays no role.
• Content-relevant causation: Beliefs cause behavior partly in virtue of their content. Even here, Plantinga argues, false beliefs could produce adaptive behavior just as readily as true ones.

Under each scenario, Plantinga concludes that we have no grounds for confidence in our cognitive reliability.

Counterarguments and Rebuttals

Evolution Does Track Truth

The most powerful response is that Plantinga dramatically underestimates the connection between true beliefs and adaptive behavior. Philosophers like Elliot Sober and Branden Fitelson have argued that in practice, the simplest and most reliable way for an organism to produce consistently adaptive behavior is to represent the environment accurately. Plantinga’s alternative scenarios - the gazelle that runs from lions because it finds them sexually attractive - are logically possible but biologically absurd. They require elaborate, gerrymandered belief-desire systems that are far less likely to arise and far less robust across varied environments than simple truth-tracking systems.

Consider the evolutionary epistemology literature: organisms with more accurate perceptual systems consistently outcompete those with less accurate ones. The independent evolution of camera-type eyes across multiple lineages demonstrates that natural selection powerfully drives toward environmental accuracy. While evolution does not produce perfect cognition - our many cognitive biases attest to that - it reliably produces cognition that is accurate enough to ground further reasoning.

The Bizarre Belief Scenarios Are Not Equiprobable

Plantinga treats his fanciful alternative belief systems as if they were equally probable to truth-tracking belief systems. But philosophers like James Beilby have pointed out that this ignores the enormous space of possible false beliefs relative to the narrow set that would produce adaptive behavior. For any given adaptive behavior, there may be one or a few true belief-desire combinations that produce it, but the false belief-desire combinations that happen to produce the same behavior are wildly specific and improbable. The likelihood of evolution stumbling upon a complex system of false beliefs that just happens to generate correct behavior across a wide range of environments is negligible compared to the likelihood of it producing approximately true beliefs.

The Self-Defeat Problem Applies to Theism Too

If Plantinga’s skeptical argument about cognitive reliability is taken seriously, it threatens theistic belief as well. How do we know that God designed our minds for truth rather than for obedience, worship, or comfort? We would need reliable cognitive faculties to evaluate whether our cognitive faculties were designed for truth - but that is the very capacity in question. As philosopher Troy Nunley and others have argued, the EAAN generates a universal skepticism that does not selectively spare theism.

Naturalism Can Be Confirmed Empirically

Even if Plantinga is right that P(R|N&E) is initially low, this does not mean naturalists are trapped. Naturalized epistemology, developed by W.V.O. Quine, holds that we can bootstrap our way to justified confidence in our cognitive faculties through empirical testing. Science works by generating predictions and checking them against observations. The spectacular success of science - landing rovers on Mars, predicting eclipses to the second, curing diseases - provides strong evidence that our cognitive faculties are reliable, regardless of the prior probability assigned by a philosophical argument. Plantinga’s a priori probability calculation is overwhelmed by this a posteriori evidence.

Historical Background

Plantinga first sketched the EAAN in Chapter 12 of Warrant and Proper Function (1993), though the seeds of the idea go back much further. Charles Darwin himself expressed a version of the worry in an 1881 letter: “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.” C.S. Lewis made a related argument in Miracles (1947), contending that naturalism undermines the rational foundations it needs to support itself.

Plantinga’s distinctive contribution was to formalize the argument using probability theory and to embed it within his broader Reformed epistemology framework. He refined the argument in response to critics in Naturalism Defeated? (2002, edited by James Beilby), which collected responses from supporters and critics alike, and again in Where the Conflict Really Lies (2011), where he argued that the real conflict in science and religion is not between theism and science but between naturalism and science.

Modern Developments

The EAAN continues to generate active debate in epistemology and philosophy of religion. Several developments are worth noting:

Bayesian reformulations. Philosophers have attempted to formalize Plantinga’s probability claims using Bayesian epistemology. These formalizations have generally been unfavorable to the EAAN, showing that even modest connections between true beliefs and adaptive behavior are sufficient to make P(R|N&E) reasonably high.

Evolutionary debunking arguments. The broader family of evolutionary debunking arguments - which use evolutionary origins to cast doubt on moral beliefs, mathematical beliefs, or religious beliefs - has grown substantially. Ironically, some of these arguments are used against theism: if religious beliefs evolved because they were adaptive (promoting group cohesion and cooperation), that gives us reason to doubt their truth, not affirm it.

Cognitive science of religion. Research in the cognitive science of religion suggests that humans are naturally predisposed to detect agency and purpose in the environment - a tendency called the “hyperactive agency detection device.” This research cuts both ways: it explains why theistic beliefs are widespread (they are cognitively natural) but also suggests they may be products of cognitive biases rather than accurate perception of reality.

Reliabilism and proper function. Plantinga’s own proper function epistemology holds that beliefs are warranted when produced by cognitive faculties functioning properly according to a design plan aimed at truth. Naturalists argue that evolution provides exactly such a design plan, making proper function epistemology compatible with naturalism despite Plantinga’s protests.

Relationship to Other Arguments

The EAAN is closely related to the Argument from Reason, which makes a broader case that naturalism undermines the reliability of rational thought. Where the EAAN focuses specifically on the evolutionary origins of cognition, the Argument from Reason (following C.S. Lewis) targets the more general question of whether physical causation can ground rational inference. The two arguments share the same basic strategy - arguing that naturalism is self-defeating - and face many of the same objections.

The Argument from Consciousness offers a complementary approach from a different angle. Rather than questioning whether evolved minds can be reliable, it questions whether physical processes can produce conscious experience at all. Together, these three arguments form a cluster of challenges centered on the nature of mind.

The Moral Argument for God employs a parallel structure in the ethical domain: if moral beliefs evolved for survival rather than tracking objective moral truth, then evolutionary naturalism undermines moral realism. All four arguments share the strategy of claiming that naturalism cannot ground some important feature of our mental lives - reason, consciousness, morality, or cognitive reliability.

Critics often point out that the Burden of Proof Argument applies forcefully here: the EAAN does not provide positive evidence for God’s existence but instead tries to undermine naturalism, leaving theism as the last option standing. An argument that merely weakens the competition, without providing its own evidence, is inherently limited. The Occam’s Razor principle also applies - invoking a supernatural designer to explain cognitive reliability introduces additional complexity that requires justification.

Common Misconceptions

“The EAAN says evolution is false.” It does not. Plantinga accepts evolution. The argument targets the combination of naturalism and evolution, claiming they are jointly self-defeating. Plantinga believes evolution is true but guided by God.

“The EAAN proves God exists.” The argument is defensive, not constructive. Even if fully successful, it would only show that naturalism is self-defeating - it would not positively establish God’s existence. Other non-naturalist worldviews (panpsychism, idealism, neutral monism) would also escape the self-defeat charge.

“Plantinga claims we cannot trust science.” Plantinga’s claim is conditional: if naturalism is true, then we cannot trust our faculties. He believes we can trust our faculties because he believes theism (not naturalism) is true. The argument is an attack on naturalism, not on science or on evolution.

Our Scoring

The soundness score of 20 reflects that the EAAN, while cleverly constructed, rests on a deeply questionable central premise - that natural selection cannot produce reliable cognitive faculties. The argument’s fanciful scenarios of false beliefs producing adaptive behavior are logically possible but biologically implausible. Evolutionary epistemology provides strong reasons to think that truth-tracking is the most efficient and robust cognitive strategy for surviving in complex environments, and the empirical success of science provides overwhelming a posteriori evidence that our cognitive faculties are substantially reliable. The argument also faces the bootstrapping objection: its skeptical conclusion, if taken seriously, would undermine the very reasoning used to construct the argument in the first place, creating a self-referential paradox. Most philosophers working in epistemology and philosophy of mind consider the EAAN an interesting but ultimately unsuccessful challenge to naturalism.

The Creator score of 50 is the highest, alongside Higher Power, because the EAAN - if sound - most directly suggests that our cognitive faculties were designed for truth by an intelligent being. The idea of a designer who built minds to be reliable maps naturally onto a Creator concept. However, the argument does not specify what kind of creator or provide evidence of specific creative acts, which limits the score.

The Higher Power score of 50 matches the Creator score because the argument equally supports a vaguer notion of some rational principle or consciousness behind reality that ensures our minds can track truth. A Higher Power that grounds the rational order of the universe would satisfy the argument’s requirements without needing to be a personal being with specific attributes.

The Personal God score of 45 is slightly lower because the EAAN, even if sound, establishes only that something designed our minds for reliability - not that this designer is omniscient, omnibenevolent, or personally invested in human affairs. The gap between “a rational designer ensured cognitive reliability” and “the God of classical theism who hears prayers and intervenes in history” is substantial. Plantinga himself bridges this gap through his broader Reformed epistemology, but the EAAN alone does not provide the resources to do so. The score is still moderate because a personal God who cares about truthful relationship with humans would have strong motivation to equip us with reliable cognitive faculties.