ris3n's Apologetics Codex

Argument

Argument from the Reliability of Reason

Intro

There are ads on our codex that pay for hosting and keep the codex free. If you can, please consider whitelisting ris3n.com or allowing scripts to support the work.

Sponsored

If your brain is just an organ shaped by blind evolution to keep your ancestors alive long enough to reproduce, why should you trust what it tells you about big-picture questions like how the universe started or whether God exists?

That is the strange corner the atheist worldview backs itself into. Evolution selects for survival, not truth. Most of the time those two go together: see the snake, jump back, live another day. But survival and truth are not the same thing. A caveman who runs from every rustle, whether it is a tiger or the wind, lives just as long as one who only runs from real tigers. Beliefs that are useful do not have to be accurate.

So if your reasoning apparatus was built only to keep your genes around, you have no strong reason to trust its conclusions on subjects evolution did not care about. And one of those subjects is whether evolution is the only thing that built me. That conclusion is produced by the same brain whose reliability you just questioned. The whole picture wobbles.

The Christian view sidesteps this. Our minds were built by a Mind, on purpose, to track truth. We are not perfect at it, but the basic equipment is designed for the job. Reason can be trusted, in a careful and humble way, because it comes from somewhere reasonable.

C.S. Lewis sketched this in Miracles in 1947. Plantinga gave it formal teeth in his Evolutionary Argument Against Naturalism (EAAN) in the 1990s. The argument does not directly prove God exists. It shows that naturalism plus evolution undercuts its own ground to stand on.

In full

A reductio against the conjunction of naturalism and evolution (N&E). If naturalism is true and our cognitive faculties are produced by unguided evolution, the probability that those faculties are reliably truth-tracking is low or inscrutable. So the naturalist who recognizes this has a defeater for any belief produced by their faculties, including the belief in N&E itself. Therefore N&E is self-defeating; one cannot rationally affirm it. Plantinga's Evolutionary Argument Against Naturalism (EAAN) is the most rigorous formulation; Lewis's Miracles (1947) is the classical-apologetic predecessor. This page is structured as debate prep, each premise carries a second-order positive case, anticipated objections, rebuttals, a live-cite kit, and tactical notes for live engagement. Sister page: Argument from Reason (Lewis-style ground-consequent version).

Argument structure

# Premise
P1 If naturalism is true, our cognitive faculties were produced by unguided evolution.
P2 Unguided evolution selects for survival-fitness, not truth-tracking; the two can systematically decouple.
P3 P(R | N&E), the probability that our cognitive faculties are reliable, given naturalism + evolution, is low or inscrutable.
P4 If you accept N&E and recognize P3, you have a defeater for any belief produced by your faculties.
P5 This defeater includes the belief in N&E itself (the defeater is self-applying).
C Therefore, N&E is self-defeating; rationally, one cannot accept it.

Form

Probabilistic-deductive defeater (reductio). The argument does not directly establish theism; it establishes that the joint affirmation of naturalism and evolution rationally undermines itself. Plantinga's contribution is to give Lewis's intuition rigorous probabilistic form: the key step is the probability claim (P3), defended via considerations about the relation between adaptive behavior and belief content under naturalist constraints. The conclusion is self-defeat, not direct refutation. The natural further move (theism best explains why our faculties are reliable) is abductive and made separately.

P1, If naturalism is true, our cognitive faculties were produced by unguided evolution

Affirmative case (second-order arguments)

  1. Naturalism's official commitment. Mainstream naturalism holds there is no God, no transcendent designer, no top-down telos in the natural order. Cognitive faculties are biological systems; biological systems on naturalism are products of evolution by natural selection (the only known game in town for explaining adaptive complexity without design).
  2. No alternative on offer. No naturalist has proposed a non-evolutionary alternative for the origin of cognitive faculties that does not collapse into design or theism. The conjunction N&E is what every actual contemporary naturalist defends.

Anticipated objections

  1. "Some naturalists hold that consciousness is fundamental (panpsychism, neutral monism)."
  2. "You're attacking a strawman of naturalism."

Rebuttals

  1. Panpsychist naturalists are still committed to evolutionary origins for human cognitive faculties. Even if proto-consciousness is fundamental, the specific organization that yields human-level reasoning is evolutionary. The argument runs unchanged.
  2. The argument targets exactly the naturalism most naturalists defend. Plantinga himself is careful: the argument is against the conjunction N&E, not against every conceivable naturalism. If a naturalist disclaims evolution as the origin of faculties, they have moved to a fringe position that owes its own account.

Live-cite kit

  • Scripture: Genesis 1:27 (imago Dei); Psalm 139:14
  • Scholarly: Plantinga (Where the Conflict Really Lies, 2011, ch. 10); Reppert (C. S. Lewis's Dangerous Idea, 2003)
  • Aphorism: "Naturalism owes us an account of the mind. Evolution is its only candidate."

Tactical notes

  • This premise rarely needs defense in live debate; opponents grant it. Move quickly to P2.

P2, Unguided evolution selects for survival-fitness, not truth-tracking

Affirmative case (second-order arguments)

  1. Standard Darwinism: only behavior matters to selection. Natural selection acts on phenotypic traits that affect reproductive success. Beliefs are evolutionarily relevant only insofar as they cause survival-enhancing behavior. Selection has no direct grip on the truth of beliefs, only on the survival-conducive behavior they produce.
  2. Truth and survival can decouple in principle. A creature could survive perfectly well with massively false beliefs, as long as those beliefs cause appropriate actions. Plantinga's "Paul running from tigers" example: Paul could believe tigers are friendly playmates and that running is the best way to engage them, and survive by running away, without ever forming a true belief about tigers.
  3. Most useful beliefs are simplifications, heuristics, or approximations rather than truths. Newtonian physics survived evolutionary selection long after relativity showed it false. Folk biology, folk psychology, folk physics, all serve evolutionary purposes by being adaptive, not by being true.
  4. Semantic content is causally inert on standard physicalism. If brain-states cause behavior in virtue of their physical (not semantic) properties, the content of a belief is not what selection sees. Selection sees the behavior; the belief-content is hitchhiking. There is no mechanism that ties content-truth to selection.

Anticipated objections

  1. "Truth-tracking IS survival-conducive, accurate perception of reality survives better than inaccurate perception." The standard intuitive reply.
  2. "Even granting truth and survival can decouple in principle, they will tend to converge in practice (the 'truth tends to be useful' line)."
  3. "This is the same argument from incredulity, selection HAS produced reliable cognition; we're here aren't we?"

Rebuttals

  1. The objection conflates behavior-environment coupling with belief-environment coupling. Selection requires the former (frog catches fly, antelope flees lion); it does not require the latter (frog truly believes about the fly). The naturalist needs to argue that natural selection privileges content-sensitive belief-action coupling. But on physicalism, semantic content doesn't have causal powers, it's the physical states that cause, and contents come along for the ride. The objection assumes content-sensitivity that physicalism cannot deliver. Failure mode: conflating behavior-tracking with belief-tracking.
  2. Convergence in practice is empirically false in the deep cases. We can grant that for immediate practical matters (food, predators, mates) truth and survival often align. But for deep metaphysical truths, naturalism, theism, the structure of reality, the alignment claim is precisely what's at issue. Selection has no grip on metaphysics; the cognitive faculty for forming beliefs about deep matters was not directly selected for. Yet N&E is itself a deep metaphysical belief. Failure mode: smuggling in the reliability that needs to be derived.
  3. "We're here, aren't we?" confuses practical effectiveness with truth-tracking. We can grant cognition is good enough for survival without conceding it is reliable about deep metaphysical truths. The argument is not that science doesn't work; it's that the naturalist's argument for naturalism is not the kind of cognitive achievement evolution can underwrite. Failure mode: shifting the goalposts from metaphysics to engineering.

Live-cite kit

  • Scripture: Romans 1:18-21 (general revelation grasp-able by reason); Romans 12:2 (renewing of the mind)
  • Scholarly: Plantinga (Warrant and Proper Function, 1993, ch. 12; Where the Conflict Really Lies, 2011, ch. 10); Reppert (2003); Stephen Stich (The Fragmentation of Reason, 1990, secular but useful, argues evolution doesn't deliver classical rationality)
  • Aphorism: "Selection cares whether you survive, not whether you're right."

Tactical notes

  • The "Paul and the tigers" example is the canonical illustration. Use it. It is concrete and forces the abstract point through.
  • Don't argue against evolution itself, that is a separate, distracting battle, and EAAN is strongest when it grants evolution. Plantinga himself accepts common-descent evolution; the argument runs against the conjunction with naturalism, not against evolution alone.
  • The opponent will frequently invoke "but evolution selects for truth because truth helps survival." Have the content-vs-behavior distinction memorized and ready.

P3, P(R | N&E) is low or inscrutable

Affirmative case (second-order arguments)

  1. Plantinga's formal probability argument. Given N&E, beliefs are produced by physical brain-states selected for behavioral consequences. Many false-belief / true-action mappings are equally adaptive (the frog catches the fly equally well whether it believes "tasty bug to seize" or "lethal threat to flee from", both produce tongue-extension if action-content is uncoupled from belief-content). Therefore, given N&E alone, we should expect most beliefs to be false or unreliable about reality, even when adaptive. P(R | N&E) is low.
  2. Inscrutability variant. Even if the probability isn't low, it is inscrutable, we have no good way to assign it. If we cannot rationally judge that our faculties are reliable, we have no warrant to trust them. The defeater fires either way (low probability or inscrutable probability).
  3. Convergent defenders. Patricia Churchland herself wrote that "what evolution does is select for being adequate to the task of survival, and we have no reason to think that this is the same as being correct about reality" (paraphrased; Brain-Wise, 2002). Naturalist Stephen Stich (The Fragmentation of Reason, 1990) makes a parallel case from inside the camp.

Anticipated objections

  1. "The probability calculation is malformed; we have no good way to assign priors (Fitelson, Sober)."
  2. "P(R | N&E) is HIGH, because reliable cognition is itself adaptive."
  3. "Even if low, the defeater doesn't follow, we can rationally trust our faculties without proof of reliability."

Rebuttals

  1. The objection is itself defeated by EAAN. To object to the probability calculation, the objector uses cognitive faculties whose reliability is precisely what's in question. You cannot rationally object to EAAN using faculties EAAN has shown to be unreliable. The regress closes off the escape route. Failure mode: self-undermining objection.
  2. The "reliable cognition is adaptive" reply is the P2 conflation again. Behavior is adaptive; belief-content is hitchhiking. Selection cannot tune content if content has no causal grip. Plantinga's response is detailed in Warrant and Proper Function ch. 12 and in his exchange with Sober (Where the Conflict Really Lies, 2011, ch. 10).
  3. Properly basic trust requires a non-defeater. It is true we ordinarily trust our faculties without proof, but only as long as we have no defeater. EAAN is the defeater. Once the defeater is in play, continued trust without addressing it is irrational, like continuing to trust a witness after evidence of deception. The defeater must be defeated, not ignored.

Live-cite kit

  • Scripture: John 1:9 ("the true Light… enlightens every man"); Psalm 36:9 ("in Your light we see light")
  • Scholarly: Plantinga (Warrant and Proper Function, 1993, ch. 12; "Reply to Beilby's Cohorts" in Naturalism Defeated?, 2002, ed. Beilby); Beilby (ed.) (Naturalism Defeated?, 2002, collected EAAN debates)
  • Aphorism: "Either God designed your faculties for truth, or you have no reason to trust them."

Tactical notes

  • The probability claim is the most technical part of the argument. In live debate, you can run the intuitive version (Paul-and-the-tigers) without defending the formal probabilities, the intuition does the work.
  • If the opponent challenges Plantinga's probability calculation specifically, be prepared with the inscrutability fallback: "Even if P(R | N&E) isn't low, it's inscrutable, and that's enough for the defeater."
  • The most rigorous engagements are in Naturalism Defeated? (Beilby ed., 2002). Cite it when the discussion gets technical.

P4, If you accept N&E and recognize P3, you have a defeater for any belief produced by your faculties

Affirmative case (second-order arguments)

  1. The defeater logic (Plantinga, Warrant: The Current Debate, 1993). A defeater is a reason for thinking a belief is false or unreliably formed. If I learn that my belief-forming process is unreliable, I have a reason to doubt any belief produced by it. Recognizing P3 is recognizing the unreliability; the defeater follows.
  2. The defeater applies broadly. Not just to specific beliefs but to all beliefs produced by the faculty. If the faculty is globally unreliable, every belief from the faculty inherits the unreliability, there is no special category of beliefs immune from the defeater.

Anticipated objections

  1. "I can compartmentalize, trust my faculties for daily life without trusting them for metaphysics."
  2. "The defeater is too strong, it would defeat all knowledge whatsoever."

Rebuttals

  1. Compartmentalization is ad hoc. The compartmentalizer needs to specify which beliefs are reliable and which aren't, on principled grounds, but EAAN's force is that the underlying process is uniformly unreliable on N&E. There is no principled cut.
  2. The defeater does defeat all knowledge whatsoever, for the naturalist. This is the point of EAAN, not a refutation. The naturalist who accepts N&E ends up with no warranted beliefs. The way out is to reject N&E; that's the argument's intended conclusion. Failure mode: mistaking the conclusion for an objection.

Live-cite kit

  • Scholarly: Plantinga (Warrant: The Current Debate, 1993; Warranted Christian Belief, 2000); Bergmann (Justification Without Awareness, 2006)
  • Aphorism: "An unreliable mind is unreliable about its own unreliability, and that doesn't make it reliable."

Tactical notes

  • Don't get stuck explaining defeater theory in detail, keep it simple: "If you have reason to think your eyes are unreliable, you can't trust what your eyes report; same with the mind."

P5, This defeater includes the belief in N&E itself

Affirmative case (second-order arguments)

  1. Self-application is automatic. N&E is a belief produced by the very faculties whose reliability has been undermined. There is no exempt category. The defeater for "all beliefs from this faculty" includes "this belief about all beliefs from this faculty."
  2. The reasoning that supports N&E is also under defeat. Even the inferences by which the naturalist arrived at N&E (the philosophical arguments, the scientific evidence, the explanatory parsimony arguments) are products of faculties that, on N&E's own terms, are not reliably truth-tracking.

Anticipated objections

  1. "This is just self-defeat by definition; it's a verbal trick."
  2. "Theism is no better, God could give us false beliefs (a trickster God scenario)."

Rebuttals

  1. It's not a verbal trick, it's the substantive consequence of the structure. Self-defeat is what you get when a position's truth implies the unreliability of the very faculty by which the position is affirmed. This is the same structure that makes "this sentence is false" a paradox and "I never tell the truth" a self-undermining claim. Naturalists object to applying this to their own metaphysics, but the structural parallel is exact. Failure mode: dismissing structural consequences as verbal tricks.
  2. A truthful God grounds expectation of reliability; a deceiver God is ruled out by Christian theism's specific commitments. The objection conflates "any theism" with "Christian theism." Christian theism includes God is truth (John 14:6, 17:17), God cannot lie (Hebrews 6:18, Titus 1:2), God created humans in His image (Gen 1:27, including for truth-tracking). These commitments rule out the trickster scenario for the Christian. The argument doesn't claim God cannot give false beliefs in some bizarre possible world; it claims God's nature grounds our expectation of generally-truth-tracking cognition. Naturalism has no such grounding. Failure mode: conflating distinct theisms; ignoring Christian-specific commitments.

Live-cite kit

  • Scripture: John 14:6; John 17:17 ("Your word is truth"); Hebrews 6:18 ("impossible for God to lie"); Titus 1:2; Genesis 1:27
  • Scholarly: Plantinga (Warranted Christian Belief, 2000); Bergmann (2006); Descartes (Meditations IV, the original move from divine truthfulness to faculty-reliability)
  • Aphorism: "A liar can have unreliable witnesses. A truthful God designs reliable witnesses."

Tactical notes

  • The "self-defeat is just a verbal trick" reply is the most common deflection. Have the structural-consequence frame ready, point out that this is the same logic that makes paradoxes paradoxes.
  • The "trickster God" reply is also common. Have the Christian-specific commitments memorized: God is truth; cannot lie; designed humans for knowing.

Conclusion

Therefore, N&E is self-defeating; rationally, one cannot accept it. Some other view that grounds the reliability of reason, most plausibly, theism (in which a rational, truthful God created humans with truth-tracking minds in His image), is required. The argument does not directly establish theism; it eliminates a major competitor and creates the dialectical opening for theism's positive case. Combined with the Argument from Reason (Lewis-style), Argument from Consciousness, and the natural-theological arguments, the case strengthens.

Master objections to the whole argument

  1. "EAAN proves too much, it would defeat all knowledge." Reply: only for the naturalist. The argument's whole point is that if N&E, then global cognitive defeat. The way out is to reject N&E. Misreading the conclusion as an objection.
  2. "Reliable cognition would itself be selected for, survival often depends on truth." Reply: see P2 rebuttal. Conflates behavior-tracking with belief-tracking; relies on the content-causation that physicalism cannot deliver.
  3. "Our reasoning does work, we built civilization, science, technology." Reply: see P3 rebuttal. Conflates practical effectiveness with truth-tracking about deep metaphysics.
  4. "Theism is no better, God could give us false beliefs." Reply: see P5 rebuttal. Conflates Christian theism with generic theism; ignores God-is-truth commitment.
  5. "Plantinga is a Christian philosopher with a stake in the conclusion." Reply: ad hominem; the argument stands or falls on its merits. Note that secular philosophers (Stich, Churchland in moments) make adjacent points.
  6. "This is god-of-the-gaps applied to epistemology." Reply: not gaps, structural mismatch between unguided selection and truth-tracking faculties. The argument identifies a principled feature of the relationship, not a temporary lacuna.

Tactical opening / closing lines

Opening line: "If your mind was produced by an unguided process that selects only for survival, then you have no reason to trust what your mind tells you about anything that doesn't directly affect survival, including the belief that your mind is the product of an unguided process. Want to walk through why?"

Closing landing strip: "EAAN doesn't compel Christian theism; it eliminates the joint affirmation of naturalism and evolution as a rationally affirm-able position. Once that's gone, the question becomes which worldview can underwrite the reliability of the very mind doing the asking. Christianity's answer, a truthful God designed minds for truth, has the structural fit naturalism cannot replicate."

Connection to Scripture

Patristic / scholarly note

Classical / patristic / medieval:

  • Augustine (De Trinitate 14-15; De Magistro), divine illumination doctrine
  • Aquinas (ST I, q. 84, a. 5), human intellect knows truth by analogous participation in divine intellect
  • Descartes (Meditations IV), the original move from divine truthfulness to faculty-reliability

Modern:

  • C. S. Lewis (Miracles, 1947, esp. chs. 3-5), classical apologetic predecessor
  • Alvin Plantinga (Warrant and Proper Function, 1993; Warranted Christian Belief, 2000; Where the Conflict Really Lies, 2011), the EAAN
  • James Beilby, ed. (Naturalism Defeated? Essays on Plantinga's Evolutionary Argument Against Naturalism, 2002), collected debates
  • Victor Reppert (C. S. Lewis's Dangerous Idea: In Defense of the Argument from Reason, 2003)
  • William Hasker (The Emergent Self, 1999)
  • J. P. Moreland (The Recalcitrant Imago Dei, 2009)
  • Edward Feser (Philosophy of Mind, 2005)
  • Michael Bergmann (Justification Without Awareness, 2006), defeater theory

Critical / engagement:

  • William Ramsey (Stanford Encyclopedia of Philosophy entry on EAAN), surveys debate
  • Daniel Dennett (Darwin's Dangerous Idea, 1995), naturalist counter
  • Patricia Churchland (Brain-Wise, 2002)
  • Stephen Stich (The Fragmentation of Reason, 1990), secular but EAAN-adjacent

See also